39 Articular and ocular complications of inflammatory bowel disease
TIMOTHY R. ORCHARD AND DEREK P. JEWELL
Introduction
Articular and ocular complications of inflammatory bowel disease (IBD) are relatively common, and were recognized as far back as thel9tli century. Despite causing a significant morbidity in IBD patients they have not been widely studied and their pathogenesis remains unclear. Recent studies have produced a clearer classification, and this in turn has led to new pathogenetic insights.
Arthritis
Historical considerations and classification Peripheral arthritis
Hale White described postmortem abnormalities in the joints of patients with ulcerative colitis (UC) in 1895, and Bargen described arthritis in association with UC in the 1920s. Initially this was thought to be coincident rheumatoid arthritis, but in 1958 Bywaters and Ansell published a paper demonstrat- ing that the arthritis was inflammatory with a lymphocytic infiltration of the synovium, but that it was not erosive or deforming and produced a negative rheumatoid factor [1].
Two large studies in the 1960s examined periph- eral arthritis in UC: a retrospective study of 675 patients from Oxford demonstrated a polyarthritis [2], whereas a prospective study of 269 patients from Leeds showed a large joint arthritis in 11.5% of patients [3]. Subsequent studies in Crohn's disease (CD) found a similar distribution of disease [4].
However, many studies did not differentiate between arthritis and arthralgia. This is potentially a very important distinction, as was demonstrated by a study from Israel [5]: this compared arthritis and arthralgia in a group of 54 CD patients with age- and sex-matched controls. Forty-four percent of the CD patients complained of arthralgia, but 46% of
controls also had joint pains. When objective evidence of joint inflammation was sought 7.4%) of CD patients had arthritis, compared to none of the controls. Thus arthralgia is a common problem in both CD patients and healthy controls, but inclusion of patients with arthralgia in studies of arthritis may considerably distort the results.
Recently a large study from Oxford studied arthri- tis in 976 UC patients and 483 CD patients, studying only those with objective evidence of articular inflammation [6]. This found two distinct forms of peripheral arthritis associated with IBD, both sero- negative, with differing patterns of joint involvement and natural histories.
1. Type 1 (pauciarticular) affects less than five joints including a weightbearing joint and is asymme- trical. The swelling is acute and self-limiting and associated with relapse of the IBD in the majority of cases. It lasts for a maximum of 10 weeks, although, like reactive arthritis (ReA), 10-20%
will develop persistent problems. There is a strong association with other extraintestinal features such as erythema nodosum and uveitis.
2. Type 2 (polyarticular) affects five or more joints, and affects a wide range of joints but particularly the metacarpophalangeal (MCP) joints. It may cause persistent problems with a m e d i a n duration of 3 years. It is associated with uveitis, but not erythema nodosum.
The onset of arthritis may occur at any time during the course of the IBD or before it becomes clinically manifest. It is not related to disease extent in UC but in CD it is possibly more common in colonic disease.
In both types of arthropathy there is little or no joint destruction and patients are seronegative.
Stephan R. Targan, Fergus Shanahan andLoren C. Karp (eds.), Inflammatory Bowel Disease: From Bench to Bedside, 2nd Edition, 747-756.
© 2003 Kluwer Academic Publishers. Printed in Great Britain
In the Oxford series of 1459 patients the preva- lence of type 1 was 3.6% in UC and 6.0% in CD and for type 2 was 2.5% in UC and 4.0% in CD. These figures are shghtly lower than in some other studies, which may reflect the strict entry criteria and the retrospective nature of the study. In total between 5%
and 10% of UC patients develop peripheral arthritis and 15-20% of CD patients. This study demonstrates that careful clinical characterization of patients may lead to the detection of new clinical entities. This in turn may lead to new pathogenic insights.
Axial arthritis
In some ways the association of axial arthritis with IBD is clearer than peripheral arthritis. Both ankylosing spondylitis and isolated sacroiliitis are known to be associated with IBD, but there remain a number of questions unanswered.
Ankylosing spondylitis (AS) is a seronegative inflammatory arthropathy afl'ecting the vertebral column, characterized by sacroiliitis and progressive ankylosis (fusion) of the vertebral facet joints. This progressive fusion leads to a characteristic 'question mark' posture and may lead to respiratory embar- rassment secondary to poor chest expansion and upper lobe pulmonary fibrosis. It is associated with peripheral arthritis in about 30% of cases. Its prevalence in the general population is between 0.25% and 1%, with a male:female ratio of 3:1 [7, 8].
