Gut 1995; 37: 708-711

Ranitidine, diarrhoea, and lymphocytic colitis

L Beaugerie, N Patey, N Brousse

AbstractA 69 year old woman developed chronicdiarrhoea while being treated with rani-tidine. Sigmoidoscopy was performedafter six weeks and showed typical histo-logical features of lymphocytic colitis.Ranitidine was withdrawn and the diar-rhoea resolved. Eight months later, a 72hour oral rechallenge period ofranitidine,was performed immediately preceded(period 1) and followed (period 2) by sig-moidoscopy and biopsy. Diarrhoearecurred during the rechallenge periodand resolved again within one day afterdrug withdrawal. The mean (SEM)intraepithelial lymphocyte count was notsignificantly different between periods 1and 2 (11.9 (0.6) and 13.1 (0.4) per 100,respectively). An immunopathologicalstudy of 30 serial sections ofbiopsy speci-mens was performed for both periods 1and 2. The expression ofHLA-DR by therectal epithelium was mild or absent in allsections from period 1, and was consider-able in 25 of 30 sections from period 2(p<10-9). It is suggested that the oralintake of ranitidine was responsible forthe diarrhoea and induced the immuno-pathological signs of activation of therectal mucosal immune system during therechallenge period.(Gut 1995; 37: 708-711)

Keywords: ranitidine, diarrhoea, lymphocyiic colitis.

Department ofGastroenterology,H6pital Rothschild,Paris, FranceL Beaugerie

Department ofPathology, H6pital desEnfants Malades,Paris, FranceN PateyN Brousse

Correspondence to:Dr L BeaugerieHepatogastroenterologie,H6pital Rothschild, 33boulevard de Picpus, 75571Paris cedex 12, France.

Accepted for publication26 April 1995

Lymphocytic colitis, previously termed 'micro-scopic colitis', 1-3 is a clinicopathological syn-drome of chronic water diarrhoea, diffusecolonic mucosal inflammatory changes, and an

increased intraepithelial lymphocyte countdespite normal endoscopic tests.4The cause of most cases of lymphocytic

colitis is unknown. Lymphocytic colitis maybe drug induced, however, as in the threecases of lymphocytic colitis with protracteddiarrhoea secondary to the longterm use ofCyclo 3 Fort, a phlebotonic drug used inFrance, which we reported previously.5 Wedescribe here the case of a woman who devel-oped chronic diarrhoea while being treatedwith ranitidine and who, after six weeks, wasfound to have lymphocytic colitis. Eightmonths after the withdrawal of ranitidine andthe cessation of diarrhoea, we performed a 72hour rechallenge test with ranitidine givenorally, preceded and followed by rectalbiopsy. Based on the clinicopathologicalresults of this test, we suggest that the diar-rhoea was drug induced. Furthermore, we

show that the oral intake of ranitidine in itselfcan induce an activation of the rectal mucosal

immune system. We cannot conclude, how-ever, whether lymphocytic colitis was rani-tidine induced or ranitidine exacerbated, aspersistant histological features of lymphocyticcolitis were seen before the drug rechallenge.

Case reportA 69 year old woman was admitted to ourdepartment in February 1993 for intense epi-gastric pain. She had a past history of subtotalthyroidectomy for toxic multinodular goitre atage 16, and she had been treated intermittentlyfor 10 years with non-steroidal anti-inflamma-tory drugs for psoriatic arthritis. At the timeof admission, she was receiving diclofenac100 mg daily for low back pain. Her physicalexamination was unremarkable, except for epi-gastric pain provoked by palpation of theabdomen. An upright film of the abdomen wasnormal. Upper gastrointestinal endoscopyshowed a duodenal ulcer. Non-steroidalinflammatory drugs were definitively with-drawn. The patient received omeprazole20 mg daily and reported relief of epigastricpain within the first 72 hours of this treatment.Six weeks later, omeprazole was replaced withranitidine 150 mg daily in the evening. Duringall this period, the patient continued to receivelevothyroxin for postsurgical thyroid insuffi-ciency and paracetamol for lumbar pain.At the beginning of the ranitidine treat-

ment, the patient had two to four formedstools per week with no past history of inter-mittent or sustained diarrhoea. After one weekreceiving ranitidine, her stools became softand more frequent. After 10 days of treat-ment, she had persistent diarrhoea (five toseven liquid stools per day, including twoto three stools during the night). An initialflexible sigmoidoscopy was performed sixweeks after the onset of diarrhoea (period 0).Endoscopic examination showed normalcolonic mucosa. Three biopsy samples weretaken from the rectum and fixed in Bouin'ssolution. Paraffin wax sections were stainedwith haematoxylin and eosin. The intraepi-thelial lymphocyte count was assessed byestimating their number per 100 epithelialcells, 100 to 300 enterocytes being counted ateach section level. The intraepithelial lympho-cyte count was evaluated in three sectionsfrom each biopsy specimen - that is, for ninesection levels. The subepithelial collagen layerwas measured in 10 intercryptal spaces in ninewell oriented sections.The mean (SEM) intraepithelial lymphocyte

count was 20-4 (1-2) per 100 epithelial cells.Patchy surface epithelial detachment was seenin most of the sections. There was moderateinfiltration of the lamina propria by mono-nuclear cells, but not by neutrophils or

