Summary
Background
Treatment of children with hypertrophic cardiomyopathy might be improved if the risk of death or heart transplantation could be predicted by risk factors present at the time of diagnosis.
Methods
The Pediatric Cardiomyopathy Registry collected longitudinal data on 1085 children with hypertrophic cardiomyopathy from 1990 to 2006. The primary outcome was death or heart transplantation. Our goal is to understand how patient factors measured at the time of diagnosis will predict the subsequent risk of death or heart transplantation. The Kaplan-Meier method was used to calculate time-to-event rates from the time of diagnosis to the earlier of heart transplantation or death for children in each subgroup. Cox proportional-hazards regression was used to identify both univariable and multivariable predictors of death or heart transplantation within each aetiologic subgroup.
Findings
The poorest outcomes (death or transplant) were observed among children with inborn errors of metabolism, for whom the estimated rate of death or heart transplantation was 57% (95% CI: 45%, 69%) at 2 years. Children with mixed functional phenotypes also did poorly, with rates of death or heart transplantation of 45% (95% CI: 33%, 57%) at 2 years for children with mixed hypertrophic and dilated cardiomyopathy, and 38% (95% CI: 24%, 52%) at 2 years for children with mixed hypertrophic and restrictive cardiomyopathy. Excellent outcomes were observed among the 407 children who received a diagnosis of idiopathic hypertrophic cardiomyopathy at 1 year of age or older, with rates of death or heart transplantation of 3% (95% CI: 1%, 5·0%) at 2 years.
The risk factors for poor outcomes varied according to hypertrophic cardiomyopathy subgroup, but they generally included age, weight, congestive heart failure, lower left ventricular (LV) fractional shortening, or higher LV end-diastolic posterior wall thickness or end-diastolic ventricular septal thickness at the time of cardiomyopathy diagnosis. For all hypertrophic cardiomyopathy subgroups, the risk of death or heart transplantation is significantly increased when two or more risk factors are present and also as the number of risk factors increases.
Interpretation
Among children with hypertrophic cardiomyopathy, the risk of death or heart transplantation is greater for those who present as infants or with inborn errors of metabolism or with mixed hypertrophic and dilated or restrictive cardiomyopathy. Risk stratification by subgroup of cardiomyopathy, by characteristics such as lower weight, congestive heart failure, or abnormal echocardiographic findings, and by the presence of multiple risk factors allows for more informed clinical decision-making and prognosis at the time of diagnosis.
Introduction
Cardiomyopathy is a rare but serious condition in infants and children. Hypertrophic cardiomyopathy is a heterogeneous group of disorders in childhood.1–10 Paediatric hypertrophic cardiomyopathy encompasses conditions with diverse genetic origins and clinical phenotypes, including associations with inborn errors of metabolism, neuromuscular disorders, and malformation syndromes.7,11 Limited data are available to predict which patients with paediatric hypertrophic cardiomyopathy will progress to end-stage pump failure with subsequent death or need for heart transplant or who will die from sudden cardiac death.12–14 Despite the relative rarity of this disorder, the incidence, morbidity, and mortality are higher in the first year of life by a factor of 10 as compared to the rest of childhood.9,10,12 There is a need for increased understanding of the factors contributing to a poor outcome in the various aetiological subgroups of paediatric hypertrophic cardiomyopathy.
Using data from the Pediatric Cardiomyopathy Registry (PCMR) funded by the National Heart, Lung, and Blood Institute of the USA National Institutes of Health, we have previously reported the demographic and clinical features of children with pure hypertrophic cardiomyopathy, as well as their clinical outcomes.12 In that report, all available annual follow-up data were used to categorize children into aetiologic subgroups in as scientifically rigorous a manner as possible. Clinical outcome and survival were found to be dependent on the aetiology of hypertrophic cardiomyopathy and patient age. Although this report provided the most complete estimates of the aetiologies of the disease, the impact of clinical characteristics at the time of diagnosis on outcome were not evaluated. Physicians treating children with hypertrophic cardiomyopathy must base their initial course of care on information collected at the time of diagnosis. Therefore, we have reanalysed the PCMR database to quantify clinical outcomes and their predictors on the basis on aetiologic subgroups (including age at diagnosis and mixed phenotype hypertrophic cardiomyopathy) in children with hypertrophic cardiomyopathy at the time of diagnosis. Our goal is to understand how patient factors measured at the time of diagnosis will predict the subsequent risk of death or heart transplantation. Our hope is that this information will allow clinicians to more effectively risk-stratify patients resulting in improved family counselling and clinical planning development for these children resulting in reduced morbidity and mortality.
Methods
This study utilized PCMR data from patients seen between January 1990 and February 2009 at 98 participating paediatric cardiology centres in the United States and Canada. The design and implementation of the PCMR are described elsewhere.9,15 Briefly, patients less than 18 years of age who received a recent diagnosis of cardiomyopathy at participating centres, were eligible for inclusion. Children with specific secondary aetiologies of ventricular hypertrophy such as pulmonary parenchymal or vascular disease, endocrine disease (including maternal gestational diabetes), rheumatic disease, immunologic disease, cardiotoxic exposures, systemic hypertension, or congenital cardiovascular malformations independent of a malformation syndrome, were excluded.
Each case of hypertrophic cardiomyopathy was classified morphologically according to specific criteria: either ventricular septal or left ventricular (LV) posterior wall thickness for body-surface area more than 2 SD different from the value for a normal population of infants, children, and adolescents with similar body-surface area; or the presence of localized LV hypertrophy. Cases of hypertrophic cardiomyopathy were categorized by phenotype as pure (isolated), mixed hypertrophic and dilated cardiomyopathy, and mixed hypertrophic with restrictive (or other) cardiomyopathy. The mixed hypertrophic and dilated cardiomyopathy phenotype was defined as HCM (based on the criteria above) with echocardiographic evidence of LV dilation LV end-diastolic dimension Z score > 2) and depressed ventricular function (LV fractional shortening Z score < −2). The mixed hypertrophic and restrictive cardiomyopathy phenotype was defined as HCM with enlargement of one or both atria relative to a ventricle of normal or small size with evidence of impaired diastolic filling in the absence of significant valvar heart disease. The pure hypertrophic cases were further subdivided on the basis of age at diagnosis and presumed aetiology at the time of diagnosis into the following subgroups: idiopathic hypertrophic cardiomyopathy diagnosed at less than 1 year of age, idiopathic hypertrophic cardiomyopathy diagnosed at 1 year of age or older, hypertrophic cardiomyopathy with malformation syndromes, hypertrophic cardiomyopathy with inborn errors of metabolism, hypertrophic cardiomyopathy with neuromuscular disease, and familial hypertrophic cardiomyopathy.15 Clearly, some children initially diagnosed as idiopathic HCM may, upon further diagnostic evaluation, may be assigned a specific aetiology, so that the idiopathic subgroup is not homogenous.
