Background: In a case–control study, aspirin use was associated with a lower risk of a common prostate cancer molecular subtype, the TMPRSS2:ERG gene fusion. We sought to validate this finding in a prospective cohort.

Methods: In the Health Professionals Follow-up Study, 49,395 men reported on aspirin use on biennial questionnaires and were followed for prostate cancer incidence over 23 years. TMPRSS2:ERG status was assessed by IHC for presence of ERG on archival tumor specimens for 912 patients with prostate cancer, of whom 48% were ERG-positive.

Results: In multivariable models, we found no association between regular use of aspirin and risk of ERG-positive prostate cancer (HR, 1.02; 95% confidence interval, 0.85–1.23), nor any association with duration or frequency of aspirin use. In restricting to cases with either high Gleason grade or advanced stage disease, there remained no association with aspirin use.

Conclusions: Data from this prospective study with repeated assessments of aspirin use do not support the hypothesis that aspirin use is associated with a lower risk of ERG-positive prostate cancer.

Impact: Aspirin use is unlikely to lower the risk of this common molecular subtype of prostate cancer. However, there is emerging data supporting the role of other lifestyle and genetic factors underlying the development of the TMPRSS2:ERG fusion. Cancer Epidemiol Biomarkers Prev; 27(10); 1231–3. ©2018 AACR.

The TMPRSS2:ERG gene fusion is the most common somatic event in primary prostate cancer, with an estimated 100,000 U.S. patients diagnosed with TMPRSS2:ERG-positive cancer annually. Our group and others have reported on associations between lifestyle and inherited genetic factors specifically associated with TMPRSS2:ERG-defined disease (1–4). In a retrospective case–control study, current aspirin use was associated with a lower risk of TMPRSS2:ERG-positive cancer [OR, 0.63; 95% confidence interval (CI), 0.43–0.93], whereas there was no association with cancers that lacked TMPRSS2:ERG (OR, 0.99; 95% CI, 0.69–1.42; ref. 5). The authors speculated that aspirin may protect against TMPRSS2:ERG-positive cancer through reduction in cellular stress, inflammation, and DNA damage. We sought to validate this association in a prospective cohort of men with longitudinal measures of aspirin use and 23 years of follow-up for prostate cancer incidence.

This study was nested in the Health Professionals Follow-up Study (HPFS), a cohort of 51,529 male health professionals age 40 to 75 years at baseline in 1986 (6). For this study, we excluded men with cancer diagnoses other than nonmelanoma skin cancer before 1986 (n = 2,076), missing age or diagnosis date (n = 42), and implausible diagnosis or death dates (n = 16). Participants responded to biennial questionnaires on lifestyle, diet (every 4 years), diagnoses, and medication use. Biennial follow-up rates exceeded 93%. Patients with incident prostate cancer were followed with specific questionnaires. Clinical data were abstracted from medical records and pathology reports.

On biennial questionnaires, participants reported current regular aspirin use (with example brand names provided). If participants did not return a specific questionnaire, their prior response was carried forward. Starting in 1992, men reported categories of frequency of use, and we defined regular use as ≥2 days/week.

We characterized TMPRSS2:ERG status on tumor tissue microarrays from men who underwent radical prostatectomy (n = 912) using a genomically validated immunohistochemistry method for the ERG protein (7). A case was scored ERG-positive if at least one core had positive ERG staining within cancer cells.

Cox proportional hazards models, adjusted for predefined covariates, were used to estimate HRs and two-sided 95% CIs for total, advanced (stage, ≥T3b/N1, or M1 at any time), and high-grade (Gleason grade, ≥4 + 3) cancers, each according to ERG status.

Among 49,395 men, 14,547 (29.4%) were current aspirin users at baseline in 1986. In 2008, 47.2% and 36.3% of the remaining 28,355 participants were current and past aspirin users, respectively. A total of 6,189 participants (12.5%) were diagnosed with incident prostate cancer (Table 1). From 2,332 patients treated with prostatectomy, ERG status was available for 912 tumors.

Table 1.

