Correlation of Biochemical Recurrence With Adverse Late Toxic Events Following Prostate Radiation Therapy (COBALT Study): An Individual Patient Data Meta-Analysis of 7 Randomized Trials.

PURPOSE: The association between late toxicity and biochemical recurrence (BCR) after prostate radiation therapy is unclear. We set out to characterize the relationship between late gastrointestinal (GI) and genitourinary (GU) toxicity and BCR among patients receiving conventionally fractionated (CF) or moderately hypofractionated (MHF) prostate radiation therapy.

METHODS AND MATERIALS: This was an individual patient data meta-analysis that identified randomized phase 3 trials of CF or MHF prostate radiation therapy in the MARCAP (Meta-Analysis of Randomized trials in Cancer of the Prostate) Consortium that had individual-level late toxicity and BCR data available. Data were provided to MARCAP by study investigators. The associations between BCR and late (>3 months after radiation therapy) grade ≥2 GI and GU toxicities were assessed using Fine-Gray subdistribution hazard models with an 18-month landmark to address immortal time bias.

RESULTS

Seven of 26 available trials met all eligibility criteria. A total of 6761 patients were included (CF: n = 4333; MHF: n = 2428). Median follow-up was 72 months (IQR, 61-94 months). BCR occurred in 17.0% of patients (1142/6732). The rate of late grade ≥2 GI toxicity was 14.3% (965/6761), although the rate of grade ≥2 GU toxicity was 15.5% (1045/6761). BCR was inversely associated with late grade ≥2 GI toxicity (subdistribution hazard ratio, 0.64; 95% CI, 0.43-0.96; P = .03). BCR was not significantly associated with late grade ≥2 GU toxicity (subdistribution hazard ratio, 1.06; 95% CI, 0.70-1.60; P = .78).

CONCLUSIONS

Late grade ≥2 GI toxicity was significantly associated with lower rates of BCR. We hypothesize that this may be related to the impact of prostatic motion during treatment, specifically anterosuperior motion of the prostate that would increase the dose to the rectum and to posterior dominant intraprostatic lesions. Late grade ≥2 GU toxicity did not appear to be associated with BCR.