Study Population
This is a retrospective, longitudinal study from the database of HCM subjects evaluated at Children’s Hospital of Philadelphia between 2003 to 2021. All subjects were aged between 0-21 years with primary diagnosis of HCM or referred for HCM screening. HCM was defined as per the American College of Cardiology /American Heart Association (ACC/AHA) guidelines for unexplained left ventricular hypertrophy in the absence of another cardiac or systemic disease [14]. The inclusion criteria for the subjects were the availability of ECG and echocardiograms from our institutional database. Subjects with HCM secondary to metabolic disorders, systemic hypertension or endocrine disorders were excluded.
The variants identified by genetic testing were classified as pathogenic or variant of uncertain significance (VUS) in genes associated with sarcomeric or non-sarcomeric cardiomyopathy. For our study, in the sarcomeric group subjects with pathogenic variants and subjects with VUS were grouped together to create a combined, variant-positive (V+P+) group, as described in previous studies [12]. They were grouped together because both groups had clear-cut clinically diagnosed HCM (including the VUS group). For statistical analysis, the V+P+ group was subdivided into subjects with actual pathogenic mutations in genes encoding sarcomere proteins (G+P+) and subjects with VUS (VUS+P+) (Figure 2).
The subclinical variant carriers (V+P-) were comprised of subjects positive for the gene variant (G+/VUS+) but without the presence of LV hypertrophy (P-)
All subjects in the non-sarcomeric group were previously diagnosed with Friedreich’s ataxia (FA) or RASopathies. The non-sarcomeric group includes only subjects with FA and RASopathies because they constituted the dominant causes of HCM in our institutional database. RASopathies consist of a group of phenotypically related disorders arising from variants in genes within the Ras/mitogen-activated protein kinase (Ras/MAPK) pathway and in our study consisted of subjects with Noonan syndrome and Noonan syndrome with multiple lentigines (NSML).
Classification of variants and genotyping
Per institutional practice, each patient had undergone either HCM panel testing or cardiomyopathy next generation sequencing panel testing. The updated American College of Medical Genetics and Genomics guidelines [15] were used to classify variants in HCM-associated genes. Variant analysis was completed as previously published [13]. The composition of representatives panel is shown in the Supplementary Table 1. Abstracted data from clinical notes were evaluated by a clinical geneticist, and any benign or likely benign variants were excluded. The impact of multiple VUS was not assessed, as this was not the focus of our study.
Variants in MYH7, MYBPC3, and TNNI3, TTN and TNNT2 (Troponin complex) were the most commonly identified. In addition, there were multiple variants identified in: FXN, RYR1, PTPN11, RAF1, SOS1, PKP2, PRKAG2, CASQ2, ACTC, LAMA, and ABCD. For subjects with FA, we took into consideration the number of repeats to further explain their genetic status. Genomic location of the variant and population frequency were obtained from a large genomic dataset (gnomAD version 3.0) [16].
Study outcomes
In our study, adverse outcomes were defined as occurrence of any one of the following events: appropriate insertion of Implantable Cardioverter Defibrillator (ICD), myomectomy, heart transplant and death. Based on a convention established in previous studies these individual outcomes were grouped together in a single group labeled as Major Adverse Cardiac Events (MACE). [11]
Echocardiography
LVH was defined as the maximal septal or LV posterior wall thickness exceeding the pediatric Boston z score of 2, corrected for age, gender, and body surface area (BSA), [17]. Other conventional measurements were abstracted from official clinical reports. In this study the primary role of echocardiography was to confirm the diagnosis of HCM and detect any LVOT obstruction.
Electrocardiogram
Conventional 15-lead resting ECGs were obtained in the supine position. All the electrocardiograms were re-analyzed by three senior cardiologists (AB, VLV, and PS) blinded to the clinical and echocardiographic data. In V+P+ and non-sarcomeric groups, the first ECG at the time of disease manifestation was included. The second ECG in this group was recorded when the patient showed maximal LV hypertrophy during follow-up. The ECGs at the first encounter and the last follow-up were included in the V+P- group. The criteria for defining ECG abnormalities are summarized in Table 1.
Leads V1-V4 were considered as anterior leads, leads I, V5, V6 and aVL were considered as lateral leads and II, III, aVF were considered as inferior leads. For our study, pathologic Q wave was defined as ≥ 4mm in two or more leads. This is an institutional preference and contrasts with the more restrictive definition of ≥ 3mm proposed by the Seattle criteria [18].
T wave abnormalities included T wave inversion, flat T, biphasic T, and notched T waves. Inverted T waves in lateral, or inferior leads were considered abnormal, but not in leads aVL, V1-V2. T wave inversion can vary with age, gender, race/ethnicity. LVH was defined by two pediatric-based criteria (Table1).
In this study, ECG features corresponding to repolarization abnormalities were extracted and grouped into two categories. For statistical analysis, subjects were classified having ST segment abnormalities, T-wave abnormalities, or grouped together as ST and T wave abnormalities, termed as “combined” repolarization abnormalities (defined in Table 1).
The study was approved by the Institutional Review Board of the Children's Hospital of Philadelphia.
Statistical Analysis
Normally distributed data are presented as mean ± standard deviation (SD), whereas nonparametric data are presented as median with interquartile ranges (IQRs). Unpaired samples were analyzed by the unpaired 2-tailed Student’s t-test, whereas the Mann-Whitney U test was utilized for nonparametric data. Logistic regression was used to evaluate the association between ECG features and MACE. The analysis was adjusted for age, sex, and race. Categorical variables were compared between groups using Fisher’s exact test. Timed outcome endpoints were assessed with Kaplan-Meier survival curves using univariate Cox regression to compare cumulative outcome rates between groups. The primary endpoint for Kaplan-Meier analysis was the occurrence of MACE. For all significance testing, a difference was considered significant if p value < 0.05. The statistical analysis was performed using StataCorp. 2013. Stata Statistical Software: Release 13. College Station, TX: StataCorp LP.