Professional Documents
Culture Documents
J Thorsurg 2018 07 001
J Thorsurg 2018 07 001
D i ffi c u l t H i a t a l H e r n i a
Matthew Rochefort, MDa, Jon O. Wee, MDb,*
KEYWORDS
Paraesophageal hernia Mesh Gastroplasty Fundoplication
KEY POINTS
Repair of the hernias requires assessment of the esophageal length and crural tension.
Shortened esophageal length can be address with aggressive esophageal mobilization, but may
require a lengthening procedure.
Mesh reinforcement of the hiatus can result in short term benefit, but may not improve long term
outcomes.
A hiatal hernia refers to the herniation of an intra- operative intervention only if they become symp-
abdominal organ or organs into the thoracic cavity tomatic, as in conjunction with an antireflux pro-
through the space between the left and right crus cedure. Type II-IV hiatal hernias are commonly
of the diaphragm. The vast majority of hiatal her- repaired once identified because of the concern
nias are referred to as type I or sliding hernias. In for acute gastric volvulus and catastrophic stran-
this situation, weakness of the phrenoesophageal gulation of the stomach,1,2 in addition to symptom
ligament allows herniation of the gastroesopha- correction, such as postprandial chest discomfort,
geal junction into the thoracic cavity thereby dysphagia, or reflux. Acute presentations can lead
bringing the cardia of the stomach above the dia- to the need for emergency detorsion and a well-
phragmatic hiatus. A type II hiatal hernia is a para- documented increased morbidity and mortality
esophageal hernia in which the gastroesophageal compared with elective laparoscopic repair. It is
junction is fixed and another portion of the stom- estimated that the risk of serious complication
ach herniates through the diaphragmatic hiatus due to the presence of a paraesophageal hiatal
into the chest alongside the esophagus. A type III hernia is approximately 1% per year.3 Type III-IV
hernia is a combination of type I and II whereby together have been referred to as “giant paraeso-
both the gastroesophageal junction and another phageal hernias,” whereby at least 30% of the
portion of the stomach have herniated through stomach is contained within the thoracic cavity,4
the diaphragmatic hiatus. A type IV hiatal hernia in- and the surgical management of this clinical situa-
dicates that an intra-abdominal organ in addition tion is the focus of this review.
to the stomach has herniated through the hiatus,
commonly the colon or the small bowel but may SURGICAL TREATMENT OF HIATAL HERNIAS
include the spleen or pancreas.
Type I or sliding hiatal hernias are becoming The history of operative repair of hiatal hernias be-
increasingly common, likely due to the increasing gins in the early twentieth century, when open
incidence of obesity; however, they usually require abdominal laparotomy incisions were used to
thoracic.theclinics.com
Disclosure: Dr M. Rochefort has nothing to disclose. Dr J.O. Wee is a consultant for Medtronic.
a
Division of Thoracic Surgery, Brigham & Women’s Hospital, 75 Francis Street, Boston, MA 02115, USA;
b
Esophageal Surgery, Division of Thoracic Surgery, Brigham & Women’s Hospital, 75 Francis Street, Boston,
MA 02115, USA
* Corresponding author.
