Abstract Background Left ventricular (LV) trabeculation are increasingly observed in patients with hypertrophic cardiomyopathy (HCM), but its clinical significance remains controversial. This study aims to clarify the characteristics of LV trabeculation and evaluate its prognostic value in HCM. Methods We evaluated 1028 patients with HCM undergoing cardiac magnetic resonance. For each patient, thickness of compacted and trabeculated myocardium was measured at 16 segments of LV. The extent of LV trabeculation was expressed as the maximal trabeculation/compaction (T/C) ratio in any of 16 segments. The primary endpoint was major adverse cardiovascular events (MACEs). The secondary endpoints were heart failure, thromboembolic events, and ventricular arrhythmias. There were 689 patients undergoing whole-exome sequencing. Results LV trabeculation predominantly located in midventricular-to-apical area, anterior, and lateral free walls in HCM. A greater extent of LV trabeculation was correlated with a higher prevalence of female, lower LV ejection fraction, and higher prevalence of extensive late gadolinium enhancement. During a median 4.8-year follow-up, a greater extent of LV trabeculation was associated with an increased risk of MACEs (adjusted hazard ratio [HR] 1.214, P = 0.005), heart failure (adjusted HR 1.372, P = 0.006), thromboembolic events (adjusted HR 1.242, P = 0.032), and ventricular arrhythmias (adjusted HR 1.240, P = 0.047). No gene was significantly associated with trabeculation at genome-wide level. Conclusions The distribution pattern of LV trabeculation was inhomogeneous and asymmetric in HCM. A greater extent of LV trabeculation was associated with poor prognosis. The progression of LV trabeculation might be the natural course of HCM. Graphical Abstract The distribution pattern of LV trabeculation in patients with HCM and the association between the extent of LV trabeculation and poor prognosis. CMR, cardiac magnetic resonance; HCM, hypertrophic cardiomyopathy; MACEs, major adverse cardiovascular events; LGE, late gadolinium enhancement; LV, left ventricular; LVEF, LV ejection fraction; WES, whole-exome sequencing [40]graphic file with name 12916_2025_4142_Figa_HTML.jpg Supplementary Information The online version contains supplementary material available at 10.1186/s12916-025-04142-7. Keywords: Hypertrophic cardiomyopathy, Left ventricular trabeculation, Prognosis, Magnetic resonance imaging, Whole-exome sequencing Background Left ventricular (LV) trabeculation clinically presents as a two-layered structure: an inner loose interwoven meshwork and an outer compact myocardium [[41]1]. Excessive trabeculation of the myocardium (termed “hypertrabeculation”) as newly recommended by the Task Force of the European Society of Cardiology [[42]2]) forms deep intratrabecular recesses that fill with blood directly from the ventricular cavity. This condition increases the risk of myocardial dysfunction, thromboembolic events, and arrhythmias [[43]1, [44]3–[45]5]. Currently, controversies exist regarding the concept of LV hypertrabeculation. Traditionally termed LV noncompaction, this nomenclature has faced significant challenges and is now discouraged by the latest guidelines due to insufficient evidence supporting the existence of a compaction process in the human heart [[46]6, [47]7]. Moreover, whether LV hypertrabeculation is an independent disease entity or merely represents a phenotypic trait remain a matter of debate. LV trabeculation is frequently observed by imaging studies in association with cardiac diseases of different origins, including hypertrophic cardiomyopathy (HCM) [[48]6, [49]8]. Clarifying the prognostic significance of LV trabeculation in patients with HCM is necessary because of the effect of hypertrabeculation on multiple cardiovascular risks. Authors of previous studies have made inconsistent conclusions on the prognostic significance of LV trabeculation in patients with HCM [[50]9–[51]11], but these studies are limited by their small sample size and inadequate analysis for all segments of the LV in detail. Cardiac magnetic resonance (CMR) imaging is the preferred method for identifying LV trabeculation because of better visualization of the trabeculae [[52]12, [53]13]. In this study, we comprehensively evaluated trabeculation of the LV by CMR in a large cohort of adult HCM. We aimed to determine the characteristics of LV trabeculation in patients with HCM, and whether the extent of LV trabeculation affects the prognosis. Furthermore, the majority of patients underwent whole-exome sequencing to investigate the etiology of LV trabeculation in patients with HCM. Methods Study population This observational study recruited 1064 patients with HCM who underwent CMR at Fuwai Hospital, Chinese Academy of Medical Sciences, between 2009 and 2022. HCM was confirmed by echocardiographic and/or CMR