Characteristics of participants of the HPFS by aspirin use at baseline, standardized to the age distribution of the study population

Baseline characteristics, 1986Nonusers of aspirinCurrent users of aspirin
N 34,848 14,547 
Frequency of aspirin use, mean (SD; d/mo) 0.0 8.6 (0.0)a 
Age, mean (SD; y)b 53.7 (9.7) 56.5 (9.7) 
BMI, mean (SD; kg/m225.5 (3.3) 25.7 (3.5) 
Family history of prostate cancer 12.0% 11.8% 
Smoking status 
 Never smoker 45.7% 41.8% 
 Past smoker 39.6% 47.3% 
 Current smoker 9.3% 10.4% 
 Missing 5.4% 0.5% 
Total physical activity, mean (SD; METS-h/wk) 18.8 (26.5) 18.6 (26.1) 
Diabetes diagnosis 3.0% 3.6% 
Cumulative incidence, by 2009   
 Prostate cancer diagnosis 12.5% 12.7% 
 Prostate cancer death 1.3% 1.1% 
 Overall mortality 27.7% 30.8% 
Baseline characteristics, 1986Nonusers of aspirinCurrent users of aspirin
N 34,848 14,547 
Frequency of aspirin use, mean (SD; d/mo) 0.0 8.6 (0.0)a 
Age, mean (SD; y)b 53.7 (9.7) 56.5 (9.7) 
BMI, mean (SD; kg/m225.5 (3.3) 25.7 (3.5) 
Family history of prostate cancer 12.0% 11.8% 
Smoking status 
 Never smoker 45.7% 41.8% 
 Past smoker 39.6% 47.3% 
 Current smoker 9.3% 10.4% 
 Missing 5.4% 0.5% 
Total physical activity, mean (SD; METS-h/wk) 18.8 (26.5) 18.6 (26.1) 
Diabetes diagnosis 3.0% 3.6% 
Cumulative incidence, by 2009   
 Prostate cancer diagnosis 12.5% 12.7% 
 Prostate cancer death 1.3% 1.1% 
 Overall mortality 27.7% 30.8% 

Abbreviations: BMI, body mass index; METS, metabolic equivalent tasks.

aThis increased to 26.3 days/month (mean; SD, 4.6) among current aspirin users in 2010.

bNot adjusted for age.

There was no statistically significant association between current regular aspirin use and the risk of ERG-positive (HR, 1.02; 95% CI, 0.85–1.23) or ERG-negative prostate cancer (HR, 1.09; 95% CI, 0.91–1.30; Pheterogeneity = 0.69 by ERG status), nor for total prostate cancer including all cases (HR, 1.05; 95% CI, 0.99–1.10) in fully adjusted models. Dose–response analyses according to cumulative duration or frequency of aspirin use were null (Table 2). Results for ERG-positive cancer were also null for age-adjusted models for advanced and high-grade cancer.

Table 2.

ERG-positive and ERG-negative incident prostate cancer by aspirin use (fully adjusted modela)

No. of casesHR (95% CI)
ERG-positiveERG-negative
Total439473ERG-positiveERG-negative
Categories of use 
 Never user 147 138 1 (ref) 1 (ref) 
 Past user 107 114 1.02 (0.79–1.31) 1.04 (0.82–1.33) 
 Current user 185 221 1.03 (0.83–1.28) 1.11 (0.89–1.37) 
    Pheterogeneity = 0.88 
Current use 
 Never/past user 254 252 1 (ref) 1 (ref) 
 Current user 185 221 1.02 (0.85–1.23) 1.09 (0.91–1.30) 
    Pheterogeneity = 0.63 
Ever use 
 Never user 147 138 1 (ref) 1 (ref) 
 Ever user 292 335 1.03 (0.84–1.25) 1.08 (0.89–1.33) 
    Pheterogeneity = 0.69 
Duration of use since baseline 
 Non-aspirin user 147 138 1 (ref) 1 (ref) 
 Aspirin use <5 years 134 149 1.00 (0.79–1.26) 1.17 (0.94–1.46) 
 Aspirin use 5–<10 years 90 92 1.11 (0.85–1.44) 0.98 (0.76–1.28) 
 Aspirin use 10 years+ 68 94 0.96 (0.71–1.30) 1.03 (0.77–1.37) 
    Pheterogeneity = 0.48 
 Per year of use 439 473 1.00 (1.00–1.00) 1.00 (1.00–1.00) 
    Pheterogeneity = 0.62 
Frequency of use 
 Aspirin use <2 d/wk (never/past user) 254 252 1 (ref) 1 (ref) 
 Aspirin use 2–<6 d/wk 72 81 0.94 (0.73–1.21) 1.05 (0.83–1.33) 
 Aspirin use 6+ d/wk 113 140 1.09 (0.87–1.37) 1.12 (0.90–1.39) 
    Pheterogeneity = 0.82 
 Per d/wk of use 439 473 1.02 (0.98–1.05) 1.01 (0.98–1.05) 
    Pheterogeneity = 0.93 
No. of casesHR (95% CI)
ERG-positiveERG-negative
Total439473ERG-positiveERG-negative
Categories of use 
 Never user 147 138 1 (ref) 1 (ref) 
 Past user 107 114 1.02 (0.79–1.31) 1.04 (0.82–1.33) 
 Current user 185 221 1.03 (0.83–1.28) 1.11 (0.89–1.37) 
    Pheterogeneity = 0.88 
Current use 
 Never/past user 254 252 1 (ref) 1 (ref) 
 Current user 185 221 1.02 (0.85–1.23) 1.09 (0.91–1.30) 
    Pheterogeneity = 0.63 
Ever use 
 Never user 147 138 1 (ref) 1 (ref) 
 Ever user 292 335 1.03 (0.84–1.25) 1.08 (0.89–1.33) 
    Pheterogeneity = 0.69 
Duration of use since baseline 
 Non-aspirin user 147 138 1 (ref) 1 (ref) 
 Aspirin use <5 years 134 149 1.00 (0.79–1.26) 1.17 (0.94–1.46) 
 Aspirin use 5–<10 years 90 92 1.11 (0.85–1.44) 0.98 (0.76–1.28) 
 Aspirin use 10 years+ 68 94 0.96 (0.71–1.30) 1.03 (0.77–1.37) 
    Pheterogeneity = 0.48 
 Per year of use 439 473 1.00 (1.00–1.00) 1.00 (1.00–1.00) 
    Pheterogeneity = 0.62 
Frequency of use 
 Aspirin use <2 d/wk (never/past user) 254 252 1 (ref) 1 (ref) 
 Aspirin use 2–<6 d/wk 72 81 0.94 (0.73–1.21) 1.05 (0.83–1.33) 
 Aspirin use 6+ d/wk 113 140 1.09 (0.87–1.37) 1.12 (0.90–1.39) 
    Pheterogeneity = 0.82 
 Per d/wk of use 439 473 1.02 (0.98–1.05) 1.01 (0.98–1.05) 
    Pheterogeneity = 0.93 

