Ovarian Crohnʼs disease: Contiguous or metastatic?

To the Editor: Although Crohn and Yarnis1 reported two cases of regional ileitis associated fistulas to the fallopian tubes in 1958, Wlodarski and Trainer2 are credited with the first report of Crohn's disease (CD) affecting the ovary. They described granulomatous inflammation of an ovary from an ileocecal-tubo-ovarian mass. Since then, a small number of cases of resections of ovarian-associated Crohn's masses have been reported,3,–5 in which the ovarian granulomas contained foreign matter or vegetable material, suggesting direct fistulization. We report a case in which granulomas of a macroscopically abnormal ovary were identified in the presence of an intact ovarian capsule, raising the possibility that CD of the ovary may not simply occur by direct extension but may be metastatic or a true extra-intestinal manifestation of the disease. A 28-year-old white woman presented with a 1 year history of lower abdominal pain, loose unformed stools, and 9 kg of weight loss. Her past medical and surgical history was otherwise unremarkable, and she had no family history of inflammatory bowel disease. General physical and abdominal examinations were unremarkable. The initial investigations showed hemoglobin of 11.1 g/dL, white cell count of 12.1 × 109/L, and erythrocyte sedimentation rate of 49 mm/hr. Liver function test, urea, electrolytes, endomysial antibodies detection, and stool culture were all within the normal limits. Esophagogastroduodenoscopy, Helicobacter test, and biopsy of the duodenum were unremarkable. However, small bowel enema demonstrated a terminal ileal stricture with an ileo-sigmoid fistula. Rectal biopsy suggested CD. She was begun on metronidazole and continued on prednisone daily and mesalasine 1 g twice daily. She underwent 10 months of conservative management, during which time her weight increased and her diarrheal symptoms resolved. The entero-sigmoid fistula failed to heal, infliximab was not considered, and she was therefore referred for consideration for surgery. At laparotomy, a markedly thickened segment of terminal ileum was adherent to the sigmoid and the right edematous enlarged ovary (7 × 3 cm). The left ovary, fallopian tubes, and uterus appeared normal. After discussion with a gynecologic consultant who judged the right ovary to be possibly affected by CD, an en block right salpingo-oophorectomy and a standard ileocecal resection was performed. The ileo-sigmoid fistula tract was resected, and primary repair of the sigmoid colon with omental interposition was performed. Her postoperative recovery was uncomplicated, and she was discharged on the sixth postoperative day. Microscopically, sections from the small intestine confirmed CD, with patchy mucosal ulcerations, fissure ulcers, submucosal fibrosis, transmural inflammation, lymphoid aggregates, and epithelioid granulomas. The right ovary showed extensive necrotizing epithelioid granulomatous inflammation with the presence of multinucleated giant cells consistent with CD (Fig. 1). Although the granulomatous inflammation was present close to the surface of the ovary, the ovarian capsule and surface of the ovary were intact and histologically normal. The fallopian tube showed patchy lymphocytic infiltration. Stains for actinomycosis and acid-fast bacilli were negative. At 3 years follow-up, she remains symptom free on azathioprine and mesalasine. Right ovary showed extensive necrotizing epithelioid granulomatous inflammation with presence of multinucleated giant cells consistent with Crohn's disease. CD can affect any part of the intestinal tract and is associated with extra-intestinal manifestations.1 Involvement of the uterus, fallopian tubes, or ovary occurs in up to 24% of women, with 25% of this group having an adnexal mass. This frequency of adnexal involvement appears high, but lacks histologic confirmation, and fails to rule out concomitant pathology such as cyst or inflammation. It has been generally accepted that ovarian CD results by a direct extension of a fistula between the diseased bowel segments to its adjacent ovary, confirmed by the presence of material of intestinal origin in the ovary. To date, histologically confirmed ovarian CD has been reported in the English literature in only nine cases;2,–5 despite this, it ranks as the second most common cause of ovarian granulomata in Western society.5 As with our case, Wlodarski and Trainer2 reported a direct extensive adhesion of CD to the right ovary without fistulization. They reported granulomata deep in the ovarian tissue in addition to the ovarian cortex. Although they did confirm the absence of foreign material in the granulomata, they did not comment upon whether the capsule was intact and interpreted their findings as a case of direct involvement of the ovary rather than metastatic disease. Honoré3 reported a 30-year-old woman with a complex right iliac fossa inflammatory mass involving the ileum, caecum fallopian tube, and right ovary. Macroscopically, there was evidence of direct fistulization from the bowel to the tube and ovary, with ovarian abscesses. Microscopically, there were foreign-body type granulomata and foreign material of intestinal origin within the ovarian abscesses. This case obviously represents direct involvement of the ovary by fistulizing CD. Goldberg et al4 described a left iliac fossa mass involving the sigmoid colon, left fallopian tube, and ovary associated with a colon-vesical fistula. They did not comment on the pathologic findings in the resected specimen but did suggest a fistulous connection between the bowel and ovary. McCluggage and Allen5 reviewed their series of granulomatous disease of the ovary, in which they reported four patients with ovarian granulomata and CD. Three of these women (2 with bilateral involvement) had foreign body granulomata, with evidence of intestinal material in micro-abscesses within the ovary, which was adherent to the affected bowel. The fourth patient showed several small granulomata near the cortical surface of the affected ovary. No foreign material was identified, and the authors concluded that the histologic change may not be related to the CD and may be idiopathic cortical granulomata. They further described five other cases of “idiopathic” ovarian granulomata, listing similar histologic features of few cortical granulomata, which were small, well circumscribed, and without central necrosis. Our patient does not demonstrate the histologic features of either fistula-related disease because we found an intact ovarian capsule with no evidence of intestinal material within the ovary or idiopathic granulomatous disease caused by the multiplicity of the granulomata and the presence of central necrosis. We therefore suggest that our findings are consistent with a diagnosis of metastatic CD of the ovary, and this may represent the first reported case. The implications of CD affecting the ovary are twofold: first, because extra-intestinal CD may improve with control of active bowel disease, the colorectal surgeon should make every effort to salvage even a markedly abnormal ovary because function may improve after intestinal resection; second, the gynecologist must be aware that the ovary may give symptoms in women with CD. We would suggest a joint approach between colorectal surgeons and gynecologists when managing such patients and would not support oophorectomy unless the ovary is involved in a complex mass and is not salvageable.