demonstration of LV hypertrophy with a maximal LV wall thickness of ≥ 15 mm (or ≥ 13 mm in patients with a family history of HCM) in the absence of any other cardiac or systemic disease capable of producing such a magnitude of hypertrophy (e.g., uncontrolled hypertension, cardiac valve disease, and phenocopies) [[54]14]. Patients with poor image quality (n = 20) and patients who did not participate in any phase of follow-up (n = 16) were excluded, and 1028 patients were included in the final analysis. A flowchart of this study is shown in Fig. [55]1. Fig. 1. Fig. 1 [56]Open in a new tab Flowchart of this study. CMR, cardiac magnetic resonance; HCM, hypertrophic cardiomyopathy; MACEs, major adverse cardiovascular events Informed consent was obtained from all patients in accordance with the principles of the Declaration of Helsinki. The study was approved by the Ethics Committee of Fuwai Hospital. Measurement of LV trabeculation CMR evaluation was performed on a 1.5 T scanner (Magnetom Avanto, Siemens Healthcare) or a 3 T scanner (Magnetom Skyra, Siemens; Ingenia, Philips; and Discovery MR750, GE Healthcare) using electrocardiographic and respiratory gating. Typical imaging parameters were as follows: 3 T scanner, repetition time (ms)/echo time (ms) of 2.9–3.4/1.5–1.7; and 1.5 T scanner, 2.8–3.0/1.1–1.5. All CMR images were analyzed using commercially available postprocessing software (Argus, Siemens; and CVI 42 v.5.9.0, Circle Cardiovascular Imaging Inc.) Three long-axis (two, three, and four chambers) cine images at end-diastole were used for measuring the thickness of compacted and trabeculated myocardium at the center of 16 segments of the American Heart Association model. A compacted myocardium was defined as a myocardial layer of homogeneous moderate signal intensity that was distinctly separate from the blood pool. A trabeculated myocardium was defined as a meshwork of moderate signal intensity on the endocardial surface of the compacted myocardium with interspersion of higher signal intensity from the blood pool. Papillary muscles were excluded from the measurements. Short-axis views were used in conjunction with long-axis images to aid the identification of papillary muscles. Segment 17 (apex) was typically thin with a prominent trabeculation, and was thus excluded from analysis. LV end-diastolic volume index (LVEDVi), LV end-systolic volume index and LV ejection fraction (LVEF) were determined using an Argus workstation (Siemens Medical Solutions) and short-axis cine images. Using a validated method, the endocardial border was drawn to include papillary muscle and exclude LV trabeculation to calculate LV volumes [[57]15, [58]16]. In this study, we used the trabeculation to compaction (T/C) ratio instead of the noncompaction to compaction ratio because “noncompaction” is currently controversial. The ratio of T/C was calculated for each segment. The extent of LV trabeculation was expressed as the maximal T/C ratio in any of the 16 segments of the LV. Median, tertiles and quartiles of the maximal T/C ratio were utilized to categorize patients with HCM into different extents of LV trabeculation. Follow-up and outcomes The follow-up lasted from the first evaluation until death from any cause or the last known contact date. All patients were followed up annually until December 2023 by clinic visit or telephone interview. The primary endpoint of this study was major adverse cardiovascular events (MACEs), which were a composite of cardiovascular death, unplanned hospitalization for heart failure (HF), heart transplantation, non-fatal stroke, resuscitated cardiac arrest, ventricular fibrillation, sustained ventricular tachycardia, and appropriate implantable cardioverter-defibrillator therapy. The secondary endpoints of this study were as follows: (1) HF (i.e., HF-related death, unplanned hospitalization for HF, and heart transplantation), (2) thromboembolic events (i.e., embolic stroke-related death and non-fatal embolic stroke), and (3) ventricular arrhythmias (i.e., sudden cardiac death, resuscitated cardiac arrest, ventricular fibrillation, sustained ventricular tachycardia, and appropriate implantable cardioverter-defibrillator therapy). Detection of genetic variants associated with LV trabeculation A total of 689 patients underwent whole-exome sequencing as previously described [[59]17]. Gene-based burden test for rare variants was performed in the study. Rare variants were defined by a cutoff of a minor allele frequency smaller than 0.0001 across population studies, including the 1000 Genomes Project, NHLBI Exome Sequencing Project Database, Genome Aggregation Database, and Exome Aggregation Consortium databases. Regarding the rare variants, we then filtered for missense, nonsense, splicing, or frameshift variants with a Genomic Evolutionary Rate Profiling score > 2.0 and a Combined Annotation Dependent Depletion