Efficacy characteristics of different therapeutic modalities for locally advanced prostate cancer: a Bayesian network meta-analysis of randomized controlled trials.

Abstract

Though previous studies have investigated the efficacy characteristics of several different therapeutic modalities for locally advanced prostate cancer (LAPCa) patients, the available results remained unestablished. Therefore, the aim of this meta-analysis was conducted to clarify such differences. The online PubMed, EMBASE and Web of Science were comprehensively searched for relevant studies published before September 1st, 2017, and eventually eleven relevant studies met the inclusion criteria. The hazard odds ratios (HRs) with 95% credible interval (CI) were utilized to evaluate the efficacy characteristics of several different therapeutic modalities for LAPCa patients by Markov chain Monte Carlo methods. Five different therapeutic modalities were ultimately enrolled to shed light on the efficacy characteristics for LAPCa patients and seven different clinical outcomes were finally analyzed in this study. The cumulative rank probability of overall survival (OS) or cancer-specific survival (CSS) from best to worst was radiotherapy (RT) + orchiectomy, RT + long-term androgen deprivation therapy (LTADT), RT + short-term androgen deprivation therapy (STADT), LTADT and RT; RT + LTADT, RT + orchiectomy, RT + STADT, LTADT and RT, respectively. Meanwhile, in the terms of progression-free survival (PFS), biochemical failure rate (BFR), disease-free survival (DFS), local progression rate (LPR) and metastasis rate (MR), RT + LTADT as well as RT + STADT had a higher, whereas RT alone or LTADT had a relatively lower treatment effect. All in all, our results indicated that RT + LTADT or RT + orchiectomy was among the best two therapeutic regimens in the prognostic aspects of the patients with LAPCa. Furthermore, in consideration of reducing invasive treatment of eligible patients, RT + LTADT could yield better survival benefit of LAPCa patients, compared with others. In addition, the results of our analysis might provide a reference in the clinical selection. Larger sample sizes of strictly designed randomised controlled trials (RCTs) were wanted to validate our findings.