The cause of ankylosing spondylitis is unknown. For many years, it was considered a variant of rheumatoid arthritis but is now known to differ in several important respects. There is no strong evidence of an autoimmune mechanism, and rheumatoid factor is negative except in approximately 5% of cases. It is well known, however, that genetic factors play a significant role. The disease is 40 times more common among first-degree relatives than in the general population, and the concordance rate in affected identical twins is 70%.¹ Family pedigrees suggest that the gene for ankylosing spondylitis is inherited as a mendelian dominant trait with 70% penetrance in males and 10% in females.² It also is known that the frequency of ankylosing spondylitis is high among certain American Indian groups such as the Haida, Bella Coola, and Pima and relatively rare among blacks.^3,4

Histocompatibility (HLA) antigens are cell surface determinants present on most human cell membranes. Their specificity in humans is governed by a group of genes located on the sixth chromosome. Each gene determines a cell surface antigen so that an individual will have four cell surface antigens that are determined genetically. The frequency of an HLA type can be determined by testing a patient’s lymphocytes with various different antisera.

Although HLA typing has been used principally to match donors and recipients in organ transplantation, the association of HLA types with specific disease entities is now receiving considerable attention. In 1973, Schlosstein et al⁵ in Los Angeles and Brewerton et al⁶ in England reported a strikingly high incidence of HLA-B27 in patients with spondylitis. They found the antigen (which has a frequency of 4%–9% in control populations) in 95% of patients with ankylosing spondylitis, and comparably high percentages have been reported by other investigators.

Ankylosing spondylitis is common in families that propagate the HLA-B27 antigen. It has been estimated that if the risk that an antigen-negative female will have spondylitis is one, then the risk in a B27-positive female is 276 times greater. For a B27-positive male, the risk is 1,937 times greater.

The significance of this high frequency of HLA-B27 in patients affected with ankylosing spondylitis is uncertain. The antigen also is found in 90% of patients with a spondylitis that occurs after infection with Yersinia enterocolitica. This is a Gram-negative bacterium that causes a transient inflammatory bowel disorder that is sometimes followed by polyarthritis and sacroiliitis. In patients who recover uneventfully from the infection, the incidence of HLA-B27 is the same as for the general population, but of those in whom arthritis develops, B27 is found in approximately 90%. This suggests that spondylitis can follow an infectious disease in a genetically susceptible individual.^6B Whether this is because of a defective immune response or a hypersensitivity mechanism is not known. An infectious disease process also may be the basis of Reiter syndrome and even of ankylosing spondylitis.

As with ankylosing spondylitis, most patients with Reiter syndrome (90%) have the histocompatibility antigen HLA-B27. The reason for this association is not known. The first possibility is that the B27 antigen may act as a favorable receptor site for certain microbial pathogens, several of which, especially Chlamydia, Mycoplasma, and Shigella, have been irregularly associated with Reiter syndrome. A second possibility is that infectious agents that cause Reiter syndrome and other rheumatoid diseases share common antigens with the host’ own connective tissues. This could result in the host’s inability to recognize and combat the invading microbe, or alternatively it could result in an immune response directed against the host’s tissues as well as against the infectious agent. A third possibility, and the most widely accepted, is that HLA-B27 is linked in the genetic material to an immune-response gene and that this gene is responsible for the clinical syndrome. The immune response may be directed appropriately against the infecting agent or inappropriately against the patient’s own tissues.

There are only a few immunologic abnormalities in patients with Reiter syndrome. Cellular immunity is intact, although lymphocyte transformation has been shown with chlamydial and prostatic antigens.⁷ In some patients with Reiter syndrome, cell-mediated immunity to autologous immunoglobulin G (IgG) also has been found.⁸

In one of many attempts that have been made to implicate a microorganism in Reiter syndrome, Schachter et al⁹ propagated a chlamydial agent that they had isolated from synovial fluid, synovial membranes, urethras, and conjunctivae of patients with the syndrome. When they recovered the organism from the synovial membrane of an affected patient and inoculated it into the anterior chambers of rabbits, ocular disease (consisting of papillary conjunctivitis, corneal edema, corneal opacities, corneal neovascularization, and iritis) developed.¹⁰ Several cases of Reiter syndrome also have been reported in association with Shigella dysentery and other enteric infections, and it may be that several infectious agents can produce the syndrome in genetically susceptible individuals.¹¹

The antigenic stimulus that initiates joint inflammation in rheumatoid arthritis still is a mystery. Certain microorganisms have been implicated, but their causal roles still are uncertain. Viruses, particularly the slow viruses, may play a role, although no virus particles have been certainly identified. The results of attempts to isolate Mycoplasma and rubella virus from the joints of patients with rheumatoid arthritis have been inconclusive.¹² During the prodromal stage of hepatitis, hepatitis B virus can form immune complexes and produce a syndrome resembling serum sickness with polyarthralgia and vasculitis. This virus seems to be the first virus known to produce a chronic rheumatic disorder in humans, and, like many other infectious agents, it can give rise to rheumatoid factor in serum. The clinical syndrome, however, bears no resemblance to either rheumatoid arthritis or systemic lupus erythematosus (SLE).