708

on October 19, 2021 by guest. P

rotected by copyright.http://gut.bm

j.com/

Gut: first published as 10.1136/gut.37.5.708 on 1 N

ovember 1995. D

ownloaded from

Ranitidine, diarrhoea, and lymphocytic colitis

eosinophils (Fig 1 (A)). The thickness of thesubepithelial collagen layer was within thenormal range, from 0 to 7 ,um in all thesections. All these histological features wereconsistent with the diagnosis of lymphocyticcolitis.

Ranitidine was withdrawn, whereas levo-thyroxin and paracetamol were continued.Diarrhoea disappeared within 48 hours and thebowel habit returned to its previous normalstate. Omeprazole 20 mg three times a weekwas then introduced as a maintenance treat-ment for the duodenal ulcer. Eight monthslater, the patient was seen at our unit in anambulatory setting, and reported no recur-rence of diarrhoea. We suggested a 72 hourranitidine rechallenge period with flexible sig-moidoscopy and biopsy specimens obtainedimmediately before (period 1) and after(period 2) the rechallenge. The aims of the testwere to assess the state of the colon by histo-logical examination after the drug had beenwithdrawn for some time and to detect the firstimmunopathological events induced by thedrug on rechallenge. The patient gave herinformed consent to the test. Omeprazole waswithdrawn, and ranitidine given orally at adose of 150 mg daily for three days. Duringeach sigmoidoscopy, three biopsy specimenswere obtained and fixed in Bouin's solution.Paraffin wax sections were cut and stained withhaematoxylin and eosin. The intraepitheliallymphocyte count was assessed in 10 sectionsfrom each biopsy specimen - that is, in 30 sec-tions per period - as described above, and inone section from rectal specimens of 10 adultcontrols with a histologically normal colon.The pathologist was unaware whether the sec-tions were from period 1 or 2. The meanintraepithelial lymphocyte counts were com-pared using variance analysis, and differencesbetween the pairs of means compared by theNewman-Keuls test.

During each sigmoidoscopy in periods 1 and2, three additional biopsy specimens from therectum were taken and immediately frozen,and then stored at -80°C until cryostat sec-tions were obtained. A three stage indirectimmunoperoxidase staining technique usinganti-CD 25 and anti-HLA-DR monoclonalantibodies was used on these sections. CD 25is an interleukin 2 receptor and its expressionis a marker of activation of a number ofcell types, including T cells, B cells, andmacrophages. HLA-DR is normally expressedby antigen-presenting cells - that is, macro-phages of the lamina propria - but it is not orbarely expressed by colonic epithelial cells.6 7

The expression of HIA-DR by the epithelialcells in crypts and intercryptic spaces wasscored as absent or mild (no or a few positivecells) or considerable (numerous positive cells)in 30 different sections for each time period.Again, the pathologist was unaware of theperiod. Similarly, the expression of CD 25 bymononuclear cells of the lamina propria wasscored as absent or mild (none or a few positivecells) or considerable (numerous positive cells)and assessed in 30 different sections for eachperiod. The expression of HLA-DR and CD

A

Figure 1: Histological appearance oft. (A) the rectal mucosaafter six weeks of treatment with ranitidine and withcontinued diarrhoea (period 0); (B) the rectal mucosa eightmonths after ranitidine withdrawal (period 1)(haematoxylin and eosin; orginal magnfication X 25).Note the partial regression of the lamina propria infiltrationby mononuclear cells between periods 0 and 1.

25 was compared between periods 1 and 2 bythe x2 test.

Diarrhoea recurred on the second and thirdday of ranitidine rechallenge, and disappearedagain within one day after drug withdrawal.The endoscopic appearance of rectal mucosawas normal before as well as after rechallenge.

