Professional Documents
Culture Documents
Thoracic Post Op Complications PDF
Thoracic Post Op Complications PDF
Thoracic Post Op Complications PDF
and Management
of Postoperative
Complications
Editor
JOHN D. MITCHELL
THORACIC
SURGERY CLINICS
www.thoracic.theclinics.com
Consulting Editor
M. BLAIR MARSHALL
http://www.thoracic.theclinics.com
This periodical and the individual contributions contained in it are protected under copyright by Elsevier, and the following
terms and conditions apply to their use:
Photocopying
Single photocopies of single articles may be made for personal use as allowed by national copyright laws. Permission of the
Publisher and payment of a fee is required for all other photocopying, including multiple or systematic copying, copying for
advertising or promotional purposes, resale, and all forms of document delivery. Special rates are available for
educational institutions that wish to make photocopies for non-profit educational classroom use. For information on how to
seek permission visit www.elsevier.com/permissions or call: (144) 1865 843830 (UK)/(11) 215 239 3804 (USA).
Derivative Works
Subscribers may reproduce tables of contents or prepare lists of articles including abstracts for internal circulation within their
institutions. Permission of the Publisher is required for resale or distribution outside the institution. Permission of the Publisher is
required for all other derivative works, including compilations and translations (please consult www.elsevier.com/permissions).
Notice
No responsibility is assumed by the Publisher for any injury and/or damage to persons or property as a matter of products
liability, negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in
the material herein. Because of rapid advances in the medical sciences, in particular, independent verification of diagnoses
and drug dosages should be made.
Although all advertising material is expected to conform to ethical (medical) standards, inclusion in this publication does not
constitute a guarantee or endorsement of the quality or value of such product or of the claims made of it by its manufacturer.
Thoracic Surgery Clinics (ISSN 1547-4127) is published quarterly by Elsevier Inc., 360 Park Avenue South, New York, NY 10010-
1710. Months of publication are February, May, August, and November. Business and editorial offices: 1600 John F. Kennedy
Boulevard, Suite 1800, Philadelphia, PA 19103-2899. Periodicals postage paid at New York, NY, and additional mailing offices.
Subscription prices are $350.00 per year (US individuals), $453.00 per year (US institutions), $165.00 per year (US Students),
$435.00 per year (Canadian individuals), $585.00 per year (Canadian institutions), $225.00 per year (Canadian and international
students), $465.00 per year (international individuals), and $585.00 per year (international institutions). Foreign air speed delivery
is included in all Clinics’ subscription prices. All prices are subject to change without notice. POSTMASTER: Send address
changes to Thoracic Surgery Clinics, Elsevier Health Sciences Division, Subscription Customer Service, 3251 Riverport Lane,
Maryland Heights, MO 63043. Customer Service (orders, claims, online, change of address): Telephone: 1-800-654-
2452 (U.S. and Canada); 314-447-8871 (outside U.S. and Canada). Fax: 314-447-8029. E-mail: journalscustomerservice-
usa@elsevier.com (for print support); journalsonlinesupport-usa@elsevier.com (for online support).
Reprints. For copies of 100 or more, of articles in this publication, please contact Commercial Rights Department, Elsevier Inc.,
360 Park Avenue South, New York, NY 10010-1710. Tel: 212-633-3874; Fax: 212-633-3820; E-mail: reprints@elsevier.com.
Thoracic Surgery Clinics is covered in MEDLINE/PubMed (Index Medicus), EMBASE/Excerpta Medica, Science Citation Index
Expanded (SciSearch), Journal Citation Reports/Science Edition, and Current Contents/Clinical Medicine.
Contributors
CONSULTING EDITOR
EDITOR
AUTHORS
Contents
Preface: Reducing the Footprint of Postoperative Complications xiii
John D. Mitchell
Cardiovascular events after thoracic surgery can result in increased morbidity, mor-
tality, length of stay, and increased overall cost. The prevention of postoperative
cardiovascular complications is an area of intense study, and the body of evidence
guiding clinicians continues to grow. Early diagnosis and management of cardiovas-
cular events can minimize the consequences of these complications.
The Prevention and Management of Air Leaks Following Pulmonary Resection 411
Bryan M. Burt and Joseph B. Shrager
Alveolar air leaks are a common problem in the daily practice of thoracic surgeons.
Prolonged air leak following pulmonary resection is associated with increased
morbidity, increased length of hospital stay, and increased costs. This article
reviews the evidence for the various intraoperative and postoperative options to
prevent and manage postoperative air leak.
Empyema after anatomic lung resection is rare but causes serious morbidity, partic-
ularly if associated with a bronchopleural fistula. Careful assessment of preoperative
risk factors and proper surgical technique can minimize risks. Empyema after
segmentectomy or lobectomy may respond to simple drainage and antibiotics, or
may require decortication with or without muscle transposition. After pneumonec-
tomy, treatment principles include initial drainage of the intrathoracic space, closure
of the fistula if present, and creation of an open thoracostomy, which is packed and
later closed. Success rates can exceed 80%.
viii Contents
Complications Following Surgery for Gastroesophageal Reflux Disease and Achalasia 485
Samad Hashimi and Ross M. Bremner
Surgical procedures to treat reflux disease are common, but good outcomes rely on
both a thorough preoperative workup and careful surgical techniques. Although
complications are uncommon, surgeons should recognize these and possess the
skills to overcome them in clinical practice.
Nerve injuries can cause substantial morbidity after thoracic surgical procedures.
These injuries are preventable, provided that the surgeon has a thorough understand-
ing of the anatomy and follows important surgical principles. When nerve injuries
occur, it is important to recognize the options available in the immediate and postop-
erative settings, including expectant management, immediate nerve reconstruction,
or auxiliary procedures. This article covers the basic anatomy and physiology of
nerves and nerve injuries, an overview of techniques in nerve reconstruction, and a
guide to the nerves most commonly involved in thoracic operative procedures.
The main indications for chest wall resection continue to be tumors, infection, and
radiation injury. Complications surrounding chest wall resection procedures include
x Contents
Index 529
Prevention and Management of Postoperative Complications xi
RELATED INTEREST
Radiologic Clinics, Volume 52, Issue 1 (January 2014)
Thoracic Imaging
Jane P. Ko, Editor
Available at: www.radiologic.theclinics.com
P re f a c e
R e d u c i n g th e F o o t p r i n t o f
Postoperative Complications
No matter what measures are taken, doctors leak, are specific to particular operations. In
will sometimes falter, and it isn’t reasonable each article, though, the contributing authors
to ask that we achieve perfection. What is have tried to emphasize both prevention and
reasonable is to ask that we never cease to management strategies to minimize patient
aim for it. morbidity. I believe they have been successful
—Atul Gawande, Complications: A in achieving this goal.
Surgeon’s Notes on an Imperfect Science I’d like to thank each of the contributing authors
for their expertise and for their outstanding contri-
If you operate, expect things to occasionally go butions to this issue. A special thanks goes to
wrong. This is a fact surgeons should learn early Blair Marshall, the Consulting Editor of Thoracic
in their career—that no one is exempt from patient Surgery Clinics, for her gracious invitation to serve
morbidity. It is possible, though, to minimize the as guest editor and her gentle guidance after my
occurrence of complications through thoughtful acceptance of her offer. Finally, I’d like to express
preventive measures in the perioperative time my appreciation to John Vassallo and Susan
frame, and to reduce the impact of complications Showalter of Elsevier for their help and support
when they do occur through early recognition throughout the preparation of this issue.
and decisive patient management. Surgeons
adept at these strategies are often the most suc- John D. Mitchell, MD, FACS
cessful with the best patient outcomes. Section of General Thoracic Surgery
This issue of Thoracic Surgery Clinics is dedi- Division of Cardiothoracic Surgery
cated to the prevention and management of University of Colorado School of Medicine
complications that may occur following thoracic Academic Office 1, C-310
surgery. Some of the topics included, such as 12631 East 17th Avenue
adverse cardiovascular events, respiratory fail- Aurora, CO 80045, USA
ure, and chronic pain syndromes, are generic to
any thoracic procedure. Other complications, E-mail address:
such as prolonged air leak and anastomotic John.Mitchell@ucdenver.edu
thoracic.theclinics.com
KEYWORDS
Atrial fibrillation Myocardial infarction Pulmonary embolism
KEY POINTS
Cardiovascular complications occur in up to 30% of patients following noncardiac thoracic surgery.
In patients with recent percutaneous coronary intervention, the risk of stent thrombosis must be
weighed against the risk of bleeding if antiplatelet therapy is continued through surgery.
Venous thromboembolism prophylaxis after thoracic surgery is optimized by the use of mechanical
and pharmacologic prophylaxis.
Several medications reduce the incidence of atrial fibrillation after thoracic surgery but may have
significant side effects requiring surgeons to individualize the consideration for prophylaxis.
When treating new-onset atrial fibrillation, it is critical to consider a potential process that is driving
the rhythm change, such as pulmonary embolus or tension pneumothorax.
Risk Factors and Risk Stratification MI.11,12 This point is particularly relevant as many
patients may be at increased risk for bleeding
The American College of Cardiology (ACC) and
because of the need for perioperative anticoagula-
the American Heart Association (AHA) both
tion, as outlined in the following sections.
consider noncardiac thoracic surgery to be an in-
termediate risk for postoperative cardiovascular
death and nonfatal MI,5 which is consistent with Current Guidelines for Preoperative Cardiac
the reported range in the general thoracic surgical Testing
literature (1%–5%).6,7 Preoperative risk assess-
Provocative cardiac testing (stress test) has the
ment for MI in thoracic patients may be performed
potential to identify coronary artery disease that
using the Revised Goldman Cardiac Risk Index
places patients at a higher risk for perioperative
(RCRI). The RCRI was derived using a broad
MI. However, the desire to minimize MI risk must
noncardiac surgery patient population (12% gen-
be tempered with the hazards of performing tests
eral thoracic) and is one of the most validated and
that are not indicated. Several specialty-specific
widely used risk models for postoperative events.
societies have attempted to rectify the potential
This model uses 6 risk factors to predict the rate
for risk modulation (chance the test will help the
of cardiac death, nonfatal MI, and nonfatal car-
patients) with resource utilization (chance the test
diac arrest after noncardiac surgery as shown in
was not helpful). The 2013 American College of
Table 1.8,9 In 2010 Brunelli and colleagues10 pro-
Cardiology Foundation (ACCF) consensus state-
posed a modified version of the RCRI that was
ment has addressed preoperative cardiac evalua-
specific for patients undergoing lobectomy or
tion by risk of procedure (Fig. 1).13,14 A brief
pneumonectomy. In this scoring system, serum
synopsis is given next.
creatinine greater than 2 mg/dL, history of cardiac
The following scenarios should proceed to sur-
ischemia, cerebrovascular ischemia, and pneu-
gery without further cardiac testing:
monectomy as a surgical procedure were associ-
ated with an increased risk of major cardiac All emergent procedures
complications. Elective low-risk procedures (ie, endoscopic,
In addition to preoperative considerations, sur- superficial procedures)
geons should attempt to minimize perioperative Elective intermediate- to high-risk procedures
blood loss, as several studies have implicated (ie, lobectomy, pneumonectomy, esophagec-
transfusion as a risk factor for perioperative tomy) in patients without symptoms that can
Table 1
RCRI and the risk of major cardiac events
RCRI
Predictors/risk factors High-risk surgery (vascular, thoracic, open intraperitoneal)
History of ischemic heart diseasea or history of coronary revascularization
with current IHD symptoms
History of heart failureb
History of cerebrovascular diseasec
Insulin-dependent diabetes mellitus
Preoperative serum Cr >2.0 mg/dL
Rate of major cardiac No. of Risk Factors
events 0: 0.5%
1: 1.3%
2: 4.0%
3: 9.0%
Fig. 1. Evaluation and care for thoracic surgery with cardiac risk factors. MET, metabolic equivalent. a Acute car-
diac condition: MI/acute coronary syndrome. b Low-risk surgery: endoscopic, superficial, ambulatory surgery.
c
Proceed with surgery if work-up is negative, if patients are stabilized, and if there is an elevated risk of adverse
outcomes with delaying surgery. d Delay surgery if patients have unstable coronary disease, decompensated heart
failure, severe arrhythmia, or valvular heart disease or if patients undergo percutaneous coronary intervention.
(Adapted from Report of the American College of Cardiology/American Heart Association Task Force on Practice
Guidelines (Writing Committee to Revise the 2002 Guidelines on Perioperative Cardiovascular Evaluation for
Noncardiac Surgery): developed in collaboration with the American Society of Echocardiography, American Soci-
ety of Nuclear Cardiology, Heart Rhythm Society, Society of Cardiovascular Anesthesiologists, Society for Cardio-
vascular Angiography and Interventions, Society for Vascular Medicine and Biology, and Society for Vascular
Surgery. Circulation 2007;116:e423.)
Table 2
Metabolic equivalents and questions concerning activity
if the results would change management (ie, if with known COPD and asthma,5 newer protocols
the team reacts to a positive stress test, with a with adenosine have been shown safe and effec-
coronary intervention, or direct patients to a tive for thoracic surgery patients.22–24 It is also
lower-risk procedure) worth noting that theophylline (as well as caffeine)
within 24 hours of a stress test can interfere with
With the increased use of computed tomogra- the effectiveness of pharmacologic stress test
phy (CT) scanning, coronary calcifications are agents and, therefore, should be avoided.14
increasingly identified preoperatively in the
thoracic surgery population. At this time, there is
not a consensus with regard to the need for further Perioperative Management of Antiplatelet
testing for incidentally found coronary calcifica- Medications for Coronary Stents
tions on thoracic imaging. The reported risk It is not uncommon for general thoracic surgery pa-
associated with incidentally found coronary calci- tients to be taking antiplatelet therapy for prior coro-
fications is inconsistent but may be significant in nary stenting (given that smoking is a risk factor for
the setting of type II diabetes.15–17 both thoracic malignancies and coronary artery dis-
ease). Antiplatelet medications may increase the
Echocardiography
likelihood of perioperative bleeding (which is itself
The 2014 ACC/AHA’s guidelines recommend pre-
a risk for perioperative MI) creating a dilemma for
operative resting echocardiography only be per-
thoracic surgeons. The risk of perioperative coro-
formed if pursuing an established or suspected
nary events varies in accordance with the type of
diagnosis (ie, to evaluate a murmur or worsening
stent that was placed as well as the interval between
dyspnea). Clinically stable patients with previous
stent placement and the thoracic surgery proce-
left ventricular dysfunction may benefit from an
dure. The ACC/AHA’s 2014 guidelines recommend
echocardiogram if there has not been an assess-
continued dual antiplatelet therapy (eg, aspirin and
ment within a year before surgery.1 Patients with
clopidogrel [Plavix]) for at least 12 months after
known or suspected pulmonary hypertension (eg,
drug-eluting stents (DES) placement and at least 4
marked pulmonary artery [PA] enlargement and
to 6 weeks for bare-metal stents (BMS) to reduce
impaired diffusion) may benefit from a preoperative
the rate of major cardiac events.1 The most common
echocardiogram to better assess PA pressure and
medications used as dual antiplatelet therapy are
right heart function for perioperative management.18
aspirin and clopidogrel, although several newer
Stress test considerations medications are increasingly being used (Table 3).25
The most appropriate stress test will depend on There are currently no data-driven recommen-
many factors, including the patients’ ability to exer- dations to guide the perioperative management
cise, the resting electrocardiogram (EKG), the clin- of stent-related antiplatelet therapy that are spe-
ical indication for the test, the patients’ body cific for the general thoracic surgical population.
habitus, and any history of prior revascularization. The surgeon must balance the risk for bleeding
Most patients in whom a stress test is indicated with the likelihood and potential consequences
should undergo an exercise stress EKG (unless un- of coronary stent thrombosis. There is a stent-
able to exercise). Cardiopulmonary exercise testing specific (BMS or DES) correlation between the
duration the stent has been in place and the likeli-
may be considered for patients with unknown func-
tional status. A low anaerobic threshold with a VO2 hood of perioperative stent thrombosis. The Amer-
max less than 10 mL O2/kg/min is associated with ican College of Chest Physicians (ACCP) has
increased cardiovascular complications including recommended that patients needing an operation
MI.1,19 If patients have an underlying EKG abnor- within 6 weeks of a placement of a BMS or within
mality (eg, left bundle branch block), are paced, or 6 months to 1 year of DES placement should have
have structural cardiac abnormalities, then stress antiplatelet therapy continued, especially for low-
test with imaging is recommended. and intermediate-risk of bleeding.26 Other factors
the surgeon must consider include surgical chal-
Cautions with stress test and bronchospasm Patients lenges that could increase the bleeding risk (exten-
with significant bronchospasm may be at sive prior surgery, prior pleurodesis, and so forth),
increased risk during pharmacologic stress coronary anatomy that could affect the conse-
testing. Intravenous dipyridamole and adenosine quences of stent thrombosis (stent location, collat-
have been shown to cause bronchospasm in pa- erals, and so forth), and the options for nonsurgical
tients with asthma and chronic obstructive pulmo- treatment (eg, stereotactic radiosurgery for lung
nary disease (COPD).20–22 Although the ACCF’s cancer). Fig. 2 provides a flowchart regarding the
2013 guidelines discourage the use of adenosine management of dual antiplatelet therapy for pa-
as a pharmacologic agent in stress testing patients tients who have received a DES.25
Cardiovascular Complications 375
Table 3
Antiplatelet medications and characteristics
Time to Recover
Platelet Function
Mechanism of After Drug Platelet Administration
Drug Action Half-life Withdrawal Inhibition Route
Aspirin Cox-1 inhibition 12–20 min 4d Irreversible Oral
Clopidogrel P2Y12 receptor 7–9 h 7–10 d Irreversible Oral
(Plavix) inhibition
Prasugrel P2Y12 receptor 7h 2–3 d Irreversible Oral
(Effient) inhibition
Ticagrelor P2Y12 and (partly) 7–9 h 3–4 d Reversible Oral
(Brilinta) P2Y1 receptor
inhibition
Cangrelor ATP analogue 3–6 min Rapid (minutes to Reversible IV
(Kengrexal) with a high hours)
affinity for the
P2Y12 receptor
Abciximab Glycoprotein IIb/ 10–15 min 12 h Reversible IV
(ReoPro) IIIa receptor
inhibitor
Eptifibatide Glycoprotein IIb/ 2.5 h 2–4 h Reversible IV
(Integrilin) IIIa receptor
inhibitor
Tirofiban Glycoprotein IIb/ 2h 2–4 h Reversible IV
(Aggrastat) IIIa receptor
inhibitor
Holding antiplatelet therapy bridge antiplatelet therapy for coronary stents with
If the surgeon elects to hold antiplatelet therapy, the shorter-acting agents (eg, low-molecular-weight
duration of the preoperative time interval required heparin).26 The antiplatelet therapy should be re-
for platelet function to normalize varies according started after the surgery has taken place as soon
to the half-life of the particular antiplatelet agent as possible (once the surgeon is comfortable that
(see Table 3). At present, there is no indication to the bleeding risk is no longer prohibitive).27
Performing surgery with patients on pathology (eg, coronary artery disease) and the
antiplatelet therapy risk of MI.
There are numerous reports of performing thoracic If an MI is highly suspected (and massive hemor-
procedures without discontinuing antiplatelet ther- rhage is not suspected), patients should be given
apy (lobectomy,28,29 esophagectomy29,30). It is aspirin 325 mg and morphine and should be evalu-
also the authors’ experience that with a bit more ated by cardiology for possible catheterization and
attention to hemostasis, it is typically possible to coronary intervention. The clinical impact of a peri-
safely operate without discontinuing antiplatelet operative MI may be highly variable, depending on
therapy (without increasing transfusion require- patients’ starting ventricular function and the terri-
ments or other bleeding complications). However, tory involved in the event. The decision of whether
if the surgeon performs a thoracic procedure or not to take patients for percutaneous interven-
without discontinuing antiplatelet therapy, the sur- tion is an individual one recognizing any percuta-
geon must be aware of the potential reversibility of neous coronary intervention is likely to require
specific antiplatelet agents in the event of signifi- antiplatelet therapy (which has certainly been
cant bleeding (see Table 3). done successfully after thoracic surgical proce-
dures37). Similar to the general acute coronary syn-
drome patient population (eg, not just patients
Prophylaxis with Beta-Blockade
recovering from surgery), the time to revasculariza-
Several factors in the postoperative setting (pain, tion is crucial and must be weighed against the risk
anemia, fluid volume shifts, and so forth) may in- of perioperative bleeding.5,13 For this reason, the
crease cardiac work and perfusion demands. use of fibrinolytics is not commonly used for post-
Beta-blockers may blunt some of the effects of post- operative MIs. The use of beta-blockers, angio-
operative cardiac stimulation,31–33 which has tensin-converting enzyme inhibitors, and aspirin
motivated the empirical use of perioperative beta- are recommended for all patients with an MI and
blockers to reduce cardiac events. At this time, there low ejection fraction (per the ACC/AHA’s guidelines
is considerable debate over the role of beta- for ST segment elevation MI and unstable angina5).
blockade in the perioperative setting to prevent
MI.34–36 In 2006, Lai and colleagues32 illustrated Cardiac Herniation
the use of beta-blockers in elderly esophagectomy Cardiac herniation through a defect in the pericar-
patients prevented postoperative cardiac events. dium is a very rare, potentially fatal complication of
Several subsequent meta-analyses illustrated a thoracic procedures that involves a resection of a
decrease in postoperative MI with beta-blockers portion of the pericardium. This complication has
started 1 day before surgery. However, they noted most commonly been described following a pneu-
an increase in stroke, hypotension, bradycardia, monectomy that includes a pericardial resection.
and even death.34 At the time of this article, it is the The average size of the defect associated with
position of the ACC and AHA (per the 2014 guide- this complication is 4 cm.38–47 The herniated
lines) that there are insufficient data regarding the ventricle can become constricted by the surround-
safety and efficacy of initiating (de novo) beta- ing pericardium and become ischemic. The mor-
blockers in the perioperative period and further tality for this complication reaches 50% for
research is needed. Patients chronically taking recognized cases and 100% for unrecognized
beta-blockers (before their surgical evaluation) cases.48 For these reasons, surgeons should
should continue through surgery. consider closing a pericardial defect created dur-
ing a pneumonectomy. Fenestrated Gore-Tex
Treatment of Postoperative Myocardial (W.L. Gore and Associates, Inc, Newark, DE) has
Infarction been frequently used for this purpose if the defect
is too large to be closed primarily.49 Patients pre-
The presentation of perioperative MI may be senting with suspected cardiac herniation require
masked by chest pain from incisions and limitations urgent evaluation with a low threshold to reoperate
in the patients’ activity level. The diagnosis is typi- to prevent further cardiovascular collapse.
cally made secondary to a change in clinical status
that is accompanied by EKG changes. Within the dif-
DEEP VEIN THROMBOSIS/PULMONARY
ferential diagnosis are acute hemorrhage and pro-
EMBOLISM
found hypoxic respiratory failure both of which, if
Introduction
severe, could cause cardiac ischemia (which may
or may not be distinguishable by EKG). It must be Venous thromboembolic events (VTEs) are impor-
kept in mind that a negative stress test does not tant considerations after general thoracic surgery.
eliminate patients’ chances of having coronary Not only are many thoracic patients predisposed
Cardiovascular Complications 377
to develop VTEs (hypercoagulable from cancer, reach the floor, compliance with pneumatic
trauma, lack of mobility) but the patients’ ability compression devices can vary significantly
to tolerate a significant event after thoracic surgery (53%–75%),64 which can affect the effectiveness
may also be affected by reductions in pulmonary of this strategy. Using both mechanical and phar-
reserve. For example, in one study the 18-month macologic prophylaxis together seems to further
survival after pneumonectomy complicated by decrease the incidence of DVT after thoracic sur-
VTE was only 13%.50 gery operations and is currently recommended
Similar to the previous section, the manage- by the ACCP.64
ment of VTEs typically involves anticoagulation
(and bleeding risk). This treatment can be particu- Pharmacologic
larly challenging when the cause of patients’ Pharmacologic prophylaxis by way of low-dose
status change (the underlying diagnosis) is un- unfractionated heparin (LDUH) or low-molecular-
clear, as empirical treatment can lead to life- weight heparin (LMWH) decreases the incidence
threatening hemorrhage. Therefore, great of DVT and VTE in thoracic surgical pa-
emphasis must be placed on preventing and diag- tients.51,64,70 The risk of surgical bleeding from
nosing VTEs in patients recovering from general pharmacologic prophylaxis seems to be out-
thoracic surgery. weighed by the far greater benefit of the VTE
reduction.64,70,71 There does not seem to be a sig-
Incidence in Thoracic Surgery nificant difference in effectiveness of prophylactic
The reported incidence of VTE after general low dose unfractionated heparin or LWMH.64,72
thoracic surgical procedures is on the order of Several studies have attempted to optimize
5%. There is significant variability in VTE incidence pharmacologic prophylaxis by varying the specific
(1%–26%) between different procedures, various regimens. For example, several studies have
surgical approaches, and the use of prophy- examined the dose of pharmacologic prophylaxis
laxis.51–54 The Society of Thoracic Surgery (STS) given. There did not seem to be a difference be-
databases report a 1.0% VTE incidence after tween 10,000 units per day and 15,000 units per
lung cancer surgery2 and 2.5% VTE incidence af- day of LDUH.70,73 The dosage for LWMH used
ter esophagectomy.3 In addition to the traditional for VTE prophylaxis is 40 mg/d. Although weight-
risk factors for VTE (eg, Virchow triad of stasis, hy- based LMWH prophylaxis (0.5 mg/kg/d) has
percoagulable state, and endothelial injury), the been found to be efficacious in preventing VTE in
use of induction chemotherapy,53,55,56 the use of obese bariatric patients,74,75 the routine use of
a thoracotomy,57–60 and smoking cessation50,61,62 weight-based LMWH for thoracic surgery patients
have all been suggested to increase the incidence has not been fully studied.
of VTEs after general thoracic surgery. At one time The duration of pharmacologic prophylaxis has
video-assisted thoracic surgery (VATS) was also been studied, and there does not seem to
thought to increase the risk for VTE by restricting be a benefit to continuing prophylaxis after pa-
venous return63; however, this observation has tients have been discharged. Extended VTE
not been validated by the large database studies prophylaxis (longer than postoperative hospital
that have examined this question.58 stay) has only been shown beneficial for patients
undergoing major abdominal or gynecologic sur-
Prophylaxis gery but has not been appreciated in thoracic
surgery.64
General thoracic surgery patients benefit from VTE The guidelines set forth by the ACCP are sum-
prophylaxis. The ACCP’s guidelines are specific to marized later. In addition to these guidelines, early
the thoracic surgical population64 for VTE prophy- postoperative ambulation is imperative to help
laxis, which include early postoperative mobiliza- prevent DVT and VTE.65
tion65,66 and mechanical66 and pharmacologic Guidelines for VTE prophylaxis for thoracic sur-
prophylaxis.64,67 gery patients (adopted from the ACCP and the UK
Royal College of Physicians)62,64 are as follows:
Mechanical
Mechanical prophylaxis via pneumatic compres- Start mechanical VTE prophylaxis at admis-
sion devices or stockings is an effective strategy sion before surgery (intermittent pneumatic
to reduce VTEs (reducing the rate of DVT by up compression devices elastic stockings).
to 60%).68 It is important to place compression de- Patients with low-risk for bleeding add phar-
vices or stocking at admission and in the operating macologic prophylaxis: LDUH (10,000–
room before surgery (with many studies advo- 15,000 units per day subcutaneous) or
cating before anesthesia).64,66,69 Once patients LMWH (40 mg subcutaneous daily).
378 Keshava & Boffa
Patients with high-risk of bleeding (eg, pleural of the significant mortality associated with
pneumonectomy) add mechanical prophy- untreated VTEs.
laxis only and add pharmacologic prophylaxis
(LDUH or LMWH) when the risk of bleeding Diagnosis
diminishes. The diagnostic evaluation for suspected PE should
Continue mechanical and pharmacologic pro- include a CT scan of the chest (specific PE proto-
phylaxis until patients are ambulatory (w5– col) if patients’ renal function permits. A ventilation
7 days). perfusion scan may be helpful in patients in which
Inferior vena-cava filters a PE is suspected but who cannot undergo CT
The primary role of inferior vena-cava (IVC) filters in scanning (because of impaired renal function or
the surgical population has been to prevent PEs contrast allergy with insufficient time to premedi-
(new or additional) in patients diagnosed with cate patients). The downside of ventilation perfu-
VTEs in whom anticoagulation was contraindi- sion scanning is the significant variability
cated.50,55 IVC filters enable anticoagulation to between interpreting radiologists (resulting in vari-
be held perioperatively in thoracic surgery patients able sensitivity and specificity). There are high
and are often placed in patients who are diag- numbers of indeterminate scans and nondiagnos-
nosed with a VTE before surgery. Patients should tic studies (particularly in patients with COPD77,78).
have a retrievable IVC filter placed to allow the filter A transthoracic or transesophageal echocardio-
to be removed to prevent filter-related complica- gram may also be used in patients who are either
tions (w3–4 months in thoracic surgery unable to have a CT scan or in patients with acute
patients).64,76 hemodynamic compromise in whom the diagnosis
is unclear. Right heart strain is the main diagnostic
parameter considered for these patients, but it is
Clinical Presentation also possible at times to detect thrombus within
The diagnosis of VTE can be challenging in the the heart or the pulmonary artery. The main diag-
thoracic surgical population. The classic signs nostic imaging modality for a DVT is a duplex ultra-
and symptoms of a VTE (eg, asymmetric leg sound, which can be used to assess for a source
swelling, acute tachycardia, acute hypoxia) of subsequent PEs.
certainly should prompt a consideration for VTE
Treatment
evaluation. However, general thoracic surgery pa-
tients may experience events that mirror the Prompt diagnosis and treatment is crucial for
classic presentation of a VTE. For example, favorable outcomes for patients with VTE. PE re-
thoracic patients often have tachycardia from peri- quires the most urgent treatment (because of the
carditis, pain, postoperative dysrhythmia, and so risk of hemodynamic collapse). Treatment of any
forth. Thoracic patients may be hypoxic from a VTE must reconcile bleeding risk (with anticoagu-
mucous plug, splinting from incompletely lation) and the risk of untreated VTE.
controlled pain, a COPD flare, aspiration pneumo- The patients’ hemodynamic status is a critical
nitis, or pneumonia. That being said, clinicians determinant of the initial treatment of PE (Fig. 3).
should have a relatively low threshold to pursue Unstable patients (those in shock or profoundly
the VTE diagnosis (focused imaging evaluation) hypotensive) are unlikely to be transported for im-
in light of these other potential diagnoses because aging studies; therefore, a suspected PE diagnosis
is unlikely to be confirmed. A portable echocardio- major limitation to this form of treatment in the
gram demonstrating right heart strain could sup- postoperative setting.
port the diagnosis of PE, but treatment should
not be held for this study if suspicion is high in he- Discharge and Follow-up
modynamically compromised patients. The clini-
The ACCP recommends at least 3 months of anti-
cian must weigh the risks of bleeding from
coagulation for patients with an isolated single PE
empirical anticoagulation (therapeutic intravenous
or DVT following surgery.83 In patients with
heparin or subcutaneous LMWH) against the high
advanced cancer, there is some evidence to
mortality from an untreated PE in unstable pa-
continue anticoagulation indefinitely.83 Recent
tients. If patients’ hemodynamics do not improve
studies by Agnelli and colleagues84,85 have shown
with resuscitation, then consideration may be
a benefit in long-term anticoagulation for prevent-
given to thrombolysis (see later discussion) or
ing further VTE in patients receiving chemotherapy
(rarely) surgical embolectomy.
or with stage 4 cancer particularly in lung and
For hemodynamically stable patients, anticoa-
gastrointestinal malignancies. There are no data
gulation is the cornerstone of treatment. Anticoa-
for empirical follow-up imaging for treated PE or
gulation after a PE effectively reduces mortality
DVT unless clinically indicated. (Although there is
and recurrent embolic events. Acute treatment be-
minimal risk associated with an extremity ultra-
gins with unfractionated heparin or LMWH over the
sound for DVT, radiation exposure should be
first 5 to 10 days. Subsequent treatment may
considered if clinicians are inclined to follow PE re-
include an oral vitamin K antagonist (eg, warfarin)
absorption in asymptomatic patients.)
or continuation of LMWH as an outpatient.69 No
difference has been appreciated with the treat-
Pulmonary Artery Thrombosis After Vascular
ment of PE with unfractionated heparin and
Sleeve Resection
LMWH. The advantage of intravenous (IV) unfrac-
tionated heparin drip is the ability to quickly inter- A vascular sleeve resection of the PA carries a
rupt and restore anticoagulation for necessary theoretic risk of artery thrombosis secondary to
procedures or signs of hemorrhage. The absorp- the period of clamping and trauma to the artery.
tion of subcutaneous (SQ) injections of LMWH Some investigators estimate this to be around
may be variable in morbidly obese patients, which 2%,86,87 whereas others have not experienced
could also favor IV unfractionated heparin. For pa- an increase incidence of thrombosis after vascular
tients being treated managed with vitamin K an- sleeve resection.88,89
tagonists, a heparin bridge is used to prevent The use of prophylactic heparin in vascular
any complications from the initial procoagulant sleeve resections has been debated, with some
phase of vitamin K antagonism. centers using routine anticoagulation,86 whereas
Newer anticoagulants, such as factor Xa inhib- others not routinely using prophylactic anticoagu-
itors (rivaroxaban, apixaban, edoxaban) and lation.87–89 At this point in time, there are no firm
direct thrombin inhibitors (dabigatran), have recommendations; but the authors currently
been shown efficacious for the treatment of PE administer 5000 units of IV heparin before clamp-
in nonsurgical patients.79–81 Table 4 describes ing the PA for a vascular sleeve resection (as
common and newer anticoagulants used in the long as there are no contraindications).
treatment of VTE, including risks, benefits, and The diagnosis of a PA thrombosis may be de-
reversing agents. There is a paucity of data layed, as the initial clinical indicators (eg, lobar
regarding the safety and efficacy of using these opacity on chest radiograph, fever, increase in
agents to treat PE occurring after thoracic sur- white blood cell count) are nonspecific and could
gery. The lack of reversal agents may lead to represent mucus plugging or pneumonia. An IV
greater blood loss should a bleeding complication contrast CT scan is the most common route to
arise. establish PA thrombosis.86–88 The management
depends on the patients’ status and the suspicion
of lobar infarct. If patients also underwent sleeve
Thrombolysis resection of the airway, the bronchial arterial sup-
Thrombolysis restores pulmonary perfusion faster ply to the lung has almost certainly been elimi-
than anticoagulation with heparin alone. The nated, resulting in a pulmonary infarct. Making a
maximum benefit occurs if patients undergo determination of pulmonary infarct is challenging,
thrombolysis within 48 hours of presentation. and bronchoscopy may not accurately reflect the
Streptokinase, urokinase, and tissue plasminogen status of the parenchyma. If the suspicion for
activator have been used with a reduction in mor- infarct is high, then reoperation and inspection of
tality and recurrent PE.82 The risk of bleeding is the the lung is indicated. If the remaining lobe is not
380
Keshava & Boffa
Table 4
Types of pharmacologic anticoagulation treatment of venous thromboembolism and atrial fibrillation
Abbreviations: Afib, atrial fibrillation; aPTT, activated partial thromboplastin time; CrCl, creatine clearance; FFP, fresh frozen plasma; HIT, heparin induced thrombocytopenia; INR,
international normalized ratio; n/a, not applicable; postop, postoperatively; UFH, unfractionated heparin.
a
Fixed 5 fixed dosing based on specific manufacturer guidelines and creatine clearance.
b
High creatine clearance 5 30–50 mL/min.
c
May need longer duration of anticoagulation for patients with cancer.
Adapted from Frendl G, Sodickson AC, Chung MK, et al. 2014 AATS guidelines for the prevention and management of perioperative atrial fibrillation and flutter for thoracic sur-
gical procedures. J Thorac Cardiovasc Surg 2014;148:e185; and Konstantinides SV, Torbicki A, Agnelli G, et al. 2014 ESC guidelines on the diagnosis and management of acute pul-
Cardiovascular Complications
monary embolism. Eur Heart J 2014;35:3054.
381
382 Keshava & Boffa
viable, then resection should take place to prevent Prophylaxis for Postoperative Atrial
further complications. Fibrillation
Several randomized trials and multiple meta-
ATRIAL ARRHYTHMIAS analyses have evaluated the efficacy of atrial fibril-
lation prophylaxis after thoracic surgery, with
Atrial arrhythmias are among the more common
mixed results. The STS’ 2012 and the American
complications after thoracic surgery; although
Association of Thoracic Surgeons’ (AATS) 2014
rarely life threatening, they may cause a significant
guidelines for the prevention and management of
derailment of an otherwise smooth recovery. The
atrial fibrillation and flutter for thoracic surgery pro-
incidence varies significantly by the procedure
cedures were based on these data and give
performed (Table 5). Several investigators have
recommendations for prophylaxis outlined
evaluated the risk factors for postoperative atrial
next.90,98–103
arrhythmias after thoracic surgery with mixed re-
sults. Some factors that have been identified to in-
crease the risk of postoperative atrial fibrillation Beta-blockers
include preoperative atrial fibillration,90 increased Class 1 evidence is recognized by both the STS’
age,91–95 male sex,91,96 and coronary artery dis- and AATS’ guidelines for continuing patients on
ease/congestive heart failure.94,97–99 preoperative beta-blocker before thoracic surgery
to prevent potential beta-blocker withdrawal syn-
drome and increased risk of postoperative atrial
Causes of Atrial Fibrillation
fibrillation. The administration of prophylactic
Most atrial fibrillation after general thoracic surgery beta-blockers for patients who are beta-blocker
is presumed to be the result of surgical stress, naı̈ve is unclear. There is some evidence pointing
anatomic manipulations, and other deviations to prophylactic beta-blocker use and the preven-
from patients’ normal physiologic state. However, tion of postoperative atrial fibrillation; however,
there are important precipitating events in the gen- the incidence of postoperative hypotension,
eral thoracic population that must be considered in bradycardia, and stroke-related mortality is
patients with new-onset atrial fibrillation, such as high.1,100,102 For this reason, prophylactic beta-
pneumothorax, hemothorax, PE, aspiration, anas- blockers for beta-blocker–naı̈ve patients is not
tomotic leak, and MI. Although atrial fibrillation currently recommended by the STS or the
carries risk to patients’ recovery, failure to recog- AATS.98,99 The STS’ guidelines state that postop-
nize a precipitating complication could be erative administration of a new beta-blocker can
devastating. be used to prevent postoperative atrial
Table 5
Risk of postoperative atrial fibrillation after thoracic surgery
patients with substantial preexisting lung disease, and other patients who cannot undergo a TEE, 4
primarily because of the risk of pulmonary toxicity to 6 weeks of anticoagulation before cardioversion
with amiodarone that may occur in up to 10% of is recommended.99
patients.98
Rate-control agents should be continued for 4 to Anticoagulation
8 weeks after discharge if the heart rate is in sinus Anticoagulation for postoperative atrial fibrillation
rhythm. If patients are still in rate-controlled atrial is indicated to prevent thrombus formation in the
fibrillation, cardiology consultation is warranted left atrium and subsequent thromboembolism
for further assessment and possible cardioversion (eg, cerebrovascular accident). One predictive
for rhythm control.98,99 model commonly used to assess the risk for
thromboembolic events from patients in atrial
Cardioversion fibrillation is called the CHA2DS2-VASc score
Within the first 48 hours of onset, hemodynamical- (congestive heart failure, hypertension, age 75
ly unstable patients should be cardioverted years, diabetes mellitus, prior stroke, vascular dis-
without the need for anticoagulation. ease, age 65–74 years, female sex category). The
For patients in atrial fibrillation for more than CHA2DS2-VASc score is a clinical prediction tool
48 hours, cardioversion, chemical cardioversion estimating the annual risk of stroke in patients
with flecainide or amiodarone, or DC cardioversion with nonrheumatic atrial fibrillation using points
is reasonable in hemodynamically stable patients. allocated to known risk factors. A higher
However, the STS and AATS recommend obtain- CHA2DS2-VASc score indicates an increasing
ing a transesophageal echocardiogram (TEE) to risk of stroke (Tables 6 and 7). The AATS’ 2014
rule out left atrial thrombus. If a thrombus is iden- recommendations for anticoagulation depend on
tified, patients should be rate controlled and anti- the timing of postoperative atrial fibrillation, the pa-
coagulated for 4 to 6 weeks before attempting tients’ CHA2DS2-VASc score, the duration of post-
cardioversion.98,99 For esophagectomy patients operative atrial fibrillation, and the risk of
Table 6
Stroke risk stratification using the CHA2DS2-VASc score
Abbreviations: LV, left ventricular; PVD, peripheral vascular disease; TIA, transient ischemic attack.
a
Adjusted for patients receiving aspirin to give annual risk of stroke without therapy.
Adapted from Lip GYH. Implications of the CHA2DS2-VASc and HAS-BLED scores for thromboprophylaxis in atrial fibril-
lation. Am J Med 2011;124:112; with permission.
386 Keshava & Boffa
Table 7
Commonly used rate/rhythm control agents for postoperative atrial fibrillation
Mechanism of Rate/Rhythm
Agent Action Dosage Control Duration Side Effects
Metoprolol Beta-blocker 2.5–5.0 mg Rate 4–8 wk Bradycardia,
IV 3 doses hypotension,
Followed by bronchospasm,
12.5–100 mg heart failure
oral every 6– exacerbation
12 h
Diltiazem Calcium channel 0.25 mg/kg IV Rate 4–8 wk Hypotension,
blocker loading dose bradycardia,
over 2 min, then heart failure
5–15 mg/h IV exacerbation
continuous
infusion
Followed by 30–
60 mg every 6–
8h
Amiodarone Class III 150 mg/10 min Rate/Rhythm 4–8 wk Bradycardia, QTc
antiarrhythmic IV; then prolongation,
1 mg/min for pulmonary
6 h; then toxicity, thyroid
0.5 mg/min IV toxicity,
for 18 h pancreatitis
Followed by
400 mg twice
oral daily
Digoxin Na1/K 1 ATPase Total IV dose of Rate 4–8 wk Nausea, vomiting,
in myocardium 1.25–1.5 mg in anorexia,
the first 24 h confusion, AV
Followed by block
0.125–0.5 mg
oral daily
Flecainide Class IC 200–300 mg Rhythm 4–8 wk Dizziness, blurred
antiarrhythmic single oral dose vision, sinus
Followed by 50– bradycardia, AV
150 mg oral block,
every 12 h contraindicated
with low left
ventricular
function or
coronary artery
disease
Procainamide Class IA 20–50 mg/min Rhythm Until Afib Hypotension, QTc
antiarrhythmic IV continuous, terminated prolongation,
until Afib torsades de
terminated pointes,
OR 100 mg IV contraindicated
every 5 min un- with low left
til Afib ventricular
terminated function
postoperative bleeding.99 For patients with atrial ventricular systolic pressure greater than 70 mm
fibrillation of less than 48 hours duration, anticoa- Hg or a mean PA pressure of greater than or equal
gulation is typically not recommended (see Fig. 4). to 50 mm Hg or right heart catheterization showing
For patients with atrial fibrillation extending PA hypertension.109,111
beyond 48 hours, anticoagulation should be For patients with severe pulmonary hyperten-
considered (particularly if there is consideration sion, even procedures ordinarily considered to be
for cardioversion). Unfractionated heparin, low risk may be dangerous. For example,
LMWH, and warfarin result in equivalent risk numerous patients with pulmonary hypertension
reduction of thrombosis.99 Warfarin and LMWH have been described to have severe bleeding
are the mainstay for long-term anticoagulation, from wedge resections.18 This postoperative
with LMWH having restrictions on patients with bleeding potentially reflects the hemostatic capa-
renal insufficiency and its high cost. bility of the staples, which are not designed for
Table 4 describes both common and some of elevated pressures. Although no specific guide-
the newer anticoagulants used for atrial fibrillation lines exist for performing thoracic surgery on
along with risks, benefits, and reversing agents. patients with pulmonary hypertension, a multidis-
Recent guidelines from the AATS consider the ciplinary approach with cardiology, anesthesi-
data in support of these newer agents to warrant ology, and thoracic surgery is crucial for optimum
a class II recommendation as an alternative to diagnosis, treatment, and management.
warfarin.99 These agents should be used with
caution in the perioperative setting as their antico- SUMMARY
agulation effects are not immediately reversible.
Anticoagulation should be continued for 4 to Cardiovascular events after thoracic surgery can
6 weeks once patients have returned to sinus result in increased morbidity, mortality, length of
rhythm. Longer anticoagulation may be needed stay, and increased overall cost. The prevention
for patients with a higher CHA2DS2-VASc score. of postoperative cardiovascular complications is
For patients that remain in atrial fibrillation, a TEE an area of intense study, and the body of evidence
may be performed after 4 to 6 weeks of anticoagu- guiding clinicians continues to grow. Early diag-
lation to rule out atrial thrombus. If no thrombus is nosis and management of cardiovascular events
identified, patients may be DC cardioverted to si- can minimize the consequences of these
nus rhythm. Patients who fail electrical and phar- complications.
macologic cardioversion and remain in atrial
fibrillation should remain anticoagulated to reduce REFERENCES
the risk of a cerebral vascular accident.
1. Fleisher LA, Fleischmann KE, Auerbach AD, et al.
2014 ACC/AHA guideline on perioperative cardio-
OTHER CONSIDERATIONS
vascular evaluation and management of patients
Pulmonary Hypertension and Lung Surgery
undergoing noncardiac surgery: a report of the
Preoperative pulmonary hypertension is associ- American College of Cardiology/American Heart
ated with an increase in morbidity and mortality Association Task Force on Practice Guidelines.
after noncardiac thoracic surgery.108,109 Hemody- J Am Coll Cardiol 2014;64:e77–137.
namic compromise has been suggested to be the 2. Boffa DJ, Allen MS, Grab JD, et al. Data from the
result of changes in intrathoracic pressures, lung Society of Thoracic Surgeons general thoracic sur-
volumes, and oxygenation that can acutely in- gery database: the surgical management of pri-
crease pulmonary vascular resistance and lead mary lung tumors. J Thorac Cardiovasc Surg
to right ventricular failure.109,110 Currently, specific 2008;135:247–54.
guidelines do not exist for general thoracic surgery 3. Wright CD, Kucharczuk JC, O’Brien SM, et al. Pre-
in patients with pulmonary hypertension because dictors of major morbidity and mortality after
of the paucity of data. esophagectomy for esophageal cancer: a Society
Symptoms for pulmonary hypertension are of Thoracic Surgeons general thoracic surgery
nonspecific (eg, dyspnea). Echocardiography is database risk adjustment model. J Thorac Cardio-
an important screening tool if pulmonary hyperten- vasc Surg 2009;137:587–96 [discussion: 596].
sion is suspected, yet right heart catheterization is 4. Marret E, Miled F, Bazelly B, et al. Risk and protec-
the gold standard for diagnosis and assessment of tive factors for major complications after pneumo-
vasodilator response in pulmonary hypertension. nectomy for lung cancer. Interactive Cardiovasc
Specific risk factors for surgical complications Thorac Surg 2010;10:936–9.
associated with pulmonary hypertension include 5. Fleisher LA, Beckman JA, Brown KA, et al. ACC/AHA
an echocardiogram demonstrating a right 2007 guidelines on perioperative cardiovascular
388 Keshava & Boffa
evaluation and care for noncardiac surgery: a report multimodality appropriate use criteria for the detec-
of the American College of Cardiology/American tion and risk assessment of stable ischemic heart dis-
Heart Association Task Force on Practice Guidelines ease: a report of the American College of Cardiology
(Writing Committee to Revise the 2002 Guidelines on Foundation Appropriate Use Criteria Task Force,
Perioperative Cardiovascular Evaluation for Noncar- American Heart Association, American Society of
diac Surgery): developed in collaboration with the Echocardiography, American Society of Nuclear Car-
American Society of Echocardiography, American diology, Heart Failure Society of America, Heart
Society of Nuclear Cardiology, Heart Rhythm Soci- Rhythm Society, Society for Cardiovascular Angiog-
ety, Society of Cardiovascular Anesthesiologists, raphy and Interventions, Society of Cardiovascular
Society for Cardiovascular Angiography and Inter- Computed Tomography, Society for Cardiovascular
ventions, Society for Vascular Medicine and Biology, Magnetic Resonance, and Society of Thoracic Sur-
and Society for Vascular Surgery. Circulation 2007; geons. J Nucl Cardiol 2014;21:192–220.
116:e418–99. 15. Ovrehus KA, Jasinskiene J, Sand NP, et al. Coro-
6. Rentz J, Bull D, Harpole D, et al. Transthoracic nary calcification among 3477 asymptomatic and
versus transhiatal esophagectomy: a prospective symptomatic individuals. Eur J Prev Cardiol 2015.
study of 945 patients. J Thorac Cardiovasc Surg [Epub ahead of print].
2003;125:1114–20. 16. Loffroy R, Bernard S, Serusclat A, et al. Noninva-
7. Bailey SH, Bull DA, Harpole DH, et al. Outcomes sive assessment of the prevalence and character-
after esophagectomy: a ten-year prospective istics of coronary atherosclerotic plaques by
cohort. Ann Thorac Surg 2003;75:217–22 [discus- multidetector computed tomography in asymptom-
sion: 222]. atic type 2 diabetic patients at high risk of signifi-
8. Lee TH, Marcantonio ER, Mangione CM, et al. Deri- cant coronary artery disease: a preliminary study.
vation and prospective validation of a simple index Arch Cardiovasc Dis 2009;102:607–15.
for prediction of cardiac risk of major noncardiac 17. Becker A, Leber AW, Becker C, et al. Predictive
surgery. Circulation 1999;100:1043–9. value of coronary calcifications for future cardiac
9. Devereaux PJ, Goldman L, Cook DJ, et al. Periop- events in asymptomatic patients with diabetes mel-
erative cardiac events in patients undergoing litus: a prospective study in 716 patients over 8
noncardiac surgery: a review of the magnitude of years. BMC Cardiovasc Disord 2008;8:27.
the problem, the pathophysiology of the events 18. Ross AF, Ueda K. Pulmonary hypertension in
and methods to estimate and communicate risk. thoracic surgical patients. Curr Opin Anaesthesiol
CMAJ 2005;173:627–34. 2010;23:25–33.
10. Brunelli A, Varela G, Salati M, et al. Recalibration of 19. Hall A, Older P. Cardiopulmonary exercise testing
the revised cardiac risk index in lung resection accurately predicts risk of major surgery including
candidates. Ann Thorac Surg 2010;90:199–203. esophageal resection: letter 1. Ann Thorac Surg
11. Devereaux PJ, Xavier D, Pogue J, et al. Character- 2009;87:670–1.
istics and short-term prognosis of perioperative 20. Husain Z, Palani G, Cabrera R, et al. Hemody-
myocardial infarction in patients undergoing namic response, arrhythmic risk, and overall
noncardiac surgery: a cohort study. Ann Intern safety of regadenoson as a pharmacologic stress
Med 2011;154:523–8. agent for myocardial perfusion imaging in chronic
12. Kamel H, Johnston SC, Kirkham JC, et al. Associa- obstructive pulmonary disease and bronchial
tion between major perioperative hemorrhage and asthma patients. Int J Cardiovasc Imaging
stroke or Q-wave myocardial infarction. Circulation 2012;28:1841–9.
2012;126:207–12. 21. Leitha T, Gwechenberger M, Falger-Banyai S. Does
13. Fihn SD, Blankenship JC, Alexander KP, et al. 2014 dyspnoea during dipyridamole cardiac stress
ACC/AHA/AATS/PCNA/SCAI/STS focused update testing indicate bronchospasm and is the pretest
of the guideline for the diagnosis and management clinical history predictive of this side-effect? Eur J
of patients with stable ischemic heart disease: a Nucl Med 1995;22:1408–10.
report of the American College of Cardiology/Amer- 22. Balan KK, Critchley M. Is the dyspnea during aden-
ican Heart Association Task Force on Practice osine cardiac stress test caused by broncho-
Guidelines, and the American Association for spasm? Am Heart J 2001;142:142–5.
Thoracic Surgery, Preventive Cardiovascular 23. Fricke E, Esdorn E, Kammeier A, et al. Respiratory
Nurses Association, Society for Cardiovascular resistance of patients during cardiac stress testing
Angiography and Interventions, and Society of with adenosine: is dyspnea a sign of broncho-
Thoracic Surgeons. J Am Coll Cardiol 2014;64: spasm? J Nucl Cardiol 2008;15:94–9.
1929–49. 24. Reyes E, Loong CY, Wechalekar K, et al. Side ef-
14. Ronan G, Wolk MJ, Bailey SR, et al. ACCF/AHA/ASE/ fect profile and tolerability of adenosine myocardial
ASNC/HFSA/HRS/SCAI/SCCT/SCMR/STS 2013 perfusion scintigraphy in patients with mild asthma
Cardiovascular Complications 389
75. Jamal MH, Corcelles R, Shimizu H, et al. Thrombo- 88. Alifano M, Cusumano G, Strano S, et al. Lobectomy
embolic events in bariatric surgery: a large with pulmonary artery resection: morbidity, mortal-
multi-institutional referral center experience. Surg ity, and long-term survival. J Thorac Cardiovasc
Endosc 2015;29:376–80. Surg 2009;137:1400–5.
76. Sarosiek S, Crowther M, Sloan JM. Indications, 89. Shrager J, Lambright E, McGrath C, et al. Lobec-
complications, and management of inferior vena tomy with tangential pulmonary artery resection
cava filters: the experience in 952 patients at an without regard to pulmonary function. Ann Thorac
academic hospital with a level I trauma center. Surg 2000;70:234–9.
