In search of a dose-response relationship with radiotherapy in the management of recurrent rectal carcinoma in the pelvis: A systematic review

Purpose: A systematic review of the literature was undertaken to address the question: What is the most effective dose fractionation schedule for the relief of symptoms in patients with pelvic recurrence from rectal or colorectal carcinoma? Methods and Materials: Cancerlit/Medline-computerized databases were searched between the years 1966-1996. Studies that explored the response to radiotherapy in patients with pelvic recurrence from rectal/rectosigmoid carcinoma were included. Factors that may contribute to differences in results were postulated in advance and the variations encountered between articles were presented. Articles with data applicable to recurrent disease only were included in the primary analysis. The effect of including articles that reported outcomes of recurrences with unresectable primaries and residual disease was presented as a sensitivity analysis. Results: Only retrospective series (level V evidence) were available. The many sources of potential bias inherent in retrospective analyses make the data suitable for hypothesis generation only. Comparison of response was made between lower vs. higher doses, using 45-50 Gy as the dividing dose, base on the primary analysis. There were no significant differences observable in terms of initial response and the proportion maintaining a response at 6 months, within the range of doses employed. When data from articles that reported outcomes of recurrent disease with primary untreated cancers and postoperative residual disease were included, there was a suggestion for a more favorable response with higher doses. This requires cautious interpretation within the methodological limitations of the data. Conclusion: The optimal dose fractionation schedule for the palliation of pelvic recurrence from rectal carcinoma remains undefined. Well-designed randomized studies, with study arms that are sufficiently diverse biologically to allow the detection of a dose-response relationship if one existed, equipped with suitable symptom control end points, are necessary to provide a clinically relevant answer.