Asian Journal of Andrology (2012), 1–3ß 2012 AJA, SIMM & SJTU. All rights reserved 1008-682X/12 $32.00
Incidence rate of prostate cancer in men treated forerectile dysfunction with phosphodiesterase type 5inhibitors: retrospective analysis
Anthony H Chavez1, K Scott Coffield1, M Hasan Rajab2,3 and Chanhee Jo2
The purpose of this study was to determine the incidence rate of prostate cancer among men with erectile dysfunction (ED) treated withphosphodiesterase type 5 inhibitors (PDE-5i) over a 7-year period vs. men with ED of the same age and with similar risk factors who werenot treated with PDE-5i. In a retrospective review of electronic medical records and billing databases between the years 2000 and2006, men with ED between the ages of 50 and 69 years and no history of prostate cancer prior to 2000 were identified. Theseindividuals were divided into two groups: 2362 men who had treatment with PDE-5i, and 2612 men who did not have treatment. Demographic data in each group were compared. During the study period, 97 (4.1%) men with ED treated with PDE-5i were diagnosedwith prostate cancer compared with 258 (9.9%) men with ED in the non-treated group (P,00001). A higher percentage of AfricanAmericans were treated with PDE-5i vs. those who were not (10.5% vs. 7.1%; P,0.0001). The PDE-5i group had lower documenteddiagnosis of elevated prostate-specific antigen (10.0% vs. 13.1%; P50.0008) and higher percentage of benign prostatic hyperplasia(38.4% vs. 35.1%; P50.0149). Men with ED treated with PDE-5i tended to have less chance (adjusted odds ratio: 0.4; 95%confidence intervals: 0.3–0.5; P,0.0001) of having prostate cancer. Our data suggest that men with ED treated with PDE-5i tended tohave less of a chance of being diagnosed with prostate cancer. Further research is warranted. Asian Journal of Andrology advance online publication, ** ** 2012; doi:10.1038/aja.2012.162
Keywords: erectile dysfunction (ED); phosphodiesterase 5 inhibitors (PDE-5i); prostate cancer; prostatic neoplasms
classification of disease (ICD9) code, we identified men between the
Several recent studies have suggested an association between sexual
ages of 50 and 69 diagnosed with ED with no history of prostate cancer
activity and risk of prostate cancer in older men.1,2 Some have pro-
prior to 2000. Age was determined at the date of enrollment. Study
posed that increased sexual activity was an indicator of increased
participants were divided into two groups: those who had received
androgenic activity and thus a higher risk for prostate cancer.3
treatment with PDE-5i and those who had not. PDE-5i included tada-
However, Leitzmann et al.,2 in a large prospective study, found that
high ejaculation frequency was related to decreased risk of total and
Prostate cancer was confirmed by prostate biopsy with histological
organ-confined prostate cancer. One theory is that infrequent ejacu-
diagnosis of prostate adenocarcinoma documented in the EMR.
lation leads to retained carcinogenic secretions in the prostatic acini.4
Demographics and baseline characteristics including age, ethnicity,
Erectile dysfunction (ED) is defined as ‘the inability to attain and/or
benign prostatic hyperplasia (BPH) and diagnosis of elevated pro-
maintain penile erection sufficient for satisfactory sexual perfor-
state-specific antigen (PSA) (.4.0 ng dl21) were compared between
mance’. In patients with ED, ejaculatory frequency is expected to be
the two study groups. BPH was diagnosed by the physician docu-
decreased.5 It would be valuable to assess if treating men with ED with
mentation in the EMR, or by current procedural terminology
phosphodiesterase type 5 inhibitors (PDE-5i) has any effect on the
that the physician submitted for billing. As in actual practice, this
incidence of prostate cancer based on Leitzmann’s findings. Our hypo-
was a clinical diagnosis, not based on pathology or specific objective
thesis was that the incidence rate of prostate cancer would be higher in
the PDE-5i-treated group due to detection bias.