Other possible infectious causes of rheumatoid arthritis have been investigated. Rheumatoid synovial cells show a diminished sensitivity to infection with either Newcastle disease virus or rubella virus. Mycoplasma antibodies have been isolated from patients with rheumatoid arthritis, especially from those with long-standing disease, and recently a slow-growing infectious agent with some of the properties of Mycoplasma has been isolated from the synovial fluid of such patients.¹³ In other studies, however, investigators have failed to find any evidence of previous Mycoplasma infection in patients with rheumatoid arthritis.¹⁴

Various immunologic abnormalities have been found in patients with rheumatoid arthritis, and there now is considerable evidence that the disease is caused by an autoimmune process. Antibodies against IgG are formed in the patient’s blood and synovial fluid. Immune complexes that are formed and deposited in the joints and other tissues activate the complement system through the classic and alternate pathways.¹⁵ Activation of the complement system results in several inflammatory phenomena, including chemotaxis of leukocytes, histamine release, and cell lysis. Enzymes released by the synovial leukocytes produce inflammatory changes in the joints and the destruction of normal structures. The inflammatory response is amplified by the various humoral amplification systems.

The humoral immune system appears to be highly active and important in the pathogenesis of rheumatoid arthritis (Fig. 33.1). The number of synovial B lymphocytes, which are precursors of antibody-producing plasma cells, often is abnormally high. More than 50% of the synovial plasma cells produce IgG rheumatoid factor, an antibody directed against other IgG molecules.¹⁰ Immune complexes may be found within plasma cells of the synovial membrane, a finding unparalleled in any other immunopathologic disorder. Although a greater-than-normal number of peripheral B cells usually are found, the number is hard to estimate because of the antilymphocyte antibodies present.¹⁶ When these antibodies are removed, the number of peripheral B cells may in fact be reduced.

The following antibodies also have been identified in the sera of patients with rheumatoid arthritis: (a) antibodies to double-stranded DNA; (b) antibodies to human native and denatured collagen; and (c) antinuclear antibodies (ANAs).^17,18 Recently, it has been shown that IgG molecules in the sera of patients with rheumatoid arthritis have a conformational anomaly in the hinge region.¹⁹ This altered IgG may be recognized as abnormal by B-lymphocyte receptors, leading to an autoimmune response directed against IgG.

The following defects in cellular immunity also have been associated with rheumatoid arthritis:

This T-cell response to IgG antigens and the fact that soluble mediators of lymphocytes contribute to the inflammatory changes that take place in the rheumatoid joint strongly suggest that there is a cell-mediated immune component in rheumatoid arthritis.

A third population of lymphocytes, lacking conventional B- and T-cell markers, may be important in the pathogenesis of rheumatoid disease. This population, known as null cells, may include the so-called killer lymphocytes (K cells) that are cytotoxic to IgG-coated target cells.²² Null cells may be responsible for the formation of “rheumatoid rosettes,” which are formed by the interaction of lymphocytes and IgG-coated indicator erythrocytes. In addition, peripheral blood leukocytes collected from patients with rheumatoid arthritis may be cytotoxic for synovial cells.²³

Antigen preparations of uvea-retina, synovial membrane, and articular cartilage inhibit the migration of leukocytes obtained from patients with rheumatoid arthritis.²⁴ In ankylosing spondylitis, inhibition is induced only by synovial membrane antigens. Lymphocytes from blood and synovial fluid of patients with rheumatoid arthritis also show a markedly diminished blastogenic response to phytohemagglutinin and pokeweed mitogen.²⁵

The reason for depressed cellular immunity in rheumatoid arthritis is unknown. It may result from a preoccupation of the host’s immune mechanism with cell-mediated immune reactions related to the pathogenesis of the disease, or it may be related to a systemic viral infection. Alternatively, depressed cellular immunity may be caused by immune complex formation or by rheumatoid arthritis therapy. The HLA-Dw4 allele occurs with high frequency and DQw7 influences severity.²⁶

Although rheumatoid factor is prevalent in adult rheumatoid arthritis, it is not commonly found in patients with JRA, with only 10% to 20% testing positively compared with 50% to 85% of those with the adult disease. Conversely, ANAs are found in 20% to 40% of children with JRA (most with pauciarticular and polyarticular JRA but a few with Still disease) and in 88% of JRA patients who have chronic iridocyclitis.^27,28 The ANAs are predominantly of the IgG immunoglobulin class, and the titers are 1:50 or higher in most patients. The consistent homogeneity of the nuclear fluorescence pattern is typical of reactivity against deoxyribonucleoprotein.²⁹ Antibodies to DNA and RNA usually are not demonstrable, but antibodies against double-stranded RNA recently have been reported in children with JRA and iridocyclitis.³⁰ A positive ANA test result usually precedes the onset of iridocyclitis and may be useful in identifying patients with JRA who are likely to have chronic iridocyclitis develop. The reason for a positive ANA test result in this disease is not known, but in general both infectious processes and host immune defects are associated with the formation of ANA and other autoantibodies.

In patients with JRA and iridocyclitis, smooth muscle antibodies, usually of the immunoglobulin M (IgM) class, are found in 15% to 23%, which is the same frequency with which they are found in control groups. The HLA antigens of patients with JRA have been examined, and some investigators have reported a high incidence of HLA-B27.^31,32 Others have not found this to be the case, however, and suggest that the earlier studies included patients with ankylosing spondylitis, which, in its early stages, can mimic JRA.^33,34

The clinical picture of Still disease suggests the possibility of a disseminated infection. Evidence indicating an infectious etiology is scant, however. A rise in antibody titer to Coxsackie B3 and A9 viruses and the isolation of adenovirus 7 have both been reported.²⁹ The titer of rubella virus antibody has been found to be elevated, and rubella virus antigen has been isolated from synovial fluid. Perhaps a viral infection, combined with defective immunity, allows an infectious agent to persist in eye and joint tissues, which leads to the development of JRA.