2022 Volume 86 Issue 12 Pages 1943-1949
Background: There has been no nationwide survey on the prognosis of pediatric restrictive cardiomyopathy (RCM) in Japan; therefore, this retrospective multicentered study was designed to investigate the long-term survival rate of pediatric patients with RCM in Japan.
Methods and Results: A multicentered, retrospective observational study was performed between 1990 and 2014 and included patients diagnosed with RCM who were aged <18 years from 18 Japanese institutions. A total of 54 patients were diagnosed with RCM. The median age at diagnosis was 4.4 years, and the median duration of observation was 2.2 years at the time of this study. Of these patients, 54% had symptoms, including heart failure. Twelve patients died without heart transplantation, mostly due to heart failure. The median time to death from diagnosis was 2.5 years. Freedom from death at 1, 5, and 10 years was 91%, 68%, and 62%, respectively. Death occurred within 5 years of diagnosis in most patients. Twenty-two patients underwent heart transplantation. Freedom from heart transplantation at 1, 5, and 10 years was 77%, 58%, and 53%, respectively. Freedom from death or heart transplantation at 1, 5, and 10 years was 72%, 40%, and 34%, respectively. The presence of symptoms was a risk factor for death or transplantation.
Conclusions: The prognosis of pediatric RCM is poor, and the heart transplantation rate is low in Japan.
Restrictive cardiomyopathy (RCM) is a rare form of heart muscle disease characterized by decreased ventricular diastolic function with restrictive physiology and normal or near-normal systolic function.1,2 The prognosis of pediatric RCM is very poor, with approximately 50% mortality within 2 years of diagnosis.3–11 Patients may deteriorate or die suddenly; therefore, it is difficult to predict each patient’s outcome. In population-based studies from the United States, freedom from death or heart transplantation at 10 years of diagnosis was only 20%.12 Heart transplantation is an option for management once the diagnosis of RCM is made, even when mild symptoms are presented.13–15 There has been no nationwide survey on the prognosis of pediatric RCM in Japan to the best of our knowledge, where the number of pediatric organ donors is limited,16 compared to that in other countries.17 Therefore, we investigated the long-term survival rate of Japanese pediatric patients with RCM in a retrospective multicentered study.
We performed a retrospective multicentered observational study on the outcomes of pediatric cardiomyopathy in Japan. Cardiomyopathy was categorized according to the World Health Organization classification.18 Patients diagnosed with cardiomyopathy who were aged <18 years between 1990 and 2014 were included. Eighteen institutions participated in the nationwide survey (Appendix).
RCM was diagnosed based on echocardiographic findings and/or cardiac catheterization data indicating diastolic dysfunction of the left ventricle (LV), in the absence of severely impaired systolic function.4–9 Given that some echocardiographic parameters of diastolic function, such as an increased mitral E/A wave ratio, cannot be adopted in children,19–21 only left atrial dilatation that was disproportionate to aortic or ventricular size without mitral regurgitation, with left atrial dimension/aortic dimension >1.5, is used as an echocardiographic criterion of RCM.12 Echocardiographic measurements were normalized according to the body surface area. No specific requirements were set for the thickness of the LV walls. Absence of severely impaired systolic function was defined as fractional shortening of the LV >0.20. Clinically stable patients underwent cardiac catheterization and the criteria of diastolic dysfunction were defined as pulmonary artery wedge pressure or LV end-diastolic pressure >12 mmHg.7 Children with cardiomyopathy secondary to neuromuscular disorder, inborn errors of metabolism with multiple organ involvement, and significant structural heart diseases were excluded. Family history was considered present if there was an affected first- or second-degree relative with cardiomyopathy. Genotype information of cardiomyopathy was not included in the present study.
The following parameters were collected from medical records: age at diagnosis, medications, initial symptoms, family history, laboratory data, electrocardiographic (ECG) abnormalities, cardiothoracic ratio on chest roentgenogram at onset, device therapy including cardiac resynchronization therapy (CRT), implantable cardioverter-defibrillator (ICD), extracorporeal membrane oxygenation (ECMO), and ventricular assist device (VAD). Symptoms of heart failure were defined as feeding difficulty, failure to thrive, and respiratory distress in infants and easy fatiguability in older children. The diagnostic criteria for LV hypertrophy, right ventricular hypertrophy, right atrial enlargement, and left atrial enlargement on ECG were based on the guideline published previously.22 The primary endpoint was death or heart transplantation. All participating hospitals had institutional review board approval. Patients were enrolled following the opt-out approach via a proxy. The study protocol and institutional review board approval have been made available on the Tokyo Women’s Medical University website. The procedures followed were in accordance with the Declaration of Helsinki, and the ethical standards of the responsible committee on human experimentation.