score > 15 [[60]18, [61]19]. These variants were aggregated into gene sets, and the Spearman correlation test was used to assess the associations between the maximal T/C ratio and the carrier status. Genes in which the carrier status was associated with the maximum T/C ratio with raw P < 0.05 were used to perform Kyoto Encyclopedia of Genes and Genomes pathway analysis by ClusterProfiler Package (version 4.10.0 in R). Enriched Kyoto Encyclopedia of Genes and Genomes pathways were determined on the basis of a false discovery rate of < 0.05. In addition, eight sarcomeric genes (MYH7, MYBPC3, TNNT2, TNNI3, MYL2, MYL3, TPM1, and ACTC1) were evaluated. Variants were classified by pathogenicity as pathogenic (P), likely pathogenic (LP), variant of unknown significance (VUS), likely benign (LB), or benign (B) according to American College of Medical Genetics and Genomics criteria. A genotype-positive state was defined as presence of any P or LP variants in the above-mentioned sarcomeric genes. Statistical analysis Statistical analyses were performed using R (Version 4.3.2). Continuous data are expressed as the mean ± standard deviation or median with interquartile range. Categorical variables are expressed as the total number (percentage). The Student t test or Mann–Whitney U test was used to compare continuous variables, and the chi-square test was used to compare categorical variables. Univariable and multivariable linear regression analyses were used to evaluate the relationships between LV trabeculation and clinical parameters. Variables demonstrating significant associations with the maximal T/C ratio in the univariable models were subsequently incorporated into the multivariable model. In survival analyses, univariable and multivariable Cox proportional hazards regression analyses were performed to calculate the hazard ratios (HRs) and 95% confidence intervals (CIs), and to evaluate the associations between LV trabeculation and outcomes. A competing risk analysis was used to evaluate the associations between LV trabeculation and the secondary endpoints, where death not related to each secondary endpoint was modelled as a competing event. Multivariable models were adjusted for age, sex, New York Heart Association functional class, syncope, atrial fibrillation, non-sustained ventricular tachycardia, coronary artery disease, valvular heart disease, carotid artery disease, chronic kidney disease, history of stroke, family history of sudden cardiac death, treatment strategies (septal reduction therapy or medical therapy alone), maximal LV wall thickness, LVEF, left atrial diameter, LV end-diastolic diameter, LVEDVi, LV outflow tract gradient, late gadolinium enhancement (LGE) extent, and LV apical aneurysm. A collinearity test was conducted to detect the presence of collinearity in the multivariable model. No collinearity (defined as coefficient of correlation greater than 0.7) was found amongst the variables in our multivariable models (Additional file 1: Figure S1). Survival curves were constructed using the Kaplan–Meier method to compare the cumulative incidence of the primary and secondary endpoint between patients with HCM with low and high extent of LV trabeculation, as categorized by the median maximal T/C ratio using the log-rank test. Additional survival curves were constructed to assess the cumulative incidence of the study endpoints between patients with HCM with and without extensive LGE (defined as LGE ≥ 15% of the myocardial mass). Two-sided P-values < 0.05 were considered statistically significant. Results Study population characteristics A total of 1028 patients with HCM were analyzed in this study (Table [62]1). The mean age of the study population was 47.2 ± 14.3 years and 64.7% were men. A New York Heart Association functional class greater than I was observed in 670 (65.2%) patients, and 21 (2.0%) patients presented with a reduced LVEF (< 50%). A positive family history of sudden cardiac death was observed in 113 (11.0%) patients. Table 1. Clinical characteristics in 1028 patients with HCM Total HCM with MACEs HCM without MACEs P-value N = 1028 N = 119 N = 909 Age (years) 47.2 ± 14.3 52.3 ± 15.2 46.5 ± 14.1  < 0.001 Sex (male), n (%) 665 (64.7) 72 (60.5) 593 (65.2) 0.310 Maximal T/C ratio 1.47 (0.98-2.10) 1.77 (1.22-2.33) 1.44 (0.94-2.07) 0.001 Syncope, n (%) 132 (12.8) 17 (14.3) 115 (12.7) 0.616 NYHA greater than I, n (%) 670 (65.2) 92 (77.3) 578 (63.6) 0.003 Atrial fibrillation, n (%) 187 (18.2) 39 (32.8) 148 (16.3)  < 0.001 NSVT, n (%) 95 (9.2) 17 (14.3) 78 (8.6) 0.043 Hypertension, n (%) 342 (33.3) 42 (35.3) 300 (33.0) 0.618 Diabetes, n (%) 72 (7.0) 10 (8.4) 62 (6.8) 0.525 Hyperlipidemia, n (%) 259 (25.2) 36 (30.3) 223 (24.5) 0.177 Coronary artery disease, n (%) 105 (10.2) 52 (10.1) 53 (10.3) 0.918 Valvular heart disease, n (%) 45 (4.4) 25 (4.9) 20 (3.9) 0.446 Carotid artery disease, n (%) 53 (5.2) 