Long-term Exposure to Testosterone Therapy and the Risk

of High Grade Prostate Cancer

Jacques Baillargeon,* Yong-Fang Kuo, Xiao Fang and Vahakn B. Shahinian
From the Departments of Preventive Medicine and Community Health (XF), and Internal Medicine, Sealy Center on Aging,
University of Texas Medical Branch (JB, YFK), Galveston, Texas, and Department of Internal Medicine, University of
Michigan (VBS), Ann Arbor, Michigan

Purpose: To our knowledge no population based studies have been done to

Abbreviations
and Acronyms
ADT androgen deprivation
therapy
PSA prostate specific antigen
Accepted for publication May 29, 2015.
Study received University of Texas Medical
Branch institutional review board approval.
* Correspondence: Department of Preventive
Medicine and Community Health, University of
Texas Medical Branch, Galveston, Texas 77555
(telephone: 409-772-4534; FAX: 409-772-8931;
e-mail: jbaillar@utmb.edu).
See Editorial on page 1527.
examine whether long-term exposure to testosterone therapy is associated with
an increased risk of high grade prostate cancer. We examined whether exposure
to testosterone during a 5-year period was associated with an increased risk of
high grade prostate cancer and whether this risk increased in a dose-response
fashion with the cumulative number of testosterone injections.
Materials and Methods: Using SEER (Surveillance, Epidemiology and End
Results)-Medicare linked data we identified 52,579 men diagnosed with incident
prostate cancer between January 1, 2001 and December 31, 2006 who had
a minimum of 5 years continuous enrollment in Medicare before the cancer
diagnosis. We excluded patients diagnosed at death or after autopsy, those
enrolled in a health maintenance organization in the 60 months before diagnosis
and those with unknown tumor grade or tumor stage. In the 5 years before
diagnosis 574 men had a history of testosterone use and 51,945 did not.
Results: On logistic regression adjusting for demographic and clinical character-
istics exposure to testosterone therapy was not associated with an increased risk of
high grade prostate cancer (OR 0.84, 95% CI 0.67e1.05) or receipt of primary
androgen deprivation therapy following diagnosis (OR 0.97, 95% CI 0.74e1.30).
In addition the risk of high grade disease did not increase according to the total
number of testosterone injections (OR 1.00, 95% CI 0.98e1.01).
Conclusions: Our finding that testosterone therapy was not associated with an
increased risk of high grade prostate cancer may provide important information
regarding the risk-benefit assessment for men with testosterone deficiency
considering treatment.

Key Words: prostatic neoplasms, testosterone, androgens,

neoplasm grade, SEER program

TESTOSTERONE prescriptions for older association between testosterone

men in the United States have
increased more than threefold in the
last decade.1 Despite this rapid growth
to our knowledge there have been
no large, long-term, randomized, con-
trolled trials of testosterone therapy to
establish its safety. Currently there
is considerable debate about the
therapy and prostate cancer.2e16
Research dating back to the 1940s
established the testosterone depen-
dence of prostate cancer and led to
widespread concern that testosterone
therapy may increase the risk of
prostate cancer or an aggressive form
of the disease.6,17 Although several

0022-5347/15/1946-1612/0 http://dx.doi.org/10.1016/j.juro.2015.05.099

1612 j www.jurology.com
THE JOURNAL OF UROLOGY
2015 by AMERICAN UROLOGICAL ASSOCIATION EDUCATION AND RESEARCH, INC.
Vol. 194, 1612-1616, December 2015
Printed in U.S.A.

