Testosterone Therapy does not Increase the Risks of Prostate Cancer Recurrence or Death after Definitive Treatment for Localized Disease - Full Text Article

Background: The safety of testosterone therapy (TT) after definitive treatment for localized prostate cancer remains undefined. We analyzed the risks of biochemical recurrence and mortality in men receiving TT after treatment for localized prostate cancer.

Methods: Cohort analysis using the national US Veterans Affairs Informatics and Computing Infrastructure. We identified 69,984 patients with localized prostate cancer diagnosed from 2001 to 2015 treated with surgery or radiation. We coded receipt of TT after treatment as a time-dependent covariate; used the National Death Index to identify cause of death; and defined biochemical recurrence as PSA > 0.2 ng/mL after surgery and nadir + 2 ng/mL after radiation. We analyzed recurrence and mortality using cumulative incidence curves, Fine–Gray competing risk regression, and Cox regression.

Results: This cohort included 28,651 surgery patients and 41,333 radiation patients, of whom 469 (1.64%) and 543 (1.31%), respectively, received TT with a median follow-up of 6.95 years. Comparing testosterone users to nonusers, there were no between-group differences in biochemical recurrence, prostate cancer-specific mortality, or overall mortality after surgery [hazard ratios (HR): 1.07; HR: 0.72 (p = 0.43); and HR: 1.11 (p = 0.43), respectively] or radiation [HR: 1.07; HR: 1.02 (p = 0.95); and HR: 1.02 (p = 0.86), respectively]. Limitations included lack of detailed data on TT duration and serum testosterone concentrations.

Conclusions: In this multi-ethnic national cohort, TT did not increase the risks of biochemical recurrence or prostate cancer specific or overall mortality after surgery or radiation. These data suggest that TT is safe in appropriate men after definitive treatment of localized prostate cancer.


Testosterone deficiency affects 30% of men who age 40–79 years.1-2 Testosterone therapy (TT) may ameliorate testosterone deficiency symptoms including fatigue, decreased libido, erectile dysfunction, depressed mood, and decreased lean body mass3,4. Many patients with localized prostate cancer may benefit from TT. However, the safety of TT in these men remains undefined, even in those who are disease-free after definitive treatment. While clinical data are limited, principles of androgen-driven prostate carcinogenesis5 and androgen deprivation therapy (ADT) for advanced prostate cancer6 have fueled speculation that TT may increase the risks of disease progression and recurrence. These concerns prompted the US Federal Drug Administration to issue a black box warning —its most serious advisory—recommending against TT in men with known prostate cancer, regardless of disease status.7

Still, cohort analyses in prostate cancer patients receiving TT after surgery8-10 or radiation10, 11 suggest that TT does not increase the risks of biochemical recurrence, salvage ADT, or mortality.12 Accordingly, the European  Association of Urology has recommended consideration of TT in patients with symptomatic testosterone deficiency and localized low-to-intermediate-risk prostate cancer in patients who remained disease-free for more than 1 year after prostatectomy.13 The American Urological Association has similarly noted that there is insufficient evidence to demonstrate associations of TT with prostate cancer risk.14

Since randomized clinical trials of TT after definitive prostate cancer treatment would lack equipoise and feasibility, observational studies provide the most robust safety data. Most prior studies of TT in this population involved relatively small cohorts with shorter term follow-up.8,9,11 Additional survival analyses with longer-term follow-up in larger and more diverse populations would thus substantively inform care. Using the Veterans Affairs Informatics and Computing Infrastructure (VINCI) database, a national cohort of US military veterans, we assessed associations of TT with biochemical recurrence, prostate cancer-specific mortality, and overall mortality. We hypothesized that there would be no differences between TT users and nonusers for recurrence or mortality.


Data source

The VINCI database includes electronic health records of more than 20 million veterans from the years 2000–2015 who sought healthcare at ~1400 VA outpatient facilities and 152 medical centers in the US.15 VINCI includes tumor registry data collected by registrars who follow protocols issued by the American College of Surgeons. For mortality data, VINCI was linked with the US National Death Index. The San Diego VA Institutional Review Board approved this study. RRS, SHP, BSR, RD, and AK accessed the patient records.

