Beyond Wall Thickness: Clinical Predictors of Genotype Positivity in Hypertrophic Cardiomyopathy

Abstract

Background: Genetic testing in hypertrophic cardiomyopathy (HCM) yields variable positivity rates. Identifying clinical predictors of positive genetic tests could improve pre-test counseling and refine expectations about diagnostic yield. Methods: We analyzed consecutive genotyped HCM probands from a contemporary multicenter cohort across four Italian tertiary centers. Genotype positivity was defined as the presence of ≥1 pathogenic or likely pathogenic variant (ACMG classes 4–5). Multivariable logistic regression identified predictors of genotype positivity. Sensitivity analyses assessed the incremental value of left atrial volume index (LAVI) ≥ 34 mL/m² and the mode of first clinical presentation. Results: Among 274 genotyped probands (median age at diagnosis 54 years; 62% male), 86 (31%) were genotype-positive (38% MYBPC3, 29% MYH7). Age at diagnosis <40 years (OR 2.38, 95%CI 1.26–4.51, p = 0.008), family history of sudden cardiac death/major ventricular arrhythmias (OR 2.34, 95%CI 1.16–4.84, p = 0.019) and family history of non-ischemic cardiomyopathy (OR 1.92, 95%CI 1.04–3.54, p = 0.038), were independently associated with genotype positivity whereas arterial hypertension was inversely associated (OR 0.42, 95%CI 0.23–0.77). Maximal left ventricular wall thickness > 20 mm and gender were not predictive of genotype positivity. Inclusion of LAVI modestly improved the model performance (AUC 0.769, p = 0.016, ΔAUC +0.024; DeLong p = 0.016) but without leading to meaningful patient reclassification. Conclusions: Genotype positivity in HCM links to earlier onset and family history; traditional severity markers and initial presentation may not independently suggest genetic causality. These findings may help shape a personalized approach to genetic counseling in HCM.

1. Introduction

Hypertrophic cardiomyopathy (HCM) is the most common inherited cardiac disease, often due to pathogenic sarcomeric variants. Current guidelines [1,2,3] advocate genetic testing in all clinically diagnosed HCM patients to clarify etiology and enable cascade screening, but only about 30–50% test positive, with substantial inter-cohort variability [4]. MYBPC3 and MYH7 are the most frequently affected genes [5], accounting for roughly 45% and 35% of genotype-positive individuals, respectively. Genotype-positive patients are typically younger at diagnosis, have more often a family history of cardiomyopathy or sudden cardiac death (SCD) [6], and accumulate disease-related complications earlier, supporting a biologically distinct trajectory [7]. A detailed review of the complex genetic architecture of HCM has recently been published [8]. Genotype-negative individuals usually present later and have more cardiovascular comorbidities, suggesting alternative or multifactorial mechanisms for left ventricular hypertrophy, potentially including polygenic risk from multiple single-nucleotide polymorphisms (SNPs) [9].

Clinical predictors of genotype positivity have been explored, yielding tools like the Toronto Hypertrophic Cardiomyopathy Genotype Score [10] and the Mayo Clinic genotype prediction score [11]. These perform reasonably in derivation cohorts, but external validation is limited [12]. Recently, machine-learning approaches with broader clinical, electrocardiographic, and imaging data show promise but face interpretability and generalizability challenges [13].

Using a contemporary multicenter HCM registry, we aimed to identify independent clinical predictors of genotype positivity and assess the incremental contribution of echocardiographic/clinical features. Our goal was to provide a pragmatic, phenotype-driven framework to inform genetic testing strategies in everyday HCM practice.

2. Materials and Methods

2.1. Study Design and Population

This study was conducted within a contemporary multicenter registry of HCM patients, including four tertiary referral centers in the Piedmont region (Italy), one of which is the regional heart transplantation center. Consecutive adult patients with a clinical diagnosis of HCM who underwent genetic testing were prospectively enrolled starting from January 2020. Relatives identified through cascade or family screening were excluded to avoid familial clustering, non-independence of observations, and enrichment of genotype-positive cases.

HCM was diagnosed according to current international guidelines [1,2,3], requiring left ventricular hypertrophy (maximal wall thickness ≥15 mm in adult probands) unexplained by the loading conditions. Patients with phenocopies (e.g., amyloidosis, Fabry disease, or other infiltrative or metabolic cardiomyopathies) were excluded.

Genetic testing was performed according to local clinical practice and guideline recommendations, using the Illumina True Sight One panel, with analysis involving an extended panel of 51 genes involved in HCM (full list provided in the Supplemental Materials). Variant interpretation followed the American College of Medical Genetics and Genomics (ACMG) criteria [14].

The primary outcome was genotype positivity, defined as the presence of at least one pathogenic (P) or likely pathogenic (LP) variant (ACMG class 4 or 5) in an HCM-related gene.

