Review article: joint involvement in inflammatory bowel disease

Transcription

1 Aliment Pharmacol Ther 2004; 20 (Suppl. 4): Review article: joint involvement in inflammatory bowel disease M. DE VOS Department of Gastroenterology, Ghent University Hospital, Belgium SUMMARY Peripheral involvement of the joints, including pauciarticular, asymmetrical, transitory and migrating synovitis and enthesiopathy, is observed in 10 20% of affected inflammatory bowel disease patients. Recurrence is common and frequently coincides with a flareup of intestinal disease. The true prevalence of axial involvement is less well established. Sacroiliitis is a hallmark of spondylitis, but is under-reported due to its insidious onset and sometimes asymptomatic nature. Radiographic evidence of sacroiliitis is present in about 20 25% of patients. Ankylosing spondylitis, as defined by the Rome criteria, is present in 3 10% of inflammatory bowel disease patients, and is thought to have a different genetic predisposition in these patients compared with ÔclassicÕ ankylosing spondylitis: whereas the human leucocyte antigen B27 phenotype is present in 90% of patients with ÔclassicÕ ankylosing spondylitis, the prevalence decreases to only 30% in patients with ankylosing spondylitis secondary to Crohn s disease. Polymorphisms involving CARD15 appear to be a possible genetic trigger: 78% of patients with Crohn s disease and symptomatic or asymptomatic sacroiliitis carry at least one mutation, compared with only 48% of control Crohn s disease patients. Moreover, in other forms of spondyloarthropathy, a similar association has been reported: 42% of patients with spondyloarthropathy and associated asymptomatic chronic gut inflammation, who are considered likely to develop Crohn s disease and ankylosing spondylitis, are carriers of at least one CARD15 mutation, compared with only 7% of patients with normal histology. In addition to genetic markers, clinical features support the relationship between gut and joint pathophysiology. In cases of spondyloarthropathy, a very rapid, substantial and sustained improvement in symptoms has been reported following treatment with infliximab, suggesting an essential role for tumour necrosis factor-a in spondyloarthropathy, similar to that observed in Crohn s disease. INTRODUCTION Inflammatory peripheral arthropathy, as observed in 10 20% of patients with inflammatory bowel disease, is mostly pauciarticular, asymmetrical, transitory and migrating. The clinical diagnosis requires the presence of synovitis, eventually associated with an enthesitis. Axial involvement may vary from asymptomatic sacroiliitis to clinical evidence of inflammatory low back pain, associated with decreased spine mobility, to Correspondence to: Professor M. De Vos, University Hospital, De Pintelaan 185, 9000 Ghent, Belgium. martine.devos@ugent.be classic ankylosing spondylitis. The prevalence varies from 7% to 25% of patients with inflammatory bowel disease and may or may not be associated with peripheral arthritis. Both types of arthropathy fall within the concept of spondyloarthropathy (Figure 1). 1 EVIDENCE FOR GUT JOINT ENTEROPATHY IN PERIPHERAL ARTHRITIS Inflammatory peripheral arthropathy in inflammatory bowel disease is similar to reactive arthritis secondary to intestinal infection. In the latter, the list of possible aetiological agents is small, including intracellular (either obligate or facultative), aerobic, invasive, 36 Ó 2004 Blackwell Publishing Ltd