E-mail address: jwee@bwh.harvard.edu
gain access to the esophageal hiatus. At that time, scarring and fibrosis secondary to repeated expo-
the openings in the diaphragmatic hiatus were sure of the distal esophagus to gastric con-
closed directly around the organs that naturally tents.4,11,12 With each swallow, the esophagus
passed through, namely the esophagus.5 During shortens and applies a constant force against the
the second half of the twentieth century, the focus hiatal opening.13 Radial tension is due to the
transitioned to procedures performed through the chronic dilation of the hiatal opening and the force
thoracic cavity whereby the hernia sac could be required to bring the edges of the crura back
reduced into the abdomen, the diaphragmatic together. The crural pillars are also usually quite
opening could be narrowed under direct vision, thin in large hernias and made of attenuated mus-
and extensive mobilization of the esophagus could cle fibers and not fascia that can lead to sutures
be performed. The subsequent development of pulling through when tied under tension.6,10
the fundoplication and the ability to treat reflux dis- To address these 2 forms of tension, the surgical
ease by wrapping the stomach around the distal repair of hiatal hernias via a laparoscopic
esophagus led to the broad incorporation of a fun- approach has some well-accepted tenets. These
doplication as an integral portion of repair of a par- include thorough preoperative testing, meticulous
aesophageal hernia.5 atraumatic technique, routine division of the short
The advent of minimally invasive surgery has gastric vessels, complete reduction and resection
dramatically increased the number of reflux and hi- of the hernia sac, circumferential dissection of the
atal hernia operations being performed. Laparos- esophagus up into the mediastinum to provide a
copy provides improved visualization of the 2-cm to 3-cm length of tension-free intra-abdom-
hiatus, dissection of the esophagus and the hernia inal esophagus, careful preservation of the vagus
sac can be performed well up into the medias- nerves, maintenance of the peritoneal lining
tinum under direct vision, and is associated with covering the crura, posterior primary crural closure
a significantly shorter length of stay, less need of over a bougie, and an associated gastric fundopli-
a nasogastric tube in place, less postoperative cation.4,6,11,12,14–18 Preoperative planning includes
pain, and decreased morbidity compared with a barium esophagram, upper endoscopy, and
open repairs.6 However laparoscopic repairs esophageal manometry, and for all patients with
have higher reported recurrence rates. In one high- a significant component of gastroesophageal
ly quoted study out of the University of Southern reflux, a pH probe.16 The lack of complete reduc-
California, the rate of radiologic recurrences tion of the hernia sac from the mediastinum is
following laparoscopic suture cruroplasty repair associated with increased frequency of postoper-
for paraesophageal hernias was 42%.7 Possible ative seroma formation and increased risk of
reasons include decreased adhesion generation recurrence of the hernia.1,16 The addition of the
following laparoscopy, less meticulous medias- fundoplication addresses the gastroesophageal
tinal dissection of the hernia sac, and less tactile reflux symptoms that most patients with giant par-
feedback and recognition of crural closures under aesophageal hernias present with and theoreti-
tension. As longer-term follow-up is available, it cally helps to anchor the gastroesophageal
has also been demonstrated that the prevalence junction below the level of the diaphragm.1 Poste-
of hernia recurrence increases over time with rior suture cruroplasty is the preferred method of
greater long-term follow-up.8,9 crural closure; however, occasional anterior su-
tures may be required to close the crural defect
RISK FACTOR FOR RECURRENCE without inducing undue angulation of the esoph-
agus as it passes through the hiatus.13,16 In rare
Some of the factors that contribute to the recur- circumstances in which the preoperative manom-
rence of hiatal hernias after repair include the dy- etry demonstrates a severe esophageal motility
namic nature of the esophageal hiatus, with disorder, more than 90% of failed peristalsis,
continuous movement secondary to respiratory then a partial fundoplication may be favored to
excursion, esophageal peristalsis, and cardiac ac- decrease the risk of postoperative dysphagia.16
tivity, and the pressure gradient between the peri- The addition of a fundoplication, although useful
toneum and the thorax that is heightened during as an antireflux procedure, has not been proven
coughing or sneezing.6,10 Any repair to a hernia to reduce the rate of reherniation following laparo-
defect in the hiatus is subject to 2 forms of tension. scopic hernia repair.19
Axial tension is generated due to displacement of
the gastroesophageal junction into the medias- PERMANENT MESH
tinum. This displacement could be in response to
an intrinsically shortened esophagus due to The advent of synthetic mesh and popularization
longitudinal collagen contraction resulting from of tension-free repair for inguinal and ventral
Management of the Difficult Hiatal Hernia 535
hernias with demonstration of reduced recurrence significant limitation in that the length of follow-
led to the application of synthetic mesh reinforce- up was significantly longer for patients with sim-
ment of hiatal hernia repairs. ple cruroplasty than for the mesh-reinforced re-
Frantzides and colleagues,20 from Rush, per- pairs. In contrast to the group from Rush, they
formed a randomized prospective study of mesh reported that only half of the radiographic recur-
repair versus suture crural closure on 72 patients rences were symptomatic, and that there was no
with hiatal hernias larger than 8 cm. In the pros- significant difference in the symptomatic
thetic mesh group, they used an onlay of oval outcome between the mesh repair and the sim-
mesh with a radial defect that was used to form ple cruroplasty group.
