The erythropoietic effect of androgens has been researched since the 1960s; however, it was not until the 1990s that polycythemia as a risk of testosterone therapy became more widely known.^1,2 It is currently understood that erythrocytosis secondary to testosterone therapy is multifactorial (Fig. 1),³ with testosterone’s effects on hepcidin, erythropoietin set point, and estradiol levels identified as culprits. The risk of polycythemia is now well recognized by the American Urological Association, and physicians are recommended to test for hematocrit levels before and during treatment.⁴ Despite the current guidelines, the true risk of untreated erythrocytosis in patients undergoing testosterone therapy has not been entirely quantified. In a systematic review and meta-analysis by the University of Ottawa Heart Institute, Elliott et al compiled 4 randomized control trials and found no significant association with erythrocytosis.⁵ Furthermore, they found that no association existed between testosterone therapy and any adverse cardiovascular event.

Figure 1. Proposed erythropoietic mechanisms of testosterone.

To better understand whether any cardiovascular risk exists with patients who developed polycythemia after initiating testosterone therapy, we conducted a large, population-level retrospective study.⁶ In an analysis of the TriNetX Research Network, which is a federated network of hospital electronic health record systems containing over 70 million patients, we compared patients on testosterone therapy who developed secondary polycythemia to those who did not. After adjusting for potentially confounding variables that increased cardiovascular risk and excluding patients with recent cardiac events, we found that patients who developed polycythemia while undergoing testosterone therapy were at a statistically significant increased risk of cardiac events (Fig. 2). However, the clinical significance of the absolute value of this increase (1.3%) is questionable.

Figure 2. Kaplan-Meier estimate of cardiovascular events in polycythemia patients.

Figure 3. Forest plot of hematocrit (Hct) change by testosterone formulation compared against placebo.

With a myriad of options available for patients seeking testosterone therapy, focus has shifted to the different pharmacokinetic effects that the various formulations of testosterone may have.⁷ As the erythropoietic effect of testosterone depends on serum levels and doses, both of which are affected by the modality of testosterone therapy, there has been interest in evaluating the unique adverse effect profiles of each formulation.^8,9 Although some data exist comparing formulations against one another, the vast majority of randomized control trials focus on comparing testosterone against placebo. Due to questions that remained regarding the physiological variations that result from different testosterone formulations and routes of administration, we analyzed 29 randomized control trials in a network meta-analysis.¹⁰ We specifically sought to determine the actual percent increase of hematocrit that can be expected from each testosterone therapy modality. Ultimately, we showed that all testosterone therapies should be expected to increase hematocrit, but none had a mean increase greater than 4.3% (Fig. 3). In a comparison of treatment arms, intramuscular testosterone enanthate/cypionate resulted in a significantly greater than testosterone patches, but otherwise there were no differences between formulations. These results support the current American Urological Association guidelines regarding testing patients and monitoring on an individual basis.

Based on current and past data, we would recommend careful hematocrit monitoring of patients being prescribed testosterone therapy, with special care and attention for those with preexisting cardiovascular risk factors.