Gut, 1975, 16, 249-254

Oral manifestations of Crohn's diseaseM. K. BASU, P. ASQUITH, R. A. THOMPSON, AND W. T. COOKE

From the Departments of Oral Pathology and ofExperimental Pathology, Birmingham University,the Regional Immunology Laboratory, East Birmingham Hospital, and the Nutritional andIntestinal Unit, General Hospital, Birmingham

SUMMARY In a systematic study of 100 patients with Crohn's disease, 100 with ulcerative colitis,and of 100 normal subjects matched for age, sex, and denture status, nine patients with Crohn'sdisease, two with ulcerative colitis, and one normal control were found to have oral lesions. InCrohn's disease, the macroscopic and histological appearances resembled those encountered else-where in the gastrointestinal tract and their incidence was related to the activity of the disorder.The lesions in the other two groups were different macroscopically and histologically.

Production of salivary IgA was found to be reduced in Crohn's patients with active bowel disease.It is suggested that the occurrence of oral lesions in patients with Crohn's disease might representa local immunological reaction to oral antigens.

There have been a number of reports of 'apparentlyspecific' oral lesions in Crohn's disease (Dudeneyand Todd, 1969; Issa, 1971; Schiller, Golding,Peebles, and Whitehead, 1971; Bishop, Brewster,and Antonioli, 1972; Ellis and Truelove, 1972;Stankler, Ewen, and Kerr, 1972), which macro-scopically and histologically resemble those foundin the gastrointestinal tract. A retrospective analysisby Croft and Wilkinson (1972) of 332 Crohn'spatients also showed that 6% of such patients hadgiven a history of oropharyngeal ulceration. How-ever, patients without Crohn's disease commonlysuffer from recurrent 'aphthous type' ulceration,which was found by Sircus, Church, and Kelleher(1957) to have an incidence rate of 19-3% in a sampleof 1783 patients attending hospital outpatient andgeneral practitioner clinics.

Oral lesions in general are often associated withlow serum iron and folate levels (Jacobs and Cavill,1968; Dolby, 1972), findings which are common inpatients with Crohn's disease (Swan, 1969; Dyer,Child, Mollin, and Dawson, 1972). It has also beensuggested that immunological mechanisms could berelevant to the aetiology of mouth ulcers (Lehner,1972). Consequently, the finding of alterations inhumoral and cellular reactivity in many such patients(Kraft and Kirsner, 1971) may be relevant.The present study was carried out to compare the

incidence of oral lesions in patients with Crohn'sdisease with that in controls and to study theirReceived for publication 10 February 1975.

macroscopic and histological features; second!v, tostudy the interrelationship of mouth lesions withother manifestations of Crohn's disease; and lastlyto correlate them with haematological and bio-chemical values and the results of salivary IgAestimations with a view to establishing their possibleaetiology and pathogenesis.

Materials and Methods

One hundred patients with Crohn's disease (38 menand 62 women) were examined for the presence oforal lesions and their incidence was compared withthat in two further groups, one consisting of 100normal control subjects attending the ConservationDepartment of the Dental Hospital, matched forage, sex, and denture status, and the other consist-ing of 100 patients with ulcerative colitis (42 menand 58 women). Based on the macroscopic appear-ances corresponding with the previous descriptionsof oral Crohn's lesions, four lesions were considered-cobblestone, ulcerative, tags, and erythema

migrans. Recurrent aphthous type ulcers were notincluded.A biopsy of each oral lesion was carried out under

local anaesthesia. Step serial 8 ,t-thick sections wereprepared from paraffin-embedded specimens andstained with haematoxylin and eosin for routinehistological examination, with Gram's iodine withfast green counterstain for microorganisms, withPAS to disclose Candida albicans, with Weigert's

2492

M. K. Basu, P. Asquith, R. A. Thompson, and W. T. Cooke

stain for internal elastic lamina of the blood vessels,and with the silver cyanate reaction for peripheralnerves (Rowles and Brain, 1960).