a “keyhole” around the esophagus. All of the pa- Despite the positive results from these studies,
tients received a 3-cm floppy Nissen fundoplica- there is still a significant hesitancy to place syn-
tion in additional to the crural closure. They thetic mesh during hiatal hernia repairs due to
found that at up to 6 years of follow-up there the potential for significant complications when
were no recurrences in the mesh group and there placed near the dynamic esophageal hiatus,
was a 22% recurrence rate in the simple cruro- including the possibility of erosion, ulceration,
plasty group, and that all of the patients with radio- stricture, and obstruction.6,7,9,10,16,17 In a review
graphic recurrences were symptomatic.20 They of the complications associated with permanent
did not identify any patients with either stricture mesh when placed at the hiatus, there were 6 pa-
or erosion following repair with polytetrafluoro- tients who required an esophagectomy and an
ethylene (PTFE), but they acknowledge that re- additional patient who required a gastrectomy to
pairs with polypropylene should be avoided treat the complications of mesh erosion into the
given the risk for erosion at the hiatus. They lumen.10 It is thought that erosions occur from
concluded that the addition of the mesh is respon- the constant movement of the esophagus against
sible for reducing the tension on the repair and the solid edge of the mesh, with the mesh cutting
therefore allows for improved healing. into the esophagus over time. This happens even
Granderath and colleagues,21 from Austria, per- more frequently when the mesh is used to bridge
formed a randomized control trial comparing sim- the crural defect, as this results in increased con-
ple cruroplasty following laparoscopic Nissen tact between the esophagus and the mesh.2,23
fundoplication to prosthetic closure with a poly- The other commonly reported complication was
propylene mesh. They similarly performed the the creation of a dense band of fibrosis
mesh repair as an onlay but did not include the around the implanted prosthesis, which results in
radial defect and did not encircle the esophagus significant dysphagia.10,18 These complications
with the mesh. The mesh was only placed poste- occurred over the course of months to years with
rior to the esophagus. They found that the rate of an average time to complication of 23 months.
intrathoracic wrap migration in the group repaired
with a prosthetic mesh was significantly less (1%) BIOLOGIC MESH
than the group with a simple cruroplasty (6%).
They did notice a significantly higher rate of The concern over the relatively high recurrence
dysphagia in the patients with the prosthetic rates associated with pure suture repair of the
wrap at the 3-month time point; however, at hiatus in conjunction with concern over potential
1-year follow-up, there was no longer a significant complications associated with a permanent
difference between the 2 groups. They surmised foreign material near the hiatus led to the investi-
that with the mesh placed in the posterior position, gation of biologic material reinforcement of the hi-
it does not come in contact with the esophagus at atus. Additional theoretic benefits of biologic mesh
all, but instead with the posterior portion of the include resistance to infection if placed in a
fundic wrap, thereby providing some buffer to contaminated field, decreased inflammatory reac-
esophageal erosions. tion, availability for neovascularization, and a
Similarly, Muller-Stitch and colleagues,22 from reconstruction that ultimately is made of natural
Switzerland, repaired 16 patients with a butterfly tissue.6,24 For patients with large hiatal defects
shaped mesh placed in the posterior location, that cannot easily be closed without tension, the
and they noted that the rate of radiographic use of a small intestine submucosa and non–
recurrence was reduced from 19% in the simple cross-linked human collagen matrix can be used
cruroplasty group to 0% in the mesh-reinforced to buttress the repair.16
group. The mesh used in these repairs was a Oelschlager and colleagues6 performed a multi-
combination of Prolene and Vicryl, which the center prospective randomized trial, of patients
group believed would reduce the potential for with type II-IV paraesophageal hernias, comparing
mesh-related complications. This study had a suture closure of the hiatus to suture closure of the
536 Rochefort & Wee
hiatus reinforced with a U-shaped piece of porcine of a keyhole reinforcement that is circumferential
small intestinal submucosa (Cook Surgical, Bloo- around the esophagus. As this scars over time, it
mington, IN). The mesh was used to buttress the may lead to increased dysphagia. With the
posterior hiatal closure and the limbs extended posterior-only reinforcement, there was no in-
anteriorly along either side of the esophagus, and crease in rates of dysphagia between those
was specifically not placed in a circumferential repaired with mesh and those not repaired with
