For 80 years the dominant story has been that testosterone fuels prostate cancer. A new 2026 study from MD Anderson published in the Journal of Urology pushes the narrative the other way — and adds a careful caveat that matters for any man on active surveillance.
In 924 men with localized prostate cancer monitored without immediate treatment, those with baseline total testosterone at or below 300 ng/dL had 1.61 times the hazard of progressing to grade group 3 or higher disease compared with men whose testosterone was higher. The signal survived adjustment for age, race, PSA, and prostate volume.
Key Takeaways
- 924 men in active surveillance at MD Anderson, 2001–2024
- Median follow-up: 46.1 months
- Low baseline T (≤300 ng/dL) → 1.61x hazard of progressing to grade group 3+ (HR 1.61, 95% CI 1.03–2.51, P=.04)
- Low T was not associated with progression to grade group 2
- Lead author: Cesar E. Lawen, MD; senior author: Justin R. Gregg, MD (MD Anderson)
- Published February 2026, Journal of Urology (DOI 10.1097/JU.0000000000004986)
- Association, not causation — and it does not test TRT directly
What the Study Actually Did
The MD Anderson team pulled clinical and pathological data on 924 men enrolled in their active surveillance (AS) program between 2001 and 2024. All men had localized prostate cancer that did not require immediate treatment under standard criteria.
The exposure was baseline total testosterone, dichotomized at the American Urological Association deficiency cutoff of 300 ng/dL. The primary outcomes were progression to grade group 2 or higher and progression to grade group 3 or higher — the cancer-aggressiveness markers that typically trigger a move from surveillance to active treatment.
Cohort Snapshot
| Parameter |
Detail |
| Center |
MD Anderson Cancer Center |
| Sample size |
924 men on active surveillance |
| Enrollment window |
2001–2024 |
| Median follow-up |
46.1 months |
| Exposure |
Baseline total testosterone (≤300 vs >300 ng/dL) |
| Primary outcome 1 |
Progression to grade group ≥2 |
| Primary outcome 2 |
Progression to grade group ≥3 |
| Adjustments |
Age, race, PSA, prostate volume |
| Lead author |
Cesar E. Lawen, MD |
| Senior author |
Justin R. Gregg, MD |
| Publication |
Journal of Urology, Feb 2026 |
The retrospective design and single-center cohort are limitations — but the cohort size and the use of a clinically meaningful cutoff make the result harder to wave off than older signals from smaller studies.
What They Found
The headline number: men with baseline testosterone at or below 300 ng/dL had 61% higher hazard of progressing to grade group 3 or higher (HR 1.61, 95% CI 1.03–2.51, P=.04) compared with men whose testosterone was above 300 ng/dL.
The finding was specific to high-grade progression. Low T was not associated with progression to grade group 2, which represents a milder change in cancer biology.
| Outcome |
Hazard Ratio (Low vs Normal T) |
95% CI |
P value |
| Progression to grade group ≥2 |
Not significant |
— |
NS |
| Progression to grade group ≥3 |
1.61 |
1.03–2.51 |
.04 |
That selectivity matters. Grade group 3+ is where biology shifts toward more aggressive disease. The fact that the signal only appears at the high-grade endpoint suggests baseline androgen environment may interact with the tumor's progression to clinically meaningful disease, not the early bumpiness of grade group reclassification.
Why Low Testosterone Might Track With Aggressive Cancer
The authors are careful: this is association data, not causation. There are three plausible explanations, and the truth is probably a blend.
Explanation 1: Metabolic health is the upstream driver. Low testosterone, obesity, insulin resistance, and chronic inflammation cluster together. The same metabolic dysfunction that suppresses the HPG axis also creates an inflammatory tumor microenvironment that selects for more aggressive cancer cell populations. In this view, low T is a marker of a hostile metabolic state — not a direct cancer driver. The recent NHANES analysis on diabetes and low T showed how strongly metabolic health predicts low testosterone in American men, and the same metabolic factors are well-documented prostate cancer aggressors.
Explanation 2: The saturation model, inverted. The classical Huggins-Hodges model says reducing testosterone shrinks prostate cancer. Morgentaler's saturation hypothesis updated this: prostate androgen receptors are fully saturated at low testosterone levels (~250 ng/dL), so adding more T above that threshold doesn't drive more growth. A possible corollary: in some tumors, a chronically low-androgen environment may select for cell populations that have adapted to grow without much androgen — a precursor pattern to castration-resistant biology. This is speculative but mechanistically plausible.
Explanation 3: Confounding. Men with low T tend to be older, sicker, and more inflamed. Even after adjusting for age, PSA, race, and prostate volume, residual confounding from comorbidities, BMI, and lifestyle factors not captured in the dataset could account for some of the signal.
The honest answer is we don't yet know which explanation dominates. The clinical takeaway is the same regardless: baseline testosterone may be a useful risk-stratification marker on top of standard active surveillance variables.
What This Does NOT Say
This is where most coverage gets sloppy. Three things this study does not do:
It does not test TRT. The exposure is baseline endogenous testosterone — what the men were producing on their own. None of the men in the analysis were on testosterone replacement therapy at baseline. The study cannot say whether raising testosterone with TRT in a hypogonadal man on active surveillance would help, hurt, or do nothing.
