We have confirmed the findings of several previous investigators and shown that serum IgA is raised in AS.5"8 Our further investigations have shown that these elevated IgA levels are associated predominantly with active phases of AS as measured by ESR and CRP. There is also a comparatively smaller but definite rise in serum IgG in active phases of AS. We have previously found that both ESR and CRP correlate with clinical assessment of disease activity.12 More striking, however, is the observation that there was no elevation in the mean IgA of patients with inactive disease as measured by laboratory parameters.
The origin or site of the inflammatory activity in AS remain obscure. Inflammation obviously must occur along the axial skeleton and in peripheral joint and eye tissues, but this is unlikely to explain the marked elevation in serum IgA, a class of im-munoglobulin produced mainly within the mucosal associated lymphoid tissues of the gastrointestinal tract.9
This finding strongly suggests that some external triggering
factor is acting across a mucosal surface such as the lower gastrointestinal
tract, initiating the rise in IgA and possibly the associated inflammatory disease.
We have previously demonstrated an increased isolation of Klebsiella pneumoniae
from the faeces of AS patients with active joint and eye disease2
3 and have also found that elevated ESR and CRP both independently
correlated with positive faecal cultures for this micro-organism.12
Lymphangiographic studies indicate that regional lymph nodes draining the pelvis, pelvic colon, and rectum, the presacral and para-aortic lymph nodes, become enlarged during the active or progressive phase of AS and before the development of radio-graphic changes.8
Inflammatory changes in the large bowel have been demonstrated
in AS by rectal biopsy,13 and an association between inflammatory
bowel disease and AS has been well established." Even if the large bowel
does have an aetiological role in the patho-genesis of AS, it is still unclear
why distal sites such as the spine, peripheral joints, and the eye are affected.
The lesions may be caused by release of arthritogenic bacterial debris from
the gut such as peptidoglycans,15 but that hypothesis alone cannot
explain the strong predisposition by individuals with the tissue antigen HLA
B27 for developing AS.
One possible explanation is provided by the cross-reactivity or molecular mimicry theory,16 which suggests that some micro-organisms may carry surface antigens having stereochemical similarity to self-antigen of the host such as HLA B27.17-20
Persistent infection may lead to the production of cross-reacting antimicrobial antibodies which can also act as self-damaging autoantibodies. If these antimicrobial antibodies bind to self-antigen, then tissue damage may follow complement activation, migration of leucocytes to inflamed regions, and release of lysosomal enzymes from inflammatory cells. Such tissue damage may lead to increased production of IgG, as has been shown in other conditions,21 and this could explain the slight rise in serum IgG in our AS patients. A similar mechanism may operate in the reactive arthritis of acute rheumatic fever or in the B27 positive arthritis following infection with salmonella, shigella, or Yersinia enrerocolitica. In these examples raised antibody levels have been found against the primary infecting micro-organisms.
Whether such anti-microbial antibodies can be demonstrated in the raised IgA of patients with active AS remains to be seen. An increase in serum IgA immunoglobulin during active disease would appear to be consistent with the hypothesis that microbial infection in the gut could act as an initiating agent or trigger in the development of AS.
This study was supported by the Arthritis and Rheumatism Council. P. C. was in receipt of an SRC studentship, and we thank Mrs C. Padamsey for typing the manuscript. We are grateful to Dr A. C. Boyle for his help and support.