ORIGINAL ARTICLE

Oro-Facial Granulomatosis: Crohn’s Disease or a New Inflammatory Bowel Disease? Jeremy Sanderson,* Carlo Nunes,* Michael Escudier,† Kate Barnard,† Penelope Shirlaw,† Edward Odell,† Catherine Chinyama,‡ and Stephen Challacombe†

Background: Oro-facial granulomatosis (OFG) is a rare chronic inflammatory disorder presenting characteristically with lip swelling but also affecting gingivae, buccal mucosa, floor of mouth, and a number of other sites in the oral cavity. Histologically, OFG resembles Crohn’s disease (CD), and a number of patients with CD have oral involvement identical to OFG. However, the exact relationship between OFG and CD remains unknown.

Methods: Thirty-five patients with OFG and no gut symptoms were identified from a combined oral medicine/gastroenterology clinic. All underwent a standardized assessment of the oral cavity and oral mucosal biopsy to characterize the number of sites affected and the type of inflammation involved. Hematological and biochemical parameters were also recorded. All 35 patients underwent ileocolonoscopy and biopsy to assess the presence of coexistent intestinal inflammation.

Results: Ileal or colonic abnormalities were detected in 19/35 (54%) cases. From gut biopsies, granulomas were present in 13/19 cases (64%). An intestinal abnormality was significantly more likely if the age of OFG onset was less than 30 years (P = 0.01). Those with more severe oral inflammation were also more likely to have intestinal inflammation (P = 0.025), and there was also a correlation between the histologic severity of oral inflammation and the histologic severity of gut inflammation (P = 0.047). No relationship was found between any blood parameter and intestinal involvement. Conclusions: Endoscopic and histologic intestinal abnormalities are common in patients with OFG with no gastrointestinal symptoms. Younger patients with OFG are more likely to have concomitant intestinal involvement. In these patients, granulomas are more frequent in endoscopic biopsies than reported in patients with documented CD. OFG with associated intestinal inflammation may represent a separate entity in which granulomatous inflammation

Received for publication July 6, 2005; accepted July 7, 2005. From the *Department of Gastroenterology and Oral Medicine, †Department of Pathology, and ‡Department of Histopathology, King’s College London, United Kingdom. Reprints: Jeremy D. Sanderson, MD, FRCP, Department of Gastroenterology, 1st Floor, College House, St. Thomas’ Hospital, London SE1 7EH, United Kingdom (e-mail: [email protected]) Copyright Ó 2005 by Lippincott Williams & Wilkins

840

occurs throughout the gastrointestinal tract in response to an unknown antigen or antigens. Key Words: Crohn’s disease, granuloma, inflammatory bowel disease, oral cavity, oro-facial granulomatosis (Inflamm Bowel Dis 2005;11:840–846)

O

ro-facial granulomatosis (OFG) is a chronic inflammatory disorder presenting characteristically with lip swelling but also affecting gingivae, buccal mucosa, floor of mouth, and a number of other sites in the oral cavity. Wiesenfeld et al1 originally described the term OFG in those patients with oral lesions resembling Crohn’s disease (CD) but without any evident intestinal involvement. Most cases of OFG occur as a separate clinical entity, often presenting to dental or maxillo-facial surgeons, but a proportion present in association with established CD of the small or large intestine. Rarely, OFG may be a manifestation of sarcoidosis or occur as part of the Melkersson-Rosenthal syndrome.2–4 Although CD and OFG share a number of clinical and histologic features, the exact relationship between the 2 conditions is unknown. Nonspecific oral lesions are common in CD, usually reflecting periods of nutritional deficit or active disease.5,6 Conversely, OFG-like changes, usually lip swelling or buccal involvement, are uncommon in established gut CD.7 Only 1 previous study has investigated the presence of coexistent gut inflammation in OFG without gut symptoms.8 This study used rigid sigmoidoscopy and barium studies and reported evidence of intestinal CD in 37% of cases. This would be likely to underestimate the true prevalence of gut inflammation. No previous study has reported on the use of ileocolonoscopy in patients with OFG. The aim of this study was therefore to undertake ileocolonoscopy, to characterize in detail the intestinal features in a series of patients with OFG without gut symptoms, and to relate any findings to the severity and presentation of oral disease.

