Prevalence and predictive factors of testosterone-induced erythrocytosis: a retrospective single center study.

Aim: This study analyzes the prevalence and predictive factors of testosterone-induced erythrocytosis (TIE) in patients receiving testosterone replacement therapy (TRT).

Methods: Retrospective single-center observational study.

Results: 247 patients were included; median age was 47.0 years (interquartile range (IQR) 32-60) and median follow-up years 2.9 (1.0-5.5). The most common indication for TRT was central hypogonadism (51%) followed by primary hypogonadism (26%). TRT was carried out with testosterone undecanoate (TU) n=194, testosterone enanthate (TE) n=18 and testosterone gel (n=35). Compared to baseline, hematocrit (HCT) values at last follow-up (LFU) increased significantly by +0.04 (95% confidence interval (CI) [0.027, 0.050], p=<0.0001) in all patients (n=92) and +0.06 (95%CI [0.031, 0.057], p<0.0001) in the TU group (n=71). 57% of the patients reached an HCT value>0.46, 23% >0.5 and 5%>0.54. 46% of the patients who have reached an HCT value >0.46 have had their highest HCT measurement within the first year of TRT application. Logistic regression analysis indicated that body mass index (BMI) was significantly associated with the development of an HCT ≥0.5 (p=0.013) and HCT ≥0.46 (p=0.008). There was an association between the baseline HCT measurement and the outcome of a HCT measurement ≥0.46 (p=0.025), patients with high starting values were more likely to develop TIE.

Conclusions: TIE appears to be frequent and does not only present within the first year of therapy which indicates a close follow-up of laboratory values within the first year followed by annual controls. Baseline BMI and baseline HCT measurement should be considered in risk stratification of TIE development.