Ileocolonoscopy in Seronegative Spondylarthropathy

Size: px

Start display at page:

Download "Ileocolonoscopy in Seronegative Spondylarthropathy"

David Simpson
5 years ago
Views:

1 GASTROENTEROLOGY 1989;96: Ileocolonoscopy in Seronegative Spondylarthropathy M. DE vas, c. CUVELIER, H. MIELANTS, E. VEYS, F. BARBIER, and A:ELEWAUT Departments of Gastroenterology, Pathology, and Rheumatology, State University of Ghent, Ghent, Belgium A prospective endoscopic and histologic study of terminal ileum and colorectum in 211 patients with seronegative spondylarthropathy revealed macroscopic inflammatory lesions varying from erythema to superficial erosions in 30% of the patients and microscopic inflammation in 61%. Two types of inflammation were observed: an acute inflammation resembling an infectious enterocolitis and a chronic inflammation. In idiopathic reactive arthritis both types of inflammation were equally present, whereas chronic inflammation predominated in patients with ankylosing spondylitis. In 32% of patients with chronic inflammation, the lesions particularly resembled early Crohn's disease. Repeat ileocolonscopy on 19 patients demonstrated a parallel evolution of joint symptoms and histologic lesions. All patients with acute inflammation went into clinical and histologic remission, whereas lesions persisted in patients with Crohn-like inflammation. In patients with chronic inflammation, remission and persistence were observed equally. This study identified a group of patients with seronegative spondylarthropathy which, even in the absence of gastrointestinal symptoms, showed evidence of gut inflammation, probably inducing an increased gut permeability with transgression of the oral tolerance and absorption of provocative antigens into the circulation. It is also possible that both diseases reflect a common underlying process. The concept of the seronegative spondylarthropathies (1) includes different disease entities such as ankylosing spondylitis (AS), psoriatic arthritis, reactive arthritis (ReA), some forms of juvenile chronic arthritis and arthritis associated with inflammatory bowel disease, Whipple's disease, and idiopathic acute anterior uveitis. Common features link these diseases: the absence of rheumatoid factor, the presence of peripheral inflammatory arthritis frequently associated with sacroiliitis and enthesopathies, the clinical overlap between the different diseases, the tendency toward familial aggregation, and the significant association with HLA B27 antigen (2,3). The peripheral arthritis of ReA is mostly pauciarticular and asymmetrical, involving the large and small joints mainly of the lower extremities. There is often an accompanying enthesopathy and sacroiliitis. The term enthesopathy includes all pathological changes around the insertion of the ligament into the bone. The arthritis is mostly triggered by urogenital or bowel infection (Table 1) but the causative microorganism cannot be isolated from the joint (4). In a third group of patients evidence of infection is lacking: the term idiopathic ReA is used here (5,6). The diagnosis of AS is based on the Rome criteria (7). Although axial involvement is the most common clinical and radiologic expression of the disease, peripheral joint involvement occurs in up to 35% of the patients (8). In previous studies we demonstrated evidence of gut inflammation in patients with ReA or AS (9,10). The study of a large number of biopsy specimens, endoscopically taken at terminal ileum, ileocecal valve, and colon, revealed two types of inflammation: an acute inflammation with or without ulcerations and a chronic inflammation sometimes remarkably identical to early Crohn's disease. These results and their clinical importance are reported. Materials and Methods An ileocolonoscopy was performed in 136 patients with ReA, in 75 patients with AS, and in 65 control patients, divided into 2 groups: an articular control group Abbreviations used in this paper: AS, ankylosing spondylitis; ReA, reactive arthritis; CD, Crohn's disease by the American Gastroenterological Association /89/$ 3.50

