Circulation
RESEARCH LETTER
Evaluation of 2 Existing Diagnostic Scores for
Heart Failure With Preserved Ejection Fraction
Against a Comprehensively Phenotyped Cohort
H
eart failure with preserved ejection fraction (HFpEF) is increasingly common Timothy W. Churchill ,
and presents challenges for accurate diagnosis. Two recent publications MD*
proposed HFpEF diagnostic tools: the HFA-PEFF algorithm1 and the H2FPEF Shawn X. Li, MD, MBA*
score. Efforts to validate these scores in external populations have been limited by
2
Lisa Curreri, RDCS
lack of a gold standard phenotypic definition of HFpEF, and no previous validation Emily K. Zern, MD
has compared these scores against invasive hemodynamic criteria. We sought to Emily S. Lau, MD
evaluate these diagnostic approaches against a hemodynamic definition of HFpEF Elizabeth E. Liu , BS
in individuals undergoing cardiopulmonary exercise testing. Robyn Farrell, BS
We studied patients with chronic dyspnea and preserved left ventricular ejec- Mark W. Shoenike, BS
tion fraction (≥50%) who underwent clinically indicated cardiopulmonary exercise John Sbarbaro, BA
testing with invasive hemodynamic monitoring3 between 2006 and 2017. After Rajeev Malhotra, MD
applying exclusion criteria,3 there were 156 individuals with data required for both Matthew Nayor , MD,
scores and available echocardiogram within 1 year of cardiopulmonary exercise MPH
Downloaded from http://ahajournals.org by on November 13, 2024
testing date. All participants underwent comprehensive testing including resting Carsten Tschöpe, MD
and exercise hemodynamic measurements, cardiopulmonary exercise testing gas Rudolf A. de Boer , MD,
exchange, and NT-proBNP (N-terminal pro–B-type natriuretic peptide). We defined PhD
our reference standard (invasive HFpEF [HFpEFinv]) using the following criteria: el- Gregory D. Lewis , MD†
evated pulmonary capillary wedge pressure (PCWP) at rest (≥15 mm Hg) or during Jennifer E. Ho , MD†
exercise (≥15 mm Hg) coupled with an abnormally steep change in PCWP relative
to cardiac output (ΔPCWP/Δcardiac output >2.0 mm Hg·L−1·min−1), based on the
established physiologic and prognostic significance of this threshold from our pre-
vious work.3,4 Echocardiograms were reviewed per society guidelines.5 Risk strati-
fication by each score was compared against HFpEFinv. HFA-PEFF analysis included
evaluation of noninvasive score alone along with the algorithm’s invasive criteria.
The study was approved by Massachusetts General Hospital’s institutional review
board, and all participants provided informed consent. Study data will be shared
on reasonable request to the corresponding author.
A total of 156 participants (59±16 years of age, 67% women) were included.
Across both tools, individuals with higher scores were older and had more comor-
bidities including diabetes, hypertension, and hyperlipemia (P<0.01 for all). For the
HFA-PEFF algorithm, 44 (28%) participants had rule-out scores (0 to 1 point; Figure *Drs Churchill and Li contributed
[A]). Within this low-risk group, 11 (25%) were found to have HFpEFinv. Conversely, equally as co–first authors.
22 (14%) had rule-in scores (≥5 points), of whom 4 (18%) did not have HFpEFinv, †Drs Lewis and Ho contributed equally
bringing the number misclassified by the noninvasive portion to 15/66 (23%). Af- as co–senior authors.
ter addition of the algorithm’s invasive criteria, there was agreement in classifica- Key Words: diagnostic techniques,
tion when compared with HFpEFinv in 126/156 (81%), with total misclassification cardiovascular ◼ echocardiography
◼ exercise test ◼ heart failure ◼ heart
of 30 patients (19%). failure, diastolic ◼ hemodynamics
For the noninvasive score, the negative predictive value of scores of 0 to 1 was
© 2021 American Heart Association, Inc.
