Abstract
Aims
H2FPEF and HFA‐PEFF scores have demonstrated prognostic value in heart failure (HF) with preserved ejection fraction. This study aimed to explore the value of the H2FPEF and HFA‐PEFF scores for HF risk stratification in patients with hypertrophic cardiomyopathy (HCM).
Methods and results
In this cohort study, 1068 HCM patients were included. Then the H2FPEF and HFA‐PEFF scores were calculated to categorize patients into low, intermediate, and high score groups. The primary endpoint was a composite of the first HF hospitalization and all‐cause death. 594 (55.6%) patients were classified discordantly. After a follow‐up period of 3.1 ± 2.1 years, 85 (8.0%) patients were admitted for HF for the first time, and 62 (5.8%) patients died. Rates of first HF hospitalization and all‐cause death per 1000 person‐years for the low, intermediate, and high H2FPEF score groups were 25.0 (95% confidence interval [CI]: 14.5–35.4), 52.0 (95% CI: 41.6–62.3), and 148.1 (95% CI: 77.7–218.5), respectively. For the low‐intermediate and high HFA‐PEFF score groups, rates were 19.3 (95% CI: 11.6–27.0) and 69.3 (95% CI: 56.4–82.1), respectively. Intermediate H2FPEF score (hazard ratio [HR]: 1.820, 95% CI: 1.135–2.919; P = 0.013), high H2FPEF score (HR: 3.464, 95% CI: 1.774–6.765; P < 0.001), and high HFA‐PEFF score (HR: 2.414, 95% CI: 1.501–3.882; P < 0.001) were each independently associated with an increased risk of the primary endpoint. Intermediate‐high H2FPEF score demonstrated an equal risk for the primary endpoint compared to the high HFA‐PEFF score (HR: 0.826, 95% CI: 0.636–1.072; P > 0.05). Obesity (HR: 1.958, 95% CI: 1.140–3.363; P = 0.015), atrial fibrillation (HR: 1.686, 95% CI: 1.071–2.654; P = 0.024), pulmonary hypertension (HR: 1.613, 95% CI: 1.032–2.521; P = 0.036) of the H2FPEF score, and the morphological major criterion (HR: 1.601, 95% CI: 1.084–2.364; P = 0.018) and functional major criterion (HR: 2.340, 95% CI: 1.442–3.797; P < 0.001) of the HFA‐PEFF score were independent predictors of the primary endpoint. A new algorithm was constructed using the independent predictors from both scores, with the functional major criterion weighted as 2 points and the others as 1 point. The H2FPEF score, HFA‐PEFF score, and the new algorithm demonstrated C‐indices of 0.594, 0.651, and 0.681, respectively.
Conclusions
There is discordance in the classification of patients with HCM using the H2FPEF and HFA‐PEFF scores. Both scores demonstrated prognostic value in risk stratification for HF hospitalization and all‐cause death in HCM patients. Future studies should develop and validate a new algorithm integrating both scores.
Keywords: Hypertrophic cardiomyopathy, Heart failure, H2FPEF, HFA‐PEFF, Risk stratification
1. Introduction
Hypertrophic cardiomyopathy (HCM) is the most common inherited cardiac disease, with a prevalence of 1:200–1:500 in the general population, characterized by myocardial hypertrophy, disarray, and interstitial fibrosis, 1 , 2 typically with preserved left ventricular (LV) systolic function. 3 Current studies have shown that patients with HCM are at an increased risk for heart failure (HF), particularly heart failure with preserved ejection fraction (HFpEF). 4 Appropriate risk stratification is crucial for the treatment of HCM patients and for reducing mortality. However, the risk stratification for HF is relatively underdeveloped compared to that for sudden cardiac death and cardiac arrhythmias. 5
Recently, H2FPEF 6 and HFA‐PEFF 7 scores have been developed to aid in the diagnosis of HFpEF, and both have been validated against invasive gold standards. 8 H2FPEF score is a weighted composite score derived from several demographic, clinical, and echocardiographic parameters. 6 The HFA‐PEFF score, developed based on an expert consensus recommendation from the Heart Failure Association of the European Society of Cardiology, is composed of three domains that utilize echocardiographic data and laboratory tests. 7 Additionally, both the H2FPEF and HFA‐PEFF scores have been confirmed to have prognostic value in patients with unexplained dyspnoea, 9 in patients with HFpEF caused by cardiac amyloidosis, 10 and in patients undergoing transcatheter aortic valve implantation. 11 , 12 In community patients with unexplained dyspnoea, both scores were found to be equivalent in terms of prognosis. 9 However, differences between the two scores were reported in HFpEF associated with cardiac amyloidosis, where the HFA‐PEFF score demonstrated superiority in diagnosis and prognosis. 10 This difference likely arises from the distinct pathophysiology of HFpEF in the setting of cardiac amyloidosis. 10
A recent study by Laenens et al. showed that the increased H2FPEF score was independently associated with worse HF outcomes in patients with HCM. 5 However, the prognosis value of HFA‐PEFF score in patients with HCM in their study is lacking due to the lack of B‐type natriuretic peptide data, which is required for the calculation of HFA‐PEFF score. 5 Ethnical differences in both scores have been reported. Black individuals tend to have higher H2FPEF scores, while White individuals tend to have higher HFA‐PEFF scores. 9 This variation presents challenges in prognostic assessment, as the clinical significance of a given score may differ between ethnic groups. Furthermore, both scores have limited validation data in Asian populations, making their application in these populations uncertain. Hence, this study aimed to explore the value of the H2FPEF and HFA‐PEFF scores for HF risk stratification in Chinese patients with HCM.
2. Methods
2.1. Study population and data collection
This study protocol was approved by our institutional review board (TJ‐IRB202410062). In this cohort study, patients aged 18 years or older who were diagnosed with HCM according to the guidelines, 3 with a maximal LV wall thickness ≥15 mm in the absence of another cardiac, systemic, or metabolic condition possibly causing the LV hypertrophy, were included from January 2014 to December 2023. Given that both scores were developed for the diagnosis of HFpEF 9 , 10 , 13 and that septal reduction therapy can significantly influence the prognosis of patients with HCM, 3 patients with reduced LV ejection fraction (LVEF) (<50%), a previous history of myectomy or alcohol septal ablation, or missing values necessary for calculating the H2FPEF and HFA‐PEFF scores were excluded. A flow diagram of the current analysis is presented in Figure S1 .
All echocardiographic examination were performed and analysed strictly according to guidelines, 14 and all echocardiographic records underwent quality control review. Standard two‐dimensional and Doppler echocardiography was performed using GE Vivid E9 or Vivid E95 ultrasound machine (GE Healthcare, Horten, Norway) with an M5Sc transducer. Dimensions of the left atrial (LA) and LV were measured from the parasternal long‐axis view according to the guidelines, 15 including LA end‐systolic dimension, LV end‐diastolic dimension, end‐diastolic interventricular septal thickness, and end‐diastolic LV posterior wall thickness. The maximal wall thickness was measured at the basal, mid, and apical levels of LV from the parasternal short‐axis view at end‐diastole. LVEF was calculated using the biplane Simpson's method. 15 Mitral inflow peak early (E) and late diastolic (A) velocities, systolic peak tricuspid regurgitation velocity, and the LV outflow tract systolic peak gradient at rest and during provocation (Valsalva manoeuvre, exercise, or amyl nitrite) were measured as recommended. 16 Tissue Doppler peak early diastolic velocities (e′) at septal and lateral mitral annulus were measured and averaged to calculate E/e′ ratio. 16 Mitral regurgitation (MR) was assessed following the steps recommended by the guideline. 17 , 18 Relative wall thickness (RWT) was calculated as twice the LV posterior wall thickness divided by the LV end‐diastolic dimension. 7 LV mass was calculated using the Devereux formula 15 and indexed to body surface area (BSA) to obtain LV mass index (LVMI). Demographic and clinical data were collected from electronic medical records. Height, weight, and N‐terminal pro‐brain natriuretic peptide (NT‐proBNP) levels measured within 1 month before or after the echocardiography examination were also recorded.
