Correlations between dose–surface histograms and the incidence of long-term rectal bleeding following conformal or conventional radiotherapy treatment of prostate cancer

Abstract

Background and Purpose: In a randomized trial, the incidence of rectal bleeding among patients treated for prostate cancer using conformal radiotherapy was significantly lower ( p = 0.002) than that among those treated conventionally. Here the relationship between rectal dose distributions and incidences of bleeding is assessed. Methods and Materials: Rectal dose–surface histograms (DSHs) have been calculated for 79 trial patients. The relationship between the DSHs and incidences of Grade 1–3 bleeding has been explored using both semiempiric and biologic (parallel) model-based approaches. Results: Semiempiric analysis of the trial data suggests that it is more useful to work with DSH fractional surface areas multiplied by outlined rectal lengths than with either raw DSH fractional areas or fractional areas multiplied by absolute total outlined rectal surface area. Fitting the parallel model to length-multiplied rectal DSHs and complication data reveals the existence of a significant volume effect, the rate of Grade 1–3 bleeding falling by 1.1% (95% confidence interval [0.04, 2.2]%) for each 1% decrease in the fraction of rectal wall (outlined over an 11-cm length) receiving a dose of more than 57 Gy. Conclusion: The existence of this volume effect suggests that dose escalation can be achieved using conformal techniques, although the extent to which doses may be safely escalated cannot be reliably estimated from the trial data.