2 REVIEW: JOINT INVOLVEMENT IN IBD 37 Figure 1. Articular involvement in inflammatory bowel disease. Gram-negative bacteria, e.g. Shigella flexneri, Salmonella typhimurium, Yersinia enterocolitica and Campylobacter jejuni. However, organisms with very similar enteroinvasive properties, such as enteroinvasive Escherichia coli, have shown no association with reactive arthritis. These observations indicate that specific properties of the bacterium may be important. 2 Reactive arthritis appears to arise as a result of a T-cellmediated immune response to bacterial antigens and degradation products circulating from gut to joint. Experimental evidence supports the theory that viable bacteria colonize the joint. Pathogenic bacterial antigens or lipopolysaccharides have been detected in synovium following a response of synovial fluid T-cell clones to these antigens. 3 6 Direct evidence for the presence of bacterial DNA and RNA in synovial fluid has been demonstrated using the reverse transcriptase polymerase chain reaction technique in patients with long-term undifferentiated spondyloarthropathy and juvenile-onset spondyloarthropathy. 7 However, the sensitivity of the techniques required to demonstrate nucleic acids from the pathogens in the joints suggests that the organisms are subject to transitory recurrent trafficking rather than remaining in the joint tissue. 2 A CD4+ T-cell response to the bacteria, possibly accompanied by a CD8+ T-cell response, drives the arthritic process. Persistent delivery of bacterial antigens to the joint may perpetuate local inflammation. The clinical observation of a parallelism between the flare-up of Crohn s disease and peripheral arthritis suggests a similar aetiopathogenesis. However, to the best of our knowledge, no information is available about the possible presence of bacterial DNA, RNA or bacterial degradation products in the synovium or the synovial fluid of inflammatory bowel disease patients. Most convincing evidence for a bacterial role in Crohn s disease-related peripheral arthritis comes from the germ-free B27/b2m transgenic rat model. Transgenic mice were found to develop colitis and arthritis only after restoration of the gut flora. 8 EVIDENCE FOR GUT JOINT ENTEROPATHY IN AXIAL INVOLVEMENT Axial involvement in inflammatory bowel disease is similar to that observed in other forms of spondyloarthropathy. This involvement may vary from asymptomatic sacroiliitis, to inflammatory lower back pain, to classic ankylosing spondylitis. Asymptomatic sacroiliitis in inflammatory bowel disease has been reported by several authors, with a prevalence varying from 11% to 52% according to detection technique. We recently observed radiographic evidence of sacroiliitis (at least grade 2) in 11% of patients with Crohn s disease, 9 and confirmed previous results (Table 1). The concept of inflammatory lower back pain has been less well studied. The diagnosis includes the presence of pain during the night and at rest, improving with movement. We observed this clinical symptom in 29 of 102 Crohn s disease patients. In 14 patients, sacroiliitis and/or spondylitis was present, whereas, in 15 patients, no radiological abnormalities were detected. The clinical significance and progression of these symptoms are unknown. The diagnosis of ankylosing spondylitis includes: clinical criteria; low back pain and morning stiffness for more than 3 months associated with a decreased mobility of the lumbar spine and limitation in chest expansion and radiological criteria; and sacroiliitis of at least grade 2 bilaterally or grade 3 unilaterally. The prevalence varies from 3% to 10%. Although some alterations in gut flora have been described in ankylosing spondylitis, especially related to the presence of Klebsiella, data concerning the presence of bacterial