Descargado para Anonymous User (n/a) en ClinicalKey Espanol Colombia, Ecuador & Peru Flood Relief de ClinicalKey.es por Elsevier en julio 21, 2017.
Para uso personal exclusivamente. No se permiten otros usos sin autorizacin. Copyright 2017. Elsevier Inc. Todos los derechos reservados.
 TESTOSTERONE THERAPY AND RISK OF HIGH GRADE PROSTATE CANCER
longitudinal studies subsequently showed no
increased risk of prostate cancer incidence associated
with testosterone use,3,5,7,18e20 no population based
studies been done to examine the association of
high grade prostate cancer with testosterone ex-
posure beyond 1 year. Given the slow course of
prostate cancer development and the unknown la-
tency period associated with testosterone exposure,
examining this risk during a sufficient period is
critically important from a clinical and a public
health perspective.
Therefore, we examined whether exposure to
testosterone in a 5-year period was associated with
an increased risk of high grade prostate cancer
at diagnosis and whether this risk increased in
dose-response fashion with the total number of
testosterone injections. As an additional marker of
high risk disease we examined the receipt of pri-
mary ADT.
MATERIALS AND METHODS
Data Source
We analyzed data from SEER-Medicare, a linkage of
population based cancer registries from 20 SEER regions,
which cover approximately 28% of the population of the
United States, with Medicare administrative data.21
Approximately 94% of patients recorded in the SEER
registry have been linked to their Medicare claims for
covered health related services.
Study Cohort
Using SEER-Medicare linked data we identified 52,579
men diagnosed with incident prostate cancer between
January 1, 2001 and December 31, 2006 who had a min-
imum of 5 years of available records. We excluded from
study patients diagnosed at death or after autopsy, those
enrolled in a health maintenance organization in the
60 months before diagnosis with incident prostate cancer
and those with unknown tumor grade or stage. Among the
cohort 574 men had a history of testosterone use before
the prostate cancer diagnosis and 51,945 did not.
Testosterone Therapy
Testosterone therapy was defined using HCPCS
(Health Care Procedure Coding System) drug administra-
tion codes, including J0900dtestosterone enanthate, up
to 1 cc; J1060dtestosterone cypionate, up to 1 ml; J1070d
testosterone cypionate, up to 100 mg; J1080dtestosterone
cypionate, up to 200 mg; J3120dtestosterone enanthate,
up to 100 mg; J3130dtestosterone enanthate, up to 200
mg; J3140dtestosterone suspension, up to 50 mg; J3150d
testosterone propionate, up to 100 mg; and S0189d
testosterone pellet, 75 mg (table 1). The cumulative dose of
testosterone was assessed by summing the total number of
testosterone injections for the 5-year period.
Risk Factors
Sociodemographic characteristics examined included
age at diagnosis, marital status (currently married or
currently not married), race/ethnicity (nonHispanic white,
Descargado para Anonymous User (n/a) en ClinicalKey Espanol Colombia, Ecuador & Peru Flood Relief de ClinicalKey.es por Elsevier en julio 21, 2017.
Para uso personal exclusivamente. No se permiten otros usos sin autorizacin. Copyright 2017. Elsevier Inc. Todos los derechos reservados.
1613
Table 1. Intramuscular testosterone
Drug Name (dose) HCPCS Code
Testosterone cypionate: J1070
Up to 100 mg
1 cc, 200 mg J1080
Up to 50 mg J1090
Testosterone enanthate:
Up to 100 mg J3120
Up to 200 mg J3130
Testosterone suspension (up to 50 mg) J3140
Testosterone propionate (up to 100 mg) J3150
black, Hispanic American or other) and the percentage of
those living below the poverty line in the census tract.
Race/ethnicity was self-reported during the initial enroll-
ment with the Social Security Administration. We
included this variable in our analyses because cancer
treatment and outcomes may vary by race/ethnicity.
Clinical factors assessed included receipt of PSA tests and
number of visits with a primary care physician. Comor-
bidity was measured with an adaptation of the Charlson
comorbidity index22 using information from the diagnosis
codes in hospital and physician claims (inpatient and
outpatient) corresponding to a date of service in the year
before the cancer diagnosis.
Outcomes
The primary outcome of this study was high tumor grade
on biopsy specimen at diagnosis. As an additional marker
of high risk disease we also examined primary ADT,
defined as receiving at least 6 months of exclusive ADT in
the first 12 months after the prostate cancer diagnosis.
Tumor Grade Classification
Before 2003 cancer grade was captured in the SEER
database using certain categories, including lowd
Gleason 2-4, moderatedGleason 5-7 or highdGleason
8-10. In 2003 Gleason grade 7 was switched to the high
grade category. From 2004 and thereafter the SEER
database has captured specific Gleason score, allowing
classification into similar groupings for data prior to 2003.
Therefore, the high grade category was different in 2003
compared with the other study years. To examine the
potential impact of this classification system we per-
formed an analysis in which we removed all patients who
were diagnosed with prostate cancer in 2003.
Statistical Analysis
Differences in the proportion of study characteristics
among testosterone users and nonusers were assessed
using Pearson chi-square analysis. Logistic regression
models were then used to evaluate whether the likelihood
of being diagnosed with high grade disease varied as a
function of the cumulative testosterone dose. All analyses
were performed with SAS, version 9.1. This study was
reviewed and approved by the University of Texas Medi-
cal Branch institutional review board.
RESULTS
The supplementary table (http://jurology.com/) lists
demographic and clinical characteristics of testos-
terone therapy users and nonusers. The distribution
1614 TESTOSTERONE THERAPY AND RISK OF HIGH GRADE PROSTATE CANCER