StatisticsContinuous variables are shown as medians [interquartile range (IQR)]. Categorical data are presented as numbers and percentages. The chi-squared test was used to analyze particular probability distributions between categorical data. The Kaplan-Meier survival time utilized the time from diagnosis to death or heart transplantation. Cox regression was used to assess the multivariate analysis of the survival rate. Cumulative incidence was assessed to determine the event rate curve and calculate the subdistribution hazard ratio. A P value <0.05 was considered statistically significant. All statistical analyses were performed using STATA®SE version ver15 (StataCorp, TX, USA).
A total of 54 patients were diagnosed with RCM (Table 1). The median age at diagnosis was 4.4 years, and median duration of observation was 2.2 years at the time of this study. The age at diagnosis was <8 years in most cases (Figure 1). There was no difference in prevalence by sex. Family history was noted in only 11% of patients (Table 1). Twenty-nine (54%) patients had symptoms, including heart failure in 24 patients (44%). Symptoms of heart failure included feeding difficulty and failure to thrive in 5 patients, respiratory distress in 10 patients, and easy fatigability in 15 patients. At the initial visit, the New York Heart Association (NYHA) functional classes III and IV were 17% and 8%, respectively. Symptomatic arrhythmia was observed in only 1 patient (2%) who experienced atrial flutter and bradycardia due to atrial standstill. The reasons for the RCM diagnosis in asymptomatic patients were the detection of abnormal heart sounds, heart murmur, and electrocardiogram (ECG) abnormalities during health checks and school screenings. Among 13 patients with ECG abnormalities that led to the diagnosis of RCM, ST depression in limb-leads II, III, and aVF and/or leads V5 and V6 was noted in 10 patients, right bundle branch block was noted in 7 patients, right atrial enlargement was noted in 7 patients, left atrial enlargement was noted in 6 patients, right ventricular hypertrophy was noted in 3 patients, and left ventricular hypertrophy was noted in 2 patients. The median B-type natriuretic peptide value was 505 pg/mL, and the median cardiothoracic ratio was 58% at the initial visit. The LV end-diastolic dimension z-score was 1.2, and LV systolic function was normal in the majority of patients (Table 1). Six patients had interventricular septum and/or LV posterior wall thickness >2 z-score and these patients were considered to have a mixed phenotype of RCM and hypertrophic cardiomyopathy (HCM). Left atrial/aortic dimension ratio was 1.9 (IQR 1.7–2.2). Median pulmonary capillary wedge pressure or LV end-diastolic pressure was 23 mmHg, ranging from 13 to 37 mmHg (IQR: 20–28 mmHg) (Table 1).
| No. of patients | 54 |
| Male (n, %) | 25 (46) |
| Age at diagnosis, years (median, IQR) | 4.4 [2.5 to 7.3] |
| No. of patients diagnosed at age <1 year (n, %) | 8 (14.8) |
| Observation duration, years (median, IQR) | 2.2 [0.8 to 6.7] |
| Family history of cardiomyopathy (n, %) | 6 (11) |
| Asymptomatic (n, %) | 25 (46) |
| Abnormal ECG | 13 (24) |
| Heart murmur | 12 (23) |
| Symptomatic (n, %) | 29 (54) |
| Heart failure | 24 (44) |
| Clinical arrhythmia | 1 (2) |
| Resuscitated death | 2 (4) |
| Others | 2 (4) |
| Laboratory data (median, IQR) at initial visit | |
| BNP (pg/dL) | 505 [236 to 982] |
| CTR (%) | 58 [52 to 61] |
| Echocardiographic findings | |
| LVIDd, z-score (median, IQR) | 1.2 [−0.1 to 2.2] |
| LVPWTd, z-score (median, IQR) | 1.8 [−0.1 to 2.8] |
| IVSd, z-score (median, IQR) | 0.6 [−0.5 to 1.5] |
| FS (median, IQR) | 0.33 [0.27 to 0.37] |
| LA/Ao ratio | 1.9 [1.7 to 2.2] |
| Deceleration time of E wave (ms) | 104 [82 to 125] |
| Catherterization data | |
| Pulmonary arterial wedge or LV end-diastolic pressure (mmHg) (n=42) | 23 [20 to 28] |
BNP, B-type natriuretic peptide; CTR, cardiothoracic ratio; FS, fractional shortening; IQR, interquartile range; IVSd, interventricular septum thickness at diastole; LA/Ao ratio, left atrial / aortic dimension ratio; LV, left ventricle; LVIDd, left ventricular internal dimension at diastole; PWTd, posterior wall thickness at diastole; RCM, restrictive cardiomyopathy.
