LETTERSmight be a precipitating factor for AAU among Japanese patients with AS, but not in persons who are otherwise healthy.We agree that in those of our patients who had AAU, the onset of AS was earlier than in those who had not developed AAU; however, this difference was not statistically significant (Student's t-test). We also compared the age at onset of AS between DR8-positive and DR8-negative patients. No significant difference in the age at onset was observed between these 2 groups. In our study, the only significant difference observed was in the association of HLA-DR8 in patients with versus those without AAU. There was only 1 patient who was positive for DR8 but did not have AAU. Since the number of DR8-positive AS patients with the early-onset type of the disease in our study population was not large enough to enable statistical analysis, it was difficult to conclude whether DR8 had any relation to the age of onset of AS. More studies with more patients are needed in order to support the hypothesis that HLA-DR8 is specifically associated with the early-onset type of AS.With regard to the LMP2 genes, it is difficult to comment about which class I1 specificity occurs with the LMP;! gene for linkage disequilibrium, since this has not been studied in a Japanese population. This might be a useful area for further investigation.In conclusion, this first study with a small number of Japanese AS patients indicated a significant association of HLA-DR8 among patients who later developed AAU. A firm conclusion can be reached only when further studies reconfirm the same observation.