Article Text

Download PDFPDF

Colonic carcinoma after ureterosigmoidostomy
  1. A Huang,
  2. G A D McPherson
  1. Department of Colorectal Surgery, Wycombe Hospital, Queen Alexandra Road, High Wycombe, Buckinghamshire HP11 2TT, UK
  1. Mr A Huang, 40 York Terrace East, London NW1 4PT, UK (email:andyhuang{at}


Urinary carcinogens promote late malignant transformation of the colon after a ureterosigmoidostomy. An unusual case is presented where, despite the early removal of the latter and hence cessation of urine flow, a colonic carcinoma developed at the site of previous anastomosis. The importance of surveillance of all patients who have undergone this procedure to avoid an iatrogenic cancer is emphasised.

  • rhabdomyosarcoma
  • ureterosigmoidostomy
  • colonic carcinoma

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Colonic carcinoma arising from the site of a functioning ureterosigmoidostomy anastomosis is a recognised late complication. If the anastomosis is subsequently taken down with no further urine flow into the bowel, the risk of neoplasia is reduced. We describe a rare case where a colonic carcinoma developed coincidentally at the site of a previous ureterosigmoidostomy after a long latent period.

Case history

A 41 year old man presented with rectal bleeding without bowel habit changes. At the age of 3 he underwent cystectomy with ureterosigmoidostomy formation for a rhabdomyosarcoma. Due to recurrent pyelonephritis the ureters were divided near the sigmoid colon and implanted in an ileal conduit seven years later. Two short segments of ureters were left attached to the colon.

Physical examination was normal and a barium enema revealed a lesion in the sigmoid colon. At laparotomy a circumferential cancer was found at the exact site of previous ureteric implantation (fig 1). Histological examination revealed a moderately differentiated adenocarcinoma with two short segments of residual ureter (fig 2). Four of the 14 lymph nodes recovered contained metastatic deposits (Dukes C) and he received adjuvant chemotherapy with 5-fluorouracil and folinic acid.

Figure 1

Resected specimen demonstrating sigmoid carcinoma with two short segments of ureters (markers).

Figure 2

Photomicrograph of resected carcinoma (A = adenocarcinoma; U = ureter; haematoxylin and eosin × 30).


Colonic carcinoma developing at the site of ureteric implant was first described by Hammer in 1929.1 The incidence is 100 to 550 times that of the general population2 with an overall lifetime risk of 5%; if the diversion is performed before the age of 25 years, the risk increases to 7000-fold.3 The latency is between six to 50 years after the procedure4with the mean time at 21 years; the median age at diagnosis is 33 years.5

Urine in direct contact with colonic epithelium plays a pivotal part in the initiation of carcinogenesis at the suture line.6Stewart proposed that dietary nitrates excreted in urine come into the presence of high concentrations of secondary amines when diverted into the colon, with resultant bacterial activation of carcinogenic N-nitroso compounds.7 Constant faecal stream does not appear to be a prerequisite as carcinomas have been described arising from isolated colonic loops used as a neobladder.8 Other theories of carcinogenesis including surgical and mechanical trauma, excess concentrations of electrolytes, and chronic irritation resulting in malignant transformation have not been proved.

In the present case the ureterosigmoidostomy had been defunctioned many years previously and the subsequent development of a colonic carcinoma was likely to be a clinical coincidence. It could be argued that there was a causal link between the ureteric implantation and the bowel carcinoma, especially as the latter developed at such young age. However as there was no urine flow during this period this is unlikely.

Lifelong surveillance is recommended for all patients who undergo ureterosigmoidostomy. Starling et alsuggested that annual colonoscopy with faecal occult blood test should be started soon after ureterosigmoidostomy, with subsequent alternative urinary diversion if recurrent polyps, cancer, or dysplasia were found.9 A more complex regimen of a faecal occult blood test every three months after two years, an excretory urogram yearly after five years, and sigmoidoscopy or colonoscopy every five years has also been proposed.2

Registries of patients are often undertaken in large centres7 but the long latency to cancer development with subsequent patient movement makes tracking of these patients difficult.9 Patients and their physicians should therefore be fully informed of the risks associated with this procedure so that appropriate surveillance could be arranged. Hospital specialists should have a high index of suspicion when a patient presents with a history of urinary diversion and hence the possibility of colonic malignancy. Failure to do so would prevent the early detection of an iatrogenic bowel cancer.