2.2. H2FPEF and HFA‐PEFF score calculation
The H2FPEF score is calculated as follows: (1) history of atrial fibrillation (AF), 3 points; (2) body mass index (BMI) > 30 kg/m2, 2 points; (3) arterial hypertension, defined as treatment with 2 or more antihypertensive medications, 1 point; (4) echocardiographic pulmonary artery systolic pressure >35 mmHg, 1 point; (5) age >60 years, 1 point; and (6) echocardiographic E/e′ ratio >9, 1 point. The H2FPEF score for each patient was calculated as the sum of points for all six variables (range from 0 to 9 points). In the present study, arterial hypertension was defined based on the diagnosis of primary hypertension according to patient history. Accordingly, patients were divided into three groups: low H2FPEF score group (0–1), intermediate H2FPEF score group (2–5), and high H2FPEF score group (6–9). 6
The HFA‐PEFF score consists of functional, morphological, and biomarker domains. 7 Within each domain, a major criterion scores 2 points, and a minor criterion scores 1 point. Each domain can contribute up to 2 points if any major criterion from this domain is positive or 1 point if no major but any minor criterion is positive. The HFA‐PEFF score is calculated as the sum of the domains (range from 0 to 6 points). In the present study, we used LA end‐systolic dimension instead of LA volume index to calculate the HFA‐PEFF score. An LA end‐systolic dimension >47/43 mm (male/female) was adapted to meet the major criteria of the morphological domain according to the guidelines of the American Society of Echocardiography. 19 All parameters used to calculate HFA‐PEFF score in the present study included septal e′, E/e′ ratio, peak tricuspid regurgitation velocity, LA end‐systolic dimension, LVMI, RWT, LV wall thickness, and NT‐proBNP. Due to the LV wall thickness >15 mm in all patients with HCM, the score for the morphological domain is at least 1 point. Patients were then divided into three groups: low HFA‐PEFF score group (0–1), intermediate HFA‐PEFF score group (2–4), and high HFA‐PEFF score group (5–6) according to the clinical categories of HFA‐PEFF score. 7
2.3. Follow‐up and clinical outcomes
The time of the first visit for patients diagnosed with HCM who had complete echocardiographic records was considered as the starting point for follow‐up. Follow‐up continued until all‐cause death or the study's end in August 2024. Patients included were followed up using electronic medical records or by phone calls to the patients or their relatives. The primary endpoint was a composite of the first hospitalization for HF, defined as deterioration of New York Heart Association (NYHA) functional class to class III or IV requiring hospitalization, 20 and all‐cause death. The secondary endpoints were first hospitalization for HF or all‐cause death. Of the initial 1068 patients, 72 (6.7%) were lost to follow‐up.
2.4. Statistical analyses
Continuous data were expressed as the mean ± standard deviation, and data with a skewed distribution were expressed as the median (interquartile range). Categorical data were presented as frequencies with percentages. Student's t‐test, Mann–Whitney U‐test, chi‐square test, and analysis of variance were used to examine differences between the patient groups, as appropriate. Post hoc analysis was performed using the Bonferroni correction.
The Kaplan–Meier method was employed to determine survival curves for the primary and secondary endpoints. Survival curves were compared using log‐rank test. Patients lost to follow‐up were censored. Cox proportional hazards regression analysis was performed to identify independent predictors of the primary endpoint. Then, variables statistically significant in the univariate Cox proportional hazards analysis were selected for multivariate Cox proportional hazards analysis. Cox proportional hazards regression was employed to compare the risk between H2FPEF and HFA‐PEFF scores, with high HFA‐PEFF serving as the reference group. Sensitivity analysis was performed in three subgroups, excluding (1) patients with LV outflow tract obstruction (LVOTO) or severe MR; (2) patients with coronary heart disease; and (3) patients obtaining 2 points in the morphological domain of the HFA‐PEFF score based solely on LA end‐systolic dimension, respectively. The coefficients of multivariate Cox proportional hazards model were scaled to one another to produce integer values of relatively similar impact to create a new algorithm. 21 The effectiveness of the different models was compared using Harrell's C‐index.
R version 4.3.3 (R Project for Statistical Computing, Vienna, Austria) was used for statistical analyses. Statistical significance was defined as a two‐sided P value <0.05.
3. Results
3.1. Baseline characteristics
A total of 2132 patients were enrolled initially. After exclusion, 1068 patients were included for analysis (Figure S1 ). The number of patients with low, intermediate, and high scores were 246 (23.0%), 765 (71.6%), and 57 (5.4%), respectively, for H2FPEF scores and 29 (2.7%), 400 (37.5%), and 639 (59.8%), respectively, for HFA‐PEFF scores. Patients with low (0–1) and intermediate (2–4) HFA‐PEFF scores were pooled, given few patients in the 0–1 category (Figure S1 ). Consequently, patients were divided into two groups: low‐intermediate HFA‐PEFF score group (0–4) and high HFA‐PEFF score group (5–6).
Baseline clinical characteristics and echocardiographic and laboratory findings of study population and groups categorized by H2FPEF and HFA‐PEFF scores are presented in Table 1 . Apart from the clinical and echocardiographic variables used to calculate H2FPEF, which were significantly different among low, intermediate and high H2FPEF score groups, the high and intermediate H2FPEF score groups exhibited higher NT‐proBNP levels and greater percentages of female, coronary heart disease, and NYHA III–IV (Table 1 ). In echocardiographic data, higher H2FPEF score groups showed larger LA end‐systolic dimension and LV end‐diastolic dimension, greater LV posterior wall thickness, LV mass, mitral E velocity, and lower LVEF (Table 1 ). Similarly, significant differences existed in variables used to calculate HFA‐PEFF score between low‐intermediate and high HFA‐PEFF score groups. Additionally, the high HFA‐PEFF score group was older and had greater percentages of female, AF, NYHA III–IV, and lower BMI (Table 1 ). In echocardiographic data, the high HFA‐PEFF group had greater percentages of severe MR and LVOTO and slightly lower LVEF (64.1 ± 6.5 vs. 65.0 ± 5.7, P = 0.019) (Table 1 ).
Table 1.
Baseline clinical characteristics and echocardiographic and laboratory findings and outcome of the study population stratified by the H2FPEF or HFA‐PEFF scores
