The Effect of Route of Testosterone on Changes in Hematocrit: A Systematic Review and Bayesian Network Meta-Analysis of Randomized Trials

Abstract

ABSTRACT Introduction Testosterone therapy (TT) is known to be associated with secondary erythrocytosis. Previous studies have shown that method of administration and formulation may result in variances in hematocrit (Hct) increases. Given the increasing demand for TT and newer formulations becoming available, an updated review and meta-analysis of relative increases in hematocrit is warranted. As head-to-head trials are rare, network meta-analysis of the contemporary studies is the only way to compare hematocrit changes by testosterone type, including topical gels and patches, injectables (both short-acting and long-acting) and oral tablets. Objective We aimed to characterize hematocrit changes by testosterone type, including gel, patch, oral testosterone undecanoate (TU), intramuscular (IM) TU, and IM testosterone enanthate/cypionate (TE/C) through a review and meta-analysis of randomized control trials (RCTs) on this topic. Methods We conducted a thorough search of listed publications in Scopus, PubMed, EMBASE, Cochrane CENTRAL, and ClinicalTrials.gov. Twenty-nine placebo-controlled randomized trials (3,393 men) met inclusion criteria for analysis of mean hematocrit change after testosterone therapy. Randomized controlled trial data for the following formulations of testosterone were pooled via network meta-analysis: gel, patch, oral testosterone undecanoate, intramuscular testosterone undecanoate, and intramuscular testosterone enanthate/cypionate. Results All types of testosterone therapies result in statistically significant increases in mean hematocrit when compared with placebo. Meta-analysis revealed all formulations, including gel (3.0%, 95% CI 1.8-4.3%), oral testosterone undecanoate (4.3%, 0.7-8.0%), patch (1.4%, 0.2-2.6%), intramuscular testosterone enanthate/cypionate (4.0%, 2.9-5.1%), and intramuscular testosterone undecanoate (1.6%, 0.3-3.0%) result in statistically significant increases in mean hematocrit when compared with placebo. When comparing all formulations against one another, intramuscular testosterone cypionate/enanthate were associated with a significantly higher increase in mean hematocrit compared to patch, but no differences in hematocrit between other formulations were detected. Conclusions All types of testosterone are associated with increased hematocrit; however, the clinical concern of this increase remains questionable, warranting future studies. This is the first network meta-analysis to quantify mean hematocrit change and compare formulations, given the absence of head-to-head trials. Disclosure No