Study population

A total of 72,083 clinically localized (low, intermediate, and high) nonmetastatic prostate cancer patients were identified between 2001 and 2015 who received primary treatment with radical prostatectomy or radiation. Of these, 2099 patients were excluded for unknown information for covariates or cause of death, leaving a final analytic cohort of 69,984 (n = 28,651 surgery and n = 41,333 radiation). Through December 31, 2017, all men were followed until death or last follow-up with a VA provider. Age, year of diagnosis, body mass index (BMI), race, Gleason sum, clinical T stage, ADT, and prostate-specific antigen (PSA) were identified for each patient. Eligible ADT medications included leuprolide, goserelin, triptorelin, histrelin, flutamide, bicalutamide, nilutamide, and degarelix. Patients who initiated ADT of any sort were considered to be exposed to ADT regardless of whether ADT use occurred before or after TT use. Charlson comorbidity index scores were determined using previously described methods.16-18


Outcomes included biochemical recurrence, prostate cancer-specific mortality, and all-cause mortality. Biochemical recurrence was defined as nadir + 2 ng/ml19 after radiation PSA ≥ 0.2 ng/ml after surgery.20 Exposure ascertainment TT utilization was ascertained using VA inpatient and outpatient pharmacy records. TT was identified by: (1) testosterone formulation (testosterone cypionate, testosterone suspension, testosterone propionate, testosterone pellet, testosterone, Depo-testosterone, Androgel, Testim, Fortest, Axiron, Vogelxo, Androderm, Testred, Methyltestosterone, Striant, Aveed, Methitest, Natesto, Testopel, Androxy); or (2) Healthcare Common Procedure Coding System Codes (J0900, J1060, J1070, J1080, J2320, J3120, J3030, J3140, J3150, S0189). Androxy and Methitest are synthetic androgens. TT exposure was defined as a binary variable independent of dose or method of therapy.

Statistical analysis

Differences in covariates between TT users and nonusers were assessed using chi-square and Wilcoxon rank-sum tests. Surgical patients and radiation patients were analyzed separately. Cumulative incidence curves were used to assess unadjusted estimates of PCSM and ACM. TT exposure was coded as a time-dependent covariate in multivariable models, to account for the possibility of the immortal time bias. We classified TT exposure beginning with initiation of TT and then continuing for the remainder of the study period: that is, a patient was not considered to be exposed to TT until initiation of TT, and then was considered as exposed for the rest of follow-up. Adjusted estimates of PCSM and NCM were evaluated using multivariable Fine–Gray competing risk regression to account for the competing risk of death and ACM and BCR using multivariable Cox regression. Subdistribution hazard ratios (HRs) and were estimated for PCSM while HRs were estimated for ACM and BCR to quantify the effect a covariate has on the risk of these outcomes. All models adjusted for the covariates listed in the model building section of the methods and started evaluating mortality after RT or RP, respectively. All statistical tests were two sided, with p < 0.05 considered significant, and conducted with SAS 9.4 (SAS Institute Inc, Cary, NC).

Table 1: Patient demographics and exposure to testosterone therapy after radical prostatectomy and radiation therapy. 



For radical prostatectomy, the cohort included 28,651 patients, of whom 469 (1.64%) received TT. Median follow-up was 7.36 (95% CI: 1.19–15.62) years: 8.74 (95% CI: 2.26–16.36) and 7.32 (95% CI: 1.18–15.62) years for TT users and nonusers, respectively. Median time from surgery to initiation of TT was 3.04 (95% CI: 0.025–11.95) years. At baseline, there were no differences in clinical T stage, adjuvant ADT use, or duration of ADT use between TT users and nonusers. TT users had higher BMI and Charlson scores; lower median PSA and lower Gleason sum at time of diagnosis; and were more likely to be White (Table 1).

For radiation therapy, the cohort included 41,333 patients, of whom 543 (1.31%) received TT. Median follow-up was 6.69 years (95% CI: 1.08–14.89 years): 8.79 (95% CI: 2.62–15.72 years) and 6.66 years (95% CI: 1.08–14.89 years) for TT users and nonusers, respectively. Median time from completion of radiation treatment to TT was 3.53 (95% CI: 0.39–11.49) years. At baseline, there were no differences in Charlson comorbidity, clinical T stage, median PSA, or—among those who had received it in conjunction with radiotherapy—duration of neoadjuvant ADT between groups. TT users were younger, more likely to have higher BMI, had lower median Gleason at time of diagnosis, more likely to be diagnosed at a younger age, and more likely to have used neoadjuvant ADT therapy (Table 1).

