Article Text

PDF

Sporadic duodenal adenoma and colorectal neoplasia
  1. A C Ford,
  2. O Rotimi,
  3. S M Everett
  1. Centre for Digestive Diseases, Leeds General Infirmary, Leeds, UK
  1. Correspondence to:
    Dr A C Ford
    Centre for Digestive Diseases, Leeds General Infirmary, Great George St, Leeds LS1 3EX, UK; alexf12399yahoo.com

Statistics from Altmetric.com

We read with great interest Murray et al’s article concerning the association between sporadic duodenal adenoma and colorectal neoplasia (Gut 2004;53:261–5).

We have also noted a lack of published literature regarding sporadic duodenal adenomas, with the exception of the two reports identified by the authors,1,2 as well as a more recent case series,3 and therefore welcome the addition of their large case series and comparison group. Although there is the potential for bias in the detection of colorectal neoplasia in their duodenal adenoma cases, as the indications for colonoscopy were somewhat different between cases and controls (with almost one third of cases of duodenal adenomas being endoscoped for investigation of anaemia or melaena), overall their data support a clinically relevant association between duodenal and colonic adenomas.

These results are corroborated by our own recent experience at Leeds General Infirmary where we have studied the natural history of duodenal adenoma. We also examined histopathology records and case notes for patients with sporadic duodenal adenomas between 1990 and 2002 (excluding familial adenomatous polyposis and hereditary non-polyposis colorectal cancer). We identified 35 cases; 16 males and 19 females (mean age 65 years). The majority of these were noted incidentally at upper gastrointestinal endoscopy or endoscopic retrograde cholangiopancreatography, with a similar range of indications for examination as those noted by Murray et al. Using the criteria stipulated by the authors, 25 (71%) of the adenomas were advanced. Of the 35 patients in our series, 11 underwent colonoscopic examination. This revealed a synchronous colonic adenoma in four patients (36%), three of which were tubulovillous adenomas and therefore advanced according to the aforementioned criteria, and the fourth a tubular adenoma. This compares with a reported rate of colonic neoplasia of 25.5% in a series of 1000 consecutive unselected patients attending the Leeds General Infirmary for colonoscopy.4 Unlike Murray et al, we did not find any colonic carcinomas in our patients. This discrepancy presumably relates to the fact that both series are relatively small and the confidence intervals are likely to be wide, much as they found.

Unfortunately, Murray et al provide no follow up data relating to the duodenal adenomas in their series. We have clinical data for all our cases, with a median duration of follow up of four years. Seventeen of the 35 adenomas were removed (six surgically, four by endoscopic mucosal resection (EMR), three snared, three during biopsy, and one by argon photocoagulation), and of these, two patients had a recurrence (one treated by EMR, the other kept under surveillance), and all were alive at the time of review (median time from diagnosis 58 months). Among the 18 patients who had no therapy for their polyp, 10 died (median length of follow up 43 months). One patient who had a tubulovillous adenoma with high grade dysplasia developed a 3 cm adenocarcinoma of the first part of the duodenum five months after initial diagnosis and underwent a Whipple’s procedure but died 12 months later from metastatic disease. The other patients died from unrelated causes.

Despite published literature that suggests duodenal polyps have no prognostic importance,5 we believe that the optimal management of duodenal adenomas is early excision and follow up. We also concur that there is an association between sporadic duodenal adenomas and colonic neoplasia, and agree entirely that all patients found to have sporadic duodenal adenomas should undergo colonoscopy. However, owing to reasons of study size and bias in the comparison groups, the magnitude of this association cannot be accurately determined. Furthermore, the reasons for this association are as yet unclear, and molecular studies are required to further investigate.

References

View Abstract

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.

Linked Articles