Age at diagnosis. The distribution of age at diagnosis is <9 years old in a majority of patients.
Of these patients, 67% were on medication, including angiotensin-converting enzyme inhibitors, diuretics, and β-blockers (Table 2). Twenty-eight percent received antiplatelet and/or anticoagulation therapy. None of the patients had any symptoms of thromboembolism. Four percent of the patients had re ceived an ICD (Table 2).
| No medication | 18 (33) |
| Medication | 36 (67) |
| β-blocker | 18 (33) |
| Diuretics | 27 (50) |
| ACE inhibitor | 25 (46) |
| Aspirin | 14 (26) |
| Warfarin | 4 (7) |
| Devices | 5 (10) |
| ICD | 2 (4) |
| ECMO | 1 (2) |
| VAD | 2 (4) |
Data are presented as n (%). ACE, angiotensin-converting enzyme; ECMO, extracorporeal membrane oxygenation; ICD, implantable cardioverter defibrillator; VAD, ventricular assist device.
Twelve patients (22%) died without receiving a heart transplantation, mostly due to heart failure. The median time to death from diagnosis was 2.5 years (range 0.7–4.8 years). Freedom from death at 1, 5, 10, and 20 years was 91%, 68%, 62%, and 53%, respectively (Figure 2A). Death occurred within 5 years of diagnosis in most patients. Twenty-two patients underwent heart transplantation, and freedom from transplantation at 1, 5, 10, and 20 years was 77%, 58%, 53%, and 47%, respectively (Figure 2B). Most heart transplantations were performed within 5 years after diagnosis. Freedom from death or heart transplantation at 1, 5, 10, and 20 years was 72%, 40%, 34%, and 26%, respectively (Figure 2C). Multivariate analysis showed that the presence of heart failure symptoms was a significant risk factor for death or transplantation (Table 3). The use of diuretics significantly decreased the risk for death or transplantation. Age at diagnosis and echocardiographic parameters, including HCM phenotype, were not risk factors (Table 3). Freedom from death or heart transplantation in patients with the NYHA functional class III or IV was significantly lower than in patients with NYHA functional class I or II (Figure 3). However, no significant risk factors for death were identified (Table 4).
Overall freedom from death or heart transplantation. (A) Freedom from death at 1, 5, 10, and 20 years was 91%, 68%, 62%, and 53%, respectively. (B) Freedom from heart transplantation at 1, 5, 10, and 20 years was 77%, 58%, 53%, and 47%, respectively. (C) Freedom from death or heart transplantation at 1, 5, 10, and 20 years was 72%, 40%, 34%, and 26%.