Table 2: Survival and recurrence in men treated with radical prostatectomy. 


Outcomes: radical prostatectomy

In unadjusted analyses, TT users had lower 10-year cumulative incidences of prostate cancer-specific (1.1 vs. 3.5%, p = 0.01) and all-cause (11.7 vs. 20.0%; p < 0.001) mortality. In adjusted models, there were no differences in biochemical recurrence (HR: 1.07; 95% CI: 0.84–1.36; p = 0.59) or prostate cancer-specific (HR: 0.72; 95% CI: 0.32–1.62; p = 0.43) or all-cause (HR: 1.11; 95% CI: 0.85–1.44; p = 0.43) mortality between groups (Table 2).

Outcomes: radiation therapy

On univariable analysis, TT users had lower 10-year cumulative incidences of prostate cancer-specific (2.7 vs. 4.9%; p = 0.05) and all-cause (17.9 vs. 32.7%, p < 0.001) mortality. On multivariable analysis, there were no differences in biochemical recurrence (HR: 1.07; 95% CI: 0.90–1.27; p = 0.45) or prostate cancer-specific (HR: 1.02; 95% CI: 0.62–1.67; p = 0.95) or all-cause mortality (HR: 1.02; 95% CI: 0.84–1.24; p = 0.86) between groups (Table 3).


In this ethnically diverse population-based study of nearly 70,000 US veterans, TT therapy did not increase the risks of biochemical recurrence or prostate cancer-specific or allcause mortality after surgery or radiation for localized prostate cancer. These data suggest that TT is safe in appropriate patients after definitive treatment of localized prostate cancer.

Table 3: Survival and recurrence in men treated with radiation therapy. 


To our knowledge, this study is the most comprehensive and diverse comparative analysis to date of the safety of TT after definitive treatment for prostate cancer. In contrast to most other prior studies, it utilized a large, multi-ethnic, nationwide cohort incorporating a broad age range. A substantial strength of this study was the high prevalence of African-American men: 24% for radical prostatectomy and 28% for radiotherapy. These results provide reassurance that African-American men, who are at increased risk of death from prostate cancer, may receive TT after definitive treatment for localized disease without increased of cancer-specific mortality (HR: 0.86, CI: 0.77–0.97, p = 0.01) or all-cause mortality (HR: 0.94, CI: 0.9–0.98, p = 0.01).

Randomized clinical trials of testosterone in hypogonadal prostate cancer survivors, with incident disease recurrence or death as primary outcomes, would pose substantial ethical challenges. Observational studies such as this one are thus the most feasible method for assessing the safety of therapeutic testosterone in this population. Although testosterone promotes prostate tumor cell growth, there are no definitive data to show that normalization of serum testosterone in hypogonadal men with prostate cancer worsens disease outcomes after definitive treatment with intent to cure. While we were unable to definitely determine whether the men in this study were clinically hypogonadal, our results suggest that TT does not worsen cancer outcomes after treatment.

Prior observational studies have demonstrated similar results. In two small cohorts of patients who received TT after surgery, there was no increased risk of biochemical recurrence.8,9 In a study of 149,354 men over the age of 65 years with prostate cancer from 1992 to 2007 in the US Surveillance, Epidemiology, and End Results-Medicare registry, 1181 (0.79%) of whom received TT after diagnosis, there were no associations of TT with posttreatment use of salvage ADT, or in prostate cancer-specific or overall mortality.12 Similarly, in 38,570 patients in the National Prostate Cancer Register of Sweden identified from 2009 to 2012, of whom 1% initiated TT, there were no increased risk in those who were treated with TT compared with the control, OR: 1.03; 95% CI: 0.90–1.17.21