75% (95% CI, 62%–88%); positive predictive value of scores of 5 to 6 was 82%
(66%–98%). The area under the receiver operating characteristic curve was 0.73 https://www.ahajournals.org/journal/circ
Circulation. 2021;143:289–291. DOI: 10.1161/CIRCULATIONAHA.120.050757 January 19, 2021 289
Churchill et al Evaluation of 2 HFpEF Diagnostic Scores
CORRESPONDENCE
A B
Figure. Classification of the total sample.
Classification of the total sample (n=156) according to the HFA-PEFF algorithm (A) and the H2FPEF score (B), along with corresponding rates of invasively defined
heart failure with preserved ejection fraction (HFpEFinv). Global longitudinal strain was not performed and thus was excluded from HFA-PEFF noninvasive score; left
atrial linear dimensions were used in place of left atrial volumes. Diastolic stress test was not performed. For the H2FPEF score, predicted HFpEF prevalence was
Downloaded from http://ahajournals.org by on November 13, 2024
derived from the original score publication.2 In C, important metrics of exercise performance are shown stratified by HFA-PEFF and H2FPEF scores. *P<0.05 in pair-
wise comparison of low vs high groups using 2-tailed t tests or Wilcoxon rank-sum test, as appropriate for the data distribution, when Panalysis of variance or PKruskal-Wallis is
significant across groups. LVEF indicates left ventricular ejection fraction; and PCWP, pulmonary capillary wedge pressure.
(0.65–0.81). Sensitivity and specificity of the full algo- overall and track closely with exercise performance, but
rithm were 72% (61%–81%) and 91% (82%–96%); our findings highlight potential misclassification among
positive and negative predictive values were 89% individuals with low scores. These findings raise a note
(81%–97%) and 75% (66%–84%). of caution about the use of these clinical approaches to
The Figure (B) compares predicted versus actual rule out HFpEF in patients with dyspnea, which appears
HFpEF prevalence across H2FPEF scores. Among the to be the most salient limitation of both tools. This un-
low score group (0 to 1 points; n=50), 14 individuals derdiagnosis at low scores underscores the difficulty of
(28%) had confirmed HFpEFinv. Among 95 individu- using resting measures to define a phenotype that is
als with intermediate and 11 with high H2FPEF scores, fundamentally exertional in nature and highlights the
60% and 91% met HFpEFinv criteria, respectively. Us- potential of exercise to serve as a diagnostic probe to
ing a cut point of ≥5, sensitivity and specificity were unmask abnormal physiology.
31% (95% CI, 21% to 42%) and 92% (83% to 97%) Some argue for a fixed cutoff for PCWP during ex-
and the positive and negative predictive values were ercise, but we have chosen to use a criterion indexed
81% (67% to 95%) and 55% (46% to 64%). The area to cardiac output, given the incremental predictive abil-
under the receiver operating characteristic curve was ity of this definition beyond PCWP4 for clinical events
0.74 (0.66–0.81). and the strong physiologic rationale for incorporating
Exercise capacity was markedly impaired across the measures of flow into pressure assessment, given the
entire cohort (peak oxygen consumption 16.1±5.7 mL/ obligatory rise in filling pressures with increasing car-
kg/min) and was worse in the high score groups (Figure diac output.4 We acknowledge several limitations, in-
[C]). Higher score groups also had more impaired chro- cluding a patient cohort that was relatively young com-
notropic response and higher filling pressures. pared with many patients with HFpEF and the absence
We assessed diagnostic performance of 2 con- of a noninvasive diastolic stress test and 2 important
temporary HFpEF diagnostic tools against an invasive echocardiographic measures (global longitudinal strain
hemodynamic definition. Both tools performed well and left atrial volumes).
290 January 19, 2021 Circulation. 2021;143:289–291. DOI: 10.1161/CIRCULATIONAHA.120.050757
Churchill et al Evaluation of 2 HFpEF Diagnostic Scores
We demonstrate good overall performance of the Disclosures
HFA-PEFF algorithm and the H2FPEF score, but highlight
CORRESPONDENCE
Dr Ho has received research supplies from EcoNugenics Inc and research grants
potential misclassification, particularly at low scores. from Bayer and Gilead Sciences. The other authors report no conflicts.
