|Year : 2018 | Volume
| Issue : 2 | Page : 127-141
Nonsteroidal anti-inflammatory drugs and clinical outcomes among men with prostate cancer: A systematic review and meta-analysis
Divyesh Thakker1, Amit D Raval2, Nilixa Raval3, Ami Vyas4
1 Novartis Healthcare, Hyderabad, Telangana, India
2 Healthcore, Inc. Wilmington, Wilmington, Delaware, USA
3 Independent Researcher, Wilmington, Delaware, USA
4 College of Pharmacy, University of Rhode Island, Kingston, Rhode Island, USA
|Date of Web Publication||21-Mar-2018|
Dr. Amit D Raval
Healthcore Inc, 123 Justison Street, Suite 200, Wilmington DE-19801, Delaware
Source of Support: None, Conflict of Interest: None
Background: Nonsteroidal anti-inflammatory drugs (NSAIDs) have shown properties of inhibiting the progression of prostate cancer (PCa) in preclinical studies. However, epidemiological studies yield mixed results on the effectiveness of NSAIDs in PCa. Objective: The objective of this study was to determine the effect of NSAID use on clinical outcomes in PCa using systematic review and meta-analysis. Methods: Original articles published until the 1st week of October, 2016, were searched in electronic databases (Medline-Ovid, PubMed, Scopus, The Cochrane Library, and Web of Science) for studies on NSAID use in PCa. The main clinical outcomes for the review were: PCa-specific (PCM) and all-cause mortality (ACM), biochemical recurrence (BCR), and metastases. Meta-analysis was performed to calculate the pooled hazard ratio (pHR) and their 95% confidence interval (95% CI). Heterogeneity between the studies was examined using I2 statistics. Appropriate subgroup analyses were conducted to explore the reasons for heterogeneity. Results: Out of 4216 retrieved citations, 24 observational studies and two randomized controlled studies with a total of 89,436 men with PCa met the inclusion criteria. Overall, any NSAID use was not associated with PCM, ACM, and BCR, with significant heterogeneity. Neither precancer treatment aspirin use (pHR: 1.00, 95% CI: 0.83, 1.19, P = 0.97, 5 studies, I2: 51%) nor postcancer treatment aspirin use (pHR: 0.94, 95% CI: 0.72, 1.23, P = 0.67, 8 studies, I2: 86%) was associated with PCM. Similar findings, that is, no significant association was observed for NSAID use and ACM or BCR overall, and in subgroup by types of NSAID use, and NSAID use following radiation or surgery. Conclusion: Although NSAID use was not associated with ACM, PCM, or BCR among men with PCa, significant heterogeneity remained in the included studies even after subgroup analyses.
Keywords: Aspirin, biochemical recurrence, prostate cancer, prostate cancer-specific mortality
|How to cite this article:|
Thakker D, Raval AD, Raval N, Vyas A. Nonsteroidal anti-inflammatory drugs and clinical outcomes among men with prostate cancer: A systematic review and meta-analysis. Indian J Med Paediatr Oncol 2018;39:127-41
|How to cite this URL:|
Thakker D, Raval AD, Raval N, Vyas A. Nonsteroidal anti-inflammatory drugs and clinical outcomes among men with prostate cancer: A systematic review and meta-analysis. Indian J Med Paediatr Oncol [serial online] 2018 [cited 2018 May 23];39:127-41. Available from: http://www.ijmpo.org/text.asp?2018/39/2/127/228100
| Introduction|| |
Prostate cancer (PCa) is the second most common nonskin cancer with an estimated 508,345,355 survivors and 1,392,727 incident cases by 2020 as per the World Health Organization's report of 184 countries. PCa accounts for 15% of incident cancer cases diagnosed in men as of 2012 and is the fifth leading cause of death due to cancer in men. Along the global disease burden, the worldwide cost of cancer medications is expected to increase by 11.5% from $107 billion in 2015 to 150 billion by 2020, attributed mainly to newly approved cancer chemotherapy. With advance in the health-care management and focus on the value-based incentive payment models, there are needs to find potentially effective and affordable care among men with PCa.