26 (5.1) 27 (5.3) 0.888 Chronic kidney disease, n (%) 14 (1.4) 6 (1.2) 8 (1.6) 0.590 Stroke history, n (%) 19 (1.8) 6 (1.2) 13 (2.5) 0.105 Septal reduction therapy, n (%) 412 (40.1) 40 (33.6) 372 (40.9) 0.126 Family history of SCD, n (%) 113 (11.0) 21 (17.6) 92 (10.1) 0.014 Echocardiography and CMR parameters  LVWT (mm) 23.4 ± 5.6 24.6 ± 5.7 23.3 ± 5.6 0.019  LVEDD (mm) 44.0 ± 5.8 45.0 ± 6.7 43.9 ± 5.6 0.073  LVEDVi (mL/m^2) 74.2 (63.8–86.4) 73.4 (61.3–86.1) 75.1 (64.2–86.8) 0.116  LVEF (%) 67.4 ± 7.3 64.9 ± 9.4 67.7 ± 6.9 0.002  LVEF < 50%, n (%) 21 (2.0) 11 (9.2) 10 (1.1)  < 0.001  LAD (mm) 42.0 ± 7.2 44.8 ± 8.6 41.6 ± 6.9  < 0.001  LVOTG (mmHg) 34.0 (8.9–74.0) 31.0 (6.8–65.0) 35.0 (9.0–75.0) 0.046  LVOTO, n (%) 542 (52.7) 60 (50.4) 482 (53.0) 0.592  LV apical aneurysm, n (%) 12 (1.2) 4 (3.4) 8 (0.9) 0.055  LGE extent 3.1 (0.0–7.7) 5.8 (2.3–14.1) 2.8 (0.0–7.1)  < 0.001 [63]Open in a new tab CMR cardiac magnetic resonance, HCM hypertrophic cardiomyopathy, LAD left atrial diameter, LGE late gadolinium enhancement, LV left ventricular, LVEDD LV end-diastolic diameter, LVEDVi LV end-diastolic volume index, LVEF LV ejection fraction, LVOTG LV outflow tract (LVOT) gradient, LVOTO LVOT obstruction, defined as LVOTG ≥ 30 mmHg, LVWT, maximal left ventricular (LV) wall thickness, MACEs major adverse cardiovascular events, NSVT non-sustained ventricular tachycardia, NYHA New York Heart Association, SCD sudden cardiac death, T/C trabeculation to compaction Evaluation of the extent of LV trabeculation at CMR In our segment-based assessment of trabeculation, we found that a trabeculated myocardium was present in all 16 LV segments in HCM and was more commonly identified in the mid-anterior (segment 7), mid-anterolateral (segment 12), apical-anterior (segment 13), and apical-lateral (segment 16) segments (Fig. [64]2A and B). The myocardium in the septum and base was less likely to be trabeculated. Fig. 2. [65]Fig. 2 [66]Open in a new tab Visualization of segmental extent of LV trabeculation in patients with hypertrophic cardiomyopathy. A Histogram shows percentages of presence of trabeculation in each segment in the LV. B Diagram shows the distribution of presence of trabeculation in the LV. C Histogram shows percentages of the maximal T/C ratio in each segment in the LV. D Diagram shows the distribution of the maximal T/C ratio in the LV. LV, left ventricular or ventricle; T/C, trabeculation to compaction A quantitative analysis showed that the median maximal T/C ratio in all 1028 patients was 1.47 (IQR 0.98–2.10). The distribution of segmental locations of the maximal T/C ratio is shown in Fig. [67]2C and D. The maximal T/C ratio was mainly located at the mid-anterior (segment 7), mid-anterolateral (segment 12), and apical-lateral (segment 16) segments. Similarly, the maximal T/C ratio seldom appeared at the septum and base. The multivariable linear regression showed that a higher maximal T/C ratio was independently associated with a higher prevalence of female sex, lower LVEF, and higher prevalence of extensive LGE (Table [68]2 and Fig. [69]3). Table 2. Univariable and multivariable linear regression analyses of the relationship between LV trabeculation and clinical parameters Variants Univariable linear regression Multivariable linear regression Unadjusted β value (95% CI) Unadjusted P-value Adjusted β value (95% CI) Adjusted P-value Age  − 0.002 (− 0.007–0.002) 0.324 Sex (female vs. male) 0.172 (0.031–0.313) 0.017 0.197 (0.053–0.340) 0.007 Syncope  − 0.002 (− 0.204–0.200) 0.984 NYHA greater than I  − 0.070 (− 0.212–0.072) 0.332 Atrial fibrillation  − 0.112 (− 0.287–0.063) 0.208 NSVT  − 0.169 (− 0.402–0.064) 0.155 Coronary artery disease 0.100 (− 0.123–0.323) 0.378 Valvular heart disease 0.069 (− 0.261–0.399) 0.682 Carotid artery disease 0.061(− 0.366–0.245) 0.697 Chronic kidney disease 0.048 (− 0.534–0.631) 0.871 Stroke history 0.079 (− 0.580–0.422) 0.757 Septal reduction therapy  − 0.004 (− 0.141–0.134) 0.958 Family history of SCD 0.246 (0.030–0.461) 0.025 0.172 (− 0.043–0.386) 0.117 LVWT  − 0.005 (− 0.017–0.007) 0.456 LVEDD 0.001 (−0.001–0.012) 0.920 LVEDVi 0.002 (− 0.001–0.005) 0.163 LVEF < 50% 0.891 (0.416–1.365)  < 0.001 0.515 (0.013–1.016) 0.044 LAD  − 0.002 (− 0.012–0.007) 0.630 LVOTG  − 0.002 (− 0.004 to − 4.210e-4) 0.014  − 0.002 (− 0.004 to − 1.600e-4) 0.032 LV apical aneurysm 0.150 (− 0.479–0.778) 0.640 Extensive LGE 0.434 (0.207–0.662)  < 0.001 0.327 (0.089–0.565) 0.007 [70]Open in a new tab HCM hypertrophic cardiomyopathy, LAD left atrial diameter, LGE late gadolinium enhancement, extensive LGE defined as LGE extent ≥ 15%, LV left ventricular, LVEDD, LV end-diastolic diameter, LVEDVi LV end-diastolic volume index, LVEF LV ejection fraction, LVOTG LV outflow tract (LVOT) gradient, LVWT maximal left ventricular (LV) wall thickness, NSVT non-sustained ventricular tachycardia, NYHA New York Heart Association, SCD