3 38 M. DE VOS Table 1. Prevalence of asymptomatic sacroiliitis in the inflammatory bowel disease population Reference Population agents in axial joints are extremely rare due to low joint accessibility. The best evidence for gut joint iteropathy is the striking presence of gut alterations in approximately 60% of patients with spondyloarthropathy (without associated inflammatory bowel disease), varying from an increased number of lymphoid follicles, 10 to an asymptomatic chronic intestinal inflammation In the long term, a relationship was found between the disappearance of gut inflammation and evolution to articular remission. However, in 13% of patients with an initial asymptomatic chronic inflammation, a progression to classic Crohn s disease was reported, supporting the concept of subclinical Crohn s disease in a subpopulation of patients with spondyloarthropathy. 14 GENETIC BACKGROUND Prevalence of asymptomatic sacroiliitis (%) Method Ansel and Wright IBD 12 Rx Wright 234 IBD 18 Rx and Watkinson 44 Hyla et al IBD 12 Rx Davis et al IBD 52 Technetium scan Davis et al IBD 11 Rx Dekker-Saeys et al IBD 10 Rx McEniff et al IBD 18 Rx Scott et al CD 20 CT Queiro et al IBD 24 Rx de Vlam et al IBD 18 Rx De Vos et al CD 11 Rx CD, Crohn s disease; CT, computed tomography; IBD, inflammatory bowel disease; Rx, x-ray. In spondyloarthropathy, a very strong association is observed between the human leucocyte antigen (HLA)- B27 and axial involvement: more than 90% of patients with classic ankylosing spondylitis and 50% of patients with reactive arthritis are HLA-B27 positive. The role of linkage with major histocompatibility complex class I HLA-B27 can be extrapolated from work with the B27/ b2m transgenic rat model and from in vitro experiments: (a) Aberrant antigen presentation as a result of either the presence of specific (although unknown) arthritogenic bacterial peptides or shared antigenic recognition of B27-bound peptides between bacterial peptides and self-peptides (molecular mimicry). 15 (b) An impaired clearance of intracellular bacteria by HLA-B27, with persistent traffic of dendritic cells, and sharing of homing receptors between synovium and gut, resulting in increased cell adhesion and transmigration. 16 (c) Misfolding of the B-pocket of the peptide-binding groove occurring during the intracellular assembly process, resulting in the formation of heavy chain homodimers with potential immunogenic capacity. 17 The absence of arthritis and colitis in germ-free B27/ b2m transgenic rats and the induction of both kinds of inflammation by the reintroduction of gut bacteria suggest a central role for intestinal flora. Moreover, the manipulation of bacteria may influence intestinal immunopathology, as the bacterial load in the caecum seems to determine the severity of intestinal inflammation in rat model studies. In contrast with other forms of spondyloarthropathy, the association between axial involvement and HLA- B27 in inflammatory bowel disease patients is much less conclusive: only 25 75% of patients with Crohn s disease and ankylosing spondylitis are HLA-B27 positive. In a recent study, we demonstrated an increase in the prevalence of HLA-B27 from 0% in Crohn s disease patients with inflammatory back pain only, to over 14% in patients with sacroiliitis not corresponding to the criteria of ankylosing spondylitis, and up to 78% in patients with ankylosing spondylitis (Table 2). This suggests that sacroiliitis in Crohn s disease is not strongly associated with HLA-B27, but that evolution to ankylosing spondylitis is more likely to occur in HLA- B27-positive patients. 18 Another candidate gene is CARD15, because of the welldescribed association between CARD15 polymorphisms and Crohn s disease, and also its potential role in bacterial handling. CARD15 is located on chromosome 16 and encodes for an intracellular protein with binding affinity for bacterial muramyl dipeptide. One frame shift mutation and two missense mutations increase the risk of Crohn s disease by three-fold in heterozygous patients and 38-fold in compound heterozygote individuals.