Further validation of these tools in varied populations
and clinical settings is needed.
REFERENCES
1. Pieske B, Tschöpe C, de Boer RA, Fraser AG, Anker SD, Donal E, Edelmann F,
ARTICLE INFORMATION Fu M, Guazzi M, Lam CSP, et al. How to diagnose heart failure with pre-
served ejection fraction: the HFA-PEFF diagnostic algorithm: a consensus
The podcast and transcript are available as a Data Supplement at https://www.
ahajournals.org/doi/suppl/10.1161/CIRCULATIONAHA.120.050757. recommendation from the Heart Failure Association (HFA) of the Euro-
pean Society of Cardiology (ESC). Eur Heart J. 2019;40:3297–3317. doi:
Correspondence 10.1093/eurheartj/ehz641
2. Reddy YNV, Carter RE, Obokata M, Redfield MM, Borlaug BA. A sim-
Jennifer E. Ho, MD, Massachusetts General Hospital, 185 Cambridge St, CPZN
ple, evidence-based approach to help guide diagnosis of heart failure
#3192, Boston, MA 02114. Email jho1@mgh.harvard.edu
with preserved ejection fraction. Circulation. 2018;138:861–870. doi:
10.1161/CIRCULATIONAHA.118.034646
Affiliations 3. Ho JE, Zern EK, Wooster L, Bailey CS, Cunningham T, Eisman AS,
Cardiology Division (T.W.C., S.X.L., E.K.Z., E.S.L., E.E.L., R.F., M.W.S., J.S., R.M., Hardin KM, Zampierollo GA, Jarolim P, Pappagianopoulos PP, et al. Differ-
M.N., G.D.L., J.E.H.) and Division of Pulmonary and Critical Care Medicine ential clinical profiles, exercise responses, and outcomes associated with
(G.D.L.), Department of Medicine, and Cardiovascular Research Center (L.C., existing HFpEF definitions. Circulation. 2019;140:353–365. doi: 10.1161/
R.M., J.E.H.), Massachusetts General Hospital, Boston. Berlin Institute of Health CIRCULATIONAHA.118.039136
Center for Regenerative Therapies (C.T.) and Department of Cardiology (C.T.), 4. Eisman AS, Shah RV, Dhakal BP, Pappagianopoulos PP, Wooster L, Bailey C,
Charité, University Medicine Berlin, Campus Virchow Clinic, Germany. German Cunningham TF, Hardin KM, Baggish AL, Ho JE, et al. Pulmonary cap-
Center for Cardiovascular Research, Partner Site Berlin, Germany (C.T.). Depart- illary wedge pressure patterns during exercise predict exercise capac-
ment of Cardiology, University Medical Centre Groningen, University of Gronin-
ity and incident heart failure. Circ Heart Fail. 2018;11:e004750. doi:
gen, The Netherlands (R.A.d.B.).
10.1161/CIRCHEARTFAILURE.117.004750
5. Lang RM, Badano LP, Mor-Avi V, Afilalo J, Armstrong A, Ernande L,
Sources of Funding Flachskampf FA, Foster E, Goldstein SA, Kuznetsova T, et al. Recommen-
This work was supported by grants from the National Institutes of Health (R01- dations for cardiac chamber quantification by echocardiography in adults:
HL134893, R01-HL140224, and K24 HL153669 to Dr Ho, R01-HL142809 to Dr an update from the American Society of Echocardiography and the Eu-
Malhotra, K23-HL138260 to Dr Nayor, and R01-HL131029 to Dr Lewis) and the ropean Association of Cardiovascular Imaging. J Am Soc Echocardiogr.
American Heart Association (15GPSGC24800006 to Dr Lewis). 2015;28:1–39.e14. doi: 10.1016/j.echo.2014.10.003
Downloaded from http://ahajournals.org by on November 13, 2024
Circulation. 2021;143:289–291. DOI: 10.1161/CIRCULATIONAHA.120.050757 January 19, 2021 291