sudden cardiac death Fig. 3. [71]Fig. 3 [72]Open in a new tab Typical cardiac magnetic resonance scans of patients with hypertrophic cardiomyopathy with heavy burden of trabeculation. This is a 49-year-old female patient with hypertrophic cardiomyopathy, with a maximal T/C ratio of 5, a left ventricular ejection fraction of 33%, and LGE extent of 33%. A cine image in 2-ch, B LGE image in 2-ch, C cine image in 4-ch, D LGE image in 4-ch. LGE, late gadolinium enhancement; T/C, trabeculation to compaction Prognostic value of LV trabeculation in patients with HCM During the median follow-up of 4.8 years (IQR 3.1–6.6 years), 119 (11.6%) patients had MACEs, of which 35 (29.4%) were HF, 30 (25.2%) were thromboembolic events, and 39 (32.8%) were ventricular arrhythmias. Multivariable Cox proportional hazards regression analyses showed that a greater extent of LV trabeculation was associated with a significantly increased risk of MACEs (per 1 increase in the maximal T/C ratio: adjusted HR 1.214, 95% CI 1.059–1.391, P = 0.005). A greater extent of LV trabeculation was also associated with a significantly increased risk of HF (per 1 increase in the maximal T/C ratio: adjusted HR 1.372, 95% CI 1.094–1.722, P = 0.006), thromboembolic events (per 1 increase in the maximal T/C ratio: adjusted HR 1.242, 95% CI 1.019–1.514, P = 0.032), and ventricular arrhythmias (per 1 increase in the maximal T/C ratio: adjusted HR 1.240, 95% CI 1.003–1.533, P = 0.047) (Table [73]3). The Kaplan–Meier curves showed that patients who had greater extent of LV trabeculation (as categorized by the median value) had a higher cumulative incidence of MACEs, HF, thromboembolic events, and ventricular arrhythmias (log-rank, P < 0.001, and P = 0.003, 0.024, 0.010, respectively) (Fig. [74]4). Moreover, the multivariable Cox proportional hazards regression analyses also showed a trend of higher risk of MACEs in patients with HCM with greater median value, tertiles and quartiles of LV trabeculation (Additional file 1: Table S1). Table 3. Cox regression analysis of the association between the characteristics and outcomes in patients with HCM Variants Unadjusted HR (95% CI) Unadjusted P-value Adjusted HR (95% CI) Adjusted P-value Primary endpoint  MACEs Per 1 increase in the maximal T/C ratio 1.250 (1.109–1.410)  < 0.001 1.214 (1.059–1.391) 0.005 Secondary endpoints  HF Per 1 increase in the maximal T/C ratio 1.308 (1.096–1.562) 0.003 1.372 (1.094–1.722) 0.006  Thromboembolic events Per 1 increase in the maximal T/C ratio 1.322 (1.083–1.612) 0.006 1.242 (1.019–1.514) 0.032  Ventricular arrhythmias Per 1 increase in the maximal T/C ratio 1.275 (1.089–1.493) 0.002 1.240 (1.003–1.533) 0.047 [75]Open in a new tab Models were adjusted for age, sex, New York Heart Association functional class, syncope, atrial fibrillation, non-sustained ventricular tachycardia, coronary artery disease, valvular heart disease, carotid artery disease, chronic kidney disease, history of stroke, family history of sudden cardiac death, treatment strategies (septal reduction therapy or medical therapy alone), maximal left ventricular (LV) wall thickness, left ventricular ejection fraction, left atrial diameter, LV end-diastolic diameter, LV end-diastolic volume index, LV outflow tract gradient, LV apical aneurysm and late gadolinium enhancement extent CI confidential interval, HCM hypertrophic cardiomyopathy, HF heart failure, HR hazard ratio, MACEs major adverse cardiovascular events, T/C trabeculation to compaction Fig. 4. [76]Fig. 4 [77]Open in a new tab Kaplan–Meier survival analysis of patients with hypertrophic cardiomyopathy. Kaplan–Meier curves for A MACEs, B HF, C thromboembolic events and D ventricular arrhythmias. HF, heart failure; MACEs, major adverse cardiovascular events Furthermore, we evaluated the associations of trabeculation in individual myocardial regions with outcomes. Our results showed that a greater extent of LV trabeculation in distinct regions (i.e., apical area and inferior free wall) was associated with a higher risk of MACEs. Additionally, a greater LV apical trabeculation extent was also associated with a higher risk of HF and thromboembolic events; while the risk of ventricular arrhythmias was mostly related to the extent of LV midventricular trabeculation (Table [78]4). Table 4. Associations between LV regional trabeculation and outcomes in patients with HCM Variants MACEs HF Thromboembolic events Ventricular arrhythmias Adjusted HR (95% CI) Adjusted P-value Adjusted HR (95% CI) Adjusted P-value Adjusted HR (95% CI) P-value Adjusted HR (95% CI) Adjusted P-value Extent of LV trabeculation, regional analysis  Base 1.134 (0.943–1.363) 0.182 1.193 (0.857–1.662) 0.296 1.194 (0.726–1.964) 0.484 1.167 (0.860–1.583) 0.321  Midventricular area 1.103 (0.865–1.406) 0.430 1.010 (0.551–1.854) 0.973 0.903 (0.649–1.257) 0.546 1.633 (1.155–2.309) 0.006  Apical area 1.221 (1.072–1.391) 0.003 1.381 (1.114–1.711) 0.003 1.240 (1.004–1.531) 0.046 1.213 (0.965–1.526) 0.098  Septum 1.192 (0.947–1.502) 0.135 1.381 (0.927–2.059) 0.113 1.181 (0.766–1.821) 0.451 1.388 (1.019–1.890) 0.037  Anterior free wall 1.107 (0.917–1.337) 0.289 1.415 (1.072–1.868) 0.014 1.213 (0.859–1.713) 0.274 1.119 (0.854–1.466) 0.416  Inferior free wall 1.271 (1.103–1.465)  < 0.001 1.340 (1.064–1.687) 0.013 1.383 (1.096–1.746) 0.006 1.188 (0.884–1.596) 0.253  Lateral free wall 1.123 (0.959–1.314) 0.151 1.218 (0.883–1.681) 0.230 0.979 (0.772–1.241) 0.860 1.208 (0.946–1.542) 0.130 [79]Open in a new tab Models were adjusted for age, sex, New York Heart Association functional class, syncope, atrial fibrillation, non-sustained ventricular tachycardia, coronary artery disease, valvular heart disease, carotid artery disease, chronic kidney disease, history of stroke, family history of sudden cardiac death, treatment strategies (septal reduction therapy or medical therapy alone), maximal left ventricular (LV) wall thickness, left ventricular ejection fraction, left atrial diameter, LV end-diastolic diameter, LV end-diastolic volume index, LV outflow tract gradient, LV apical aneurysm and late gadolinium enhancement extent In addition, to further assess the direct impact of trabeculation on the study endpoints, we evaluated the independent association between the thickness of both the trabecular and the compact layer with study endpoints. Multivariable Cox models demonstrated that higher thickness of the trabecular layer increased the risk of MACEs (adjusted HR 1.074, 95% CI 1.017–1.134, P = 0.010), thromboembolic events (adjusted HR 1.103, 95% CI 1.034–1.176, P = 0.003) and ventricular arrhythmias (adjusted HR 1.141, 95% CI 1.040–1.251, P = 0.005), while the compact layer thickness modestly increased the risk of MACEs only (Additional file 1: Table S2). Association between LV trabeculation and LGE In our patients, correlation analyses showed a positive association between the trabecular and compact layer and the extent of LGE (Additional file 1: Figure S2). As LGE is a well-established risk factor in patients with HCM, Kaplan–Meier curves further revealed that extensive LGE was associated with higher incidence of MACEs, HF and ventricular arrhythmias in patients with HCM (log-rank, P < 0.001, P = 0.024, and P = 0.003, respectively) (Additional file 1: Figure S3). To clarify the prognosis of LV trabeculation in patients with HCM independent of LGE, we incorporated LGE extent in the Cox models. The results confirmed that LV trabeculation remained an independent risk factor for the primary and secondary endpoints after adjusting for the LGE extent. Genetic associations of LV trabeculation To obtain more insight into the etiology of LV trabeculation in patients with HCM, we performed a gene-based burden test for rare variants associated with the extent of LV trabeculation. The burden test did not find any gene significantly associated with the maximal T/C ratio in patients with HCM at genome-wide level. However, interestingly, the gene-based burden test identified a total of 287 genes correlated with the maximal T/C ratio with raw P < 0.05 (Additional file 1: Table S3). Further pathway enrichment analysis showed that these genes were predominantly enriched in the “hypertrophic cardiomyopathy,” “dilated cardiomyopathy,” and “arrhythmogenic right ventricular cardiomyopathy” pathways (Fig. [80]5). Fig. 5. [81]Fig. 5 [82]Open in a new tab Pathway enrichment analysis of trabeculation-associated genes. KEGG, Kyoto Encyclopedia of Genes and Genomes In addition, the prevalence of positive HCM genotype (defined as carrying at least one P/LP variants of eight sarcomeric genes, MYH7, MYBPC3, TNNT2, TNNI3, MYL2, MYL3, TPM1, and ACTC1) was 35.2% among our cohort. In line with abovementioned genome-wide result, analysis showed that HCM genotype was significantly associated with the maximal T/C ratio (maximal T/C ratio of HCM genotype-positive vs. genotype-negative: 1.61 [IQR 1.00–2.36] vs. 1.31 [IQR 0.90–1.84], P < 0.001). Yet, it should be noted that sarcomeric variants was not associated with adverse prognosis in HCM (Additional file 1: Table S4). Discussion Increased LV trabeculation is a not-so-infrequent finding in patients with HCM. Previous studies have demonstrated that LV trabeculation can be detected in over 80% of patients with HCM, with 40% exhibiting extensive trabeculation [[83]8, [84]20]. However, the clinical significance of LV trabeculation remains largely unknown. This study quantified the 16-segment extent of LV trabeculation for 1028 patients with HCM using CMR. To date, this is the largest cohort in the field of myocardial trabeculation in HCM. We showed that the distribution pattern of LV trabeculation was inhomogeneous and asymmetric in patients with HCM. A greater extent of LV trabeculation was associated with adverse