2 340 DE VOS ET AL. GASTROENTEROLOGY Vol. 96, No.2, Part 1 Table 1. Pathogens Associated With Enterogenic Forms of Reactive Arthritis Genitourinary infection (sexually acquired) Neisseria gonorrhea Chlamydia trachomatis Ureaplasma urealyticum Mycoplasma? Herpes genitalis Enteric infection (enterocolitic) Salmonella Shigella flexneri Yersinia enterocolitica Campylobacter jejuni Table 2. Sex and Age Distribution in Different Patient Populations Mean Age age Diagnosis n M/F (yr) (yr) Reactive arthritis Enterocolitic 12 9/ Urogenital 20 15/ Idiopathic / Ankylosing spondylitis Axial 29 22/ Peripheral 46 35/ Arthritis control Intestinal control 28 18/ consisting of 37 patients with other inflammatory joint diseases (27 cases of rheumatoid arthritis, 3 cases of juvenile chronic arthritis with polyarticular onset, 4 cases of psoriatic arthritis, and 3 cases of systemic lupus erythematosus) and an intestinal control group consisting of 28 patients with noninflammatory abdominal diseases (5 cases of colon polyps, 8 cases of chronic obstipation, and 15 cases of irritable bowel disease). All patients gave verbal consent to the endoscopy and the investigations were approved by the local committee on ethics. Age and sex ratio were comparable in the four groups of patients (Table 2). The endoscopy was performed with an Olympus CF colonoscope (Olympus Corporation of America, New Hyde Park, N.Y.). In only a few cases «3%) the ileum could not be reached. These cases were not included in the study. In each patient, biopsy specimens (mean, 13 specimens) were taken in terminal ileum, ileocecal valve, and colorectum at sites of obvious abnormality if present or blindly if the mucosa appeared to be intact. The tissues were fixed in 10% formalin or in sublimate formaldehyde mixture for 3-4 h at room temperature. After fixation the specimens were routinely processed and enibedded in paraplast. Tissue sections were cut at 5 nm and stained with haematoxylin-eosin. Both the endoscopist and the pathologist remained uninformed about the clinical diagnosis. In 19 patients (17 patients with ReA and 2 with AS) a second ileocolonoscopy was performed after a mean interval of 12.7 mo (range, 4-18 mol. At the time of the second ileocolonoscopy 10 patients were in clinical and biological remission. The histopathological spectrum had the following features: I" o NORMAL MUCOSA ~ ERYTHEMA GRANULARITY lei EROSION ULCERATION Figure 1. Distribution of endoscopic lesions in patients with reactive arthritis (ReA), ankylosing spondylitis (AS), articular control group, and intestinal control group. ENT, enterocolitic; UROG, urogenital; IDIOP, idiopathic; AX, axial ankylosing spondylitis; PJ AS, ankylosing spondylitis with peripheral joint involvement. 10 II NIT CONTROl INTEST 00. UROG. I ANKYlOSING SPONDYLmS I