Inclusion criteria for the PDE-5i treatment group required the me-
dication documented in all encounters over an one-year minimum per-
iod. Due to the limitations of the retrospective design, more strict criteria
This was an institutional review board-approved retrospective review
for determining inclusion criteria of the PDE-5i group were not feasible.
of electronic medical records (EMR) and billing databases at our in-
For categorical variables, a chi-square test was used to compare the
stitution between the years 2000 and 2006. Using the international
patients with and without PDE-5i. Continuous variables were first
1Departments of Urology, Scott & White Healthcare, Temple, Texas 76508, USA; 2Departments of Biostatistics, Scott & White Healthcare, Temple, Texas 76508, USA and3Deparment of Biostatistics, College of Medicine, Alfaisal University, Riyadh 11533, Saudi ArabiaCorrespondence: Dr KS Coffield (firstname.lastname@example.org)Received: 30 March 2012; Revised: 6 September 2012; Accepted: 11 December 2012
Table 1 General patient characteristics by PDE-5i treatment status
Abbreviations: BPH, benign prostatic hyperplasia; PDE-5i, phosphodiesterase type 5 inhibitors; PSA, prostate-specific antigen.
*P value for African American’s only.
examined for normality and then compared using two-sample t-test
for independent samples or Wilcoxon non-parametric two-sample
In this retrospective study among men with ED, the use of PDE-5i was
test, as appropriate. We used multiple logistic regression in the ana-
associated with a decreased incidence rate of prostate cancer. The use
lysis of the study, having prostate cancer between 2000 and 2006. A P
of PDE-5i, PSA and advanced age were significant predictors of pro-
value of less than 0.05 indicated a statistical significance and SAS
state cancer. Interestingly, there were more African-American men
version 9.2 (SAS Institute Inc., Cary, NC, USA) was used for the
in the PDE-5i treatment group, predisposing these patients to an
increased risk of prostate cancer. The number of patient prostatebiopsies varied with PSA elevation and abnormal digital prostate
exam. We were unable from this database to determine whether the
A total of 4974 men were included in the study, 2362 (47.5%) of
patients in the treatment group who had BPH had more severe symp-
whom had a record of being prescribed a PDE-5i. Mean age was
toms of BPH prior to PDE-5i treatment.
61.264.85 years, ranging from 50 to 69 years. Sixty-nine percent were
It is clear from the more recent literature that BPH is common and
Caucasians, 8.8% were Africa Americans, 6.6% were Hispanic and
that the PDE-5i treatment has been acknowledged as an acceptable
15.7% were classified as others. Mean ages for the treated (PDE-5i)
alternative in this frequently diagnosed disease.6 There is no preexist-
and non-treated groups were 61.064.8 vs. 61.464.9 years, respectively
ing data to suggest that PDE-5i treatment increases BPH risk and these
(P50.0044). There was a higher percentage of African Americans in
data do not suggest that. Although Krain3 proposed that increased
the treated (PDE-5i) group (10.5% vs. 7.1%; P,0.0001) (Table 1).
sexual activity was an indicator of increased androgenic activity and
Fewer patients in the PDE-5i group had a documented diagnosis of
thus a higher risk for prostate cancer, our results showed that the risk
elevated PSA (10.0% vs. 13.1%; P50.0008). Prostate biopsies were
for prostate cancer decreased in those who used PDE-5i.