JAMA Intern Med 2013;173:513–7. 90. Amar D. Prevention and management of periopera-
77. Srivastava SD, Eagleton MJ, Greenfield LJ. Diag- tive arrhythmias in the thoracic surgical population.
nosis of pulmonary embolism with various imaging Anesthesiol Clin 2008;26:325–35, vii.
modalities. Semin Vasc Surg 2004;17:173–80. 91. Roselli EE, Murthy SC, Rice TW, et al. Atrial fibrilla-
78. Hartmann IJ, Hagen PJ, Melissant CF, et al. Diag- tion complicating lung cancer resection. J Thorac
nosing acute pulmonary embolism: effect of chronic Cardiovasc Surg 2005;130:438–44.
obstructive pulmonary disease on the performance 92. Amar D. Postthoracotomy atrial fibrillation. Curr
of D-dimer testing, ventilation/perfusion scintig- Opin Anaesthesiol 2007;20:43–7.
raphy, spiral computed tomographic angiography, 93. Mason DP, Marsh DH, Alster JM, et al. Atrial fibrilla-
and conventional angiography. ANTELOPE Study tion after lung transplantation: timing, risk factors,
Group. Advances in New Technologies Evaluating and treatment. Ann Thorac Surg 2007;84:1878–84.
the Localization of Pulmonary Embolism. Am J Re- 94. Ivanovic J, Maziak DE, Ramzan S, et al. Incidence,
spir Crit Care Med 2000;162:2232–7. severity and perioperative risk factors for atrial
79. Schulman S, Kearon C, Kakkar AK, et al. Dabiga- fibrillation following pulmonary resection. Interac-
tran versus warfarin in the treatment of acute tive Cardiovasc Thorac Surg 2014;18:340–6.
venous thromboembolism. N Engl J Med 2009; 95. Neragi-Miandoab S, Weiner S, Sugarbaker DJ.
361:2342–52. Incidence of atrial fibrillation after extrapleural
80. Lee AY, Bauersachs R, Janas MS, et al. CATCH: a pneumonectomy vs. pleurectomy in patients with
randomised clinical trial comparing long-term tin- malignant pleural mesothelioma. Interactive Cardi-
zaparin versus warfarin for treatment of acute ovasc Thorac Surg 2008;7:1039–42.
venous thromboembolism in cancer patients. 96. Vaporciyan AA, Correa AM, Rice DC, et al. Risk
BMC Cancer 2013;13:284. factors associated with atrial fibrillation after
81. Loffredo L, Perri L, Del Ben M, et al. New oral anti- noncardiac thoracic surgery: analysis of 2588 pa-
coagulants for the treatment of acute venous tients. J Thorac Cardiovasc Surg 2004;127:779–86.
thromboembolism: are they safer than vitamin K 97. Stougard J. Cardiac arrhythmias following pneu-
antagonists? A meta-analysis of the interventional monectomy. Thorax 1969;24:568–72.
trials. Intern Emerg Med 2015;10(4):499–506. 98. Fernando HC, Jaklitsch MT, Walsh GL, et al. The
82. Kameyama K, Huang CL, Liu D, et al. Pulmonary Society of Thoracic Surgeons practice guideline
embolism after lung resection: diagnosis and treat- on the prophylaxis and management of atrial fibril-
ment. Ann Thorac Surg 2003;76:599–601. lation associated with general thoracic surgery:
83. Guyatt GH, Akl EA, Crowther M, et al. Introduction executive summary. Ann Thorac Surg 2011;92:
to the ninth edition: antithrombotic therapy and pre- 1144–52.
vention of thrombosis, 9th ed: American College of 99. Frendl G, Sodickson AC, Chung MK, et al. 2014
Chest Physicians evidence-based clinical practice AATS guidelines for the prevention and manage-
guidelines. Chest 2012;141:48s–52s. ment of perioperative atrial fibrillation and flutter
84. Agnelli G, George DJ, Kakkar AK, et al. Semulo- for thoracic surgical procedures. J Thorac Cardio-
parin for thromboprophylaxis in patients receiving vasc Surg 2014;148:e153–93.
chemotherapy for cancer. N Engl J Med 2012; 100. Flynn BC, Vernick WJ, Ellis JE. b-Blockade in the
366:601–9. perioperative management of the patient with car-
85. Agnelli G, Verso M. Thromboprophylaxis during diac disease undergoing non-cardiac surgery. Br
chemotherapy in patients with advanced cancer. J Anaesth 2011;107(Suppl 1):i3–15.
Thromb Res 2010;125(Suppl 2):S17–20. 101. Bagshaw SM, Galbraith PD, Mitchell LB, et al. Pro-
86. Ma Q, Liu D, Guo Y, et al. Surgical techniques and phylactic amiodarone for prevention of atrial fibrilla-
results of the pulmonary artery reconstruction for tion after cardiac surgery: a meta-analysis. Ann
patients with central non-small cell lung cancer. Thorac Surg 2006;82:1927–37.
J Card Surg 2013;8:219. 102. Bouri S, Shun-Shin MJ, Cole GD, et al. Meta-anal-
87. Rendina E, Venuta F. Sleeve resection and pros- ysis of secure randomised controlled trials of
thetic reconstruction of the pulmonary artery for beta-blockade to prevent perioperative death in
lung cancer. Ann Thorac Surg 1999;68:995–1001. non-cardiac surgery. Heart 2014;100:456–64.
392 Keshava & Boffa
103. Cook RC, Yamashita MH, Kearns M, et al. Prophy- 107. Amar D, Zhang H, Heerdt PM, et al. Statin use is
lactic magnesium does not prevent atrial fibrillation associated with a reduction in atrial fibrillation after
after cardiac surgery: a meta-analysis. Ann Thorac noncardiac thoracic surgery independent of C-
Surg 2013;95:533–41. reactive protein. Chest 2005;128:3421–7.
104. Amar D, Roistacher N, Rusch VW, et al. Effects of 108. Lai HC, Lai HC, Wang KY, et al. Severe pulmonary
diltiazem prophylaxis on the incidence and clinical hypertension complicates postoperative outcome
outcome of atrial arrhythmias after thoracic sur- of non-cardiac surgery. Br J Anaesth 2007;99:
gery. J Thorac Cardiovasc Surg 2000;120:790–8. 184–90.
105. Dehghani MR, Kasianzadeh M, Rezaei Y, et al. 109. Ramakrishna G, Sprung J, Ravi BS, et al. Impact of
Atorvastatin reduces the incidence of postopera- pulmonary hypertension on the outcomes of
tive atrial fibrillation in statin-naive patients noncardiac surgery: predictors of perioperative
undergoing isolated heart valve surgery: a morbidity and mortality. J Am Coll Cardiol 2005;
double-blind, placebo-controlled randomized 45:1691–9.
trial. J Cardiovasc Pharmacol Ther 2015;20(5): 110. Pritts CD, Pearl RG. Anesthesia for patients with
465–72. pulmonary hypertension. Curr Opin Anaesthesiol
106. Zheng H, Xue S, Hu ZL, et al. The use of statins to 2010;23:411–6.
prevent postoperative atrial fibrillation after coro- 111. Minai OA, Yared JP, Kaw R, et al. Perioperative risk
nary artery bypass grafting: a meta-analysis of 12 and management in patients with pulmonary hy-
studies. J Cardiovasc Pharmacol 2014;64:285–92. pertension. Chest 2013;144:329–40.
Pain M anagement
F o l l o w i n g Th o r a c i c
Surgery
Brett Elmore, MD, Van Nguyen, MD, Randall Blank, MD,
Kenan Yount, MD, Christine Lau, MD*
KEYWORDS
Postoperative pain Thoracic surgery Pain management Chronic pain
KEY POINTS
Managing postoperative pain is critical in reducing postoperative respiratory.
Postoperative pain results from multiple etiologic factors. There is no one modality that addresses
each contributing factor.
Optimizing pain control while minimizing sedation and respiratory depression are challenging and
competing goals, and neuraxial or regional techniques are strongly preferred over primary paren-
teral analgesia in the immediate postoperative period.
Epidural anesthesia is the gold standard for treatment; paravertebral nerve blocks are gaining
popularity, but can be technically difficult to perform for an inexperienced anesthesiologist.
Chronic pain complicates all types of thoracic procedures; once established, chronic postthoracot-
omy pain is difficult to treat. Preventive approaches include regional and neuraxial analgesia and
careful surgical technique.
Division of Thoracic & Cardiovascular Surgery, Department of Anesthesia, University of Virginia, Charlottesville,
VA 22908, USA
* Corresponding author.
E-mail address: cll2y@virginia.edu
or route of administration that addresses every indi- stretching of the incision. Without adequate anal-
vidual contributor to pain, and thus treatment regi- gesia, the resulting intense pain frequently leads
mens should be multimodal and tailored to the patients to breathe shallowly.
patient and procedure. Treating these individual Rib fractures can exacerbate an already painful
contributors to achieve patient satisfaction is an incision. Patients in this population are often aging
important primary perioperative goal, and failure to and have poor bone mineral density, which can
address acute postoperative pain may lead to the place them at increased risk for fractures. A careful
development of chronic pain syndromes. surgeon takes measures to avoid unintended frac-
It is now well-established that effective pain pro- tures, but vigorous rib spreader insertion and
phylaxis and treatment regimens begin with retraction can lead to such injuries. The extensive
regional and neuraxial anesthesia techniques. exposure sometimes required to operate in the
Thoracic epidural anesthesia (TEA) is still consid- thoracic cavity may necessitate sectioning or
ered the gold standard for treating postoperative excising segments of rib (ie, “shingling”). This
pain. However, the complications of TEA are can certainly reduce postoperative pain, but the
many and, in rare cases, can be devastating. periosteal compromise itself is painful
Although paravertebral nerve block (PVB) has nonetheless.
gained popularity as a treatment modality, too Thoracostomy tubes, frequently placed away
few anesthesiologists are comfortable with PVBs. from the primary incision, are used to drain blood
If performed by inexperienced personnel, the pa- and reduce iatrogenic pneumothorax. The addi-
tient can suffer from complications including tional incision is an obvious contributor to pain,
epidural and intrathecal spread of local anesthetic, but the tube also makes contact with the highly
pneumothorax, nerve injury, and inadequate pain innervated visceral and parietal pleura. During
relief. Furthermore, the advent of diverse and inspiration, the tube can mechanically irritate pari-
numerous oral anticoagulant and antiplatelet etal and more sensitive visceral pleura, resulting in
agents have increased the potential for neuraxial intense discomfort. If the tube is inserted too
hematoma and, hence, the contraindications for deeply or malpositioned, the resultant friction can
epidural and paravertebral techniques. Although be excruciating. In severe cases, lung injury can
systemic opioids are repeatedly vilified in this pop- occur.
ulation, they may become the cornerstone of the Ipsilateral shoulder pain is a nearly ubiquitous
treatment regimen when the risk of catastrophic complaint after all types of thoracic operations
bleeding prohibits the use of regional techniques. involving thoracotomy or thoracoscopy incisions
This article outlines systemic agents, regional and may be the result of phrenic nerve injury or
techniques (and attendant complications), etiol- diaphragmatic trauma. Although these two etiol-
ogies of pain following thoracic procedures, and ogies are especially sinister given their impact on
the development and treatment of chronic pain. respiratory function and complications postopera-
tively, it is unlikely that they are solely responsible
ETIOLOGIES OF PAIN for all cases of shoulder pain. The elderly
frequently have shoulder pathologies that are
The major component of acute postthoracotomy easily worsened by lateral decubitus positioning.
pain is attributed to the intercostal incision that Chest tube irritation can also be referred to the
spans the skin, subcutaneous tissue, muscle shoulder area. Brachial plexopathies are common
layers (including intercostal muscles, latissimus in supine procedures, but even more common in
dorsi, serratus anterior, and the pectoralis major), the lateral position if pressure points and exagger-
and parietal pleura.2 These layers are innervated ated shoulder extension are not closely monitored
by nerves with unique origins. Skin, subcutaneous intraoperatively. It is prudent to perform a focused
tissues, and intercostal muscles are innervated by neurologic examination before and after thoracic
the intercostal nerves. The latissimus dorsi and surgery.
serratus anterior are supplied by the brachial
plexus (thoracodorsal and long thoracic nerves,
Thoracotomy Versus Video-Assisted Thoracic
respectively). The parietal pleura has contributions
Surgery
from the intercostal nerves and the phrenic nerve.
Intraoperatively, the incision is forcibly retracted, Minimally invasive thoracic surgical techniques
which can lead to crushing of cutaneous or inter- have improved considerably in the past two de-
costal nerves and further muscle trauma.3 cades. In general, patients recovering from VATS
Although VATS operations decrease the extent of operations have fewer respiratory complications
the incision, identical tissue layers are traversed. and have lower pain scores postoperatively.
During inspiration, the chest wall expands causing Generally, length of stay following lobectomy via
Pain Management Following Thoracic Surgery 395
thoracotomy for lung cancer averages 4 to 5 days, enough to address pain while the patient sleeps,
but only 3 to 4 days following VATS lobectomy.4 participates in physical therapy, or practices
Many advocate thoracic epidural use for VATS lo- necessary pulmonary toilet measures (eg, incen-
bectomies, although its utility when compared with tive spirometry, deep-breathing, coughing, and
systemic analgesia remains controversial. Devel- so forth) However, recent work7 suggests that
opment of PTPS is somewhat similar in the two PCAs, when paired with nonopioid adjuvant med-
groups.5 Muscle-sparing thoracotomy ap- ications, can be effective in providing adequate
proaches have also gained popularity, but data analgesia without increasing the risk of PTPS.
are conflicting regarding the development of Nonetheless, if systemic opioids are to be used
PTPS for this approach when compared with in place of regional anesthesia, the patient should
posterolateral approach.6 be monitored at even more frequent intervals in a
critical care setting with tight titration and liberal
Systemic Versus Regional Analgesia use of nonopioid adjuvants.
Although TEA is recognized to be the gold stan-
dard for pain prevention and treatment among Nonsteroidal Anti-inflammatory Drugs
thoracic surgery patients, this modality fails to
address all the components of acute pain following Adjunctive analgesic medications that do not sup-
thoracic surgery. For example, pain contributions press ventilatory drive are of particular interest in
from thoracic dermatomes can be reliably muted this population. Nonsteroidal anti-inflammatory
with thoracic epidural or multilevel paravertebral drugs (NSAIDs) provide analgesia by reversibly in-
analgesia, but shoulder discomfort and pleural hibiting cyclooxygenase (COX), which reduces
discomfort often persist. Other therapeutic agents prostaglandin synthesis. NSAIDs have varying de-
may also be useful in treating pain following grees of specificity for inhibiting the subtypes of
thoracic surgery. Whether systemically or region- COX inhibitors: COX-1 and COX-2. COX-1, pre-
ally administered, these therapies possess benefi- sent in several tissues at baseline, is central in
cial and potentially detrimental qualities that must the synthesis of various prostaglandins from
be weighed when tailoring a treatment regimen arachidonic acid, including those responsible for
(Tables 1 and 2). regulating inflammation, platelet aggregation,
renal vascular vasodilation, and gastric acid secre-
tion. COX-2 is upregulated in states of inflamma-
SYSTEMIC AND PARENTERAL THERAPIES
tion and was of considerable interest as a
Opioids
pharmaceutical target in hopes of reducing inflam-
Whether administered intravenously, orally, or mation with fewer side effects than traditional
neuraxially, opioids are an important component nonselective COX-inhibitors. Unfortunately, COX-
of most treatment regimens following thoracic op- 2 inhibitors (eg, celecoxib and rofecoxib) were
erations. Both systemic and neuraxial routes pro- found to significantly increase risk for adverse car-
vide reliable analgesia, although systemic opioids diac and cerebrovascular events, thus limiting
are more likely to increase the risk of pulmonary their therapeutic role in postthoracotomy pain. At
complications, such as atelectasis, pneumonia, least one study8 suggests an increase in patient
and hypoxic or hypercarbic respiratory failure. satisfaction with analgesia when celecoxib was
Opioid overdose can be devastating in this popu- combined with TEA, but the study was not suffi-
lation. Serious morbidity can result even with min- ciently powered to evaluate cardiac morbidity.
imal opioid-induced respiratory depression. The Nonselective NSAIDs are more routinely used in
adverse effects of opioids are not well tolerated supplementing opioids perioperatively. When
in patients with obstructive pulmonary disease, administered concurrently with acetaminophen,
particularly following lung resection. Unless con- NSAIDs were noted to reduce opioid requirements
traindicated, the benefits of regional and neuraxial by roughly 30% to 35%.9 Indomethacin adminis-
techniques using a combination of local anesthetic tration was shown10 to significantly reduce opioid
and opioid outweigh the risks and should be consumption and reduce pain scores after thora-
considered superior to a parenteral opioid cotomy. IV ketorolac, if given preoperatively, was
technique. noted to reduce morphine consumption by 36%
Intravenous (IV) patient-controlled analgesia when compared with placebo, although it had no
(PCA) systems have become a mainstay of treat- effect on pulmonary function postoperatively.11 If
ment of postsurgical pain in the modern era. used, significant attention must be given to their
Although this therapy does allow the patient to potential side effects including not only increased
titrate relief to perceived pain level and reduce risk of cardiovascular events, but also bleeding,
the risk of overdose, PCAs are not dynamic renal impairment, dyspepsia, and gastric
396 Elmore et al
Table 1
Nonopioid analgesics
Systemic
Analgesics Benefits Risks Recommendations
Acetaminophen Safe, effective analgesic Liver toxicity Recommended in
and antipyretic combination with other
Reduces pain scores and analgesics
opioid requirements
No increased incidence in
hemorrhage, gastric
ulceration,
cardiovascular, and
renal adverse effects
Has “ceiling effect”
NSAIDs Improves pain relief Impaired coagulation, Recommended in
Reduces opioid gastric irritation, renal combination with other
consumption by 30% dysfunction, and analgesics
and decreases opioid- cardiovascular adverse
related adverse effects effects
COX-2 inhibitors Improves pain scores, Potential gastric Recommended in
decreases opioid irritation, renal combination with other
consumption, and dysfunction, and analgesics
reduces opioid-related cardiovascular adverse
adverse effects effects
Similar efficacy as NSAIDs
No effects on platelet
function and
perioperative bleeding
Glucocorticoids Reduces inflammation, Increase blood glucose Recommended as an
(dexamethasone) improves pain relief, levels up to 24 h, but adjunct
prolongs time to first may not be clinically
analgesic, and modest relevant
reduction in opioid
requirements
Ketamine Analgesic properties Sympathomimetic and Not recommended for
without respiratory neurocognitive side routine use
depressive effects, effects
reduces pain scores,
and opioid
consumption, and
prolongs time to first
analgesic
Optimal dose and
duration of
administration remain
controversial
Gabapentinoids Reduced pain scores and Sedation, dizziness, and Not recommended for
(gabapentin and opioid requirements visual disturbances routine use
pregabalin) Optimal dose and
duration of
administration remain
controversial
Table 2
Regional analgesia techniques for thoracic surgery
Regional
Anesthesia
Techniques Benefits Risks Recommendations
Thoracic Superior dynamic analgesia Epidural spread of Recommended
paravertebral during coughing and local anesthetic with
analgesia physical therapy associated risks, vascular
Improved postoperative injury, and pleural injury
outcome Potential for catastrophic
Equally effective as TEA neurologic complications
Trend toward lower is remote
incidence of major
complications compared
with TEA and lower block
failure rate
Limited value with single-
shot injection
TEA Superior analgesia during High (15%) failure rate, Recommended
coughing and physical complicates
therapy, and improved postoperative
postoperative outcome anticoagulation,
hypotension, nausea,
urinary retention,
pruritus, accidental
intrathecal spread,
epidural hematoma, and
epidural abscess
Intrathecal Better static and dynamic Risk of respiratory Recommended, if
opioid pain scores compared depression, pruritus, paravertebral block or
analgesia with systemic opioid urinary retention, TEA is contraindicated
analgesia specifically in nausea, and vomiting or not possible
first 24 h postoperatively
Intercostal Simple and easy to Systemic local toxicity, Recommended in
analgesia perform, superior pain better pain scores with combination with
scores, reduced opioid continuous catheter, or nonopioid analgesics,
requirements, and multiple injections such as acetaminophen
improved postoperative and NSAIDs or COX-2-
outcome specific inhibitors, if
paravertebral block or
TEA is contraindicated
or not possible
Interpleural Easy to perform but not Potential of local Not recommended
analgesia efficacious anesthetic toxicity
Intercostal nerve Effective in perioperative Implicated in increasing Not recommended
cryoanalgesia period in improving pain incidence of chronic pain
scores compared with
placebo
Abbreviations: COX, cyclooxygenase; NSAIDs, nonsteroidal anti-inflammatory drugs.
From Romero A, Garcia JE, Joshi GP. The state of the art in preventing postthoracotomy pain. Semin Thorac Cardiovasc
Surg 2013;25(2):188; with permission.
pulmonary complications. It also avoids the adverse ropivacaine and morphine. Forty-nine patients
side effects of NSAIDs, such as renal function, were randomized to low-dose ketamine infusion
gastric ulceration, or platelet aggregation. Its use is versus saline placebo. Patients receiving low-
only limited by its potential for liver toxicity and, dose ketamine demonstrated lower pain scores
accordingly, patients with hepatic impairment at 24 hours, 48 hours, 1 month, and 3 months. If
should receive it only sparingly. Recently, an IV regional anesthesia is contraindicated, providers
formulation of acetaminophen has generated should strongly consider ketamine as an adjunct
considerable interest in the perioperative use of because of its favorable side effect profile and
acetaminophen. Administration in critically ill pa- demonstrated analgesic efficacy in this context.20
tients has shown to decrease time to extubation, Although ketamine is known to sometimes lead to
reduce opioid-related side effects (nausea and intense dysphoria, this is less likely at doses tradi-
sedation), and decrease meperidine consump- tionally used as an adjunct for pain treatment (0.5–
tion.15 In a recent, double-blind, randomized, 1 mg/kg dose or <10 mg/h infusion).21–23 Still, this
placebo-controlled study, Mac and colleagues16 adverse effect should be respected in an aging
demonstrated that 48 hours of perioperative acet- population that is already at increased risk for
aminophen significantly reduced ipsilateral shoulder postoperative delirium.
pain compared with placebo. Given its relatively safe
profile and proved benefit when used in multimodal Dexmedetomidine
analgesia, acetaminophen should be incorporated
Dexmedetomidine is a selective a2-agonist that
into most multimodal pain regimens. When used,
has a sedative effect without affecting respiratory
our typical IV regimen is 1000 mg every 6 hours
drive. Wahlander and colleagues24 assessed
(maximum, 4000 mg in 24 hours).
whether an IV infusion of dexmedetomidine
reduced epidural use in postthoracotomy patients.
Benzodiazepines
Although there was no difference in pain scores or
Not surprisingly, patients can have debilitating anx- epidural use compared with placebo, the placebo
iety if suffering from severe pain that is exacerbated group required more supplemental epidural fenta-
simply with inhalation and coughing. Although ben- nyl. A study by Ramsay and colleagues25 had
zodiazepines can be used for severe cases of anxi- similar results: pain scores were similar in the dex-
ety (especially in patients previously on daily medetomidine infusion group and place group, but
regimens), this category of drugs has fallen out of the dexmedetomidine group used 41% less
favor because of its strong association with postop- opioids, comparatively. Limiting its usefulness is
erative delirium. The elderly are particularly sensitive the association of dexmedetomidine with hypo-
to the effects, which can be long-lasting and exag- tension and bradycardia. This population is
gerated in this subgroup.17 Furthermore, benzodiaz- already at risk for severe hypotension secondary
epines are known to cause respiratory depression, to epidural-induced sympathectomy, almost uni-
especially when used in combination with systemic versal b-blocker administration for tachyar-
opioids. Although benzodiazepines may be useful rhythmia prophylaxis, and fluid restriction.
in attenuating the dysphoric effects of ketamine, Administration should be limited to continuous
the use of true analgesics and multimodal adjuncts low-dose infusions to avoid hypotension.
should be considered preferable to the routine use
of benzodiazepines in this context. Gabapentinoids
Gabapentin and pregabalin are the two clinically
Ketamine
available gabapentinoids used in the management
Ketamine is an N-methyl-D-aspartate receptor of neuropathic pain. In a recent review, Humble
antagonist that has long been used to reduce and colleagues26 found that gabapentinoids
opioid requirements. Ketamine is particularly use- reduced postoperative pain in cases without
ful as an adjunct in this population because it is concomitant epidural anesthesia. Kinney and col-
known to stimulate respiratory drive even with leagues27 in a double-blind, randomized,
low-dose administration. A systematic review by controlled trial showed that a single dose of gaba-
Laskowski and colleagues18 suggests that keta- pentin (600 mg) preoperatively had no effect on
mine, when used intraoperatively, shows pain scores or opioid consumption. Studies by
improved quality of pain control independent of Omran and Mohamed28 and Solak and co-
timing, dose, and route of opioid administration workers29 noted that longer preoperative and
in major operations. Its role in reducing PTPS is postoperative gabapentinoid regimens did result
still unclear. Suzuki and colleagues19 evaluated in reductions in opioid requirements. The role of
ketamine-potentiation of epidural analgesia with gabapentinoids on PTPS is discussed later.
Pain Management Following Thoracic Surgery 399
catheter not in the epidural space, or incomplete disruption of the spinal column by previous surgery
block per Ready37). Epidurals also can complicate or instrumentation, duration of catheterization, and
the postoperative plan for anticoagulation. Our coexisting sources of infection. Patients undergo-
typical practice is to schedule prophylactic ing thoracic operations frequently possess one or
enoxaparin at 18:00 daily (single-dose) for all more of these risk factors. Presenting symptoms
patients with epidurals in place to facilitate their of epidural abscess show remarkable inconsis-
timely removal and optimal management of tency, despite traditional teaching that a patient
anticoagulation. will classically present with new neurologic symp-
Catastrophic complications of TEA, such as toms, fever, and back pain.47 The importance of
epidural hematoma and abscess, are rare. Howev- vigilance and a high index of suspicion for these
er, studies suggest that the incidence of epidural complications in patients with a recent neuraxial
hematoma has increased over the past two de- intervention cannot be overstated. Many reported
cades. A multicenter, retrospective study in North cases presented well after the patient was dis-
America academic hospitals by Bateman and col- charged from inpatient care, so patients should
leagues38 indicates that incidence of epidural he- be instructed to self-monitor for concerning signs
matoma requiring emergent laminectomy is and symptoms.47
between 1 in 4330 and 1 in 22,189 placements. The benefits of TEA in reducing postsurgical pul-
In four of the seven documented epidural hema- monary complications over the past few decades
tomas (out of 62,450 patients), current American have been eroded by a combination of improved
Society of Regional Anesthesia (ASRA) anticoagu- surgical technique, prophylactic antibiotics, early
lation guidelines were not followed. Time to devel- mobilization, and more aggressive pulmonary
opment of symptoms in this study varied widely physiotherapy.48 TEA is still the most preferred
from 11 to 71 hours after epidural placement. Hor- thoracic pain management technique, but its
locker and Kopp39 note that prior studies from routine use has been increasingly questioned.49–51
1993 early show a much lower incidence (in one
case, <1 in 150,000). This is at least in part caused
Paravertebral Analgesia
by current medical practice of aggressive venous
thromboembolic prophylaxis with modern potent Although TEA has long been considered the gold
anticoagulants. This warrants serious concern in standard for thoracic operations, PVB has been
patients receiving TEA (or even PVB) after thoracic increasing steadily in popularity. Reasons for the
operations, because they often require concurrent increasing popularity of paravertebral analgesia
anticoagulation therapy for other conditions, such include the emergence of ultrasound guidance to
as atrial fibrillation, coronary stents, and history of facilitate easier and more accurate paravertebral
deep vein thrombosis. The ASRA has developed catheter placement (PVB has been traditionally
consensus guidelines to assist physicians in man- been placed using landmark/loss of resistance
aging regional anesthesia techniques concurrently technique, and, intraoperatively, under direct visu-
with anticoagulation therapy.40 These guidelines alization), and a more favorable side effect profile
should be adhered to whenever possible in this as compared with TEA. Multiple reviews and
population to minimize the likelihood of epidural meta-analysis have shown no difference in pain
hematoma. Table 3 summarizes some of these scores, postoperative pulmonary function, and
recommendations when using common modern postoperative pulmonary complications in pa-
anticoagulants. Development of new or exagger- tients undergoing thoracic procedures between
ated neurologic symptoms any time after place- PVB and TEA.32,52
ment should warrant further neurologic Joshi and colleagues32 concluded in their sys-
examination, neuraxial imaging, or consultation temic reviews of randomized controlled trials that
of neurosurgery, especially when anticoagulants continuous PVB provides comparable pain relief
are administered in high-risk patients or patients to continuous TEA when local anesthetic alone
with coagulopathy. was used in both infusions. The evidence for supe-
Epidural abscess is extraordinarily uncommon, riority between the two modalities was inconclusive
but it can be more sinister because symptoms are when it came to continuous infusions containing
often nonspecific.43 Incidence of epidural abscess both local anesthetic and opioid. A more recent
complicating epidural placement is difficult to esti- randomized controlled trial by Grider and col-
mate, with studies ranging from 1 in 1000 to 1 in leagues53 studied patients undergoing thoracot-
100,000.44–46 Grewal and colleagues44 assert that omy receiving either continuous PVB with
certain risk factors may predispose to epidural ab- bupivacaine alone, TEA with bupivacaine alone,
scess: compromised immunity (ie, diabetes or TEA with bupivacaine and hydromorphone.
mellitus, immunosuppression therapy, cancer), Although the groups receiving PVB with
Pain Management Following Thoracic Surgery 401
bupivacaine and TEA with bupivacaine alone had bolus in the paravertebral space is more suscepti-
similar VAS scores, the group with TEA with bupiva- ble to epidural spread compared with small,
caine and hydromorphone had statistically signifi- repeated boluses or continuous infusions through
cant lower VAS scores than the other two groups. a catheter) and with the use of ultrasound guid-
This suggests that although PVB and TEA modal- ance.58 Another potential risk of PVB is pleural
ities may offer similar pain control, bupivacaine injury resulting in pneumothorax, which can be
and hydromorphone can work synergistically in minimized under intraoperative, direct visualiza-
the epidural space to provide better analgesia. tion. The use of ultrasound does not guarantee
Multiple studies have demonstrated that PVB safe, extrapleural PVB catheter placement,59 but
results in better postoperative pulmonary function likely reduces the incidence of pneumothorax in
and less adverse effects on hemodynamics the hands of an experienced anesthesiologist.
compared with TEA. The prospective, multicenter, Vascular injury is also a concern given the highly
observational trial by Powell and colleagues54 vascular nature of the paravertebral space. How-
comparing the effect of PVB versus TEA on major ever, because of the large volume of the paraver-
postoperative complications in patients undergo- tebral space, the likelihood of a hematoma
ing pneumonectomy concluded that the PVB causing significant adverse effects is remote.
group experienced lower incidence of hypotension It should be noted that a continuous infusion
requiring inotropes, arrhythmia requiring antiar- through a paravertebral catheter should be
rhythmics, respiratory complications requiring preferred over a single injection for patients under-
ventilator support, and need for surgical re- going thoracic operations given its ability extend
exploration. From a technical standpoint, place- the duration of pain management. If only a single
ment of a paravertebral catheter is unlikely to paravertebral injection is possible, longer-acting
result in accidental dural puncture that can lead local anesthetics, such as bupivacaine or ropiva-
to postdural puncture headache, a complication caine, should be considered to give the patient
infrequently associated with the placement of longer duration of pain control. The commonly
thoracic epidural catheters. Position of the para- used local anesthetics lidocaine, bupivacaine,
vertebral catheter outside the rigid epidural space and ropivacaine seem equally efficacious when
also means that the possibility of catastrophic infused continuously in the paravertebral space.
neurologic injuries associated with TEA, such as Unlike TEA, the addition of opioids to local anes-
epidural hematoma or epidural abscess, is less thetics has not been shown to improve pain con-
likely. Still, the most recent ASRA guidelines39,40 trol in PVB, and therefore is not recommended.60
recommend that the same anticoagulation guide- Several studies have demonstrated that PVB is
lines for neuraxial procedures also be applied to superior to TEA in reducing the incidence of hypo-
“deep peripheral” blocks (such as PVB). Strict tension without compromising quality of anal-
application of these guidelines limits the use of gesia.61–63 PVB was also shown to lead to less
PVB when TEA is also contraindicated because nausea/vomiting and urinary retention than TEA.61
of existing coagulopathy. Table 3 shows current Although it can be argued that TEA and PVB are
recommendations for the discontinuation of com- both viable options for many unilateral thoracic
mon anticoagulant and antiplatelet therapies procedures, certain operations favor TEA because
before neuraxial and paravertebral procedures.55 of its bilateral and arguably more complete anal-
However, in the setting of thrombocytopenia, gesia. Examples include bilateral lung transplant
PVB has been used safely.56 It should be noted or heart-lung transplant (bilateral thoracotomy
that many anesthesiologists are more comfortable with sternal transection), Ivor-Lewis esophagec-
placing epidural catheters than PVBs. However, tomy (right thoracotomy with midline laparotomy),
the reported failure rate of PVB (w6% per57) is and thoracoabdominal aneurysm repair (left thora-
lower than that of TEA (10%–15% per36). cotomy with variable midline incision). Although
Injection of medication within the paravertebral more labor intensive, bilateral PVB still remains
space may spread less reliably when compared an option for these procedures. Richardson and
with that in the epidural space. Up to 70% of pa- colleagues64 reviewed 12 studies that used bilat-
tients with PVB have some degree of epidural eral PVB and concluded that this method was a
spread. Spread of infusion from the paravertebral reasonable alternative to epidural anesthesia with
space into the epidural space through the verte- complications, such as hypotension and pneumo-
bral foramen has limited side effects, such as tran- thorax (3.6% and 1%, respectively; N 5 196),
sient hypotension and/or contralateral numbness, deemed low-risk. Ostensibly, level of comfort
which may mimic the effects of TEA. The incidence with performing PVB considerably limits use of
of epidural spread of PVB seems to be higher with bilateral PVB and likely raises the risk of
larger volume of injectate (single large-volume complications.
402
Elmore et al
Table 3
Anticoagulation guidelines for regional anesthesia and analgesia: Wake Forest University RAAPM recommendations to avoid increasing the risk of
neuraxial hematoma following neuraxial analgesic/anesthetic procedures
Intravenous (1.5 h)
Heparin (unfractionated)b (1.5 h) Pro–antithrombin III (anti-II.X) No restriction,40 caution during No restriction40 c
Initiate 2 h postremovala
Rivaroxaban (Xarelto) (5–13 h) Anti-Xa 48–72a 6 h, or 24 h if traumatic insertion c,d
catheter42 Initiate 2 h
postremovala
Abciximab (Reopro) (30 min) Glycoprotein IIb/IIIa inhibitor 48 h40 Not recommended with —
catheter40 Initiate 2 h
postremovala
Eptifibatide (Integrilin) (2.5 h) Glycoprotein IIb/IIIa inhibitor 8 h40 Not recommended with —
catheter40 Initiate 2 h
postremovala
Tirofiban (Aggrastat) (2 h) Glycoprotein IIb/IIIa inhibitor 8 h40 Not recommended with —
catheter40
Initiate 2 h postremovala
Bivalrudin (Augiomax) Thrombin (II) Insufficient data40 Insufficient data40 —
Desirudi (Iprivask) Inhibitor (IV) Neuraxial techniques not
Argatroban (Acova) recommendeda
Note: Recommendations are based on single drug use, combinations increase risk. Caution if traumatic neuraxial technique. Recommendation compliance does not eliminate the risk
for neuraxial hematoma.
Abbreviations: ADP, adenosine diphosphate; aPTT, activated partial thromboplastin time; BID, twice daily; CRI, continuous rate infusion; INR, international normalized ratio;
LMWH, low-molecular-weight heparin; QD, every day; SQ, subcutaneous; TID, three times a day; WNL, within normal limits.
a
Our current practice, no current published guidelines.
b
Patients receiving unfractionated heparin should have platelet count checked after 4 days to monitor for possible heparin-induced thrombocytopenia.
c
Caution with CRI, low weight, elderly.
d
T 1/2 doubled with strong CYP3A4 inhibitors (antifungals, antiretrovirals).
From Henshaw DS, Jaffe JD, Weller RS. Quick Reference Guide for Regional Anesthesia in the Anticoagulated Patient. Available at: http://www.nysora.com/newsletterz/2014/4313-
july-2014-newsletter.html. Accessed July 27, 2015; with permission.
403
404 Elmore et al
intercostal muscle flap to reduce PTPS are sup- 2. Ochroch EA, Gottschalk A. Impact of acute pain and
ported by Cerfolio and colleagues,76,77 who arrive its management for thoracic surgical patients.
at the same conclusion in separate studies. Thorac Surg Clin 2005;15(1):105–21.
3. Sapkota R, Shrestha UK, Sayami P. Intercostal mus-
Treatment cle flap and intracostal suture to reduce post-
thoracotomy pain. Asian Cardiovasc Thorac Ann
The treatment of PTPS, like other neuropathic pain 2013;22(6):706–11.
syndromes, can be challenging. There are few 4. Flores RM, Park BJ, Dycoco J, et al. Lobectomy by
studies designed to specifically investigate the video-assisted thoracic surgery (VATS) versus thora-
treatment of PTPS, and most suffer from a variety cotomy for lung cancer. J Thorac Cardiovasc Surg
of methodologic flaws. Broadly speaking, the 2009;138(1):11–8.
management of PTPS is divided into pharmaco- 5. Landreneau RJ, Mack MJ, Hazelrigg SR, et al. Prev-
logic and interventional modalities. alence of chronic pain after pulmonary resection by
thoracotomy or video-assisted thoracic surgery.
Pharmacologic
J Thorac Cardiovasc Surg 1994;107(4):1079–85.
Because chronic persistent surgical pain, such as
6. Rogers ML, Duff JP. Surgical aspects of chronic
PTPS, is thought to be neuropathic in nature, sys-
post-thoracotomy pain. Eur J Cardiothorac Surg
temic opioid therapy may not be effective. Gaba-
2000;18:711–6.
pentinoids, however, theoretically have benefits.
7. Tiippana E, Nelskylä K, Nilsson E, et al. Managing
Solak and colleagues29 showed in their prospec-
post-thoracotomy pain: epidural or systemic analgesia
tive, randomized, controlled study of 40 patients
and extended care: a randomized study with an “as
with PTPS that those receiving gabapentin
usual” control group. Scand J Pain 2015;5(4):240–7.
showed significantly lower pain scores after
8. Senard M, Deflandre EP, Ledoux D, et al. Effect of
60 days with treatment when compared with those
celecoxib combined with thoracic epidural anal-
receiving naproxen. The authors concluded that
gesia on pain after thoracotomy. Br J Anaesth
gabapentin is safe and effective in treatment of
2010;105(2):196–200.
PTPS, with high patient compliance and tolera-
9. Ong CK, Seymour RA, Lirk P, et al. Combining para-
bility. Pregabalin, like gabapentin, is a structural
cetamol (acetaminophen) with nonsteroidal antiin-
analog of g-aminobutyric acid and is commonly
flammatory drugs: a qualitative systematic review
used to treat neuropathic pain. Studies have
of analgesic efficacy for acute postoperative pain.
demonstrated that pregabalin can reduce opioid
Anesth Analg 2010;110(4):1170–9.
consumption perioperatively; however, it has
10. Pavy T, Medley C, Murphy DF. Effect of indometh-
high incidence of visual disturbances. It should
acin on pain relief after thoracotomy. Br J Anaesth
be noted that both gabapentin and pregabalin
1990;65(5):624–7.
can cause drowsiness and dizziness.
11. Boussofara M, Mtaallah MH, Bracco D, et al. Co-
Other medications typically used in the treatment
analgesic effect of ketorolac after thoracic surgery.
of PTPS include tricyclic antidepressant, serotonin-
Tunis Med 2006;84(7):427–31.
norepinephrine reuptake inhibitors, tramadol, and
12. Peduto VA, Ballabio M, Stefanini S. Efficacy of prop-
lidocaine. Unfortunately, their effectiveness is
acetamol in the treatment of postoperative pain:
mostly anecdotal and further research is necessary.
morphine-sparing effect in orthopedic surgery.
Interventional Acta Anaesthesiol Scand 1998;42(3):293–8.
Beyond pharmacologic treatment, case reports and 13. Hernández-Palazón J, Tortosa JA, Martı́nez-Lage JF,
retrospective studies have suggested that interven- et al. Intravenous administration of propacetamol re-
tions, such as ICNB, pulsed radiofrequency of the duces morphine consumption after spinal fusion sur-
dorsal root ganglion, and spinal cord and peripheral gery. Anesth Analg 2001;92:1473–6.
nerve stimulation, can provide some benefits in the 14. Khalili G, Janghorbani M, Saryazdi H, et al. Effect of
treatment of PTPS. In a randomized controlled trial, preemptive and preventive acetaminophen on post-
the use of acupuncture for the treatment of PTPS operative pain score: a randomized, double-blind
was found to be ineffective.78 Epidural steroid injec- trial of patients undergoing lower extremity surgery.
tion, a technique common in the management of J Clin Anesth 2013;25(3):188–92.
neuropathic pain, does not have any formal literature 15. Memis D, Inal MT, Kavalci G, et al. Intravenous para-
supporting its use in the treatment of PTPS. cetamol reduced the use of opioids, extubation time,
and opioid-related adverse effects after major sur-
gery in intensive care unit. J Crit Care 2010;25(3):
REFERENCES
458–62.
1. Gerner P. Postthoracotomy pain management prob- 16. Mac TB, Girard F, Chouinard P, et al. Acetaminophen
lems. Anesthesiol Clin 2008;26:355–67. decreases early post-thoracotomy ipsilateral
Pain Management Following Thoracic Surgery 407
shoulder pain in patients with thoracic epidural anal- 31. Cui W, Turgeon AF, Côté D, et al. Systemic adminis-
gesia: a double-blind placebo-controlled study. tration of lidocaine reduces morphine requirements
J Cardiothorac Vasc Anesth 2005;19(4):475–8. and postoperative pain of patients undergoing
17. Pisani MA, Murphy TE, Araujo KL, et al. Benzodiaz- thoracic surgery after propofol-remifentanil-based
epine and opioid use and the duration of intensive anaesthesia. Eur J Anaesthesiol 2010;27(1):41–6.
care unit delirium in an older population. Crit Care 32. Joshi GP, Bonnet F, Shah R, et al. A systematic re-
Med 2009;37(1):177–83. view of randomized trials evaluating regional tech-
18. Laskowski K, Murphy TE, Araujo KL, et al. niques for postthoracotomy analgesia. Anesth
A systematic review of intravenous ketamine for Analg 2008;107:1026–40.
postoperative analgesia. Can J Anaesth 2011; 33. Bauer C, Hentz JG, Ducrocq X, et al. Lung function
58(10):911–23. after lobectomy: a randomized, double-blinded trial
19. Suzuki M, Haraguti S, Sugimoto K, et al. Low-dose comparing thoracic epidural ropivacaine/sufentanil
intravenous ketamine potentiates epidural analgesia and intravenous morphine for patient-controlled
after thoracotomy. Anesthesiology 2006;105(1):111–9. analgesia. Anesth Analg 2007;105:238–44.
20. Mendola C, Cammarota G, Netto R, et al. S (1) 34. Ali M, Winter DC, Hanly AM, et al. Prospective ran-
ketamine for control of perioperative pain and pre- domized, controlled trial of thoracic epidural or
vention of post-thoracotomy pain syndrome: a ran- patient-controlled opiate analgesia on perioperative
domized, double-blind study. Minerva Anestesiol quality of life. Br J Anaesth 2010;104:292–7.
2012;78:757–66. 35. Oka T, Ozawa Y, Ohkubo Y. Thoracic epidural bupi-
21. Himmelseher S, Durieux ME. Ketamine for perioper- vacaine attenuates supraventricular tachyarrhyth-
ative pain management. Anesthesiology 2005; mias after pulmonary resection. Anesth Analg
102(1):211–20. 2001;93:153–9.
22. Guillou N, Tanguy M, Seguin P, et al. The effects of 36. Hermanides J, Hollmann MW, Stevens MF, et al.
small-dose ketamine on morphine consumption in Failed epidural: causes and management. Br J
surgical intensive care unit patients after major Anaesth 2012;109(2):144–54.
abdominal surgery. Anesth Analg 2003;97:843–7. 37. Ready LB. Acute pain: lessons learned from 25,000
23. Chia YY, Liu K, Liu YC, et al. Adding ketamine in a patients. Reg Anesth Pain Med 1999;24:499–505.
multimodal patient-controlled epidural regimen re- 38. Bateman BT, Mhyre JM, Ehrenfeld J, et al. The risk
duces postoperative pain and analgesic consump- and outcomes of epidural hematomas after periop-
tion. Anesth Analg 1998;86:1245–9. erative and obstetric epidural catheterization: a
24. Wahlander S, Frumento RJ, Wagener G, et al. report from the Multicenter Perioperative Outcomes
A prospective, double-blind, randomized, placebo- Group Research Consortium. Anesth Analg 2013;
controlled study of dexmedetomidine as an adjunct 116(6):1380–5.
to epidural analgesia after thoracic surgery. 39. Horlocker T, Kopp S. Epidural hematoma after
J Cardiothorac Vasc Anesth 2005;19(5):630–5. epidural blockade in the United States: it’s not just
25. Ramsay MA, Newman KB, Leeper B, et al. Dexme- low molecular heparin following orthopedic surgery
detomidine infusion for analgesia up to 48 hours af- anymore. Anesth Analg 2013;116(6):1195–7.
ter lung surgery performed by lateral thoracotomy. 40. Horlocker T, Wedel DJ, Rowlingson JC, et al.
Proc (Bayl Univ Med Cent) 2014;27(1):3–10. Regional anesthesia in the patient receiving antith-
26. Humble SR, Dalton AJ, Li L. A systematic review of rombotic or thrombolytic therapy: American Society
therapeutic interventions to reduce acute and chronic of Regional Anesthesia and Pain Medicine
post-surgical pain after amputation, thoracotomy or evidence-based guidelines (third edition). Reg
mastectomy. Eur J Pain 2015;19(4):451–65. Anesth Pain Med 2010;35(1):64–101.
27. Available at: http://www.ncbi.nlm.nih.gov/pubmed/ 41. ASRA Safety Announcement: Updated recommen-
21676165. dations to decrease risk of spinal column bleeding
28. Omran AF, Mohamed AE. A randomized study of the and paralysis in patients on low molecular weight
effects of gabapentin versus placebo on post- heparins, 11/6/13.
thoracotomy pain and pulmonary function. Eg J 42. Regional anesthesia and Antithrombotic agents:
Anaesth 2005;21:277–81. recommendations of the European Society of
29. Solak O, Metin M, Esme H, et al. Effectiveness of ga- Anesthesiology.
bapentin in the treatment of chronic post- 43. Chan YC, Dasey N. Iatrogenic spinal epidural ab-
thoracotomy pain. Eur J Cardiothorac Surg 2007; scess. Acta Chir Belg 2007;107(2):109–18.
32:9–12. 44. Grewal S, Hocking G, Wildsmith JAW. Epidural ab-
30. Vigneault L, Turgeon AF, Côté D, et al. Perioperative scesses. Br J Anaesth 2006;96(3):292–302.
intravenous lidocaine infusion for postoperative pain 45. Grieve JP, Ashwood N, O’Neill KS, et al. A retrospective
control: a meta-analysis of randomized controlled study of surgical and conservative treatment for spinal
trials. Can J Anaesth 2011;58(1):22–37. extradural abscess. Eur Spine J 2000;9:67–71.
408 Elmore et al
46. Phillips JM, Stedeford JC, Hartsilver E, et al. an updated meta-analysis. PLoS One 2014;9(5):
Epidural abscess complicating insertion of epidural e96233.
catheters. Br J Anaesth 2002;89:778–82. 62. Baidya DK, Khanna P, Maitra S. Analgesic efficacy
47. Darouiche RO. Spinal epidural abscess. N Engl J and safety of thoracic paravertebral and epidural
Med 2006;355(19):2012–20. analgesia for thoracic surgery: a systematic review
48. Popping DM, Elia N, Marret E, et al. Protective and meta-analysis. Interact Cardiovasc Thorac
effects of epidural analgesia on pulmonary compli- Surg 2014;18(5):626–35.
cations after abdominal and thoracic surgery: a 63. Okajima H, Tanaka O, Ushio M, et al. Ultrasound-
meta-analysis. Arch Surg 2008;143:990–9. guided continuous thoracic paravertebral block pro-
49. Low J, Johnston N, Morris C. Epidural analgesia: vides comparable analgesia and fewer episodes of
First do no harm. Anaesthesia 2008;63:1–3. hypotension than continuous epidural block after
50. Freise H, Van Aken HK. Risks and benefits of thoracic lung surgery. J Anesth 2015;29(3):373–8.
epidural anaesthesia. Br J Anaesth 2011;107:859–68. 64. Richardson J, Lönnqvist PA, Naja Z. Bilateral
51. Rawal N. Epidural technique for postoperative pain thoracic paravertebral block: potential and practice.
gold standard no more? Reg Anesth Pain Med Br J Anaesth 2011;106(2):164–71.
2012;37:310–7. 65. Meylan N, Elia N, Lysakowski C, et al. Benefits and
52. Kotze A, Scally A, Howell S. Efficacy and safety of risks of intrathetcal morphine without local anaes-
different techniques of paravertebral block for anal- thetic in patients undergoing major surgery: meta-
gesia after thoracotomy: a systematic review and analysis of randomized trials. Br J Anaesth 2009;
metaregression. Br J Anaesth 2009;103:626–36. 102:156–67.
53. Grider JS, Mullet TW, Saha SP, et al. A random- 66. Dango S, Harris S, Offner K, et al. Combined para-
ized double-blind trial comparing continuous vertebral and intrathecal vs thoracic epidural anal-
thoracic epidural bupivacaine with and without gesia for post-thoracotomy pain relief. Br J
opioid in contrast to a continuous paravertebral Anaesth 2013;110:443–9.
infusion of bupivacaine for post-thoracotomy 67. Suksompong S, Pongpayuha P, Lertpaitoonpan W,
pain. J Cardiothorac Vasc Anesth 2012;26:83–9. et al. Low-dose spinal morphine for post-
54. Powell ES, Cook D, Pearce AC, et al. A prospective, thoracotomy pain: a prospective randomized study.
multicentre, observational cohort study of analgesia J Cardiothorac Vasc Anesth 2013;27:417–22.
and outcome after pneumonectomy. Br J Anaesth 68. Meierhenrich R, Hock D, Kuhn S, et al. Analgesia
2011;106:364–70. and pulmonary function after lung surgery: Is a sin-
55. Henshaw D, Jaffe J, Weller R. Quick reference guide gle intercostal nerve block plus patient-controlled
for regional anesthesia in the anticoagulated patient. intravenous morphine as effective as patient-
New York School of Regional Anesthesia; 2014. controlled epidural anesthesia? A randomized non-
Available at: http://www.nysora.com/newsletterz/2014/ inferiority clinical trial. Br J Anaesth 2011;106:580–9.
4313-july-2014-newsletter.html. Accessed July 27, 69. Khanbhai M, Yap KH, Mohamed S, et al. Is cryoanal-
2015. gesia effective for post-thoracotomy pain? Interact
56. Nguyen VH, de Souza DG, Blank RS, et al. Ultra- Cardiovasc Thorac Surg 2014;18:202–9.
sound-guided thoracic paravertebral catheter 70. Doan LV, Augustus J, Androphy R, et al. Miti-
placement in a patient undergoing thoracotomy gating the impact of acute and chronic post-
who had the relative contraindication of thrombocy- thoracotomy pain. J Cardiothorac Vasc Anesth
topenia to epidural placement. J Cardiothorac Vasc 2014;28:1060–8.
Anesth 2012;26:666–8. 71. Nomori H, Horio H, Suemasu K. Anterior limited tho-
57. Naja Z, Lönnqvist PA. Somatic paravertebral nerve racotomy with intrathoracic illumination for lung can-
blockade. Incidence of failed block and complica- cer: its advantages over anteroaxillary and
tions. Anaesthesia 2001;56(12):1184–8. posterolateral thoracotomy. Chest 1999;115:874–80.
58. Chelly JE. Paravertebral blocks. Anesthesiol Clin 72. Nomori H, Horio H, Suemasu K. Intrathoracic light-
2012;30:75–90. assisted anterior limited thoracotomy in lung cancer
59. Kus A, Gurkan Y, Gul Akgul A, et al. Pleural puncture surgery. Surg Today 1999;29:606–9.
and intrathoracic catheter placement during ultra- 73. Landreneau RJ, Pigula F, Luketich JD, et al. Acute
sound guided paravertebral block. J Cardiothorac and chronic morbidity differences between
Vasc Anesth 2013;27(2):11–2. muscle-sparing and standard lateral thoracotomies.
60. Romero A, Garcia JE, Joshi GP. The state of the art J Thorac Cardiovasc Surg 1996;112:1346–50.
in preventing postthoracotomy pain. Semin Thorac 74. Khan IH, McManus KG, McCraith A, et al. Muscle
Cardiovasc Surg 2013;25:116–24. sparing thoracotomy: a biomechanical analysis con-
61. Ding X, Jin S, Niu X, et al. A comparison of the anal- firms preservation of muscle strength but no
gesia efficacy and side effects of paravertebral improvement in wound discomfort. Eur J Cardio-
compared with epidural blockade for thoracotomy: thorac Surg 2000;18:656–61.
Pain Management Following Thoracic Surgery 409
75. Wildgaard K, Ravn J, Kehlet H. Chronic post- 77. Cerfolio RJ, Bryant AS, Maniscalco LM.
thoracotomy pain: a critical review of pathogenic A nondivided intercostal muscle flap further reduces
mechanisms and strategies for prevention. Eur J pain of thoracotomy: a prospective randomized trial.