3 February 1989 ILEOCOLONOSCOPY IN SERONEGATIVE SPONDYLARTHROPATHY Acute inflammation with well-preserved normal mucosal architecture. The surface and crypt epithelium were infiltrated by neutrophils or eosinophils without a significant increase of lymphocytes. Small superficial ulcers covered with fibrin and neutrophils overlying hyperplastic lymphoid follicles sometimes occurred. The lamina propria was edematous and contained polymorphonuclear cells (10). 2. Chronic inflammation was also revealed. Epithelial changes consisted of partial villous atrophy without significant increase of intraepithelial lymphocytes or crypt distortion, increased mononuclear cell content in the lamina propria, and presence of lymphoid follicles in the basal lamina propria (10). 3. Lesions indistinguishable from Crohn's disease (CD) were observed: aphthoid ulcers, villous blunting and fusion or atrophy, crypt distortion, branching and pyloric metaplasia associated with sarcoidlike granulomas, discontinuous inflammation in the lamina propria, cryptitis, and crypt abscesses with neutrophils and eosinophils (10). In 216 patients HLA-B27 phenotype was determined by mean of the microlymphocytotoxicity test according to Terasaki and McLelland (11). The results were statistically analyzed according to the ~ method. Results No endoscopic lesions were present in ileum or colon in the control group or in patients with a urogenital form of ReA (Figure 1). In contrast, macroscopic lesions were observed in half of the patients with an enterogenic form of ReA (6 of 12 patients, p < 0.001) and in 35% of the patients with an idio- pathic form of ReA (32 of 72 patients, p < 0.001) (Figure 1). The frequency of macroscopic abnormalities was also significantly (p < 0.001) elevated in patients with AS associated with peripheral arthritis (16 of 46 patients or 35%) and in patients with axial involvement alone (7 of 29 patients or 24%) (Figure 1). The macroscopic lesions mainly involved the terminal ileum and ileocecal valve (80%). Aphthoid ulcerations were seen in 33 patients. Cobblestones were present in 4 patients. Recurrent intestinal complaints (mild diarrhea, abdominal pain, and sometimes mild hematochezia) were reported in 59 patients with AS and ReA. Only 20 of these 59 patients had macroscopic gut lesions. Consequently, 41 patients with endoscopic abnormalities never complained of recurrent intestinal symptoms. The histology of the specimens of all intestinal control patients and 36 of 37 rheumatic control patients was normal. Only 1 patient with rheumatoid arthritis had histologic evidence of a mild acute colitis (Figure 2). In 3 of 20 patients with a urogenital form of ReA, acute inflammation was present in the terminal ileum and chronic inflammation was lacking (Figure 2). The incidence of histologic inflammation was significantly elevated (p < 0.001) in patients with other forms of ReA and in patients with AS (84 of 116 patients or 72% and 42 of 75 patients or 56%, respectively) (Figure 2). In 60% of the cases the inflammation was confined to the ileum and the ileocecal valve. In 35%, colon and ileum were in- D NORMAL MUCOSA bill ACUTE INFLAMMATION ~ CHRONIC INFLAMMATION ~ CO-UKE LESIONS 40 ART INTEST ENT UROG. IDIOP. /IJ( AS PJ AS CONTROl REACTIVE ARTHRITIS ANKYLOSING SPONDYLITIS Figure 2. Distribution of histologic lesions in patients with reactive arthritis (ReA). ankylosing spondylitis (AS), articular control group, and intestinal control group. ENT, enterocolitic: UROG, urogenital:!diop, idiopathic: AX AS, axial ankylosing spondylitis; PJ AS, ankylosing spondylitis with peripheral joint involvement.