performed in patients with elevated PSA and abnormal digital prostate
Several theories may explain the potential association between
exams. There was a higher percentage of patients with BPH in the
PDE-5i and decreased risk of prostate cancer. Patients with ED treated
PDE-5i treatment group (38.4% vs. 35.1%; P50.0149). Out of the
with PDE-5i are expected to have a higher ejaculation frequency than
2362 men in the treated (PDE-5i) group, 97 (4.1%) had a record of
those who are untreated. Leitzmann et al.’s large prospective cohort
prostate cancer vs. 258 (9.9%) out of the 2612 in the non-treated group
study2 demonstrated a protective effect of increased ejaculation fre-
(P,0.0001). In the multiple logistic regression analysis, use of PDE-5i,
quency with regard to prostate cancer diagnosis. Furthermore, there
age and elevated PSA were significant predictors of prostate cancer
may be a protective effect from the physiological and molecular acti-
(Table 2). On the other hand, race and BPH were not significant
vity of the PDE-5i. Zenzmaier et al.7 studied the in vitro effects of PDE-
predictors of prostate cancer though BPH was on the borderline
5i on prostate cells. They reported that tadalafil at 2.5 mmoll21 had
pronounced effect on proliferation of prostate stromal cells though
Men with ED who were treated with PDE-5i tended to have less
required 25 mmoll21 concentration that notably reduced proliferation
chance (adjusted odds ratio: 0.4; 95% confidence intervals: 0.3–0.5;
of prostate epithelial cells.7 Zenzmaier et al.7 confirmed that PDE-5
P,0.0001) of having prostate cancer. Treatment univariate and mul-
was highly expressed in the stromal compartment of the prostate but
tiple logistic regression analyses are summarized in Tables 1–3,
not the epithelium. Hypoxia is common in prostate tumors, promot-
Table 3 Logistic regression analysis on incidence of prostate cancer
Table 2 Logistic regression analysis on incidence of prostate cancer
Abbreviations: BPH, benign prostatic hyperplasia; PDE-5i, phosphodiesterase type
Abbreviations: BPH, benign prostatic hyperplasia; PDE-5i, phosphodiesterase type
5 inhibitors; PSA, prostate-specific antigen.
5 inhibitors; PSA, prostate-specific antigen.
Prostate cancer and PDE-5 inhibitorsAH Chavez et al
ing tumor progression.8 There is evidence that hypoxia plays a pivotal
role in cancer adaptive responses leading to poor clinical outcomes for
Our data suggest that men with ED who were treated with PDE-5i had
patients with prostate cancer.9 Khandrika et al.10 recently presented
a reduced risk of being diagnosed with prostate cancer. We believe that
data demonstrating the effects of intermittent hypoxia on prostate
cancer. These data suggest that intermittent hypoxia-reoxygenationpromotes an aggressive phenotype of prostate cancer via hypoxia-
induced increased androgen receptor activity, increased androgen
AHC conceived the study, participated in the study design and drafted
receptor responsive promoter activity independent of androgens
the manuscript. KSC coconceived the study, participated in the study
and increased hypoxia-inducible factor-1a levels and activity.10 It is
design and revised the manuscript. MHR participated in the study
well known that PDE-5i facilitate circulatory perfusion and reduce
design, revised the manuscript and performed statistical analysis. CJ
hypoxia by decreasing the degradation of cyclic guanosine monopho-
reviewed the statistical analysis of the data. All authors read and
sphate.11 PDE-5i may alter the development and progression of pro-
state cancer by decreasing intermittent hypoxia-reoxygenation effectand prevention of the formation of a tumor-reactive stroma.
Hamilton et al.12 reported the first study to characterize cGMP- and
All authors declare that there are no competing financial interests.
cAMP-dependent PDE activity in prostate cancer cell lines. Flow cyto-metry was used to determine the effects of PDE inhibition on surface
< major histocompatibility complex class I-related chain A, a natural
killer cell-activating ligand in a mouse model to evaluate the in vivoeffects of PDE inhibition on the growth of human prostate cancer
Giles GG, Severi G, English DR, McCredie MR, Borland R et al. Sexual factors and
cells.12 PDE (PDE5 and PDE11) inhibition completely reversed hyp-
prostate cancer. BJU Int 2003; 92: 211–6.
oxia-induced shedding of the immune-stimulatory molecule, major
Leitzmann MF, Platz EA, Stampfer MJ, Willett WC, Giovannucci E. Ejaculation
histocompatibility complex class I-related chain A, and attenuated the
frequency and subsequent risk of prostate cancer. JAMA 2004; 291: 1578–86.
Krain LS. Some epidemiologic variables in prostatic carcinoma in California. Prev Med
growth of human prostate tumor xenografts (Du-145 cells) in a natu-
ral killer cell-competent murine model.