The prevalence of AS in IBD is 1-6% [9, 10], varying according to the study population, and the propor- tion of female patients is higher than in idiopathic AS, accounting for up to 50% of patients in some series. Otherwise the clinical features are identical to idiopathic AS and it runs a course independent of the IBD.
Whilst AS, with sacroiliitis, back pain, restriction of movement and respiratory embarrassment, is progressive over a number of years, isolated sacro- iliitis, without these features, has also been described in IBD. Its prevalence is largely dependent upon the means of diagnosis, and it is often asymptomatic.
Radiographic surveys suggest a prevalence of up to 18% [10], but more recent studies have suggested a higher prevalence. Diagnosis on radiographic grounds alone is hampered by a large degree of inter-observer and intra-observer error [11]. Compu- terized tomographic (CT) imaging studies have detected sacroiliitis in 32% of patients with IBD [12]
and studies using radioisotope scintigraphy have found uptake abnormalities in up to 52% of patients with CD and 42% with UC [13]. However, the
significance of these findings is not clear, and long- term follow-up studies to demonstrate progression to AS have not been undertaken. It may be that in the majority of patients this is a non-progressive condi- tion.
Pathogenesis Spondyloarthropathies
Before considering the IBD-associated arthro- pathies it is worth considering the spondyloarthro- pathies, with which they are often classified. AS is the model for this group of rheumatological conditions.
It includes reactive arthritis (post enteric and uro- genital), psoriatic arthritis and IBD-associated arthritis [14]. They are all seronegative inflammatory arthritides, and they share important clinical and pathogenic features. These include the presence of i n f l a m m a t o r y low back p a i n , an i n c r e a s e d prevalence of AS, an association with erythema nodosum and uveitis, and (with the exception of IBD peripheral arthritis) an association with HLA- B27. This association was first recognized in AS, where it is strongest, with 94% of patients possessing HLA-B27 in a recent study, compared to 10% of the general population [15]. The association is weaker, but still significant, in the peripheral arthritides: In reactive arthritis approximately 70% of patients possess HLA-B27, although the prevalence varies widely between studies [16]. In the peripheral arthri- tides it appears that it is those patients who are HLA- B27-positive who develop the long-term complica- tions of disease - namely sacroiliitis, acute uveitis or recurrent arthritis [17, 18].
Behcet's disease is sometimes included in the spondyloarthropathy group, but its inclusion remains controversial [19]. The reported prevalence of HLA-B27 is only modestly increased, and the association with sacroiliitis is weak. However, there is an increase in mucocutaneous complications such as erythema nodosum and uveitis (although this is characteristically posterior, in contrast to uveitis in spondyloarthropathy which is anterior).
IBD arthritis
IBD arthritis is likely to be caused by the interaction
of genetic and environmental factors.
Genetic factors Axial arthritis
As mentioned above, AS associated with IBD appears clinically to be identical to idiopathic AS, and it runs a course independent of the IBD. Some authors have suggested that it is generally milder than idiopathic AS, but ascertainment bias is a problem in these studies, and currently there is no consensus on this point.
The HLA- B27 association of IBD AS is different, being considerably weaker: 50-80% of IBD AS patients possess HLA-B27 [20-22] compared to 94%
of idiopathic AS patients [15]. The pathogenic significance of this difference is unclear, but some possibilities are discussed later.
The genetic associations of isolated sacroiliitis are less clear. This is partly because it is not clear what proportion of patients with sacroiliitis have early progressive disease (and will ultimately develop AS), and what proportion have true isolated sacroiliitis.
Large follow-up studies are required to address this question.
The only study of sacroiliitis performed so far is a small study in 136 patients with CD, which studied patients with symptomatic back pain [23]. This demonstrated that 29%) of these patients had evidence of sacroiliitis, including 7%o who had AS.
Those patients with sacroiliitis in the absence of AS were less likely to be HLA-B27-positive. However, a full clinical assessment was not undertaken to differentiate inflammatory from mechanical back pain, and the study did not examine patients with asymptomatic sacroiliitis.
Thus it is probable that patients with isolated sacroiliitis do not have an increased prevalence of HLA-B27, and that HLA-B27 is a marker of progressive axial disease rather than sacroiliitis per se, but clearly long-term follow-up studies are required.
Peripheral arthritis
Most studies of the peripheral arthritis of IBD have been small and have not made any distinction between different patterns of disease. Possibly as a result of this they failed to find any association between HLA-B27 and arthritis.
In a recent study of the immunogenetics of IBD arthritis, patients were subdivided according to the clinical classification described above. By doing this, distinct HLA associations were discovered [24].
