Mucinous rectal adenocarcinoma i n perianal Crohn's disease fistulas

Mucinous rectal adenocarcinoma in perianal Crohn's disease fistulas. Can J Gastroenterol 1991;5(2):59-61.A47-year-old male with longstanding Crohn's disease complicated by perianal fistulous disease is described. In this patient, presentation with a large acute ischiorectal abscess resulted in diagnosis of a complicating mucinous adenocarcinoma. Despite the relatively high frequency of perianal complications in C rohn's disease, diagnosis of carcinoma in chis setting is difficult, and survival remains poor.


C ROI IN'S DISEASE IS ASSOCIATED
with an inc reased incidence of gastrointesti na l tract carci noma. Most reported cancers in Crohn 's disease have appeared in the small intestine or in bypassed sma ll in testinal loops. A lthough perianal d isease is a common complication of Crohn's d isease, car-c inoma arising in the anorectum is rare. In rhe presen t paper, a case of adcnocarcinoma of the rectum b described in a rat ient wit h Crohn 's d isease rresenting with pcriana l fist ulas and abscess. Even though carci noma complicating perianal C rohn's disease is a n uncommon occurrence, awareness of this complica- Deparrment

CASE PRESENTATION A 4 7-year-old ma n had a diagnos is
of Cmhn's disease esrnblished in 1970, following prcsen tntion with right lower quadrant rain ,md presumed appendicitis. At l;1panHomy, a sma ll hnwel resecu on was pe rformed. The patient suhsequen rly Jevdoped an abdominal wall fist ula that required eight 111011 1 hs to heal. In 1980, he h:1d an ahdominal wall ahsces~ drained and furthe r resection of 19. 5 cm o( small intest ine. Suhseq uently, the patien t remained well except for mterm ittent diarrhe,1 and a small anal fisrula in itially detected in 1986 because of a slight ana l d ischarge. T h is resolved wi t ho ut treatment.
T he patient was fi rst referred m January J 990 because of recurrent dia rrhea. Exammat ion was no rma l except for a sligh tly tender nonfluc lUant area of induration at the a norccta l verge. No fiss ure, fistu la or abscess was present.
Flexible sigmo ido~copy co 35 c m d id not show any mass lesion, and the mucosa was normal. Recta l b iopsy was normal. Other investigations, includi ng bloodwork and stool ~tudks for bacteria and parasites, were normal. Because a small pe rianal abscess was suspected, the pat ient was treated with oral metronidazole 250 mg nd. In Ma rc h 1990, mmimal interminem drainage fro m the ..   ( 4 ). Indeed, in one report up to 30% of all cases of sma ll howe l carcino ma in C rohn's disease occurred in exclu<leJ segments (5). The incidenccofcolorectal carcinoma in Crohn's d isease is also believed to be increased fro m four L o 20 times the incidence of de novo colorecrnl carcinoma observed in the general popu la tio n (6)(7)(8). ln a comparati ve study of C rohn's disease and ulcerative coli Lis, the incidence of colorectal carcinoma in Crohn's d isease was estimated to be comparable to that in ulce rati ve coli tis pa tients wi th left. sided disease alone (6). In July 1990, the patient presented to the emergency room with fou r days of feve r and seve re perianal pa in . Perianal dra inage had increased significantly <luring the previo us one to two weeks. Examination revealed a temperat ure of 40.S'>C. Three fistulous openings were present in the perianal region with a large, L ender, horseshoeshaped, fluctuant, 6 to 7 cm ischio rectal abscess extending around the posterior rectum. Surgical drainage of the abscess was required; excisional biopsy of the pcrianal fistula revealed a mucinous ad enocarc inoma (Figures 1,2 ). Pre· operative evaluation with colonoscopy and small bowe l barium radiographs revealed a focal area of induration just inside the anal verge and narrowing of the ileocolic anastomosis; anastomotic 60 Duration of Crohn's before cancer diag!)osis (years) Presenting feature 30 Perianal fistula 22 Rec tovaglnal fistula 17 Perinea! fistulas 14 Anal stricture 27 Rectal bleeding 22 Perianal fistula 28 Perianal fistula 10 Rectal ulcer 20 Perinea! fistulas biopsy demonstrated acuLe inflammation but no mher evidence for active Crohn's disease. An abdominal-perianal resectio n was done. Patho logical examin;ition showed the prox imal and distal resection margins to be free from tumour, but extension of the a<lenocarcinoma into the muscularis propria was present. One of four regional lym.ph nodes was posi· tive fo r carcinoma.

