Author + information
- Ricardo H. Pignatelli, MD∗ (, )
- Karen M. Texter, MD,
- Susan W. Denfield, MD,
- Michelle A. Grenier, MD,
- Carolyn A. Altman, MD,
- Nancy A. Ayres, MD and
- Subash Chandra-Bose Reddy, MD
- ↵∗Division of Cardiology, MC 19345 C, Texas Children's Hospital, Houston, Texas 77030.
Left ventricular noncompaction (LVNC) is a distinct primary myocardial disease characterized by abnormally prominent trabeculations in the ventricular myocardium, and is reported to coexist with congenital heart diseases like Ebstein's anomaly (EA) and others (1). The clinical course of LVNC with EA is unknown in the pediatric literature. We report a pediatric cohort of LVNC and EA, with emphasis on the natural course and the outcome.
We conducted a retrospective search of our institutional database from 2002 to 2007 for patients with EA and LVNC. This cohort was divided into 2 groups: group 1, patients with EA and LVNC; and group 2, patients with EA alone. We reviewed patients' medical records and collected information on the age at diagnosis, clinical presentation, electrocardiographic features, echocardiographic severity of the LVNC based on the extent of distribution of abnormal myocardial trabeculations (the extent of the distribution of abnormal left ventricular [LV] myocardial trabeculations to the lateral wall, apex, and the ventricular septum were labeled as segments 1, 2, and 3, respectively), LV end-diastolic dimension indexed to the body surface area (with z scores), ejection fraction (by Simpson biplane method) at initial diagnosis and at latest follow-up, management (medical therapy, catheter interventions, and surgeries), length of follow-up (in months), and the outcome (alive or dead). All echocardiograms were reviewed for LVNC, as defined by the standard echocardiographic criteria, by 2 independent observers blinded to clinical data.
Sixty-one patients were identified; 10 patients (16%) showed EA and LVNC (group 1) (Fig. 1). The remaining 51 patients (84%) showed EA alone (group 2). Nine of 10 patients (90%) in group 1 were diagnosed at birth with a suspicion of LVNC on the fetal echocardiogram. In addition, 4 of 10 patients (40%) showed cyanosis at birth, with right-to-left shunting across the patent foramen ovale. All patients in this group showed involvement of LV segments 1 and 2, with an additional 3 patients showing involvement of LV segment 3. There were no significant differences in the electrocardiographic findings: incidence of right bundle branch block, ventricular pre-excitation, or supraventricular tachycardia among these 2 groups.
The risk for an adverse clinical outcome was higher in group 1. Five of 10 patients (50%) in this group developed progressive LV dysfunction, of whom 3 (30%) died due to refractory congestive heart failure. Of the patients surviving beyond the neonatal period, the risk for progressive LV dysfunction was higher in group 1 (5 of 10 patients; 50%) compared with group 2 (4 of 51 patients; 8%) (relative risk: 6.375; 95% confidence interval: 2.06 to 19.66). The risk of death in group 1 was 30% (3 of 10 patients) versus 13% (7 of 51 patients) in group 2 (relative risk: 2.185; 95% confidence interval: 0.67 to 7.04).
We looked at the effect of LVNC on the outcome over short-term follow-up. Coexistent LVNC led to early diagnosis, often in utero as fetal hydrops or in the neonatal period as cyanosis, and presented with refractory heart failure leading to death in 30% in the neonatal period. In addition, over 50% of those surviving past the neonatal period developed progressive LV dysfunction requiring either medical therapy or additional interventions. In contrast, the mortality was relatively low (13%) in group 2, with 44 surviving patients and 2 patients (5%) developing progressive LV dysfunction due to dilated cardiomyopathy and requiring cardiac transplantation. Excluding deaths in the neonatal period, the relative risk for adverse outcome—either development of LV dysfunction, congestive heart failure, or death—was higher in group 1 versus group 2. Overall, patients with EA with no evidence of LVNC surviving past the neonatal period did well, with 95% remaining asymptomatic with normal ventricular function.
In summary, we noted a trend toward early detection and adverse outcome in group 1 patients. However, this study was limited by small numbers (10 patients), and we did not look at the severity of EA as a contributory factor. LVNC is known to show a variable genotypic–phenotypic clinical expression, with mutations involving several genes, and whether any of these mutations represents a specific marker for EA or for a poor outcome is unknown. Our case series extends support for in-depth studies for better understanding of these 2 conditions (1).
Please note: Drs. Pignatelli and Texter contributed equally to this work and are both first authors of the manuscript.
- American College of Cardiology Foundation