Cardiothorac Surg 2009;36:170–80. Ann Thorac Surg 2008;85:1901–6.
76. Cerfolio RJ, Bryant AS, Patel B, et al. Intercostal 78. Deng G, Rusch V, Vickers A, et al. Randomized
muscle flap reduces the pain of thoracotomy: a pro- controlled trial of a special acupuncture technique
spective randomized trial. J Thorac Cardiovasc Surg for pain after thoracotomy. J Thorac Cardiovasc
2005;130:987–93. Surg 2008;136:1464–9.
T h e Pre v e n t i o n a n d
M a n a g e m e n t o f A i r Le a k s
F o l l o w i n g Pu l m o n a r y R e s e c t i o n
Bryan M. Burt, MDa, Joseph B. Shrager, MDb,c,*
KEYWORDS
Pulmonary resection Postoperative Air leak Alveolar pleural fistula
KEY POINTS
Based on preoperative risk factors, selected patients should be considered for intraoperative tech-
niques to minimize air leaks and the residual spaces that predispose to prolonged air leaks,
including pleural tenting and pneumoperitoneum.
There is insufficient evidence for the routine use of surgical sealants following pulmonary resection;
we recommend buttressing staple lines in nonanatomic pulmonary resections for patients with
moderate to severe emphysema (forced expiratory volume in 1 second <60% of predicted) to pre-
vent prolonged air leaks.
In postoperative patients with less than a large air leak and no more than a small pneumothorax,
algorithms incorporating no applied external suction or alternating suction likely reduce the duration
of air leak.
Initial evaluation of digital drainage systems suggest that their use may result in shorter duration of
air leak, duration of chest tube, and length of stay.
Most prolonged alveolar air leaks resolve with time and tube drainage alone, and a trial of a few
weeks of watchful waiting incorporating a Heimlich valve is reasonable in the outpatient setting.
a
Division of Thoracic Surgery, Baylor College of Medicine, One Baylor Plaza, BCM 390, Houston, TX 77030,
USA; b Division of Thoracic Surgery, Stanford Hospitals and Clinics, Stanford University School of Medicine,
300 Pasteur Drive, Falk Building CV-207, Stanford, CA 94305, USA; c Division of Thoracic Surgery, VA Palo
Alto Health Care System, Palo Alto, CA, USA
* Corresponding author. Division of Thoracic Surgery, Stanford Hospitals and Clinics, Stanford University
School of Medicine, 300 Pasteur Drive, Falk Building CV-207, Stanford, CA 94305.
E-mail address: shrager@stanford.edu
with serum. Three randomized trials in patients severe emphysema undergoing lobectomy. How-
undergoing lung resection have shown that pleural ever, as far as we know, this is purely theoretic,
tenting performed at the time of lung resection because the sealants have never been studied in
decreases chest tube duration and mean hospital this population. However, the fact that 1 study
stay5; decreases incidence of postoperative air shows substantially reduced air leaks in patients
leak6; and decreases air leak duration, chest after LVRS on the side treated with a sealant
tube duration, length of hospital stay, and hospital versus the control side not treated with sealants
costs.7 is suggestive.15
Table 2
Randomized trials evaluating no external suction (NES) algorithms following pulmonary resection
CXR Benefit of
Author Algorithm N Resections Evaluation NES Comments
Cerfolio NES on POD 2 33 Lobectomy, Yes Yes Greater sealing of
et al,25 2001 after 20 cm H2O sublobar AL by POD 3
Marshall NES after 20 cm 68 Lobectomy, Yes Yes Reduced duration
et al,26 2002 H2O while in OR sublobar of AL
Brunelli et al,27 NES on POD 1 145 Lobectomy No No Increased
2004 after 20 cm H2O complications
with NES
Brunelli Alternating 10 94 Lobectomy No Yes Fewer PALs, shorter
et al,28 2005 cm H2O (day) tube duration
and NES (night) and LOS
vs full-time NES,
after 10 cm H2O
Alphonso Immediate NES 239 Lobectomy, No Yes No differences in
et al,29 2005 sublobar PAL or tube
duration but
increased
mobilization
with NES
Leo NES or 15 cm 500 Lobectomy, Yes No (there Reduced PAL
et al,30 2013 H2O on POD 1, sublobar was a duration with
all tubes to benefit 15 cm H2O
Heimlich valve to 15 cm suction in the
on POD 3 H2O) anatomic
resection
subgroup
Abbreviations: AL, air leak; CXR, chest radiograph; OR, operating room; PAL, prolonged air leak; POD, postoperative day.
Data from Refs.25–30
suction at night with no suction during the day) or be difficult to differentiate a true parenchymal air
the application of no external suction after a brief leak from evacuation of a small residual pleural
period of low suction (either in the operating room space in the absence of an ongoing leak, and
only or overnight for the first night, are reasonable).1 from a momentum leak, which is the appearance
The senior author’s chest tube management has of leak created by momentum of the fluid column
evolved, for nearly all patients having lobectomy, in patients who are able to generate an unusually
to 10 cm H2O of suction for the first night strong cough. Although these momentum leaks
following surgery, then full-time water seal begin- typically are present only with coughing and not
ning the following morning regardless of air leak, with normal tidal breathing, and often are
unless the air leak is subjectively large. Patients observed only during the first several coughs a
who have air leaks undergo a chest radiograph 2 patient is asked to perform, these can cause
to 4 hours after water seal is initiated; they return confusion and delay chest tube removal and thus
to 10 cm H2O of suction only if there is a pneu- discharge to home.
mothorax more than 20% in size, increasing sub- Different companies have produced objective
cutaneous emphysema, or clinical signs of failure systems capable of precisely measuring the
of water seal (eg, new atrial fibrillation, dyspnea), amount of airflow through the chest tube. These
all of which rarely occur. devices express air leak in quantitative metrics
collected over longer periods of time, rather than
Pleural drainage systems by observation of bubbles, and they have the
In traditional chest drainage systems, air leaks are capability to record and retrieve information that
evaluated by detecting bubbles of air in the air leak may ultimately make it possible to standardize
chamber during forced expiratory maneuvers chest tube management across different surgeons
or cough. Using these systems, it can sometimes and institutions. These systems have been shown
416 Burt & Shrager
to significantly reduce the variability in deciding home safely on an outpatient drainage device
when to remove a chest drain31 and to decrease such as a Heimlich valve as early as postoperative
the duration of chest drain and hospital stay in ran- day 4.39,41
domized trials.32–34 Further, when using air leak
grading systems, the amount of air leak identified Noninvasive Management of Prolonged Air
in the early postoperative period can be effective Leak
in quantifying the risk of having persistent air leak
in the later postoperative period and may predict It is rare for aggressive reinterventions to be
which patients will not tolerate a no-external- required to treat PALs. In several published
suction algorithm.25,35 studies, the incidence of reoperation or other
Traditional pleural drainage systems deliver a aggressive reinterventions to treat this complica-
fixed level of suction independent from the level tion is less than 2%.1 The treatment strategy of
of intrapleural pressure, which can be variable watchful waiting, often as an outpatient, is largely
depending on several factors, including the col- successful. Approximately 95% of PALs that
umn of fluid in the pleural drainage system tubing. permit a no-external-suction algorithm resolve
Moreover, wide oscillations in the early postopera- within a few weeks of operation with chest tube
tive pleural pressures are associated with a high drainage alone, with only rare development of
risk of a PAL.36 Further, regulated pressure deliv- empyema.1
ered by the digital pleural drainage devices, in Heimlich valve
which the intrapleural pressure is maintained at a For patients with no more than a small, stable, and
consistent level within 0.1 cm H2O by a pressure asymptomatic pneumothorax on water seal, PALs
sensor, is capable of stabilizing the pressure in can be managed in the outpatient setting using a
the pleural cavity with minimal oscillations, and 1-way valve attached to the drain. If necessary to
this may promote accelerated sealing of air differentiate true air leak from residual space evac-
leak.37 In a single-institution randomized trial, uation, the patients can undergo a provocative
regulated suction ( 11 to 20 cm H2O) compared clamping trial; most of these patients are able to
with regulated seal ( 2 cm H2O) showed that regu- safely have their chest tubes removed. A report
lated seal is effective and as safe as regulated by Cerfolio and colleagues42 on 199 patients dis-
suction, with a trend toward decreased duration charged home with a persistent air leak and chest
of air leak in the regulated suction group.38 A sepa- tube placed to a suctionless portable drainage de-
rate multicenter and international randomized trial vice suggested that the air leak seals in the outpa-
of digital versus traditional drainage devices tient setting in almost all patients. For 9 patients
was recently completed. In patients undergoing with a persistent air leak after 2 weeks of outpa-
anatomic resection, patients randomized to digital tient management with a Heimlich valve, all pa-
drainage systems had significantly shorter air leak tients had their chest tubes removed without
duration, duration of chest tube placement, and sequela, some after a provocative clamping trial.
postoperative lengths of stay.32 If a period of watchful waiting for several weeks
(we have waited as long as 4 weeks) is unsuccess-
Chest tube removal ful in treating a PAL, or if no external suction is not
Chest tubes are traditionally removed when an air tolerated because of a larger leak, then the clini-
leak has resolved and when the pleural fluid cian must consider active interventions to me-
drainage is at an acceptable value to the thoracic chanically seal the site of the leak. Most of these
surgeon, which, for many surgeons, has been options are supported by expert consensus with
estimated to be in the range of 250 to 350 mL variable amounts of published data.
per day. Regarding the volume of fluid drainage,
there are now several studies that suggest that Pleurodesis and autologous blood patch
chest drains can be removed safely at much higher If the residual lung is fully expanded, chemical
levels of effluent. Data from 2 separate institutions pleurodesis with instillation via the thoracostomy
have recently shown that it is safe to remove chest tube of tetracycline, doxycycline, or talc can pro-
tubes following pulmonary resection when mote pleural symphysis and leak closure. These
drainage is as high as 450 to 500 mL per day, treatments all seem to be effective for pleurodesis
with exceedingly low rates of readmissions for in small cases series.43,44
symptomatic effusions.39,40 For patients who An autologous blood patch is another simple and
have a persistent air leak but are otherwise ready often effective treatment of PAL. Although some
for hospital discharge (and without large or symp- reports suggest an associated increased risk of
tomatic pneumothorax on water seal), it has been intrathoracic infection (in 1 study, 1 of 10 patients
shown that these patients can be discharged developed empyema45), several prospective data
Air Leaks Following Pulmonary Resection 417
suggest that an autologous blood patch has supe- on. A bronchoscopy should be done to rule out a
rior outcomes compared with the conservative bronchial rather than a parenchymal fistula. If the
management of PAL (summarized in Ref.46). Two residual lung is sufficiently normal, the leak can
of these studies were randomized trials that sug- be restapled or oversewn with good results, and
gested that intrapleural instillation of autologous the approach can be thoracoscopic or open. Other
blood (120 mL) decreased the duration of PAL, options include thoracoscopic applications of
time to chest tube removal, and time to hospital topical sealants,53 with which we have no experi-
discharge.45 ence. Decortication of surrounding lung may be
An additional randomized trial suggested that a required to facilitate full lung expansion. Chemical
100-mL instillation of autologous blood decreased pleurodesis or parietal pleurectomy/mechanical
the time to air leak cessation more than a 50-mL pleurodesis can be added when pleural apposition
instillation.47 We typically perform a blood patch can be achieved.54 If a residual space is present,
by the sterile instillation of approximately 150 mL that space should be obliterated with either
of blood freshly withdrawn into a 50-mL syringe muscle or omental transposition, although most
from a large-bore intravenous line, and instilled of the data on this are derived from the broncho-
immediately into the thoracostomy tube. Heparin pleural fistula literature. Following sublobar resec-
is typically not used because it prevents the tion, completion lobectomy is necessary on rare
necessary intrathoracic clotting of blood that occasions. Thoracoplasty or the creation of an
forms the patch, and we administer periprocedural open window can be considered under extreme
antibiotics to minimize the risk of infection. circumstances.
volume reduction surgery: lessons learned from the 18. Miller JI Jr, Landreneau RJ, Wright CE, et al.
National Emphysema Treatment Trial. Ann Thorac A comparative study of buttressed versus nonbut-
Surg 2006;82(1):197–206 [discussion: 206–7]. tressed staple line in pulmonary resections. Ann
5. Okur E, Kir A, Halezeroglu S, et al. Pleural tenting Thorac Surg 2001;71(1):319–22 [discussion: 323].
following upper lobectomies or bilobectomies of 19. Rocco G. Pleural partition with intrathoracic muscle
the lung to prevent residual air space and transposition (muscle tent) to manage residual
prolonged air leak. Eur J Cardiothorac Surg 2001; spaces after subtotal pulmonary resections. Ann
20(5):1012–5. Thorac Surg 2004;78(4):e74–6.
6. Allama AM. Pleural tent for decreasing air leak 20. Widmer MK, Krueger T, Lardinois D, et al.
following upper lobectomy: a prospective rando- A comparative evaluation of intrathoracic latissimus
mised trial. Eur J Cardiothorac Surg 2010;38(6): dorsi and serratus anterior muscle transposition.
674–8. Eur J Cardiothorac Surg 2000;18(4):435–9.
7. Brunelli A, Al Refai M, Monteverde M, et al. Pleural 21. Shrager JB, Wain JC, Wright CD, et al. Omentum is
tent after upper lobectomy: a randomized study of highly effective in the management of complex
efficacy and duration of effect. Ann Thorac Surg cardiothoracic surgical problems. J Thorac Cardio-
2002;74(6):1958–62. vasc Surg 2003;125(3):526–32.
8. Rocco G. Intraoperative measures for preventing re- 22. Gomez-Caro A, Calvo MJ, Lanzas JT, et al. The
sidual air spaces. Thorac Surg Clin 2010;20(3):371–5. approach of fused fissures with fissureless tech-
9. Toker A, Dilege S, Tanju S, et al. Perioperative pneu- nique decreases the incidence of persistent air
moperitoneum after lobectomy – bilobectomy opera- leak after lobectomy. Eur J Cardiothorac Surg
tions for lung cancer: a prospective study. Thorac 2007;31(2):203–8.
Cardiovasc Surg 2003;51(2):93–6. 23. Cooper JD, Patterson GA. Lung-volume reduction
10. Carboni GL, Vogt A, Kuster JR, et al. Reduction of surgery for severe emphysema. Chest Surg Clin N
airspace after lung resection through controlled Am 1995;5(4):815–31.
paralysis of the diaphragm. Eur J Cardiothorac 24. Cooper JD, Patterson GA, Sundaresan RS, et al.
Surg 2008;33(2):272–5. Results of 150 consecutive bilateral lung volume
11. Belda-Sanchis J, Serra-Mitjans M, Iglesias reduction procedures in patients with severe emphy-
Sentis M, et al. Surgical sealant for preventing sema. J Thorac Cardiovasc Surg 1996;112(5):1319–
air leaks after pulmonary resections in patients 29 [discussion: 1329–30].
with lung cancer. Cochrane Database Syst Rev 25. Cerfolio RJ, Bass C, Katholi CR. Prospective
2010;(1):CD003051. randomized trial compares suction versus water seal
12. Allen MS, Wood DE, Hawkinson RW, et al. Prospec- for air leaks. Ann Thorac Surg 2001;71(5):1613–7.
tive randomized study evaluating a biodegradable 26. Marshall MB, Deeb ME, Bleier JI, et al. Suction vs
polymeric sealant for sealing intraoperative air leaks water seal after pulmonary resection: a randomized
that occur during pulmonary resection. Ann Thorac prospective study. Chest 2002;121(3):831–5.
Surg 2004;77(5):1792–801. 27. Brunelli A, Monteverde M, Borri A, et al. Compar-
13. Merritt RE, Singhal S, Shrager JB. Evidence-based ison of water seal and suction after pulmonary lo-
suggestions for management of air leaks. Thorac bectomy: a prospective, randomized trial. Ann
Surg Clin 2010;20(3):435–48. Thorac Surg 2004;77(6):1932–7 [discussion:
14. Malapert G, Hanna HA, Pages PB, et al. Surgical 1937].
sealant for the prevention of prolonged air leak after 28. Brunelli A, Sabbatini A, Xiume F, et al. Alternate suc-
lung resection: meta-analysis. Ann Thorac Surg tion reduces prolonged air leak after pulmonary lo-
2010;90(6):1779–85. bectomy: a randomized comparison versus water
15. Moser C, Opitz I, Zhai W, et al. Autologous fibrin seal. Ann Thorac Surg 2005;80(3):1052–5.
sealant reduces the incidence of prolonged air 29. Alphonso N, Tan C, Utley M, et al. A prospective ran-
leak and duration of chest tube drainage after lung domized controlled trial of suction versus non-
volume reduction surgery: a prospective random- suction to the under-water seal drains following
ized blinded study. J Thorac Cardiovasc Surg lung resection. Eur J Cardiothorac Surg 2005;
2008;136(4):843–9. 27(3):391–4.
16. Stammberger U, Klepetko W, Stamatis G, et al. But- 30. Leo F, Duranti L, Girelli L, et al. Does external
tressing the staple line in lung volume reduction sur- pleural suction reduce prolonged air leak after
gery: a randomized three-center study. Ann Thorac lung resection? Results from the AirINTrial after
Surg 2000;70(6):1820–5. 500 randomized cases. Ann Thorac Surg 2013;
17. Hazelrigg SR, Boley TM, Naunheim KS, et al. Effect 96(4):1234–9.
of bovine pericardial strips on air leak after stapled 31. Varela G, Jimenez MF, Novoa NM, et al. Postopera-
pulmonary resection. Ann Thorac Surg 1997;63(6): tive chest tube management: measuring air leak
1573–5. using an electronic device decreases variability in
Air Leaks Following Pulmonary Resection 419
the clinical practice. Eur J Cardiothorac Surg 2009; Ann Thorac Surg 2009;87(6):1690–4 [discussion:
35(1):28–31. 1694–6].
32. Pompili C, Detterbeck F, Papagiannopoulos K, et al. 43. Kilic D, Findikcioglu A, Hatipoglu A. A different
Multicenter international randomized comparison of application method of talc pleurodesis for the treat-
objective and subjective outcomes between elec- ment of persistent air leak. ANZ J Surg 2006;76(8):
tronic and traditional chest drainage systems. Ann 754–6.
Thorac Surg 2014;98(2):490–6 [discussion: 496–7]. 44. Read CA, Reddy VD, O’Mara TE, et al. Doxycycline
33. Brunelli A, Salati M, Refai M, et al. Evaluation of a new pleurodesis for pneumothorax in patients with AIDS.
chest tube removal protocol using digital air leak moni- Chest 1994;105(3):823–5.
toring after lobectomy: a prospective randomised trial. 45. Shackcloth MJ, Poullis M, Jackson M, et al. Intra-
Eur J Cardiothorac Surg 2010;37(1):56–60. pleural instillation of autologous blood in the treat-
34. Cerfolio RJ, Bryant AS. The benefits of continuous ment of prolonged air leak after lobectomy: a
and digital air leak assessment after elective pulmo- prospective randomized controlled trial. Ann Thorac
nary resection: a prospective study. Ann Thorac Surg 2006;82(3):1052–6.
Surg 2008;86(2):396–401. 46. Chambers A, Routledge T, Bille A, et al. Is blood
35. Cerfolio RJ, Bryant AS, Singh S, et al. The manage- pleurodesis effective for determining the cessation
ment of chest tubes in patients with a pneumothorax of persistent air leak? Interact Cardiovasc Thorac
and an air leak after pulmonary resection. Chest Surg 2010;11(4):468–72.
2005;128(2):816–20. 47. Andreetti C, Venuta F, Anile M, et al. Pleurodesis with
36. Brunelli A, Cassivi SD, Salati M, et al. Digital mea- an autologous blood patch to prevent persistent air
surements of air leak flow and intrapleural pressures leaks after lobectomy. J Thorac Cardiovasc Surg
in the immediate postoperative period predict risk of 2007;133(3):759–62.
prolonged air leak after pulmonary lobectomy. Eur J 48. Carbognani P, Spaggiari L, Solli PG, et al. Postoper-
Cardiothorac Surg 2011;39(4):584–8. ative pneumoperitoneum for prolonged air leaks and
37. Brunelli A, Beretta E, Cassivi SD, et al. Consensus residual spaces after pulmonary resections.
definitions to promote an evidence-based approach J Cardiovasc Surg 1999;40(6):887–8.
to management of the pleural space. A collaborative 49. De Giacomo T, Rendina EA, Venuta F, et al. Pneumo-
proposal by ESTS, AATS, STS, and GTSC. Eur J Car- peritoneum for the management of pleural air space
diothorac Surg 2011;40(2):291–7. problems associated with major pulmonary resec-
38. Brunelli A, Salati M, Pompili C, et al. Regulated tions. Ann Thorac Surg 2001;72(5):1716–9.
tailored suction vs regulated seal: a prospective ran- 50. Kovitz KL, French KD. Endobronchial valve place-
domized trial on air leak duration. Eur J Cardiothorac ment and balloon occlusion for persistent air leak:
Surg 2013;43(5):899–904. procedure overview and new current procedural ter-
39. Cerfolio RJ, Bryant AS. Results of a prospective al- minology codes for 2013. Chest 2013;144(2):661–5.
gorithm to remove chest tubes after pulmonary 51. Travaline JM, McKenna RJ Jr, De Giacomo T, et al.
resection with high output. J Thorac Cardiovasc Treatment of persistent pulmonary air leaks using
Surg 2008;135(2):269–73. endobronchial valves. Chest 2009;136(2):355–60.
40. Bjerregaard LS, Jensen K, Petersen RH, et al. Early 52. Firlinger I, Stubenberger E, Muller MR, et al. Endo-
chest tube removal after video-assisted thoracic sur- scopic one-way valve implantation in patients with
gery lobectomy with serous fluid production up to prolonged air leak and the use of digital air leak
500 ml/day. Eur J Cardiothorac Surg 2014;45(2): monitoring. Ann Thorac Surg 2013;95(4):1243–9.
241–6. 53. Thistlethwaite PA, Luketich JD, Ferson PF, et al.
41. McKenna RJ Jr, Mahtabifard A, Pickens A, et al. Ablation of persistent air leaks after thoracic proce-
Fast-tracking after video-assisted thoracoscopic dures with fibrin sealant. Ann Thorac Surg 1999;
surgery lobectomy, segmentectomy, and pneumo- 67(2):575–7.
nectomy. Ann Thorac Surg 2007;84(5):1663–7 [dis- 54. Suter M, Bettschart V, Vandoni RE, et al. Thoraco-
cussion: 1667–8]. scopic pleurodesis for prolonged (or intractable)
42. Cerfolio RJ, Minnich DJ, Bryant AS. The removal of air leak after lung resection. Eur J Cardiothorac
chest tubes despite an air leak or a pneumothorax. Surg 1997;12(1):160–1.
B ro n c h o p l e u r a l Fi s t u l a
a n d Em p y e m a A f t e r
Anatomic Lung Resection
Giorgio Zanotti, MDa, John D. Mitchell, MDb,*
KEYWORDS
Empyema Bronchopleural fistula Eloesser flap Decortication Clagett procedure
KEY POINTS
Bronchopleural fistula (BPF) and empyema remain rare but serious complications after anatomic
lung resection, particularly pneumonectomy.
Careful attention to identified risk factors and proper surgical technique can minimize the risk of
BPF/empyema in most cases.
Management of BPF/empyema after surgical resection must address both issues of bronchial
integrity and the infected plural space for a successful outcome.
After pneumonectomy, resolution may require multiple, additional surgical interventions.
Empyema after anatomic pulmonary resection re- ETIOLOGY, RISK FACTORS, AND PREVENTION
mains a rare but serious complication, often
leading to major morbidity and increased mortality. A BPF may arise either from dehiscence or disrup-
In a modern series of 1023 patients undergoing tion of a bronchial closure after anatomic lung
anatomic resection, empyema occurred postoper- resection (segmentectomy, lobectomy, pneumo-
atively in 1.1%.1 It occurs more commonly after nectomy), or from anastomotic dehiscence after
pneumonectomy, particularly after surgery for bronchoplastic resection. Postoperative BPF is
benign disease.2 The reported incidence of de- classified based on the time of onset after surgery
pends in part on the postoperative surveillance as early (within the first week), intermediate
protocols and diagnostic techniques used.3,4 The (between 7 and 30 days), and late (after 30 days).8
associated mortality rate may exceed 10%5; even There are a number of predisposing factors that
if the patient survives, the recurrence rate of infec- may place the patient at increased risk of developing
tion can be as high as 38%.6 Importantly, up to a fistula and subsequent empyema. Malnutrition,
80% of cases of procedure-related empyemas various immunosuppressive therapies (steroids,
are associated with bronchopleural fistula (BPF)4 antimetabolites), prior thoracic radiation therapy,
and fewer than 20% of these can be expected to poorly controlled pulmonary or pleural infection,
close spontaneously.7 The presence or absence active smokers, and the use of induction chemo-
of BPF in the setting of postoperative pleural empy- therapy have all been implicated in the development
ema defines 2 clinical cohorts that are distinct with of fistula. Interestingly, induction therapy has been
respect to etiology, risk factors, and treatment cited as a risk factor after pneumonectomy,9 but
algorithm. not after bronchoplastic procedures.10–12
thoracic.theclinics.com
Early fistulas are most commonly owing to surgi- (exudative, acute), exudative fluid is present, the
cal technical problems. It is well-established that visceral pleural remains elastic and the dimensions
right pneumonectomy is associated with a greater of the chest cavity are maintained. Stage 2 (fibrino-
risk (up to 13.2%) of BPF compared with left pneu- purulent, subacute) is characterized by the pres-
monectomy (up to 5.0%).13 There are 2 main ence of infected or frankly purulent fluid, and
reasons accounting for this; first, the most com- fibrin deposition creates septations and locula-
mon anatomic variant of bronchial arterial supply tions within the pleural cavity. Lung compliance
is composed of 1 artery on the right, whereas a may also be reduced owing to thick fibrin deposi-
dual arterial supply is the most common configura- tions. In stage 3 (consolidative, chronic), granula-
tion on the left. Second, the left main stem bron- tion tissue formally replaces the pleural space,
chus is protected under the aortic arch and and the lung becomes completely entrapped by
surrounded by its vascularized mediastinal tissue, a fibrinous peel. Late in the course, organization
whereas the right bronchial stump has no such of the inflammatory tissue causes contraction of
coverage. Overzealous mediastinal lymphadenec- the affected hemithorax with ipsilateral shift of
tomy,14 bronchial stump greater than 25 mm in the mediastinum, elevation of the diaphragm,
diameter,15 long bronchial stump, residual malig- and narrowing of intercostal spaces.
nancy at the bronchial margin,5 requirement for 4 This time course, characteristic of a common or
or more units of intraoperative packed red blood postpneumonic empyema, is usually altered by
cell transfusions,3 completion pneumonectomy,3 the postoperative medical attention these patients
and tension along the anastomosis are associated receive, particularly with the presence of a BPF. It
with stump ischemia and are well-described risk is imperative that the treating surgeon be
factors for early failure. Stapler misfiring, improper acquainted with the often subtle symptoms and
tissue apposition, or poorly secured sutures are signs that can lead to early diagnosis and treatment.
also common technical causes of early bronchial
anastomotic breakdown. CLINICAL PRESENTATION AND DIAGNOSIS
Postoperatively, the main risk factor for BPF is
positive pressure mechanical ventilation16; for Postoperative empyema of the pleural space is
this reason, extubation at the end of the case is associated with a constellation of signs and symp-
typically a priority. toms that are dictated mainly by the presence of
Late fistulas are typically secondary to patient- an infected pleural cavity and a BPF. If a BPF
related factors causing poor healing: age greater has indeed opened, its size and timing of formation
than 60 years, malnutrition, ongoing pulmonary are major determinants of the clinical picture. The
or pleural infection, and recurrence of malignancy. duration of illness is also an important determinant
Empyema after anatomic lung resection in of the clinical manifestations.
absence of BPF is most commonly caused by intra- Minor fistulae may be occult or minimally symp-
operative contamination by aerobic bacteria. Zah- tomatic and are usually detected if a postoperative
eer and colleagues17 and Eerola and associates18 bronchoscopic screening for BPF is performed
found that Staphylococcus aureus is the most com- routinely.22 A persistent (>7 days) air leak, espe-
mon organism isolated, followed by Streptococcus cially if brisk, without a history of visceral pleural
pneumoniae. Spillage of infected bronchial secre- dissection or injury at the time of surgery, new
tions, active plural infection at the time of surgery, evidence of pneumomediastinum, a decrease in
and esophagopleural or gastropleural fistulas are the fluid level in the ipsilateral pleural cavity, or a
the most common causes. Less frequently, a pri- new air–fluid level (“meniscus sign”) at the height
mary infection of pleural space, because it may of the bronchial stump on chest imaging after
occur after chest trauma with chest wall penetration lung resection should raise the suspicion for BPF
or hemothorax and mycobacterial infection of the (Fig. 1). In cases where the BPF is larger than a
pleural cavity,19 may lead to bronchial stump break- few millimeters, respiratory distress may be a
down. There is scant evidence that hematogenous prominent finding and is caused by either spillage
infection of the pleural space can occur from a of pleural fluid through the fistula into the contralat-
distant infection site (classically osteomyelitis) eral lung or by “dead space” ventilation into the
without an intermediate lung infection, which in empty pleural space. Worsening dyspnea and pro-
turn contaminates the pleural cavity.20 ductive cough with frothy or purulent sputum her-
ald the loss of integrity of the bronchial closure.
PATHOLOGY Expectoration typically worsens with the patient
lying on the side opposite to the one involving
It is worth reviewing the classic time course and the fistula. The flooding of the contralateral lung
stages of empyema development.21 In stage 1 may lead to, if not overt pneumonia, an alveolar
Bronchopleural Fistula and Empyema 423
Fig. 1. Occurrence of an empyema associated with bronchopleural fistula after right pneumonectomy. In this
supine patient, note the air–fluid interface at the level of the right mainstem stump closure.
injury similar to respiratory distress syndrome that diagnosing selected cases of BPF but require sub-
is almost invariably a lethal condition. Fever, stantial time and cooperation from a nonintubated
increasing white blood cell count, and indications patient and can be inconclusive in the setting of
of systemic inflammation are often present small fistulas or underlying lung disease such
because of the infected pleural space; pleuritic as chronic obstructive pulmonary disease.25 In
chest pain, night sweats, and chills may also be almost all cases, bronchoscopy is used to assess
observed and resemble symptoms of pneumonia. and confirm the presence of a BPF. A loss of integ-
Interestingly, in some patients decompression of rity of or fine bubbling at the anastomosis after
the infected contents of the pleural space into instillation of saline solution into the stump is
the airway may temporize some features of the pathognomonic. Delayed or less extensive bron-
systemic inflammatory response, thus making chial disruptions may present diagnostic prob-
the diagnosis more difficult. lems, even at the time of bronchoscopy. Rarely,
Early after surgery, the diagnosis of BPF is sug- surgical reexploration is needed as a final step to
gested by a combination of clinical and radiologic confirm or rule out the diagnosis.
clues and is eventually confirmed by visualization
of the fistula on bronchoscopy. Computed tomog- MANAGEMENT
raphy of the chest is quite sensitive at demon-
strating abnormalities related to the presence of Empyema of the chest is one of the oldest known
BPF, but its sensitivity at demonstrating the pres- general thoracic surgery conditions. Hippocrates
ence and the location of the fistula (at least after (460–370 BC) first described the drainage of
lobectomy) is approximately 50%23 and therefore patients with pleural empyema more than 2000
is of little to no use in the diagnostic process. In years ago. He thought that drainage by either an
these instances, the use of noninvasive 133xenon intercostal incision or rib resection followed by
ventilation scintigraphy has been used by some recurrent plural irrigation was necessary.26,27 Ga-
authors24 to diagnose occult BPFs by visualizing lenus (130–200 BC) and Celsus (25 BC-50 AD) in
equilibration of radioactive gas tracer into the the Roman age devised metal tubes to drain the
empty pleural space. Alternatively, ventilation purulent collection, a teaching that ruled the Mid-
scintigraphy with other radioactive tracers such dle Ages, during which barber–surgeons refined
as 81krypton and 99technetium has also been re- the art of drainage by using solutions with mildly
ported. These advanced imaging techniques that acidic pH (often wine).27 The magnitude of Hippo-
use radioactive isotopes may be useful in crates’s contribution to the understanding of chest
424 Zanotti & Mitchell
empyema is underscored by the fact that few true debilitated, an open thoracostomy (Eloesser flap)
advances in either the diagnosis or the cure of this may be used to treat the infected postlobectomy re-
condition were made in the following 2000 years. sidual space. After pneumonectomy, these same
In the modern era, the management of postop- options for tissue transposition or thoracostomy
erative empyema of the pleural cavity depends window are present, although these authors prefer
on the etiology (including the presence or absence the latter because it is much simpler and is associ-
of a BPF), the chronicity, the state of the underlying ated with less morbidity.
lung (if any is left), and the patient’s clinical If a BPF is present, concurrent management of
and nutritional status. Importantly, the protean na- the empyema as well as BPF repair is considered
ture of the disease has made it difficult to draft the current standard of care. The basic principles
evidence-based guidelines for the treatment of include initial drainage (described elsewhere in
this condition. Therefore, the optimal treatment this article), antibiotics and optimization of nutri-
strategy needs to be individualized in each case tion; closure of the fistula, typically with autolo-
based on the aforementioned evidence, the dura- gous tissue buttressing; and appropriate
tion of illness, and the healing potential of the management of the infected space. This is usually
infected pleural cavity. To minimize morbidity performed through a posterolateral thoracotomy,
and mortality, the treatment must proceed with a although a BPF after pneumonectomy may be ap-
sense of urgency. proached preferentially via sternotomy. In poor
The initial management of an acute, post lung surgical candidates, chronic empyema can be
resection empyema consists of placement of a managed with a large-bore chest tube to be slowly
large-bore (32–36F) thoracostomy tube and posi- retracted over the course of weeks, whereas a
tioning of the patient in reverse Trendelenburg po- small air leak is managed expectantly or with
sition. In the case of BPF, the affected side should open-window thoracostomy. Open-window thora-
be “down” in the most dependent position to costomy was first described by Robinson29 in
prevent spillage of pleural fluid into the contralat- 1916 for nontuberculous empyema and subse-
eral lung. When mechanical ventilation is needed, quently revised in 1935 by Eloesser30 for tubercu-
particularly after pneumonectomy, consideration lous empyema. Several iterations of the Eloesser
should be given to selective intubation of the unin- flap have been described since that time; the
volved side to minimize barotrauma to the bron- authors favor a modified “H”-type incision with un-
chial stump. derlying resection of segments of 2 to 3 ribs. The
If the presence of a BPF is known or if skin flaps created by the incision are then used
purulent fluid is found, initiation of intravenous, to epithelialize the entryway into the pleural space.
broad-spectrum antibiotic therapy is undertaken. The opening should be placed, if possible, low and
The British Thoracic Society recently published anterior in the chest to facilitate drainage. Often,
the most updated guidelines on how to direct anti- this approach needs to be modified owing to prior
biotic therapy in this complex arena.28 Antibiotic thoracotomy incisions. Additionally, the authors
treatment should be tailored based on the results preserve the serratus anterior muscle under the
of cultures and their sensitivity profile. cephalad skin flap for use at the time of thoracos-
Treatment options diverge from this point tomy closure (Fig. 2).
depending on the presence or absence of an asso- At the time of thoracotomy, the bronchial stump
ciated BPF. In the absence of BPF, treatment is inspected with particular attention paid to its
options closely mirror conventional management length. If the stump is found to be long, it should
of simple empyema: drainage and intravenous be resected back to its origin, reclosed either
antibiotics—and, if reoperation is deemed neces- with suture or staples, and buttressed with autolo-
sary, debridement of the pleural space and mini- gous tissue. If flush with its origin, it needs to be
mizing the residual pleural space. Of these, it is reclosed using interrupted, absorbable suture
the residual pleural space, obligatory after anatomic and then covered with a vascularized flap. Dissec-
lung resection, that may prove problematic. In the tion of a pneumonectomy stump for reclosure can
setting of segmentectomy, this is rarely a concern. be hazardous, given the dense fibrosis typically
After lobectomy, the residual intrathoracic space present and the proximity of the ligated pulmonary
may be addressed with adjuncts such as temporary artery. If the stump cannot be reclosed, options
phrenic nerve paresis or pneumoperitoneum. In re- include a pedicled flap that can be sewn directly
fractory (or chronic) cases, the use of muscle or to the stump and function as a plug17 or a central
omental transposition with or without thoracoplasty bronchoplastic procedure, such as carinal
may be used to obliterate the infected intrathoracic resection.
space. If the options regarding tissue transposi- The choice of the autologous tissue for
tion are limited, or if the patient is significantly coverage is important for optimal results, and in
Bronchopleural Fistula and Empyema 425
SUMMARY
part depends on the patient’s prior surgical his-
Postoperative empyema is a relatively rare compli-
tory. Most commonly, rotational muscle flaps are
cation after anatomic lung resection, but when pre-
used involving the latissimus dorsi, the pectoralis
sent is associated with considerable morbidity. For
major, or intercostal muscle. Free flaps are used
many patients, the presence or absence of a BPF
in patients that are not candidates for pedicled
makes the difference between recovery, chronic
tissue transfer.31,32 The rotational muscle flaps
illness, and death. Despite this, the principles un-
are typically harvested before placement of the
derlying treatment have not changed: source
retractor. Intercostal muscle flaps are versatile,
control of the infection with drainage (tube thora-
and are particularly well-suited to be combined
costomy and/or video-assisted thoracoscopic sur-
with open thoracostomy, because the small foot-
gery vs open drainage), closure of the fistula if
print of the intercostal en route to the bronchus
present, and sterilization of the pleural cavity either
will not interfere with postoperative packing of
via decortication, filling the space with autologous
the intrathoracic cavity.
tissue or with the use of the Clagett procedure.
Pleural flaps,33 although being frequently used
There is a lack of evidence over optimal treatment
owing to their simplicity of harvest, have the disad-
strategy because no large or randomized trials
vantage of being extremely thin and potentially
have been performed to date. The clinical scenario
lacking adequate blood supply. Their use is, there-
and surgeon’s acquaintance with the various treat-
fore, questionable for more critical purposes such
ment techniques are the main factors influencing
as buttressing a postpneumonectomy stump.
the treatment strategy in a case-tailored approach.
Some authors reported the use of diaphragmatic34
or omental flaps35 with good results; these authors
have found the latter particularly useful in the REFERENCES
setting of poorly controlled infection. Brewer and
associates36 and Taghavi and colleagues37 re- 1. Allen MS, Darling GE, Pechet TT, et al. Morbidity and
ported the use pericardial flaps for bronchial mortality of major pulmonary resections in patients
stump coverage with remarkable results. This, with early-stage lung cancer: initial results of the
however, exposed the mediastinum to the infected randomized, prospective ACOSOG Z0030 trial.
milieu of the pleural cavity with inherent risk of Ann Thorac Surg 2006;81(3):1013–9 [discussion:
phrenic nerve injury during the harvest. 1019–20].
The management of an open thoracostomy is 2. Shapiro M, Swanson SJ, Wright CD, et al. Predictors of
routine, packing the space once or twice daily with major morbidity and mortality after pneumonectomy
gauze typically sprinkled with quarter-strength utilizing the society for thoracic surgeons general
426 Zanotti & Mitchell
thoracic surgery database. Ann Thorac Surg 2010; 18. Eerola S, Virkkula L, Varstela E. Treatment of post-
90(3):927–35. pneumonectomy empyema and associated bron-
3. Deschamps C, Bernard A, Nichols FC 3rd, et al. chopleural fistula. Experience of 100 consecutive
Empyema and bronchopleural fistula after pneumo- postpneumonectomy patients. Scand J Thorac Car-
nectomy: factors affecting incidence. Ann Thorac diovasc Surg 1988;22(3):235–9.
Surg 2001;72(1):243–7 [discussion: 248]. 19. Pomerantz M, Madsen L, Goble M, et al. Surgical
4. Vallieres E. Management of empyema after lung re- management of resistant mycobacterial tuberculosis
sections (pneumonectomy/lobectomy). Chest Surg and other mycobacterial pulmonary infections. Ann
Clin N Am 2002;12(3):571–85. Thorac Surg 1991;52(5):1108–11 [discussion: 1112].
5. Liberman M, Cassivi SD. Bronchial stump dehis- 20. Sherman MM, Subramanian V, Berger RL. Manage-
cence: update on prevention and management. ment of thoracic empyema. Am J Surg 1977;133(4):
Semin Thorac Cardiovasc Surg 2007;19(4):366–73. 474–9.
6. Weber J, Gräbner D, al-Zand K, et al. Empyema af- 21. Sahn SA. Diagnosis and management of parapneu-
ter pneumonectomy–empyema window or thoraco- monic effusions and empyema. Clin Infect Dis 2007;
plasty? Thorac Cardiovasc Surg 1990;38(6):355–8. 45(11):1480–6.
7. Shields T. General thoracic surgery. Baltimore (MD): 22. Bylicki O, Vandemoortele T, Orsini B, et al. Incidence
Williams and Wilkins; 1994. and management of anastomotic complications af-
8. Varoli F, Roviaro G, Grignani F, et al. Endoscopic ter bronchial resection: a retrospective study. Ann
treatment of bronchopleural fistulas. Ann Thorac Thorac Surg 2014;98(6):1961–7.
Surg 1998;65(3):807–9. 23. Westcott JL, Volpe JP. Peripheral bronchopleural fis-
9. Martin J, Ginsberg RJ, Abolhoda A, et al. Morbidity tula: CT evaluation in 20 patients with pneumonia,
and mortality after neoadjuvant therapy for lung can- empyema, or postoperative air leak. Radiology
cer: the risks of right pneumonectomy. Ann Thorac 1995;196(1):175–81.
Surg 2001;72(4):1149–54. 24. Pigula FA, Keenan RJ, Naunheim KS, et al. Diag-
10. Burfeind WR Jr, D’Amico TA, Toloza EM, et al. Low nosis of postpneumonectomy bronchopleural fistula
morbidity and mortality for bronchoplastic proce- using ventilation scintigraphy. Ann Thorac Surg
dures with and without induction therapy. Ann 1995;60(6):1812–4.
Thorac Surg 2005;80(2):418–21 [discussion: 422]. 25. Gaur P, Dunne R, Colson YL, et al. Bronchopleural
11. Ohta M, Sawabata N, Maeda H, et al. Efficacy and fistula and the role of contemporary imaging.
safety of tracheobronchoplasty after induction ther- J Thorac Cardiovasc Surg 2014;148(1):341–7.
apy for locally advanced lung cancer. J Thorac Car- 26. Aboud FC, Verghese AC. Evarts Ambrose Graham,
diovasc Surg 2003;125(1):96–100. empyema, and the dawn of clinical understanding
12. Rendina EA, Venuta F, De Giacomo T, et al. Safety of negative intrapleural pressure. Clin Infect Dis
and efficacy of bronchovascular reconstruction after 2002;34(2):198–203.
induction chemotherapy for lung cancer. J Thorac 27. Miller JI Jr. The history of surgery of empyema,
Cardiovasc Surg 1997;114(5):830–5 [discussion: thoracoplasty, Eloesser flap, and muscle flap
835–7]. transposition. Chest Surg Clin N Am 2000;10(1):
13. Darling GE, Abdurahman A, Yi QL, et al. Risk of a 45–53, viii.
right pneumonectomy: role of bronchopleural fistula. 28. Davies HE, Davies RJ, Davies CW, et al. Manage-
Ann Thorac Surg 2005;79(2):433–7. ment of pleural infection in adults: British Thoracic
14. Deslauriers J, Demers P. Postpneumonectomy Society Pleural Disease Guideline 2010. Thorax
empyema and bronchopleural fistula. In: Yang SC, 2010;65(Suppl 2):ii41–53.
Cameron DE, editors. Current therapy in thoracic 29. Robinson S. The treatment of chronic non-tuberculous
and cardiovascular Surgery. 1st. Edition. Philadelphia: chronic empyema. Surg Gynecol Obstet 1916;22:257.
Mosby; 2004. p. 304–6. 30. Eloesser L. An operation for tuberculous empyema.
15. Hollaus PH, Setinek U, Lax F, et al. Risk factors for Surg Gynecol Obstet 1935;60:1096.
bronchopleural fistula after pneumonectomy: stump 31. Walsh MD, Bruno AD, Onaitis MW, et al. The role of
size does matter. Thorac Cardiovasc Surg 2003; intrathoracic free flaps for chronic empyema. Ann
51(3):162–6. Thorac Surg 2011;91(3):865–8.
16. Wright CD, Wain JC, Mathisen DJ, et al. Postpneumo- 32. Takanari K, Kamei Y, Toriyama K, et al. Management
nectomy bronchopleural fistula after sutured bronchial of postpneumonectomy empyema using free flap
closure: incidence, risk factors, and management. and pedicled flap. Ann Thorac Surg 2010;89(1):
J Thorac Cardiovasc Surg 1996;112(5):1367–71. 321–3.
17. Zaheer S, Allen MS, Cassivi SD, et al. Postpneumonec- 33. Anderson TM, Miller JI Jr. Surgical technique and
tomy empyema: results after the Clagett procedure. application of pericardial fat pad and pericardio-
Ann Thorac Surg 2006;82(1):279–86 [discussion: phrenic grafts. Ann Thorac Surg 1995;59(6):
286–7]. 1590–1.
Bronchopleural Fistula and Empyema 427
34. Mineo TC, Ambrogi V. Early closure of the postpneu- fat graft reinforcement. J Thorac Surg 1953;26(5):
monectomy bronchopleural fistula by pedicled 507–32.
diaphragmatic flaps. Ann Thorac Surg 1995;60(3): 37. Taghavi S, Marta GM, Lang G, et al. Bronchial stump
714–5. coverage with a pedicled pericardial flap: an effec-
35. Okada M, Tsubota N, Yoshimura M, et al. Surgical tive method for prevention of postpneumonectomy
treatment for chronic pleural empyema. Surg Today bronchopleural fistula. Ann Thorac Surg 2005;79(1):
2000;30(6):506–10. 284–8.
36. Brewer LA 3rd, King EL, Lilly LJ, et al. Bronchial 38. Clagett OT, Geraci JE. A procedure for the manage-
closure in pulmonary resection: a clinical and ment of postpneumonectomy empyema. J Thorac
experimental study using a pedicled pericardial Cardiovasc Surg 1963;45:141–5.
Management of
Postoperative
R e s p i r a t o r y Fa i l u re
Michael S. Mulligan, MDa,*, Kathleen S. Berfield, MDa,
Ryan V. Abbaszadeh, MDb
KEYWORDS
Postoperative pneumonia Postpneumonectomy pulmonary edema
Acute respiratory distress syndrome (ARDS) Pulmonary embolism
KEY POINTS
Prevention of postoperative respiratory failure begins in the preoperative period.
Postoperative respiratory failure is associated with high morbidity and mortality.
Early recognition of signs and symptoms of postoperative respiratory failure is key to management
and successful outcomes.
a
Division of Cardiothoracic Surgery, University of Washington, Seattle, WA, USA; b Department of General
Surgery, University of Washington, Seattle, WA, USA
* Corresponding author. University of Washington Medical Center, 1959 Northeast Pacific Street, Box 356310,
Seattle, WA 98195-6310.
E-mail address: msmmd@u.washington.edu
mobility. Video-assisted thoracoscopic surgery ap- dyspnea, tachypnea, and new or progressive infil-
proaches are minimally invasive and do not cause trates on chest radiography.
the degree of respiratory muscle loss and lung func-
tion decline that open thoracotomy requires.2 Prevention
Postoperatively, an intensive pulmonary hygiene
Postoperative Factors
protocol, including targeted chest physiotherapy,
Adequate pain control without causing overseda- incentive spirometry, and early ambulation,
tion is necessary in the postoperative period to should be used to prevent atelectasis and pro-
facilitate early mobilization, clearance of secre- mote clearance of secretions. Atelectasis itself is
tions, and prevention of atelectasis. Use of an a known complication associated with pulmonary
epidural catheter, which is placed preoperatively resection and most atelectasis seen in the post-
as a measure of preemptive pain control for use operative period is subsegemental and of little
in the immediate postoperative period, is the clinical consequence. However, segmental atel-
standard of care for thoracic surgical procedures ectasis is less unlikely to improve with routine
at the authors’ institution. Local analgesic agents incentive spirometry and usually requires bron-
(bupivacaine or ropivacaine) and opiates can be choscopy. There is debate about the efficacy of
used in combination as continuous infusions with bronchoscopy in patients with abundant secre-
or without additional boluses administered by tions. Some studies have shown that little benefit
nursing or directly by the patient. The epidural was achieved when attempting to expel secre-
catheter is usually kept in place until the chest tions with bronchoscopy, although there may be
tubes are removed. In addition to effective anal- benefit for those patients whose presentation is
gesia, chest physiotherapy and aggressive pulmo- refractory to chest physiotherapy and intensive
nary hygiene also play important roles in clearing pulmonary hygiene.7
the airway and preventing atelectasis. The judi-
cious use of fluids, particularly in thoracic surgery, Treatment
can prevent fluid overload, pulmonary edema, and
The diagnosis and management of POP is similar
eventual respiratory failure.1–3
to that for hospital-acquired pneumonia. When
suspected, lower respiratory tract cultures should
POSTOPERATIVE PNEUMONIA be obtained and prompt initiation of broad-
Incidence, Mortality, and Risk Factors spectrum antibiotics should occur. POP is often
The incidence of postoperative pneumonia (POP) polymicrobial in nature, with Enterobacter, Staph-
has been cited as 2.2% to 6%4 in the literature ylococcus, and Streptococcus being the most
but this is a moving target because it can be diffi- common organisms. Anaerobic coverage should
cult to differentiate POP from the expected physi- also be considered following thoracoabdominal
ologic and radiographic responses to thoracic procedures. Close respiratory monitoring should
surgery normally seen in the postoperative period, be implemented in a patient with pneumonia and
such as fever, hypoxemia, atelectasis, and radio- supplemental oxygen administered if the patient
graphic abnormalities. Regardless, POP is associ- cannot appropriately oxygenate on room air. Me-
ated with significant mortality ranging from 20% to chanical ventilation may be required in cases of
50%5,6 as well as increased hospital length of stay. respiratory distress and respiratory failure.8
Patient-specific risk factors for development of
POP include preoperative hospitalization, immu- POSTPNEUMONECTOMY PULMONARY
nocompromised state, extent of the procedure, EDEMA
poor underlying cardiopulmonary reserve, smok-
ing history, and presence of atelectasis. Postpneumonectomy pulmonary edema (PPE) can
occur following pulmonary resection and is defined
as an acute, hypoxemic respiratory failure that is
Clinical Presentation
not cardiogenic in nature. It most commonly oc-
The onset of POP usually occurs within the first 5 curs following right pneumonectomy but can occur
postoperative days and there is some question following less extensive pulmonary resections,
whether the onset of POP is related to cessation including lobectomy. Its presentation is similar to
of routine postoperative antibiotic prophylaxis, acute respiratory distress syndrome (ARDS) and
which is generally continued through the first the same criteria are used for its diagnosis: bilateral
24 hours postoperatively.6 Initial signs and symp- infiltrates on chest radiograph, PaO2, fraction of
toms include the development of fever, hypoxemia, inspired oxygen less than 200, and pulmonary
leukocytosis, increased oropharyngeal secretions, capillary wedge pressure less than 18.9–11
Management of Postoperative Respiratory Failure 431
Incidence, Mortality, and Risk Factors associated with increased incidence of PPE as
well as disruption of lymphatic drainage due to
The incidence of PPE ranges from 2.2% to 7%.
the extent of dissection.15 Both hyperoxia and hyp-
More recent data suggest the average incidence
oxia have been implicated in the development of
following pneumonectomy is 3% to 4%, whereas
PPE. Hyperoxia of the contralateral lung during
lobectomy can be as low as 2%. Mortality in PPE
single-lung ventilation with the production of reac-
is high and has consistently been reported as
tive oxygen species and reactive nitrogen species
50% to 100%. Risk factors associated with
are implicated in ARDS. Additionally, hypoxic pul-
increased incidence of PPE are right pneumonec-
monary vasoconstriction and ischemia reperfu-
tomy, perioperative fluid overload, intraoperative
sion–type injury to the ipsilateral lung, which
transfusion of fresh frozen plasma, and high intra-
contribute to increased circulation of inflammatory
operative airway pressures.10–12
mediators, are thought to contribute to progression
of PPE or ARDS-like syndromes.14
Clinical Presentation or Pathogenesis
The onset of PPE is variable; however, most Treatment
patients will present within the first 72 hours post- As mentioned previously, the mortality rate of PPE
operatively.13 Early indications of PPE include an is extremely high. Thus, early identification and
increased oxygen requirement, tachypnea, management is essential. Inevitably, due to refrac-
tachycardia, and fever. Radiographic signs of tory hypoxia and hypercarbia, patients with PPE
PPE lag behind the clinical presentation and range require reintubation, mechanical ventilatory sup-
from patchy interstitial infiltrates to frank alveolar port, and initiation of lung protective ventilation
edema. Evidence of disease progression and with low tidal volumes and high positive end-
increasing dead space is hypercarbia and wors- expiratory pressure.16 Careful attention to peak
ening hypoxemia refractory to increasing oxygen airway pressures and avoidance of high pressures
levels. Early on, signs and radiographic findings are key due to the risk of further barotrauma and
may be similar to early POP, which may delay bronchial stump dehiscence. Several small studies
initial diagnosis of PPE.14 Pathophysiology of have also shown inhaled nitric oxide to be benefi-
PPE has been found to mirror ARDS with initial cial in the treatment of PPE due to its pulmonary
loss of endothelial integrity followed by capillary vasodilatory effects and improvement in arterial
leak and associated extravasation of proteins oxygenation. Additionally, inhaled nitric oxide has
and inflammatory cells into the interstitial space. been associated with improvement of acute pul-
This inflammatory reaction eventually leads to fi- monary hypertension and associated right ventric-
broproliferative and obliterative changes at the ular dysfunction that can be seen with ARDS.17
alveolar and microvascular level. This results in
interstitial fibrosis and remodeling of the pulmo-
PULMONARY EMBOLISM
nary vascular bed.14 Ongoing hypoxia and alveolar
damage leads to hypoxic pulmonary vasoconstric- Immobility and malignancy are common features
tion and development of acute pulmonary hyper- seen in patients undergoing lung resection or other
tension, which can lead to right ventricular strain intrathoracic procedures, predisposing patients to
and possible failure, which complicates the man- the development of pulmonary embolism (PE).
agement of PPE.14 Therefore, PE must be considered in the differen-
tial diagnosis when evaluating a patient with new
Prevention or ongoing postoperative respiratory failure.