These are shown in Table 1.
Table 1. HLA associations of peripheral artiiropathy in IBD (percentages)
HLA type HLA-B27 HLA-B35 HLA-B44 HLA-DR103
Type1 arthropathy
(A? = 3 0 )
27*
33**
13 40™
Type 2 arthropathy
(A7 = 3 0 )
3 7
63***0
Controls (n = 603)
7+
15++
31+++
3++++
Type 1 vs controls: + p = 0.001, RR = 4.0; + + p = 0.01, RR2.2;
+ + + + p < 0 . 0 0 0 1 , R R = 12.1.
Type 2 vs controls: + + + Pc = 0.01, RR = 2.1.
Type 1 vs type 2: *p = 0.03, RR = 8.0; **p = 0.02, RR = 5.0;
***p = 0.0001, RR = 4.8; ****p = 0.0001, RR = incalculable.
Type 1 (the large joint arthritis) is associated with HLA-B27, but type 2 has a distinct association, with HLA-B44. The association of type 1 with HLA-B27 is, perhaps, unsurprising, as type 1 peripheral arthri- tis (PeA) is clinically very similar to ReA, in which approximately 70% of patients are HLA-B27-posi- tive. In addition in type 1 there was a very strong association with the rare HLA class II allele HLA- DR103 (DRB1*0103). This association is also seen in postenteric ReA. Interestingly in ReA the HLA- B27 association is twice as strong as in type 1 IBD PeA, whereas the HLA-DR103 association is twice as strong in type 1 IBD PeA.
The pathogenic significance of these associations (and any functional significance) is difficult to gauge, and the associations described may simply represent linkage disequilibrium with pathogenic genes located close by. This possibility is well illustrated by type 2 PeA. In this condition there is a strong association with HLA-B44, but there is a stronger association with the nearby gene MICA (MHC class I chain-like gene A). MICA codes for a non-classical HLA molecule which is known to be expressed on the gastrointestinal epithelium, and is up-regulated under conditions of cellular stress. It interacts with y5 T cells, but does not appear to require exogenous antigen for this interaction. Its role appears then, to be in regulating the immune response, and so is a good c a n d i d a t e for involvement in i m m u n e - mediated inflammation. Ninety-nine percent (44/
45) of patients studied with type 2 PeA possessed the
MICA*008 allele compared to 72% of IBD controls
and 73%) of healthy controls (p = 0.001).
The associations described above provide good evidence that genes in the MHC region of chromo- some 6 play an important role in both axial and peripheral arthritis associated with IBD. However, the exact nature of this role remains unclear.
Environmental factors
There is little known about the environmental factors involved in the pathogenesis of arthritis in IBD, but some clues can be gained from studying similar conditions in humans and animal models.
It has been suggested that idiopathic AS is asso- ciated with antibodies to Klebsiella species [25], and that an abnormal immune response involving HLA- B27 is of pathogenetic importance. In AS associated with IBD the association with HLA-B27 is less marked, and this might suggest that in the presence of an inflamed gut, with increased permeability, more antigen may be presented, allowing other HLA-B alleles to act pathogenically. However, in IBD no associations between AS and bacteria have been demonstrated and in idiopathic AS the evidence in favor of Klebsiella as opposed to other gut flora is unconvincing.
In contrast, the association between postdysen- teric ReA and bacterial infection is well established.
Gram-negative enterobacteria such as Salmonella, Escherichia coli, Yersinia, Klebsiella and Campylobac- ter are all associated with arthritis, and bacterial antigens may be isolated from the affected joints.
These conditions are clinically very similar to type 1 PeA seen in IBD, and so it seems likely that presenta- tion of bacterial antigen may be important in the initiation of type 1 IBD arthritis. This is plausible given the association between type 1 arthritis and active disease. In these circumstances the gut is inflamed, and therefore more permeable; a similar situation to acute bacterial enterocolitis. ProHfera- tive T cell responses to the relevant bacteria have been demonstrated from the synovial fluid of patients with ReA, but interestingly most of these appear to be HLA-DR restricted rather than HLA-B restricted [26-28]. This suggests that the HLA class II alleles may be more important than HLA-B27 in the peripheral arthritides associated with entero- pathy.
