A PSA that climbs fivefold in eight months usually means you missed something. Sometimes it means the patient forgot to mention his testosterone injections.

The Clinical Problem

This week I saw a 79-year-old man whose PSA rose from 4.0 to 21.0 ng/mL in eight months. By the time of biopsy, it had already eased to 16.

The workup was quieter than the PSA. MRI showed a PI-RADS 3 lesion. Biopsy showed Gleason 3+4 = 7, Grade Group 2. PSMA PET showed uptake in the prostate and nothing beyond it. His Decipher score came back at 0.34, low risk. 

Favorable intermediate-risk disease. Yet that PSA velocity nagged me. Were we missing something more aggressive?

Then, near the end of the visit, he mentioned it. He had started testosterone cypionate right around the time his PSA began to climb.

My concerns subsided. The testosterone didn’t create his cancer. It turned up the volume on the PSA and revealed a low-risk cancer that is good to know about.

Why This Matters — To You and Your Patient

Testosterone therapy is everywhere. Many men start it through wellness clinics, telehealth, or primary care, and many don’t think to mention it.

When exogenous testosterone drives the PSA, velocity stops measuring tumor behavior and starts measuring androgen exposure. Miss that history and you may chase an aggressive cancer that isn’t there.

The stakes for this man are real. Taken at face value, a PSA above 20 would push him into a high-risk category and toward treatment. Read in context, he has favorable intermediate-risk disease and a reasonable path to keep his testosterone and his quality of life.

How It Works — Mechanism & Physiology

PSA is an androgen receptor readout. Think of the androgen receptor (AR) as a lock and testosterone as the key. When the key turns, AR switches on a set of genes, and KLK3, the PSA gene, is one of them.⁷ 

So PSA measures how loudly AR is talking in prostate tissue, benign or malignant. It is not a direct measure of cancer aggressiveness.

The saturation model explains this PSA. Prostate tissue is exquisitely sensitive to testosterone at very low levels. But AR binding becomes nearly maximal well below the normal physiologic range. Above that point, more testosterone adds little.^3,4 

Picture a dry sponge. The first pour soaks in completely. Once it’s saturated, pouring more does almost nothing.

If this man was hypogonadal before treatment, testosterone cypionate refilled a dry sponge. Benign and malignant tissue both responded, and PSA surged. That is a one-time step up, not a sign of an accelerating cancer.

Once the tissue is saturated, PSA should settle. His drift from 21 to 16 before biopsy fits that pattern. On continued TRT, I expect PSA to find a stable plateau. What would concern me is a PSA that keeps climbing after it should have leveled off.

The saturation model also explains the bigger picture. Castration shrinks prostate cancer, yet testosterone levels in the normal range don’t predict who gets it.^3,5 

Why Decipher matters here. Decipher is a 22-gene RNA test run on the biopsy tissue itself, scoring the tumor’s likelihood of metastasis. Because it reads tumor biology directly rather than serum PSA, it gives us an androgen-independent check. In biopsy validation work, it added prognostic value beyond clinical risk tools.⁸ 

His score of 0.34 is low risk. Grade Group 2 pathology, low-risk genomics, and organ-confined PSMA all point the same way. Only the PSA disagreed, and the PSA has an explanation.

The Therapeutic Goal

The goal is not the lowest possible PSA. It is the best possible life, with enough vigilance to catch anything we got wrong.

For this man, that meant offering to restart TRT if he chose to, for the energy and quality of life he valued. Paired with it: close PSA and imaging follow-up over the next two years or so.

His age doesn’t shorten that horizon as much as you might think. His father lived to 95 and his mother to 100. He may have 20 years ahead, which is exactly why the monitoring matters.

What the Evidence Says

Testosterone levels don’t predict prostate cancer. A pooled analysis of 18 prospective studies (3,886 cases, 6,438 controls) found no association between serum testosterone or DHT and prostate cancer risk.⁵ A later meta-analysis agreed: each 5 nmol/L increase in testosterone carried a relative risk of 0.99, essentially no effect.⁶ 

TRT in higher-risk men. Pozzi and colleagues (2025) compared 623 men with testosterone deficiency and a family history of prostate cancer who received TRT against 623 matched men who did not. Over 10 years, neither diagnosis (HR 0.81, 95% CI 0.51–1.28) nor active treatment (HR 0.55, 95% CI 0.29–1.03) differed significantly.² 

One caution: the curves crossed late, with 10-year diagnosis estimates of 21.7% versus 12.4%, and wide confidence intervals.² The authors also note small studies suggesting TRT may be safe in men on active surveillance.² 

Low testosterone may be the bigger red flag. Lawen and colleagues (2026) followed 924 men on active surveillance. Baseline testosterone at or below 300 ng/dL was associated with a 61% higher risk of progression to Grade Group 3 or higher (HR 1.61, 95% CI 1.02–2.54).¹ 

That interval runs from 2% to 154%, so the signal is real but imprecise. Low testosterone was not significantly tied to Grade Group 2 progression, and men with low testosterone also had higher BMI and more statin use.¹ 

The likely biology: tumors that grow in a low-androgen environment may adapt by making their own androgens and amplifying AR signaling.^1,10 Keeping a man hypogonadal is not automatically the safe choice.

Evidence quality and conflicts. Pozzi and Lawen are retrospective. Pozzi lacks Gleason data; Lawen’s cohort was mostly Grade Group 1.^1,2 Decipher validation studies were largely conducted with involvement from the test’s developer.⁸ Conflict-of-interest disclosures for each study should be confirmed before publication.