Demographic data, cardiovascular risk factors, family history, clinical presentation, and prior cardiac events were collected at enrollment, coinciding with the initial evaluation and genotyping at participating centers. Family history (first- and second-degree relatives) was specifically assessed for SCD or major ventricular arrhythmic events, as well as for non-ischemic cardiomyopathy. Resting 12-lead electrocardiograms were analyzed for rhythm, atrioventricular and intraventricular conduction abnormalities, fragmented QRS, pathologic Q waves, repolarization abnormalities, and voltage criteria for left ventricular hypertrophy. A positive family history for non-ischemic cardiomyopathy (NICM) was defined as either documented evidence of HCM in a first- or second-degree family member (based on a review of clinical investigations) or a highly convincing patient report suggestive of NICM, independently of the age of the family member at diagnosis. A positive family history for SCD was defined as unexplained death in first- or second-degree relatives <80 years of age. Comprehensive transthoracic echocardiography was performed in all patients at enrollment, following current recommendations. Measured parameters included left ventricular dimensions and volumes, left ventricular ejection fraction, maximal wall thickness, diastolic indices (including E/e′ ratio), left atrial volume index (LAVI), right ventricular size and function, and the presence of left ventricular outflow tract obstruction at rest or with provocation. Left atrial enlargement was defined as LAVI ≥ 34 mL/m². Clinical follow-up was prospectively collected starting from 2020, with periodic reassessment at each participating center. Follow-up data included arrhythmic events, heart failure progression, hospitalizations, device implantation, advanced therapies, and vital status. For the present analysis, follow-up was used primarily to ensure contemporary phenotypic characterization rather than time-to-event modeling. Clinical and genetic data were collected using a standardized case report form shared across participating centers.

2.2. Statistical Analysis

Continuous variables are reported as medians (interquartile range) and were compared using the Mann–Whitney U test. Categorical variables, presented as counts and percentages, were compared using the χ² test (with Yates’ correction where applicable). Multivariable logistic regression was used to identify independent clinical and phenotypic predictors of genotype positivity at enrollment. Candidate variables were selected based on reported clinical relevance and univariable associations. Treatment-related or treatment-influenced variables, such as ICD and medical treatment, were excluded from the model because they reflect prior management rather than only patient characteristics. Multivariable models were fitted using a complete-case approach; thus, only participants with non-missing data for all included covariates contributed to each model. Due to the biological relevance, interactions between sex and each predictor included in the multivariable model were tested and sex-stratified analyses were performed.

As a sensitivity, pre-specified analysis, we explored whether atrial remodeling, as reflected by an enlarged LAVI or mode of first clinical presentation, improved genotype prediction when added to the main multivariable model (independent from the results of the univariable assessment). Three alternative models were constructed by individually including arrhythmic presentation (sustained or non-sustained ventricular arrhythmias or supraventricular arrhythmias), syncope, or heart failure as the first clinical manifestation. Each variable was entered separately into the main model to avoid collinearity. Model discrimination was assessed using the C-statistic (area under the receiver operating characteristic curve, AUC). Discriminative performance of extended models was compared with the main model using the DeLong test for correlated ROC curves on aligned samples. Model calibration was evaluated using calibration plots across deciles of predicted risk and the Brier score. Incremental predictive performance of extended models was explored using category-free net reclassification improvement (NRI) and integrated discrimination improvement (IDI).

Two further sensitivity analyses were conducted: one restricting genotype positivity to pathogenic/likely pathogenic variants in core sarcomere genes (MYBPC3, MYH7, TNNT2, TNNI3, TPM1, MYL2, MYL3, ACTC1), and another treating maximal left ventricular wall thickness as a continuous variable.

All statistical analyses were performed in Python (version 3.13) using reproducible scripts.

3. Results

3.1. Baseline Characteristics of the Study Population

A total of 274 genotyped HCM probands were analyzed, including 86 (31%) genotype positive (

Table 1. Baseline characteristics of the study population by genotype.

