PMID: 30476289
DOI: 10.1093/rheumatology/key328
Objective: FMF is an inherited autoinflammatory syndrome, characterized by attacks of painful periodic fever caused by diffuse serositis and risk of secondary amyloidosis due to IL-1β-mediated inflammation. The disease appears to be transmitted through autosomal recessive mutations in the MEFV gene encoding the pyrin protein Although more than 300 variants have been reported worldwide so far, their association with symptom severity, the relative frequencies in different populations and the disease penetrance are far from being completely understood. We investigated genotype-phenotype correlations in two large nuclear families and verified whether commonly used web-based tools can usefully predict variant pathogenicity in FMF.
Methods: Peripheral blood samples were obtained from 15 patients of two families who had been diagnosed with FMF according to international criteria. The entire MEFV coding region was sequenced in all subjects, and 179 MEFV variants were surveyed with five different pathogenicity predictors.
Results: The inheritance of FMF could not be explained by traditional autosomal recessivity in both families. In silico tools demonstrated a significant association of variants’ pathogenicity with their position along the coding sequence but not with variants’ frequency.
Conclusion: By describing two large families with paradigmatic complexity of FMF genetics, we conclude that established concepts in assessing the causative role of variants identified in mutation screening cannot be easily translated into appropriate genetic counselling in FMF. Furthermore, we demonstrate that variants frequently associated with severe disease are not predicted to significantly impact protein function using in silico algorithms.