Multiple factors have been implicated in the patho-
Incidence, Mortality, and Risk Factors
genesis of PPE and involve elements of patient
management throughout the perioperative period. Venous thromboembolism (VTE) is a major cause
Perioperative volume overload has long been of morbidity and mortality. Hospital-acquired
thought to precipitate the development of PPE, deep vein thrombosis (DVT) is reported to be
although studies have not yet demonstrated a clear 15% to 40% among general surgical patients
causal relationship. Regardless, judicious use of and is the second-most common postoperative
intravenous fluids continues to be the rule after medical complication.18 Patient-specific risk fac-
pneumonectomy and, certainly, volume overload tors include increasing patient age, history of
can exacerbate preexisting pulmonary edema by malignancy, prior VTE, smoking, hypertension,
increasing pulmonary capillary hydrostatic pres- obesity, and history of a hypercoagulable disorder.
sure.12 Surgical factors, such right-sided pneumo- The 2012 American College of Chest Physicians
nectomy and carinal pneumonectomy, have been (ACCP) guidelines stratifies surgical patients into
432 Mulligan et al
3 groups: low, moderate, and high risk for VTE. role for thrombolytic therapy should be considered
Generally, patients undergoing thoracic surgery for hemodynamically unstable patients with refrac-
fall into the moderate risk group and their risk of tory hypoxemia, perfusion defects, refractory
VTE in the absence of prophylaxis is estimated to hypotension, and acute right ventricular dysfunc-
be approximately 3%.19 tion. However, there are many contraindications
to thrombolytic therapy and these must be taken
Clinical Presentation into account before initiation of this therapy on a
The presentation of PE can vary widely and often case-by-case basis. Contraindications include
presents in an atypical fashion. The most classic prior intracranial hemorrhage, suspected aortic
presentation consists of the abrupt onset of pleuritic dissection, active bleeding, and ischemic stroke
chest pain, dyspnea, and hypoxia.19 The timing of within 3 months of the embolism.
and progression of symptoms is unpredictable Treatment of PE in a hemodynamically stable
and patients may asymptomatic or present with patient is systemic anticoagulation with LMWH is
progressive dyspnea or acute hemodynamic deteri- preferred to the use of intravenous or subcutane-
oration as the case may be for large pulmonary ous UFH. Concomitantly, an oral anticoagulant
emboli. Additionally, atypical signs and symptoms, such as warfarin should be initiated at the time of
such as fever, wheezing, hemoptysis, abdominal diagnosis. The LMWH or UFH should not be dis-
pain, orthopnea, and lower extremity swelling, continued until the international normalized ratio
may be present. When PE is suspected and the pa- is 2.0 for a minimum of 24 hours and not sooner
tient is hemodynamically normal, computed tomog- than 5 days after initiation. If it is the first episode
raphy (CT) pulmonary angiogram is the imaging of PE and there is an identifiable and temporary
modality of choice. However, in the patient who is risk factor such as surgery, a 3-month duration
hemodynamically unstable and cannot undergo of therapy is recommended. In the case of recur-
CT scan, empiric treatment with systemic anticoa- rent or unprovoked PE, lifelong anticoagulation
gulation is recommended and duplex ultrasonogra- may be indicated. However, this decision must
phy of the lower extremities to look for DVT or be weighed carefully and the patient’s risk of
echocardiogram to evaluate for signs of right heart bleeding and medical compliance taken into
strain can be used.20 account.22,23
Routine placement of inferior vena cava (IVC)
Prevention filter is not recommended for prevention or man-
agement of PE. However, in the presence of a
Primary prevention of perioperative VTE or PE contraindication, prior complication, or recurrent
is multifaceted. Pharmacologic and mechanical PE in the setting of therapeutic anticoagulation,
prophylaxes are essential in the prevention of an IVC filter should be considered.24 In special
VTE and recommended in the postoperative circumstances, when ongoing thrombolytic ther-
setting. The 2012 ACCP guidelines recommend apy is either contraindicated or failed, or if an
the use of routine pharmacologic prophylaxis embolus is seen within the right atrium or proximal
with low-dose unfractionated heparin (UFH), low pulmonary artery, surgical embolectomy may be
molecular weight heparin (LMWH), or fondapari- appropriate.20
nux.19 The typical regimen of 5000 units of sub-
cutaneous UFH initiated preoperatively, within
2 hours of incision and continued every 8 hours SUMMARY
postoperatively has been shown to reduce the Postoperative respiratory failure after thoracic
rate of DVT from 22.4% to 9.0%, and PE from surgery is multifactorial and is associated with
2.0% to 1.3%. In patients with malignancy who significant morbidity and mortality. Early recogni-
undergo an operation, a 4-week course of phar- tion and treatment of respiratory complications is
macologic prophylaxis, as opposed to 1 week, key to optimizing patient outcomes. Manage-
was shown to reduce the incidence of VTE ment is labor intensive and in the setting of
from 12.0% to 4.8%.21 postpneumonectomy edema and ARDS can be
protracted and resource intensive. The use of
Treatment
ancillary services is not discussed but is of the
When PE is diagnosed, anticoagulation is indi- utmost importance in the progression of patient
cated as long as the risk of bleeding is not prohib- care. Nurses, pain management specialists,
itively high. In the hemodynamically unstable respiratory therapists, and physical and occupa-
patient, without contraindication to systemic anti- tional therapists all play key roles in progressing
coagulation, thrombolysis or, in rare circum- patient care and preventing avoidable respiratory
stances, embolectomy may be considered. The complications.
Management of Postoperative Respiratory Failure 433
REFERENCES 14. Jordan S, Mitchell JA, Quinlan GJ, et al. The patho-
genesis of lung injury following pulmonary resection.
1. Agostini P, Cieslik H, Rathinam S, et al. Postopera- Eur Respir J 2000;15(4):790–9. Available at: http://
tive pulmonary complications following thoracic sur- www.ncbi.nlm.nih.gov/pubmed/10780775.
gery: are there any modifiable risk factors? Thorax
15. Alvarez JM, Bairstow BM, Tang C, et al. Post-lung
2010;65(9):815–8.
resection pulmonary edema: a case for aggressive
2. Canet J, Gallart L. Postoperative respiratory failure:
management. J Cardiothorac Vasc Anesth 1998;
pathogenesis, prediction, and prevention. Curr
12(2):199–205.
Opin Crit Care 2014;20(1):56–62.
3. Gupta H. Development and validation of a risk 16. Ventilation with lower tidal volumes as compared
calculator predicting postoperative respiratory fail- with traditional tidal volumes for acute lung injury
ure. Chest 2011;140(5):1207. and the acute respiratory distress syndrome.
4. Lee JY, Jin S-M, Lee C-H, et al. Risk factors of post- N Engl J Med 2000;342(18):1301–8.
operative pneumonia after lung cancer surgery. 17. Mathisen DJ, Kuo EY, Hahn C, et al. Inhaled nitric ox-
J Korean Med Sci 2011;26(8):979–84. ide for adult respiratory distress syndrome after pul-
5. Kollef M. Prevention of postoperative pneumonia. monary resection. Ann Thorac Surg 1998;66(6):
Hosp Physician 2007;64:47–60. 1894–902.
6. Schussler O, Alifano M, Dermine H, et al. Postoper- 18. Venous thromboembolism: reducing the risk of
ative pneumonia after major lung resection. Am J venous thromboembolism (deep vein thrombosis
Respir Crit Care Med 2006;173(10):1161–9. and pulmonary embolism) in patients admitted to
7. Kazaure HS, Martin M, Yoon JK, et al. Long-term re- hospital. (n.d.). National Clinic Guideline Centre.
sults of a postoperative pneumonia prevention pro-
19. Guyatt GH, Akl EA, Crowther M, et al. Introduction to
gram for the inpatient surgical ward. JAMA Surg
the ninth edition: Antithrombotic Therapy and Pre-
2014;149(9):914.
vention of Thrombosis, 9th ed: American College
8. Niederman M, Craven D. Guidelines for the manage-
of Chest Physicians Evidence-Based Clinical Prac-
ment of adults with hospital acquired, ventilator-asso-
tice Guidelines. Chest 2012;141(2 Suppl):48S–52S.
ciated, and healthcare-associated pneumonia. Am J
Crit Care Med 2005;171:388–416. 20. Rosenberger P, Shernan SK, Mihaljevic T, et al.
9. Martinez F, Medina J, Ojeda D, et al. Postoperative Transesophageal echocardiography for detecting
acute respiratory distress syndrome after lung extrapulmonary thrombi during pulmonary embolec-
resection. Arch Bronconeumol 2007;43(11):623–7. tomy. Ann Thorac Surg 2004;78(3):862–6.
10. Samano M, Sancho L, Beyruti R, et al. Postpneumo- 21. Zurawska U, Parasuraman S, Goldhaber SZ. Pre-
nectomy pulmonary edema. J bras pneumol 2005; vention of pulmonary embolism in general surgery
31(1):69–75. patients. Circulation 2007;115(9):302–8.
11. Shapiro M, Swanson SJ, Wright CD, et al. Predictors
22. Kovacs MJ, Hawel JD, Rekman JF, et al. Ambulatory
of major morbidity and mortality after pneumonec-
management of pulmonary embolism: a pragmatic
tomy utilizing the Society for Thoracic Surgeons
evaluation. J Thromb Haemost 2010;8(11):2406–11.
General Thoracic Surgery Database. Ann Thorac
Surg 2010;90(3):927–34. 23. Nijkeuter M, Söhne M, Tick LW, et al. The natural
12. Alvarez JM, Panda RK, Newman MA, et al. Postpneu- course of hemodynamically stable pulmonary embo-
monectomy pulmonary edema. J Cardiothorac Vasc lism: clinical outcome and risk factors in a large pro-
Anesth 2003;17(3):388–95. spective cohort study. Chest 2007;131(2):517–23.
13. Turnage W, Lunn J. Postpneumonectomy pulmonary 24. Crowther MA. Inferior vena cava filters in the man-
edema: a retrospective analysis of associated vari- agement of venous thromboembolism. Am J Med
ables. Chest 1993;103(6):1646–50. 2007;120(10 Suppl 2):S13–7.
C o m p l i c a t i o n s Fo l l o w i n g
Carinal Resections and
Sleeve Resections
Luis F. Tapias, MD, Harald C. Ott, MD, Douglas J. Mathisen, MD*
KEYWORDS
Carinal resection Sleeve resection Bronchopleural fistula Airway anastomosis
Airway stenosis
KEY POINTS
Careful patient selection is important for the success of carinal and sleeve resections.
Attention to established technical issues is fundamental, including preservation of bronchial blood
supply, proper tension-free anastomotic technique correcting size mismatch, and buttressing with
vascularized tissue.
Prevention of complications by means of adequate surgical technique is critical; once complica-
tions appear, their management is difficult.
Bronchoscopic evaluation before discharge is useful for the early identification of potential anasto-
motic problems.
Aggressive management of complications adhering to established principles should be imple-
mented as soon as the diagnosis is made.
Division of Thoracic Surgery, Department of Surgery, Massachusetts General Hospital, Harvard Medical School,
55 Fruit Street, Blake 15, Boston, MA 02114, USA
* Corresponding author.
E-mail address: dmathisen@mgh.harvard.edu
more than 60 years, recent advances in anesthetic of positive N1 nodes, as it is associated with lower
techniques, as well as improved patient selection morbidity and mortality as well as comparable
and postoperative care have lead to acceptable long-term results when compared with pneumo-
rates of postoperative morbidity and mortality, nectomy.6,7 Some conditions involve only the
making this a valid therapeutic option in patients mainstem bronchus, including low-grade neo-
with involvement of the central airways. plasms (eg, carcinoid tumors, mucoepidermoid
Carinal and sleeve lung resections represent carcinomas, fibrous histiocytomas, and adenoid
true challenges for thoracic surgeons. These cystic carcinomas) and benign stenosis of infec-
patients require careful and thorough evaluation, tious (eg, histoplasmosis), inflammatory, trau-
strict attention to technical details in the operating matic, iatrogenic, or idiopathic etiologies.9 The
room, a keen understanding of the principles of same surgical principles hold for these broncho-
resection and reconstruction of the tracheobron- plastic procedures as for sleeve lobectomy: care-
chial tree, and dedicated postoperative care. ful technique, adjustment for size discrepancy,
These procedures hold great potential for signifi- avoidance of devascularization, avoidance of
cant postoperative morbidity given their impact excessive tension, and the use of pedicled vascu-
on cardiopulmonary physiology and the risk for larized tissue flaps.
development of anastomotic complications.8
Here, we aim to summarize key aspects to be Contraindications
considered when performing carinal or sleeve re-
sections, and when dealing with postoperative Absolute contraindications are related to poor
complications. surgical candidacy due to significant medical co-
morbidities, very low predicted postoperative pul-
monary function, or extensive local invasion of the
INDICATIONS AND CONTRAINDICATIONS tumor precluding reconstruction. The presence of
Indications mediastinal lymph node involvement (N2 disease
Carinal resections and bronchoplastic procedures or greater) is considered a relative contraindication
are most commonly indicated for the surgical given the poor long-term results obtained in these
treatment of centrally located tumors involving patients. The need for resection and reconstruction
the orifice of the lobar bronchus or extending into of vascular structures (ie, superior vena cava or pul-
intermediate or main bronchi or the carina. Most monary artery) is not considered a contraindication.
frequently, these are non–small cell lung cancers
(NSCLC), particularly squamous cell carcinomas. Special Situations
However, they are also useful in the surgical treat- There are a few important special situations to
ment of centrally located low-grade malignancies, carinal surgery and sleeve resections that deserve
such as carcinoid tumors and salivary type tumors mention:
(eg, mucoepidermoid carcinoma and adenoid
cystic carcinomas). Additionally, bronchoplastic Chronic use of corticosteroids: This can lead
procedures can be applied to treat benign steno- to impairment of airway anastomotic healing.
sis, particularly of the mainstem bronchi, of trau- Ideally, patients should be weaned from ste-
matic, infectious, or idiopathic etiologies.9 roids at least 2 to 4 weeks before surgery. If
In the case of carinal resections, a right carinal necessary, coring out the obstructing tumor
pneumonectomy is the most frequent procedure, is beneficial until such time as resection can
especially in patients with NSCLC originating be done.
from the right upper lobe orifice and extending Anticipated need for prolonged postoperative
into the lateral aspect of the carina and lower mechanical ventilation: Positive-pressure
trachea. The tumor length should not exceed ventilation can put stress on the airway anas-
4 cm of trachea, as this poses anatomic limits to tomosis. Additionally, the need for postopera-
resection. Alternatively, carinal resections may be tive mechanical ventilation has been strongly
indicated when there is involvement of the bron- associated with postoperative morbidity.10
chial margin after a standard pneumonectomy. The anesthetic plan should include every
Sleeve lobectomy is indicated when tumors are effort to permit extubation at the end of the
located at the origin of the lobar bronchus, but do procedure.
not infiltrate the remaining lobes to justify a pneu- Neoadjuvant therapy: Patients who have
monectomy. Additionally, it may be indicated received neoadjuvant therapy, particularly ra-
when there is direct infiltration from metastatic diation, are at a higher risk for anastomotic
peribronchial nodes or a positive bronchial margin. complications due to relative tissue ischemia.
Sleeve lobectomy is still indicated in the presence These patients should receive an anastomotic
Carinal Resections and Sleeve Resections 437
wrap with vascularized tissue. Coverage can true in cases of right carinal pneumonectomy, as
be achieved with pedicled intercostal muscle the left bronchus is restrained by the aortic arch,
flaps or omentum in these circumstances.11,12 limiting its cephalad migration. Release maneu-
vers are helpful in reducing anastomotic tension
SURGICAL PRINCIPLES for all bronchoplastic procedures.
Fig. 2. Technique for sleeve lobectomy. (A) Traction sutures placed in the mid-lateral portion proximally and
distally. (B) First anastomotic suture placed posteriorly. (C) Completed anastomosis.
The low rate of bronchial anastomotic complica- and oxidative stress leading to increased pulmo-
tions in our institution is a direct result of routinely nary vascular permeability.45
avoiding tension on the anastomosis by perform-
ing a hilar release maneuver whenever necessary,
precise dissection with preservation of the bron- Prevention
chial blood supply, and placement of vascularized Every effort should be made to limit the adminis-
tissue between the bronchial anastomosis and the tration of fluids and blood products intraopera-
pulmonary artery. tively. The anesthesiologist should be cautioned
to avoid overdistension of the lung to avoid
POSTPNEUMONECTOMY PULMONARY barotrauma injury. The use of a bolus of methyl-
EDEMA prednisolone before ligation of the pulmonary
Presentation artery has been suggested to reduce the incidence
of postpneumonectomy pulmonary edema,48 but
Postpneumonectomy pulmonary edema is an
requires further investigation.
acute syndrome that develops within the first
72 hours after resection.45 This syndrome repre-
sents a continuum spanning acute lung injury Treatment
and ARDS based on current diagnostic criteria. It
is reported to complicate 3% to 7% of all pneumo- Treatment remains supportive to ensure adequate
nectomies, although it has been described in as oxygen delivery by optimizing ventilator settings,
many as 16% of patients. It presents more hemodynamics, and pulmonary toilet, as well as
frequently after right-sided resections.45 It has nutrition and antibiotic coverage. Inhaled nitric
been reported in 4% to 14% of cases involving a oxide has been reported to improve oxygena-
carinal resection.10,18,19,21,46,47 The clinical picture tion,49,50 but these finding have not been repli-
is that of tachypnea, hypoxia, hypercapnia, and cated in animal studies.51 In severe refractory
“ground glass” infiltrates on chest imaging. cases, the use of extracorporeal membrane
Usually there is rapid deterioration mandating oxygenation should be considered.52,53
endotracheal intubation and mechanical ventila-
tory support. Multiple factors have been impli-
Prognosis
cated in its pathophysiology, including fluid
overload, barotrauma, disruption of lymphatic Postpneumonectomy pulmonary edema is a cata-
drainage, the perioperative use of blood products, strophic complication, as it carries a mortality risk
occult microaspiration, systemic inflammation, of 50% to 100%.45
440 Tapias et al
Fig. 3. Technique for airway anastomosis. (A) Placement of traction sutures proximally and distally in the mid-
lateral portion and placement of the first posterior anastomotic suture. (B) Completed anastomotic sutures
(without traction sutures) with proper spacing to control for bronchus size discrepancy.
Table 1
Summary of complications after carinal and sleeve resections in published surgical series with n >50
Anastomotic Complications, %
Bronchopleural
Fistula/
Mortality, Morbidity, Dehiscence/ Bronchovascular Local Empyema, 5-y
Author, Year n % % Ischemia Fistula Stenosis Recurrence % Survival, %
Carinal resection
Tedder et al,8 1992 1915 20.9 — 10.1 2.9 — 4.2 8.6 —
Mitchell et al,10 134 12.7 38.8 17.2b — — — 2.2 —
1999a
Mitchell et al,18 60 15.0 45.0 16.7b — — 3.3 — 42
2001a
Porhanov et al,19 231 16.0 35.5 21.6 — 7.4 5.0 14.7 25
2002
Regnard et al,20 65 7.7 50.8 10.8 — 4.6 — 7.7 27
2005
de Perrot et al,21 119 7.6 47.1 10.1 — 2.5 4.2 5.0 44
2006
441
442
Table 1
(continued )
Tapias et al
Anastomotic Complications, %
Bronchopleural
Fistula/
Mortality, Morbidity, Dehiscence/ Bronchovascular Local Empyema, 5-y
Author, Year n % % Ischemia Fistula Stenosis Recurrence % Survival, %
Terzi et al,30 2002 160 11.3 24.4 8.1 1.9 5.6 — — 39
Hollaus et al,31 2003 108 5.5 26.8 2.8 — — — 2.8 —
Burfeind et al,32 73 2.7 37.0 1.4 0.0 9.6 2.7 — —
2005
Lausberg et al,33 171 1.8 — 0.6 — 0.0 — 0.0 43–46
2005
Ludwig et al,34 2005 116 4.3 — 6.9 — 0.9 — 1.7 39
Takeda et al,35 2006 62 4.8 45.2 3.2 — — 9.7 9.7 54
Yildizeli et al,36 218 4.1 22.9 3.7 0.9 1.8 5.3 1.8 53
2007
Rea et al,37 2008 199 4.5 17.9 1.0 2.0 15.1 — 1.0 40
Yamamoto et al,38 201 1.5 39.8 2.0 — 1.5 2.5 3.0 58
2008
Merritt et al,39 196 2.0 34.6 2.0 0.0 0.0 4.8 2.0 44
2009a
Milman et al,40 64 3.1 45.3 0.0 1.6 1.6 4.7 — 41–48
2009
Gómez-Caro et al,41 58 3.4 34.5 0.0 1.7 0.0 1.7 3.4 62
2011
Storelli et al,42 2012 103 2.9 23.3 0.0 0.0 1.0 7.8 — 63
Gonzalez et al,43 99 3.0 50.5 2.0 — 3.0 — — 28–45
2013
Bylicki et al,44 2014 108 — — 9.3 — 12.0 — — —
a
Experience at Massachusetts General Hospital.
b
Includes all anastomotic complications.
Carinal Resections and Sleeve Resections 443
In their classic review in 1992, Tedder and col- remove granulation tissue, and obtain tissue to
leagues8 reported 16 cases of bronchovascular rule out tumor recurrence as the etiology. Bron-
fistula yielding an incidence of 2.6% after broncho- choscopic resection of granulomas at the suture
plastic procedures. According to this review, the line is usually curative. Patients with benign anas-
result was invariably fatal whenever the outcomes tomotic strictures are best managed initially with
were specified. After this review, we found 11 bronchoscopic balloon dilation.55 If the stricture
surgical series with a sample size of more than recurs, repeated bronchoscopic dilations can be
50 patients that specifically stated the occurrence performed. However, if the patency of the anasto-
of bronchovascular fistulas after sleeve lobec- mosis cannot be practically sustained by dilation,
tomies.25,26,28,30,32,36,37,39–42 Among 1462 patients the use of endobronchial silicone stents should
included in these reports, there were 16 (1.1%) be considered, but this could be difficult after
cases of bronchovascular fistula, resulting in the sleeve lobectomy, given that the remaining main-
death of 14 patients (mortality: 88%). Specific stem bronchus is generally very short and
management was detailed for 3 patients.28,36,40 segmental bronchi are very close to the anasto-
All underwent completion pneumonectomy (2 cari- mosis inferiorly. Indications for intervention are
nal pneumonectomies). One patient developed repeated infections, shortness of breath, or
ARDS after the reintervention leading to death. collapse of the remaining lung. Patients should
The remaining 2 patients survived. Although infre- be evaluated with pulmonary function tests and
quent, these results stress the severity of this ventilation perfusion scans. If symptoms are
complication and the need to avoid it. absent, watchful waiting may be prudent. In cases
of severe anastomotic strictures, segmental
Anastomotic Stricture/Stenosis resection of the stricture with reanastomosis is
Presentation favored,9 but this decision must be balanced
Benign strictures or stenosis, including suture with the structural integrity of the remaining lung
granuloma formation, have an insidious onset parenchyma. These operations are quite difficult
and are the result of ischemia at the anastomosis secondary to scarring between the airway and
leading to scar formation. Most frequently, steno- vascular structures. Therefore, completion
sis is diagnosed during routine postoperative sur- pneumonectomy may be necessary if segmental
veillance bronchoscopy. If bronchoscopy reveals resection is not possible or when the remaining
the bronchial anastomosis appears ischemic but lung is structurally compromised by postobstruc-
intact, the patient should be kept in hospital until tive effects.25 One series reported that the most
the bronchus declares itself. If it starts to break common indication for completion pneumonec-
down, surgical intervention is warranted to avoid tomy after bronchial sleeve resections was the
a fistula or separation. If the bronchus remains occurrence of benign anastomotic stenosis.56
intact, the patient can be observed with frequent
Local Recurrence
bronchoscopies. However, patients may present
with postobstructive pneumonitis, pneumonia, Presentation
dyspnea, hemoptysis, coughing, or wheezing,54 Tumor recurrence at the anastomotic suture line is
which may prompt further interventions. Routine usually discovered during surveillance bronchos-
surveillance imaging studies may also suggest copy. It is to be expected in cases with positive
narrowing of the bronchial lumen or may show tracheobronchial margins. The liberal use of intra-
postobstructive pneumonia. Anastomotic stric- operative frozen-section analysis of the resected
tures have been described to occur in 2.5% to specimen should minimize the risk of residual
7.4% of patients after carinal resections19–21 and malignancy. Some series have reported tumor
in 0% to 15.1% of patients undergoing sleeve recurrence at the anastomosis even when nega-
resections5,8,25–28,30,32–34,36–44 (see Table 1). The tive margins were confirmed on permanent section
occurrence of suture granulomas has reportedly analysis.25 Local tumor recurrence has been re-
decreased with the decreased use of silk or ported in fewer than 5% to 10% of cases of carinal
nonabsorbable sutures when constructing the and sleeve resections (see Table 1), with isolated
airway anastomosis.8,16,17 Also, a continuous reports of very high rates up to 20%.29
suture technique using nonabsorbable material
has been associated with benign circumferential Treatment
stenosis.17 Recurrence at the anastomosis after a sleeve lo-
bectomy in patients without evidence of extensive
Management nodal or metastatic disease may be treated with
Surveillance bronchoscopy is necessary to completion pneumonectomy.8,25,56 The same is
monitor the anastomosis, aspirate secretions, true for patients who are found to have a positive
Carinal Resections and Sleeve Resections 445
bronchial margin on the final histologic evaluation.56 11. Muehrcke DD, Grillo HC, Mathisen DJ. Reconstruc-
The limitation usually lies in the fitness of the tive airway operation after irradiation. Ann Thorac
patients to withstand further resection of lung Surg 1995;59(94):14–8.
parenchyma. Poor surgical candidates for comple- 12. Shrager JB, Wain JC, Wright CD, et al. Omentum is
tion pneumonectomy can be treated with radiation highly effective in the management of complex
therapy. The use of lasers, brachytherapy, photo- cardiothoracic surgical problems. J Thorac Cardio-
dynamic therapy, or expandable metallic stents vasc Surg 2003;125(3):526–32.
may be required as palliation of airway obstruction. 13. Gaissert HA, Grillo HC, Shadmehr MB, et al. Long-
term survival after resection of primary adenoid
cystic and squamous cell carcinoma of the trachea
SUMMARY and carina. Ann Thorac Surg 2004;78:1889–97.
14. Honings J, Gaissert HA, Weinberg AC, et al. Prog-
Careful patient selection and attention to technical
nostic value of pathologic characteristics and resec-
details should minimize postoperative complica-
tion margins in tracheal adenoid cystic carcinoma
tions following bronchoplastic procedures. Compli-
[Internet]. Eur J Cardiothorac Surg 2010;37(6):
cations when they do occur are often devastating,
1438–44.
and require early detection and aggressive treat-
15. Wright CD, Grillo HC, Wain JC, et al. Anastomotic
ment. Unsuccessful management of complications
complications after tracheal resection: prognostic
often leads to death.
factors and management. J Thorac Cardiovasc
Surg 2004;128(5):731–9.
16. Frist WH, Mathisen DJ, Hilgenberg AD, et al. Bron-
REFERENCES
chial sleeve resection with and without pulmonary
1. Lanuti M, Mathisen DJ. Carinal resection. Thorac resection [Internet]. J Thorac Cardiovasc Surg
Surg Clin 2004;14:199–209. 1987;93(3):350–7. Available at: http://www.ncbi.
2. Predina JD, Kunkala M, Aliperti LA, et al. Sleeve nlm.nih.gov/pubmed/3821144. Accessed February
lobectomy: current indications and future directions. 15, 2015.
Ann Thorac Cardiovasc Surg 2010;16(5):310–8. 17. Hsieh CM, Tomita M, Ayabe H, et al. Influence of su-
3. Belsey R. Resection and reconstruction of the intra- ture on bronchial anastomosis in growing puppies
thoracic trachea [Internet]. Br J Surg 1950;38(150): [Internet]. J Thorac Cardiovasc Surg 1988;95(6):
200–5. Available at: http://www.ncbi.nlm.nih.gov/ 998–1002. Available at: http://www.ncbi.nlm.nih.
pubmed/14791963. Accessed February 15, 2015. gov/pubmed/3374164. Accessed February 15,
4. Thomas C. Conservative resection of the bronchial tree 2015.
[Internet]. J R Coll Surg Edinb 1956;1(3):169–86. 18. Mitchell JD, Mathisen DJ, Wright CD, et al. Resec-
Available at: http://www.ncbi.nlm.nih.gov/pubmed/ tion for bronchogenic carcinoma involving the ca-
13307666. Accessed February 15, 2015. rina: long-term results and effect of nodal status on
5. Gaissert HA, Mathisen DJ, Moncure AC, et al. Survival outcome. J Thorac Cardiovasc Surg 2001;121:
and function after sleeve lobectomy for lung cancer. 465–71.
J Thorac Cardiovasc Surg 1996;111:948–53. 19. Porhanov VA, Poliakov IS, Selvaschuk AP, et al. Indi-
6. Ferguson MK, Lehman AG. Sleeve lobectomy or cations and results of sleeve carinal resection. Eur J
pneumonectomy: optimal management strategy us- Cardiothorac Surg 2002;22:685–94.
ing decision analysis techniques. Ann Thorac Surg 20. Regnard JF, Perrotin C, Giovannetti R, et al. Resec-
2003;76(03):1782–8. tion for tumors with carinal involvement: technical
7. Deslauriers J, Grégoire J, Jacques LF, et al. Sleeve aspects, results, and prognostic factors. Ann Thorac
lobectomy versus pneumonectomy for lung cancer: Surg 2005;80:1841–6.
a comparative analysis of survival and sites or recur- 21. De Perrot M, Fadel E, Mercier O, et al. Long-term re-
rences. Ann Thorac Surg 2004;77:1152–6. sults after carinal resection for carcinoma: does the
8. Tedder M, Anstadt MP, Tedder SD, et al. Current benefit warrant the risk? J Thorac Cardiovasc Surg
morbidity, mortality, and survival after bronchoplas- 2006;131(1):81–9.
tic procedures for malignancy [Internet]. Ann Thorac 22. Roviaro G, Vergani C, MacIocco M, et al. Tracheal
Surg 1992;54(2):387–91. sleeve pneumonectomy: long-term outcome. Lung
9. Bueno R, Wain JC, Wright CD, et al. Bronchoplasty Cancer 2006;52:105–10.
in the management of low-grade airway neoplasms 23. Eichhorn F, Storz K, Hoffmann H, et al. Sleeve pneu-
and benign bronchial stenoses [Internet]. Ann monectomy for central non-small cell lung cancer:
Thorac Surg 1996;62(3):824–9. indications, complications, and survival [Internet].
10. Mitchell JD, Mathisen DJ, Wright CD, et al. Clinical Ann Thorac Surg 2013;96(1):253–8.
experience with carinal resection. J Thorac Cardio- 24. Ruffini E, Parola A, Papalia E, et al. Frequency and
vasc Surg 1999;117:39–53. mortality of acute lung injury and acute respiratory
446 Tapias et al
distress syndrome after pulmonary resection for 38. Yamamoto K, Miyamoto Y, Ohsumi A, et al. Sleeve
bronchogenic carcinoma. Eur J Cardiothorac Surg lung resection for lung cancer: analysis according
2001;20:30–6 [discussion: 36–7]. to the type of procedure [Internet]. J Thorac Cardio-
25. Kawahara K, Akamine S, Takahashi T, et al. Manage- vasc Surg 2008;136(5):1349–56.
ment of anastomotic complications after sleeve 39. Merritt RE, Mathisen DJ, Wain JC, et al. Long-term
lobectomy for lung cancer [Internet]. Ann Thorac results of sleeve lobectomy in the management of
Surg 1994;57(6):1529–32 [discussion: 1532–3]. non-small cell lung carcinoma and low-grade neo-
26. Icard P, Regnard JF, Guibert L, et al. Survival and plasms [Internet]. Ann Thorac Surg 2009;88(5):
prognostic factors in patients undergoing paren- 1574–82.
chymal saving bronchoplastic operation for primary 40. Milman S, Kim AW, Warren WH, et al. The incidence
lung cancer: a series of 110 consecutive cases. Eur of perioperative anastomotic complications after
J Cardiothorac Surg 1999;15:426–32. sleeve lobectomy is not increased after neoadjuvant
27. Tronc F, Grégoire J, Rouleau J, et al. Long-term chemoradiotherapy [Internet]. Ann Thorac Surg
results of sleeve lobectomy for lung cancer. Eur J 2009;88(3):945–51.
Cardiothorac Surg 2000;17:550–6. 41. Gómez-Caro A, Garcı́a S, Jiménez MJ, et al. Lung
28. Fadel E, Yildizeli B, Chapelier AR, et al. Sleeve lo- sparing surgery by means of extended broncho-
bectomy for bronchogenic cancers: factors angioplastic (sleeve) lobectomies. Arch Bronconeu-
affecting survival. Ann Thorac Surg 2002;74:851–9. mol 2011;47(2):66–72.
29. Mezzetti M, Panigalli T, Giuliani L, et al. Personal 42. Storelli E, Tutic M, Kestenholz P, et al. Sleeve resec-
experience in lung cancer sleeve lobectomy and tions with unprotected bronchial anastomoses are
sleeve pneumonectomy. Ann Thorac Surg 2002;73: safe even after neoadjuvant therapy. Eur J Cardio-
1736–9. thorac Surg 2012;42:77–81.
30. Terzi A, Lonardoni A, Falezza G, et al. Sleeve lobec- 43. Gonzalez M, Litzistorf Y, Krueger T, et al. Impact of
tomy for non-small cell lung cancer and carcinoids: induction therapy on airway complications after
results in 160 cases. Eur J Cardiothorac Surg 2002; sleeve lobectomy for lung cancer [Internet]. Ann
21:888–93. Thorac Surg 2013;96(1):247–52.
31. Hollaus PH, Wilfing G, Wurnig PN, et al. Risk factors 44. Bylicki O, Vandemoortele T, Orsini B, et al. Incidence
for the development of postoperative complications and management of anastomotic complications
after bronchial sleeve resection for malignancy: a after bronchial resection: a retrospective study
univariate and multivariate analysis. Ann Thorac [Internet]. Ann Thorac Surg 2014;98:1961–7. Avail-
Surg 2003;75:966–72. able at: http://linkinghub.elsevier.com/retrieve/pii/
32. Burfeind WR, D’Amico TA, Toloza EM, et al. Low S0003497514015082.
morbidity and mortality for bronchoplastic proce- 45. Villeneuve PJ, Sundaresan S. Complications of pul-
dures with and without induction therapy. Ann monary resection: postpneumonectomy pulmonary
Thorac Surg 2005;80:418–22. edema and postpneumonectomy syndrome. Thorac
33. Lausberg HF, Graeter TP, Tscholl D, et al. Broncho- Surg Clin 2006;16:223–34.
vascular versus bronchial sleeve resection for cen- 46. Rea F, Marulli G, Schiavon M, et al. Tracheal sleeve
tral lung tumors. Ann Thorac Surg 2005;79:1147–52. pneumonectomy for non small cell lung cancer
34. Ludwig C, Stoelben E, Olschewski M, et al. Compar- (NSCLC): short and long-term results in a single
ison of morbidity, 30-day mortality, and long-term institution. Lung Cancer 2008;61:202–8.
survival after pneumonectomy and sleeve lobec- 47. Roviaro G, Varoli F, Romanelli A, et al. Complications
tomy for non-small cell lung carcinoma. Ann Thorac of tracheal sleeve pneumonectomy: personal expe-
Surg 2005;79:968–73. rience and overview of the literature. J Thorac Cardi-
35. Takeda SI, Maeda H, Koma M, et al. Comparison of ovasc Surg 2001;121:234–40.
surgical results after pneumonectomy and sleeve lo- 48. Cerfolio RJ, Bryant AS, Thurber JS, et al. Intraopera-
bectomy for non-small cell lung cancer. Trends over tive solumedrol helps prevent postpneumonectomy
time and 20-year institutional experience. Eur J Car- pulmonary edema. Ann Thorac Surg 2003;76:
diothorac Surg 2006;29:276–80. 1029–35.
36. Yildizeli B, Fadel E, Mussot S, et al. Morbidity, mor- 49. Mathisen DJ, Kuo EY, Hahn C, et al. Inhaled nitric
tality, and long-term survival after sleeve lobectomy oxide for adult respiratory distress syndrome after
for non-small cell lung cancer. Eur J Cardiothorac pulmonary resection. Ann Thorac Surg 1998;
Surg 2007;31:95–102. 66(98):1894–902.
37. Rea F, Marulli G, Schiavon M, et al. A quarter of a 50. Rabkin DG, Sladen RN, DeMango A, et al. Nitric ox-
century experience with sleeve lobectomy for non- ide for the treatment of postpneumonectomy pul-
small cell lung cancer. Eur J Cardiothorac Surg monary edema. Ann Thorac Surg 2001;72(01):
2008;34:488–92. 272–4.
Carinal Resections and Sleeve Resections 447
51. Filaire M, Fadel E, Decante B, et al. Inhaled nitric Belg 1998;98(6):269–72. Available at: http://www.
oxide does not prevent postpneumonectomy pulmo- ncbi.nlm.nih.gov/pubmed/9922817. Accessed
nary edema in pigs. J Thorac Cardiovasc Surg March 3, 2015.
2007;133:770–4. 54. Farkas EA, Detterbeck FC. Airway complications af-
52. Dünser M, Hasibeder W, Rieger M, et al. ter pulmonary resection. Thorac Surg Clin 2006;16:
Successful therapy of severe pneumonia- 243–51.
associated ARDS after pneumonectomy with 55. Sheski FD, Mathur PN. Long-term results of fiberop-
ECMO and steroids [Internet]. Ann Thorac Surg tic bronchoscopic balloon dilation in the manage-
2004;78(1):335–7. Available at: http://www.ncbi. ment of benign tracheobronchial stenosis. Chest
nlm.nih.gov/pubmed/15223462. Accessed March 1998;114:796–800.
3, 2015. 56. Van Schil PE, Brutel de la Rivière A, Knaepen PJ, et al.
53. Verhelst H, Vranken J, Muysoms F, et al. The use of Completion pneumonectomy after bronchial sleeve
extracorporeal membrane oxygenation in postpneu- resection: incidence, indications, and results. Ann
monectomy pulmonary oedema [Internet]. Acta Chir Thorac Surg 1992;53:1042–5.
A n a s t o m o t i c Le a k a g e
Following
Esophagectomy
Carolyn E. Jones, MD, Thomas J. Watson, MD*
KEYWORDS
Anastomotic leaks Esophagectomy Esophagogastrostomy Esophageal stents
Esophageal cancer
KEY POINTS
Anastomotic leaks following esophagectomy remain a major source of morbidity.
Esophagogastric anastomotic leaks are associated with a spectrum of clinical presentations, lead-
ing to multiple treatment options tailored to the specific needs.
Systemic, local, and technical factors may contribute to the cause of leaks following esophagec-
tomy with esophagogastrostomy.
Esophageal stenting has been successful at managing a significant number of anastomotic leaks
following esophagectomy and has decreased the need for reoperation.
When reoperation is necessary to treat an esophagogastric anastomotic leak, techniques to main-
tain esophagogastric continuity should be considered and usually are successful.
Division of Thoracic and Foregut Surgery, Department of Surgery, University of Rochester Medical Center, 601
Elmwood Avenue, Box Surgery, Rochester, NY 14642, USA
* Corresponding author.
E-mail address: Thomas_watson@urmc.rochester.edu
Fig. 1. Endoscopic views of an esophagogastric anastomosis revealing subtle ischemic changes without a gross
leak. (A) Antegrade view. (B) Retrograde view also demonstrating the gastric fundic tip.
452 Jones & Watson
underlying esophageal cancer or motility disorder, multivariate analysis, heart failure, hypertension,
or the effects of recent chemotherapy or radiation. renal insufficiency, and type of procedure were
The value of preoperative nutritional support is predictive of leak.14 Another retrospective,
debated, with most studies supporting its use only population-based review from the United Kingdom
in cases of severe nutritional depletion.17 Although found low body mass index and neoadjuvant
the definition of severe malnutrition is not exact, a chemotherapy to be significantly associated with
loss of greater than 20% of usual body weight or a esophagogastric anastomotic leak.23 In addition,
serum albumin less than 3.0 g/dL are commonly a study assessing patients undergoing transhiatal
accepted criteria.18 esophagectomy for esophageal cancer found
The duration of nutritional therapy, when indi- weight loss, low forced expiratory volume, and
cated, is also a matter of controversy. In theory, low preoperative albumin to be associated with
preoperative enteral or parenteral nutritional supple- postoperative anastomotic leak.3
mentation should improve healing and lower anasto-
motic leak rates, although this must be balanced
Local Factors
against the urgency of surgical intervention, particu-
larly for malignancy, the likelihood of successful The most important local factor impacting esopha-
repletion, and the potential for complications, such gogastric anastomotic integrity is the adequacy of
as sepsis, especially when the parenteral route is the perfusion to the gastric conduit and esophageal
used. Preoperative nutrition is aided by placement remnant. The blood supply to the esophagus is not
of an esophageal stent in cases of bulky tumors usually a concern, given the segmental nature of its
causing luminal obstruction and dysphagia, arterial inflow and venous drainage, the extensive
although concerns exist regarding the negative submucosal vascular plexus within the esophageal
impact such stents may have on subsequent tumor wall, and the short length of esophagus typically re-
resectability. Anorexia may limit oral intake even maining after esophageal resection. On occasion,
when esophageal luminal patency has been re- particularly if the inferior thyroid arteries have been
established with a stent. Feeding gastrostomy or je- ligated previously (eg, as part of a thyroidectomy)
junostomy tubes may be placed to provide enteral and a long portion of the esophagus has been mobi-
nutrition in cases when an oral diet is inadequate. To- lized, the vasculature of the esophagus may be
tal parenteral nutrition, on occasion, may be neces- impaired, predisposing to anastomotic leak.
sary before esophagectomy. The rich blood supply to the stomach makes it a
Of major importance in the perioperative suitable esophageal substitute. After gastric mobili-
period is the prevention of hypotension and hypo- zation for esophageal replacement, the perfusion of
volemia.18–21 Although the definition of hypoten- the fundus is compromised, because it is a water-
sion and the assessment of hypovolemia are not shed zone of blood flow from the short gastric, left
precise, many surgeons are liberal with the admin- gastric, and right gastroepiploic arteries, the former
istration of fluids during and immediately after sur- two typically being ligated as part of the procedure.
gery to maximize splanchnic perfusion and Because the fundus is the most common site for
minimize vasoconstriction within the gastric anastomosis to the residual esophagus, the
vasculature. A retrospective review of 137 patients ischemia inherent to gastric mobilization is a major
with anastomotic leaks following gastrointestinal risk factor for poor anastomotic healing. Similarly,
surgery (pancreatectomy, esophagectomy, or co- venous obstruction from any of a variety of causes,
lectomy) found low serum albumin (<3.5 g/dL), including disrupted tributaries, scarring, extrinsic
anemia (hemoglobin <8 g/dL), hypotension, the compression, twisting, or distention, may adversely
use of inotropes, and blood transfusion to be inde- impact anastomotic integrity.
pendent predictors of their occurrence.22 The risk Care must be exercised in mobilization of the
of anastomotic leak was four times greater in pa- gastric conduit to preserve the blood supply,
tients who required inotropic support in the periop- derived mainly from the right gastroepiploic ves-
erative period, and three times higher in patients sels with lesser contributions from the right gastric
who developed hypotension. artery and vein, and to minimize trauma to the
The recent Society of Thoracic Surgeons smaller vessels comprising the intramural collat-
Thoracic database analysis of anastomotic leak erals within the gastric wall. Prior gastric surgery
following esophagectomy found obesity, heart with resultant chronic inflammation and disruption
failure, coronary disease, vascular disease, hyper- of the gastric blood supply, both macroscopic and
tension, steroids, diabetes, renal insufficiency, to- microscopic, also may adversely impact esopha-
bacco use, procedure duration greater than gogastric anastomotic healing.
5 hours, and type of procedure to be associated Gastric distention likely is a factor compromising
with leak (P<.05) on univariate analysis. On esophagogastric anastomotic integrity. Many
Anastomotic Leakage Following Esophagectomy 453
esophageal surgeons routinely decompress the thoracic inlet may impair healing of anastomoses
stomach with a nasogastric tube for several days created in the neck. In addition, increased esoph-
after an esophagectomy to prevent aspiration of re- ageal motion during flexion and extension of the
tained gastric contents and to minimize gastric neck also has been theorized to negatively impact
distention that might impair anastomotic healing. cervical anastomoses, especially when the esoph-
In addition, local cervical, mediastinal, or abdom- ageal remnant is short.
inal infection, which can result from intraoperative An extensive body of literature has evolved
soilage of spilled esophageal or gastric contents, regarding leak rates depending on the location of
can prevent healing if in the vicinity of the anasto- the esophagogastric anastomosis (Table 1). The
mosis. Finally, an esophagogastric anastomosis available data support the contention that rates
placed in a preoperative radiation field has been are higher in the neck than in the chest, although
found to be a strong predictor of anastomotic recent reports suggest the gap may be narrow-
leakage in patients with esophageal cancer treated ing.14 The consequences of leakage in the neck,
with neoadjuvant chemoradiation.24,25 however, tend to be less than for leakage into
the chest. Cervical leaks usually can be managed
Technical Factors adequately with a previously placed surgical drain
or with opening and packing of the neck wound,
Tension is detrimental to the healing of any wound, whereas intrathoracic leaks may require place-
particularly an anastomosis. Adequate mobiliza- ment of image-guided percutaneous tubes or
tion and proper construction of the gastric tube reoperation for drainage, closure, containment,
are critical to the success of esophagogastric or diversion. Anastomotic leakage into the chest
anastomotic healing. An appropriately prepared has the potential for unilateral or bilateral pleural
gastric conduit should reach to the cervical level contamination, mediastinitis, and erosion into the
without excessive tension. The narrower the membranous airway, pulmonary parenchyma, or
gastric conduit, the longer its length, although major vascular structures, such as the aorta, with
excessive narrowing can lead to an impaired blood lethal consequences. Although intrathoracic
supply from disruption of submucosal collaterals. leakage traditionally has been associated with sig-
Balancing conduit length and perfusion, some sur- nificant morbidity and mortality, the ability to
geons have recommended a gastric tube trans- rescue the patient from the consequences of
verse diameter of 4 to 5 cm as being optimal. such leaks has improved in recent years.26
Cervical esophageal anastomoses are thought
to be more prone to leakage than those performed
Anastomotic Technique
in the chest, because of the greater length of the
gastric conduit necessary to reach the neck, the Methods of constructing the esophagogastric
resultant anastomotic tension, and the decreased anastomosis vary and include hand-sewn (contin-
blood supply at the fundic tip (used for cervical uous or interrupted, single- or double-layer, and
anastomoses) compared with a more distal loca- absorbable or nonabsorbable sutures), stapled (cir-
tion along the gastric body (used for intrathoracic cular or linear), and hybrid approaches combining
anastomoses). The difference in final position be- sutures and staples. In addition, the anastomosis
tween cervical and intrathoracic anastomoses, can be performed in an end-to-end, end-to-side,
however, may only be a few centimeters. Thus, or side-to-side fashion. Whichever technique is
factors other than conduit length may be impor- used, the anastomosis needs to be constructed
tant. Extrinsic compression by the relatively rigid carefully, incorporating all layers of the esophageal
spine, sternum, and trachea at the level of the and gastric walls while avoiding excessive tissue
Table 1
Population-based analyses of esophagogastric anastomotic leak rates based on anastomotic location
strangulation, and creating a “watertight” closure The concept of division of the left gastric and short
not under excessive tension. A mastery of several gastric vessels at the time of laparoscopic staging
approaches is optimal for the surgeon to be able before esophagectomy holds appeal, in that such
to apply the best one in any specific situation. an approach would not necessarily increase the
Multiple reports have compared anastomotic cost or morbidity to the overall treatment paradigm
leak rates among various techniques. A recent for an esophageal malignancy if such staging
meta-analysis27 found lower leak rates with a already were being contemplated. The timing be-
linear stapled esophagogastric anastomosis tween partial devascularization and subsequent
compared with a completely hand-sewn tech- esophagectomy is a matter of debate, with one an-
nique. A separate meta-analysis by the same imal study showing increased gastric neovascular-
authors found no difference in leak rates between ity 30 days, but not 7 days, after division of the left,
linear stapled and circular stapled esophagogas- right, and short gastric vessels.32 The optimal
tric anastomoses.28 Regardless of the technique, extent of devascularization also remains a subject
the incidence of leaks is higher for anastomoses for further investigation. A separate animal study
performed in the neck compared with those in found relative preservation of gastric blood flow
the chest, as confirmed in two recent meta- to the fundic tip with preoperative ligation of both
analyses.29,30 the short gastric and left gastric vessels, although
The route of transposition of the gastric conduit not with ligation of the short gastric vessels alone.33
also is a factor in the propensity for an anastomotic A recent review article assessed the published liter-
leak to occur. Out of 1030 patients undergoing ature regarding ischemic preconditioning before
transhiatal esophagectomy at the University of esophagectomy and concluded that the available
Michigan in whom the stomach was positioned in evidence does not support its use outside of a clin-
the posterior mediastinum, the anastomotic leak ical research protocol.34
rate was 13%, whereas out of seven patients un- Another surgical adjunct for decreasing leak
dergoing a retrosternal gastric pull-up, the leak rates following esophagectomy is the use of a
rate was 86%.31 Several factors potentially pedicled omental flap for wrapping the esophago-
contribute to the higher leak rate with the retroster- gastric anastomosis. Several reports, including
nal route, including the longer distance the stom- one systematic review of randomized control tri-
ach must traverse, and the relative lack of als, have assessed outcomes with this technique
surrounding soft tissue divestments. With a retro- and have shown lower leak rates compared with
sternal conduit, the anastomosis sits in a subcu- unbuttressed anastomoses.5,35–37
taneous position and is unsupported during a Biocompatible sealants have been used during
cough or Valsalva maneuver. When the conduit is various types of operations, such as lung resections
within the posterior mediastinum, the anastomosis or vascular procedures. No approved sealants are
is buttressed posteriorly by the prevertebral fascia, currently on the market for prevention or treatment
laterally by the carotid sheath, and anteriorly by of esophageal perforations or anastomotic leaks,
the membranous trachea. These surrounding and there are no studies assessing their utility in
structures not only aid healing, but also support adults. Two studies of fibrin glue for the prevention
the anastomosis during internal pressurization. of esophagogastric anastomotic leaks in children
suggested a benefit to its use.38,39 The efficacy of
ADJUNCTS FOR PREVENTION various sealants following esophagectomy is a
topic in need of further investigation.
Despite an awareness of the factors predisposing Other adjuncts for prevention of esophagogastric
to esophagogastric anastomotic leakage following anastomotic leaks include pleural tenting after Ivor
esophagectomy, and efforts to optimize modifi- Lewis esophagectomy,40 intraoperative assess-
able risks, the incidence of leaks remains signifi- ment of gastric graft perfusion using laser-assisted
cant. As a result, various adjunctive measures fluorescent-dye angiography,41 and planned delay
have been introduced to help prevent or control of oral intake following surgery.42 Perhaps a combi-
anastomotic leaks and their sequelae. nation of preventative measures will lead to a mean-
Ischemia of the mobilized and partially devascu- ingful decrease in leak rates over time.
larized gastric fundus is an important cause of
esophagogastric leaks. Ischemic preconditioning MANAGEMENT
is based on the hypothesis that the vascularity of
the gastric fundus can be improved, and anasto- The management of an esophagogastric anasto-
motic leaks reduced, by partial devascularization motic leak requires considerable clinical judgment
in advance of esophagectomy to improve the and depends on the location of the anastomosis;
gastric microcirculation (the “delay phenomenon”). the extent of the anastomotic disruption; the
Anastomotic Leakage Following Esophagectomy 455
adequacy of conduit perfusion; the involvement of Moderate to large disruptions are more chal-
adjacent structures, such as the airway or lung; the lenging. Surgical repair in such cases is fraught
degree of sepsis; and the hemodynamic stability of with difficulties because the factors resulting in
the patient. The principles governing treatment of the initial leak, such as compromised blood flow
anastomotic leakage are the same as those for and anastomotic tension, are usually still present
esophageal perforation in general (Box 4). at the time of a reoperation. Additional factors,
such as contamination of the surgical field, hemo-
Nonoperative Approaches dynamic or respiratory compromise, and the ef-
Occult anastomotic leaks (grade I) occurring in pa- fects of systemic sepsis, further complicate
tients not exhibiting clinical signs or symptoms, reintervention. Repair of the anastomosis, when
and detected as part of routine postoperative undertaken, should include buttressing with viable
endoscopy or imaging studies, can be managed tissue, such as omentum, pericardial fat, or mus-
with a delay in institution of oral feeding and cle flap around the suture line.
enteral nutritional support via a jejunostomy or If repair is not feasible at the time of re-
nasogastric/nasoduodenal feeding tube. Signs of exploration, management options include wide
infection should prompt the early administration local drainage, placement of a large, exteriorized
of broad-spectrum antibiotics. Spillage of saliva T-tube across the defect, or takedown of the anas-
or gastric secretions are controlled with previously tomosis with cervical/cervicothoracic esophagos-
placed surgical drains, if present, or the introduc- tomy (with or without resection of the remaining
tion of percutaneous drains into the involved areas intrathoracic stomach). The latter commits the pa-
of the pleural space and mediastinum. tient to a subsequent foregut reconstruction, a
sizable undertaking with associated morbidity
Surgical Options and the potential for mortality, assuming sufficient
recovery to allow reoperation.
