Our knowledge of spinal muscular atrophy (SMA) has advanced considerably since the earliest descriptions done by Werdnig, in 1891, and Hoffmann, in 1893. We know that it is due to a defect in the translation of the telomeric survival motor neuron (SMN) protein, which seems to play a role in several essential cellular functions (RNA metabolism, processing, and splicing) and other functions more specifically related to the survival of alpha motor neurons (apoptosis, axonal transport) in the anterior horn of the spinal cord. The SMN protein is encoded by the SMN1 and SMN2 genes. The literature describes a worldwide incidence of approximately 1 in 10,000 live births, with 1 out of 40 people being carriers of the disease.

The classical clinical picture of SMA is predominantly proximal muscle weakness and atrophy, with ages of onset and severity that vary depending on the clinical type of SMA. Traditionally, the classification of SMA comprised 3 types: infantile, juvenile, and adult. At present, the International Consensus Statement for Standard Care in SMA1,2 classifies it into more types based on the age of onset and clinical course, dividing type III into subtypes according to the age of onset. Type IV was added for adult-onset cases, and type 0 for cases with prenatal onset, which result in death within the first weeks of life (Table 1). While there is clinical variability among individuals in each type and up to 25% of the patients cannot be categorised, this classification is useful for clinical practice and prognosis.

The gene involved in SMA, discovered in 1995,3 is located in the long arm of chromosome 5 (5q11.1-13.3). It has been named SMN for “survival motor neuron”. Humans have two nearly identical copies of this gene that have been called SMN1 and SMN2. They differ in a single nucleotide at the beginning of exon 7 (C in SMN1 and T in SMN2) that is important in the splicing of the SMN RNA. We do not know why this ubiquitous protein causes this highly selective neuronal disease. Deletions in SMN1 cause SMA. Approximately 95% of SMA patients have homozygous absence of exons 7 and 8 of the SMN1 gene, and about 5% are compound heterozygous for absence of exons 7 and 8 in one SMN1 allele and a point mutation in the other. Deletions in the SMN2 gene do not cause the disease; instead, it is the number of copies of this gene, which may vary, that has an effect on the phenotype and determines the severity of SMA, with a greater number of SMN2 copies correlating to a milder phenotype, although there are other factors at play.4,5 In this regard, it has been hypothesised that the disease could be treated with gene therapy by induced overexpression of the SMN2 gene, and the day when this terrible disease can be mitigated may not be far away.6,7Another genetic approach that was successful in an experimental study was the insertion of the SMN gene by means of viral vectors in the genome of a mouse model of the disease.8

Other forms of disease with clinical features that overlap with SMA, with subtle differences, are associated to different genes. Cases associated with chromosome 11 alterations have been described (spinal muscular atrophy with respiratory distress type 1 [SMARD1])9,10 as well as cases of early-onset recessive X-linked disease in males (Kennedy's disease).11,12

The aim of this study was to determine the present incidence of SMA in our population, its genotype distribution, and its epidemiological and clinical characteristics in the past 25 years in the context of our current knowledge, and to assess the level of care and the outcomes of SMA in our setting based on the international consensus.1,2

We performed a retrospective descriptive study by reviewing the clinical histories of patients with a clinical and neurophysiological diagnosis of spinal muscular atrophy assessed in our hospital in the past 25 years (from 1987 to early 2013).

We collected data for epidemiological, clinical and genetic variables, the supportive care received, and survival. The variables analysed were age at clinical onset and age at death, sex, family history of relatives that had the disease, initial symptoms, neurophysiological examinations, genetic testing, type of SMA, subsequent prenatal diagnostic examinations, supportive care received, and survival.

The molecular analysis involved amplification of exons 7 and 8 of the SMN1 gene by polymerase chain reaction from blood DNA samples from patients with suspected SMA. The SMN1 and SMN2 genes were differentiated by restriction fragment length polymorphism analysis. The absence of one of the 3 restriction fragments was interpreted as a homozygous deletion of exon 7 of the SMN1 gene. The homozygous deletion of exon 8 of the SMN1 gene was evinced by the absence of the 189-bp band of the SMN1 gene that was not digested in the presence of the 2 restriction fragments of the SMN2 gene.

In cases in which homozygous deletions were not detected, the number of copies of the SMN gene was quantified by multiplex ligation-dependent probe amplification. In patients with a single copy of the SMN1 gene (the other copy may be inactivated by a mutation) we sequenced the DNA of the coding part of the SMN1 gene to detect potential deletions, duplications and point mutations.