4 REVIEW: JOINT INVOLVEMENT IN IBD 39 Table 2. Prevalence of sacroiliitis and ankylosing spondylitis in the inflammatory bowel disease (IBD) population Reference Prevalence of sacroiliitis associated with IBD Prevalence of ankylosing spondylitis associated with IBD Hyla et al. 45 1/11 (9%) 3/4 (75%) Dekker-Saeys et al. 47 1/11 (9%) 2/4 (50%) Enlow et al. 51 0/2 (0%) 4/10 (40%) Queiro et al. 50 3/15 (20%) de Vlam et al. 1 3/24 (13%) 2/8 (25%) De Vos et al. 18 0/11 (0%) 3/9 (33%) Previous genotype phenotype studies have not demonstrated any association with extra-intestinal manifestations of Crohn s disease. However, data for these studies were gathered from questionnaires and patients clinical records, relying mostly on Crohn s disease activity index scores. We recently performed a prospective clinical and radiological evaluation of 102 Crohn s disease patients, and found a striking association between CARD15 polymorphisms and the presence of sacroiliitis: 78% of patients with sacroiliitis were carriers of a polymorphism, compared with 48% of patients without sacroiliitis (P ¼ 0.01). 9 In patients with other forms of spondyloarthropathy, we also found an association with CARD15 polymorphisms. 18 Although several studies have failed to show an association between the prevalence of CARD15 polymorphisms and the presence of spondyloarthropathy, we were able to demonstrate an association in a subgroup of patients: 48% of patients with spondyloarthropathy and chronic gut inflammation were carriers of a mutation, compared with 28% of spondyloarthropathy patients without gut inflammation (P ¼ 0.01). Moreover, 92% of the carriers had evidence of chronic gut inflammation vs. 36% of patients with the wild-type genotype (P ¼ 0.002). Recently, an association between CARD15 polymorphisms and a psoriatic subtype of spondyloarthropathy has also been demonstrated, 23 supporting the need for careful subanalysis of patients and the possibility that CARD15 may play a role in gut joint iteropathy of spondyloarthropathy. PROPOSED AETIOPATHOGENIC MECHANISMS A number of clinical markers of spondyloarthropathy support increased antigen handling and presentation, similar to that observed in Crohn s disease, including an augmented number of intestinal follicles, an increased number of dendritic cells expressing CD11c-CD11a and vascular cell adhesion molecule-1 and mucosal infiltration by CD68+ and CD163+ macrophages. 10 A number of bacterial agents, including adherent, invasive, Gramnegative E. coli and Gram-negative anaerobic rods of the genera Bacteroides and Fusobacterium, have been implicated in the aetiopathogenesis of Crohn s disease. Similarly, bacterial agents are involved in the aetiopathogenesis of spondyloarthropathy, either by the presence of an imbalance in the microflora or by the presence of arthritogenic strains. Intestinal bacteria may reach dendritic cells and macrophages in a number of different ways: (i) transport by M cells; (ii) uptake and processing by epithelial cells; (iii) direct access to dendrites located between epithelial cells; and (iv) direct access to antigenpresenting cells through breaks in the epithelial barrier. The roles of CARD15 and HLA-B27 in this process remain unclear. The possible role of HLA-B27 has been described above. CARD15 protein is expressed by monocytes, granulocytes, dendritic cells and epithelial cells. In vitro, this protein induces the activation of the nuclear factor jb pathway after recognition of muramyl dipeptide (a compound of peptidoglycan not recognized by toll-like receptor 2), and may help to protect the gut wall against aggression. Genetic modulation of CARD15 expression may lead to disturbed handling of bacterial products, and hence inappropriate elimination. Interaction between microorganisms and germ lineencoded receptors, such as toll-like receptors, on mucosal epithelial cells, monocytes, macrophages and dendritic cells induces the secretion of a variety of mediators and triggers lymphocyte activation. Binding of T-cell receptors to major histocompatibility complex class II peptide complexes on the surface of antigenpresenting cells results in the production of cytokines and the subsequent regulation of the immune response: (a) interleukin-12 (IL12), IL23 and IL18 induce the differentiation of the T-helper-1 response; (b) IL4 and 1L10 induce the differentiation of the T-helper-2 response; (c) simultaneously, regulatory T cells are induced. T cells in the lamina propria of Crohn s disease and spondyloarthropathy patients produce interferon-c, resulting in the production of pro-inflammatory cytokines (tumour necrosis factor-a, IL6 and IL1). 24

