In a matched retrospective analysis reported in The Lancet Oncology, Zhao et al identified and validated a 24-gene predictor of response to postoperative radiotherapy in prostate cancer. Felix Y. Feng, MD, of the University of California, San Francisco, is the corresponding author of The Lancet Oncology article.
The analysis involved data from patients from five published U.S. studies (cohort, case-cohort, and case-control studies) of patients with prostate adenocarcinoma who had radical prostatectomy with or without postoperative radiotherapy. They also had gene-expression analysis of the tumor, long-term follow-up, and complete clinicopathologic data. Postsurgical treatment was at the treating physician’s discretion.
In the training cohort (n = 196 from 1 study) and pooled validation cohort (n = 330 from 4 remaining studies), patients who had postoperative radiotherapy were matched with patients who had not for Gleason score, prostate-specific antigen level, surgical margin status, extracapsular extension, seminal vesicle invasion, lymph node invasion, and androgen-deprivation therapy. In the training cohort, a 24-gene Post-Operative Radiation Therapy Outcomes Score (PORTOS) was developed to predict response to postoperative radiotherapy. The 24-gene set included 6 genes related only to DNA-damage response, 4 genes related to both DNA-damage and radiation response, and several genes involved in immune response (including IL1B, IL7R, PTPN22, and HCLS1). The primary endpoint was development of distant metastasis.
Predictive Performance
Among 39 patients in the training cohort with a high PORTOS, the 10-year rate of distant metastasis was 5% in those who had radiotherapy (n = 20) vs 63% in those who did not (hazard ratio [HR] = 0.12, P < .0001). Among 157 patients with a low PORTOS, the 10-year rate of distant metastasis was 57% in those who had radiotherapy (n = 78) vs 31% in those who did not (HR = 2.5, P < .0001; P < .0001 for interaction).
Among 82 patients in the validation cohort with a high PORTOS, the 10-year rate was 4% in those who had radiotherapy (n = 57) vs 35% in those who did not (HR = 0.15, P = .0020). Among 248 patients with a low PORTOS, the 10-year rate was 32% in those who had radiotherapy (n = 108) vs 32% in those who did not (HR = 0.92, P = .76; P = .016 for interaction).
The conventional prognostic tools Decipher, CAPRA-S, and microarray version of the cell-cycle progression signature did not significantly predict response to radiotherapy (P > .05 for interaction for all).
The investigators concluded: “Patients with a high PORTOS who had postoperative radiotherapy were less likely to have metastasis at 10 years than those who did not have radiotherapy, suggesting that treatment with postoperative radiotherapy should be considered in this subgroup. PORTOS should be investigated further in additional independent cohorts.” ■
