Dosimetric Predictors of Toxicity and Quality of Life Following Single Fraction High Dose-Rate Prostate Brachytherapy.

Abstract

PURPOSE/OBJECTIVE(S)\nLittle is known about optimal dose constraints when high dose rate (HDR) brachytherapy is used as monotherapy for prostate cancer. Most dose constraints for HDR monotherapy were extrapolated from predictors of toxicity when HDR brachytherapy was combined with external beam radiotherapy as a boost. We sought to determine clinical and dosimetric predictors of toxicity and health related quality of life (HRQOL) in men treated with HDR brachytherapy as monotherapy.\n\n\nMATERIALS/METHODS\nEligible patients were treated with single fraction HDR brachytherapy as monotherapy on two prospective clinical trials at a single institution. Patients in the first trial (HDR-mono) received a single fraction of 19 Gy without a dominant intraprostatic lesion (DIL) boost, and patients from the second trial (MARS) received 19 Gy in a single fraction with an MRI-guided DIL boost to ≥ 23 Gy. ADT was not used. Univariable and multivariable logistic regression was used to evaluate clinical (age, IPSS score, prostate volume, alpha blocker use at baseline, and receipt of a DIL boost) and dosimetric (prostate V100, V150, V200, D90, Urethral Dmax and D10, and Rectal Dmax and V80) predictors of CTCAE v4 acute/late toxicity and HRQOL changes measured with expanded prostate index composite (EPIC). Three classifications of late minimally clinically important changes (MCICs) were evaluated: small (> 0.5 standard deviation [SD]), moderate (> 1.0 SD) and severe (> 2.0 SD) declines compared to baseline.\n\n\nRESULTS\n147 patients were included (87 from HDR-mono, 60 from MARS). Median follow-up was 62.8 months. Only increasing prostate size predicted acute GU toxicity ≥ grade 2 (OR 1.05, 95% CI 1.01-1.09, P\u202f=\u202f0.021). On multivariable regression, predictors of late GU toxicity ≥ grade 2 were not receiving a DIL boost (OR 3.78, 95% CI 1.88-7.83, P < 0.001) and higher baseline IPSS score (OR 1.89, 95% CI 1.89-3.18, P\u202f=\u202f0.015). Rectal and urethral dose constraints were not associated with late GU/GI toxicity ≥ grade 2. Contrary to our hypothesis, small (OR 0.91, 95% CI 0.81-0.98, P\u202f=\u202f0.032) and moderate (OR 0.91, 95% CI 0.80-0.98, P\u202f=\u202f0.037) urinary MCICs were less frequent in those with higher urethral Dmax. No impact of urethra D10 on urinary MCICs was seen. Rectal Dmax and V80 did not predict bowel MCIC changes at any threshold.\n\n\nCONCLUSION\nWe were unable to identify dose constraints that were predictive of late toxicity or relevant HRQOL changes in single fraction HDR brachytherapy. This suggests dose parameters used in these trials were safe, and that minor variability in plan dosimetry likely has little clinical significance. While increasing urethral Dmax paradoxically predicted for less frequent urinary MCICs, we do not believe this finding is clinically relevant. Further research exploring optimal dose constraints to be used in HDR brachytherapy as monotherapy for prostate cancer is warranted.