Demographic and clinical data relevant to cardiomyopathy, including echocardiographic measurements, family history, and clinical findings, were collected at diagnosis and annually thereafter. Primary clinical outcomes were heart transplantation or death. Death was analysed as all-cause mortality. For purely descriptive purposes, cause of death, defined as 1) pure cardiac, 2) cardiac causes contributing with another pre-existing condition, 3) non-cardiac, or 4) unknown, was determined for each HCM subgroup.
Quality control systems were developed to ensure the accuracy and completeness of data collected for the PCMR. Our method for determining patient eligibility had clearly defined echocardiographic measurements that could be verified by the data management system. Patients’ echocardiographic values in the data management system were compared with previously defined echocardiographic norms.15,16 If the patient’s values did not meet the study inclusion criteria, the PCMR cardiologists reviewed the data and a final determination of eligibility was made. Edit reports were sent to the individual clinical centers for resolution. The data collection forms: built-in redundancies to substantiate the data contained therein. All PCMR data collectors had an intensive 2-day training seminar in data abstraction from medical records. They were trained in medical terminology and purposes of various cardiac procedures and tests, classification of cardiomyopathies, inclusion and exclusion criteria, graphic interpretation of echocardiographic parameters, and data form completion. Any questions about data interpretation or inclusion and exclusion criteria were discussed with the PCMR study site cardiologist.
Patient confidentiality was assured through the use of coded patient identifiers without transfer of patient names and addresses outside of their primary institution. All participating centres obtained Institutional Review Board approval.
Statistical analysis
Body-mass index (BMI, the weight in kilograms divided by the square of the height in meters) Z scores were calculated for children of appropriate size who were 2 years of age or older with the use of the USA Centers for Disease Control and Prevention (CDC) tables, and weight-for-height Z scores (from the CDC tables) were substituted for children who did not meet the BMI age and size requirements.17 Echocardiographic measurements were transformed into Z scores adjusted for body-surface area or age relative to normal children.16 Descriptive demographic and clinical characteristics of the children at the time of diagnosis of hypertrophic cardiomyopathy are presented as percentages for categorical data and as means and standard deviations for continuous data.
The Kaplan-Meier method was used to calculate time-to-event rates from the time of diagnosis to the earlier of heart transplantation or death for children in each subgroup. Only a few children with hypertrophic cardiomyopathy underwent heart transplantation. The results of the analyses with death as the end point did not differ statistically from the results of the analyses with death or transplantation as the end point, and therefore only the analyses using the combined end point are presented. Data were censored at the last follow-up visit, and children without follow-up data were removed from the analyses.
Cox proportional-hazards regression was used to identify both univariable and multivariable predictors of death or heart transplantation within each aetiologic subgroup. To avoid collinearity and over-fitting, the multivariable models initially included age, sex, weight Z score, presence of congestive heart failure, and the most statistically significant univariable echocardiographic predictors. For each subgroup, the predictive value of the Cox regression model is illustrated by taking the predictors from the multivariable model and, for each patient, counting the number of risk factors present. Continuous predictors were considered to be present if the patient was in either the highest or lowest quartile of the range. We then estimated 2-year mortality using the overall Kaplan-Meier curve, altered according to the hazard ratio calculated from the Cox regression. A bar graph is shown to compare the 2-year risk of death or heart transplantation according to the number of risk factors: none versus one versus two or more.
All analyses were conducted with the use of the Statistical Analysis System, Version 9·1, and S-Plus 6·1. A p value less than 0·05 was considered to indicate statistical significance.
Role of the funding source
The funding source had no role in the study design, data collection, data analysis, data interpretation, or writing of the report. The corresponding author had full access to all the data in the study and had final responsibility for the decision to submit for publication.
Results
Study participants
Of the 1085 children with hypertrophic cardiomyopathy in the database, 69 also had characteristics of dilated cardiomyopathy, 58 also had characteristics of restrictive cardiomyopathy, and the remaining 958 had pure hypertrophic cardiomyopathy at the time of initial diagnosis. We divided the cohort of patients with pure hypertrophic cardiomyopathy into subgroups according to aetiology. In the subgroups of patients with neuromuscular disorders and familial hypertrophic cardiomyopathy, there were too few clinical outcomes to proceed with risk-factor modelling, so neither group was included in further analyses.
At the time of diagnosis, the remaining 788 children with pure hypertrophic cardiomyopathy were of relatively normal height and weight, and few presented with congestive heart failure (15%; table 1). However, they had echocardiographic evidence of substantially elevated LV fractional shortening, ventricular septal thickness, LV posterior wall thickness and LV ventricular mass, as well as diminished LV ventricular end-diastolic dimension. In contrast, children whose hypertrophic cardiomyopathy was mixed with restrictive cardiomyopathy or dilated cardiomyopathy had lower mean weight and height, more commonly presented with congestive heart failure and a family history of cardiomyopathy, had below-normal or normal LV fractional shortening, and had elevated or normal LV end-diastolic dimension. Children with mixed functional phenotypes also had elevated end-diastolic ventricular septal thickness and LV posterior wall thicknesses and LV mass.
Table 1.
