The 85-Year Error: How One Patient's Data Made Generations of Men Fear Testosterone

In 1941, a surgeon at the University of Chicago injected testosterone into three men with metastatic prostate cancer and measured their acid phosphatase levels. In two of the three, the marker rose. He then castrated them and the marker fell. From this — two data points moving in a direction, in a single intact patient observed for fourteen days — Charles Huggins concluded that testosterone feeds prostate cancer and castration starves it.

He won the Nobel Prize in 1966. The castration half was correct. The testosterone half was not. But by then it didn't matter. The idea that testosterone causes prostate cancer had calcified into medical orthodoxy, embedding itself in guidelines, consent forms, and clinical reflexes for the next eight decades.

This article is about what happened when someone actually went back and read the original paper.

The Foundation: One Patient, Fourteen Days

Huggins and Hodges published their landmark paper in Cancer Research in 1941. The study involved three men with metastatic prostate cancer who received testosterone propionate injections. The full picture:

Patient 2 had been previously castrated — he was receiving exogenous testosterone to replace what surgery had removed. His acid phosphatase response tells us nothing about whether testosterone initiates or promotes cancer in an intact man. Patient 3's data was not reported in a way that supported the conclusion. The entire thesis that testosterone promotes prostate cancer rested on Patient 1, observed for two weeks, using a surrogate marker (acid phosphatase) that has since been abandoned.

Abraham Morgentaler identified this in 2006. He went back to the original paper — something apparently no one had done in 65 years — and found that the emperor's evidence was, at best, a preliminary observation. At worst, it was a single anecdote elevated to universal truth by a Nobel Prize.

The Saturation Model

Morgentaler and Traish formalized the saturation model in 2006, with updated evidence in 2009. The core insight: androgen receptors in prostate tissue have a finite binding capacity. Once those receptors are saturated — which occurs at a serum testosterone of approximately 250 ng/dL — additional testosterone has no further effect on prostate growth or cancer behavior.

Below saturation, testosterone is rate-limiting. Castration works for advanced cancer because it removes testosterone from a range where receptors are not saturated. Above saturation, exogenous testosterone is adding fuel to a fire that's already burning as hot as it can.

The mechanism was confirmed at the tissue level. Page et al. (2006) and Mostaghel et al. (2016) showed that when you castrate men and then add back exogenous testosterone, something unexpected happens to intraprostatic hormone levels:

This is the mechanism behind saturation. The prostate has its own androgen economy, partly decoupled from serum testosterone. It synthesizes DHT locally from adrenal precursors — dehydroepiandrosterone and androstenedione — via intracrine pathways. Flooding the serum with more testosterone does not proportionally increase what reaches the androgen receptor in prostate tissue.

The Kim Critique and Its Resolution

In 2020, Kim and colleagues published a critique of the saturation model in European Urology Focus, raising four objections: that Morgentaler cited studies out of context, that Ho et al.'s methodology was flawed, that Wright's Figure 2B contradicted their claims, and that Bhasin's data was incomplete. Morgentaler and Traish responded point-by-point, rebutting all four. Their conclusion: the saturation model is "an accurate framework for understanding the relationship between androgens and prostate cancer," not a hypothesis.

What is notable about the Kim critique is not that it was made — scientific models should be challenged — but that it was the strongest formal challenge published against the saturation model, and it was refuted using the challengers' own cited sources.

The Evidence Reversal: Low Testosterone Is the Danger

If testosterone promoted prostate cancer, we would expect men with higher testosterone to have worse cancers. The data shows the opposite.

The PSA Blind Spot

Fattahi et al. proposed a mechanism for this paradox: low testosterone suppresses PSA production. PSA is the primary screening tool for prostate cancer. If testosterone is low, PSA stays deceptively low even as cancer grows — leading to delayed diagnosis and worse stage at detection. The cancer isn't more aggressive because testosterone is low; it's diagnosed later because the alarm system is muted.

This has a practical implication: multiparametric MRI, unlike PSA, is not affected by testosterone level. For hypogonadal men, MRI-based screening may catch cancers that PSA misses. This is not yet reflected in guidelines.

What 11,161 Men Actually Show

The question of whether TRT causes prostate cancer has now been tested across multiple meta-analyses with aggregate samples in the tens of thousands. The results are unanimous.

The TRAVERSE trial — the largest testosterone safety RCT ever conducted (5,246 men, mean follow-up 33 months) — reported 12 prostate cancer events in the testosterone group versus 11 in placebo. No difference. The BSSM published the first formal professional society consensus statement in 2026 (World Journal of Men's Health): "No compelling evidence linking TRT to initiation or promotion of prostate cancer."

This is not ambiguity. This is not "more research needed." Four meta-analyses, the largest RCT, and a professional consensus all say the same thing: TRT does not cause prostate cancer.

After Prostate Cancer: The Data That Shouldn't Exist

If you were trained in the Huggins paradigm, the following studies should not exist. They do.