Variable	Overall, n = 274	Gen−, n = 188	Gen+, n = 86	p-Value
Age at enrollment, years	61.0 [51.0–69.0]	62.0 [54.0–70.0]	56.0 [39.2–67.8]	0.002
Age at diagnosis, years	54.0 [38.2–64.0]	55.5 [46.0–65.0]	43.5 [31.0–58.0]	<0.001
Age at diagnosis < 40 years	72 (26.3%)	33 (17.6%)	39 (45.3%)	<0.001
Male sex	171 (62.4%)	113 (60.1%)	58 (67.4%)	0.303
Body mass index, kg/m2	25.0 [21.8–29.4]	25.5 [22.0–29.4]	24.2 [21.3–27.2]	0.142
Serum creatinine, mg/dL	1.0 [0.8–1.3]	1.0 [0.8–1.3]	1.0 [0.9–1.2]	0.782
Diabetes mellitus	38 (13.9%)	28 (15.0%)	10 (11.6%)	0.580
Hypertension	158 (57.9%)	126 (67.4%)	32 (37.2%)	<0.001
Dyslipidemia	120 (44.0%)	89 (47.6%)	31 (36.0%)	0.098
Coronary artery disease	25 (9.2%)	22 (11.8%)	3 (3.5%)	0.039
Prior stroke or TIA	9 (3.3%)	8 (4.3%)	1 (1.2%)	0.281
Prior septal myectomy	8 (3.1%)	4 (2.3%)	4 (4.9%)	0.275
History of atrial fibrillation	76 (27.8%)	54 (28.9%)	22 (25.6%)	0.675
Prior CRT implantation	9 (3.3%)	6 (3.2%)	3 (3.5%)	1.000
Prior pacemaker implantation	18 (6.6%)	15 (8.0%)	3 (3.5%)	0.197
History of syncope	24 (9.4%)	16 (9.2%)	8 (9.8%)	1.000
Prior aborted cardiac arrest	11 (4.0%)	5 (2.7%)	6 (7.0%)	0.178
Sustained VT	16 (5.9%)	7 (3.7%)	9 (10.5%)	0.055
NSVT	32 (11.8%)	19 (10.2%)	13 (15.3%)	0.318
Prior ICD implantation	43 (15.8%)	23 (12.3%)	20 (23.3%)	0.033
Race/ethnicity				0.390
Caucasian	261 (95.3%)	177 (94.1%)	84 (97.7%)
Black	11 (4.0%)	9 (4.8%)	2 (2.3%)
First clinical presentation				0.670
- Asymptomatic	104 (38.4%)	69 (37.3%)	35 (40.7%)
- Heart failure	39 (14.4%)	30 (16.2%)	9 (10.5%)
- Chest pain	45 (16.6%)	35 (18.9%)	10 (11.6%)
- Supraventricular arrhythmias	30 (11.1%)	19 (10.3%)	11 (12.8%)
- Syncope	16 (5.9%)	9 (4.9%)	7 (8.1%)
- NSVT	3 (1.1%)	2 (1.1%)	1 (1.2%)
- Sustained VAs	5 (1.8%)	3 (1.6%)	2 (2.3%)
- AV blocks	4 (1.5%)	2 (1.1%)	2 (2.3%)
- Other	25 (9.2%)	16 (8.6%)	9 (10.5%)
Smoking status				0.212
Active	52 (19.1%)	40 (21.5%)	12 (14.0%)
Previous	71 (26.1%)	44 (23.7%)	27 (31.4%)
NYHA functional class				0.659
I	165 (61.3%)	109 (59.6%)	56 (65.1%)
II	80 (29.7%)	56 (30.6%)	24 (27.9%)
III	22 (8.2%)	17 (9.3%)	5 (5.8%)
IV	2 (0.7%)	1 (0.5%)	1 (1.2%)

Legend: AV: atrioventricular, CRT: cardiac resynchronization therapy, Gen−: genotype negative, Gen+: genotype positive, NSVT: non-sustained ventricular tachycardia, TIA: transient ischemic attack, VAs: ventricular arrhythmias, VT: ventricular tachycardia. Squared blankets include IQR.

). Most patients were male (113, 61%). Median age at enrollment was 61 years (IQR 51–69). Genotype-positive patients were significantly younger both at diagnosis (median 44 vs. 56 years, p < 0.001) and at enrollment (median 56 vs. 62 years, p = 0.002). Accordingly, diagnosis before 40 years was significantly more frequent among genotype-positive patients (45.3% vs. 17.6%, p < 0.001).

Traditional cardiovascular risk factors were common in the overall cohort; however, hypertension was significantly less prevalent among genotype-positive patients (37.2% vs. 67.4%, p < 0.001). The family history of non-ischemic cardiomyopathy was more frequent in genotype-positive patients (51.2% vs. 24.9%, p < 0.001), as was family history of SCD or major ventricular arrhythmic events (31.4% vs. 13.7%, p < 0.001,

Table 2. Family history of the study population by genotype.

Variable	Overall n = 274	Gen− n = 188	Gen+ n = 86	p-Value
FH of atherosclerotic CV disease				0.064
No	231 (85.2%)	156 (84.3%)	75 (87.2%)
First degree	38 (14.0%)	29 (15.7%)	9 (10.5%)
Second degree	2 (0.7%)	0 (0.0%)	2 (2.3%)
FH of NICM				<0.001
No	181 (66.8%)	139 (75.1%)	42 (48.8%)
First degree	87 (32.1%)	44 (23.7%)	43 (50.0%)
Second degree	3 (1.1%)	2 (1.1%)	1 (1.2%)
FH of SCD/major VAs				<0.001
No	217 (80.7%)	158 (86.3%)	59 (68.6%)
First degree	37 (13.8%)	20 (10.9%)	17 (19.8%)
Second degree	15 (5.6%)	5 (2.7%)	10 (11.6%)

Legend: FH: family history, Gen−: genotype negative, Gen+: genotype positive, NICM: non-ischemic cardiomyopathy, VAs: ventricular arrhythmias.