| Univariate analysis | Multivariate analysis | |||||
|---|---|---|---|---|---|---|
| HR | 95% CI | P value | HR | 95% CI | P value | |
| Age (years) | ||||||
| <1 | Ref. | |||||
| 1–5 | 1.813 | (0.603–5.458) | 0.290 | |||
| 6–12 | 0.732 | (0.214–2.504) | 0.619 | |||
| >12 | 1.376 | (0.366–5.175) | 0.637 | |||
| Male (vs. female) | 0.81 | (0.402–1.631) | 0.555 | |||
| Echocardiography | ||||||
| LVIDd (z-score) | 0.752 | (0.610–0.926) | 0.007 | |||
| IVSd (z-score) | 0.914 | (0.661–1.263) | 0.586 | |||
| PWTd (z-score) | 1.078 | (0.861–1.349) | 0.514 | |||
| FS | 0.589 | (0.004–82.54) | 0.834 | |||
| LA/Ao ratio | 0.242 | (0.009–5.953) | 0.385 | |||
| Catheterization data | ||||||
| Pulmonary arterial wedge or LV end-diastolic pressure (mmHg) |
1.060 | (0.983–1.137) | 0.129 | |||
| Laboratory data at initial visit | ||||||
| BNP, pg/mL | 1.001 | (1.000–1.001) | 0.039 | |||
| CTR (%) | 1.029 | (0.981–1.079) | 0.237 | |||
| Presence of symptoms at initial visit | 2.165 | (1.017–4.606) | 0.045 | |||
| Resuscitated death | 0.681 | (0.092–5.021) | 0.707 | |||
| Heart failure | 2.857 | (1.361–5.996) | 0.006 | 3.706 | (1.665–8.249) | 0.001 |
| Medication | ||||||
| β-blocker | 0.412 | (0.184–0.923) | 0.031 | 0.582 | (0.237–1.426) | 0.236 |
| ACE inhibitor | 0.668 | (0.329–1.358) | 0.265 | |||
| Diuretics | 0.328 | (0.159–0.676) | 0.003 | 0.342 | (0.149–0.782) | 0.011 |
HR, hazard ratio. Other abbreviations as in Tables 1,2.
Freedom from death or heart transplantation. The freedom from death or heart transplantation in patients with New York Heart Association (NYHA) functional class III/IV was significantly lower than that in patients with NYHA I/II.
| Univariate analysis | |||
|---|---|---|---|
| HR | 95% CI | P value | |
| Age at diagnosis (years) | |||
| <1 | Ref. | ||
| 1–5 | 0.647 | (0.142–2.947) | 0.573 |
| 6–12 | 0.415 | (0.083–2.074) | 0.284 |
| >12 | 0.914 | (0.148–5.643) | 0.923 |
| Male (ref: female) | 1.011 | (0.325–3.145) | 0.985 |
| Echocardiographic parameters | |||
| LVIDd (z-score) | 0.924 | (0.558–1.531) | 0.759 |
| IVSd (z-score) | 1.287 | (0.364–4.559) | 0.695 |
| PWTd (z-score) | 0.895 | (0.525–1.524) | 0.682 |
| FS <0.10 | 0.005 | (0–408.0) | 0.358 |
| LA/Ao ratio | 0.001 | 0–NA | 1.000 |
| Catheterization data | |||
| Pulmonary arterial wedge or LV end-diastolic pressure (mmHg) |
1.090 | (0.975–1.236) | 0.121 |
| Laboratory data and chest X-ray | |||
| Initial BNP (pg/mL) | 1.003 | (0.999–1.007) | 0.152 |
| Initial CTR (%) | 1.014 | (0.929–1.106) | 0.755 |
| Presence of symptoms | 1.108 | (0.353–3.48) | 0.861 |
| Resuscitated death | 2.117 | (0.267–16.776) | 0.478 |
| Heart failure | 1.755 | (0.555–5.552) | 0.338 |
| Management | |||
| β-blocker | 0.423 | (0.113–1.589) | 0.203 |
| ACE inhibitor | 0.695 | (0.218–2.217) | 0.539 |
| Diuretics | 0.501 | (0.158–1.588) | 0.24 |
Abbreviations as in Tables 1–3.
To the best of our knowledge, this study is the first report of a nationwide multi-center study regarding the survival rate of pediatric RCM in Japanese children.
Clinical CharacteristicsIn previous reports, most patients were symptomatic at diagnosis.4,11 In a population-based study from the United States, however, the proportion of patients with heart failure symptoms was only 37%.12 In accordance with the data from a US registry, 54% of patients had symptoms including heart failure in the present study (Table 1). Other clinical presentations of RCM in the present study (Table 1) were similar to that reported in the United States.12
Medication use in Japan (Table 2) was similar to that reported in the United States.9 In the study by Weller et al,9 58% of patients were treated with anti-congestive medications. In the present study, the use of diuretics significantly decreased the risk for death or transplantation (Table 3); however, because of the low number of patients, the efficacy of medical therapy for RCM cannot be concluded from the present study.
Thromboembolism was observed in 15% of patients in the study by Weller et al.9 In the present study, 33% of patients had taken antiplatelet or anticoagulant drugs, and there were no symptoms related to thromboembolism.