| Overall (n = 1068) | H2FPEF score 0–1 (n = 246) | H2FPEF score 2–5 (n = 765) | H2FPEF score 6–9 (n = 57) | HFA‐PEFF score 0–4 (n = 429) | HFA‐PEFF score 5–6 (n = 639) | |
|---|---|---|---|---|---|---|
| Clinical characteristics | ||||||
| Age, years | 57.5 ± 12.8 | 48.9 ± 10.2 | 59.4 ± 12.2* | 69.2 ± 9.8* , # | 54.8 ± 12.5 | 59.3 ± 12.6* |
| Age > 60 years | 457 (42.8) | 14 (5.7) | 392 (51.2)* | 51 (89.5)* , # | 140 (32.6) | 317 (49.6)* |
| Female | 361 (33.8) | 63 (25.6) | 280 (36.6)* | 18 (31.6) | 96 (22.4) | 265 (41.5)* |
| BSA, m2 | 1.77 ± 0.19 | 1.78 ± 0.17 | 1.76 ± 0.19 | 1.78 ± 0.23 | 1.82 ± 0.18 | 1.73 ± 0.18* |
| BMI, kg/m2 | 25.3 ± 3.8 | 24.4 ± 2.7 | 25.4 ± 3.9* | 26.4 ± 6.0* | 25.9 ± 3.4 | 24.8 ± 4.0* |
| BMI > 30 kg/m 2 | 86 (8.1) | 0 (0.0) | 78 (10.2)* | 8 (14.0)* | 44 (10.3) | 42 (6.6)* |
| Hypertension | 675 (63.2) | 32 (13.0) | 595 (77.8)* | 48 (84.2)* | 271 (63.2) | 404 (63.2) |
| Diabetes | 236 (22.1) | 35 (14.2) | 187 (24.4)* | 14 (24.6) | 98 (22.8) | 138 (21.6) |
| Coronary heart disease | 484 (45.3) | 73 (29.7) | 381 (49.8)* | 30 (52.6)* | 185 (43.1) | 299 (46.8) |
| Atrial fibrillation | 114 (10.7) | 0 (0.0) | 57 (7.5)* | 57 (100.0)* , # | 29 (6.8) | 85 (12.3)* |
| NYHA III–IV | 79 (7.4) | 10 (4.1) | 51 (6.7) | 18 (31.6)* , # | 8 (1.9) | 71 (11.1)* |
| Echocardiographic characteristics | ||||||
| LA dimension, mm | 40.2 ± 7.1 | 37.8 ± 6.5 | 40.5 ± 6.7* | 46.8 ± 9.3* , # | 38.2 ± 5.7 | 41.6 ± 7.5* |
| LA dimension > 46/43 mm (male/female) | 198 (18.5) | 24 (9.8) | 144 (18.8)* | 30 (52.6)* , # | 33 (7.7) | 165 (25.8)* |
| LV dimension, mm | 45.5 ± 4.8 | 44.7 ± 4.7 | 45.7 ± 4.8* | 47.0 ± 5.2* | 45.6 ± 4.3 | 45.5 ± 5.2 |
| IVS, mm | 15.8 ± 5.0 | 16.1 ± 5.7 | 15.6 ± 4.6 | 16.5 ± 6.4 | 13.6 ± 3.0 | 17.3 ± 5.4* |
| LV posterior wall thickness, mm | 11.7 ± 2.6 | 11.2 ± 2.5 | 11.8 ± 2.5* | 12.2 ± 3.7* | 10.9 ± 1.7 | 12.2 ± 2.9* |
| Maximal wall thickness, mm | 17.7 ± 4.1 | 18.1 ± 4.7 | 17.5 ± 3.7 | 18.6 ± 6.2 | 16.0 ± 1.9 | 18.9 ± 4.7* |
| RWT | 0.52 ± 0.14 | 0.51 ± 0.14 | 0.52 ± 0.13 | 0.53 ± 0.18 | 0.48 ± 0.08 | 0.54 ± 0.16* |
| RWT > 0.42 | 859 (80.4) | 187 (76.0) | 628 (82.1) | 44 (77.2) | 327 (76.2) | 532 (83.3)* |
| LV mass, g | 254 ± 101 | 245 ± 100 | 254 ± 94 | 291 ± 168* , # | 212 ± 62 | 282 ± 112* |
| LVMI, g/m2 | 144 ± 54 | 139 ± 57 | 144 ± 50 | 162 ± 80* , # | 116 ± 31 | 162 ± 58* |
| LVMI ≥ 149/122 g/m 2 (male/female) | 479 (44.9) | 96 (39.0) | 353 (46.1) | 30 (52.6) | 64 (14.9) | 415 (64.9)* |
| LVEF, % | 64.4 ± 6.2 | 65.2 ± 5.8 | 64.4 ± 6.3 | 61.7 ± 5.9* , # | 65.0 ± 5.7 | 64.1 ± 6.5* |
| Mitral E velocity, cm/s | 72.8 ± 24.0 | 67.9 ± 21.8 | 73.0 ± 23.9* | 90.3 ± 26.1* , # | 69.4 ± 19.1 | 75 ± 26.5* |
| Septal e′ velocity, cm/s | 5.0 ± 1.8 | 5.8 ± 2.0 | 4.7 ± 1.7* | 5.4 ± 2.0 # | 5.9 ± 2.0 | 4.4 ± 1.5* |
| Septal e′ velocity < 7 cm/s | 873 (81.7) | 162 (65.9) | 667 (87.2)* | 44 (77.2) | 277 (64.6) | 596 (93.3)* |
| E/e′ ratio | 16.2 ± 7.2 | 13.1 ± 6.3 | 17.0 ± 7.2* | 18.4 ± 6.6* | 12.7 ± 4.5 | 18.4 ± 7.7* |
| E/e′ ratio > 9 | 947 (88.7) | 165 (67.1) | 727 (95.0)* | 55 (96.5)* | 336 (78.3) | 611 (95.6)* |
| E/e′ ratio ≥ 15 | 505 (47.3) | 70 (28.5) | 396 (51.8)* | 39 (68.4)* | 113 (26.3) | 392 (61.3)* |
| Peak tricuspid regurgitation velocity, m/s | 2.4 ± 0.5 | 2.2 ± 0.4 | 2.4 ± 0.5* | 2.7 ± 0.6* , # | 2.2 ± 0.4 | 2.5 ± 0.6* |
| Peak tricuspid regurgitation velocity > 2.8 m/s | 118 (11.0) | 0 (0.0) | 96 (12.5)* | 22 (38.6)* , # | 9 (2.1) | 109 (17.1)* |
| Severe MR | 57 (5.3) | 16 (6.5) | 38 (5.0) | 3 (5.3) | 8 (1.9) | 49 (7.7)* |
| LVOTO | 343 (32.1) | 75 (30.5) | 253 (33.1) | 15 (26.3) | 92 (21.5) | 251 (39.3)* |
| NT‐proBNP, pg/mL | 408 (124–1335) | 278 (88–1058) | 382 (129–1231) | 1996 (904–4341)* , # | 90 (46–175) | 950 (427–2325)* |
| NT‐proBNP > 220/660 pg/mL (SR/AF) | 656 (61.4) | 135 (54.9) | 474 (62.0) | 47 (82.5)* , # | 54 (12.6) | 602 (94.2)* |
| Events | ||||||
| Primary endpoint | 136 (12.7) | 22 (8.9) | 97 (12.7) | 17 (29.8) | 24 (5.6) | 112 (17.5) |
| Event rate (95% CI) per 1000 person‐years | 47.5 (39.5–55.5) | 25.0 (14.5–35.4) | 52.0 (41.6–62.3)* | 148.1 (77.7–218.5)* , # | 19.3 (11.6–27.0) | 69.3 (56.4–82.1)* |
| First heart failure hospitalization | 85 (8.0) | 16 (6.5) | 58 (7.6) | 11 (19.3) | 13 (3.0) | 72 (11.3) |
| Event rate (95% CI) per 1000 person‐years | 29.5 (23.2–35.8) | 18.2 (9.3–27.1) | 31.0 (23.0–38.9)* | 85.8 (35.1–136.5)* , # | 10.4 (4.8–16.1) | 44.0 (33.8–54.2)* |
| All‐cause death | 62 (5.8) | 8 (3.3) | 46 (6.0) | 8 (14.0) | 14 (3.3) | 48 (7.5) |
| Event rate (95% CI) per 1000 person‐years | 21.0 (15.8–26.2) | 8.9 (2.7–15.1) | 23.9 (17.0–30.8)* | 62.0 (19.0–104.9)* , # | 11.1 (5.3–16.9) | 28.4 (20.3–36.4)* |
Values are mean ± SD, n (%), median (Q1–Q3), or event rate (95% CI). Components for the calculation of H2FPEF and HFA‐PEFF scores were bolded.
AF, atrial fibrillation; BMI, body mass index; BSA, body surface area; CI, confidence interval; LA, left atrial; LV, left ventricular; LVEF, left ventricular ejection fraction; LVMI, left ventricular mass index; LVOTO, ventricular outflow tract obstruction; MR, mitral regurgitation; NT‐proBNP, N‐terminal pro‐brain natriuretic peptide; NYHA, New York Heart Association; RWT, relative wall thickness; SR, sinus rhythm.
P < 0.05 vs. H2FPEF score 0–1 or HFA‐PEFF 0–4.
P < 0.05 vs. H2FPEF score 2–5.
3.2. Discordance and concordance between H2FPEF and HFA‐PEFF scores
Figure 1 presents the histograms of the H2FPEF and HFA‐PEFF scores and mosaic plot of groups for both scores. For the H2FPEF score, low and intermediate scores were more frequent (Figure 1 A ), while for the HFA‐PEFF score, intermediate and high scores were more common (Figure 1 B ). The variables used to calculate the H2FPEF and HFA‐PEFF scores for the study population are summarized in Table S1 . For the H2FPEF score, there were a small number of patients with AF (3 points) or obesity (2 points), with numbers of 114 (10.7%) and 86 (8.1%), respectively (Tables 1 and S1 ), resulting in an overall low‐intermediate distribution. In contrast, for the HFA‐PEFF score, a relatively large number of patients met the major criteria, with 905 (84.7%), 494 (46.3%), and 656 (61.4%) meeting the functional, morphological, and biomarker major criteria, respectively (Tables 1 and S1 ), suggesting an overall intermediate‐high distribution.
Figure 1.
Histogram and Mosaic plot of H2FPEF and HFA‐PEFF scores. (A) Histograms of the H2FPEF scores. Large portion in low and intermediate H2FPEF score (0–5). (B) Histograms of the HFA‐PEFF scores. Large portion in high HFA‐PEFF score (5–6). (C) Mosaic plot of H2FPEF and HFA‐PEFF scores. Fifty‐two patients scored high on both H2FPEF and HFA‐PEFF, while 423 patients scored low and intermediate on both. Discordance was observed in 594 patients: 6 with high H2FPEF score and low and intermediate HFA‐PEFF score and 588 with low and intermediate H2FPEF score and high HFA‐PEFF score.