). Family history of atherosclerotic cardiovascular disease did not differ significantly between groups (12.8% vs. 15.7%, p = 0.064). The mode of first clinical presentation did not differ significantly between genotype groups (p = 0.670), with a comparable proportion of asymptomatic patients at diagnosis (41% vs. 37%; p = 0.592). In contrast, presentation with chest pain or heart failure was more frequent in genotype-negative patients (35.1% vs. 22.1%, p = 0.03). Supraventricular arrhythmias were relatively common at presentation (11% overall), with no differences between groups, whereas ventricular arrhythmias were rare.

At enrollment, occurring a median of seven years after initial diagnosis, the clinical profile had evolved. A history of major ventricular arrhythmias—defined as aborted cardiac arrest or sustained ventricular tachycardia—was significantly more frequent among genotype-positive patients (12.8% vs. 4.3%, p = 0.02), resulting in a higher rate of ICD implantation compared with genotype-negative individuals (23.3% vs. 12.3%, p = 0.033). Rates of pacemaker and cardiac resynchronization therapy implantation were similar between groups. Non-sustained ventricular tachycardia occurred at comparable frequencies (15.3% vs. 10.2%, p = 0.318), as did a prior diagnosis of atrial fibrillation (25.6% vs. 28.9%, p = 0.675), the latter showing an almost threefold increase in prevalence compared with the time of diagnosis.

On echocardiographic assessment at enrollment (

Table 3. Echocardiographic features at enrollment by genotype status.

Variable	Overall n = 274	Gen− n = 188	Gen+ n = 86	p-Value
LV end-diastolic volume, mL	93.0 [74.0–117.0]	92.0 [74.0–120.0]	96.0 [74.0–114.0]	0.731
LV end-diastolic volume index, mL/m2	52.0 [42.5–64.0]	53.0 [42.8–64.2]	50.0 [42.5–60.0]	0.204
LV end-diastolic diameter, mm	45.0 [41.0–50.0]	45.0 [41.0–50.0]	45.0 [40.2–49.5]	0.586
IV septal thickness, mm	16.0 [14.0–18.0]	16.0 [14.0–18.0]	16.0 [13.0–19.0]	0.776
Posterior wall thickness, mm	12.0 [10.0–14.0]	12.0 [11.0–14.0]	11.0 [10.0–13.0]	0.003
Maximal LV wall thickness, mm	17.0 [15.0–20.0]	17.0 [15.0–20.0]	17.5 [15.0–20.8]	0.870
Maximal LV wall thickness > 20 mm	57 (30.3%)	38 (29.2%)	19 (32.8%)	0.753
LVEF, %	60.0 [55.0–65.0]	62.0 [55.0–65.0]	60.0 [55.0–65.0]	0.536
LVEF < 50%	51 (19.8%)	35 (20.2%)	16 (19.0%)	0.955
E/A ratio	1.0 [0.8–1.5]	0.9 [0.8–1.4]	1.0 [0.8–1.8]	0.182
E/A > 2	26 (14.0%)	15 (11.5%)	11 (20.0%)	0.193
E/e′ ratio	10.0 [8.0–14.0]	10.2 [8.0–14.0]	10.0 [8.0–13.0]	0.310
E/e′ > 14	52 (26.7%)	41 (29.7%)	11 (19.3%)	0.188
LAVI, mL/m2	42.0 [33.8–58.0]	42.0 [34.0–58.0]	44.0 [33.0–59.0]	0.801
LAVI ≥ 34 mL/mq	177 (64.6%)	118 (62.8%)	59 (68.6%)	0.423
Left atrial diameter, mm	41.0 [37.0–47.0]	41.0 [38.0–46.0]	43.0 [37.0–50.0]	0.775
RV basal diameter, mm	36.0 [34.0–38.0]	36.0 [34.0–38.0]	35.0 [34.0–38.0]	0.172
TAPSE, mm	22.0 [19.0–25.0]	22.0 [19.0–24.0]	22.0 [20.0–25.0]	0.707
RV fractional area change, %	43.0 [40.0–45.0]	43.0 [40.0–46.0]	41.0 [40.0–45.0]	0.033
RV–RA gradient, mmHg	25.0 [20.0–33.2]	25.0 [20.0–31.0]	26.0 [20.0–38.0]	0.220
Estimated RAP, mmHg	5.0 [3.0–5.0]	5.0 [3.0–5.0]	5.0 [3.0–5.0]	0.714
Resting LVOT gradient, mmHg	15.0 [0.0–42.0]	20.0 [5.2–44.2]	8.0 [0.0–30.0]	0.049
Provoked LVOT gradient, mmHg	37.0 [3.8–69.2]	50.0 [10.0–72.5]	20.0 [0.0–51.0]	0.028
Obstructive HCM (LVOT ≥ 30 mmHg)	62 (26.6%)	47 (29.7%)	15 (20.0%)	0.157
Restrictive filling pattern	66 (33.0%)	49 (35.0%)	17 (28.3%)	0.450
RV dilation	1 (1.1%)	1 (1.5%)	0 (0.0%)	1.000
RV hypertrophy	38 (16.7%)	22 (14.4%)	16 (21.3%)	0.257
Moderate/severe aortic stenosis	5 (2.0%)	5 (3.0%)	0 (0.0%)	0.174
Left ventricular thrombus	2 (0.8%)	1 (0.6%)	1 (1.2%)	1.000
SAM of mitral valve	56 (24.2%)	43 (27.4%)	13 (17.6%)	0.144
Mid-ventricular obstruction	16 (6.8%)	12 (7.5%)	4 (5.4%)	0.781
Apical aneurysm	9 (3.8%)	8 (5.0%)	1 (1.4%)	0.279
Apical thrombus	0 (0.0%)	0 (0.0%)	0 (0.0%)	1.000
Apical HCM	21 (10.1%)	17 (12.3%)	4 (5.8%)	0.221

