Abstract
This case series investigates the rate of prostate cancer diagnoses among transgender women treated in the Veterans Affairs health system.
Transgender women retain their prostate even after gender-affirmation surgery and thus remain at risk of prostate cancer. Studies to date are limited to case reports.1 We describe a large case series of transgender women with prostate cancer within the Veterans Affairs (VA) health system.
Methods
After Durham VA institutional review board approval with a waiver of informed consent, VA records for all adults with an International Classification of Diseases code for prostate cancer and at least 1 code for transgender identity at any time from January 2000 to November 2022 were queried. Detailed chart review was performed to confirm transgender identity and prostate cancer diagnosis. The number of cases per year was approximated using previous estimates of the median year of transgender identity entry in VA records (2011), corresponding to approximately 11 years of follow-up (2011-2022).2 Data from community clinicians were included if available in VA notes. Data at diagnosis were collected, including gender-affirming hormone therapies used, prostate-specific antigen (PSA), PSA density (measure of prostate cancer aggressiveness [lower risk, <0.15 ng/mL/g]), bilateral orchiectomy status, biopsy grade group (grades 1-5, with 5 being most severe), and clinical stage (T1-T4, with T4 being most advanced; eMethods in Supplement 1). Because estrogen, which lowers testosterone and suppresses prostate cancer,3 was the most common feminization therapy, patients were grouped by never used estrogen, formerly used estrogen but stopped prior to prostate cancer diagnosis, or actively used estrogen at diagnosis. Due to limited sample sizes, analyses were only descriptive.
Results
Among 449 patients with prostate cancer and transgender identity codes, 155 (35%, an estimated 14 cases per year) were confirmed transgender women with prostate cancer: 116 never used estrogen, 17 formerly used estrogen, and 22 actively used estrogen at diagnosis. All prostate cancer diagnoses were screen detected. At diagnosis, the median age was 61 years, 88% were White, median PSA was 6.8 ng/mL, the median duration of estrogen use was 32 months among former and active estrogen users, 43% (50 of 115) were biopsy grade group 1, and 45% (49 of 108) were clinical stage T1 (Table); 98% (152 of 155) had not undergone bilateral orchiectomy. Biopsy grade group 1 or 2 was found in 58 of 82 patients (71%) with no prior estrogen use, 9 of 16 (56%) with former estrogen use, and 9 of 17 (53%) with estrogen use at diagnosis.
Table. Patient Characteristics at Prostate Cancer Diagnosisa.
| No prior estrogen (n = 116) | Formerly receiving estrogen (stopped before prostate cancer diagnosis) (n = 17) | Taking estrogen at prostate cancer diagnosis (n = 22) | Overall (N = 155) | |
|---|---|---|---|---|
| Age at prostate cancer diagnosis, median (IQR), y | 60 (56-66) | 66 (61-71) | 61 (59-65) | 61 (56-67) |
| Race and ethnicity, No. (%)b | ||||
| No. | 96 | 17 | 16 | 129 |
| African American/Black | 8 (8) | 1 (6) | 1 (6) | 10 (8) |
| White | 84 (88) | 15 (88) | 15 (94) | 114 (88) |
| Other | 4 (4) | 1 (6) | 0 | 5 (4) |
| Year of diagnosis, median (IQR) | 2010 (2005-2013) |
2018 (2014-2019) |
2017 (2010-2019) |
2011 (2007-2016) |
| PSA at diagnosis, No. (%), ng/mL | ||||
| No. | 90 | 17 | 18 | 125 |
| 0-4 | 4 (4) | 1 (6) | 3 (17) | 8 (6) |
| >4-10 | 62 (69) | 11 (65) | 10 (56) | 83 (66) |
| >10-20 | 13 (14) | 5 (29) | 3 (17) | 21 (17) |
| >20 | 11 (12) | 0 | 2 (11) | 13 (10) |
| PSA, median (IQR), ng/mL | 6.7 (5.4-10.9) |
8.3 (5.0-11.3) |
7.0 (4.7-11.2) |
6.8 (5.3-11.1) |
| PSA density, median (IQR), ng/mL/gc | 0.21 (0.14-0.32) |
0.26 (0.19-0.30) |
0.31 (0.13-0.36) |
0.22 (0.14-0.33) |
| Type of gender-affirming hormone therapy ever used before diagnosis, No. (%) | ||||
| Estrogen | 0 | 17 (100) | 22 (100) | 39 (25) |
| Other | 9 (8) | 13 (76) | 15 (68) | 37 (24) |
| Duration receiving estrogen at diagnosis, mo | ||||
| Median (IQR) | 44 (11-78) | 25 (4-70) | 32 (8-72)d | |
| Duration not receiving estrogen at diagnosis, mo | ||||
| Median (IQR) | 34 (18-84) |
0 | NA | |
| Orchiectomy status at diagnosis, No. (%) | ||||
| No. | 116 | 17 | 22 | 155 |
| No orchiectomy | 116 (100) | 16 (94) | 20 (91) | 152 (98) |
| Before prostate cancer diagnosis | 0 | 1 (6) | 2 (9) | 3 (2) |
| Biopsy grade group at diagnosis, No. (%)e | ||||
| No. | 82 | 16 | 17 | 115 |
| 1 | 40 (49) | 5 (31) | 5 (29) | 50 (43) |
| 2 | 18 (22) | 4 (25) | 4 (24) | 26 (23) |
| 3 | 5 (6) | 3 (19) | 2 (12) | 10 (9) |
| 4 | 9 (11) | 3 (19) | 1 (6) | 13 (11) |
| 5 | 10 (12) | 1 (6) | 5 (29) | 16 (14) |
| Clinical stage at diagnosis, No. (%)f | ||||
| No. | 77 | 15 | 16 | 108 |
| T1 | 36 (47) | 6 (40) | 7 (44) | 49 (45) |
| T2 | 41 (53) | 9 (60) | 9 (56) | 59 (55) |
| T3-T4 | 0 | 0 | 0 | 0 |
Abbreviations: NA, not available; PSA, prostate-specific antigen (historical reference range for cisgender men, 0-4 ng/mL).