Abbreviation: ref, reference category.

aAdjusted for age, calendar time, race (Caucasian, other), family history of prostate cancer in father or brother (yes, no), height (≤68, >68–70, >70–72, >72 inches), body mass index (<21, 21–<25, 25–<30, 30+ kg/m2), body mass index at age 21 years (<20, 21–<25, 25–<30, 30+ kg/m2), physical activity (quintiles of metabolic equivalents-hours/week), smoking (never, former/quit >10 years ago, former/quit ≤10 years ago, current), history of diabetes (yes, no), time-varying current statin use (yes, no), PSA testing in the 2 years prior to the questionnaire date (yes, no; lagged by one period to avoid counting diagnostic PSA tests as screening), and PSA testing in >50% of possible time periods (yes, no; lagged by one period to avoid counting diagnostic PSA tests as screening).

In this prospective study with updated information on aspirin, we found no association between regular aspirin use and risk of ERG-positive prostate cancer. Similarly, we found no association between duration or frequency of aspirin use and ERG-positive disease, including for clinically significant high-risk cancers. Our findings are in contrast with those of Wright and colleagues’ (5), who reported a strong inverse association in their case–control study, which included 346 cases (49% ERG-positive) and 942 controls.

Differences in results may partly be due to differences in study design. First, our study was nested in a prospective cohort, whereas the prior study collected data from cases after diagnosis and used random digit dialing to select controls free from prostate cancer. Second, genetic and environmental factors are associated with ERG status (1–4) and could lead to confounding if not controlled for (5). In our study population, however, adjusted and unadjusted estimates were nearly identical. Third, misclassification of aspirin exposure is expected to be nondifferential in HPFS, where medical professionals repeatedly reported on medication use before cancer diagnosis. Recall bias in the prior study cannot account for differences in risk according to ERG status, which was unknown to participants. It is unlikely that ERG assessment via IHC or FISH would have biased either study's result (7). Finally, differences in results may be due to chance. Our study had >99% power to detect an HR of 0.63, corresponding to the previously reported effect size (5).

In summary, our data do not support the hypothesis that aspirin use lowers the risk of TMPRSS2:ERG-positive prostate cancer. Emerging data suggest other modifiable etiologic and prognostic factors for this common molecular subtype (1–4, 8).

No potential conflicts of interest were disclosed.