Legend: HCM: hypertrophic cardiomyopathy, Gen−: genotype negative, Gen+: genotype positive, IV: interventricular, LAVI: left atrial volume index, LV: left ventricular, LVEF: left ventricular ejection fraction, LVOT: left ventricular outflow tract, RA: right atrium, RAP: right atrial pressure, RV: right ventricle, SAM: systolic anterior motion (of the mitral valve), TAPSE: Tricuspid Annular Plane Systolic Excursion. Squared blankets include IQR.

), interventricular septum thickness as well as maximal left ventricular wall thickness did not differ significantly between genotype-positive and genotype-negative patients. The posterior wall was modestly but significantly thicker among genotype-negative patients (median 12 vs. 11 mm, p = 0.003). Left ventricular systolic function was generally preserved in the majority of patients in both groups. Although genotype-negative patients had higher resting and provoked LVOT gradients, the prevalence of obstructive HCM (LVOT gradient ≥ 30 mmHg) did not differ significantly between groups. Diastolic indices, left atrial dimensions (LAVI 42 vs. 44 mL/mq, p = 0.423) and right ventricular (RV) dimensions were also similar.

Electrocardiographic features at enrollment did not differ significantly between groups (

Table 4. ECG features and ongoing therapy at enrollment by genotype status.

Variable	Overall n = 274	Gen− n = 188	Gen+ n = 86	p-Value
ECG features
First-degree AV	20 (7.4%)	13 (7.0%)	7 (8.1%)	0.930
High-degree AV	10 (3.7%)	8 (4.3%)	2 (2.4%)	0.729
Fragmented QRS	25 (12.9%)	17 (13.1%)	8 (12.5%)	1.000
Pathological Q waves	18 (9.3%)	10 (7.7%)	8 (12.5%)	0.411
Low QRS voltages	3 (1.5%)	2 (1.5%)	1 (1.6%)	1.000
ECG signs of LV hypertrophy	75 (38.7%)	52 (40.0%)	23 (35.9%)	0.697
Pre-excitation	1 (0.5%)	0 (0.0%)	1 (1.6%)	0.330
Negative T-waves in lateral leads	53 (27.5%)	37 (28.7%)	16 (25.0%)	0.713
Therapy at enrollment
Beta-blocker therapy	239 (87.2%)	170 (90.4%)	69 (80.2%)	0.032
RAAS inhibitor therapy	111 (40.5%)	89 (47.3%)	22 (25.6%)	0.001
MRA	48 (20.3%)	37 (22.8%)	11 (14.9%)	0.216
SGLT2i	24 (10.2%)	16 (9.9%)	8 (11.0%)	0.983
Vericiguat	1 (1.0%)	0 (0.0%)	1 (3.7%)	0.270
Loop diuretics	43 (18.4%)	32 (19.8%)	11 (15.3%)	0.527
Nitrates	6 (2.5%)	5 (3.1%)	1 (1.4%)	0.668
Oral anticoagulation	47 (19.8%)	35 (21.5%)	12 (16.2%)	0.444
Amiodarone	11 (4.6%)	8 (4.9%)	3 (4.1%)	1.000
Ranolazine	8 (3.4%)	5 (3.1%)	3 (4.1%)	0.707
Mexiletine	1 (0.4%)	0 (0.0%)	1 (1.4%)	0.314
Disopyramide	17 (7.2%)	13 (8.0%)	4 (5.4%)	0.593
Verapamil	5 (2.1%)	5 (3.1%)	0 (0.0%)	0.328
Diltiazem	1 (0.4%)	1 (0.6%)	0 (0.0%)	1.000
Mavacamten	1 (0.4%)	1 (0.6%)	0 (0.0%)	1.000

Legend: AV: atrioventricular, Gen−: genotype negative, Gen+: genotype positive, LV: left ventricle, MRA: mineral receptor antagonist, RAAS: renin–angiotensin–aldosterone system, SGLT2i: Sodium-Glucose Cotransporter 2 Inhibitors.

