Objective: Homozygous deletions/mutations of the SMN1 gene cause infantile spinal muscular atrophy (SMA). The presence of at least one SMN2 gene copy is required for normal embryogenesis. Lack of SMN protein results in degeneration of motor neurons, while extraneuronal manifestations have been regarded as a chance association with SMA. We report on heart defects in the subgroup of congenital SMA type I patients.
Methods: Data were recruited from 65 unselected SMA I patients whose diagnosis had been confirmed genetically within the first 6 months of age. SMN2 copy numbers were analysed retrospectively and correlated with clinical findings including heart malformations.
Results: Four (6%) patients had one copy of SMN2, 56 (86%) had two and five (8%) had three SMN2 copies. Three out of four (75%) patients with a single SMN2 copy had congenital SMA with haemodynamically relevant atrial or ventricular septal defects.
Conclusions: Previous case reports of SMA I patients with congenital heart defects did not clarify whether the cardiac malformations were coincidental. Given the respective incidences of congenitally lethal SMA with a single SMN2 copy and of cardiac septal defects in humans, a chance association of both conditions would occur in less than one out of 50 million individuals. Our findings suggest that the SMN protein is relevant for normal cardiogenesis.
Statistics from Altmetric.com
Infantile spinal muscular atrophy (SMA) is characterised by motor neuron degeneration of the spinal cord and brain stem, resulting in progressive muscle weakness and atrophy. Patients are classified into three types, SMA I–III, based upon age of onset and clinical severity. More than 90% of all patients show homozygous absence of the SMN1 gene, and clinical severity is modified by the number of SMN2 gene copies.
Severe SMA along with large atrial septal defects were first been reported in three Norwegian sibs before the era of molecular genetic diagnosis.1 Over the past 12 years, several cases of genetically confirmed SMA type I patients with heart defects, mostly atrial and ventricular septal defects, have been published. However, the observation of a congenital heart defect in SMA patients has generally been regarded as a chance association, given the high incidence of both disorders. Our study provides strong evidence that severe SMN deficiency results in abnormal cardiogenesis.
PATIENTS AND METHODS
We conducted a retrospective study on the natural history of SMA type I patients in Germany in order to define a historical control group for future therapeutic trials.2 Inclusion criteria for the study were date of birth between 2000 and 2005 and molecular genetic diagnosis of functionally absent SMN1 in the first 6 months of life by one of the participating laboratories (Aachen, Cologne, Würzburg). Recruitment was through the authors contacting the referring physicians of all eligible patients rather than waiting for parents to react to a general invitation to participate in this study. The latter would have created a bias for families who are more prone to take the initiative possibly because their child is more or less severely affected. Parents completed a questionnaire on the disease course, agreed in DNA studies and gave consent to review medical histories. Consent was obtained according to the Declaration of Helsinki, and the ethical committees of the medical faculties RWTH Aachen and University of Cologne have approved the study.
Genomic DNA of the patients was isolated from peripheral lymphocytes by a simple salting out procedure. Polymerase chain reaction (PCR) based testing for the homozygous SMN1 deletion was performed and a homozygous deletion of SMN1 was present in all patients. Dosage analysis to determine the number of SMN2 copies was done by real-time PCR3 or multiplex ligation dependent probe amplification (MLPA)4 as part of this study.
Questionnaires were sent out to referring physicians of 174 patients; data sets were obtained for 66 SMA I patients (38% return rate), 62 unrelated patients, and two sibships. The reasons for no return included “unknown”, “unknown contact details of family”, and “parents or physicians emotionally upset by death of child”. SMN2 copy numbers were assessed in 65 of 66 patients; one DNA sample failed to be amplified for further analysis. In four (6%) of the 65 SMA I patients, only a single copy of SMN2 was seen. Fifty-six patients (86%) had two SMN2 copies and five patients (8%) had three SMN2 copies.
Age at onset was within the first 5 months of life in all patients (median/mean 1.2/1.4 months). Nine patients (14%) achieved head control, and none was able to maintain a sitting position. Median/mean age at death was 6.7/9.0 months.