outcomes, including MACEs, HF, thromboembolic events, and ventricular arrhythmias. More importantly, the gene-based burden test showed that the LV trabeculation in HCM was not correlated with a mutation phenotype. The distribution of trabeculated myocardium has certain patterns in patients with HCM. We found that trabeculation was mainly located in the mid-cavity, lateral free wall, and apical area of the LV, and these were also the most common regions where the maximal T/C ratio was located. In contrast, the septal and basal regions were the least likely to be trabeculated. The mechanism underlying this distribution pattern of trabeculation is still controversial. One explanation of this mechanism is based on the sequential compaction of the trabeculated layer from base to apex, and from the septal to the lateral wall [[85]21]. Trabeculation is considered to be caused by the arrest of compaction; therefore, it is more likely to occur in regions where the compaction process is latterly involved [[86]8, [87]20]. However, this explanation has been recently challenged because a trabeculated and compacted myocardium shows an independent growth pattern with different rates [[88]6, [89]22, [90]23]. Additionally, this explanation does not explain trabeculation in adults whose cardiomyocytes are terminally differentiated without embryonic identity, where the formation of new cardiac structure is unlikely [[91]2, [92]6, [93]7]. Further studies are required to address the mechanism behind the phenomenon that trabeculation preferentially involves certain regions of the LV in the adult population. There have been ongoing debates on the prognostic significance of LV trabeculation. While studies in healthy population and athletes did not show a definite association between LV trabeculation and unfavorable outcome, trabeculation in the context of myocardial disorders is a different picture. HF, thromboembolic events, and ventricular arrhythmias are the most frequent cardiovascular complications related to trabeculation [[94]1, [95]3–[96]5]. A meta-analysis by Aung et al. showed that the prognosis of patients with hypertrabeculation was comparable to that of patients with dilated cardiomyopathy [[97]5]. Three previous studies reported the prognostic value of trabeculation in adult patients with HCM, but their conclusions were inconsistent. Casanova et al. [[98]10] found that a higher extent of trabeculated myocardium did not increase the incidence of cardiovascular mortality in 211 patients with HCM. In contrast, Kawamura et al. [[99]11] reported that a higher extent of trabeculated myocardium was associated with a higher incidence of cardiovascular events, including cardiovascular mortality, admission for worsening HF, and sustained ventricular tachycardia/fibrillation, in their HCM population (n = 172). Wang et al. [[100]24] analyzed 378 patients with HCM and showed that patients with increased apical trabeculation as quantified by fractal analysis had a higher risk of mortality. These previous studies had small sample sizes without detailed segmental evaluation, and they also did not investigate the association between trabeculation and several highly trabeculation-related endpoints, such as thromboembolic events. Our study enrolled a relatively large number of patients with HCM with detailed segmental trabeculation measurements. We found that an overall greater extent of LV trabeculation was associated with an increased risk of MACEs, HF, thromboembolic events, and ventricular arrhythmias in patients with HCM. In addition, LV trabeculation in different regions carried different types of risk. Since the management of hypertrabeculation has been a debated topic yet, our results suggest that more attention should be paid to patients with HCM who have a relatively high extent of trabeculation, and anticoagulation therapy may be beneficial because of their increased thromboembolic risk. Ongoing fibrosis, impaired systolic function, and thrombi formation within the recesses might be the underlying reasons accounting for worse outcomes in patients with increased trabeculation. However, the relationship between trabeculation and myocardial fibrosis has been controversial. While the study of Casanova et al. identified an inverse correlation between LGE and trabeculation, our study showed that a greater extent of LV trabeculation was significantly associated with lower systolic function and more severe fibrosis, which was consistent with Captur et al., who found a positive correlation between the extent of myocardial scarring and fractal dimension (an alternative indicator of LV trabeculation) [[101]10, [102]25]. Cardiac fibrosis can be increased by impaired myocardial perfusion [[103]1]. Sub-endocardial ischemia may stimulate the development of trabeculation [[104]1]. Yet the causal association between trabecular myocardium and myocardial fibrosis