Nonsteroidal anti-inflammatory drugs (NSAIDs) have been the first line of treatment to relieve pain and fever – two common indicators of majority of diseases. With aspirin being the first NSAID, NSAIDs have seven chemical classes: salicylates, fenamates, para-aminophenol, acetic acid, enolic acid and propionic acid derivatives, and diaryl heterocyclic. Biologically, NSAIDs inhibit development prostanoids by blocking the activity of the cyclooxygenase (COX) enzymes. Blockage of COX enzyme activity leads to a cascade of beneficial reactions inhibiting inflammatory response in cancer. For example, preclinical studies have found that NSAIDs inhibit platelet activation which in turn inhibits the development of aggressive cancers and metastases. Such studies have demonstrated that activated platelet could lead to carcinogenesis via releasing angiogenic factors, forming platelet–tumor cell aggregates, and evading immune surveillance in the blood., In addition, NSAIDs could also play a role by inhibiting cancer-related inflammation such as the infiltration of white blood cells; tumor-associated macrophages (TAMs); cytokines such as interleukin (IL)-1, IL-6, tumor necrosis factor-α; chemokines such as (C-C motif) ligand 2 (CCL2) and CXCL8; acceleration of cell cycle progression and cell proliferation; evasion from apoptosis; and stimulation of tumor angiogenesis., Therefore, the NSAID may serve as a novel therapeutic option to manage PCa.
Epidemiological studies on NSAIDs yielded mixed results on the association between NSAIDs and clinical outcomes among men with PCa. Previously Liu et al. conducted a systematic review of eight observational studies and found beneficial association of NSAID and PCa-specific mortality (PCM) for certain subgroups using published literature until 2013. However, there still remains the need for future research to explore significant heterogeneity in the pooled study estimates with limited subgroup analyses. In addition, Liu et al. did neither examine association between the NSAID use and biochemical recurrence (BCR), metastases, or all-cause mortality (ACM) among men with PCa, nor examine the effect of primary cancer treatment in men with PCa. With available recent publications, there is a need to re-evaluate the impact of NSAIDs and PCa-related outcomes. There have been publications of many studies since the date of search of previous two systematic reviews. Therefore, we carried out a systematic review and meta-analysis to examine the effect of NSAID on the clinical outcomes such as PCM, ACM, BCR, and metastases among men with PCa.
| Methods|| |
We followed the standard guidelines of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement  and the Strengthening the Reporting of Observational studies in Epidemiology statement  to conduct and report the current systematic review and meta-analysis.
Criteria for study selection
Our systematic review included both prospective randomized controlled trials (RCTs), as well as prospective and retrospective nonrandomized a.k.a observational studies examining the effects of NSAIDs among men with PCa. However, we excluded experimental studies a. k. a cell lines,in vitro and animal studies, and studies with shorter duration (≤6 months) of follow-up. The main outcomes of interest for our review were PCM, ACM, BCR, and development of metastases.
Data sources and searches
We searched electronic databases (Medline [Ovid], Scopus, and the Cochrane library) to identify published articles on topic of our interest from the inception of each database to the 2nd week of March 2016. In addition, we also searched the Web of Science (WOS) to identify gray literature related to conference abstracts from the inception of WOS to the 3rd week of July 2016. We searched these databases using keywords such as “Non-steroidal anti-inflammatory drugs,” “Aspirin,” “prostate neoplasm,” and “prostate cancer.” We reported the details of search strategy for each database in Appendix 1 with keywords and number of retrieved citations per string. Further, we created a weekly alter for new citations' electronic databases. As of now, we included articles available in weekly search until October 10, 2016. Furthermore, we also scanned through the reference lists of identified studies for additional relevant studies.
Data extraction and quality assessment
Two authors (NR and DT) independently assessed the retrieved articles and gray literature for inclusion of articles in the review. We also checked the agreement for inclusion and exclusion of studies between two authors using the kappa statistic. In case of discrepancies about the inclusion or exclusion between two authors, a third author (ADR) resolved the issues with consensus. Three authors (ADR, NR, and DT) independently extracted information from the included studies using a data extraction template. The data extraction template has information on study design, country of participants, year of publication, sample size, inclusion and exclusion criteria of individual studies, PCa stage and severity-related variables, duration of NSAID use, and type and other baseline characteristics. In addition, we also extracted reported outcomes from each study on BCR, metastases, ACM, and PCM with details on statistical parameters such as number of events, median time to outcomes, unadjusted rates of outcomes, and unadjusted and adjusted hazard ratios (HRs).