MATERIALS AND METHODS From a larger series of patients attending a joint Oral Medicine-Gastroenterology clinic, 35 patients were identified Inflamm Bowel Dis  Volume 11, Number 9, September 2005

Inflamm Bowel Dis  Volume 11, Number 9, September 2005

Oro-Facial Granulomatosis

with the clinical features of OFG and in whom there were no gastrointestinal symptoms. In each case, a biopsy of affected oral mucosa had been taken, and all biopsy specimens were reviewed as part of this study. OFG was defined by typical clinical presentation and/or histological presence of noncaseating granulomatous inflammation.9,10 In 5 cases, the diagnosis was made on clinical grounds without granulomas on oral biopsy but the presence of other typical but histologically nonspecific microscopic features such as edema, fibrosis, and pattern of lymphoplasmacytic infiltration. An additional 35 patients with no gut symptoms undergoing surveillance colonoscopy for a family history of colorectal cancer were recruited as comparative asymptomatic controls for the presence or absence of gut inflammation.

infiltrating cells, edema, and fibrosis, either by presence/absence or using subjective semiquantitative scoring.

Clinical Assessment, Blood Tests, and Patch Testing All patients with OFG underwent a detailed assessment of the oral cavity to characterize the number of sites affected and the type of inflammation involved. A standardized oral severity score was used in each case, which combined the number of oral sites involved and severity of inflammation derived from a similar scoring system for mucous membrane pemphigoid and lichen planus.11 In most cases, blood was taken for analysis of full blood count, ESR, C-reactive protein, vitamin B12 and folate, iron parameters, and serum angiotensin converting enzyme. Where clinically indicated, a chest x-ray was performed in each patient to look for any features suggestive of sarcoidosis.

Ileocolonoscopy All 35 patients with OFG and the 35 controls underwent ileocolonoscopy and biopsy to assess the presence of intestinal inflammation. In 6 cases, an upper gastrointestinal endoscopy was also performed. Intravenous sedation was administered in each case using pethidine and midazolam. Endoscopic procedures were performed using Olympus EVIS series video endoscopes.

Histologic Analysis Oral mucosal biopsy specimens A total of 39 oral biopsy specimens were examined from 35 patients in the series. All were surgical biopsy specimens of flat mucosa except for 2 small mucosal tags, 1 punch biopsy taken from the outer aspect of the lip, and 1 surgical lip reduction. If no granulomatous inflammation was identified on initial sections, specimens were cut at a minimum of 3 levels, and an average of 24 mm length of mucosa was examined for each specimen. Hematoxylin and eosin–stained slides from all levels were scored by one pathologist (E.O.) for granuloma frequency, structure and level, component cells and relationship to epithelium, salivary ducts, and lymphatics, nonspecific inflammation pattern, and q 2005 Lippincott Williams & Wilkins

Gastrointestinal tract biopsy specimens Histologic assessment of intestinal biopsies was carried out by a single gastrointestinal histopathologist (C.C.). All mucosal biopsies were fixed in formalin, embedded in paraffin, sectioned, and stained with hematoxylin and eosin using standard methods. The histopathologist was blinded to case details. In each case, mucosal biopsies were scored for the presence of epithelial ulceration, neutrophilic, lymphocytic, and eosinophilic infiltration (none, mild, moderate, or severe in each case) and for the presence of epithelioid granulomata. All patients gave written informed consent, and the study was approved by the local research ethics committee.

Statistical Methods The x2 test and Spearman rank coefficient of correlation were used to examine the significance of any relationship between oral cavity variables, age of onset of oral disease, and the presence or absence and severity of any intestinal inflammation.

RESULTS Of the 35 patients studied, 18 (51%) were women, and the median age was 24 years (range, 6–74 yr). The median age at onset of oral disease was 20 years, and the median duration of OFG was 3 years. Of the 35 controls studied, 24 (69%) were women; median age was 40 years (range, 22–62 yr). The sites affected in the oral cavity are shown in Fig. 1. Lip swelling (Fig. 2) was the most common (95%) presentation, but buccal and gingival inflammation was also common. Swelling (91%) and erythema (70%) were the most common types of macroscopic inflammatory lesion encountered. Cobblestoning was seen in 49% and fissuring in 37% of patients, usually in the buccal mucosa or sulcus (Fig. 3). Aphthous-like ulceration (15%) and deeper linear-type ulcers (12%) were uncommon. The results of hematological and other laboratory parameters are shown in Table 1. No consistent abnormality was detected in the full blood count, ESR, or any hematinic deficiency that could be related to the presence or absence of intestinal abnormalities in the patients studied. Sixteen of the 30 patients had a chest x-ray performed, and this was normal in each case.