The trajectory is clear. Applewhite et al. wrote in The Journal of Sexual Medicine (2025) that TRT on active surveillance represents "a paradigm shift." Le Guevelou et al. asked in Andrology (2025): "Time to take the leap?" Saffati and Khera (Baylor) published a comprehensive review endorsing the shift. The conversation is no longer whether TRT is safe after prostate cancer treatment — it is how quickly practice can catch up with evidence.

The Irony: Using Testosterone to Treat Prostate Cancer

The most dramatic reversal of Huggins' legacy is bipolar androgen therapy (BAT) — the deliberate administration of supraphysiologic testosterone to men with castration-resistant prostate cancer. Not as palliation. As treatment.

The mechanism is counterintuitive but mechanistically sound. In CRPC, cancer cells have adapted to an androgen-depleted environment by overexpressing androgen receptors. When these hyper-sensitized cells are suddenly flooded with testosterone, the excess AR accumulation becomes toxic. Supraphysiologic testosterone forces topoisomerase IIβ (TOP2b) to create DNA double-strand breaks, triggering apoptosis. The cancer's adaptation becomes its vulnerability.

The TRANSFORMER finding is the most clinically important: BAT alone didn't outperform enzalutamide, but the sequence mattered enormously. Men who received BAT first and then switched to enzalutamide had a combined PFS2 of 28.2 months, versus 19.6 months for those who went straight to enzalutamide. BAT appears to resensitize tumors to subsequent anti-androgen therapy — a reset button for drug resistance.

TP53 mutations and homologous recombination deficiency (HRD) are enriched in deep BAT responders, pointing toward biomarker-driven patient selection. The BAT + olaparib combination exploits this directly: BAT suppresses HRR gene expression, creating synthetic lethality with PARP inhibitors even in tumors without inherent HRR mutations. This is not brute-force therapy — it is mechanistically targeted.

The Cost of Androgen Deprivation

While the fear of testosterone has been guiding men away from TRT, the treatment built on Huggins' insight — androgen deprivation therapy — has been inflicting its own damage. ADT's metabolic side effects are severe, quantifiable, and connect directly to the feedback loops I mapped in The Metabolic Trap.

ADT doesn't just suppress testosterone — it activates every loop in the metabolic trap simultaneously. Insulin resistance, fat gain, lean mass loss, dyslipidemia. The HOMA-IR ratio reaches 17.0 in ADT patients versus 6.0 in controls. Men treated for prostate cancer are being metabolically damaged by the treatment that was supposed to save them, based on a thesis derived from one patient.

Not All ADT Is Equal

The HERO trial showed that relugolix — a GnRH antagonist — produced 54% fewer major adverse cardiovascular events than leuprolide (a GnRH agonist). In the cardiovascular history subgroup: 3.6% vs 17.8%. This was a signal, not a fluke.

The REVELUTION trial (JAMA Cardiology, SUO 2025) provided the biological mechanism: leuprolide caused significant coronary artery plaque progression — specifically non-calcified plaque volume — while relugolix did not. Three MACE events (9.7%) occurred in the leuprolide arm versus zero in the relugolix arm. REPLACE-CV (NCT05605964) is now specifically powered to confirm this in a larger sample.

And testosterone recovery after ADT cessation is not guaranteed. Amorim et al. (AUA 2025) published a meta-analysis showing that while 75.8% of men recover testosterone overall, only 41.8% of GnRH agonist patients do — using a generous threshold of just 240 ng/dL. ADT duration did not predict recovery. The damage may be to the hypothalamic-pituitary axis itself — connecting this to the broader landscape of drug-induced hypogonadism I have previously mapped.

Where the Evidence Stops

The reversal is real but not complete. The honest accounting of what remains unknown:

The Timeline of an Error

What This Means

The prostate cancer scare has shaped clinical practice for 85 years. Men with symptomatic hypogonadism have been denied testosterone replacement because of a fear derived from one patient's fourteen-day observation. Clinicians have withheld treatment that could have improved quality of life, metabolic health, and possibly even cancer outcomes — all to avoid a risk that four meta-analyses and over 11,000 randomized patients say does not exist.

The correction is happening, but slowly. The BSSM consensus is the first professional society to formally state the obvious. NCT06733350 will provide the first prospective randomized evidence for TRT on active surveillance. BAT trials are rewriting the treatment landscape for castration-resistant disease.

But the gap between evidence and practice remains wide. Many urologists still reflexively contraindicate TRT in any man with a prostate cancer history. Many consent forms still list prostate cancer as a risk. The inertia of an 85-year error does not reverse overnight.

What Huggins got right — that castration helps in advanced disease — was genuinely important and saved lives. What he got wrong — that testosterone causes prostate cancer — was a single observation inflated beyond its evidence. The Nobel Prize sealed it. Sixty-five years of uncritical repetition embedded it. And now, study by study, the record is being corrected.

The irony is complete: testosterone, the hormone feared to cause prostate cancer, is now being used as a weapon against it. Huggins would not have predicted that. But then, he only had one patient.