709

on October 19, 2021 by guest. P

rotected by copyright.http://gut.bm

j.com/

Gut: first published as 10.1136/gut.37.5.708 on 1 N

ovember 1995. D

ownloaded from

Beaugerie, Patey, Brousse

Intraepithelial lymphocyte count of rectal mucosa in the patientfor the three periods, and in10 controls

Intraepithelial lymphocyte count(°o, mean (SEM))

PatientAt the time of diarrhoea (period 0, n=9 sections)Just before rechallenge (period 1, n= 10 sections)At day 3 of rechallenge (period 2, n= 10 sections)

Controls (n= 10)

20-4 (1-2)*1 1-9 (0-6)t13-1 (0.4)t7.4 (1 1)§

*Significantly different from t and f; t not different from *, t, and * significantly different from §.

The Table gives the mean intraepithelial lym-phocyte counts. Intraepithelial lymphocytecounts for periods 1 and 2 did not differ signifi-cantly, but both were significantly lower thanthe period 0 intraepithelial lymphocyte count(p< 10-5), and significantly higher than con-trols (p<00l). There was a mild lamina pro-pria infiltration by mononuclear cells in bothperiods (Fig 1 (B)). The subepithelial collagenlayer thickness was within the normal range,from 0 to 7 ,um in all the sections from bothperiods 1 and 2.The expression of HLA-DR by the epithe-

lium was mild or absent in all 30 sectionsduring period 1 (Fig 2 (A)) and was consider-able in 25 of the 30 sections from period 2 -that is, after ranitidine rechallenge (Fig 2 (B)).This difference was highly significant(p< 10-9). CD 25 expression by mononuclearcells of the lamina propria was more oftenmarked during period 2 (21 of 30 sections)than during period 1 (13 of 30 sections,p<0O05).

DiscussionMicroscopic colitis is now regarded as an

'umbrella' term covering any case of colitis inwhich there is histological but not colonoscopicor barium enema abnormality.89 The histo-logical abnormalities encountered when our

patient was taking ranitidine met all the diag-nostic criteria of lymphocytic colitis and notthose of collagenous colitis.10 At this point, thediagnosis of idiopathic lymphocytic colitis wasthe most likely for the following reasons. Firstly,like most patients with lymphocytic colitis,our patient was a middle aged woman, withassociated arthritic symptoms. Secondly, therewas no evidence of drug induced lymphocyticcolitis. As a matter of fact, the longterm use ofnon-steroidal anti-inflammatory drugs has beenpointed out as a possible cause of collagenouscolitis11 12 but not lymphocytic colitis.

Eight months after stopping ranitidine, we

performed a 72 hour rechallenge test with oralranitidine preceded and followed by rectalbiopsy. Although the patient had experiencedno recurrence of diarrhoea after the ranitidinewas stopped, the rectum was still histologi-cally abnormal before the drug rechallenge.This result shows that lymphocytic colitis canbe seen histologically in the absence of diar-rhoea. Clinicopathological discrepancy hasbeen reported in the case of collagenouscolitisl3 but, to our knowledge, not in the case

of lymphocytic colitis.Diarrhoea recurred during the 72 hour

rechallenge period and it resolved again soon

Figure 2: HLA-DR antigen expression by rectal mucosacells immediately before (A, period 1) and after (B, period2) a 72 hour period ofdrug rechallenge. Note the absence ofHLA-DR antigen expression by epithelial cells beforerechallenge (A) and the considerable HLA-DR antigenexpression of epithelial cells after rechallenge (B)(immunoperoxidase staining with monoclonal antibodydirected against HLA-DR non-polymorphic determinants,original magnification X20).

after withdrawal of ranitidine. This featurestrongly suggested that ranitidine was the causeof the diarrhoea. This suggestion is not in itselforiginal, as diarrhoea was previously reported asthe most common side effect of ranitidine.14

A

B

710

on October 19, 2021 by guest. P

rotected by copyright.http://gut.bm

j.com/

Gut: first published as 10.1136/gut.37.5.708 on 1 N

ovember 1995. D

ownloaded from

Ranitidine, diarrhoea, and lymphocytic colitis 711

The immunopathological study of rectalbiopsy specimens performed immediatelybefore and after the 72 hour rechallenge periodallowed us to show the existence of earlyimmunopathological changes induced by theoral intake of ranitidine. These changes - thatis, augmentation of CD 25 expression bylamina propria mononuclear cells and appear-ance of diffuse HLA-DR expression by epithe-lial cells - were consistent with activation ofmucosal immune cells.'5 16 The site and themechanism of this activation are not specifiedby our model, nor could we discover if the drugitself or its metabolites are responsible for thisactivation. Nevertheless, it can be hypothe-sised that the histological appearance of lym-phocytic colitis seen after six weeks ofdiarrhoea and ranitidine intake was partly ortotally related to the chronic immune activa-tion of the rectal immune system secondary toranitidine. It could not be stated in our case,however, whether the lymphocytic colitis wasdrug induced as has been shown for Cyclo 3Fort,5 or simply a 'drug exacerbed' idiopathiccolitis. As a matter of fact, histological featuresconsistent with a minor aspect of lymphocyticcolitis were still seen in our patient eightmonths after ranitidine had been withdrawn.Moreover, our patient exhibited an underlyingcondition (that is, middle age and arthriticdisease), usually seen in patients with idio-pathic lymphocytic colitis.'0Our finding suggests that some drugs could