Discordance and concordance between H2FPEF and HFA‐PEFF scores are summarized in Table S2 . A total of 51 patients (4.8%) had both high H2FPEF and high HFA‐PEFF scores, and 423 patients (39.6%) had low and intermediate scores for both H2FPEF and HFA‐PEFF, which together accounted for 44.4% of the study population (Table S2 and Figure 1 C ). The discordance occurred in 6 patients with high H2FPEF score and low and intermediate HFA‐PEFF score and in 588 patients with low and intermediate H2FPEF score and high HFA‐PEFF score, accounting for 55.6% of the study population in total (Table S2 and Figure 1 C ). Patients with high H2FPEF score and low and intermediate HFA‐PEFF score were older and had greater BSA and BMI, higher percentages of AF, and lower interventricular septal thickness, LVMI, and NT‐proBNP levels compared to those with low and intermediate H2FPEF score and high HFA‐PEFF score (Table S3 ). This indicates that the discordance primarily arises from differences in the variables used to calculate each score.
In the subgroup of patients with AF (n = 114), concordance was significantly higher, with 51 patients (44.7%) having both high H2FPEF and HFA‐PEFF scores and 23 patients (20.2%) having low or intermediate scores for both, accounting for 64.9% of this subgroup (Table S2 and Figure S2 A). Discordance occurred in six patients with a high H2FPEF score and a low or intermediate HFA‐PEFF score and in 34 patients with a low or intermediate H2FPEF score and a high HFA‐PEFF score, representing 35.1% of the subgroup (Table S2 and Figure S2 A).
In the subgroup of patients with obesity (BMI > 30 kg/m2, n = 86), slightly but not significantly higher concordance was observed. Four patients (4.7%) had both high H2FPEF and HFA‐PEFF scores, and 40 patients (46.5%) had low or intermediate scores for both, accounting for 51.2% of this subgroup (Table S2 and Figure S2 B). Discordance occurred in four patients with a high H2FPEF score and a low or intermediate HFA‐PEFF score and in 38 patients with a low or intermediate H2FPEF score and a high HFA‐PEFF score, representing 48.8% of the subgroup (Table S2 and Figure S2 B).
In the subgroup of patients with elevated NT‐proBNP levels (n = 656), 47 patients (7.2%) had both high H2FPEF and high HFA‐PEFF scores, and 54 patients (8.2%) had low or intermediate scores for both H2FPEF and HFA‐PEFF, accounting for 15.4% of the patients classified as consistent (Table S2 and Figure S2 C). The discordance occurred in 555 patients with low or intermediate H2FPEF scores and high HFA‐PEFF scores, accounting for 84.6% of the subgroup Table S2 and Figure S2 C). These findings suggest that the concordance between H2FPEF and HFA‐PEFF scores is influenced by patient characteristics, particularly the presence of AF and elevated NT‐proBNP levels.
3.3. Follow‐up outcomes
After a follow‐up period of 3.1 ± 2.1 years, 85 (8.0%) patients were admitted for HF for the first time, and 62 (5.8%) patients died. The outcomes are presented in Table 1 . Event rates for first HF hospitalization and all‐cause death were higher in patients with higher H2FPEF scores (Table 1 ) or high HFA‐PEFF score (Table 1 ) compared to those with lower H2FPEF or low HFA‐PEFF scores. Patients in the intermediate and high H2FPEF score groups demonstrated lower event‐free survival compared to the low H2FPEF score group for the primary endpoint (Figure 2 A ), first HF hospitalization (Figure 2 B ), and all‐cause death (Figure 2 C ). Similar trends were observed in the HFA‐PEFF score groups, where event‐free survival was lower in the high HFA‐PEFF score group compared to the low HFA‐PEFF score group for the primary endpoint (Figure 2 D ), first HF hospitalization (Figure 2 E ), and all‐cause death (Figure 2 F ).
Figure 2.
Kaplan–Meier curves of HFA‐PEFF and H2FPEF score groups. Primary endpoint is defined as first HF hospitalization and all‐cause death.
In subgroups stratified by gender, male patients in the intermediate and high H2FPEF score groups, as well as those in the high HFA‐PEFF score group, experienced worse outcomes for the primary endpoint (Figure S3A,C ). While no significant difference was observed between the low and intermediate H2FPEF score groups for female patients, those with high H2FPEF and high HFA‐PEFF scores experienced worse outcomes for the primary endpoint (Figure S3B,D ). These findings suggest that both scores are generally robust across genders, although some differences in the magnitude of effect are apparent.
Sensitivity analysis excluding patients with LVOTO or severe MR (n = 709) confirmed that the intermediate and high H2FPEF score groups, and the high HFA‐PEFF score group, had worse outcomes for the primary endpoint (Figure 3 A,D ), first HF hospitalization (Figure S4A,D ), and all‐cause death (Figure S5A,D ). In patients without coronary heart disease (n = 584), sensitivity analysis yielded similar findings for the primary endpoint (Figure 3 B,E ), first HF hospitalization (Figure S4B,E ), and all‐cause death (Figure S5B,E ). Sensitivity analysis excluding patients obtaining 2 points in the morphological domain of the HFA‐PEFF score based solely on LA end‐systolic dimension (n = 988) showed similar results for the primary endpoint (Figure 3 C,F ), first HF hospitalization (Figure S4C,F ), and all‐cause death (Figure S5C,F ).
Figure 3.
Sensitivity analysis of Kaplan–Meier curves for the primary endpoint in subgroups. Sensitivity analysis was performed in three subgroups. (A,D) Excluding patients with left ventricular outflow tract obstruction or severe mitral regurgitation. (B,E) Excluding patients with coronary artery disease. (C,F) Excluding patients obtaining 2 points in the morphological domain of the HFA‐PEFF score based solely on left atrial end‐systolic dimension.
3.4. Prognostic value of H2FPEF and HFA‐PEFF scores
The results of univariable and multivariable Cox regression analyses for the primary endpoint are presented in Table 2 . In univariable Cox regression analysis, female sex, NYHA III–IV, LVEF, maximal wall thickness, LVOTO, and both scores were associated with the primary endpoint. All components of H2FPEF and HFA‐PEFF scores were associated with the primary endpoint, except for hypertension and age >60. In multivariable Cox regression analysis, NYHA III–IV (hazard ratio [HR] 1.724, 95% confidence interval [CI] 1.081–2.749; P = 0.022), maximal wall thickness (HR 1.042, 95% CI 1.002–1.084; P = 0.040), intermediate H2FPEF score (HR 1.820, 95% CI 1.135–2.919; P = 0.013), high H2FPEF score (HR 3.464, 95% CI 1.774–6.765; P < 0.001), and high HFA‐PEFF score (HR 2.414, 95% CI 1.501–3.882; P < 0.001) remained independently associated with the primary endpoint (Table 2 ). Comparison of risk for the intermediate and high H2FPEF scores vs. the high HFA‐PEFF score is presented in Table 3 . The intermediate H2FPEF score demonstrated lower risks for the primary endpoint (HR 0.747, 95% CI 0.569–0.980; P < 0.05) and first HF hospitalization (HR 0.705, 95% CI 0.499–0.997; P < 0.05) compared to the high HFA‐PEFF score. However, the risk for all‐cause death was comparable between the intermediate H2FPEF score and the high HFA‐PEFF score (Table 3 ). The high H2FPEF score showed significantly higher risks for the primary and secondary endpoints compared to the high HFA‐PEFF score (Table 3 ). Overall, the risks associated with the intermediate‐high H2FPEF score for the first HF hospitalization and all‐cause death were equivalent to those of the high HFA‐PEFF score (Table 3 ). Additionally, the high H2FPEF score presented higher risks for the first HF hospitalization and all‐cause death than high HFA‐PEFF score (Table 3 ).
Table 2.
