The complexity of genetic defects associated with NMD and high relative frequency of DM1 and DM2 in Poland, 394 and 441 families respectively [24, current data from studies conducted but not published] necessitated to design a diagnostic approach based on a comprehensive analysis using different molecular techniques. In this study: (1) NGS analysis; (2) DM1 and DM2 genotyping; (3) MLPA assays; and (4) Sanger sequencing have been performed. A targeted 89 NGS gene panel was applied among 52 Polish patients suffering from NMDs. In the tested group, the preliminary clinical diagnoses of myotonia syndromes, muscular dystrophies, or myopathies were established. In total, 29 of them reached a genetic diagnosis after using TGP, placing its effectiveness at 55.8%. Regardless of the NGS data, we identified a dynamic mutation in the CNBP gene in three patients and confirmed a gross deletion in the CAPN3 gene in one individual. Altogether, the diagnostic rate of the established approach reached 57.7% (30 patients).

The most common entity identified in patients was myotonia congenita with variants in the SCN4A and CLCN1 genes. In this study only a recessive form of myotonia congenita caused by pathogenic variants in the CLCN1 gene was detected. The recent study, evaluating the functional significance of 95 different CLCN1 variants, suggests that variants resulting in dominant functional features are clustered in the first half of the transmembrane domain and alter voltage dependence of channel activation, whereas variants with recessive functional features without a shift in voltage dependence of activation are clustered in the second half of transmembrane domain of the skeletal muscle chloride channel 1 – CLCN1 protein [24]. Although the c.2680C > T variant has been widely implicated in both dominant and recessive forms of Thomsen-Becker myotonia, according to our results and population frequency data (0.3% in the European non-Finnish population) we suppose that the most common variant c.2680C > T (p.Arg894*) in the CLCN1 gene cannot be inherited as a dominant one.

One of the common variants of the CAPN3 gene: c.1746-20C > G was identified as a heterozygous in 4 patients with LGMD phenotype. Its high frequency in Poland has been previously described [25]. Until recently, its intronic variant has been considered a variant with conflicting interpretation of pathogenicity. However, Mroczek et al. (2022) showed that this variant is hypomorphic causing LGMDR1 when occurs in trans position with another pathogenic/likely pathogenic variant [26]. Many studies confirm that this variant is causal when occurs in the compound heterozygous state [25, 27, 28]. According to these findings, we can hypothesize that one of our patients, in whom compound heterozygous CAPN3:c.[700G > A];[1746-20C > G] variants together with heterozygous POLG likely pathogenic variant: c.2243G > C (p.Trp748Ser) were identified can be diagnosed with LGMDR1. However, to confirm its pathogenicity a segregation analysis in the family is necessary.

In the presented study, a gross deletion encompassing exons 2–8 of the CAPN3 gene has been also identified by MLPA in a patient, in whom the CAPN3:c.319G > A (p.Glu107Lys) variant was found by NGS. CAPN3:c.319G > A (p.Glu107Lys) variant has been described previously as a causative pathogenic variant in a heterozygous, compound heterozygous as well as together with variants in the FKRP gene [29, 30]. On the other hand, its frequency in the gnomAD database is high and reaches 1.3% within non-Finnish population. Also, numerous ClinVar submitters reported this variant as a benign or likely benign. Based on the literature, databases and our findings we assume that CAPN3:c.319G > A identified alone, even in a homozygous state, cannot be classified as a pathogenic one. However, together with another pathogenic variant, it might be implicated in LGMD. To confirm this assumption, a functional study should be performed. Since only a DNA sample was collected from one individual in the family, we have not been able to perform segregation study or functional testing to date. We are aware of this limitation. Here, we aim to note that both variants of the CAPN3 gene: c.319G > A and deletion of exons 2–8 may together be responsible for the patient’s clinical signs. However, further investigation should be carried out when possible. Moreover, skeletal muscle MRI findings are widely recognized as a useful tool in the diagnosis and clinical management of LGMDR1. Unfortunately, no patient underwent muscle MRI prior to genetic testing. We would like to emphasize that MLPA analysis is worth performing in every patient with the CAPN3 variant and a questionable diagnosis of LGMD.

Furthermore, during the study, we identified two individuals with co-occurrence of DM2 and LGMD. In one patient, CNBP dynamic mutation and CAPN3 homozygous variant have been detected (Patient 17), whereas in another individual the CNBP expansion was present together with DYSF variants (Patient 23). Presently, the patient’s phenotype corresponds with LGMDR1 rather than DM2 (Patient 17). The segregation analysis in his family showed that both parents were carriers of a variant in CAPN3 gene, whereas an expansion in the CNBP gene was maternally inherited (Patient 17). In patient 23, the segregation analysis was not available. A similar phenomenon has been already described in several individuals, who harbored point pathogenic variants in the CLCN1 [31] or SCN4A [32] genes together with expansion in the CNBP gene, and therefore, all our patients with or without point pathogenic variants in these genes were tested for DM1 and DM2.

In the studied group of 52 patients, the variants in the CLCN1, followed by CAPN3, SCN4A and SGCA genes were most frequently identified. The genetic spectrum of neuromuscular disorders varies, greatly depending on the population and/or country, the size of a tested cohort and their homogeneity or heterogeneity. In the Dutch, the most common genes related to LGMD spectrum were CAPN3, SGCA/B/G/D, ANO5 accounting for nearly 70%, whereas the remaining genes included FKRP, EMD, GMPPB, contraction of D4Z4 repeat, SMN1, FLNC, MICU1, TRIM32 [28]. In Austria, the most frequent cause of limb-girdle muscular weakness and hereditary myopathy were pathogenic variants in CAPN3, FKRP, ANO5, DYSF, SGCA [33]. However, in China and Turkey the most common cause of LGMD were variants in the DYSF and CAPN3 genes, followed by pathogenic variants in SGCA, LMNA, and other genes (DNAJB6, FKRP, SGCB, SGCD, TRIM32, POMT1, ANO5) [34], and SGCA, CAPN3, and DYSF [35]. Moreover, the presence of homozygous and compound heterozygous variant in the SGCA gene: c.850C > T (p.Arg284Cys) reported by Özyilmaz et al. [35] and this study broadens the genetic spectrum of this gene.

Among 32 different variants identified in this study, four are newly discovered and broaden the mutational spectrum of particular genes, including: (1) DMD:c.6630del (p.Asn2211Ilefs*10); (2) COL6A1:c.1029_1032delinsTTG; (3) SGCA:c.747G > A (p.Leu249 =); and (4) DYSF:c.5356del (p.Glu1786Argfs*77).

Since all genetic testing methods have their limitations, there is no single comprehensive one, suitable for all purposes. Even advanced techniques such as WES/WGS in some cases may turn to be unavailable. Furthermore, epidemiological factors may also influence a diagnostic strategy. For instance, in some countries DM2 is as prevalent as DM1 or may have a high incidence as in Finland [36]. In Poland, the incidence of DM2 is even higher than DM1 and patients present several unspecific symptoms, according to authors’ published and unpublished data [37]. In the study, we implemented a developed panel to study a group of patients with clinical diagnosis of the spectrum of neuromuscular disorders. The assessment of the targeted gene panel enriched with other methods resulted in effective diagnostics of genetic disorders in this group of patients, expanding the mutational spectrum of the genes implicated in NMDs and maximizing the diagnostic utility.