How did ankylosing spondylitis become a separate disease?

Individual patients whose disease in retrospect is compatible with a diagnosis of ankylosing spondylitis (AS) began to be described in the 19th century, at a time when "rheumatism" comprised an undefined conglomeration of ailments. In the 1890s, rheumatoid arthritis (RA) began to be extricated from rheumatic fever and gout. But what criteria should delimit the diagnosis of RA? The first assistance came with the introduction of radiology in the first decade of the new century. By the 1930s, objective radiologic distinctions between RA and A S were being made, beginning with the preferential involvement of the sacroiliac joints in AS. The first useful serologic test was developed in the 1950s: "rheumatoid factor" that eventually is present in about three-fourths of cases of RA, but is absent in AS. In the 1970s discovery of clinical associations with specific histocompatibility antigens finalized the distinction between RA and AS with the discovery that one antigen, B-27, is associated ten times as frequently with AS than with RA, while it occurs no more frequently with RA than in the general population. Associations between B-27 and certain radiologic appearances has further been mutually confirmatory of their diagnostic significance.