AndroGel (Testosterone Topical) Erythrocytosis: When It Doesn't Go Away

What the label says to do when hematocrit rises

AndroGel is a topical testosterone replacement product [7]. The polycythemia section of the AndroGel 1.62% label says to check hematocrit before starting treatment, that it would be appropriate to re-evaluate hematocrit 3 to 6 months after starting, and then annually. If hematocrit becomes elevated, the label says to stop therapy until hematocrit decreases to an acceptable concentration, and it notes that an increase in red blood cell mass may increase the risk of thromboembolic events [13]. The generic testosterone gel 1% label has the same wording [1]. Neither attaches a hematocrit value to "elevated" or "acceptable" [1][13].

How much does a testosterone gel raise hematocrit?

Across 29 placebo-controlled randomized trials covering 3,393 men, every testosterone formulation raised mean hematocrit compared with placebo. Gel raised it by 3.0% (95% CI 1.8 to 4.3) and intramuscular enanthate or cypionate by 4.0% (2.9 to 5.1). The only significant difference between formulations was intramuscular esters versus patch, so gel and injections were not statistically separated in this analysis [4].

In the 180-day phase 3 trial in the generic testosterone gel 1% label, a treatment-related category called "lab test abnormal", which covered elevated hemoglobin or hematocrit among other findings in nine patients, was reported in 6% of men on 50 mg, 5% on 75 mg and 3% on 100 mg. In the 3-year extension study, 9.3% of 162 men had a treatment-related lab abnormality, again a pooled category [1].

A retrospective single-center review of 178 hypogonadal men followed for three years found hematocrit above 50% in 12.8% of men on gels, 66.7% on injectable testosterone and 35.1% on pellets. The injection group was younger (mean 42.5 versus 54.1 years) and started with a higher hematocrit, and the study was not randomized [5].

A 2024 retrospective study of 247 men (194 on testosterone undecanoate, 18 on enanthate and 35 on gel) found that 57% of patients reached a hematocrit above 0.46, 23% above 0.5 and 5% above 0.54. Among patients whose hematocrit exceeded 0.46, 46% had their highest recorded hematocrit within the first year of treatment [8].

When the count stays high after the gel is stopped

A case series of five men on long-term testosterone who developed erythrocytosis after empagliflozin was added for type 2 diabetes is the closest reported follow-up. Their erythrocytosis resolved after testosterone was discontinued or the dose reduced, or after scheduled phlebotomy, and the report gives no time to resolution. Four of the five men carried one copy of the HFE H63D variant [2].

A 2025 narrative review of testosterone adverse-effect management, drawing on American Urological Association and Endocrine Society guidance, states that therapeutic phlebotomy is indicated when hematocrit exceeds 52% [3]. That threshold is given for hematocrit during testosterone treatment; the review does not address what to do when a count remains high after stopping the gel.

What else is associated with a high count during treatment

In the 2024 cohort, a higher baseline hematocrit was associated with reaching a hematocrit of at least 0.46, while a higher body mass index was associated with reaching at least 0.46 and at least 0.5 [8]. The Endocrine Society guideline recommends against starting testosterone therapy in men who already have an elevated hematocrit [9].

A second drug is also associated with higher counts. In a cohort of 53,971 US veterans with type 2 diabetes taking an SGLT-2 inhibitor, testosterone, or both, 1.4% developed a hematocrit above 54% within 365 days; men on both drugs had higher adjusted odds than men on testosterone alone (OR 2.49, 95% CI 1.51 to 3.59) [10]. These findings are about counts during treatment; neither addresses why a count would stay high after the gel is stopped.

Does a high count on the gel mean a higher risk of heart attack or stroke?

TRAVERSE randomized 5,246 men aged 45 to 80 with hypogonadism and existing or high cardiovascular risk to daily 1.62% testosterone gel or placebo gel [7][6]. Mean treatment was 21.7 months and mean follow-up 33 months. A major adverse cardiovascular event (cardiovascular death, nonfatal heart attack or nonfatal stroke) occurred in 7.0% of men on testosterone versus 7.3% on placebo, hazard ratio 0.96 (95% CI 0.78 to 1.17) [6]. TRAVERSE tested a gel, not injections [6].

Venous thromboembolism was numerically higher with the gel in the figures reported in the generic gel 1% label: 1.7% versus 1.2%, with pulmonary embolism 0.9% versus 0.5% [1].

Whether the elevated hematocrit itself is what causes clots is unresolved. A 2018 review found the risk of erythrocytosis with testosterone therapy well established, with short-acting injectables having the highest incidence, and described the evidence tying erythrocytosis to thromboembolic events as limited [11]. A 2015 review reported a 315% greater risk of erythrocytosis (hematocrit above 0.52) versus controls and called the link to venous thromboembolism inconclusive [12].