We utilized the Newcastle Ottawa scale (NOS) tool to examine the risk of bias in included observational studies. The NOS allots up to nine points for the least risk of bias in three domains: (1) selection of study groups (four points); (2) comparability of groups (two points); and (3) ascertainment of exposure and outcomes (three points) for cohort studies. The risk of bias or poor quality was considered as “high” with one or four score total scores, “fair” with a total score of 4–6, and “good” with a total score 7 or more. In addition, we used the Cochrane Risk of Bias assessment tool to evaluate the risk of bias for performance, selection, reporting, and detecting bias domain for RCTs.
Data synthesis and analysis
We computed a pooled hazard ratio (pHR) with 95% confidence interval (CI) for all clinical outcomes reported in the included studies using random-effects models. We used the Cochrane Chi-square (Cochran Q) statistic and the I 2 test to analyze heterogeneity across included studies. In the presence of heterogeneity of pooled estimates, we performed subgroup analyses by study design, countries of studies, cancer stage, primary cancer treatment, types of NSAIDs, timing of NSAID exposures, and potential adjusted confounders. We also determined the presence of publication bias for observational studies using Egger's method (Kendall's Tau) and using a contour-enhanced funnel plot to determine other causes of publication bias by examining the symmetry of the plot. All the analyses were performed using Review Manager (RevMan) [Computer program]. Version 5.3. Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration, 2014.
| Results|| |
We identified 4219 citations through electronic databases and other resources. Out of the 4219 citations, we removed 875 duplicates with 3344 citations eligible for first-level screening. We excluded 3068 citations in the first pass based on title and abstract and 237 citations in the second level screening based on full-text information. We excluded the following studies: animal models,in vitro studies, reviews, RCTs on interventions other than NSAIDs, RCTs or observational studies on NSAID use in noncancer population, and studies assessing the risk of PCa with the use of NSAIDs. Finally, a total of 26 studies (39 references) met inclusion criteria for our review. [Figure 1] depicts the study selection process as per the PRISMA framework from the retrieved citations.
|Figure 1: Systematic review and meta-analyses flow chart for study selection for the systematic review on nonsteroidal anti-inflammatory drugs and clinical outcomes in men with prostate cancer. RCT: Randomized Controlled Trials|
Click here to view
Characteristics of included studies
[Table 1] describes the general characteristics of included studies published between 2001 and 2016. We identified 24 observational studies and two randomized controlled studies with a total pooled cohort of 89,436 men with PCa. The included studies had a total of 69,247 men from 20 retrospective cohort studies,,,,,,,,,,,,,,,,,,,, 13,855 men from two prospective cohort studies,, 1619 men from one case cohort study  and 4715 men from one nested case–control study, and 300 men from two prospective randomized controlled studies., Seventeen studies were carried out in the United States,,,,,,,,,,,,,,,,, three in the United Kingdom (UK),,, two in Canada,, one each in Belgium, Greece, Finland, and Norway. The sample size of the study cohort ranged between 74 and 11,779. With respect to types of NSAID use, 23 of the included studies reported aspirin as the NSAID, one of the each studies reported Exisulind and Ketorolac, and two studies did not specify the type of NSAIDs. The proportion of men with NSAID use ranged from 9.7% to 66%. All the included studies had at least 1 year of median follow-up period with a maximum median duration of 9.25 years in a study.
Seven of the included studies restricted their study cohort to localized PCa ,,,,,, and one study included men with advance PCa, whereas the rest of studies did not restrict their cohort based on PCa stage. Six studies included men with PCa treated with radiation therapy (RT),,,,,, five studies restricted the study cohort to men with PCa treated with radical prostatectomy (RP),,,,, one study had men with active surveillance, and the rest of the included men were diagnosed with PCa without restricting them to any primary PCa treatment such as RP/RT.