5 40 M. DE VOS Effector T cells need to migrate from inductive sites to effector sites. The expression of adhesion molecules regulates cell traffic. Lymphocytes are involved in gut joint enteropathy as manifest lymphocytic infiltrate is observed in inflammatory synovial lesions of spondyloarthropathy. 25 Intestinally activated T cells may enter the synovium, either through the presence of cognate antigens at both sites or by homing of lymphocytes primed in gut by a4b7 and vascular adhesion protein 1 (VAP1). 26, 27 The discovery of identical T-cell expansions in colon mucosa, synovium and blood support this concept. 28 TREATMENT OF JOINT DISEASE Conventional treatments for acutely inflamed joints include nonsteroidal anti-inflammatory drugs (NSAIDs), joint aspiration and (after exclusion of septic arthritis) injection of depot steroids intra-articularly. Analgesia and physiotherapy will also help to maintain range of motion. The clinical response to NSAIDs is rapid for both peripheral arthritis and axial disease, but relapses are frequent after arrest of intake. Most gastroenterologists are reluctant to use NSAIDs in inflammatory bowel disease as evidence exists that they may exacerbate intestinal disease. A large case-controlled study, including 200 patients admitted to hospital as emergencies with colitis, demonstrated an association with current or recent NSAID use, with an adjusted odds ratio of In a second, smaller, case-controlled study, including 60 consecutive inflammatory bowel disease patients admitted to hospital due to exacerbation or onset of inflammatory bowel disease, 31% of patients were on NSAIDs, compared with 2% in the control group. 30 However, current data are inconsistent and the risk may be less than generally claimed. Moreover, careful analysis of some studies has suggested that paracetamol use is associated with an even higher risk of colitis than classic NSAIDs. The question of whether selective cyclo-oxygenase-2 inhibitors offer increased safety remains unanswered. There is evidence that cyclo-oxygenase-2-derived prostaglandins are involved in repair processes in the colon. In animal models, the effect of cyclo-oxygenase-2 inhibitors differs according to the type of induced colitis. In humans, only a small retrospective study is available, reporting an aggravation of inflammatory bowel disease in 7.4% of patients treated with celecoxib or rofecoxib. 31 However, interpretation of this study is limited due to the exclusion of a control group. Although the bacterial flora seems to play a central role in the pathogenesis of spondyloarthropathy, no randomized controlled trials with antibiotics have shown any benefit, even in reactive arthritis. 2 Disease-modifying agents, such as sulfasalazine, have been shown to be beneficial in the treatment of peripheral joint involvement in spondyloarthropathy patients. 32, 33 Data regarding the effect of mesalazine (mesalamine) alone are limited to two small, open trials, which demonstrated a significant effect of Pentasa (mesalazine, Ferring SA, Denmark) on several clinical measures, such as pain, number of swollen joints and degree of stiffness, but no effect on axial flexibility. Methotrexate has a similar effect on peripheral joint involvement in spondyloarthropathy, although data are scarce. 34 The major breakthrough in the treatment of spondyloarthropathy was the dramatic efficacy of tumour necrosis factor-a blockade, with clinical effects similar to those observed in Crohn s disease. This approach was brought into focus by a report on the associated beneficial effect of infliximab on gut and joint symptoms in Crohn s disease patients. 34 To date, 12 studies have been published on the use of tumour necrosis factor-a-blocking agents in spondyloarthropathy: eight open-labelled studies and four placebo-controlled studies The majority of studies included patients with ankylosing spondylitis or undifferentiated spondyloarthropathy. One study has been published on patients with psoriatic arthritis. Infliximab was used in nine of these studies, whereas the other three studies involved etanercept (a soluble receptor binding monoclonal antibody). A dramatic improvement of all clinical and laboratory variables was demonstrated in all studies. Both axial and peripheral symptoms, as well as enthesitis, demonstrated improvement. In spondyloarthropathy, as in Crohn s disease, the clinical effect of the drug was rapid and sustained when administered on a regular basis. In contrast with rheumatoid arthritis, the effect of infliximab on articular evolution has not been studied in spondyloarthropathy. In rheumatoid arthritis, an important retardation in the radiologically detectable destructive processes and a decrease in serum cartilage oligomeric matrix protein, as a marker of reduced cartilage turnover, have been observed. 40, 41 No longterm studies on radiological evolution are currently available in ankylosing spondylitis.

