An unsafe hematocrit threshold for men receiving testosterone therapy (TT) has never been tested. This study seeks to determine whether secondary polycythemia among men receiving TT confers an increased risk of major adverse cardiovascular events (MACE) and venous thromboembolic events (VTE).

Using a multi-institutional database of 74 million patients, we identified 2 cohorts of men with low testosterone (total testosterone <350 ng/dl) who received TT and subsequently either developed polycythemia (5,887) or did not (4,2784). Polycythemia was defined as hematocrit ≥52%. As a secondary objective, we identified 2 cohorts of hypogonadal men without polycythemia, who either did (26,880) or did not (27,430) receive TT. Our primary outcome was the incidence of MACE and VTE in the first year after starting TT. We conducted a Kaplan-Meier survival analysis to assess differences in MACE and VTE survival time, and measured associations following propensity score matching.

A total of 5,842 men who received TT and developed polycythemia were matched and compared to 5,842 men who did not develop polycythemia. Men with polycythemia had a higher risk of MACE/VTE (number of outcomes: 301, 5.15%) than men who had normal hematocrit (226, 3.87%) while on TT (OR 1.35, 95% CI 1.13-1.61, p <0.001). In hypogonadal men who received testosterone, no increased risk of MACE and VTE was identified as compared to hypogonadal men naïve to TT.

Developing polycythemia while on TT is an independent risk factor for MACE and VTE in the first year of therapy. Future research on the safety of TT should include hematocrit as an independent variable.