There were two patterns of TT utilization in this cohort which were similar to observations in prior studies, and likely reflect prevailing treatment principles for these patients: time intervals between cancer treatment and TT; and disease characteristics of patients receiving TT. First, the median time from completion of cancer therapy to initiation of TT was 6 months longer for the radiation group—a disparity potentially attributable to verification of an appropriate nadir 12 months after radiotherapy prior to beginning TT.11, 22-24 Second, in both the surgery and radiation groups, patients with Gleason ≤ 7 (Grade Group 3) were more likely to receive TT than those with Gleason ≥ 8 (Grade Group 4). These patterns, consistent with prior studies,21,25 may have reflected reluctance to initiate TT in patients with higher-risk disease. The higher prevalence of TT among radiotherapy (17%) compared with surgery (12%) patients with high-risk disease may have reflected a higher prevalence of persistent posttreatment testosterone deficiency resulting from neoadjuvant ADT.

There are some potential limitations to this study. First, VINCI lacks detailed TT duration, dose, and formulation data. Inferences regarding testosterone dose–response and prostate cancer risk, therefore, cannot be drawn. Second, VINCI lacks data on patient blood testosterone concentrations. Thus, testosterone deficiency diagnosis could not be independently verified with serum testosterone, and baseline and on-treatment testosterone data were not available. Yet the prevalence of TT in this cohort was similar to other studies in this population,12, 21 suggesting that appropriate, standard-of-care principles were applied for diagnosis and treatment of testosterone deficiency; and that differential misclassification bias was unlikely. Finally, pathologic staging was not consistently available for all patients in VINCI, and could not be incorporated into the analyses. However, since the models adjusted for Gleason sum and clinical stage, it is unlikely that pathologic stage data would have further informed these analyses.


In men with localized prostate cancer who underwent prior have undergone treatment with surgery or radiation, TT was not associated with increased risks of biochemical recurrence, prostate cancer-specific mortality, or all-cause mortality. These results suggest that TT is safe in appropriate patients after definitive management of localized prostate cancer.

Acknowledgments: This work was supported by the National Institutes of Health (Grant Number TL1TR001443 to RRS).

Compliance with ethical standards

Conflict of interest: The authors declare that they have no conflict of interest.
Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Authors: Reith R. Sarkar1,2, Sunil H. Patel2,3, J. Kellogg Parsons3, Rishi Deka1, Abhishek Kumar1,2, John P. Einck1, Arno J. Mundt1, A. Karim Kader3, Christopher J. Kane3, Paul Riviere1,2, Rana McKay4, James D. Murphy1,2, Brent S. Rose1,2,3

  1. Department of Radiation Medicine and Applied Sciences, University of California San Diego, La Jolla, CA 92093, USA
  2. Veteran Affairs Healthcare, San Diego, CA, USA
  3. Department of Urology, University of California San Diego, La Jolla, CA 92093, USA
  4. Division of Hematology-Oncology, Department of Internal Medicine, University of California San Diego, La Jolla, CA 92093, USA