Patient characteristics and outcomes by hypertrophic cardiomyopathy phenotype category*
| Variable | Pure Hypertrophic Cardiomyopathy | Mixed Hypertrophic with Dilated Cardiomyopathy |
Mixed Hypertrophic with Restrictive or Other Cardiomyopathy |
|---|---|---|---|
| Sample size | 788 | 69 | 58 |
| Female sex-% | 32 | 48 | 47 |
| Age at diagnosis-year | 6·0±6·3 | 3·8±5·8 | 7·0±6·0 |
| BMI Z score ‡ | 0·17±1·6 | −1·26±1·8 | −0·19±1·5 |
| Weight Z score | 0·18±1·7 | −1·02±1·3 | −0·47±1·8 |
| Height Z score | −0·32±1·5 | −0·93±1·4 | −0·84±1·7 |
| Congestive heart failure at diagnosis-% | 15 | 65 | 33 |
| Left ventricular fractional shortening Z score | 3·2±5·2 | −6·9±5·1 | 0·1±4·2 |
| End-diastolic ventricular septal thickness Z score | 3·3±2·6 | 1·9±2·5 | 2·5±2·5 |
| Left ventricular end-diastolic dimension Z score | −2·0±3·3 | 2·3±2·3 | −0·6±2·4 |
| Left ventricular mass | 1·9±2·9 | 4·2±3·5 | 2·1±2·0 |
| Death-no./total no. (%) | |||
| All children | 124/788 (16) | 22/69 (32) | 11/58 (19) |
| Idiopathic hypertrophic cardiomyopathy diagnosed at <1 year | 49/252 (19) | ||
| Idiopathic hypertrophic cardiomyopathy diagnosed at ≥1 year | 21/407 (5) | ||
| Malformation syndrome | 13/60 (22) | ||
| Inborn errors of metabolism | 41/69 (59) | ||
| Sudden Death-no ./total no. (%) | |||
| All children | 19/788 (2) | 0/69 (0) | 1/58 (2) |
| Idiopathic hypertrophic cardiomyopathy diagnosed at <1 year | 7/252 (3) | ||
| Idiopathic hypertrophic cardiomyopathy diagnosed at ≥1 year | 10/407 (2) | ||
| Malformation syndrome | 1/60 (2) | ||
| Inborn errors of metabolism | 1/69 (1) | ||
| Death or heart transplantation-no./total no. (%) | |||
| All children | 142/788 (18) | 28/69 (41) | 24/58 (41) |
| Idiopathic hypertrophic cardiomyopathy diagnosed at <1 year | 56/252 (22) | ||
| Idiopathic hypertrophic cardiomyopathy diagnosed at ≥1 year | 28/407 (7) | ||
| Malformation syndrome | 14/60 (23) | ||
| Inborn errors of metabolism | 44/69 (64) |
Plus-minus values are means ± standard deviation.
Race or ethnic group was determined by enrolling physician notation or by medical chart review.
BMI denotes body-mass index (the weight in kilograms divided by the square of the height in meters).
Patient outcomes
Table 1 and figure 1 show the estimated rates of death or heart transplantation within our six analytic subgroups of children, divided according to functional type and aetiology. Of the 157 deaths, 13% were classified as sudden, 67% were non-sudden, and 20% were unknown cause of death (table 2). The poorest outcomes were observed among children with inborn errors of metabolism, for whom the estimated rate of death or heart transplantation was 48% (95% confidence interval [CI]: 36%, 60%) at 1 year and 57% (95% CI: 45%, 69%) at 2 years. Children with mixed functional phenotypes also did poorly, with rates of death or heart transplantation of 43% (95% CI: 31%, 55%) at 1 year and 45% (95% CI: 33%, 57%) at 2 years for children with mixed hypertrophic and dilated cardiomyopathy, and 30% (95% CI: 18%, 42%) at 1 year and 38% (95% CI: 24%, 52%) at 2 years for children with mixed hypertrophic and restrictive cardiomyopathy.
Figure 1.
The survival rates, using the Kaplan-Meier method, from diagnosis of hypertrophic cardiomyopathy to death or heart transplantation of subgroups of children with hypertrophic cardiomyopathy based on age at diagnosis, aetiology, or phenotype. Panel A shows the rates for 252 patients with pure hypertrophic cardiomyopathy diagnosed at under 1 year of age (black solid line), 407 patients with pure hypertrophic cardiomyopathy diagnosed at 1 year of age or older (red dashed line), and 60 patients with pure hypertrophic cardiomyopathy with malformation syndromes (green dotted line). Panel B shows the rates for 69 patients with pure hypertrophic cardiomyopathy with inborn errors of metabolism (black solid line), 58 patients with mixed hypertrophic and restrictive or other cardiomyopathies (red dashed line), and 69 patients with mixed hypertrophic and dilated cardiomyopathy (green dotted line).
Table 2.
