Erectile function recovery in men treated with phosphodiesterase type 5 inhibitor administration after bilateral nerve-sparing radical prostatectomy: a systematic review of placebo-controlled randomized trials with trial sequential analysis

The impact of phosphodiesterase type 5 inhibitor (PDE5I) treatment modality (on-demand vs. daily), PDE5I half-life and time from surgery to PDE5I prescription on the achievement of drug-assisted erectile function (EF) recovery is uncertain. We systematically reviewed published randomized clinical trials (RCTs). We performed meta-analyses of data on 2317 men treated with PDE5Is after nerve-sparing radical prostatectomy (NSRP). A PubMed and SCOPUS search was performed for trials published from 1 January 1969 to 30 June 2016. PDE5Is are effective in achieving drug-assisted recovery of erectile function (EF). From a statistical standpoint, these studies were subjected to Trial Sequential Analysis to determine whether the pooled data were adequately powered to verify the study outcomes. On-demand treatment with PDE5Is was significantly better than daily treatment in recovering drug-assisted EF. This effect was maintained even when the drugs were stratified according with half-life. Although not based on head-to-head trials, Avanafil used on-demand was the most effective PDE5I in recovering drug-assisted EF. Whereas tadalafil was equally effective when used both on-demand and daily, vardenafil significantly improved drug-assisted EF recovery only when used on-demand. The start of PDE5I treatment six months or more after surgery compared to treatment started earlier did not negatively affect the rate of drug-assisted EF recovery or the possibility to have successful intercourse based on the Sexual Encounter Profile question-3 (SEP-3). Current trials do not support the hypothesis that PDE5I use recovers drug-unassisted EF, although chronic low-dose tadalafil administration may help to preserve erectile tissue integrity. Potential shortcomings in the trials design may partially explain these disappointing results and several questions concerning the recovery of drug-unassisted EF remain unanswered. Thus, there is a need for well-designed new RCTs requiring changes in the timing of PDE5I administration as well as in the dose and the treatment duration.