Further information regarding the role of bacteria in initiating arthritis in the presence of gut inflamma- tion has been gained from the study of the HLA-B27 transgenic animal models. Whilst the rat and mouse models differ in some respects they provide useful
working models. These animals spontaneously develop a colitis and arthritis when reared under normal conditions [29, 30]. However, if reared in a germfree environment, the gut and joint inflamma- tion is abrogated [30]. Furthermore Rath and colleagues have demonstrated that different bacteria induce gut and joint inflammation with differing efficiency, with Bacteroides vulgatus and a cocktail of bacteria isolated from CD patients being the most efficient, whilst E. coli is ineffective [29, 31].
Thus it appears likely that bacteria are important in the pathogenesis of the type 1 PeA of IBD and possibly IBD-associated AS, but the mechanisms by which they interact with the immune system are unclear.
The exact site within the gut where this interaction occurs is also a matter for debate, and in this area animal models may also provide some useful infor- mation: Rath et al. have demonstrated in the HLA- B27 transgenic rat that diversion of the fecal stream away from the cecum abrogates distant inflamma- tion, whilst leaving the colitis unaffected [32]. They have postulated a role for bacterial overgrowth in the cecum in the pathogenesis of extracolonic inflamma- tion. In humans the cecum is relatively much smaller than in the rat, and so these data cannot be extra- polated directly. However, in a study of 434 patients with CD it has been shown that there is a significant decrease in the incidence of new joint complications after resection of the ileocecal region - from one complication for every 89 years of follow-up to one complication every 701 years of follow-up [33]. This is illustrated as Kaplan-Meier survival curves in Fig.
1. This is highly significant, even when correcting for the time spent in remission from CD after surgery.
Although this study provides circumstantial evi- dence that the ileocecal region may be important in the development of arthritis in IBD it does not help determine whether it is stasis proximal to the ileo- cecal valve or cecal bacteria that are of importance.
Neither is it clear whether it is the interaction of bacteria with up-regulated HLA class II molecules in the ileum or with newly induced class II molecules in the colon (or another mechanism) that is important.
Taking the genetic and environmental data
together it is possible to hypothesize concerning the
mechanisms involved in the arthritides associated
with IBD. Classical type 1 peripheral arthritis of
IBD occurs in genetically susceptible individuals,
and may be caused by the interaction between gut
bacteria and HLA class II molecules in the context of
1 -
.8 -
13 -6 -
>
1
00 B
A -
.2 -
0 -
p J 1 1 1 1 1 1 1 I
V
J—I
I
1 1 1 1 1 1 1 L _
ICR
No ICR
p=0.0001
200 250
Time (months)
Figure 1. Kaplan-Meier curves of survival free of joint complications in Crohn's disease patients who had never undergone surgery and patients after ileocecal resection (ICR).
an actively inflamed, and therefore abnormally permeable, gut. Whether this interaction is proximal or distal to the ileocecal valve is unclear, and further work is required.
In contrast, the axial compHcations, such as AS, seem more likely to be mediated by typical or atypical interactions between HLA class I molecules and the immune system, although intestinal inflam- mation may make the conditions more favorable for the development of axial inflammation. This would be consistent with the observation that the HLA-B27 association in IBD is weaker than in idiopathic AS, suggesting that the presence of intestinal inflamma- tion may allow the development of joint disease even in patients without the usual genetic predisposition.
The pathogenetic mechanisms in type 2 peripheral arthritis are less clear, and although there is a strong HLA class I association this may be secondary to linkage disequilibrium with other genes in the region such as the non-classical class I gene MICA.
All the hypotheses described above remain largely speculative, and there are others, including the suggestion that arthritis and other extraintestinal manifestations are an autoimmune reaction to self
antigens such as isoforms of tropomyosin found in the gut, eye and joint [34]. However, these do not account for the HLA associations found, and the evidence has not yet been widely replicated.
Clinical features Diagnosis
Axial disease
Symptomatic disease
Low back pain is a common complaint in both IBD and the general population. In IBD it may represent inflammatory arthritis of the sacroiliac joints (sacro- iliitis) or progressive AS. It has been suggested that the course of AS associated with IBD is less severe than idiopathic AS, but this is not universally accepted. In a patient presenting with back pain it is therefore important to distinguish between mechan- ical low back pain and the inflammatory back pain associated with sacroiliitis and AS.
The clinical features of inflammatory low back
pain are an insidious onset over months, morning
stiffness and exacerbation of pain by rest and pain
radiating into the buttocks (rather than central back pain). It tends to occur in patients under 40 years of age. Mechanical back pain may be of sudden onset, is often central and is better after rest, occurring generally in older patients. Plain radiographs of the sacroiliac joints are the conventional means of diag- nosis, but X-ray changes occur only after several months and magnetic resonance imaging scanning (with or without gadolinium enhancement) is more sensitive and avoids the necessity of a radiation dose.