Univariable and multivariable Cox regression for the association of variables with the primary endpoint
| Variable | Univariable analysis | Multivariable analysis | ||
|---|---|---|---|---|
| Hazard ratio (95% CI) | P value | Hazard ratio (95% CI) | P value | |
| Female sex | 1.459 (1.036–2.055) | 0.031 | 1.234 (0.866–1.738) | 0.238 |
| Diabetes | 1.075 (0.717–1.613) | 0.726 | ||
| Coronary heart disease | 0.897 (0.636–1.265) | 0.535 | ||
| NYHA III–IV | 2.864 (1.855–4.422) | <0.001 | 1.724 (1.081–2.749) | 0.022 |
| LVEF | 0.963 (0.937–0.990) | 0.008 | 0.980 (0.954–1.008) | 0.154 |
| LV end‐diastolic dimension | 1.001 (0.965–1.038) | 0.974 | ||
| Maximal wall thickness | 1.076 (1.042–1.112) | <0.001 | 1.042 (1.002–1.084) | 0.040 |
| LVOTO | 1.579 (1.118–2.231) | 0.010 | 1.099 (0.748–1.617) | 0.630 |
| Severe MR | 0.930 | |||
| HFA‐PEFF components | 0.964 (0.425–2.188) | |||
| Function (0–2) | 1.744 (1.152–2.642) | 0.009 | ||
| Morphology (0–2) | 2.599 (1.821–3.709) | <0.001 | ||
| Biomarker (0–2) | 1.885 (1.450–2.450) | <0.001 | ||
| HFA‐PEFF | <0.001 | |||
| 0–4 | Ref. | Ref. | ||
| 5–6 | <0.001 | 2.414 (1.501–3.882) | <0.001 | |
| H2FPEF components | 3.591 (2.310–5.582) | |||
| BMI > 30 | 1.820 (1.079–3.072) | 0.025 | ||
| Hypertension | 1.417 (0.992–2.024) | 0.055 | ||
| Atrial fibrillation | 2.154 (1.384–3.351) | <0.001 | ||
| Pulmonary artery systolic pressure >35 mmHg | 2.471 (1.620–3.769) | <0.001 | ||
| Age >60 years | 0.878 (0.615–1.255) | 0.476 | ||
| E/e′ > 9 | 2.575 (1.203–5.512) | 0.015 | ||
| H2FPEF | <0.001 | |||
| 0–1 | Ref. | Ref. | ||
| 2–5 | 2.375 (1.452–3.887) | <0.001 | 1.820 (1.135–2.919) | 0.013 |
| 6–9 | 6.650 (3.444–12.840) | <0.001 | 3.464 (1.774–6.765) | <0.001 |
Abbreviations as in Table 1 .
Table 3.
Comparison of risk of the primary and secondary endpoints for H2FPEF and HFA‐PEFF scores
| High HFA‐PEFF score (n = 639) | Intermediate H2FPEF score (n = 765) | High H2FPEF score (n = 57) | Intermediate‐high H2FPEF score (n = 822) | |
|---|---|---|---|---|
| Primary endpoint | ||||
| Event rate (95% CI) per 1000 person‐years | 69.3 (56.4–82.1) | 52.0 (41.6–62.3)* | 148.1 (77.7–218.5)* | 57.5 (47.0–68.1) |
| Hazard ratio (95% CI) | Ref. | 0.747 (0.569–0.980)* | 2.105 (1.263–3.507)* | 0.826 (0.636–1.072) |
| First heart failure hospitalization | ||||
| Event rate (95% CI) per 1000 person‐years | 44.0 (33.8–54.2) | 31.0 (23.0–38.9)* | 85.8 (35.1–136.5)* | 34.5 (26.3–42.6) |
| Hazard ratio (95% CI) | Ref. | 0.705 (0.499–0.997)* | 1.954 (1.036–3.748)* | 0.785 (0.564–1.093) |
| All‐cause death | ||||
| Event rate (95% CI) per 1000 person‐years | 28.4 (20.3–36.4) | 23.9 (17.0–30.8) | 62.0 (19.0–104.9)* | 26.3 (19.3–33.3) |
| Hazard ratio (95% CI) | Ref. | 0.834 (0.557–1.250) | 2.122 (1.004–4.488)* | 0.917 (0.621–1.353) |
P < 0.05 vs. high HFA‐PEFF score group; abbreviations as in Table 1 .
Furthermore, components of both scores with P < 0.05 in univariate Cox regression analyses, including obesity, hypertension, AF, pulmonary hypertension, and E/e′ > 9 from the H2FPEF score, and all major criteria of the HFA‐PEFF score (Table 2 ), were entered into a multivariate Cox regression model to select variables for developing a new algorithm integrating the two scores. The results of the multivariate Cox regression analysis are summarized in Table 4 . Obesity, AF, pulmonary hypertension, the morphological major criterion, and the biomarker major criterion remained independently associated with the primary endpoint (Table 4 ). A new algorithm integrating these two scores was developed, with elevated NT‐proBNP levels scoring 2 points and other independent predictors scoring 1 point (Table 4 ). The new algorithm was superior to H2FPEF and HFA‐PEFF scores, with a C‐index of 0.681 (Figure 4 ).
Table 4.
Multivariable Cox analysis of the components of the H2FPEF and HFA‐PEFF scores and assigned points for the new algorithm
| Hazard ratio (95% CI) | Coefficients | Assigned points | P value | |
|---|---|---|---|---|
| BMI > 30 kg/m2 | 1.958 (1.140–3.363) | 0.67 | 1 | 0.015 |
| Hypertension | 1.401 (0.975–2.014) | 0.34 | 0.069 | |
| Atrial fibrillation | 1.686 (1.071–2.654) | 0.52 | 1 | 0.024 |
| Pulmonary artery systolic pressure > 35 mmHg | 1.613 (1.032–2.521) | 0.48 | 1 | 0.036 |
| E/e′ > 9 | 1.554 (0.680–3.551) | 0.44 | 0.296 | |
| Septal e′ < 7 cm/s or E/e′ ≥ 15 or TR velocity > 2.8 m/s | 1.355 (0.704–2.610) | 0.30 | 0.364 | |
| LA > 46/43 mm (m/w) or LVMI ≥ 149/122 g/m2 (m/w) and RWT > 0.42 | 1.601 (1.084–2.364) | 0.47 | 1 | 0.018 |
| NT‐proBNP > 220/660 pg/mL (SR/AF) | 2.340 (1.442–3.797) | 0.85 | 2 | <0.001 |
Abbreviations as in Table 1 .
Figure 4.
Effectiveness of H2FPEF and HFA‐PEFF scores and the new integrating algorithm for the primary endpoint.
In patients without LVOTO or severe MR, sensitivity analysis confirmed that the intermediate H2FPEF score (HR 1.937, 95% CI 1.051–3.571; P = 0.034), the high H2FPEF score (HR 4.401, 95% CI 1.926–10.058; P < 0.001), and the high HFA‐PEFF score (HR 2.813, 95% CI 1.570–5.040; P < 0.001) were independently associated with the primary endpoint (Table S4 ). Additionally, sensitivity analysis confirmed the independent association of the intermediate H2FPEF score (HR 1.967, 95% CI 1.114–3.474; P = 0.020), the high H2FPEF score (HR 7.615, 95% CI 3.263–17.769; P < 0.001), and the high HFA‐PEFF score (HR 1.889, 95% CI 1.055–3.380; P = 0.032) in patients without coronary heart disease (Table S5 ). In patients excluding those obtaining 2 points in the morphological domain of the HFA‐PEFF score based solely on LA end‐systolic dimension, sensitivity analysis also demonstrated that the intermediate H2FPEF score (HR 1.979, 95% CI 1.207–3.246; P = 0.007), the high H2FPEF score (HR 3.084, 95% CI 1.415–6.722; P = 0.005), and the high HFA‐PEFF score (HR 2.323, 95% CI 1.439–3.764; P < 0.001) were independently associated with the primary endpoint (Table S6 ).
4. Discussion
In the present study, we assessed the value of the H2FPEF and HFA‐PEFF scores for HF risk stratification in patients with HCM. The main findings of this study are that, in HCM patients, (1) although significant discordance exists in classifying HCM patients using the H2FPEF and HFA‐PEFF scores, both higher H2FPEF scores and high HFA‐PEFF score were associated with an increased risk of HF hospitalization and all‐cause death; (2) H2FPEF and HFA‐PEFF score values were independent predictors of HF hospitalization and all‐cause death; (3) intermediate‐high H2FPEF and high HFA‐PEFF scores demonstrated comparable risks for HF hospitalization and all‐cause death; and (4) integrating components of both scores that are independent predictors for the primary endpoint with different weights could enhance the performance of risk stratification.