In Crohn's patients the overall incidence of orallesions was correlated with age, sex, primary site ofthe disease, and the presence of perianal and skinlesions, and also with the serum levels of folate, B12,iron, albumin, and seromucoids.

Finally, in 20 patients with Crohn's disease and 20age- and sex-matched volunteers (hospital staff andlaboratory personnel), a timed sample of saliva wascollected in an ice-cooled test tube from the rightparotid duct under acid drop stimulation and itsvolume measured. The saliva was immediatelycentrifuged at 8000 g for 10 minutes at 4°C and adrop of 1 in 100 merthiolate in distilled water addedto the supernatant before being stored at -20°C.The level of total IgA was estimated within onemonth by a modified Mancini single radial diffusiontechnique (Fahey and McKelvey, 1965) using amonospecific IgA antiserum raised in rabbits and anormal control reference serum as standard. Theparotid IgA secretion rate was then calculated.

Fig I

Fig 3

rig 4

Fig 3 Ulcerative lesion on palate of a patient withulcerative colitis. Compare shape, size, and margin ofthe ulcer with figure 2.

Fig 4 Tags on the inner side of the cheek of Crohn'spatient. Tongue at lower right corner.

Fig 2

Fig 1 Cobblestone lesion on the labial gingiva abovecentral and lateral incisor teeth of Crohn's patient.

Fig 2 Ulcerative lesion on lateral margin of thetongue ofa Crohn's patient, showing linear shape,hyperplastic margins, and depth.

250)

Oral manifestations of Crohn's disease

Results

CLINICAL FINDINGSAn increased incidence (p < 0-05) of mouth lesionswas found in Crohn's disease (9/100) compared withthat in patients with ulcerative colitis (2/100) andcontrols (1/100). The ages of the nine patients withCrohn's disease ranged from 21 to 72 years and sixwere female. Involvement of both the small and largeintestine was present in eight patients, whilst one hadCrohn's colitis only: perianal disease was found insix, and one had an eczematous skin lesion.

MACROSCOPIC APPEARANCESFour patients with Crohn's disease had areas ofovert hyperplasia of the oral mucous membrane withan inflamed and deeply fissured surface, producinga cobblestone appearance (fig 1). The oral ulcers seenin three patients with Crohn's disease were deep,linear in shape, and with a nodular hyperplasticmargin and clean base (fig 2). In contrast, the ulcers

Fig 5 Epithelioid cell follicle in submucosa of oralmucous membrane. Cobblestone lesion from Crohn'spatient (x 200, H & E).

seen in two patients with ulcerative colitis wereirregular in shape, with rolled margins, and werecovered by a greyish-white slough (fig 3). The oraltags seen in one patient with Crohn's disease werefleshy and inflamed (fig 4) and quite distinct to thatseen in the normal control, having the typicalappearance of leukoedema. The appearance oferythema migrans in one Crohn's patient wascharacteristic of that condition.

HISTOLOGYOf the nine oral lesions in Crohn's disease only oneshowed epithelioid cell follicles in the lamina propriaand submucosa (fig 5), some with giant cells. Al-though lymphoid follicles were found in one otherbiopsy only (fig 6), the lamina propria and sub-mucosa in all biopsies showed abnormal focalcollections of lymphocytes and perivascular in-filtrates of mononuclear cells. Other non-specificsigns of chronic inflammation, such as neuronalhyperplasia, lymphangiectasia, and oedema of the

Fig 6 Secondary lymphoid follicle with germinalcentre in submucosa ofanother cobblestone lesion fromCrohn's patient ( x 200 H & E).

251

M. K. Basu, P. Asquith, R. A. Thompson, and W. T. Cooke

Oral Lesions Serum Folate Vitamin B12 Iron Seromucoid Albumin

Mean Percentage Mean Percentage Mean Percentage > 150 mgll00 ml < 4 g/100 ml< 4 nglml < 160 pglml < 60 ug/ml

With 12-68 ± 1-33 22 479 55 ± 3665 11 71-08 ± 3 58 44 75 78Without 15-69 ± 5 78 16 413-34 ± 171-4 5 71-86 ± 13-85 37 26 34

Table Incidence ofabnormal serum levels in Crohn's disease with and without oral lesions

lamina propria and submucosa, were found regularly.In addition, of particular interest was the presenceof focal collections of lymphocytes within thestromal connective tissue (fig 7) and signs of acinaratrophy and duct hyperplasia in the minor salivaryglands included in the biopsies.