If radiologic evidence of sacroiliitis is found then a further assessment should be made to detect evidence of progressive axial disease. This should include the modified Schober test of lumbar flexion, lateral lumbar flexion and chest expansion.
The most useful blood test in diagnosis is H LA- BI? status, although in IBD a negative HLA-B27 status does not preclude the diagnosis of AS, and patients with low back pain or decreased spinal mobility in the presence of sacroiliitis should be treated as having AS .
A symptomatic disease
Up to 20% of IBD patients may have asymptomatic sacroiliitis detectable by plain radiology, and it may be diagnosed on the basis of routine abdominal X- rays. In many cases there is no history of inflamma- tory back pain even on direct questioning. If sacro- ilitis is diagnosed a clinical assessment of spinal mobility should be undertaken along with HLA-B27 status. If there is evidence of decreased spinal mobility patients should be treated as having early AS.
Peripheral arthritis (PeA)
As mentioned above, there are two distinct forms of PeA, and in addition arthralgia may occur. This is often in conjunction with reducing doses of cortico- steroids or the c o m m e n c e m e n t of i m m u n o - suppressants such as azathioprine. These normally settle with time and rarely require specific treatment.
Both forms of arthritis are rheumatoid factor- negative and are not generally erosive or deforming.
If there is evidence of erosive disease then further investigation by a rheumatologist is required.
Diagnosis of IBD PeA is largely based upon clinical grounds, but other joint disease should be excluded.
For type 1 arthritis the differential diagnosis includes gout and pseudogout, septic arthritis and arthritis following genitourinary infections such as
Chlamydia and gonorrhea. These causes should be actively sought, particularly where there is no appar- ent relation to activity of the bowel disease, and serum urate and calcium should be checked routinely. Joint aspiration and examination should be performed in cases where there is any doubt. For type 2 arthritis a rheumatoid factor and autoanti- body screen should be performed to exclude sero- positive rheumatic disease. Arthritis associated with IBD is very rarely erosive or deforming and is not associated with other manifestations of rheumatoid arthritis such as subcutaneous nodules. X-rays of the most severely aff'ected joints should therefore be performed if symptoms are persistent, to exclude erosive disease.
Management of IBD arthritis
General points
The management of arthritis in IBD depends upon the nature and duration of the arthritis, as discussed below. The aims of treatment are to alleviate symp- toms and preserve joint function for the future, particularly in axial disease. Because of the relatively small number of patients involved no controlled trials of treatment have been conducted and most treatment modalities rely on general principles.
Physical treatments (rest, range of movement exercises and physiotherapy) are important compo- nents of management and should not be ignored.
These may be enhanced by simple measures such as splinting affected joints and the use of assistive devices such as a walking stick.
Analgesia is a key element in management, and is a potentially difficult area, as many analgesics have undesirable effects on the gut. A stepped approach to the prescription of analgesics should be used, starting with simple analgesia such as paracetamol, progressing to stronger analgesics as required.
Analgesics containing opioids may cause proble- matic constipation, particularly in patients with resistant distal colitis.
Non-steroidal anti-inflammatory drugs (NSAIDs)
should be avoided if at all possible. They may cause
enterocolitis in their own right and may trigger or
exacerbate relapse of pre-existing IBD. This may be
associated with significant lower gastrointestinal
bleeding. They should not be used at all in patients
with active disease, and in these situations other
therapies should be used. The recent advent of
cyclo-oxygenase2 (COX-2) specific NSAIDs has
raised the prospect of fewer unwanted gastro- intestinal problems, particularly in the stomach.
However, experiments in animals have demonstrated that the enterocolitis of NSAIDs is not mediated by the cyclo-oxygenase pathway, and that the role of the COX isoforms in inflammation may be quite complex. It is therefore likely that COX-2 specific NSAIDs may not improve the tolerability of these drugs in IBD patients, and they should still be avoided.
Axial disease
Patients with AS should be managed in conjunction with a rheumatologist. Physical therapies are of particular importance and all patients should take regular exercise to maintain the mobility of the spine.
This may include swimming and spinal exercises with regular physiotherapy if necessary. These forms of exercise should also be recommended to patients with sacroiliitis associated with any form of low back pain or decreased spinal mobility. For patients with severely reduced spinal mobility or severe low back pain injection of steroid into the sacroiliac joints may help, particularly in association with a period of intensive inpatient physiotherapy. The period of relief from sacroiliac injection alone may be brief.