Oral Mucosal Histology Granulomatous inflammation was found in 38 specimens from the 35 patients included in the series (Fig. 4). No granulomas were present in 5 patients from the series. Of those specimens with granulomas, these were sparse in 9 (requiring at least 3 levels for diagnosis). In 10 specimens, granulomas

841

Sanderson et al

Inflamm Bowel Dis  Volume 11, Number 9, September 2005

FIGURE 1. Oral sites affected among 35 patients with OFG.

were evident in 1 section but less frequently than 1 per 5-mm surface length; granulomas were numerous in 19 specimens. The majority of specimens, 25 in total, contained discrete noncaseating granulomas, and in 4, these merged to form confluent granulomatous sheets of macrophages and giant cells. Nine specimens contained only loose or poorly formed granulomas or small clusters of macrophages. The granulomas were found at all levels. Multinucleate giant cells were present

FIGURE 3. Involvement of the buccal mucosa and sulcus in OFG. There is thickening, fissuring and mucosal tags in the buccal mucosa. In the sulcus, there is linear ulceration.

in all but 6 specimens. Granulomatous lymphangitis could be identified in only 6 specimens. Nonspecific inflammation was present in the background in all but 4 cases and was usually a patchy lymphoplasmacytic infiltrate of the corium. When it extended to the deeper tissue, it was frequently perivascular. Fibrosis was present in 31 specimens, and in 13, it was marked. Fibrosis was usually associated with presence of granulomas, and in only 1 case with marked fibrosis were granulomas absent. Edema or a fibrin-rich exudate was present in 13 specimens.

Results of Ileocolonoscopy

FIGURE 2. OFG presenting with involvement of the lower lip with swelling and nodularity. There is also angular cheilitis and some erythematous gingivitis.

842

In each control case, apart from adenomatous or hyperplastic polyps (6 cases), the macroscopic findings in this group were entirely normal. Ileal or colonic abnormalities were detected in 19/35 (54%) of OFG cases. Of these, 17 had macroscopic abnormalities, and 2 were abnormal on biopsy alone (Table 2). Macroscopically, scattered aphthoid-type ulceration (Fig. 5) was the most common abnormality seen at ileocolonoscopy. One case had more extensive ulceration and q 2005 Lippincott Williams & Wilkins

Inflamm Bowel Dis  Volume 11, Number 9, September 2005

TABLE 1. Additional Studies of 35 Cases of OFG Study

No. Performed

Full blood count ESR

35 35

Vitamin B12

26

Folate

24

Ferritin sACE Chest x-ray

23 24 15

Results Mild anemia in 1 case only Normal in all cases ,20 in all other cases Low in 3 cases unrelated to gut abnormality Low in 3 cases unrelated to gut abnormality Low in 1 case Normal in all cases Normal in all cases

erythema in the cecum more typical of classic CD (Fig. 6). Two additional cases showed larger ulcers, up to 8 mm in size, in the colon or ileum. Total colonoscopy was achieved in 33/35 cases (94%). Ileoscopy was performed in 29 of 35 cases (83%), and abnormalities were evident in the ileum alone in 3 cases.

Intestinal Mucosal Histology No abnormal inflammatory infiltrate, ulceration, or granulomas were detected in any of the 35 control cases apart from 5 cases in which a moderate excess of eosinophils was

Oro-Facial Granulomatosis

detected. In addition, melanosis coli was seen in 3 cases (8%) and pseudolipomatosis in 2 cases (5%). Mucosal biopsies from the intestine of 19 cases of OFG were histologically abnormal. Of these abnormalities, 7 were present in the colon alone, 3 in the ileum alone, and 9 in both ileum and colon. Superficial mucosal ulceration was observed in only 3 cases. In all abnormal cases, a variable increase in chronic inflammatory cells was present. In most cases (15/19 [79%]), this was accompanied by a variable but lesser degree of acute inflammation (neutrophil infiltration). Granulomas were present in 13/19 (68%) cases. In 11 of these 13 cases, granulomas were discrete and infrequent (Fig. 7). Two cases had marked granulomatosis. A moderate excess of eosinophils was present in 8 cases, 5 of which had other inflammatory changes. Upper gastrointestinal endoscopy was also performed in 6 cases. Four cases were abnormal. Two showed macroscopic abnormality (aphthous duodenitis in 1 case, erythematous gastritis in the other). Microscopically, all 4 abnormal cases showed a Helicobacter pylori–negative, mild chronic gastritis, with granulomas present in 1 of these cases.