6 REVIEW: JOINT INVOLVEMENT IN IBD 41 An important difference between Crohn s disease and spondyloarthropathy is the effect of etanercept. Crohn s disease is intractable to etanercept treatment, in contrast with spondyloarthropathy, where the drug acts in a similar way to infliximab. A possible explanation for this discrepancy is the difference in apoptotic effect: in intestinal disease, apoptosis is a probable mode of action, whereas, in synovial tissue, apoptosis seems not to be induced; only a reduction in the number of inflammatory cells was observed. 42 CONCLUSIONS In conclusion, although current therapeutic trials in Crohn s disease pay only minimal attention to the effect on extra-intestinal manifestations, a shift in focus will be necessary in the future. In order to optimize the expected therapeutic benefit and improve the cost benefit ratio of treatment, targeting different therapeutic biological molecules to well-defined subgroups of patients is required. Further elucidation of the role of the intestine in the articular and other extra-intestinal manifestations of the disease will lead to more refined therapeutic alternatives. Moreover, investigation of the subclinical inflammation in the intestine of spondyloarthropathy patients may further our understanding of the very early immunopathology of Crohn s disease. REFERENCES 1 de Vlam K, Mielants H, Cuvelier C, De Keyser F, Veys EM, De Vos M. Spondyloarthropathy is underestimated in inflammatory bowel disease: prevalence and HLA association. J Rheumatol 2000; 27: Gaston JSH, Lillicrap MS. Arthritis associated with enteric infection. Best Pract Res Clin Rheumatol 2003; 17: Granfors K, Jalkanen S, von Essen R, et al. Yersinia antigens in synovial fluid cells from patients with reactive arthritis. N Engl J Med 1989; 320: Granfors K, Jalkanen S, Lindberg AA, et al. Salmonella lipopolysaccharide in synovial cells of patients with reactive arthritis. Lancet 1990; 335: 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 arthritis. Clin Exp Rheumatol 1989; 75: Nikkari S, Rantakokko K, Ekman P, et al. Salmonella triggered reactive arthritis: use of polymerase chain reaction, immunocytochemical staining and gas-chromatography mass spectrometry in the detection of the bacterial components from synovial fluid. Arthritis Rheum 1999; 42: Pacheco-Tena C, Alvarado De La Barrera C, Lopez-Vidal Y, et al. Bacterial DNA in synovial fluid cells of patients with juvenile spondyloarthropathies. Rheumatology 2001; 40: Taurog JD, Richardson JA, Croft JT, et al. The germfree state prevents development of gut and joint inflammatory disease in HLA-B27 transgenic rats J Exp Med 1994; 180: Peeters H, Vander Cruyssen B, Laukens D, et al. Radiological sacroiliitis, a hallmark of spondylitis, is linked with CARD15 gene polymorphisms in Crohn s disease. Ann Rheum Dis 2004; in press. 10 Demetter P, Van Huysse JA, De Keyser F, et al. Increase in lymphoid follicles and leukocyte adhesion molecules emphasizes a role for the gut in spondyloarthropathy pathogenesis. J Pathol 2002; 198: De Vos M, Cuvelier C, Mielants H, Veys E, Barbier F, Elewaut A. Ileocolonoscopy in seronegative spondyloarthropathy. Gastroenterology 1989; 96: Leirisalo Repo M, Turunen U, Stenman S. High frequency of silent inflammatory bowel disease in spondyloarthropathy. Arthritis Rheum 1994; 37: Simenon G, Van Gossum A, Adler M, Rickaert