Drug treatments
Simple analgesics should be used if possible.
NSAIDs are the drugs of choice in idiopathic AS, and if there is active spinal disease in the absence of active IBD then it is reasonable to use NSAIDs.
However, they should be stopped if the IBD flares up. Sulfasalazine can be used for both the IBD and joint symptoms, but it is most effective in patients with associated peripheral joint problems. Other 5- aminosalicylic (5-ASA) drugs are not as effective, as it is thought to be the sulfapyridine component that confers the articular effects rather than the 5-ASA.
If oral steroids are required for active disease then they will also have a beneficial effect on the spinal disease. However long-term steroid therapy for spinal disease alone should be avoided. This may be achieved by using methotrexate, which may be effective in both gut and spinal disease.
In severe progressive disease unresponsive to the measures outlined above radiotherapy remains a last resort; however, the increased risk of hematologic malignancy associated with this form of treatment means it is rarely used.
The most important part in the successful manage- ment of patients with AS and IBD is good commu- nication between patient, gastroenterologist and rheumatologist to maximize the effectiveness of the available therapeutic options.
Isolated sacroiliitis should be treated sympto- matically with simple analgesia, and regular exercise should be encouraged. However, further treatment should not be required in the absence of decreased spinal mobility or severe pain.
Periplieral joint disease Type 1 arthritis
This is usually self-limiting, so treatment is largely symptomatic.
Resting the joint is important, and use of a walking stick or splint to take pressure off'the joint may lead to a significant improvement. Range of movement exercises should be performed to minimize any periarticular muscle atrophy and to prevent contrac- tures. In severe cases formal physiotherapy may be required to improve function.
Analgesia should be with simple analgesia, and, as type 1 arthritis is usually associated with active bowel disease, NSAIDs should not be used. A good, but relatively seldom used, therapy is intra-articular injection of steroid. This may provide very effective symptom relief and may remove the requirement for other treatments. If oral steroids are used to treat the active bowel disease then these will normally treat the arthritis effectively. If not, an empirical change of 5-ASA drug to sulfasalazine may give good symptom relief. If sulfasalazine fails then low doses of oral steroid specifically for the joint disease may be effective, but this should not be prolonged for more than a few weeks.
Long-term treatment is not usually required, although maintenance with sulfasalazine as the 5- ASA of choice may be appropriate, particularly in patients at risk of recurrent disease such as those who are HLA-DR103-positive. In the minority of patients with persistent problems the treatment options are those of type 2 arthritis (see below).
Type 2 arthritis
These patients generally have persistent problems and may require long-term treatment. Again, as the disease is usually non-erosive and non-deforming symptomatic relief is the major aim.
Again, splinting of affected joints and rest are
important components of management, but the
persistent and polyarthritic nature of this condition makes this harder to achieve than in type 1 arthritis.
Simple analgesia should be tried initially. NSAIDs should be considered only in patients with quiescent disease unresponsive to simple or combination analgesia, but if there is any evidence of an increase in the activity of the bowel disease NSAIDs should be stopped.
For patients with persisting problems sulfasala- zine, or low-dose prednisolone, may be used. How- ever, prolonged courses of oral steroids should be avoided. In patients with active bowel disease it may be appropriate to use methotrexate as the first-line immunosuppressant rather than azathioprine, in order to treat both gut and joints.
Patients who show evidence of erosive joint disease, a positive rheumatoid factor, or who do not respond to the measures outlined above, should be managed jointly with a rheumatologist.
New therapies for IBD arthritis
The major possible new therapy for IBD-associated arthritis is the use of infliximab. This is a monoclonal antibody to the proinflammatory cytokine Tumor necrosis factor alpha. This drug is hcensed for use in resistant CD and rheumatoid arthritis, where it has been demonstrated to be effective in improving symptoms and decreasing inflammation. Small open-label trials in spondyloarthropathy have demonstrated that it is again effective in reducing inflammation and symptoms [35, 36]. In IBD-asso- ciated arthritis the published experience is limited to case reports but, unsurprisingly here also a reduction in inflamed joints and inflammatory markers has been reported [37]. Thus infliximab may well be an alternative treatment in patients with severe resistant joint disease associated with IBD. As in other
rheumatic diseases this may be combined with other immunosuppressants such as methotrexate [38].
However, further studies are required to confirm its efficacy and the best treatment regimens.
Ocular manifestations of IBD
Epidemiology and clinical features
Ocular inflammation in IBD was first documented by Crohn in 1925. If left untreated it is potentially a cause of blindness; however, prompt treatment with topical steroids can minimize this risk.