Characteristics of men with prostate cancer
[Table 2] describes the demographic, comorbidity, and lifestyle characteristics of men among the included studies by NSAID use. Majority of the included studies had 60–66 years of median or mean age of men with PCa. Five of the included studies had 1%–2% of cohort with African-American race distributed evenly between NSAID users and nonusers,,,,, whereas one study  included only African-American men with PCa. Eleven studies reported the status of chronic conditions.,,,,,,,,,, Five studies reported significant greater rate scores of Charlson Comorbidity Index or Adult Comorbidity Score-27 among NSAID users as compared to nonusers,,,,, whereas six studies reported greater rates of difference in types of chronic conditions, specifically diabetes, heart disease, and hypertension among NSAID users as compared to nonusers.,,,,, With regard to lifestyle characteristics, eight studies reported body mass index of the study cohort and majority of the studies had greater proportions of overweight or obese men in both NSAID users and nonusers.,,,,,,, Six studies reported had the smoking status of study cohort and out of those all the studies had more than one-third of past or current smokers across NSAID users and nonusers.,,,,,
|Table 2: Demographic, comorbidity, and lifestyle characteristics of study population by nonsteroidal anti-inflammatory drugs use among the included studies|
Click here to view
[Table 3] describes PCa-related clinical and cancer-related treatment characteristics by NSAID use. Fourteen studies reported the status of prostate-specific antigen (PSA) levels ,,,,,,,,,,,,, and those studies did not find significant difference in PSA levels between NSAID users and nonusers. Sixteen studies reported Gleason score,,,,,,,,,,,,,,,,, of which two studies included >50% of men with Gleason score >7. Thirteen studies reported tumor stage,,,,,,,,,,,,, of which two studies , had significant differences in the proportion of men with T2/3 stages among NSAID users and nonusers. Nine studies ,,,,,,,, reported the National Comprehensive Cancer Network scores of which one study  had significant difference in medium- and high-risk PCa among NSAID users and nonusers. Fifteen studies reported androgen deprivation therapy (ADT),,,,,,,,,,,,,, of which two studies , did not have any men with ADT users among NSAIDs users and no users.
|Table 3: Prostate cancer and its treatment-related characteristics among the included studies|
Click here to view
Quality assessment of included studies
We used NOS tool to determine the quality of the included observational studies and the Cochrane risk of bias tool for RCT. Twenty-four of the included observational studies had a fair or good quality as per the NOS scale and two of the included observational studies had low risk for selection, performance bias, and unclear risk for reporting bias. Details on each of the three domains for risk of bias selection, ascertainment of exposure and outcomes, and comparability for observational studies are provided in Appendix 1.1-1.3 [Additional file 1] [Additional file 2] [Additional file 3] and the risk of bias table for RCT are provided in Appendix 1.4. [Additional file 4] Out of the 24 observational studies, ten utilized single institutional data,,,,,,,,,, seven utilized survey or prospective cohort data,,,,,,, and six studies utilized cancer registry-linked electronic medical records or administrative claim databases.,,,,, The coding algorithms for PCa were not defined with their diagnostic accuracy in single institutional data. With respect to exposure of interest, five studies ,,,, reported the use of NSAID postcancer or postcancer therapy, five studies ,,,, reported the precancer period or precancer treatment period NSAID exposure, and rest of the studies measured NSAID at any time during the study period. Except one study, all other observational studies conducted multivariable regression analysis controlling for the potential confounders while examining association between NSAID use and clinical outcomes. However, only Fourteen studies ,,,,,,,,,,,, controlled for co-medications.
Nonsteroidal anti-inflammatory drug use and clinical outcomes
We conducted meta-analyses of the included studies to evaluate the association between NSAID use and several clinical outcomes in men with PCa. Due to the presence of significant heterogeneity in the pooled estimates, we performed several subgroup analyses by types of NSAIDs, pre-, post- or any NSAID use, and type of PCa treatment.
Prostate cancer-specific mortality
Eleven studies reported the association between NSAID use and PCM,,,,,,,,,,, of which three studies found an inverse relationship between aspirin use and PCM,,, while two studies found an increased risk of PCM associated with aspirin use,, whereas rest of the six studies did not find any association between NSAID use and PCM.
Overall, any type of NSAID use was not associated with the risk of PCM (pHR: 0.95; 95% CI: 0.79, 1.05; P = 0.19; I 2 = 62%) with significant heterogeneity [Figure 2]. In the subgroup analyses by timing of aspirin use, we found that aspirin use, irrespective of prior to or following PCa diagnosis, was not associated with PCM (pHR: 0.95; 95% CI: 0.77, 1.16; P = 0.60; studies = 11, I 2 = 88%). Furthermore, aspirin use prior to PCa diagnosis or primary treatment was not significantly associated with PCM (pHR: 1.00; 95% CI: 0.83, 1.19; P = 0.97; studies = 5; I 2 = 51%), and so was for ever (pHR: 1.34; 95% CI: 0.71, 2.51; P = 0.36; studies = 2; I 2 = 97%) or postdiagnosis or postcancer aspirin use (pHR: 0.94; 95% CI: 0.72, 1.23; P = 0.67; studies = 8; I 2 = 86%). In contrast, postdiagnosis or post-treatment NSAID use was associated with 2.5 times increased risk of PCM (pHR: 2.50; 95% CI: 1.75, 3.57; P < 0.0001; studies = 1). However, one study reported a nonsignificant association between precancer diagnosis NSAID use and PCM (pHR: 1.03; 95% CI: 0.79, 1.34; P = 0.83).
