On the practice of therapy decision in locally limited prostate cancer: surgery vs. radiation-who benefits? : Allocation and results of a monocentric, cumulative long-term study

Abstract

The goal of this two-armed observational study was to map the clinical therapy effectiveness of radical prostatovesiculectomy (RPVE) and external beam radiation therapy (EBRT) in locally limited prostate cancer (PCA) in direct comparison over 20years under clinical conditions. Retrospectively, the various variables and predictors for the individual therapy decision were identified, and the preference was to compared with studies on survival and recurrence characteristics. The presentation of toxicity was not the focus of this work. In all, 743patients from a single center were enrolled according to biopsy/staging chronologically in the sequence of the initial consultation after clarification and informed consent: 494patients were in the RPVE arm and 249patients in the EBRT arm. We used retrospective data analysis with univariate and multivariate comparisons in the alternative therapy arms. Multivariate logical regression models were developed to objectify the allocation process. Univariate processing of survival analyses, the comparison of tumor- and comorbidity-specific mortality rates was co-founded. Predictive variables for RPVE vs. EBRT therapy decision are significantly age, Gleason score, D'Amico index, Charlson index, prostate-specific antigen (PSA), and prostate volume. There was no significance level for the biopsy score. The age gap was in the median 67(RPVE) and 73(EBRT) years. Overall survival (n = 734, 20years, all risks) in the RPVE arm was 56.8% (95% confidence interval [CI] 45.1-67.0%) and in the EBRT arm 19.2% (95%CI 9.2-31.8%). Comorbid risk was highly significantly (p < 0.0001) different (27.1% [95%CI 18.0-36.1%] in the RPVE arm, and 60.4% [95%CI 47.3-73.5%] in the EBRT arm). The risk of tumor-specific death at 16.2% (95%CI 8.1-24.4%) after RPVE and 20.5% (95%CI 11.7-29.3%) after EBRT was not significantly different (p = 0.2122, overlapping 95%CI). After stratification, aclear advantage can be demonstrated for the high-risk tumors after allocation to the RPVE arm. The complexity of the predictive variables of the PCA further complicates the individual therapy decision. According to our data, the higher D'Amico score, the rather low Charlson index, ahigh Gleason score and ahigher organ volume speak for avalid therapy for RPVE.