The prevalence of ocular inflammation varies widely between studies - from 2% to 13% depending
on the population and methodology. Various forms have been described including iritis, episcleritis, scleritis and anterior uveitis. In our retrospective study of 1459 patients (976 UC and 483 CD patients) 3% of UC and 5% of CD patients had eye complica- tions. The commonest were iritis (60%), episcleritis (30%) and uveitis (10%). However, in the vast major- ity of cases the inflammation caused no lasting ocular damage [39].
The female:male ratio was 2.9:1, the eye complica- tions were present at or before diagnosis in 19% of cases, and in 72% of cases the eye complications were associated with active IBD. Thirty percent of patients went on to have recurrent episodes of ocular inflam- mation.
In common with other studies patients who suffered eye complications were more likely to suffer other mucocutaneous complications, notably arthritis and erythema nodosum.
Pathogenesis
The cHnical observation that ocular inflammation occurs more commonly in those patients with other extraintestinal manifestations has led to several hypotheses to explain it.
The first suggested that the eye, the joints, the skin and the liver expressed a particular isoform of tropomyosin, to which an autoimmune reaction was generated [40]. However, although this antigen was found in the target organs it was not found in the components that became inflamed. Thus it was expressed in chondrocytes but not synovium, and in ciliary muscles but not the iris. In addition this hypothesis fails to explain why in some patients ocular inflammation may occur as a single manifes- tation and in others as part of a complex of extra- intestinal manifestations (EIM).
This phenomenon of the EIM being distinct but overlapping clinical entities may be explained by a genetic-based hypothesis. This suggests that the presence of the diff'erent EIM is determined by different genes located in the same region. Linkage disequilibrium between these genes would mean that the genes for different manifestations are inherited together more frequently than would be expected by chance, leading to the clinical clustering of the EIM.
Given the HLA associations described above we
have studied the HLA-B and DR regions in 52
patients with ocular complications of IBD. This
study found that there are associations between
ocular complications and HLAB27 (40% vs 9% in
IBDcontrols,/?<0.0001), HLA-B58 (12% vs 1%,;? = 0.002) and HLA-DR103 (20% vs 8%, p = 0.001).
These associations appear independent of the HLA- B27 and DR103 associations seen in arthritis [41].
This suggests strongly that there are genes in the HLA region of chromosome 6 which determine the different EIM and that linkage disequilibrium between these genes may account for the clinical observation of distinct but overlapping clinical syndromes.
Management of ocular manifestations
For a proper diagnosis to be made the patient should undergo a full ophthalmologic assessment including sht-lamp examination. Treatment normally consists of topical steroid treatment initially frequently (hourly) but reducing subsequently, although a course of treatment lasting 6-8 weeks in total is usual. A cycloplegic agent such as atropine is often added to prevent the formation of p o s t e r i o r synechiae. Occasionally oral steroids may be required for acute anterior uveitis, and often in scleritis. Oral or topical NSAIDs may be useful in anterior scleritis, but have no role in anterior uveitis.
Their use carries the risks to the IBD described above.
Thus patients with acute ocular inflammation should have a rapid assessment by an ophthal- mologist, who should guide their treatment, which normally consists of topical or oral corticosteroid therapy. Failure to treat the inflammation effectively may lead to long term complications or even blind- ness, although these situations are rare.
References
1. Bywaters E, Ansell B. Arthritis associated with ulcerative coHtis: a cHnical and pathological study. Ann Rheum Dis 1958; 17: 169-83.
2. Edwards F, Truelove S. The course and prognosis of ulcerative colitis. III. Complications. Gut 1964; 5: 1-15.
3. Wright V, Watkinson G. The arthritis of ulcerative colitis. Br Med J 1965; 2: 670-5.
4. Greenstein A, Janowitz H, Sachar D. Extra-intestinal com- plications of Crohn's disease and ulcerative colitis. Medicine 1976; 55: 401-12.
5. Stein H, Volpin G, Shapira D, Suir E, Stevenberg A, Eidel- man S. Musculoskeletal manifestations of Crohn's disease.
Bull Hosp Jt Dis 1993; 53: 17-20.
6. Orchard TR, Wordsworth BP, Jewell DR Peripheral arthro- pathies in inflammatory bowel disease: their articular dis- tribution and natural history. Gut 1998; 42: 387-91.
7. van der Linden S, van der Heijde DM. Clinical and epidemiologic aspects of ankylosing spondylitis and spondy- loarthropathies. Curr Opin Rheumatol 1996; 8: 269-74.