The clinical course of the four unrelated patients with a single SMN2 copy (table 1) was most severe with prenatal onset of muscle weakness, congenital contractures and respiratory distress from birth (fig 1). Lifespan did not exceed a few months, and all infants had survived only with the assistance of mechanical ventilation. Three patients with a single SMN2 gene copy had either atrial or ventricular septal defects (table 1) that were not explained by intrauterine muscle weakness, respiratory insufficiency at birth or other causes. Diagnosis of cardiac anomalies was confirmed by repeated echocardiographic examinations in all patients. Metaphase chromosome analysis was performed in three patients and gave normal results (table 1).
In six of 56 patients with two SMN2 copies, minor cardiac anomalies were reported which resolved spontaneously: a patent foramen ovale (PFO) was recorded in four infants (associated with a hypertrophic septum in one), a patent ductus arteriosus (PDA) in one patient, combined with a PFO in another patient. A small apical VSD along with PDA was seen in one patient who had an otherwise classical SMA I phenotype and died at 11 months. She was the child of consanguineous parents who had lost four other children due to alleged sudden infant death syndrome. No cardiac malformation was documented in five patients with three SMN2 copies.
In our patient cohort of 65 SMA type I patients with homozygous deletion of the SMN1 gene, major cardiac septal defects were documented in three out of the four patients, who were retrospectively analysed to have a single SMN2 copy. No major cardiovascular malformation was observed in the fourth patient with a single SMN2 copy or in any of the 61 patients with two or three SMN2 copies. The incidence of SMA I is about one in 20 000,5 and the proportion of SMA I patients with one SMN2 copy varied in different studies between 7–19% (mean 8.5%; equivalent to an incidence for SMA type I patients with a single SMN2 copy of approximately 1: 250 000).6 Thus, patients with one SMN2 copy are mostly seen as exceptional cases in smaller case studies. Congenital heart defects are present in about 1% of livebirths, and septal defects have an incidence of about four in 1000 newborns.7 PFO and PDA are frequent observations in healthy newborns and are mostly asymptomatic. Combining the incidences of SMA I with one SMN2 copy and cardiac septal defects, a chance association would occur in less than one out of 50 million individuals. Hence, the observation that three out of four patients with one SMN2 copy had a large ASD, VSD or complete AVSD suggests a causal relationship.
In contrast to previous case reports8–16 on genetically confirmed infantile SMA with cardiac malformations (table 2) which were prone to selection bias, our study attempted to define the natural history of SMA I depending on the number of SMN2 gene copies. Since quantitative analysis of SMN copy numbers is a relatively recent technique, and since quantification of SMN2 copy numbers is not part of the routine analysis, SMN2 copy numbers are largely unknown in the previously reported cases. The clinical picture with prenatal or congenital onset, congenital contractures, respiratory distress from birth, and a very short life span suggests that in most reported patients with heart defects (table 2) only one SMN2 copy might have been present.
Cardiac malformations have not yet been documented in any animal models of SMA. Trans-genic mice in whom the murine Smn has been replaced by human SMN2 genes develop SMA of different severity.19 It was observed that the vast majority of mice with a single SMN2 gene die as embryos or shortly after birth.20 It is not known whether some of these animals have congenital anomalies. Surviving mice with higher levels of SMN expression might not develop extra-neuronal manifestations. Conditional knockout mice that were created by French investigators have shed some light on the SMN function in non-neuronal tissues. While mice with SMN deficiency restricted to the spinal cord (neuronal mutant) developed SMA,21 other cell specific pathological manifestations were induced when exon 7 of the murine Smn gene was removed specifically in muscle22 or liver.23 It was concluded that knocking out SMN in any cell would be detrimental to its survival.
It is likely that SMN has additional functions since numerous binding partners have been identified and SMN has been found in various cellular compartments. Other neurodegenerative disorders are associated with extra-neuronal manifestations in their most severe expression. In this context it is worth noting that an abnormal fatty acid metabolism was found in infants with severe SMA but not in milder forms or controls.24 Further studies are required to investigate the impact of SMN on different steps of cardiogenesis and more specifically of cardiac septation.
We are grateful to parents and physicians of patients with SMA who contributed to this work.
Funding: BW was supported by grants of the Deutsche Forschungsgemeinschaft, the Center for Molecular Medicine Cologne, the Initiative “Forschung und Therapie für SMA” and Families of SMA (USA).
Competing interests: None.
Patient consent: Parental consent obtained.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.