Isaacs JT. Prostatic structure and function in relation to the etiology of prostatecancer. Prostate 1983; 4: 351–66.
The progression of prostate cancer is associated with malignant
Steidle CP, McCullough AR, Kaminetsky JC, Crowley AR, Siegel RL et al. Early
cellular phenotype characterized by increased tumor cell invasion,
Sildenafil dose optimization and personalized instruction improves the frequency,
metastasis and escape from immune destruction. The acquisition of
flexibility, and success of sexual intercourse in men with erectile dysfunction. Int JImpot Res 2007; 19: 154–60.
these malignant traits is influenced by genetic and epigenetic factors,
Egerdie RB, Auerbach S, Roehrborn CG, Costa P, Garza MS et al. Tadalafil 2.5 or 5 mg
but also by microenvironmental stressors such as hypoxia.12 Though
administered once daily for 12 weeks in men with both erectile dysfunction and signsand symptoms of benign prostatic hyperplasia: results of a randomized, placebo-
the relationship between PDE-5i therapy and development and pro-
controlled, double-blind study. J Sex Med 2012; 9: 271–81.
gression of prostate cancer remains to be determined, the use of PDE-5i
Zenzmaier C, Sampson N, Pernkopf D, Plas E, Untergasser G et al. Attenuated
therapy may play a role in this commonly occurring cancer.
proliferation and trans-differentiation of prostatic stromal cells indicate suitabilityof phosphodiesterase type 5 inhibitors for prevention and treatment of benign
To our knowledge, this is the first report seeking an association
prostatic hyperplasia. Endocrinology 2010; 151: 3975–84.
between prostate cancer and PDE-5i therapy. This study includes
Rudolfsson SH, Bergh A. Hypoxia drives prostate tumour progression and impairs theeffectiveness of therapy, but can also promote cell death and serve as a therapeutic
the limitations associated with its single-center location, retrospective
target. Expert Opin Ther Targets 2009; 13: 219–25.
design, potential variation in physician documentation in the EMR,
Movsas B, Chapman JD, Greenberg RE, Hanlon AL, Horwitz EM et al. Increasing levels
lack of strict criteria for prescribed PDE-5i medication and lack of
of hypoxia in prostate carcinoma correlate significantly with increasing clinical stageand patient age: an Eppendorf pO(2) study. Cancer 2000; 89: 2018–24.
family history documentation. The limitation of subjects with varying
Khandrika L, Lieberman R, Koul S, Kumar B, Meacham R et al. Hypoxia-reoxygenation
lengths of exposure to PDE-5i therapy, because of entering the study at
promotes aggressive phenotype in prostate cancer cells. (Abstract) J Urol 2008; 179(Suppl): 192.
different times during the 7-year period of data collection, is acknow-
Corbin JD, Francis SH. Molecular biology and pharmacology of PDE-5-inhibitor
ledged. This limitation and others are best addressed by a prospective
therapy for erectile dysfunction. J Androl 2003; 24: 538–41.
trial. However, data from this study suggest that an association
Hamilton TK, Hu N, Kolomitro K, Bell EN, Maurice DH et al. Potential therapeuticapplications of phosphodiesterase inhibition in prostate cancer. World J Urol; e-pub
between the rate of incidence of prostate cancer and PDE-5i may exist.
ahead of print 2 March 2012; doi:10.1007/s00345-012-0848-7.
Journal: Asian Journal of AndrologyPaper: aja2012162Title: Incidence rate of prostate cancer in men treated for erectile dysfunction with phosphodiesterase type 5 inhibitors:retrospective analysis
Please confirm the full form of CPT.
Please confirm the full form of MHC.
Please provide acknowledgments section.
PRODUCT BRIEF DEVELOPMENT S.W.O.T. Analysis In a few words: If you know your strengths and weaknesses and understand the opportunities and threats you have, then you can do something about them. In its simplest form, a SWOT analysis can be understood as the examination of an organization's internal strengths and weaknesses, and its environments opportunities, and threats. I