Correlation Between Oral and Intestinal Features Intestinal involvement was significantly more likely if the age of OFG onset was less than 30 years (18/24 versus

FIGURE 4. Low power of oral mucosa showing superficial dense lymphocytic infiltrate in the epithelium and corium, the latter showing marked scarring and containing noncaseating granulomata with multinucleate giant cells (A). High power showing well-organized typical granuloma (B) and remote granulomas in deep underlying muscle 1 cm below the surface (C). q 2005 Lippincott Williams & Wilkins

843

Inflamm Bowel Dis  Volume 11, Number 9, September 2005

Sanderson et al

TABLE 2. Endoscopic Findings Among 35 Cases of OFG Total number Complete colonoscopy Ileoscopy Gut abnormal Colon only Ileum only Both Macroscopic abnormality Aphthous ulceration Erythema only Pleomorphic ulcers Abnormal on biopsy only

35 33 29 19 (54%) 6 4 9 17 (49%) 10 4 3 2 (6%)

3/11, x2 test, P = 0.01). Those with more severe clinical oral inflammatory scores were more likely to have intestinal inflammation (P = 0.025), and there was also a correlation between the histologic severity of oral inflammation and the histologic severity of gut inflammation, either as numbers of granulomas (P = 0.047, r = 0.338) or as chronic inflammation (P = 0.042, r = 0.333). The number of oral sites affected or type of inflammation present in the oral cavity did not predict intestinal involvement. There was no relationship between any blood parameter abnormality and intestinal involvement.

FIGURE 5. Patches of erythema and a single aphthous ulcer seen in the transverse colon at colonoscopy in a patient with OFG.

844

FIGURE 6. Colonoscopy in a patient with OFG showing erythema and more extensive ulceration in the cecum and ileocecal valve fold.

DISCUSSION The term OFG has been used collectively to describe a number of diseases restricted to the oro-facial region, with granulomatous inflammation as the histologic hallmark. This includes sarcoidosis, Melkersson-Rosenthal syndrome, Cheilitis granulomatosa, and oral CD. However, there remains confusion regarding the correct terminology, and in clinical practice, the majority of OFG cases present as a separate clinical entity, often difficult to treat and of unknown etiology.

FIGURE 7. Terminal ileal biopsy from a patient with OFG showing an inflammatory cell infiltrate with loosely formed noncaseating epithelioid granuloma. q 2005 Lippincott Williams & Wilkins

Inflamm Bowel Dis  Volume 11, Number 9, September 2005

Oro-Facial Granulomatosis

It is not clear whether patients with OFG who later develop CD should retain the diagnosis of OFG or be more appropriately labeled with oral CD.12 The association of oral lesions with CD has been well described. This includes granulomatous inflammation affecting various sites in the oral cavity but also aphthous ulceration, gingivitis, angular cheilitis, glossitis, candidiasis, and dental caries.13–15 Many of these features such as aphthous ulcers and glossitis are considered secondary to the effects of active inflammation elsewhere or to hematinic deficiency.5,6 The true incidence of oral CD is difficult to define but may be in the region of 6%.16,17 It is clear that patients with gut CD may present with oral inflammation many years before the onset of symptomatic gut inflammation. In 1 study of 29 patients defined as having oral CD, the oral disease predated the detection of intestinal disease in 9 cases.18 Likewise, Scully et al8 described a group of patients with oral CD in whom 63% had no gut symptoms. These patients were also studied by sigmoidoscopy and barium radiology, and evidence of intestinal CD was shown in 37% of patients. However, barium studies and limited endoscopic examination would clearly underestimate the prevalence of any gut inflammation. Our study represents the first detailed description of the oral and intestinal features of OFG. We have shown that, among a carefully defined group of patients with OFG, 54% have intestinal abnormalities detectable by endoscopy, despite an absence of any preexisting gut symptoms. These intestinal abnormalities are generally discrete, with only 2 cases in this series showing endoscopic features typical of CD. Furthermore, granulomas were shown in a high proportion of cases (68%), much greater than the 17% to 30% reported in studies of endoscopic biopsies from patients with CD.19–23 In gut CD, finding granulomas on histology indicate a more aggressive course and independently predict surgical treatment, but there is no other evidence to suggest the presence or absence of granulomas indicates any difference in disease behavior or in diagnosis.22 However, the high proportion of granulomas in the patients studied with OFG may suggest that OFG and CD are different conditions. Patients presenting with OFG at a younger age are much more likely to have intestinal inflammation. No other oral or intestinal features differed with age of onset. Young onset OFG, particularly as children, may differ in etiology from OFG seen in older adults. Alternatively, associated gut inflammation may regress with time. Recent interest has focused on the potential role of diet in OFG with reports of benefit from cinnamon- and benzoate-free diets and from elemental diet.24,25 Personal observation suggests response is more likely to diet in children with OFG as is the case for CD. The exact nature of the relationship between diet and OFG needs further detailed study, but it seems likely that OFG, particularly in children and young adults, has a dietary trigger as the primary cause.

