The Fix That Might Not Be One
A 45-year-old on injectable testosterone whose hematocrit has climbed to 55% — and a genuine dispute over whether the standard fix, phlebotomy, is actually the safer move or a well-intentioned one the evidence doesn't yet support.
Ray K. has worked construction framing crews for over twenty years and started injectable testosterone eighteen months ago for confirmed hypogonadism after a workplace injury sidelined him for several months and left him deconditioned and symptomatic. The treatment worked as intended — his energy and strength both recovered, and he went back to full crew duty within four months. His hematocrit, checked at his routine six-month interval, has climbed steadily since: 46% at three months, 51% at nine months, and 55% at his most recent check, now over the 54% threshold most guidelines flag for intervention.
Injectable testosterone carries the highest erythrocytosis risk of any route of administration, reported by Dobs and colleagues at 44% for injectable against roughly 15% for transdermal, because peak serum levels after each injection drive erythropoietin production more forcefully than the steadier release a gel or patch produces. The mechanism runs through two separate pathways: testosterone stimulates erythropoietin directly, and it independently suppresses hepcidin, the master regulator that normally limits iron availability for new red cell production — both push in the same direction, and neither slows down on its own once hematocrit is already elevated.
The 2018 American Urological Association and Endocrine Society guidelines both name 54% as the threshold at which therapy should be withheld or hematocrit actively lowered, and both name therapeutic phlebotomy as an accepted management option alongside dose reduction. A 2024 systematic review and a companion mechanistic paper published the same year in Endocrine Connections complicate that recommendation directly: the evidence actually supporting phlebotomy's efficacy and safety specifically for testosterone-induced erythrocytosis, as opposed to erythrocytosis from other causes, is thin and inconsistently reported across the small studies that exist. More pointedly, the mechanistic argument the same authors raise is that phlebotomy lowers tissue oxygen tension and depletes iron stores, both of which activate hypoxia-inducible factor signaling — the same pathway testosterone itself uses to drive erythropoiesis — raising a real, if unproven, possibility that repeated phlebotomy could paradoxically increase rather than decrease his actual thrombotic risk, even while it successfully lowers the number on the lab report that everyone is reacting to.
Follow-up visit, arguing over which fix actually helps
Fifty-five percent is over the 54% threshold both AUA and Endocrine Society name for intervention, and phlebotomy gets him back into a safer range immediately rather than waiting weeks for a dose reduction to bring the number down gradually. Elevated hematocrit raises blood viscosity and real clot risk — that part isn't controversial, and phlebotomy is the guideline-endorsed direct answer to it.
The threshold and the viscosity concern aren't what I'm disputing. What the 2018 guidelines don't reflect is a real 2024 review, published in Endocrine Connections, showing the evidence that phlebotomy actually helps specifically in testosterone-induced erythrocytosis is thinner than the confidence it's usually recommended with. The same authors raise a genuine mechanistic concern: phlebotomy depletes iron and lowers tissue oxygen tension, both of which activate hypoxia-inducible factor signaling — the identical pathway testosterone itself uses to drive red cell production.
Lowering the number on his lab report isn't automatically the same as lowering his actual clot risk, and it's at least plausible we'd be treating the report rather than the patient.
There's a path that doesn't require settling whose reading of the phlebotomy evidence is right. Injectable dosing is driving this through peak-level erythropoietin stimulation more than the drug itself does — switching him to transdermal, which carries roughly a third the erythrocytosis incidence in the comparative literature, or simply reducing his injected dose, addresses the actual mechanism rather than the downstream number. Recheck hematocrit in six weeks; phlebotomy stays available if it hasn't come down, but we're not reaching for it as the first move on a genuinely contested evidence base.
Agreed: transdermal formulation switch with a dose equivalent adjustment, hematocrit rechecked in six weeks, phlebotomy named explicitly as the next step if the route change doesn't bring the number down on its own.
Not agreed, and stated plainly rather than resolved: the hematologist's underlying concern that phlebotomy may not be the safe default intervention the guidelines currently present it as. The primary care physician's read of the same 2024 evidence is that it raises a real question without yet overturning the guideline recommendation, and Ray's plan was built to avoid needing an answer to that question today rather than to settle it.