Endocrine Abstracts

Introduction Some published studies suggest acromegalic patients have an increased risk of colonic tumours or pre-malignant dysplastic polyps. We assessed the prevalence of colonic polyps in Oxford acromegalic patients.

Methods Patient database identified all 80 acromegalic patients who had had a colonoscopy.

Control Patients Our data and published data on two autopsy and eight colonoscopic screening studies in asymptomatic non-acromegalic subjects were compared.

Results 114 colonoscopies were performed on 80 patients(41male, mean age 56yrs). No colonic adenocarcinomas were detected. 12(15%) patients had 19 dysplastic adenomas(size1-40mm, 74% left-sided, histology-8 tubular,6 tubulo-villous,1 tubulo-papillary,2 villous, 2 mildly adenomatous)and 11(13.8%)had metaplastic polyps. Average age for dysplastic adenomas was 60.7 plus/minus 13years and metaplastic polyps 59 plus/minus 19years. M:F was 2:1

At the first colonoscopy 21(26%) patients had polyps.10(12.5%)were dysplastic adenomas, 8(10%)were metaplastic polyps(1/21 patients had one of each)and 4 were others(inflammatory bowel/diverticular disease, mild atypia).

34 patients had follow-up colonoscopies between 1-5yrs later. 4/34 had dysplastic adenomas with reoccurrence in 2 patients(polyp size 5mm/40mm, repeat colonoscopy 1 and 24 months later and IGF-1 didn't predict reoccurrence). 8/12 patients with dysplastic polyps underwent repeat colonoscopies(between 1-3yrs). 2/12 had reoccurrences.1/12 had a dysplastic polyp on the second colonoscopy(normal first).

8/12 patients with dysplastic adenomas, 9/11 with metaplastic polyps and 38/57 with normal colonoscopies had active acromegaly. In the respective groups the IGF-1 was comparable and mean growth hormone(milli international units per litre) was 67mu/l, 65mu/l and 67mu/l at diagnosis and 9.3mu/l, 5.3mu/l and 12mu/l at colonoscopy. Adenomas were not related to duration of disease.

Conclusion We demonstrate a 15 % prevalence rate for colonic dysplastic adenomas with no colon cancers. The control data is suboptimal. Comparing our findings to published screening studies our prevalence rate is comparable to similar aged asymptomatic patients. We have not demonstrated an increased risk of colonic neoplasia in our acromegalic patients.