SEPTEMBER 7, 2021

VOLUME 78

NUMBER 10

Listen to this issue’s audio summary by JACC Editor-in-Chief Dr. Valentin Fuster.
 SEPTEMBER 7, 2021

VOLUME 78

NUMBER 10

CONTENTS

ORIGINAL INVESTIGATIONS Metoprolol in Critically Ill Patients With COVID-19 1001

Agustín Clemente-Moragón, Juan Martínez-Milla, Eduardo Oliver, Arnoldo Santos,
Javier Flandes, Iker Fernández, Lorena Rodríguez-González, Cristina Serrano del Castillo,
Ana-María Ioan, María López-Álvarez, Sandra Gómez-Talavera, Carlos Galán-Arriola,
Valentín Fuster, César Pérez-Calvo, Borja Ibáñez

Coronavirus disease-2019 (COVID-19) can progress to acute respiratory distress syndrome (ARDS).
Although there is a lack of specific therapies, the disease is known to involve alveolar infiltration by
activated neutrophils. Metoprolol ameliorates exacerbated inflammation in other clinical settings by
targeting neutrophils. The MADRID-COVID (Intravenous Metoprolol in Respiratory Distress Due to
COVID-19) randomized clinical trial tested the potential benefits of metoprolol repurposing for ARDS
in COVID-19 patients. Three days of intravenous metoprolol therapy (15 mg/day) significantly
reduced lung neutrophil infiltration, significantly improved oxygenation, and showed a trend toward
fewer days on invasive mechanical ventilation. Although further studies are needed, metoprolol is a
safe and cheap intervention that may be beneficial in severe COVID-19.

SEE ADDITIONAL CONTENT ONLINE

n EDITORIAL COMMENT
Beta-Blockers in the Critically Ill: Friend or Foe? 1012
Mourad H. Senussi

Risk of Cardiovascular Hospital Admission After Exposure to 1015

Fine Particulate Pollution
Yi Zhang, Runmei Ma, Jie Ban, Feng Lu, Moning Guo, Yu Zhong, Ning Jiang, Chen Chen,
Tiantian Li, Xiaoming Shi

The authors conducted a case-crossover study aimed at exploring the association between heavy fine
particulate matter, particulate matter with an aerodynamic diameter #2.5 mm (PM2.5), pollution
events and hospital admission for cardiovascular diseases. The OR associated with extremely heavy
pollution events lasting for 3 days or more for total cardiovascular disease was 1.085 (95% CI:
1.077-1.093). Heavy and extremely heavy PM2.5 pollution events resulted in substantial increased
hospital admission risk for cardiovascular disease. With higher PM2.5 concentration and longer
duration of heavy PM2.5 pollution events, a greater risk of cardiovascular hospital admission was
observed. Health protection measures should be implemented against the heavy and extremely
heavy PM2.5 pollution events.

SEE ADDITIONAL CONTENT ONLINE

n EDITORIAL COMMENT
How Tweets Can Cause a Revolution 1025
Sanjay Rajagopalan, Robert D. Brook, Sadeer Al-Kindi

JACC CME/MOC/ECME is available online.

Go to http://www.acc.org/jacc-journals-cme.

Articles have accompanying

audio accessible online at
www.onlineJACC.org.
CONTENTS
SEPTEMBER 7, 2021 VOLUME 78, NUMBER 10

Effect of Heart Rate on the Outcome of Renal Denervation in Patients With 1028
Uncontrolled Hypertension
Michael Böhm, Konstantinos Tsioufis, David E. Kandzari, Kazuomi Kario,
Michael A. Weber, Roland E. Schmieder, Raymond R. Townsend,
Saarraaken Kulenthiran, Christian Ukena, Stuart Pocock, Sebastian Ewen,
Joachim Weil, Martin Fahy, Felix Mahfoud

The authors evaluated the association between baseline heart rate and blood pressure reduction
following renal denervation in the absence of antihypertensive medications in the SPYRAL HTN-OFF
MED Pivotal trial. Patients with office systolic blood pressure (SBP) 150-180 mm Hg were
randomized 1:1 to renal denervation (RDN) or sham control and separated into groups with baseline
office heart rate <70 or $70 beats/min. Treatment differences at 3 months between RDN and sham
control groups with baseline heart rate $70 beats/min for mean office, 24-hour, daytime, and
nighttime SBP were greater compared with patients with baseline heart rate <70 beats/min.

SEE ADDITIONAL CONTENT ONLINE

n EDITORIAL COMMENT
Baseline Heart Rate Predicts the Blood Pressure Response to Renal Denervation in 1039
Untreated Hypertension
Brent M. Egan

Compression-Only Versus Rescue-Breathing Cardiopulmonary Resuscitation After 1042

Pediatric Out-of-Hospital Cardiac Arrest
Maryam Y. Naim, Heather M. Griffis, Robert A. Berg, Richard N. Bradley, Rita V. Burke,
David Markenson, Bryan F. McNally, Vinay M. Nadkarni, Lihai Song, Kimberly Vellano,
Victoria Vetter, Joseph W. Rossano

In these analyses of pediatric out-of-hospital cardiac arrests (OHCA), compression only-

cardiopulmonary resuscitation (CO-CPR) was the most common type of bystander CPR in the
United States. Rescue breathing-CPR (RB-CPR) and CO-CPR were associated with better outcomes
compared with no bystander-CPR (NO-CPR), RB-CPR had superior outcomes compared with CO-
CPR. In age-stratified analyses, RB-CPR was associated with better outcomes compared with NO-
CPR in all age groups. CO-CPR was associated with better outcomes compared with NO-CPR in
children and adolescents, but not in infants. These results support current guidelines for RB-CPR
for pediatric OHCA.

SEE ADDITIONAL CONTENT ONLINE

n EDITORIAL COMMENT
Chest Compression-Only Cardiopulmonary Resuscitation in Pediatric 1053
Out-of-Hospital Cardiac Arrest: (Don’t) Take My Breath Away
Gene Yong-Kwang Ong
 CONTENTS
SEPTEMBER 7, 2021 VOLUME 78, NUMBER 10

THE PRESENT
AND FUTURE

JACC STATE-OF-THE- Chronic Fatigue Syndrome and Cardiovascular Disease: JACC State-of-the-Art Review 1056
ART REVIEW Benjamin H. Natelson, Danielle L. Brunjes, Donna Mancini

Myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS) is an unexplained illness characterized

by severe fatigue limiting normal daily activities for at least 6 months. Although rarely considered to
have cardiac dysfunction, ME/CFS patients frequently have reduced stroke volume and a significant
inverse relation between cardiac output and severity of postexertional malaise (PEM)—a prominent
symptom of ME/CFS. Magnetic resonance imaging of ME/CFS patients compared with normal
control subjects found significantly reduced stroke volume and cardiac mass. Autonomic
dysfunction is frequently observed with postural orthostatic tachycardia and/or hypocapnia. Two
cardiopulmonary stress tests done 24 hours apart may provide metabolic data substantiating PEM.

JACC REVIEW TOPIC Myocardial Viability Assessment Before Surgical Revascularization in 1068
OF THE WEEK Ischemic Cardiomyopathy: JACC Review Topic of the Week
Julio A. Panza, Lukasz Chrzanowski, Robert O. Bonow

In ischemic cardiomyopathy, noncontractile myocardium may be viable and improve function with
surgical revascularization. Although observational studies reported that only patients with
substantial amounts of viability had better outcomes following revascularization, prospective trials
have not confirmed such an interaction between myocardial viability and benefit from this
procedure. In addition, recovery of left ventricular function is not the principal mechanism by
which revascularization improves prognosis. A more contemporary application of viability testing
therefore rests on the alternative concept that the main goal of surgical revascularization is to
prevent further damage by protecting the residual viable myocardium from subsequent acute
coronary events.

CARDIOVASCULAR Cardiovascular Disease Fellowship Interviews 2021: An Evolution in Process 1078

MEDICINE AND SOCIETY Lisa J. Rose-Jones, Mustafa M. Ahmed, Benjamin H. Freed, Andrew M. Kates,
Meera Kondapaneni, Jonathan R. Salik, Victor Soukoulis, Helga Van Herle,
Gaby Weissman

LETTERS ONLINE FEATURE Sudden Death: Thinking Beyond the Heart e61
Susan K. Keen, Ross J. Simpson, Jr.

ONLINE FEATURE Reply

Zian H. Tseng, Santo Ricceri, James W. Salazar

THESE ARTICLES ONLY APPEAR IN THE ONLINE VERSION OF THE ISSUE.

 JOURNAL OF THE AMERICAN COLLEGE OF CARDIOLOGY VOL. 78, NO. 10, 2021

ª 2021 THE AUTHORS. PUBLISHED BY ELSEVIER ON BEHALF OF THE AMERICAN

COLLEGE OF CARDIOLOGY FOUNDATION. THIS IS AN OPEN ACCESS ARTICLE UNDER

THE CC BY-NC-ND LICENSE (http://creativecommons.org/licenses/by-nc-nd/4.0/).

ORIGINAL INVESTIGATIONS

Metoprolol in Critically Ill Patients

With COVID-19
Agustín Clemente-Moragón, BSC,a,* Juan Martínez-Milla, MD, PHD,a,b,* Eduardo Oliver, PHD,a,c
Arnoldo Santos, MD, PHD,d,e Javier Flandes, MD, PHD,f Iker Fernández, MD,f Lorena Rodríguez-González, TECH,g,h
Cristina Serrano del Castillo, MD,i Ana-María Ioan, MD,d María López-Álvarez, RN,b,c Sandra Gómez-Talavera, MD,a,b,c
Carlos Galán-Arriola, DVM, PHD,a,c Valentín Fuster, MD, PHD,a,j César Pérez-Calvo, MD, PHD,d
Borja Ibáñez, MD, PHDa,b,c

ABSTRACT

BACKGROUND Severe coronavirus disease-2019 (COVID-19) can progress to an acute respiratory distress syndrome
(ARDS), which involves alveolar infiltration by activated neutrophils. The beta-blocker metoprolol has been shown to
ameliorate exacerbated inflammation in the myocardial infarction setting.

OBJECTIVES The purpose of this study was to evaluate the effects of metoprolol on alveolar inflammation and on
respiratory function in patients with COVID-19–associated ARDS.

METHODS A total of 20 COVID-19 patients with ARDS on invasive mechanical ventilation were randomized to meto-
prolol (15 mg daily for 3 days) or control (no treatment). All patients underwent bronchoalveolar lavage (BAL) before and
after metoprolol/control. The safety of metoprolol administration was evaluated by invasive hemodynamic and elec-
trocardiogram monitoring and echocardiography.

RESULTS Metoprolol administration was without side effects. At baseline, neutrophil content in BAL did not differ
between groups. Conversely, patients randomized to metoprolol had significantly fewer neutrophils in BAL on day 4
(median: 14.3 neutrophils/ml [Q1, Q3: 4.63, 265 neutrophils/ml] vs median: 397 neutrophils/ml [Q1, Q3: 222, 1,346 neu-
trophils/ml] in the metoprolol and control groups, respectively; P ¼ 0.016). Metoprolol also reduced neutrophil extra-
cellular traps content and other markers of lung inflammation. Oxygenation (PaO2:FiO2) significantly improved after
3 days of metoprolol treatment (median: 130 [Q1, Q3: 110, 162] vs median: 267 [Q1, Q3: 199, 298] at baseline and day 4,
respectively; P ¼ 0.003), whereas it remained unchanged in control subjects. Metoprolol-treated patients spent fewer
days on invasive mechanical ventilation than those in the control group (15.5  7.6 vs 21.9  12.6 days; P ¼ 0.17).

CONCLUSIONS In this pilot trial, intravenous metoprolol administration to patients with COVID-19–associated
ARDS was safe, reduced exacerbated lung inflammation, and improved oxygenation. Repurposing metoprolol for
COVID-19–associated ARDS appears to be a safe and inexpensive strategy that can alleviate the burden of the COVID-19
pandemic. (J Am Coll Cardiol 2021;78:1001–1011) © 2021 The Authors. Published by Elsevier on behalf of the
American College of Cardiology Foundation. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).

Listen to this manuscript’s

audio summary by
From the aCentro Nacional de Investigaciones Cardiovasculares (CNIC), Madrid, Spain; bCardiology Department, IIS-Fundación
Editor-in-Chief
Jiménez Díaz University Hospital, Madrid, Spain; cCIBER de Enfermedades Cardiovasculares, Madrid, Spain; dIntensive Care Unit,
Dr. Valentin Fuster on
e
IIS-Fundación Jiménez Díaz University Hospital, Madrid, Spain; CIBER de Enfermedades Respiratorias, Madrid, Spain;
JACC.org.
f
Department of Pulmonary Medicine, IIS-Fundación Jiménez Díaz University Hospital, Madrid, Spain; gPathology Department, IIS-
Fundación Jiménez Díaz University Hospital, Madrid, Spain; hBiobank Patform-PT20/00141, IIS-Fundación Jiménez Díaz Hospital,
Madrid, Spain; iFlow Citometry Laboratory, IIS-Fundación Jiménez Díaz University Hospital, Madrid, Spain; and the jCardio-
vascular Institute, Icahn School of Medicine at Mount Sinai, New York, New York, USA. *Drs Clemente-Moragón and Martínez-
Milla contributed equally to this work.
Christie Ballantyne, MD, served as Guest Editor-in-Chief for this paper.

ISSN 0735-1097 https://doi.org/10.1016/j.jacc.2021.07.003

1002 Clemente-Moragón et al.
MADRID-COVID Pilot Trial
C
ABBREVIATIONS oronavirus disease-2019 (COVID-19),
AND ACRONYMS caused by severe acute respiratory
syndrome-coronavirus-2 (SARS-CoV-2)
ARDS = acute respiratory
distress syndrome
infection, is an ongoing pandemic affecting
more than 145 million people worldwide
BAL = bronchoalveolar lavage
and responsible for more than 3 million
COVID-19 = coronavirus
disease-2019 deaths to date. An estimated 6%-18% of
ICU = intensive care unit
COVID-19 cases progress to an acute respira-
tory distress syndrome (ARDS) requiring
IMV = invasive mechanical
ventilation intensive care unit (ICU) admission and inva-
NET = neutrophil extracellular sive mechanical ventilation (IMV) (1). There
trap is currently a lack of specific therapies for
COVID-19–associated ARDS.
SEE PAGE 1012
In the early stages of SARS-CoV-2 infection, the
host immune system is activated to block disease
progression. However, in some cases rapid replication
of SARS-CoV-2 in the respiratory tract triggers an
exacerbated inflammatory response and a cytokine
storm (2). This situation leads to progression to ARDS
together with other clinical complications, such as
septic shock, microthrombi, coagulopathy, and mul-
tiple organ dysfunction (3).
ARDS of different etiologies (4), including SARS-
CoV-2 infection (5,6), is highly dependent on the ac-
tion of neutrophils. Activated neutrophils contribute
to alveolar injury by releasing prestored inflamma-
tory mediators (reactive oxygen species and myelo-
peroxidase [MPO]) and by interacting with other
cells, such as platelets, to induce microthrombi. In
addition, the formation of neutrophil extracellular
traps (NETs) and highly injurious histones activates
the inflammasome and triggers the release of pro-
inflammatory cytokines (7). NETs released from
alveolar-infiltrated activated neutrophils increase
pulmonary inflammation and serum levels of proin-
flammatory cytokines, leading to extensive lung
damage and microthrombotic events in COVID-19
patients (2,3,8,9).
Despite the massive worldwide impact of
COVID-19, there is a shortage of effective therapies to
prevent transition from moderate to severe disease
and to improve prognosis. Given the intense pressure
COVID-19 is placing on ICUs worldwide, there is an
urgent need to identify therapies to reduce the
number of days in the ICU. The most sought-after
interventions are those able to mitigate COVID-19–
The authors attest they are in compliance with human studies committees and animal welfare regulations of the authors’
institutions and Food and Drug Administration guidelines, including patient consent where appropriate. For more information,
visit the Author Center.
Manuscript received April 28, 2021; revised manuscript received June 17, 2021, accepted July 1, 2021.
JACC VOL. 78, NO. 10, 2021
SEPTEMBER 7, 2021:1001–1011
associated immune dysregulation (10). An attractive
candidate approach is to use host-directed therapies,
which have emerged in recent years as an adjuvant
strategy to limit damage during infectious or sterile
exacerbated inflammation.
Beta-adrenergic receptor antagonists (b-blockers)
have been used for many decades to treat cardiovas-
cular conditions such as hypertension, arrhythmias,
and myocardial infarction (11). Observational retro-
spective studies have established a link between
b -blocker therapy and increased survival in critically
ill patients caused by different conditions, such as
sepsis (12-14), acute respiratory failure (15), severe
traumatic brain injury (16,17), and others (18,19).
Recent findings show that the b 1-selective blocker
metoprolol has a direct effect on neutrophils, damp-
ening their deleterious effects during exacerbated
inflammation (20). In the context of ischemia/
reperfusion (acute myocardial infarction), metoprolol
targeting of neutrophils has been shown to have a
strong cardioprotective effect, both in animal models
and in patients (20-23). More recently, our group
demonstrated that metoprolol (but not other
clinically available intravenous b -blockers) abrogates
neutrophil-driven exacerbated inflammation,
neutrophil-platelet interaction, and NETs formation
in a mouse model of LPS-induced acute lung injury
(24). These experimental data prompted us to inves-
tigate whether treatment with intravenous (IV)
metoprolol could ameliorate lung inflammation—and
eventually improve prognosis—in patients with
COVID-19–associated ARDS.
METHODS
STUDY DESIGN AND POPULATION. The MADRID-
COVID (Intravenous Metoprolol in Respiratory
Distress Due to COVID-19) pilot trial was approved by
the Fundación Jiménez Díaz University Hospital
ethics committee (Eudract registry number 2020-
002310-41). All patients, or a close relative, gave
written consent to participate. Inclusion criteria were
age 18-80 years, rt-PCR–confirmed SARS-CoV-2
infection (in either nasal swab or bronchoalveolar
lavage), invasive mechanical ventilation #72 hours,
heart rate $60 beats/min, and invasive systolic blood
pressure $120 mm Hg. Exclusion criteria included
prolonged hospital admission (>5 days) before
JACC VOL. 78, NO. 10, 2021
SEPTEMBER 7, 2021:1001–1011
enrollment, concomitant acute heart failure, left
ventricular ejection fraction <50%, right ventricular
systolic dysfunction, concomitant pulmonary embo-
lism, moderate-severe peripheral artery disease,
moderate-severe valvular heart disease, moderate-
severe COPD, or active treatment with b -blockers
before enrollment. A total of 20 patients with ARDS
secondary to SARS-CoV-2 infection under IMV were
enrolled and randomized to IV metoprolol tartrate
(Recordati) (3  5 mg boluses, 2 minutes apart, daily
for 3 days; n ¼ 12) or control (no treatment; n ¼ 8).
Two minutes after each bolus, blood pressure and
heart rate were measured, and if they were above the
limits set, the next bolus was injected.
Randomization was stratified by age (#59 years vs
>59 years), history of hypertension (yes/no), and
circulating neutrophil counts (<6,000 vs $6,000).
Bronchoalveolar lavage (BAL) fluid and blood samples
were obtained from patients at randomization (base-
line) and 24 hours after the third metoprolol dose/
control (day 4). The main study goal was to assess the
effect of metoprolol on inflammatory markers (mainly
neutrophil infiltration and NETs). The main secondary
goals were to assess the effect of metoprolol on days on
invasive mechanical ventilation and days in the ICU
after randomization, as well as pulmonary function.
The main safety outcome measure was hemodynamic
complications (cardiogenic shock, severe hypoten-
sion, or severe bradycardia/atrioventricular block).
Because this was a pilot trial, sample size was
calculated based on the capacity of identifying
changes in lung inflammation (neutrophil infiltra-
tion). Based on previous experimental studies, we
speculated that 20 patients would be enough to
detect a significant biological effect of metoprolol in
this context.
FLOW CYTOMETRY OF BAL SAMPLES. For flow
cytometry (FCM) studies, BAL samples (8 mL) were
previously inactivated with 2 mL of a cellular antigen
stabilization reagent containing formaldehyde
(TransFix, Cytomark Ltd). Samples were then centri-
fuged (5 minutes at 540g), the supernatant discarded,
and the cell pellet resuspended in 200 m L phosphate-
buffered saline. Afterwards, 100 m L of cell suspension
was stained for 15 minutes at room temperature with
the following color combination: antihuman CD15-
fluorescein isothiocyanate, CD33-phycoerythrin, and
CD3-V-450 and CD45-V-500 (Becton Dickinson Bio-
sciences). After staining, 2 mL of FACS lysing solution
(Becton/Dickinson Biosciences) was added, and after
5 minutes incubation, the sample was centrifuged
and resuspended in 100 m L phosphate-buffered sa-
line. Before acquisition, the fluorescent dye DRAQ5
Clemente-Moragón et al. 1003
MADRID-COVID Pilot Trial
(Biostatus Limited) (25,26) and Perfect-COUNT
microspheres (Cytognos SL) (27) were added for the
selection of DNA-positive cells and cell count,
respectively. Samples were run on a FACSCanto II
flow cytometer (Becton Dickinson Biosciences)
equipped with FACSDiva software (Becton Dickinson
Biosciences), and information was acquired about all
events corresponding to nucleated cells present in
the stained sample aliquot.
Data were analyzed with INFINICYT software
(Cytognos SL). FCM analysis included a first-step
identification of nucleated cells by DRAQ5 staining.
Leukocyte populations were identified with a gating
strategy based on forward scatter, side scatter, and
CD45 expression. Neutrophils and macrophages were
identified from their relatively higher light-scattering
properties, their unique pattern of CD45 expression,
and the expression of CD15 (neutrophils) and CD33
(alveolar macrophages). Lymphocytes were also
identified according to their CD45 expression and
forward and side scatter properties. Neutrophil,
macrophage, and lymphocyte populations were
quantified as the percentage of total CD45 events.
CHEMOKINE ELISA ASSAYS. Samples were inacti-
vated by incubation in a final concentration of 0.2%
SDS per 0.1% Tween-20 and heat treatment at
60 C for 15 minutes. Plasma and cell-free BAL sam-
ples were analyzed with human ELISA kits for von
Willebrand factor (RAB0556-1KT, Sigma) and the
chemokines monocyte chemoattractant protein
(MCP)-1 (orb315028, Biorbyt), interleukin (IL)-6
(orb219452, Biorbyt), and IL-8 (orb315028, Biorbyt).
NETosis MARKERS. A total of 3 NETosis biomarkers
were measured: citrullinated histone-3 (Cit-H3),
MPO-DNA complexes, and cell-free DNA. For Cit-H3
and MPO-DNA ELISA, samples were first inactivated
by suspension in 0.2% SDS per 0.1% Tween-20 and
heat treatment at 60  C for 15 minutes. For cell-free
DNA measurement, samples were inactivated by
heat treatment at 60 C for 1 hour.
Cit-H3 was measured with an ELISA kit (clone 11D3,
Cayman, 501620). Quantification of MPO-DNA com-
plexes was based on a previously described protocol
(28,29) that uses several reagents from the Cell Death
Detection ELISA Kit (Roche, 11544675001) but in-
cludes a high-binding EIA/RIA 96-well plate differ-
ently coated overnight at 4  C with antihuman MPO
antibody (Bio-Rad, 0400-0002). Cell-free DNA was
measured using the Quant-iT PicoGreen dsDNA Assay
Kit (Invitrogen, Thermo Fisher Scientific, P11496).
NEUTROPHIL AND NET VISUALIZATION IN BAL. NETs
were visualized by Giemsa staining of BAL samples
1004 Clemente-Moragón et al.
MADRID-COVID Pilot Trial
(30,31). BAL samples were centrifuged for 10 minutes
at 2,500 revolutions/min. The pellet was resuspended
and spread for staining with Giemsa solution. Sam-
ples were then inactivated and fixed for 10 minutes at
room temperature with an alcohol-based spray fixa-
tive for cyto-diagnosis (M-Fix spray fixative). For
image analysis, fixed samples were digitalized with a
scanner (Nanozoomer-RS C110730, Hamamatsu) and
analyzed using NDP view image analysis soft-
ware (Hamamatsu).
STATISTICAL ANALYSIS. Data were analyzed with
Graphpad Prism version 8.4 and RStudio. Due to the
small sample size, all distributions were considered
non-normal, and nonparametric tests were applied
for statistical analyses. Paired comparisons between
pretreatment and post-treatment samples (basal and
4 days) were by Wilcoxon matched pairs signed rank
test. Comparisons between treatment conditions
(vehicle vs metoprolol) at baseline or after treatment
were made by unpaired Mann-Whitney U test. For
hemodynamics and functional parameters during
metoprolol administration, differences at baseline or
pre-post boluses were calculated by the nonpara-
metric chi-square Friedman test with correction by
the Durbin-Conover test for pairwise comparisons.
For categorical data, percentages were compared by
exact methods. Differences were deemed statistically
significant at P values below 0.05.
RESULTS
PATIENT CHARACTERISTICS. Between October 19,
2020, and January 19, 2021, a total of 20 patients were
enrolled; 12 were randomized to metoprolol and 8 to
control. There were no between-group differences in
baseline characteristics (Table 1). All patients were
treated during ICU admission with corticosteroids
(dexamethasone 6 mg daily), anticoagulants, mela-
tonin, and acetylcysteine. Before enrollment in the
trial, all patients (except 1 in the metoprolol group)
were treated with bolus and maintenance dose of
corticosteroids (methylprednisolone and/or dexa-
methasone) in the ward before admission to the ICU
without differences between groups.
Of the patients randomized to metoprolol, 11
received all scheduled IV doses (15 mg daily for
3 days). The remaining patient received 15 mg of
metoprolol on the first 2 days but not the third
because the heart rate was <50 beats/min caused by
intensified sedation (initiation of propofol). BAL was
conducted without complications in all patients
before and 24 hours after treatment. Clinical
JACC VOL. 78, NO. 10, 2021
SEPTEMBER 7, 2021:1001–1011
laboratory analyses at baseline and after treatment
are presented in Supplemental Table 1.
CARDIOVASCULAR SAFETY OF INTRAVENOUS
METOPROLOL ADMINISTRATION TO ARDS PATIENTS
ON MECHANICAL VENTILATION. Administration of
IV b-blockers has largely been proven to be
safe except for patients with acute pump failure.
Given the cardiovascular effects of metoprolol,
patients were monitored invasively and by echocar-
diography before and on every day after metoprolol
injection/control. As expected, metoprolol signifi-
cantly reduced heart rate (P < 0.01) and invasively
measured systolic blood pressure (P < 0.05),
although both remained within the physiological
range (Supplemental Table 2). Echocardiography
showed no deterioration of cardiac function parame-
ters after metoprolol treatment (Supplemental
Table 3). Overall, metoprolol intravenous adminis-
tration was shown to be safe and without side effects
in severe COVID-19 patients with ARDS on IMV.
METOPROLOL ADMINISTRATION ATTENUATES
NEUTROPHIL-DRIVEN LUNG EXACERBATED
INFLAMMATION. To assess the ability of metoprolol
to ameliorate neutrophil-mediated exacerbated lung
inflammation, we analyzed leukocyte populations in
BAL samples by flow cytometry at baseline and on
day 4. At baseline, the metoprolol and control groups
showed no differences in BAL neutrophil content
(Supplemental Figure 1). In contrast, on day 4 (after
3 days of metoprolol/control treatment), neutrophil
content was significantly lower in BAL from patients
in the metoprolol group than in those randomized to
control (median: 14.3 neutrophils/m l [Q1, Q3: 4.63, 265
neutrophils/m l] vs median: 397 neutrophils/ ml [Q1, Q3:
222, 1,346 neutrophils/ m l]; P ¼ 0.016). Day 4 BAL from
metoprolol-treated patients also had lower total
inflammatory-cell content and lower monocyte/
macrophage content, whereas lymphocytes did not
differ between groups (Figure 1A). We further
explored the impact of metoprolol on MCP-1 in BAL,
because this chemokine has been shown to promote
pulmonary fibrosis in late-stage ARDS (32,33). MCP-1
in cell-free BAL was significantly attenuated after
3 days of metoprolol treatment (median: 298 pg/mL
[Q1, Q3: 236, 350 pg/mL] vs median: 203 pg/mL [Q1,
Q3: 175, 258 pg/mL] for baseline and day 4, respec-
tively; P ¼ 0.009), whereas it remained unchanged in
control patients (Figure 1B). Conversely, changes in
IL-8 and -6 in cell-free BAL did not differ between
treatment groups (Supplemental Figure 2).
Excessive neutrophil activation in the lungs is
associated with NET formation and the release of
JACC VOL. 78, NO. 10, 2021 Clemente-Moragón et al. 1005
SEPTEMBER 7, 2021:1001–1011 MADRID-COVID Pilot Trial

T A B L E 1 Patient Characteristics at Randomization

All Metoprolol Control P Value

Age, y 60 (53.8, 68) 60 (57.8, 68.5) 58.5 (43.3, 65.8) 0.354

Male 13 (65.0) 8 (66.7) 5 (62.5) 1.000
BMI, kg/m2 27.1 (25.3, 31.1) 26.8 (25.1, 30.4) 27.1 (26.2, 31.5) 0.422
Comorbidities
Hypertension 6 (30.0) 4 (33.3) 2 (25.0) 1.000
Diabetes 2 (10.0) 2 (16.7) 0 (0.0) 0.648
Smokers 3 (15.0) 1 (8.3) 2 (25.0) 0.701
Dyslipidemia 6 (30.0) 4 (33.3) 2 (25.0) 1.000
Previous treatment
RAS inhibitors 5 (25.0) 3 (25.0) 2 (25.0) 1.000
Anticoagulants 0 (0.0) 0 (0.0) 0 (0.0) 1.000

Values are median (Q1, Q3) or n (%).

BMI ¼ body mass index; RAS ¼ renin-angiotensin system.

reactive oxygen species and proteolytic enzymes,
which can drive severe epithelial and endothelial
injury (2,34). To study whether the inflammation-
disrupting effect of metoprolol reduced the produc-
tion of these neutrophil activation byproducts,
we measured the NETosis markers Cit-H3 and
MPO-DNA complexes. Levels of both markers were
decreased in day 4 BAL from metoprolol-treated
patients (P ¼ 0.005 and P ¼ 0.086 vs baseline,
respectively), whereas no changes were observed in
BAL from control patients (Figure 1C). Lower NET
formation and inflammatory content in the meto-
prolol group was confirmed by Giemsa staining
(Figure 1D). We found no differences in cell-free DNA
content (Supplemental Figure 3), probably reflecting
its nonspecific nature as a NETosis biomarker (3).
To determine if attenuated immune-cell infiltra-
tion in the lungs was associated with a systemic
effect, we assessed changes in circulating levels of
chemokines known to be markedly elevated in severe
COVID-19 patients (34). The 3-day treatment with
metoprolol was associated with a significant reduc-
tion in the circulating concentrations of the pro-
inflammatory cytokine IL-8 (median: 94.4 pg/mL
[Q1, Q3: 72.1, 168 pg/mL] vs median: 80.1 pg/mL
[Q1, Q3: 69.5, 85.2 pg/mL] for baseline and day 4,
respectively; P ¼ 0.003), whereas no changes were
observed in controls (Supplemental Figure 4). Meto-
prolol had no significant effect on circulating levels
of IL-6 (Supplemental Figure 4) or NETosis markers
(Supplemental Figure 5).
METOPROLOL TREATMENT IMPROVES
FUNCTION. Oxygenation was measured as the ratio
between arterial oxygen partial pressure and frac-
tional inspired oxygen (PaO 2:FiO 2). Baseline and post-
treatment oxygenation parameters are shown in
Table 2. At baseline, oxygenation was worse in pa-
tients randomized to metoprolol than in the control
group, despite higher FiO 2. After the 3-day metoprolol
treatment, PaO 2 significantly improved (median:
87.5 mm Hg [Q1, Q3: 78.8, 110 mm Hg] vs median:
108 mm Hg [Q1, Q3: 98.3, 139 mm Hg] for baseline and
day 4, respectively; P ¼ 0.017). Metoprolol treatment
also significantly improved PaO 2:FiO 2 (median: 130
[Q1, Q3: 110, 162] vs median: 267 [Q1, Q3: 199, 298] at
baseline and day 4, respectively; P ¼ 0.007).
Conversely, in control subjects, PaO 2 and PaO 2:FiO 2
both deteriorated, although the change did not reach
statistical significance (P ¼ 0.107 and P ¼ 0.363 vs
baseline, respectively) (Figures 2A and 2B).
Patients randomized to metoprolol spent fewer
days on mechanical ventilation, although this differ-
ence did not reach statistical significance (15.5 
7.6 days vs 21.9  12.6 days in the metoprolol and
control groups, respectively; P ¼ 0.17). A similar trend
was observed for days of ICU admission after enroll-
ment (14.5  7.2 days vs 21.4  13.4 days in the
metoprolol and control groups, respectively; P ¼ 0.15)
(Figure 2C). All patients were discharged from the ICU,
and 1 patient in each group died before hospital
discharge.
DISCUSSION
The COVID-19 pandemic and associated ARDS is
placing an immense burden on health care systems.
In addition to high mortality, COVID-19–associated
ARDS results in prolonged ICU admission, contrib-
RESPIRATORY uting to morbidity among survivors and high hospital
expenditure. The current approach with these
patients is mainly based on protective IMV (35,36),
which ensures sufficient gas exchange while causing
minimal alveolar damage. With the exception of
1006 Clemente-Moragón et al. JACC VOL. 78, NO. 10, 2021

MADRID-COVID Pilot Trial SEPTEMBER 7, 2021:1001–1011

F I G U R E 1 Metoprolol Disrupts COVID-19–Associated Exacerbated Lung Inflammation

A Cells Neutrophils Macrophages Lymphocytes

4,000
P = 0.05

3,000
Pl

*
No. of Cells/P

2,000

P = 0.15
P = 0.05
1,000

0
l l l l l l l l
n tro r olo ntro r olo ntro r olo ntro r olo
Co t op Co t op Co t op Co t op
Me Me Me Me

B MCP-1 D
Control Metoprolol
500
**
400
MCP-1 (pg/ml)

300

200

100

0
Control Metoprolol
Baseline Day-4

C Cit-H3 MPO-DNA
35 35
30 30
Complexes (Pg/ml)

P = 0.086
Cit-H3 (ng/ml)

25 25
MPO-DNA

**
20 20
15 15
10 10
5 5
0 0
Control Metoprolol Control Metoprolol
Baseline Day-4

(A) Day 4 inflammatory cell populations in BAL from control and metoprolol-treated severe COVID-19 patients. Dots represent individuals and bars and error bars show
mean values (boxes)  SD (error bars). *P < 0.05 by unpaired Student’s t-test. (B) Attenuation of MCP-1 in cell-free BAL from metoprolol-treated patients. (C)
Attenuation of neutrophil hyperactivation biomarkers (Cit-H3 and MPO-DNA complexes) in cell-free BAL from metoprolol-treated patients. Data are presented as
individuals’ (dots) paired data between days 1 and 4. **P < 0.01 by paired Student’s t-test. (D) Representative images of Giemsa-stained BAL samples from control and
metoprolol-treated patients at day 4. Scale bar, 50 mm. Control, n ¼ 8; metoprolol, n ¼ 12. BAL ¼ bronchoalveolar lavage; Cit-H3 ¼ citrullinated histone-3;
COVID-19 ¼ coronavirus disease-2019; MCP ¼ monocyte chemoattractant protein; MPO ¼ myeloperoxidase.
JACC VOL. 78, NO. 10, 2021 Clemente-Moragón et al. 1007
SEPTEMBER 7, 2021:1001–1011 MADRID-COVID Pilot Trial

T A B L E 2 Baseline and Post-Treatment Ventilation Parameters

Baseline Day 4

Metoprolol Control P Value Metoprolol Control P Value

PaO2, mm Hg 87.5 (78.8, 110.0) 104.0 (93.0, 122) 0.105 108.0 (98.3, 139.0) 83.5 (77.3, 92.5) 0.004
PaCO2, mm Hg 48.5 (43.8, 52.5) 47 (41.5, 48.8) 0.562 51.0 (46.5, 53.3) 47.0 (45.3, 50.5) 0.353
PEEP, cm H2O 12.0 (10.0, 12.5) 13.0 (10.0, 14.0) 0.625 10.0 (9.00, 12.0) 11.0 (10.0, 12.0) 0.666
FiO2 0.60 (0.5, 0.75) 0.48 (0.44, 0.60) 0.241 0.40 (0.39, 0.53) 0.43 (0.40, 0.57) 0.634
PaO2/FiO2, 130 (110, 162) 223 (188, 242) 0.076 267 (199, 298) 163 (145, 209) 0.037
Lactic acid, mmol/L 1.3 (1.2, 1.8) 1.2 (0.98, 2.00) 0.785 1.4 (1.20, 1.73) 1.9 (1.50, 2.05) 0.094
pH 7.41 (7.38, 7.42) 7.42 (7.37, 7.45) 0.485 7.43 (7.40, 7.46) 7.41 (7.38, 7.44) 0.461

Values are median (Q1, Q3). Bold indicates statistical significance.

FiO2 ¼ fraction of inspired oxygen; PaCO2 ¼ partial pressure of carbon dioxide; PaO2 ¼ partial pressure of oxygen; PEEP ¼ positive end-expiratory pressure.

dexamethasone, which showed promising results in
an early trial (37), there are no therapies specifically
targeting exacerbated inflammation in ARDS (38).
In this study, we present the effects of 3-day
intravenous metoprolol administration
inflammation in COVID-19 patients with ARDS. The
MADRID-COVID pilot trial shows the following: 1) IV
administration of the clinically approved b-blocker
metoprolol tartrate is safe in this clinical context; 2)
metoprolol treatment abrogates the exacerbated lung
inflammation associated with the disease; and 3) the
disruptive effect on exacerbated inflammation is
associated with better oxygenation and, conse-
quently, fewer days on IMV and in the ICU (Central
Illustration). These data suggest that metoprolol
repurposing for the treatment of ARDS in COVID-19
patients is a safe and inexpensive strategy with the
potential to improve outcomes.
The present study stems from our extensive
experience in the field of myocardial ischemia/
reperfusion injury. We previously demonstrated that
metoprolol protects the heart
myocardial infarction by stunning neutrophils and
abrogating exacerbated inflammation (20,24). The
identification of this cardioprotective mechanism
created an opportunity to repurpose metoprolol for
other acute conditions in which exacerbated inflam-
mation plays a role, as is the case for COVID-19–
associated ARDS. The present study highlights the
importance of knowing the mechanism of action of
long-established drugs to identify other potential
indications.
Patients with severe COVID-19 present with bilat-
eral pneumonia that can lead to respiratory distress
requiring IMV. COVID-19–associated ARDS is charac-
terized by active neutrophil infiltration into the
alveolar space, which perpetuates
inflammation, leading to a cytokine storm and
hypoxemia (8,34). Neutrophil infiltration is thus a
major contributing factor to the poor prognosis of
these patients. Mitigation of immune dysregulation is
therefore a major therapeutic avenue for the treat-
on lung ment and prevention of severe COVID-19.
Several studies have tested the potential benefits of
b-blockers in sepsis/septic shock. Retrospective
observational data have suggested that patients
admitted with septic shock and previously on main-
tenance b-blocker therapy have a better vital prog-
nosis than those who were not on b-blockers before
admission (13). In addition, small prospective clinical
trials have tested the benefits of IV b -blockers in sepsis
patients (12,39-41). The conclusion of most of these
trials is that b-blockers seem to offer a clinical benefit.
In an analysis of diverse experimental models of
exacerbated inflammation, we very recently showed
that not all b1-selective blockers exert the same ef-
fects on neutrophil biology. Of all tested b-blockers,
only metoprolol significantly attenuated exacerbated
inflammation and reduced neutrophil infiltration and
during ongoing interaction with other cell types (24). Those results
position metoprolol as the b-blocker of choice in the
context of exacerbated inflammation.
The present study shows that 3-day treatment with
IV metoprolol reduces exacerbated inflammation in
critically ill COVID-19 patients with associated ARDS.
This was evidenced by the attenuation of infiltration
by immune cells, especially neutrophils, and reduced
levels of their related pro-inflammatory and NETosis
byproducts (Figure 1), which are potential drivers of
severe epithelial and endothelial injury. Lower
neutrophil infiltration in metoprolol-treated patients
was accompanied by a significant reduction in circu-
lating levels of the pro-inflammatory IL8, which ex-
erts chemotactic and activating functions on
exacerbated neutrophils, suggesting a systemic anti-inflammatory
effect of this treatment (Supplemental Figure 4).
1008 Clemente-Moragón et al. JACC VOL. 78, NO. 10, 2021

MADRID-COVID Pilot Trial SEPTEMBER 7, 2021:1001–1011

F I G U R E 2 Metoprolol Rescues Pulmonary Function in ICU Patients With Severe COVID-19

A PaO2 B PaO2:Fio2
300 800
**
*

PAFI Index (pO2:FiO2)

600
pO2 (mm Hg)

200

400

100
200

0 0
Control Metoprolol Control Metoprolol
Baseline Day-4 Baseline Day-4

C
Hospital admission (ward) Intensive care unit admission
Enrollment

METOPROLOL 2.5 [1.8-4.0] 3.2 [2.0-3.3] 11.5 [9.5-18.8] P = 0.152

CONTROL 2.5
1.5 [0.0-2.3] 17.0 [11.3-29.3]
[1.8-3.0]

Invasive METOPROLOL 12.0 [10.0-19.8] P = 0.175

mechanical
ventilation CONTROL 18.5 [11.8-27.5]

Data are expressed as days, median [25th-75th percentile]

(A and B) Improved oxygenation (PaO2 and PaO2:FiO2) in patients receiving metoprolol, but not those in the control group. Data are presented as individuals’ (dots)
paired data between days 1 and 4. *P < 0.05, **P < 0.01 by paired Student’s t-test. (C) Days spent by severe COVID-19 patients on IMV and in the ICU according to
allocation to IV metoprolol or control (no treatment). Control, n ¼ 8; metoprolol, n ¼ 12. FiO2 ¼ fraction of inspired oxygen; ICU ¼ intensive care unit; IMV ¼ invasive
mechanical ventilation; PAFI ¼ PaO2:FiO2 ratio; other abbreviations as in Figure 1.

Systemic markers of NETosis were unaffected at day
4; however, an effect over a longer time window after
metoprolol treatment cannot be
(Supplemental Figure 5). The ameliorative effect of
metoprolol on pulmonary inflammation of COVID-19
patients with ARDS was associated with strong in-
dicators of clinical benefit, demonstrated by a sig-
nificant improvement in oxygenation (PaO 2:FiO2 ) not
seen in control patients (Figure 2). These results are
very encouraging, but further large-scale trials are
needed to validate the clinical benefits of metoprolol
in this context. Given that neutrophils play a major
role in the pathophysiology of ARDS of many causes
(not only COVID-19 related), further large validation
discarded studies might include a wide spectrum of patients
with this condition.
The MADRID-COVID pilot trial has demonstrated
that IV administration of the clinically approved
b -blocker metoprolol to critically ill patients with
ARDS caused by COVID-19 is safe and disrupts the
exacerbated lung inflammation associated with the
disease. The beneficial effects on exacerbated
inflammation were associated with better oxygena-
tion and a nonsignificant reduction in the number of
JACC VOL. 78, NO. 10, 2021 Clemente-Moragón et al. 1009
SEPTEMBER 7, 2021:1001–1011 MADRID-COVID Pilot Trial

C ENTR AL I LL U STRA T I O N Metoprolol Repurposing for Treating ARDS in Critically Ill COVID-19 Patients

Clemente-Moragón, A. et al. J Am Coll Cardiol. 2021;78(10):1001–1011.