In ulcerative colitis, the biopsies were quitedistinct with the lamina propria and submucosacontaining a polymorphonuclear infiltrate and a

large number of eosinophils with evidence of peri-vascular deposition of fibrin. That from the controlindividual showed parakeratosis and acanthosis ofthe surface epithelium with slight inflammation ofthe superficial layers of the lamina propria typicalof leukoedema (Shafer, Hine, and Levy, 1974).Finally bacteria and Candida were not seen in thehistological specimens.

HAEMATOLOGY AND BIOCHEMISTRYNo significant differences were found in the meanvalues or in the incidence of low levels for serumfolate, B12, or serum iron between Crohn's patientswith and without oral lesions (see table). In contrast,an increased number of patients with oral lesionshad raised serum seromucoid levels and low serumalbumin levels compared with the incidence inpatients without oral lesions (p < 0-05) (table I).

SALIVARY IgA LEVELThe parotid secretion rates for healthy controls andfor Crohn's patients showed no significant differ-ence. However, Crohn's patients with raised serumseromucoid and low serum albumin levels hadsignificantly lower (p < 0 05) IgA secretion ratesthan patients with low serum seromucoid and high

Regression of Partid IgA secretionrate on

IgA

Fig 7 Minor salivary glandfrom Crohn's patient.Focal lymphocytic infiltration ofstromal connectivetissue with acinar atrophy and duct dilatation(x 200, H & E).

Serum Seromucoidin Crohns Patients

Serum Albumin

in Crohns Patients01 .2-50 3-00 3-50 4-00 4-50 5-.0

g 1100 ml Serum Albumin

Fig 8 Regression ofparotid IgA secretion rate onserum seromucoid and serum albumin in Crohn'spatients.

252

Oral manifestations of Crohn's disease 253

serum albumin levels when compared with healthycontrols. The parotid IgA secretion rate of patientswith Crohn's disease also showed a significantregression (p < 0-05) when plotted against theirserum seromucoid and serum albumin levels (fig 8).

Discussion

In this study, oral lesions occurred more commonlyin patients with Crohn's disease compared with thesituation in normal controls and patients with ulcera-tive colitis. The age range, sex distribution, and siteof disease in Crohn's patients with oral lesions weresimilar to those reported in the literature (Dudeneyand Todd, 1969; Issa, 1971; Schiller et al, 1971;Bishop et al, 1972; Bottomley, Giorgini, andJulienne, 1972; Stankler et al, 1972; Varley, 1972;Eisenbud, Katzka, and Platt, 1972) and to those inthe total sample of Crohn's patients examined,suggesting that these factors were unimportant ininfluencing the incidence of mouth lesions.

In several respects the oral lesions macroscopic-ally and histologically resembled those seen in thegastrointestinal tract in Crohn's disease suggestingthat they were due to the same disease process.However, typical sarcoid-like granulomas were notseen. Such granulomas are thought to be the mostuseful histological sign of Crohn's disease, occurringin 60% of resected bowel specimens (Williams, 1964;Morson, 1971; Cook, 1972). The failure to find themin the biopsies of the mouth lesions may be relatedto the small size of the biopsy. Also the incidence ofgranulomas varies from tissue to tissue with only19% occurring in rectal biopsies from patients withCrohn's disease (Dyer, Stansfeld, and Dawson,1970).In the absence of sarcoid-like granuloma, an