|Figure 2: Forest plot of comparison: Nonsteroidal anti-inflammatory drug users versus nonusers for prostate cancer-specific mortality|
Click here to view
Nine studies reported the association between NSAID use and PCM among men who had RP, RT, or other therapies.,,,,,,,, In this subgroup, NSAID use was not significantly associated with PCM (pHR: 1.02; 95% CI: 0.79, 1.30; P = 0.90; studies = 10; I 2 = 91%). Among men with PCa who had RT as the primary treatment, NSAID use was not significantly associated with PCM (pHR: 0.91; 95% CI: 0.65, 1.27; P = 0.58). Two studies , which controlled for time of NSAID exposure using time-dependent NSAID use reported a lower though nonsignificant risk of PCM due to NSAID use (pHR: 0.84; 95% CI: 0.25, 2.76; P = 0.77; studies = 2; I 2 = 91%).
Three studies reported the risk of PCM by NSAID dose.,, As compared to nonusers, either high-dose or low-dose NSAID was associated with 5% reduction in the PCM, though the association was not significant.
Nonsteroidal anti-inflammatory drug use and all-cause mortality
Three studies reported ACM as an outcome.,, NSAID use was not associated with ACM in the pooled analysis with significant heterogeneity (pHR: 0.82; 95% CI: 0.54, 1.24; P = 0.19; I 2 = 73%) [Figure 3]. To find the cause of heterogeneity, leave-one-out sensitivity analysis was performed which did not alter the study finding.
|Figure 3: Forest plot of comparison: Nonsteroidal anti-inflammatory drug users versus nonusers for all-cause mortality|
Click here to view
Nonsteroidal anti-inflammatory drug use and biochemical recurrence
The hazard of BCR was reported in five studies.,,,, Overall, NSAID use was not significantly associated with reducing the hazards of BCR using random-effects model (pHR: 0.82, 95% CI: 0.54, 1.24, P = 0.34, 5 studies, I 2: 73%) with significant heterogeneity. We carried out subgroup analyses to explore the possible reasons of heterogeneity. NSAID use in men who had RP (pHR: 0.79, 95% CI: 0.45, 1.38, P = 0.41, 3 studies, I 2: 69%) or RT (pHR: 0.85, 95% CI: 0.35, 2.03, P = 0.71, 2 studies, I 2: 88%) was not significantly associated with reducing the risk of BCR [Figure 4].
|Figure 4: Forest plot of comparison: Nonsteroidal anti-inflammatory drug users versus nonusers for biochemical recurrence|
Click here to view
Nonsteroidal anti-inflammatory drug use and metastases
Two studies , examined the hazards of metastases with the use of NSAID. Both the studies found a beneficial effect of NSAID in the univariate analysis and did not have a sufficient sample size to conduct multivariable analyses.
| Discussion|| |
Our review aimed to generate evidence on the effects of NSAIDs on cancer prognosis among men with PCa. We found 24 observational studies and two RCTs examining the impact of NSAIDs on different cancer prognostic outcomes. We found that NSAID use was not associated with BCR, metastases, ACM, and PCM in pooled estimates with significant heterogeneity. To account for heterogeneity, we conducted various subgroup analyses and observed significant findings across those subgroup analyses which reduced the heterogeneity in some cases.
First, as majority of the included studies used aspirin as the main NSAID class, we did a subgroup analysis by types of NSAID as aspirin users and nonaspirin NSAID. We did not find any conclusive evidence on the beneficial effect of NSAIDs and PCM. Our findings are also similar to a previous systematic review by Elwood et al. examining the association between NSAID and PCM. Elwood et al. found no association between NSAID and PCM in a pooled estimate of nine studies with significant heterogeneity. However, in that study, they observed a significant association with omission of one study. This difference in our findings could be due to negative or no association observed in the newer additional studies included in our review.