Percentages may not sum to 100 due to rounding and are based on number of patients with available data for each category.
Race and ethnicity were self-classified by each patient at entry into the Veterans Affairs health system. Race and ethnicity were assessed in this study due to varying incidence, morbidity, and mortality in prostate cancer that have been associated with race. Other includes Native Hawaiian or Other Pacific Islander.
PSA density of less than 0.15 ng/mL/g is associated with lower risk of prostate cancer (and lower risk of aggressive disease for individuals with prostate cancer).
Among individuals who used estrogen before prostate cancer diagnosis (ie, former and active estrogen users).
Biopsy grade group ranges from 1 to 5, with grade 5 representing the most severe disease.
Clinical stage ranges from T1 to T4, with T4 representing the most advanced disease.
The percentage of patients with biopsy grade group 4 or 5 was 23% (19 of 82) of patients who never used estrogen, 10 (12%) of whom were grade group 5; 25% (4 of 16) of former users, 1 (6%) of whom was grade group 5; and 35% (6 of 17) of active users, 5 (29%) of whom were grade group 5. The median PSA density was 0.21 ng/mL/g for never estrogen users, 0.26 ng/mL/g for former users, and 0.31 ng/mL/g for active users.
Discussion
This case series demonstrated that prostate cancer occurs in transgender women and is not as rare as published case reports might suggest. However, rates were lower than expected based on prior prostate cancer incidence estimates in cisgender male veterans.4 Using data from 2005 to 2019,4 the age-weighted cisgender mean number of cases was 331 per 100 000 in 2011 (median year of transgender identity entry in the VA2). Given estimates of 10 000 transgender women in the VA,2 33 cases per year would be expected. Instead, only about 14 per year were observed. Lower rates may stem from less PSA screening owing to barriers including lack of prostate cancer risk awareness or stigma,5 the suppressive effects of estrogen on prostate cancer development, or prostate cancers being missed in transgender women because of misinterpretation of “normal” PSA levels among those receiving gender-affirming hormone therapies.
Although limited numbers precluded formal statistical comparisons, transgender women receiving estrogen at diagnosis had the most aggressive disease (highest PSA density and proportion of biopsy grade group 5), suggesting delayed diagnosis or early selection of cancer cells resistant to androgen deprivation, which tend to be more aggressive. Among transgender women, 25% overall and 35% receiving estrogen had biopsy grade group 4 or 5 vs only 16% among cisgender male veterans.6 Only 8% of transgender women with prostate cancer vs 29% of cisgender male veterans with prostate cancer were Black,6 suggesting additional disparities at the intersection of race and gender identity.
Limitations include being a case series without a formal comparison group, restriction to the VA population, and the inability to access records outside the VA unless documented in notes, which may have underestimated prostate cancer numbers. Associations between characteristics such as estrogen use and markers of disease severity may reflect selection on disease diagnosis (an index event). Future work is needed to optimize prostate cancer detection strategies in transgender women.
Section Editors: Jody W. Zylke, MD, Deputy Editor; Kristin Walter, MD, Senior Editor.
eMethods
eReferences
Data Sharing Statement
References
- 1.Deebel NA, Morin JP, Autorino R, Vince R, Grob B, Hampton LJ. Prostate cancer in transgender women: incidence, etiopathogenesis, and management challenges. Urology. 2017;110:166-171. doi: 10.1016/j.urology.2017.08.032 [DOI] [PubMed] [Google Scholar]
- 2.Nik-Ahd F, Waller J, De Hoedt AM, et al. Seeing the unseen: how can we best identify transgender women within the Veterans Affairs healthcare system’s electronic medical record? J Sex Med. 2023;qdac033. doi: 10.1093/jsxmed/qdac033 [DOI] [PubMed] [Google Scholar]
- 3.Malkowicz SB. The role of diethylstilbestrol in the treatment of prostate cancer. Urology. 2001;58(2)(suppl 1):108-113. doi: 10.1016/S0090-4295(01)01252-3 [DOI] [PubMed] [Google Scholar]
- 4.Bryant AK, Lee KM, Alba PR, et al. Association of prostate-specific antigen screening rates with subsequent metastatic prostate cancer incidence at US Veterans Health Administration facilities. JAMA Oncol. 2022;8(12):1747-1755. doi: 10.1001/jamaoncol.2022.4319 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Sterling J, Garcia MM. Cancer screening in the transgender population: a review of current guidelines, best practices, and a proposed care model. Transl Androl Urol. 2020;9(6):2771-2785. doi: 10.21037/tau-20-954 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Makarov DV, Ciprut S, Walter D, et al. Association between guideline-discordant prostate cancer imaging rates and health care service among veterans and Medicare recipients. JAMA Netw Open. 2018;1(4):e181172. doi: 10.1001/jamanetworkopen.2018.1172 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
eMethods
eReferences
Data Sharing Statement