aggregated pattern of lymphocytic infiltration ofthe lamina propria and the submucosa, somelymphocytes with germinal centres, perivascularinfiltration of lymphohistocytic cells, and neuronalhyperplasia was found in the oral lesions; thesefeatures have been considered as good indicators ofCrohn's disease (Morson and Dawson, 1972; Cookand Dixon, 1973). On this basis the histologicalfeatures of the mouth lesions in our patients wereconsistent with a Crohn's lesion. Of additionalinterest were the findings in minor salivary glands.These were similar to changes described in auto-immunity such as Sjogren's disease, rheumatoidarthritis, and lupus erythematosus (Whaley,Williamson, Chisholm, Webb, Mason, andBuchanan, 1973), and, as far as we can ascertain,they have not been reported previously in Crohn'sdisease.To explain the extraintestinal lesions of Crohn's

disease, various mechanisms have been proposed.Crohn and Yarnis (1958) suggested that perianallesions may be brought about by faecal contamina-tion and/or by the effects of gravitation. A retro-grade flow in the lymphatic channels connecting theaggregations of lymphoid tissue of the ileum and theperianal region has also been suggested (Fielding,1972). However, such mechanisms cannot be re-sponsible for the development of either the meta-static lesions of the skin (McCallum and Kinmont,1968; Mountain, 1970) or of the oral lesions. InCrohn's disease low serum albumin and high serumseromucoid levels are considered good indices ofdisease activity (Cooke, Fowler, Cox, Gaddie, andMeynell, 1958). Hence the present finding of asignificant correlation between the presence of orallesions in Crohn's patients and low serum albuminand high serum seromucoid levels suggests that theyare related to active Crohn's disease. In contrast,the lesions were not related to a deficiency of folate,B12, or iron, as reflected in serum levels of thesesubstances.

Finally, reduced secretion rates of parotid salivaryIgA have been found with increasing activity of thebowel disease. This can be explained if it is postu-lated that in active disease a transient transepithelialblock to IgA develops in the parotid glands, similarto that which has recently been demonstrated indiseased epithelium of resected bowel specimens(Green and Fox, 1973). Alternatively in activedisease part of the parotid IgA could exist aspolymers which would result in falsely low IgAestimates with the technique used. Finally, it couldalso be postulated that in patients with activedisease IgA plasma cells destined for the parotidgland are diverted to the site of maximum antigenicstimulation in the bowel. As IgA is thought to beimportant in the defence of mucosal surfaces(Tomasi and Bienenstock, 1968), then the relativelack of IgA in the mouth of patients with activebowel disease may allow invasion of the oral mucousmembrane by the food or bacterial antigens in themouth (Ivanyi and Lehner, 1970), so producing anoral lesion.

References

Bishop, R. P., Brewster, A. C., and Antonioli, D. A. (1972). Crohn'sdisease of the mouth. Gastroenterology, 62, 302-306.

Bottomley, W. K., Giorgini, G. L., and Julienne, C. H. (1972). Oralextension of regional enteritis (Crohn's disease). Oral Surg., 34,417-420.

Cook, M. G. (1972). The size and histological appearance of mesen-teric lymph nodes in Crohn's disease. Gut., 13, 970-972.

Cook, M. G., and Dixon, M. F. (1973). An analysis of the reliabilityof detection and diagnostic value of various pathologicalfeatures in Crohn's disease and ulcerative colitis. Gut, 14,255-262.

Cooke, W. T., Fowler, D. I., Cox, E. V., Gaddie, R., and Meynell,M. J. (1958). The clinical significance of seromucoids in

254 M. K. Basu, P. Asquith, R. A. Thompson, and W. T. Cooke

regional ileitis and ulcerative colitis. Gastroenterology, 34,910-919.

Croft, C. B., and Wilkinson, A. R. (1972). Ulceration of the mouth,pharynx, and larynx in Crohn's disease of the intestine. Brit.J. Surg., 59, 249-252.

Crohn, B. B., and Yarnis, H. (1958). Regional Ileitis, 2nd ed. Gruneand Stratton, New York.

Dolby, A. E. (1972). Advances in dentistry. Practic., 209, 551-557.Dudeney, T. P., and Todd, I. P. (1969). Crohn's disease of the mouth.

