Bf polymorphism and ankylosing spondylitis

The authors have typed 44 Italian A.S. patients for Bf (all carried the B27 antigen, none were B27 homozygous): 27 had Bf SS, 14 SF, 2 S1S, and 1 FF. Phenotype frequencies of the Bf locus are in the Hardy-Weinberg equilibrium and in good agreement with our control group (χ25 6.30, 0.25<p<0.50) and with those published for a healthy, White population (χ23 3.61, 0.25<p<0.50). Family studies allowed us to define HLA haplotype in 19 patients: 17 were B27-Bf(S) and 2 B27-Bf(F). This ratio does not differ from the expected value based on the hypothesis of no association between the traits (0.25<p<0.50). The phenotypes of the remaining 25 individuals without available family members were 16 SS and 9 SF. The former 16 SS cases, though obviously B27-Bf(S), were not included in the previous sample with known haplotype for the chi square calculation. This was to avoid distortion of the S/F ratio due to the fact that in the 9 SF heterozygotes we could not determine which Bf allele shares the B27 haplotype. Nonetheless, the finding of at least 2 cases with a B27-Bf(F) association plus the fact that Bf gene frequencies in the patients were not different from those in a healthy population, suggest that the Bf locus can not be used, at least in our Italian patients, as a marker for the 'A.S. susceptibility gene'.