Reduced lung inflammation was associated with a significant improvement in oxygenation and fewer days on mechanical ventilation and of intensive care unit
admission. Repurposing metoprolol for the treatment of acute respiratory distress syndrome associated with coronavirus disease-2019 (COVID-19) appears to be a
safe and inexpensive strategy that can alleviate the burden of the COVID-19 pandemic.
1010 Clemente-Moragón et al. JACC VOL. 78, NO. 10, 2021

MADRID-COVID Pilot Trial SEPTEMBER 7, 2021:1001–1011

days on mechanical ventilation and in the ICU.
Intravenous metoprolol appears as a promising
intervention that could improve the prognosis of
critically ill COVID-19 patients. Although these data
need to be corroborated in a larger sample, meto-
prolol is a clinically available and cheap drug (daily
treatment costs <2V) that can improve outcomes in
patients with severe COVID-19.
and Federico Rojo at the Fundación Jiménez Díaz.
Simon Bartlett provided English editing.
FUNDING SUPPORT AND AUTHOR DISCLOSURES
Mr Clemente-Moragón is supported by a fellowship from the Minis-
terio de Ciencia e Innovación (FPU2017/01932). The CNIC is supported
by the ISCIII, the Ministerio de Ciencia e Innovación, and the Pro
CNIC Foundation. Dr Ibáñez is supported by the European Commis-

STUDY LIMITATIONS. The main limitation of this
study is the small sample size. The study was pow-
ered to detect differences in lung inflammation and
not clinical events. Another limitation is the single-
center nature of the study. This was an open-label
study, and treating physicians were not blinded to
treatment allocation. Finally, we cannot rule out a
sion (ERC-CoG grant No 819775) and by the Spanish Ministry of Sci-
ence and Innovation (MCN; “RETOS 2019” grant No PID2019-
107332RB-I00). Dr Oliver is supported by funds from the Comunidad
de Madrid Programa de Atracción de Talento (2017-T1/BMD-5185). All
other authors have reported that they have no relationships relevant
to the contents of this paper to disclose.
ADDRESS FOR CORRESPONDENCE: Dr Borja Ibáñez,

selection bias resulting in patients with very poor
condition according to physicians not considered
for inclusion.
CONCLUSIONS
Our results show that IV administration of metoprolol
to patients with severe COVID-19–associated ARDS is
safe and abrogates the exacerbated lung inflamma-
tion associated with the disease. Reduced lung
inflammation was associated with a significant
improvement in oxygenation and with a trend toward
fewer days on mechanical ventilation and of ICU
admission. Metoprolol repurposing for the treatment
of ARDS associated with COVID-19 is a safe and cheap
intervention that can help to alleviate the massive
personal and health care burden associated with the
pandemic.
ACKNOWLEDGMENTS The authors thank the following
for their important support during this study: Noemí
Escalera, Rocío Escudero, and Antonio de Molina-
Iracheta at the CNIC; and Luis Nieto, Ana María
Venegas, Jose Tuñón, Ignacio Cornejo, Sandra Zazo,
Translational Laboratory for Cardiovascular Imaging
and Therapy, Centro Nacional de Investigaciones
Cardiovasculares (CNIC) and IIS-Fundación Jiménez
Díaz University Hospital, c/ Melchor Fernandez
Almagro, 3. 28029 Madrid, Spain. E-mail: bibanez@
cnic.es. Twitter: @Borjaibanez1, @CNIC_CARDIO.
PERSPECTIVES
COMPETENCY IN PATIENT CARE AND
PROCEDURAL OUTCOMES: In critically ill patients
with COVID-19 on mechanical ventilatory support,
intravenous administration of metoprolol upon
admission to the ICU is safe and improves pulmonary
function and clinical outcome.
TRANSLATIONAL OUTLOOK: Future studies with
larger sample sizes are needed to confirm the benefit
of metoprolol in critically ill patients with COVID-19
and potentially other inflammatory etiologies of
ARDS.

REFERENCES

1. Ritter M, Ott DVM, Paul F, Haynes JD, Ritter K.
COVID-19: a simple statistical model for predicting
intensive care unit load in exponential phases of
the disease. Sci Rep. 2021;11:5018.
2. Middleton EA, He XY, Denorme F, et al.
Neutrophil extracellular traps contribute to
immunothrombosis in COVID-19 acute respiratory
distress syndrome. Blood. 2020;136:1169–1179.
3. Zuo Y, Yalavarthi S, Shi H, et al. Neutrophil
extracellular traps in COVID-19. JCI Insight. 2020;5.
4. Thompson BT, Chambers RC, Liu KD. Acute
respiratory distress syndrome. N Engl J Med.
2017;377:562–572.
5. Cavalcante-Silva LHA, Carvalho DCM, Lima EA,
et al. Neutrophils and COVID-19: the road so far.
Int Immunopharmacol. 2021;90:107233.
6. Borges L, Pithon-Curi TC, Curi R, Hatanaka E.
COVID-19 and neutrophils: the relationship be-
tween hyperinflammation and neutrophil extra-
cellular traps. Mediators Inflamm. 2020;2020:
8829674.
7. Standiford TJ, Ward PA. Therapeutic targeting
of acute lung injury and acute respiratory distress
syndrome. Transl Res. 2016;167:183–191.
8. Chiang CC, Korinek M, Cheng WJ, Hwang TL.
Targeting neutrophils to treat acute respiratory
distress syndrome in coronavirus disease. Front
Pharmacol. 2020;11:572009.
9. Tomar B, Anders HJ, Desai J, Mulay SR.
Neutrophils and neutrophil extracellular traps
drive necroinflammation in COVID-19. Cells.
2020;9:1383. https://doi.org/10.3390/cells90
61383
10. Schulte-Schrepping J, Reusch N, Paclik D,
et al. Severe COVID-19 is marked by a dysregu-
lated myeloid cell compartment. Cell. 2020;182:
1419–1440.e23.
11. Martinez-Milla J, Raposeiras-Roubin S, Pasc-
ual-Figal DA, Ibanez B. Role of beta-blockers in
JACC VOL. 78, NO. 10, 2021
SEPTEMBER 7, 2021:1001–1011
cardiovascular disease in 2019. Rev Esp Cardiol
(Engl Ed). 2019;72:844–852.
12. Chacko CJ, Gopal S. Systematic review of use
of beta-blockers in sepsis. J Anaesthesiol Clin
Pharmacol. 2015;31:460–465.
13. Tan K, Harazim M, Tang B, McLean A, Nalos M.
The association between premorbid beta blocker
exposure and mortality in sepsis-a systematic re-
view. Crit Care. 2019;23:298.
14. Oliver E, Mayor F Jr, D’Ocon P. Beta-blockers:
historical perspective and mechanisms of action.
Rev Esp Cardiol (Engl Ed). 2019;72(10):853–862.
15. Noveanu M, Breidthardt T, Reichlin T, et al.
Effect of oral beta-blocker on short and long-term
mortality in patients with acute respiratory failure:
results from the BASEL-II-ICU study. Crit Care.
2010;14:R198.
16. Cotton BA, Snodgrass KB, Fleming SB, et al.
Beta-blocker exposure is associated with improved
survival after severe traumatic brain injury.
J Trauma. 2007;62:26–33. discussion 33-5.
17. Inaba K, Teixeira PG, David JS, et al. Beta-
blockers in isolated blunt head injury. J Am Coll
Surg. 2008;206:432–438.
18. Christensen S, Johansen MB, Tonnesen E, et al.
Preadmission beta-blocker use and 30-day mor-
tality among patients in intensive care: a cohort
study. Crit Care. 2011;15:R87.
19. van der Jagt M, Miranda DR. Beta-blockers in
intensive care medicine: potential benefit in acute
brain injury and acute respiratory distress syn-
drome. Recent Pat Cardiovasc Drug Discov. 2012;7:
141–151.
20. Garcia-Prieto J, Villena-Gutierrez R, Gomez M,
et al. Neutrophil stunning by metoprolol reduces
infarct size. Nat Commun. 2017;8:14780.
21. Garcia-Ruiz JM, Fernandez-Jimenez R, Garcia-
Alvarez A, et al. Impact of the timing of meto-
prolol administration during STEMI on infarct size
and ventricular function. J Am Coll Cardiol.
2016;67:2093–2104.
22. Pizarro G, Fernandez-Friera L, Fuster V, et al.
Long-term benefit of early pre-reperfusion meto-
prolol administration in patients with acute
myocardial infarction: results from the
METOCARD-CNIC Trial (Effect of Metoprolol in
Cardioprotection During an Acute Myocardial
Infarction). J Am Coll Cardiol. 2014;63:2356–2362.
23. Ibanez B. Intravenous beta-blockers in STEMI:
what you are about to do, do it quickly. Eur Heart J
Acute Cardiovasc Care. 2020;9:459–461.
24. Clemente-Moragon A, Gomez M, Villena-
Gutierrez R, et al. Metoprolol exerts a non-class
effect against ischaemia-reperfusion injury by
abrogating exacerbated inflammation. Eur Heart J.
2020;41:4425–4440.
25. Smith PJ, Wiltshire M, Errington RJ. DRAQ5
labeling of nuclear DNA in live and fixed cells. Curr
Protoc Cytom. 2004 Chapter 7:Unit 7.25.
26. Martin RM, Leonhardt H, Cardoso MC. DNA
labeling in living cells. Cytometry A. 2005;67:45–
52.
27. Mandy F, Brando B. Enumeration of Absolute
Cell Counts Using Immunophenotypic Techniques.
Curr Protoc Cytom. 2000;13:6.8.1–6.8.26.
28. Kessenbrock K, Krumbholz M,
Schonermarck U, et al. Netting neutrophils in
autoimmune small-vessel vasculitis. Nat Med.
2009;15:623–625.
29. Zuo Y, Zuo M, Yalavarthi S, et al. Neutrophil
extracellular traps and thrombosis in COVID-19.
J Thromb Thrombolysis. 2021;51:446–453.
30. Malachowa N, Kobayashi SD, Freedman B,
Dorward DW, DeLeo FR. Staphylococcus aureus
leukotoxin GH promotes formation of neutrophil
extracellular traps. J Immunol. 2013;191:6022–
6029.
31. Vargas A, Boivin R, Cano P, Murcia Y, Bazin I,
Lavoie JP. Neutrophil extracellular traps are
downregulated by glucocorticosteroids in lungs in
an equine model of asthma. Respir Res. 2017;18:
207.
32. Hartl D, Griese M, Nicolai T, et al. A role for
MCP-1/CCR2 in interstitial lung disease in children.
Respir Res. 2005;6:93.
Clemente-Moragón et al. 1011
MADRID-COVID Pilot Trial
33. He C, Carter AB. C(C)Learing the Role of Che-
mokines in Pulmonary Fibrosis. Am J Respir Cell
Mol Biol. 2020;62:546–547.
34. Pandolfi L, Fossali T, Frangipane V, et al.
Broncho-alveolar inflammation in COVID-19 pa-
tients: a correlation with clinical outcome. BMC
Pulm Med. 2020;20:301.
35. Haas CF. Lung protective mechanical ventila-
tion in acute respiratory distress syndrome. Respir
Care Clin N Am. 2003;9:363–396.
36. Gong MN, Ferguson ND. Lung-protective
ventilation in acute respiratory distress syndrome.
How soon is now? Am J Respir Crit Care Med.
2015;191:125–126.
37. Villar J, Ferrando C, Martinez D, et al. Dexa-
methasone treatment for the acute respiratory
distress syndrome: a multicentre, randomised
controlled trial. Lancet Respir Med. 2020;8:267–
276.
38. Baudouin SV. Manipulation of inflammation in
ARDS: achievable goal or distant target? Thorax.
2006;61:464–465.
39. Schmittinger CA, Dunser MW, Haller M, et al.
Combined milrinone and enteral metoprolol ther-
apy in patients with septic myocardial depression.
Crit Care. 2008;12:R99.
40. Morelli A, Donati A, Ertmer C, et al. Micro-
vascular effects of heart rate control with esmolol
in patients with septic shock: a pilot study. Crit
Care Med. 2013;41:2162–2168.
41. Morelli A, Ertmer C, Westphal M, et al. Effect
of heart rate control with esmolol on hemody-
namic and clinical outcomes in patients with septic
shock: a randomized clinical trial. JAMA. 2013;310:
1683–1691.
KEY WORDS acute care, ARDS, COVID,
metoprolol
A PP END IX For supplemental tables and
figures, please see the online version of this
paper.
JOURNAL OF THE AMERICAN COLLEGE OF CARDIOLOGY
ª 2021 BY THE AMERICAN COLLEGE OF CARDIOLOGY FOUNDATION
PUBLISHED BY ELSEVIER
EDITORIAL COMMENT
Beta-Blockers in the Critically Ill
Friend or Foe?*
Mourad H. Senussi, MD, MS
B eta-blockers are a commonly prescribed medi-
cation for patients with cardiovascular comor-
bidities, such as ischemia, heart failure,
hypertension, and arrhythmias. Beta-blockers inhibit
the sympathomimetic stimulation of b-1 adrenergic
receptors primarily found on cardiomyocytes as well
as the sinoatrial and atrioventricular nodes. This re-
sults in both a negative chronotropic and inotropic ef-
fect on the heart that leads to decreased blood
pressure, heart rate, and myocardial work (1). Beta-
blockers along with their beneficial cardiac profile
also exhibit important metabolic and immunomodu-
latory effects (2,3) that are often overlooked.
Beta-blockers have become ubiquitous in the
treatment of patients with heart failure and ischemic
heart disease. The 2013 American College of Cardiol-
ogy/American Heart Association Heart Failure guide-
lines recommend b -blockade in patients with a
history of myocardial infarction or acute coronary
syndrome and reduced ejection fraction to reduce
mortality (4) and use of beta-blockers in all patients
with reduced ejection fraction to prevent symptom-
atic heart failure and hospitalization.
During critical illness, the initial adrenergic stimu-
lation may be beneficial; however, sympathetic over-
activity, if persistent, may become deleterious, leading
to further oxygen consumption, catabolism, hyper-
glycemia, hypercoagulability, cardiac dysfunction,
modulation of systemic inflammatory cytokines, and
ensuing multiorgan failure (2,5-7). There has been an
*Editorials published in the Journal of the American College of Cardiology
reflect the views of the authors and do not necessarily represent the
views of JACC or the American College of Cardiology.
From the Baylor St. Luke’s Medical Center, Baylor College of Medicine,
Houston, Texas, USA.
Christie Ballantyne, MD, served as the Guest Editor for this paper.
The author attests they are in compliance with human studies commit-
tees and animal welfare regulations of the author’s institution and Food
and Drug Administration guidelines, including patient consent where
appropriate. For more information, visit the Author Center.
ISSN 0735-1097/$36.00
VOL. 78, NO. 10, 2021
ever-increasing interest in the role of beta-blockers in
critically ill patients in blunting this adrenergic
response (3). There is a growing body of literature
regarding the role of beta-blockers in a wide variety of
critically ill patients with sepsis and before major
surgery, acute respiratory distress syndrome, and
traumatic brain injury (8-13).
Several observational studies have shown a
reduced mortality in patients in the intensive care
unit who were on preadmission beta-blockers (14,15)
and specifically those with acute respiratory failure
(10). Furthermore, discontinuation of these drugs
during their hospital stay was associated with higher
mortality rates independent of the cause of respira-
tory failure.
The use of continuous infusion of esmolol, an
intravenous short-acting beta-blocker, showed a
decreased need for fluids and vasopressors as well as
improved 28-day mortality in patients with sepsis
(11). However, conflicting evidence exists in other
critically ill populations (16), with individuals who
received beta-blockers 30 days before elective cardiac
surgery showing increased risk of death and stroke
associated with an increased incidence of hypoten-
sion, bradycardia, and bleeding (17).
There remains a considerable gap in knowledge in
the understanding of the complex balance between
beneficial sympathetic stimulation and the multitude
of derangements caused by persistent sympathetic
overactivity in critically ill individuals. The use of
beta-blockers has become ubiquitous in the treatment
of patients with cardiac disease; however, there
remains considerable variation on the optimal timing
of the initiation, withholding, and restarting of
beta-blocker therapy in critically ill patients.
The repurposing of medications to combat an
ever-growing COVID-19 pandemic has reached fever
pitch, with a variety of drug classes being used, with
the vast majority showing unclear benefit. In this issue
of the Journal, Clemente-Moragón et al (18) explore
https://doi.org/10.1016/j.jacc.2021.07.006
JACC VOL. 78, NO. 10, 2021 Senussi 1013
SEPTEMBER 7, 2021:1012–1014 Friend or Foe?

the use of intravenous metoprolol on alveolar inflam-
mation and respiratory function in critically ill pa-
tients with COVID-19–related hypoxemic respiratory
failure. Although the study has a small sample size, we
commend the authors, who attempt to shed light on
the important pathophysiologic underpinnings that
help establish biological plausibility for this inexpen-
sive, safe, and widely available medication.
SEE PAGE 1001
Host response to disease plays a central and
pivotal role in critical illness and can lead to
significant organ dysfunction and ensuing mortality.
Different treatment modalities targeting
inflammatory response have been an intense area of
research as well as a source of contention. The more
important question is not its disease-specific effects
on COVID-19 per se but, rather, its beneficial effects
in critical illness in general due to sympathetic
blunting and the subsequent immune-modulating
effects of this drug class.
In this study, beta-blockers were instituted early in
the disease course, which may play an important role
in sympathetic blunting before the effects of the
dysregulated inflammatory cascade that may lead to
organ dysfunction. Patients in this study who
received beta-blockers showed less lung inflamma-
tion, improved PaO 2:FiO2 ratios and decreased venti-
lator days. However, there is clear selection bias
because only those patients who are hemodynami-
cally stable enough can receive beta-blockers. Several
confounding factors include exposure to corticoste-
roids before randomization and analgesia/sedation
practices that may blunt the inflammatory response.
An important consideration is that a large propor-
tion of patients admitted to the intensive care
unit have a high burden of cardiac
Cardioprotective medications have improved out-
comes, but the management of these medications in
critical illness remains unclear. There is clear evi-
dence that beta-blocker exposure before and during
ICU admission is protective, and abrupt withdrawal
causes harm. One source of clinical variability that
may lead to suboptimal treatment and poorer out-
comes is the management of cardioprotective medi-
cations. Accordingly, there remains a considerable
knowledge gap regarding the complex interplay be-
tween cardioprotective medications and critical
illness, with no clear guidelines or consensus state-
ments on the management of these medications.
this Although a small-sized, single-center study amid a
multitude of others exploring potential treatment
modalities for COVID-19—this study uses a readily
available, safe, and inexpensive medication; has a
simple study design; and, most importantly, shows
biological plausibility. Although observed in patients
with COVID-19, this sets the groundwork for further
research in the use of beta-blockade in the critically
ill. Further studies are needed to elucidate and
identify where along the inflammatory spectrum
these critically ill patients lie, which patients would
benefit from beta-blockers, and at what time point
during their hospital stay.
FUNDING SUPPORT AND AUTHOR DISCLOSURES
The author has reported that he has no relationships relevant to the
contents of this paper to disclose.
ADDRESS FOR CORRESPONDENCE: Dr Mourad
Senussi, Baylor St. Luke’s Medical Center, 6720
Bertner Avenue, P631, Houston, Texas 77005, USA.
E-mail: Mourad.Senussi@bcm.edu. Twitter:
disease. @msenussiMD.

REFERENCES

1. Gorre F, Vandekerckhove H. Beta-blockers: 2010;38(2):388–394. https://doi.org/10.1097/ 6. Hagiwara S, Iwasaka H, Maeda H, Noguchi T.

focus on mechanism of action. Which beta-
blocker, when and why? Acta Cardiol.
2010;65(5):565–570. https://doi.org/10.1080/ac.
65.5.2056244
2. Coppola S, Froio S, Chiumello D. b-blockers in
critically ill patients: from physiology to clinical
evidence. Crit Care. 2015;19:119. https://doi.org/
10.1186/s13054-015-0803-2
3. Ackland GL, Yao ST, Rudiger A, et al. Car-
dioprotection, attenuated systemic inflammation,
and survival benefit of beta1-adrenoceptor
blockade in severe sepsis in rats. Crit Care Med.
CCM.0b013e3181c03dfa
4. Yancy CW, Jessup M, Bozkurt B, et al. 2013
ACCF/AHA guideline for the management of heart
failure: a report of the American College of Car-
diology Foundation/American Heart Association
Task Force on Practice Guidelines. J Am Coll Car-
diol. 2013;62(16):e147–e239.
5. Oberbeck R, van Griensven M, Nickel E,
Tschernig T, Wittwer T, Pape H-C. Influence of
beta-adrenoceptor antagonists on hemorrhage-
induced cellular immune suppression. Shock.
2002;18(4):331–335.
Landiolol, an ultrashort-acting beta1-adrenoceptor
antagonist, has protective effects in an LPS-
induced systemic inflammation model. Shock.
2009;31(5):515–520. https://doi.org/10.1097/
SHK.0b013e3181863689
7. Ginsberg F. b-blockers: more good news? Crit
Care Med. 2012;40(10):2901–2902. https://doi.
org/10.1097/CCM.0b013e31825f757b
8. Chacko CJ, Gopal S. Systematic review of use of
b-blockers in sepsis. J Anaesthesiol Clin Pharmacol.
2015;31(4):460–465. https://doi.org/10.4103/
0970-9185.169063
1014 Senussi
Friend or Foe?
9. Herndon DN, Hart DW, Wolf SE, Chinkes DL,
Wolfe RR. Reversal of catabolism by beta-
blockade after severe burns. N Engl J Med.
2001;345(17):1223–1229. https://doi.org/10.1056/
NEJMoa010342
10. Noveanu M, Breidthardt T, Reichlin T, et al.
Effect of oral b-blocker on short and long-term
mortality in patients with acute respiratory
failure: results from the BASEL-II-ICU study. Crit
Care. 2010;14(6):R198. https://doi.org/10.1186/
cc9317
11. Morelli A, Ertmer C, Westphal M, et al. Effect of
heart rate control with esmolol on hemodynamic
and clinical outcomes in patients with septic
shock: a randomized clinical trial. JAMA.
2013;310(16):1683–1691. https://doi.org/10.1001/
jama.2013.278477
12. van der Jagt M, Miranda DR. Beta-blockers in
intensive care medicine: potential benefit in acute
brain injury and acute respiratory distress syn-
drome. Recent Pat Cardiovasc Drug Discov.
2012;7(2):141–151.
13. Cotton BA, Snodgrass KB, Fleming SB, et al.
Beta-blocker exposure is associated with improved
survival after severe traumatic brain injury.
J Trauma. 2007;62(1):26–35. https://doi.org/10.
1097/TA.0b013e31802d02d0
14. Christensen S, Johansen MB, Tønnesen E, et al.
Preadmission beta-blocker use and 30-day mor-
tality among patients in intensive care: a cohort
study. Crit Care. 2011;15(2):R87. https://doi.org/
10.1186/cc10085
15. Macchia A, Romero M, Comignani PD, et al.
Previous prescription of b-blockers is associated
with reduced mortality among patients hospital-
ized in intensive care units for sepsis. Crit Care
Med. 2012;40(10):2768–2772. https://doi.org/10.
1097/CCM.0b013e31825b9509
JACC VOL. 78, NO. 10, 2021
SEPTEMBER 7, 2021:1012–1014
16. Al Harbi SA, Al Sulaiman KA, Tamim H, Al-
Dorzi HM, Sadat M, Arabi Y. Association between
b-blocker use and mortality in critically ill pa-
tients: a nested cohort study. BMC Pharmacol
Toxicol. 2018;19(1):22. https://doi.org/10.1186/
s40360-018-0213-6
17. POISE Study Group, Devereaux PJ, Yang H,
et al. Effects of extended-release metoprolol
succinate in patients undergoing non-cardiac sur-
gery (POISE trial): a randomised controlled trial.
Lancet. 2008;371(9627):1839–1847. https://doi.
org/10.1016/S0140-6736(08)60601-7
18. Clemente-Moragón A, Martínez-Milla J,
Oliver E, et al. Metoprolol in critically ill patients
with COVID-19. J Am Coll Cardiol. 2021;78(10):
1001–1011.
KEY WORDS acute care, ARDS, COVID,
metoprolol

Listen to this manuscript’s
audio summary by
Editor-in-Chief
Dr. Valentin Fuster on
JACC.org.
JOURNAL OF THE AMERICAN COLLEGE OF CARDIOLOGY VOL. 78, NO. 10, 2021
ª 2021 BY THE AMERICAN COLLEGE OF CARDIOLOGY FOUNDATION
PUBLISHED BY ELSEVIER
Risk of Cardiovascular Hospital
Admission After Exposure to
Fine Particulate Pollution
Yi Zhang, MPH,a Runmei Ma, MSC,a Jie Ban, MD,a Feng Lu, PHD,b Moning Guo, MD,b Yu Zhong, MSC,a
Ning Jiang, BSMED,a Chen Chen, MSC,a Tiantian Li, PHD,a Xiaoming Shi, MDa
ABSTRACT
BACKGROUND Heavy fine particulate matter with an aerodynamic diameter #2.5 mm (PM2.5) pollution events continue
to occur frequently in developing countries.
OBJECTIVES The authors conducted a case-crossover study aimed at exploring the association between heavy PM2.5
pollution events and hospital admission for cardiovascular diseases.
METHODS Hospital admissions for cardiovascular diseases were observed by Beijing Municipal Commission of Health
and Family Planning Information Center from 2013 to 2017. Air pollution data were collected from the Beijing Municipal
Environmental Monitoring Center. Distinct definitions were used to identify heavy and extremely heavy fine particulate
pollution events. A conditional logistic regression model was used. The hospital admission burdens for cardiovascular
disease were also estimated.
RESULTS A total of 2,202,244 hospital admissions for cardiovascular diseases and 222 days of extremely heavy PM2.5
pollution events (PM2.5 concentration $150 mg/m3) were observed. The ORs associated with extremely heavy PM2.5
pollution events lasting for 3 days or more for total cardiovascular disease, angina, myocardial infarction, ischemic stroke,
and heart failure were 1.085 (95% CI: 1.077-1.093), 1.112 (95% CI: 1.095-1.130), 1.068 (95% CI: 1.037-1.100), 1.071
(95% CI: 1.053-1.090), and 1.060 (95% CI: 1.021-1.101), respectively. The numbers and days of cardiovascular disease
hospital admission annually related to extremely heavy PM2.5 pollution events lasting for 1 day or more were 3,311
(95% CI: 2,969-3,655) and 37,020 (95% CI: 33,196-40,866), respectively.
CONCLUSIONS Heavy and extremely heavy PM2.5 pollution events resulted in substantial increased hospital admission
risk for cardiovascular disease. With higher PM2.5 concentration and longer duration of heavy PM2.5 pollution events, a
greater risk of cardiovascular hospital admission was observed. (J Am Coll Cardiol 2021;78:1015–1024) © 2021 by the
American College of Cardiology Foundation.
D espite improvements in fine particulate
matter with an aerodynamic
#2.5 mm (PM 2.5) pollution among developed
countries, the State of Global Air/2019 report pub-
diameter
East Asia continue to suffer the impact of a high con-
centration of PM2.5 in 2017 (1). Heavy PM2.5 pollution
lasting for consecutive days occurred in China from
2013-2017 (2). According to the Report on the State of
lished by the Health Effects Institute (HEI) in the Ecological Environment in China/2017, 3113 days of
United States demonstrates that low-income and extremely heavy air pollution (air quality
middle-income countries in South Asia, Western sub- index $200) were observed in 338 cities in total,
Saharan Africa, North Africa, the Middle East, and among which 74.2% of the air pollution was caused
From the aChina CDC Key Laboratory of Environment and Population Health, National Institute of Environmental Health, Chinese
Center for Disease Control and Prevention, Beijing, China; and the bBeijing Municipal Health Commission Information Center,
Beijing, China.
The authors attest they are in compliance with human studies committees and animal welfare regulations of the authors’
institutions and Food and Drug Administration guidelines, including patient consent where appropriate. For more information,
visit the Author Center.
Manuscript received April 22, 2021; revised manuscript received June 3, 2021, accepted June 24, 2021.
ISSN 0735-1097/$36.00 https://doi.org/10.1016/j.jacc.2021.06.043
1016 Zhang et al.
Heavy Pollution and CVD
ABBREVIATIONS by a high concentration of PM2.5 (3). This situ-
AND ACRONYMS ation provided a rare opportunity to conduct
this study on the health effects of heavy
ICD-10 = International
Classification of Diseases-
PM 2.5 pollution events in China.
10th Revision Several episode studies conducted previ-
PM2.5 = fine particulate matter ously have estimated the association be-
with an aerodynamic tween heavy air pollution and health effects
diameter £2.5 mm
(4-6). Bell et al (4) conducted a reassessment
of London fog from 1952 with 1400 m g/m 3 as the
average concentration of total suspended matter
during the episode, and found an 8% (95% CI: 5%-
10%) increase of mortality risk for exposure per
100 m g/m 3 of total suspended matter. Chen et al (5)
and Ferreri et al (6) investigated the associations be-
tween heavy PM2.5 pollution event in Beijing during
2013 (231 m g/m 3 PM 2.5 as episode average) and medical
visits, reporting that cardiovascular disease hospital
admission and emergency medical visits increased
127% (95% CI: 12%-362%) and 55% (95% CI:
10% to
168%), respectively, during the episode. However,
both studies only focused on 1 extremely heavy PM2.5
pollution event, and as such, their results could not
be generalized to other heavy PM2.5 pollution events.
Hospital admission data from the single hospital used
by previous studies cannot represent the large city,
and furthermore, it may lead to selection bias.
SEE PAGE 1025
In contrast to the relative infrequency of single
heavy PM2.5 pollution events in developed countries
within Europe and North America, heavy PM2.5
pollution events occurred frequently in developing
countries (1,4,7,8). High concentrations of PM 2.5 in
the heavy PM2.5 pollution events may cause large
amounts of toxic components of PM2.5 entering into
the lungs in acute exposure (9), and then directly or
indirectly damage the cardiovascular system (10).
Prolonged exposure will lead to the collapse of the
body’s ability to compensate and eventually cause
the incidence of diseases (11,12). Previous epidemi-
ology studies demonstrated that the cardiovascular
system is the most vulnerable system of the human
body when exposed to PM 2.5 (13-19) in the short term.
Excessive cardiovascular disease admission due to
PM 2.5 may furthermore lead to a high subsequent fa-
tality burden and substantial medical expenditure
(20-23). However, because of the lack of research on
the health risks of frequently occurring heavy PM2.5
pollution events, this critical public health burden
has failed to attract sufficient attention from the
government, marked by a pronounced lack of health
protection measures to combat heavy PM2.5 pollution
events. Accordingly, without this quantitative evi-
dence, necessary legislation and early warning
JACC VOL. 78, NO. 10, 2021
SEPTEMBER 7, 2021:1015–1024
systems have not been established (6-8). Without a
robust response, this situation will eventually result
in serious harm to public health. Thus, it is urgent to
fully address the health risks of regularly occurring
heavy PM 2.5 pollution events.
Beijing, the capital of China, is located in the center
of the Beijing-Tianjin-Hebei (Jing-Jin-Ji) typical air
pollution area, and experienced heavy PM 2.5 pollu-
tion events from 2013 to 2017 (24,25). Based on the
high-quality case-level hospital admission data of
Beijing, this study aims to analyze the association
between heavy PM2.5 pollution events and cardio-
vascular disease hospital admissions, specifically
examining whether the number of lasting days of
heavy pollution could affect the risk value, and
identify which specific cardiovascular disease results
in the greatest hospital admission burden.
METHODS
DATA COLLECTION. Air pollution data from 35
monitoring stations distributed in all 16 districts in
Beijing were obtained from the Beijing Municipal
Environmental Monitoring Center. The daily
average PM2.5 , O 3, NO 2, SO 2, and CO levels were
calculated from the hourly data of the surveillance
stations. Daily temperature and relative humidity
were obtained from the Chinese Meteorological Data
Network of the Chinese Meteorological Bureau.
Hospital admission records for cardiovascular dis-
ease (International Classification of Diseases 10th
Revision codes [ICD-10]: I00-I99) from January 1, 2013,
to December 31, 2017, were collected from the Beijing
Municipal Health Commission Information Center, a
citywide hospital admission surveillance system in the
Chinese capital. The surveillance system covered all
the hospitals, with the exception of first-level hospi-
tals (primary hospitals that directly provide compre-
hensive medical, prevention, rehabilitation, and
health care services to communities, and are primary
health care institutions). Patients who resided in Bei-
jing for <6 months were excluded. Age, sex, race,
marital status, and days in hospital were also assessed.
The study was approved by the ethical committee at
the National Institute of Environmental Health, Chi-
nese Center for Disease Control and Prevention.
Furthermore, we extracted the different specific
causes of hospital admissions for cardiovascular dis-
ease as outcomes and their corresponding ICD-10
codes, including myocardial infarction, ICD-10 code
I21-I23; chronic ischemic heart disease, ICD-10 code
I25; arrhythmia, ICD-10 code I44-I49; heart failure,
ICD-10 code I50; angina, ICD-10 code I20; ischemic
stroke, ICD-10 code I63; hemorrhagic stroke, ICD-10
JACC VOL. 78, NO. 10, 2021 Zhang et al. 1017
SEPTEMBER 7, 2021:1015–1024 Heavy Pollution and CVD

T A B L E 1 Frequency and Percentage of Basic Information of Hospital Admission for Cardiovascular Disease in Beijing, 2013-2017

Chronic
Ischemic Other
Cardiovascular Myocardial Heart Heart Hemorrhagic Ischemic Hypertensive Pulmonary Cardiovascular
Disease Infarction Disease Arrhythmia Failure Angina Stroke Stroke Diseases Embolism Disease

Total, person-time 2,202,244 131,993 157,645 149,778 88,767 513,002 60,745 412,321 126,843 21,243 539,907
Age, mean  SD, y 64  16 64  14 65  13 59  20 74  13 65  12 60  15 68  13 62  16 59  20 58  18
<65 y 50.79 51.47 49.07 55.94 19.81 50.08 61.51 40.67 54.90 53.90 61.88
$65 y 49.21 48.53 50.93 44.06 80.19 49.92 38.49 59.33 45.10 46.10 38.12
Sex
Men 58.48 71.55 57.19 52.17 50.22 60.87 62.00 60.6 48.65 42.33 56.42
Women 41.52 28.45 42.81 47.83 49.78 39.13 38.00 39.4 51.35 57.67 43.48
Marriage status
Unmarried 3.46 1.74 0.96 8.95 2.27 1.03 3.96 1.67 3.56 8.09 6.69
Married 91.82 93.27 94.04 87.47 88.17 95.02 89.52 89.90 91.93 87.51 89.95
Divorced 0.95 0.10 0.71 0.62 1.57 0.65 1.73 1.34 1.02 0.88 0.85
Widowed 2.84 2.74 2.70 2.22 6.91 2.46 3.06 3.91 2.75 2.77 1.86
Han nationality 94.62 95.00 94.16 94.14 96.68 95.86 96.11 94.25 94.49 95.13 93.08

Values are % unless otherwise indicated.

code I60-I61; hypertensive diseases, ICD-10 code I10-
I15; and pulmonary embolism, ICD-10 code I26-I27.
STUDY DESIGN. To identify the different effects of
heavy and extremely heavy PM 2.5 pollution events
lasting for different consecutive days, we used 3
definitions of heavy and extremely heavy PM 2.5
pollution events: definition 1, PM2.5 concentration
75-149 m g/m 3
lasting for 1 day or more is a
heavy PM2.5 pollution event, and PM2.5 concentra-
tion $150 m g/m 3
lasting for 1 day or more is an
extremely heavy PM2.5 pollution event; definition 2,
PM 2.5 concentration $75 m g/m lasting for 2 days or
3 non-heavy pollution days were then compared be-
more, but not reaching $150 m g/m lasting for 2 days3
or more, is a heavy PM 2.5 pollution event (using 75-
149 m g/m 3 lasting for 2 days or more to represent this
class), and PM2.5 concentration $150 m g/m3 lasting
for 2 days or more is an extremely heavy PM2.5
pollution event; definition 3, PM2.5 concentration
$75 m g/m 3 lasting for 3 days or more, but not
reaching $150 m g/m 3 lasting for 3 days or more, is a
heavy PM2.5 pollution event (using 75-149 m g/m 3
lasting for 3 days or more to represent this class),
and PM 2.5 concentration $150 m g/m 3 lasting for
3 days or more is an extremely heavy PM2.5 pollu-
tion event.
We used a time-stratified case-crossover design to
explore the association between different lasting
days of PM2.5 heavy pollution events and cardiovas-
cular disease admission (26). In this study, each car-
diovascular disease patient was recognized as a case
on cardiovascular disease admission day and as a
matched control during the time before (or after)
cardiovascular disease admission. The control days
were matched to the case days on the same weekday,
the same month, and the same year. For each case, we
compared whether a heavy (or extremely heavy)
pollution event day occurred during and before (or
after) the cardiovascular disease admission. For
example, if a patient was admitted to hospital for
cardiovascular disease on May 20, 2015, Wednesday,
then the case day consisted of May 20, 2015, and the
control days were May 6, May 13, and May 27, 2015.
Heavy (or extremely heavy) pollution event days and
tween these cases and the control days in all analyses.
STATISTICAL ANALYSIS. A conditional logistic
regression model stratified by each cardiovascular
disease admission was used through R software
version 3.6.3 (R Foundation for Statistical
Computing). The case was assigned 1, and the control
was assigned 0. Whether it was the heavy pollution
event day (or extremely heavy pollution event day)
was contrasted in the model as the independent
variable. The daily mean temperature was controlled
as a covariate using the natural spline (3 degrees of
freedom). In addition, the daily humidity was
controlled as a linear covariate, and holiday was
controlled as the categorical covariate in the main
analysis (main model). The adjusted OR and 95% CI
were calculated.
We ran the models one by one for each definition of
heavy and extremely heavy PM2.5 pollution events.
References for ORs were days of non-heavy and
nonextremely heavy pollution events under the
1018 Zhang et al.
Heavy Pollution and CVD
F I G U R E 1 Days of Heavy PM 2.5 Pollution Events Under Different Definitions
Definition 1 Definition 2
300
Number
200
100
0
2013 2014 2015 2016 2017 2013 2014 2015 2016 2017
Year
Blue represents days of extremely heavy PM2.5 pollution events under different definitions; red represents days of heavy PM2.5 pollution events under
different definitions; black represents days that were not heavy and extremely heavy PM2.5 pollution events under different definitions. PM2.5 ¼ fine
particulate matter with an aerodynamic diameter #2.5 mm.
certain definition. We also examined the lag effects
for the associations of heavy or extremely heavy
PM 2.5 pollution events and cardiovascular disease
under different definitions, and the results are shown
in Supplemental Figure 1. In addition to the total
cardiovascular disease admission, we used specific
diseases (myocardial infarction, chronic ischemic
heart disease, heart failure, angina, ischemic stroke,
hemorrhagic stroke, hypertensive diseases, pulmo-
nary embolism, and arrhythmia) to rerun the main
model to estimate the ORs for the specific causes.
We used estimated ORs from the main model with
heavy (or extremely heavy) PM2.5 pollution events
lasting 1 day or more, the average number of daily
patient admissions during the study period (to
represent a normal condition), the number of heavy
(or extremely heavy) PM2.5 pollution event days in a
certain year (2013, 2014, 2015, 2016, 2017), and
average days in hospital for each hospital admission
to calculate the increased number of hospital admis-
sion patients (“In” in Equation 1) and increased days
in hospital (“Id” in Equation 2) due to the heavy (or
extremely heavy) PM 2.5 pollution events (13).
In ¼ ðOR
1Þ  A  H (Equation 1)
Where OR is estimated risk with heavy (or extremely
heavy) PM2.5 pollution events; A is the average number
of daily patient admissions during the study period; and
JACC VOL. 78, NO. 10, 2021
SEPTEMBER 7, 2021:1015–1024
Definition 3
2013 2014 2015 2016 2017
H is number of heavy (or extremely heavy) PM2.5
pollution events days in a certain year.
Id ¼ ðOR
1Þ  A  H  D (Equation 2)
Where OR is estimated risk with PM2.5 heavy (or
extremely heavy) pollution events; A is the average
number of daily patient admissions during the study
period; H is the number of heavy (or extremely heavy)
PM2.5 pollution events days in a certain year; and D is
average days in hospital for each hospital admission.
To test the robustness of the main analysis, we
added O 3, NO 2, SO 2, and CO as covariates one by one
into the main model as the sensitivity analysis
models.
RESULTS
In total, 2,202,244 hospital admissions were observed
for cardiovascular disease from 2013 to 2017 in Beijing
(Table 1). The average age was 64 years (standard
deviation 16 years). More than one-half of the pa-
tients were men (58.48%). Angina was the most
common specific cause of hospital admission for
cardiovascular disease (513,002 person-times). More
than nine-tenths of patients were married. The major
nationality among participants was Han.
For each definition of heavy and extremely heavy
PM2.5 pollution events (Figure 1), the number of
JACC VOL. 78, NO. 10, 2021 Zhang et al. 1019
SEPTEMBER 7, 2021:1015–1024 Heavy Pollution and CVD

T A B L E 2 Acute Effects of Heavy or Extremely Heavy PM 2.5 Pollution Events on Cardiovascular Disease Under Different Definitions a

Heavy PM2.5 Pollution Events

Definition 1 Definition 2 Definition 3

75-149 mg/m3 $150 m g/m3 75-149 mg/m3 $150 mg/m 3 75-149 m g/m3 $150 mg/m 3

Cardiovascular disease 1.011 (1.007-1.015) 1.062 (1.055-1.068) 1.017 (1.013-1.021) 1.072 (1.065-1.079) 1.002 (0.997-1.006) 1.085 (1.077-1.093)
Angina 1.024 (1.015-1.032) 1.09 (1.077-1.104) 1.033 (1.025-1.042) 1.101 (1.086-1.116) 1.009 (1-1.018) 1.112 (1.095-1.13)
Myocardial infarction 1.024 (1.008-1.041) 1.061 (1.036-1.087) 1.014 (0.997-1.03) 1.066 (1.039-1.093) 1.004 (0.987-1.021) 1.068 (1.037-1.1)
Chronic ischemic heart disease 1.01 (0.995-1.025) 1.072 (1.048-1.096) 1.014 (0.999-1.029) 1.081 (1.056-1.107) 0.992 (0.977-1.007) 1.109 (1.078-1.14)
Ischemic stroke 1.004 (0.995-1.013) 1.05 (1.036-1.065) 1.004 (0.995-1.014) 1.058 (1.043-1.074) 0.998 (0.989-1.008) 1.071 (1.053-1.09)
Hemorrhagic stroke 0.992 (0.969-1.016) 0.98 (0.946-1.016) 0.986 (0.963-1.01) 0.992 (0.956-1.03) 0.97 (0.947-0.995) 0.994 (0.952-1.039)
Heart failure 1.03 (1.01-1.05) 1.052 (1.021-1.084) 1.019 (0.999-1.039) 1.061 (1.028-1.095) 1.011 (0.99-1.032) 1.06 (1.021-1.101)
Arrhythmia 1.013 (0.998-1.028) 1.033 (1.009-1.058) 1.019 (1.003-1.035) 1.031 (1.005-1.057) 1.001 (0.985-1.017) 1.057 (1.026-1.089)
Hypertensive heart disease 0.998 (0.981-1.014) 1.053 (1.027-1.08) 1.005 (0.989-1.022) 1.074 (1.046-1.103) 0.982 (0.965-0.999) 1.086 (1.052-1.12)
Pulmonary embolism 1.013 (0.973-1.055) 1.047 (0.985-1.114) 1.017 (0.976-1.059) 1.048 (0.981-1.119) 0.998 (0.957-1.041) 1.046 (0.969-1.13)

Values are OR (95% CI). aReferences for ORs were days of non-heavy and non-extremely heavy pollution events under the certain definition.
PM2.5 ¼ fine particulate matter with an aerodynamic diameter #2.5 mm.