With the use of risk stratification and implantable defibrillators in HCM patients, sudden deaths have been reduced measurably. 22 HF, instead, has gained increasing recognition in HCM patients. 1 HF with preserved systolic function is one of the most significant complications of HCM and can occasionally progress to systolic dysfunction, referred to as end‐stage HCM. 23 Multiple factors contribute to HF in HCM patients, including myocardial hypertrophy, myocardial stiffness and fibrosis, myocardial ischemia, LVOTO, abnormal vasomotor response, and vascular remodelling. 24 More challenges are present in the diagnosis of HFpEF in HCM patients, where exertional dyspnoea exists in 50% of patients, often without clinical signs of congestion. 25 Risk stratification for HF is crucial for the management of HCM patients, especially in relatively older patients. 26 Echocardiography is the primary imaging modality for screening, diagnosing, prognostic stratification, and follow‐up in HCM patients. 3 Previous studies have found that several echocardiographic parameters have prognostic value for incident HF or HF progression, including LVOTO ≥ 30 mmHg at rest or ≥50 mmHg (provoked), global longitudinal strain ≤16%, systolic annular lateral wall velocity <4 cm/s, and elevated filling pressures E/e′ > 10 ratio, 24 some of which are components of the H2FPEF and HFA‐PEFF scores as well. Although originally developed for the diagnosis of HFpEF, the H2FPEF and HFA‐PEFF scores have demonstrated prognostic value in several cardiac diseases, 9 , 10 , 13 which share similarities with HCM, such as diastolic dysfunction, AF, and mitral or tricuspid regurgitation. 11 , 27 In patients diagnosed with HFpEF, both the H2FPEF and HFA‐PEFF scores are highly prognostic. 24 However, a recent study found that the HFA‐PEFF score exhibited greater diagnostic and prognostic utility compared to the H2FPEF score in patients with HFpEF caused by cardiac amyloidosis. 10 Furthermore, the H2FPEF and HFA‐PEFF scores showed significant discordance in classifying patients with unexplained dyspnoea in the community. 25 Based on these findings, further in‐depth research is warranted to explore and compare the prognostic value of the H2FPEF and HFA‐PEFF scores in patients with HCM. The H2FPEF score has been proven to have prognostic value for HF risk stratification in HCM patients, as a recent study found that a high H2FPEF score was independently associated with worse HF outcomes. 5 However, HFA‐PEFF score was not evaluated in the study by Laenens et al., 5 and a comparative analysis between H2FPEF and HFA‐PEFF scores was lacking. Furthermore, the previous study also demonstrated that Asian ethnicity is an independent risk predictor for HF and all‐cause death in HCM patients, 5 indicating potential differences in applying the H2FPEF score to Asian HCM patients. In the present study of Asian HCM patients, we found that both the H2FPEF and HFA‐PEFF scores served as a promising tool for risk prediction in patients with HCM with intermediate‐high H2FPEF and high HFA‐PEFF scores having equivalent risks.
The discordance between the H2FPEF and HFA‐PEFF scores is not surprising, as noted in previous studies. 9 , 10 The H2FPEF and HFA‐PEFF scores were developed on different foundations. The H2FPEF score was created using variables and strength of association by beta coefficients from logistic regression, 21 with invasive exercise testing as the gold standard. 6 In contrast, the HFA‐PEFF score was a consensus recommendation, consisting of several major and minor criteria. 7 In terms of the components of both scores, the HFA‐PEFF score involves nine echocardiographic variables, compared to two for the H2FPEF score, along with brain natriuretic peptide levels, which did not independently predict HFpEF in the development of the H2FPEF score. 6 In this study, we observed that the discordance was due to differences in the variables used to calculate both scores (Table S3 ). Upon comparing patients stratified discordantly by these two scores, we found that H2FPEF score is heavily weighted by AF (3 points) and BMI (2 points), while the HFA‐PEFF was more strongly associated with LVMI and NT‐proBNP levels (Table S3 ; all P < 0.05). In patients with AF, the discordance between the two scores was significantly lower. Conversely, patients with elevated NT‐proBNP levels showed significantly higher discordance (Table S2 ). These findings further indicate that the discordance between the H2FPEF and HFA‐PEFF scores is affected by patient characteristics, particularly the presence of AF and elevated NT‐proBNP levels. The significantly lower agreement in patients with elevated NT‐proBNP highlights the impact of including this variable in the H2FPEF score and warrants further investigation.
However, some variables, including E/e′ ratio and peak tricuspid regurgitation velocity, were utilized for the calculation of both H2FPEF and HFA‐PEFF scores, which may partly explain prognostic value for HCM patients in both scores. The variables used to calculate the HFA‐PEFF score, including LA end‐systolic dimension, NT‐proBNP levels, LV mass, E/e′ ratio, and peak tricuspid regurgitation velocity, were higher and septal e′ was lower in the intermediate and high H2FPEF score groups compared to the low H2FPEF score group, suggesting a higher HFA‐PEFF score (Table 1 ). Meanwhile, in the high HFA‐PEFF score group, the variables used to calculate the H2FPEF score, including age, peak tricuspid regurgitation velocity, E/e′ ratio, and percentage of AF, were similarly elevated compared to those in the low HFA‐PEFF score group, indicating a higher H2FPEF score (Table 1 ). Those findings may also explain both H2FPEF and HFA‐PEFF scores have the prognostic value in risk stratification for HCM patients. Notably, the high H2FPEF group constituted a relatively small proportion in the present study (n = 57, 5.3%). This may be attributed to the relatively small number of patients with AF, which contributes 3 points to the H2FPEF score. In the study by Laenens et al., 5 the high H2FPEF group included 105 patients (11.0%), while the patients with AF totalled 219 (22.9%). In contrast, the present study included 114 patients with AF (10.7%).
When comparing these two scores, both the high HFA‐PEFF score and the intermediate‐high H2FPEF score demonstrated comparable risk stratification for first HF hospitalization and death. A previous community‐based study of patients with unexplained dyspnoea found that high H2FPEF and HFA‐PEFF scores conferred equivalent risk for HF and death compared to known HFpEF. 9 However, in HFpEF caused by cardiac amyloidosis, a high HFA‐PEFF score, but not H2FPEF, independently predicted death. This difference may be attributable to the distinct HFpEF phenotype observed in cardiac amyloidosis, which often presents with hypotension, weight loss, and a younger age in patients with immunoglobulin light‐chain amyloidosis. 10 The situation in patients with HCM appears similar to that of the community‐based cohort study. 9 HF is a long‐term burden in HCM, predominantly affecting older patients, 28 and tending towards the classical HFpEF phenotype (elderly, obesity, hypertension, and atrial fibrillation). 29 This may partially explain the equivalent predictive value of the two scores in HCM patients.
Nowadays, comprehensive echocardiography and laboratory tests, including BNP and NT‐proBNP, are commonly used in the diagnosis and clinical decision‐making for patients with HCM, as recommended by the guidelines. 3 , 30 This implies that the H2FPEF and HFA‐PEFF scores do not impose an additional burden but rather capitalize on existing test results to assess patient prognosis. However, using two tools with significant discordance in clinical practice is often difficult and confusing. Given that both scores offer risk assessment value, we proposed a new algorithm integrating both scores, which demonstrated superior effectiveness for the primary endpoint. Notably, we only preliminarily selected and assigned weights to the components of both scores. The major criteria of HFA‐PEFF are composite and consist of multiple variables. Finding predictive indicators among them might further improve model performance. Future research is needed to further develop and validate a new algorithm for HF risk stratification in patients with HCM before implementing both scores in clinical practice.
4.1. Limitations
Firstly, the data used to calculate both scores were obtained retrospectively. Secondly, the present study was conducted at a single centre in central China, which may weaken its generalizability to other populations, especially those in multicentre and Western cohorts, where variations in demographics, healthcare systems, and cultural factors could influence the applicability of our results. Moreover, some studies have indicated that ethnicity also affect the prognosis of patients with HCM. 5 Notably, ethnical differences have been reported in both scores beyond Asian populations. Black individuals were associated with a higher H2FPEF score tertile, while White individuals were associated with a higher HFA‐PEFF score. 9 This presents challenges in applying both scores for prognosis, as the risk represented by a higher H2FPEF score may be lower in Black populations compared to White populations, for example. Therefore, the results need to be validated prospectively through multicentre studies involving diverse ethnic groups.
The potential impact of confounders, including sex, coronary heart disease, and LVOTO, while not directly involved in the calculation of either score, is associated with the development of HF 3 , 5 and could influence our findings. Nonetheless, after adjusting for confounders, the H2FPEF and HFA‐PEFF scores remained independently associated with the primary endpoint (Table 2 ). Furthermore, subgroup analysis stratified by gender showed consistent significant differences in survival between high and low score groups (Figure S3 ). Sensitivity analysis indicated that excluding patients with coronary heart disease and LVOTO did not change our conclusions, suggesting that our primary conclusions are robust to the potential influence of these specific confounders.
