Objective: Familial Mediterranean fever (FMF) is a genetically recessive inflammatory disease caused by mutations in the MEFV gene. Most patients of non-Ashkenazi Jewish ancestry or those who are homozygous for M694V manifest a severe disease course, but some express a mild form of the disease. We therefore searched for other genes which could possibly be implicated in the disease phenotype. We tested MICA (major histocompatibility complex class I chain-related gene A) because it has been associated with a number of other inflammatory disorders.

Methods: One hundred fifty FMF probands and their family members were evaluated. The MEFV gene was screened by a combination of denaturing gradient-gel electrophoresis, restriction fragment length polymorphism, and amplification refractory mutation system. The MICA transmembrane polymorphism in exon 5 was analyzed after biotin-labeled polymerase chain reaction products were loaded onto sequencing gels and subjected to autoradiography.

Results: The contribution of MICA to the FMF phenotype was confirmed after adjustment for the patient's ancestry and for the MEFV genotype. MEFV was individually the most important prognostic factor for the disease. However, the impact of M694V homozygosity on the age at disease onset (OR 2.3) was aggravated if patients also inherited MICA-A9 (OR 6.3). In contrast, the frequency of attacks was found to be dramatically reduced (OR 0.16) in patients with MICA-A4.

Conclusion: We have identified the first FMF modifier locus, MICA. FMF is the first model of a Mendelian disease associated with MICA. These results clarify, at least partly, the inconsistent phenotype-MEFV correlation in FMF.