Contributors to the substantial variation in on-treatment testosterone levels in men receiving transdermal testosterone gels in randomized trials.

There is substantial inter-individual variability in serum testosterone levels in hypogonadal men treated with testosterone gels. We aimed to elucidate participant-level factors that contribute to inter-individual variability in testosterone levels during testosterone therapy. An exploratory aim was to determine whether polymorphisms in genes encoding testosterone-metabolizing enzymes could explain the variation in on-treatment testosterone concentrations in men who were randomized to testosterone arm in TOM Trial. We used data from three randomized trials that used 1% transdermal testosterone gels and had testosterone levels measured 2-4 weeks after randomization for dose adjustment: Testosterone in Older Men with Mobility Limitation (TOM), Effects of Testosterone on Pain Perception (TAP), and Effects of Testosterone on Atherosclerosis Progression (TEAAM). Forty-seven percent, 38%, and 9% of participants in TAP, TEAAM, and TOM trials, respectively, failed to raise testosterone levels >400 ng/dL; 6, 8, and 30% of participants had on-treatment testosterone levels >1000 ng/dL. Even after dose adjustment, there was substantial variation in on-treatment levels at subsequent study visits. Baseline characteristics (age, height, weight, baseline testosterone, SHBG, hematocrit, and creatinine) accounted for only a small fraction of the variance (<8%). Polymorphisms in SHBG and AKR1C3 genes were suggestively associated with on-treatment testosterone levels. To conclude, baseline participant characteristics account for only a small fraction of the variance in on-treatment testosterone levels investigated. Multiple dose titrations are needed to maintain on-treatment testosterone levels in the target range. The role of SHBG and AKR3C1 polymorphisms as contributors to variations in on-treatment testosterone levels should be investigated.