). Betablocker and RAAS inhibitor use (Table 4) was more prevalent in genotype-negative patients.

Among genotype-positive patients, pathogenic/likely pathogenic variants predominantly involved sarcomeric genes, with MYBPC3 (n = 33; 38.4%) and MYH7 (n = 25; 29.1%) accounting for most cases (Figure 1, blue columns). Overall, variants of uncertain significance (VUS) were identified in 67 probands (24% of the total cohort), most frequently in MYBPC3 (n = 13; 19%), MYH7 (n = 11; 16%) and MYH6 (n = 9; 13%) (Figure 1, red columns).

3.2. Predictors of a Positive Genetic Test

In the multivariable logistic regression model (

Table 5. Multivariable logistic regression model.

	OR	95%CI	p-Value
Age < 40 years at diagnosis	2.38	1.26–4.51	0.008
FH of SCD/major VAs	2.34	1.16–4.84	0.019
FH of NICM	1.92	1.04–3.54	0.038
Hypertension	0.42	0.23–0.77	0.005
LV MWT > 20 mm	1.13	0.49–2.61	0.776
Male gender	1.34	0.74–2.43	0.340

Legend. FH: family history, LV: left ventricle, MWT: maximal wall thickness, NICM: non-ischemic cardiomyopathy, SCD: sudden cardiac death, VAs: ventricular arrhythmias.

), genotype positivity was independently associated with age at diagnosis <40 years (OR 2.38; 95% CI 1.26–4.51, p = 0.008), a family history of SCD or major ventricular arrhythmias (OR 2.34; 95% CI 1.16–4.84, p = 0.019), and a family history of non-ischemic cardiomyopathy (OR 1.92; 95% CI 1.04–3.54, p = 0.038).

Arterial hypertension was inversely associated with genotype positivity (OR 0.42; 95% CI 0.23–0.77, p = 0.005). In contrast, maximal LV wall thickness >20 mm (OR 1.13; 95% CI 0.49–2.61, p = 0.776) and male sex (OR 1.34; 95% CI 0.74–2.43, p = 0.340) were not independently associated with genotype positivity (Figure 2). The main clinical model demonstrated adequate discrimination for genotype positivity, with an AUC of 0.746 (95% CI 0.680–0.810) and a Brier score of 0.178 (Figure 3).

In sex-stratified analyses (Tables S1 and S2), age at diagnosis < 40 years was significantly associated with genotype positivity among women (OR 6.48, p = 0.021) and showed a similar direction of effect among men (OR 2.51), although not reaching statistical significance (p = 0.086). Notably, no other independent predictors of a positive genotype were found in females, while only a borderline significance (p = 0.057) was found in men for a FH of SCD/major VAs. These differences are likely attributable to reduced statistical power within strata rather than true biological effect modification. Overall, discrimination remained comparable in women and men (AUC ~0.77 vs. ~0.74). No statistically significant interactions were found between sex and any predictor in the multivariable model, including age at diagnosis < 40 years (p = 0.328).

In a sensitivity analysis extending the primary model with left atrial enlargement (LAVI ≥ 34 mL/m²), the magnitude and direction of associations of the main predictors remained consistent; left atrial enlargement was independently associated with genotype positivity (OR 1.94, 95% CI 1.02–3.70, p = 0.043). The LAVI-extended model showed a modest increase in discrimination (AUC 0.769, 95% CI 0.710–0.828, Table S3) with a similar Brier score (0.177). Comparative ROC analysis in the aligned complete-case sample confirmed a small but statistically significant increase in AUC (ΔAUC +0.024; DeLong p = 0.016) (Figure 4, Table S4).

Incremental predictive performance beyond discrimination was limited, with a modest and not significant category-free net reclassification improvement (continuous NRI 0.121, p = 0.311) and integrated discrimination improvement (IDI 0.01; p = 0.162, Table S4). Calibration plots demonstrated overall good agreement between predicted and observed event rates across deciles for both models, without meaningful improvement after inclusion of LAVI ≥ 34 mL/m² (Figure S1). Overall, these metrics suggest that LAVI may help refine risk estimation rather than fundamentally reclassifying patients across probability strata. Accordingly, LAVI distribution in genotype-positive versus genotype-negative patients, with further stratification by hypertension and age, did not reveal a discernible trend.

VUS carriers, compared to P/LP carriers (Table S3), exhibited a significantly lower (halved) prevalence of a family history of both SCD/major VAs and NICM, a less pronounced early-onset profile, and a higher prevalence of hypertension. Structural parameters like maximal wall thickness and left atrial size were generally similar across groups.

Additional sensitivity analyses explored the incremental value of the mode of first clinical presentation. Arrhythmic presentation (OR 1.32, 95% CI 0.58–2.98, p = 0.51), syncope as first manifestation (OR 1.28, 95% CI 0.41–4.00, p = 0.67), and HF presentation (OR 0.73, 95% CI 0.32–1.70, p = 0.47) were not independently associated with genotype positivity when added individually to the primary model (Figure S2, Table S4). Discrimination remained similar across these models, indicating that the first clinical presentation did not provide incremental predictive value beyond the main clinical predictors (Tables S5 and S6).