Patient outcomes by hypertrophic cardiomyopathy phenotype category
| Pure Hypertrophic Cardiomyopathy | Mixed Hypertrophic Cardiomyopathy | |||||
|---|---|---|---|---|---|---|
| Idiopathic Diagnosed at <1 year |
Idiopathic Diagnosed at ≥1 year |
Malformation Syndromes |
Inborn Errors of Metabolism |
With Dilated Cardiomyopathy |
With Restrictive Cardiomyopathy |
|
| Number of Children | 252 | 407 | 60 | 69 | 69 | 58 |
| Deaths and Transplants | 56/252 (22%) | 28/407 (7%) | 14/60 (23%) | 44/69 (64%) | 28/69 (41%) | 24/58 (41%) |
| Sudden Deaths | 7/252 (3%) | 10/407 (2%) | 1/60 (2%) | 1/69 (1%) | 0/69 (0%) | 1/58 (1%) |
| Deaths | 49/252 (19%) | 21/407 (5%) | 13/60 (22%) | 41/69 (59%) | 22/69 (32%) | 11/58 (19%) |
| Causes of Death | ||||||
| Pure Cardiac | 24/49 (49%) | 14/21 (67%) | 5/13 (38%) | 15/41 (37%) | 12/22 (55%) | 7/11 (64%) |
| Cardiac Causes Contributing | 15/49 (31%) | 3/21 (14%) | 6/13 (46%) | 9/41 (22%) | 3/22 (14%) | 2/11 (18%) |
| Non-Cardiac | 2/49 (4%) | 0/21 (0%) | 0/13 (0%) | 5/41 (12%) | 3/22 (14%) | 1/11 (9%) |
| Unknown | 8/49 (16%) | 4/21 (19%) | 2/13 (15%) | 12/41 (29%) | 4/22 (18%) | 1/11 (9%) |
| Pre-Existing Conditions | ||||||
| Neurological | 7/49 (14%) | 1/21 (5%) | 2/13 (15%) | 7/41 (17%) | 3/22 (14%) | 1/11 (9%) |
| Congestive Heart Failure | 27/49 (55%) | 3/21 (14%) | 5/13 (38%) | 13/41 (32%) | 7/22 (32%) | 7/11 (64%) |
| Respiratory Failure | 22/49 (45%) | 3/21 (14%) | 6/13 (46%) | 6/41 (15%) | 5/22 (23%) | 6/11 (55%) |
| Multi-Organ Failure | 11/49 (22%) | 2/21 (10%) | 4/13 (31%) | 9/41 (22%) | 4/22 (18%) | 3/11 (27%) |
Excellent outcomes were observed among the 407 children who received a diagnosis of idiopathic hypertrophic cardiomyopathy at 1 year of age or older, with rates of death or heart transplantation of 1% (95% CI: 0·2%, 2%) at 1 year and 3% (95% CI: 1%, 5·0%) at 2 years. Only 18 of 788, children with pure hypertrophic cardiomyopathy underwent heart transplantation (2%), as compared with 6 of 69 children with mixed hypertrophic and restrictive cardiomyopathy (9%) and 12 of 58 children with mixed hypertrophic and dilated cardiomyopathy (21%).
Univariable predictors of death or heart transplantation
In general, younger, smaller children and those with congestive heart failure were at the highest risk for poor outcomes (table 3). Echocardiographic measures were important prognostic factors in all aetiologic subgroups except for children with malformation syndromes. However, the significant echocardiographic measures varied according to the aetiology of hypertrophic cardiomyopathy. Year at presentation, U.S. region, race-ethic group, or health insurance status were not statistically significant predictors of poor outcomes.
Table 3.
Univariate predictors of death or heart transplantation by hypertrophic cardiomyopathy age at diagnosis, aetiologic or phenotypic subgroup
| Pure Hypertrophic Cardiomyopathy | Mixed Hypertrophic Cardiomyopathy | |||||
|---|---|---|---|---|---|---|
| Predictor | Idiopathic Diagnosed at <1 year |
Idiopathic Diagnosed at ≥1 year |
Malformation Syndromes |
Inborn Errors of Metabolism |
With Restrictive Cardiomyopathy |
With Dilated Cardiomyopathy |
| Demographic | ||||||
| Female sex (versus male) | 1·2 (p=0·56) | 2·2 (p=0·04) | 0·53 (p=0·29) | 1·1 (p=0·83) | 1·9 (p=0·13) | 1·4 (p=0·45) |
| Physical | ||||||
| Age at diagnosis (per year) | 0·21 (p=0·01) | 1·05 (p=0·23) | 0·70 (p=0·09) | 0·88 (p=0·01) | 0·92 (p=0·03) | 0·99 (p=0·70) |
| Weight Z score (per 1 SD) | 0·69 (p=0·007) | 0·88 (p=0·38) | 1·0 (p=0·91) | 0·77 (p=0·09) | 0·85 (p=0·30) | 0·78 (p=0·16) |
| Height Z score (per 1 SD) | 0·96 (p=0·75) | 1·1 (p=0·74) | 0·82 (p=0·55) | 0·95 (p=0·72) | 0·89 (p=0·48) | 0·88 (p=0·49) |
| BMI Z score (per 1 SD)† | 0·71 (p=0·005) | 0·80 (p=0·19) | 0·96 (p=0·90) | 0·80 (p=0·03) | 0·73 (p=011) | 0·92 (p=0·52) |
| Cardiac | ||||||
| Congestive heart failure (versus no congestive heart failure) | 4·3 (p<0·001) | 2·6 (p=0·11) | 13·5 (p<0·001) | 2·1 (p=0·02) | 4·1 (p<0·001) | 2·6 (p=0·04) |
| Tricuspid valve regurgitation | 1·4 (p=0·61) | 3·7 (p=0·053) | 0·51 (p=0·55) | 1·9 (p=0·24) | 0·48 (p=0·30) | 1·2 (p=0·79) |
| Mitral valve regurgitation | 0·88 (p=0·83) | 0·91 (p=0·89) | 2·3 (p=0·36) | 1·3 (p=0·58) | 0·62 (p=0·48) | 0·83 (p=0·78) |
| Left ventricular fractional shortening Z score (per 1 SD) | 0·95 (p=0·06) | 0·86 (p=0·001) | 0·88 (p=0·051) | 0·95 (p=0·09) | 0·87 (p=0·03) | 0·92 (p=0·13) |
| Left ventricular mass Z score (per 1 SD) | 1·12 (p=0·02) | 1·20 (p=0·06) | 0·99 (p=0·98) | 1·10 (p=0·13) | 1·12 (p=0·40) | 1·07 (p=0·27) |
| End-diastolic ventricular septal thickness Z score (per 1 SD) | 1·02 (p=0·81) | 1·05 (p=0·58) | 1·26 (p=0·09) | 1·12 (p=0·18) | 0·89 (p=0·18) | 1·23 (p=0·02) |
| Left ventricular end-diastolic posterior wall thickness Z score (per 1 SD) | 1·16 (p=0·003) | 1·21 (p=0·02) | 1·02 (p=0·89) | 1·14 (p=0·047) | 1·30 (p=0·01) | 1·06 (p=0·32) |
| Left ventricular end-systolic dimension Z score (per 1 SD) | 1·07 (p=0·053) | 1·09 (p=0·24) | 1·17 (p=0·28) | 0·99 (p=0·93) | 1·15 (p=0·19) | 1·12 (p=0·13) |
| Left ventricular end-diastolic dimension Z score (per 1 SD) | 1·07 (p=0·16) | 1·05 (p=0·63) | 0·85 (p=0·18) | 0·95 (p=0·18) | 0·95 (p=0·64) | 1·06 (p=0·55) |
| Family history | ||||||
| Cardiomyopathy | 0·16 (p=0·07) | 0·92 (p=0·86) | No events* | 0·90 (p=0·83) | 0·24 (p=0·06) | 0·93 (p=0·86) |
| Sudden death | 0·74 (p=0·69) | 1·1 (p=0·88) | No events* | 0·70 (p=0·45) | 0·48 (p=0·48) | 1·2 (p=0·67) |
| Heart disease | No events* | 1·3 (p=0·71) | 4·7 (p=0·06) | 0·95 (p=0·93) | 0·44 (p=0·43) | 0·37 (p=0·34) |
| Genetic syndrome | 1·4 (p=0·60) | 2·8 (p=0·17) | No events* | 0·88 (p=0·79) | 2·8 (p=0·33) | No events* |
BMI denotes body-mass index (the weight in kilograms divided by the square of the height in meters).