What It Does NOT Do

The saturation model explains a PSA step-up. It does not guarantee that every PSA rise on TRT is benign. No randomized trial has tested TRT in men with untreated, low-risk prostate cancer. Reassuring data come from observational and small surveillance studies.^2,9 

Lawen shows an association between low testosterone and upgrading. It does not prove that raising testosterone prevents progression.¹ 

A low Decipher score lowers the odds of aggressive biology. It does not eliminate biopsy sampling error or replace follow-up imaging.

Safety, Practical Use & Clinical Considerations

Ask about testosterone every time. Injections, gels, pellets, clomiphene, and telehealth or wellness-clinic products. Line up the start date against the PSA curve.

Expect a new PSA plateau on TRT. Set a new on-treatment baseline once PSA stabilizes, and don’t let pre-TRT and on-TRT values share one trend line. A continued rise after the plateau is the signal that warrants a closer look.

Monitor closely for the first two years. Serial PSA with scheduled MRI, and repeat biopsy if imaging or PSA kinetics change. Track hematocrit as with any TRT patient.

Document the reasoning. Record the pre-TRT PSA, the Decipher score, PSMA findings, and the shared decision to resume TRT. This is individualized care built on limited data, and the chart should show it.⁹ 

Final Thoughts

One question asked late in the visit changed the whole picture. Testosterone didn’t make this cancer dangerous; it made the PSA loud. 

Evidence first, context always, and a monitoring plan that lets a 79-year-old with possibly 20 good years ahead actually enjoy them.

Actionable Clinical Takeaways

1. Ask every man with a rising PSA about testosterone and other androgen-modifying products, and match the start date to the PSA curve.
2. Use the saturation model to read a PSA surge after starting TRT in a hypogonadal man: expect a step up and then a plateau, and worry about a PSA that keeps climbing.
3. When PSA and pathology disagree, let androgen-independent signals break the tie: Grade Group, Decipher, and PSMA staging.
4. For low-risk or favorable intermediate-risk disease, TRT can be a reasonable shared decision when paired with structured PSA and imaging follow-up.
5. Don’t assume a hypogonadal state is protective; low baseline testosterone was associated with more upgrading to Grade Group 3 or higher on surveillance.

Go Deeper: Testosterone & Prostate Cancer on the Podcast

The testosterone question has been a running thread on The Dr. Geo Prostate Podcast. If this issue raised questions for your practice, these four conversations pick up where it leaves off.

The Shocking Truth Between Testosterone and Prostate Cancer with Dr. Helen Bernie — Episode 108 

Does Testosterone Fuel or Protect Against Prostate Cancer? With Dr. Mohit Khera — Episode 154   

Dr. Morgentaler Debunks the Myth Between Testosterone and Prostate Cancer — Episode 169   

References

1. Lawen, T, Garcia, RS, Smith, MT, Pyo, J, Higgason, NM, Robert, MJM, Chapin, BF, Chery, LS, Davis, JW, Ward, JF, Pisters, LL, Pettaway, CA, Hahn, AW, Frigo, DE, Wei, P, Daniel, CR, & Gregg, JR (2026). Low testosterone levels and grade group progression among localized prostate cancer patients on active surveillance: A retrospective cohort study. The Journal of Urology. Advance online publication. doi:10.1097/JU.0000000000004986
2. Pozzi, E, Able, CA, Kohn, T, Kava, BR, Montorsi, F, and Salonia, A (2025). Incidence of prostate cancer in men with testosterone deficiency and a family history of prostate cancer receiving testosterone therapy: A comparative study. BMJ Oncology, 4(1), e000520. doi:10.1136/bmjonc-2024-000520
3. Morgentaler, A, and Traish, AM (2009). Shifting the paradigm of testosterone and prostate cancer: The saturation model and the limits of androgen-dependent growth. European Urology, 55(2), 310–320.
4. Khera, M, Crawford, D, Morales, A, Salonia, A, and Morgentaler, A (2014). A new era of testosterone and prostate cancer: From physiology to clinical implications. European Urology, 65(1), 115–123.
5. Roddam, AW, Allen, NE, Appleby, P, Key, TJ, and Endogenous Hormones and Prostate Cancer Collaborative Group. (2008). Endogenous sex hormones and prostate cancer: A collaborative analysis of 18 prospective studies. Journal of the National Cancer Institute, 100(3), 170–183.
6. Boyle, P, Koechlin, A, Bota, M, d’Onofrio, A, Zaridze, DG, Perrin, P, Fitzpatrick, JM, and Burnett, AL (2016). Endogenous and exogenous testosterone and the risk of prostate cancer and increased prostate-specific antigen (PSA) level: A meta-analysis. BJU International, 118(5), 731–741.
7. Heinlein, CA, and Chang, C (2004). Androgen receptor in prostate cancer. Endocrine Reviews, 25(2), 276–308.
8. Nguyen, PL, Haddad, Z, Ross, AE, Martin, NE, Deheshi, S, Lam, LLC, et al. (2017). Ability of a genomic classifier to predict metastasis and prostate cancer-specific mortality after radiation or surgery based on needle biopsy specimens. European Urology, 72(5), 845–852.
9. Kaplan, AL, Hu, JC, Morgentaler, A, Mulhall, JP, Schulman, CC, and Montorsi, F (2016). Testosterone therapy in men with prostate cancer. European Urology, 69(5), 894–903.
10. Mostaghel, EA (2014). Steroid hormone synthetic pathways in prostate cancer. Frontiers in Oncology, 4, 60.