of age, race, diagnosis year and marital status was therapy during a 5-year period was not associated
comparable between testosterone users and nonusers. with an aggressive form of the disease. Moreover
However, testosterone users had a higher degree of the risk of high grade prostate cancer did not in-
comorbid disease (total Charlson score 3 or greater), crease according to the total number of testosterone
a higher prevalence of conditions for which testos- injections. To our knowledge this is the first popu-
terone therapy was indicated (fatigue, hypogonadism, lation based study to determine whether the risk of
osteoporosis and erectile dysfunction), a greater
number of PSA tests in the 2 years before diagnosis
and a greater number of visits with the primary care Table 3. Adjusted OR of prostate tumor grade at diagnosis
associated with testosterone use
physician in the 2 years before diagnosis.
Table 2 lists the tumor grade distribution at OR (95% CI)
diagnosis according to testosterone therapy use and Testosterone therapy:
the total number of testosterone injections. Testos- No 1.00 (referent)
Yes 0.84 (0.67e1.05)
terone users were less likely to have high grade Age at prostate Ca diagnosis:
tumors at diagnosis. No consistent patterns of 70e74 1.00 (referent)
tumor grade at diagnosis were observed across the 75e79 1.16 (1.10e1.22)
80e84 1.54 (1.45e1.63)
cumulative dose categories. 85 or Greater 2.34 (2.17e2.51)
Table 3 shows the results of logistic regression Race:
predicting tumor grade at the diagnosis of high Hispanic 1.00 (referent)
White 1.15 (1.04e1.27)
grade prostate cancer. After adjusting for potential Black 1.21 (1.08e1.37)
confounders patients with prostate cancer who Other 1.14 (1.00e1.29)
received testosterone therapy in the 5 years before Diagnosis yr:
2005 1.00 (referent)
diagnosis were not at increased risk for high grade 2004 0.95 (0.89e1.02)
disease at diagnosis (OR 0.84, 95% CI 0.67e1.05). 2003* 2.40 (2.24e2.56)
We also performed multivariable analyses to assess 2002 1.20 (1.12e1.28)
2001 1.33 (1.24e1.42)
a dose-response association among testosterone Marital status:
users. The risk of high tumor grade did not increase Not married 1.00 (referent)
with an increasing number of testosterone in- Married 0.98 (0.94e1.02)
Education quartile (% less than high school):
jections (OR 1.00, 95% CI 0.99e1.01). Analyses 1 (less than 8.3) 1.00 (referent)
excluding patients diagnosed in 2003 yielded 2 (8.3eless than 14.7) 1.03 (0.97e1.10)
findings consistent with those of the overall 3 (14.7eless than 24.6) 1.05 (0.98e1.13)
4 (24.6 or greater) 1.03 (0.95e1.12)
analyses (receipt of testosterone therapy OR 0.86, Unknown 0.99 (0.79e1.24)
95% CI 0.66e1.12 and total number of testosterone Poverty quartile (% adults below poverty line):
injections OR 1.00, 95% CI 0.98e1.01). Because the 1 (less than 4.1) 1.00 (referent)
2 (4.1eless than 7.6) 0.98 (0.92e1.04)
sample size of men who received testosterone ther- 3 (7.6eless than 14.9) 1.02 (0.95e1.09)
apy was relatively small, we performed a power 4 (14.9 or greater) 1.08 (1.00e1.17)
analysis. Given that the incidence of high grade Comorbidity index
0 1.00 (referent)
prostate cancer was 20.1% among the 574 testos- 1 1.09 (1.04e1.15)
terone users and 27.0% among the 51,945 nonusers, 2 1.12 (1.03e1.21)
our combined sample had 99% statistical power 3 1.18 (1.07e1.31)
Indication:
to detect the observed association (OR 0.85) with a Hypogonadism 1.01 (0.87e1.17)
2-sided a level of 0.05. Fatigue 1.03 (0.99e1.08)
Erectile dysfunction 1.22 (1.15e1.33)
Psychosexual dysfunction 1.06 (0.95e1.19)
Osteoporosis 1.12 (1.03e1.22)
DISCUSSION Tumor stage at diagnosis:
In this population based study of 52,579 men diag- T1 1.00 (referent)
nosed with prostate cancer exposure to testosterone T2 1.52 (1.45e1.58)
T3 6.37 (5.61e7.33)
T4 9.73 (8.83e10.71)
Table 2. Clinical characteristics by tumor grade at diagnosis
No. PSA tests in 2 prior yrs:
No. Low/Moderately 0 1.00 (referent)
Differentiated (%) No. High Grade (%) 1 0.93 (0.88e0.98)
2 0.87 (0.80e0.95)
Testosterone user: 3 or Greater 0.95 (0.87e1.05)
Yes 458 (79.8) 116 (20.1) No. primary care physician visits in 2 prior yrs:
No 37,922 (73.0) 14,023 (27.0) 0 1.00 (referent)
Testosterone dose (No. injections): 1 0.92 (0.86e0.99)
1e2 212 (80.6) 51 (19.4) 2 0.91 (0.85e0.98)
3e7 105 (82.7) 22 (17.3) 3 or Greater 0.89 (0.85e0.94)
8 or Greater 141 (76.6) 43 (23.4)
* High grade category differed compared to other study years.

