PORTOS gene signature as a predictor of risk of adverse events after dose-escalated vs. lower-dose prostate radiation therapy in NRG/RTOG 0126.

375 Background: Dose-escalated radiation therapy is standard treatment for patients with prostate cancer. Dose-escalation improves cancer control but also increases the risk of treatment adverse effects. We hypothesized RNA-based tumor gene expression recapitulates normal tissue gene expression and therefore could identify patients at increased risk of adverse events after dose-escalated radiation. We specifically evaluated the 24-gene PORTOS score which characterizes response to DNA damage and radiation. Methods: PORTOS scores were calculated from biopsy samples obtained from 215 patients treated on the NRG/RTOG 0126 clinical trial that randomized patients with intermediate-risk prostate cancer between 70.2 Gy and 79.2 Gy delivered in 1.8 Gy fractions. In this trial, adverse events were categorized using RTOG criteria. Fine-Gray multivariable analysis of continuous and categorical PORTOS (tertiles) were used to calculate subdistribution hazard ratios (sHR), treating death without events as a competing risk, adjusting for age. Results: Median age was 70 years [IQR 65-74]. Fifty percent received 70.2 Gy (n=107), 50% received 79.2 Gy (n=108) and median follow up was 12.8 years. Patient and treatment characteristics were well balanced across treatment arms (all p>0.05) and across PORTOS groups (all p>0.05). Forty-five percent (n=97) of patients experienced grade 2 or higher adverse events after treatment. In patients receiving standard dose 70.2 Gy radiation, PORTOS was not associated with grade 2 or higher adverse events. However, in patients receiving dose-escalated 79.2 Gy radiation, higher PORTOS score was associated with a higher rate of grade 2 or higher adverse events (sHR = 1.12 [95% CI 1.03-1.22], p=0.01). There was a statistically significant interaction between continuous PORTOS scores and treatment arm for grade 2 or higher adverse events (p=0.01). Regarding treatment arm effects by PORTOS tertile, we observed that for patients with higher tertile PORTOS scores, dose-escalated radiation is more likely to cause grade 2 or higher adverse events compared to lower-dose radiation (sHR = 2.15 [1.04 - 4.44], p = 0.04; five-year cumulative incidence adverse events of 61% after 79.2 Gy vs. 36% after 70.2 Gy). In contrast, risk of grade 2 or higher adverse events was similar after treatment with dose-escalated vs. lower-dose radiation for patients with lower (p=0.41) or mid-tertile (p=0.78) PORTOS scores. Conclusions: HigherPORTOS scores were associated with an increased risk of adverse events after administration of dose-escalated radiation compared to standard-dose radiation. PORTOS is the first radiation sensitivity biomarker to be validated for toxicity with data from a phase III randomized trial and could be used to help personalize radiation therapy dose for patients to limit risk of treatment toxicity.