Introduction Chronic gastrointestinal (GI) symptoms after radical pelvic radiotherapy are common. There is no evidence whether medical intervention helps. Most affected patients are never referred to specialists. We developed a comprehensive, peer-reviewed management algorithm for patients with new onset GI symptoms after pelvic radiotherapy. A prospective three arm randomised controlled trial was performed to test two hypotheses: (1) intervention using our algorithm provides benefit at 6 months after randomisation compared to no intervention; (2) outcomes do not differ when patients are managed by nurse or doctor. Other end points include: cost-effectiveness of intervention; effect on non-GI symptoms; outcomes after 12 months.

Methods Consenting people who had completed pelvic radiotherapy >6 months previously with persisting GI symptoms were randomised to see a GI nurse or gastroenterologist, both following our algorithm, or to receive the MacMillan booklet “Pelvic radiotherapy: possible late effects”. After 6 months patients in the booklet arm with persisting symptoms could see the gastroenterologist. Patients in the nurse arm, were transferred to the gastroenterologist if they had problems beyond the algorithm's scope. The primary end point was change in the modified Inflammatory Bowel Disease Questionnaire-bowel sub score (IBDQ-B). The trial was designed with 80% power to answer the 1st hypothesis after randomising 196 patients and the 2nd after closing the booklet arm, and randomising 22 more patients to gastroenterologist or nurse.

Results This 1st analysis includes 152 men, 44 women randomised to the three arms and followed for 6 months: booklet (n=68) vs combined treatment arms (66 nurse, 62 gastroenterologist). Median age was similar in both groups (69 years range 29–87); 25 patients had radiotherapy for GI, 30 gynaecological, 141 urological cancer. 18 (9%) withdrew/were withdrawn from the trial; 26 (38%) from the booklet group and 5 (8%) from the nurse arm crossed to the gastroenterologist. Intention to treat analysis showed a non-significant (p=0.056) improvement in IBDQ-B score of 2.8 points (95% CI 6.5 to −0.1). Planned per protocol analysis in 158 patients with complete data sets showed significant (p=0.041) improvement in IBDQ-B between treated and non-treated arms of 3.4 points (95% CIs 6.7 to 0.1).

Conclusion Medical intervention can ameliorate radiotherapy-induced GI symptoms. A 2nd analysis in December 2012 will address the other end points and the 2nd hypothesis. This study was funded by RFPB, NIHR.

Competing interests None declared.