Multivariable predictors of death or heart transplantation
Children with idiopathic hypertrophic cardiomyopathy, diagnosed before the age of 1 year, were at the highest risk for poor outcomes if at diagnosis they also had congestive heart failure or low weight (table 4). Of children with two or more of these risk factors (including young age and elevated LV end-diastolic posterior wall thickness), 38% died or underwent heart transplantation, whereas only 4% of children with none of these risk factors had poor outcomes (figure 2). Among children older than 1 year of age with idiopathic hypertrophic cardiomyopathy, female sex and depressed LV fractional shortening significantly predicted death or heart transplantation. When these two risk factors were combined with congestive heart failure and older age, the percentage of children with poor outcomes ranged from 3% for children with no risk factors to 13% for children with two or more risk factors (figure 2).
Table 4.
Multivariable regression models for time to death or heart transplantation by hypertrophic cardiomyopathy age at diagnosis, aetiologic or phenotypic subgroup
| Pure Hypertrophic Cardiomyopathy | Mixed Hypertrophic Cardiomyopathy | |||||
|---|---|---|---|---|---|---|
| Predictor | Idiopathic Diagnosed at year |
<1 Idiopathic Diagnosed at ≥1 year |
Malformation Syndromes |
Inborn Errors of Metabolism |
With Restrictive Cardiomyopathy |
With Dilated Cardiomyopathy |
| Female sex (versus male) | 2·4 (p=0·03) [1·09, 5·17] | |||||
| Weight Z score | 0·70 (p=0·01) [0·54, 0·90] | 0·76 (p=0·08) [0·55, 1·04] | 0·75 (p=0·14) [0·52, 1·10] | |||
| Age at diagnosis (year) | 0·27 (p=0·07) [0·06, 1·14] | 1·08 (p=0·07) [0·99, 1·17] | 0·80 (p=0·17) [0·58, 1·10] | 0·90 (p=0·03) [0·81, 0·99] | 0·96 (p=0·36) [0·89, 1·05] | |
| Congestive heart failure (versus no congestive heart failure) | 2·9 (p<0·001) [1.63, 5.23] | 2·0 (p=0·27) [0·58, 7·06] | 17·3 (p<0·001) [4·65, 64·5] | 1·2 (p=0·58) [0·60, 2·45] | 3·1 (p=0·01) [1·27, 7·66] | 3·1 (p=0·03) [1·15, 8·44] |
| Left ventricular fractional shortening Z score (per 1 SD) | 0·87 (p=0·002) [0·80, 0·95] | 0·82 (p=0·01) [0·70, 0·96] | ||||
| End-diastolic ventricular septal thickness Z score (per 1 SD) | 1·29 (p=0·01) [1·06, 1·57] | |||||
| Left ventricular end-diastolic posterior wall thickness Z score (per 1 SD) | 1·11 (p=0·052) [0·99, 1·22] | 1·01 (p=0·85) [0·88, 1·17] | 1·18 (p=0·13) [0·95, 1·46] | |||
Figure 2.
The probability of death or heart transplantation (Tx) two years after the diagnosis (Dx) of hypertrophic cardiomyopathy (HCM) according to the number of risk factors across the six subgroups of children with hypertrophic cardiomyopathy based on age at diagnosis, aetiology, or phenotype. The sample sizes and number of deaths or Tx for each HCM subgroup are: 1) Idiopathic HCM < 1 year at Dx (N=128, Death or Tx N=23); 2) Idiopathic HCM ≥ 1 year at Dx (N=316, Death or Tx N=23); 3) HCM with malformation syndrome (N=42, Death or Tx N=9); 4) HCM with inborn error of metabolism (N=34, Death or Tx N=22); 5) Mixed HCM with restrictive cardiomyopathy (N=39, Death or Tx N=19); 6) Mixed HCM with dilated cardiomyopathy (N=36, Death or Tx N=18). The threshold values for scoring a point in the risk scores by HCM subgroup are: 1) Idiopathic HCM <1 year at DX (weight Z score < 1·4; age at Dx <0·01 years; end-diastolic posterior wall thickness (EDPT) Z score > 4·42); 2) Idiopathic HCM ≥1 year at Dx: (age at Dx > 14·4 years; fractional shortening (FS) Z score < 1·85); 3) HCM with malformation syndrome (age at Dx <0·2 years; FS Z score < 3·45); 4) HCM with inborn errors of metabolism (weight Z score < −1·8; age at Dx < 0·2 years; EDPT Z score < 6·7); 5) Mixed HCM with restrictive cardiomyopathy (age at Dx < 0·6 years; EDPT Z score >3·9); 6) Mixed HCM with dilated cardiomyopathy (weight Z score < 1·8 years; end-diastolic septal thickness Z score > 3·1). Binomial risk factors (gender, CHF status) do not have threshold values.