has not been elucidated, and controversies exist regarding the myocardial perfusion status in the trabecular and compact layer. Future study is required to explore this puzzle. Additionally, thrombi formation attributable to sluggish blood flow in the deep intratrabecular recesses not only increases the risk of thromboembolic events, but has also been speculated to play a role as a substrate for arrhythmias through local inflammatory response and mechanical stimulation of the ventricular wall [[105]1, [106]26]. Further research is needed to better elucidate the mechanism of arrhythmias in patients with LV trabeculation. The classification of LV hypertrabeculation has gained great interest since it was first defined. Whether the excessive trabeculation is merely a phenotypic trait or a cardiomyopathy with its distinct disease origins is still controversial. Genetic analysis is an effective approach that helps to determine the underlying etiology of LV hypertrabeculation. However, our gene-based burden test did not find any evidence supporting the LV hypertrabeculation to be an isolated disease in HCM. When the significance threshold was relaxed to 0.05, the pathway enrichment analysis showed that LV trabeculation-associated mutated genes were predominantly enriched in the pathways of HCM, dilated cardiomyopathy, and arrhythmogenic right ventricular cardiomyopathy. This intriguing result suggests that the extent of LV trabeculation in HCM is disease context-dependent. Subsequent analysis of sarcomeric variants revealed that LV trabeculation was associated with the intrinsic disease background of HCM. Therefore, the progression of LV trabeculation might be secondary to the natural course of established cardiomyopathies. Moreover, the results of survival analyses have provided new perspective on the role of LV trabeculation in patients with HCM. We propose that presence of trabeculation in adult HCM should not be viewed as a separate disease entity with a definite diagnostic threshold. Instead, it should be treated as a quantitative characteristic of HCM itself. Our findings suggest that LV trabeculation, at least in most patients with HCM, tends to be a phenotypic trait rather than a concomitant cardiomyopathy, which supports the new viewpoint from the most recent European Society of Cardiology guidelines [[107]2]. Conclusions In conclusion, although there is controversy regarding the concept of hypertrabeculation, the morphological manifestation of myocardial trabeculation is well recognized. The extent of trabeculation is closely related to MACEs, HF, thromboembolic events, and ventricular arrhythmias in patients with HCM. Therefore, LV hypertrabeculation requires clinical attention once such a phenomenon is observed in HCM. We speculate that avoiding additional stress on the myocardium, such as avoiding strenuous physical activity and using anticoagulant drugs, may help to improve the prognosis. More importantly, our findings suggest that LV trabeculation in patients with HCM tends to be a phenotypic trait following pathophysiological changes in HCM, and this is consistent with the new viewpoint in current guidelines. Limitations There are some limitations in this study. First, this was a single-center study that involved patients from a tertiary referral hospital, which may have limited the generalizability of the findings. Second, studies regarding LV trabeculation are controversial. Although our results strived to clearly and objectively demonstrate the clinical significance of LV trabeculation, our cohort did not have data on the dynamic alteration of the trabecular and compact layer over the natural course of HCM or myocardial perfusion of the two layers. In addition, the presence of a thrombus in the trabecular layer was not assessed. Another limitation was that results of the genetic analyses were exploratory, but current findings pointed out potential direction for future studies and should be tested in larger cohorts. Finally, whole-genome sequencing was not performed in all of our patients. Therefore, the possible functions of introns could not be investigated. Supplementary Information [108]12916_2025_4142_MOESM1_ESM.docx^ (466.9KB, docx) Additional file 1: Figure S1–S3, Table S1–S4. Figure S1. The correlation coefficient analysis between variables. Table S1. Association between LV trabeculation extent and endpoints in 1028 patients with HCM. Table S2. Association between trabecular and compact layer with endpoints in 1028 patients with HCM. Figure S2. Correlation between the thickness of trabecular layer and compact layer with LGE extent. Figure S3. Kaplan-Meier curves of extensive LGE in 1028 patients with HCM. Table S3. Genes associated with LV trabeculation in patients with HCM with raw P <0.05. Table S4. Association between sarcomeric mutations and endpoints in patients with HCM. Acknowledgements