Abstract Background Testosterone replacement therapy (TRT) with transdermal gel is widely used in men with hypogonadism. Guidelines recommend once-daily application with monitoring of serum testosterone at peak levels. However, real-world clinical experience suggests that peak–trough fluctuations in some patients, potentially affecting hormonal stability and treatment tolerability. The impact of dose fractionation remains unclear. Objectives To compare hormonal profiles and biochemical safety parameters between once-daily and twice-daily (fractionated) testosterone gel administration in hypogonadal men receivingh the same total daily dose. Methods We retrospectively identified hypogonadal men treated with 2% testosterone gel who transitioned from once-daily to twice-daily dosing, maintaining the same total daily dose. Eligible patients had serum total testosterone (TT) measurements available at both peak (3 h post-application) and nadir (pre-dose). Hormonal and laboratory parameters under the two regimens were compared. Results Twelve patients met inclusion criteria. Fractionated dosing was associated with lower peak TT levels 9.0 (5.3–13.0) vs. 5.7 (3.6–7.1) ng/mL, paired difference − 3.6 ng/mL (95% CI: −4.9 to − 1.8) and higher nadir TT levels 1.4 (1.1–1.8) vs. 3.2 (2.6–5.4) ng/mL, paired difference 2.6 ng/mL (95% CI: 1.2 to 5.1), resulting in a reduced difference between peak-to-pre-dose difference within the observed sampling window. The proportion of patients achieving target TT levels at peak increased (33% vs. 83%). Calculated free testosterone showed a similar pattern. No clinically relevant differences were observed in luteinizing hormone, prostate-specific antigen, haematocrit, or haemoglobin. Conclusion In this real-world cohort, fractionated testosterone gel administration was associated with a higher proportion of patients achieving biochemical targets and with a reduced peak-to-pre-dose difference within the observed sampling window, without short-term biochemical safety signals. These findings should be considered hypothesis-generating.
Maltese et al. (Sat,) studied this question.