OutcomePrevious reports showed a poor natural history with an actuarial 2-year survival rate of <50% after diagnosis;9,11 however, in the US registry, the survival rate was better; freedom from death was 81% at 1 year and 79% at 5 years.12 In the present study, following the US data, freedom from death or transplantation was 72% at 1 year and 40% at 5 years, and freedom from death was 91% at 1 year and 68% at 5 years (Figure 2A).
In this study, freedom from transplantation was 77% at 1 year and 58% at 5 years. In the US registry of patients with RCM without the hypertrophic cardiomyopathy phenotype, freedom from transplantation was 58% at 1 year and 32% at 5 years.12 Thus, although the survival rate in Japan was similar to that in the US, children in the US received more benefit from heart transplantation than children in Japan. Zangwill et al14 reported that 81% of children with RCM listed for transplantation underwent heart transplantation within 1 year in the US. In the study by Anderson et al,8 0 of the 8 patients diagnosed between 1975 and 1993 underwent heart transplantation, and 7 of 12 patients diagnosed between 1994 and 2013 underwent heart transplantation. They reported that the 5-year survival rate improved from 38% to 80%. The number of pediatric heart transplants in Japan remains lower than that in other countries, such as the United States.17 In Japan, the organ transplant law was revised in 2010 so that brain dead children can be donors; however, there have been only 1–3 annual pediatric organ donations between 2010 and 2014.16 Although the number of pediatric organ donors has been increasing and reached 15 in 2019, patients usually have to wait a few years on the list for heart transplantation.16 As effective treatments for RCM are few, other than heart transplantation, strategies to further increase the number of pediatric organ donors should be considered to improve the prognosis of pediatric RCM.
Risk FactorsIn the US registry, lower LV fractional shortening and heart failure symptoms were associated with an increased risk of death or transplantation.12 In accordance with these studies,12 heart failure symptoms were risk factors for death or transplantation in the present study. However, in this study, low LV fractional shortening was not identified as a risk factor for death or transplantation. The reason why fractional shortening was not a risk factor in this study remains unclear; however, it was not reduced in most patients (median: 33%), whereas this parameter in the US registry was slightly reduced (mean z-score: −0.36). In this study, no risk factors were identified for death. This suggests that the prediction of death without transplantation is difficult in patients with RCM. Previous studies also reported difficulty in predicting poor outcome in each patient.6,12
Study LimitationsThis study had several limitations. First, the study had a small sample size. Second, due to its retrospective nature, serial echocardiographic data regarding LV function during follow up were not included. A prospective registration system for Japanese pediatric cardiomyopathy patients should be established to clarify the prevalence, mortality, and risk factors.
In conclusion, this study showed that the prognosis of pediatric RCM was poor, and that the heart transplantation rate was relatively low in Japan.
This work was supported by the Health and Labour Sciences Research Grant from the Ministry of Health and Labour (Grant number H25-102).
The authors declare no conflicts of interest.
The Medical Ethics Committee of the Tokyo Women’s Medical University (Clinical Study #2714) approved this study.
Following is a list of investigators and institutions that participated in this study:
Hiroki Mori, Mikito Shimizu-Ishido, Toshio Nakanishi (Tokyo Women’s Medical University); Tadahiro Yoshikawa, In-Sum Park (Sakakibara Heart Institute); Hiroshi Ono, Hitoshi Kato (National Center for Child Health and Development); Yasuo Ono, Masaki Nii (Shizuoka Children’s Hospital); Tadahiro Shindo, Ryo Inuzuka (Tokyo University); Hitoshi Horigome (Tsukuba University); Fukiko Ichida, Keiichi Hirono (Toyama University); Tomio Kobayashi (Gunma Children’s Medical Center); Shigetoyo Kogaki (Osaka University); Masaru Miura (Tokyo Metropolitan Children’s Medical Center); Junko Shiono (Ibaraki Children’s Hospital); Motoki Takamuro (Hokkaido Medical Center for Child Health and Rehabilitation); Mamoru Ayusawa (Nihon University); Minako Hoshiai (Yamanashi University); Shunichi Ogawa (Nihon Medical School); Kiyoshi Ogawa (Saitama Children’s Medical Center); Toshiki Kobayashi (Saitama Medical University International Medical Center); Hiromichi Hamada (Tokyo Women’s Medical University Yachiyo Medical Center); Hiroyuki Matsuura (Toho University Omori Medical Center); Isao Shiraishi (National Cerebral and Cardiovascular Center).