4 342 DE VOS ET AL. GASTROENTEROLOGY Vol. 96, No.2, Part 1 Table 3. Distribution of Gut Inflammation in Seronegative Spondylarthropathic Patients With and Without HLA-B27 Phenotype Reactive arthritis Ankylosing spondylitis Histologic Enterogenic Urogenital Idiopathic Ax AS PJ AS inflammation (HLA-B27) Normal Acute Chronic " 18 7 Total Ax AS, axial ankylosing spondylitis; PJ AS, ankylosing spondylitis with peripheral joint involvement. " p < 0.01 (y method). volved, whereas in 5% the inflammation was confined to the colon. In 10 of 12 patients (83%) with an enterocolitic form of ReA and 74 of 104 (71 %) with an idiopathic form of ReA, acute or chronic inflammation was present in equal frequency (50% vs. 50% and 47% vs. 53%, respectively) (Figure 2). The incidence of acute inflammation in patients with AS (9 of 75 patients or 12%) was not significantly elevated. In contrast, chronic inflammation was present in 33 of 75 patients with AS (44%, p < 0.001). The frequency of lesions was significantly (p < 0.01) lower in the group of AS patients with axial involvement than in the group with associated peripheral joint disease (Figure 2). The presence of HLA-B27 antigen influenced the results only in patients with AS. Chronic inflammation was significantly (p < 0.01) more frequent in axial AS patients without the phenotype than in patients with HLA-B27 antigen. In contrast, in ReA, HLA-B27 phenotype had no influence on the presence or the type of the inflammatory gut lesions (Table 3). Endoscopic lesions were not observed in histologic normal mucosa. The incidence of endoscopic lesions was significantly (p < 0.05) higher in patients with chronic inflammation (25 of 53 patients or 47%) than in patients with acute inflammation (14 of 52 patients or 27%). Endoscopic lesions consisted mostly of erosions or ulcerations, and were present in 92% of the patients with CD-like lesions (Table 4). At the second ileocolonoscopy no lesions were seen in patients who went into clinical remission. In contrast, lesions persisted or were aggravated in 6 of the 9 patients with persisting active articular disease. The same parallel evolution was present on histology: inflammation was absent in 9 of 10 patients in clinical remission but persisted in patients with active arthritis (Figure 3). All but 1 patient with acute inflammation on first ileocolonoscopy went into remission. In contrast, inflammation and arthritis persisted in all patients with CD-like lesions. Half of the patients with chronic inflammation went into remission, whereas gut and articular inflammation persisted in the other half (Figure 3). Discussion In 71 % of the patients with idiopathic ReA and in 56% of those with AS, inflammatory lesions were seen on specimens from ileum, ileocecal valve, and colon obtained during ileocolonoscopy. In onethird of these patients macroscopic abnormalities were also present at endoscopy. Only a minority had intestinal complaints. These findings suggest that gut inflammation has a role in the pathogenesis of these seronegative spondylarthropathies. They also support the existence of a link between the different diseases belonging to this concept. Acute inflammation, with lesions as in acute'selflimiting enterocolitis, was present in one-third of the patients with idiopathic ReA. In contrast it was only present in 12% of the patients with AS. Although an intensive search for causative microorganisms in stool and in serum remained negative, the histologic findings suggested an infectious organism as responsible for this type of inflammation. The clinical and histologic remission observed at second ileocolonoscopy suggested that this triggering was time-limited and perhaps sensible to therapy with nonsteroi- Table 4. Frequency of Endoscopic Lesions in Microscopically Inflamed Ileum or Colon. Subdivision of Histopathological Lesions Into Acute, Chronic, and Crohn's Disease-like Types of Inflammation Histologic Endoscopic staging inflammation Stage 0 Stage 1 Stage 2 Acute Chronic CD like CD, Crohn's disease. Endoscopic stage 0 refers to normal mucosa; stage 1 represents erythema, friability; stage 2 includes erosions or ulcerations.

5 February 1989 ILEOCOLONOSCOPY IN SERONEGATIVE SPONDYLARTHROPATHY 343 I ACTIVE JOINT DISEASE I I REMISSION I HISTOlOGY Of LESIONS HISTOlOGY Of LESIONS CD-Uke chronic acute acute normal tstlleocolo 2nd ILEOCOLO nonnal Figure 3. Evolution of histologic lesions at second ileocolonoscopy in patients with persistent active arthritis and in patients with clinical remission. CD, Crohn's disease. dal anti-inflammatory drugs or sulfasalazine, or both. Chronic inflammation was observed in 38% of the patients with idiopathic ReA and in 44% of the patients with AS. A chronic or repetitive irritation of the gut or a disturbed immunologic status of the gut wall could be implicated in the etiopathogenesis of this type of inflammation and arthritis. In 32% of the patients with chronic inflammation, the ileal localization and the patchy, discontinuous distribution of the lesions and the microscopic features of the inflammation suggested a relationship with CD (12). Moreover, the gastrointestinal complaints and endoscopic lesions were significantly more frequent and more pronounced in this subgroup of patients. No clinical or histologic remission could be induced by therapy. The similarity with inflammatory bowel disease in this patient population suggested the existence of CD with a subclinical gut inflammation and arthritis as first and sole expression of the disease. Although the incidence of HLA-B27 phenotype is elevated in reactive arthritis (60%-80%) (6,13) and in AS (96%) (14), it seemed unessential for the presence or type of gut inflammation in ReA. In contrast, in AS, inflammation was predominantly present in patients with axial AS without HLA-B27 phenotype and in patients with associated peripheral joint involvement independently of their B27 phenotype. The high frequency of chronic inflammation in B27-negative AS patients with axial involvement only supports our assumption that this group of patients may have a subclinical CD. In the other inflammatory joint disorders, particularly rheumatoid arthritis, inflammatory gut lesions were completely absent, suggesting a different etiopathogenetic mechanism. As almost all rheumatoid arthritis patients were being treated with nonsteroidal anti-inflammatory drugs, the absence of lesions ruled out a direct relationship between drug intake and gut inflammation. In summary, this study demonstrated a clear relationship between inflammatory joint diseases and the gut, with a high incidence of acute or chronic inflammatory lesions in ileum and colon of patients with ReA and AS. In patients with acute inflammatory lesions the evolution is generally favorable; in patients with chronic lesions striking similarities with inflammatory bowel disease are observed. Two mechanisms may be responsible for the occurrence of gut inflammation in patients with ReA and AS. The first hypothesis suggests that both diseases are the reflection of an unique underlying process. An alternative, and in our opinion more attractive hypothesis suggests that the gut lesions initiate or trigger the joint disease: an intestinal inflammation, resulting in a transgression of the oral tolerance and an increase in gut permeability, leads to an absorption of exogenous antigens into the circulation and initiates joint inflammation in genetically predisposed patients. References 1. Wright V, Moll JHM, eds. Seronegative polyarthritis. Amsterdam: North-Holland Publishing, 1976.