Likewise, few studies have addressed the immunologic characteristics of the inflammation in OFG. Lim et al26 reported a restricted T-cell repertoire and clonal T-cell expansion in the lesional vicinity not observed in the normal oral mucous, suggesting a delayed hypersensitivity-type response to an unidentified antigen. Gibson et al27 also suggested an allergic basis for OFG. This group showed that 50% of patients with OFG had HLA-A2 and 25% had HLA-A11. A consistent HLA relationship has not been shown in CD.28,29 Studies from our group have shown raised serum IgA antibodies to Saccharomyces cerevisiae in patients with intestinal involvement but not in OFG alone.30 Mycobacterium paratuberculosis retains a possible role in the etiology of CD, with many studies reporting detection in inflamed gut.31,32 One study in OFG, however, did not detect Mycobacterium paratuberculosis in any case examined.33 Overall, therefore, while limited data exist from which to draw firm conclusions, OFG and CD would seem more likely to be 2 separate entities. OFG defines a group of conditions in which granulomatous inflammation occurs in the oral cavity, especially the lip. In some cases, the inflammation seems limited to the oral cavity. However, in a large proportion of OFG cases, in particular those presenting at a younger age, the oral inflammation may be part of a more generalized, discrete granulomatous enteritis. The intestinal inflammation detectable in OFG would conventionally raise the possibility of CD. However, OFG with associated gut inflammation may represent an entirely separate entity in which discrete granulomatous inflammation occurs throughout the gastrointestinal tract in response to an unknown, but perhaps dietary, antigen or antigens. Further studies are required to determine the immunologic, genetic, and microbiological characteristics of the inflammation in OFG and, furthermore, to determine to natural history with respect to possible later presentation with more typical CD.

q 2005 Lippincott Williams & Wilkins

REFERENCES 1. Wiesenfield D, Ferguson MM, Mitchell DN, et al. Orofacial granulomatosis- a clinical and pathological analysis. Q J Med. 1985;54:101–113. 2. Rogers RS. Melkersson-Rosenthal syndrome and orofacial granulomatosis. Dermatol Clin. 1996;14:371–379. 3. Zimmer WM, Rogers RS III, Reeve CM, et al. Orofacial manifestations of Melkersson-Rosenthal syndrome. A study of 42 patients and review of 220 cases from the literature. Oral Surg Oral Med Oral Pathol. 1992; 74:610–619. 4. Takada K, Ina Y, Noda M, et al. The clinical course and prognosis of patients with severe, moderate or mild sarcoidosis. J Clin Epidemiol. 1993;46:359–366. 5. Sundh B, Emilson CG. Salivary and microbial conditions and dental health in patients with Crohn’s disease: a 3 year study. Oral Surg Oral Med Oral Pathol. 1989;67:286–290. 6. Halme L, Meurman JH, Laine P, et al. Oral findings in patients with active or inactive Crohn’s disease. Oral Surg Oral Med Oral Pathol. 1993; 76:175–181. 7. Field EA. Oral lesions in IBD. Inflamm Bowel Dis Monit. 2001;2:66–72. 8. Scully C, Cochran KM, Russel R, et al. Crohn’s disease of the mouth: an indicator of intestinal involvement. Gut. 1982;23:198–201.