Descargado para Anonymous User (n/a) en ClinicalKey Espanol Colombia, Ecuador & Peru Flood Relief de ClinicalKey.es por Elsevier en julio 21, 2017.
Para uso personal exclusivamente. No se permiten otros usos sin autorizacin. Copyright 2017. Elsevier Inc. Todos los derechos reservados.
 TESTOSTERONE THERAPY AND RISK OF HIGH GRADE PROSTATE CANCER
high grade prostate cancer is associated with
testosterone exposure in an extended period and
whether the risk of high grade disease increases
with number of testosterone injections. Given the
slow growth of prostate cancer and the unknown
latency period associated with testosterone expo-
sure, this investigation offers novel and clinically
relevant information to clinicians, patients and the
general public.
To date only 1 population based study has been
done to examine whether exposure to testosterone
therapy increases the risk of high grade prostate
cancer. In a study of 149,354 men diagnosed with
prostate cancer from 1992 to 2007 Kaplan and Hu
found that up to 1 year of testosterone therapy was
not associated with more aggressive prostate cancer
at diagnosis.20 The similar finding in our study for a
5-year exposure period offers important additional
information on long-term exposure for men with
testosterone deficiency who are assessing the bene-
fits and risks of treatment. In addition our findings
are generally consistent with those of a number of
previous studies showing no significant increase in
prostate cancer incidence associated with testos-
terone therapy.3,5,18
Our study has several limitations. 1) Information
on risk factors was obtained from diagnosis codes
included in charges for outpatient and hospitaliza-
tion services. Such diagnoses are not always accu-
rate or complete.22 2) Medicare claims during the
study period provided no data on other formulations
of testosterone therapy (oral, transdermal patch or
gel). Therefore, we were unable to assess the extent
to which use of these formulations contributed to
the outcome. However, given that the prevalence of
these 3 formulations combined is estimated to be
less than 1% of the male population of the United
States during this period,1 it is unlikely that this
would have caused a high degree of misclassification
in our study cohort. 3) Prescription claims data do
not capture information on pharmaceutical agents
purchased outside the plan. Given the perceived
REFERENCES
1. Baillargeon J, Urban RJ, Ottenbacher KJ et al:
Trends in androgen prescribing in the United
States, 2001-2011. JAMA Intern Med 2013;
173: 1465.
4. Fernandez-Balsells MM, Murad MH, Lane M
2. Bhasin S, Cunningham GR, Hayes FJ et al:
Testosterone therapy in men with androgen
deficiency syndromes: an Endocrine Society
clinical practice guideline. J Clin Endocrinol
Metab 2010; 95: 2536.
5. Fenely MR and Carruthers M: Is testosterone
3. Calof OM, Singh AB, Lee ML et al: Adverse
events associated with testosterone re-
placement in middle-aged and older men:
Descargado para Anonymous User (n/a) en ClinicalKey Espanol Colombia, Ecuador & Peru Flood Relief de ClinicalKey.es por Elsevier en julio 21, 2017.
Para uso personal exclusivamente. No se permiten otros usos sin autorizacin. Copyright 2017. Elsevier Inc. Todos los derechos reservados.
1615
social stigma associated with receiving testosterone
therapy, some men may choose to seek treatment
outside their usual health care setting. 4) Because
we did not have access to patient laboratory
testosterone values before or after the initiation of
testosterone treatment, we could not determine
whether patients reached low, normal or supra-
physiological testosterone levels or to assess the
impact of such testosterone levels on subsequent
prostate cancer risk. 5) Because our study was
restricted to a sample of patients with prostate
cancer, we could not make any inferences about
whether testosterone therapy was associated with
an increased risk of prostate cancer in general. This
finding could have increased the absolute risk of
high grade prostate cancer among testosterone
users.
Finally given the retrospective nature of this
study, it is possible that undetected selection bias
may have affected the findings. For example, older
men who received testosterone therapy may have
