Commentary on “Metabolic syndrome, levels of androgens, and changes of erectile dysfunction and quality of life impairment 1 year after radical prostatectomy”

Neuzillet et al.1 present a timely paper that aims to describe both quality of life (QoL) and erectile function (EF) changes as well as their potential preoperative predictors, specifically metabolic factors, after radical prostatectomy. The group utilizes patient data from the AndroCan trial, a prospective four-institution longitudinal cohort study of newly diagnosed patients with localized prostate cancer referred for robot-assisted radical prostatectomy. It is important to note that 1-year postoperative data were satisfactorily obtained in only 378 (28.1%) of the 1343 men who participated in the trial. The sharp drop-off was surmised by the authors to be potentially due to the nature of the EF assessment tool (the short-form International Index of Erectile Function 5-item [IIEF-5]), with men who indicated that they could not provide IIEF answers being an older cohort with worse baseline IIEF-5 scores and lower preoperative bioavailable testosterone – implying an underestimation of the IIEF outcomes as a whole.

The other main follow-up questionnaire used was the Aging Male Study (AMS) for QoL. The choice of the AMS questionnaire is interesting – the AMS scale is often used as a screening tool for testosterone deficiency rather than a global postcancer QoL measure, like the expanded prostate cancer index composite (EPIC) or the quality of life questionnaire core 30 (QLQC-30).2 This may be due to the focus on sexual function in this paper, but as such, it potentially neglects the potential impacts of postoperative urinary and bowel function on QoL and sexual outcomes.3 Another key issue is the lack of prostatectomy procedure information, specifically the nerve-sparing extent, which would have a significant impact on the outcome measures.4 Furthermore, it is unclear if any of these patients had penile rehabilitation postoperatively or used adjuncts such as phosphodiesterase type 5 (PDE5) inhibitors or other erectogenic agents, or any further cancer treatment within the year, especially as a relationship was said to have been observed between pathological aggressiveness and QoL on the global AMS score.

The authors noted a few findings that have been previously described5 – namely that there is a worsening of EF after a year (by a mean of 8 points in this case), with a worsening of QoL – assessed here by the global AMS score, chiefly driven by its sexual domain. The non-Caucasian population was said to have “significantly higher scores for erectile dysfunction” at 1 year postoperatively although looking at the data, perhaps the better description would be higher mean IIEF-5 scores (2.5; 95% confidence interval [CI]: 0.8–4.1) indicating better outcomes. Whether this was related to better baseline IIEF-5 in the non-Caucasian cohort was not made clear. Lower baseline IIEF and higher age are known to be associated with worsening IIEF outcomes at 1 year, which has been shown in other trials in radical prostatectomy4 as well as other treatments such as focal therapy.6 While the baseline IIEF was shown to be a significant EF predictor on regression modeling in this paper, the model did not take into account other confounders shown to be potentially significant such as age, which were found to be correlated with EF scores. Surprisingly, both height and body mass index (BMI) were included in the IIEF regression model, despite the obvious interaction between the two parameters. Furthermore, the authors stated “follicle-stimulating hormone (FSH), luteinizing hormone (LH), dehydroepiandrosterone (DHEA), androstenediol (D5), and DHEA-sulfate are significantly correlated with IIEF-5, AMS psychological subscore, or AMS global score” without further elaboration – which would be important, especially as this was said to be a hypothesis-generating study, justifying no correction (such as Bonferroni’s) to probability values. By a similar measure, it is unclear why an increased cholesterol level would be associated with better EF results.

While the authors have certainly analyzed a crucial amount of important data, there remain flaws in both methodology and gaps in interpretation of the results that would merit deeper discussion and analysis.

COMPETING INTERESTS

The author declares no competing interests.