Finally, for the purpose of the current study, some variables used to calculate the HFA‐PEFF and H2FPEF scores were modified. We utilized LA end‐systolic dimension instead of LA volume index. While this substitution is justified by guideline recommendations, it may limit comparability with other studies. However, the impact of this substitution appears limited, as only a relatively small number of patients (80, 7.5%) obtained 2 points in the morphological domain of the HFA‐PEFF score solely based on LA end‐systolic dimension. A sensitivity analysis excluding these patients yielded similar results.
5. Conclusions
In patients with HCM, there is significant discordance in the classifications of the H2FPEF and HFA‐PEFF scores. Both scores demonstrate value in identifying risks for HF hospitalization and all‐cause death, with the intermediate‐high H2FPEF score and the high HFA‐PEFF score being equivalent in risk stratification. However, the discordance would hinder the application of both scores in clinical practice. Integrating components of both scores demonstrated an improved effectiveness, suggesting future research to further explore integrating both scores to develop a new algorithm, which should be validated in multicentre and diverse populations, ultimately creating a more applicable risk stratification tool for patients with HCM.
Funding
None.
Conflict of interest
None declared.
Supporting information
Table S1. Variables for the calculation of H2FPEF and HFA‐PEFF scores of the study population.
Table S2. Discordance and concordance between H2FPEF and HFA‐PEFF scores in the study population and subgroups.
Table S3. Baseline Clinical Characteristics, Echocardiographic and Laboratory Findings of the Study Population Stratified Discordantly by the H2FPEF and HFA‐PEFF Scores.
Table S4. Sensitivity analysis of univariable and multivariable Cox regression for the association of variables with the primary endpoint in patients without left ventricular outflow tract obstruction or severe mitral regurgitation.
Table S5. Sensitivity analysis of univariable and multivariable Cox regression for the association of variables with the primary endpoint in patients without coronary heart disease.
Table S6. Sensitivity analysis of univariable and multivariable Cox regression for the association of variables with the primary endpoint in patients scoring 2 points in morphological domain without sole left atrial criterion.
Figure S1. Flowchart of Study Population. NT‐proBNP, N‐terminal pro‐brain natriuretic peptide.
Figure S2. Mosaic plot of H2FPEF and HFA‐PEFF scores in subgroups with atrial fibrillation, obesity, and elevated NT‐proBNP levels. NT‐proBNP, N‐terminal pro‐brain natriuretic peptide.
Figure S3. Kaplan–Meier curves for the primary endpoint in subgroups stratified by gender.
Figure S4. Sensitivity analysis of Kaplan–Meier curves for the first heart failure hospitalization in subgroups. Sensitivity analysis was performed in 3 subgroups. (A,D) Excluding patients with left ventricular outflow tract obstruction or severe mitral regurgitation. (B,E) Excluding patients with coronary artery disease. (C,F) Excluding patients obtaining 2 points in the morphological domain of the HFA‐PEFF score based solely on left atrial end‐systolic dimension.
Figure S5. Sensitivity analysis of Kaplan–Meier curves for all‐cause death in subgroups. Sensitivity analysis was performed in 3 subgroups. (A,D) Excluding patients with left ventricular outflow tract obstruction or severe mitral regurgitation. (B,E) Excluding patients with coronary artery disease. (C,F) Excluding patients obtaining 2 points in the morphological domain of the HFA‐PEFF score based solely on left atrial end‐systolic dimension.
Gao, Y.‐P. , Liu, H.‐Y. , Bi, X.‐J. , Sun, J. , Zhu, Y. , Zhou, W. , Fan, Y.‐T. , Cheng, X.‐Q. , Huang, P.‐N. , Liu, Y.‐N. , and Deng, Y.‐B. (2025) H2FPEF and HFA‐PEFF scores for heart failure risk stratification in hypertrophic cardiomyopathy patients. ESC Heart Failure, 12: 2225–2238. 10.1002/ehf2.15247.
Contributor Information
Hong‐Yun Liu, Email: yani.liu@163.com.
Ya‐Ni Liu, Email: ybdeng2007@hotmail.com, Email: yani.liu@163.com.
You‐Bin Deng, Email: ybdeng2007@hotmail.com.
References
- 1. Maron BJ. Clinical course and management of hypertrophic cardiomyopathy. N Engl J Med 2018;379:655‐668. doi: 10.1056/NEJMra1710575 [DOI] [PubMed] [Google Scholar]
- 2. Maron BJ, Desai MY, Nishimura RA, Spirito P, Rakowski H, Towbin JA, et al. Diagnosis and evaluation of hypertrophic cardiomyopathy: JACC state‐of‐the‐art review. J Am Coll Cardiol 2022;79:372‐389. doi: 10.1016/j.jacc.2021.12.002 [DOI] [PubMed] [Google Scholar]
- 3. Ommen SR, Ho CY, Asif IM, Balaji S, Burke MA, Day SM, et al. AHA/ACC/AMSSM/HRS/PACES/SCMR guideline for the management of hypertrophic cardiomyopathy: a report of the American Heart Association/American College of Cardiology Joint Committee on Clinical Practice Guidelines. Circulation 2024;2024:149. doi: 10.1161/CIR.0000000000001250 [DOI] [PubMed] [Google Scholar]
- 4. Olivotto I, Udelson JE, Pieroni M, Rapezzi C. Genetic causes of heart failure with preserved ejection fraction: emerging pharmacological treatments. Eur Heart J 2023;44:656‐667. doi: 10.1093/eurheartj/ehac764 [DOI] [PubMed] [Google Scholar]
- 5. Laenens D, Zegkos T, Kamperidis V, Wong RCC, Li TYW, Sia CH, et al. Heart failure risk assessment in patients with hypertrophic cardiomyopathy based on the H2FPEF score. Eur J Heart Fail 2024;26:2173‐2182. doi: 10.1002/ejhf.3413 [DOI] [PubMed] [Google Scholar]
- 6. Reddy YNV, Carter RE, Obokata M, Redfield MM, Borlaug BA. A simple, evidence‐based approach to help guide diagnosis of heart failure with preserved ejection fraction. Circulation 2018;138:861‐870. doi: 10.1161/CIRCULATIONAHA.118.034646 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Pieske B, Tschöpe C, De Boer RA, Fraser AG, Anker SD, Donal E, et al. How to diagnose heart failure with preserved ejection fraction: the HFA–PEFF diagnostic algorithm: a consensus recommendation from the Heart Failure Association (HFA) of the European Society of Cardiology (ESC). Eur J Heart Fail 2020;22:391‐412. doi: 10.1002/ejhf.1741 [DOI] [PubMed] [Google Scholar]
- 8. Reddy YNV, Kaye DM, Handoko ML, van de Bovenkamp AA, Tedford RJ, Keck C, et al. Diagnosis of heart failure with preserved ejection fraction among patients with unexplained dyspnea. JAMA Cardiol 2022;7:891‐899. doi: 10.1001/jamacardio.2022.1916 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Selvaraj S, Myhre PL, Vaduganathan M, Claggett BL, Matsushita K, Kitzman DW, et al. Application of diagnostic algorithms for heart failure with preserved ejection fraction to the community. JACC Heart Fail 2020;8:640‐653. doi: 10.1016/j.jchf.2020.03.013 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Tomasoni D, Aimo A, Merlo M, Nardi M, Adamo M, Bellicini MG, et al. Value of the HFA‐PEFF and H2FPEF scores in patients with heart failure and preserved ejection fraction caused by cardiac amyloidosis. Eur J Heart Fail 2022;24:2374‐2386. doi: 10.1002/ejhf.2616 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Ludwig S, Pellegrini C, Gossling A, Rheude T, Voigtländer L, Bhadra OD, et al. Prognostic value of the H2FPEF score in patients undergoing transcatheter aortic valve implantation. ESC Heart Fail 2021;8:461‐470. doi: 10.1002/ehf2.13096 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Seoudy H, von Eberstein M, Frank J, Thomann M, Puehler T, Lutter G, et al. HFA‐PEFF score: prognosis in patients with preserved ejection fraction after transcatheter aortic valve implantation. ESC Heart Fail 2022;9:1071‐1079. doi: 10.1002/ehf2.13774 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Verbrugge FH, Reddy YNV, Sorimachi H, Omote K, Carter RE, Borlaug BA. Diagnostic scores predict morbidity and mortality in patients hospitalized for heart failure with preserved ejection fraction. Eur J Heart Fail 2021;23:954‐963. doi: 10.1002/ejhf.2142 [DOI] [PubMed] [Google Scholar]