8. CaHn A. Ankylosing spondylitis. In: Maddison PJ, Isenberg DA, Woo P, Glass DN, eds. Oxford Textbook of Rheuma- tology. Oxford: Oxford University Press, 1998: 1058-70.
9. Dekker-Saeys B, Meuwissen S, VandenBerg-Loonen E, De Haas W, Agenant D, Tytgat G. Prevalence of peripheral arthritis, sacroiliitis and ankylosing spondylitis in patients suffering from inflammatory bowel disease. Ann Rheum Dis 1978; 37: 33-5.
10. Wright V, Watkinson G. Sacroiliitis and ulcerative colitis. Br Med J 1965; 2: 675-80.
11. Hollingsworth P, Cheah P, Dawkins R, Owen E, Calin A, Wood P. Observer variation in grading sacroiliac radio- graphs in HLA-B27 positive individuals. J Rheumatol 1983;10:247-54.
12. McEnififN, Eustace S, McCarthy C, O'Malley M, Morain C, Hamilton S. Asymptomatic sacroiliitis in inflammatory bowel disease. Assessment by computed tomography. Clin Imaging 1995; 19: 258-62.
13. Agnew JE, Pocock DG, Jewell DP. Sacroiliac joint uptake ratios in inflammatory bowel disease: relationship to back pain and to activity of bowel disease. Br J Radiol 1982; 55:
821.
14. Dougados M, Linden S, Juhlin R et al. The European Spondyloarthropathy Study Group preliminary criteria for the classification of spondyloarthropathy. Arthritis Rheum 1991; 34: 1218-27.
15. Brown MA, Pile KD, Kennedy LG et al. HLA class I associations of ankylosing spondylitis in the white popula- tion in the United Kingdom. Ann Rheum Dis 1996; 55:
268-70.
16. Leirisalo-Repo M, Suoranta H. Ten year follow-up study of patients with Yersinia arthritis. Arthritis Rheum 1988; 31:
533-7.
17. Calin A, Fries JF. An 'experimental' epidemic of Reiter's syndrome revisited. Follow-up evidence on genetic and environmental factors. Ann Intern Med 1976; 84: 564-6.
18. Thomson GT, DeRubeis DA, Hodge MA, Rajanayagam C, Inman R D Vosi-Salmonella reactive arthritis: late clinical sequelae in a point source cohort. Am J Med 1995; 98: 13- 21.
19. Olivieri I, Salvarani C, Cantini F. Is Behget's disease part of the spondyloarthritis complex? [Editorial]. J Rheumatol 1997; 24: 1870-2.
20. Dekker-Saeys B, Meuwissen S, Berg-Loonen EVD et al.
Clinical characteristics and results of histocompatibitility typing (HLA B27) in 50 patients with both ankylosing spondylitis and inflammatory bowel disease. Ann Rheum Dis 1978; 37: 36-41.
21. Brewerton D, Caffery M, Nicholls A, Walters D, James D.
HL-A27 and arthropathies associated with ulcerative colitis and psoriasis. Lancet 1974; 1: 956-8.
22. Mallas EG, Mackintosh P, Asquith P, Cooke WT. Histo- compatibility antigens in inflammatory bowel disease. Their clinical significance and their association with arthropathy with special reference to HLA-B27 (W27). Gut 1976; 17:
906-10.
23. Steer S, Jones H, Hibbert J, Gibson T, Sanderson J. CT defined sacroiliitis and HLA-B27 in Crohn's disease. Gut 1999;44(Suppll.): A41.
24. Orchard TR, Thiyagaraja S, Welsh KI, Wordsworth BP, Hill Gaston JS, Jewell DP. Clinical phenotype is related to HLA genotype in the peripheral arthropathies of inflammatory bowel disease. Gastroenterology 2000; 118: 274-8.
25. Ebringer A. Ankylosing spondylitis is caused by Klebsiella.
Evidence from immunogenetic, microbiologic, and serologic studies. Rheum Dis Clin N Am 1992; 18: 105-21.
26. Gaston J, Life P, Granfors K et al. Synovial T lymphocyte
recognition of organisms that trigger reactive arthritis. Clin
Exp Immunol 1989; 76: 348-53.
27. Hermann E, Fleischer B, Mayet WJ, Poralla T, Meyer zum Buschenfelde KH. Response of synovial fluid T cell clones to Yersinia enterocolitica antigens in patients with reactive
Yersinia arthritis. Clin Exp Immunol 1989; 75: 365-70.