Leaks associated with cervical anastomoses
A single institution, retrospective review of 761
generally are treated by wound opening and pack-
patients undergoing esophagectomy over a 10-
ing. The risk of extensive mediastinitis is less
year time period (1993–2003) found 48 patients
compared with intrathoracic leaks, because the
(6.3%) with anastomotic leak.43 Of the 47 patients
contamination is generally confined to the cervical
available for analysis, 27 (57%) were managed
soft tissues. Leaks in the chest, although less
nonoperatively and 20 (43%) required operative
common than those in the neck, can result in se-
intervention. Primary anastomotic repair was
vere sepsis and mediastinitis requiring additional
possible in 14 patients, reinforcement of the anas-
closed drainage with chest tubes or surgical
tomosis with viable tissue was possible in six pa-
washout.
tients, and esophageal diversion was necessary
The percentage of circumferential disruption of
in two patients. Median hospitalization was
the anastomosis and the extent of conduit necro-
20 days in the nonoperative group and 31 days
sis also determine management options. Small
in the operative group (P 5 .0037). The authors
disruptions are managed with delayed oral intake,
concluded that contained leaks could be managed
enteral or parenteral nutritional support, drainage
nonoperatively and that esophagogastric continu-
of adjacent fluid collections, antibiotics, and time.
ity could be preserved in most cases.
Box 4
Principles of management of esophageal Stenting
perforations
Esophageal stenting has been introduced in
Systemic antibiotics recent years as an option for treatment of clinically
Close or occlude the defect as soon as possible significant esophageal perforations and esopha-
gogastric anastomotic leaks.44–50 Stents can be
Drain associated fluid collections
placed expeditiously, even in an intensive care
Prevent distal obstruction setting, avoiding many of the issues associated
Ensure lack of factors keeping perforation with surgical reintervention. The goal of stenting
open (eg, tumor, foreign body, persistent is closure of the anastomotic defect, realizing
infection) that associated fluid collections may require addi-
Esophageal diversion or resection if sepsis tional image-guided percutaneous or operative
poorly controlled with more conservative drainage. If endoscopy reveals significant gastric
measures ischemia, stenting may not be appropriate
because conduit excision may be required.
456 Jones & Watson
Anastomotic leaks involving less than 30% of stent is left in place for a prolonged period of
the circumference have been shown to be appro- time.
priate for stent placement. Properly positioned, A study of 17 patients undergoing stent place-
stents can seal the area of disruption and allow ment for acute intrathoracic anastomotic leaks
for healing. A fully covered self-expanding metal, following esophagectomy found successful leak
plastic, or hybrid stent of a large diameter is the occlusion in all.48 Fourteen patients (82%) were
ideal choice, because subsequent removal is able to resume an oral diet within 72 hours of stent
necessary (Fig. 2). Uncovered or partially covered placement. Stent migration occurred in three pa-
stents are not appropriate, because they may be tients (18%), requiring repositioning in two and
difficult to remove at a later date. Larger- replacement in one. All stents were subsequently
diameter stents typically are used to seal perfora- removed at a mean of 17 9 days.
tions compared with stents placed for palliation of A separate study assessed 18 (11.3%) anasto-
obstructing esophageal cancers. motic leaks out of 160 minimally invasive esopha-
A common problem with stents placed across gectomies performed at a single institution.51
esophagogastric anastomoses is migration. Leaks were managed with a variety of techniques
Stent placement into a gastric conduit with a including neck drainage (N 5 4), esophageal diver-
relatively large diameter prevents adherence of sion (N 5 2), thoracoscopic drainage with or
the stent to the gastric wall; the stent is seated without T-tube placement (N 5 3), or endoscopic
mainly via its interface with the esophageal mu- stenting with or without percutaneous drainage
cosa. Because fully covered stents are necessary (N 5 9). Leaks were successfully controlled in
for anastomotic leaks, adherence to the mucosa 89% of the patients treated surgically, and 100%
is further compromised compared with stents of the patients treated with stenting. No 60-day
with an uncovered proximal portion, which per- or in-hospital mortality was noted in either group.
mits tissue ingrowth. Adjuncts to stent place- The authors noted a shift over time to endoscopic
ment, such as endoscopic clipping or suturing, management of leaks at their institution.
have been advocated to assist in the prevention
of stent migration in this scenario. Other compli-
Aerodigestive Fistula Management
cations associated with stents include inade-
quate coverage of the leak; plugging; and The close proximity of the esophagogastric
erosion into surrounding structures, such as the anastomosis and the gastric conduit to the lung
airway or major blood vessels, particularly if the parenchyma and membranous airway when the
stomach is positioned within the posterior medias- or jejunal interposition with or without “super-
tinum predisposes to aerodigestive fistula forma- charging,” depending the vascularity of the
tion. Such fistulization often requires emergent conduit and the length needed to reach the
intervention because of the associated aspiration remnant esophagus.52,53
of pharyngeal and gastric secretions into the
lungs, leading to pneumonia/pneumonitis and res- SUMMARY
piratory compromise. In these cases, a temporary
endoesophageal stent may be placed across the Anastomotic leaks remain a significant clinical
anastomosis to control contamination (Fig. 3). challenge following esophagectomy with foregut
Stenting, however, is a poor long-term solution, reconstruction. Despite an increasing under-
because the fistula requires eventual operative standing of the multiple contributing factors,
intervention. advancements in perioperative optimization of
The principles behind surgical management of modifiable risks, and improvements in surgical,
anastomotic fistulae to the airway include primary endoscopic, and percutaneous management
repair of the anastomosis, if possible, and closure techniques, leaks remain a source of major
of the defect in the airway. The latter may be aided morbidity associated with esophageal resection.
by the use of pericardium or aortic homograft as a Fortunately, the mortality resulting from anasto-
buttress or patch (Dr Doug Mathisen, Massachu- motic leaks seems to be improving. As with
setts General Hospital, Boston, MA, personal most disease processes, prevention is best,
communication, 2015). Vascularized soft tissue, although even a perfectly executed esophageal
such as omentum or a muscle flap, should be resection in an otherwise healthy individual can
interposed between the esophagogastric and result in anastomotic leakage. The surgeon
airway suture lines to prevent refistulization. If the should be well versed in the principles underlying
patient requires a thoracotomy to complete the the cause of leaks and strategies to minimize their
repair and single-ventilation is not tolerated, the occurrence. Similarly, when confronted with a
procedure can be performed under venovenous leak, the surgeon must exercise sound judgment
extracorporeal membrane oxygenation support. in deciding on a treatment paradigm that is best
The degree of anastomotic disruption and asso- for the circumstances of the individual. Appropri-
ciated gastric ischemia is often extensive enough ately diagnosed and managed, most anastomotic
to require conduit excision and the creation of a leaks following esophagectomy can be brought to
cervical or cervicothoracic esophagostomy. If the a successful resolution.
patient survives, delayed reconstruction can be
considered after a full recovery, usually months REFERENCES
later. At the time of subsequent reconstruction, a
conduit other than the stomach is necessary. 1. Jancin B. Esophagectomy cases have been steadily
Options include of a substernally transposed colon rising. Thorac Surg News 2013;9(1):6.
458 Jones & Watson
2. Conners RC, Reuben BC, Neumayer LA, et al. 17. Dutta S, Fullarton GM, Forshaw MJ, et al. Persistent
Comparing outcomes after transthoracic and tran- elevation of C-reactive protein following esophago-
shiatal esophagectomy: a 5-year prospective cohort gastric cancer resection as a predictor of postoper-
of 17,395 patients. J Am Coll Surg 2007;205:735–40. ative surgical site infectious complications. World J
3. Tabatabai A, Hashemi M, Mohajeri G, et al. Inci- Surg 2011;35:1017–25.
dence and risk factors predisposing anastomotic 18. Dewar L, Gelfand G, Finley RJ, et al. Factors
leak after transhiatal esophagectomy. Ann Thorac affecting cervical anastomotic leak and stricture for-
Med 2009;4:197–200. mation following esophagogastrectomy and gastric
4. Ben-David K, Sarosi GA, Cendan JC, et al. tube interposition. Am J Surg 1992;163:484–9.
Decreasing morbidity and mortality in 100 consecu- 19. Perracchia A, Bardini R, Ruol A, et al. Esophago-
tive minimally invasive esophagectomies. Surg visceral anastomotic leak: a prospective statistical
Endosc 2012;26:162–7. study of predisposing factors. J Thorac Cardiovasc
5. Dai JG, Zhang ZY, Min JX, et al. Wrapping of the Surg 1988;95:685–91.
omental pedicle flap around esophagogastric anas- 20. Buzby GP, Blouin G, Colling CL, et al. Perioperative
tomosis after esophagectomy for esophageal total parenteral nutrition in surgical patients. N Engl
cancer. Surgery 2011;149:404–10. J Med 1991;325:525–32.
6. Muller JM, Erasmi H, Stelzner M, et al. Surgical ther- 21. Baker JP, Detsky AS, Wesson DE, et al. Nutritional
apy of oesophageal carcinoma. Br J Surg 1990;77: assessment: a comparison of clinical judgment
845–57. and objective measurements. N Engl J Med 1982;
7. Vigneswaran WT, Trastek VF, Pairolero PC, et al. 306:969–72.
Transhiatal esophagectomy for carcinoma of the 22. Choudhuri AH, Uppal R, Kumar M. Influence of non-
esophagus. Ann Thorac Surg 1993;56:838–46. surgical risk factors on anastomotic leakage after
8. Chasseray VM, Kiroff GK, Buard JL, et al. Cervical or major gastrointestinal surgery: audit from a tertiary
thoracic anastomosis for esophagectomy for carci- care teaching institute. Int J Crit Illn Inj Sci 2013;3:
noma. Surg Gynecol Obstet 1989;169:55–62. 246–9.
9. Tilanus HW, Hop WC, Langenhorst BL, et al. Esoph- 23. Escofet X, Manjunath A, Twine C, et al. Prevalence
agectomy with or without thoracotomy: is there any and outcome of esophagogastric anastomotic leak
difference? J Thorac Cardiovasc Surg 1993;105: after esophagectomy in a UK regional cancer
898–903. network. Dis Esophagus 2010;23:112–6.
10. Patil PK, Patel SG, Mistry RC, et al. Cancer of the 24. Juloori A, Tucker SL, Komaki R, et al. Influence of
esophagus: esophagogastric anastomotic leak: a preoperative radiation field on postoperative leak
retrospective study of predisposing factors. J Surg rates in esophageal cancer patients after trimodality
Oncol 1992;49:163–7. therapy. J Thorac Oncol 2014;9:534–40.
11. Shahian DM, Neptune WB, Ellis FH Jr, et al. Trans- 25. Hulscher JB, van Sandick JW, de Boer AG, et al.
thoracic versus extrathoracic esophagectomy: mor- Extended transthoracic resection compared to
tality, morbidity and long-term survival. Ann Thorac limited transhiatal resection for adenocarcinoma of
Surg 1986;41:237–46. the esophagus. N Engl J Med 2002;347:1662–9.
12. King RM, Pairolero PC, Trastek VF, et al. Ivor Lewis 26. Martin LW, Swisher SG, Hofstetter W, et al. Intratho-
esophagogastrectomy for carcinoma of the esoph- racic leaks following esophagectomy are no longer
agus: early and late functional results. Ann Thorac associated with increased mortality. Ann Surg
Surg 1987;44:119–22. 2005;242:392–402.
13. Urschel JD. Esophagogastrostomy anastomotic 27. Deng XF, Liu QX, Zhou D, et al. Hand-sewn vs line-
leaks complicating esophagectomy: a review. Am arly stapled esophagogastric anastomosis for
J Surg 1995;169:634–40. esophageal cancer: a meta-analysis. World J
14. Kassis ES, Kosinski AS, Ross P Jr, et al. Predictors of Gastroenterol 2015;21:4757–64.
anastomotic leak after esophagectomy: an analysis 28. Zhou D, Liu QX, Deng XF, et al. Comparison of two
of the Society of Thoracic Surgeons General Thoracic different mechanical esophagogastric anastomosis
Database. Ann Thorac Surg 2013;96:1919–26. in esophageal cancer patients: a meta-analysis.
15. Swanson JO, Levine MS, Redfern RO, et al. Useful- J Cardiothorac Surg 2015;10:67.
ness of high-density barium for detection of leaks af- 29. Hulscher JB, Tijssen JG, Obertop H, et al. Transtho-
ter esophagogastrectomy, total gastrectomy, and racic versus transhiatal resection for carcinoma of
total laryngectomy. AJR Am J Roentgenol 2003; the esophagus: a meta-analysis. Ann Thorac Surg
181:415–20. 2001;72:306–13.
16. Solomon DG, Sasaki CT, Salem RR. An evaluation of 30. Markar SR, Arya S, Karthikesalingam A, et al. Tech-
the routine use of contrast radiography as a screening nical factors that affect anastomotic integrity
test for cervical anastomotic integrity after esophagec- following esophagectomy: systematic review and
tomy. Am J Surg 2012;203:467–71. meta-analysis. Ann Surg Oncol 2013;20:4274–81.
Anastomotic Leakage Following Esophagectomy 459
31. Orringer MB, Marshall B, Iannettoni MD, et al. 42. Bolton JS, Conway WC, Abbas AE. Planned delay of
Transhiatal esophagectomy for treatment of benign oral intake after esophagectomy reduces the cervi-
and malignant disease. World J Surg 2001;25: cal anastomotic leak rate and hospital length of
196–203. stay. J Gastrointest Surg 2014;18:304–9.
32. Perry KA, Banarjee A, Liu J, et al. Gastric ischemic 43. Crestanello JA, Deschamps C, Cassivi SD, et al.
conditioning increases neovascularization and re- Selective management of intrathoracic anastomotic
duces inflammation and fibrosis during gastro- leak after esophagectomy. J Thorac Cardiovasc
esophageal anastomotic healing. Surg Endosc Surg 2005;129:254–60.
2013;27:753–60.
44. Freeman RK, Van Workom JM, Ascioti AJ. Esopha-
33. Beck SM, Malay MB, Gagne DJ, et al. Experimental geal stent placement for the treatment of iatrogenic
model of laparoscopic gastric ischemic precondi- intrathoracic esophageal perforation. Ann Thorac
tioning prior to transhiatal esophagectomy. Surg
Surg 2007;83:2003–8.
Endosc 2011;25:2470–7.
45. Hunerbein M, Stroszczynski C, Moesta KT, et al.
34. Yuan Y, Duranceau A, Ferraro P, et al. Vascular con-
Treatment of thoracic anastomotic leaks after esoph-
ditioning of the stomach before esophageal recon-
agectomy with self-expanding plastic stents. Ann
struction by gastric interposition. Dis Esophagus
Surg 2004;240:801–7.
2012;25:740–9.
35. Yuan Y, Zeng X, Hu Y, et al. Omentoplasty for oesopha- 46. Chak A, Singh R, Linden PA. Covered stents for the
gogastrostomy after oesophagectomy. Cochrane treatment of life-threatening cervical esophageal
Database Syst Rev 2014;(10):CD008446. anastomotic leaks. J Thorac Cardiovasc Surg
36. Sepesi B, Swisher SG, Walsh GL, et al. Omental 2011;141:834–44.
reinforcement of the thoracic esophagogastric anas- 47. D’Cunha J, Rueth NM, Groth SS, et al. Esophageal
tomosis: an analysis of leak and reintervention rates stents for anastomotic leaks and perforations.
in patients undergoing planned and salvage esoph- J Thorac Cardiovasc Surg 2011;142:39–46.
agectomy. J Thorac Cardiovasc Surg 2012;144: 48. Freeman RK, Vyverberg A, Ascioti AJ. Esophageal
1146–50. stent placement for the treatment of acute intratho-
37. Yoshida N, Baba Y, Watanabe M, et al. Triangulating racic anastomotic leak after esophagectomy. Ann
stapling technique covered with the pedicled Thorac Surg 2011;92:204–8.
omental flap for esophagogastric anastomosis: a 49. Wilson JL, Louie BE, Farivar AS, et al. Fully covered
safe anastomosis with fewer complications. J Am self-expanding metal stents are effective for benign
Coll Surg 2015;220:e13–6. esophagogastric disruptions and strictures.
38. Saldana-Cortes JA, Larios-Arceo F, Prieto-Diaz- J Gastrointest Surg 2013;17:2045–50.
Chavez E, et al. Role of fibrin glue in the prevention
50. Eizaguirre E, Larburu S, Asensio JI, et al. Treatment of
of cervical leakage and strictures after esophageal
anastomotic leaks with metallic stent after esophagec-
reconstruction of caustic injury. World J Surg 2009;
tomies. Dis Esophagus 2015. [Epub ahead of print].
33:986–93.
39. Upadhyaya VD, Gopal SC, Gangopadhyaya AN, 51. Nguyen NT, Rudersdorf PD, Smith BR, et al. Man-
et al. Role of fibrin glue as a sealant to esophageal agement of gastrointestinal leaks after minimally
anastomosis in cases of congenital esophageal invasive esophagectomy: conventional treatments
vs. endoscopic stenting. J Gastrointest Surg 2011;
atresia with tracheo-esophageal fistula. World J
Surg 2007;31:2412–5. 11:1952–60.
40. Asteriou C, Barbetakis N, Lalountas M, et al. Modi- 52. Kesler KA, Pillai ST, Birdas TJ, et al. “Supercharged”
fied pleural tenting for prevention of anastomotic isoperistaltic colon interposition for long-segment
leak after Ivor Lewis esophagogastrectomy. Ann esophageal reconstruction. Ann Thorac Surg 2013;
Surg Oncol 2011;18:3737–42. 95:1162–8.
41. Zehetner J, DeMeester SR, Alicuben ET, et al. Intra- 53. Blackmon SH, Correa AM, Skoracki R, et al. Super-
operative assessment of perfusion of the gastric charged pedicled jejunal interposition for esopha-
graft and correlation with anastomotic leaks after geal replacement: a 10 year experience. Ann
esophagectomy. Ann Surg 2014;262(1):74–8. Thorac Surg 2012;94:1104–11.
Ma na gement of C on d u it
Necrosis Followin g
Esophagectomy
Karen J. Dickinson, MBBS, BSc, MD, FRCS,
Shanda H. Blackmon, MD, MPH*
KEYWORDS
Esophagectomy Conduit Complication Necrosis Reconstruction
KEY POINTS
Prevention of conduit ischemia or necrosis is better than conduit loss and esophageal diversion.
Intraoperative assessment of conduit ischemia by the surgeon clinically is poor; specialized tech-
niques, for example, Doppler fluorescence, may be useful.
Intraoperative suspicion of conduit ischemia should always be acted on, and an alternative conduit
may be necessary.
Primary esophageal defunctioning may be necessary for intraoperative graft necrosis (with venting
gastrostomy and feeding jejunostomy).
Vigilance is required to detect postoperative gastric conduit necrosis as clinical signs may be
nonspecific.
esophagus. An ischemic conduit may progress to of conduit necrosis and morbidity/mortality from
necrosis and may present early in the postopera- surgery will be prohibitive, and patients and fam-
tive course with an anastomotic leak or weeks to ilies should be counseled accordingly. Patients
months after surgery with a stricture. with significant comorbidity have been shown to
have an odds ratio (OR) of 2.2 (1.1–4.3, P 5 .023)
Consequences for the development of conduit ischemia.15 Addi-
The consequences of conduit necrosis are grave, tionally, conduit ischemia was associated with an
both for the patients in terms of potential mortality OR of 5.5 (2.5–12.10) for anastomotic leak and
but also in terms of quality of life. Mortality after 4.4 (2.0–9.6) for stricture development.15 Patient
conduit necrosis can exceed 90%.3–6 Focal necro- factors, such as smoking, neoadjuvant therapy,
sis of the conduit is most likely to occur in the and preoperative weight, were not, however,
region of the anastomosis, as this has the most associated with increased ischemia. Intuitively,
tenuous blood supply. It is essential that any when contemplating esophageal reconstruction,
gastrointestinal anastomosis should be created the authors are concerned about patients with
without tension and with a good blood supply. diabetes, hypertension, and peripheral arterial
Focal ischemia of the anastomosis is, therefore, disease. In comorbid patients a strategy to deal
likely to cause an anastomotic leak. Anastomotic with this is to perform delayed conduit formation.
leaks occur in around 10% of esophagectomies Of the 37 patients in whom this was used, 6 had
nationally7 and can be graded from I to IV (grade diabetes, 18 hypertension, 16 coronary artery dis-
I, radiological leak only; grade II, minor contained ease, 10 were obese (body mass index >30), and 2
leak; grade III, major leak with evidence of sepsis; had undergone previous pneumonectomy.16 Pre-
grade IV, conduit necrosis).8,9 Leaks without sig- operative risk assessment of any patient undergo-
nificant clinical sequelae can be managed conser- ing esophageal resection is essential and should
vatively or endoscopically, for example, with a include a full history and examination including
covered stent and endoscopically sutured in place assessment of cardiorespiratory function (ie,
to prevent migration.10,11 Necrosis of the entire supervised exercise in clinic, cardiopulmonary ex-
conduit or a large section will require reoperation ercise testing, pulmonary function testing, echo-
for resection and an end esophagostomy for diver- cardiography, coronary angiography where
sion, in combination with a venting gastrostomy indicated). For hypertensive patients control
(where applicable), drainage of collections, and should be assessed and end organ damage iden-
feeding jejunostomy. The patients’ clinical status tified, for diabetic patients assessment of diabetic
must then be optimized before reconstruction control is critical (hemoglobin A1c) and, in the
can be considered.12–14 The consequences for pa- presence of previous abdominal surgery or where
tients are significant. Although an end esophagos- colonic interposition is considered mesenteric
tomy on the chest rather than the neck can be angiography should be performed.
hidden under clothes and patients will still be
able to eat, there is no denying the physical and Intraoperative: Ischemic Preconditioning,
psychological effects of being dependent on jeju- Assessment of Conduit, Blood Pressure
nostomy feeding and the anticipation of further Support
major surgery for reconstruction.
A late consequence of conduit ischemia is the Ischemic preconditioning
development of a stricture, either in the conduit it- Ischemic preconditioning of the gastric conduit
self or at the anastomosis. Conduit ischemia, before esophageal reconstruction has been pro-
anastomotic leak, and stricture are intertwined posed in an attempt to reduce conduit necrosis
clinical entities.15 Strictures will present with dys- and anastomotic leak rates. Urschel and
phagic symptoms and will affect patients’ quality colleagues17,18 first described this in 1997. Subse-
of life. Multiple dilatations will often be required quently Schröder and colleagues19,20 demon-
in these patients. strated that the gastric conduit microcirculation
takes 4 days to return to preoperative levels after
esophagectomy. The mechanism of effect is un-
PREVENTION
clear, but neovascularization of the stomach and
Preoperative: Individual Risk Assessment
release of humoral factors to improve the blood
Of key importance to the management of conduit supply to the fundal area have both been pro-
necrosis is the identification of risk factors and posed. The basis for this concern is related to
high-risk patients. Preoperative and intraoperative studies in cadaveric specimens after esophagec-
strategies can be used to optimize patients, their tomy in which it has been shown that 20% of the
anatomy, and physiology. In some cases the risks blood flow to the top of the gastric conduit is
Conduit Necrosis Following Esophagectomy 463
derived from the mucosal capillary network rather conventional angiography, and laser-induced fluo-
than a named vessel.21 rescence of indocyanine green28 (Tables 1 and 2).
Preconditioning may take the form of radiological Arguably the most commonly used intraopera-
embolization of the arterial supply of the stomach or tive assessment of conduit perfusion, when
involve preresection laparoscopic ligation of the left used, are Doppler ultrasound and intravenous
gastric artery and/or short gastric arteries. Initial (IV) fluorescent dye; they have reported sensitivity
experience with preoperative arterial embolization of up to 60%.29 These techniques are easily used
was not successful and complicated by high rates in the operating room, with a handheld Doppler or
of pancreatitis and splenic infarction.22 However, with the SPY system (Fluorescence Imaging Sys-
this was likely because the splenic and short gastric tem [Novadaq, USA]) in which a fluorescent agent
arteries were embolized in addition to the left (indocyanine green) enables visualization of the
gastric artery. When preresectional laparoscopy perfusion of the proposed conduit.30 Murawa
is performed, a celiac lymph node dissection and and colleagues31 were the first to publish use of
inspection for omental, peritoneal, and liver metas- the SPY system in the assessment of the gastric
tases are also possible. To elucidate the mecha- conduit; several other studies have suggested
nism by which preconditioning may reduce utility of this technique.28,32,33 These studies are
anastomotic complications, one group monitored all small, however, and not powered to detect a
vascular endothelial growth factor (VEGF) levels in clinically significant difference in anastomotic
patients undergoing laparoscopic ligation of the leak rates. The SPY system may be particularly
left gastric artery compared with those without.23 useful when a complex esophageal reconstruction
No significant difference was observed between is considered to assess the adequacy of super-
the two groups, which may be because of an early charging the conduit and to assess jejunal or
increase in VEGF levels or because neoangiogene- colonic blood supply (Fig. 1).
sis is not the mechanism responsible. A recent sys- Although not all intraoperative assessments give
tematic review of factors affecting anastomotic real-time information with regard to the adequacy
integrity following esophagectomy demonstrated of conduit blood flow (eg, SPECT), information
that in 12 studies with 1215 patients, no difference gained from these studies can be used to guide
in anastomotic leak rates was observed in those intraoperative decision making. The authors
being preconditioned (8.83%) compared with recommend using these tools in conjunction with
those not (14.11%) (pooled OR 5 0.73, 95% confi- clinical acumen to decide whether the conduit
dence interval 0.5–1.06, P 5 .1).24 There was no ev- blood supply is suitable or whether another
idence of statistical heterogeneity or bias. When conduit must be sought.
sensitivity analysis was performed for studies
comparing preoperative vessel embolization, no Perfusion pressure
difference between leak rates was seen in those Patients undergoing esophagectomy are predis-
patients with and without intervention. The same posed to hemodynamic changes, secondary to
was true for studies comparing laparoscopic vessel fluid shifts, intraoperative fluid loss, epidural anes-
ligation with no ligation. The present clinical evi- thesia, IV fluid administration, and the potential use
dence does not favor preoperative preconditioning of vasopressor agents. It is common for patients
of the gastric conduit. In fact, this may be associ- undergoing esophagectomy to have a thoracic
ated with increased cost given the additional sur- epidural for analgesia. It can be postulated that
gery and hospital admission. epidural anesthesia causes increased blood flow
in the gastric conduit by way of sympathetic block;
Intraoperative assessment of conduit in fact, a small human study has demonstrated
vascularity increased gastric conduit blood flow 1 hour and
The intraoperative clinical impression of con- 18 hours following thoracic epidural anesthesia.34
duit perfusion is not particularly accurate.25–27 Despite this, administration of an intraoperative
Techniques exist to assess blood supply to the thoracic epidural bupivacaine bolus has been
conduit and can be applied to gastric, jejunal, shown to reduce blood flow at the tip of the gastric
and colonic reconstructions. These techniques conduit during esophagectomy.35 Administration
include fluorescence angiography (eg, Woods of IV adrenaline reversed this reduced blood flow.
lamp), handheld Doppler, laser Doppler flowmetry Many surgeons have reservations with regard to
and spectrophotometry, transmucosal oxygen the intraoperative administration of vasopressor
saturation measurement, hydrogen clearance, agents and are concerned about reducing perfu-
visible light spectroscopy, single-photon emis- sion to the conduit. In animal models, the adminis-
sion computed tomography (SPECT), esophago- tration of vasopressors, noradrenaline in swine,36
gastroduodenoscopy (EGD), CT angiography, has been associated with reduced perfusion to
464 Dickinson & Blackmon
Table 1
Comparison of technologies used to assess perfusion of the gastric conduit in esophageal
reconstructive surgery
the conduit. However, in a small human study, the associated with increased incidence of adult respi-
administration of IV phenylephrine following a bu- ratory distress syndrome.38 This may be a surro-
pivacaine bolus via thoracic epidural increased gate marker of patient comorbidity and
the flux (as measured in perfusion units) to the perioperative instability.
anastomotic end of the gastric conduit.37 Intrao- An alternative strategy to maintain perfusion
perative use of vasopressors has, however, been pressure to the tip of the gastric conduit is
Conduit Necrosis Following Esophagectomy 465
Table 2
Esophageal tube conduits
Abbreviations: AV, arteriovenous; LIMA, left internal mammary artery; LIMV, left internal mammary vein; SMA, superior
mesenteric artery.
From Blackmon SH, Dickinson KJ. Atlas of esophageal intervention. New York: Springer; 2015; with permission.
aggressive fluid resuscitation. The administration norepinephrine to maintain a mean arterial pres-
of intraoperative volume allows the patients’ blood sure of 65 mm Hg or greater reduced the incidence
pressure and, hence, conduit perfusion pressure of pneumonia and respiratory complications with
to be maintained. This strategy may, however, no increase in anastomotic leak or incidence of
necessitate a large volume of fluid, which conduit ischemia.39
may have implications in terms of pulmonary With regard to conduit perfusion, current
edema and bowel edema responsible for postop- evidence does not support one approach over
erative ileus. It has been shown that a restrictive in- the other when fluid resuscitation or vasopressor
traoperative fluid regimen coupled with administration are used to correct the decrease
in blood pressure seen in 40% to 60% with general
anesthetic agent administration. This is similar to
the current evidence regarding free flaps in plastic
surgery.40–43 Recently, however, it has been
shown that intraoperative vasopressor use is not
uncommon during free tissue transfer flaps and
this did not seem to adversely affect patient
outcomes.44 Although care should be taken
extrapolating these results to the often-pedicled
gastrointestinal reconstructions after esophagec-
tomy, it may be that judicious use of vasopressors
during esophageal resections is not always harm-
ful. In fact, hypotension with concomitant reduced
conduit perfusion is not benign. Clear communica-
tion between the surgical and anesthetic team is
Fig. 1. SPY imaging of jejunal conduit demonstrating required to form a management strategy tailored
perfusion after it has been tunneled. to each patient.
466 Dickinson & Blackmon
Management of Acute Intraoperative conduit. The jejunum and the colon should be as-
Conduit Loss sessed for suitability. The jejunum is preferred in
this situation as the colon is unlikely to be
Intraoperative conduit loss is dreaded. This loss
prepped. Consideration should be given to super-
may occur because of injury to the right gastroepi-
charging the graft, and intraoperative consultation
ploic artery (GEA) during dissection, inadequate
with plastic surgical colleagues should be ar-
perfusion from the right GEA caused by previous
ranged. Reconstruction with the jejunum or colon
surgery, or because of patient factors, such as
can proceed in the usual fashion after resection
arterial disease or intraoperative hemodynamic
of the ischemic stomach conduit. The authors
instability. The key factor in appropriate manage-
advocate preserving as much of the stomach
ment is timely recognition of intraoperative conduit
conduit as possible as this can then be used in
loss. Conduit loss can be occult. In this situation
the reconstruction.
methods of assessing conduit perfusion may
When there is concern for the viability of the
suggest inadequate blood flow, or clinically the
conduit, particularly in patients with significant co-
conduit may look hypoperfused. If action is not
morbidity, an alternative option to an end esopha-
taken at this point, definite and irreversible conduit
gostomy and gastrostomy has been described.
loss will occur. Supercharging the conduit and
Oezcelik and colleagues16 describe the technique
enabling adequate arterial supply and venous
of delayed esophagogastrectomy for patients
drainage should be considered if feasible (ie, for je-
with conduit ischemia intraoperatively. A cervical
junal or colonic conduits). Gastric conduits are not
esophagostomy was performed and the gastric
usually supercharged. If supercharging of the
conduit brought to the neck either through the pos-
conduit can be performed and perfusion improves,
terior mediastinum or retrosternally. The conduit
that is, peristalsis and pulsatile blood flow in the
was then secured in the neck, but no anastomosis
mesenteric arcade, the intestinal anastomosis
was performed. At 90 days postoperatively, by re-
can be completed. If there is any doubt as to the
opening the cervical incision, the gastric conduit
viability of the conduit, anastomosis should not
was identified by locating a marker suture. When
be performed. When this is the case, or in the
the conduit looked healthy, anastomosis was per-
case of frank conduit ischemia or even necrosis,
formed, in some cases requiring resection of the
a damage-control strategy should be used. If he-
left half of the manubrium, left sternoclavicular
modynamic instability in patients has contributed
joint, and the medial portion of the first rib. This
to or resulted from the conduit failure, it may be
technique was used in 37 out of 554 patients over
appropriate to simply resect the ischemic/necrotic
a 7-year period; 35 patients were reconstructed,
area and close the patients. This approach allows
and 9% experienced dysphagia postoperatively.
transfer to an intensive care facility and optimiza-
tion. This solution is not durable, and patients
Management of Delayed Conduit Loss
must be appropriately drained before closure (ie,
nasogastric [NG], gastrostomy when appropriate) The key principle to managing delayed conduit
and brought back to the operating room within ischemia/necrosis is to identify this early. Patients
24 to 48 hours for definitive management. The with conduit necrosis may be extremely unwell
definitive surgical approach would be defunction- and deteriorate suddenly. More often there is an
ing esophagostomy, venting gastrostomy, feeding insidious course. Patients may develop a tachy-
jejunostomy, and preserving as much of the cardia, pyrexia, leukocytosis, or unexplained dete-
conduit as possible while resecting all ischemic/ rioration in condition in the absence of pneumonia
necrotic tissue. or an anastomotic leak. In these patients, it is
Intraoperative conduit loss in stable patients important to consider and aggressively investigate
may be approached differently. In the case of conduit ischemia to enable this to be managed
loss of the gastric conduit as a consequence of appropriately. Even with this in mind, mortality of
local anatomy or vascular injury, other conduits this condition approaches 90%.
should be considered. Patient factors and length If patients are systemically unwell and the diag-
of surgery should be taken into consideration. If, nosis apparent, for example, bloody NG aspirate,
for example, the surgery has already taken many lactic acidosis, hemodynamic instability, it is
hours, it may be better to formally defunction pa- appropriate to proceed to the operating room.
tients (esophagostomy, venting gastrostomy, and Patients should be resuscitated aggressively and
feeding jejunostomy). If the operative time is rela- broad-spectrum IV antibiotics commenced.
tively short and there is minimal edema of the Before incision, EGD can be performed to confirm
bowel, reconstruction can be performed at the the diagnosis; however, in this situation, the
same time as resection of the ischemic/necrotic authors would argue video-assisted thoracic
Conduit Necrosis Following Esophagectomy 467
surgery/thoracotomy/reopening of the neck inci- EGD. In a manner similar to the grading of anasto-
sion (depending on esophagectomy approach) is motic leaks following esophagectomy, a grading
required to visualize the conduit. If the conduit is system for ischemic changes and leak apparent
unsalvageable, this should be resected. An end on EGD has been proposed.50 Grade I ischemia
esophagostomy should be performed, with a vent- represents ‘dusky bluish colored mucosa in the vi-
ing gastrostomy (if some stomach is salvageable) cinity of the anastomosis covered with tenacious
and feeding jejunostomy. metallic appearing mucous that could not be
washed off’; grade II ‘partial anastomotic break-
Investigation of conduit ischemia after surgery down with equivocal viability of normal pink mu-
In patients in whom diagnosis is in doubt, investi- cosa margins’; grade III ‘complete circumferential
gation is required to determine the viability of the breakdown with pink mucosal margins’; and grade
conduit. Several studies have been used to eval- IV ‘complete conduit ischemia manifested by
uate conduits following esophagectomy and necrotic black mucosa throughout the gastric
include meglumine diatrizoate (Gastrografin) swal- conduit with the anastomosis still intact.’ The
low, CT, and EGD. Early detection is critical to sur- degree of ischemia observed on EGD can guide
vival in these patients. management. For example, those patients with
More traditionally, upper gastrointestinal con- milder ischemic changes, which may be associ-
trast swallows have been used in order to assess ated with a small anastomotic leak, may be
the esophageal conduit postoperatively for anas- managed successfully with stenting, venting,
tomotic leak. Gastrografin, barium, and iohexol enteral feeding, and antibiotics. Page and col-
(Omnipaque) can all be used. Barium can persist leagues51 have described how all patients under-
for some time and interfere with subsequent imag- going esophagectomy safely underwent routine
ing, whereas Gastrografin is water soluble and endoscopy in the first postoperative week to
does not. Gastrografin, however, can cause pneu- detect anastomotic leakage after esophagectomy.
monitis; Omnipaque may be preferred in this situ- In these patients, clinical care was guided by the
ation as it is less caustic if aspirated. EGD findings and postoperative conduit perfusion
It can be argued that a leak is only significant if problems were preempted.
there is evidence of clinical deterioration. Up to
40% of anastomotic leaks are detected using Gas-
Optimization of Patients Before
trografin swallow in asymptomatic patients.45 In
Reconstruction
addition, the mucosa cannot be adequately visual-
ized during contrast swallow. It is possible to Having survived the necrotic conduit and subse-
see ischemic-appearing changes (cobblestoning quent surgical intervention, it is essential for
surface changes or change to a dark color) of the patients to recuperate before considering recon-
mucosa with conduit ischemia, but at this point struction. During the in-hospital postoperative
patients are likely to be critically unwell. The course, sepsis should be managed appropriately,
administration of oral contrast is also associated with IV antibiotics depending on culture results
with a risk of aspiration and is not possible in intu- and drainage of collections as appropriate. During
bated patients.46,47 Contrast could be cautiously this time, enteral nutrition should be established.
administered through an NG tube but again is Dietician and physiotherapy input are essential
associated with aspiration risk. For these reasons, during this period. Patient education with regard
several studies have shown routine esophageal to their esophagostomy is critical. Patients should
contrast swallow to have potential dangers and be taught how to care for their skin and their appli-
suggested their use should be limited.46–49 ance. They should be educated as to when to re-
CT with oral contrast may be considered more present to their physician, for example, stenosis
sensitive than contrast swallow and is able to of the esophagostomy, cellulitis.
give added information with regard to pneumonia,
pleural effusion, and other thoracic or abdominal
Reconstruction Options
abnormality. Despite this, aspiration risk persists;
it has been demonstrated that a normal CT exam- There are several ways to create a neo-esophagus
ination does not exclude ischemia of the gastric when resection of the esophagus is required. The
conduit.50 stomach is always the first choice of esophageal
The use of EGD to assess the conduit after reconstruction, when available. In order to ensure
esophagectomy has been viewed with caution adequate blood flow to the conduit and reduce
because of the concern of disrupting the fresh the risk of conduit necrosis, careful mobilization
anastomosis. Subtle mucosal ischemia that would of the stomach is required, paying attention not
not be radiologically apparently can be seen on to damage the gastroepiploic vessels and use
468 Dickinson & Blackmon
minimal handling of the conduit to protect the sub- if the right colon is less than 8 cm in diameter, the
mucosal arcade. right colonic artery is not dominant over the
When the gastric conduit is not suitable or ileocolic artery, and preoperative mesenteric angi-
frankly ischemic/necrotic, the esophagus can be ography does not demonstrate a dominant
reconstructed using jejunum14 or colon.52 Short je- ascending branches of the inferior mesenteric ar-
junal grafts can be used (eg, the Merendino proce- tery. The substernal route is often preferred for
dure for a gastrointestinal junction gastrointestinal colonic interposition, and the conduit may also
stromal tumor resection and reconstruction) but be supercharged. As with the supercharged jeju-
are not suitable for reconstruction after gastric nal grafts, resection of the left hemi-manubrium,
conduit necrosis. When using the jejunum to clavicle, and first rib is required to allow space
reconstruct the esophagus, this can be done in a for the conduit, to perform the vascular anastomo-
roux-en-Y fashion if the esophageal remnant is of ses, and to avoid obstruction and venous
sufficient length and vascularity. If the esophagus engorgement of the conduit caused by entrap-
is shorter and the reconstruction must travel a ment. At long-term follow-up in expert hands, the
greater distance, a supercharged jejunal graft quality of life as assessed by the RAND 36-Item
can be used to reconstruct the esophagus. Super- Short Form Health Survey and assessment of
charging the jejunum was first described in 1947, gastrointestinal function demonstrated that 89%
but it only recently has been more widely used.53 of patients experienced no dysphagia, 84% were
The authors recommend supercharging the jejunal free of regurgitation, and 84% free of heartburn.
graft, and this surgery should be performed in Importantly, 90% of patients were within normal
collaboration with plastic surgical colleagues. body mass index limits.55 In this series, 7 of the
The microvascular arterial and venous anastomo- 79 patients underwent reoperation for colonic
ses are critical to the survival of this graft, which redundancy.
is often run on the second jejunal branch of the su-
perior mesenteric artery, the first preserved to Postreconstruction Management
ensure adequate blood supply to the biliopancre- Careful follow-up of patients in whom esophageal
atic jejunal limb. An indicator flap is left exterior- reconstruction has been necessary for conduit ne-
ized so that in the postoperative course of the crosis is essential. This follow-up is not only to
surgery the blood supply to the conduit can be in- detect any evidence of tumor recurrence for those
ferred and monitored. The principles of manage- who underwent esophageal resection for malig-
ment of plastic surgical flaps are applied in the nancy. Quality-of-life assessment is critical to pro-
perioperative period with avoidance of hypoten- vide the best outcomes for these patients. Medical
sion and most advocate avoidance of vasopressor management, for example, of reflux symptoms, is
agents. The perioperative risk in these patients is often all that is necessary to improve functional
more than those undergoing a primary esopha- outcomes. However, in some, revision surgery
gectomy with gastric conduit reconstruction. Early will be required, for example, if redundancy of
complications have been reported with pneu- the conduit is encountered. Complex esophageal
monia in 30% of patients and a higher anastomotic reconstruction is not commonplace, and patient
leak rate of 32% and a 5% incidence of graft ne- support groups provide an invaluable source of
crosis.13 The 90-day mortality was 10% in this se- empathy and advice to these patients.
ries. This finding is a reflection of the more tenuous
blood supply to the graft, even with supercharging, REFERENCES
and the complex nature of this surgery. Despite
this, postoperative quality of life and graft function 1. DeMeester TD, Johansson K-E, Franze I, et al. Indi-
is generally good. Eighty-three percent of patients cations, surgical technique, and long-term func-
were able to return to eating their regular diet, and tional results of colon interposition or bypass. Ann
segmental peristalsis was observed on follow-up Surg 1988;208:460–74.
manometry. 2. Wang LS, Huang MH, Huang BS, et al. Gastric sub-
When colonic conduits are used to reconstruct stitution for resectable carcinoma of the esophagus:
the esophagus after ischemia or necrosis of the an analysis of 368 cases. Ann Thorac Surg 1992;53:
gastric conduit, the left or the right colon may be 289–94.
used. Advocates of the left colon for esophageal 3. Horvath OP, Lukacs L, Cseke L. Complications
reconstruction suggest that the smaller diameter following esophageal surgery. Recent Results Can-
allows less distension and redundancy and less cer Res 2000;155:161–73.
variation in the blood supply and the conduit ulti- 4. Gaissart HA, Mathisen DJ. Short segment colon and
mately can be a longer length.1,5,54 Kesler and col- jejunal interposition. Semin Thorac Cardiovasc Surg
leagues52 suggest that the right colon can be used 1992;4:328–35.
Conduit Necrosis Following Esophagectomy 469
5. Cerfolio RJ, Allen MS, Deschamps C, et al. Esopha- 19. Schröder W, Beckurts KT, Stahler D, et al. Microcir-
geal replacement by colon interposition. Ann Thorac culatory changes associated with gastric tube
Surg 1995;59:1382–4. formation in the pig. Eur Surg Res 2002;34:411–7.
6. Fujita H, Yamana H, Sueyoshi S, et al. Impact on 20. Schröder W, Holscher AH, Bludau M, et al. Ivor-
outcome of additional microvascular anastomosis- Lewis esophagectomy with and without laparo-
supercharge-on colon interposition for esophageal scopic conditioning of the gastric conduit. World J
replacement: comparison and multivariate analysis. Surg 2010;34:738–43.
World J Surg 1997;21:998–1003. 21. Liebermann-Meffert DMI, Meier R, Siewert JR.
7. Kassis ES, Kosinski AS, Ross P Jr, et al. Predictors of Vascular anatomy of the gastric tube used for
anastomotic leak after esophagectomy: an analysis esophageal reconstruction. Ann Thorac Surg 1992;
of the society of thoracic surgeons general thoracic 54:1110–5.
database. Ann Thorac Surg 2013;96:1919–26. 22. Akiyama S, Ito S, Sekiguchi H, et al. Preoperative
8. Lerut T, Coosemans W, Decker G, et al. Anastomotic embolization of gastric arteries for esophageal can-
complications after esophagectomy. Dig Surg 2002; cer. Surgery 1996;120:542–6.
19:92–8. 23. Bludau M, Holscher AH, Vallbohmer D, et al.
9. Low DE, Alderson D, Cecconello I, et al. Interna- Vascular endothelial growth factor expression
tional consensus on standardization of data following ischemic conditioning of the gastric
collection for complications associated with esopha- conduit. Dis Esophagus 2013;26:847–52.
gectomy: esophagectomy complications consensus 24. Markar SR, Arya S, Karthikesalingam A, et al. Tech-
group (ECCG). Ann Surg 2015;262(2):286–94. nical factors that affect anastomotic integrity
10. Schaheen L, Blackmon SH, Nason KS. Optimal following esophagectomy: systematic review and
approach to the management of intrathoracic meta-analysis. Ann Surg Oncol 2013;20:4274–81.
esophageal leak following esophagectomy: a sys- 25. Bulkley GB, Zuidema GD, Hamilton SR, et al. Intrao-
tematic review. Am J Surg 2014;208:536–43. perative determination of small intestinal viability
11. Stephens EH, Correa AM, Kim MP, et al. Classifica- following ischemia injury: a prospective, controlled
tion of esophageal stent leaks: leak presentation, trial of two adjuvant methods (Doppler and fluores-
complications and management. Ann Thorac Surg cein) compared with standard clinical judgment.
2014;98:297–303. Ann Surg 1981;93:628–37.
12. Baker CR, Forshaw MJ, Gossage JA, et al. Long- 26. Karliczek A, Benaron DA, Bass PC, et al. Intraoper-
term outcome and quality of life after supercharged ative assessment of microperfusion with visible light
jejunal interposition for oeosphageal replacement. spectroscopy in esophageal and colorectal anasto-
Surgeon 2014;13(4):187–93. moses. Eur Surg Res 2008;41:303–11.
13. Blackmon SH, Correa AM, Skoracki R, et al. Super- 27. Karliczek A, Harlaar NJ, Zeebregts CJ, et al. Sur-
charged pedicled jejunal interposition for esopha- geons lack predictive accuracy for anastomotic
geal replacement: a 10-year experience. Ann leakage in gastrointestinal surgery. Int J Colorectal
Thorac Surg 2012;94:1104–11. Dis 2009;24:569–76.
14. Gaur P, Blackmon SH. Jejunal graft conduits after 28. Pacheco PE, Hill SM, Henriques SM, et al. The novel
esophagectomy. J Thorac Dis 2014;6(Suppl 3): use of intraoperative laser-induced fluorescence of
S333–40. indocyanine green tissue angiography for evaluation
15. Briel JW, Tamhanker AP, Hagen JA, et al. Preva- of the gastric conduit in esophageal reconstructive
lence and risk factors for ischemia, leak, and stric- surgery. Am J Surg 2013;205:349–52.
ture of esophageal anastomosis: gastric pull up 29. Ballard JL, Stone WM, Hallett JW, et al. A critical
versus colon interposition. J Am Coll Surg 2004; analysis of adjuvant techniques used to assess
198:536–42. bowel viability in acute mesenteric ischemia. Am
16. Oezcelik A, Banki F, DeMeester SR, et al. Delayed Surg 1993;59:309–11.
esophagogastrectomy: a safe strategy for manage- 30. Gurtner GC, Jones GE, Neligan PC, et al. Intraoper-
ment of patients with ischemic gastric conduit at ative laser angiography using the SPY system: re-
time of esophagectomy. J Am Coll Surg 2009;208: view of the literature and recommendations for
1030–4. use. Ann Surg Innov Res 2013;7:1.
17. Urschel JD, Antkowiak JG, Delacure MD, et al. 31. Murawa D, Hunerbein M, Spychala A, et al. Indocya-
Ischemic conditioning (delay phenomenon) im- nine green angiography for the evaluation of gastric
proves esophagogastric anastomotic healing in the conduit perfusion during esophagectomy – first
rat. J Surg Oncol 1997;66:254–6. experience. Acta Chir Belg 2012;122:275–80.
18. Urschel JD. Ischemic conditioning of the stomach 32. Kumagi Y, Ishiguro T, Haga N, et al. Hemodynamics
may reduce the incidence of esophagogastric anas- of the reconstructed gastric tube during esophagec-
tomotic leaks complicating esophagectomy: a hy- tomy: assessment of outcomes with indocyanine
pothesis. Dis Esophagus 1997;10:217–9. green fluorescence. World J Surg 2014;38:138–43.
470 Dickinson & Blackmon
33. Rino Y, Yukawa N, Sato T, et al. Visualization of 44. Kelly DA, Reynolds M, Crantford C, et al. Impact of
bloody supply route to the reconstructed stomach intraoperative vasopressor use in free tissue transfer
by indocyanine green fluorescence imaging during for head, neck and extremity reconstruction. Ann
esophagectomy. BMC Med Imaging 2014;22:14–8. Plast Surg 2014;72:S135–8.
34. Michelet P, Roch A, D’Journo XB, et al. Effect of 45. Schaible A, Sauer P, Hartwig W, et al. Radiologic
thoracic epidural anesthesia on gastric blood flow versus endoscopic evaluation of the conduit after
after oesophagectomy. Acta Anaesthesiol Scand esophageal resection: a prospective, blinded intra-
2007;51:587–94. individually controlled diagnostic study. Surg
35. Al-Rawi OY, Pennefather SH, Page RD, et al. The ef- Endosc 2014;28:2078–85.
fect of thoracic epidural bupivacaine and an intrave- 46. Sauvanet A, Baltar J, Le Mee J, et al. Diagnosis and
nous adrenaline infusion on gastric tube flow during conservative management of intrathoracic leakage
esophagectomy. Anesth Analg 2008;106:884–7. after oesophagectomy. Br J Surg 1998;85:1446–9.
36. Theodorou D, Drimousis PG, Larentzakis A, et al. 47. Griffin S, Lamb P, Dresner S, et al. Diagnosis and
The effect of vasopressors on perfusion of the management of a mediastinal leak following radical
gastric graft after esophagectomy. An experimental oesophagectomy. Br J Surg 2001;88:1346–51.
study. J Gastrointest Surg 2008;12:1497–501. 48. Fan S, Lau W, Yip W, et al. Limitations and dangers
of Gastrografin swallow after esophageal and upper
37. Pathak D, Pennefather SH, Russell GN, et al. Phen-
gastric operations. Am J Surg 1988;160:322–3.
ylephrine infusion improves blood flow to the stom-
49. Page R, Shackcloth M, Russell G, et al. Surgical
ach during oesophagectomy in the presence of
treatment of anastomotic leaks after oesophagec-
thoracic epidural anesthesia. Eur J Cardiothorac
tomy. Eur J Cardiothorac Surg 2005;27:337–43.
Surg 2013;44:130–3.
50. Oezcelik A, Banki F, Ayazi S, et al. Detection of
38. Paul DJ, Jamieson GG, Watson DI, et al. Periopera-
gastric conduit ischemia or anastomotic breakdown
tive risk analysis for acute respiratory distress syn-
after cervical esophagogastrectomy: the use of
drome after elective oesophagectomy. ANZ J Surg
computed tomography scan versus early endos-
2011;81:700–6.
copy. Surg Endosc 2010;24:1948–51.
39. Buise M, Van Bommel J, Mahra M, et al. Pulmonary
51. Page RD, Asmat A, McShane J, et al. Routine
morbidity following esophagectomy is decreased af-
endoscopy to detect anastomotic leakage after
ter introduction of a multimodal anesthetic regimen.
esophagectomy. Ann Thorac Surg 2013;95:292–8.
Acta Anaesthesiol Belg 2008;59:257–61.
52. Kesler KA, Pillai ST, Birdas TJ, et al. ‘Supercharged’
40. Monroe MM, McClelland J, Swide C, et al. Vaso- isoperistaltic colon interposition for long-segment
pressor use in free tissue transfer surgery. Otolar- esophageal reconstruction. Ann Thorac Surg 2013;
yngol Head Neck Surg 2010;142:169–73. 95:1162–9.
41. Cordeiro PG, Santamaria E, Hu GY, et al. Effects of 53. Longmire WP. A modification of the roux technique
vasoactive medications on the blood flow of island for antethoracic esophageal reconstruction: an
musculocutaneous flaps in swine. Ann Plast Surg anastomosis of the mesenteric and internal mam-
1997;39:524–31. mary blood vessels. Surgery 1947;22:94–100.
42. Massey MF, Gupta DK. The effects of systemic 54. Thomas P, Fuentes P, Guidicelli R, et al. Colon inter-
phenylephrine and epinephrine on pedicle artery position for esophageal replacement: current indica-
and microvascular perfusion in a pig model of myoa- tions and long-term function. Ann Thorac Surg 1997;
dipocutaneous rotational flaps. Plast Reconstr Surg 64:757–64.