- 14. Mitchell C, Rahko PS, Blauwet LA, Canaday B, Finstuen JA, Foster MC, et al. Guidelines for performing a comprehensive transthoracic echocardiographic examination in adults: recommendations from the American Society of Echocardiography. J Am Soc Echocardiogr 2019;32:1‐64. doi: 10.1016/j.echo.2018.06.004 [DOI] [PubMed] [Google Scholar]
- 15. Lang RM, Badano LP, Mor‐Avi V, Afilalo J, Armstrong A, Ernande L, et al. Recommendations for cardiac chamber quantification by echocardiography in adults: an update from the American Society of Echocardiography and the European Association of Cardiovascular Imaging. J Am Soc Echocardiogr 2015;28:1‐39.e14. doi: 10.1016/j.echo.2014.10.003 [DOI] [PubMed] [Google Scholar]
- 16. Nagueh SF, Smiseth OA, Appleton CP, Byrd BF, Dokainish H, Edvardsen T, et al. Recommendations for the evaluation of left ventricular diastolic function by echocardiography: an update from the American Society of Echocardiography and the European Association of Cardiovascular Imaging. J Am Soc Echocardiogr 2016;29:277‐314. doi: 10.1016/j.echo.2016.01.011 [DOI] [PubMed] [Google Scholar]
- 17. Otto CM, Nishimura RA, Bonow RO, Carabello BA, Erwin JP, Gentile F, et al. 2020 ACC/AHA guideline for the management of patients with valvular heart disease: executive summary: a report of the American College of Cardiology/American Heart Association Joint Committee on Clinical Practice Guidelines. Circulation 2021;143:e35‐e71. doi: 10.1161/CIR.0000000000000932 [DOI] [PubMed] [Google Scholar]
- 18. Lancellotti P, Tribouilloy C, Hagendorff A, Popescu BA, Edvardsen T, Pierard LA, et al. Recommendations for the echocardiographic assessment of native valvular regurgitation: an executive summary from the European Association of Cardiovascular Imaging. Eur Heart J Cardiovasc Imaging 2013;14:611‐644. doi: 10.1093/ehjci/jet105 [DOI] [PubMed] [Google Scholar]
- 19. Stefano GT, Zhao H, Schluchter M, Hoit BD. Assessment of echocardiographic left atrial size: accuracy of M‐mode and two‐dimensional methods and prediction of diastolic dysfunction. Echocardiography 2012;29:379‐384. doi: 10.1111/j.1540-8175.2011.01643.x [DOI] [PubMed] [Google Scholar]
- 20. Caraballo C, Desai NR, Mulder H, Alhanti B, Wilson FP, Fiuzat M, et al. Clinical implications of the New York Heart Association classification. J Am Heart Assoc 2019;8:e014240. doi: 10.1161/JAHA.119.014240 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Sullivan LM, Massaro JM, D'Agostino RB. Presentation of multivariate data for clinical use: the Framingham study risk score functions. Stat Med 2004;23:1631‐1660. doi: 10.1002/sim.1742 [DOI] [PubMed] [Google Scholar]
- 22. Maron BJ, Rowin EJ, Casey SA, Link MS, Lesser JR, Chan RHM, et al. Hypertrophic cardiomyopathy in adulthood associated with low cardiovascular mortality with contemporary management strategies. J Am Coll Cardiol 2015;65:1915‐1928. doi: 10.1016/j.jacc.2015.02.061 [DOI] [PubMed] [Google Scholar]
- 23. Maron BJ, Maron MS. Hypertrophic cardiomyopathy. Lancet 2013;381:242‐255. doi: 10.1016/S0140-6736(12)60397-3 [DOI] [PubMed] [Google Scholar]
- 24. Mandeş L, Roşca M, Ciupercă D, Popescu BA. The role of echocardiography for diagnosis and prognostic stratification in hypertrophic cardiomyopathy. J Echocardiogr 2020;18:137‐148. doi: 10.1007/s12574-020-00467-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Rowin EJ, Maron MS, Chan RH, Hausvater A, Wang W, Rastegar H, et al. Interaction of adverse disease related pathways in hypertrophic cardiomyopathy. Am J Cardiol 2017;120:2256‐2264. doi: 10.1016/j.amjcard.2017.08.048 [DOI] [PubMed] [Google Scholar]
- 26. Kitaoka H, Kubo T, Okawa M, Takenaka N, Sakamoto C, Baba Y, et al. Tissue Doppler imaging and plasma BNP levels to assess the prognosis in patients with hypertrophic cardiomyopathy. J Am Soc Echocardiogr 2011;24:1020‐1025. doi: 10.1016/j.echo.2011.05.009 [DOI] [PubMed] [Google Scholar]
- 27. Geske JB, Sorajja P, Nishimura RA, Ommen SR. Evaluation of left ventricular filling pressures by Doppler echocardiography in patients with hypertrophic cardiomyopathy: correlation with direct left atrial pressure measurement at cardiac catheterization. Circulation 2007;116:2702‐2708. doi: 10.1161/CIRCULATIONAHA.107.698985 [DOI] [PubMed] [Google Scholar]
- 28. Ho CY, Day SM, Ashley EA, Michels M, Pereira AC, Jacoby D, et al. Genotype and lifetime burden of disease in hypertrophic cardiomyopathy: insights from the Sarcomeric Human Cardiomyopathy Registry (SHaRe). Circulation 2018;138:1387‐1398. doi: 10.1161/CIRCULATIONAHA.117.033200 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Packer M. Do most patients with obesity or type 2 diabetes, and atrial fibrillation, also have undiagnosed heart failure? A critical conceptual framework for understanding mechanisms and improving diagnosis and treatment. Eur J Heart Fail 2020;22:214‐227. doi: 10.1002/ejhf.1646 [DOI] [PubMed] [Google Scholar]
- 30. Authors/Task Force members , Elliott PM, Anastasakis A, Borger MA, Borggrefe M, Cecchi F, et al. ESC guidelines on diagnosis and management of hypertrophic cardiomyopathy: the task force for the diagnosis and Management of Hypertrophic Cardiomyopathy of the European Society of Cardiology (ESC). Eur Heart J 2014;2014:2733‐2779. doi: 10.1093/eurheartj/ehu284 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Table S1. Variables for the calculation of H2FPEF and HFA‐PEFF scores of the study population.
Table S2. Discordance and concordance between H2FPEF and HFA‐PEFF scores in the study population and subgroups.
Table S3. Baseline Clinical Characteristics, Echocardiographic and Laboratory Findings of the Study Population Stratified Discordantly by the H2FPEF and HFA‐PEFF Scores.
Table S4. Sensitivity analysis of univariable and multivariable Cox regression for the association of variables with the primary endpoint in patients without left ventricular outflow tract obstruction or severe mitral regurgitation.
Table S5. Sensitivity analysis of univariable and multivariable Cox regression for the association of variables with the primary endpoint in patients without coronary heart disease.
Table S6. Sensitivity analysis of univariable and multivariable Cox regression for the association of variables with the primary endpoint in patients scoring 2 points in morphological domain without sole left atrial criterion.
Figure S1. Flowchart of Study Population. NT‐proBNP, N‐terminal pro‐brain natriuretic peptide.
Figure S2. Mosaic plot of H2FPEF and HFA‐PEFF scores in subgroups with atrial fibrillation, obesity, and elevated NT‐proBNP levels. NT‐proBNP, N‐terminal pro‐brain natriuretic peptide.
Figure S3. Kaplan–Meier curves for the primary endpoint in subgroups stratified by gender.
Figure S4. Sensitivity analysis of Kaplan–Meier curves for the first heart failure hospitalization in subgroups. Sensitivity analysis was performed in 3 subgroups. (A,D) Excluding patients with left ventricular outflow tract obstruction or severe mitral regurgitation. (B,E) Excluding patients with coronary artery disease. (C,F) Excluding patients obtaining 2 points in the morphological domain of the HFA‐PEFF score based solely on left atrial end‐systolic dimension.
Figure S5. Sensitivity analysis of Kaplan–Meier curves for all‐cause death in subgroups. Sensitivity analysis was performed in 3 subgroups. (A,D) Excluding patients with left ventricular outflow tract obstruction or severe mitral regurgitation. (B,E) Excluding patients with coronary artery disease. (C,F) Excluding patients obtaining 2 points in the morphological domain of the HFA‐PEFF score based solely on left atrial end‐systolic dimension.