F, Appelboom T. Macroscopic and microscopic gut lesions in seronegative spondyloarthropathies. J Rheumatol 1990; 17: De Vos M, Mielants H, Cuvelier C, et al. Long-term evolution of gut inflammation in patients with spondyloarthropathy. Gastroenterology 1996; 110: Hermann E, Yu D, Meyer zum Büschenfelde K, Fleischer B. HLA-B27 restricted CD8 T cells derived from synovial fluids of patients with reactive arthritis and ankylosing spondylitis. Lancet 1993; 342: Laitio P, Virtala M, Salmi M, et al. HLA-B27 modulates the survival of Salmonella enteriditis on human monocytic cells. Eur J Immunol 1997; 27: Allen RL, O Callaghan CA, McMichael AJ, Bowness P. Cutting edge: HLA-B27 can form a novel b2-microglobulin free heavy chain homodimer structure. J Immunol 1999; 162: De Vos M, Laukens D, Marichal D, et al. CARD15 mutations in patients with spondyloarthropathy are linked with disease progression and evolution to Crohn s disease. Gastroenterology 2003; 124(Suppl.): A Crane AM, Bradbury L, van Heel DA, et al. Role of NOD2 variants in spondyloarthritis. Arthritis Rheum 2002; 46: Miceli-Richard C. CARD15/NOD2 analyses in spondyloarthropathy. Arthritis Rheum 2002; 46: Ferreiros-Vidal I, Amarelo J, Barros F, Carracedo A, Gomez- Reino JJ, Gonzalez A. Lack of association of ankylosing spondylitis with the most common NOD2 susceptibility alleles to Crohn s disease. J Rheumatol 2003; 30: van der Paardt M, Crusius JB, de Koning MH, et al. CARD15 gene mutations are not associated with ankylosing spondylitis. Genes Immun 2003; 4: Rahman P, Bartlett S, Siannis F, et al. CARD15: a pleiotropic autoimmune gene that confers susceptibility to psoriatic arthritis. Am J Hum Genet 2003; 73: Van Damme N, De Keyser F, Demetter P, et al. The proportion of Th1 cells which prevail in gut mucosa is decreased in

7 42 M. DE VOS inflammatory bowel syndrome. Clin Exp Immunol 2001; 125: Baeten D, Demetter P, Cuvelier C, et al. A comparative study of the synovial histology in rheumatoid arthritis, spondyloarthropathy and osteoarthritis: influence of disease duration and activity. Ann Rheum Dis 2000; 59: Elewaut D, De Keyser F, Van Den Bosch F, et al. Enrichment of T cells carrying beta7 integrins in inflamed synovial tissue from patients with early spondyloarthropathy, compared to rheumatoid arthritis. J Rheumatol 1998; 25: Salmi M, Jalkanen S. Human leucocyte subpopulations from inflamed gut bind to joint vasculature using distinct sets of adhesion molecules. J Immunol 2001; 166: May E, Marker-Hermann E, Wittig BM, Zeitz M, Meyer zum Buschenfelde KH, Duchmann R. Identical T-cell expansions in the colon mucosa and the synovium of a patient with enterogenic spondyloarthropathy. Gastroenterology 2000; 119: Evans JM, McMahon AD, Murray FE, McDevitt DG, MacDonald TM. Non-steroidal anti-inflammatory drugs are associated with emergency admission to hospital for colitis due to inflammatory bowel disease. Gut 1997; 40: Felder JB, Korelitz BI, Rajapakse R, Schwarz S, Horatagis AP, Gleim G. Effects of nonsteroidal antiinflammatory drugs on inflammatory bowel disease: a case control study. Am J Gastroenterol 2000; 95: Mahadevan U, Loftus EV Jr, Tremaine WJ, Sandborn WJ. Safety of selective cyclooxygenase-2 inhibitors in inflammatory bowel disease. Am J Gastroenterol 2002; 97: Clegg DO, Reda DJ, Weisman MH, et al. Comparison of sulfasalazine and placebo in the treatment of