845

Sanderson et al

Inflamm Bowel Dis  Volume 11, Number 9, September 2005

9. Plauth M, Jenss H, Meyle J. Oral manifestations of Crohn’s disease. J Clin Gastroenterol. 1991;13:9–37. 10. Eveson JW. Granulomatous disorders of the oral mucosa. Sem Diagn Pathol. 1996;13:118–127. 11. Setterfield J, Shirlaw PJ, Kerr-Muir M, et al. Mucous membrane pemphigoid: a dual circulating antibody response with IgG and IgA signifies a more severe and persistent disease. Br J Dermatol. 1998;138:602–610. 12. Challacombe SJ. Oro-facial granulomatosis and oral Crohn’s disease: are they specific diseases and do they predict systemic Crohn’s disease? Oral Dis. 1997;3:127–129. 13. Dudeney TP, Todd IP. Crohn’s disease of the mouth. Proc R Soc Med. 1969;62:1237. 14. Varley EW. Crohn’s disease of the mouth: report of three cases. Oral Surg Oral Med Oral Pathol. 1972;33:570–578. 15. Field EA, Tyldesley WR. Oral Crohn’s disease revisited—a ten year review. Br J Oral Surg. 1989;27:1114–1123. 16. Basu MK, Asquith P. Oral manifestations of inflammatory bowel disease. Clin Gastroenterol. 1980;9:307–321. 17. Williams AJK, Wray D, Ferguson A. The clinical entity of orofacial Crohn’s disease. Q J Med. 1991;79:451–458. 18. Scully C, Eveson JW. Oral granulomatosis (leading article). Lancet. 1991;338:20–21. 19. Le Berre N, Heresbach D, Kerbaol M, et al. Histological discrimination of idiopathic inflammatory bowel disease from other types of colitis. J Clin Pathol. 1995;48:749–753. 20. Surawicz CM, Haggitt RC, Husseman M, et al. Mucosal biopsy diagnosis of colitis: acute self-limited colitis and idiopathic inflammatory bowel disease. Gastroenterology. 1994;107:755–763. 21. Potzi R, Walgram M, Lochs H, et al. Diagnostic significance of endoscopic biopsy in Crohn’s disease. Endoscopy. 1989;21:60–62. 22. Heresbach D, Alexandre JL, Branger B, et al. Frequency and significance of granulomas in a cohort of incident cases of Crohn’s disease. Gut. 2005;54:215–222.

23. Schmitz-Moormann P, Pittner PM, Malchow H, et al. The granuloma in Crohn’s disease. A bioptical study. Pathol Res Pract. 1984;178:67– 76. 24. Lomer M, White A, Escudier M, et al. Cinnamon and benzoate free diet as primary treatment of oro-facial granulomatosis. Gut. 2004;53(Suppl 3):A32. 25. Sweatman MC, Tasker R, Warner JO, et al. Oro-facial granulomatosis. Response to elemental diet and provocation by food additives. Clin Allergy. 1986;16:331–338. 26. Lim SH, Stephens P, Cao Q-X, et al. Molecular analysis of T cell receptor b variability in a patient with orofacial granulomatosis. Gut. 1997; 40:683–686. 27. Gibson J, Neilly JB, Wray AP, et al. 99Tcm-HMPAO leucocyte labelling in orofacial granulomatosis and gastrointestinal Crohn’s disease in childhood and early adulthood. Nucl Med Commun. 2000;21:155–158. 28. Satsangi J, Welsh KI, Bunce M, et al. Contribution of genes of the major histocompatibility complex to susceptibility and disease phenotype in inflammatory bowel disease. Lancet. 1996;347:1212–1217. 29. Naom I, Lee J, Ford D, et al. Analysis of the contribution of HLA genes to genetic predisposition in inflammatory bowel disease. Am J Hum Genet. 1996;59:226–233. 30. Savage NW, Barnard K, Shirlaw PJ, et al. Serum and salivary IgA antibody responses to Saccharomyces cerevisiae, Candida albicans and Streptococcus mutans in orofacial granulomatosis and Crohn’s disease. Clin Exp Immunol. 2004;135:483–489. 31. Sanderson JD, Moss MT, Tizard ML, et al. Mycobacterium paratuberculosis DNA in Crohn’s disease tissue. Gut. 1992;33:890–896. 32. Fidler HM, Thurrell W, Johnson NM, et al. Specific detection of Mycobacterium paratuberculosis DNA associated with granulomatous tissue in Crohn’s disease. Gut. 1994;35:506–510. 33. Riggio MP, Gibson J, Lennon A, et al. Search for Mycobacterium paratuberculosis DNA in orofacial granulomatosis and oral Crohn’s disease tissue by polymerase chain reaction. Gut. 1997;41:646–650.

846

q 2005 Lippincott Williams & Wilkins