manner to avoid postoperative dysphagia as the mesh at a follow-up of 6 months.17
mesh contracts over time. At 6 months, the pa- In a meta-analysis of 22 articles dealing with bio-
tients in the study underwent an upper gastroin- logic mesh reinforcement of hiatal hernia repairs,
testinal (GI) barium swallow that demonstrated a the 2 most commonly used mesh repairs were
decreased radiographic recurrence rate in those with SIS and human acellular cadaveric dermis
patients treated with the biologic mesh, from (HACD). These were most commonly used in a
24% in primary repair group to 9% in the biologic U-shape or pantaloon fashion placed posterior to
mesh group. The patients with radiographic recur- the esophagus with the limbs of the mesh encir-
rence were more likely to have chest pain and early cling the esophagus. The graft could be anchored
satiety but did not have significantly different rates with sutures, tacks, and fibrin glue. There is level
of heartburn or dysphagia according to symptom 1b data that support lower recurrence rate with
evaluation and quality of life.6 biologic mesh reinforcement over the short term;
The same group published their long-term re- however, with longer-term follow-up, there is level
sults at 5 years and interestingly there was no dif- 2b data that this benefit is lost.24
ference in recurrence rates between the simple
crural closure group and the biologic mesh group AUTOLOGOUS TISSUE REINFORCEMENT
at 5 years. The primary repair group had a recur-
rence rate of 59%, whereas the small intestine With the documented success of permanent mesh
submucosa group had a recurrence rate of 54% reinforcement of the hiatal repair in reducing recur-
(P 5 .7).9 In addition, there were no statistically sig- rences but the ultimate increase in potentially
nificant differences in frequency or severity of up- life-threatening complications, there has been
per GI symptoms or in quality of life measures investigation into the use of autologous tissue to
between the 2 groups, but there were significant reinforce the hiatus during hiatal hernia repair.
improvements in symptoms and quality of life One potential source of this autologous tissue
compared with the preoperative baseline.9 that has been recommended is the well-
AlloMax is an alternative to porcine small intes- vascularized falciform ligament. In the technique
tine submucosa (SIS) and is a sterile non–cross- described by the group from the University of
linked human collagen matrix that supports Louisville, the falciform ligament is mobilized
cellular in growth and revascularization. The group from its attachments to the anterior abdominal
out of the University of Southern California has wall, staying as close to the abdominal wall as
begun to regularly use it to reinforce their crural possible to protect the vascular supply of the
closure.2 They routinely use mesh reinforcement pedicle flap. The flap can then be delivered
for those with a hernia larger than 5 cm, patients beneath the left lobe of the liver and used to cover
with thin or atrophic crural pillars, and in all pa- the posterior cruroplasty as an onlay mesh rein-
tients undergoing reoperation.2 The group from forcement.14 The investigators report having per-
Swedish Medical Center similarly began using formed this procedure on 15 patients with no
AlloMax for all of their hiatal hernia repairs in significant complications or need for reintervention
2007. Initially the mesh was placed as a for the hiatal hernias.
posterior-only onlay reinforcement and evolved
over time to a U and then to a reverse C-shaped RELAXING INCISIONS IN THE DIAPHRAGM
mesh reinforcement, as there were a number of re-
currences noted anteriorly and to the left of the The radial tension created from the splaying of the
esophagus. They had no episodes of allograft crural fibers that leads to a paraesophageal hernia
erosion into the esophagus during the study, no contributes to a significant source of hernia recur-
evidence of persistent dysphagia, and minimal dif- rences. The effect of tension on hernia recurrence
ficulty with adhesions at reoperation.25 has been well demonstrated in both inguinal and
There is some concern for shrinkage of biologic ventral abdominal hernias, leading to the practice
mesh over time as the mesh is replaced with vas- of creating a relaxing incision to reduce some of
cularized tissue and undergoes scar formation. For the tension on the actual repair.13 This incision is
this reason, many investigators advocate a poste- made in an adjacent, but less critical area of fascia
rior-only17 or a U-shaped reinforcement8 instead or muscle, and allows the tissue in the area of
Management of the Difficult Hiatal Hernia 537
interest to come together with less tension.13 This shortened esophagus into 3 groups: those with
concept when extended to the crural separation of apparent short esophagus, when the esophagus