extremely heavy pollution events (PM 2.5 concentra-
tion $150 mg/m 3) decreased year by year. There were
59, 52, 48, 40, and 23 days of extremely heavy
pollution events in 2013, 2014, 2015, 2016, and 2017,
respectively, and 222 days in total. Further informa-
tion detailing air pollutants and meteorological fac-
tors is shown in Supplemental Tables 1 and 2.
Table 2 reports the results of OR estimated values
for 3 different definitions of heavy or extremely
heavy PM2.5 pollution events. For total cardiovascular
disease, the ORs were 1.062 (95% CI: 1.055-1.087),
1.072 (95% CI: 1.065-1.079), and 1.085 (95% CI: 1.077-
1.093) under definition 1, definition 2, and definition 3
for extremely heavy PM2.5 pollution events respec-
tively. Most of the specific causes of hospital admis-
sions for cardiovascular disease significantly were
associated with heavy PM 2.5 pollution events or
extremely heavy PM 2.5 pollution events, except hem-
orrhagic stroke. As shown in Supplemental Table 3,
after introducing gaseous pollutants one by one into
the main model, the estimated values of the models
were similar. Therefore, our results are robust.
The average increased admission numbers for
cardiovascular disease annually was 1,322 (95% CI:
839-1,806) and 3,311 (95% CI: 2,969-3,655) admissions
during heavy and extremely heavy PM 2.5 pollution
events lasting for 1 day or more in Beijing with a PM 2.5
concentration 75-149 mg/m 3 and $150 m g/m 3, respec-
tively (Table 3). Increased admission days related to
heavy and extremely heavy pollution events lasting
for 1 day or more were 14,781 days (95% CI: 9,381-
20,193 days) and 37,020 days (95% CI: 33,196-40,866
days), respectively. Increased admission numbers
and days staying in hospital for specific causes of
hospital admission for cardiovascular disease are
shown in Figure 2.
DISCUSSION
This study identified heavy (or extremely heavy)
PM2.5 pollution events as an independent risk factor
and comprehensively analyzed their association with
cardiovascular disease admission over the several
years of the study period. We found significant acute
effects of heavy and extremely heavy PM2.5 pollution
events on total cardiovascular disease hospital
admission and specific causes of hospital admission
for cardiovascular disease (Central Illustration). Heavy
(or extremely heavy) PM2.5 pollution events exposure
led to a high risk of cardiovascular disease admission
in contrast to non-heavy PM2.5 pollution days.
Furthermore, extremely heavy PM2.5 pollution events
lasting 3 days or more resulted in more serious
adverse health effects. Heavy and extremely heavy
PM2.5 pollution events resulted in thousands of extra
cardiovascular disease admissions per year, with
angina representing the greatest cause of hospital
admission among all the specific causes of hospital
admission for cardiovascular disease.
We found that extremely heavy PM2.5 pollution
events had adverse effects on hospital admissions for
cardiovascular disease. Previous studies reported the
adverse health effects of serious air pollution events
in developed countries, such as the London smog in
1952, the Meuse Valley event in 1930, and the Donora
PA event in 1948 (4,7,8). However, due to the absence
1020 Zhang et al. JACC VOL. 78, NO. 10, 2021

Heavy Pollution and CVD SEPTEMBER 7, 2021:1015–1024

T A B L E 3 Increased Hospital Admission Numbers and Days for Cardiovascular Disease With Heavy and Extremely Heavy PM 2.5 Pollution
Events Lasting 1 Day or More (Definition 1) a

Increased Admission Numbers Increased Admission Days

Year 75-149 mg/m 3 $150 mg/m3 75-149 m g/m3 $150 m g/m3

2013 1,327 (843-1,814) 3,865 (3,466-4,267) 15,793 (10,033-21,589) 45,998 (41,249-50,782)

2014 1,545 (981-2,111) 3,619 (3,246-3,995) 17,836 (11,325-24,370) 41,778 (37,472-46,119)
2015 1,303 (827-1,780) 3,584 (3,215-3,957) 14,652 (9,300-20,016) 40,303 (36,153-44,497)
2016 1,387 (881-1,896) 3,214 (2,883-3,548) 15,060 (9,566-20,587) 34,899 (31,305-38,525)
2017 947 (602-1,295) 1,903 (1,706-2,100) 9,807 (6,234-13,410) 19,707 (17,666-21,747)
Average of study years 1,322 (839-1,806) 3,311 (2,969-3,655) 14,781 (9,381-20,193) 37,020 (33,196-40,866)

Values are estimated value (95% CI). aReference for increased hospital admission numbers or days was PM2.5 concentration <75 mg/m3.
PM2.5 ¼ fine particulate matter with an aerodynamic diameter #2.5 mm.

of health information technology in the early 20th
century, these studies cannot distinguish the specific
disease of mortality and morbidity nor the specific
particle size of particulate pollution. Chen et al (5)
reported the adverse effects of extremely heavy
pollution events in 2013 in China using the episode
study design and observed extra cardiovascular
disease–related hospital admissions, which is consis-
tent with our study. Recently conducted, largescale
multicenter studies for PM 2.5 and hospital admission
showed that, except for total cardiovascular disease,
PM 2.5 concentration increase was significantly asso-
ciated with specific causes, such as heart failure,
myocardial infarction, ischemic stroke, and
arrhythmia (13,14,17), and similar affected specific
diseases were found in our study. Tian et al (14) also
demonstrated that ischemic heart disease, ischemic
stroke, and arrhythmia had higher hospital admission
risk estimate values than the total cardiovascular
hospital admission. Unfortunately, high-risk–specific
cardiovascular disease information was not available
in previous heavy PM2.5 pollution event studies. More
detailed disease information is necessary for targeted
health protection.
Among the different definitions of heavy PM2.5
pollution events, we found the higher concentra-
tion and longer lasting heavy PM2.5 pollution
resulted in a greater hospital admission risk for
cardiovascular disease. This may be because under
extremely PM 2.5 heavy pollution, more PM2.5 enters
the human body and engenders more damage to
the cardiovascular system through oxidative stress
and inflammation, abnormal activation of the he-
mostatic system, disturbance of autonomic nervous
system, and injury of endothelial cells (27). Ex-
periments in mice have shown that continuous
exposure to high concentrations of PM2.5 made the
heart decompensate, which ultimately lead to
pathological changes and decreased heart function
(11,12). Therefore, it is reasonable that at the onset
of pollution events, some individuals maintained a
relatively normal level of cardiovascular function in
the short term through the mechanism of
compensation. As the duration of the heavy pollu-
tion events increased, the ability to compensate
was exhausted, leading to an increase in cardio-
vascular disease and more cardiovascular disease
hospital admissions.
For each specific cardiovascular disease, under 2 or
more heavy PM 2.5 pollution definitions, exposure to
heavy (or extremely heavy) PM 2.5 pollution events
showed significant positive associations, except
hemorrhagic stroke. Among which, angina had the
highest OR estimated value and caused the greatest
burden among the specific causes, followed by
ischemic stroke, during extremely heavy PM2.5
pollution events. The possible mechanism is that
short term exposure to PM2.5 resulted in increased
blood pressure, heart rate, plaque vulnerability, and
thrombosis, and eventually induced angina and
ischemic stroke attack (28-31). The sensitive specific
cardiovascular diseases (myocardial infarction,
ischemic heart disease, ischemic stroke, heart failure,
and arrhythmia) were consistent with previous
studies of PM2.5 concentration conducted in China
(14,32). After conducting a comprehensive analysis
for PM2.5 concentration exposure among 214 mutually
exclusive disease groups, Wei et al (13) found signif-
icant association between PM 2.5 concentration and
congestive heart failure, acute myocardial infarction,
and cardiac dysrhythmias in the United States. They
did not observe a risk of ischemic stroke hospital
admission in the main analysis, which suggests that
ischemic stroke is sensitive to multiple exposure to
excessive PM2.5 pollution, and was overlooked in
developed countries with low PM2.5 concentrations.
JACC VOL. 78, NO. 10, 2021 Zhang et al. 1021
SEPTEMBER 7, 2021:1015–1024 Heavy Pollution and CVD

F I G U R E 2 Increased Hospital Admissions Related to Heavy PM 2.5 Pollution Events

A Increased Admission Numbers per Year

Angina

Ischemic Stroke

Chronic Ischemic Heart Disease

Hypertensive Heart Disease

Myocardial Infarction

Arrhythmia

Heart Failure

Pulmonary Embolism

0 500 1,000
Increased Admission Number

B Increased Admission Days per Year

Angina

Ischemic Stroke

Chronic Ischemic Heart Disease

Hypertensive Heart Disease

Myocardial Infarction

Arrhythmia

Heart Failure

Pulmonary Embolism

0 4,000 8,000
Increased Admission Days

(A) Increased admission numbers for specific cardiovascular disease per year; (B) increased admission days for specific cardiovascular disease
per year. Blue represents extremely heavy PM2.5 pollution events (PM2.5 concentration $150 mg/m3); red represents heavy PM2.5 pollution
events (PM2.5 concentration 75-149 mg/m3). Reference for increased hospital admission numbers or days in hospital was PM2.5
concentration <75 mg/m3. PM2.5 ¼ fine particulate matter with an aerodynamic diameter #2.5 mm.
1022 Zhang et al. JACC VOL. 78, NO. 10, 2021

Heavy Pollution and CVD SEPTEMBER 7, 2021:1015–1024

C E N T R A L IL LU ST R A T I O N Heavy PM 2.5 Pollution Events Increased the Risks of Hospital Admission for
Cardiovascular Diseases

Zhang, Y. et al. J Am Coll Cardiol. 2021;78(10):1015–1024.

The framework and main findings of this study on association between heavy PM2.5 pollution event and hospital admission for cardiovascular disease. The
odds ratios were estimated by using a conditional logistic regression model in time-stratified case-crossover design. PM2.5 ¼ fine particulate matter.

Based on the results of our study, health protection besides the average concentration of PM2.5 needing to
measures should be implemented against the adverse be reduced, the peak concentration of PM 2.5 should be
effects from the heavy and extremely heavy PM 2.5 urgently eliminated, in order to reduce the burden of
pollution events. With regard to the policymakers, heavy PM 2.5 pollution events. The policymakers
JACC VOL. 78, NO. 10, 2021
SEPTEMBER 7, 2021:1015–1024
would furthermore need to establish an early warning
system for health protection during heavy PM 2.5
pollution events, especially to warn the
vulnerable populations to take appropriate protection
measures. With respect to hospitals and doctors, it is
necessary to strengthen the education and awareness
of preparing for the increase of cardiovascular ad-
missions before heavy PM 2.5 pollution events, such as
adding cardiovascular hospital beds and paying more
attention to the occurrence of acute cardiovascular
events. Lastly, it is necessary for patients to be better
educated about the risks for them of adverse health
effects during the heavy PM 2.5 pollution events, so
that they know how to minimize their exposure to
heavy PM2.5 pollution events in order to prevent the
occurrence of cardiovascular disease, and seek timely
medical attention to address symptoms
discomfort.
STUDY STRENGTHS. This study encompasses several
strengths. The case-crossover design and the condi-
tional logistic regression model employed here
allowed for controlling for unchanged confounders,
such as age, sex, and long-term smoking history, over
a short time period (33). Furthermore, by using the
case-crossover design and our heavy pollution events
definitions, we generalized the heavy PM2.5 pollution
events together, which occurred each year during the
study period, and can be used for reference by other
developing countries with frequently occurring heavy
PM 2.5 pollution events. In addition, Beijing is home to
the best hospitals in China, and its residents are un-
likely to have hospital admissions elsewhere.
Accordingly, it can be considered that we have
observed all the hospital admissions of Beijing resi-
dents, which reduced the selection bias. Lastly, the
hospital admission surveillance data consisted of ac-
curate and nonmissing ICD-10 codes, covering all
types of cardiovascular diseases.
STUDY LIMITATIONS. Despite these numerous
strengths, there are several limitations in our study.
First, we used the city-level average concentrations
of PM2.5 obtained from fixed surveillance monitoring
stations, which may have resulted in information bias
for the exposure assessment (34). However, we
collected PM2.5 concentration data only to identify
whether it was a heavy (or extremely heavy) pollution
event, and then used the heavy (or extremely heavy)
pollution events as the exposure. Thus, it will reduce
the impact of information bias on research results.
Zhang et al. 1023
Heavy Pollution and CVD
Second, a case-crossover study is suitable for acute
events of disease. However, considering that hospital
most admission for chronic diseases may suggest an acute
exacerbation, we included them in this study.
Because we cannot confirm whether each case was
indeed an acute exacerbation, it may cause selection
bias. Third, we conducted the single-city study in
Beijing; the results can only be generalized to cities of
similar socioeconomic development status. Multi-
center studies focus on the association between
heavy PM2.5 pollution events and hospital admission
should be further conducted.
CONCLUSIONS
Heavy and extremely heavy PM2.5 pollution events
resulted in substantially increased hospital admission
and
risk for cardiovascular disease. With higher PM2.5
concentration and longer duration of heavy PM2.5
pollution events, a greater risk of cardiovascular
hospital admission was observed.
FUNDING SUPPORT AND AUTHOR DISCLOSURES
This work was funded by National Research Program for Key Issues in
Air Pollution Control grant DQGG0401, the Beijing Natural Science
Foundation grant 7172145, and the National Natural Science Foun-
dation of China grant 92043301. The authors have reported that they
have no relationships relevant to the contents of this paper to
disclose.
ADDRESS FOR CORRESPONDENCE: Dr Tiantian Li,
National Institute of Environmental Health, Chinese
Center for Disease Control and Prevention, No. 7,
Panjiayuan Nanli, Chaoyang District, Beijing 100021,
China. E-mail: litiantian@nieh.chinacdc.cn. OR Dr
Xiaoming Shi, National Institute of Environmental
Health, Chinese Center for Disease Control and Pre-
vention, No. 7, Panjiayuan Nanli, Chaoyang District,
Beijing 100021, China. E-mail: shixm@chinacdc.cn.
PERSPECTIVES
COMPETENCY IN MEDICAL KNOWLEDGE: Heavy and
extremely heavy PM2.5 pollution events are associated with
substantial increase in hospital admission for cardiovascular
disease.
TRANSLATIONAL OUTLOOK: Further studies are needed to
explore the association between components or sources of PM2.5
and cardiovascular disease.
1024 Zhang et al.
Heavy Pollution and CVD
REFERENCES
1. Health Effects Institute. State of Global Air /
2019. Special Report. Boston, Massachusetts:
Health Effects Institute; 2019.
2. Ministry of Ecology and Environment. Report on
the State of Ecological Environment in China.
Ministry of Ecology and Environment of the Peo-
ple’s Republic of China; 2018.
3. Ministry of Ecology and Environment. Report on
the State of Ecological Environment in China/2017.
Ministry of Ecology and Environment of the Peo-
ple’s Republic of China; 2018.
4. Bell ML, Davis DL. Reassessment of the lethal
London fog of 1952: novel indicators of acute and
chronic consequences of acute exposure to air
pollution. Environ Health Persp. 2001;109(Suppl
3):389–394.
5. Chen R, Zhao Z, Kan H. Heavy smog and hos-
pital visits in Beijing. Am J Resp Crit Care.
2013;188:1170–1171.
6. Ferreri JM, Peng RD, Bell ML, Liu Y, Li T,
Anderson GB. The January 2013 Beijing “air-
pocalypse” and its acute effects on emergency and
outpatient visits at a Beijing hospital. Air Qual
Atmos Health. 2018;11(3):301–309. https://doi.
org/10.1007/s11869-017-0538-0
7. Davis DL, Bell ML, Fletcher T. A look back at the
London smog of 1952 and the half century since.
Environ Health Persp. 2002;110(12):A734–A735.
https://doi.org/10.1289/ehp.110-a734
8. Anderson HR. Air pollution and mortality: a
history. Atmos Environ. 2009;43:142–152.
9. Hunt A, Abraham JL, Judson B, Berry CL. Toxi-
cologic and epidemiologic clues from the charac-
terization of the 1952 London smog fine
particulate matter in archival autopsy lung tissues.
Environ Health Persp. 2003;111:1209–1214.
10. Fiordelisi A, Piscitelli P, Trimarco B, Coscioni E,
Iaccarino G, Sorriento D. The mechanisms of air
pollution and particulate matter in cardiovascular
diseases. Heart Fail Rev. 2017;22:337–347.
11. Li D, Zhang R, Cui L, et al. Multiple organ injury
in male C57BL/6J mice exposed to ambient par-
ticulate matter in a real-ambient PM exposure
system in Shijiazhuang, China. Environ Pollut.
2019;248:874–887.
12. Su X, Tian J, Li B, et al. Ambient PM2.5 caused
cardiac dysfunction through FoxO1-targeted car-
diac hypertrophy and macrophage-activated
fibrosis in mice. Chemosphere. 2020;247:125881.
13. Wei Y, Wang Y, Di Q, et al. Short term exposure
to fine particulate matter and hospital admission
risks and costs in the Medicare population: time
stratified, case crossover study. Brit Med J.
2019;367:l6258.
14. Tian Y, Liu H, Wu Y, et al. Association between
ambient fine particulate pollution and hospital
admissions for cause specific cardiovascular dis-
ease: time series study in 184 major Chinese cities.
Brit Med J. 2019;367:l6572.
15. Atkinson RW, Kang S, Anderson HR, Mills IC,
Walton HA. Epidemiological time series studies of
PM2.5 and daily mortality and hospital admissions:
a systematic review and meta-analysis. Thorax.
2014;69:660–665.
16. Dominici F, Peng RD, Bell ML, et al. Fine par-
ticulate air pollution and hospital admission for
cardiovascular and respiratory diseases. JAMA.
2006;295:1127–1134.
17. Lanzinger S, Schneider A, Breitner S, et al.
Ultrafine and fine particles and hospital admis-
sions in central Europe. results from the UFIREG
study. Am J Resp Crit Care. 2016;194:1233–1241.
18. Bell ML, Ebisu K, Peng RD, et al. Seasonal and
regional short-term effects of fine particles on
hospital admissions in 202 US counties, 1999-
2005. Am J Epidemiol. 2008;168:1301–1310.
19. Liu C, Chen R, Sera F, et al. Ambient particulate
air pollution and daily mortality in 652 cities.
N Engl J Med. 2019;381:705–715.
20. Dudas K, Lappas G, Rosengren A. Long-term
prognosis after hospital admission for acute
myocardial infarction from 1987 to 2006. Int J
Cardiol. 2012;155:400–405.
21. Fujimi K, Imaizumi T, Suematsu Y, et al. Dif-
ferential prognostic impact between completion
and non-completion of a 5-month cardiac reha-
bilitation program in outpatients with cardiovas-
cular diseases. Int J Cardiol. 2019;292:13–18.
22. Hirst NG, Whitty JA, Synnott RL, Eley DS,
Scuffham PA. Predictors of government subsidized
pharmaceutical use in patients with diabetes or
cardiovascular disease in a primary care setting:
evidence from a prospective randomized trial.
J Med Econ. 2011;14:698–704.
23. Hajat C, Stein E. The global burden of multiple
chronic conditions: a narrative review. Prev Med
Rep. 2018;12:284–293.
24. Fan F, Lei Y, Li L. Health damage assessment
of particulate matter pollution in Jing-Jin-Ji region
of China. Environ Sci Pollut Res Int. 2019;26(8):
7883–7895. https://doi.org/10.1007/s11356-018-
04116-8
25. Beijing Municipal Ecology and Environment
Bureau. Beijing Ecology and Environment
JACC VOL. 78, NO. 10, 2021
SEPTEMBER 7, 2021:1015–1024
Statement. Beijing Municipal Ecology and Envi-
ronment Bureau; 2018.
26. Rich DQ, Kipen HM, Zhang J, Kamat L,
Wilson AC, Kostis JB. Triggering of transmural in-
farctions, but not nontransmural infarctions, by
ambient fine particles. Environ Health Persp.
2010;118:1229–1234.
27. Wang C, Tu Y, Yu Z, Lu R. PM2.5 and car-
diovascular diseases in the elderly: an over-
view. Int J Environ Res Public Health.
2015;12(7):8187–8197. https://doi.org/10.3390/
ijerph120708187
28. Lim YH, Bae HJ, Yi SM, Park E, Lee BE,
Hong YC. Vascular and cardiac autonomic function
and PM(2.5) constituents among the elderly: a
longitudinal study. Sci Total Environ. 2017;607-
608:847–854.
29. Zanobetti A, Canner MJ, Stone PH, et al.
Ambient pollution and blood pressure in cardiac
rehabilitation patients. Circulation. 2004;110:
2184–2189.
30. Suwa T, Hogg JC, Quinlan KB, Ohgami A,
Vincent R, van Eeden SF. Particulate air pollution
induces progression of atherosclerosis. J Am Coll
Cardiol. 2002;39:935–942.
31. An Z, Jin Y, Li J, Li W, Wu W. Impact of par-
ticulate air pollution on cardiovascular health. Curr
Allergy Asthma Rep. 2018;18(3):15. https://doi.
org/10.1007/s11882-018-0768-8
32. Gu J, Shi Y, Chen N, Wang H, Chen T. Ambient
fine particulate matter and hospital admissions for
ischemic and hemorrhagic strokes and transient
ischemic attack in 248 Chinese cities. Sci Total
Environ. 2020;715:136896.
33. Maclure M. The case-crossover design: a
method for studying transient effects on the risk
of acute events. Am J Epidemiol. 1991;133:144–
153.
34. Mannshardt E, Sucic K, Jiao W, et al.
Comparing exposure metrics for the effects of fine
particulate matter on emergency hospital admis-
sions. J Expo Sci Environ Epidemiol. 2013;23:627–
636. https://doi.org/10.1038/jes.2013.39
KEY WORDS cardiovascular disease, fine
particulate matter, heavy pollution event,
hospital admission
A PPE NDI X For a supplemental figure and
tables, please see the online version of this
paper.
JOURNAL OF THE AMERICAN COLLEGE OF CARDIOLOGY
ª 2021 BY THE AMERICAN COLLEGE OF CARDIOLOGY FOUNDATION
PUBLISHED BY ELSEVIER
EDITORIAL COMMENT
How Tweets Can Cause a Revolution*
Sanjay Rajagopalan, MD,a,b Robert D. Brook, MD,c Sadeer Al-Kindi, MDa,b
A lmost 13 years ago, in 2008, a solitary air
pollution monitor located at the roof top of
the U.S. embassy in the Chaoyang district in
Beijing, China (@BeijingAirTwitter) began tweeting
hourly air pollution levels (1). The sensor was meant
to provide warnings of bad-air days to U.S. citizens
at the time of the Beijing Olympics. On a particularly
smoggy day in Beijing in November 2010, the nor-
mally restrained programmer responsible for the
Twitter feed used the epithet “Crazy bad” to describe
a value that was beyond the range of the U.S. Envi-
ronmental Protection Agency’s air quality index, a
score typically ranging from 1 to 500 (2). The Twitter
description of air quality in Beijing nearly sparked a
diplomatic furor. The existence of the embassy’s
monitor and its continual feed had been an ongoing
irritant for Chinese Ministry officials, since before
the 2008 Olympics, when sensor readings routinely
undercut official rosy pronouncements of “blue sky”
days. Chinese officials had asked the U.S. Embassy
to halt the Twitter feed. They claimed that the data
were not only confusing but incorrect and that
continued channeling of air pollution levels through
social media could lead to “social consequences.”
Subsequent corroboration by us and others did
indeed prove that the levels reported by the U.S. Em-
bassy monitor were correct (Figures 1A and 1B) and
attested to the poor air quality in Beijing (3). What
happened within 2 years of that fateful night in
November was certainly consequential from a societal
*Editorials published in the Journal of the American College of Cardiology
reflect the views of the author and do not necessarily represent the views
of JACC or the American College of Cardiology.
From the aDivision of Cardiovascular Medicine, Harrington Heart and
Vascular Institute, University Hospitals, Cleveland, Ohio USA; bDivision
of Cardiovascular Medicine, Cardiovascular Research Institute, Case
Western Reserve University, Cleveland, Ohio USA; and the cDivision of
Cardiology, Wayne State University, Detroit, Michigan USA.
The authors attest they are in compliance with human studies commit-
tees and animal welfare regulations of the authors’ institutions and Food
and Drug Administration guidelines, including patient consent where
appropriate. For more information, visit the Author Center.
ISSN 0735-1097/$36.00
VOL. 78, NO. 10, 2021
standpoint. Governmental policy toward air pollution
regulation witnessed a seismic shift, with creation of
a national monitoring program, stringent controls,
and a massive ongoing overhaul of energy sources.
The city of Beijing established the Beijing Environ-
mental Monitoring Center, a network of 36 air quality
monitoring sites in Beijing.
SEE PAGE 1015
From this context, the study reported in this
issue of the Journal by Zhang et al (4) on cardio-
vascular hospitalizations in Beijing area hospitals
between 2013 and 2019 was conducted during a
highly consequential period. In general, the findings
were entirely consistent with the known effects of
air pollution, with both dose and duration of heavy
antecedent exposure increasing the risk for hospi-
talization. Hundreds of time series, case crossover,
and prospective cohort studies have indeed
demonstrated a robust relationship between partic-
ulate matter #2.5 m m (PM 2.5) air pollution and car-
diovascular events, including myocardial infarction,
stroke, cardiovascular mortality, and heart failure.
During the same period as the study, we conducted
studies in Beijing that provided a mechanistic basis
for increased cardiovascular events, including in-
creases in blood pressure, vascular stiffness, and
elevated biomarkers of plaque vulnerability (2,3,5).
Zhang et al (4) used 3 different definitions of air
pollution on a graded incremental scale that
included degree and duration of severity. Heavy air
pollution was defined in 3 ways as daily average
PM 2.5 $75 mg/m 3 lasting $1, 2, or 3 days, whereas
very heavy air pollution was defined as $150 m g/m3
lasting $1, 2, or 3 days. To provide perspective, U.S.
National Ambient Air Quality Standards mandate
daily levels <35 m g/m 3. Thus, although the level
cutoffs used by Zhang et al (4) are arbitrary, they
were >2- to 5-fold higher than acceptable air quality
standards in the United States.
The study used a time-stratified referent selection
(the day before or after the event) to explore the
https://doi.org/10.1016/j.jacc.2021.07.009
1026 Rajagopalan et al. JACC VOL. 78, NO. 10, 2021

Twitter and Revolution SEPTEMBER 7, 2021:1025–1027

F I G U R E 1 Air Pollution Quality in Beijing, China

A 60
300
50

200
40
Black Carbon (µg/m3)

100

PM2.5 (µg/m3)
30

0
20

−100
10

0 −200

−10 −300
14

3
4

/0

/0
/1

/2

/1
/0

/1

/2

/2

/2
/2

/0

/1

/0

/1
/

07
06
03
02

04

05

06
03
02

04

05

07
06
03

04

05

Date: February 14, 2012 – July 14, 2012

PM2.5 - U.S. Embassy PM2.5 - BEMC PM2.5 - Tsinghua BC - Personal BC - PUMC

B Average Daily Air Quality Index (AQI) Values

250
201-300
Very
Unhealthy
200
151-200
Moderate
150
AQI Scale

101-150
Unhealthy for
Sensitive Groups
100
51-100
Moderate
50
0-50
Good
0
Jan 2015 Jan 2016 Jan 2017 Jan 2018 Jan 2019 Jan 2020 Jan 2021

(A) Time series of particulate matter #2.5 mm (PM2.5) and black carbon (BC) levels in Beijing during a period of hazardous levels of air pollution in 2012. PM2.5 was
monitored using Twitter data from the U.S. Embassy and at Tsinghua University and Beijing Environmental Monitoring Center (BEMC) (represents a blended average
across stations). From Zhao et al (3). (B) Average daily air quality index in Beijing between January 2015 and January 2021. Data derived from (7). PUMC ¼ Peking Union
Medical College.
JACC VOL. 78, NO. 10, 2021 Rajagopalan et al. 1027
SEPTEMBER 7, 2021:1025–1027 Twitter and Revolution

association between episodes characterized as heavy hospitalizations attributable to decreases in air

or very heavy air pollution and cardiovascular disease
admissions. The strengths of this study include the
relatively large number of cardiovascular hospitali-
zations (n ¼ 2,202,244), multipollutant sensitivity
analyses, and cause-specific attribution of cardio-
vascular hospitalizations using International Classi-
fication of Diseases-10th Revision (ICD-10) codes. The
number of excess hospitalizations and the duration in
days of these hospitalizations were also calculated.
Additional sensitivity analyses were conducted using
other co-pollutant measures to test the robustness of
the association with PM 2.5 .
Angina was the leading cause of cardiovascular
hospitalization followed by ischemic stroke, resulting
in the greatest burden of attributable days of hospi-
talization. Extremely heavy air pollution levels
lasting $3 days were associated with an 8% increase
in hospitalizations resulting from a range of cardio-
vascular causes. The effect estimates in this study
were consistent with the known exposure-response
relationship between cardiovascular events and air
pollution levels (6).
This study has several limitations, including the
fact that the monitors were regional stations at
fixed locations, with attribution of “exposure” level
for each index case to the closest monitor. Not only
is this not a true depiction of exposure, but also,
importantly, it may result in exposure misclassifi-
cation. However, most of the time, this misclassifi-
cation has the effect of attenuating effect estimates,
rather than amplifying them. ICD-10 estimates were
used, which in many cases are well known not to
correspond to the index event that precipitated the
hospitalization. A missed opportunity in this study
may have been to compare and contrast other
noncardiovascular causes, to delineate differential
effects of air pollution, and to compare and contrast
effect estimates. Additionally, there could have
been an attempt to examine the reduction in
pollution levels over the duration of the study,
given the higher levels earlier in the study and the
lower levels later.
The most important contribution of the paper may
be that it highlights China’s success in reducing air
pollution levels within less than a decade. As of 2020,
China’s air pollution levels had improved dramati-
cally, with annual averages across China of 33 m g/m 3
compared with 81 m g/m3 in 2015. Average annual
levels in Beijing have fallen to approximately
38.7 m g/m 3, with daily peak average air quality index
scores in 2020 reaching 262, well below “crazy bad.”
Although this by no means suggests that air in Beijing
or the rest of China is clean, the rapidity of progress
and commitment to reducing greenhouse gas emis-
sions and air pollution are noteworthy. As part of the
Paris Agreement and additional announcements by
Chinese Premier Xi Jinping, China has committed to
peaking its carbon emissions by 2030 and be carbon
neutral by 2060. Whether these goals are accom-
plished remains to be seen, given the trade-off be-
tween making up for lost growth after COVID-19 and
China’s commitment to its citizens’ social and envi-
ronmental well-being. At least for 2021, the Chinese
government has set air quality targets at a higher
level, 34.5 m g/m 3, than the 2020 annual average PM2.5
level of 33.3 mg/m 3. Forgoing an opportunity to
regulate air pollution levels would only mean that
China may be farther away from achieving a healthy
environment for its citizenry.
FUNDING SUPPORT AND AUTHOR DISCLOSURES
The authors have reported that they have no relationships relevant to
the contents of this paper to disclose.
ADDRESS FOR CORRESPONDENCE: Dr Rajagopalan,
Harrington Heart & Vascular Institute, University Hos-
pitals, 11100 Euclid Avenue, Mailstop Lakeside 5038,
Cleveland, Ohio 44106 USA. E-mail: sxr647@case.edu.

REFERENCES

1. Kintisch E. Rooftop sensors on U.S. embassies are
warning the world about ‘crazy bad’ air pollution.
2018. Science; 2018. Accessed June 30, 2021. https://
www.sciencemag.org/news/2018/04/rooftop-
sensors-us-embassies-are-warning-world-about-
crazy-bad-air-pollution
2. Rajagopalan S, Brauer M, Bhatnagar A,
et al. Personal-level protective actions
against particulate matter air pollution
exposure: a scientific statement from the
American Heart Association. Circulation.
2020;142:411–431.
3. Zhao X, Sun Z, Ruan Y. Yet al. Personal black car-
bon exposure influences ambulatory blood pressure:
air pollution and cardiometabolic disease (AIRCMD-
China) study. Hypertension. 2014;63:871–877.
4. Zhang Y, Ma R, Ban J, et al. Risk of cardiovas-
cular hospital admission after exposure to fine
particulate pollution. J Am Coll Cardiol. 2021;78:
1015–1024.
5. Xu H, Wang T, Liu S, et al. Extreme levels
of air pollution associated with changes in
biomarkers of atherosclerotic plaque
vulnerability and thrombogenicity in healthy
adults: the Beijing AIRCHD study. Circ Res.
2019;118:313948.
6. Rajagopalan S, Al-Kindi SG, Brook RD. Air
pollution and cardiovascular disease. J Am Coll
Cardiol. 2018;72:2054–2070.
7. AirNow. Available at: https://www.airnow.gov/.
KEY WORDS air pollution, cardiovascular,
China, hospitalization

Listen to this manuscript’s
audio summary by
Editor-in-Chief
Dr. Valentin Fuster on
JACC.org.
JOURNAL OF THE AMERICAN COLLEGE OF CARDIOLOGY VOL. 78, NO. 10, 2021
ª 2021 BY THE AMERICAN COLLEGE OF CARDIOLOGY FOUNDATION
PUBLISHED BY ELSEVIER
Effect of Heart Rate on the Outcome of
Renal Denervation in Patients With
Uncontrolled Hypertension
Michael Böhm, MD,a Konstantinos Tsioufis, MD,b David E. Kandzari, MD,c Kazuomi Kario, MD, PHD,d
Michael A. Weber, MD,e Roland E. Schmieder, MD,f Raymond R. Townsend, MD,g Saarraaken Kulenthiran, MD,a
Christian Ukena, MD,a Stuart Pocock, PHD,h Sebastian Ewen, MD,a Joachim Weil, MD,i Martin Fahy, MS,j
Felix Mahfoud, MDa
ABSTRACT
BACKGROUND Sham-controlled trials demonstrated safety and efficacy of renal denervation (RDN) to lower blood
pressure (BP). Association of baseline heart rate with BP reduction after RDN is incompletely understood.
OBJECTIVES The purpose of this analysis was to evaluate the impact of baseline heart rate on BP reduction without
antihypertensive medications in the SPYRAL HTN-OFF MED (Global Clinical Study of Renal Denervation With the Sym-
plicity Spyral Multi-electrode Renal Denervation System in Patients With Uncontrolled Hypertension in the Absence of
Antihypertensive Medications) Pivotal trial.
METHODS Patients removed from any antihypertensive medications were enrolled with office systolic blood pressure
(SBP) $150 and <180 mm Hg and randomized 1:1 to RDN or sham control. Patients were separated according to baseline
office heart rate <70 or $70 beats/min. BP changes from baseline to 3 months between treatment arms were adjusted
for baseline SBP using analysis of covariance.
RESULTS Scatter plots of 3-month changes in 24-hour and office SBP illustrate a wide range of changes in SBP for
different baseline heart rates. Treatment difference at 3 months between RDN and sham control with baseline office
heart rate $70 beats/min for 24-hour SBP was
6.2 mm Hg (95% CI:
9.0 to
3.5 mm Hg) (P < 0.001) and for baseline
office heart rate <70 beats/min it was
0.1 mm Hg (
3.8 to 3.6 mm Hg) (P ¼ 0.97) with an interaction P value of 0.008.
Results were similar for changes in office, daytime, and nighttime SBP at 3 months, with a greater reduction in SBP with
baseline office heart rate $70 beats/min.
CONCLUSIONS Reduction in mean office, 24-hour, daytime, and nighttime SBP for RDN at 3 months was greater with
baseline office heart rate $70 than <70 beats/min, suggesting an association between baseline heart rate and BP
reduction after RDN. (SPYRAL PIVOTAL—SPYRAL HTN-OFF MED Study; NCT02439749)
(J Am Coll Cardiol 2021;78:1028–1038) © 2021 by the American College of Cardiology Foundation.
From the aKlinik für Innere Medizin III, Universitätsklinikum des Saarlandes, Saarland University, Homburg/Saar, Germany;
b
National and Kapodistrian University of Athens, Hippocratio Hospital, Athens, Greece; cPiedmont Heart Institute, Atlanta,
Georgia, USA; dJichi Medical University School of Medicine, Tochigi, Japan; eSUNY Downstate College of Medicine, Brooklyn, New
York, USA; fUniversitätsklinikum Erlangen, Erlangen, Germany; gPerelman School of Medicine, University of Pennsylvania,
Philadelphia, Pennsylvania, USA; hLondon School of Hygiene and Tropical Medicine, London, United Kingdom; iDepartment of
Cardiology, Sana Cardiomed Heart Center, Lübeck, Germany; and jMedtronic, Santa Rosa, California, USA.
The authors attest they are in compliance with human studies committees and animal welfare regulations of the authors’
institutions and Food and Drug Administration guidelines, including patient consent where appropriate. For more information,
visit the Author Center.
Manuscript received May 14, 2021; revised manuscript received June 25, 2021, accepted June 28, 2021.
ISSN 0735-1097/$36.00 https://doi.org/10.1016/j.jacc.2021.06.044
JACC VOL. 78, NO. 10, 2021
SEPTEMBER 7, 2021:1028–1038
H METHODS
eart rate is regulated by the interaction of
the sympathetic and parasympathetic ner-
vous systems (1) adjusting cardiac output
according to the peripheral demands of the body (2).
However, the association between heart rate and
more direct measures of sympathetic activation has
not been demonstrated (3,4), suggesting that the para-
sympathetic nervous system may also have strong
control over heart rate in concert with the sympathetic
nervous system. Hypertension is the most prevalent
risk factor for mortality (5) and is associated with
increased risk for cardiovascular, all-cause, and coro-
nary death in patients with stable coronary artery dis-
ease (6), high cardiovascular risk (7), and metabolic
disease (8). Increased heart rate is a marker for poor
cardiovascular prognosis in the general population
(9) and in patients with hypertension (10), diabetes
(11), stroke (12), and heart failure with reduced ejection
fraction (13,14) or preserved ejection fraction (15,16).
SEE PAGE 1039
This association has also been demonstrated with
other chronic noncardiovascular conditions such as
chronic obstructive pulmonary disease (17) and cancer
(18), and acute conditions such as hypertensive emer-
gency (19) or shock (20). The threshold resting heart
rate to identify patients with higher risk has been
shown to be approximately $70 beats/min (9-16). In
hypertension, overactivity of the sympathetic ner-
vous system contributes to the development and
progression of arterial hypertension and might be
associated with further aggravation of cardiovascular
risk (21,22). Catheter-based renal denervation (RDN)
reduced blood pressure (BP) in patients without
(23) or with (24) antihypertensive drugs in pilot
studies as well as in a large real-world international
registry (25). Recently, the powered, multicenter,
randomized, sham-controlled SPYRAL HTN-OFF
MED (Global Clinical Study of Renal Denervation
With the Symplicity Spyral Multi-electrode Renal
Denervation System in Patients With Uncontrolled
Hypertension in the Absence of Antihypertensive
Medications) Pivotal trial reported a significant BP
reduction over sham in patients not receiving
concomitant medication (26). As heart rate and BP
interact with the sympathetic nervous system, it
was hypothesized that higher heart rate could lead
to greater BP reduction after RDN. The first evidence
of this was derived from a retrospective analysis of
the GSR (Global SYMPLICITY Registry) (27) and also
the SPYRAL HTN-OFF MED pilot trial (28). In the
current analysis, we evaluate whether higher base-
line office heart rate $70 beats/min results in greater
BP reduction after RDN.
Böhm et al. 1029
Renal Denervation and Heart Rate
ABBREVIATIONS
AND ACRONYMS
STUDY DESIGN. The SPYRAL HTN-OFF MED
BP = blood pressure
Pivotal trial was a multicenter, international,
RDN = renal denervation
prospective, single-blinded randomized,
SBP = systolic blood pressure
sham-controlled trial conducted in accor-
dance with the declaration of Helsinki, and the trial
protocol was approved by the Institutional Review
Boards or ethics committee from 44 study sites in
Australia, Austria, Canada, Germany, Greece, Ireland,
Japan, the United Kingdom, and the United States.
The trial is registered at ClinicalTrials.gov
(NCT02439749). Inclusion and exclusion criteria were
published previously (29). The primary efficacy
endpoint was BP reduction based on 24-hour ambu-
latory BP measurement at 3 months (26,29,30). In this
analysis, we explored reduction in systolic BP in this
trial according to baseline office heart rate <70 beats/
min vs $70 beats/min.
RANDOMIZATION, BLINDING, AND PROCEDURES.
The details of RDN and sham procedures have been
previously described (29,30). All patients were
required to refrain from all antihypertensive medi-
cations and were informed that testing for the pres-
ence of antihypertensive drugs and their metabolites
in serum and urine would be performed (29). BP and
heart rate measurements were obtained with an
automated BP monitor (Omron, Omron Healthcare
Inc), and 24-hour ambulatory blood pressure moni-
toring (Mobil-O-Graph IEM GmbH) was supplied to
evaluate 24-hour mean BP.
STATISTICAL ANALYSIS. Statistical analyses were
performed according to the intention-to-treat princi-
ple. Means and SDs of continuous variables are pre-
sented by treatment group. BP changes between
treatment arms were adjusted for baseline systolic
blood pressure (SBP) using analysis of covariance
models. Daytime is defined as 7:00 AM to 10:00 PM and
nighttime defined as 10:00 PM to 7:00 AM. Interaction P
values between heart rate groups (baseline office
heart rate <70 and $70 beats/min) were determined
from a multivariable model with the following cova-
riates: baseline SBP, baseline heart rate group, treat-
ment arm, and a treatment  baseline heart rate
interaction. The relationship between baseline heart
rate as a continuous measure and SBP change at
3 months was also assessed by fitting a regression
model and using interaction testing to compare the
slopes in the RDN and sham control groups. The
interaction model included the following covariates:
baseline SBP, baseline heart rate, treatment arm, and
a treatment  baseline heart rate interaction. SAS for
Windows version 9.4 or higher (SAS Institute) was
used for all statistical analyses.
1030 Böhm et al. JACC VOL. 78, NO. 10, 2021

Renal Denervation and Heart Rate SEPTEMBER 7, 2021:1028–1038

F I G U R E 1 Patient Flow Diagram

Patients Eligible for Randomization

(N = 366)

Renal denervation Sham Control

(N = 182) (N = 184)

Office heart rate Office heart rate Office heart rate Office heart rate
<70 beats/min ≥70 beats/min <70 beats/min ≥70 beats/min
(N = 72) (N = 110) (N = 60) (N = 124)

3 months 3 months 3 months 3 months

Office SBP (N = 64) Office SBP (N = 106) Office SBP (N = 50) Office SBP (N = 114)
ABPM (N = 60) ABPM (N = 93) ABPM (N = 49) ABPM (N = 98)
Day SBP (N = 60) Day SBP (N = 95) Day SBP (N = 49) Day SBP (N = 100)
Night SBP (N = 61) Night SBP (N = 100) Night SBP (N = 50) Night SBP (N = 101)

Of the 366 patients eligible for randomization in the SPYRAL HTN-OFF MED Pivotal trial, 182 were randomized to renal denervation and 184 to sham
control. Patients were then divided into 2 groups with an office heart rate <70 and $70 beats/min. Numbers of patients with available office heart rate
and complete ABPM data at 3 months are shown. ABPM ¼ ambulatory blood pressure monitoring; SBP ¼ systolic blood pressure.

RESULTS characteristics are shown in Table 1. In the high

In this analysis, a total of 366 patients were ran-
domized to RDN (n ¼ 182) or sham control (n ¼ 184).
Patients were divided into groups based on baseline
heart rate <70 or $70 beats/min (Figure 1). Baseline
baseline office heart rate group, there were more
current smokers.
The relationship between baseline heart rate and 3-
month changes in 24-hour and office SBP were
examined by fitting a regression model and using

T A B L E 1 Baseline Characteristics

Heart Rate <70 beats/min Heart Rate $70 beats/min

RDN Sham Control RDN Sham Control

(n ¼ 72) (n ¼ 60) (n ¼ 110) (n ¼ 124)

Age, y 55.4  10.8 55.8  9.1 50.7  10.4 51.2  10.3

Male 73.6 (53) 85.0 (51) 58.2 (64) 62.1 (77)
Body mass index, kg/m2 30.5  5.9 30.8  4.5 31.6  6.1 31.1  5.9
Diabetes type II 5.6 (4) 3.3 (2) 2.7 (3) 7.3 (9)
Current smoker 6.9 (5) 5.0 (3) 23.6 (26) 21.0 (26)
Obstructive sleep apnea 5.6 (4) 6.7 (4) 10.0 (11) 7.3 (9)
Peripheral artery disease 0.0 (0) 0.0 (0) 0.9 (1) 0.0 (0)
Coronary artery disease 0.0 (0) 5.0 (3) 0.0 (0) 4.0 (5)
Myocardial infarction/ACS 0.0 (0) 0.0 (0) 0.0 (0) 1.6 (2)
Stroke/TIA 0.0 (0) 0.0 (0) 0.9 (1) 0.0 (0)
Office SBP, mm Hg 163.8  7.4 163.6  8.6 162.1  8.0 162.9  7.2
Office DBP, mm Hg 99.4  5.9 99.4  5.7 102.3  7.6 103.4  7.7
Office heart rate, beats/min 63.2  5.3 62.5  5.8 80.4  7.9 79.7  7.0
Mean 24-h SBP, mm Hg 152.6  7.8 152.4  7.4 150.3  7.8 150.7  7.6
Mean 24-h DBP, mm Hg 97.7  6.6 98.9  7.6 97.6  8.6 99.5  7.5

Values are mean  SD or % (n).