ankylosing spondylitis. Arthritis Rheum 1996; 39: Dougados M, vam der Linden S, Leirisalo-Repo M, et al. Sulfasalazine in the treatment of spondyloarthropathy. A randomized, multicenter, double-blind, placebo-controlled study. Arthritis Rheum 1995; 38: Altan L, Bingol U, Karakoc Y, Aydiner S, Yurtkuran M, Yurtkuran M. Clinical investigation of methotrexate in the treatment of ankylosing spondylitis. Scand J Rheumatol 2001; 30: Van den Bosch F, Kruithof E, De Vos M, De Keyser F, Mielants H. Crohn s disease associated with spondyloarthropathy: effect of TNFalpha blockade with infliximab on articular symptoms. Lancet 2000; 356: Van Den Bosch F, Kruithof E, Baeten D, et al. Randomized double-blind comparison of chimeric monoclonal antibody to tumor necrosis factor alpha (infliximab) versus placebo in active spondyloarthropathy. Arthritis Rheum 2002; 46: Braun J, Brandt J, Listing J, et al. Treatment of active ankylosing spondylitis with infliximab: a randomised controlled multicentre trial. Lancet 2002; 359: Mease PJ, Goffe BS, Metz J, et al. Etanercept in the treatment of psoriatic arthritis and psoriasis. Lancet 2002; 356: Gorman J, Sack K, Davis J. Treatment of ankylosing spondylitis by inhibition of tumor necrosis factor. N Engl J Med 2002; 346: Lipsky PE, van der Heijde DM, St Clair EW, et al. Infliximab and methotrexate in the treatment of rheumatoid arthritis. Anti-Tumor Necrosis Factor Trial in Rheumatoid Arthritis with Concomitant Therapy Study Group. N Engl J Med 2000; 343: Crnkic M, Mansson B, Larsson L, Geborek P, Heinegard D, Saxne T. Serum cartilage oligomeric matrix protein decreases in rheumatoid arthritis patients treated with infliximab or etanercept. Arthritis Res Ther 2003; 5: R Smeets TJ, Kraan MC, van Loon ME, Tak PP. Tumor necrosis factor alpha blockade reduces the synovial cell infiltrate early after initiation of treatment, but apparently not by induction of apoptosis in synovial tissue. Arthritis Rheum 2003; 48: Ansel BM, Wright RAD. Arthritic manifestations in regional enteritis. Ann Rheum Dis 1964; 23: Wright V, Watkinson G. Sacroiliitis and ulcerative colitis. Br Med J 1965; 2: Hyla JF, Franck WA, Davis JS. Lack of association of HLA B27 with radiographic sacroiliitis in inflammatory bowel disease. J Rheumatol 1976; 3: Davis P, Thomson AB, Lentle BC. Quantitative sacroiliac scintigraphy in patients with Crohn s disease. Arthritis Rheum 1978; 21: Dekker-Saeys BJ, Meuwissen SG, Van den Berg-Loonen EM, De Ha WH, Agenant D, Tytgat GN. Ankylosing spondylitis and inflammatory bowel disease. Prevalence of peripheral arthritis, sacroiliitis and ankylosing spondylitis in patients suffering from inflammatory bowel disease. Ann Rheum Dis 1978; 37: McEniff N, Eustace S, McCarthy C, O Malley M, O Morain A, Hamilton S. Asymptomatic sacroiliitis in inflammatory bowel disease. Assessment by computed tomography. Clin Imaging 1995; 19: Scott WW, Fishman EK, Kuhlman JE, O Brien JJ, Bayless TM. Computed tomography evaluation of the sacroiliac joints in Crohn disease. Radiologic/clinical correlation. Skeletal Radiol 1990; 19: Queiro R, Mainz O, Intxausti J, et al. Subclinical sacroiliitis in inflammatory bowel disease: a clinical and follow-up study. Clin Rheumatol 2000; 19: Enlow RW, Bias WB, Arnett FC. The spondylitis of inflammatory bowel disease. Evidence for a non-hla linked axial arthropathy. Arthritis Rheum 1980; 23: