Please see attached CAPN3 PVS1 flowchart. In addition, follow the SVI Working Group’s recommendations (Walker et al. 2023; PMID: 37352859) for any variant for which an impact on splicing has been experimentally demonstrated OR is predicted to affect splicing but occurs outside of a donor/acceptor +/-1,2 dinucleotide position.

For missense variants for which the amino acid change is the expected mechanism of disease, apply at Strong for 1 pathogenic or 2 likely pathogenic variants resulting in the same amino acid change. The likely pathogenic or pathogenic variant(s) must have been classified using LGMD VCEP specifications, and potential splice effects must be excluded for the missense variant under curation and the variant(s) resulting in the same amino acid change (SpliceAI score ≤0.10 or experimental evidence of normal splicing). PS1 can potentially be applied to multiple nucleotide changes at the same residue as long as the variant classification that determines the strength level does not depend on PS1 application.

For missense variants encoded by the first or last 3 nucleotides of an exon, PS1 should be considered only in the context of altered splicing (see below), unless a splice effect has been experimentally ruled out for the variant under curation and the variant(s) resulting in the same amino acid change. 

For variants for which the nucleotide change is the expected mechanism of disease (altered splicing), follow SVI Working Group recommendations (Walker et al. 2023; PMID: 37352859).

Use the SVI Working Group’s recommended point system (see supplementary file “de novo table”). Probands meeting the criteria for at least PP4_Moderate or PP4_Strong can be counted as “Phenotype highly specific for gene”. Probands meeting the criteria for PP4 (Supporting) can be counted as “Phenotype consistent with gene but not highly specific”. Probands who do not meet the criteria for PP4 (Supporting) should not be counted. Maternity and paternity should be confirmed by trio WES or other testing. PS2 should be applied for ≥ 2 pts but < 4 pts.

Variant-specific animal models are rare but may be assessed on a case-by-case basis and PS3 applied at a strength level that reflects how well the model recapitulates features observed in human patients. Apply PS3 at Strong for a variant-specific animal model that meets all of the following conditions, regardless of species: 

• signs of myopathy or dystrophy are present in skeletal muscle 
• an effect on gene or protein function is demonstrated (e.g., decreased protein expression, impaired membrane localization, or other functional abnormality)
• behavioral signs of muscle weakness
• progression over time

For CAPN3, functional studies in heterologous systems are hard to conduct and rare in the literature. Therefore, PS3 is not applicable at this time for in vitro assays for variants in CAPN3. Assays that may be considered in the future include titin ectopic expression, titin degradation assay, baculovirus-based titin cleavage assay, and assays of autolytic activity.

For any variant type, experimental evidence for altered splicing should be scored under PVS1 in accordance with the decision tree for RNA splicing assay results outlined in Walker et al. 2023 (PMID: 37352859).

Apply PS3 only once, for the piece of evidence that meets the highest possible strength level.

Use without disease-specific modification if case-control studies are available. 
While case-control studies could potentially be considered for a few pathogenic variants with high minor allele frequency, the VCEP is unaware of any such studies being conducted for CAPN3. Any case-control study would require careful selection of an appropriate control population given the potential for late onset and mild disease.

Use the SVI Working Group’s recommended point system to determine PM3 strength (see supplementary file “PM3 table”). PM3_Strong should be applied for ≥ 2 pts but < 4 pts.

Apply only for missense variants for which the amino acid change is the expected mechanism of disease. For the missense variant under curation and the variant(s) resulting in a different amino acid change, exclude potential splice effects (SpliceAI score ≤0.10 or experimental evidence of normal splicing) and confirm the applicability of PP3. Missense changes at the same residue must be classified according to LGMD VCEP specifications, and no benign missense variation should be present at the residue. Do not apply for missense variants encoded by the first or last 3 nucleotides of an exon unless a splice effect has been experimentally ruled out for the variant under curation and the variant(s) resulting in the same amino acid change. Apply at Strong for 2 pathogenic or 3 likely pathogenic variants resulting in different amino acid changes at the same residue as the variant under curation.

PM5 can potentially be applied to multiple amino acid changes at the same residue as long as the variant classification that determines the strength level does not depend on PM5 application.

Use the SVI Working Group’s recommended point system (see supplementary file “de novo table”). Probands meeting the criteria for at least PP4_Moderate or PP4_Strong can be counted as “Phenotype highly specific for gene”. Probands meeting the criteria for PP4 (Supporting) can be counted as “Phenotype consistent with gene but not highly specific”. Probands who do not meet the criteria for PP4 (Supporting) should not be counted. Maternity and paternity should be confirmed by trio WES or other testing. PM6_Strong should be applied for ≥ 2 pts but < 4 pts.

Segregations should be counted across families, with the total number of segregations determining the strength level. Apply PP1_Strong for 3 affected segregations (in addition to proband) across ≥2 families. When applied together, PP1 and PP4 cannot exceed 5 Bayesian pts (Supporting + Strong or Moderate + Moderate).

Use the PP4 table (see supplementary file “PP4 table CAPN3”) to determine the appropriate PP4 strength level. Apply PP4 only once, for a patient meeting the highest possible strength level. When applied together, PP1 and PP4 cannot exceed 5 Bayesian pts (Supporting + Strong or Moderate + Moderate).