Congestive heart failure and decreased LV fractional shortening predicted death or heart transplantation among children with malformation syndromes. The percentage of children with poor outcomes in this subgroup ranged from 0% among those with no risk factors to 50% among children with two or three risk factors, with younger age included in the model. Younger children with inborn errors of metabolism were at significantly higher risk for poor outcomes. Although congestive heart failure, increased LV end-diastolic posterior wall thickness, and low weight were significant in the adjusted model, all were retained in the model represented in figure 2. Even among children with no risk factors, 55% of those with inborn errors of metabolism either died or underwent heart transplantation. This percentage increased to 89% among children with two or three risk factors.
Among children with mixed hypertrophic and restrictive cardiomyopathy, congestive heart failure was the only significant predictor of death or heart transplantation. However, younger age at diagnosis and elevated LV end-diastolic posterior wall thickness were included in the same model. As shown in figure 2, among children with mixed hypertrophic and restrictive cardiomyopathy, 25% of those with none of these three risk factors and 71% of those with two or three of these risk factors died or underwent heart transplantation.
Both congestive heart failure and increased end-diastolic ventricular septal thickness were significant predictors of death or heart transplantation in children with mixed hypertrophic and dilated cardiomyopathy. When lower weight was added to the model, the percentage of children who died or underwent heart transplantation was 57% in those with two or three risk factors but 0% in those with no risk factors.
Across all six subgroups, 60% of the children had one or more risk factors and accounted for 85% of all deaths and heart transplantations.
Discussion
Predicting the clinical outcome for paediatric patients with hypertrophic cardiomyopathy is challenging because of the marked heterogeneity of the population. This study expands the earlier PCMR report on the epidemiologic features and outcomes of paediatric hypertrophic cardiomyopathy to include the influence of demographic and clinical factors at the time of diagnosis on the subsequent outcomes of death and heart transplantation.12 Specifically, the results of this analysis utilize data available to paediatric cardiologists for clinical decision-making at the time of presentation of a child with cardiomyopathy.
The prognosis is worse in children with hypertrophic cardiomyopathy presenting at less than 1 year of age, in those with inborn errors of metabolism, malformation syndromes, or with hypertrophic cardiomyopathy in combination with other cardiomyopathy phenotypes.12,18,19 Decreased weight and BMI Z scores and congestive heart failure at diagnosis were also significant predictors of death or heart transplantation.
Previous studies have shown that the prognosis for children with hypertrophic cardiomyopathy is worse for those diagnosed as infants.12,18,20,21 Patients with inborn errors of metabolism had the worst overall prognosis, with approximately half of these children dying or requiring heart transplantation within 2 years after diagnosis. Regarding the 20 deaths classified as sudden, there were too few deaths in any of the diagnostic subgroups to analyse subgroup based risk factors for sudden death.
The presence of congestive heart failure at the time of diagnosis of hypertrophic cardiomyopathy was a strong predictor of death or heart transplantation in those who are diagnosed at less than 1 year of age, children with malformation syndromes, and children with mixed hypertrophic and dilated or restrictive cardiomyopathy. This finding has been mentioned in prior studies and is further substantiated here.2,21,22 Heart transplantation is a standard therapy for congestive heart failure refractory to medical therapy in children with hypertrophic cardiomyopathy, and outcomes are comparable to those for other forms of cardiomyopathy among patients who survive the initial post-heart transplantation period.20,22,23 However, some children with hypertrophic cardiomyopathy may not be eligible for transplantation, such as those with progressive degenerative extracardiac disease (e.g. certain inborn errors of metabolism) or significant hepatic, pulmonary, neurologic or renal dysfunction from a variety of extracardiac primary diagnoses.
A previous study of echocardiographic predictors of outcomes in the children with hypertrophic cardiomyopathy indicated that various echocardiographic features were associated with poor outcomes, although the sample size was small and no outcome analyses by aetiologic subgroup were included.18 Consistent with this study, in multivariable models, we found associations between echocardiographic findings and clinical outcomes. Increased LV fractional shortening was associated with improved survival in patients who received the diagnosis of hypertrophic cardiomyopathy at 1 year of age or older and those with malformation syndromes. Increased end-diastolic ventricular septal thickness was associated with increased risk in children with mixed hypertrophic and dilated cardiomyopathy. Echocardiographic features were not associated with survival among children in the other aetiologic subgroups.
Nearly all analyses of risk stratification for hypertrophic cardiomyopathy have been conducted in primarily adult populations, with sudden death as the primary outcome, and risk factors that included previous cardiac arrest, syncope, ventricular arrhythmias, family history of sudden death, extreme LV hypertrophy and a blunted blood pressure response to exercise.24–27 These risk factors are not readily applicable to the assessment of younger patients with hypertrophic cardiomyopathy who are at higher risk for non-sudden cardiac death and who would be the most likely to benefit from early, accurate risk stratification. LV outflow tract obstruction, LV systolic dysfunction, and heart failure have been reported to be risk factors for non-sudden cardiac death in children and adults with hypertrophic cardiomyopathy although the positive predictive value of these factors is unknown.24,28–30
The most striking result of our analysis is the clear demonstration of the predictive value of multiple risk factors present at diagnosis in predicting death or heart transplantation in children with hypertrophic cardiomyopathy. For all forms of paediatric hypertrophic cardiomyopathy, the risk of death or heart transplantation is markedly increased when two or more risk factors are present. This finding is particularly important in paediatric hypertrophic cardiomyopathy associated with malformation syndromes or inborn errors of metabolism, idiopathic hypertrophic cardiomyopathy presenting at under 1 year of age, or mixed cardiomyopathy phenotypes. Not only is the absolute risk of death or heart transplantation significantly higher when multiple risk factors are present, but the risk increases significantly as the number of risk factors increases. Our results emphasize the importance of risk factor analysis and risk stratification at the time of diagnosis in order to better inform and individualize the management plan.