2007;120:1289–99. 55. Greene C, De Meester SR, Augustin F, et al. Long-
43. Harris L, Goldstein D, Hofer S, et al. Impact of vaso- term quality of life and alimentary satisfaction after
pressors on outcomes in head and neck free tissue esophagectomy with colon interposition. Ann Thorac
transfer. Microsurgery 2012;32:15–9. Surg 2014;98:1713–20.
Functional Conduit
D i s o rde r C o m p l i c a t i n g
Esophagectomy
Kamran Mohiuddin, MDa, Donald E. Low, MD, FRCS(C)b,*
KEYWORDS
Esophagectomy Disorders Functional conduit disorder Esophageal cancer
KEY POINTS
Multimodality therapy as well as early detection of esophageal cancer has increased long-term sur-
vival, making postoperative quality of life an important issue in a larger portion of patients following
esophagectomy.
Functional problems after esophagectomy can dramatically affect quality of life.
Anastomosis placed in the mid and lower chest can increase the incidence of delayed gastric
emptying and reflux.
Delayed gastric emptying, anastomotic stricture, dumping, and reflux are common sequelae of
esophagectomy.
Surgeons should be committed to long-term follow-up of these patients and develop strategies for
treating these functional disorders.
a
Virginia Mason Medical Center, 1100 Ninth Avenue, Seattle, WA 98101, USA; b Digestive Disease Institute
Esophageal Center of Excellence, Ryan Hill Research Foundation, Virginia Mason Medical Center, 1100 Ninth
Avenue, Seattle, WA 98101, USA
* Corresponding author.
E-mail address: donald.low@virginiamason.org
survival rate in patients amenable to definitive potential to reduce conduit dysfunction. The resto-
treatment ranges from 19% to 30%.14 Barrett’s ration of foregut function after esophagectomy
surveillance programs have increased detection greatly affects patient satisfaction and continues
of early-stage cancer, increasing the potential for to challenge esophageal surgeons. This review fo-
cure and making the maintenance of quality of cuses specifically on functional conduit disorders
life increasingly important. Esophagectomy has after esophagectomy.
the potential to be a life-altering operation. Pa-
tients can lose up to 15% to 20% of their body DELAYED GASTRIC EMPTYING
weight from the time of diagnosis through the first
6 months after the surgery, but this trend typically After esophagectomy, the stomach is commonly
stabilizes after 6 months. Most patients adapt to used to restore the continuity of the upper gastro-
smaller, more frequent, meals. Simple sugars intestinal tract.20,21 However, functional conduit
and fluids at mealtime may need to be avoided un- disorders, such as DGE, can occur, which signifi-
til the function of the conduit is established. It is cantly impacts postoperative nutrition and quality
important to match the surgical approach accord- of life. DGE puts patients at increased risk of aspi-
ing to the tumor and physiologic issues. Other fac- ration pneumonia, malnutrition, decreased patient
tors that can affect functional outcome include satisfaction, prolonged hospital stay, and read-
choice of reconstructive conduit, and location, as missions.22,23 The current literature reports the
well as technique of anastomosis. Short-term incidence of DGE as ranging between 10% and
conduit function will vary but can be impacted by 50%.24–27 However, documenting the actual
timing of nasogastric (NG) tube removal, timing incidence is complex because the definition of
of resumption of oral diet, and utilization of post- DGE varies among institutions. Most of the time
operative jejunostomy feeding tubes. patients complain of reflux, regurgitation, early
The reconstructive method of choice for most satiety, pain, and bloating while eating.
surgeons after esophagectomy is gastric interposi- DGE may result from a number of causes: va-
tion (>90% of the cases). Colon interposition is an gotomy, torsion of the stomach in the right chest,
appropriate alternative but in many centers the compression of distal gastric conduit at the hiatus,
colon is used when the stomach is unavailable gastric conduit redundancy, and pyloric obstruc-
due to tumor extension or previous surgery.15,16 Ac- tion. The incidence of DGE appears to be higher
cording to the Society of Thoracic Surgeons (STS) in patients with intrathoracic anastomosis due to
guidelines, the gastric tube is the preferred esoph- the increased potential of gastric conduit redun-
ageal substitute. Alternatively, in some cases, the dancy above the level of the diaphragm. Early
small intestine, pedicled Roux-en-Y reconstruction satiety after esophagectomy is common, and re-
(typically appropriate to the level of the inferior pul- sults from diminished motor function and loss of
monary vein), free graft (requires microvascular gastric reservoir. Immediately after esophagec-
anastomosis), or pedicled skin-muscle flaps, can tomy, the gastric conduit functions as a nonmotile
be selectively used. Either thoracic or cervical anas- tube, and ingested food must empty by gravity
tomoses are applied for gastric tube reconstruc- alone (Fig. 1). Patients should be routinely advised
tion. The creation of the gastric neo-esophagus to initiate oral nutrition with multiple (5–6) small
is associated with substantial alteration to the portions throughout the day, rather than attempt
stomach blood supply. Ligation of left gastric, short to consume 3 regular meals in the first month
gastric, and left gastroepiploic arteries typically re- after reconstruction. Gastric contractility is not
sults in significant potential for ischemia at the tip of completely lost after esophagectomy and the
the conduit, which is typically the location of the denervated stomach may recover some motor
anastomosis.17 Anastomotic methods can include function over time. This is one of the issues that
hand-sewn anastomoses (continuous and interrup- surgeons use to support a stepwise resumption
ted sutures, single-layer or double-layer sutures, of oral intake following esophagectomy and the
absorbable or nonabsorbable stitches), stapling use of temporary jejunostomy tubes. The motor
(circular and linear), and combined hand-sewn activity of the gastric tube is affected by the size,
and stapled anastomoses.18,19 shape, and location of the neo-esophagus. The
Delayed gastric emptying (DGE), dumping syn- gastric conduit can be placed in 3 locations: (1)
drome, anastomotic stricture/leak, and reflux are the original esophageal bed in posterior medias-
recognized postoperative complications that can tinum, (2) retrosternal space, or (3) tunneled sub-
contribute to nutritional problems and impact cutaneously anterior to the sternum. There have
postoperative quality of life. There is no one surgi- been no studies to date that show significant dif-
cal approach that can eliminate any one of these ferences in conduit emptying between these
complications, but certain techniques have the various pathways.28–30
Functional Disorder, Esophagectomy 473
The stomach has historically been the preferred reconstruction, the greatest challenge is the
conduit for esophageal replacement. The most preparation of the esophageal conduit, completion
desirable width of gastric conduit has been of the esophageal anastomosis, and revisional
debated. Every effort should be made to construct conduit ischemia. The risk of ischemia is related
the gastric conduit to provide a direct route to conduit type, length of conduit, comorbidities,
through the hiatus and into the small bowel. and/or operative technique.
Most surgeons strongly recommend avoiding Our current approach involves producing a rela-
placement of the anastomosis in the lower chest. tively narrow 3 to 4 cm conduit, targeting anasto-
Conduit redundancy above the diaphragm allows mosis in the upper chest 3 to 4 cm above the
food to accumulate above the diaphragm. This azygous vein or in the cervical region. Special
accumulation can result in retention of food and attention is paid toward producing a straight
fluids within the conduit, promoting reflux, regurgi- (nonredundant) pathway for the conduit into the
tation, and ultimately dilation of the conduit. abdomen (Figs. 3 and 4). We do not routinely
Patients who present with DGE and are found on perform pyloroplasty or pyloromyotomy unless
radiologic examination to have significant redun- there is a documented pyloric stricture. A Kocher
dancy and food retention, should undergo esoph- maneuver is performed to bring the pylorus up to
agogastroduodenoscopy (EGD) to assess the 2 to 4 cm from the diaphragmatic hiatus, which
pylorus with consideration of dilatation and Botox also straightens the pathway into duodenum. We
injection. Patients should be advised to eat in routinely insert a jejunostomy tube at the time of
small volumes and a “witnessed” upper gastroin- esophagectomy. Low-volume jejunostomy feeds
testinal series (member of the surgical team is begin on postoperative day 1, in association with
present during the upper gastrointestinal study to an early mobilization protocol. Patients typically
direct positioning of the patient to discover which undergo a swallow study on postoperative day 3
orientation best promotes conduit emptying) can or 4 to examine the anastomosis and, equally
be used to determine the optimal postprandial importantly, to assess gastric emptying (Figs. 5
positioning for promoting the gastric emptying. and 6). The findings of these studies are catego-
Revision surgery may be required, which can rized into 2 groups: (1) gastric emptying is immedi-
involve mobilization of the distal gastric conduit ate, which results in the NG tube being removed
to place additional length below the diaphragm. and limited oral intake being initiated the next
If significant conduit dilation is present, the gastric morning, or (2) gastric emptying is delayed, which
conduit may need revision to remove the supra- results in the NG tube remaining in place and the
diaphragmatic redundancy and improve conduit patient is placed on erythromycin. After 48 hours,
emptying. This revisional surgery must be done we repeat the swallow study and, if gastric
with careful attention paid to the preservation of emptying is still delayed, the NG tube is removed
the conduit blood supply (Fig. 2). According to and we perform endoscopy with pyloric dilation
Wormuth and colleagues,38 during esophageal (to 1.5 cm maximum diameter) and also inject
Fig. 2. (A) A 68-year-old female patient 2 years after Ivor Lewis esophagectomy for cancer. Severe problem with
early satiety, regurgitation, and weight loss. Chest radiograph shows a dilated redundant conduit with an air
fluid level high (see arrow) in the right chest. (B) Contrast study of the same patient showing supra-
diaphragmatic reservoir in a redundant gastric conduit. (C) Patient did not respond to prokinetic agents, pyloric
dilation, and Botox injection. She ultimately underwent repeat right thoracotomy with resection of gastric
conduit redundancy with repeat contrast study showing more regular conduit contour and improved emptying.
Functional Disorder, Esophagectomy 475
Botox. Our current target day for discharge is day while having episodes of abdominal cramping
7. At time of hospital discharge, patients are on and bloating, flushing, and explosive diarrhea.
liquid or pureed oral intake (1 cup per hour) along Based on the presentation of symptoms, dumping
with nocturnal jejunostomy feeds. We provide syndrome can be divided into Early (10–30 minutes
structured dietary advancement for the next after eating) and Late (1–3 hours after eating), with
6 weeks with decreasing jejunostomy feeds. the patients experiencing a combination of both
presentations. Early symptoms of dumping syn-
drome postesophagectomy are more common,
Dumping
most frequently in women and in a younger age
One of the common clinical complications of group. Decreased capacity of the gastric conduit
altered gastric function postesophagectomy is and changes in pyloric orientation and function
dumping syndrome. It is estimated that 5% to can increase the tendency for dumping, especially
68% of patients show some manifestation of in the months after esophagectomy. Most cases
dumping syndrome after esophagectomy.39–42 are mild and controlled with dietary modification.
However, 5% to 10% of patients show symptoms The pathophysiology of dumping syndrome is
with moderate severity, and 1% to 5% of patients multifactorial. Dumping syndrome manifests as a
experience severe symptoms.33 Patients typically variety of gastrointestinal and vasomotor symp-
report feeling dizzy, nauseated, and diaphoretic, toms. It is usually a clinical diagnosis but can be
Fig. 5. A 72-year-old-male patient 3 days after Ivor Lewis esophagectomy undergoing “witnessed” upper gastro-
intestinal study (water-soluble contrast followed by thin barium) showing (A) intact anastomosis in the upper
chest (arrow) and (B) immediate emptying through the pylorus, which allows NG tube removal and initiation
of oral intake.
confirmed by documenting symptoms of hypogly- Eating small, frequent meals, avoiding liquids
cemia after oral glucose challenge test. Early with meals, and avoiding eating foods containing
dumping, 10 to 30 minutes after ingestion, is typi- sugar and other simple carbohydrates can
cally due to the rapid transit of hyperosmolar decrease the incidence of dumping. Separating
gastric contents into the upper gastrointestinal liquids and solids during meals can prevent, and
tract.43 Late dumping is diagnosed by symptoms may also decrease, the severity of dumping. To
occurring 1 to 3 hours after eating.44 The rapid meet daily caloric needs, intake of protein and fat
transport of carbohydrate to the small intestine should be increased.47–49 Most dumping syn-
triggers the release of insulin and leads to subse- dromes improve over time. Nonsurgical options
quent hypoglycemic symptoms.45,46 Although for management include administration of phar-
effective gastric emptying is the preferred goal macologic therapy. In severe cases, drugs include
after esophagectomy, no current strategy has tincture of opium, diphenoxylate, beta blockers,
been developed to standardize the objective methysergide, acarbose, and octreotide. Although
assessment of gastric emptying after esophageal they have not demonstrated conclusive benefit in
resection. previous reports, they can be considered.50–53
Patients with severe dumping syndrome are rarely
unable to attain adequate nutrition orally. In these
patients, placement of jejunostomy tube or central
venous access for total parenteral nutrition can be
required.
Anastomotic Stricture/Dysphagia
Anastomotic stricture after esophagectomy is an
important postoperative event that impacts quality
of life and the ability to resume a normal pattern of
eating. Anastomotic stricture causes dysphagia;
however, dysphagia after esophagectomy may
not necessarily be associated with an anastomotic
stricture. Most postoperative dysphagia can be
classified as anastomotic or functional due to
oropharyngeal dysfunction. Similarly, strictures
can be classified into scar contracture and/
or stricture associated with localized leak or
ischemia. Anastomotic strictures are reasonably
Fig. 6. Another patient showing delayed gastric common after esophagectomy, with a mean re-
emptying on day 3, which delays NG tube removal ported prevalence of 30% (range, 9%–48%).54–56
and treatment with erythromycin. Previous reports have identified several risk
Functional Disorder, Esophagectomy 477
factors for developing anastomotic strictures, and the moderate stricture can resolve spontane-
including postoperative anastomotic leakage, ously. Early treatment of strictures consists of
poor vascularization of the gastric tube, and a sta- dilation using a through-the-scope balloon or
pled rather than a hand-sewn anastomosis, and wire-directed bougie. The use of bougie dilators
in patients with preoperative cardiac disease, involves 2 mechanisms of dilation: a shearing force
hypoalbuminemia, increased intraoperative blood and a radial force; whereas, balloon dilation in-
loss, hypotension, hypoxemia, and diabetes melli- volves only a radial force. For stricture dilation, it
tus.57 In recent years, the incidence of complica- is essential to first assess the luminal diameter,
tions after esophagectomy has decreased, but and then dilate up to 3 levels at a time. For
the rate of occurrence of anastomotic stricture example, if the stricture diameter is 0.8 cm, initial
has remained relatively stable. dilatation should involve sequential 27, 30, and
Endoscopy and barium swallow examinations 33 French dilation.
are the initial assessments for investigating early For strictures that do not respond to an initial
and late dysphagia after esophagectomy. The dilation, subsequent dilations can be done with
esophagus is a muscular tube with elastic walls. synchronous injection of steroids. The esoph-
In the resting state, the lumen is in a collapsed ageal dilation is routinely performed in an outpa-
state, whereas during the swallowing process, tient setting and intralesional steroid injection
the muscles of the esophagus relax to accommo- is combined with the esophagogastric dilation
date the food bolus (anterior-posterior diameter, procedure. Triamcinolone acetate or acetonide
up to 2 cm; lateral diameter, up to 3 cm). However, 10 mg/mL is the most common dose used, and it
esophagogastric anastomotic healing produces a can be increased up to 40 mg/mL.60–62 The vol-
scar resulting in an inelastic segment associated ume of corticosteroid used per injection has varied
with the scar around the anastomosis. The size from 0.5 to 2.8 mL. In a randomized study, Ram-
of the anastomosis varies depending on the diam- age and colleagues61 standardized the dosage to
eter of the stapler, the anastomotic method, and 0.5-mL aliquots of 20 mg each of 40 mg/mL triam-
an individual’s degree of scar contraction. There- cinolone acetonide. Ramboer and colleagues63
fore, patients who consume solid food that forms used a betamethasone preparation: one 1-mL
a bolus exceeding the diameter of the anasto- vial containing 5 mg betamethasone as a dipropi-
mosis will experience periodic dysphagia and the onate suspension diluted into usually 5 mL, and
potential for bolus obstruction. If the patient is sometimes 10 mL, of normal saline solution and in-
masticating the food adequately, an anastomotic jected as 0.5-mL to 1.0-mL aliquots. Miyashita and
luminal diameter of 1 cm should be adequate to colleagues64 used a total of 8 mg dexamethasone
consume a diversified diet. (2 mL) injected endoscopically into 4 sites (2 mg/
One-layer hand-sewn anastomosis has lower 0.5 mL per site) at the anastomotic site immedi-
rates of stricture formation than 2-layer hand- ately after dilation. However, there were no differ-
sewn anastomosis.58 Similarly, end-to-end circu- ences among the reported studies with regard to
lar stapler anastomosis has a higher rate of the response outcome after using different steroid
stricture formation versus side-to-side longitudinal formulations. The number of injection sessions in
semimechanical anastomosis, which creates a the reported series also varies from only 1, to
larger cross-sectional anastomotic area.55 It is as many sessions as the number of dilations.
recommended to use at least a 25-mm load Kochhar and Makharia60 used a maximum of 4
when using a circular stapler. Most cases with sessions, whereas Rupp and colleagues65 carried
anastomotic stricture respond to early dilation, out a maximum of 5 sessions and Gandhi and
and when performed by an experienced endo- colleagues66 conducted as many as 13 sessions.
scopist, dilation can be safely done as early as The number of sessions is an issue that has not
postoperative day 3 to 5. Early anastomotic stric- yet been settled and requires standardization.
tures are a major contributor to the occurrence Our approach is to use steroid injection in no
of anastomotic leak. Strictures are also common more than 2 dilations and, if unsuccessful in
sequelae of anastomotic leak and there is evi- restoring normal swallowing, move to other treat-
dence to suggest that elective dilation in patients ment approaches.
who have experienced an anastomotic leak after Electroincision is an alternative option to dilation
esophagectomy can decrease the incidence and in patients with refractory or recurrent esophago-
severity of anastomotic stricture. gastric anastomotic strictures. The optimal stric-
Trentino and colleagues59 reported that after a ture for this procedure is short and membranous,
mean of 3.6 dilations, there was an 83% success rather than long and tapered. This treatment
rate. However, several studies have also sug- option is safe and effective, with reports showing
gested that the anastomosis continues to mature results superior to dilation alone.67 Historically,
478 Mohiuddin & Low
patients with refractory strictures have been SEPS over SEMS, which include ease of retrieval,
taught to self-dilate. Although still an acceptable limited local tissue reaction, and lower cost.71–78
option, it requires extensive time for patient in- The self-expandable plastic stent Polyflex
struction and not all patients are suitable. (Boston Scientific, Natick, MA) is a silicone device
Approximately 50% of strictures are resolved with an encapsulated monofilament braid made
with a single dilation.68 If strictures are refractory of polyester and can be used to manage diffi-
to dilation and corticosteroid injection, the next cult stricture after esophagectomy. The mesh is
option is to consider use of a removable self- completely covered by a silicone layer with a
expandable metallic (SEMS) or plastic (SEPS) smooth inner surface and a more structured outer
stent. Stents are not considered first-line treat- surface to decease the incidence of migration.79
ment for early strictures. They are reserved only The edges are protected with silicone to avoid
for patients who have failed previous therapy (usu- impaction and/or tissue damage at the proximal
ally multiple sessions of dilation).69 In rare cases, and distal ends. The radiopaque markers are
the dilation does not open the stricture at all, or located at the proximal, midpoint, and distal
the stricture opens but then closes within a short ends of the stent to help in positioning at the
period of time. In those cases in which the stricture time of insertion. Potential complications include
closes within a short period of time, the goal is to stent migration, airway compression, aspiration,
place the stent to hold the stricture open for pain, nausea, and vomiting. There have also
an extended prolonged period of time, causing been several cases of stricture development at
the scar tissue to remodel around the stent the end of the stent flare, especially in partially
to decrease the incidence of recurrence. We covered stents. These strictures are most
currently recommend leaving the stent in place commonly due to granulation tissue proliferation.
for 2 to 3 weeks (Fig. 7). The granulation tissue and stricture typically
There are 2 major types of esophageal stents: resolve on stent removal. The introduction of pro-
SEMS and SEPS. The metal stents are used for ton pump inhibitors, improved stent technology,
palliation of malignant conditions, and plastic and the development of better and safer endos-
stents are used for the management of benign copy and dilation techniques, has resulted in a
esophageal conditions, such as tracheoesopha- major decrease in the complications secondary
geal fistulas, esophageal strictures, esophageal to the treatment of anastomotic stricture.
perforations, and also for postoperative esopha- Our approach in early dysphagia is to perform a
gectomy leaks. Until recently, there were no swallow study to assess for swallowing mecha-
SEMS approved for use in benign conditions by nism disorder, anastomotic leak, or stricture. If a
the US Food and Drug Administration (FDA), stricture is identified, we perform EGD 2 to 8 weeks
although there are published reports of SEMS be- after esophagectomy with balloon dilation up to
ing used in benign strictures. There are multiple 1.0 cm. In late dysphagia secondary to stricture,
types of prostheses that are available from various we perform EGD and, depending on degree of ste-
manufacturers. The metal stents are made of nosis, we perform wire-directed bougie dilation. If
nitinol (alloy of nickel and titanium) and plastic the stricture is 0.5 cm in diameter, we dilate to 18,
stents are made of polyester. There are 3 types 21, and 24, and then repeat the EGD in 2 weeks to
of SEMS: fully covered, partially covered, and un- assess the stricture. If recurrence persists after 2
covered. There is evidence that covered SEMS are dilations, we use steroid injection of triamcinolone
better than uncovered stents in benign conditions. acetate 10 mg/mL into 4 quadrants (total dose
A fully covered nitinol stent has recently been 40 mg/mL). If stricture persists, we consider either
approved by the FDA (Niti-S; TaeWoong Medical, needle-knife electroincision or will insert an
Seoul, Korea) to be used in benign and malignant expandable removable stent with elective removal
esophageal strictures. Recently, SEPS have plan 2 to 3 weeks later (see Fig. 7).
been increasingly used in the treatment of benign
esophageal diseases that include esophageal
Reflux
strictures, fistulas, perforation, and anastomotic
leaks.70 The SEPS are made of polyester netting Esophageal resection and reconstruction results
embedded in a silicone membrane, creating a in removal of the gastroesophageal antireflux
polyester mesh outer cover with a smooth silicone mechanism, increasing the risk for prolonged
inner lining that is present for the entire length exposure to gastric acid and duodenal juice. Re-
of the stent. The middle and distal portions are of flux is a common problem after esophagectomy,
the same diameter, whereas the proximal end of occurring in 60% to 80% of patients and impacts
the stent is flared in an attempt to prevent distal quality of life.80 Clinical manifestations may
migration. There are potential advantages of include cervical heartburn or with regurgitation,
Functional Disorder, Esophagectomy 479
Fig. 7. A 60-year-old female patient discovered to have localized leak on day 5 following thoraco-abdominal
resection with cervical anastomosis. (A) A Gastrografin swallow study confirming localized anastomotic leak
which was treated with antibiotics and holding oral intake with resolution. (B) Endoscopy of the same patient
demonstrating a tight anastomotic stricture that had been treated with multiple dilations. (C) Endoscopy
following the insertion of a Polyflex stent across the stricture (Boston Scientific). (D) Endoscopic picture immedi-
ately after stent removal 4 weeks after insertion showing widely patent anastomosis.
aspiration, bile-induced and gastric acid-induced together to promote reflux across the anasto-
laryngitis, vomiting, chronic cough, hoarseness, mosis. The pyloric drainage procedures may pro-
inability to lie in a supine position, choking, and mote duodenogastric reflux, leading in turn to
recurrent pneumonias. DGE also can increase bile reflux into the proximal esophagus.
the severity of reflux symptoms. It has been assumed that gastric acid secretion
Symptoms after esophagectomy are typically typically decreases after esophagectomy due to
more troublesome when patients are supine. In vagal interruption. Gutschow and colleagues81
some cases this can lead to sleep deprivation. studied 91 patients who had undergone tubular
The negative intrathoracic pressure in the chest stomach reconstruction associated with esopha-
and the positive intra-abdominal pressure work geal resection by monitoring intraluminal gastric
480 Mohiuddin & Low
pH and bile. There were 3 groups based on the the incidence and severity of postoperative func-
length of follow-up (group 1 1 year, group tional problems.
2 5 1–3 years, group 3 3 years) and the results Surgeons must also be able to identify the signs
showed that 32.3% had normal acid secretion and symptoms of DGE, dumping, and anasto-
within 1 year after esophagectomy, 81.5% by motic strictures, be prepared to advise patients
2 years, and 97.6% after 3 years. regarding lifestyle and medical management and,
The location of the anastomosis, as well as other when appropriate, initiate endoscopic and surgical
technical conduit factors, is an important issue in intervention.
the prevention of reflux after esophagectomy. An
anastomosis below the aortic arch, utilization of REFERENCES
whole stomach, and having a significant portion
of the gastric conduit in the abdomen can increase 1. Siegel R, Ma J, Zou Z, et al. Cancer statistics, 2014.
the reflux symptoms after esophagectomy.82,83 As CA Cancer J Clin 2014;64:9–29.
a result, placement of the anastomosis in the up- 2. O’Rourke I, Tait N, Bull C, et al. Oesophageal can-
per chest (above the aortic arch), tubulization of cer: outcome of modern surgical management.
gastric conduit, and mobilizing of the pylorus to a Aust N Z J Surg 1995;65:11–6.
position within 2 to 8 cm of the diaphragmatic 3. Lewis I. The surgical treatment of carcinoma of the
hiatus is recommended. Also, patients should be oesophagus; with special reference to a new opera-
encouraged to change eating habits, which will tion for growths of the middle third. Br J Surg 1946;
lead to less chance of reflux, especially in the 34:18–31.
early stages after esophagectomy. These should 4. McKeown KC. Total three-stage oesophagectomy
include eating 6 small nutritious meals a day, tak- for cancer of the oesophagus. Br J Surg 1976;63:
ing small bites, chewing food well, sitting in an 259–62.
upright position for 45 to 60 minutes after eating, 5. Orringer MB, Sloan H. Esophagectomy without
avoiding lying in the supine position immediately thoracotomy. J Thorac Cardiovasc Surg 1978;76:
after meals, walking after meals, and sleeping 643–54.
with their head elevated 20 to 30 . The STS 6. Ochsner A, DeBakey M. Surgical aspects of carci-
evidence-based surgical guidelines also recom- noma of esophagus: review of literature and report
mend a high intrathoracic anastomosis above the of four cases. J Thorac Surg 1941;10:401–45.
azygous vein or a cervical anastomoses as the 7. Adelstein DJ, Rice TW, Becker M, et al. Use of con-
preferred anastomotic location, and avoiding a current chemotherapy, accelerated fractionation ra-
low intrathoracic anastomosis. diation, and surgery for patients with esophageal
Some surgeons routinely keep patients on H2 carcinoma. Cancer 1997;80:1011–20.
blocker or proton pump inhibitors after esopha- 8. Keller SM, Ryan LM, Coia LR, et al. High dose
gectomy. There are surgical strategies that have chemoradiotherapy followed by esophagectomy
been proposed to prevent reflux after esophagec- for adenocarcinoma of the esophagus and gastro-
tomy, such as creating an anastomotic fundopli- esophageal junction: results of a phase II study of
cation, intercostal muscle flap, or diaphragmatic the Eastern Cooperative Oncology Group. Cancer
slings. They are complex to create and their effi- 1998;83:1908–16.
cacy remains unproven, and therefore these oper- 9. Kelsen DP, Ginsberg R, Pajak TF, et al. Chemo-
ations have not been widely accepted. therapy followed by surgery compared with surgery
alone for localized esophageal cancer. N Engl J
Med 1998;339:1979–84.
SUMMARY
10. Wright CD, Wain JC, Lynch TJ, et al. Induction
Esophagectomy is a complex surgical procedure therapy for esophageal cancer with paclitaxel and
that routinely impacts quality of life. Improved hyperfractionated radiotherapy: a phase I and II
results associated with early diagnosis in Barrett’s study. J Thorac Cardiovasc Surg 1997;114:811–5.
screening programs and improved neoadjuvant 11. Ando N, Ozawa S, Kitagawa Y, et al. Improvement in
protocols increase the potential for cure in the results of surgical treatment of advanced squa-
patients with esophageal cancer who are treated mous esophageal carcinoma during 15 consecutive
surgically. years. Ann Surg 2000;232:225–32.
Surgeons must recognize that there are 12. Law S, Wong KH, Kwok KF, et al. Predictive factors
important technical issues associated with gastric for postoperative pulmonary complications and mor-
conduit construction, anastomosis construction tality after esophagectomy for cancer. Ann Surg
and location, and approaches to intraoperative py- 2004;240:791–800.
loric management and postoperative nutritional 13. Siewert JR, Stein HJ, Feith M, et al. Histologic tumor
management, and that these issues will impact type is an independent prognostic parameter in
Functional Disorder, Esophagectomy 481
esophageal cancer: lessons from more than 1,000 carcinoma: comparison between the anterior and
consecutive resections at a single center in the posterior mediastinal approaches. Hepatogastroen-
Western world. Ann Surg 2001;234:360–7. terology 1998;45:2224–7.
14. American Cancer Society. Cancer facts and figures 30. van Lanschot JJ, van BM, Oei HY, et al. Randomized
2014. Atlanta (GA): American Cancer Society; comparison of prevertebral and retrosternal gastric
2014. Available at: http://www.cancer.org/research/ tube reconstruction after resection of oesophageal
cancerfactsstatistics/cancerfactsfigures2014/#. Ac- carcinoma. Br J Surg 1999;86:102–8.
cessed May 21, 2014. 31. Cheung HC, Siu KF, Wong J. Is pyloroplasty neces-
15. Hiebert C, Bredenberg C. Selection and placement sary in esophageal replacement by stomach? A pro-
of conduits. In: Pearson F, Cooper J, Deslauriers R, spective, randomized controlled trial. Surgery 1987;
et al, editors. Esophageal surgery. 2nd edition. 102:19–24.
New York: Churchill Livingstone; 2002. p. 794–801. 32. Urschel JD, Blewett CJ, Young JE, et al. Pyloric
16. Muller JM, Erasmi H, Stelzner M, et al. Surgical ther- drainage (pyloroplasty) or no drainage in gastric
apy of oesophageal carcinoma. Br J Surg 1990;77: reconstruction after esophagectomy: a meta-
845–57. analysis of randomized controlled trials. Dig Surg
17. Stelzner F, Kunath U. Results of esophago-intestinal 2002;19:160–4.
anastomoses and studies of gastric perfusion in the 33. Burt M, Scott A, Williard WC, et al. Erythromycin
stomach mobilized for that purpose. Chirurg 1977; stimulates gastric emptying after esophagectomy
48:651–6 [in German]. with gastric replacement: a randomized clinical trial.
18. Beitler AL, Urschel JD. Comparison of stapled and J Thorac Cardiovasc Surg 1996;111:649–54.
hand-sewn esophagogastric anastomoses. Am J 34. Bagheri R, Fattahi SH, Haghi SZ, et al. Botulinum
Surg 1998;175:337–40. toxin for prevention of delayed gastric emptying af-
19. Kim RH, Takabe K. Methods of esophagogastric ter esophagectomy. Asian Cardiovasc Thorac Ann
anastomoses following esophagectomy for cancer: 2013;21:689–92.
a systematic review. J Surg Oncol 2010;101:527–33. 35. Cerfolio RJ, Bryant AS, Canon CL, et al. Is botulinum
20. Gockel I, Gonner U, Domeyer M, et al. Long-term toxin injection of the pylorus during Ivor Lewis [cor-
survivors of esophageal cancer: disease-specific rected] esophagogastrectomy the optimal drainage
quality of life, general health and complications. strategy? J Thorac Cardiovasc Surg 2009;137:
J Surg Oncol 2010;102:516–22. 565–72.
21. Sutcliffe RP, Forshaw MJ, Tandon R, et al. Anasto- 36. Eldaif SM, Lee R, Adams KN, et al. Intrapyloric
motic strictures and delayed gastric emptying after botulinum injection increases postoperative esopha-
esophagectomy: incidence, risk factors and man- gectomy complications. Ann Thorac Surg 2014;97:
agement. Dis Esophagus 2008;21:712–7. 1959–64.
22. Fok M, Cheng SW, Wong J. Pyloroplasty versus no 37. Salameh JR, Aru GM, Bolton W, et al. Electrostimula-
drainage in gastric replacement of the esophagus. tion for intractable delayed emptying of intrathoracic
Am J Surg 1991;162:447–52. stomach after esophagectomy. Ann Thorac Surg
23. Mannell A, McKnight A, Esser JD. Role of pyloro- 2008;85:1417–9.
plasty in the retrosternal stomach: results of a pro- 38. Wormuth JK, Heitmiller RF. Esophageal conduit ne-
spective, randomized, controlled trial. Br J Surg crosis. Thorac Surg Clin 2006;16:11–22.
1990;77:57–9. 39. Geer RJ, Richards WO, O’Dorisio TM, et al. Efficacy
24. Collard JM, Otte JB, Reynaert M, et al. Quality of life of octreotide acetate in treatment of severe postgas-
three years or more after esophagectomy for cancer. trectomy dumping syndrome. Ann Surg 1990;212:
J Thorac Cardiovasc Surg 1992;104:391–4. 678–87.
25. Hinder RA, Esser J, DeMeester TR. Management of 40. McLarty AJ, Deschamps C, Trastek VF, et al. Esoph-
gastric emptying disorders following the Roux-en-Y ageal resection for cancer of the esophagus: long-
procedure. Surgery 1988;104:765–72. term function and quality of life. Ann Thorac Surg
26. Kraft R, Fry W, Deweese M. Postvagotomy gastric 1997;63:1568–72.
atony. Arch Surg 1964;88:865–74. 41. Orringer MB, Stirling MC. Esophageal resection
27. Miedema BW, Sarr MG, van Heerden JA, et al. Com- for achalasia: indications and results. Ann Thorac
plications following pancreaticoduodenectomy. Cur- Surg 1989;47:340–5.
rent management. Arch Surg 1992;127:945–9. 42. Wang LS, Huang MH, Huang BS, et al. Gastric sub-
28. Gawad KA, Hosch SB, Bumann D, et al. How impor- stitution for resectable carcinoma of the esophagus:
tant is the route of reconstruction after esophagec- an analysis of 368 cases. Ann Thorac Surg 1992;53:
tomy: a prospective randomized study. Am J 289–94.
Gastroenterol 1999;94:1490–6. 43. Woodward ER, Bushkin FL. The early postprandial
29. Imada T, Ozawa Y, Minamide J, et al. Gastric dumping syndrome: prevention and treatment. Ma-
emptying after gastric interposition for esophageal jor Probl Clin Surg 1976;20:14–27.
482 Mohiuddin & Low
44. Woodward ER, Neustein CL. The late postprandial 61. Ramage JI Jr, Rumalla A, Baron TH, et al.
dumping syndrome. Major Probl Clin Surg 1976; A prospective, randomized, double-blind, placebo-
20:28–33. controlled trial of endoscopic steroid injection ther-
45. Andreasen JJ, Orskov C, Holst JJ. Secretion of apy for recalcitrant esophageal peptic strictures.
glucagon-like peptide-1 and reactive hypoglycemia Am J Gastroenterol 2005;100:2419–25.
after partial gastrectomy. Digestion 1994;55:221–8. 62. Zein NN, Greseth JM, Perrault J. Endoscopic intrale-
46. Shultz KT, Neelon FA, Nilsen LB, et al. Mechanism of sional steroid injections in the management of re-
postgastrectomy hypoglycemia. Arch Intern Med fractory esophageal strictures. Gastrointest Endosc
1971;128:240–6. 1995;41:596–8.
47. Gitzelmann R, Hirsig J. Infant dumping syndrome: 63. Ramboer C, Verhamme M, Dhondt E, et al. Endo-
reversal of symptoms by feeding uncooked starch. scopic treatment of stenosis in recurrent Crohn’s
Eur J Pediatr 1986;145:504–6. disease with balloon dilation combined with local
48. Harju E, Larmi TK. Efficacy of guar gum in prevent- corticosteroid injection. Gastrointest Endosc 1995;
ing the dumping syndrome. JPEN J Parenter Enteral 42:252–5.
Nutr 1983;7:470–2. 64. Miyashita M, Onda M, Okawa K, et al. Endoscopic
49. Kneepkens CM, Fernandes J, Vonk RJ. Dumping dexamethasone injection following balloon dilatation
syndrome in children. Diagnosis and effect of gluco- of anastomotic stricture after esophagogastrostomy.
mannan on glucose tolerance and absorption. Acta Am J Surg 1997;174:442–4.
Paediatr Scand 1988;77:279–86. 65. Rupp T, Earle D, Ikenberry S, et al. Randomized trial
50. Chandos B. Dumping syndrome and the regulation of savary dilation with/without intralesional steroids
of peptide YY with verapamil. Am J Gastroenterol for benign gastroesophageal reflux strictures. Gas-
1992;87:1530–1. trointest Endosc 1995;41:357.
51. Christoffersson E, Wallensten S. Drug therapy in the 66. Gandhi RP, Cooper A, Barlow BA. Successful man-
dumping syndrome. Nord Med 1971;86:1589–90 agement of esophageal strictures without resection
[in Norwegian]. or replacement. J Pediatr Surg 1989;24:745–9.
52. Shibata C, Funayama Y, Fukushima K, et al. Effect of 67. Simmons DT, Baron TH. Electroincision of refractory
steroid therapy for late dumping syndrome after total esophagogastric anastomotic strictures. Dis Esoph-
gastrectomy: report of a case. Dig Dis Sci 2004;49: agus 2006;19:410–4.
802–4. 68. Park JY, Song HY, Kim JH, et al. Benign anastomotic
53. Sigstad H. Effect of tolbutamide on the dumping strictures after esophagectomy: long-term effective-
syndrome. Scand J Gastroenterol 1969;4:227–31. ness of balloon dilation and factors affecting recur-
54. Dewar L, Gelfand G, Finley RJ, et al. Factors rence in 155 patients. AJR Am J Roentgenol 2012;
affecting cervical anastomotic leak and stricture for- 198:1208–13.
mation following esophagogastrectomy and gastric 69. Urschel JD, Urschel DM, Miller JD, et al. A meta-
tube interposition. Am J Surg 1992;163:484–9. analysis of randomized controlled trials of route of
55. Law S, Fok M, Chu KM, et al. Comparison of hand- reconstruction after esophagectomy for cancer. Am
sewn and stapled esophagogastric anastomosis after J Surg 2001;182:470–5.
esophageal resection for cancer: a prospective ran- 70. Sharma P, Kozarek R. Role of esophageal stents in
domized controlled trial. Ann Surg 1997;226:169–73. benign and malignant diseases. Am J Gastroenterol
56. Orringer MB, Marshall B, Iannettoni MD. Eliminating 2010;105:258–73.
the cervical esophagogastric anastomotic leak with 71. Barthel JS, Kelley ST, Klapman JB. Management of
a side-to-side stapled anastomosis. J Thorac Cardi- persistent gastroesophageal anastomotic strictures
ovasc Surg 2000;119:277–88. with removable self-expandable polyester silicon-
57. Bailey SH, Bull DA, Harpole DH, et al. Outcomes af- covered (Polyflex) stents: an alternative to serial
ter esophagectomy: a ten-year prospective cohort. dilation. Gastrointest Endosc 2008;67:546–52.
Ann Thorac Surg 2003;75:217–22. 72. Evrard S, Le MO, Lazaraki G, et al. Self-expanding
58. Zieren HU, Muller JM, Pichlmaier H. Prospective plastic stents for benign esophageal lesions. Gas-
randomized study of one- or two-layer anastomosis trointest Endosc 2004;60:894–900.
following oesophageal resection and cervical oeso- 73. Fukumoto R, Orlina J, McGinty J, et al. Use of Poly-
phagogastrostomy. Br J Surg 1993;80:608–11. flex stents in treatment of acute esophageal and
59. Trentino P, Pompeo E, Nofroni I, et al. Predictive gastric leaks after bariatric surgery. Surg Obes Relat
value of early postoperative esophagoscopy for Dis 2007;3:68–71.
occurrence of benign stenosis after cervical esoph- 74. Garcia-Cano J. Dilation of benign strictures in the
agogastrostomy. Endoscopy 1997;29:840–4. esophagus and colon with the polyflex stent: a
60. Kochhar R, Makharia GK. Usefulness of intralesional case series study. Dig Dis Sci 2008;53:341–6.
triamcinolone in treatment of benign esophageal 75. Holm AN, de la Mora Levy JG, Gostout CJ, et al.
strictures. Gastrointest Endosc 2002;56:829–34. Self-expanding plastic stents in treatment of benign
Functional Disorder, Esophagectomy 483
esophageal conditions. Gastrointest Endosc 2008; strictures: less cost and more efficiency. Ann Thorac
67:20–5. Surg 2008;86:436–40.
76. Papachristou GI, Baron TH. Use of stents in benign 80. Aly A, Jamieson GG. Reflux after oesophagectomy.
and malignant esophageal disease. Rev Gastroen- Br J Surg 2004;91:137–41.
terol Disord 2007;7:74–88. 81. Gutschow C, Collard JM, Romagnoli R, et al.
Denervated stomach as an esophageal substitute
77. Pennathur A, Chang AC, McGrath KM, et al. Polyflex
recovers intraluminal acidity with time. Ann Surg
expandable stents in the treatment of esophageal
2001;233:509–14.
disease: initial experience. Ann Thorac Surg 2008;
82. Borst HG, Dragojevic D, Stegmann T, et al. Anasto-
85:1968–72.
motic leakage, stenosis, and reflux after esophageal
78. Wong RF, Adler DG, Hilden K, et al. Retrievable replacement. World J Surg 1978;2:861–4.
esophageal stents for benign indications. Dig Dis 83. McKeown KC. Total oesophagectomy—three-
Sci 2008;53:322–9. staged resection. In: Jamieson G, editor. Surgery
79. Martin RC, Woodall C, Duvall R, et al. The use of of the oesophagus. Edinburgh (United Kingdom):
self-expanding silicone stents in esophagectomy Churchill Livingstone; 1988. p. 677–85.
C o m p l i c a t i o n s Fo l l o w i n g
Surgery for
G a s t ro e s o p h a g e a l R e f l u x
D i s e a s e an d A c h a l a s i a
Samad Hashimi, MD, Ross M. Bremner, MD, PhD*
KEYWORDS
Achalasia Antireflux surgery Complications Nissen fundoplication Toupet fundoplication
Treatment of complications
KEY POINTS
A thorough preoperative workup is essential for good outcomes in surgery for achalasia or reflux
disease.
An understanding of the evolution and advances of laparoscopic surgery paired with careful surgi-
cal technique improve patient satisfaction and outcomes.
Although unusual, complications do occur. The surgeon must be adept at managing these
problems.
Reoperations for functional esophageal disorders are far more complicated than first-time opera-
tions and should be performed in high-volume centers that have extensive experience performing
reoperations.
Disclosure Statement: Neither the authors nor their institution have any personal, financial, or institutional in-
terest in any of the materials or devices described in this article.
Department of Thoracic Disease and Transplantation, Norton Thoracic Institute, St. Joseph’s Hospital and Med-
ical Center, 500 W. Thomas Road, Suite 500, Phoenix, AZ 85013, USA
* Corresponding author.
E-mail address: Ross.Bremner@dignityhealth.org
treatment of patients with achalasia, hiatal hernia, to the hernia itself, including anemia, chest pain or
and GERD has evolved into what should now be discomfort, early satiety, and bloating. Although
recognized as highly successful, cost-effective surgery for these patients is often complicated, it
management for these diseases. As reported by can still usually be performed laparoscopically.
Stefanidis and colleagues,5 the results of at least Specific techniques that should be considered for
7 randomized controlled trials (RCTs) have been these cases are discussed below.
published that show antireflux surgery to be supe-
rior to medical management for alleviating symp- COMPLICATIONS OF ANTIREFLUX SURGERY
toms of GERD. As with any surgical intervention,
complications do occur. Some are unpredictable; Major complications associated with antireflux
however, surgeons who pay close attention to a surgery are uncommon, and perioperative mortal-
patient’s initial symptoms, possess a clear under- ity rates are very low. In 2006, Dominitz and col-
standing of the underlying pathophysiology of that leagues8 reported a perioperative mortality rate
particular patient, and apply modern laparoscopic of 0.8% after antireflux surgery, but a more recent
techniques will reliably achieve good outcomes. study using the Nationwide Inpatient Sample of
Complications occur at least twice as frequently more than 15,000 patients showed a much lower
in reoperations.6 Reoperations should therefore mortality rate of 0.08%.9 Moreover, this report
be performed at centers of excellence by experi- also demonstrated that, despite the progressive
enced surgeons.7 age increase in patients undergoing antireflux
Unfortunately, the esophagus is a relatively surgery and their larger number of comorbidities,
unforgiving organ and the best time to repair a the operative mortality of antireflux surgery has
functional disorder is at the time of the first sur- decreased by 50% over the past decade.9
gery. Bearing this in mind, we cannot emphasize
enough that great care must be exercised before Intraoperative Complications
functional surgery is offered to a patient who pre-
sents with foregut symptoms. Aspiration during intubation
Patients undergoing antireflux surgery usually have
a hiatal hernia and the lower esophageal sphincter
ANTIREFLUX SURGERY (LES) is frequently defective. Great care should be
Preoperative Examination exercised when these patients are intubated for an
A considerable amount of research has been pub- operation, as they are at very high risk of aspiration.
lished regarding the necessary tests that should Rapid-sequence induction with cricoid pressure
be performed before antireflux surgery is consid- and inclining the torso 45 are useful techniques to
ered. At Norton Thoracic Institute in Phoenix, AZ, avoid aspiration. If aspiration is suspected, a bron-
we routinely perform endoscopy, manometry, choscopy should be performed immediately to
and barium esophagograms, gleaning comple- evaluate the extent of pulmonary soilage and to clear
mentary information from each study. This helps the airways. If gross soilage is suspected, it may be
us to understand the whole picture, to better prudent to cancel the procedure and reschedule it
comprehend the patient’s symptoms, and to for a later time, after the patient has had an opportu-
select the best course of treatment. Impedance nity to recover from the aspiration event.
and pH studies should be carried out when signif- Viscus injury from trocar insertion
icant diagnostic uncertainty exists or when a Intestinal injuries usually can be repaired laparos-
diagnosis of GERD cannot be made based on copically. Trocar injuries are now quite infrequent
endoscopy. The pH studies are useful to rule out as a result of increased surgeon experience and
reflux as a potential source of the problem in pa- blunt-tipped trocars, some of which offer camera
tients with complex symptoms. visualization of the layers of the abdominal wall.
Proton pump inhibitors (PPIs) can alleviate Once the camera port is inserted, all other port
symptoms for most patients with mild GERD or a insertions should be performed under camera
minimal hiatal hernia. The widespread use of PPIs visualization. Some surgeons still prefer the use
has resulted in a shift in the presenting symptoms of a traditional spring-loaded Veress needle. In
of patients referred for surgery over the past 2 de- all cases, it is critical that the surgeon pay close
cades, as PPIs have been shown to control heart- attention if the patient has had previous abdominal
burn symptoms relatively well, even for extended surgery, as intra-abdominal adhesions may exist.
periods of time. Patients are now more commonly
referred for surgical correction of a large paraeso- Complications after retractor placement
phageal hernia or for an intrathoracic stomach. Retractors are usually needed to retract the liver,
These patients frequently have symptoms related but injury to the liver may result from their use.10
Gastroesophageal Reflux Disease and Achalasia 487
With modern increased rates of obesity, a large, Perforation of the esophagus by the bougie
fatty liver may impede access to the hiatus. To that is commonly used during fundoplication pro-
improve hiatal access, the diaphragmatic liga- cedures also has been described; in one study,
ments that hold the left lobe of the liver over the hi- the bougie was associated with esophageal injury
atus may need to be taken down. Small tears and in 1.2% of cases.13 Injury can occur anywhere
injury to the liver caused by instrumentation can along the length of the esophagus and should
usually be controlled with electrocautery or hemo- be repaired immediately, if possible. We use
static agents and the application of pressure. Car- a bougie only in select cases, and it is usually
diac tamponade after a retractor-caused injury unnecessary for Toupet or partial fundoplica-
also has been described.11 tions. However, in the case of a Collis lengthening
procedure, a bougie is necessary to prevent
Capnothorax overnarrowing of the gastric tube. A Collis opera-
One or both pleural spaces are frequently entered tion also has been associated with leaks from the
during dissection in the mediastinum. This is more staple line, but this is an unusual complication
common with reoperations or with large hiatal her- occurring up to 2.7% of the time in experienced
nias, and it occurs while the surgeon is resecting hands.14
the sac of the hernia. The capnothorax may cause
a mild hemodynamic effect intraoperatively, as Intra-abdominal catastrophe
intra-abdominal pressure is transmitted into the A severe esophageal tear or a delayed diagnosis of
chest. Fluid administration by the anesthesiologist an esophageal leak causing sepsis and a hostile
can improve venous return and stabilize hemo- inflammatory phlegmon are complex problems,
dynamics. Occasionally, intraperitoneal pressure and may be impossible to repair acutely. Adequate
needs to be decreased. Once the pneumoperito- drainage and stents have been used successfully
neum is released at the end of the procedure, the to treat these complications, but an esophageal
anesthesiologist should give several large positive diversion may be necessary. If the esophagus is
pressure breaths to reexpand the lung before extu- to be stapled at the gastroesophageal (GE) junc-
bation. Although a significant capnothorax may be tion or just above it, it is useful to have a drain in
evident on the postoperative chest radiograph, it the lumen of the distal esophagus (the drain exits
rarely requires any intervention other than oxygen through an abdominal site), as this can help drain
administration if the patient is otherwise stable. salivary secretions and can aid in later reoperation
Carbon dioxide dissolves very rapidly, and the cap- for esophageal reconstruction.
nothorax usually will resolve within hours. Gastric necrosis necessitating gastric resection
has been described in cases of giant hiatal her-
Esophageal perforation nias.15 Gastric necrosis is thought to result from
Esophageal perforation occurs occasionally, but devascularization of the proximal stomach from
it is more commonly associated with dissection injury to the left gastric artery or from takedown
of giant hiatal hernias than with smaller, simple of the short gastric vessels and excessive skeleto-
hernias. Traction injury to the esophagus also nization of the greater curvature, especially if some
can occur when the esophagus is directly re- of these vessels are taken down before a giant her-
tracted or is retracted via a Penrose drain, which nia is reduced. At Norton Thoracic Institute, we
encircles the esophagus. We suggest always hav- have treated several patients who were referred
ing a nasogastric tube or an endoscope in the to us after gastric necrosis. Adequate debridement
esophagus during retraction. We routinely use an and drainage, maintenance of nutritional support,
intraoperative endoscope, as this facilitates identi- and a delayed reconstruction months later is the
fication of the esophagus (especially in cases of best approach. A Roux-en-Y esophagojejunos-
reoperation). The endoscope allows visualization tomy is a sound surgical option that reconstructs
of the esophagogastric mucosal integrity when the alimentary tract while obviating any future GE
injury is suspected. Intraoperative repair of a small reflux, but a more extensive operation using a co-
esophageal laceration usually is successful, but an lon interposition may be required.
overlooked perforation can lead to devastating
mediastinitis and sepsis.12 Sutures through the Splenic injury
esophagus anchoring the Nissen also can be In the former era of open surgery, between 1% and
passed too deeply, resulting in a small mucosal 20% of antireflux procedures were complicated
tear that can leak later. The esophageal lumen is by the need for splenectomy.16,17 In the laparo-
home to multiple bacteria, including oral flora scopic era, however, this number has fallen to
and anaerobes, and mediastinitis and peritonitis less than 1%,15 possibly as a result of the use of
can be severe. improved thermocoagulation devices such as the
488 Hashimi & Bremner
Harmonic scalpel (Ethicon, Cincinnati, OH) and the potential sites of bleeding during an antireflux
decreased incidence of traction injury. Occasion- operation, as summarized in Box 1.
ally a portion of the spleen may infarct after take-
down of the short gastric vessels, and this Early Postoperative Complications
infarction may be noted intraoperatively or inciden-
tally on a postoperative computed tomography Dysphagia
(CT) scan, but it is usually of no clinical conse- Dysphagia in the early postoperative period is
quence. Bleeding from the short gastric vessels common and is usually due to edema of the fundo-
or from a minor tear of the capsule can still occur in- plication, but in some cases it indicates a narrow
traoperatively despite the use of thermocoagula- fundoplication or a tight crural closure. Patients
tion devices, but usually can be controlled by can usually easily swallow liquids within a day of
applying gentle pressure with an intra-abdominal surgery and generally remain on a liquid diet for
sponge and the use of endoclips. a week or 2 after surgery until the edema subsides.
Mild dysphagia usually resolves a few weeks after
Bleeding surgery, so the patient and surgeon alike are
Although it is uncommon, severe intraoperative urged to be patient. Mild dysphagia also may be
bleeding does occur occasionally, and it is more related to poor perioperative motility. A patient
likely to occur in reoperation than in initial surgery. with known poor peristalsis preoperatively is
Bleeding also can occur in the immediate postop- more likely to be patient with “expected” mild
erative period, and postoperative hypotension is a dysphagia postoperatively.
red flag, as a significant amount of blood can be If the patient has significant difficulty swallow-
hidden in the abdomen, especially in obese pa- ing liquids, barium studies may be warranted
tients. Postoperative bleeding may necessitate a and the surgeon may consider endoscopy. If
return to the operating suite, and an open opera- there has been a technical error, an early opera-
tion may be required to control this bleeding. tion to remove a crural stitch or loosen the wrap
Fig. 1 shows the anatomy of the hiatus and is far easier than a delayed operation in a
Fig. 1. Anatomy of the hiatus and sites of potential hemorrhage that can occur during antireflux surgery.