have a role in the pathogenesis or the exacerba-tion of some cases of lymphocytic colitis, orboth. We propose that all recent and concur-rent drug treatments should be noted whenlymphocytic colitis is diagnosed. Withdrawalshould be attempted if possible when the roleof a drug is suspected because ofprevious find-ings. When a drug has not been implicated pre-viously in published works, we suggest that ashort-term rechallenge should be considered,using our immunopathological model. In that

way, early immunopathological changesinduced by the drug may be detected, longafter the drug has been stopped.The authors wish to thank Mrs Odile Le Pelletier for her experttechnical assistance, and Dr David Levy for his great help inpreparing the manuscript.

1 Read NW, Krejs GJ, Read MG, Santa Ana CA, MorawskiSG, Fordtran JS. Chronic diarrhea of unknown origin.Gastroenterology 1980; 78: 264-71.

2 Kingham JG, Levison DA, Ball JA, Dawson AM.Microscopic colitis - a cause of chronic watery diarrhoea.BMJ 1982; 285: 1601-4.

3 Bo Linn GW, Vendrell DD, Lee E, Fordtran JS. An evalua-tion of the significance of microscopic colitis in patientswith chronic diarrhoea. Jf Clin Invest 1985; 75: 1559-69.

4 Lazenby AJ, Yardley JH, Giardiello FM, Jessurun J, BaylessTM. Lymphocytic ('microscopic') colitis: a comparativehistopathologic study with particular reference to collage-nous colitis. Hum Pathol 1989; 20: 18-28.

5 Beaugerie L, Luboinski J, Brousse N, Cosnes J, ChateletFP, Gendre JP, et al. Drug induced lymphocytic colitis.Gut 1994; 35: 426-8.

6 Hirata I, Austin LL, Blackwell WH, Weber JR,Dobbins WO. Immunoelectron microscopic localizationof HLA-DR antigen in control small intestine and colonin inflammatory bowel disease. Dig Dis Sci 1986; 31:1317-30.

7 Mayer L, Eisenhardt D, Salomon P, Bauer W, Plous R,Piccinini L. Expression of class II molecules on intestinalepithelial cells in humans. Gastroenterology 1991; 100:3-12.

8 Levison DA, Lazenby AJ, Yardley JH. Microscopic colitiscases revisited. Gastroenterology 1993; 105: 1593-5.

9 Bogomoletz WV, Flejou JF. Microscopic colitis: a 'trans-atlantic' unifying concept. Gastroenterology 1994; 106:1727.

10 Giardiello FM, Lazenby AJ, Bayless TM, Levine EJ, BiasWB, Ladenson PW, et al. Lymphocytic (microscopic)colitis. Clinicopathological study of 18 patients and com-parison to collagenous colitis. Dig Dis Sci 1989; 34:1730-8.

11 Giardiello FM, Hansen III FC, Lazenby AJ, Hellman DB,Milligan FD, Bayless TM, et al. Collagenous colitis in set-ting of nonsteroidal inflammatory drugs and antibiotics.Dig Dis Sci 1990; 35: 257-60.

12 Riddell RH, Tanaka M, Mazzoleni G. Non-steroidal anti-inflammatory drugs as a possible cause of collagenouscolitis: a case-control study. Gut 1992; 33: 683-6.

13 Wang KK, Perrault J, Carpenter HA, Schroeder KW,Tremaine WJ. Collagenous colitis: a clinicopathologiccorrelation. Mayo Clin Proc 1987; 62: 665-71.

14 Wormsley KG. Safety profile of ranitidine. A review. Drugs1993; 46: 976-85.

15 MacDonald TT, Weinel A, Spencer J. HLA-DR expressionin fetal human small intestinal epithelium. Gut 1988; 19:1342-8.

16 Choy MY, Walker-Smith JA, Williams CB, MacDonaldTT. Differential expression of CD25 (interleukin-2 recep-tor) or lamina propria T cells and macrophages in theintestinal lesions in Crohn's disease and ulcerative colitis.Gut 1990; 31: 1365-70. on O

ctober 19, 2021 by guest. Protected by copyright.

http://gut.bmj.com

/G

ut: first published as 10.1136/gut.37.5.708 on 1 Novem

ber 1995. Dow

nloaded from