Conclusions
Although children with hypertrophic cardiomyopathy comprise the smallest proportion of children with cardiomyopathy listed for heart transplantation, selected patient groups may benefit from early listing for heart transplantation (unless they have specific contraindications to transplantation such as certain inborn errors of metabolism) given the malignant course and high mortality observed in certain subgroups.18,21 Specifically, children diagnosed with hypertrophic cardiomyopathy at less than 1 year of age, those with mixed phenotypes of hypertrophic cardiomyopathy, those with a lower weight or BMI Z score, congestive heart failure, and those who had more abnormal echocardiographic measures of LV structure or function had the worst prognosis. The presence of multiple risk factors at diagnosis had the worst outcomes in most aetiologic subgroups. Children presenting with these risk factors should be evaluated for early heart transplantation and followed more closely than children without these characteristics. A limitation of this analysis is that some children initially diagnosed as idiopathic HCM, upon further diagnostic evaluation, may be assigned a specific aetiology, so that the idiopathic subgroup is probably not homogenous. However, as previously stated, our results are based on the best data available to paediatric cardiologists at the time a child presents with cardiomyopathy.
For those who are eligible for heart transplantation, our results may be helpful in counselling families at the time of diagnosis regarding their child's prognosis. Our results may inform improved risk stratification at the time of diagnosis for children presenting with hypertrophic cardiomyopathy, thereby enabling more informed clinical decision-making which could result in better patient outcomes.
Research in Context [Panel].
Systematic Review
We searched PubMed (English) from January 1970 to July 2013, for reports using the search terms “pediatric or child” and “hypertrophic cardiomyopathy” and one or more of “survival analysis”, or “outcomes”. We eliminated any reports only in adult patients, reports or outcomes from surgical interventions, heart transplantation, descriptive genetic reports, redundant review articles by the same authors or which reiterated published original research, and reports restricted to patients with congenital syndromes not associated with cardiomyopathy. The remaining ten articles are referenced in our report.
Interpretation
Children diagnosed with hypertrophic cardiomyopathy at less than 1 year of age, mixed phenotypes of hypertrophic cardiomyopathy, a lower weight or BMI Z score, congestive heart failure, and those who had abnormal echocardiographic measures of LV structure or function at diagnosis have the worst prognosis. The presence of multiple risk factors at diagnosis predicts the worst clinical outcomes; those children should be evaluated for early heart transplantation listing or heightened clinical monitoring.
Acknowledgments
Funding: This study was supported by the National Heart, Lung, and Blood Institute of the National Institutes of Health, USA (HL53392) (Principal Investigator: Steven E Lipshultz) and the Children’s Cardiomyopathy Foundation, Tenafly, NJ, USA (Principal Investigator: Steven E Lipshultz)
Note: The conclusions and opinions expressed in this article are those of the authors and do not necessarily reflect those of the USA National Institutes of Health or the USA Department of Health and Human Services.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Contributors:
SEL designed the study, searched the published work, interpreted the data and wrote the article. EJO designed the study, analysed and interpreted the data and wrote the article. JDW designed the study, searched the published work, interpreted the data, and wrote the article. JAT designed the study, interpreted the data and wrote the article. JEM collected the data and wrote the article. AML analysed and interpreted the data and wrote the article. LAS analysed and interpreted the data and wrote the article. SDC designed the study, interpreted the data, and wrote the article. The other authors interpreted the data and wrote the article.
Conflicts of Interest
The authors declared no conflicts of interest
REFERENCES
- 1.Richardson P, McKenna W, Bristow M, et al. Report of the 1995 World Health Organization International Society and Federation of Cardiology Task Force on the definition and classification of cardiomyopathies. Circulation. 1996;93:841–842. doi: 10.1161/01.cir.93.5.841. [DOI] [PubMed] [Google Scholar]
- 2.Maron BJ, Tajik AJ, Ruttenberg HD, et al. Hypertrophic cardiomyopathy in infants: clinical features and natural history. Circulation. 1982;65:7–17. doi: 10.1161/01.cir.65.1.7. [DOI] [PubMed] [Google Scholar]
- 3.Suda K, Kohl T, Kovalchin JP, Silverman NH. Echocardiographic predictors of poor outcome in infants with hypertrophic cardiomyopathy. Am J Cardiol. 1997;80:595–600. doi: 10.1016/s0002-9149(97)00428-1. [DOI] [PubMed] [Google Scholar]
- 4.Schaffer MS, Freedom RM, Rowe RD. Hypertrophic cardiomyopathy presenting before 2 years of age in 13 patients. Pediatr Cardiol. 1983;4:113–119. doi: 10.1007/BF02076335. [DOI] [PubMed] [Google Scholar]
- 5.Bruno E, Maisuls H, Juaneda E, Moreyra E, Alday LE. Clinical features of hypertrophic cardiomyopathy in the young. Cardiol Young. 2002;12:147–152. doi: 10.1017/s1047951102000306. [DOI] [PubMed] [Google Scholar]
- 6.McKenna WJ, Deanfield JE. Hypertrophic cardiomyopathy: an important cause of sudden death. Arch Dis Child. 1984;59:971–975. doi: 10.1136/adc.59.10.971. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Schwartz ML, Cox GF, Lin AE, et al. Clinical approach to genetic cardiomyopathy in children. Circulation. 1996;94:2021–2038. doi: 10.1161/01.cir.94.8.2021. [DOI] [PubMed] [Google Scholar]