(Courtesy of Norton Thoracic Institute, Phoenix, AZ.)
Gastroesophageal Reflux Disease and Achalasia 489
of DVT and PE. Ambulation on the day of surgery habit of swallowing sometimes continues, trap-
also is vital. ping small amounts of swallowed air in the stom-
ach, exacerbating bloating. Carbonated
Later Postoperative Complications beverages compound the problem and should
Dysphagia be avoided. Patients may report an increase in flat-
As seen in Box 2, there are many possible ulence; for these patients, simethicone is available
causes of persistent or delayed-onset dysphagia. over the counter and can be useful, but the symp-
Although it is uncommon, dysphagia can signifi- tom usually resolves over time. Gas bloat, espe-
cantly inhibit a patient’s lifestyle. It is important to cially troublesome for patients with concomitant
note that antireflux surgery is usually associated irritable bowel syndrome, is generally less com-
with an improvement in preoperative dysphagia, mon after a partial fundoplication, because some
but surgery also can trigger dysphagia as a new belching may be possible.25
symptom for a patient who has undergone fundo- Vagal damage and delayed gastric emptying are
plication. Reported incidences vary greatly in the possible complications of difficult surgical proce-
literature, but larger studies have shown the inci- dures, reoperations, or reduction of large paraeso-
dence of mild dysphagia to be less than 20% phageal hernias. A gastric emptying study can be
1 year postoperatively, with only 5% to 8% of useful, but a gastric bezoar or retained gastric
patients suffering from long-term dysphagia.25 food seen on endoscopy after an overnight fast
Postoperative dysphagia is less common after a usually indicates poor emptying. Prokinetic agents
partial fundoplication than after a 360 complete can be used, but metoclopramide should not be
fundoplication.25 Some series report up to 6% of used for extended lengths of time, and erythro-
patients requiring esophageal dilation to treat mycin can have other intestinal side effects. Pyloric
dysphagia, and this is usually successful.26 Reop- dilation or endoscopic injection of botulinum toxin A
eration may be necessary to modify a long or too- can help alleviate symptoms.28 Patients with a his-
tight fundoplication. A recent article noted a trend tory of major depression have been shown to have a
toward greater dysphagia in patients with a longer higher incidence of this symptom postoperatively.
Nissen (3 cm) than in patients with a shorter Nissen Toupet fundoplication is advised in these cases.29
(1.5 cm).26 Significant dysphagia associated with
reherniation or with a low wrap or “slipped” Nissen Recurrence of hiatal hernia
may require revision surgery. Perhaps the most frustrating long-term complica-
tion for surgeons and patients alike is the recurrence
Gas bloat of a hiatal hernia, reported in up to 40% of patients
Patients with GERD are habitual swallowers, and undergoing repair of large paraesophageal her-
some patients swallow twice as frequently as pa- nias.30 Possible causes of recurrence are a fibrosed,
tients without reflux disorders.27 Fundoplication shortened esophagus; weak diaphragmatic tissue;
limits belching, yet the patient’s subconscious very large defects that require tension for closure;
or postoperative stressors on the repair, such as
trauma, vomiting, chronic cough, or obesity. Some
Box 2 investigators tout the liberal use of the Collis gastro-
Causes of persistent dysphagia plasty; however, even in some of the largest re-
ported series, recurrence rates remain the same
Tight crural closure
between Collis gastroplasty and standard fundopli-
Long or narrow/tight fundoplication cation.14,31 The use of a patch appears to help, but
Problems with a Collis (if performed) the investigators of one study of an absorbable bio-
Fibrotic reaction or erosion of pledgets or mesh logic mesh showed that although early recurrence
decreased,32 long-term recurrence did not.32,33 Per-
Persistent or new-onset esophageal stricture
manent mesh placement has been associated with
(reflux- or pill-induced)
some catastrophic complications, such as erosion
Underlying motility disorder (especially into the esophagus34; therefore, its use should be
hypomotility) avoided. Our own practice has been to use an
Concomitant cricopharyngeal disorder (eg, Zen- absorbable onlay patch of a biocompatible synthetic
ker diverticulum) tissue scaffold (Gore Bio-A Tissue Reinforcement;
Recurrent hiatal hernia (especially paraesopha- W. L. Gore and Associates, Flagstaff, AZ) to
geal component) encourage ingrowth of collagen I. The early results
“Slipped” Nissen (a Nissen riding on the top of of our cohort of more than 800 patients are encour-
the stomach) aging, but long-term follow-up results are not yet
available.
Gastroesophageal Reflux Disease and Achalasia 491
Recurrences are possible, and this is not surpris- Recurrence of reflux and regurgitation
ing when one considers all of the forces on the dia- It is possible for a fundoplication to unravel and
phragm at the hiatus, especially when the hiatus has cause a recurrence of GERD and typical heart-
been weakened by a large hernia. Chronic cough- burn. Patients undergoing a partial fundoplication
ing, heaving, or retching; abdominal trauma; and may experience a progressive worsening of heart-
weight gain all contribute to increased long-term burn, as the fundoplication may loosen over time
transdiaphragmatic pressure gradients and subse- (Fig. 2A). This loosening is more common after
quent reherniation. Reoperation should be consid- an anterior rather than a posterior fundoplication,
ered only for symptomatic patients: it is critical to and the anterior fundoplication should be avoided
be mindful of the difference between radiographic for patients with reflux disease.37 In the absence of
and symptomatic recurrence, as many of the former recurrent hiatal hernia, it is often sufficient to
remain innocuous and asymptomatic. The overall manage most of these patients with medication.
reoperation rate for hiatal hernias ranges between
2% and 5%.35,36 A small recurrence often can be Mesh or pledget erosion
managed conservatively, as these hernias rarely Endoscopy for dysphagia or chest discomfort after
become large enough to warrant surgery, although a Nissen fundoplication occasionally reveals a
scant data exist to show the progression of these foreign body in the distal esophagus or proximal
small hernias with time. Fig. 2 shows several exam- stomach. This may be a remnant from a previous
ples of failure or loosening of fundoplications and pledget or a previously placed permanent mesh
recurrence of hernias, as seen on endoscopy. (Fig. 3).34 A suture or small piece of pledget can
Fig. 2. (A) A patient with recurrent heartburn. Endoscopic image shows that the previously placed Toupet is
intact but has loosened over time. Retroflex maneuver showing the opening of the fundoplication with easy
eructation during insufflation. (B, C) A fundoplication that has herniated into the chest. (D–G) Endoscopic images
depicting a paraesophageal recurrence of hiatal hernia with partial dehiscence of the diaphragmatic closure.
(Courtesy of Norton Thoracic Institute, Phoenix, AZ.)
492 Hashimi & Bremner
Fig. 3. (A) Foreign material eroding into the esophageal lumen associated with a tight stricture. This patient
required resection of this area with a Roux-en-Y esophagojejunostomy. (B) Endoscopic view showing a perma-
nent mesh that has eroded into the esophageal lumen. (C) The permanent mesh resected during reoperation.
(Courtesy of Norton Thoracic Institute, Phoenix, AZ; and Sumeet Mittal, Omaha, NE; with permission.)
be removed endoscopically, but this finding gener- more anxious about surgery, and the outcomes
ally necessitates reoperation (Fig. 3C). The GE for reoperation are less successful than first-
junction often cannot be salvaged in this situation, time procedures.39 Barium studies, pH testing,
and despite reports of gallant treatment modalities, manometry, and gastric emptying are all tests
such as stenting after removal, resection of either that play a role in the workup of these patients.
the esophagus or the proximal stomach is often We prefer to do an endoscopy as a standalone
required.34 If the esophagus still has some peri- procedure to better understand what procedures
stalsis and if intestinal metaplasia is not present, a have been done previously and to visualize the
distal esophagectomy and proximal gastrectomy anatomy. It can be difficult for the untrained eye
with a Roux-en-Y esophagojejunostomy is a good to detect small paraesophageal recurrences, a
option, as it avoids removal of the entire esophagus low or slipped Nissen (especially in the presence
while obviating any future reflux problems. If the pa- of Barrett esophagus) (Fig. 4), or to evaluate the
tient has an aperistaltic esophagus, intestinal meta- stability of a previously placed fundoplication. A
plasia (certainly with any degree of dysplasia), or clear antegrade view to look for rugal folds above
the mesentery of the small bowel does not reach a fundoplication and a retroflexed view with gastric
to the healthy intrathoracic esophagus, an esopha- air distention are critical, informative maneuvers.
gectomy may be needed. An intrathoracic anasto- The surgeon should also note any foreign material
mosis should be performed in this circumstance or sutures in the gastric lumen. A transoral inci-
because the proximal stomach is usually some- sionless fundoplication has a typical appearance
what devitalized and cervical anastomoses have as shown in Fig. 5.
been associated with a high rate of complications Reoperation involves takedown of the previous
after a previous antireflux procedure.38 fundoplication (with extreme care to avoid vagal
injury if possible), reduction of a hernia, rerepair
of a diaphragmatic dehiscence or an onlay patch
Reoperation
of absorbable mesh, and refundoplication. Reop-
Reoperation deserves special mention, as patients erations have higher conversion rates to open
who undergo reoperation present a special series surgery than first-time surgery, and the operations
of challenges. Not only is their symptomatology last longer.40,41 As mentioned previously, if any
complex; these patients are understandably vagal injury is perceived or if the vagus nerves
Gastroesophageal Reflux Disease and Achalasia 493
Fig. 5. (A–D) Foreign material that may be seen after endoscopic antireflux procedures such as the transoral
incisionless fundoplication procedure. (Courtesy of Norton Thoracic Institute, Phoenix, AZ.)
494 Hashimi & Bremner
the esophagus and 2 to 3 cm onto the stom- be expected in some patients. Fortunately, PPIs
ach.55,56 Failed myotomy is commonly associated are usually very effective and we advise our pa-
with inadequate extension onto the stomach. An tients that they may be needed long-term after
intraoperative endoscope can ensure adequacy myotomy.
of the myotomy and the postoperative mucosal
integrity.47 Inadequate myotomy with persistent The peroral endoscopic myotomy
dysphagia can be managed with postoperative Discussed briefly earlier in this article, the POEM is
pneumatic dilation, but reoperation may be a new, evolving myotomy technique that accesses
warranted. natural orifices to treat achalasia. This technique
involves opening the mucosa endoscopically in
Reflux esophagitis the mid esophagus and using the endoscope
Surgical treatment of achalasia balances relief and a small electrocautery device (microknife) to
of outflow obstruction of the esophagus at the develop a submucosal tunnel to access the thick-
expense of an increased risk of GE reflux. In the ened circular fibers of the spastic LES. These mus-
era before PPIs, there was much debate about cle fibers are then cut endoscopically with the
how long to extend the myotomy to palliate the microknife. After the scope is withdrawn, the
patient without causing severe reflux and perhaps mucosal entry point is closed with endoclips. A
even peptic stricture.11,54 The laparoscopic steep learning curve has been associated with
approach used now generally includes extension the POEM.60
onto the stomach, and the most common long- First described by Inoue and colleagues61 in
term complication after an effective modified 2009, centers in Japan appear to have the greatest
Heller myotomy is GE reflux. Esophageal exposure experience with the POEM.62 There are now
to refluxed gastric juice is worsened by a several centers in the United States that are
depressed clearance mechanism of the aperistal- increasing the experience with the procedure,
tic esophageal body. and although the early results seem promising,
It is now generally accepted that some form of long-term results are unavailable. A potentially
fundoplication is a necessary part of most surgical high incidence of postoperative GERD is a
procedures. In a large meta-analysis, Campos and concern, as no antireflux measure is performed
colleagues57 showed that adding an antireflux pro- as part of the procedure. Proponents of the proce-
cedure to myotomy decreased reflux symptoms dure indicate that preservation of the longitudinal
from 31% to 9%, and objective abnormal pH muscle fibers may provide some form of antireflux
testing dropped from 42% to 15%. A recent review barrier.
of 2 meta-analyses, 3 RCTs, and 3 prospective Familiari and colleagues63 recently published a
trials has shown that some form of antireflux pro- review of 100 Italian patients who underwent the
cedure is necessary to decrease reflux.58 The POEM with a mean follow-up of 11 months. Clin-
optimal fundoplication type, however, is still being ical success was documented in 94.5% of pa-
debated. A complete Nissen fundoplication has tients, and 24-hour pH monitoring documented
been associated with an increased incidence of GERD in 53.4% of patients. However, only a
dysphagia, and it is generally agreed that a partial minority of patients had heartburn (24.3%) or
fundoplication is preferable.58 One advantage of esophagitis (27.4%), and these complications
an anterior Dor fundoplication is that it does not were successfully treated with PPIs. Wang and
require posterior esophageal dissection or take- colleagues64 noted that a modified POEM with
down of the short gastric vessels, and it can cover shorter myotomy under endotracheal anesthesia
the exposed myotomy, sealing any potential de- and CO2 insufflations resulted in a safe procedure
layed leaks. A Toupet fundoplication holds the with excellent short-term efficacy in a series of 46
myotomized muscle edges apart and may provide patients treated for achalasia. A drawback of the
better long-term reflux control. However, a Toupet procedure is that many patients develop a pneu-
fundoplication does require posterior esophageal moperitoneum, and CO2 used to insufflate the
dissection and takedown of the short gastric tunnel appears to be absorbed more rapidly than
vessels. Although Mayo and colleagues’ review58 air. A decompression needle in the abdomen
showed no difference between the Toupet and may be required during the procedure.
the Dor procedures in terms of symptom manage- Although their study was not prospectively ran-
ment, Wei and colleagues59 noted that the Dor domized, Bhayani and colleagues65 compared
fundoplication may be associated with higher POEM to laparoscopic myotomy with partial fun-
rates of postoperative GERD. At the very least, doplication with similar early results. The POEM
a partial fundoplication is necessary after an also has been successful in treating patients with
effective cardiomyotomy, and some reflux should previous failed Heller myotomy.66
496 Hashimi & Bremner
As noted, POEM is still an emerging technique, 10. Pasenau J, Mamazza J, Schlachta CM, et al. Liver
and few centers in the United States have signifi- hematoma after laparoscopic Nissen fundoplication:
cant experience. Little is known about long-term a case report and review of retraction injuries.
complications associated with the procedure, but Surg Laparosc Endosc Percutan Tech 2000;10(3):
short-term results are positive and the procedure 178–81.
seems to be safe when performed by experi- 11. Firoozmand E, Ritter M, Cohen R, et al. Ventricular
enced practitioners. Long-term GERD may be an laceration and cardiac tamponade during laparo-
ongoing problem for some patients. scopic Nissen fundoplication. Surg Laparosc En-
dosc 1996;6(5):394–7.
12. Rantanen TK, Salo JA, Sipponen JT. Fatal and life-
SUMMARY
threatening complications in antireflux surgery: anal-
Surgery for GERD and achalasia, usually per- ysis of 5,502 operations. Br J Surg 1999;86(12):
formed laparoscopically, has excellent outcomes. 1573–7.
Complications are unusual, but the surgeon 13. Patterson EJ, Herron DM, Hansen PD, et al. Effect
should be prepared to manage them when they of an esophageal bougie on the incidence of
occur. Reoperations are sometimes needed, but dysphagia following Nissen fundoplication: a pro-
these difficult procedures should be performed spective, blinded, randomized clinical trial. Arch
at centers with sufficient experience. Surg 2000;135(9):1055–61 [discussion: 1061–2].
14. Nason KS, Luketich JD, Awais O, et al. Quality of life
REFERENCES after Collis gastroplasty for short esophagus in pa-
tients with paraesophageal hernia. Ann Thorac
1. Attwood SE, Lundell L, Ell C, et al. Standardization of Surg 2011;92(5):1854–60 [discussion: 1860–1].
surgical technique in antireflux surgery: the LOTUS 15. Pohl D, Eubanks TR, Omelanczuk PE, et al. Manage-
Trial experience. World J Surg 2008;32(6):995–8. ment and outcome of complications after laparo-
2. Peters MJ, Mukhtar A, Yunus RM, et al. Meta-anal- scopic antireflux operations. Arch Surg 2001;
ysis of randomized clinical trials comparing open 136(4):399–404.
and laparoscopic anti-reflux surgery. Am J Gastro- 16. DeMeester TR, Bonavina L, Albertucci M. Nissen
enterol 2009;104(6):1548–61 [quiz: 1547, 1562]. fundoplication for gastroesophageal reflux disease.
3. Mahon D, Rhodes M, Decadt B, et al. Randomized Evaluation of primary repair in 100 consecutive pa-
clinical trial of laparoscopic Nissen fundoplication tients. Ann Surg 1986;204(1):9–20.
compared with proton-pump inhibitors for treatment 17. Rogers DM, Herrington JL Jr, Morton C. Incidental
of chronic gastro-oesophageal reflux. Br J Surg splenectomy associated with Nissen fundoplication.
2005;92(6):695–9. Ann Surg 1980;191(2):153–6.
4. Mehta S, Bennett J, Mahon D, et al. Prospective trial 18. Iqbal A, Kakarlapudi GV, Awad ZT, et al. Assess-
of laparoscopic Nissen fundoplication versus proton ment of diaphragmatic stressors as risk factors for
pump inhibitor therapy for gastroesophageal reflux symptomatic failure of laparoscopic Nissen fundo-
disease: seven-year follow-up. J Gastrointest Surg plication. J Gastrointest Surg 2006;10(1):12–21.
2006;10(9):1312–6 [discussion: 1316–7]. 19. Oelschlager BK, Yamamoto K, Woltman T, et al.
5. Stefanidis D, Hope WW, Kohn GP, et al. Guidelines Vagotomy during hiatal hernia repair: a benign
for surgical treatment of gastroesophageal reflux esophageal lengthening procedure. J Gastrointest
disease. Surg Endosc 2010;24(11):2647–69. Surg 2008;12(7):1155–62.
6. Wang YR, Dempsey DT, Richter JE. Trends and peri- 20. Lindeboom MY, Ringers J, van Rijn PJ, et al. Gastric
operative outcomes of inpatient antireflux surgery in emptying and vagus nerve function after laparo-
the United States, 1993-2006. Dis Esophagus 2011; scopic partial fundoplication. Ann Surg 2004;
24(4):215–23. 240(5):785–90.
7. Funch-Jensen P, Bendixen A, Iversen MG, et al. 21. Hinder RA, Stein HJ, Bremner CG, et al. Relationship
Complications and frequency of redo antireflux sur- of a satisfactory outcome to normalization of de-
gery in Denmark: a nationwide study, 1997-2005. layed gastric emptying after Nissen fundoplication.
Surg Endosc 2008;22(3):627–30. Ann Surg 1989;210(4):458–64 [discussion: 464–5].
8. Dominitz JA, Dire CA, Billingsley KG, et al. Compli- 22. Klaus A, Hinder RA, DeVault KR, et al. Bowel
cations and antireflux medication use after antireflux dysfunction after laparoscopic antireflux surgery:
surgery. Clin Gastroenterol Hepatol 2006;4(3): incidence, severity, and clinical course. Am J Med
299–305. 2003;114(1):6–9.
9. Funk LM, Kanji A, Scott Melvin W, et al. Elective anti- 23. Hazan TB, Gamarra FN, Stawick L, et al. Nissen fun-
reflux surgery in the US: an analysis of national doplication and gastrointestinal-related complica-
trends in utilization and inpatient outcomes from tions: a guide for the primary care physician.
2005 to 2010. Surg Endosc 2014;28(5):1712–9. Southampt Med J 2009;102(10):1041–5.
Gastroesophageal Reflux Disease and Achalasia 497
24. Jamieson GG, Watson DI, Britten-Jones R, et al. 38. Shen KR, Harrison-Phipps KM, Cassivi SD, et al.
Laparoscopic Nissen fundoplication. Ann Surg Esophagectomy after anti-reflux surgery. J Thorac
1994;220(2):137–45. Cardiovasc Surg 2010;139(4):969–75.
25. Varin O, Velstra B, De Sutter S, et al. Total vs partial 39. Furnee EJ, Draaisma WA, Broeders IA, et al. Surgi-
fundoplication in the treatment of gastroesophageal cal reintervention after failed antireflux surgery: a
reflux disease: a meta-analysis. Arch Surg 2009; systematic review of the literature. J Gastrointest
144(3):273–8. Surg 2009;13(8):1539–49.
26. Mickevicius A, Endzinas Z, Kiudelis M, et al. Influ- 40. Iqbal A, Awad Z, Simkins J, et al. Repair of 104
ence of wrap length on the effectiveness of Nissen failed anti-reflux operations. Ann Surg 2006;244(1):
and Toupet fundoplication: a prospective random- 42–51.
ized study. Surg Endosc 2008;22(10):2269–76. 41. Oelschlager BK, Lal DR, Jensen E, et al. Medium-
27. Bremner RM, Hoeft SF, Costantini M, et al. Pharyn- and long-term outcome of laparoscopic redo fundo-
geal swallowing. The major factor in clearance of plication. Surg Endosc 2006;20(12):1817–23.
esophageal reflux episodes. Ann Surg 1993;218(3): 42. Awais O, Luketich JD, Reddy N, et al. Roux-en-Y
364–9 [discussion: 369–70]. near esophagojejunostomy for failed antireflux
28. Hooft N, Smith M, Huang J, et al. Gastroparesis is operations: outcomes in more than 100 patients.
common after lung transplantation and may be Ann Thorac Surg 2014;98(6):1905–11 [discussion
ameliorated by botulinum toxin-A injection of the 1911–3].
pylorus. J Heart Lung Transplant 2014;33(12): 43. Heller E. Extramukose cardioplastik beim chroni-
1314–6. schen cardiospasmus mit dilatation des oesoph-
29. Kamolz T, Granderath FA, Pointner R. Does major agus. Mitt Grenzeg Med Chir 1913;27:141.
depression in patients with gastroesophageal reflux 44. Parise P, Santi S, Solito B, et al. Laparoscopic Heller
disease affect the outcome of laparoscopic antire- myotomy plus Dor fundoplication in 137 achalasic
flux surgery? Surg Endosc 2003;17(1):55–60. patients: results on symptoms relief and successful
30. Hashemi M, Peters JH, DeMeester TR, et al. Laparo- outcome predictors. Updates Surg 2011;63(1):11–5.
scopic repair of large type III hiatal hernia: objective 45. Bredenoord AJ, Fox M, Kahrilas PJ, et al. Chicago
followup reveals high recurrence rate. J Am Coll classification criteria of esophageal motility disorders
Surg 2000;190(5):553–60 [discussion: 560–1]. defined in high resolution esophageal pressure topog-
31. Luketich JD, Nason KS, Christie NA, et al. Outcomes raphy. Neurogastroenterol Motil 2012;24(Suppl 1):
after a decade of laparoscopic giant paraesopha- 57–65.
geal hernia repair. J Thorac Cardiovasc Surg 2010; 46. Muller M. Impact of high-resolution manometry on
139(2):395–404, 404.e1. achalasia diagnosis and treatment. Ann Gastroen-
32. Oelschlager BK, Pellegrini CA, Hunter J, et al. Bio- terol 2015;28(1):3–9.
logic prosthesis reduces recurrence after laparo- 47. Chuah SK, Chiu CH, Tai WC, et al. Current status in
scopic paraesophageal hernia repair: a multicenter, the treatment options for esophageal achalasia.
prospective, randomized trial. Ann Surg 2006; World J Gastroenterol 2013;19(33):5421–9.
244(4):481–90. 48. Nau P, Rattner D. Laparoscopic Heller myotomy
33. Oelschlager BK, Pellegrini CA, Hunter JG, et al. Bio- as the gold standard for treatment of achalasia.
logic prosthesis to prevent recurrence after laparo- J Gastrointest Surg 2014;18(12):2201–7.
scopic paraesophageal hernia repair: long-term 49. Smith CD, Stival A, Howell DL, et al. Endoscopic
follow-up from a multicenter, prospective, random- therapy for achalasia before Heller myotomy results
ized trial. J Am Coll Surg 2011;213(4):461–8. in worse outcomes than Heller myotomy alone. Ann
34. Stadlhuber RJ, Sherif AE, Mittal SK, et al. Mesh Surg 2006;243(5):579–84 [discussion: 584–6].
complications after prosthetic reinforcement of hiatal 50. Galvani C, Gorodner MV, Moser F, et al. Laparo-
closure: a 28-case series. Surg Endosc 2009;23(6): scopic Heller myotomy for achalasia facilitated
1219–26. by robotic assistance. Surg Endosc 2006;20(7):
35. Luketich JD, Raja S, Fernando HC, et al. Laparo- 1105–12.
scopic repair of giant paraesophageal hernia: 100 51. Shaligram A, Unnirevi J, Simorov A, et al. How does
consecutive cases. Ann Surg 2000;232(4):608–18. the robot affect outcomes? A retrospective review of
36. Pierre AF, Luketich JD, Fernando HC, et al. Results open, laparoscopic, and robotic Heller myotomy for
of laparoscopic repair of giant paraesophageal her- achalasia. Surg Endosc 2012;26(4):1047–50.
nias: 200 consecutive patients. Ann Thorac Surg 52. Tanaka E, Murata H, Minami H, et al. Anesthetic
2002;74(6):1909–15 [discussion: 1915–6]. management of peroral endoscopic myotomy for
37. Engstrom C, Lonroth H, Mardani J, et al. An anterior esophageal achalasia: a retrospective case series.
or posterior approach to partial fundoplication? J Anesth 2014;28(3):456–9.
Long-term results of a randomized trial. World J 53. Luckey AE 3rd, DeMeester SR. Complications of
Surg 2007;31(6):1221–5 [discussion: 1226–7]. achalasia surgery. Thorac Surg Clin 2006;16(1):95–8.
498 Hashimi & Bremner
54. Frobese AS, Hawthorne HR, Nemir P Jr. The surgical 61. Inoue H, Minami H, Satodate H, et al. First clinical
management of achalasia of the esophagus. Ann experience of submucosal endoscopic esophageal
Surg 1956;144(4):653–69. myotomy for esophageal achalasia with no skin inci-
55. Chen Z, Bessell JR, Chew A, et al. Laparoscopic sion. Gastrointest Endosc 2009;69(5):AB122.
cardiomyotomy for achalasia: clinical outcomes 62. Minami H, Inoue H, Haji A, et al. Per-oral endoscopic
beyond 5 years. J Gastrointest Surg 2010;14(4): myotomy: emerging indications and evolving tech-
594–600. niques. Dig Endosc 2015;27(2):175–81.
56. Oelschlager BK, Chang L, Pellegrini CA. Improved 63. Familiari P, Gigante G, Marchese M, et al. Peroral
outcome after extended gastric myotomy for acha- endoscopic myotomy for esophageal achalasia: out-
lasia. Arch Surg 2003;138(5):490–5 [discussion: comes of the first 100 patients with short-term follow-
495–7]. up. Ann Surg 2014. [Epub ahead of print].
57. Campos GM, Vittinghoff E, Rabl C, et al. Endoscopic 64. Wang J, Tan N, Xiao Y, et al. Safety and efficacy
and surgical treatments for achalasia: a systematic of the modified peroral endoscopic myotomy with
review and meta-analysis. Ann Surg 2009;249(1): shorter myotomy for achalasia patients: a prospec-
45–57. tive study. Dis Esophagus 2014. [Epub ahead of
58. Mayo D, Griffiths EA, Khan OA, et al. Does the addi- print].
tion of a fundoplication improve outcomes for patients 65. Bhayani NH, Kurian AA, Dunst CM, et al.
undergoing laparoscopic Heller’s cardiomyotomy? A comparative study on comprehensive, objective
Int J Surg 2012;10(6):301–4. outcomes of laparoscopic Heller myotomy with
59. Wei MT, He YZ, Deng XB, et al. Is Dor fundoplication per-oral endoscopic myotomy (POEM) for achalasia.
optimum after laparoscopic Heller myotomy for Ann Surg 2014;259(6):1098–103.
achalasia? A meta-analysis. World J Gastroenterol 66. Onimaru M, Inoue H, Ikeda H, et al. Peroral endo-
2013;19(43):7804–12. scopic myotomy is a viable option for failed surgical
60. Kurian AA, Dunst CM, Sharata A, et al. Peroral endo- esophagocardiomyotomy instead of redo surgical
scopic esophageal myotomy: defining the learning Heller myotomy: a single center prospective study.
curve. Gastrointest Endosc 2013;77(5):719–25. J Am Coll Surg 2013;217(4):598–605.
Prevention and
Management of Nerve
Injuries in Thoracic Surgery
Hugh G. Auchincloss, MD, MPHa, Dean M. Donahue, MDb,*
KEYWORDS
Peripheral nerves Anatomy Physiology Thoracic surgery Complications
Peripheral nerve injury
KEY POINTS
Understanding of the basic anatomy and physiology of the peripheral nerves of the chest is impor-
tant for thoracic surgeons.
Peripheral nerves may be injured during an operation by partial or complete transection, thermal
insult, crush, stretch, or exposure to a toxic environment.
Progress has been made in the field of nerve repair and regeneration, but there is no treatment that
is as reliable as avoiding these injuries.
a
Department of Thoracic Surgery, Massachusetts General Hospital, Blake 1570, Boston, MA 02114, USA;
b
Department of Thoracic Surgery, Harvard Medical School, Massachusetts General Hospital, Blake 1570, 55
Fruit Street, Boston, MA 02114, USA
* Corresponding author.
E-mail address: ddonahue@partners.org
network that supplies individual neurons. The cannot repair itself and undergoes a process
viability of the epineurial layer is a key determinant of Wallerian degeneration, whereby the distal
to how well a nerve may recover after injury, with or part of the axon degenerates and the resultant
without repair. products are scavenged by macrophages.
Neurons can be classified in multiple ways. Neu- Renervation of the target organ can be
rons belonging to the somatic nervous system achieved by collateralization or axonal regen-
provide voluntary skeletal motor control and eration. Collaterization occurs if only some of
precise sensory information, whereas neurons the axons leading to a target organ are
belonging to the autonomic nervous system pro- damaged, allowing the tissue innervated by
vide involuntary motor and sensory information the remaining axons to hypertrophy and pre-
to and from the viscera and smooth muscle. Within serve the function of the organ. If the majority
the autonomic nervous system, a further subdivi- of axons have been destroyed, then regener-
sion exists between sympathetic (“fight or flight”) ation must take place for function to be
and parasympathetic (“rest and digest”) neurons. maintained. This is possible because the con-
These distinctions are of little importance to sur- nective tissue scaffolding of the nerve remains
geons with respect to the handling of individual intact. After a period of Wallerian degenera-
peripheral nerves, because most peripheral nerves tion, axons begin to regrow at rate of 1 to
contain neurons of all different types enclosed 3 mm per day along their endoneurial tubes,
within a single epineurial sheath. Injury to a nerve nourished by the surrounding epineurium. If
is likely to affect all neurons equally. they reach the target organ within 12 to
18 months, renervation can occur, albeit with
somewhat abnormal conduction. Beyond
Classification of Peripheral Nerve Injury
that window, a process of denervation-
A peripheral nerve may be damaged in several related fibrosis makes muscle recovery
ways during the course of an operation, including unlikely; however, some sympathetic and
by partial or complete transection, thermal insult, sensory nerve targets have been shown to
crush, stretch, or exposure to a toxic environment. maintain renervation potential as far out as
The pattern of injury that results from these injuries 3 years.
exists along a spectrum with important implica- Neurotmesis (Sunderland class III/V injury) is
tions for immediate management and prognosis axonal damage along with some degree of
for recovery. The severity of peripheral nerve injury connective tissue damage. If the endoneurium
is determined first by whether or not there is alone is disrupted, then axonal regeneration
axonal damage, and then by the degree to which may be possible. However, if damage extends
there is connective tissue damage. A classification to the epineurium (ie, complete transection of
of peripheral nerve injuries was originally proposed the nerve), surgical repair is required.
by Seddon in 1943, expanded by Sunderland in
the 1950s, and is still widely in use today. Special consideration exists for nerve injuries
caused by thermal insult, local anesthetic toxicity,
Neurapraxia (Sunderland class I injury) refers or exposure to damaging environmental factors.
to temporary interruption of normal nerve con- Thermal injuries are unique in that the degree of
duction without axonal or connective tissue destruction to both the axons and the connective
damage. Neuropraxial injuries are the mildest tissue scaffolding of the nerve may be more severe
form of peripheral nerve trauma. Usually the than is perceived by the surgeon. Total disruption
cause is mild crush, stretch, or thermal injury of the epineurial layer with subsequent fibrosis
to the nerve that in turn causes demyelination. may occur in the absence of obvious physical
The mechanism by which demyelination is signs of damage. For this reason, thermal nerve in-
triggered is not understood fully, but may be juries are often not identified at the time of surgery.
related to transient ischemia or local inflam- If they are, however, the safest course is to resect
matory response. However, because the the affected nerve along with proximal and distal
axon itself remains intact, full recovery of margin and reconstruct the remaining tissue. The
nerve function is expected in days or weeks. severity of local anesthetic toxicity may also be
Axonotmesis (Sunderland class II injury) is difficult to judge in the immediate period. Infiltra-
defined as axonal injury without damage to tion of local anesthetic agent directly into the epi-
the connective tissue scaffolding of the nerve. neurial layer typically results in neuropraxial
Typically, this injury is caused by prolonged injury. However, if high enough concentrations
stretch, crush, or ischemia with resultant are achieved axonal death and even connective
axonal damage. An axon that is damaged tissue damage are possible. Last, experimental
Prevention and Management of Nerve Injuries 511
models suggest that certain environments, partic- assume particular importance when the operation
ularly extreme acidity, can result in severe nerve becomes challenging or stressful; hemostasis
injury. This theoretic risk has led the manufacturers achieved at the expense of damage to an impor-
of certain topical hemostatic agents whose mech- tant peripheral nerve may proved to be a pyrrhic
anism of action produces local acidity to caution victory.
against the use of their products in the area of
important peripheral nerves. However, there are Phrenic Nerve
no data to suggest that this concern is justified,
and in practice topical hemostatic agents are often The phrenic nerve provides motor, sensory, and
used as a way to avoid the need for electrocautery sympathetic input to and from the diaphragm. It
around peripheral nerves. also provides some of the innervation to the medi-
astinal pleura. The phrenic nerve arises from the
ventral rami of the third, fourth, and fifth cervical
Repair of Peripheral Nerves
nerves. It follows a similar course on the right
Peripheral nerves may be repaired in several ways. and left sides. In the neck, the phrenic nerve is
If a nerve is transected and the bridging distance is found on the anterior aspect of the anterior
short, a primary repair is used. For longer defects scalene muscle. It then course inferiorly and
and situations in which tension limits primary passes medial to the insertion of the anterior
repair, an interposition graft using either native scalene onto the first rib. As the nerve exits the
nerve (eg, sural nerve) or synthetic material may thoracic inlet and enters the superior mediastinum,
be used. Peripheral nerve repair is beyond the it is found anterior to the subclavian artery and
scope of practice for thoracic surgeons and posterior to the confluence of the subclavian and
should be done by plastic surgeons trained in internal jugular veins. It then travels on the lateral
microsurgery. Repair is typically done using fine aspect of the superior vena cava (on the right)
nonabsorbable monofilament suture under an and the proximal aortic arch anterior to the first
operating microscope. intercostal vein (on the left) before continuing
along the mediastinal pleura anterior to the pulmo-
NERVE INJURIES IN THORACIC SURGERY nary hilum. It is accompanied at this level by the
pericardiophrenic artery and vein, which are
Despite the progress made in the field of periph- branches of the internal mammary artery and
eral nerve repair and reconstruction, the results vein, respectively. The nerve then inserts on the
remain modest and avoidance of nerve injuries re- central portion the diaphragm. Occasionally, an
mains paramount for the thoracic surgeon. accessory phrenic nerve arises from the cervical
Consideration begins with careful positioning of nerve roots and follow a similar course.
the patient. In the supine position, attention should Injury to the phrenic nerve is of considerable
be paid to padding of the elbows and gentle supi- concern to the thoracic surgeon for 2 reasons.
nation of the forearms and wrists. Upper extrem- First, the location of the nerve in the neck and
ities should not be extended, abducted, or the anterior hilum put it at risk during the majority
externally rotated beyond 90 . In the lateral decu- of thoracic operative procedures in the neck, hem-
bitus position, care is taken to avoid overextension ithorax, or mediastinum. Second, the phrenic
of the upper shoulder and unevenly distributed nerve provides the only source of motor control
pressure on the dependent shoulder. to the diaphragm, meaning that neuropraxial injury
Sophisticated knowledge of the anatomy of the to the nerve places the patient at high risk for post-
peripheral nerves of the thorax along with their operative respiratory complications, and signifi-
common anatomic variants is essential to the cant injury results in unilateral diaphragmatic
safe practice of surgery. Extra care must be taken paralysis. Protection of the phrenic nerve depends
when the patient’s pathology distorts normal anat- on careful identification and sharp dissection (as
omy. Important structures should be identified with anterior scalenectomy during supraclavicular
during dissection and this identification should first rib resection), extreme caution (as with thy-
be explicit, particularly when operating with mectomy, where the nerve frequently appears
trainees whose perception of the situation may around the superior pole of the thymus in a decep-
differ from that of an experienced surgeon. When tively anterior location), or outright avoidance (as
operating in the vicinity of an important peripheral with decortication for empyema, during which
nerve or when the relevant anatomy is ambiguous, stripping of the mediastinal pleura should be un-
sharp dissection or bipolar cautery is favored over dertaken judiciously). The phrenic nerve can also
standard cautery. Care is taken to avoid excessive be injured during cryoablation or pulmonary vein
retraction or handling of nerves. These points isolation for the treatment of atrial arrhythmias,
512 Auchincloss & Donahue
radiofrequency ablation of lung tumors, or by with marginal effect on patient outcomes. Injury to
excessive cooling of the heart during cardiopulmo- both vagi proximal to the cardiac plexus would
nary bypass. theoretically result in unopposed sympathetic
A full-thickness injury to the phrenic nerve innervation to the heart, as seen after cardiac
recognized at the time of surgery should be transplantation; however, this scenario is exceed-
treated with repair or interposition graft. For an ingly unlikely. In practice, most surgeons would
injury too extensive to permit reconstruction, not repair an injured vagus nerve.
some surgeons consider performing plication of
the diaphragm at the same operation. More often,
Recurrent Laryngeal Nerve
however, phrenic nerve injury is unrecognized until
the postoperative period when a chest radiograph The RLN is a branch of the vagus nerve and pro-
demonstrates an elevated hemidiaphragm. In this vides innervation to all the muscles of the ipsilat-
case, there are few immediate surgical options. eral larynx with the exception of the cricothyroid.
The best course is to wait a period of months to Additionally, the nerve carries sensory information
observe whether diaphragmatic function returns. from the larynx and cervical esophagus. On the
After 12 to 18 months, if there has been no change right side, the RLN arises at the level of the subcla-
in function and the patient is symptomatic, dia- vian artery, hooks around the artery posteriorly,
phragmatic plication should be considered. and then courses superiorly in the tracheoesopha-
geal groove. On the left, the RLN arises at the level
of the aortic arch, hooks around the arch posteri-
Vagus Nerve
orly lateral to the ligamentum arteriosum, and
The vagus nerve (cranial nerve X) constitutes the runs superiorly in the tracheoesophageal groove.
parasympathetic innervation of the cervical, Both nerves insert on the inferior lateral larynx at
thoracic, and abdominal viscera. Importantly, it the approximate level of the inferior parathyroid
also carries motor and sensory fibers that inner- gland. However, considerable variability exists.
vate the larynx via the superior and recurrent laryn- The course of the RLN is said to be more consis-
geal nerves (RLNs). The vagus nerve arises in the tent on the left than on the right. An infrequent
brainstem, exits the skull through the jugular fora- anatomic variant is the nonrecurrent laryngeal
men, and enters the neck in the posterior of the ca- nerve. This occurs predominantly on the right in
rotid sheath along with the internal jugular vein. It association with vascular anomalies (ie, right sub-
enters the chest anterior to the subclavian artery clavian arising directly from the aortic arch). In this
and posterior to the innominate vein, in a more case, the nerve travels along either the superior or
medial position than the phrenic nerve. On the inferior thyroid artery.
right the RLN branches off the vagus nerve at the Injury to the RLN can occur in the chest during
level of the subclavian, whereas on the left it arises vascular procedures on the right subclavian artery
at level of the aortic arch. In the superior medias- or aortic arch, during sampling of aortopulmonary
tinum, the vagus nerve runs posteroinferiorly along lymph nodes, or by aggressive biopsy technique
the trachea, lateral to the vena cava (on the right) or use of cautery on the left side of the trachea dur-
and the aortic arch (on the left) before passing pos- ing mediastinoscopy. However, injury to the RLN
terior to the pulmonary hilum. In the upper and is more common in the neck during procedures
middle thorax, the vagus nerve gives rise to involving the thyroid or parathyroid glands, tra-
extensive cardiac, pulmonary, and esophageal chea, or cervical esophagus. Considerable debate
plexuses. Caudally, the left and right vagi pass exists among endocrine surgeons regarding the
through the esophageal hiatus anterior and poste- best way to avoid RLN injury during thyroidectomy
rior to the esophagus, respectively, although the or parathyroidectomy, with some favoring contin-
communication between the 2 nerves is extensive. uous nerve monitoring to facilitate identification
The vagus nerve may be at risk for injury during of the nerve and others arguing that the RLN
extensive posterior hilar dissection or sampling of should be avoided altogether. During tracheal
subcarinal or paratracheal lymph nodes. The left resection and reconstruction, the RLN is avoided
vagus is also in the immediate vicinity during left by keeping the dissection as close to the trachea
upper lobectomy or sampling of aortopulmonary as possible. For procedures on the cervical esoph-
lymph nodes. The latter is significant because agus such as cricopharyngeal myotomy for Zen-
injury to the vagus nerve at this level also compro- ker’s diverticulum, the esophagus is approached
mises the left RLN. Distal to this, however, injury to posterolaterally to reflect the RLN anteriorly and
the vagus nerve (particularly unilateral injury) is avoid injury.
probably of little consequence. Indeed, both vagi Unilateral injury to the RLN results in paralysis of
are routinely sacrificed during esophageal surgery the ipsilateral vocal cord and hoarseness. Bilateral
Prevention and Management of Nerve Injuries 513
RLN injury results in potentially catastrophic with bodies located in the spinal cord, these post-
airway compromise if unrecognized, and should synaptic nerve fibers form splanchnic nerves that
prompt immediate tracheostomy if the injury is terminate in the cardiac, pulmonary, or esopha-
felt to be significant. If injury to the RLN is sus- geal plexus, or rejoin spinal nerves and provide
pected during the course of surgery and circum- motor innervation to the blood vessels, hair folli-
stances permit (eg, total thyroidectomy for small cles, and sweat glands of the body wall and upper
papillary thyroid cancer), surgery on the contralat- extremity. The stellate ganglion is located on the
eral thyroid lobe should be abandoned. anterior surface of the transverse process of the
The routine use of continuous nerve monitoring C7 vertebral body immediately superior to the
during thyroid surgery has led to increased intrao- posterior insertion of the first rib. Extending
perative detection of RLN injuries. The majority of caudally, each successive vertebral body is asso-
these injuries are neuropraxial and do not require ciated with a ganglion of the sympathetic chain
further intervention. For more severe injuries, with connections to the ganglia located above
including those that involve complete transection and below. The sympathetic ganglion immediately
of the nerve, there is controversy regarding the inferior to the stellate ganglion may also communi-
most appropriate management. Classic teaching cate directly with the intercostal nerve above via
held that repair of the nerve was contraindicated the nerve of Kuntz. In the posterior thorax, the
because of the potential for laryngeal synkinesis. sympathetic trunk is found deep to the parietal
Synkinesis is the process of misdirected renerva- pleura with each ganglion located anterior and
tion of the complex musculature of the larynx, superior to the costovertebral junction of its asso-
leading to paradoxic motion and in rare cases ciated rib.
airway compromise. In contrast, spontaneous Temporary or permanent interruption of the stel-
nerve regeneration (which can occur for distal in- late ganglion or sympathetic trunk may be desir-
juries even if the connective tissue layers of the able in certain circumstances. Regional block of
nerve are damaged) was thought to lead to better the stellate ganglion has been used with limited
functional outcome. Newer data suggest that this success for temporary palliation of upper extrem-
is probably not the case, and the current practice ity Raynaud syndrome, complex regional pain dis-
of most surgeons is to perform nerve repair if sig- orders, and refractory angina. Permanent division
nificant injury is recognized. Alternatively, the ansa of the sympathetic trunk at the level of T3 or T4
cervicalis may be located on the lateral aspect of is a well-described treatment for palmar hyperhi-
the carotid sheath and anastomosed to the distal drosis. In these instances, the negative effects of
end of the transected nerve. This does not restore sympathectomy are weighed carefully against
function to the larynx, but does seem to provide potential benefit; the same cannot be said in the
tone to the associated muscles and improve the case of inadvertent injury. The stellate ganglion in
functional outcome of spontaneous regeneration. particular can be damaged during resection of
Management of a unilateral RLN nerve injury de- the first rib for thoracic outlet syndrome if division
tected postoperatively is an extensive topic. There of the posterior rib occurs too medially. It may also
are multiple techniques for restoring maximal be involved directly by a superior sulcus (Pan-
airway patency and phonation, including gelfoam coast) tumor or be damaged during tumor resec-
injection, cord medicalization, and laryngoplasty. tion. The sympathetic chain is more protected in
None of these techniques should be used in the its posterior location in the thorax. However, injury
immediate postoperative period because the po- is possible during pleurectomy for benign or malig-
tential for delayed nerve recovery is substantial. nant disease, or during anterior exposure of the
Current recommendations are that for a stable pa- thoracic spine. Reconstruction of the sympathetic
tient with no significant aspiration risk, no irrevers- chain with sural nerve graft has been described,
ible procedures should be performed in the first 12 but is probably not indicated in most cases.
to 18 months. Once injured, no reconstruction options exist
for the stellate ganglion because it contains
the neuronal cell and innumerable synaptic
Stellate Ganglion/Sympathetic Trunk
connections.
The stellate (or cervicothoracic) ganglion is formed Disruption of the stellate ganglion results in ipsi-
by the fusion of the inferior cervical ganglion and lateral Horner syndrome (a triad of ptosis, meiosis,
superior thoracic ganglion. It is the most superior and facial anhydrosis caused by a lack of sympa-
ganglion in the thoracic sympathetic trunk. The thetic input to the face). Injury to the inferior
sympathetic trunk contains the postsynaptic cell sympathetic trunk produces variable results de-
bodies of the sympathetic autonomic nervous sys- pending on the level at which the injury occurs.
tem. After receiving input from presynaptic nerves Sympathectomy performed at the level of T2 or
514 Auchincloss & Donahue
higher seems to result in some changes in cardiac however, the deficit is occasionally permanent. In-
physiology, namely depression in resting heart traoperatively, the brachial plexus is encountered
rate, shortening of QTc interval, and possibly during procedures involving the thoracic outlet.
blunting of exercise response. Cases of life- This includes first rib resection and scalenectomy
threatening bradycardia have been reported, but and resection of superior sulcus tumors. First rib
for the most part these changes are mild and resection involves careful identification and dissec-
clinically unimportant. Compensatory truncal or tion of the 3 trunks of the brachial plexus with judi-
gustatory sweating (caused by disruption of cious use of bipolar cautery; monopolar cautery
excessive sudomotor function in the ipsilateral up- should never be used. Paralytic agents are not
per extremity) is more apparent and probably used during the procedure so that inadvertent stim-
occurs in more than one-half of patients after ulation of the plexus (or the adjacent phrenic nerve)
sympathectomy for hyperhidrosis. It is not clear is immediately apparent. Special care is taken to
whether these results can be applied to patients avoid excessive retraction on the nerve trunks dur-
with normal preoperative sudomotor function. ing exposure of the first rib. As a general rule, if
exposure is excellent, retraction is too vigorous.
Superior sulcus tumors represent another chal-
Brachial Plexus
lenge for the surgeon because the patient’s tumor
The brachial plexus is the most anatomically com- may invade the plexus directly and the surrounding
plex nerve structure routinely encountered by anatomy may be fibrotic from neoadjuvant chemo-
thoracic surgeons. Typically, the plexus is formed radiation. The surgeon may choose to sacrifice a
by the ventral rami the spinal nerves originating portion of the brachial plexus as part of a thorough
from C5, C6, C7, C8, and T1. Occasionally it is oncologic operation. This possibility should be
formed from the spinal nerves of C4 through C8 discussed with the patient preoperatively.
or C6 through T2, leading to a prefixed or postfixed
brachial plexus, respectively. The 5 nerve roots of SUMMARY
the plexus join to form upper, middle, and lower
trunks, followed by divisions, cords, and ultimately A nerve injury after thoracic surgery often feels like
terminal branches (the musculocutaneous, axil- a technical failure to surgeon and patient alike. For
lary, radial, median, and ulnar nerve). At its origin, the surgeon, there is the knowledge that the rele-
the brachial plexus emerges between the anterior vant anatomy was misidentified or poor surgical
and middle scalene muscles and follows a course technique was used and a preventable error re-
along with the subclavian artery that leads superior sulted; for the patient, there is temporary or lasting
to the first rib, through the axilla, and into the functional deficit that may result in substantial
medial aspect of the upper extremity. Proximal morbidity and decrease quality of life. Progress
branches such as the long thoracic nerve (C5, has been made in the field of nerve repair and
C6, and C7), dorsal scapular nerve (C5), median regeneration; prompt identification and appro-
pectoral nerve (C8 and T1), and thoracodorsal priate management can significantly mitigate the
nerve (C6, C7, and C8) exit the plexus superior to morbidity of nerve injuries. However, the surgeon
the first rib and run along the lateral chest wall. should be mindful that there is no current remedy
The brachial plexus provides essentially the for these injuries that is as reliable as avoidance.
entire motor and sensory innervation to the upper
extremity as well as some of the muscles of the FURTHER READINGS
chest wall. Traumatic or iatrogenic injury to the
main plexus or any of its branches results in defi- Arcasoy SM, Jett JR. Superior pulmonary sulcus tumors
cits that range from minor sensory loss to moder- and Pancoast’s syndrome. N Engl J Med 1997;
ate functional difficulty (eg, scapula alata or 337(19):1370–6.
“winged scapula” from long thoracic nerve injury) Illig K, Thompson R, Frieschlag J, et al. Thoracic outlet
to extreme disability (eg, “Erb’s palsy” from avul- syndrome. 1st edition. New York: Springer-Verlag;
sion of the upper brachial plexus). 2013.
The brachial plexus may be damaged during any Kwong KF, Cooper LB, Bennett LA, et al. Clinical experi-
procedure by poor patient positioning. Abduction, ence in 397 consecutive thoracoscopic sympathec-
external rotation, or overextension of the shoulder, tomies. Ann Thorac Surg 2005;80(3):1063–6
particularly when coupled with contralateral head [discussion: 1066].
flexion, can cause traction injury to the superior Macchiarini P, Dartevelle P, Chapelier A, et al.
nerve roots. An unsupported and dependent shoul- Technique for resecting primary and metastatic
der in the lateral decubitus position is at particularly nonbronchogenic tumors of the thoracic outlet.
high risk. The resulting palsy is usually temporary; Ann Thorac Surg 1993;55(3):611–8.
Prevention and Management of Nerve Injuries 515
Menorca RM, Fussell TS, Elfar JC. Nerve physiology: Selke F, del Nido P, Swanson S. Sabiston and Spencer
mechanisms of injury and recovery. Hand Clin surgery of the chest. 8th edition. Philadelphia:
2013;29(3):317–30. Elsevier/Saunders; 2010.
Randolph G. Surgery of the thyroid and parathyroid Shields T, Locicero J III, Reed C, et al. General thoracic
glands. 2nd edition. Philadelphia: Elsevier/Saun- surgery. 7th edition. Philadelphia: Lippincott Wil-
ders; 2013. liams and Wilkins; 2009.
Seddon HJ, Medawar PB, Smith H. Rate of regeneration Sunderland S. A classification of peripheral nerve in-
of peripheral nerves in man. J Physiol 1943;102(2): juries producing loss of function. Brain 1951;74(4):
191–215. 491–516.
C h e s t Wa l l R e s e c t i o n a n d
Reconstruction
Management of Complications
Kweku Hazel, MD, Michael J. Weyant, MD*
KEYWORDS
Chest wall resection Prosthetic reconstruction Complications of chest wall resection
KEY POINTS
The main complications after chest wall resection and reconstruction include respiratory complica-
tions and wound/prosthetic complications.
The main risk factors for complications after chest wall resection include age, size of defect, and
concomitant lung resection.
It is critical to provide adequate tissue coverage over the prosthesis to minimize wound
complications.
A rigid prosthesis should be considered to reconstruct the largest of anterolateral chest wall
defects.
Section of General Thoracic Surgery, Division of Cardiothoracic Surgery, University of Colorado Denver School
of Medicine, 12631 East 17th Avenue, MS C310, Aurora, CO 80045, USA
* Corresponding author.