DISCUSSION
The association of Crohn 's disease with carcinoma was initially described by W arren and Sommers in 1948 (l). T his was fo llowed by repo rts fro m G inzburg et al (2) and Kornfeld ct al (3). S ince these earlier reports, carcino ma nf the small bowel or colon has been increasingly recognized as a complication of Crohn's disease. The incidence of cancer in t his seLting is Ca n ce r has occas io n a ll y been reported in association with various chronic fistulas in C rohn's disease. It has been suggested that chronic inflammation and irrita tion plus disordered immunoregulati o n in infla mmat0ry bowel disease may predispose to car· cinogenesis. However, despite the high freq uency of perianal disease in patients with Crohn 's disease, perianal cancer is not frequently observed. A lo ng term follow-up study fro m the Mayo Clinic recorded only a single case of ca rci noma at this site (9). Similarly, a quarter of a century of ex perience at St Mark\ Hospi tal in Lo ndon tota ls o nl y two cases ( 10). lndeeJ, a review of the English literature fount! l>nl y eight well descri bed c,1scs of anorecrnl adenocarcmoma in associarion with C rohn's disease. In addition, five cases of squ,11nous cell care inoma and th rec cases of cloacogen ic cancer h ave hee n described. Table 1 summari zes the c.ises of anorecta l adenocarcinomas reported in Croh n's d isease.
As in the present patient, a ll cases had Crohn's disease of long duration prior to d iagnosis of ca ncer, with d1e average pntient age at diagnosis be111g only 41 year~. Mo~t of these patient~ haJ periana l/pe rinca l fi stu las. The pathology not infrequently sh nweJ mucinous adenocarcinllma, as mited previously, as adenocarcinomas arising m the background llf inflammatory bowel disease have a higher percentage of colloid adcnocarcinoma compared to noncolitic colorec tal cancer ( 11). The most recently reported patients had fisculous disease present for a prolonged period before diagnosis of cancer. In rhe present case, a perianal fiswla with minor in term it ten t dra inage wa s present fo r seve ral years; no olwious mass le~lll n was evident unul the patient' :-. rather Jramaric presenrarion with fever, pcrianal pa111, fistulous disease, and a large, rather acu te 1schiorectal a bscess that required su rg ical exc ision a nd drainage. Except for the deocl1 li c ana~toml1sis that reveal ed some intl a mmawry change, evidence for active Crnhn's d isease was not pre. ent elsewhere 1n the small or large bowel. Although colorecrnl carunoma may occ ur at the site of anive Crnhn\ dise,1'e, including fo.tulas, the observation in the present p,itient 1s consistent with previous reports. Hamdron (1 2) observed colorectal cancers located in normal bowel in up to one-third of cases of C rohn's disease. Greenstein et al ( 13) fun her stre se<l this ohcrvation, and in their report of rectal carcinomas in C rohn's disease, the colon was otherwise normal by endoscopic, pathol ogica l o r radiological eva luation.
Even though the association bet ween C rohn'~ disease and carcintllna 1s recognized, there is considerable d,f. ficu lty in formulating a surveill ance program. Occult hlnod sc reenmg is un-reliable, and the prescnce of surgical artefact ;md strict urc makes cndllscnpic exa minattons difficult. Rad1ograph1c evalu,nions arc difficult to interpret hecause of st ricrure, inflammatory change,, bypassed lo o p :, and fi stu las. Furthermore, for colorcctal carcinoma, even though colonoscopic hiopsies may he possi ble, the relationship between muclisal dysplasia and carcinoma -as hypothesized for the dysplas ia-cancer sequence in ulcerative colitis -is not yet fi rmly established (18).
It remains ro be determined if there are ::iny risk facto rs that can be identified in longstanding Crohn's disease, particularly with fistulous disease, that wi ll increase the likel ihood of carcinoma. Although pcrianal disease and fowlas :: uc common, the presence of carc inoma is rare. Unfo rtunately, detection of carcinoma at present, appears w be dependent upon hcighreneJ suspicion a nd performance of liberal biopsies. In this se tting, biopsy of fis-tula~ may nl1t k idea l. Moreover, the prognosis is unive rsally poor if the diagnmis 1s made, with , urvival less than tw11 years.