Prevalence, incidence and carrier frequency of 5q–linked spinal muscular atrophy – a literature review

Spinal muscular atrophy (SMA) is characterised by degeneration of the alpha motor neurons of the spinal cord anterior horn cells, leading to progressive proximal muscle weakness and atrophy and, in the most severe types, paralysis.

The clinical phenotype of SMA is heterogeneous, ranging from a severe to a mild phenotype. It is generally divided into three main subtypes: type I (also called Werdnig Hoffmann disease), type II and, type III (also called Kugelberg Welander disease). However, these phenotypes are seen more as a continuum rather than as distinct subtypes and sometimes further subtypes at both ends of the spectrum are observed. SMA type 0 is a very severe form with onset in utero, reduced or absent movements, contractures, and requirement for mechanical ventilation support at birth and death before six months of age, while SMA type IV is a mild late (adult) onset form that has a normal life expectancy [1, 2]. An overview of the different subtypes is given in Table 1.

SMA is inherited in an autosomal recessive manner. In most cases it is caused by mutations in the survival motor neuron 1 (SMN1, SMN ^T , telomeric) gene, located on chromosome 5q13.2 [3]. In rare cases (~4%) SMA is caused by mutation in another gene (non-5q SMA). The majority of the patients (92%) have a homozygous deletion of SMN1. In the remaining patients small mutations that abolish the production of the SMN protein are found, mostly in a combination with an SMN1 deletion (~4%) [4, 5]. A centromeric homologue of the gene, SMN2, (previously also called SMN ^C or ^C BCD541) is present in humans. SMN2 differs from SMN1 by five nucleotides of which only one (an 840C➔T transition at exon 6–7) lies in the coding sequence and is transitionally silent. This change and a change in intron 7 cause exon 7 of the SMN2 transcript to be poorly recognized by the splicing machinery, resulting in the skipping of this exon in the majority of transcripts. This results in a frame-shift and production of a protein with a different C-terminal end, which is unstable and non-functional [3, 6]. Since exon 7 is sometimes included in SMN2 transcripts, some full-length SMN protein can be produced, albeit as very low levels (~10–20%) that are insufficient to prevent disease. The number of SMN2 copies varies within the general population, and is inversely associated with disease severity as having more SMN2 copies ensures that the absolute amount of SMN protein that is produced is higher. Notably, SMN2 defects in isolation do not seem to cause the disease [7‐9]. Other modifiers that might play a role are NAIP, H4F5, GTF2H2 and PLS3 [10‐15]. NAIP, H4F5 and GTF2H2 are thought to be a modifiers due to their proximity to the SMN1 gene and NAIP also shows homology to apoptosis inhibitory proteins [12, 14, 16]. PLS3 restores the function of the neuromuscular junction, by stabilizing F-actin-dependent endocytosis [17].

The first therapy for SMA, Spinraza (IONIS-SMNRx, nusinersen), has recently been approved by the Food and Drug Administration (FDA) in the US [18] and by the European Medicines Agency (EMA) in Europe [19]. Clinical trials for other potential therapies are progressing. As such, the knowledge about the frequency of the disease becomes even more important. This review provides an overview of what is currently known about the prevalence, incidence and carrier frequency of SMA.

Published literature on prevalence, incidence or carrier frequency of SMA was identified through PubMed searches. Search terms were ‘spinal muscular atrophy’ OR ‘Werdnig Hoffmann’ OR ‘Kugelberg Welander’ AND ‘prevalence’ OR ‘incidence’, OR ‘carrier frequency’. No restrictions for language were used; however articles in other languages than English may be missed, due to the use of English search terms. Retrieved literature was scanned and all available articles performing a prevalence, incidence or carrier frequency study were used for this review. Additional publications were identified from references in the articles. Available literature published through 6th December 2016 was taken into account; no start date was used. For prevalence and incidence studies, all studies had determining the prevalence and/or incidence as primary goal. For carrier frequency studies also studies in which carrier frequency was determined for other purposes were included. All articles were appraised critically for accurate use of terminology and were reassigned if needed. For detailed methods on the analysis of carrier frequency differences between ethnic groups see Additional file 1.

To date, only a few studies have been performed to assess the prevalence and incidence of SMA. Most of these have been conducted before 1995, when the disease causing gene was identified, therefore using clinical rather than genetic diagnosis as an inclusion criterion. Generally, an estimation of the incidence of all types of SMA of around 10 in 100,000 (1 in 10,000) live births is cited [20, 21].

Since SMA is a recessive disease, there are also unaffected, heterozygous carriers of the disease. Carriers fall into four main groups of genotypes (Fig. 1). The most common one is the ‘1 + 0’ genotype (one normal, functional allele and a SMN1 deleted, disease allele). A much less common category is the ‘2 + 0’ genotype with two functional genes on one chromosome and none on the other. Furthermore, there are also ‘1 + 1^D’ and ‘2 + 1^D’ genotypes, which have one or two functional genes on one chromosome and a non-functional gene due to either a point mutation or a microdeletion on the other. These last two genotypes are very rare [56, 57]. Four or even more copies of the SMN1 gene have also been found, indicating a ‘2 + 2’ or possibly a ‘3 + 1’ genotype. This suggests ‘3 + 0’ or ‘3 + 1^D’ carrier genotypes might also be possible, however these will be even rarer.

No signs of disease have been associated with being a carrier for SMA. However, some studies suggest abnormal SMN1 copy numbers (either deletions or duplications) may increase the risk and severity of sporadic amyotrophic lateral sclerosis (ALS), although other studies have been unable to confirm this association (for a review see Butchbach et al., 2016 [58]). Furthermore, it was suggested that in the rare disorder progressive muscular atrophy (PMA) SMN1 duplications might be associated with a more severe clinical phenotype [59].

After the discovery of mutations in SMN1 as the cause of SMA, several studies into the carrier status of SMA have been performed. In contrast to the prevalence/ incidence studies, most studies have been performed outside of Europe. Some of these are population screening programmes, whereas others are large samples of the general population [46, 60‐81]. There are also studies where small population samples were analysed or the carrier frequency was estimated from healthy controls screened for SMN1 for other purposes [8, 30, 82‐99]. As mentioned before, frequencies estimated from a small population sample are less accurate. An overview of all studies is given in Additional file 2.

SMA is a severe, heterogeneous, neuromuscular disorder. The few available prevalence and incidence studies mainly predate genetic testing and were performed in small geographical areas, mainly in Europe. This highlights the need for larger, more generalizable prevalence studies.

Good epidemiological data is needed to gain insight into health care needs and for research studies and clinical trials. This is especially important in rare diseases where clinical trials require a careful planning. Furthermore, newborn screening will become increasingly important, especially now a drug has been approved and other new therapies are in advanced clinical trial stages. The introduction of new therapies is also likely to impact on the prevalence of SMA and as such may have significant resource implications for health care planning.