been more likely than their peers to have engaged in
positive health behaviors (eg diet and exercise).
However, our statistical adjustment for sociodemo-
graphic behaviors (race, education and income) and
preventive behaviors (PSA testing) may help
address some of these biases.
Despite these limitations we believe that this
study has important strengths, including a large
sample size, long followup and inclusion of a clini-
cally and socioeconomically diverse cohort.
CONCLUSIONS
In view of the large increase in the use of testos-
terone therapy in recent years1 examining the
short-term and long-term risks of testosterone
therapy holds increasing clinical and public health
relevance. Future research, particularly large-scale
randomized clinical trials, should be performed to
more definitively examine the association of testos-
terone therapy and prostate cancer.
a meta-analysis of randomized, placebo- 6. Fowler JE and Whitmore WF: The response of
controlled trials. J Gerontol A Biol Sci Med metastatic adenocarcinoma of the prostate to
Sci 2005; 60: 1451. exogenous testosterone. J Urol 1981; 126: 372.
et al: Adverse effects of testosterone therapy 7. Grech A, Breck J and Heidelbaugh J: Adverse
in adult men: a systematic review and meta- effects of testosterone replacement therapy: an
analysis. J Clin Endocrinol Metab 2010; update on the evidence and controversy. Ther
95: 2560. Adv Drug Saf 2014; 5: 190.
treatment good for the prostate? Study of safety 8. Klotz L: Testosterone therapy and prostate
during long-term treatment. J Sex Med 2012; cancer-safety concerns are well founded. Nat
9: 2138. Rev Urol 2015; 12: 48.
1616 TESTOSTERONE THERAPY AND RISK OF HIGH GRADE PROSTATE CANCER
9. Morgentaler A: Testosterone replacement ther-
apy and prostate cancer. Urol Clin North Am
2007; 34: 555.
10. Morgentaler A: Testosterone and prostate
cancer: an historical perspective on a modern
myth. Eur Urol 2006; 50: 935.
11. Morgentaler A: Testosterone and prostate
cancer: what are the risks for middle-aged men?
Urol Clin North Am 2011; 38: 119.
12. Rhoden EL and Morgentaler A: Risks of
testosterone-replacement therapy and recom-
mendations for monitoring. N Engl J Med 2004;
350: 482.
13. Coward R, Simhan J and Carson C: Prostate
specific antigen changes and prostate cancer in
hypogonadal men treated with testosterone
replacement therapy. BJU Int 2008; 103: 1179.
Descargado para Anonymous User (n/a) en ClinicalKey Espanol Colombia, Ecuador & Peru Flood Relief de ClinicalKey.es por Elsevier en julio 21, 2017.
Para uso personal exclusivamente. No se permiten otros usos sin autorizacin. Copyright 2017. Elsevier Inc. Todos los derechos reservados.
14. Spitzer M, Huang G, Basaria S et al: Risks and
benefits of testosterone therapy in older men.
Nat Rev Endocrinol 2013; 9: 414.
15. Barqawi A and Crawford ED: Testosterone
replacement therapy and the risk of prostate
cancer: is there a link? Int J Impot Res 2006;
18: 323.
16. Gould DC and Kirby RS: Testosterone replace-
ment therapy for late onset hypogonadism: what
is the risk of inducing prostate cancer. Prostate
Cancer Prostatic Dis 2006; 9: 14.
17. Huggins C and Hodges CV: Studies on prostate
cancer I. The effect of castration, of estrogen
and of androgen injection on serum phospha-
tases in metastatic carcinoma of the prostate.
Cancer Res 1941; 1: 293.
18. Wang C, Cunningham G, Dobs A et al: Long-term
testosterone gel treatment maintains beneficial
effects on sexual function and mood, lean and
fat mass, and bone mineral density in hypo-
gonadal men. J Clin Endocrinol Metab 2004;
89: 2085.
19. Shabsigh R, Crawford ED, Nehra A et al:
Testosterone therapy in hypogonadal men and
potential prostate cancer risk: a systematic re-
view. Int J Impot Res 2009; 21: 9.
20. Kaplan AL and Hu JC: Use of testosterone
replacement therapy in the United States and its
effect on subsequent prostate cancer outcomes.
Urology 2013; 82: 321.
21. Warren JL, Klabunde CN, Schrag DS et al:
Overview of the SEER-Medicare data: content,
research applications, and generalizability to the
United States elderly population. Med Care
2002; 40: IV-3.
22. Klabunde CN, Warren JL and Legler JM:
Assessing comorbidity using claims data: an
overview. Med Care 2002; 40: IV-26.