There has been significant heterogeneity in the pooled estimates even after subgroup analyses, which suggest the variation in the analytic samples among the included studies. Most of the studies had utilized secondary preexisting data and the study was based on a data-driven design approach rather than design-driven data collection approach. For example, there have been great variations on timing of NSAID use as pre/postdiagnosis, or pre/post-treatment NSAID use, or ever use of NSAID. Both pre- and post-diagnosis NSAID could not serve as meaningful treatment to inhibit progression of cancer if men do not adhere to medications or used for symptomatic pain relief. One of the major biases with NSAIDs is bias by treatment indications. NSAID can be used for different indications at different stages of prostate cancer having low or high comorbidity burden. For example, aspirin can be used to prevent primary prevention of heart disease in men with localized PCa as well as to relieve pain in men with metastatic PCa, which have drastically different prognosis.
Second, the timing of aspirin in the postdiagnosis or post-treatment suffers from two major biases: immortal time bias and lag time bias. With respect to immortal time bias, men using NSAID in the follow-up period should be alive until they receive the NSAID postdiagnosis. This leads to misclassification of unexposed person-time as exposed person-time. In addition, if NSAID is initiated few days before death due to palliative pain relief, then there would be very short NSAID exposure time for fatal event. Therefore, such exposure may not provide conclusive findings due to design limitations. Although two studies have controlled for time of NSAID exposure using time-dependent NSAID use, both the studies yielded mixed results on the NSAID use and PCM. Such results may be again due to variation in the population.
Further, we also explored relationship between NSAID and clinical outcomes by high-risk PCa. Again, study findings among those with high-risk PCa had mixed findings. These could be due to effect modification due to hormone therapy use in high-risk PCa with aspirin. A study found that aspirin was associated with abnormal liver function and the serum level of aspirin is also expected to be higher due to low level of testosterone which leads to inhibit the metabolism of aspirin. Eventually, men using NSAID are at risk of abnormal liver function  and affect the adherence to hormone therapy in men with PCa. Therefore, the proportion of patients using hormone therapy could serve as a potential confounder for the association between NSAID use and clinical outcomes in men with PCa.
Strengths and limitations of our study findings suggest the need for well-designed observational studies to examine the association between NSAID use and clinical outcomes among homogeneous PCa population, i.e., stage or primary PCa treatment controlling for important covariate including adherence and persistence or duration of NSAID use.
| Conclusions|| |
Although we did not find association between NSAID use and clinical outcomes in men with PCa in our systematic review, our study findings highlight the need to consider variation in the population of NSAID users requiring need for better design in future observational studies.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
| References|| |
Thun MJ, Blackard B. Pharmacologic effects of NSAIDs and implications for the risks and benefits of long-term prophylactic use of aspirin to prevent cancer. Recent Results Cancer Res 2009;181:215-21.
Goubran HA, Burnouf T, Radosevic M, El-Ekiaby M. The platelet-cancer loop. Eur J Intern Med 2013;24:393-400.
Gay LJ, Felding-Habermann B. Contribution of platelets to tumour metastasis. Nat Rev Cancer 2011;11:123-34.
Colotta F, Allavena P, Sica A, Garlanda C, Mantovani A. Cancer-related inflammation, the seventh hallmark of cancer: Links to genetic instability. Carcinogenesis 2009;30:1073-81.
Kundu JK, Surh YJ. Inflammation: Gearing the Journey to Cancer. Vol. 659. Proceedings 5th
International Conference Environmental Mutagens Humam Population ICEMHP; 2008. p. 15-30.
Liu Y, Chen JQ, Xie L, Wang J, Li T, He Y, et al.
Effect of aspirin and other non-steroidal anti-inflammatory drugs on prostate cancer incidence and mortality: A systematic review and meta-analysis. BMC Med 2014;12:55.
Moher D, Liberati A, Tetzlaff J, Altman DG, PRISMA Group. Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement. PLoS Med 2009;6:e1000097.
von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP, et al.
Strengthening the reporting of observational studies in epidemiology (STROBE) statement: Guidelines for reporting observational studies. BMJ 2007;335:806-8.
van Houwelingen HC, Arends LR, Stijnen T. Advanced methods in meta-analysis: Multivariate approach and meta-regression. Stat Med 2002;21:589-624.