It does not say low T causes prostate cancer. Earlier evidence — including the TRAVERSE trial — found no increased prostate cancer incidence with TRT. The Lawen study is about progression of already-diagnosed cancer, not initial diagnosis.
It does not change the 80-year prostate cancer myth picture for general TRT decisions. The Huggins-era assumption that more T equals more cancer growth still does not hold up in healthy men or men post-prostatectomy in whom TRT has been studied. The Lawen finding is a narrow, specific signal in one population: men with active, untreated, localized cancer.
How This Should Change Your Thinking
If you don't have prostate cancer, this study is mostly informational. The mainstream evidence base on TRT and prostate risk has not shifted — it remains that TRT is safe in selected hypogonadal men with appropriate monitoring.
If you have a prostate cancer diagnosis or are on active surveillance, three changes are worth discussing with your urologist:
- Add baseline testosterone to your AS workup. It's a cheap, standard test. If your total T is at or below 300 ng/dL, ask your urologist whether your monitoring cadence should be tightened.
- Treat metabolic health as a cancer modifier, not just a TRT issue. If you're carrying excess body fat, pre-diabetic, or inflamed, that's an upstream driver of both low T and aggressive cancer biology. A 12-week metabolic reset before any TRT decision is rarely a wrong move.
- Don't pursue TRT through a generic telehealth clinic if you're on active surveillance. This is a urologist conversation, not a 15-minute video visit. Look for clinics that coordinate with your urologic oncologist. See how to choose a TRT clinic for the criteria that matter, and TRT clinic red flags for what disqualifies a clinic from this conversation.
How This Fits the FDA's 2026 TRT Push
This study lands in the same window as the FDA's potential TRT label expansion for idiopathic hypogonadism. The two stories pull in opposite directions for clinicians.
The FDA move would broaden TRT access for men whose low T has no obvious structural or genetic cause. The Lawen data argues that in a subset of men — those with localized prostate cancer on active surveillance — baseline low T is a risk signal that should not be ignored. Both can be true: TRT is overall safer than the 1941 narrative suggests, and there are specific clinical populations where low baseline T is a flag rather than just a symptom to be treated.
For a vetted TRT clinic the operational shift is small: ask about prostate cancer history at intake, and don't enroll men on active surveillance through a standard hypogonadism workflow. For patients, the operational shift is also small: if you have a prostate cancer history, your TRT decision is multidisciplinary, not unilateral.
Limits of the Study
Three caveats worth flagging.
Single center. MD Anderson's AS program self-selects for men referred to a high-volume cancer center. Generalization to community-practice AS cohorts is plausible but not guaranteed.
Retrospective design. Baseline T was measured at varying times relative to enrollment, with assay differences over a 23-year window. The signal is robust enough to clear statistical significance but the magnitude has uncertainty.
No metabolic adjustment. The published model adjusts for age, race, PSA, and prostate volume — not BMI, fasting glucose, or inflammatory markers. Those are the most likely confounders, and they were not in the model.
None of these change the clinical takeaway, which is conservative: in men on active surveillance, baseline testosterone is worth measuring and may help stratify monitoring intensity.
Bottom Line
The Lawen 2026 MD Anderson cohort is the strongest signal yet that baseline testosterone matters in active surveillance prostate cancer. Men with total T at or below 300 ng/dL had 61% higher hazard of progressing to grade group 3 or higher disease, even after standard adjustments.
It does not say testosterone replacement is dangerous. It does not say low T causes prostate cancer. It says, narrowly: in men with localized cancer being monitored, baseline androgen status is a clinical signal worth paying attention to.
For most men reading this, the study is a reminder that prostate health and metabolic health travel together. For men on active surveillance, it's a prompt to add baseline testosterone to the workup and have a real conversation with a urologist — not a generic clinic — about what that number means for monitoring.
Related Reading
References
- Lawen CE, Gregg JR, et al. Low Testosterone Levels and Grade Group Progression Among Localized Prostate Cancer Patients on Active Surveillance: A Retrospective Cohort Study. Journal of Urology. February 2026. DOI: 10.1097/JU.0000000000004986.
- Testosterone replacement therapy following definitive treatment for prostate cancer: a scoping review of safety and efficacy. International Journal of Impotence Research. 2025. DOI: 10.1038/s41443-025-01206-3.
- Testosterone Replacement Therapy in Hypogonadal Men with a Prostate Cancer Diagnosis: A British Society for Sexual Medicine Consensus Statement. World Journal of Men's Health. 2025.
- Lincoff AM, et al. Cardiovascular Safety of Testosterone-Replacement Therapy (TRAVERSE). New England Journal of Medicine. 2023;389:107-117.
- American Urological Association. Evaluation and Management of Testosterone Deficiency Guideline. 2018, updated 2024.
- Morgentaler A, Traish AM. Shifting the paradigm of testosterone and prostate cancer: the saturation model and the limits of androgen-dependent growth. European Urology. 2009;55(2):310-321.