Pathological complete response following pre-operative chemoradiotherapy in rectal cancer: analysis of phase II/III trials

Abstract

Pathological complete response (pCR) has been used as a marker for the efficacy of pre-operative chemoradiotherapy (CRT) schedules in rectal cancer. To date there have been no randomized trials comparing CRT regimens in rectal cancer. Prospective phase II and CRT arms of randomized trials reported up to January 2004 were included, providing they defined the following minimum variables: drugs employed during CRT, radiotherapy dose and pCR rate. Multivariate analysis was used to examine the relationship of these variables on the pCR rate. In addition, the use of neoadjuvant chemotherapy, the type of publication (peer reviewed vs meeting abstract) and whether the tumours were stated to be unresectable/clinically fixed or to have threatened circumferential margins were investigated. The method of analysis was weighted linear modelling of the pCR rate which was normalized by the arcsine transformation. Phase II and phase III trials were identified including a total of 3157 patients. On multivariate analysis only the use of continuous infusion 5FU (p = 0.01), the use of a second drug (p = 0.001) and radiation dose (p = 0.02) were associated with higher rates of pCR. The use of a two drug regimen, the mode of delivery of 5FU and the radiation dose appear to be related to the incidence of pCR following CRT for rectal cancer. These results may generate hypotheses for future randomized trials. Important factors not considered in this analysis include the variability in pathological examination and in the time interval between CRT and surgery. In addition, the toxicity of the CRT regimens requires further investigation.