the diaphragm would suggest that a tension-free was actually normal in length but appeared short-
closure of the hiatus would result in the best ened secondary to displacement of the gastro-
long-term outcomes. Bradley and colleagues26 esophageal junction, those with a truly shortened
demonstrated that the width of the crural defect but reducible esophagus, and those with a short-
is not necessarily correlative with the degree of ened but irreducible esophagus. There is a subset
tension required to close the hernia; however, of patients who have significant enough esopha-
the shape of the defect did correlate. The oval- geal shortening that they will require an
shaped defects had significantly higher tension esophageal-lengthening procedure.11 There are
then defects that were slits or tear-dropped. One 5% to 8% of patients who have moderate esoph-
mechanism to achieve this tension-free closure is ageal shortening and can benefit from an
to perform a relaxing incision in the right dia- extended mediastinal dissection that can effec-
phragm when the crural pillars cannot be brought tively lengthen the esophagus such that the
together without undue tension.2 This relaxing inci- required 2 to 3 cm of tension-free intra-abdominal
sion is made just antero-lateral to the right crus, esophagus can be achieved.11,12 In the study by
parallel to the inferior vena cava, and usually Swanstrom and colleagues,15 looking only at pa-
does not need to be more than 1 to 2 cm in tients with preoperative indications placing them
length.13 Some investigators recommend entering at higher risk for a shortened esophagus, 30%
the ipsilateral pleural space before performing the had normal-length esophagus, 50% had short
relaxing incision to ensure that there is not any esophagus that could be lengthened with medias-
lung adhesed to the diaphragm in that area.13 If tinal dissection alone, and 20% had significant
the right crus is fibrotic or there is not adequate shortening requiring a Collis gastroplasty. This
distance between the hiatus and the inferior vena supports the general conclusion that the determi-
cava, then a left-sided relaxing incision can be per- nation of esophageal length and need for addi-
formed, which is created along the course of the tional lengthening procedures must be made by
seventh rib and not radially to avoid injury to the visualization in the operating room.12 By perform-
left-sided phrenic nerve.2,13 The left-sided incision ing an extended mediastinal dissection, which
often must be much longer to provide adequate was defined as more than 5 cm and often up to
tension-free closure of the hiatus. The diaphrag- 10 cm into the mediastinum along the esophagus,
matic defect created by the relaxing incision can in patients who had moderate esophageal short-
then be repaired with a permanent mesh patch, ening, it is possible to achieve comparable radio-
such as PTFE, and the crural closure reinforced graphic recurrence rates to patients with a
with a biologic mesh13 Some investigators have standard-length esophagus.11 This reserves Collis
described repairing right-sided relaxing incisions gastroplasty to only those patients in whom 2 to
with a U-shaped biologic mesh that incorporates 3 cm of tension-free esophagus cannot be ob-
the reinforcement of the hiatus and the closure of tained after extended dissection. The avoidance
the relaxing incision with a single mesh.16 These of the gastroplasty decreases the potential com-
different relaxing incisions can reduce the tension plications of the operation, namely gastric leak
on the crural closure by approximately 50%.26 and complications of ectopic gastric mucosa
within the neoesophagus.11 The addition of
EXTENDED TRANSMEDIASTINAL DISSECTION the extended transmediastinal dissection was
able be performed without any significant
Axial tension caused by a preoperative shortened added morbidity compared with the standard
esophagus is one of the primary causes of failure dissection.11
of repair of paraesophageal hernias. This short-
ening is thought to be secondary to collagen COLLIS GASTROPLASTY
contraction and fibrosis from exposure to gastric
contents refluxing into the distal esophagus.12 It is a violation of one of the tenants of surgery to
This shortening can lead to disruption of the perform a hernia repair under tension and there-
wrap or herniation of the wrap into the medias- fore is a similar violation to perform a fundoplica-
tinum, and can lead to chest pain, dysphagia, tion around an esophagus that is under tension.
and even strangulation.12 Patients who are at This will inherently expose the fundoplication
increased risk include those with large hernias, and the hiatus to axial tension, which can lead
larger than 5 cm, and those with paraesophageal to wrap herniation or wrap disruption.12 When
hernias.11 Horvath and colleagues12 grouped pa- an appropriate length of intra-abdominal esoph-
tients with preoperative imaging concerning for a agus cannot be achieved without tension, then
538 Rochefort & Wee