E-mail address: Michael.weyant@ucdenver.edu
Table 1
Morbidity and mortality rates of chest wall resection and reconstruction
The most frequent complications described are model of controlled chest wall resection
respiratory in nature and are reported to occur in comparing a group who had reconstruction with
up to 24% of patients.1,6 prosthetic mesh and a group who had only skin
closure. In this study, the size of the chest wall
RESPIRATORY COMPLICATIONS resection could be tightly controlled. The group
of animals who did not undergo any chest wall
Given the high incidence and level of morbidity reconstruction had significantly poorer respiratory
associated with respiratory complications after mechanics after the procedure.
chest wall resection, a great deal of attention has Other investigators have attempted to directly
been paid to postulating the mechanism behind compare the type of prosthesis used and assess
these risks. Much debate exists as to whether outcomes. Kilic and colleagues11 reported a series
there are superior reconstructive techniques that of patients comparing a rigid prosthesis (MMMM)
may lead to a decreased incidence of pulmonary versus PTFE. The investigators reported reduced
complications. Randomized trials do not exist paradoxic respiratory motion, lower operative
comparing types of prosthetic materials; popula- morbidity, and shorter length of stay in patients
tions of patients undergoing chest wall resection having the rigid prosthesis in this nonrandomized
are quite heterogeneous, making level 1 evidence study. In a series of patients having MMMM recon-
nearly impossible to obtain. The large retrospec- struction, Lardinois and colleagues12 report no dif-
tive series available suggest that over time ference in preoperative and postoperative forced
morbidity from respiratory complications seems expiratory volume in the first second of expiration
to be diminishing, with one of the most recent se- even after concomitant lung resection.12 The
ries demonstrating a respiratory complication rate finding that respiratory function can be unaltered
of 11%6 (see Table 1). These findings raise the with the use of a rigid prosthesis is enticing, but
question as to whether reconstructing the chest without an adequate comparison with nonrigid
wall with a truly rigid prosthesis, such as a methyl reconstruction the question remains open.
methacrylate marlex mesh composite (MMMM), is Given the lack of prospective randomized trials
more beneficial than a prosthetic mesh alone, it is unclear whether the surgical technique or
such as polytetrafluoroethylene (PTFE) or polypro- type of prosthesis clearly influences the incidence
pylene mesh (PPM). It is important to note that of respiratory complications. Advances in anes-
even though a lower rate of respiratory complica- thetic techniques or critical care may also be play-
tions was reported in that study, 7 of the 10 deaths ing a role. Overall it is reasonable to consider the
in the series were caused by respiratory complica- use of a rigid prosthesis in carefully selected pa-
tions indicating a need for continued study to tients with large anterolateral chest wall defects.
reduce these events.
The available data that suggest that prosthetic WOUND AND PROSTHETIC COMPLICATIONS
reconstruction of the chest wall is beneficial is in-
ferred from historical literature. In 1978, Thomas Wound complications and complications related
and colleagues8 reported on the possible benefit to the prosthesis are second in incidence to respi-
of surgical stabilization of the chest wall in patients ratory complications after chest wall resection and
with severe flail chest. Lardinois and colleagues9 reconstruction. These complications are reported
also reported results in 55 patients with severe an- to occur in 7% to 20% of patients.6,7 The main
terolateral flail chest who had improved respiratory complications related to the wound and prosthesis
mechanics after surgical chest wall stabilization. are wound infection, wound dehiscence, and
Niwa and colleagues10 reported data in a canine dislodgement or fracture of the prosthesis
Chest Wall Resection and Reconstruction 519
operating room. Flap loss from ischemia occurs in resected by stating the number of ribs resected
up to 5% of patients and can be a difficult problem in the specimen. Weyant and colleagues6 used
to treat, requiring alternative methods of soft tis- the area (square centimeters) measurements
sue coverage.5 Donor site complications are rare from pathology reports to more accurately des-
but include donor site hernias and infection. cribe the size of the lesions. For example, the
mean number of ribs removed in all of the pros-
PREDICTORS OF COMPLICATIONS thesis groups was 3, yet there were significant
differences in the size based on area measure-
Very few studies have provided analysis if predic- ments. Using more precise size measurements
tors of complications after chest wall resection should allow for more accurate description of
and reconstruction. The two series with greater complications.
than 100 patients who have performed predictive The report by Weyant and colleagues6 reports
analysis are summarized in Box 1. Many univariate that concomitant lung resection is a significant
predictors have been evaluated, but relatively few predictor of postoperative complications. This
turn out to be significant predictors of postopera- finding is not entirely surprising given the addi-
tive complications. The most significant predictors tional respiratory compromise produced by the
of complications seem to be the size of the lesion, added lung resection. In their series, concomitant
patient age, concomitant lung resection, ulcera- lung resection was performed in 141 (54%) of pa-
tion of the chest wall lesion before resection, and tients. The increased risk of postoperative compli-
the use of omentum in the reconstruction. cations seemed to be specific to anatomic lung
Regarding the size of the lesion, most historical resections (segmentectomy, lobectomy, bilobec-
studies have quantified the size of the lesions tomy, pneumonectomy), whereas the addition of
a wedge resection did not increase the risk of res-
piratory complications. Importantly, although mor-
Box 1 tality predictors cannot be reliably analyzed
Multivariate predictors of complications after because of the small sample size, it seems that
chest wall resection there is significant risk of death when combining
Factors Analyzed in Univariate Analysis chest wall resection with pneumonectomy. The
report from Weyant and colleagues6 contained 9
Type of prosthesis
patients who had combined pneumonectomy
Anatomic location with chest wall resection. These patients had a
Sternal resection postoperative mortality rate of 44% (4 of 9), indi-
Concomitant lung resection
cating the need for caution in deciding on opera-
tive therapy for these patients.
Medical comorbidity The finding of ulceration of a chest wall defect to
Prior radiotherapy be a predictor of postoperative wound complica-
Prior chemotherapy tions is not surprising, but it helps to illustrate
possible ways of preventing these complications.
Reoperation
Ulcerations can be produced by a tumor eroding
Soft tissue transfer through the skin or necrosis of skin harmed by ra-
Drains diation therapy. Lans and colleagues5 report the
Age incidence of ulcerating chest wall wounds to be
14% in their series. It is important to note that in
Size (cm2)
chest wall defects produced by radiation therapy,
Tumor type the ulcerated area is the proverbial tip of the
Ulceration iceberg; in planning for these resections, a wider
Use of omentum
area than is visible usually needs to be resected.
An additional anecdotal preventive measure in pa-
Significant Multivariate Predictors of tients with ulcerating chest wounds is to consider
Complications antibiotic therapy and local wound debridement
Age before the definitive resection.
Size of defect
Concomitant lung resection RECENT ADVANCES
Ulceration of chest wall lesion The high incidence of complications after chest
Use of omentum wall resection has led to a continued interest in
improving materials and surgical techniques to
Chest Wall Resection and Reconstruction 521
KEYWORDS
Outcomes Complications Tracheal resection
KEY POINTS
Tracheal resections are generally performed at specialized centers with experienced teams of
thoracic surgeons, anesthesiologists, otolaryngologists, and intensivists.
Serious complications following tracheal surgery may be avoided by using intraoperative tech-
niques that preserve adequate blood supply and eliminate tension on the tracheal end-to-end
anastomosis.
Careful preoperative evaluation for selecting appropriate surgical candidates and determining
optimal timing of surgery is critical to avoiding complications.
Primary preoperative factors that increase the risk of anastomotic complications after tracheal sur-
gery are reoperation, increased length of resection (with need for release maneuvers), and need for
preoperative tracheostomy.
Strategies to avoid or detect postoperative complications early center on slow diet advancement to
prevent aspiration, aggressive clearance of secretions, voice rest (for laryngotracheal resections),
guardian chin stitch, and surveillance bronchoscopy.
malignancies invading the trachea usually involvement, and vocal cord function, as well as
arise from bronchogenic or thyroid presence of secretions, infection, inflammation,
neoplasms.3,4 granulation, or obstruction. Biopsy is indicated
Idiopathic laryngotracheal stenosis (ILTS) (lar- for any endobronchial lesions. Nebulizers and
yngotracheal resection) is relatively rare but is intravenous antibiotics may be used to treat signs
seen with increasing frequency. This disease of purulent secretions and active infection, and
almost entirely affects younger women and any airway surgery should be postponed in this
primarily involves the cricoid.5 scenario. Patients should not be ventilator depen-
Tracheoesophageal and tracheoinnominate dent, so as to prevent exposure of a fresh anasto-
fistulas, also a consequence of long-term mosis to positive pressure ventilation, foreign
high-pressure endotracheal tube and trache- body, and airway colonization postoperatively.
ostomy cuffs, which cause areas of granula- Active airway inflammation may be treated with a
tion, malacia, and eventual erosion into course of corticosteroid therapy, but if possible,
surrounding structures. Fistulas may also steroids should be discontinued well before
occur due to penetrating trauma to the tracheal surgery. Tracheostomy stomas should
airway.6 be mature or well healed before surgery. Radiation
Congenital and postinfectious lesions are therapy in the surgical field should likewise be
relatively rare and can represent a wide spec- avoided preoperatively, because it greatly in-
trum of disease from short- or long-segment creases the risk of anastomotic separation.9 No tri-
stenosis, cartilaginous fibrosis, or calcified als demonstrate any role for neoadjuvant radiation
nodules. in the treatment of tracheal malignancies, although
there is evidence of benefit in the palliative or adju-
Although many of these indications are serious vant setting, especially in the case of a positive
and if left untreated can progress to detrimental margin or incomplete resection.10 Any plans for
airway compromise, almost none of them signify reoperative tracheal surgery should be delayed
true emergencies. It is important to emphasize at least several months after the initial operation,
this point because careful and extensive preoper- if possible, while the airway may be stabilized
ative evaluation and determining the appropriate with a T-tube.11
timing of surgery are key measures to avoiding
complications downstream.7 The exception is tra-
cheoinnominate fistula, for which immediate and SURGICAL TECHNIQUE
definitive intervention may be necessary to control The first attempt at tracheal resection and recon-
hemorrhage. Otherwise, the following temporizing struction is often the best chance at achieving a
measures can be used until the optimal time for positive outcome for the patient. The operation
elective tracheal resection and reconstruction: can divided into 4 phases12:
Rigid bronchoscopy with dilation or tumor
coring for airway obstruction Induction
Flexible bronchoscopy with balloon dilation of The procedure begins with flexible/rigid bron-
airway stenosis choscopy to inspect the airway and clear secre-
Therapeutic flexible bronchoscopy for clear- tions. The patient is positioned supine with the
ance of secretions or hemoptysis neck extended if the approach is via a low collar
Tracheostomy, particularly in the situation of incision or median sternotomy, although lesions
tracheoesophageal fistula (TEF) such that the near the carina require a right-sided posterolateral
cuff is inflated and occlusive distal to the fis- thoracotomy in the left lateral decubitus position.
tula to protect the airway Usually, the airway is intubated under direct
visualization distal to the lesion if possible with
Of note, placement of a covered expandable
a flexible, armored single-lumen endotracheal
stent in the trachea is not a preferred option for
tube. A total intravenous anesthetic is preferable
bridging a patient to surgery because of the in-
throughout the operation, as opposed to inhaled
crease in inflammation and granulation and
agents.
should be mainly used in palliative circumstances
when resection and reconstruction are not
Resection
possible.8
Preoperative evaluation and selection of pa- The region of interest of the trachea is exposed
tients centers on computed tomographic (CT) im- and dissected carefully so as not to cause injury
aging and flexible and rigid bronchoscopy to to the recurrent laryngeal nerves or innominate
assess the length of diseased trachea, laryngeal artery or devascularize the trachea. The oral
Complications Following Tracheal Resection 501
endotracheal tube is then withdrawn, and once In summary, during tracheal surgery, it is crucial
the airway is transected, cross-table ventilation to abide by 3 main principles:
involving a flexible, armored single-lumen endotra-
cheal tube or jet catheter is initiated via the distal Preservation of blood supply to the trachea
airway. The appropriate length of trachea is then (by limiting dissection to just beyond the
resected by trimming each end to healthy mucosa area of resection)
and cartilage. Approximately 4.5 cm of the cervical Elimination of tension (by judicious airway
trachea can be resected safely with neck flexion length resection and use of release
and pretracheal mobilization alone in ideal pa- maneuvers)
tients. Up to 50% of the total tracheal length may Construction of an air-/water-tight anasto-
be resected at maximum with the use of more mosis (by meticulous attention to detail)
complex release maneuvers, although the risk of Postoperative management entails slow diet
anastomotic separation increases markedly with advancement to prevent aspiration, aggressive
increasing length resected because of excess pulmonary toilet to clear secretions, and voice
tension. rest (if a laryngotracheal resection was performed).
Laryngeal manipulation or release can affect swal-
Reconstruction lowing function, and these patients should un-
Once the resection is complete, 2-0 vicryl full- dergo formal evaluation before starting oral
thickness traction sutures are placed in the midline intake. Antibiotics and steroids are used selec-
laterally on both sides of the trachea. These su- tively. Surveillance bronchoscopy at 1 week or
tures are brought together to assess tension. sooner is the key to recognizing anastomotic com-
Release maneuvers such as (1) neck flexion, (2) plications early because ischemia can be detected
blunt finger dissection in the pretracheal plane by endoscopy earlier than when it presents with a
down to the carina (as is done for a mediastino- wound infection.
scopy) for pretracheal mobilization, (3) suprahyoid
laryngeal release,13 (4) takedown of the inferior CLINICAL OUTCOMES AND COMPLICATIONS
pulmonary ligament, (5) division of the left main-
By far, the largest experience in a review of anas-
stem bronchus, or (6) intrapericardial hilar release
tomotic complications of tracheal surgery comes
may be used to bring the ends of the trachea
from the Massachusetts General Hospital (MGH),
together without tension. The anastomosis is
published in 2004 by Wright and colleagues.14
then performed using lubricated 4-0 vicryl sutures
This experience includes 901 patients from 1975
placed in an interrupted manner 3 mm apart, and
to 2003 who underwent tracheal resection and
3 mm from the cut edge of the trachea. Once the
reconstruction for indications of postintubation
sutures are placed, the oral endotracheal tube is
stenosis (PITS), tumor, TEF, and idiopathic ILTS.
then advanced distal to the anastomosis, and the
Overall morbidity and mortality were reported at
sutures are tied down sequentially (traction su-
18.2% and 1.2%, respectively. The highest rates
tures first) with the knots on the exterior of the
of morbidity and mortality were observed in pa-
airway. A pedicled tissue flap, most often one of
tients with TEF followed by those with tracheal tu-
the strap muscles, may be used to buttress the
mors, whereas the lowest rates were seen in
anastomosis and seal it from surrounding struc-
patients with ILTS. This result is likely attributable
tures such as the esophagus and innominate ar-
to the difference in patient populations, as well
tery. A closed suction drain is left in the wound
as the length and complexity of repairs. Patients
before closure.
with ILTS tend to be young healthy women
requiring short resections for focal disease,
Extubation
whereas those with TEF are generally sicker with
The goal is to extubate the patient at the end of the a tracheostomy in place, have poor tissue quality,
procedure. If there are concerns regarding the and require a more lengthy resection and repair.
airway or postoperative edema and swelling, a Patients with tracheal malignancies likewise tend
small tracheostomy may be placed proximal or to be older and also may require longer resections
distal to the anastomosis as necessary. A guardian to achieve negative margins.
stitch is placed between the chin and chest to Overall, anastomotic complications were seen
keep the neck in a slightly flexed to neutral position in 9% of patients, and the trend was similar with
to prevent neck hyperextension as additional the highest rates of anastomotic complications in
anastomotic protection. Patients are typically patients with TEF and PITS. A complication with
monitored in the intensive care unit for 24 to the anastomosis results in significant morbidity
48 hours postoperatively. and prolonged hospitalization, and it is the single
502 Sihag & Wright
single-institution series reported in the literature proliferation is focal. Injection with corticosteroids
are summarized in Table 1. has also been described with some effect. Laser
Nonanastomotic, airway-related complications ablation and brachytherapy are reasonable treat-
such as glottic edema and/or vocal cord edema ment strategies for benign granulation tissue of
and need for temporary tracheostomy occur at the airway from tracheal stents and tracheos-
an exceedingly low rate, but all are more prevalent tomies in nonsurgical candidates; however, safety
following laryngotracheal resection by an order of and efficacy following airway reconstruction has
twofold or more compared with straightforward not been extensively studied. Silicone Montgom-
tracheal resection. Laryngeal reconstruction can ery T-tubes (Hood Laboratories, Pembroke, MA)
predictably generate some degree of glottic can be used if the proliferation is severe and recur-
edema and laryngeal dysfunction. Both of these rent and the threat of reobstruction is high.
adversely impact the normal swallowing mecha- Restenosis of the airway following reconstruc-
nism and pose a greater risk of aspiration. Severe tion is more common than granulation tissue for-
upper airway swelling with stridor and airway mation but still occurs at a rate of less than 1%.
compromise necessitates temporary tracheos- As with granulation tissue formation, patients
tomy placement. Injury to the recurrent laryngeal with a history of PITS are at highest risk. Residual
nerve with vocal cord paralysis was observed in stenosis is possible due to inadequate resection
less than 2% of patients. of initial tracheal stenosis, especially if presenta-
tion is earlier than expected. However, low-grade
ischemia to the anastomosis is the more likely
MANAGEMENT OF POSTOPERATIVE
cause. The time course of restenosis is more
COMPLICATIONS
insidious, and symptoms may develop over
Anastomotic Complications
several months. The diagnosis is confirmed using
Granulation tissue formation at the site of the flexible bronchoscopy, and the mainstay of man-
anastomosis is relatively rare as reported in the agement is dilation (Fig. 4). Either graduated rigid
MGH experience (7 of 901 patients, or .007%). bronchoscopes or a pneumatic balloon dilator
However, other series report the incidence to be may be used to dilate the anastomosis because
as high as 33%, despite using a similar technique both techniques are effective. If dilations are
to that described by Grillo and colleagues25 at required more often than every 3 to 6 months, re-
MGH for construction of the anastomosis. Howev- operation should be considered. Results of initial
er, the number of patients analyzed in this series is operative intervention for tracheal stenosis (idio-
considerably smaller at only 12. Some degree of pathic or postintubation) are quite good with suc-
granulation tissue is to be expected postopera- cess rates reported as high as greater than 95%. If
tively, but in the worst-case scenario, granulation there is adequate amount of residual trachea, re-
tissue can cause obstruction of the airway. The operation for postoperative recurrent tracheal ste-
time course of developing symptomatic granula- nosis also carries a remarkably high success rate
tion tissue is generally days to weeks, and patients of 92% despite the increased risk of anastomotic
present with stridor. As in the case of any airway complications if deferred for 6 months to 1 year al-
symptoms following tracheal surgery, prompt lowing for complete resolution of postoperative
evaluation with flexible bronchoscopy is indicated healing.26 Use of silicone stents or a Montgomery
to assess the problem (Figs. 2 and 3). Manage- T-tube may be necessary for long-term manage-
ment options include rigid bronchoscopy with ment of restenosis that is not amenable to
direct debridement of granulation tissue if the reresection.
Table 1
Rates of major complications following tracheal reconstruction
Impaired wound healing of the anastomosis is a seem promising, long-term outcomes still need
complication that is often recognized early on, at to be examined.
the initial surveillance bronchoscopy at 1 week or
less from surgery (Fig. 7). Presumably, wound
Fistulas
healing problems such as focal areas of ischemic
or necrotic cartilage or even partial separation Tracheoinnominate fistula is the most feared and
eventually lead to the more severe long-term anas- catastrophic complication following tracheal sur-
tomotic complications of restenosis and dehis- gery but happens with extraordinarily low fre-
cence, and possibly even granulation tissue quency. In the MGH experience, only 3 patients
formation. At the authors’ institution, hyperbaric developed this complication, although 2 of them
oxygen therapy administered at 2 atm; 100% oxy- died. A small-scale, so-called sentinel bleed may
gen for at least 90 minutes daily over 5 to 14 days present before signs of significant hemorrhage
has led to a substantial improvement in healing as into the airway are apparent. The main tenet in
monitored by frequent bronchoscopy.27 An initial dealing with this issue is avoiding it in the first
series of 5 patients was reported in 2014, and hy- place, because this complication is usually the
perbaric oxygen therapy continues to be an result of technical missteps. The surgical princi-
ongoing area of study. While short-term findings ples that must be adhered to in order to prevent
this problem are (1) not dissecting out the artery
completely so as to leave the adventitia exposed,
(2) covering any anastomosis in proximity to the ar-
tery with a vascularized muscle flap, and (3) not
overresecting the airway so as to leave the re-
construction at greater tension and threat of
dehiscence. If a sentinel bleed is perceived,
bronchoscopy and CT scan can be pursued to rule prompt bronchoscopic evaluation to make the
out communication with the innominate artery if diagnosis (Fig. 8). Mild-to-moderate edema of
the patient remains hemodynamically stable. Usu- the vocal cords can be easily managed with voice
ally, some degree of anastomotic separation will rest, steroids, aspiration precautions, diuretics,
be evident on bronchoscopy. In the setting of epinephrine nebulizers, and most importantly,
catastrophic hemorrhage into the airway, the patience. An assessment of whether or not the pa-
wound may need to be reopened at the bedside tient has an adequate airway must be made,
and digital pressure applied while the patient is because significant glottic edema can occasion-
emergently transported to the operating room for ally lead to obstruction. In this instance, tracheos-
reexploration. Securing the airway in this scenario tomy placement above or below the anastomosis
is always challenging. Once in the operating room, is indicated. Recurrent laryngeal nerve injury only
the prior incision is reopened and a partial (or com- occurs at a rate of less than 2% in most reported
plete) sternotomy is needed to obtain adequate series, even in laryngotracheal resection despite
exposure. The innominate artery should be the proximity of the nerve to the field of dissection.
clamped and divided proximal to the carotid- The approach among surgeons at the authors’
subclavian bifurcation. Each end should be over- institution is to dissect in the peritracheal planes
sewn and reinforced with a vascularized tissue immediately adjacent to the airway and to deli-
flap. The risk of cerebral ischemia is very low, berately avoid violating the tracheoesophageal
although a bypass graft should be considered groove or trying to identify the path of the nerve.
if changes in the electroencephalogram are Hoarseness of the voice and weak cough are usu-
observed. In addition, repairing and buttressing ally fairly easy to discern postoperatively. A laryng-
of the tracheal anastomosis should also be per- ologist may be consulted to assist with diagnosis
formed at the same time. and management. Awake bedside flexible laryn-
TEF, on the other hand, is a consequence of goscopy is performed to inspect the vocal cords.
breakdown of the membranous wall of the anasto- Minor cord dysfunction is likely to improve over
mosis. This breakdown creates an infected field time and may be a consequence of traction injury
posteriorly and allows for fistula formation to the to the nerve. However, if there is complete cord
esophagus. Unrecognized injury to the esophagus paralysis either in the abducted or adducted posi-
during the index operation or resection of part of tion, a temporizing laryngoplasty procedure (vocal
the wall of the esophagus to obtain a negative cord injection with collagen) should be undertaken
margin during tumor resection may also be during the early postoperative course to prevent
possible causes. Cough, dysphagia, and aspira- aspiration. Bilateral recurrent laryngeal nerve injury
tion pneumonia are the hallmark clinical signs.
Both bronchoscopy and barium swallow can
confirm the diagnosis. This complication is simi-
larly quite uncommon and was detected again in
only 3 of 901 patients in the MGH series, although
all were successfully salvaged. Ultimately, opera-
tive exploration with tracheal reresection and
anastomosis, primary repair of the esophagus in
2 layers, and placement of a vascularized strap
muscle or omental flap between suture lines repre-
sents definitive treatment of this problem. How-
ever, in the short term, tracheostomy placement
with an occlusive cuff distal to the fistula, enteral
feeding, drainage, and broad-spectrum antibiotics
are indicated to clear any infection before an
attempt at repair.
Laryngeal Complications
The incidence of glottic edema is relatively high
following laryngotracheal resection and recon- Fig. 8. This bronchoscopic image was from 1 week after
struction but is rare following tracheal resection. a high laryngotracheal resection in a patient who had
Rates of glottic edema are reported at 5% in pa- severe vocal cord edema. She was aphonic and had
tients who underwent laryngotracheal reconstruc- mild stridor. She improved with intravenous dexameth-
tion. Symptoms of hoarseness and stridor should asone and with time without further intervention.
Complications Following Tracheal Resection 507
is an indication for tracheostomy due to glottic 2. Grillo HC, Donahue DM. Postintubation tracheal ste-
obstruction. nosis. Chest Surg Clin N Am 1996;6(4):725–31.
Swallowing dysfunction is not common after 3. Grillo HC, Mathisen DJ. Primary tracheal tumors:
tracheal surgery, but needs to be appropriately as- treatment and results. Ann Thorac Surg 1990;
sessed if suspected. The assessment should 49(1):69–77.
begin preoperatively with a thorough evaluation 4. Gaissert HA, Honings J, Grillo HC, et al. Segmental
of glottic function before laryngotracheal or laryngotracheal and tracheal resection for invasive
tracheal resection. Several patients with PITS, for thyroid carcinoma. Ann Thorac Surg 2007;83(6):
example, may have glottic injury from traumatic 1952–9.
endotracheal intubation. Elderly patients, patients 5. Grillo HC. Management of idiopathic tracheal steno-
who have been intubated for a prolonged period, sis. Chest Surg Clin N Am 1996;6(4):811–8.
and those who have undergone a suprahyoid 6. Muniappan A, Wain JC, Wright CD, et al. Surgical
laryngeal release procedure are most at risk. Elic- treatment of nonmalignant tracheoesophageal fis-
iting the help of a speech and swallow pathology tula: a thirty-five year experience. Ann Thorac Surg
specialist is recommended. A modified barium 2013;95(4):1141–6.
swallow study can reveal the extent of swallowing 7. Mathisen DJ. Complications of tracheal surgery.
dysfunction. The swallowing mechanism should Chest Surg Clin N Am 1996;6(4):853–64.
improve with time and exercises, but enteral 8. Gaissert HA, Grillo HC, Wright CD, et al. Complica-
feeding may be required via a gastrostomy tube; tion of benign tracheobronchial strictures by self-
any suspected significant aspiration event should expanding metal stents. J Thorac Cardiovasc Surg
be investigated further with bronchoscopy. In the 2003;126(3):744–7.
MGH series, 24 of 901 patients exhibited swallow- 9. Muehrcke DD, Grillo HC, Mathisen DJ. Reconstruc-
ing dysfunction with evidence of an aspiration tive airway operation after irradiation. Ann Thorac
event postoperatively. Severe aspiration pneu- Surg 1995;59(1):14–8.
monia with progression to acute respiratory 10. Chow DC, Komaki R, Libshitz HI, et al. Treatment of
distress syndrome is obviously rare, but the occa- primary neoplasms of the trachea. The role of radia-
sional patient has required prolonged ventilatory tion therapy. Cancer 1993;71(10):2946–52.
support, and even venovenous extracorporeal 11. Gaissert HA, Grillo HC, Mathisen DJ, et al. Tempo-
membrane oxygenation. rary and permanent restoration of airway continuity
with the tracheal T-tube. J Thorac Cardiovasc Surg
1994;107(2):600–6.
SUMMARY
12. Grillo HC. Surgery of the trachea and bronchi. Ham-
Strict adherence to the principles of tracheal ilton (Ontario): BC Decker Inc; 2004.
surgery by preserving adequate blood supply, 13. Montgomery WW. Suprahyoid release for tracheal
eliminating tension, and using meticulous anasto- anastomosis. Arch Otolaryngol 1974;99(4):255–60.
motic technique represents the best strategy to 14. Wright CD, Grillo HC, Wain JC, et al. Anastomotic
avoid complications of tracheal surgery. Further- complications after tracheal resection: prognostic
more, careful preoperative evaluation and appro- factors and management. J Thorac Cardiovasc
priate selection of patients also play important Surg 2004;128(5):731–9.
roles in achieving a positive outcome. In experi- 15. Wright CD, Graham BB, Grillo HC, et al. Pediatric
enced centers, rates of significant morbidity and tracheal surgery. Ann Thorac Surg 2002;74(2):308–
mortality approximate 10% and 2%, respectively. 13 [discussion: 314].
The range of postoperative complications after 16. Bibas BJ, Terra RM, Oliveira Junior AL, et al. Predic-
tracheal reconstruction includes edema, granula- tors for postoperative complications after tracheal
tion, fistula, separation, and restenosis. Anasto- resection. Ann Thorac Surg 2014;98(1):277–82.
motic complications, in particular, lead to an 17. Grillo HC. Primary reconstruction of airway after
increased risk of mortality and must be treated resection of subglottic laryngeal and upper tracheal
aggressively to secure the airway. Early detection stenosis. Ann Thorac Surg 1982;33(1):3–18.
via bronchoscopy and effective airway manage- 18. Pearson FG, Cooper JD, Nelems JM, et al. Primary
ment with dilation, T-tube stenting, or tracheos- tracheal anastomosis after resection of the cricoid
tomy often yields successful results. cartilage with preservation of recurrent laryngeal
nerves. J Thorac Cardiovasc Surg 1975;70(5):
REFERENCES 806–16.
19. Couraud L, Jougon J, Velly JF, et al. Iatrogenic ste-
1. Cooper JD, Grillo HC. Analysis of problems related noses of the respiratory tract. Evolution of therapeu-
to cuffs on intratracheal tubes. Chest 1972;62(2 tic indications. Based on 217 surgical cases. Ann
Suppl):21S–7S. Chir 1994;48(3):277–83 [in French].
508 Sihag & Wright
20. Rea F, Callegaro D, Loy M, et al. Benign tracheal for benign strictures in adults. Interact Cardiovasc
and laryngotracheal stenosis: surgical treatment Thorac Surg 2008;7(2):227–30 [discussion: 230].
and results. Eur J Cardiothorac Surg 2002;22(3): 25. Marques P, Leal L, Spratley J, et al. Tracheal resec-
352–6. tion with primary anastomosis: 10 years experience.
21. Amoros JM, Ramos R, Villalonga R, et al. Tracheal Am J Otolaryngol 2009;30(6):415–8.
and cricotracheal resection for laryngotracheal ste-
nosis: experience in 54 consecutive cases. Eur J 26. Donahue DM, Grillo HC, Wain JC, et al. Reoperative
Cardiothorac Surg 2006;29(1):35–9. tracheal resection and reconstruction for unsuc-
22. D’Andrilli A, Ciccone AM, Venuta F, et al. Long-term cessful repair of postintubation stenosis. J Thorac
results of laryngotracheal resection for benign ste- Cardiovasc Surg 1997;114(6):934–8 [discussion:
nosis. Eur J Cardiothorac Surg 2008;33(3):440–3. 938–9].
23. Cordos I, Bolca C, Paleru C, et al. Sixty tracheal re- 27. Stock C, Gukasyan N, Muniappan A, et al. Hyper-
sections–single center experience. Interact Cardio- baric oxygen therapy for the treatment of anasto-
vasc Thorac Surg 2009;8(1):62–5 [discussion: 65]. motic complications after tracheal resection and
24. Marulli G, Rizzardi G, Bortolotti L, et al. Single- reconstruction. J Thorac Cardiovasc Surg 2014;
staged laryngotracheal resection and reconstruction 147(3):1030–5.
Postoperative
Chylothorax
Nicola Martucci, MD, Maura Tracey, RN, Gaetano Rocco, MD, FRCSEd*
KEYWORDS
Chylothorax Thoracic duct Chylous effusion Postoperative complication
KEY POINTS
Chylothorax is an unusual but serious complication of thoracic surgical procedures, and is associ-
ated with considerable morbidity if not addressed in a timely fashion.
Thoracic surgeons should be familiar with the anatomy of the thoracic duct, the causes of chylo-
thorax, and the implications of a prolonged chyle leak.
Postoperative chylothorax is typically treated conservatively with reasonable success rates. The
timing of direct intervention with percutaneous embolization of the cisterna chyli or thoracoscopic
ligation of the thoracic duct is controversial, and direct intervention is often reserved for refractory
cases.
chyli, or receptaculum chyli, described in 1651 dividuals. The presence of anatomic variations
Division of Thoracic Surgery, Department of Thoracic Surgery and Oncology, Istituto Nazionale Tumori, Fonda-
zione Pascale, IRCCS, Via Semmola 81, Naples 80131, Italy
* Corresponding author.
E-mail address: Gaetano.Rocco@btopenworld.com
together with the presence of multiple small cause of iatrogenic chylothorax, with an incidence
lymphatic vessel tributaries, shown by several ranging from 0.2% to 10.5%,2 and seems to be
anatomic studies, is probably responsible of the more frequent with a transhiatal or thoracoscopic
incidence of postoperative chylothorax, irrespec- approach to the esophagus. Most of the lesions
tive of the surgeon’s meticulous attention and skill. occur near the arch of the aorta and azygos, where
The primary role of the thoracic duct is to drain the esophagus and the thoracic duct are the
the lymph and the chyle that originate from the closest.12 In lung surgery, chylothorax has been
gastrointestinal tract: the chyle contains large reported in 2.1% of the resections.13 Reportedly,
amounts of cholesterol, triglycerides, chylomi- chylothorax after lung surgery is more common
crons, fat-soluble vitamins, and albumin, whereas on the right side and after pneumonectomy
the lymph consists of immunoglobulins, enzymes, and it is probably related to mediastinal lympha-
digestive products, and white blood cells, most of denectomy after pulmonary resection for
which are lymphocytes.7 In adults, the thoracic malignancy.13,14 Recently, the incidence of
duct transports up to 4 L of a chyle per day and postoperative chylothorax after video-assisted
its flow depends on several factors, such as diet, thoracic surgery (VATS) lung resection and medi-
medications, and gastrointestinal absorption. The astinal nodal dissection for primary lung cancer
flow is maintained by the combined effect of was estimated to be 2.6%.15
intra-abdominal and intrathoracic pressures, arte-
rial compression of the neighboring vessels, and CLINICAL FEATURES
the contraction of the muscles of the duct.8
Rupture of the thoracic duct leads to a signifi- In the postoperative period the presence of chylo-
cant loss of fat, protein, and T lymphocytes with thorax should be suspected when an abundant
consequent disorders of the immunologic and pleural collection of more than 500 mL/d with milky
nutritional profiles together with disorders related features occurs between the second and tenth
to the accumulation of large quantities of fluid in postoperative days. Usually, the effusion is evident
the pleural space.9,10 In general, after an acute when the patient resumes oral intake.10 The
phase related to mechanical compression of the different clinical aspects of chylothorax depend
intrathoracic organs caused by accumulation of on the amount of chyle lost, as well as its cause.
fluid, complications related to chylothorax depend A rapid and abundant loss is generally associated
on the consequences of the chronic loss of chyle, with hypovolemia and respiratory distress from the
whereas immunosuppression occurs because of accumulation of chyle in the pleural space. Malnu-
the loss of lymphocytes and immunoglobulins, trition is caused by the loss of proteins, fat, and
making these patients vulnerable to infection. vitamins, whereas loss of electrolytes can cause
The loss of electrolytes can result in hypovolemia, hyponatremia and hypocalcemia.9 The significant
hyponatremia, hypocalcemia, and metabolic loss of immunoglobulins, T cells, and proteins
acidosis, whereas malnutrition is the result of can cause immunosuppression, which predis-
chronic loss of fatty acids and proteins. All these poses the patient to systemic opportunistic infec-
events result in the high mortality in untreated pa- tions, whereas infection of the pleural effusion is
tients with chylothorax.2,11 rare because chyle is bacteriostatic.16 Digoxin
and amiodarone may be ineffective when adminis-
CAUSES tered to patients with chylothorax because they
are expelled through the chyle in the pleural
The causes of chylothorax can be divided into 3 space.1 Pleural effusion may be 1-sided, either
main categories: congenital, neoplastic, and trau- right (50%), left (33%), or bilateral (16.66%), and
matic. The traumatic cause is the most common, its location depends on the point of duct injury.
and can be blunt trauma or vertebral fractures, Lesions above the fifth thoracic vertebra deter-
as well as injuries during certain diagnostic proce- mine left pleural effusion, whereas lesions below
dures, such as catheterization of the subclavian this level lead to right pleural effusion.9,17
vein, central venous catheterization, or surgery.
Iatrogenic injuries may involve the entire thoracic DIAGNOSIS
duct; lesions of the abdominal tract can occur dur-
ing sympathectomy or lymph node dissection, Diagnostic tests for suspected chylothorax begin
whereas lesions of the thoracic tract may result with the confirmation of the diagnosis by chemical
from lung and esophageal procedures. In addition, analysis of pleural fluid: a high content of lympho-
lesions of the cervical portion of the duct may cytes and chylomicrons can confirm the diag-
develop during lymph node dissection or neck sur- nosis.18 Radiological studies, such as traditional
gery.10 Esophageal surgery is the most common chest radiograph and computed tomography
Postoperative Chylothorax 525
scan, do not allow diagnosis of chylothorax; they include talc pleurodesis, which can be performed
only indicate the presence of persistent postoper- by either talc slurry or by means of uniportal
ative pleural effusion. The definitive diagnosis is VATS in patients who are nonintubated and only
obtained by means of thoracentesis and subse- sedated,27 yielding a success rate of 73%, espe-
quent laboratory analysis of collected pleural fluid. cially if full pulmonary reexpansion is obtained.28
Laboratory analysis is necessary since the macro- In addition, pleurodesis with OK-432, a lyophilized
scopic features of the pleural collection may be preparation of Streptococcus pyogenes, has
misleading since the classic milky white color is shown an efficacy of up to 87%.29,30 Conservative
found in only 50% of cases, because it is often se- treatment including pleurodesis has been advo-
rous fluid or bloody serum, or blood.19 Therefore, cated as the first line of treatment of chylothorax
chemical examination is necessary to make a resulting after lung surgery.13
diagnosis based on the presence of chylomicrons Lymphangiography per se has been suggested to
in the pleural fluid. Chylomicrons are molecular have a therapeutic effect in refractory postoperative
complexes of proteins and lipids synthesized in chylothorax.31 PL with subsequent catheterization
the jejunum and transported through the thoracic of cisterna chyli or thoracic duct is another treat-
duct into the circulation. Pedal lymphography ment option that allows percutaneous embolization
(PL) is considered the gold standard for the diag- and direct injection of sclerosing agents.21,32
nosis of chylothorax and is a technique that allows Reportedly, the success rate in identifying the leak
the precise localization of the leak in the thoracic with PL ranges between 64% and 86%.21
duct.10,20,21 This diagnostic method also has Percutaneous embolization of the cisterna chyli
some therapeutic effects, probably caused by tis- or the thoracic duct is usually performed with mi-
sue sclerosis induced by iodinated contrast mate- crocoils delivered through microcatheters and
rial.20 It is still debated whether to perform pedal positioned both proximal and distal to the leak.
lymphography only in those patients undergoing In addition, cyanoacrylate glue is added to ensure
surgical repair of the leak or in all patients with complete occlusion of the duct.21 According to the
postoperative chylothorax for diagnostic pur- literature, the thoracic duct is successfully cannu-
poses.14,22 The alternative to PL is intranodal lated in only two-thirds of the patients with postop-
lymphography (IL), which uses ultrasonography- erative chylothorax. However, in these patients,
guided detection of accessible nodes (ie, inguinal) successful occlusion of the thoracic duct is
and subsequent injection of these nodes with lipio- obtained in 90% of cases.21 Main procedure-
dol. The advantages of the intranodal injection related complications include parenchymal
include bypassing of the lower extremities, thus bleeding, lipiodol embolization in the pulmonary
yielding reduced duration of the procedure, as system, and infection of the injection site.21
well as decreased radiation doses and volumes The timing of surgical ligation is still debated,
of contrast medium.21 As a consequence, IL is with some surgeons suggesting that loss of more
preferentially used in children because of the small than 1 L of chyle per day for a week is evidence
size of their lymphatic vessels and the commonly of failure of conservative treatment,33 others re-
enlarged inguinal nodes.21 sorting to surgery after 5 day of an output greater
than 1 L per day,34 and a few surgeons waiting
TREATMENT for more than 2 weeks and intervening only if there
are severe metabolic and nutritional imbal-
Treatment of postoperative chylothorax can be ances.1,35 According to some investigators,29
either conservative or surgical. Conservative early surgical treatment may be indicated if the
therapy consists of reducing the flow of chyle by drainage through the chest tube is greater than
means of complete elimination of fat intake by 500 mL during the first 24 hours after initiation of
establishing enteral or, preferably, parenteral complete fasting and total parenteral nutrition.
feeding, which is considered the nutrition of choice Surgical treatment of postoperative chylothorax
in many studies.10,23 Use of somatostatin or octreo- includes multiple options, among which the first is
tide has been shown to be useful in the conserva- the direct ligation of the duct by means of a trans-
tive treatment of chylothorax because these thoracic access, proposed by Lampson36 in 1948.
hormones reduce the production of intestinal chyle, The surgical ligation of the thoracic duct requires
decreasing the flow through the thoracic duct.24–26 the location of the leak by means of a preoperative
Drainage of the pleural cavity is an indispens- test, such as PL, which should help to plan the sur-
able procedure in conservative management and gical procedure. Some investigators use preoper-
a chest tube without suction is always placed ative or intraoperative administration of fats,
because suction does not allow the closure of through a nasogastric tube, which can facilitate
the fistula.10 Conservative treatment may also the visualization of the leak in the thoracic duct
526 Martucci et al
by increasing the flow of lymph.37 Besides the showed no difference in the incidence of chylo-
need to locate the leak, other challenges may be thorax in the event of prophylactic ligation, which
caused by the postoperative inflammation, with also had an adverse effect on overall survival on
pleural thickening, which makes surgical dissec- multivariate analysis.45
tion as well as suturing the inflamed tissue diffi- In refractory cases and in cases of malignancy
cult.10 Once the leak is identified it can be with patients unsuitable for surgical treatment,
sutured or clipped, even if good results have some investigators have described the use of a
been reported with the use of biological glues.38 pleuroperitoneal shunt, which was first used in pe-
Surgery can be performed by means of thoracot- diatric cases48 and then in adults.49 This device
omy, or more recently by VATS or a robotic ensures a 1-way communication between the
approach. VATS has become the treatment of pleura and the peritoneum, through a pump sys-
choice to reduce the invasiveness of the surgical tem that is activated by a simple pressure by the
repair of the defect in the thoracic duct.39,40 In patient. The shunt minimizes immunologic and
cases in which the loss of chyle is not identified nutritional deficiency and when the loss of chyle
either in thoracoscopy or thoracotomy a mass ends it can easily be removed.8,23,50
ligation of the duct may be indicated above the An alternative therapy for the treatment of post-
esophageal hiatus between the aorta, vertebral operative chylothorax is radiotherapy, proposed
bodies, and the pericardium. This ensures duct because of the positive experience in the treat-
ligation at its entry in the chest, sealing all the ment of inguinal lymphatic fistulas51 and in the
accessory ducts that could be the source of the management of malignant chylothorax.52 Radia-
chylothorax. The mass ligation of the thoracic tion therapy, in combination with a fat-free diet,
duct is usually performed by means of thoraco- may be an effective alternative to traditional surgi-
scopy, especially if performed early.41 Some in- cal treatments but it still lacks reliable supportive
vestigators suggest that the ligation of the data.53
thoracic duct should be performed immediately
behind the cisterna chyli using an abdominal SUMMARY
approach.42
Chylothorax after esophagectomy can be an Postoperative chylothorax is a challenging compli-
ominous event.43 A literature and institutional re- cation of major lung surgery and esophagectomy.
view of the incidence of chylothorax in these pa- Although it is still not certain whether it is best to
tients was published in 2013 by Kranzfelder and proceed to concurrent thoracic duct ligation dur-
colleagues.44 Overall, there was a 2% incidence ing esophagectomy or whether and when to begin
of postoperative chylothorax at their institution conservative management, the treatment options
and 2.6% from the systematic review of the litera- are many and potentially effective in more than 1
ture.44 Early reoperation was the institutional clinical scenario. The association of diet and pleu-
preferred method to treat postoperative chylo- rodesis, the use of lymphangiography and percu-
thorax, whereas the literature was split between taneous microcatheters to deliver microcoils and
studies favoring conservative treatment and those glue to achieve thoracic duct occlusion, and thor-
supporting a surgical approach.44 No statistically acoscopic/robotic ligation all represent minimally
significant difference was noted in terms of mortal- invasive alternatives to highly morbid repeat
ity and between transhiatal and transthoracic thoracotomies.
techniques.44
Cerfolio4 reported a mortality of approximately REFERENCES
16% for patients undergoing surgical treatment
of chylothorax after esophagectomy, compared 1. Nair SK, Petko M, Hayward MP. Aetiology and man-
with an 80% mortality after conservative treat- agement of chylothorax in adults. Eur J Cardiothorac
ment: these data alone seem to argue in favor of Surg 2007;32:362–9.
early surgical repair of leaks of the thoracic 2. Wemyss-Holden SA, Launois B, Maddern GJ. Man-
duct.4,39 agement of thoracic duct injuries after oesophagec-
One controversial issue is the concept of pro- tomy. Br J Surg 2001;88:1442–8.
phylactic ligation of the duct commonly performed 3. Sieczka EM, Harvey JC. Early thoracic duct ligation
during esophagectomy.45,46 According to an anal- for postoperative chylothorax. J Surg Oncol 1996;
ysis of the literature performed in 2012, enough ev- 61:56–60.
idence supporting prophylactic ligation during 4. Cerfolio RJ. Chylothorax after esophagogastrec-
esophagectomy could be gathered.47 However, tomy. Thorac Surg Clin 2006;16:49–52.
recent evidence on 989 patients who underwent 5. Régnier C. Jean Pecquet (1622-1674) and his
prophylactic ligation of the thoracic duct clearly cistern. Rev Prat 1999;49:125–8.
Postoperative Chylothorax 527
6. Van Pernis PA. Variations of the thoracic duct. Sur- 25. Kalomenidis I. Octreotide and chylothorax. Curr
gery 1949;26:806–9. Opin Pulm Med 2006;12:264–7.
7. Schlierf C, Falor WH, Wood PD, et al. Composition of 26. Kelly RF, Shumway SJ. Conservative management of
human chyle chylomicrons following single fat feed- postoperative chylothorax using somatostatin. Ann
ings. Am J Clin Nutr 1969;22:79–86. Thorac Surg 2000;69:1944–5.
8. Paes ML, Powell H. Chylothorax: an update. Br J 27. Rocco G, Martucci N, La Manna C, et al. Ten-year
Hosp Med 1994;51:482–90. experience on 644 patients undergoing single-port
9. McGrath EE, Blades Z, Anderson PB. Chylothorax: (uniportal) video-assisted thoracoscopic surgery.
aetiology, diagnosis and therapeutic options. Respir Ann Thorac Surg 2013;96:434–8.
Med 2010;104:1–8. 28. Akin H, Olcmen A, Isgorucu O, et al. Approach to
10. Chalret du Rieu M, Baulieux J, Rode A, et al. Man- patients with chylothorax complicating pulmonary
agement of postoperative chylothorax. J Visc Surg resection. Thorac Cardiovasc Surg 2012;60:135–9.
2011;148:e346–52. 29. Shimizu K, Yoshida J, Nishimura M, et al. Treatment
11. Wasmuth-Pietzuch A, Hansmann M, Bartmann P, et al. strategy for chylothorax after pulmonary resection
Congenital chylothorax: lymphopenia and high risk of and lymph node dissection for lung cancer.
neonatal infections. Acta Paediatr 2004;93:220–4. J Thorac Cardiovasc Surg 2002;124:499–502.
12. Guo W, Zhao YP, Jiang YG, et al. Prevention of post- 30. Takuwa T, Yoshida J, Ono S, et al. Low-fat diet man-
operative chylothorax with thoracic duct ligation dur- agement strategy for chylothorax after pulmonary
ing video-assisted thoracoscopic esophagectomy resection and lymph node dissection for primary
for cancer. Surg Endosc 2012;26:1332–6. lung cancer. J Thorac Cardiovasc Surg 2013;146:
13. Cho HJ, Kim DK, Lee GD, et al. Chylothorax compli- 571–4.
cating pulmonary resection for lung cancer: effec- 31. Kawasaki R, Sugimoto K, Fujii M, et al. Therapeutic
tive management and pleurodesis. Ann Thorac effectiveness of diagnostic lymphangiography for
Surg 2014;97:408–13. refractory postoperative chylothorax and chylous
14. Le Pimpec-Barthes F, D’Attellis N, Dujon A, et al. ascites: correlation with radiologic findings and pre-
Chylothorax complicating pulmonary resection. ceding medical treatment. AJR Am J Roentgenol
Ann Thorac Surg 2002;73:1714–9. 2013;201:659–66.
15. Liu CY, Hsu PK, Huang CS, et al. Chylothorax 32. Itkin M, Kucharczuk JC, Kwak A, et al. Non operative
complicating video- assisted thoracoscopic surgery thoracic duct embolization for traumatic thoracic
for non-small cell lung cancer. World J Surg 2014; duct leak: experience in 109 patients. J Thorac Car-
38:2875–81. diovasc Surg 2010;139:584–9.
16. Dumont AE, Mayer DJ, Mulholland JH. The suppres- 33. Patterson GA, Todd TR, Delarue NC, et al. Supra-
sion of immunologic activity by diversion of thoracic diaphragmatic ligation of the thoracic duct in intrac-
lymph. Ann Surg 1964;160:373–83. table chylous fistula. Ann Thorac Surg 1981;32:44–9.
17. Bessone LN, Ferguson TB, Burford TH. Chylothorax. 34. Platis IE, Nwogu CE. Chylothorax. Thorac Surg Clin
Ann Thorac Surg 1971;12:527–50. 2006;16:209–14.
18. Agrawal V, Doelken P, Sahn SA. Pleural fluid analysis 35. Selle JG, Snyder WH 3rd, Schreiber JT. Chylothorax:
in chylous pleural effusion. Chest 2008;133:1436–41. indications for surgery. Ann Surg 1973;177:245–9.
19. Rahman NM, Chapman SJ, Davies RJ. Pleural effu- 36. Lampson RS. Traumatic chylothorax; a review of the
sion: a structured approach to care. Br Med Bull literature and report of a case treated by mediastinal
2005;72:31–47. ligation of the thoracic duct. J Thorac Surg 1948;17:
20. Boffa DJ, Sands MJ, Rice TW, et al. A critical evalu- 778–91.
ation of a percutaneous diagnostic and treatment 37. Shackcloth MJ, Poullis M, Lu J, et al. Preventing of
strategy for chylothorax after thoracic surgery. Eur chylothorax after oesophagectomy by routine pre-
J Cardiothorac Surg 2008;33:435–9. operative administration of oral cream. Eur J Cardio-
21. Lee EW, Shin JH, Ko HK, et al. Lymphangiography thorac Surg 2001;20:1035–6.
to treat postoperative lymphatic leakage: a technical 38. Inderbitzi RG, Krebs T, Stirneman T, et al. Treatment
review. Korean J Radiol 2014;15:724–32. of postoperative chylothorax by fibrin glue applica-
22. Vallieres E, Shamji FM, Todd TR. Postpneumonec- tion under thoracoscopic view with use of local
tomy chylothorax. Ann Thorac Surg 1993;55(4): anesthesia. J Thorac Cardiovasc Surg 1992;104:
1006–8. 209–10.
23. Riquet M, Assouad J, Le Pimpec Barthes F. 39. Rocco G. Chylothorax. In: Sabiston DC,
Traitment du chylothorax. EMC-Chirurgie 2004; Spencer EC, editors. Surgery of the chest. 8th edi-
6:662–81. tion. Philadelphia: WB Saunders; 2010. p. 427–30.
24. Al-Zubairy SA, Al-Jazairi AS. Octreotide as a thera- 40. Thompson KJ, Kernstine KH, Grannis FW Jr, et al.
peutic option for management of chylothorax. Ann Treatment of chylothorax by robotic thoracic duct
Pharmacother 2003;37:679–82. ligation. Ann Thorac Surg 2008;85:334–6.
528 Martucci et al
41. Kent RB 3rd, Pinson TW. Thoracoscopic ligation of the incidence of post-operative chylothorax? Int J
the thoracic duct. Surg Endosc 1993;7:52–3. Surg 2012;10:203–5.
42. Mason PF, Ragoowansi RH, Thorpe JA. Post-thora- 48. Azizkhan RG, Canfield J, Alford BA, et al. Pleuroper-
cotomy chylothorax - a cure in the abdomen? Eur itoneal shunts in the management of neonatal chylo-
J Cardiothorac Surg 1997;11:567–70. thorax. J Pediatr Surg 1983;18:842–50.
43. Shah RD, Luketich JD, Schuchert MJ, et al. Posteso- 49. Milsom JW, Kron IL, Rheuban KS, et al. Chylothorax:
phagectomy chylothorax: incidence, risk factors, an assessment of current surgical management.
and outcomes. Ann Thorac Surg 2012;93:897–903. J Thorac Cardiovasc Surg 1985;89:221–7.
50. Gupta D, Ross K, Piacentino V 3rd, et al. Use of
44. Kranzfelder M, Gertler R, Hapfelmeier A, et al. Chy-
LeVeen pleuroperitoneal shunt for refractory high-
lothorax after esophagectomy for cancer: impact of
volume chylothorax. Ann Thorac Surg 2004;78:
the surgical approach and neoadjuvant treatment:
e9–12.
systematic review and institutional analysis. Surg
51. Dietl B, Pfister K, Aufschläger C, et al. Radiotherapy
Endosc 2013;27:3530–8.
of inguinal lymphorrhea after vascular surgery. A
45. Hou X, Fu JH, Wang X, et al. Prophylactic thoracic retrospective analysis. Strahlenther Onkol 2005;
duct ligation has unfavorable impact on overall sur- 181:396–400 [in German].
vival in patients with resectable esophageal cancer. 52. Gerstein J, Kofahl-Krause D, Frühauf J, et al. Com-
Eur J Surg Oncol 2014;40:1756–62. plete remission of a lymphoma-associated chylo-
46. Lai FC, Chen L, Tu YR, et al. Prevention of chylo- thorax by radiotherapy of the celiac trunk and
thorax complicating extensive esophageal resection thoracic duct. Strahlenther Onkologie 2008;184:
by mass ligation of thoracic duct: a random control 484–7.
study. Ann Thorac Surg 2011;91:1770–4. 53. Sziklavari Z, Allgäuer M, Hübner G, et al. Radio-
47. Choh CT, Khan OA, Rychlik IJ, et al. Does ligation of therapy in the treatment of postoperative chylo-
the thoracic duct during oesophagectomy reduce thorax. J Cardiothorac Surg 2013;8:72.
Index
Note: Page numbers of article titles are in boldface type.