Conception and design: M.K. Downer, A. Pettersson, L.A. Mucci

Development of methodology: M.K. Downer, S. Finn, R.E. Graff

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): S. Finn, R.T. Lis, L.A. Mucci

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): K.H. Stopsack, A.G. Gonzalez-Feliciano, M.K. Downer, A. Pettersson, M. Loda, P.W. Kantoff, T.U. Ahearn, L.A. Mucci

Writing, review, and/or revision of the manuscript: K.H. Stopsack, A.G. Gonzalez-Feliciano, S.F. Peisch, M.K. Downer, R.A. Gage, S. Finn, R.T. Lis, R.E. Graff, A. Pettersson, C.H. Pernar, M. Loda, P.W. Kantoff, T.U. Ahearn, L.A. Mucci

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): S.F. Peisch, M.K. Downer, R.T. Lis

Study supervision: L.A. Mucci

We would like to thank the participants and staff of the Health Professionals Follow-up Study for their valuable contributions as well as the following state cancer registries for their help: AL, AZ, AR, CA, CO, CT, DE, FL, GA, ID, IL, IN, IA, KY, LA, ME, MD, MA, MI, NE, NH, NJ, NY, NC, ND, OH, OK, OR, PA, RI, SC, TN, TX, VA, WA, WY. The authors assume full responsibility for analyses and interpretation of these data. In particular, we would like to recognize the contributions of Liza Gazeeva, Siobhan Saint-Surin, Robert Sheahan, and Betsy Frost-Hawes. The Health Professionals Follow-up Study was supported by NIH grant U01 CA167552. This research was funded in part by the Dana-Farber/Harvard Cancer Center Specialized Programs of Research Excellence program in Prostate Cancer (5P50 CA090381), the NCI (R01CA136578 to L.A. Mucci; T32CA09001 to M.K. Downer and C.H. Pernar; and R25CA112355 to R.E. Graff), and the NIH/NCI Cancer Center Support Grants P30 CA008748 and P30 CA06516. K.H. Stopsack, S. Finn, and L.A. Mucci are Prostate Cancer Foundation Young Investigators.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Graff
RE
,
Meisner
A
,
Ahearn
TU
,
Fiorentino
M
,
Loda
M
,
Giovannucci
EL
, et al
Pre-diagnostic circulating sex hormone levels and risk of prostate cancer by ERG tumour protein expression
.
Br J Cancer
2016
;
114
:
939
44
.
2.
Egbers
L
,
Luedeke
M
,
Rinckleb
A
,
Kolb
S
,
Wright
JL
,
Maier
C
, et al
Obesity and prostate cancer risk according to tumor TMPRSS2:ERG gene fusion status
.
Am J Epidemiol
2015
;
181
:
706
13
.
3.
Graff
RE
,
Pettersson
A
,
Lis
RT
,
Ahearn
TU
,
Markt
SC
,
Wilson
KM
, et al
Dietary lycopene intake and risk of prostate cancer defined by ERG protein expression
.
Am J Clin Nutr
2016
;
103
:
851
60
.
4.
Penney
KL
,
Pettersson
A
,
Shui
IM
,
Graff
RE
,
Kraft
P
,
Lis
RT
, et al
Association of prostate cancer risk variants with TMPRSS2:ERG status: evidence for distinct molecular subtypes
.
Cancer Epidemiol Biomarkers Prev
2016
;
25
:
745
9
.
5.
Wright
JL
,
Chery
L
,
Holt
S
,
Lin
DW
,
Luedeke
M
,
Rinckleb
AE
, et al
Aspirin and NSAID use in association with molecular subtypes of prostate cancer defined by TMPRSS2:ERG fusion status
.
Prostate Cancer Prostatic Dis
2016
;
19
:
53
6
.
6.
Giovannucci
E
,
Liu
Y
,
Platz
EA
,
Stampfer
MJ
,
Willett
WC
. 
Risk factors for prostate cancer incidence and progression in the health professionals follow-up study
.
Int J Cancer
2007
;
121
:
1571
8
.
7.
Pettersson
A
,
Graff
RE
,
Bauer
SR
,
Pitt
MJ
,
Lis
RT
,
Stack
EC
, et al
The TMPRSS2:ERG rearrangement, ERG expression, and prostate cancer outcomes: a cohort study and meta-analysis
.
Cancer Epidemiol Biomarkers Prev
2012
;
21
:
1497
509
.
8.
Pettersson
A
,
Lis
RT
,
Meisner
A
,
Flavin
R
,
Stack
EC
,
Fiorentino
M
, et al
Modification of the association between obesity and lethal prostate cancer by TMPRSS2:ERG
.
J Natl Cancer Inst
2013
;
105
:
1881
90
.