At the end of the 25 years under study, the incidence in the reference population and year was similar to those described in other series, which suggests a stable genetic penetrance for this disease in our population.13 SMA type I accounted for 70.3% of the cases; the genetic defect in most of these patients (31/36) was a homozygous deletion in the SMN1 gene; the mean survival was less than one year for type I, somewhat below what is reported in the literature; and there was a predominance of male patients in our series.

Hypotonia in infancy continues to be the most common clinical presentation (Fig. 2), although in some cases (3 in our series) the presenting symptom is respiratory distress. The age of clinical onset in patients with SMA type I was 2.2 months. While the diagnosis was made relatively early, at present it is believed that this is not early enough to achieve the benefits expected of emerging therapeutic approaches.14–17 The usefulness of neonatal screening for offering specific patients participation in clinical trials is being considered. This approach has already been implemented and there are studies on the subject.18

We found a spontaneous miscarriage rate of 19.1% in the families of patients with SMA type I, at the upper limit of the range described for the general population (10–20%). However, this figure may underestimate the actual rate of spontaneous miscarriage associated with this disease, as we have no information on the occurrence of subsequent miscarriages in families that were not followed up in our hospital. Although this has been described before, there are very few published studies on the association of SMA with spontaneous miscarriage, and we did not find any systematic reviews on this matter. A published case associated foetal death to foetal akinesia-deformation sequence.19

Only one of our patients had a form of SMA unrelated to the SMN gene. The patient was diagnosed with SMARD1. The polymorphisms found in this patient had never been described before. We assumed they were involved in the aetiology based on their location in chromosome 11 and the patient's clinical features. The management of this case involved early ventilatory support with tracheostomy due to its onset with severe respiratory distress for which no initial clinical diagnosis was made, as the patient followed the natural course of this disease in its early stages with severe respiratory problems and predominantly distal involvement.20 Several publications debate the use of invasive respiratory support in SMA type I, which poses an ethical dilemma that has yet to be resolved.21 Since 2007 there is a standard of care consensus22 that aims to improve the quality of life of SMA patients, although according to several studies in different countries its introduction has raised controversy.23

In our experience, implementing this standard of care, as we have done in recent years, requires the coordinated effort of the various specialties involved, as well as adequate development of at-home hospital care. It requires, at least, the universalisation of at-home oxygen therapy, gastrostomy and noninvasive ventilatory support during the course of the disease, in addition to the optimisation of orthopaedic measures to prevent bizarre deformities, while adjusting care to the wishes and expectations of the parents. It remains to be seen whether this approach increases survival, and while this is not the goal of the care, it could have this logical result. Patient associations support this standard of care, which must be taken into account by paediatricians. The median survival for type I patients was 8 months, lower than the figures published to date, which was probably due to our late introduction of active nutritional and respiratory support. This precludes us from drawing conclusions about overall survival, although this approach seems to alleviate the anxiety of parents in regards to the suffering of their child.

In our series, tracheostomy was performed in only two patients. One was the patient with SMARD1, who had it for the reasons stated above, and the other one a patient with SMA type II, in whom it was more likely to be indicated given the slower course of the disease. Nutritional support was only provided to one patient, after the parents agreed, through a scheduled gastrostomy at 16 months of age. A cough assist machine was used in two patients with SMA type I at 8 and 23 months of age, resulting in different degrees of satisfaction, as in one case it worsened the patient's condition and the parents decided to discontinue its use.

The genetic mutations found in our series are similar to those reported in the literature (Table 2). In 2011, the only 2 families in which the truncated copy of the SMN1 gene with the deletion of exons 1–6 was detected came from the same geographical region, and in both cases the copy was inherited from the father. Analysis of the polymorphic markers linked to the SMN1 locus and of multilocus markers (272 and 212) revealed that the fathers of the two patients had the same haplotype, which suggested that both chromosomes shared a common ancestor.

To conclude, our series shows a classical clinical course with a relatively early onset and a stable population incidence of approximately 1 case per 10,000 live births per year. Most patients had a typical genetic basis, and there was a predominance of males. Approximately 1 in 10 patients had a mutation other than the classical mutation in the SMN gene. Only one of the 37 patients in our series had a mutation in a gene other than the SMN gene that corresponded to a case of SMARD1. The overall survival in our series was below that described previously. The level of care has increased in line with social and welfare demands in recent years.

We believe that multicentre cooperation is needed to adequately address the research and management of this disease, for which there is an increasingly hopeful future.

The authors have no conflicts of interest to declare.