Finally, two more sensitivity analyses, one restricted to carriers of pathogenic/likely pathogenic variants in core sarcomeric genes and another one treating maximal left ventricular wall thickness as a continuous variable, were performed. Results are detailed in the Supplementary Materials (Tables S7 and S8). Briefly, none led to significantly different results, suggesting robustness of our findings and demonstrating that results were not driven by genes with less established HCM causality nor by dichotomization of maximal LV wall thickness.

4. Discussion

The present study assesses clinical predictors of genotype positivity in HCM using variables readily available in routine clinical practice and derived from a contemporary multicenter cohort of genotyped HCM probands.

Our findings confirm a relatively modest yield of genetic testing in HCM (approximately 30–35%), with earlier age at diagnosis and family history of SCD/ventricular arrhythmia and of NICM emerging as the strongest predictors of genotype positivity, whereas traditional markers of disease severity and the mode of first clinical presentation did not independently contribute to genotype prediction in this cohort. While LAVI, traditionally a marker of HCM progression, offered a slight improvement to the model, neither it nor the mode of first clinical presentation significantly enhanced patient classification.

4.1. Yield of Genetic Testing in HCM

Genetic testing yields in HCM remain highly variable across studies, with pathogenic or likely pathogenic variants identified in approximately 30–50% of index cases, depending on proband selection, diagnostic criteria, ancestry, and panel breadth [8,15,16,17].

While expanded gene panels may slightly increase the diagnostic yield, this frequently occurs at the cost of a higher burden of variants of uncertain significance (VUS), particularly when evidence supporting disease causality is limited [18]. Consistent with prior literature, the genetic yield observed in our study is lower (31%) than early single-center cohorts [15] but comparable to more recent multicenter experiences reflecting broader inclusion criteria and contemporary referral patterns [17]. Gene distribution of pathogenic or likely pathogenic variants was also consistent with the literature, with MYBPC3 and MYH7 accounting for the majority of genotype-positive cases [8]. Notably, more than one-third of VUS were detected in the same genes, underscoring the importance of a periodic systematic variant reclassification [18] and potentially segregation analysis [3] of the most suspicious cases, also based on the currently identified predictors. This intermediate phenotype displayed by VUS in our population suggests that many VUS may not behave as fully penetrant sarcomeric variants. The reduced familial aggregation and increased hypertension underscore the need for cautious interpretation of VUS and emphasize the importance of integrating clinical context, particularly age at onset, family history and competitive causes for LV hypertrophy, into variant classification.

4.2. Age at Diagnosis and Family History as the Strongest Positive Predictors

The strongest predictors of genotype positivity were early age at diagnosis and a positive family history for SCD/ventricular arrhythmias and NICM. Diagnosis before 40 years of age yielded a more than two-fold higher likelihood of genotype positivity, supporting a tight connection between genotype and the biological timing of disease expression, as already suggested by the landmark predictive models of HCM genetic test positivity of the Toronto and the Mayo groups in 2013 [10] and in 2014 [11], respectively. Our results confirm these observations in a modern, multicenter population with many patients asymptomatic or mildly affected at diagnosis.

Age at diagnosis was remarkably consistent across studies, as was the proportion of male patients (approximately 60%), underscoring the comparability of the populations analyzed. Across all three cohorts, including ours, approximately half of genotype-positive patients had a family history of cardiomyopathy (defined as NICM in our study), whereas a consistent proportion of about 25% of genotype-negative patients also reported a positive family history. This observation further supports the notion that HCM inheritance is often complex and not strictly Mendelian, likely reflecting reduced penetrance, variable expressivity, and potentially oligogenic or polygenic contributions [8,17].

The association between family history of cardiomyopathy and genotype positivity was consistent across studies. However, family history definitions varied: we considered only first- and second-degree relatives, while the Mayo study did not specify relatedness, and the Toronto study included all family members. To minimize underreporting, we intentionally included reported family history of NICM, rather than restricting it to confirmed HCM, which may have reduced specificity.

While family history of SCD (prevalence 10–20% across studies) was independently associated with positive genetics in our cohort and the Mayo study [11], this was not observed in the Toronto study [10], likely due to varying prevalence and definitions. The association in our study may have been underestimated because autopsy data were limited, potentially including non-HCM-related SCD cases, especially in older relatives (60–80 years). A standardized definition of family history of NICM/SCD across studies is needed.

4.3. Echocardiographic Predictors

In our cohort, traditional markers of disease severity, such as maximal left ventricular wall thickness > 20 mm, did not independently predict genotype positivity. This contrasts with earlier models [6,10,11] based on selected cohorts, likely due to differences in patient populations, assessment timing, and alternative causes of hypertrophy. Our findings suggest that moderate hypertrophy, though clinically relevant, may be a downstream effect rather than a reliable indicator of monogenic cause when considering age and family history. Alternatively, the degree of hypertrophy in early disease stages may be influenced by gene- and variant-specific factors [19].