6 344 DE VOS ET AL. GASTROENTEROLOGY Vo!' 96, No.2, Part 1 2. Wright V. An unifying concept for the spondylarthropathies. Clin Orthop 1979;143: Calin A. The spondylarthropathies. Scientific American Rheum 1983;9: Ahvonen p, Sievers K, Aho K. Arthritis associated with Yersinia enterocolitica infection. Acta Rheum Scand 1969;15: Hind CRK. Reactive arthritis. Postgrad Med J 1982;58: Keat A. Reiter's syndrome and reactive arthritis in perspective. N Engl J Med 1983;309: Kellgren JH, Jeffrey MR. Ball J. Clinical criteria for ankylosing spondylitis. In: Kellgren JW, Jeffrey MR, Ball J, eds. The epidemiology of chronic rheumatism. Oxford: Blackwell Scientific, 1963: Calin A. Ankylosing spondylitis. In: Tanayi GS, ed. Clinics in rheumatic diseases: the seronegative spondylarthropathies. London: WB Saunders, 1985:11: Mielants H, Veys E, Cuvelier C, De Vos M, Botelberghe L. HLA B27 related arthritis and bowel inflammation. Part 2. Ileocolonoscopy and bowel histology in patients with HLA B27 related arthritis. J Rheumatol 1985;12: Cuvelier C, Barbatis C, Mielants H, De Vos M, Veys E, Roels H. Histopathology of intestinal inflammation related to reactive arthritis. Gut 1987;28: Terasaki PI. McLelland JD. Microdoplet assay of human serum cytotoxine. Nature 1964;204: Morson BC. The early histological lesions of Crohn's disease. Proc R Soc Med 1972;65: Aho K, Ahvonen P, Lassus A. HLA antigen 27 and reactive arthritis. Lancet 1973;ii: Woodrow JC. Histocompatibility antigens and rheumatic diseases. Semin Arthritis Rheum 1977;6: Received December 18, Accepted August 22, Address requests for reprints to: Martine De Vos, Department of Gastroenterology, State University of Ghent, De Pintelaan 185, B-9000 Ghent, Belgium.

Normal small intestine Variations

THE HISTOPATHOLOGICAL DIAGNOSIS OF THE TERMINAL ILEUM BIOPSY K. Geboes, MD, PhD, AGAF, dr. H.C. Belgium Normal small intestine Variations Duodenum Ileum 1 Proliferative compartment Normal variations 2