ACS ¼ acute coronary syndrome; DBP ¼ diastolic blood pressure; RDN ¼ renal denervation; SBP ¼ systolic blood pressure; TIA ¼ transient ischemic attack.
JACC VOL. 78, NO. 10, 2021 Böhm et al. 1031
SEPTEMBER 7, 2021:1028–1038 Renal Denervation and Heart Rate

F I G U R E 2 Changes in 24-Hour SBP and Office SBP

A 24-Hour Systolic Blood Pressure Change

Renal Denervation Control
30 30
24-Hour SBP Change at 3 Months (mm Hg)

20 20

10 10

0 0

–10 –10

–20 –20

–30 –30

–40 Interaction P = 0.031 –40

40 50 60 70 80 90 100 110 40 50 60 70 80 90 100 110

Baseline Office Heart Rate, beats/min Baseline Office Heart Rate, beats/min

B Office Systolic Blood Pressure Change

Renal Denervation Control
50 50
40 40
Office SBP Change at 3 Months (mm Hg)

30 30
20 20
10 10
0 0
–10 –10
–20 –20
–30 –30
–40 –40
–50 –50
Interaction P = 0.032
–60 –60

40 50 60 70 80 90 100 110 40 50 60 70 80 90 100 110

Baseline Office Heart Rate, beats/min Baseline Office Heart Rate, beats/min

24-hour SBP change at 3 months (A) and office SBP change at 3 months (B) is plotted against baseline office heart rate in patients after renal denervation
(left, blue) and control (right, red). Interaction P values were significant indicating that there is an association of 24-hour SBP change or office SBP
change to baseline office heart rate. SBP ¼ systolic blood pressure.
1032 Böhm et al. JACC VOL. 78, NO. 10, 2021

Renal Denervation and Heart Rate SEPTEMBER 7, 2021:1028–1038

F I G U R E 3 3-Month Changes in 24-Hour, Office, Daytime, and Nighttime SBP

A Change in 24-Hour SBP

<70 beats/min ≥70 beats/min
0
0.0 (N = 98)
–2
3-Month SBP Change (mm Hg)

–1.9 (N = 60) –1.8 (N = 49)

–4
' –0.1 (–3.8, 3.6) mm Hg
P = 0.97
–6
–6.3 (N = 93)
–8
' –6.2 (–9.0, –3.5) mm Hg
–10 P < 0.001

–12
Interaction P = 0.008
–14

B Change in Office SBP

<70 beats/min ≥70 beats/min
0

–2
3-Month SBP Change (mm Hg)

–2.3 (N = 50) –2.3 (N = 114)

–4
–4.7 (N = 64)
–6

' –1.8 (–6.7, 3.1) mm Hg

–8 P = 0.47

–10

–12
–12.0 (N = 106)
Interaction P = 0.009
–14 ' –9.9 (–13.4, –6.4) mm Hg
P < 0.001
Renal Denervation Control

Change in 24 hour SBP (A), change in office SBP (B), change in daytime SBP (C) and change in nighttime SBP (D) according to heart rate.
Follow-up treatment differences between renal denervation and sham control were determined using analysis of covariance models adjusting
for baseline values. Changes in 24-hour SBP, office SBP, daytime SBP, and nighttime SBP were significantly greater for renal denervation
compared with sham control patients with baseline office heart rate $70 beats/min. The interaction P values suggest that the effects of renal
denervation on these blood pressure reductions differ based on baseline office heart rate. SBP ¼ systolic blood pressure.

Continued on the next page

interaction testing to compare the slopes in the RDN baseline heart rate, the slope of the RDN arm regres-
and sham control groups (Figure 2) and show how 3- sion line was significantly different from sham con-
month SBP changes vary for different baseline heart trol (interaction P value [P int ] ¼ 0.031) (Figure 2A),
rate values. For 3-month change in 24-hour SBP vs suggesting a relationship between higher baseline
JACC VOL. 78, NO. 10, 2021 Böhm et al. 1033
SEPTEMBER 7, 2021:1028–1038 Renal Denervation and Heart Rate

F I G U R E 3 Continued

C Change in Daytime SBP

<70 beats/min ≥70 beats/min
0
0.5 (N = 100)
–2 –1.5 (N = 60)
3-Month SBP Change (mm Hg)

–1.9 (N = 49)

–4
' 0.2 (–3.8, 4.1) mm Hg
P = 0.94
–6
–5.8 (N = 95)

–8
' –6.5 (–9.4, –3.6) mm Hg
P < 0.001
–10

–12
Interaction P = 0.007
–14

D Change in Nighttime SBP

<70 beats/min ≥70 beats/min
0
–0.4 (N = 101)
–2 –1.1 (N = 50)
3-Month SBP Change (mm Hg)

–2.4 (N = 61)
–4
' –0.8 (–4.9, 3.3) mm Hg
–6 P = 0.70

–8 –7.5 (N = 100)

–10 ' –6.5 (–9.8, –3.2) mm Hg

P < 0.001
–12
Interaction P = 0.033
–14
Renal Denervation Control

heart rate and greater SBP reduction. Similarly, for
3-month change in
heart rate, slope of the RDN regression line was
also significantly different
(P int ¼ 0.032) (Figure 2B).
The 3-month treatment differences between RDN
and sham control patients for mean 24-hour, office,
daytime, and nighttime SBP were greater for patients
with baseline heart rate $70 beats/min compared
with <70 beats/min, with significant interaction
P values for all comparisons (Figure 3). We character-
office SBP vs baseline ized patients as responders using the following defi-
nitions: 1) 3-month office SBP reduction of $10 mm Hg;
from sham control and 2) 3-month 24-hour SBP reduction of $5 mm Hg.
We then calculated the percentage of responders
using each definition for the RDN and sham control
groups with baseline heart rate <70 and $70 beats/min.
For each definition, where the BP response is now
a binary outcome (responder/nonresponder), the
interaction P values remain significant, similar to the
1034 Böhm et al.
Renal Denervation and Heart Rate
analyses where BP change was analyzed as a contin-
uous outcome (see Supplemental Table 1).
Differences in 24-hour ambulatory SBP from base-
line to 3 months for RDN and sham control groups are
shown in Figure 4. There is a greater separation
between the curves for RDN patients with baseline
office heart rate $70 beats/min, compared with pa-
tients with baseline office heart rate <70 beats/min.
DISCUSSION
The present analysis of the SPYRAL HTN-OFF MED
Pivotal multicenter, randomized, sham controlled
trial has shown a significant BP reduction in patients
with hypertension not receiving antihypertensive
drugs (24). Patients with baseline office heart
rate $70 beats/min showed greater SBP reductions
after RDN when compared with patients with baseline
office heart rate <70 beats/min; there were no effects
on SBP in the sham group. This strongly suggests that
heart rate mirrors neuroendocrine activation identi-
fying patients with a higher likelihood of response to
RDN. BP reduction was also greater in patients with
baseline office heart rate $70 beats/min over the full
circadian cycle, including daytime and nighttime
(Central Illustration).
Overactivity of the sympathetic nervous system has
been shown to be involved in the maintenance and
development of hypertension (21,22) as well as in the
regulation of renal function (31). After early experi-
ences with surgical splanchnicectomy, which reduced
BP and improved survival in patients with severe hy-
pertension (32), a minimally invasive technique was
developed for treatment of resistant hypertension.
This procedure produced BP reductions in controlled
studies without sham (33) as well as pilot sham-
controlled studies in the presence (24) and absence
(23) of antihypertensive agents using radiofrequency
energy. Recently, an adequately powered pivotal trial
confirmed the efficacy of RDN to reduce BP (26).
However, predictors of BP reduction are not well
characterized. In previous studies from a large real-
world registry (27) and in the SPYRAL HTN-OFF MED
pilot trial (28), RDN provided evidence that it also re-
duces heart rate and that heart rate might be a pre-
dictor of BP response to RDN (28), reflecting its close
association with the sympathetic and parasympathetic
nervous systems (1). Given the influence of the para-
sympathetic nervous system on heart rate (34) and the
association demonstrated in this report between heart
rate and BP reduction after RDN, it is possible that
RDN may also enhance parasympathetic restraint to
reduce BP, but this has not been studied yet. Baseline
BP has been suggested as the only consistent predictor
JACC VOL. 78, NO. 10, 2021
SEPTEMBER 7, 2021:1028–1038
of response, which might, however, be affected by
regression to the mean, and furthermore, might be
applicable to any kind of hypertension treatment (35).
Other predictors, such as biomarkers, muscle sympa-
thetic nerve activity, baroreceptor sensitivity, and
electrical renal nerve stimulation, are technically
challenging and not accessible for routine use (36).
Herein, we extend those previous findings regarding
heart rate in an adequately powered population
without concomitant antihypertensive treatment, and
therefore not confounded by potentially heart rate–
altering drugs, particularly beta-blockers. As BP
reduction throughout 24 hours has been demon-
strated for radiofrequency (28,29,37) and ultrasound
devices (38), a constant “always on” reduction in BP
throughout 24 hours appears to be characteristic of
RDN (39).
RDN, as probably with any effective device ther-
apy, could be useful to overcome the problem of
nonadherence in hypertension (40). A consistent
reduction in BP over 24 hours might even be helpful
when only occasional doses are missed. Interestingly,
RDN reduces BP and heart rate also at nighttime,
when presumably patients are resting or sleeping,
suggesting that baseline sympathetic activity is
reduced. This is in agreement with exercise studies
post-RDN where not only were exercise-induced BP
and heart rate peaks attenuated, but also baseline
heart rate and BP were reduced, whereas the relative
increases in BP and heart rate was unaffected (41).
In hypertensive patients, higher heart rate is
associated with more adverse cardiovascular events
(10,42) and is associated with complications in hy-
pertensive emergencies (19). In heart failure patients,
the reduction of heart rate with ivabradine is depen-
dent on baseline heart rate (13) and is associated with
a reduction of cardiovascular events (43). Further-
more, high heart rates are associated with vascular
damage (44,45), and heart rate reduction is associ-
ated with improvement of endothelial dysfunction
(45) and collateral flow (46). For heart failure pa-
tients, risk increases for heart rate >70 beats/min
(13,18), and risk reduction with selective heart rate
reduction was observed for this cutoff (43), thus
identifying heart rate $70 beats/min as a modifiable
risk factor (13,43). The potential benefit of heart rate
modulation for hypertension and other conditions
should be scrutinized in prospective clinical trials
(19,41). Nevertheless, a BP reduction of just 5 to
10 mm Hg or more has been shown to be associated
with a reduction of the composite of cardiovascular
death, myocardial infarction, and stroke by 26% (47).
Therefore, the observed BP reduction together with a
probable beneficial effect of direct sympathetic
JACC VOL. 78, NO. 10, 2021 Böhm et al. 1035
SEPTEMBER 7, 2021:1028–1038 Renal Denervation and Heart Rate

F I G U R E 4 Hourly 24-Hour Ambulatory SBP From Baseline to 3 Months

A Baseline Office HR <70 beats/min

Renal Denervation: Office HR <70 beats/min Control: Office HR <70 beats/min
165
Systolic Blood Pressure (mm Hg)

160

155

150

145

140

135

130
12-1am
1-2am
2-3am
3-4am
4-5am
5-6am
6-7am
7-8am
8-9am
9-10am
10-11am
11-12pm
12-1pm
1-2pm
2-3pm
3-4pm
4-5pm
5-6pm
6-7pm
7-8pm
8-9pm
9-10pm
10-11pm
11-12am

12-1am
1-2am
2-3am
3-4am
4-5am
5-6am
6-7am
7-8am
8-9am
9-10am
10-11am
11-12pm
12-1pm
1-2pm
2-3pm
3-4pm
4-5pm
5-6pm
6-7pm
7-8pm
8-9pm
9-10pm
10-11pm
11-12am
B Baseline Office HR ≥70 beats/min
Renal Denervation: Office HR ≥70 beats/min Control: Office HR ≥70 beats/min
165
Systolic Blood Pressure (mm Hg)

160

155

150

145

140

135

130
12-1am
1-2am
2-3am
3-4am
4-5am
5-6am
6-7am
7-8am
8-9am
9-10am
10-11am
11-12pm
12-1pm
1-2pm
2-3pm
3-4pm
4-5pm
5-6pm
6-7pm
7-8pm
8-9pm
9-10pm
10-11pm
11-12am

12-1am
1-2am
2-3am
3-4am
4-5am
5-6am
6-7am
7-8am
8-9am
9-10am
10-11am
11-12pm
12-1pm
1-2pm
2-3pm
3-4pm
4-5pm
5-6pm
6-7pm
7-8pm
8-9pm
9-10pm
10-11pm
11-12am
Changes in ambulatory systolic blood pressure from baseline to 3 months are examined for renal denervation and sham control patients with baseline heart rate <70
beats/min (A) and $70 beats/min (B) in patients after renal denervation (left, blue) and in control (right, red). Hourly blood pressure plots for renal denervation
patients with baseline heart rate $70 beats/min show a reduction in hourly SBP measurements at 3 months compared with baseline, but less-effective SBP reduction at
heart rate <70 beats/min and no meaningful effects in control subjects. Abbreviations as in Figure 3.

antagonism merits prospective investigations CONCLUSIONS

defining RDN’s role in clinical outcome reduction.
STUDY LIMITATIONS. This is a secondary observa- There was greater reduction at 3 months in mean of-
tional analysis, which is not subject to randomization fice, 24-hour, daytime, and nighttime SBP for
and, therefore, might be affected by invisible con- RDN patients with baseline office heart rate
founders. However, it is derived from the largest $70 beats/min. A consistent lowering of BP
available prospective, sham-controlled RDN trial, throughout 24 hours might be of particular relevance
providing reassurance of data reliability. The present for endpoint reduction, because it covers the night
“off-medication” study design avoids potential con- and early morning high-risk periods and could be
founding influences of antihypertensive drug ther- particularly helpful in patients who are nonadherent
apy, particularly beta-blockers, on the results. to medical treatments. Given its safety and efficacy,
1036 Böhm et al. JACC VOL. 78, NO. 10, 2021

Renal Denervation and Heart Rate SEPTEMBER 7, 2021:1028–1038

C E N T R A L IL L U ST R A T I O N Association of Baseline Heart Rate With Blood Pressure Reduction

Böhm, M. et al. J Am Coll Cardiol. 2021;78(10):1028–1038.

Renal denervation interrupts efferent and afferent renal nerves to reduce sympathetic nervous system activity with the goal of lowering
blood pressure (left); ambulatory blood pressure measurements showed greater systolic blood pressure (SBP) reduction in 24-hour, office,
daytime, and nighttime SBP in patients at a heart rate $70 compared to <70 beats/min. This is confirmed by hourly 24-hour blood pressure
measurements at 3 months. 3M ¼ 3-month; HR ¼ heart rate; RDN ¼ renal denervation.

RDN might be one option beyond lifestyle modifica-
tions and drugs to be offered to patients to improve
BP control.
ACKNOWLEDGMENTS The authors thank Beth Ferri
and Jessica Dries-Devlin for editorial assistance;
Manuela Negoita for contributions to trial design;
Graeme Hickey for statistical analysis oversight;
Denise Jones, Pamela McKenna, trial oversight;
and Marianne Wanten for clinical trial oversight;
and Sidney A. Cohen for expert review, all of
Medtronic.
FUNDING SUPPORT AND AUTHOR DISCLOSURES
The trial is sponsored by Medtronic and was designed in collaboration
with the U.S. Food and Drug Administration by the steering
JACC VOL. 78, NO. 10, 2021 Böhm et al. 1037
SEPTEMBER 7, 2021:1028–1038 Renal Denervation and Heart Rate

committee and sponsor. Profs Böhm, Ukena, Ewen, and Mahfoud are
supported by the Deutsche Forschungsgemeinschaft (SFB TTR219, S- ADDRESS FOR CORRESPONDENCE: Dr Michael Böhm,
01, M-03, M-05). Prof. Böhm has received consulting fees from Abbott
Universitätsklinikum des Saarlandes, Saarland Univer-
Vascular, Bayer AG, Amgen, AstraZeneca, Servier, Medtronic, Vifor,
and Boehringer Ingelheim. Dr Tsioufis has received honoraria for
sity, Klinik für Innere Medizin III, Kardiologie, Angiolo-
advisory boards and lectures from Medtronic, Servier, Bayer, gie und Internistische Intensivmedizin, Kirrberger
Menarini, Novartis, AstraZeneca, Boehringer Ingelheim, Pfizer, Straße 1, 66421 Homburg/Saar, Germany. E-mail:
Pythagoras, Sanofi, and Amgen. Dr Kandzari has received institu-
michael.boehm@uks.eu. Twitter: @FelixMahfoud.
tional research/grant support from Medtronic CardioVascular and
Ablative Solutions; and has received personal consulting honoraria
from Medtronic CardioVascular. Prof. Kario has received scientific
support and speaker honoraria from Daiichi-Sankyo, Sanwa Chemical, PERSPECTIVES
Boehringer Ingelheim, Omron Healthcare, A & D Inc, Fukudadenshi
Inc, Medtronic, and ReCor Medical. Prof. Weber has received
consulting fees from Medtronic, ReCor, and Ablative Solutions. Prof. COMPETENCY IN PATIENT CARE AND PROCEDURAL
Schmieder has received consultant fees from Medtronic and ReCor;
SKILLS: Heart rates $70 beats/min are associated with greater
and has received grant support from Medtronic, ReCor, and Ablative
Solutions. Prof. Townsend has received consultant fees from Med-
BP reduction after renal denervation, possibly caused by
tronic, Axio, and Regeneron. Dr Pocock has received consultant fees increased sympathetic tone.
from Medtronic. Prof. Weil has received honoraria from Medtronic,
Novartis, ReCor, Cardinal Health, Bayer, and AstraZeneca. Mr Fahy is TRANSLATIONAL OUTLOOK: Additional studies are required
an employee of and shareholder for Medtronic. Prof. Mahfoud is
to elucidate the interplay between indices of renal sympathetic
supported by Deutsche Gesellschaft für Kardiologie; and has received
scientific support and speaker honoraria from Bayer, Boehringer activation on BP reduction after renal denervation and cardio-
Ingelheim, Medtronic, and ReCor Medical. All other authors have vascular outcomes in patients with uncontrolled hypertension.
reported that they have no relationships relevant to the contents of
this paper to disclose.

REFERENCES

1. Levy MN. Sympathetic-parasympathetic in-
teractions in the heart. Circ Res. 1971;29:437–445.
2. Jurca R, Jackson AS, LaMonte MJ, et al.
Assessing cardiorespiratory fitness without per-
forming exercise testing. Am J Prev Med. 2005;29:
185–193.
3. Grassi G, Cattaneo BM, Seravalle G,
Lanfranchi A, Mancia G. Baroreflex control of
sympathetic nerve activity in essential and sec-
ondary hypertension. Hypertension. 1998;31:68–
72.
4. Grassi G, Vailati S, Bertinieri G, et al. Heart rate
as marker of sympathetic activity. J Hypertens.
1998;16:1635–1639.
5. Williams B, Mancia G, Spiering W, et al. 2018
ESC/ESH guidelines for the management of arte-
rial hypertension. Eur Heart J. 2018;39:3021–
3104.
6. Vidal-Petiot E, Ford I, Greenlaw N, et al., for the
CLARIFY Investigators. Cardiovascular event rates
and mortality according to achieved systolic and
diastolic blood pressure in patients with stable
coronary artery disease: an international cohort
study. Lancet. 2016;388:2142–2152.
7. Böhm M, Schumacher H, Teo KK, et al. Achieved
blood pressure and cardiovascular outcomes in
high-risk patients: results from ONTARGET and
TRANSCEND trials. Lancet. 2017;389:2226–2237.
8. Ferrannini E, Cushman WC. Diabetes and hy-
pertension: the bad companions. Lancet.
2012;380:601–610.
9. Kannel WB, Kannel C, Paffenbarger Jr RS,
Cupples LA. Heart rate and cardiovascular mor-
tality: the Framingham Study. Am Heart J.
1987;113:1489–1494.
10. Palatini P, Thijs L, Staessen JA, et al., for the
Systolic Hypertension in Europe (Syst-Eur) Trial
Investigators. Predictive value of clinic and
ambulatory heart rate for mortality in elderly
subjects with systolic hypertension. Arch Intern
Med. 2002;162:2313–2321.
11. Böhm M, Schumacher H, Teo KK, et al. Resting
heart rate and cardiovascular outcomes in diabetic
and non-diabetic individuals at high cardiovascular
risk analysis from the ONTARGET/TRANSCEND
trials. Eur Heart J. 2020;41:231–238.
12. Böhm M, Cotton D, Foster L, et al. Impact of
resting heart rate on mortality, disability and
cognitive decline in patients after ischaemic
stroke. Eur Heart J. 2012;33:2804–2812.
13. Böhm M, Swedberg K, Komajda M, et al., for
the SHIFT Investigators. Heart rate as a risk factor
in chronic heart failure (SHIFT): the association
between heart rate and outcomes in a randomised
placebo-controlled trial. Lancet. 2010;376:886–
894.
14. Libhaber E, Sliwa K, Bachelier K, Lamont K,
Böhm M. Low systolic blood pressure and high
resting heart rate as predictors of outcome in
patients with peripartum cardiomyopathy. Int J
Cardiol. 2015;190:376–382.
15. Böhm M, Perez AC, Jhund P, et al, on behalf of
the I-Preserve Committee and Investigators.
Relationship between heart rate and mortality and
morbidity in the irbesartan in patients with heart
failure and preserved systolic function trial (I-
Preserve). Eur J Heart Fail. 2014;16:778–787.
16. Castagno D, Skali H, Takeuchi M, et al., for the
CHARM Investigators. Association of heart rate
and outcomes in a broad spectrum of patients with
chronic heart failure: results from the CHARM
(Candesartan in Heart Failure: Assessment of
Reduction in Mortality and morbidity) program.
J Am Coll Cardiol. 2012;59:1785–1795.
17. Omlor AJ, Trudzinski FC, Alqudrah M, et al., for
the German COSYCONET Cohort. Time-updated
resting heart rate predicts mortality in patients
with COPD. Clin Res Cardiol. 2020;109(6):776–
786.
18. Böhm M. Heart rate: from heart failure to
chronic diseases and cancer. Is there a role for
supportive care by heart rate reduction? Eur J
Heart Fail. 2017;19:250–252.
19. Al Bannay R, Böhm M, Husain A. Heart rate
differentiates urgency and emergency in hyper-
tensive crisis. Clin Res Cardiol. 2013;102:593–
598.
20. Wilhelm J, Hettwer S, Schuermann M, et al.
Severity of cardiac impairment in the early stage of
community-acquired sepsis determines worse
prognosis. Clin Res Cardiol. 2013;102:735–744.
21. DiBona GF. Sympathetic nervous system and
hypertension. Hypertension. 2013;61:556–560.
22. Parati G, Esler M. The human sympathetic
nervous system: its relevance in hypertension and
heart failure. Eur Heart J. 2012;33:1058–1066.
23. Townsend RR, Mahfoud F, Kandzari DE, et al.,
for the SPYRAL HTN-OFF MED trial Investigators.
Catheter-based renal denervation in patients with
uncontrolled hypertension in the absence of anti-
hypertensive medications (SPYRAL HTN-OFF
MED): a randomised, sham-controlled, proof-of-
concept trial. Lancet. 2017;390:2160–2170.
24. Kandzari DE, Böhm M, Mahfoud F, et al, for
the SPYRAL HTN-ON MED Trial Investigators. Ef-
fect of renal denervation on blood pressure in the
presence of antihypertensive drugs: 6-month
1038 Böhm et al.
Renal Denervation and Heart Rate
efficacy and safety results from the SPYRAL HTN-
ON MED proof-of-concept randomised trial. Lan-
cet. 2018;391:2346–2355.
25. Mahfoud F, Mancia G, Schmieder R, et al.
Renal denervation in high-risk patients with hy-
pertension. J Am Coll Cardiol. 2020;75:2879–
2888.
26. Böhm M, Kario K, Kandzari DE, et al, for the
SPYRAL HTN-OFF MED Pivotal Investigators. Ef-
ficacy of catheter-based renal denervation in the
absence of antihypertensive medications (SPYRAL
HTN-OFF MED Pivotal): a multicentre, rando-
mised, sham-controlled trial. Lancet. 2020;395:
1444–1451.
27. Böhm M, Ukena C, Ewen S, et al, for the Global
SYMPLICITY Registry Investigators. Renal dener-
vation reduces office and ambulatory heart rate in
patients with uncontrolled hypertension: 12-
month outcomes from the global SYMPLICITY
registry. J Hypertens. 2016;34:2480–2486.
28. Böhm M, Mahfoud F, Townsend RR, et al.
Ambulatory heart rate reduction after catheter-
based renal denervation in hypertensive patients
not receiving anti-hypertensive medications: data
from SPYRAL HTN-OFF MED, a randomized, sham-
controlled, proof-of-concept trial. Eur Heart J.
2019;40:743–751.
29. Kandzari DE, Kario K, Mahfoud F, et al. The
SPYRAL HTN Global Clinical Trial Program:
rationale and design for studies of renal
denervation in the absence (SPYRAL HTN OFF-
MED) and presence (SPYRAL HTN ON-MED) of
antihypertensive medications. Am Heart J.
2016;171:82–91.
30. Böhm M, Townsend RR, Kario K, et al. Ratio-
nale and design of two randomized sham-
controlled trials of catheter-based renal denerva-
tion in subjects with uncontrolled hypertension in
the absence (SPYRAL HTN-OFF MED Pivotal) and
presence (SPYRAL HTN-ON MED Expansion) of
antihypertensive medications: a novel approach
using Bayesian design. Clin Res Cardiol. 2020;109:
289–302.
31. DiBona GF. Neural control of the kidney: past,
present, and future. Hypertension. 2003;41:621–
624.
32. Smithwick RH, Thompson JE. Splanchnicec-
tomy for essential hypertension; results in 1,266
cases. J Am Med Assoc. 1953;152:1501–1504.
33. Esler MD, Krum H, Sobotka PA, et al., for the
Symplicity HTN-2 Investigators. Renal sympa-
thetic denervation in patients with treatment-
resistant hypertension (The Symplicity HTN-2
Trial): a randomised controlled trial. Lancet.
2010;376:1903–1909.
34. Robinson BF, Epstein SE, Beiser GD,
Braunwald E. Control of heart rate by the auto-
nomic nervous system. Studies in man on the
interrelation between baroreceptor mechanisms
and exercise. Circ Res. 1966;19:400–411.
35. Messerli FH, Rexhaj E. Of headwind and tail-
wind, regression to the mean and Wilder’s princi-
ple. J Hypertens. 2019;37:4–5.
36. Mahfoud F, Azizi M, Ewen S, et al. Pro-
ceedings from the 3rd European Clinical
Consensus Conference for clinical trials in
device-based hypertension therapies. Eur Heart
J. 2020;41:1588–1599.
37. Kario K, Yamamoto E, Tomita H, et al, for the
SYMPLICITY HTN-Japan Investigators. Sufficient
and persistent blood pressure reduction in the
final long-term results from SYMPLICITY HTN-
Japan- safety and efficacy of renal denervation
at 3 years. Circ J. 2019;83:622–629.
38. Azizi M, Schmieder RE, Mahfoud F, et al., for
the RADIANCE-HTN Investigators. Six-month re-
sults of treatment-blinded medication titration for
hypertension control following randomization to
endovascular ultrasound renal denervation or a
sham procedure in the RADIANCE-HTN SOLO Trial.
Circulation. 2019;139:2542–2553.
39. Kario K, Weber MA, Mahfoud F, et al. Changes
in 24-Hour patterns of blood pressure in hyper-
tension following renal denervation therapy. Hy-
pertension. 2019;79:244–249.
JACC VOL. 78, NO. 10, 2021
SEPTEMBER 7, 2021:1028–1038
40. A Burnier M, Egan BM. Adherence in hyper-
tension. Circ Res. 2019;124:1124–1140.
41. Ukena C, Mahfoud F, Kindermann I, et al.
Cardiorespiratory response to exercise after renal
sympathetic denervation in patients with resistant
hypertension. J Am Coll Cardiol. 2011;58:1176–
1182.
42. Paul L, Hastie CE, Li WS, et al. Resting heart
rate pattern during follow-up and mortality in
hypertensive patients. Hypertension. 2010;55:
567–574.
43. Swedberg K, Komajda M, Böhm M, et al.
Ivabradine and outcomes in chronic heart failure
(SHIFT): a randomised placebo-controlled study.
Lancet. 2010;376:875–885.
44. Custodis F, Schirmer SH, Baumhäkel M,
Heusch G, Böhm M, Laufs U. Vascular patho-
physiology in response to increased heart rate.
J Am Coll Cardiol. 2010;56:1973–1983.
45. Custodis F, Baumhäkel M, Schlimmer N,
et al. Heart rate reduction by ivabradine re-
duces oxidative stress, improves endothelial
function, and prevents atherosclerosis in apoli-
poprotein E-deficient mice. Circulation. 2008;
117:2377–2387.
46. Schirmer SH, Degen A, Baumhäkel M, et al.
Heart-rate reduction by If-channel inhibition with
ivabradine restores collateral artery growth in
hypercholesterolemic atherosclerosis. Eur Heart J.
2012;33:1223–1231.
47. Ettehad D, Emdin CA, Kiran A, et al. Blood
pressure lowering for prevention of cardiovascular
disease and death: a systematic review and meta-
analysis. Lancet. 2016;387:957–967.
KEY WORDS blood pressure, heart rate,
hypertension, renal denervation,
sympathetic activation
A PPE NDI X For a supplemental table, please
see the online version of this paper.
JOURNAL OF THE AMERICAN COLLEGE OF CARDIOLOGY
ª 2021 BY THE AMERICAN COLLEGE OF CARDIOLOGY FOUNDATION
PUBLISHED BY ELSEVIER
EDITORIAL COMMENT
Baseline Heart Rate Predicts the Blood
Pressure Response to Renal Denervation
in Untreated Hypertension*
Brent M. Egan, MD
I n this issue of the Journal, Böhm et al. (1) docu-
ment that baseline office heart rate predicts the
antihypertensive response 3 months after renal
denervation (RDN) in adults with untreated hyper-
tension (1,2). However, baseline heart rate was unre-
lated to blood pressure (BP) responses 1 year after
RDN in 846 adults with treated hypertension in a
prior report (3). Thus, the utility of baseline heart
rate as a predictor of the antihypertensive response
to RDN may apply only to adults with untreated hy-
pertension. As the authors noted, study participants
were not randomized according to baseline heart
rate, so unmeasured confounders may contribute to
the association (1).
SEE PAGE 1028
POTENTIAL APPROACHES TO IMPROVING
PREDICTION OF BP RESPONSE TO RDN
Although baseline heart rate is a statistically significant
and clinically relevant predictor of the antihypertensive
response to RDN among adults with untreated hyper-
tension, sensitivity and specificity were limited at w70%
and w45%, respectively (1). Sensitivity and specificity for
predicting BP responses to RDN might be improved by
the following: 1) elucidating factors through which heart
rate is predictive; 2) multivariable models, which could
include patient sex and smoking status that were
*Editorials published in the Journal of the American College of Cardiology
reflect the views of the authors and do not necessarily represent the
views of JACC or the American College of Cardiology.
From the Department of Medicine, University of South Carolina School of
Medicine–Greenville, South Carolina, USA; and Improving Health Out-
comes, American Medical Association, Greenville, South Carolina, USA.
The author attests they are in compliance with human studies
committees and animal welfare regulations of the author’s institution
and Food and Drug Administration guidelines, including patient consent
where appropriate. For more information, visit the Author Center.
ISSN 0735-1097/$36.00
VOL. 78, NO. 10, 2021
different between responder and nonresponder groups
(1); and 3) utilizing machine learning approaches on large
databases of patients who have undergone RDN, which
could illuminate predictors not currently evident in
theoretically-based exploration. Validation of any single
predictor or predictive models in future studies is an
important prelude to broad clinical adoption.
DOES RDN LOWER BP BY INHIBITING
SYMPATHETIC DRIVE OR ARE OTHER
MECHANISMS OPERATIVE?
Office heart rate in the Global Registry report fell
similarly in subjects irrespective of b-blocker therapy
(3). This observation reinforces evidence that RDN
may reduce sympathetic drive while enhancing para-
sympathetic restraint (1). The authors also noted
that the fall in BP after RDN was similar during the
daytime and nighttime, the latter a period of lesser
sympathetic activation. There are 2 most probable
potential explanations: 1) sufficient sympathetic
activation persists during the nighttime that RDN
could lower nocturnal BP predominantly through
reduced sympathetic drive; or 2) enhanced para-
sympathetic restraint plays a substantial role. Collec-
tively, these observations suggest that RDN lowers BP
in adults with hypertension by diminishing sympa-
thetic drive while augmenting parasympathetic re-
straint. Complementary evidence suggests that RDN
reduces premature ventricular and atrial beats
in adults with treatment-resistant hypertension by
improving cardiac sympatho-vagal balance (4).
THE UTILITY OF HEART RATE AS A MARKER
FOR CARDIAC SYMPATHETIC DRIVE AND
GENERAL SYMPATHETIC ACTIVATION
IN HYPERTENSION
Heart rate in adults with essential hypertension is a
relatively poor marker for more direct measures of
https://doi.org/10.1016/j.jacc.2021.07.019
1040 Egan
Heart Rate Predicts BP Responses to Renal Denervation
sympathetic drive, including muscle sympathetic
nerve activity (MSNA) (5). Other evidence indicates
that both sympathetic activation and reduced para-
sympathetic restraint participate in the faster heart
rates in adults with hypertension (6).
The relatively weak relationship between heart
rate and MSNA may reflect evidence that substantial
regional variation of sympathetic drive exists in
hypertension. For example, the increase in MSNA
appears greater than the increase in cardiac norepi-
nephrine turnover in adults with hypertension,
especially those with obesity and hypertension (7).
Yet, in lean and obese adults with hypertension,
cardiac norepinephrine turnover is greater than in
weight-matched individuals with normal BP. Thus,
the elevated heart rate often seen among adults with
hypertension probably reflects, at least partially,
enhanced sympathetic drive.
THE IMPORTANCE OF THE KIDNEY IN LONG-TERM
ANTIHYPERTENSIVE RESPONSES
The Guytonian model of BP regulation, which iden-
tifies the kidney as the predominant, long-term
regulator of arterial BP, has been progressively vali-
dated (8,9). This model suggests that any factor or
factors that lead to sustained elevations or reductions
in BP must raise or lower, respectively, the threshold
for pressure natriuresis. Renal sympathetic nerves
can enhance sodium retention through direct effects
on the renal tubules, activation of renin-angiotensin-
aldosterone system, and potentially through effects
on renal medullary blood flow and renal interstitial
fluid pressure, key mediators of pressure natriuresis.
Although evidence indicates that neurogenic activa-
tion and various central nervous system lesions can
affect BP long-term, these modalities must alter the
threshold for pressure natriuresis to sustain BP
change.
RENAL SYMPATHETIC EFFERENT AND
AFFERENT NERVES AND THE
ANTIHYPERTENSIVE RESPONSE TO RDN
Both renal sympathetic afferent and efferent nerves have
effects that could sustain hypertension (10). Animal
studies indicate that sympathetic afferent and efferent
reinnervation occurs within weeks to months after RDN
with functional restoration after several months (11). In
humans, although there is histological evidence for
reinnervation after kidney transplantation, renal re-
sponses to lower body negative pressure remain impaired
2 years after transplantation, suggesting sustained func-
tional impairment of renal sympathetic efferent nerves.
Moreover, the antihypertensive response to RDN is
JACC VOL. 78, NO. 10, 2021
SEPTEMBER 7, 2021:1039–1041
sustained after 3 years (12), which also suggests sustained
functional denervation.
HEART RATE AS A RISK MARKER FOR
ADVERSE CARDIOVASCULAR OUTCOMES
As the authors note, faster heart rates are associated with
increased risk for hypertension, cardiovascular disease,
and all-cause mortality in the general population,
and worse clinical outcomes in adults with untreated and
treated hypertension, myocardial infarction, stroke, and
heart failure (1). In some studies, heart rate is equivalent
to systolic BP in predicting cardiovascular events. These
observations raise the question of why heart rate does not
receive greater clinical attention. The answer significantly
reflects the dearth of evidence that selective lowering of
heart rate improves clinical outcomes (13). Yet, lowering
heart rate, largely by b-blockers, which also have heart
rate–independent effects, has been linked with attenua-
tion of atherosclerosis in animals and humans and
improved outcomes in patients with myocardial infarc-
tion and heart failure. And, selective heart rate reduction
with ivabradine in patients with heart failure and reduced
ejection fraction improves clinical outcomes.
In summary, the current report indicates that
baseline heart rate, which is inexpensive and readily
attainable, is a significant predictor of the antihyper-
tensive response to RDN in adults with untreated
hypertension. Given the invasiveness of RDN and the
heterogeneous BP response, clinically significant
predictors are important. Further information on
other readily available physiological, biochemical, or
molecular markers that improve prediction of the BP
response to RDN, either individually or in multivari-
able models, would be useful for adults with
untreated and treated hypertension. The authors have
summarized an extensive body of published reports
on the utility of heart rate for predicting adverse car-
diovascular outcomes. Future studies to assess the
benefits of selective heart rate reduction among
various at-risk groups may reduce residual risk for
cardiovascular events, which persists despite imple-
mentation of current evidence-based guidelines.
FUNDING SUPPORT AND AUTHOR DISCLOSURES
Dr Egan has received royalties from UpToDate. The statements and
conclusions in this report are those of the author and do not neces-
sarily represent the official position of the American Medical
Association.
ADDRESS FOR CORRESPONDENCE: Dr Brent M.
Egan, American Medical Association, 2 West Wash-
ington Street, Greenville, South Carolina 29601, USA.
E-mail: brent.egan@ama-assn.org.
JACC VOL. 78, NO. 10, 2021
SEPTEMBER 7, 2021:1039–1041
REFERENCES
1. Böhm M, Tsioufis K, Kandzari DE, et al. Effect of
heart rate on the outcome of renal denervation in
patients with uncontrolled hypertension. J Am Coll
Cardiol. 2021;78:1028–1038.
2. Böhm M, Mahfoud F, Townsend R, et al.
Ambulatory HR reduction after catheter-based
renal denervation in hypertensive patients not
receiving anti-hypertensive medications: Data
from SPYRAL HTN-OFF MED, a randomized, sham-
controlled, proof-of-concept trial. Eur Heart J.
2019;40:743–751.
3. Böhm M, Ukene C, Ewen S, et al. Renal dener-
vation reduced office and ambulatory HR in pa-
tients with uncontrolled hypertension: 12-month
outcomes from the global SYMPLICITY registry.
J Hypertens. 2016;34:2480–2486.
4. Tioufis C, Papdemetriou V, Tsiachris D, et al.
Drug-resistant hypertensive patients responding
to multielectrode renal denervation exhibit
improved HR dynamics and reduced arrhythmia
burden. J Hum Hypertens. 2014;28:587–593.
5. Grassi G, Bertinieri SV, Servalle G, Stella JL,
Dell-Oro R, Mancia G. HR as marker of sympathetic
activity. J Hypertens. 1989;16:1635–1639.
6. Mancia G, Grassi G. The autonomic nervous
system and hypertension. Circ Res. 2014;114:
1804–1814.
7. Grassi G, Mark A, Esler M. The sympathetic
nervous system alterations in human hyperten-
sion. Circ Res. 2015;116:976–990.
8. Guyton AC, Coleman TG, Granger HJ. Circula-
tion: overall regulation. Annual Rev Physiol.
1972;34:13–46.
9. Williams JM, Sarkis A, Lopez B, Ryan RP,
Flasch AK, Roman RJ. Elevations in renal
interstitial hydrostatic pressure and 20-
hydroxyeicosatetraenoic acid contribute to
pressure natriuresis. Hypertension. 2007;49(Part
2):687–694.
10. DiBona GF. Sympathetic nervous system and
hypertension. Hypertension. 2013;61:556–560.
Egan 1041
Heart Rate Predicts BP Responses to Renal Denervation
11. Singh RR, Denton KM. Renal denervation:
a treatment for hypertension and chronic
kidney disease. Hypertension. 2018;72:528–
536.
12. Mahfoud F, Böhm M, Schmieder R, et al. Ef-
fects of renal denervation on kidney function and
long-term outcomes: 3-year follow-up from the
Global SYMPLICITY Registry. Eur Heart J. 2019;40:
3474–3482.
13. Palatini P, Agabiti Rosei E, Casiglia E, et al.
Management of the hypertensive patient with
elevated HR. Statement of the Second
Consensus Conference endorsed by the Euro-
pean Society of Hypertension. J Hypertens.
2016;34:813–821.
KEY WORDS blood pressure, heart rate,
hypertension, renal denervation, sympathetic
activation

Listen to this manuscript’s
audio summary by
Editor-in-Chief
Dr. Valentin Fuster on
JACC.org.
JOURNAL OF THE AMERICAN COLLEGE OF CARDIOLOGY VOL. 78, NO. 10, 2021
ª 2021 BY THE AMERICAN COLLEGE OF CARDIOLOGY FOUNDATION
PUBLISHED BY ELSEVIER
Compression-Only Versus
Rescue-Breathing Cardiopulmonary
Resuscitation After Pediatric
Out-of-Hospital Cardiac Arrest
Maryam Y. Naim, MD, MSCE,a,b Heather M. Griffis, PHD,c Robert A. Berg, MD,b Richard N. Bradley, MD,d
Rita V. Burke, PHD, MPH,e David Markenson, MD,f Bryan F. McNally, MD, MPH,g Vinay M. Nadkarni, MD, MS,b
Lihai Song, MS,c Kimberly Vellano, MPH,g Victoria Vetter, MD, MPH,a Joseph W. Rossano, MD, MSa,h
ABSTRACT
BACKGROUND There are conflicting data regarding the benefit of compression-only bystander cardiopulmonary
resuscitation (CO-CPR) compared with CPR with rescue breathing (RB-CPR) after pediatric out-of-hospital cardiac arrest
(OHCA).
OBJECTIVES This study sought to test the hypothesis that RB-CPR is associated with improved neurologically favor-
able survival compared with CO-CPR following pediatric OHCA, and to characterize age-stratified outcomes with CPR
type compared with no bystander CPR (NO-CPR).
METHODS Analysis of the CARES registry (Cardiac Arrest Registry to Enhance Survival) for nontraumatic pediatric
OHCAs (patients aged #18 years) from 2013-2019 was performed. Age groups included infants (<1 year), children (1 to 11
years), and adolescents ($12 years). The primary outcome was neurologically favorable survival at hospital discharge.
RESULTS Of 13,060 pediatric OHCAs, 46.5% received bystander CPR. CO-CPR was the most common bystander CPR
type. In the overall cohort, neurologically favorable survival was associated with RB-CPR (adjusted OR: 2.16; 95% CI:
1.78-2.62) and CO-CPR (adjusted OR: 1.61; 95% CI: 1.34-1.94) compared with NO-CPR. RB-CPR was associated with a
higher odds of neurologically favorable survival compared with CO-CPR (adjusted OR: 1.36; 95% CI: 1.10-1.68). In age-
stratified analysis, RB-CPR was associated with better neurologically favorable survival versus NO-CPR in all age groups.
CO-CPR was associated with better neurologically favorable survival compared with NO-CPR in children and adolescents,
but not in infants.
CONCLUSIONS CO-CPR was the most common type of bystander CPR in pediatric OHCA. RB-CPR was associated with
better outcomes compared with CO-CPR. These results support present guidelines for RB-CPR as the preferred CPR
modality for pediatric OHCA. (J Am Coll Cardiol 2021;78:1042–1052) © 2021 by the American College of Cardiology
Foundation.
From aThe Cardiac Center, Children’s Hospital of Philadelphia and The University of Pennsylvania Perelman School of Medicine,
Philadelphia, Pennsylvania, USA; bDepartment of Anesthesiology and Critical Care Medicine, Children’s Hospital of Philadelphia
and The University of Pennsylvania Perelman School of Medicine, Philadelphia, Pennsylvania, USA; cDepartment of Biomedical
and Health Informatics, Data Science and Biostatistics Unit, Children’s Hospital of Philadelphia, Philadelphia, Pennsylvania, USA;
d
Division of Emergency Medicine, University of Texas Health Science Center, Houston, Texas, USA; eChildren’s Hospital of Los
Angeles, Keck School of Medicine, University of Southern California, Los Angeles, California, USA; fSky Ridge Medical Center,
Lone Tree, Colorado, USA; gDepartment of Emergency Medicine, Emory University, Atlanta, Georgia, USA; and the hLeonard Davis
Institute, The University of Pennsylvania, Philadelphia, Pennsylvania, USA.
The authors attest they are in compliance with human studies committees and animal welfare regulations of the authors’
institutions and Food and Drug Administration guidelines, including patient consent where appropriate. For more information,
visit the Author Center.
Manuscript received April 20, 2021; revised manuscript received June 7, 2021, accepted June 29, 2021.
ISSN 0735-1097/$36.00 https://doi.org/10.1016/j.jacc.2021.06.042
JACC VOL. 78, NO. 10, 2021
SEPTEMBER 7, 2021:1042–1052
S urvival following out-of-hospital cardiac arrest
(OHCA) among children remains poor
with <10% surviving to hospital discharge in
most communities (1). Bystander cardiopulmonary
resuscitation (CPR) is associated with better survival
in children following OHCA (1). Compression only-
cardiopulmonary resuscitation (CO-CPR) appears to
be as effective as conventional CPR with rescue
breathing (RB-CPR) for adults with OHCA.
Therefore, since 2010, the American Heart
Association (AHA) and European Resuscitation
Council have recommended CO-CPR for untrained
lay rescuers and those unwilling to perform rescue
breaths for adults with OHCA (2-7). However, CO-
CPR may be less effective for children whose OHCAs
are most commonly from asphyxia (8,9). Therefore,
the AHA and European Resuscitation Council recom-
mend RB-CPR for pediatric OHCA instead of CO-
CPR. However, if a bystander is unable to perform
rescue breathing, CO-CPR is recommended, which is
preferable to no bystander-cardiopulmonary resusci-
tation (NO-CPR) (10-12).
In the United States, national and statewide efforts
have focused on teaching CO-CPR to improve
bystander CPR rates, and currently, CO-CPR training
is mandated for high school graduates in 39 states and
the District of Columbia (2,13). Although these efforts
have increased bystander CPR rates and survival in
adult OHCA (2), the impact on pediatric OHCA is un-
known. The objectives of this study were to test the
hypothesis that RB-CPR is associated with improved
neurologically favorable survival following pediatric
OHCA compared with CO-CPR and to characterize
age-stratified outcomes with these 2 types of CPR
compared with no bystander CPR.
SEE PAGE 1053
METHODS
CARDIAC ARREST REGISTRY TO ENHANCE
SURVIVAL. Because of the sensitive nature of the
data collected for this study, requests to access the
dataset from qualified researchers trained in human
subject confidentiality protocols may be sent to
CARES (Cardiac Arrest Registry to Enhance Survival).
A secondary data analysis of the CARES database was
conducted. The CARES registry was established by
the Centers for Disease Control and Prevention in
collaboration with the Department of Emergency
Medicine at the Emory University School of Medicine.
The CARES registry includes an overall catchment
area of nearly 145 million people in 28 states across
the United States. CARES is an emergency medical
Naim et al. 1043
Compression-Only Versus Rescue Breathing CPR in Pediatric OHCA
services (EMS)-based registry for OHCA, ABBREVIATIONS
composed of a limited standard set of data AND ACRONYMS
elements from 3 sources: 911 call centers,
AED = automated external
EMS providers, and receiving hospitals (14). defibrillator
Responding EMS providers were defined as
AHA = American Heart
personnel (emergency medicine technicians Association
and paramedics) who respond to the medical CO-CPR = compression only-
emergency in an official capacity (ie, respond cardiopulmonary resuscitation
to the 911 call) as part of an organized medical CPR = cardiopulmonary
response team and were the designated resuscitation
transporter of the patient to the hospital. EMS = emergency medical
services
Detailed information on the design and
NO-CPR = no bystander-
development of this registry, as well as the
cardiopulmonary resuscitation
data elements included in the registry, is
OHCA = out-of-hospital
published elsewhere (15). cardiac arrest
CARES analysts confirmed the capture of
RB-CPR = cardiopulmonary
all cardiac arrests by each community’s 911 resuscitation with rescue
center during the data review process. Cere- breathing
bral performance category information was obtained
from receiving hospitals. The CARES registry estab-
lished a point of contact at each participating hospital
and trained the contact on CARES hospital element
definitions and the data entry process. The local
hospital contact abstracted information from the pa-
tient’s medical record and entered data. All data were
entered using a web-based platform, and an Excel file
(Microsoft Corporation) was generated with all car-
diac arrest events for the specified date range.
STUDY SAMPLE. All pediatric cases #18 years of age
with nontraumatic OHCA submitted to the registry
during the study period were eligible for inclusion.
From January 1, 2013, through December 31, 2019,
CARES captured all 911-activated nontraumatic
cardiac arrest events, defined as apnea and unre-
sponsiveness in which resuscitation with either CPR
or defibrillation was attempted. Children with
obvious signs of death (eg, rigor mortis or dependent
lividity) or for whom a “do not resuscitate” order was
respected were not included. Standardized interna-
tional Utstein definitions for defining clinical vari-
ables and outcomes were used to ensure uniformity
in reporting (16). Because the etiology of cardiac
arrest in children is difficult to determine, especially
in cases that result in death, all nontraumatic cases
were included regardless of presumed etiology,
including respiratory, cardiac, drowning, electrocu-
tion, or other. Arrests that occurred in medical facil-
ities or nursing homes, traumatic arrests, arrests with
rescue-breathing only, and 911 responder–witnessed
arrests were excluded from the analysis. This study
was approved by the Institutional Review Boards at
both Emory University and The Children’s Hospital of
Los Angeles. Given the use of deidentified data, the
1044 Naim et al.
Compression-Only Versus Rescue Breathing CPR in Pediatric OHCA
study was determined to be exempt from Institu-
tional Review Board review by the Children’s Hospital
of Philadelphia.
VARIABLES OF INTEREST AND STUDY OUTCOME.
Patient characteristics obtained from the database
included age, sex, race/ethnicity, bystander witness
status, arrest location, initial rhythm, automated
external defibrillator (AED) use, and region of arrest.
Race/ethnicity were evaluated to identify whether
CPR provision was influenced by race. Race was
assigned as considered by the child, family, or 911
emergency responder. The race category included
White, Black, Hispanic/Latino, other (American-
Indian/Alaska, Asian, and Native-Hawaiian/Pacific),
and unknown. Age groups included infants (<1
year), children (1 to 11 years), and adolescents ($12
years). Bystander CPR was defined as CPR adminis-
tered by a layperson defined as lay family member,
layperson, or layperson with medical training (ie,
physicians, nurses, or paramedics that were not part
of the organized emergency medical response). RB-
CPR was defined as CPR with chest compressions
and rescue breaths. CO-CPR was defined as CPR with
chest compressions only.
The primary outcome of interest was neurologi-
cally favorable survival, defined as a cerebral per-
formance category score of 1 (no neurologic
disability) or 2 (moderate disability) at the time of
discharge (17). Neurologically unfavorable survival
was defined as a cerebral performance category
score of 3 (severe disability), or 4 (coma or vegeta-
tive state) or death.
STATISTICAL ANALYSIS. Descriptive analyses report
overall child and arrest characteristics, characteristics
for arrests that had bystander and NO-CPR, and
characteristics for arrests that had RB-CPR and CO-
CPR. Pearson’s chi-square testing was used to
compare baseline characteristics between CPR groups
and to describe the percentage of arrests with
neurologically favorable survival. Temporal trends
across years were analyzed using the nonparametric
test for trend (nptrend in STATA software). Logistic
regression models with robust standard errors were
used to examine the associations of the predictors
with the probability of neurologically favorable sur-
vival. The analysis was adjusted for potential con-
founders including year of arrest, age, sex, race/
ethnicity, bystander witness status, arrest location,
rhythm, AED use before EMS arrival, and region.
Further analyses examined whether neurologically
favorable survival differed across bystander CPR type
within each age group. Stratified logistic regressions
are presented for each age group, with neurologically
JACC VOL. 78, NO. 10, 2021
SEPTEMBER 7, 2021:1042–1052
favorable survival as the outcome, adjusting for
covariates as described in the preceding text.
Results are expressed as ORs and marginal proba-
bilities (adjusted neurologically favorable survival)
with 95% CIs. Marginal probabilities are calculated as
the mean value of the neurologically favorable sur-
vival based on predicted values for each observation
if all observations had that value from each logistic
regression model. Associations were deemed signifi-
cant at P < 0.05 on 2-sided testing. To correct
for multiple comparisons, the Bonferroni correction
was utilized at a total alpha of 0.05 with P < 0.017
(0.05 divided by 3), for the comparison of NO-CPR,
CO-CPR, and RB-CPR. Additionally, P values and
95% CIs presented in this report have not been
adjusted for multiplicity, and therefore, inferences
drawn from these statistics may not be reproducible.
Analyses were performed using STATA software
version 14.2.
RESULTS
PATIENT CHARACTERISTICS. A total of 13,060 pedi-
atric cardiac arrests were captured in the CARES
database (Figure 1), 46.5% (6,074 of 13,060) received
bystander CPR. Following exclusions, 10,429 arrests
were evaluated, 55.6% (2,318 of 4,241) received CO-
CPR and 45.3% (1,923 of 4,241) received RB-CPR.
The presumed etiology of arrest was cardiac in
44.4% (4,634 of 10,429), respiratory in 32.8% (3,424
of 10,429), drowning in 8.8% (913 of 10,439), drug
overdose in 1.8% (182 of 10,429), electrocution in
0.2% (17 of 10,429), exsanguination in 0.2% (19 of
10,429) and other in 11.9% (1,240 of 10,429). Arrests
were more common in infants, males, and White and
Black children. The majority of arrests were unwit-
nessed, located in a home/residence, with a non-
shockable rhythm, and with no AED use before EMS
arrival (Table 1). Bystander CPR characteristics are
described in Table 1. Over the 6-year period of the
study, the rates of bystander CPR did not change, but
there was a significant increase in the proportion of
pediatric OHCAs receiving CO-CPR (Supplemental
Table 1). Bystander CPR was most commonly pro-
vided by a lay family member (71.7%, 3,040 of 4,241)
followed by lay person (21.9%, 928 of 4,241) and lay
person with medical training (6.4%, 273 of 4,241). Lay
person family members (CO-CPR 54.8%, 1,666 of
3,040) and lay persons (CO-CPR 58.9%, 547 of 928)
were more likely to perform CO-CPR, whereas lay
persons with medical training were more likely to
perform RB-CPR (61.5%, 168 of 273); P < 0.001.
ASSOCIATION OF NEUROLOGICALLY FAVORABLE
SURVIVAL WITH CO-CPR VS RB-CPR. Neurologically
JACC VOL. 78, NO. 10, 2021 Naim et al. 1045
SEPTEMBER 7, 2021:1042–1052 Compression-Only Versus Rescue Breathing CPR in Pediatric OHCA

F I G U R E 1 Flow Diagram With Inclusion and Exclusion Criteria

Pediatric out-of-hospital cardiac arrests from January 1, 2013 to

December 31, 2019
13,060

911 Responder Witnessed or

Healthcare/Nursing Home location
933

12,127

Missing CPR type 1,507

Ventilations only 52

11,988

Missing demographic, clinical or

outcome variables
139

Analysis sample 10,429

Age <1 year 5,456 Age 1-11 years 2,936 Age ≥12 years 2,037

Total number of pediatric cardiac arrests that were captured in the CARES database and those that were excluded with reasons for exclusion that included
arrest that were 911 responder–witnessed or health care/nursing home location, missing CPR type, ventilation only, and missing demographic, clinical or
outcome variables. The numbers of arrests in each group are shown at the bottom of the figure. CPR ¼ cardiopulmonary resuscitation.

favorable survival was observed in 8.6% (897 of
10,429) cardiac arrests. Over the 6-year period of the
study, there was no change in neurologically favor-
able survival (Table 2). Unadjusted for demographic
and clinical characteristics, OHCAs with RB-CPR and
CO-CPR had better outcomes compared with NO-CPR
(RB-CPR 13.4%, CO-CPR 12.2%, NO-CPR 5.8%; P <
0.001). On multivariable analysis, RB-CPR (adjusted
proportion 12.0%; 95% CI: 10.7-13.2, adjusted OR:
2.16; 95% CI: 1.78-2.62) and CO-CPR (adjusted pro-
portion 9.7%; 95% CI: 8.7-10.7; adjusted OR: 1.61;
95% CI: 1.34-1.94) were both independently associ-
ated with neurologically favorable survival compared
with NO-CPR (adjusted proportion 6.8%; 95% CI: 6.2-
7.4) (Table 2). In a separate model comparing RB-CPR
and CO-CPR excluding those who had NO-CPR,
RB-CPR was associated with a higher odds of neuro-
logically favorable survival compared with CO-CPR
(adjusted OR: 1.36; 95% CI: 1.10-1.68) (Table 3).
Additional sensitivity analyses were
excluding bystander CPR provided by lay medical
providers without a change in overall results
(Supplemental Tables 2 and 3).
Additional factors associated with neurologically
favorable survival following multivariable adjustment
included age >1 year, witnessed arrests, and shockable
rhythm. Arrests in Black children, at a home/residence,
and arrests associated with AED use before EMS arrival
had decreased neurologically favorable survival
(Tables 2 and 3).
NEUROLOGICALLY FAVORABLE SURVIVAL AND THE
TYPE OF BYSTANDER CPR ACROSS AGE GROUPS.
Neurologically favorable survival was observed in
4.6% (250 of 5,456) of infants, 10.6% (312 of 2,936)
of children, and 16.5% (335 of 2,037) of adolescents.
In infants, neurologically favorable survival was
observed in 5.2% (58 of 1,109) of arrests with CO-CPR,
6.9% (67 of 973) with RB-CPR, and 3.7% (125 of 3,374)
with NO-CPR; P < 0.001 (Table 4). In multivariable
conducted analysis, RB-CPR was associated with neurologically
favorable survival compared with NO-CPR (adjusted
1046 Naim et al. JACC VOL. 78, NO. 10, 2021

Compression-Only Versus Rescue Breathing CPR in Pediatric OHCA SEPTEMBER 7, 2021:1042–1052

T A B L E 1 Characteristics of NO-CPR, CO-CPR, and RB-CPR

NO-CPR CO-CPR RB-CPR

(n ¼ 6188) (n ¼ 2318) (n ¼ 1923) P Value

Year of arrest
2013 515 (8.3) 189 (8.2) 204 (10.6) 0.001
2014 566 (9.1) 226 (9.8) 215 (11.2)
2015 782 (12.6) 281 (12.1) 242 (12.6)
2016 833 (13.5) 328 (14.2) 262 (13.6)
2017 1,051 (17.0) 359 (15.5) 252 (13.1)
2018 1,162 (18.8) 437 (18.9) 339 (17.6)
2019 1,279 (20.7) 498 (21.5) 409 (21.3)
Age, y
<1 3,374 (54.5) 1,109 (47.8) 973 (50.6) <0.001
1-11 1,661 (26.8) 640 (27.6) 635 (33.0)
$12 1,153 (18.6) 569 (24.6) 315 (16.4)
Sex
Female 2,561 (41.4) 888 (38.3) 808 (42.0) 0.018
Male 3,627 (58.6) 1,430 (61.7) 1,115 (58.0)
Race/ethnicity
White 1,708 (27.6) 899 (38.8) 844 (43.9) <0.001
Black 2,253 (36.4) 595 (25.7) 413 (21.5)
Hispanic 731 (11.8) 237 (10.2) 133 (6.9)
Other 194 (3.1) 58 (2.5) 36 (1.9)
Unknown 1,302 (21.0) 529 (22.8) 497 (25.8)
Bystander-witnessed arrest
Unwitnessed 4,873 (78.7) 1,591 (68.6) 1,343 (69.8) <0.001
Witnessed 1,315 (21.3) 727 (31.4) 580 (30.2)
Arrest location
Nonhome/public 791 (12.8) 333 (14.4) 251 (13.1) 0.155
Home/residence 5,397 (87.2) 1,985 (85.6) 1,672 (86.9)
Shockable rhythm
Nonshockable 5,815 (94.0) 2,075 (89.5) 1,777 (92.4) <0.001
Shockable 373 (6.0) 243 (10.5) 146 (7.6)
AED used before EMS
No 4,977 (80.4) 1,842 (79.5) 1,558 (81.0) 0.424
Yes 1,211 (19.6) 476 (20.5) 365 (19.0)
Region
Midwest 1,676 (27.1) 556 (24.0) 482 (25.1) <0.001
Northeast 687 (11.1) 224 (9.7) 234 (12.2)
South 2,318 (37.5) 791 (34.1) 568 (29.5)
West 1,507 (24.4) 747 (32.2) 639 (33.2)

Values are n (%). Demographic and clinical characteristics are shown by CPR group: NO-CPR, CO-CPR, and RB-CPR. P values reported via chi-square tests.
AED ¼ automated external defibrillation; CO-CPR ¼ compression only-cardiopulmonary resuscitation; CPR ¼ cardiopulmonary resuscitation; EMS ¼ emergency medical
services; NO-CPR ¼ no bystander-cardiopulmonary resuscitation; RB-CPR ¼ cardiopulmonary resuscitation with rescue breathing.

OR: 1.65; 95% CI: 1.19-2.30); however, CO-CPR was
not associated with outcome (Table 4, Central
Illustration). In children, neurologically favorable
survival was observed in 13.9% (89 of 640) of arrests
with CO-CPR, 17.3% (110 of 635) with RB-CPR,
and 6.8% (113 of 1,661) with NO-CPR; P < 0.001. In
multivariable analysis, both RB-CPR and CO-CPR
were associated with neurologically favorable sur-
vival compared with NO-CPR.
neurologically favorable survival was observed in
23.7% (135 of 569) of arrests with CO-CPR, 25.7%
(81 of 315) with RB-CPR, and 10.3% (119 of 1153)
with NO-CPR; P < 0.001. In multivariable analysis,
both RB-CPR and CO-CPR were associated with
neurologically favorable survival compared with
NO-CPR.
DISCUSSION
This investigation has several important findings.
In adolescents, CO-CPR is the most common type of CPR for pedi-
atric OHCA in the United States. In the overall
pediatric cohort aged 0 to 18 years, both RB-CPR
and CO-CPR were associated with neurologically
JACC VOL. 78, NO. 10, 2021 Naim et al. 1047
SEPTEMBER 7, 2021:1042–1052 Compression-Only Versus Rescue Breathing CPR in Pediatric OHCA

T A B L E 2 Logistic Regression of the Association of Type of Bystander CPR With Favorable Survival

Unadjusted
Neurologically Adjusted Neurologically
Favorable Survival Favorable Survival Adjusted OR

% P Value % 95% CI OR 95% CI P Value

CPR type
NO-CPR 5.8 <0.001 6.8 6.2-7.4
CO-CPR 12.2 9.7 8.7-10.7 1.61 1.34-1.94 <0.001
RB-CPR 13.4 12.0 10.7-13.2 2.16 1.78-2.62 <0.001
Year of arrest
2013 8.2 0.487 7.7 6.2-9.2
2014 10.1 9.9 8.1-11.6 1.41 0.98-2.03 0.067
2015 8.3 8.1 6.8-9.5 1.07 0.76-1.52 0.694
2016 7.7 7.7 6.5-9.0 1.01 0.71-1.43 0.977
2017 8.4 8.6 7.4-9.8 1.16 0.84-1.62 0.369
2018 9.1 9.4 8.2-10.5 1.30 0.95-1.80 0.103
2019 8.7 8.6 7.5-9.6 1.15 0.84-1.58 0.382
Age, y
<1 4.6 <0.001 6.6 5.9-7.3
1-11 10.6 10.2 9.3-11.2 1.79 1.48-2.17 <0.001
$12 16.5 9.8 8.8-10.9 1.70 1.38-2.09 <0.001
Sex
Female 7.8 0.013 8.4 7.6-9.1
Male 9.1 8.8 8.1-9.4 1.07 0.91-1.25 0.416
Race/ethnicity
White 10.1 <0.001 9.5 8.6-10.4
Black 6.6 7.5 6.6-8.5 0.73 0.59-0.90 0.003
Hispanic 7.6 7.8 6.3-9.3 0.76 0.57-1.01 0.062
Other 9.7 8.9 6.1-11.8 0.91 0.58-1.45 0.703
Unknown 9.6 8.9 7.8-9.9 0.91 0.74-1.12 0.361
Bystander-witnessed arrest
Unwitnessed 3.7 <0.001 4.7 4.2-5.3
Witnessed 23.2 16.2 14.9-17.6 4.51 3.81-5.34 <0.001
Arrest location
Nonhome/public 24.5 <0.001 13.5 12.0-15.0
Home/residence 6.2 7.3 6.8-7.9 0.44 0.37-0.53 <0.001
Shockable rhythm
Nonshockable 6.1 <0.001 6.8 6.3-7.3
Shockable 40.9 21.7 19.1-24.3 4.71 3.84-5.77 <0.001
AED used before EMS
No 8.0 <0.001 9.0 8.4-9.5
Yes 11.2 7.5 6.6-8.4 0.78 0.64-0.94 0.010
Region
Midwest 7.6 <0.001 8.2 7.2-9.2
Northeast 8.5 8.1 6.7-9.5 0.98 0.75-1.30 0.914
South 7.9 8.6 7.7-9.5 1.07 0.87-1.33 0.513
West 10.5 9.1 8.1-10.0 1.15 0.92-1.44 0.213

Adjusted neurologically favorable survival, ORs, 95% CIs, and P values are from a logistic regression model with neurologically favorable survival as the outcome variable and
covariates age group, year of arrest, sex, race/ethnicity, bystander-witnessed arrest, arrest location, shockable rhythm, AED used before EMS, and region of arrest. No ad-
justments for multiple testing were applied.
Abbreviations as in Table 1.

favorable survival compared with NO-CPR in pedi- with NO-CPR in all age groups. CO-CPR was asso-
atric OHCA. RB-CPR was associated with a higher ciated with neurologically favorable survival
odds of neurologically favorable outcome compared compared with NO-CPR in children and adolescents,
with CO-CPR (adjusted OR: 1.36; 95% CI: 1.10-1.68). but not in infants.
In age-stratified analysis, RB-CPR was associated In 2008, in order to increase bystander CPR rates
with neurologically favorable survival compared in adults with OHCA, the American Heart
1048 Naim et al. JACC VOL. 78, NO. 10, 2021

Compression-Only Versus Rescue Breathing CPR in Pediatric OHCA SEPTEMBER 7, 2021:1042–1052

T A B L E 3 Logistic Regression Comparing the Association of CO-CPR and RB-CPR With Neurologically Favorable Survival

Adjusted Neurologically Favorable

Survival Adjusted OR

% 95% CI OR 95% CI P Value

CPR type
CO-CPR 11.6 10.4-12.7
RB-CPR 14.2 12.8-15.6 1.36 1.10-1.68 0.005
Year of arrest
2013 10.4 7.7-13.1
2014 12.8 9.9-15.7 1.36 0.82-2.25 0.233
2015 12.0 9.6-14.4 1.23 0.76-1.98 0.403
2016 11.6 9.3-13.9 1.16 0.72-1.88 0.536
2017 14.2 11.9-16.5 1.59 1.01-2.51 0.046
2018 14.6 12.4-16.8 1.66 1.06-2.59 0.026
2019 12.4 10.5-14.2 1.29 0.83-2.00 0.265
Age, y
<1 9.0 7.7-10.4
1-11 15.6 13.9-17.3 2.17 1.68-2.82 <0.001
$12 14.7 12.7-16.7 1.98 1.48-2.66 <0.001
Sex
Female 12.0 10.7-13.4
Male 13.2 12.0-14.4 1.15 0.93-1.42 0.207
Race/ethnicity
White 14.0 12.6-15.5
Black 10.7 8.9-12.6 0.67 0.50-0.90 0.008
Hispanic 11.1 8.3-14.0 0.71 0.47-1.07 0.097
Other 12.4 6.5-18.3 0.83 0.40-1.74 0.622
Unknown 13.2 11.3-15.2 0.92 0.70-1.21 0.541
Bystander-witnessed arrest
Unwitnessed 7.5 6.4-8.6
Witnessed 20.8 18.7-22.9 3.87 3.08-4.85 <0.001
Arrest location
Nonhome/public 21.1 18.2-24.1
Home/residence 10.6 9.6-11.6 0.37 0.29-0.48 <0.001
Shockable rhythm
Nonshockable 9.9 8.9-10.9
Shockable 30.3 25.8-34.9 5.04 3.79-6.70 <0.001
AED used before EMS
No 13.5 12.5-14.6
Yes 10.5 8.9-12.1 0.69 0.52-0.90 0.006
Region
Midwest 11.9 10.2-13.6
Northeast 13.7 11.0-16.3 1.23 0.86-1.77 0.261
South 13.4 11.7-15.1 1.19 0.89-1.60 0.240
West 12.4 10.7-14.0 1.06 0.78-1.44 0.698

Adjusted neurologically favorable survival, ORs, 95% CIs, and P values are from a logistic regression model with neurologically favorable survival as the outcome variable and
covariates age group, year of arrest, sex, race/ethnicity, bystander-witnessed arrest, arrest location, shockable rhythm, AED used before EMS and region of arrest. Arrests with
no bystander CPR are omitted from this model. No adjustments for multiple testing were applied.
Abbreviations as in Table 1.

Association changed its guidelines to recommend instructions to bystanders should focus on CO-CPR
CO-CPR as an acceptable alternative to RB-CPR (18). (3-6,19-21).
The subsequent stronger recommendations in 2010, This is the first report to demonstrate an increase
2015, and 2017 for CO-CPR in adult OHCA followed in the rate of CO-CPR in pediatric OHCA in the
the multiple randomized trials comparing United States. Over the 5 years of this study, the
dispatcher-assisted CPR by CO-CPR and RB-CPR, rates of bystander CPR did not change, but the pro-
which showed superior outcomes with CO-CPR, portion of CO-CPR increased with no change in
and meta-analyses that concluded dispatcher neurologically favorable survival (Supplemental
JACC VOL. 78, NO. 10, 2021 Naim et al. 1049
SEPTEMBER 7, 2021:1042–1052 Compression-Only Versus Rescue Breathing CPR in Pediatric OHCA

T A B L E 4 Association of Bystander CPR Type With Neurologically Favorable Survival, Stratified by Age Group

Unadjusted Neurologically Adjusted Neurologically

Favorable Survival Favorable Survival Adjusted OR

N % P Value % 95% CI OR 95% CI P Value P Valuea

Infants
NO-CPR 125 of 3,374 3.7 <0.001 4.0 3.4-4.7
CO-CPR 58 of 1,109 5.2 4.6 3.5-5.6 1.16 0.83-1.62 0.394
RB-CPR 67 of 973 6.9 6.1 4.8-7.5 1.65 1.19-2.3 0.003 0.072
Children
NO-CPR 113 of 1,661 6.8 <0.001 7.4 6.2-8.7
CO-CPR 89 of 640 13.9 12.5 10.2-14.8 1.94 1.41-2.68 <0.001
RB-CPR 110 of 635 17.3 16 13.4-18.6 2.73 2.00-3.72 <0.001 0.046
Adolescents
NO-CPR 119 of 1,153 10.3 <0.001 13.5 11.6-15.4
CO-CPR 135 of 569 23.7 18.6 16-21.2 1.71 1.23-2.37 0.001
RB-CPR 81 of 315 25.7 21.0 17.3-24.6 2.12 1.44-3.11 <0.001 0.290

This table presents the unadjusted percent of out-of-hospital cardiac arrests with favorable outcome by age group, with P values from chi-square tests. Adjusted neurologically favorable
survival, ORs, 95% CIs, and P values are from stratified logistic regression models (by age group) with neurologically favorable survival as the outcome variable and covariates year of arrest,
sex, race/ethnicity, bystander-witnessed arrest, arrest location, shockable rhythm, AED used before EMS, and region of arrest. No adjustments for multiple testing were applied. aThis P value
shows the difference between adjusted neurologically favorable survival between CO-CPR and RB-CPR.
Abbreviations as in Table 1.

Table 1). This increase in CO-CPR in pediatric OHCA
was also reported from 2007 to 2014 in Japan where
there was an increased rate of bystander CPR and an
increased proportion of CO-CPR without a change
in neurologically favorable survival (22). The lack
of change in neurologically favorable survival over
time may be explained by the observation in the
present study of the higher odds of neurologically
favorable survival associated with RB-CPR compared
with CO-CPR.
The current report found that both RB-CPR and CO-
CPR were associated with neurologically favorable
survival in the overall pediatric cohort, in children
aged 1 to 11 years and in adolescents. These results are
similar to 2 recent reports from Japan. In the first,
Fukada et al (23) examined pediatric OHCA from 2011
to 2012 in children and adolescents aged 1 to 17 years,
and found that both RB-CPR and CO-CPR were asso-
ciated with neurologically favorable outcome at
1 month following cardiac arrest compared with NO-
CPR. Goto et al (22) examined 6810 pediatric OHCAs
in children aged <18 years from 2007 to 2014 and
found RB-CPR was associated with better outcomes
compared with CO-CPR in children aged 1 to 17 years;
however, there was no difference between RB-CPR
and CO-CPR in children with cardiac etiology, initial
shockable rhythm, or age >8 years. For infants, there
was no difference in outcomes with RB-CPR versus
CO-CPR when there was a cardiac etiology or a wit-
nessed arrest (22). By contrast, in the present study,
neurologically favorable survival in infants was only
observed with RB-CPR, and RB-CPR
associated with greater neurologically favorable sur-
vival compared with CO-CPR in children aged 1 to 11
years. In the current analysis, in the overall cohort,
RB-CPR was associated with higher neurologically
favorable survival compared with CO-CPR. These
differences were not seen in age subgroup analysis;
however, CO-CPR was associated with neurologically
favorable survival compared with NO-CPR in children
and adolescents, but not in infants.
Additional factors associated with decreased
neurological favorable survival included Black race
and AED use before EMS arrival. In previous analyses
from the CARES registry, Black children have been
shown to have decreased bystander CPR provision
related to neighborhood social determinants of health
(24). Although shockable rhythms were associated
with neurologically favorable survival AED use before
EMS arrival was associated with decreased neuro-
logical favorable survival. This observation has been
noted in arrests related to drowning, and although
the exact reasons for this finding are unknown, it is
possible that the quality of bystander CPR was
affected during AED application and use, or that AED
use is a surrogate for other factors that influence
outcomes, including prolonged times until EMS
arrival (25).
Overall, the results of this study support current
guidelines that recommend RB-CPR for pediatric
OHCA. These results also support the use of RB-CPR
and CO-CPR in children and adolescents with pedi-
atric OHCA. However, CO-CPR was not associated
was also with neurologically favorable survival in infants;
1050 Naim et al. JACC VOL. 78, NO. 10, 2021

Compression-Only Versus Rescue Breathing CPR in Pediatric OHCA SEPTEMBER 7, 2021:1042–1052

CE N T RA L I LL U ST RA TI ON Neurologically Favorable Survival and Cardiopulmonary Resuscitation

Type in Infant, Children, and Adolescents

25 P < 0.001
Marginal Probability of Neurologically Favorable Survival (%)

NS

P = 0.001
P < 0.001
20
P = 0.046

15 P < 0.001

10 P = 0.003

NS

NS
5

0
<1 1 to 11 ≥12
Age
No Bystander-Cardiopulmonary Resuscitation
Compression Only-Cardiopulmonary Resuscitation
Cardiopulmonary Resuscitation with Rescue Breathing
Naim, M.Y. et al. J Am Coll Cardiol. 2021;78(10):1042–1052.

Adjusted neurologically favorable survival and P values are from stratified logistic regression models (by age group) with neurologically
favorable survival as the outcome variable and covariates year of arrest, sex, race/ethnicity, bystander-witnessed arrest, arrest location,
shockable rhythm, AED used before EMS, and region of arrest. No adjustments for multiple testing were applied. Blue bars indicate NO-CPR;
red bars, CO-CPR; and purple bars, RB-CPR. Error bars indicate 95% confidence intervals. CO-CPR ¼ compression only-cardiopulmonary
resuscitation; CPR ¼ cardiopulmonary resuscitation; NO-CPR ¼ no bystander-cardiopulmonary resuscitation; RB-CPR ¼ cardiopulmonary
resuscitation with rescue breathing.

therefore, RB-CPR should continue to be the recom-
mended
increased bystander CPR rates and survival in adults
who experience OHCA (2), and CO-CPR has been the
focus of public health campaigns including statewide
educational efforts (2), high school education (26,27),
and dispatcher assisted CPR (3-5,28). Although these
efforts have improved overall outcomes after OHCA
in adults, it is possible that they have disadvantaged
children, especially infants and young children. The
results of this study have important implications on
bystander CPR education and training, which should
modality. CO-CPR is associated with continue to emphasize RB-CPR for infants in cardiac
arrest and teach lay rescuers how to perform RB-CPR.
STUDY LIMITATIONS. These data are observational
and therefore may be challenged by unmeasured
confounding. Determining the etiology of a cardiac
arrest in the field is problematic, so all nontraumatic
etiologies were considered regardless of the pre-
sumed initial etiology that was chosen in the field.
Type of bystander CPR is a supplemental element in
the CARES registry and therefore was not available for
JACC VOL. 78, NO. 10, 2021
SEPTEMBER 7, 2021:1042–1052
all arrests because only a subset of agencies chose to
enter this data element. A pediatric OHCA may have
received more than 1 type of bystander CPR from 2
different bystanders (eg, CO-CPR from one bystander
and RB-CPR from another bystander). The study did
not include data on dispatcher instruction, the qual-
ity of CPR, and training of the lay rescuer although lay
persons with medical training were more likely to
perform RB-CPR compared with lay family members
and lay persons. It is possible that the group of lay
persons with medical training provided superior
chest compressions, thus improving outcomes in the
RB-CPR group independent of the potential benefits
of rescue breaths. A sensitivity analysis was per-
formed excluding arrests with CPR provided by lay
medical rescuers that continued to support the
superiority of RB-CPR compared with CO-CPR in
pediatric OHCA. Duration of CPR was not available
and time to CPR was not analyzed due to the number
of unwitnessed arrests. There are other post-cardiac
arrest in-hospital confounders, for example, tar-
geted temperature management that may have
affected outcome. In addition, there is no long-term
follow-up of survivors. The trend
bystander CPR rates has a potential for information
bias in the form of recruitment bias as each year new
agencies are added to the CARES registry. Although in
the overall cohort RB-CPR was associated with
improved outcome compared with CO-CPR in the age
subgroup analysis, these differences
observed likely secondary to inadequate power to
detect differences within the smaller groups. These
findings need to be validated in larger studies. Per-
forming a randomized control trial to compare these 2
types of CPR in pediatric OHCA is likely not feasible.
CONCLUSIONS
In the overall pediatric cohort aged 0 to 18 years,
RB-CPR was associated with a greater odds of
REFERENCES
1. Naim MY, Burke RV, McNally BF, et al. Associ-
ation of bystander cardiopulmonary resuscitation
with overall and neurologically favorable survival
after pediatric out-of-hospital cardiac arrest in the
united states: a report from the Cardiac Arrest
Registry to Enhance Survival Surveillance Registry.
JAMA Pediatr. 2017;171:133–141.
2. Bobrow BJ, Spaite DW, Berg RA, et al. Chest
compression-only CPR by lay rescuers and survival
from out-of-hospital cardiac arrest. JAMA.
2010;304:1447–1454.
3. Hallstrom A, Cobb L, Johnson E, Copass M. Car-
diopulmonary resuscitation by chest compression
alone or with mouth-to-mouth ventilation. N Engl J
Med. 2000;342:1546–1553.
Naim et al. 1051
Compression-Only Versus Rescue Breathing CPR in Pediatric OHCA
neurologically favorable survival compared with
CO-CPR. CO-CPR was associated with neurologically
favorable survival compared with NO-CPR in children
and adolescents, but not in infants. These results
support the present AHA guidelines with RB-CPR as
the preferred modality for pediatric OHCA.
FUNDING SUPPORT AND AUTHOR DISCLOSURES
The Cardiac Arrest Registry to Enhance Survival was funded by the
United States Centers for Disease Control and Prevention from 2004
to 2012. The program is now supported through private funding from
the American Red Cross, the American Heart Association, and in-kind
support from Stryker Corporation and Emory University. At present,
the Centers for Disease Control and Prevention provide technical
support and expertise for the program. This work was also supported
by the Cardiac Center Clinical Research Core at Children’s Hospital of
Philadelphia and the Scientific Advisory of the American Red Cross.
The authors have reported that they have no relationships relevant to
the contents of this paper to disclose.
ADDRESS FOR CORRESPONDENCE: Dr Maryam Y.
Naim, Division of Cardiac Critical Care Medicine, Car-
diac Center, Children’s Hospital of Philadelphia, Uni-
versity of Pennsylvania Perelman School of Medicine,
3401 Civic Center Boulevard, Philadelphia, Pennsyl-
analysis of vania 19104, USA. E-mail: naim@email.chop.edu.
Twitter: @maryam_naim.
PERSPECTIVES
were not COMPETENCY IN PATIENT CARE AND PROCEDURAL
SKILLS: Although CO-CPR is associated with higher rates of
bystander participation and patient survival in adults suffering
OHCA, in pediatric cases outcomes are better with RB-CPR than
CO-CPR.
TRANSLATIONAL OUTLOOK: Larger studies are needed to
confirm the advantage of RB-CPR over CO-CPR for infants in
cardiac arrest.
4. Rea TD, Eisenberg MS, Culley LL, Becker L. 7. Koster RW, Baubin MA, Bossaert LL, et al. Eu-
Dispatcher-assisted cardiopulmonary resuscitation ropean Resuscitation Council Guidelines for
and survival in cardiac arrest. Circulation. Resuscitation 2010 section 2. Adult basic life
2001;104:2513–2516. support and use of automated external de-
fibrillators. Resuscitation. 2010;81:1277–1292.
5. Svensson L, Bohm K, Castren M, et al.
Compression-only CPR or standard CPR in out-of- 8. Young KD, Gausche-Hill M, McClung CD,
hospital cardiac arrest. N Engl J Med. 2010;363: Lewis RJ. A prospective, population-based study
434–442. of the epidemiology and outcome of out-of-
hospital pediatric cardiopulmonary arrest. Pediat-
6. Berg RA, Hemphill R, Abella BS, et al. Part 5: adult
rics. 2004;114:157–164.
basic life support: 2010 American Heart Association
Guidelines for Cardiopulmonary Resuscitation and 9. Hickey RW, Cohen DM, Strausbaugh S,
Emergency Cardiovascular Care. Circulation. Dietrich AM. Pediatric patients requiring CPR in the
2010;122(18 Suppl 3):S685–S705. https://doi.org/10. prehospital setting. Ann Emerg Med. 1995;25:495–
1161/CIRCULATIONAHA.110.970939 501.
1052 Naim et al.
Compression-Only Versus Rescue Breathing CPR in Pediatric OHCA
10. Berg MD, Schexnayder SM, Chameides L, et al.
Part 13: pediatric basic life support: 2010 Amer-
ican Heart Association Guidelines for Cardiopul-
monary Resuscitation and Emergency
Cardiovascular Care. Circulation. 2010;122(18
Suppl 3):S862–S875. https://doi.org/10.1161/CIR-
CULATIONAHA.110.971085
11. Atkins DL, Berger S, Duff JP, et al. Part 11:
pediatric basic life support and cardiopulmonary
resuscitation quality: 2015 American Heart Asso-
ciation Guidelines Update for Cardiopulmonary
Resuscitation and Emergency Cardiovascular Care.
Circulation. 2015;132(18 Suppl 3):S519–S525.
https://doi.org/10.1161/CIR.0000000000000265
12. Biarent D, Bingham R, Eich C, et al. European
Resuscitation Council Guidelines for Resuscitation
2010 section 6. Paediatric life support. Resusci-
tation. 2010;81:1364–1388.
13. Malta Hansen C, Kragholm K, Pearson DA,
et al. Association of bystander and first-responder
intervention with survival after out-of-hospital
cardiac arrest in North Carolina, 2010-2013.
JAMA. 2015;314:255–264.
14. Johnson MA, Grahan BJ, Haukoos JS, et al.
Demographics, bystander CPR, and AED use in
out-of-hospital pediatric arrests. Resuscitation.
2014;85:920–926.
15. McNally B, Stokes A, Crouch A, Kellermann AL.
CARES: Cardiac Arrest Registry to Enhance
Survival. Ann Emerg Med. 2009;54:674–683.e2.
16. Jacobs I, Nadkarni V, Bahr J, et al. Cardiac arrest
and cardiopulmonary resuscitation outcome re-
ports: update and simplification of the Utstein
templates for resuscitation registries: a statement
for healthcare professionals from a task force of the
International Liaison Committee on Resuscitation
(American Heart Association, European Resuscita-
tion Council, Australian Resuscitation Council, New
Zealand Resuscitation Council, Heart and Stroke
Foundation of Canada, InterAmerican Heart Foun-
dation, Resuscitation Councils of Southern Africa).
Circulation. 2004;110:3385–3397.
17. Safar P. Resuscitation After Brain Ischemia.
Churchill Livingston; 1981.
18. Sayre MR, Berg RA, Cave DM, et al. Hands-only
(compression-only) cardiopulmonary resuscita-
tion: a call to action for bystander response to
adults who experience out-of-hospital sudden
cardiac arrest: a science advisory for the public
from the American Heart Association Emergency
Cardiovascular Care Committee. Circulation.
2008;117:2162–2167.
19. Hupfl M, Selig HF, Nagele P. Chest-compres-
sion-only versus standard cardiopulmonary resus-
citation: a meta-analysis. Lancet. 2010;376:1552–
1557.
20. Kleinman ME, Brennan EE, Goldberger ZD,
et al. Part 5: adult basic life support and cardio-
pulmonary resuscitation quality: 2015 American
Heart Association Guidelines Update for Cardio-
pulmonary Resuscitation and Emergency Cardio-
vascular Care. Circulation. 2015;132(18 Suppl 3):
S414–S435.
21. Kleinman ME, Goldberger ZD, Rea T, et al.
2017 American Heart Association focused update
on adult basic life support and cardiopulmonary
resuscitation quality: an update to the American
Heart Association Guidelines for Cardiopulmonary
Resuscitation and Emergency Cardiovascular Care.
Circulation. 2018;137:e7–e13.
22. Goto Y, Funada A, Goto Y. Conventional versus
chest-compression-only cardiopulmonary resusci-
tation by bystanders for children with out-of-
hospital cardiac arrest. Resuscitation. 2018;122:
126–134.
23. Fukada T, Tsuchiya Y, Iwakiri H, Ozaki M. Is the
Ambu aScope 3 Slim single-use fiberscope equally
efficient compared with a conventional broncho-
JACC VOL. 78, NO. 10, 2021
SEPTEMBER 7, 2021:1042–1052
scope for management of the difficult airway?
J Clin Anesth. 2016;30:68–73.
24. Naim MY, Griffis HM, Burke RV, et al. Race/
ethnicity and neighborhood characteristics are
associated with bystander cardiopulmonary
resuscitation in pediatric out-of-hospital cardiac
arrest in the United States: a study from CARES.
J Am Heart Assoc. 2019;8:e012637.
25. Tobin JM, Ramos WD, Pu Y, Wernicki PG,
Quan L, Rossano JW. Bystander CPR is associated
with improved neurologically favourable survival
in cardiac arrest following drowning. Resuscita-
tion. 2017;115:39–43.
26. Cave DM, Aufderheide TP, Beeson J, et al.
Importance and implementation of training in
cardiopulmonary resuscitation and automated
external defibrillation in schools: a science advi-
sory from the American Heart Association. Circu-
lation. 2011;123:691–706.
27. Lorem T, Steen PA, Wik L. High school stu-
dents as ambassadors of CPR–a model for reaching
the most appropriate target population? Resusci-
tation. 2010;81:78–81.
28. Lerner EB, Rea TD, Bobrow BJ, et al.
Emergency medical service dispatch cardiopul-
monary resuscitation prearrival instructions to
improve survival from out-of-hospital cardiac ar-
rest: a scientific statement from the American
Heart Association. Circulation. 2012;125:648–655.
KEY WORDS bystander CPR, cardiac arrest,
child, compression-only CPR, out-of-hospital
cardiac arrest, pediatric
A PPE NDI X For supplemental tables, please
see the online version of this paper.
JOURNAL OF THE AMERICAN COLLEGE OF CARDIOLOGY
ª 2021 BY THE AMERICAN COLLEGE OF CARDIOLOGY FOUNDATION
PUBLISHED BY ELSEVIER
EDITORIAL COMMENT
Chest Compression-Only
Cardiopulmonary Resuscitation in
Pediatric Out-of-Hospital Cardiac Arrest
(Don’t) Take My Breath Away*
Gene Yong-Kwang Ong, MBBS
A dult basic life support for laypersons previ-
ously included chest compressions with
ventilations, which evolved from 15 com-
pressions with 2 ventilations to 30 compressions
with 2 ventilations (1-3). “Standard cardiopulmonary
resuscitation (CPR)” for adult basic life support by
laypersons was then 30 compressions followed by 2
ventilations (2). It is acknowledged that rescue
breathing is not easy to perform, even for trained res-
cuers. To improve bystander CPR rates, chest
compression–only CPR was a community initiative
for untrained lay rescuers to provide immediate basic
life support for adults and adolescents when they wit-
ness a cardiac arrest (3). Pediatric basic life support at
that time was more complex, because it varied
depending on the number of rescuers (30 compres-
sions followed by 2 ventilations for single rescuer
and 15 compressions followed by 2 ventilations for 2
or more rescuers) (4). It was thought that bystander
CPR rates were lower in the pediatric age group
compared with adults potentially because of this (5).
To increase bystander CPR rates and simplify basic
life support, many resuscitation councils adopted a
“common layperson basic life support algorithm,”
which included a chest compression to ventilation
at 30 to 2 for all cardiac arrests, including infants
and young children.
*Editorials published in the Journal of the American College of Cardiology
reflect the views of the authors and do not necessarily represent the
views of JACC or the American College of Cardiology.
From the Department of Emergency Medicine, KK Women’s and Chil-
dren’s Hospital, Singapore; and the Duke-NUS Medical School, Singapore.
The author attests they are in compliance with human studies commit-
tees and animal welfare regulations of the author’s institution and Food
and Drug Administration guidelines, including patient consent where
appropriate. For more information, visit the Author Center.
ISSN 0735-1097/$36.00
VOL. 78, NO. 10, 2021
Over the last decade, there was good evidence that
chest compression–only CPR was not inferior to chest
compressions with ventilations for adult out-of-
hospital cardiac arrest, especially in the context of
dispatch assistance and when CPR was rendered by
untrained laypersons (6,7). Recent pandemics,
including the ongoing coronavirus disease-2019
(COVID-19), emphasized the use of chest
compression–only CPR because of infection concerns
(8). Increasingly, chest compression–only CPR is
considered the new “standard CPR” for layperson
basic life support and has been widely advocated in
community CPR training (3,8). The potential impli-
cations on its use in pediatric out-of-hospital cardiac
arrest was of concern to many pediatric resuscitation
science experts (9).
For pediatric cardiac arrests, there has been
continued emphasis on ventilation in pediatric life
support (5,9). The vast majority of pediatric cardiac
arrests were previously reported to be noncardiac in
origin (4,5,8,9). Mechanistically, if the lungs are not
normal (eg, atelectasis from asphyxiation, infection,
bronchospasm, submersion, etc), unlike in witnessed
sudden cardiac arrest, the theoretical variable air flow
generated during chest compressions might not be
able to generate sufficient tidal volumes, and may
increase risks and duration of hypoxic ischemic en-
cephalopathy during cardiac arrest events (10).
Despite this emphasis on ventilation and lack of evi-
dence for the use of chest compression only–CPR in
the pediatric population, prehospital compression-
only CPR has also been reported to be increasing for
all ages, even including infants (11,12).
SEE PAGE 1042
In this issue of the Journal, Naim et al (13) analyzed
data from the Cardiac Arrest Registry to Enhance
https://doi.org/10.1016/j.jacc.2021.07.029
1054 Ong
Chest Compression-Only CPR in Pediatric Out-of-Hospital Cardiac Arrest
Survival for 13,055 nontraumatic pediatric out-of-
hospital cardiac arrests from 2013 to 2019. The in-
vestigators studied the associations of the type of
bystander CPR rendered to 10,429 pediatric cardiac
arrest victims with favorable survival at the time of
hospital discharge. They compared the
outcome of chest compression–only CPR, chest com-
pressions with rescue breathing, and no CPR, and
performed age subgroup analyses for infants, chil-
dren 1 to 12 years of age, and adolescents 12 years of
age and above. They found that chest compressions
with ventilation was associated with a higher odds of
neurologically favorable survival compared with
chest compression–only CPR across and within all 3
pediatric age groups. Of significance was their finding
that, unlike children and adolescents, there was no
difference in the outcome between infants who
received chest compression–only CPR and who did
not receive any CPR.
The new study adds important data to the limited
published pediatric published data on this topic
outside of Japan. A 2019 systematic review and meta-
analysis of chest compression–only CPR in pediatric
out-of-hospital cardiac arrest included 5 observa-
tional studies (4 Japanese and 1 North American) and
had a combined population of 14,427 participants
(14). Overall, it found improved clinical outcomes in
favor of chest compressions with ventilations when
compared with the use of chest compression–only
CPR provided by bystanders on pediatric out-of-
hospital cardiac arrest casualties. Subgroup analysis
demonstrated that this improved outcome was sig-
nificant only in the younger children (<8 years of age)
and in those whose cardiac arrest were not of cardiac
origin. An earlier 2017 meta-analysis did not find a
survival advantage for chest compression with rescue
breathing in the infant group, contrary to the findings
of this new study (7). The comparative size of the
population of this new study is noteworthy and may
potentially change the findings, especially for infants,
if included in future meta-analyses.
Naim et al (13) reported other important observa-
tions. Of note, they found that the presumed cause of
the pediatric cardiac arrest was mainly cardiac
(44.4%) in origin, whereas 32.8% were presumed
respiratory causes. This is different from previous
reports on pediatric-out-of-hospital cardiac arrest
(1-7,9,11,12,14). The potential advantages of bystander
chest compression–only CPR were not observed in
this study. They reported that the rates of community
CPR did not increase, despite an observed increased
trend on the use of chest compression–only CPR for
the pediatric population. Of the infants who received
JACC VOL. 78, NO. 10, 2021
SEPTEMBER 7, 2021:1053–1055
CPR, the predominant CPR rendered was chest
compression–only CPR, which was concerning, given
the outcomes. Similar to other pediatric out-of-
hospital studies, they also observed that bystander
CPR was most commonly provided by a lay family
clinical member (71.7%) (11,12). This has important implica-
tions because the perceived reluctance to provide
ventilation for adult cardiac arrest victims was partly
attributed to concerns about hygiene and potential
for infections. These considerations during the pro-
vision of lifesaving measures to a young family
member may not be quite the same for adult cardiac
arrest victims. Even during the ongoing COVID-19
pandemic, the risk of the rescuers acquiring COVID-19
through provision of rescue breaths may be more
readily accepted compared to adults, as the pediatric
cardiac arrest victims are likely to be family members
or people known to the pediatric victims. This may be
the case, especially if it is known that pediatric car-
diac arrest victims are more likely to have improved
outcomes if rescue breaths are provided during chest
compressions. Furthermore, rescuers would have
likely been close contacts of pediatric cardiac arrest
victims and could have already been exposed. It was
in this context that in the interim guidelines for pe-
diatric resuscitation during the COVID-19 pandemic,
rescuers who were “trained, willing, and able” were
encouraged to provide rescue breathing for pediatric
cardiac arrest victims (5,8,9).
Although there are limitations to the evidence
provided by cohort observational studies derived
from cardiac arrest registries, realistically, controlled
trials for bystander chest compressions are unlikely
possible, especially in the pediatric population. Given
that infants were consistently reported to experience
the worst clinical outcomes, and with the finding that
bystander cardiopulmonary resuscitation with venti-
lation could potentially improve neurologically
favorable survival, it behooves us to look into this in a
timely manner (1-12,14). In light of the data presented
by Naim et al (13), there need to be important dis-
cussions on pediatric CPR recommendations and ed-
ucation for laypersons in the community, especially
for infants.
FUNDING SUPPORT AND AUTHOR DISCLOSURES
The author has reported that he has no relationships relevant to the
contents of this paper to disclose.
ADDRESS FOR CORRESPONDENCE: Dr Gene Yong-
Kwang Ong, KK Women’s and Children’s Hospital,
100 Bukit Timah Rd, Singapore 229899. E-mail: gene.
ong.y.k@singhealth.com.sg OR geneong@yahoo.com.
JACC VOL. 78, NO. 10, 2021
SEPTEMBER 7, 2021:1053–1055
REFERENCES
1. American Heart Association in collaboration
with International Liaison Committee on Resusci-
tation. Guidelines 2000 for Cardiopulmonary
Resuscitation and Emergency Cardiovascular Care:
International Consensus on Science, Part 3: Adult
Basic Life Support. Resuscitation. 2000;46:29–72.
2. Berg RA, Hemphill R, Abella BS, et al. Part 5:
adult basic life support: 2010 American Heart
Association guidelines for cardiopulmonary resus-
citation and emergency cardiovascular care. Cir-
culation. 2010;122(18 Suppl 3):S685–S705.
3. Panchal AR, Bartos JA, Cabañas JG, et al. Adult
basic and advanced life support writing group.
part 3: adult basic and advanced life support:
2020 American Heart Association guidelines for
cardiopulmonary resuscitation and emergency
cardiovascular care. Circulation. 2020;142(16
Suppl 2):S366–S468.
4. Berg MD, Schexnayder SM, Chameides L, et al.
Part 13: pediatric basic life support: 2010 American
Heart Association guidelines for cardiopulmonary
resuscitation and emergency cardiovascular care.
Circulation. 2010;122(18 Suppl 3):S862–S875.
5. Topjian AA, Raymond TT, Atkins D, et al. Pedi-
atric basic and advanced life support collabora-
tors. part 4: pediatric basic and advanced life
support: 2020 American Heart Association guide-
lines for cardiopulmonary resuscitation and
emergency cardiovascular care. Circulation.
2020;142(16 Suppl 2):S469–S523.
Chest Compression-Only CPR in Pediatric Out-of-Hospital Cardiac Arrest
6. Zhan L, Yang LJ, Huang Y, et al. Continuous
chest compression versus interrupted chest
compression for cardiopulmonary resuscitation of
non-asphyxial out-of-hospital cardiac arrest.
Cochrane Database Syst Rev. 2017;3(3):
CD010134.
7. Ashoor HM, Lillie E, Zarin W, Pham B, Khan PA,
Nincic V, et al. Effectiveness of different
compression-to-ventilation methods for cardio-
pulmonary resuscitation: a systematic review.
Resuscitation. 2017;118:112–125.
8. Edelson DP, Sasson C, Chan PS, et al. Interim
guidance for basic and advanced life support in
adults, children, and neonates with suspected or
confirmed COVID-19: from the Emergency Car-
diovascular Care Committee and Get With The
Guidelines-Resuscitation Adult and Pediatric Task
Forces of the American Heart Association. Circu-
lation. 2020;141:e933–e943.
9. Maconochie IK, Aickin R, Hazinski MF, et al., for
the Pediatric Life Support Collaborators. Pediatric
life support: 2020 International Consensus on
Cardiopulmonary Resuscitation and Emergency
Cardiovascular Care Science With Treatment Rec-
ommendations. Circulation. 2020;142(16 Suppl 1):
S140–S184.
10. Charbonney E, Grieco DL, Cordioli RL, et al.
Ventilation during cardiopulmonary resuscitation:
what have we learned from models? Respir Care.
2019;64(9):1132–1138.
Ong 1055
11. Naim MY, Burke RV, McNally BF, et al.
Association of bystander cardiopulmonary resus-
citation with overall and neurologically favorable
survival after pediatric out-of-hospital cardiac
arrest in the United States: a report from the Cardiac
Arrest Registry to Enhance Survival Surveillance
Registry. JAMA Pediatr. 2017;171:133–141.
12. Atkins DL, Everson-Stewart S, Sears GK,
Daya M, Osmond MH, Warden CR, Berg RA, for the
Resuscitation Outcomes Consortium Investigators.
Epidemiology and outcomes from out-of-hospital
cardiac arrest in children: the Resuscitation Out-
comes Consortium Epistry-Cardiac Arrest. Circu-
lation. 2009;119(11):1484–1491.
13. Naim MY, Griffis HM, Berg RA, et al.
Compression-only versus rescue-breathing car-
diopulmonary resuscitation after pediatric out-of-
hospital cardiac arrest. J Am Coll Cardiol. 2021;78:
1042–1052.
14. Zhang X, Zhang W, Wang C, et al. Chest-
compression-only versus conventional cardiopul-
monary resuscitation by bystanders for children
with out-of-hospital cardiac arrest: a systematic
review and meta-analysis. Resuscitation. 2019;134:
81–90.
KEY WORDS bystander CPR, cardiac arrest,
child, compression-only CPR, out-of-hospital
cardiac arrest, pediatric
 JOURNAL OF THE AMERICAN COLLEGE OF CARDIOLOGY VOL. 78, NO. 10, 2021

ª 2021 BY THE AMERICAN COLLEGE OF CARDIOLOGY FOUNDATION

PUBLISHED BY ELSEVIER

THE PRESENT AND FUTURE

JACC STATE-OF-THE-ART REVIEW

Chronic Fatigue Syndrome and

Cardiovascular Disease
JACC State-of-the-Art Review

Benjamin H. Natelson, MD,a Danielle L. Brunjes, PHD,b Donna Mancini, MDb

ABSTRACT

Myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS) is a medically unexplained illness characterized by severe
fatigue limiting normal daily activities for at least 6 months accompanied by problems with unrefreshing sleep, exac-
erbation of symptoms following physical or mental efforts (postexertional malaise [PEM]), and either cognitive reports or
physiological evidence of orthostatic intolerance in the form of either orthostatic tachycardia and/or hypocapnia.
Although rarely considered to have cardiac dysfunction, ME/CFS patients frequently have reduced stroke volume with a
significant inverse relation between cardiac output and PEM severity. Magnetic resonance imaging of ME/CFS patients
compared with normal control subjects found significantly reduced stroke, end-systolic, and end-diastolic volumes
together with reduced end-diastolic wall mass. Another cardiovascular abnormality is reduced nocturnal blood pressure
assessed by 24-hour monitoring. Autonomic dysfunction is also frequently observed with postural orthostatic tachycardia
and/or hypocapnia. Two consecutive cardiopulmonary stress tests may provide metabolic data substantiating PEM.
(J Am Coll Cardiol 2021;78:1056–1067) © 2021 by the American College of Cardiology Foundation.

M yalgic encephalomyelitis/chronic fatigue

syndrome (ME/CFS) is a medically unex-
plained illness characterized by severe fa-
tigue limiting normal daily activities among other
occur in these patients, including reduced cardiac
output and peripheral changes such as autonomic
dysfunction and/or muscular disease.

symptoms. Fatigue is the quintessential symptom CLINICAL IDENTIFICATION OF ME/CFS

for diagnosis, which may have many etiologies
ranging from neuromuscular to psychiatric, infec-
tious, endocrine, and/or cardiac causes. A frequent
surrogate for fatigue is a reduction in functional ca-
pacity, which is predominantly determined by cardio-
vascular response to a maximal exercise demand. In
this review, we will describe cardiac abnormalities
that occur in ME/CFS patients and whether these car-
diovascular factors contribute to this syndrome. We
will review the evidence that central cardiac changes
Chronic fatigue syndrome is not diagnosed if
accompanied by any medical cause of severe fatigue;
thus, its prevalence is low, at about 0.3%, with it
occurring in women more often than in men (1,2). A
committee of the Institute of Medicine reviewed the
illness (3); they responded to patient concerns that
the illness’ name was stigmatizing (4) and recom-
mended changing its name to ME/CFS. Prior to
examining the role of cardiovascular disease on ME/

Listen to this manuscript’s

audio summary by
Editor-in-Chief
Dr. Valentin Fuster on
JACC.org.
From the aDepartment of Neurology, Icahn School of Medicine at Mount Sinai, New York, New York, USA; and the bDivision of
Cardiology, Department of Medicine, Icahn School of Medicine at Mount Sinai, New York, New York, USA.
Barry Hurwitz, PhD, served as Guest Associate Editor for this paper. Javed Butler, MD, MPH, MBA, served as Guest Editor-in-Chief
for this paper.
The authors attest they are in compliance with human studies committees and animal welfare regulations of the authors’
institutions and Food and Drug Administration guidelines, including patient consent where appropriate. For more information,
visit the Author Center.

Manuscript received May 5, 2021; accepted June 21, 2021.

ISSN 0735-1097/$36.00 https://doi.org/10.1016/j.jacc.2021.06.045

JACC VOL. 78, NO. 10, 2021
SEPTEMBER 7, 2021:1056–1067
HIGHLIGHTS

ME/CFS, a clinical entity of unknown
etiology characterized by PEM, is easily
diagnosed when following available
guidelines.

Patients with ME/CFS typically have
small hearts, low stroke volume, and low
total blood volume, and some have OI.

More studies are needed to understand
the role of deconditioning in producing
some of the characteristics of ME/CFS.
CFS, it will be important to describe how the diag-
nosis is made. To facilitate diagnosis, patients are
presented with a visual analog scale while asking
them about their symptoms (Figure 1). ME/CFS is
defined by medically unexplained fatigue lasting at
least 6 months and severe enough to produce a
substantial (rated as at least 3 on the visual analogue
scale shown in Figure 1) decrease in activity at work,
home, social events, or in school. In addition, the
patient has to report having significant problems
with unrefreshing sleep, symptom worsening
following mild physical or emotional demands
(called postexertional malaise [PEM]), and either
cognitive reports or physiological evidence of
orthostatic intolerance (OI) in the form of either
orthostatic tachycardia and/or hypocapnia (5).
Because the diagnosis itself implies disability,
inability to work is common in these patients (6).
Individuals who report moderate effects on activity
and/or on symptom burden carry a less disabling
diagnosis—idiopathic chronic fatigue. ME/CFS pre-
sents in certain stereotypical ways. First, more than
one-third of patients report their illness starting
suddenly most often with a flu-like presentation (7),
although there is no convincing evidence of ongoing
infection. A major comorbidity is psychiatric diag-
nosis—most often major depressive disorder and/or
generalized anxiety disorder. But, the rates of these
disorders occurring (7) do not differ substantially
from those occurring during convalescence
after myocardial infarction (8). Importantly, the
diagnosis of ME/CFS appears to carry with it an
increased risk of early cardiovascular mortality
(mean was 59 years of age for ME/CFS vs 78 years of
age for the U.S. population) (9). One group of re-
searchers suggested this risk may be a function of
reduced levels of omega-3 fatty acids, which they
found in most ME/CFS patients (10).
Natelson et al. 1057
Chronic Fatigue Syndrome and Cardiovascular Disease
Because the diagnosis depends on patient ABBREVIATIONS
AND ACRONYMS
self-report across a broad range of symp-
toms, it is by definition comprised of a
HRV = heart rate variability
heterogeneous patient pool ranging from its
ME/CFS = myalgic
being a somatic manifestation of depression
encephalomyelitis/chronic
to its being a brain disease (11). This is a fatigue syndrome
classical problem in medicine, which is OI = orthostatic intolerance
solved by developing biomarkers that will PEM = postexertional malaise
allow stratification of a subgroup of pa-
POSH = postural orthostatic
tients. Until such biomarkers are available, syndrome of hypocapnia (or
clinical criteria such as sudden vs gradual hyperventilation)
illness onset can be used to try to reduce POTS = postural orthostatic
tachycardia syndrome
heterogeneity. But, because the illness is
currently a diagnosis made only when VO2VT = oxygen output at
ventilatory threshold
medical causes of severe fatigue have been
eliminated, this is a limitation toward understand-
ing pathophysiology and arriving at specific
treatments.
METHODOLOGY
This review was conducted using PRISMA (Preferred
Reporting Items for Systemic Reviews and Meta-
Analysis). Articles were identified using PubMed,
and the search was done between July 17, 2020, and
August 7, 2020. Search terms are as follows:
“chronic fatigue syndrome” AND “heart” OR “car-
diovascular” OR “cardiac function” OR post-
exertional malaise” OR “postural tachycardia syn-
drome” OR “physical activity” OR “exercise capac-
ity” OR “CPET.” After duplicates were removed, a
total of 760 papers were reviewed. A prescreen was
completed to exclude papers that were not pub-
lished in English and did not have adult pop-
ulations, as well as papers published before 1995.
Three exceptions were made for older publications
that remained relevant to this paper. The remaining
656 papers were accessed for predetermined exclu-
sion criteria, which included the following: 1)
interventional studies that focused on pain, nutri-
tional, cognitive, or behavioral changes; 2) drug or
device interventions; 3) focused on other diseases,
such as multiple sclerosis, cancer, or post-traumatic
stress disorder; 4) studies unrelated to chronic fa-
tigue syndrome or cardiovascular disease; and 5)
animal studies. Papers were included or excluded
first based on title, second on abstract, and then
third on a review of the full paper. Overall, 128 full
papers were considered (Figure 2). Papers were
further evaluated based on redundancy, relevance
to medical practice, methodological quality, and
validity of inference made.
1058 Natelson et al. JACC VOL. 78, NO. 10, 2021

Chronic Fatigue Syndrome and Cardiovascular Disease SEPTEMBER 7, 2021:1056–1067

F I G U R E 1 ME/CFS Diagnosis Analog Scale

PLEASE READ THE FOLLOWING QUESTIONS TO THE PATIENT

Part 1: Life Spheres Criteria Part 2: Symptoms Criteria

During the past 6 months, how much has On the average, how much of a problem have
your fatigue reduced your activity during: you had over the past 6 months with:
• Your time at work • Unrefreshing sleep
• Your personal life • Your symptoms getting much worse after
mild exertion
• Your social life
• Difficulty with attention or concentration [or]
• Your time in school
o Evidence of orthostatic intolerance

Grading Scale Patient fulfills criteria for ME/CFS if they have:

• 1 or more 3-5 rating for Life Spheres Criteria
0 1 2 3 4 5
and
ne

ild

e
ia

r
at

• All 3 with at least a 3 rating for Symptoms Criteria

ve

ve
nt
No

er

Se

Se
ta
od

bs

For those with ratings of <3, diagnosis is idiopathic chronic fatigue

y
M

r
Su

Ve

While asking about a patient’s symptoms, the visual analog scale shown here is used to facilitate diagnosis of myalgic encephalomyelitis/chronic fatigue syndrome
(ME/CFS).

CONNECTION TO CARDIOVASCULAR DISEASE
CARDIAC DYSFUNCTION. Reduced stroke
Eleven ME/CFS patients, selected because they had
runs of flat or inverted T waves on 24-hour Holter
(12), underwent stress multigated acquisition and
were found to have abnormalities in
fraction, wall motion, or cardiac size (13). A later
study done by Natelson’s group using impedance
cardiography found reduced stroke volume in ME/
CFS patients on the severe end of the illness
spectrum as determined by severity and frequency
of symptoms when compared with data
sedentary, matched healthy control subjects (14). Of
particular interest was the fact that a significant
inverse relation was found between cardiac output
and self-related severity of PEM such that patients
with progressively more burden with PEM had
progressively reduced cardiac output. Because blood
pressure was maintained in these patients, the
conclusion was that a low output circulatory state
existed for some patients.
Researchers in Miami, Florida, confirmed this
finding of reduced stroke index in ME/CFS patients
who were also on the severe end of the illness spec-
trum—again compared with sedentary healthy control
subjects (15). The researchers used the operational
criteria for “severe ME/CFS” that Natelson’s group
volume. had previously developed (14,16); approximately one-
half of the patients studied were in this severe group.
Importantly, the Miami group found that this reduc-
tion was a function of reduced blood volume, defined
ejection as 8% decrease from normal total blood volume and
not a problem with cardiac contractility. In their
analysis, the observed group differences for cardiac
volume were corrected by controlling for total blood
volume deficits. The frequency with which subjects
had reduced total blood volume was 63% in the se-
from vere ME/CFS group, 27% with nonsevere ME/CFS, and
0% in sedentary control subjects; thus, about one-
half of the patients had low blood volume. The au-
thors suggested that a blood volume deficit affected
oxygen delivery and nutrient supply, thus impairing
hemodynamic regulation, resulting in fatigue and
other symptoms of ME/CFS. Another group extended
these findings, reporting not only lower red blood cell
volume in ME/CFS than in control subjects and
significantly reduced stroke, end-systolic, and end-
diastolic volumes, but also reduced end-diastolic
wall mass in the ME/CFS group as determined using
cardiac magnetic resonance (17). A good correlation
between wall mass and total volume and a significant
JACC VOL. 78, NO. 10, 2021 Natelson et al. 1059
SEPTEMBER 7, 2021:1056–1067 Chronic Fatigue Syndrome and Cardiovascular Disease

F I G U R E 2 PRISMA Flowchart

Articles Identified through Additional articles identified

PubMed database search through other sources
(n = 867) (n = 60)

Articles after duplicates removed

(n = 760)

Articles excluded during pre-screen (n = 104)

• Non-English (n = 29)
• Publication date before 1995 (n = 19)
• Non-adult population (n = 50)
• Unable to obtain full article (n = 6)

Full-text articles assessed

for eligibility
(n = 656) Articles excluded during full screen using
below Exclusion criteria (n = 528)
• Supplement/Drug interventions
• Pain or Cognitive interventions
• Animal studies
• Cognitive/behavioral studies
• Sleep studies
• Unrelated to CFS or CVD
• Other Diseases: MS, Cancer, PTSD
Full article reviewed
(n = 128)
Articles excluded after further review (n = 41)
• Redundancy
• Relevance to medical practice
• Methodological quality
• Inference validity
Articles included in
Review
(n = 87)

The PRISMA flowchart shows the systematic approach and exclusion criteria used during the literature review. CFS ¼ chronic fatigue
syndrome; CVD ¼ cardiovascular disease; MS ¼ multiple sclerosis; PTSD ¼ post-traumatic stress disorder.

negative relation between plasma volume and fatigue
severity was further described by these investigators.
One possible explanation for the reduced ventric-
ular wall mass in ME/CFS may be caused by overall
heart size. Indeed, 61% of a group of Japanese ME/
CFS patients were reported to fulfill an operational
definition of small heart size—calculated cardiotho-
racic ratios #42% from A-P view of chest x-ray. Mitral
valve prolapse was identified by echocardiography in
29% of those with small hearts compared with 0% in
patients whose cardiothoracic ratio exceeded 42%
(18). Cardiological evaluation of these patients found
abnormalities: first, about one-third of patients re-
ported dyspnea on exertion or chest pain; one-fifth
had ECG evidence of right axis deviation including
vertical axis (electrical axis 90  ); and they had
significantly reduced stroke volume and cardiac in-
dexes compared with healthy control subjects (19).
Some of these patients showed clinical improvement,
at which time follow-up evaluation showed larger
1060 Natelson et al.
Chronic Fatigue Syndrome and Cardiovascular Disease
heart size and improved stroke volume/cardiac
output; the obvious inference was a relation between
cardiac dysfunction and illness severity (20). Another
group using magnetic resonance spectroscopy found
a trend to reduced creatine phosphate/adenosine
triphosphate compared with healthy control subjects;
these authors dichotomized patients into a normal
cardiac metabolic group and low creatine phosphate/
adenosine triphosphate groups, with those in the
impaired cardiac group showing significantly reduced
both maximal and initial proton efflux rates (21). And
yet, another study noted reduced cardiac mass,
stroke volume, ejection fraction, and end-diastolic
diameter in ME/CFS (22).
Importantly, most of the studies reviewed in the
previous text did not use sedentary healthy people as
control subjects for the cardiovascular changes re-
ported, and deconditioning could explain some of
these findings. This seems to be the case, at least in
part, related to the question of reduced plasma vol-
ume and small heart size. Levine’s group reported
that 2 weeks of head-down bed rest produced similar
cardiovascular changes: namely, reduction of stroke
volume, plasma volume, and ventricular mass (23).
Not using such sedentary control subjects seems to
explain the chronotropic incompetence reported
several times over the years in ME/CFS patients per-
forming a cardiopulmonary exercise test (CPET)
(24,25). Cook et al (26) matched ME/CFS patients with
healthy, sedentary control subjects based on actual
maximal VO 2 attained; with this careful matching for
fitness, heart rates during CPET were the same for
patients and control subjects (26); this effect has
recently been replicated in a larger study where 99
ME/CFS patients were similarly matched to 99
healthy control subjects—again showing no difference
in heart rate over time (Dane Cook, personal
communication, May 1, 2021). Because V02max values
were the same for the 2 groups in both these studies,
they are the first in which one can infer that decon-
ditioning is not responsible for the fatigue and other
symptoms seen in ME/CFS.
Although such data do not support the idea of
deconditioning as a causal factor in the genesis of
the symptom complex of ME/CFS, deconditioning
probably can increase any patient’s symptom
burden. Inactivity-induced deconditioning can
result in a reduction of total blood volume that af-
fects stroke volume and leads to increased sympa-
thetic and reduced parasympathetic activity to the
sinus node. Cardiac atrophy can occur in as little as
2 weeks of marked inactivity (23). In addition, lack
of activity can produce muscle atrophy, with
decrease in muscle strength and endurance,
JACC VOL. 78, NO. 10, 2021
SEPTEMBER 7, 2021:1056–1067
enzymatic changes, decrease in capillarization,
changes in body composition, changes in plasma
volume, and changes in autonomic function. More-
over, deconditioning can lead to even more inac-
tivity, which further exacerbates deconditioning
resulting in a downhill spiral. Therefore, decondi-
tioning imposed by the rest inherent in having ME/
CFS may explain some of the putative cardiac ab-
normalities previously reported.
Interestingly, a recent publication on healthy aging
and cardiovascular function describes the hemody-
namic response to exercise over 6 decades of life (27).
Functional and structural changes occur with aging,
which lead to a decrease in resting cardiac output
and, because resting heart rate is generally un-
changed, to a lower stroke volume. In healthy older
adults, peak VO 2 was reduced because of lower stroke
volume, peak heart rate, and peak exercise end-
diastolic volume in the presence of similar ventricu-
lar filling pressures, suggesting reduced inotropic
reserve or possibly reduced Frank-Sterling reserve.
Therefore, the exercise response of patients with ME/
CFS is not dissimilar from normal aging.
Oldham et al (28) described the exercise hemody-
namic response in a group of 619 patients with un-
explained dyspnea. They identified 49 patients with
low peak VO 2, with normal biventricular function but
low peak cardiac output, and without exercise-
induced pulmonary hypertension. These patients
appeared limited by their inability to increase their
filling pressures during exercise, and the authors
hypothesized that there was a failure of their Frank
Starling response possibly from inadequate venous
vasoconstriction. A subgroup of 23 patients under-
went fluid loading with normal saline followed by a
second exercise test. The fluid loading resulted in
increase in cardiac output and VO 2 in 70% of these
patients. Although these patients were not identified
as having ME/CFS, the inference is that many may
have had that diagnosis. Interestingly, these authors
report that deconditioned subjects had higher pul-
monary capillary wedge and RA pressures at matched
workloads than trained counterparts, although their
CO and SV were lower. Patients with deconditioning
did not exhibit the preload failure that was observed
in the patients who would be most similar to ME/CFS
patients.
REDUCED BLOOD PRESSURE. Blood pressure ab-
normalities have been variably described in this dis-
order. Although 1 study found no differences in blood
pressure between patients and control subjects (29),
other studies described reduced blood pressure as
assessed by 24-hour monitoring with reductions in
both systolic and diastolic pressure (30) or reduction
JACC VOL. 78, NO. 10, 2021
SEPTEMBER 7, 2021:1056–1067
in systolic pressure alone (31). This reduction in blood
pressure was more marked during the nighttime than
the daytime, and there was a modest relation be-
tween increased fatigue severity and the variability
between daytime and nighttime SBP over 24 hours
(30). These researchers concluded that “low BP and
regulation of BP particularly may contribute to the
manifestation of the symptom of fatigue” (30).
Although this group did not use sedentary healthy
people as control subjects, they did use patients with
fatigue secondary to primary biliary cirrhosis who
differed from those with ME/CFS by showing only a
modest, but significant reduction in systolic blood
pressure over the 24-hour day. Another study noted
that patients had stiffer arteries, which correlated
both with systolic blood pressure and levels of C-
reactive protein (32). The autonomic dysfunction
commonly described in these patients may also
contribute to variations in blood pressure control.
This will be discussed below.
AUTONOMIC DYSFUNCTION. There has been a lot
of interest in the role of autonomic dysfunction in
ME/CFS. Pain and fatigue experienced by patients
are often correlated to symptoms of autonomic
dysfunction. One research method of capturing
autonomic activity (33,34) is by assessing heart rate
variability (HRV) which captures the magnitude of
respiratory-coupled sinus arrythmia—a variable that
allows quantification of parasympathetic modula-
tion of heart rate. Reductions in this variable have
been shown to predict all-cause death and cardio-
vascular events (35). The relation between sympa-
thetic activity and HRV is not as clear-cut, although
some inferences can be drawn using statistical
approaches to capture slower changing bands
of HRV.
Natelson et al have done a number of studies to
evaluate HRV in CFS. In the first of these studies us-
ing paced breathing—a method that magnifies
respiratory-coupled, parasympathetic activity—ME/
CFS patients showed reductions in both sitting and
standing positions compared with healthy control
subjects (36). In a subsequent study using the ortho-
static challenge of head up tilt, ME/CFS patients had a
lower mean RR interval compared with sedentary
control subjects (37), suggesting that autonomic
dysfunction is present during times of gravitational
stress. Several studies have shown evidence of
decreased HRV during sleep in ME/CFS patients
(38,39). In the study by Togo and Natelson (38),
autonomic dysfunction during non-REM sleep was
found to correlate to unrefreshing sleep—a relation
that could exacerbate the common symptom of daily
fatigue in ME/CFS patients. A meta-analysis (40)
Natelson et al. 1061
Chronic Fatigue Syndrome and Cardiovascular Disease
supported these findings, noting that ME/CFS pa-
tients had a higher ratio between low- and high-
frequency power (thought to reflect sympathetic ac-
tivity) than control subjects as well as lower high-
frequency power. An important recent study
showed a good correlation between measures of fa-
tigue and a number of HRV variables (41). This result
suggests that HRV could be a useful objective
outcome measure in future clinical trials.
ORTHOSTATIC INTOLERANCE. Another symptom
common in ME/CFS is the report of worsening
symptoms (more cognitive problems, fatigue) and/or
blurred vision and lightheadedness while upright—ie,
self-reported orthostatic sensitivity. This symptom
may reflect both the aforementioned reduction in
blood volume as well as impairment in autonomic
control. A recent paper reported that self-reported
symptoms consistent with OI occurred in 86% of pa-
tients (42). Patients with these symptoms have been
shown to show reductions in cognitive function
following orthostatic challenge (43). Should sympa-
thetic excitation exist (a subset of those with OI),
other symptoms might include palpitations, chest
pain, and tremulousness. Although approximately
one-half of patients with these reported symptoms
have no physiological abnormalities during ortho-
static challenge, the abnormality most often reported
is postural orthostatic tachycardia syndrome (POTS).
POTS is identified as when a subject’s supine, resting
heart rate increases by 30 beats/min or if the subject’s
heart rate exceeds 120 beats/min. Several papers have
noted POTS to be common in ME/CFS related to
severity of fatigue, and are therefore important to be
recognized (42,44). In contrast, another paper re-
ported found no difference in prevalence when
compared with healthy control subjects (45).
POTS has been described in a very heterogenous
group of diseases besides ME/CFS, with multiple
different pathophysiological subtypes that include
neuropathic, hypovolemic, hyperadrenergic, and
Ehlers Danlos Syndrome-hypermobility related phe-
notypes. Although POTS is common in ME/CFS, not
every patient with POTS has ME/CFS. We view the
existence of POTS and any other physiological mani-
festations of OI as a way to stratify ME/CFS patients;
the inference we make is that those with physiolog-
ical manifestations of OI may have a different un-
derlying pathophysiological process producing their
ME/CFS symptoms compared with other patients
without OI.
Several studies have compared ME/CFS with POTS
to those without this comorbidity. One study found
that those with POTS were younger, reported less
fatigue and sleepiness and had lower depression
1062 Natelson et al.
Chronic Fatigue Syndrome and Cardiovascular Disease
scores than the non-POTS patients (46). Another
study was similar but reported higher levels of fa-
tigue, shorter illness duration, and lower systolic
blood pressure during orthostatic challenge (47). In
another study, those with POTS had higher plasma
renin activity in both supine and upright positions
(48). In contrast, another study in POTS (49) reported
that despite the lower plasma volume found in these
patients, a compensatory increase in plasma renin
and aldosterone activity was not seen. Finally, one
recent study found those with reports consistent with
OI had lower blood volumes than those without re-
ported symptoms (50).
Concerning age, the rate of POTS in adolescents
with ME/CFS approaches 100% (51), but it is seen
much less often in adults (5). Of interest is a recent
report of reduced cerebral blood flow during head up
tilt in ME/CFS—even when there was no evidence of
OI (42).
Earlier reports noted a high rate of orthostatic hy-
potension occurring late in a 45-minute head-up tilt
(52). Two groups used the same protocol but were
unable to confirm differences in blood pressure be-
tween patients and healthy control subjects during
the same 45-minute head up tilt procedure (53,54).
This orthostatic hypotension is not seen often during
standardized 10-minute orthostatic challenges except
for in patients taking antihypertensives. In our
experience, fainting is uncommon using the 10-
minute challenge.
To simplify the assessment of patients for OI, we
have adapted the 10-minute lean test originally used
by the National Aeronautics and Space Administra-
tion to evaluate returning astronauts for this problem
(55). To do this, we collect supine blood pressure,
heart rate, respiratory rate, and end-tidal CO 2 once
per minute for several minutes after giving the pa-
tient adequate time for these physiological markers to
become baseline; then, we ask the patient to lean
against a wall with feet together touching only
shoulder blades to the wall and collect these same
variables once per minute for 10 minutes.
Using this same 10-minute test, 1 recent paper has
noted that reduced pulse pressure occurs during the
latter one-half of the lean test in patients sick for <4
years; this is thought to be an indirect measure of
reduced blood volume (56). One of the reasons that
some groups have not identified POTS as a common
problem in ME/CFS may be because an alternate form
of OI can exist. Specifically, several groups have re-
ported the postural orthostatic syndrome of hypo-
capnia (POSH) (5,57). Our own study indicated that
POSH occurs more often than POTS (5). Ocon et al (58)
have shown that POSH occurs first and is the driver
JACC VOL. 78, NO. 10, 2021
SEPTEMBER 7, 2021:1056–1067
for the development of POTS. These researchers also
have shown that, while undergoing orthostatic chal-
lenge, the cognitive processing abilities of young
adults deteriorates (58) but with no change in cerebral
blood flow. An important negative is that anxiety was
ruled out as a factor in producing either POTS (59) or
POSH (5).
The mechanism of OI continues to be an important
target for research. The change in posture with
orthostatic challenge leads to a shift in plasma vol-
ume with venous pooling caused by gravity leading to
a transient decline in venous return. This results in a
decrease in cardiac output and blood pressure with 2
consequences—unloading of the baroreceptors with
subsequent sympathetic stimulation to increase heart
rate, vasoconstriction, and venous return and/or
activation of chemoreceptors that produce hyperpnea
that functions possibly to pull blood into the chest.
These compensatory mechanisms are impaired in
POTS and POSH.
One possible cause of these dysregulated re-
sponses is systemic hypovolemia, which was reported
to occur in ME/CFS many years ago by Streeten and
Scullard (60). Joyner and Masuki (61) concluded that
the physiological manifestations of OI are identical to
what is seen under conditions producing marked
deconditioning. This conclusion must be tempered by
the need for studies in which the physiological
response to orthostatic challenge in patients is
compared with that found in matched sedentary
control subjects. Importantly, data do exist that a
program of gentle physical conditioning can improve
work capacity in ME/CFS as well as self-reported
health (62).
Another risk factor for OI—especially for POTS—is
the coexistence of the hypermobile form of Ehlers-
Danlos syndrome (63). This is operationally defined
by Beighton score testing as hyperflexibility in 5 sites
of the body plus the report of arthralgia in at least 4
joints. Further assessment includes a positive family
history of hypermobility with joint pain reports as
well as a history of musculoskeletal problems (eg,
long-term pain, dislocations) and signs of faulty
connective tissue throughout the body (eg, hyper-
mobile skin, hernias, prolapses). Heart and blood
vessel conditions occurring with EDS include heart
valve and vessel dysfunction, including mitral valve
prolapse and aortic root dilation. Fatigue is
acknowledged to be a common report of patients with
Ehlers-Danlos syndrome (64), and, accordingly, over
82% of hypermobile EDS patients fulfilled diagnostic
criteria for ME/CFS (65).
As a group, ME/CFS patients have increased blood
norepinephrine levels compared with control subjects
JACC VOL. 78, NO. 10, 2021
SEPTEMBER 7, 2021:1056–1067
on orthostatic challenge (66). When baroreceptor
activation produces excessive sympathetic activa-
tion, a hypertensive response to orthostatic challenge
can occur—that is, in a patient with no history of hy-
pertension. This “hyperadrenergic” form of POTS
may have other cardiac-related symptoms (67). Gib-
bons et al (68) used the existence of small fiber neu-
ropathy and raised sensory thresholds in the foot to
define a group of POTS patients as being neuropathic.
They note that those in the non-neuropathic group
show evidence of increased sympathetic activity on
Valsalva testing.
These studies on mechanisms suggesting sub-
groups of causes of OI raise questions for future
research. One critical problem is the lack of a study
evaluating all of the possible subgroups that includes
a large series of patients with reports indicative of
orthostatic testing. For example, the possibility that
patients with normal biopsy for small fiber neuropa-
thy are those who show an exaggerated blood pres-
sure response to head up tilt—ie, the hyperadrenergic
form. And, of course, an unanswered question is if
the blood volume of patients in these groups is
normal or reduced.
POSTEXERTIONAL MALAISE. A key characteristic of
ME/CFS is worsening across the broad range of
symptoms constituting ME/CFS after relatively minor
physical or even cognitive efforts. This PEM is com-
mon in ME/CFS and is thought to be a sine qua non
for the illness (3,69). Surprisingly, most data on PEM
derive from patient self-report collected at a time
unrelated to an actual bout of PEM. In 1 study, 11% of
patients reported a delay of at least 24 hours, with
most reporting symptoms lasting over a day (70).
Another study, using an epidemiological approach,
noted that the majority of patients reported symp-
tom worsening occurring within 2 days of exertion,
and only 9% reported a delay of at least 5 days before
this happened (71). One group evaluated patients for
PEM a week after they performed CPET; the patients
retrospectively reported having had a broad increase
in ME/CFS symptoms often lasting for several days,
which was not seen in control subjects (72). Patients
who completed the Profile of Mood State question-
naire following CPET reported more fatigue, pain,
and cognitive symptoms across the 3 days of data
collection compared with values before CPET (73). In
contrast, Natelson’s group, using actigraphy (74) and
real-time report of symptoms onto a watch-like
computer (75), found a delay of 6 days before
reduced activity and symptom worsening after CPET
developed. Thus, these studies corroborated the
occurrence of PEM following exertion, but noted a
Natelson et al. 1063
Chronic Fatigue Syndrome and Cardiovascular Disease
longer delay until objective evidence of its occur-
rence was seen. The variability in results across
these studies suggests that PEM may vary both in
intensity and time course—a point raised by the
Institute of Medicine’s report on ME/CFS (3). One
recent paper has noted that PEM is worse in patients
with elevated lactate levels in the baseline condition
($2 mmol/L) (76); confirming this result will be
important. Another approach that has attracted more
attention has sought to use exercise to precipi-
tate PEM.
USING CPETs TO ASSESS PEM. Two CPETs have been
proposed as a technique by which to diagnose and
quantify PEM in ME/CFS patients (77-82). Maximal
tests are performed on 2 consecutive days within a
24-hour period. Unfamiliarity with CPET testing and
novelty inherent in doing this for the first time can
lead to variability in the physiological measures ob-
tained during this testing and improvement on repeat
testing. However, this was not the case for ME/CFS
patients, who were reported to show reduced
maximal oxygen consumption (VO 2max) and earlier
onset of the anaerobic threshold, defined as the ox-
ygen consumption at the ventilatory threshold
(VO 2VT ). The authors inferred that these changes
were a metabolic manifestation of PEM, but demon-
strating a direct relation has not as yet been done.
With exercise, increased delivery of oxygen to the
working skeletal muscle is needed. Cardiac output,
lung function, hemoglobin concentration, vaso-
dilatory capacity of the vasculature, and muscle
metabolism are all key factors in determining oxygen
availability. In healthy normal individuals, the major
limitation to exercise performance is the ability to
increase cardiac output, which occurs by heart rate
acceleration, increase in contractility from catechol-
amine stimulation, and increase in stroke volume via
the Frank Starling mechanism. In the absence of
chronotropic incompetence, or heart failure, hypo-
volemia may impact cardiac output increase during
exercise. Oxygen consumption at the anaerobic or
ventilator threshold (VO 2VT ) is the point where carbon
dioxide production increases relative to VO 2, when
lactate made via glycolysis (ie, anaerobic metabolism)
is needed to meet metabolic demands that cannot be
sustained by oxidative mechanisms (aerobic meta-
bolism) alone (83).
The reduction of VO 2max and VO 2VT on day 2 of 2-
CPET studies in ME/CFS may be a biological marker
for ME/CFS (77,79). However, many of these studies
lacked a comparable control group, which the in-
vestigators contended was not necessary because
either of these variables do not decrease on repeat
1064 Natelson et al.
Chronic Fatigue Syndrome and Cardiovascular Disease
C E N T R A L IL LU ST R A T I O N Symptomatic Schematic of Myalgic Encephalomyelitis/Chronic
Fatigue Syndrome
POTS
Hyperadrenergic
drive
CFS
Volume
Depletion
Decondi-
tioning
SV
Post
Exertional
Malaise
Natelson, B.H. et al. J Am Coll Cardiol. 2021;78(10):1056–1067.
A symptomatic schematic of myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS). POTS ¼ postural orthostatic tachycardia syndrome; PTSD ¼
post-traumatic stress disorder; SV = stroke volume.
CPET in healthy people. However, the question re-
mains whether sedentary, deconditioned control
subjects would show the same response to 2-CPET as
do more active healthy control subjects.
Although similar results with a decrease in VO 2
max and VO 2VT were found in subsequent studies
(80,81), more recent reports failed to find a reduction
in day 2 VO 2max but continued to find reductions in
the VO2VT (81,82). Most recently, a study from Norway
showed that arterial lactate on day 2 CPET was
increased at the VO2VT in comparison with day 1
CPET, whereas it was decreased in healthy control
subjects (84)—a further confirmation of a metabolic
abnormality occurring in ME/CFS. And another recent
study shows that after day 2 CPET, female patients
tended to show progressive disability that paralleled
initial illness severity (85). Some investigators (82)
attributed the inconsistency of the data to differences
in the patients’ exercise intensity, ie, failing on day 2
to maximally exercise. More clinical research is
needed to determine if serial cardiopulmonary exer-
cise testing can provide an objective tool to assess
PEM and the mechanism for exertion-induced
induced PEM remains to be determined (86).
JACC VOL. 78, NO. 10, 2021
SEPTEMBER 7, 2021:1056–1067
Depression
PTSD Immune Past
activation infection
Neuropathic
changes
Fibro-
myalgia
Psych disorders
Fatigue
STATE OF THE CURRENT SCIENCE AND
FUTURE DIRECTIONS
Because ME/CFS is diagnosed based on clinical
criteria, it is a syndrome that, much like congestive
heart failure, will be shown to have multiple causes
(Central Illustration). A reviewer of this paper sug-
gested that chronic conditions limiting activity
(widespread pain or sleeplessness) might be expected
to produce severe fatigue and perhaps OI. That
reviewer suggested that longitudinal studies be per-
formed on younger patients with ME/CFS to deter-
mine whether the cardiac changes reported here do or
do not occur in the less symptomatic patient and
what happens to any cardiovascular abnormalities as
the illness waxes and wanes over time. Only cross-
sectional studies have been done to date. The re-
sults from such a longitudinal study might eliminate
some of the confounds in our current knowledge
base.
The lack of biomarkers or organ pathology has led
to the interpretation that ME/CFS could be a somatic
manifestation of psychiatric diagnoses such as anxi-
ety or major depression. However, the organic basis
JACC VOL. 78, NO. 10, 2021 Natelson et al. 1065
SEPTEMBER 7, 2021:1056–1067 Chronic Fatigue Syndrome and Cardiovascular Disease

for most cases of ME/CFS is supported by what we are
seeing today during the coronavirus disease-2019
(COVID-19) pandemic. A report on survivors of the
2003 SARS pandemic noted that 27% fulfilled 1994
case criteria for ME/CFS (87). Consistent with that
report, we are currently seeing many patients who
were relatively mildly affected by acute COVID-19
infection but who now report symptoms consistent
with the diagnosis of ME/CFS; many of these patients
also have physiological manifestations
Comparing the post–COVID-19 patient with ME/CFS
to the ME/CFS patient whose illness is not associated
with COVID-19 infection should help us better un-
resting cardiac output, decreased total blood volume,
and altered heart rate variability with postural hy-
potension. These cardiovascular changes may
contribute to the symptoms and debilitation associ-
ated with this disorder. Further research is needed to
assess if and how these cardiovascular changes are
associated with the severe fatigue and PEM of ME/
CFS and what role deconditioning plays in symptom
maintenance over time.
of OI.
FUNDING SUPPORT AND AUTHOR DISCLOSURES
The authors have reported that they have no relationships relevant to
the contents of this paper to disclose.

derstand the pathophysiology of both.

ADDRESS FOR CORRESPONDENCE: Dr Danielle L.
CONCLUSIONS Brunjes, Icahn School of Medicine at Mount Sinai,
One Gustave L Levy Place, Box 1030, New York, New
There are cardiovascular changes that are associated York 10029, USA. E-mail: Danielle.brunjes@
with ME/CFS, such as reduced stroke volume and mountsinai.org. Twitter: @BrunjesDanielle.

REFERENCES

1. Jason LA, Richman JA, Rademaker AW, et al.
A community-based study of chronic fatigue syn-
drome. Arch Intern Med. 1999;159:2129–2137.
2. Reyes M, Nisenbaum R, Hoaglin DC, et al.
Prevalence and incidence of chronic fatigue syn-
drome in Wichita, Kansas. Arch Intern Med.
2003;163(13):1530–1536.
3. IOM. Beyond myalgic encephalomyelitis/chronic
fatigue syndrome: redefining an illness. Mil Med.
2015;180(7):721–723. https://doi.org/10.7205/
MILMED-D-15-00085
4. Green J, Romei J, Natelson BH. Stigma and
chronic fatigue syndrome. Journal of Chronic Fa-
tigue Syndrome. 1999;5(2):63–75.
5. Natelson BH, Intriligator R, Cherniack NS,
Chandler HK, Stewart JM. Hypocapnia is a bio-
logical marker for orthostatic intolerance in some
patients with chronic fatigue syndrome. Dyn Med.
2007;6:2.
6. Ciccone DS, Natelson BH. Comorbid illness in
the chronic fatigue syndrome: A test of the single
syndrome hypothesis. Psychosom Med.
2003;62(2):268–275.
7. Natelson BH, Lin J-MS, Lange G, Khan S,
Stegner A, Unger ER. The effect of comorbid
medical and psychiatric diagnoses on chronic fa-
tigue syndrome. Ann Med. 2019;51(7-8):371–378.
8. Murphy B, Le Grande M, Alvarenga M,
Worcester M, Jackson A. Anxiety and depression
after a cardiac event: prevalence and predictors.
Front Psychol. 2019;10:3010.
9. McManimen SL, Devendorf AR, Brown AA,
Moore BC, Moore JH, Jason LA. Mortality in pa-
tients with myalgic encephalomyelitis and chronic
fatigue syndrome. Fatigue. 2016;4(4):195–207.
10. Castro-Marrero J, Zaragozá MC, Domingo JC,
Martinez-Martinez A, Alegre J, von Schacky C. Low
omega-3 index and polyunsaturated fatty acid
status in patients with chronic fatigue syndrome/
myalgic encephalomyelitis. Prostaglandins Leukot
Essent Fatty Acids. 2018;139:20–24.
11. Natelson BH, Mao X, Stegner AJ, et al. Multi-
modal and simultaneous assessments of brain and
spinal fluid abnormalities in chronic fatigue syn-
drome and the effects of psychiatric comorbidity.
J Neurol Sci. 2017;375:411–416.
12. Lerner AM, Lawrie C, Dworkin HS. Repetitively
negative changing T waves at 24-h electrocar-
diographic monitors in patients with the chronic
fatigue syndrome. Left ventricular dysfunction in a
cohort. Chest. 1993;104(5):1417–1421.
13. Dworkin HJ, Lawrie C, Bohdiewicz P,
Lerner AM. Abnormal left ventricular myocardial
dynamics in eleven patients with chronic fatigue
syndrome. Clin Nucl Med. 1994;19(8):675–677.
14. Peckerman A, LaManca JJ, Dahl K, Qureishi B,
Natelson BH. Abnormal impedance cardiography
predicts symptom severity in chronic fatigue syn-
drome. Am J Med Sci. 2003;326(2):55–60.
15. Hurwitz BE, Coryell VT, Parker M, et al. Chronic
fatigue syndrome: illness severity, sedentary life-
style, blood volume and evidence of diminished
cardiac function. Clin Sci (Lond). 2009;118(2):125–
135.
16. Nelson JJ, Natelson BH, Peckerman A, et al.
Medical follow up of Gulf Veterans with severe
medically unexplained fatigue: a preliminary
study. Mil Med. 2001;166:1107–1109.
17. Newton JL, Finkelmeyer A, Petrides G, et al.
Reduced cardiac volumes in chronic fatigue syn-
drome associate with plasma volume but not
length of disease: a cohort study. Open Heart.
2016;3(1):e000381.
18. Miwa K, Fujita M. Small heart syndrome in
patients with chronic fatigue syndrome. Clin Car-
diol. 2008;31(7):328–333.
19. Miwa K, Fujita M. Cardiovascular dysfunction
with low cardiac output due to a small heart in
patients with chronic fatigue syndrome. Intern
Med (Tokyo, Japan). 2009;48(21):1849–1854.
20. Miwa K, Fujita M. Cardiac function fluctuates
during exacerbation and remission in young adults
with chronic fatigue syndrome and “small heart.”.
J Cardiol. 2009;54(1):29–35.
21. Hollingsworth KG, Jones DE, Taylor R,
Blamire AM, Newton JL. Impaired cardiovascular
response to standing in chronic fatigue syndrome.
Eur J Clin Invest. 2010;40(7):608–615.
22. Olimulder MA, Galjee MA, Wagenaar LJ, et al.
Chronic fatigue syndrome in women assessed with
combined magnetic resonance imaging. Neth
Heart J. 2016;24(12):709–716.
23. Levine BD, Zuckerman JH, Pawelczyk JA. Car-
diac atrophy after bed-rest deconditioning - a
nonneural mechanism for orthostatic intolerance.
Circulation. 1997;96(2):517–525.
24. De Becker P, Roeykens J, Reynders M,
McGregor N, De Meirleir K. Exercise capacity in
chronic fatigue syndrome. Arch Intern Med.
2000;160(21):3270–3277.
25. Davenport TE, Lehnen M, Stevens SR,
VanNess JM, Stevens J, Snell CR. Chronotropic
intolerance: an overlooked determinant of symp-
toms and activity limitation in myalgic encepha-
lomyelitis/chronic fatigue syndrome? Front
Pediatr. 2019;7:82.
26. Cook DB, Nagelkirk PR, Poluri A, Mores J,
Natelson BH. The influence of aerobic fitness and
fibromyalgia on cardiorespiratory and perceptual
responses to exercise in patients with chronic fa-
tigue syndrome. Arthritis Rheum. 2006;54(10):
3351–3362.
27. Pandey A, Kraus WE, Brubaker PH,
Kitzman DW. Healthy aging and cardiovascular
function: invasive hemodynamics during rest and
exercise in 104 healthy volunteers. J Am Coll
Cardiol HF. 2020;8(2):111–121.
1066 Natelson et al.
Chronic Fatigue Syndrome and Cardiovascular Disease
28. Oldham WM, Lewis GD, Opotowsky AR,
Waxman AB, Systrom DM. Unexplained exertional
dyspnea caused by low ventricular filling pres-
sures: results from clinical invasive cardiopulmo-
nary exercise testing. Pulm Circ. 2016;6(1):55–62.
29. Duprez DA, De Buyzere ML, Drieghe B, et al.
Long- and short-term blood pressure and RR-
interval variability and psychosomatic distress in
chronic fatigue syndrome. Clin Sci. 1998;94(1):57–
63.
30. Newton JL, Sheth A, Shin J, et al. Lower
ambulatory blood pressure in chronic fatigue
syndrome. Psychosom Med. 2009;71(3):361–365.
31. Bozzini S, Albergati A, Capelli E, et al. Cardio-
vascular characteristics of chronic fatigue syn-
drome. Biomed Rep. 2018;8(1):26–30.
32. Spence VA, Kennedy G, Belch JJ, Hill A,
Khan F. Low-grade inflammation and arterial wave
reflection in patients with chronic fatigue syn-
drome. Clin Sci (Lond). 2008;114(8):561–566.
33. Berntson GG, Bigger Jr JT, Eckberg DL, et al.
Heart rate variability: origins, methods, and
interpretive caveats. Psychophysiology.
1997;34(6):623–648.
34. Cygankiewicz I, Zareba W. Heart rate vari-
ability. Handb Clin Neurol. 2013;117:379–393.
35. Fang SC, Wu YL, Tsai PS. Heart rate variability
and risk of all-cause death and cardiovascular
events in patients with cardiovascular disease: a
meta-analysis of cohort studies. Biol Res Nurs.
2020;22(1):45–56.
36. Sisto SA, Tapp W, Drastal S, et al. Vagal tone is
reduced during paced breathing in patients with
the chronic fatigue syndrome. Clin Auton Res.
1995;5(3):139–143.
37. Yamamoto Y, LaManca JJ, Natelson BH.
A measure of heart rate variability is sensitive to
orthostatic challenge in women with chronic fa-
tigue syndrome. Exp Biol Med (Maywood).
2003;228(2):167–174.
38. Togo F, Natelson BH. Heart rate variability
during sleep and subsequent sleepiness in patients
with chronic fatigue syndrome. Auton Neurosci.
2013;176(1-2):85–90.
39. Boneva RS, Decker MJ, Maloney EM, et al.
Higher heart rate and reduced heart rate variability
persist during sleep in chronic fatigue syndrome: a
population-based study. Auton Neurosci.
2007;137(1-2):94–101.
40. Nelson MJ, Bahl JS, Buckley JD, Thomson RL,
Davison K. Evidence of altered cardiac autonomic
regulation in myalgic encephalomyelitis/chronic
fatigue syndrome: a systematic review and meta-
analysis. Medicine (Baltimore). 2019;98(43):
e17600.
41. Escorihuela RM, Capdevila L, Castro JR, et al.
Reduced heart rate variability predicts fatigue
severity in individuals with chronic fatigue syn-
drome/myalgic encephalomyelitis. J Transl Med.
2020;18(1):4.
42. van Campen CLMC, Verheugt FWA, Rowe PC,
Visser FC. Cerebral blood flow is reduced in ME/
CFS during head-up tilt testing even in the
absence of hypotension or tachycardia: A quanti-
tative, controlled study using Doppler echog-
raphy. Clin Neurophysiol Pract. 2020;5:50–58.
43. van Campen C, Rowe PC, Verheugt FWA,
Visser FC. Cognitive function declines following
orthostatic stress in adults with myalgic enceph-
alomyelitis/chronic fatigue syndrome (ME/CFS).
Front Neurosci. 2020;14:688.
44. Hoad A, Spickett G, Elliott J, Newton J.
Postural orthostatic tachycardia syndrome is an
under-recognized condition in chronic fatigue
syndrome. QJM. 2008;101(12):961–965.
45. Roerink ME, Lenders JW, Schmits IC, et al.
Postural orthostatic tachycardia is not a useful
diagnostic marker for chronic fatigue syndrome.
J Intern Med. 2017;281(2):179–188.
46. Lewis I, Pairman J, Spickett G, Newton JL.
Clinical characteristics of a novel subgroup of
chronic fatigue syndrome patients with postural
orthostatic tachycardia syndrome. J Intern Med.
2013;273(5):501–510.
47. Reynolds GK, Lewis DP, Richardson AM,
Lidbury BA. Comorbidity of postural orthostatic
tachycardia syndrome and chronic fatigue syn-
drome in an Australian cohort. J Intern Med.
2014;275(4):409–417.
48. Okamoto LE, Raj SR, Peltier A, et al. Neuro-
humoral and haemodynamic profile in postural
tachycardia and chronic fatigue syndromes. Clin
Sci (Lond). 2012;122(4):183–192.
49. Raj SR, Biaggioni I, Yamhure PC, et al. Renin-
aldosterone paradox and perturbed blood volume
regulation underlying postural tachycardia syn-
drome. Circulation. 2005;111(13):1574–1582.
50. van Campen C, Rowe PC, Visser FC. Blood
volume status in ME/CFS correlates with the
presence or absence of orthostatic symptoms:
preliminary results. Front Pedtr. 2018;6:352.
51. Stewart JM, Gewitz MH, Weldon A,
Arlievsky N, Li K, Munoz J. Orthostatic intolerance
in adolescent chronic fatigue syndrome. Pediatrics.
1999;103(1):116–121.
52. Bou-Holaigah I, Rowe PC, Kan J, Calkins H. The
relationship between neurally mediated hypoten-
sion and the chronic fatigue syndrome. JAMA.
1995;274(12):961–967.
53. LaManca JJ, Peckerman A, Walker J, et al.
Cardiovascular response during head-up tilt in
chronic fatigue syndrome. Clin Physiol. 1999;19(2):
111–120.
54. Jones JF, Nicholson A, Nisenbaum R, et al.
Orthostatic instability in a population-based study
of chronic fatigue syndrome. Am J Med. 2005;118:
1415.
55. Bungo MW, Charles JB, Johnson Jr PC. Car-
diovascular deconditioning during space flight and
the use of saline as a countermeasure to ortho-
static intolerance. Aviat Space Environ Med.
1985;56(10):985–990.
56. Lee J, Vernon SD, Jeys P, et al. Hemodynamics
during the 10-minute NASA Lean Test: evidence of
circulatory decompensation in a subset of ME/CFS
patients. J Transl Med. 2020;18(1):314.
57. Naschitz JE, Rosner I, Rozenbaum M, et al. The
capnography head-up tilt test for evaluation of
chronic fatigue syndrome. Semin Arthritis Rheum.
2000;30(2):79–86.
58. Ocon AJ, Messer ZR, Medow MS, Stewart JM.
Increasing orthostatic stress impairs neurocognitive
JACC VOL. 78, NO. 10, 2021
SEPTEMBER 7, 2021:1056–1067
functioning in chronic fatigue syndrome with
postural tachycardia syndrome. Clin Sci (Lond).
2012;122(5):227–238.
59. Masuki S, Eisenach JH, Johnson CP, et al.
Excessive heart rate response to orthostatic stress
in postural tachycardia syndrome is not caused by
anxiety. J Appl Physiol (1985). 2007;102(3):896–
903.
60. Streeten DHP, Scullard TF. Excessive gravita-
tional blood pooling caused by impaired venous
tone is the predominant non-cardiac mechanism of
orthostatic intolerance. Clin Sci (Lond). 1996;90:
277–285.
61. Joyner MJ, Masuki S. POTS versus decondi-
tioning: the same or different? Clin Auton Res.
2008;18(6):300–307.
62. Wallman KE, Morton AR, Goodman C, Grove R,
Guilfoyle AM. Randomised controlled trial of
graded exercise in chronic fatigue syndrome. Med
J Aust. 2004;180(9):444–448.
63. Roma M, Marden CL, De Wandele I,
Francomano CA, Rowe PC. Postural tachycardia
syndrome and other forms of orthostatic intoler-
ance in Ehlers-Danlos syndrome. Auton Neurosci.
2018;215:89–96.
64. De Wandele I, Rombaut L, Malfait F, De
Backer T, De Paepe A, Calders P. Clinical hetero-
geneity in patients with the hypermobility type of
Ehlers-Danlos Syndrome. Res Dev Disabil.
2013;34(3):873–881.
65. Castori M, Celletti C, Camerota F,
Grammatico P. Chronic fatigue syndrome is
commonly diagnosed in patients with Ehlers-
Danlos syndrome hypermobility type/joint hyper-
mobility syndrome. Clin Exp Rheumatol.
2011;29(3):597–598.
66. Streeten DH. Role of impaired lower-limb
venous innervation in the pathogenesis of the
chronic fatigue syndrome. Am J Med Sci.
2001;321(3):163–167.
67. Kanjwal K, Saeed B, Karabin B, Kanjwal Y,
Grubb BP. Clinical presentation and management
of patients with hyperadrenergic postural ortho-
static tachycardia syndrome. A single center
experience. Cardiol J. 2011;18(5):527–531.
68. Gibbons CH, Bonyhay I, Benson A, Wang N,
Freeman R. Structural and functional small fiber
abnormalities in the neuropathic postural tachy-
cardia syndrome. PloS One. 2013;8(12):e84716.
69. Brown A, Jason LA. Meta-analysis investi-
gating post-exertional malaise between patients
and controls. J Health Psychol. 2020;25(13-14):
2053–2071.
70. Chu L, Valencia IJ, Garvert DW, Montoya JG.
Deconstructing post-exertional malaise in myalgic
encephalomyelitis/chronic fatigue syndrome: A
patient-centered, cross-sectional survey. PloS
One. 2018;13(6):e0197811.
71. Holtzman CS, Bhatia S, Cotler J, Jason LA.
Assessment of post-exertional malaise (PEM) in
patients with myalgic encephalomyelitis (ME) and
chronic fatigue syndrome (CFS): a patient-driven
survey. Diagnostics (Basel). 2019;9(1):26. https://
doi.org/10.3390/diagnostics9010026
72. VanNess JM, Stevens SR, Bateman L, Stiles TL,
Snell CR. Postexertional malaise in women with
JACC VOL. 78, NO. 10, 2021
SEPTEMBER 7, 2021:1056–1067
chronic fatigue syndrome. J Womens Health
(Larchmt). 2010;19(2):239–244.
73. Meyer JD, Light AR, Shukla SK, et al. Post-
exertion malaise in chronic fatigue syndrome:
symptoms and gene expression. Fatigue: Biomed
Health Behav. 2013;1(4):190–209.
74. Sisto SA, Tapp WN, LaManca JJ, et al. Physical
activity before and after exercise in women with
chronic fatigue syndrome. QJM. 1998;91(7):465–
473.
75. Yoshiuchi K, Cook DB, Ohashi K, et al. A real-
time assessment of the effect of exercise in
chronic fatigue syndrome. Physiol Behav.
2007;92(5):963–968.
76. Ghali A, Lacout C, Ghali M, et al. Elevated
blood lactate in resting conditions correlate with
post-exertional malaise severity in patients with
Myalgic encephalomyelitis/Chronic fatigue syn-
drome. Sci Rep. 2019;9(1):18817.
77. VanNess JM, Snell CR, Stevens SR. Diminished
cardiopulmonary capacity during post-exertional
malaise. J Chr Fatigue Syndr. 2007;14(2):77–85.
78. Vermeulen RC, Kurk RM, Visser FC, Sluiter W,
Scholte HR. Patients with chronic fatigue syn-
drome performed worse than controls in a
controlled repeated exercise study despite a
normal oxidative phosphorylation capacity.
J Transl Med. 2010;8:93.
79. Snell CR, Stevens SR, Davenport TE,
VanNess JM. Discriminative validity of metabolic
and workload measurements for identifying peo-
ple with chronic fatigue syndrome. Phys Ther.
2013;93(11):1484–1492.
80. Keller BA, Pryor JL, Giloteaux L. Inability of
myalgic encephalomyelitis/chronic fatigue syn-
drome patients to reproduce VO(2)peak indicates
functional impairment. J Transl Med. 2014;12:104.
81. Hodges LD, Nielsen T, Baken D. Physiological
measures in participants with chronic fatigue
syndrome, multiple sclerosis and healthy controls
following repeated exercise: a pilot study. Clin
Physiol Funct Imaging. 2018;38(4):639–644.
82. Nelson MJ, Buckley JD, Thomson RL, Clark D,
Kwiatek R, Davison K. Diagnostic sensitivity of 2-
day cardiopulmonary exercise testing in myalgic
encephalomyelitis/chronic fatigue syndrome.
J Transl Med. 2019;17(1):80.
83. Brooks GA. The science and translation of
lactate shuttle theory. Cell Metab. 2018;27(4):
757–785.
Natelson et al. 1067
Chronic Fatigue Syndrome and Cardiovascular Disease
84. Lien K, Johansen B, Veierod MB, et al.
Abnormal blood lactate accumulation during
repeated exercise testing in myalgic encephalo-
myelitis/chronic fatigue syndrome. Physiol Rep.
2019;7(11):e14138.
85. van Campen CLM, Rowe PC, Visser FC. Two-
day cardiopulmonary exercise testing in females
with a severe grade of myalgic encephalomyelitis/
chronic fatigue syndrome: comparison with pa-
tients with mild and moderate disease. Healthcare
(Basel). 2020;8(3):192.
86. Bouquet J, Li T, Gardy JL, et al. Whole blood
human transcriptome and virome analysis of ME/
CFS patients experiencing post-exertional malaise
following cardiopulmonary exercise testing. PloS
One. 2019;14(3):e0212193.
87. Lam MH, Wing YK, Yu MW, et al. Mental
morbidities and chronic fatigue in severe
acute respiratory syndrome survivors: long-term
follow-up. Arch Intern Med. 2009;169(22):2142–
2147.
KEY WORDS cardiac dysfunction,
cardiopulmonary exercise test (CPET),
ME/CFS, orthostatic intolerance

Listen to this manuscript’s
audio summary by
Editor-in-Chief
Dr. Valentin Fuster on
JACC.org.
JOURNAL OF THE AMERICAN COLLEGE OF CARDIOLOGY VOL. 78, NO. 10, 2021
ª 2021 BY THE AMERICAN COLLEGE OF CARDIOLOGY FOUNDATION
PUBLISHED BY ELSEVIER
JACC REVIEW TOPIC OF THE WEEK
Myocardial Viability Assessment
Before Surgical Revascularization in
Ischemic Cardiomyopathy
JACC Review Topic of the Week
Julio A. Panza, MD,a,b Lukasz Chrzanowski, MD,c Robert O. Bonow, MDd
ABSTRACT
Ischemic cardiomyopathy results from the combination of scar with fibrosis replacement and areas of dysfunctional but
viable myocardium that may improve contractile function with revascularization. Observational studies reported that only
patients with substantial amounts of myocardial viability had better outcomes following surgical revascularization.
Accordingly, dedicated noninvasive techniques have evolved to quantify viable myocardium with the objective of
selecting patients for this form of therapeutic intervention. However, prospective trials have not confirmed the
interaction between myocardial viability and the treatment effect of revascularization. Furthermore, recent observations
indicate that recovery of left ventricular function is not the principal mechanism by which surgical revascularization
improves prognosis. In this paper, the authors describe a more contemporary application of viability testing that is
founded on the alternative concept that the main goal of surgical revascularization is to prevent further damage by
protecting the residual viable myocardium from subsequent acute coronary events.
(J Am Coll Cardiol 2021;78:1068–1077) © 2021 by the American College of Cardiology Foundation.
I schemic heart disease is the most common
cause of heart failure (HF) with reduced ejection
fraction (EF) (1) and the single most important
factor contributing to the recent and projected in-
critically important pathophysiologic
ischemic cardiomyopathy is that the impairment in
LV contractile function is usually caused by a combi-
nation of scar with fibrosis replacement and areas of
aspect of
creases in HF incidence worldwide (2). The mecha- dysfunctional but viable myocardium. The latter
nism by which ischemic heart disease leads to HF could be explained by either stunning or hiberna-
is through the development of left ventricular (LV) tion, as discussed in the next section, and offers
systolic dysfunction, usually resulting from previous the potential for improvement in contractile func-
acute myocardial infarction(s), and, alternatively, tion with revascularization. This pathophysiologic
from an insidious process of progressive decline in aspect has significant implications for the manage-
systolic function without recognizable episodes of ment of patients with ischemic cardiomyopathy.
acute coronary syndromes. Thus, the term ischemic Accordingly, the recognition of myocardial viability
cardiomyopathy describes the syndrome of HF due in regions with poor systolic function has been the
to chronic LV systolic dysfunction resulting from un- focus of intense interest and investigation in recent
derlying coronary artery disease (CAD) (3). A decades.
From the aDepartment of Cardiology, Westchester Medical Center, Valhalla, New York, USA; bNew York Medical College, Valhalla,
New York, USA; cMedical University of Lodz, Lodz, Poland; and dNorthwestern University, Chicago, Illinois, USA.
The authors attest they are in compliance with human studies committees and animal welfare regulations of the authors’
institutions and Food and Drug Administration guidelines, including patient consent where appropriate. For more information,
visit the Author Center.
Manuscript received April 6, 2021; revised manuscript received June 21, 2021, accepted July 2, 2021.
ISSN 0735-1097/$36.00 https://doi.org/10.1016/j.jacc.2021.07.004
JACC VOL. 78, NO. 10, 2021
SEPTEMBER 7, 2021:1068–1077
HIGHLIGHTS

Noncontractile ischemic myocardium may
recover function after revascularization,
but recovery of ventricular function is not
the main mechanism by which surgical
revascularization improves prognosis.

The main benefit of surgical revasculari-
zation is prevention of further damage by
subsequent acute coronary events.

Viability testing can facilitate an assess-
ment of the likelihood of successful
revascularization of viable myocardial
segments.
MECHANISMS OF DYSFUNCTIONAL BUT
VIABLE MYOCARDIUM
The fundamental definition of myocardial viability
refers to cardiac muscle that is alive, not dead. When
applied to the clinical arena, however, the concept of
myocardial viability was developed to underscore
that LV systolic dysfunction in ischemic heart disease
does not always represent irreversible damage and
that dysfunctional but viable myocardium has the
potential to improve its systolic function after
revascularization (4-7). Two basic mechanisms of
reversible ischemic dysfunction have been described:
myocardial stunning and myocardial hibernation.
Myocardial stunning was defined as “prolonged
post-ischemic ventricular dysfunction that occurs
after brief episodes of non-lethal ischemia” (8). This
phenomenon is typified by the transient LV
dysfunction commonly observed following an acute
myocardial infarction treated with prompt reperfu-
sion (Figure 1, top).
The term myocardial hibernation was first coined
(9) to hypothesize the mechanism underlying the
reversibility of contractile dysfunction following
revascularization in patients with ischemic cardio-
myopathy (10). According to this concept, chronic LV
dysfunction results from an adaptive mechanism of
the myocardium to a state of critically reduced blood
flow. Thus, a new balance of demand and supply is
established in which flow and function are matched—
both at significantly reduced levels—to avoid
ischemia and cell death (11).
An alternative mechanism is repetitive stunning
caused by recurrent episodes of reversible ischemia.
This mechanism is supported by the finding of normal
resting blood flow in areas with systolic dysfunction
(12) and validated by animal models (13). With this
Panza et al. 1069
Contemporary Use of Myocardial Viability Assessment
mechanism, the resting blood flow is normal, ABBREVIATIONS
AND ACRONYMS
but the coronary flow reserve is extremely
reduced. Consequently, any instance of
CAD = coronary artery disease
increased myocardial oxygen demand leads
CABG = coronary artery bypass
to ischemia multiple times during daily life
graft surgery
(Figure 2A). This results in chronic systolic
EF = ejection fraction
dysfunction because the myocardium lacks
HF = heart failure
sufficient time to recover its contractile force
LV = left ventricular
before another episode of ischemia occurs
LVEF = left ventricular
(Figure 2B).
ejection fraction
Finally, it has been suggested, based on
PCI = percutaneous coronary
evidence from animal models, that stunning interventions
and hibernation represent a continuum in the
PET = positron emission
development of ischemic cardiomyopathy tomography
(14). According to this proposal, repetitive SPECT = single-photon
stunning with normal basal blood flow is an emission tomography
initial stage followed by reductions in resting flow as
a result—rather than the cause—of chronic contractile
dysfunction (15).
Importantly, successful revascularization has the
potential to improve chronic LV dysfunction regard-
less of the causative mechanism. Thus, in the para-
digm of myocardial hibernation, revascularization
leads to increases in resting blood flow, thus restoring
contractile function of the affected segments (Figure 1,
bottom). At the same time, successful revasculariza-
tion increases coronary flow reserve, thus abating the
repeated episodes of myocardial ischemia that ac-
count for repetitive stunning (Figure 2C).
MYOCARDIAL VIABILITY AND RECOVERY OF
LV FUNCTION
When applied to clinical practice, the very definition
of myocardial viability has been linked to the poten-
tial for dysfunctional myocardium to improve its
contractile force after revascularization. In fact, it has
been generally accepted that dysfunctional myocar-
dium that did not improve after successful revascu-
larization was, in retrospect, not viable.
Although improvement in LV systolic function is a
salutary effect of coronary revascularization, the
clinical concept deviates from the fundamental defi-
nition of viable myocardium (ie, myocardium that is
alive) in that it requires restoration of function as
proof of viability. However, a number of different
possibilities may explain the presence of viable
dysfunctional myocardium that does not improve
function with revascularization. These include the
presence of viability limited to the subepicardial
layers of segments with subendocardial scar and the
occurrence of perioperative infarction despite
adequate protection with cardioplegia (16).
1070 Panza et al. JACC VOL. 78, NO. 10, 2021

Contemporary Use of Myocardial Viability Assessment SEPTEMBER 7, 2021:1068–1077

F I G U R E 1 Mechanisms of Dysfunctional But Viable Myocardium

Myocardial Stunning

Sudden Reduction in Coronary Blood Flow

Due to Acute Coronary Occlusion
Restoration of Blood Flow with Rapid Reperfusion
100
Percent of Normal

Recovery of Function After Rapid

Reperfusion (Days to Weeks)
Stunned Myocardium: Persistent but Reversible
Post-Ischemic Systolic Dysfunction
0
Minutes Hours Days Weeks
Rapid Impairment of Myocardial Contractile
Function Due to Ischemia

Myocardial Hibernation

Restoration of Baseline Blood

Flow with Revascularization
100
Percent of Normal

Recovery of Function After

Revascularization

Hibernating Myocardium: Chronically Reduced

Myocardial Contraction to Match the Critical
Reduction in Baseline Blood Flow
0
Months
Coronary Blood Flow Myocardial Contractile Function

Two basic mechanisms have been proposed: myocardial stunning and myocardial hibernation. Myocardial stunning (top) is a process of reversible systolic dysfunction
following an episode of transient ischemia, such as in acute myocardial infarction with rapid reperfusion. A sudden cessation of coronary blood flow (blue line) is
followed by immediate impairment of myocardial contractile function (red line). If the coronary occlusion is resolved and restoration of blood flow occurs within
minutes, the stunned myocardium will recover its function within days or weeks. Myocardial hibernation (bottom) refers to a chronic state of matched reduction in
coronary blood flow and myocardial contraction. This adaptive mechanism results in the avoidance of ischemia at the cost of chronically impaired left ventricular
systolic function. With restoration of blood flow after successful coronary revascularization, the hibernating myocardium recovers its systolic function within weeks or
months.

The requisite for improvement in systolic function
with revascularization to be the ultimate reference
standard of myocardial viability has important con-
notations. From a diagnostic standpoint, it means
that only LV segments that improve function after
revascularization should be considered viable. In fact,
all studies investigating the sensitivity and specificity
of various techniques to assess myocardial viability
have used recovery of function after revasculariza-
tion as the gold standard (17).
JACC VOL. 78, NO. 10, 2021 Panza et al. 1071
SEPTEMBER 7, 2021:1068–1077 Contemporary Use of Myocardial Viability Assessment

F I G U R E 2 Repetitive Stunning Leading to Chronic Systolic Dysfunction of Viable Myocardium

A Multiple Episodes of Ischemia Due to Increases in Myocardial

Oxygen Demands Exceeding the Ischemic Threshold

Low Ischemic Threshold Due to Severe

Coronary Stenosis
Critically
Reduced Normal Baseline Blood Flow
Coronary
Flow Reserve
Repetitive Stunning
24 Hours

B Normalization of Coronary Flow Reserve C

with Revascularization

Critically
Reduced
Coronary
Flow Reserve Chronic LV Dysfunction Due to
Repetitive Stunning Improvement in LV Function Due
to Resolution of Repetitive Ischemia

Weeks Months Years Weeks Months Years

Coronary Blood Flow Myocardial Contractile Function Myocardial Oxygen Demand

(A) Multiple episodes of ischemia during the day lead to chronic systolic dysfunction due to repetitive stunning. In the presence of severe coronary stenosis, the blood
flow (blue line) at baseline may be normal, but coronary flow reserve is reduced, resulting in ischemia every time increases in myocardial oxygen demands (gray line)
exceed the low ischemic threshold (dashed line). Each episode of ischemia leads to impairment in contractile function (red line) that is potentially transient; however,
the occurrence of multiple episodes of ischemia never allows for full recovery of systolic function. (B) Over months or years, this process results in chronic left
ventricular (LV) dysfunction. (C) With normalization of coronary flow reserve after successful revascularization, increases in myocardial oxygen demands no longer
lead to ischemia, and as a salutary result, the viable myocardium recovers contractile function.

Dedicated noninvasive techniques have evolved to
identify more accurately the presence and extent of
viable myocardium. The 4 most widely used methods
in modern clinical practice are single-photon emission
computed tomography (SPECT), dobutamine echo-
cardiography, positron emission tomography (PET),
and cardiac magnetic resonance. A detailed descrip-
tion of each method is beyond the scope of this paper
and has been reviewed extensively (18). It is note-
worthy, however, that the physiologic basis for iden-
tifying viable myocardium differs from one technique
to the other. For instance, although the use of
SPECT requires only membrane integrity for the
identification of viability, a positive finding with
dobutamine echocardiography demands a contractile
apparatus capable of evoking a mechanical response
during inotropic stimulation. This has direct implica-
tions regarding the concordance among the different
methods ultimately used for the same purpose (19).
More importantly, from a therapeutic standpoint,
the recovery of LV function has been at the center of
the treatment goals and, arguably, is the most
meaningful indicator of success of revascularization
in patients with ischemic cardiomyopathy. Indeed,
from the initial descriptions of the hibernating
myocardium, the improvement in LVEF was touted as
1072 Panza et al. JACC VOL. 78, NO. 10, 2021

Contemporary Use of Myocardial Viability Assessment SEPTEMBER 7, 2021:1068–1077

C E NT R AL IL L U STR AT IO N Conceptual Framework for the Use of Myocardial Viability Information

Patient with ischemic cardiomyopathy considered for surgical revascularization

Traditional paradigm Contemporary paradigm

Assessment of myocardial viability Assessment of myocardial viability

focused on prespecified threshold focused on distribution of viable segments

Does patient have substantial amount of viable Is there concordance between viable segments
myocardium on viability testing? and vessels suitable for revascularization?

Yes No No Yes

Decision-making process

CABG No CABG No CABG CABG

Improvement in LV Protection of viable

systolic function segments
Mechanism of benefit of
revascularization
Amelioration of Reduced risk of fatal
ischemic heart failure myocardial infarction and
ventricular arrhythmias

Increased survival

Panza, J.A. et al. J Am Coll Cardiol. 2021;78(10):1068–1077.

The traditional paradigm (left) for the decision regarding CABG in patients with ischemic cardiomyopathy is based on a binary assessment of myocardial viability. Only
patients with a substantial amount of viable myocardium, using a dichotomous classification, are considered for surgical revascularization. The resulting better
outcomes are related to the improvement of LV systolic function and the amelioration of heart failure. A more contemporary paradigm (right) is founded on an
assessment of viability aimed to determine the feasibility of revascularizing viable myocardial regions. The mechanism of benefit from CABG is the reduction in the risk
of fatal myocardial infarction and ventricular arrhythmias. In all cases, guideline-directed medical therapy is the cornerstone for better outcomes, regardless of the
extent of viable myocardium. CABG ¼ coronary artery bypass graft surgery; LV ¼ left ventricular.

the most significant achievement of coronary artery
bypass graft surgery (CABG) (9).
It follows from this paradigm that the success of
revascularization depends on the restoration of con-
tractile function of viable dysfunctional myocardium.
This would ameliorate the process of ischemic HF and
thus lead to better outcomes (Central Illustration,
left). Indeed, previous studies have shown that the
extent of recovery of LV function corresponds to the
amount of dysfunctional but viable myocardium and
to the improvement in HF symptoms following
revascularization (20).
JACC VOL. 78, NO. 10, 2021
SEPTEMBER 7, 2021:1068–1077
INTERACTION BETWEEN MYOCARDIAL VIABILITY
AND BENEFIT FROM REVASCULARIZATION
The recognition that poorly contractile but viable
myocardium has the potential to recover its function
led to the concept that discrimination between viable
and nonviable myocardium is necessary to identify
those patients most likely to benefit from revascu-
larization with CABG. Accordingly, a number of
retrospective observational studies and several meta-
analyses that pooled data from these studies collec-
tively demonstrated that only patients with
substantial amounts of viable myocardium had better
outcomes following CABG, whereas patients without
viability received no benefit or were even harmed by
the revascularization procedure (21-24).
These reports contributed to the notion that
assessment of myocardial viability in patients with
ischemic cardiomyopathy is a prerequisite for clinical
decisions regarding revascularization. However, a
number of significant limitations, largely related to
the retrospective and observational nature of the
primary studies, preclude the acceptance of these
data as conclusive demonstration or confirmation
that there is a true interaction between the results of
viability studies and the benefit of CABG. These lim-
itations include heterogeneity in the criteria for the
inclusion of patients into the different studies and the
bias associated with knowledge of the noninvasive
imaging findings that likely influenced the decision
regarding revascularization. Most important is the
lack of adequate medical therapy in patients included
in studies dating back to the 1980s and early 1990s.
Optimization of guideline-directed medical therapy
in patients with and without substantial amounts of
viable myocardium undoubtedly leads to improved
outcomes, and this has not been reflected in the re-
sults of the early studies in which beta-adrenergic
blockers, in particular, were seldom used.
A few noteworthy prospective studies with a
randomization design have addressed the viability
hypothesis. The PARR-2 (PET and Recovery Following
Revascularization 2) trial (25) randomized patients to
a PET-guided strategy or standard care without PET.
Imaging physicians issued a recommendation, and
treating physicians made the final decision. The pri-
mary analysis did not show a significant advantage of
the PET-guided strategy. Post hoc analyses restricted
to patients in whom the treatment recommendation
was adhered to (25) or including selected partici-
pating sites (26) showed improved outcomes with the
PET-guided strategy. Nevertheless, these analyses
Panza et al. 1073
Contemporary Use of Myocardial Viability Assessment
were conducted retrospectively after the main study
results did not confirm its primary hypothesis.
The HEART (Heart Failure Revascularization Trial)
randomized patients who had evidence of myocardial
viability to either conservative management or coro-
nary angiography with intent for revascularization
(27). The study was terminated prematurely and
showed no differences in mortality between the
conservative and invasive strategies. However, the
trial was clearly underpowered to address this
endpoint.
The STICH (Surgical Treatment of Ischemic
Heart Failure) trial is, to date, the only prospective
randomized study addressing the effect of CABG in
patients with ischemic cardiomyopathy and the most
significant investigative effort addressing the
viability hypothesis. The main trial demonstrated,
after an extended follow-up, that patients random-
ized to CABG had a reduced rate of all-cause mortal-
ity, cardiovascular mortality, and all-cause mortality
plus cardiovascular hospitalization compared to
those randomized to guideline-directed medical
therapy alone (28). The viability substudy was
designed prospectively to address the interaction
between the presence of viable myocardium and the
benefit of CABG (29). Approximately one-half of the
patients enrolled into STICH underwent noninvasive
studies (30). The inclusion did not follow a random-
ization scheme; however, the treatment assignment—
as per the main trial—was randomized. Despite
confirmation of the survival benefit of CABG, there
was no demonstrable interaction between the pres-
ence of substantial amounts of viable myocardium
and the benefit of revascularization, either at 5 years
or at 10 years of follow-up (30,31). The study was
limited by the inclusion of a relatively small number
of patients without viability. Whether the inclusion of
a greater number of patients could have led to a
different conclusion remains unresolved.
Thus, in contrast to the results of early retrospec-
tive reports, none of the prospective trials was able to
confirm the usefulness of myocardial viability
assessment for decisions regarding surgical revascu-
larization in patients with ischemic cardiomyopathy.
This discrepancy has been highlighted in a more
recent meta-analysis (32).
Similar to the findings observed in the viability
substudy, a separate analysis of the STICH trial
showed no interaction between the presence or
absence of inducible myocardial ischemia and the
benefit of CABG (33). In contradistinction, the pres-
ence of severe LV remodeling (ie, lower EF and larger
1074 Panza et al.
Contemporary Use of Myocardial Viability Assessment
LV volume) and more extensive CAD (ie, stenosis in
all 3 major coronary arteries) identified those patients
more likely to benefit from surgical revascularization
(34).
Additional important observations stemming
from the STICH viability substudy focus on changes
in LV function and shed insight into the mechanism
of benefit from CABG. First, although improvement
in EF at 4 months is more likely among patients
with viability than in those without viability, this
improvement is not limited to patients receiving
CABG and is also observed after the optimization of
medical treatment (31). This is consistent with pre-
vious findings of improved LV function in viable
myocardium with beta-blockers in patients with HF
(35,36). Second, no relation was observed between
changes in LV function at 4 months (with or
without CABG) and subsequent long-term outcomes.
This is also consistent with previous reports (37)
and indicates that the improvement in EF at rest is
not the only and may not be the most important
mechanism for improved outcomes following CABG.
A more recent report that analyzed the STICH trial
database identified a small subset of patients with
substantial improvement in EF ($10%) 24 months
after randomization (38). Although such improve-
ment was associated with reduced subsequent
mortality, it was not related to the mode of treat-
ment, further suggesting that improvement in LV
function is not the main mechanism by which CABG
prolongs survival. Finally, it must be acknowledged
that an apparent failure to improve EF may also be
related to the intrinsic limitations in the measure-
ments that reduce the fidelity of detecting serial
changes over time (39).
Two ongoing trials may provide further evidence to
elucidate the relationship between myocardial
viability and the benefit of revascularization. IMAGE-
HF (Imaging Modalities to Assist With Guiding and
Evaluation of Patients With Heart Failure;
NCT01288560) is a prospective comparative effec-
tiveness study that will compare the impact of
advanced imaging techniques (PET and cardiac mag-
netic resonance) on clinical outcomes of patients with
ischemic HF with those observed using standard care,
including SPECT (40). REVIVED-BCIS2 (Study of Effi-
cacy and Safety of Percutaneous Coronary Interven-
tion to Improve Survival in Heart Failure;
NCT01920048) is a prospective randomized controlled
trial designed to determine whether revascularization
with percutaneous coronary interventions (PCI) re-
duces all-cause death and hospitalization for HF
compared to optimal medical therapy alone. This trial
JACC VOL. 78, NO. 10, 2021
SEPTEMBER 7, 2021:1068–1077
will enroll patients with extensive CAD, EF of #35%,
and demonstrable myocardial viability and will be the
first controlled study effort to assess the role of PCI in
improving the outcomes of patients with ischemic
cardiomyopathy (41).
CONTEMPORARY USE OF MYOCARDIAL
VIABILITY INFORMATION
The viability hypothesis (ie, that dysfunctional
myocardium with viability shown by noninvasive
methods improves contraction after revasculariza-
tion) is still valid at the cellular, segmental, and pa-
tient levels—patients with substantial amounts of
viable myocardium benefit from revascularization.
The more difficult question is whether surgical
revascularization should be recommended to patients
who do not demonstrate a certain amount of viable
myocardium on noninvasive testing. The findings of
the randomized studies suggest that the results of
viability testing do not discern the patients who
benefit from CABG from those who do not, in contrast
to what has been suggested in retrospective studies
and meta-analyses. However, it must be recognized
that the amount of viable myocardium is a contin-
uous variable and that the dichotomous classification
of patients as “with viability” or “without viability”
used so far is based on somewhat arbitrary thresholds
that vary from one technique to another and even
from one report to another when using the same
technique.
Most importantly, one must consider the mecha-
nisms underlying the benefit of CABG. Improvement
in LVEF with revascularization is a salutary result and
the hallmark of viable myocardium, as previously
defined. However, this may not be what matters
most, because recovery of LVEF does not seem to
have a significant impact on subsequent outcomes.
If recovery of LVEF is not critical, are there other
reasons to recommend CABG in patients without
substantial amounts of viable myocardium? As
demonstrated by the analysis of the mode of death in
the STICH trial, the most important mechanism of
benefit of CABG is the protection against fatal
myocardial infarction and sudden death caused by
future acute coronary events, despite the upfront
greater risk of death from the procedure (42). Even
patients considered to be “without viability” in a
dichotomous classification have other regions of
viable myocardium that sustain their systolic func-
tion; in some patients, these viable regions are also
potentially ischemic. The most important goal of
surgical revascularization may not be related to the
JACC VOL. 78, NO. 10, 2021
SEPTEMBER 7, 2021:1068–1077
recovery of systolic function but, instead, to the
prevention of further damage. This is consistent with
the finding of an interaction between the benefit of
CABG and the extent of CAD and of LV systolic
dysfunction and remodeling. Perhaps somewhat
paradoxical at first sight, the patients who benefitted
the most from CABG in STICH were those with
extensive disease (ie, involvement of all 3 vessels)
and with worse EF and larger end-systolic volumes
(34). These patients can be described as those with
the greatest number of vulnerable plaques and with
the greatest myocardial damage from previous in-
farctions. Simply put, these patients are at greatest
risk of a future acute coronary event and, at the same
time, are least able to tolerate it. Hence, they are most
likely to benefit from CABG, whether or not there are
large areas of viable myocardium on noninvasive
testing.
Other mechanisms of benefit from CABG must also
be considered, including the amelioration
myocardial ischemia resulting from improvement in
coronary flow reserve and the potential decrease in
cumulative microinfarctions leading to ventricular
arrhythmias and progressive HF. Revascularization
may also provide functional and electrical stability to
myocytes that do not necessarily contribute to
measured LV systolic function because they are
trapped among layers of scar (43). In addition, the
benefit of CABG extends not only to prolonged sur-
vival but also to improvement in the quality of life
and the exercise capacity of patients with ischemic
cardiomyopathy (44,45).
Accordingly, patients with ischemic cardiomyopa-
thy who do not strictly meet the dichotomous criteria
for viability may also be candidates for CABG, mainly
because the benefit of surgical revascularization ex-
tends beyond the recovery of LV systolic function. A
critical factor to consider is the correspondence be-
tween the dysfunctional but viable myocardial seg-
ments and the feasibility of surgical revascularization
of the coronary artery serving that territory. Although
this important issue has not been addressed in detail
in clinical trials, it is crucial in the individualized
decision-making process, as is the issue
completeness of revascularization. In this regard, one
must note the fundamental differences between sur-
gical and percutaneous revascularization. Whereas
CABG protects the myocardium from the adverse ef-
fects of potential future rupture of flow-limiting and
non–flow-limiting atherosclerotic plaques, PCI ad-
dresses only the stenotic lesion where the stent is
placed. Thus, surgical revascularization provides a
more complete form of protection, which is most
Panza et al. 1075
Contemporary Use of Myocardial Viability Assessment
relevant in patients with multivessel disease and LV
dysfunction (46).
Thus, the contemporary application of myocardial
viability testing in patients with ischemic cardiomy-
opathy (Central Illustration, right) is founded on the
observation that the main benefit of CABG is the
prevention of subsequent fatal myocardial infarction
(42), regardless of whether a patient is classified as
“with” or “without” viability on noninvasive testing.
This requires an assessment of viability integrated to
the findings of coronary angiography, primarily to
determine the anatomic correspondence between the
viable segments and the vessels that are suitable for
revascularization. This should include an assessment
of the caliber of distal vessels, particularly in diabetic
patients, because their poor quality may limit the
treatment benefit of CABG. The decision then rests on
the likelihood of successful revascularization of the
viable myocardial regions. Finally, the presence of
of important comorbidities such as advanced age,
severity of mitral regurgitation, renal dysfunction,
and overall frailty are important determinants in the
final decision regarding surgical revascularization,
particularly considering the upfront risk associated
with CABG. Although PCI offers the advantage of
reduced procedural risk, the benefit of this form of
revascularization in patients with ischemic cardio-
myopathy has not been demonstrated.
CONCLUSIONS
The basic viability hypothesis (that dysfunctional but
viable myocardium may recover systolic contraction
with revascularization) remains valid. However, its
corollary (ie, that patients without substantial
amounts of viable myocardium do not benefit from
surgical revascularization) is not applicable to all pa-
tients. Although patients with viable myocardium on
noninvasive testing are prime candidates for CABG,
those “without viability” require a more thoughtful
and individualized approach with regard to the
constellation of factors that influence the decision-
making process.
of Hence, noninvasive assessment of myocardial
viability remains an important part of the evaluation
of patients with ischemic cardiomyopathy. The
seemingly “negative” results of the STICH viability
substudy indicate that the findings of these tests
should not be applied in a dogmatic fashion. The
decision to be made (CABG or no CABG) is binary, but
the many factors to consider in reaching that decision
are not. Most importantly, all patients with ischemic
cardiomyopathy, with or without revascularization,
1076 Panza et al.
Contemporary Use of Myocardial Viability Assessment
benefit from guideline-directed medical therapy for
LV systolic dysfunction.
FUNDING SUPPORT AND AUTHOR DISCLOSURES
The authors have reported that they have no relationships relevant to
the contents of this paper to disclose.
REFERENCES
1. Gheorghiade M, Sopko G, De Luca L, et al.
Navigating the crossroads of coronary artery dis-
ease and heart failure. Circulation. 2006;114:1202–
1213.
2. Moran AE, Forouzanfar MH, Roth GA, et al. The
global burden of ischemic heart disease in 1990
and 2010: the Global Burden of Disease 2010
study. Circulation. 2014;129:1493–1501.
3. Felker GM, Shaw LK, O’Connor CM.
A standardized definition of ischemic cardiomy-
opathy for use in clinical research. J Am Coll Car-
diol. 2002;39:210–218.
4. Diamond GA, Forrester JS, deLuz PL, Wyatt HL,
Swan HJ. Post-extrasystolic potentiation of
ischemic myocardium by atrial stimulation. Am
Heart J. 1978;95:204–209.
5. Popio KA, Gorlin R, Bechtel D, Levine JA.
Postextrasystolic potentiation as a predictor of
potential myocardial viability: preoperative ana-
lyses compared with studies after coronary bypass
surgery. Am J Cardiol. 1977;39:944–953.
6. Chesebro JH, Ritman EL, Frye RL, et al.
Regional myocardial wall thickening response to
nitroglycerin. A predictor of myocardial response
to aortocoronary bypass surgery. Circulation.
1978;57:952–957.
7. Helfant RH, Pine R, Meister SG, Feldman MS,
Trout RG, Banka VS. Nitroglycerin to unmask
reversible asynergy. Correlation with post-
coronary bypass ventriculography. Circulation.
1974;50:108–113.
8. Braunwald E, Kloner RA. The stunned myocar-
dium: prolonged, postischemic ventricular
dysfunction. Circulation. 1982;66:1146–1149.
9. Rahimtoola SH. A perspective on the three
large multicenter randomized clinical trials of
coronary bypass surgery for chronic stable angina.
Circulation. 1985;72(suppl V):V-123–V-135.
10. Braunwald E, Rutherford JD. Reversible
ischemic left ventricular dysfunction: evidence for
the “hibernating myocardium”. J Am Coll Cardiol.
1986;8:1467–1470.
11. Rahimtoola SH. The hibernating myocardium.
Am Heart J. 1989;117:211–221.
12. Vanoverschelde JL, Wijns W, Depre C, et al.
Mechanisms of chronic regional postischemic
dysfunction in humans. New insights from the
study of noninfarcted collateral-dependent
myocardium. Circulation. 1993;87:1513–1523.
13. Shen Y-T, Vatner SF. Mechanism of impaired
myocardial function during progressive coronary
stenosis in conscious pigs: hibernation versus
stunning. Circ Res. 1995;76:479–488.
ADDRESS FOR
Panza, Department of Cardiology, Westchester Med-
ical Center, 100 Woods Road, Macy Pavilion, Suite
100, Valhalla, New York 10595, USA. E-mail: julio.
panza@wmchealth.org. Twitter: @westchestermed.
14. Canty JM Jr, Fallavollita JA. Chronic hiberna-
tion and chronic stunning: a continuum. J Nucl
Cardiol. 2000;7:509–527.
15. Canty JM Jr, Fallavollita JA. Lessons from
experimental models of hibernating myocardium.
Coron Artery Dis. 2001;12:371–380.
16. Alam SR, Stirrat C, Spath N, et al. Myocardial
inflammation, injury and infarction during on-
pump coronary artery bypass graft surgery.
J Cardiothorac Surg. 2017;12:115. https://doi.org/
10.1186/s13019-017-0681-6
17. Bax JJ, Poldermans D, Elhendy A, Boersma E,
Rahimtoola SH. Sensitivity, specificity, and pre-
dictive accuracies of various noninvasive tech-
niques for detecting hibernating myocardium. Curr
Probl Cardiol. 2001;26:147–186.
18. Garcia MJ, Kwong RY, Scherrer-Crosbie M,
et al. State of the art: imaging for myocardial
viability: a scientific statement from the American
Heart Association. Circ Cardiovasc Imaging.
2020;13:e000053. https://doi.org/10.1161/HCI.
0000000000000053
19. Panza JA, Dilsizian V, Laurienzo JM, Curiel RV,
Katsiyiannis PT. Relation between thallium uptake
and the contractile response to dobutamine: im-
plications regarding myocardial viability in pa-
tients with chronic coronary artery disease and left
ventricular dysfunction. Circulation. 1995;91:990–
998.
20. Di Carli MF, Asgarzadie F, Schelbert HR, et al.
Quantitative relation between myocardial viability
and improvement in heart failure symptoms after
revascularization in patients with ischemic car-
diomyopathy. Circulation. 1995;92:3436–3444.
21. Allman KC, Shaw LJ, Hachamovitch R,
Udelson JE. Myocardial viability testing and
impact of revascularization on prognosis in pa-
tients with coronary artery disease and left ven-
tricular dysfunction: a meta-analysis. J Am Coll
Cardiol. 2002;39:1151–1158.
22. Bourque JM, Hasselblad V, Velazquez EJ,
Borges-Neto S, O’Connor CM. Revascularization in
patients with coronary artery disease, left ven-
tricular dysfunction, and viability: a meta-analysis.
Am Heart J. 2003;146:621–627.
23. Camici PG, Prasad SK, Rimoldi OE. Stunning,
hibernation, and assessment of myocardial
viability. Circulation. 2008;117:103–114.
24. Inaba Y, Chen JA, Bergmann SR. Quantity of
viable myocardium required to improve survival
with revascularization in patients with ischemic
cardiomyopathy: a meta-analysis. J Nucl Cardiol.
2010;17:646–654.
25. Beanlands RSB, Nichol G, Huszti E, et al. F-18-
fluorodeoxyglucose positron emission
JACC VOL. 78, NO. 10, 2021
SEPTEMBER 7, 2021:1068–1077
CORRESPONDENCE: Dr Julio A.
tomography imaging-assisted management of
patients with severe left ventricular dysfunction
and suspected coronary disease: a randomized,
controlled trial (PARR-2). J Am Coll Cardiol.
2007;50:2002–2012.
26. Abraham A, Nichol G, Williams KA, et al. 18F-
FDG PET imaging of myocardial viability in an
experienced center with access to 18F-FDG and
integration with clinical management teams: the
Ottawa-FIVE substudy of the PARR 2 trial. J Nucl
Med. 2010;51:567–574.
27. Cleland JGF, Calvert M, Freemantle N, et al.
The Heart Failure Revascularisation Trial (HEART).
Eur J Heart Fail. 2011;13:227–233.
28. Velazquez EJ, Lee KL, Jones RH, et al. Coro-
nary-artery bypass surgery in patients with
ischemic cardiomyopathy. N Engl J Med. 2016;374:
1511–1520.
29. Velazquez EJ, Lee KL, O’Connor CM, et al. The
rationale and design of the Surgical Treatment for
ISCHEMIC Heart Failure (STICH) trial. J Thor Car-
diovasc Surg. 2007;134:1540–1547.
30. Bonow RO, Maurer G, Lee KL, et al. Myocardial
viability and survival in ischemic left ventricular
dysfunction. N Engl J Med. 2011;364:1617–1625.
31. Panza JA, Ellis AM, Al-Khalidi HR, et al.
Myocardial viability and long-term outcomes in
ischemic cardiomyopathy. N Engl J Med. 2019;381:
739–748.
32. Orlandini A, Castellana N, Pascual A, et al.
Myocardial viability for decision-making concern-
ing revascularization in patients with left ventric-
ular dysfunction and coronary artery disease: a
meta-analysis of non-randomized and random-
ized studies. Int J Cardiol. 2015;182:494–499.
33. Panza JA, Holly TA, Asch FM, et al. Inducible
myocardial Ischemia and outcomes in patients
with coronary artery disease and left ventricular
dysfunction. J Am Coll Cardiol. 2013;61:1860–
1870.
34. Panza JA, Velazquez EJ, She L, et al. Extent of
coronary and myocardial disease and benefit from
surgical revascularization in LV dysfunction. J Am
Coll Cardiol. 2014;64:553–561.
35. Bello D, Shah DJ, Farah GM, et al. Gadolinium
cardiovascular magnetic resonance predicts
reversible myocardial dysfunction and remodeling
in patients with heart failure undergoing beta-
blocker therapy. Circulation. 2003;108:1945–1953.
36. Cleland JG, Pennell DJ, Ray SG, et al.
Myocardial viability as a determinant of the ejec-
tion fraction response to carvedilol in patients
with heart failure (CHRISTMAS trial): randomized
controlled trial. Lancet. 2003;362:14–21.
JACC VOL. 78, NO. 10, 2021
SEPTEMBER 7, 2021:1068–1077
37. Samady H, Elefteriades JA, Abbott BG,
Mattera JA, McPherson CA, Wackers FJ. Failure to
improve left ventricular function after coronary
revascularization for ischemic cardiomyopathy is
not associated with worse outcome. Circulation.
1999;100:1298–1304.
38. Perry AS, Mann DL, Brown DL. Improvement
of ejection fraction and mortality in ischaemic
heart failure. Heart. Published online August 25,
2020. https://doi.org/10.1136/heartjnl-2020-
316975
39. Pellikka PA, MD, She L, Holly TA, et al. Vari-
ability in ejection fraction measured by echocar-
diography, gated single-photon emission
computed tomography, and cardiac magnetic
resonance in patients with coronary artery disease
and left ventricular dysfunction. JAMA Netw Open.
2018;1(4):e181456.
40. O’Meara E, Mielniczuk LM, Wells GA, et al.
Alternative Imaging Modalities in Ischemic
Heart Failure (AIMI-HF) IMAGE HF Project I-A:
Panza et al. 1077
Contemporary Use of Myocardial Viability Assessment
study protocol for a randomized controlled trial. 45. Stewart RA, Szalewska D, She L, et al. Exercise
Trials. 2013;14:218. capacity and mortality in patients with ischemic
left ventricular dysfunction randomized to coro-
41. Perera D, Clayton T, Petrie MC, et al.
nary artery bypass graft surgery or medical ther-
Percutaneous revascularization for ischemic
apy: an analysis from the STICH trial (Surgical
ventricular dysfunction: rationale and design
Treatment for Ischemic Heart Failure). J Am Coll
of the REVIVED-BCIS2 trial: percutaneous
Cardiol HF. 2014;2:335–343.
coronary intervention for ischemic cardio-
myopathy. J Am Coll Cardiol HF. 2018;6:517– 46. Doenst T, Haverich A, Serruys P, et al. PCI and
526. CABG for treating stable coronary artery disease.
J Am Coll Cardiol. 2019;73:964–975.
42. Carson P, Wertheimer J, Miller A, et al. The
STICH trial (Surgical Treatment for Ischemic
Heart Failure): mode-of-death results. J Am Coll KEY WORDS coronary artery disease, heart
Cardiol HF. 2013;1:400–408. failure, ischemic cardiomyopathy, left
43. Armstrong WF. “Hibernating” myocardium: ventricular function, myocardial viability
asleep or part dead? J Am Coll Cardiol. 1996;28(2):
530–535.
Go to http://www.acc.org/
44. Mark DB, Knight JD, Velazquez EJ, et al.
jacc-journals-cme to take
Quality-of-life outcomes with coronary artery
the CME/MOC/ECME quiz
bypass graft surgery in ischemic left ventricular
for this article.
dysfunction: a randomized trial. Ann Intern Med.
2014;161:392–399.
JOURNAL OF THE AMERICAN COLLEGE OF CARDIOLOGY VOL. 78, NO. 10, 2021

ª 2021 BY THE AMERICAN COLLEGE OF CARDIOLOGY FOUNDATION

PUBLISHED BY ELSEVIER

CARDIOVASCULAR MEDICINE AND SOCIETY

Cardiovascular Disease Fellowship

Interviews 2021
An Evolution in Process

Lisa J. Rose-Jones, MD,a Mustafa M. Ahmed, MD,b Benjamin H. Freed, MD,c Andrew M. Kates, MD,d
Meera Kondapaneni, MD,e Jonathan R. Salik, MD,f Victor Soukoulis, MD, PHD,g Helga Van Herle, MD, MS, MHA,h
Gaby Weissman, MDi

THE 2021 INTERVIEW SEASON:
CREATING CERTAINTY AMID UNCERTAINTY
The COVID-19 pandemic abruptly shifted traditional
norms in graduate medical education, not the least of
which was a pivot from in-person interviews to a
virtual-only platform for the 2020 interview cycle.
Cardiovascular (CV) program directors have been
contemplating the landscape of the
recruitment season in the midst of the current COVID
pandemic as well as the increased access to vaccina-
tions and changing guidelines. Many factors will help
shape the interview process, including but not
limited to institutional guidance, local and state reg-
ulations, and national society and regulatory board
recommendations. The Undergraduate Medical Edu-
cation to Graduate Medical Education Review Com-
mittee of the Coalition for Physician Accountability,
the Alliance for Academic Internal Medicine, and the
American College of Cardiology Program Directors
and Graduate Medical Educators Leadership Council
all just recently endorsed an all-virtual recruitment
for the upcoming season (1-3). In light of these rec-
ommendations, we believe it is important to consider
the approach to the coming recruitment season.
There are both opportunities and challenges to the
virtual recruitment that affect programs and appli-
cants. Although these virtual interviews may pose
challenges for fellowship programs and feel some-
what discordant in a postvaccine world, imple-
mentation of the formal recommendations in this
paper, based on our experiences and feedback from
stakeholders, will help ensure standardization and
upcoming optimize the experience for all participants.
Given the unique challenges that virtual interviews
can pose for both fellowship programs and applicants,
it is essential to build on experiences from last year
and share best practices to promote fairness and
achieve the desired goals of virtual recruitment for
the upcoming interview season. With that, we will
consider the viewpoints of both applicant and pro-
gram as we embark on this process together.
APPLICANT VIEWPOINT:
DEFINING THE IDEAL VIRTUAL EXPERIENCE
The virtual interview presents unique opportunities
and challenges for applicants to meet faculty, interact
with current fellows, and “tour” facilities. Outside the
standard faculty interviews, programs should

From the aDivision of Cardiovascular Medicine, University of North Carolina School of Medicine, Chapel Hill, North Carolina, USA;
b
Division of Cardiovascular Medicine, University of Florida, Gainesville, Florida, USA; cDepartment of Medicine, Northwestern
University Feinberg School of Medicine, Chicago, Illinois, USA; dDivision of Cardiovascular Medicine, Washington University
School of Medicine, St. Louis, Missouri, USA; eMetroHealth Medical Center Heart, Case Western Reserve University School of
Medicine, Cleveland, Ohio, USA; fCardiology Division, Massachusetts General Hospital, Boston, Massachusetts, USA; gDivision
of Cardiovascular Medicine, University of Virginia, Charlottesville, Virginia, USA; hKeck School of Medicine at the University of
Southern California, Los Angeles, California, USA; and the iMedstar Heart and Vascular Institute, Georgetown University,
Washington, DC, USA.
The authors attest they are in compliance with human studies committees and animal welfare regulations of the authors’
institutions and Food and Drug Administration guidelines, including patient consent where appropriate. For more information,
visit the Author Center.

Manuscript received June 1, 2021; revised manuscript received July 1, 2021, accepted July 2, 2021.

ISSN 0735-1097/$36.00 https://doi.org/10.1016/j.jacc.2021.07.005

JACC VOL. 78, NO. 10, 2021
SEPTEMBER 7, 2021:1078–1081
facilitate virtual interviews that help connect appli-
cants with prospective mentors in their fields of in-
terest and those with similar life experiences when
possible. Applicants should meet with current fellows
who can highlight a program’s academic environment
and sense of community. Conversations between
applicants and fellows, without faculty, in informa-
tional and informal sessions provides an important
perspective. Limiting the number of applicants and
fellows to a given virtual room can promote
interactivity.
We may take for granted all the facilities that ap-
plicants have traditionally seen in person. Programs
therefore should create virtual tours featuring com-
munity spaces, including work and call rooms, that
exist outside of standard patient care areas (4). Pro-
grams should also ensure that all videos meet insti-
tutional privacy policies. These types of video tours
are ideal for asynchronous viewing and can offer a
more streamlined interview day that offers less
“digital fatigue.”
Technological issues, such as interview discon-
nections, increase anxiety among applicants. A sim-
ple but well-developed backup plan presented to the
applicants early on the interview day can stem this
concern. The program coordinator should be readily
available and “online” for general troubleshooting
and directions. Providing interviewers easy access to
applicants’ telephone numbers to quickly reconnect
is important (5). Some programs also had fellows
assist the coordinator with the logistics of the day.
The computer screen itself poses unique barriers.
We should break down the “fourth wall” of virtual
interviews by looking directly into the camera when
speaking. However, this can be challenging, and
subtle behavioral cues from in-person interviews may
be difficult to replicate in a virtual format. These
limitations should be kept in mind when evaluating
applicants, and discussions with interviewers about
new potential technological inequities and biases
should be addressed early on (6). Finally, neutral,
professional backgrounds can minimize distractions
and ensure that the evaluative component of an
interview is not influenced by external factors.
THE PROGRAM PERSPECTIVE:
UNIQUE CHALLENGES
There are many challenges in the virtual environment
that affect all programs. However, programs with
Rose-Jones et al. 1079
Cardiovascular Disease Fellowship Interviews 2021
fewer trainees, as well as those in less popu- ABBREVIATIONS
lous or more remote geographic regions, AND ACRONYMS
share unique challenges. For these programs
CV = cardiovascular
in particular, virtual interviews may present
new obstacles but also provide opportunities.
In previous years and traditional application cy-
cles, applicants may have bypassed small programs
because of several misperceptions. These include the
belief that such programs were unable to provide the
full breadth of training required to achieve career
goals or that programs in smaller communities or
remote locations offer fewer amenities outside of the
training environment. However, last year’s virtual
platform provided at least anecdotal evidence that
virtual interviewing can increase initial interest
because of reduced travel and time constraints (4).
Moreover, there was also a perceived reduction in the
not uncommon practice of cancellations.
Applicants may not have the familiarity with indi-
vidual programs in the virtual-only experience. As
such, the need to emphasize the unique characteris-
tics and strengths of programs is paramount. It is
particularly important to highlight the strength of the
training environment and community. The thought-
ful use of preinterview recorded materials and videos
that advertise the benefits of the training program
and the advantages of the community is key. These
and similar materials should be shared across pro-
grams within an institution and augmented by civic
and local productions aimed at recruitment of young
professionals.
Many programs proportionately increased the
number of interviews last year because of the fear of
unmatched positions in the face of initial concerns
that the ease of virtual interviews would lead to
“interview hoarding.” Interestingly, reports from
program director forums suggest similar or better
match outcomes compared to pre-COVID recruitment
years. Additionally, many smaller programs believed
that they had a greater number of underrepresented
minorities and women apply. Previous limited expe-
rience with virtual interviews have indeed confirmed
that reduced financial travel burden can promote
applicant diversity (6). Further cardiology-specific
data need to be collected to understand if the cur-
rent recruitment environment contributed to more
diversity, equity, and inclusion. This will be particu-
larly important as we look toward the future in which
virtual interviews may continue to play an important
role in recruitment.
1080 Rose-Jones et al.
Cardiovascular Disease Fellowship Interviews 2021
T A B L E 1 Key Recommendations for Program Directors to Prepare for the 2021 Virtual
Interview Season
Before Interview Day Interview Day
Prepare for an all-virtual interview season to Create a technical troubleshooting plan
promote equity and ensure consistency detailed to both interviewers and applicants
Anticipate a larger volume of applicants Ensure time between applicants and current
fellows, preferably without faculty
Use websites/social media to highlight the Facilitate interviews that connect applicants
training environment and community with prospective mentors and faculty with
similar life experiences
Develop a virtual facility tour that includes Accommodate applicants from various time
work and call rooms zones in scheduling
Consider mock interviews with faculty, Look directly into the camera and consider a
fellows, and program administration to neutral, professional background in a well-
test software lit room
PUTTING IT ALL TOGETHER: LESSONS LEARNED
Fellowship programs across the country learned
valuable lessons in recruitment during the COVID
pandemic that will serve all stakeholders well for the
future. These included improvements in online
presence, technical components to the online inter-
view, and the structure of the interview day (Table 1).
Online presence, including websites and social
media that highlight distinct attributes, is essential
even before application season. It provides a prime
opportunity to feature cultural aspects of an institu-
tion that would otherwise receive limited exposure in
the virtual environment. Programs should engage
their fellows in this process whenever possible.
Once applications become available for review,
programs should anticipate that there may be more
received than customary (7). Although this may
enhance applicant pool diversity, review committees
may require more time to accommodate this increase.
Fortunately, programs will have approximately 3
additional weeks in the 2021 season to evaluate
applications.
Technical aspects of the interview day should be
well defined in advance of the interview day. Many
different formats are available; the type of software
used is not as important as ensuring participants’
ability to use it. Mock interviews with faculty, the
program director and coordinator, and fellows are
particularly helpful to better understand all platform
features (3). Having one “host” responsible for mov-
ing all participants is important, as is structuring the
interview day to minimize applicant navigating.
Furthermore, clarity on interview duration, particu-
larly when and how each interview will end, can help
participants better transition the conversation to its
conclusion.
JACC VOL. 78, NO. 10, 2021
SEPTEMBER 7, 2021:1078–1081
Both synchronous and asynchronous interview
options can work well. Fellowship programs should
consider which type will best suit their specific needs.
Although synchronous options provide real-time in-
teractions between interviewers and applicants,
asynchronous models can involve pre-established
interview questions. Applicants then answer these
questions and submit via recorded video. Unique to
the virtual interview, accommodating different time
zones is an important consideration for synchronous
interviews.
It is also important to be mindful of both the total
number and the length of each interview during the
interview day. To avoid “digital fatigue” for both
applicants and faculty, the ideal day should last no
more than 4 to 6 hours and allow adequate “off-
camera” time between interviews for restroom and
stretch breaks. Videos detailing unique aspects of the
program, medical center, and community are prefer-
ably viewed via links that can be accessed on the
applicants’ own time.
LOOKING TO THE FUTURE
In its recent statement, the Undergraduate Medical
Education to Graduate Medical Education Review
Committee recommended “to ensure equity and
fairness, there should be ongoing study of the impact
and benefits of virtual interviewing as a permanent
means of interviewing for residency.” The virtual
platform provides certain benefits to both applicants
and CV programs by enabling greater access to all
applicants and should be considered in future
recruitment cycles. A major hurdle will be addressing
the potential cognitive bias in evaluating applicants
who travel to interviews and those who do not. This
platform may also result in a potential separation of
applicants based on socioeconomic status, which
could undo efforts that promote diversity in Cardiol-
ogy. Recognizing the desire of applicants to visit in-
stitutions and communities before committing to
years of training is also essential. The conversation
for a hybrid interview platform must begin now. The
CV fellowship community has a real opportunity to
shape this process so that it provides advantages to
both applicants and programs.
As this process continues, we are still faced with an
upcoming recruitment season that will be virtual. The
medical education community, applicants and pro-
grams alike, have learned many lessons and best
practices in just 1 year. There is no doubt that
JACC VOL. 78, NO. 10, 2021 Rose-Jones et al. 1081
SEPTEMBER 7, 2021:1078–1081 Cardiovascular Disease Fellowship Interviews 2021

building on those experiences will only help to find

the best “match” for all. ADDRESS FOR CORRESPONDENCE: Dr Lisa J. Rose-
Jones, Division of Cardiovascular Medicine, Univer-
sity of North Carolina School of Medicine, 160 Dental
FUNDING SUPPORT AND AUTHOR DISCLOSURES
Circle, CB 7075, Chapel Hill, North Carolina 27599,
The authors have reported that they have no relationships relevant to USA. E-mail: Lisa_rose-jones@med.unc.edu. Twitter:
the contents of this paper to disclose. @LisaRoseJones1.

REFERENCES

1. Initial summary report and preliminary recom-
mendations of the Undergraduate Medical Edu-
cation to Graduate Medical Education Review
Committee (UGRC). Accessed May 24, 2021. https://
physicianaccountability.org/wp-content/uploads/
2021/04/UGRC-Initial-Summary-Report-and-
Preliminary-Recommendations-1.pdf
2. American College of Cardiology. Accessed
June 27, 2021. https://www.acc.org/-/media/
Non-Clinical/Files-PDFs-Excel-MS-Word-etc/
Membership/Sections-Councils/Cardiovascular-
training/ACC-PDGME-Council-2021-Interview-Cycle-
Statement.pdf
3. Alliance for Academic Interna Medicine. Fellow-
ship application season 2021-2022. Accessed June
27, 2021. https://www.im.org/resources/ume-
gme-program-resources/resources-fellowship-
application
4. Huppert LA, Hsiao EC, Cho KC, et al. Virtual
interviews at graduate medical education
training programs: determining evidence-based
best practices. Acad Med. Published online
December 8, 2020. https://doi.org/10.1097/ACM.
0000000000003868
5. Association for American Medical Colleges. Vir-
tual interviews: tips for program directors.
Accessed May 24, 2021. https://www.aamc.org/
system/files/2020-05/Virtual_Interview_Tips_
for_Program_Directors_05142020.pdf
6. Marbin J, Hutchinson YV, Schaeffer S. Avoiding
the virtual pitfall: identifying and mitigating biases
in graduate medical education videoconference
interviews. Acad Med. Published January 12, 2021.
https://doi.org/10.1097/ACM.0000000000003914
7. Huppert LA, Santhosh L, Babik JM. Trends in US
internal medicine residency and fellowship appli-
cations during the COVID-19 pandemic vs previous
years. JAMA Netw Open. 2021;4(4):e218199.
https://doi.org/10.1001/jamanetworkopen.2021.
8199
KEY WORDS cardiology fellowship,
program directors, virtual interviews
JOURNAL OF THE AMERICAN COLLEGE OF CARDIOLOGY
ª 2021 BY THE AMERICAN COLLEGE OF CARDIOLOGY FOUNDATION
PUBLISHED BY ELSEVIER
Letters
Sudden Death
Thinking Beyond the Heart
In their recent publication, Ricceri et al (1)
contribute important findings to the epidemiologic
and pathophysiologic understanding of sudden
death. They report that 69% of resuscitated sudden
cardiac arrest (SCA) have an arrhythmic cause with
SCA survivors being more likely than SCA non-
survivors to have an arrhythmic cause (92% vs
57%). Their work extends from the contempora-
neous POST SCD (Postmortem Systematic Investi-
gation of Sudden Cardiac Death) study (2) in which
arrhythmic causes were found in only 56% of World
Health Organization (WHO)–defined sudden cardiac
death (SCD). Combined, these results suggest that
noncardiac etiologies are predictors of increasingly
poorer outcomes along the chain of survival.
We have similarly found that cardiac disease is not
common (14.8%) among all-cause sudden death vic-
tims, although cardiovascular risk factors and
related comorbidities are prevalent (3). Our 2-year
population-based study included out-of-hospital
sudden death in the socioeconomically and racially
diverse population of Wake County, North Carolina,
not attributed to trauma; suicide; overdose; or natural,
expected deaths.
We posit that sudden death should be viewed as a
syndrome in which diverse pathophysiologic pro-
cesses result in the same outcome. Differentiating
SCD from sudden death relies on imprecise defini-
tions. As Ricceri et al (1) note, WHO SCD criteria
provided only a 56% positive predictive value for
identifying actual arrhythmic death. We found that
20% of all-cause sudden deaths in our study met
criteria for WHO-defined SCD. However, current
resuscitation strategies rely on cardiopulmonary
resuscitation and defibrillation, which are unlikely to
VOL. 78, NO. 10, 2021
ISSN 0735-1097/$36.00
address the noncardiac etiologies that cause the
remaining nonarrhythmic sudden deaths. Because
the majority of those with SCA die and those who do
not survive are less likely to have cardiac causes (4),
reducing sudden death mortality will require novel
preventive strategies that address the noncardiac
causes of SCA nonsurvivors and, more broadly, sud-
den death victims.
*Susan K. Keen, MD, MSCR
Ross J. Simpson, Jr., MD, PhD
*University of North Carolina
590 Manning Drive
Chapel Hill, North Carolina 27599, USA
E-mail: suskkeen@gmail.com
Twitter: @SusanKeenMD
https://doi.org/10.1016/j.jacc.2021.06.046
The SUDDEN (Sudden Unexpected Death in North Carolina) project is funded by
individual private donations, the Heart and Vascular Division of the University
of North Carolina at Chapel Hill, and the McAllister Heart Institute. The project
described was supported by the National Center for Advancing Translational
Sciences, National Institutes of Health, through grant award number
UL1TR002489. Dr Simpson has served as a consultant for Amgen, Merck, Pfizer,
CERobs, and Innovative Science Solutions; and is currently the University of
North Carolina site principal investigator for an Amgen-sponsored observational
research project. Dr Keen has reported that she has no relationships relevant to
the contents of this paper to disclose.
The authors attest they are in compliance with human studies committees and
animal welfare regulations of the authors’ institutions and Food and Drug
Administration guidelines, including patient consent where appropriate. For
more information, visit the Author Center.
REFERENCES
1. Ricceri S, Salazar JW, Vu AA, Vittinghoff E, Moffatt E, Tseng ZH. Factors
predisposing to survival after resuscitation for sudden cardiac arrest. J Am Coll
Cardiol. 2021;77:2353–2362.
2. Tseng ZH, Olgin JE, Vittinghoff E, et al. Prospective countywide surveil-
lance and autopsy characterization of sudden cardiac death: POST SCD study.
Circulation. 2018;137:2689–2700.
3. Joodi G, Maradey JA, Bogle B, et al. Coronary artery disease and athero-
sclerotic risk factors in a population-based study of sudden death. J Gen Intern
Med. 2020;35:531–537.
4. Sasson C, Rogers MA, Dahl J, Kellermann AL. Predictors of survival from
out-of-hospital cardiac arrest: a systematic review and meta-analysis. Circ
Cardiovasc Qual Outcomes. 2010;3:63–81.
JOURNAL OF THE AMERICAN COLLEGE OF CARDIOLOGY
ª 2021 BY THE AMERICAN COLLEGE OF CARDIOLOGY FOUNDATION
PUBLISHED BY ELSEVIER
REPLY: Sudden Death
Thinking Beyond the Heart
We thank the correspondents for their interest in our
study. To study, treat, and ultimately prevent the
conditions of out-of-hospital cardiac arrest (OHCA),
sudden cardiac arrest (SCA), sudden cardiac death
(SCD), and sudden arrhythmic death (SAD) (the
only type of sudden death addressed by advanced
cardiac life support and implantable cardioverter-
defibrillators) effectively, one must start with their
definitions. Importantly, as we showed in this paper
(1), these conditions are not equivalent. OHCA re-
quires only a first responder’s primary impression of
“cardiac arrest” and does not specify suddenness
(eg, <1 hour of symptom duration before arrest); thus,
necessarily, SCAs are only a subset of all OHCAs. In
our paper, we used World Health Organization (WHO)
criteria to identify the 133 resuscitated SCAs among
the 734 emergency medical services–defined OHCAs
occurring countywide (18%) that met WHO criteria for
suddenness. Thus, it is not surprising that only 20%
of the “sudden” deaths defined purely by EMS OHCA
criteria in the paper referenced in the correspon-
dents’ letter (2) were actually sudden by WHO
criteria. In short, OHCA s SCA.
Another common conflation is that SCA is the same
as SCD because, had a victim not been resuscitated,
he or she would have died. Yet, as the correspondents
note, in our study, the further along the chain of
survival, the more likely a case was arrhythmic:
among patients with SCAs who died out of hospital,
that is, WHO-defined SCDs, approximately one-half
had arrhythmic cause compared to 92% of SCAs sur-
viving to discharge. Thus, we demonstrated that an
SCA victim is more likely to be resuscitated when they
are arrhythmic because our current resuscitation
strategies do not address the array of nonarrhythmic
causes underlying cases that did not survive. In short,
SCA s SCD.
The next misconception is that traditionally
defined SCDs are necessarily cardiac or arrhythmic,
yet EMS or epidemiologic criteria, such as those of the
WHO, cannot determine cause of death, for which
autopsy is the gold standard. In POST SCD (POst-
mortem Systematic InvesTigation of Sudden Cardiac
Death) study, we demonstrated that only one-half
(55.8%) of all WHO-defined SCDs in San Francisco
County had arrhythmic cause upon comprehensive
postmortem investigation, which included vitreous
chemistries and toxicology (3). In short, convention-
ally defined SCD should be considered presumed SCD.
VOL. 78, NO. 10, 2021
ISSN 0735-1097/$36.00
The autopsy rate was not reported in the corre-
spondents’ study but, outside of POST SCD (97%), is
typically approximately 10% for out-of-hospital
deaths (4); thus, the true prevalence of cardiac
disease among all presumed SCDs in Wake County
is unknown. Recognizing the difficulty in obtaining
autopsy for out-of-hospital deaths and leveraging
the autopsy-defined cases in POST SCD, we recently
described refinements to the WHO definition and
prediction models to assist other investigators in
better specifying true SADs among cohorts of pre-
sumed SCDs (5). In short, presumed SCD s SAD.
A multipronged approach addressing the unmet
needs in OHCA and sudden death research is needed
to enable: 1) improved SCD and SAD identification,
prediction, and prevention; 2) refinement of current
resuscitative strategies for cases with cardiac and
arrhythmic etiologies; and 3) new strategies for rapid
identification and resuscitation of cases with un-
derlying noncardiac or nonarrhythmic causes of ar-
rest. This approach must start with rigor in its
definitions.
*Zian H. Tseng, MD, MAS
Santo Ricceri, MD
James W. Salazar, MD, MAS
*Cardiac Electrophysiology Section
University of California-San Francisco
500 Parnassus Avenue, Box 1354
San Francisco, California 94143-1354, USA
E-mail: Zian.Tseng@ucsf.edu
https://doi.org/10.1016/j.jacc.2021.07.007
Dr Tseng has received grants from National Institutes of Health (NIH)/National
Heart, Lung, and Blood Institute and Centers for Disease Control outside the
submitted work. Dr Salazar has received grant support from NIH/National Heart,
Lung, and Blood Institute (award number R38HL143581). Dr Ricceri has reported
that he has no relationships relevant to the contents of this paper to disclose.
The contents of this article are solely the responsibility of the authors and do not
necessarily represent the official views of the NIH.
The authors attest they are in compliance with human studies committees and
animal welfare regulations of the authors’ institutions and Food and Drug
Administration guidelines, including patient consent where appropriate. For
more information, visit the Author Center.
REFERENCES
1. Ricceri S, Salazar JW, Vu Andrew A, Vittinghoff E, Moffatt E, Tseng ZH. Factors
predisposing to survival after resuscitation for sudden cardiac arrest. J Am Coll
Cardiol. 2021;77(19):2353–2362. https://doi.org/10.1016/j.jacc.2021.03.299
2. Keen S, Simpson RJ Jr. Sudden death: thinking beyond the heart. J Am Coll
Cardiol. 2021;78(10):e61.
3. Tseng ZH, Olgin JE, Vittinghoff E, et al. Prospective countywide surveillance
and autopsy characterization of sudden cardiac death. Circulation. 2018;137(25):
2689–2700. https://doi.org/10.1161/CIRCULATIONAHA.117.033427
4. Shojania KG, Burton EC. The vanishing nonforensic autopsy. N Engl J Med.
2008;358:873–875.
5. Tseng ZH, Salazar JW, Olgin JE, et al. Refining the World Health Organi-
zation definition. Circ Arrhythm Electrophysiol. 2019;12(7):e007171. https://
doi.org/10.1161/CIRCEP.119.007171