Proc. roy. Soc. Med., 62, 1237.Dyer, N. H., Child, J. A., Mollin, D. L., and Dawson, A. M. (1972).

Anaemia in Crohn's disease. Quart. J. Med., XLI, 419-436.Dyer, N. H., Stansfeld, A. G., and Dawson, A. M. (1970). The value

of rectal biopsy in the diagnosis of Crohn's disease. Scand. J.Gastroent., 5, 491-496.

Eisenbud, L., Katzka, I., and Platt, N. (1972). Oral manifestations inCrohn's disease. Oral Surg., 34, 770-773.

Ellis, J. P., and Truelove, S. C. (1972). Crohn's disease with mouthinvolvement. Proc. roy. Soc. Med., 65, 1080.

Fahey, J. L., and McKelvey, E. M. (1965). Quantitative determinationof serum immunoglobulins in antibody-agar plates. J.Immunol., 94, 84-90.

Fielding, J. F. (1972). Perianal lesions in Crohn's disease. J. roy.Coll. Surg. Edinb., 17, 32-37.

Green, F. H. Y., and Fox, H. (1973). An immunofluorescence study ofthe intestinal mucosa in ulcerative colitis and Crohn's disease.J. Path., 110, ii.

Issa, M. A. (1971). Crohn's disease of the mouth. Brit. dent. J., 130,247-248.

Ivanyi, L., and Lehner, T. (1970). Stimulation of lymphocyte trans-formation by bacterial antigens in patients with periodontaldisease. Arch. oral Biol., 15, 1089-1096.

Jacobs, A., and Cavill, I. (1968). The oral lesions of iron deficiencyanaemia: pyridoxine and riboflavin status. Brit. J. Haemat., 14,291-295.

Kraft, S. C., and Kirsner, J. B. (1971). Immunological apparatus ofthe gut and inflammatory bowel disease. Gastroenterology,60, 922-951.

Lehner, T. (1972). Cell-mediated immune response in oral disease:a review. J. oral Path., 1, 39-58.

McCallum, D. I., and Kinmont, P. D. C. (1968). Dermatologicalmanifestations of Crohn's disease. Brit. J. Derm., 80, 1-8.

Morson, B. C. (1971). Histopathology of Crohn's disease. Scand. J.Gastroent., 6, 573-575.

Morson, B. C., and Dawson, I. M. P. (1972). Gastrointestinal Path-ology. Blackwell, London.

Mountain, J. C. (1970). Cutaneous ulceration in Crohn's disease.Gut, 11, 18-26.

Rowles, S. L., and Brain, E. B. (1960). An improved silver method forstaining nerve fibres in decalcified sections of teeth. Arch.oral Biol., 2, 64-68.

Schiller, K. F. R., Golding, P. L., Peebles, R. A., and Whitehead, R.(1971). Crohn's disease of the mouth and lips. Gut, 12, 864-865.

Schafer, W. G., Hine, M. K., and Levy, B. M. (1974). A Textbookof Oral Pathology. Saunders, Philadelphia and London.

Sircus, W., Church, R., and Kelleher, J. (1957). Recurrent aphthousulceration of the mouth. Quart. J. Med., 26, 235-249.

Swan, C. H. J. (1969). The haematological complications of regionalenteritis. MD Thesis, University of Birmingham.

Stankler, L. Ewen, S. W. B., and Kerr, N. W. (1972). Crohn'sdisease of the mouth. Brit. J. Derm., 87, 501-504.

Tomasi, T. B., and Bienenstock, J. (1968). Immunoglobulins. Advane,Immunol., 9, 1-96.

Varley, E. W. B. (1972). Crohn's disease of the mouth: report of threecases. Oral Surg., 33, 570-578.

Whaley, K., Williamson, J.. Chisholm, D. M., Webb, J., Mason,D. K., and Buchanan, W. W. (1973). Sjogren's syndrome.Quart. J. Med., 166, 279-304.

Williams, W. J. (1964). Histology of Crohn's syndrome. Gut, 5, 510-516.