Natural history of Barth syndrome: a national cohort study of 22 patients

Abstract

Background: This study describes the natural history of Barth syndrome (BTHS).

Methods: The medical records of all patients with BTHS living in France were identified in multiple sources and reviewed.

Results: We identified 16 BTHS pedigrees that included 22 patients. TAZ mutations were observed in 15 pedigrees. The estimated incidence of BTHS was 1.5 cases per million births (95%CI: 0.2-2.3). The median age at presentation was 3.1 weeks (range, 0-1.4 years), and the median age at last follow-up was 4.75 years (range, 3-15 years). Eleven patients died at a median age of 5.1 months; 9 deaths were related to cardiomyopathy and 2 to sepsis. The 5-year survival rate was 51%, and no deaths were observed in patients ≥3 years. Fourteen patients presented with cardiomyopathy, and cardiomyopathy was documented in 20 during follow-up. Left ventricular systolic function was very poor during the first year of life and tended to normalize over time. Nineteen patients had neutropenia. Metabolic investigations revealed inconstant moderate 3-methylglutaconic aciduria and plasma arginine levels that were reduced or in the low-normal range. Survival correlated with two prognostic factors: severe neutropenia at diagnosis (<0.5 × 109/L) and birth year. Specifically, the survival rate was 70% for patients born after 2000 and 20% for those born before 2000.

Conclusions: This survey found that BTHS outcome was affected by cardiac events and by a risk of infection that was related to neutropenia. Modern management of heart failure and prevention of infection in infancy may improve the survival of patients with BTHS without the need for heart transplantation.

Figure 1

Bone marrow smear differential cell count data, including the percentage of the indicated granulocyte precursors. A) Normal bone marrow. B) Bone marrow from patients with Barth syndrome (mean of 4 bone marrow counts). A granulopoietic blockage was observed in the Barth syndrome bone marrow samples, but there were no additional morphologic abnormalities.

Figure 2

Box plots showing the distribution of cardiology parameters as determined by ultrasound according to age. Age is shown for the following categories: 0, birth to age 3 months; 0.25, between 3 months and 6 months; 0.5, between 6 months and 1 year; 1, between 1 and 1.5 years; 1.5, between 1.5 and 2 years; 2, between 2 and 3 years; 3, between 3 and 4 years; 4, between 4 and 5 years; 5, between 5 and 10 years; 10, >10 years old. The heart indicators are as follows: A) shortening fraction (SF), reported as %; B) ejection fraction (EF), reported as %; C) z-score of the left ventricular end diastolic diameter (LVEDD); D) z-score of the left ventricular mass (LV mass).

Figure 3

Plasma arginine levels. Levels are shown in samples from patients in the French Barth syndrome cohort (15 samples) and in samples from an entire hospital patient population (12,828 samples). The plot shows that 95% of the samples from Barth Sd patients fall below the lower 25-percentile of randomly selected patients. Y-axis: concentration in micromoles/litre. From left to right, arginine levels in the hospital population (“all”) and in Barth syndrome patients (“Barth”). The box plots show the median values (horizontal bars), the 50th percentile around the median (box), and the full range of the 95th percentile distribution (dotted vertical lines). Circles indicate the values from individual samples from Barth syndrome patients. p-values: Wilcoxon rank sum test.

Figure 4

Overall survival and survival according to birth year of the French Barth syndrome cohort. A) Kaplan-Meier plot and 95% confidence intervals showing the overall survival of the French Barth syndrome cohort. Time is expressed in years since birth. B) Kaplan-Meier plot showing the survival of the French Barth syndrome cohort according to birth year (before and in or after 2000). Even though the total number of patients is quite limited, the difference in survival is both important (survival at 5 years: 22% for patients born before 2000 and 70% for patients born in or after 2000) and statistically significant (p = 0.009). This suggests that recent progress in the management of heart dysfunction may improve the survival of patients with BTHS.