From a clinical standpoint, it is noteworthy that nearly half of the patients in the Mayo population [11] had obstructive HCM at enrollment, compared with only 26% of our population. Conversely, the inverse association between arterial hypertension and genotype positivity, already consistently reported [6,10,11], highlights the relevance of competing etiological pathways contributing to left ventricular hypertrophy in real-world clinical practice. The lower prevalence of hypertension among genotype-positive patients supports the concept that sarcomeric disease, particularly at onset in young adulthood, is less frequently accompanied by conventional cardiovascular risk factors, reinforcing the need for careful phenotypic contextualization when interpreting left ventricular hypertrophy in patients with suspected HCM.

Inclusion of LAVI in the current work only modestly improved the model performance but without leading to meaningful patient reclassification. Nonetheless, atrial remodeling markers may reflect genetic influences on disease expression beyond ventricular morphology. Supporting this, a recent study of early-onset atrial fibrillation patients (median age 47) found pathogenic or likely pathogenic variants in 20% of cases via NGS [20], with cardiomyopathy history, infranodal conduction disease, and elevated cardiac MRI T1 or late gadolinium enhancement being strong predictors of positive genetic tests. While ventricular phenotypes were common in dilated (54%) and arrhythmogenic (89%) cardiomyopathy variant carriers, only 40% of HCM variant carriers displayed overt ventricular phenotypes, aligning with preclinical HCM mutation carrier findings even without increased LV wall thickness or LA volume [21].

The association of HCM genes with severe isolated/predominant atrial phenotypes, while suggested, requires further investigation to elucidate the mechanisms underlying atrial myopathy in HCM [22]. Whether atrial remodeling is a primary genetic effect or an early marker of disease expression remains unclear. A systematic review [6] of 9486 HCM patients across 29 studies found no association between left atrial dimensions and genotype. Only one study [23] identified right atrial as an independent predictor of genotype positivity, predominantly sarcomeric variants, TNNT2 variants, and complex genotypes.

4.4. Mode of First Presentation

Neither arrhythmic presentation nor syncope or heart failure as the first clinical manifestation independently predicted genotype-positivity in our cohort. While three prior studies linked specific initial presentations—arrhythmic syncope [24], chest pain [25], and heart failure [26]—to genotype positivity, they lacked measures of model performance, limiting the interpretability and generalizability of these associations.

These findings support a conceptual distinction between arrhythmic burden or phenotypic severity at presentation and underlying genetic causation in HCM. While arrhythmias and heart failure-related symptoms remain central for prognostic stratification and clinical management, they appear to be poor standalone indicators of whether disease is driven by a monogenic sarcomeric variant at the individual patient level. The independent association between genotype positivity and a family history of sudden cardiac death, major ventricular arrhythmias, or non-ischemic cardiomyopathy suggested that inherited disease susceptibility is better captured by familial clustering of cardiomyopathic and arrhythmic phenotypes than by the individual patient’s initial presentation. Accordingly, a recent Swedish cohort also failed to demonstrate an association between symptoms at initial evaluations (mostly dyspnea and chest pain) and outcome [26]. Notably, coronary microvascular disfunction in HCM is highly prevalent and multifactorial [27].

5. Conclusions

Our findings suggest that genotype prediction in HCM should rely less on how patients present clinically and more on when the disease manifests and how it clusters within families. This perspective may help refine expectations around genetic testing, improve pre-test counseling, and avoid overinterpretation of arrhythmic or heart failure presentations as markers of genetic causality [28]. Future studies should clarify whether myocardial tissue characterization, particularly the presence, extent, or spatial distribution of myocardial fibrosis, can provide additional insight into the underlying genetic substrate rather than merely reflecting disease severity [29,30]. More broadly, these results further support the evolving view that the genetic architecture of HCM extends beyond the classical fully penetrant autosomal dominant model, encompassing reduced penetrance, variable expressivity, de novo variants, and potentially oligogenic contributions.

6. Limitations

The study’s pragmatic, real-world approach carries some potential limitations. First, family history variables were based on clinical reporting and were not systematically adjudicated at the genetic level. Also, we included reported family history of NICM, instead of restricting it to confirmed HCM, to minimize underreporting, potentially sacrificing specificity. Second, despite adherence to contemporary variant classification standards, reclassification of variants over time cannot be excluded and may have influenced genotype assignment. Third, although the multicenter design enhances generalizability across different clinical settings, external validation in independent cohorts will be required before routine clinical implementation. Fourth, the model was not intended to replace comprehensive genetic counseling but to complement clinical judgment by providing an objective estimate of genotype positivity, particularly to support discussions around testing prioritization and expectations. Finally, possible disease progression between initial diagnosis and enrollment/genetic analysis at referral may have influenced genotype-phenotype correlations.