- 8.Wilkinson JD, Landy DC, Colan SD, et al. The pediatric cardiomyopathy registry and heart failure: key results from the first 15 years. Heart Fail Clin. 2010;6:401–413. doi: 10.1016/j.hfc.2010.05.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Lipshultz SE, Sleeper LA, Towbin JA, et al. The incidence of pediatric cardiomyopathy in two regions of the United States. N Engl J Med. 2003;348:1647–1655. doi: 10.1056/NEJMoa021715. [DOI] [PubMed] [Google Scholar]
- 10.Nugent AW, Daubeney PEF, Chondros P, et al. The epidemiology of childhood cardiomyopathy in Australia. N Engl J Med. 2003;348:1639–1646. doi: 10.1056/NEJMoa021737. [DOI] [PubMed] [Google Scholar]
- 11.Wilkinson JD, Lowe AM, Salbert BA, et al. Outcomes in children with Noonan syndrome and hypertrophic cardiomyopathy: a study from the Pediatric Cardiomyopathy Registry. Am Heart J. 2012;164:442–448. doi: 10.1016/j.ahj.2012.04.018. [DOI] [PubMed] [Google Scholar]
- 12.Colan SD, Lipshultz SE, Lowe AM, et al. Epidemiology and cause-specific outcome of hypertrophic cardiomyopathy in children: findings from the Pediatric Cardiomyopathy Registry. Circulation. 2007;115:773–781. doi: 10.1161/CIRCULATIONAHA.106.621185. [DOI] [PubMed] [Google Scholar]
- 13.Yetman AT, Hamilton RM, Benson LN, McCrindle BW. Long-term outcome and prognostic determinants in children with hypertrophic cardiomyopathy. JACC. 1998;32:1943–1950. doi: 10.1016/s0735-1097(98)00493-8. [DOI] [PubMed] [Google Scholar]
- 14.Maron BJ, Spirito P, Ackerman MJ, et al. Prevention of sudden cardiac death with implantable cardioverter-defibrillators in children and adolescents with hypertrophic cardiomyopathy. J Am Coll Cardiol. 2013;61:1527–1535. doi: 10.1016/j.jacc.2013.01.037. [DOI] [PubMed] [Google Scholar]
- 15.Grenier MA, Osganian SK, Cox GF, et al. Design and implementation of the North American Pediatric Cardiomyopathy Registry. Am Heart J. 2000;139:S86–S95. doi: 10.1067/mhj.2000.103933. [DOI] [PubMed] [Google Scholar]
- 16.Kuczmarski RJ, Ogden CL, Grummer-Strawn LM, et al. CDC growth charts: United States. Adv Data. 2000:1–27. [PubMed] [Google Scholar]
- 17.Sluysmans T, Colan SD. Theoretical and empirical derivation of cardiovascular allometric relationships in children. J Appl Physiol. 2005;99:445–457. doi: 10.1152/japplphysiol.01144.2004. [DOI] [PubMed] [Google Scholar]
- 18.Nugent AW, Daubeney PE, Chondros P, et al. Clinical features and outcomes of childhood hypertrophic cardiomyopathy: results from a national population-based study. Circulation. 2005;112:1332–1338. doi: 10.1161/CIRCULATIONAHA.104.530303. [DOI] [PubMed] [Google Scholar]
- 19.Maskatia SA, Decker JA, Spinner JA, et al. Restrictive physiology is associated with poor outcomes in children with hypertrophic cardiomyopathy. Pediatr Cardiol. 2011;1:141–149. doi: 10.1007/s00246-011-0106-6. [DOI] [PubMed] [Google Scholar]
- 20.Gajarski R, Naftel DC, Pahl E, et al. Outcomes of pediatric patients with hypertrophic cardiomyopathy listed for transplant. J Heart Lung Transplant. 2009;28:1329–1334. doi: 10.1016/j.healun.2009.05.028. [DOI] [PubMed] [Google Scholar]
- 21.Ostman-Smith I, Wettrell G, Keeton B, Riesenfeld T, Holmgren D, Ergander U. Echocardiographic and electrocardiographic identification of those children with hypertrophic cardiomyopathy who should be considered at high risk of dying suddenly. Cardiol Young. 2005;15:632–642. doi: 10.1017/S1047951105001824. [DOI] [PubMed] [Google Scholar]
- 22.Ragni L, Biagini E, Picchio FM, et al. Heart transplantation in infants with idiopathic hypertrophic cardiomyopathy. Pediatr Transplant. 2009;13:650–653. doi: 10.1111/j.1399-3046.2008.01022.x. [DOI] [PubMed] [Google Scholar]
- 23.Dipchand AI, Naftel DC, Feingold B, et al. Outcomes of children with cardiomyopathy listed for transplant: a multi-institutional study. J Heart Lung Transplant. 2009;28:1312–1321. doi: 10.1016/j.healun.2009.05.019. [DOI] [PubMed] [Google Scholar]
- 24.Decker JA, Rossano JW, Smith EO, et al. Risk factors and mode of death in isolated hypertrophic cardiomyopathy in children. J Am Coll Cardiol. 2009;54:250–254. doi: 10.1016/j.jacc.2009.03.051. [DOI] [PubMed] [Google Scholar]
- 25.Christiaans I, van Engelen K, van Langen IM, et al. Risk stratification for sudden cardiac death in hypertrophic cardiomyopathy: systematic review of clinical risk markers. Europace. 2010;12:313–321. doi: 10.1093/europace/eup431. [DOI] [PubMed] [Google Scholar]
- 26.Sadoul N, de Chillou C, Aliot E, McKenna WJ. Evaluation of the risk of sudden death in hypertrophic cardiomyopathy. Arch Mal Coeur Vaiss. 1999;92:65–73. [PubMed] [Google Scholar]
- 27.Yetman AT, McCrindle BW. Management of pediatric hypertrophic cardiomyopathy. Curr Opin Cardiol. 2005;20:80–83. doi: 10.1097/01.hco.0000153452.45341.36. [DOI] [PubMed] [Google Scholar]
- 28.Maron MS, Olivotta I, Betocchi S, et al. Effect of left ventricular outflow tract obstruction on clinical outcome in hypertrophic cardiomyopathy. N Engl J Med. 2003;348:295–303. doi: 10.1056/NEJMoa021332. [DOI] [PubMed] [Google Scholar]
- 29.Fernández A, Vigliano CA, Casabé, et al. Comparison of prevalence, clinical course, and pathological findings of left ventricular systolic impairment versus normal systolic function in patients with hypertrophic cardiomyopathy. Am J Cardiol. 2011;108:548–555. doi: 10.1016/j.amjcard.2011.03.083. [DOI] [PubMed] [Google Scholar]
- 30.Altarabsheh SE, Dearani JA, Burkhart HM, et al. Outcome of septal myectomy for obstructive hypertrophic cardiomyopathy in children and young adults. Ann Thorac Surg. 2013;95:663–669. doi: 10.1016/j.athoracsur.2012.08.011. [DOI] [PubMed] [Google Scholar]