1. Araujo AB, Esche GR, Kupelian V, O’Donnell AB, Travison TG, Williams RE, et al. Prevalence of symptomatic androgen deficiency in men. J Clin Endocrinol Metab. 2007;92:4241–7. 
2. Wu FCW, Tajar A, Beynon JM, Pye SR, Silman AJ, Finn JD, et al. Identification of late-onset hypogonadism in middle-aged and elderly men. N Engl J Med. 2010;363:123–35. 
3. Corona G, Giagulli VA, Maseroli E, Vignozzi L, Aversa A, Zitzmann M, et al. Therapy of endocrine disease: testosterone supplementation and body composition: results from a metaanalysis study. Eur J Endocrinol. 2016;174:R99–116.
4. Hackett G. An update on the role of testosterone replacement therapy in the management of hypogonadism. Ther Adv Urol. 2016;8:147–60. 
5. Fowler JEJ, Whitmore WFJ. The response of metastatic adenocarcinoma of the prostate to exogenous testosterone. J Urol. 1981;126:372–5.
6. Perlmutter MA, Lepor H. Androgen deprivation therapy in the treatment of advanced prostate cancer. Rev Urol. 2007;9 (Suppl 1):S3–8.
7. FDA. Highlights of prescribing information on testosterone. 2009. https://www.accessdata.fda.gov/drugsatfda_docs/label/2009/021015s022lbl.pdf.
8. Kaufman JM, Graydon RJ. Androgen replacement after curative radical prostatectomy for prostate cancer in hypogonadal MEN. J Urol. 2004;172:920–2. 
9. Agarwal PK, Oefelein MG. Testosterone replacement therapy after primary treatment for prostate cancer. J Urol. 2005;173:533–6. 
10. Pastuszak AW, Khanna A, Badhiwala N, Morgentaler A, Hult M, Conners WP, et al. Testosterone therapy after radiation therapy for low, intermediate and high risk prostate cancer. J Urol. 2015;194:1271–6.
11. Sarosdy MF. Testosterone replacement for hypogonadism after treatment of early prostate cancer with brachytherapy. Cancer.
12. Kaplan AL, Trinh Q, Sun M, Carter SC, Nguyen PL, Shih YT, et al. Testosterone replacement therapy following the diagnosis of prostate cancer: outcomes and utilization trends. J Sex Med. 2014;11:1063–70. 
13. EUA. Male hypogonadism. 2012. https://uroweb.org/guideline/male-hypogonadism/#5_5.
14. Mulhall JP, Trost LW, Brannigan RE, Kurtz EG, Redmon JB, Chiles KA, et al. Evaluation and management of testosterone deficiency: AUA guideline. J Urol. 2018;200:423–32. 
15. Duvall S. What is VINCI? 2016. https://www.hsrd.research.va.gov/for_researchers/cyber_seminars/archives/video_archive.cfm?SessionID=1143. Accessed 10 Nov 2016.
16. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis. 1987;40:373–83. 
17. Quan H, Sundararajan V, Halfon P, Fong A, Burnand B, Luthi JC, et al. Coding algorithms for defining comorbidities in ICD-9-CM and ICD-10 administrative data. Med Care. 2005;43:1130–9. 
18. Klabunde CN, Potosky AL, Legler JM, Warren JL. Development of a comorbidity index using physician claims data. J Clin Epidemiol. 2000;53:1258–67. 
19. Paller CJ, Antonarakis ES. Management of biochemically recurrent prostate cancer after local therapy: evolving standards of care and new directions. Clin Adv Hematol Oncol. 2013;11:14–23.
20. Cookson MS, Aus G, Burnett AL, Canby-Hagino ED, D’Amico AV, Dmochowski RR, et al. Variation in the definition of biochemical recurrence in patients treated for localized prostate cancer: the American Urological Association Prostate Guidelines for Localized Prostate Cancer Update Panel report and recommendations for a standard in the reporting of surgical outcomes. J Urol. 2007;177:540–5.
21. Loeb S, Folkvaljon Y, Damber J-E, Alukal J, Lambe M, Stattin P. Testosterone replacement therapy and risk of favorable and aggressive prostate cancer. J Clin Oncol. 2017;35:1430–6. 
22. Ragde H, Elgamal AA, Snow PB, Brandt J, Bartolucci AA, Nadir BS, et al. Ten-year disease free survival after transperineal sonography-guided iodine-125 brachytherapy with or without 45-gray external beam irradiation in the treatment of patients with clinically localized, low to high Gleason grade prostate carcinoma. Cancer. 1998;83:989–1001.
23. Critz FA, Williams WH, Levinson AK, Benton JB, Schnell FJ, Holladay CT, et al. Prostate specific antigen bounce after simultaneous irradiation for prostate cancer: the relationship to patient age. J Urol. 2003;170:1864–7.
24. Akyol F, Ozyigit G, Selek U, Karabulut E. PSA Bouncing after short term androgen deprivation and 3D-conformal radiotherapy for localized prostate adenocarcinoma and the relationship with the kinetics of testosterone. Eur Urol. 2005;48:40–5. 
25. Murtola TJ, Kujala PM, Tammela TLJ. High-grade prostate cancer and biochemical recurrence after radical prostatectomy among men using 5alpha-reductase inhibitors and alpha-blockers. Prostate. 2013;73:923–31.

Read an Editorial by Jack Schalken, PhD: The Risks of Biochemical Recurrence and Death in Men Receiving Testosterone Therapy After Treatment for Localized Prostate Cancer - Editorial