Deeks JJ, Macaskill P, Irwig L. The performance of tests of publication bias and other sample size effects in systematic reviews of diagnostic test accuracy was assessed. J Clin Epidemiol 2005;58:882-93.
Peters JL, Sutton AJ, Jones DR, Abrams KR, Rushton L. Contour-enhanced meta-analysis funnel plots help distinguish publication bias from other causes of asymmetry. J Clin Epidemiol 2008;61:991-6.
Choe KS, Cowan JE, Chan JM, Carroll PR, D'Amico AV, Liauw SL, et al.
Aspirin use and the risk of prostate cancer mortality in men treated with prostatectomy or radiotherapy. J Clin Oncol 2012;30:3540-4.
Katz MS, Carroll PR, Cowan JE, Chan JM, D'Amico AV. Association of statin and nonsteroidal anti-inflammatory drug use with prostate cancer outcomes: Results from caPSURE. BJU Int 2010;106:627-32.
Mondul AM, Han M, Humphreys EB, Meinhold CL, Walsh PC, Platz EA, et al
. Association of statin use with pathological tumor characteristics and prostate cancer recurrence after surgery. J Urol 2011;185:1268-73.
Forget P, Tombal B, Scholtès JL, Nzimbala J, Meulders C, Legrand C, et al.
Do intraoperative analgesics influence oncological outcomes after radical prostatectomy for prostate cancer? Eur J Anaesthesiol 2011;28:830-5.
Zaorsky NG, Li T, Cohen RJ, Horwitz EM, Uzzo RG, Viterbo R, et al
. Aspirin use decreases the risk of prostate cancer recurrence after post-prostatectomy radiotherapy. J Radiat Oncol 2015;4:193-201.
Choe KS, Correa D, Jani AB, Liauw SL. The use of anticoagulants improves biochemical control of localized prostate cancer treated with radiotherapy. Cancer 2010;116:1820-6.
Flahavan EM, Bennett K, Sharp L, Barron TI. A cohort study investigating aspirin use and survival in men with prostate cancer. Ann Oncol 2014;25:154-9.
Kontraros M, Varkarakis I, Ntoumas K, Deliveliotis C. Pathological characteristics, biochemical recurrence and functional outcome in radical prostatectomy patients on statin therapy. Urol Int 2013;90:263-9.
Caon J, Paquette M, Hamm J, Pickles T. Does statin or ASA affect survival when prostate cancer is treated with external beam radiation therapy? Prostate Cancer 2014;2014:184297.
Grytli HH, Fagerland MW, Fosså SD, Taskén KA. Association between use of β-blockers and prostate cancer-specific survival: A cohort study of 3561 prostate cancer patients with high-risk or metastatic disease. Eur Urol 2014;65:635-41.
Ishak-Howard MB, Okoth LA, Cooney KA. Statin use and the risk of recurrence after radical prostatectomy in a cohort of men with inherited and/or early-onset forms of prostate cancer. Urology 2014;83:1356-61.
Jacobs CD, Chun SG, Yan J, Xie XJ, Pistenmaa DA, Hannan R, et al.
Aspirin improves outcome in high risk prostate cancer patients treated with radiation therapy. Cancer Biol Ther 2014;15:699-706.
Jacobs EJ, Newton CC, Stevens VL, Campbell PT, Freedland SJ, Gapstur SM, et al.
Daily aspirin use and prostate cancer-specific mortality in a large cohort of men with nonmetastatic prostate cancer. J Clin Oncol 2014;32:3716-22.
Assayag J, Pollak MN, Azoulay L. The use of aspirin and the risk of mortality in patients with prostate cancer. J Urol 2015;193:1220-5.
Veitonmäki T, Murtola TJ, Määttänen L, Taari K, Stenman UH, Tammela TL, et al.
Use of non-steroidal anti-inflammatory drugs and prostate cancer survival in the Finnish prostate cancer screening trial. Prostate 2015;75:1394-402.
Zaorsky NG, Buyyounouski MK, Li T, Horwitz EM. Aspirin and statin nonuse associated with early biochemical failure after prostate radiation therapy. Int J Radiat Oncol Biol Phys 2012;84:e13-7.
Agarwal G, Spuches J, Luchey A, Patel T, Pow-Sang J. The effect of NSAID use on disease progression in patients on active surveillance for prostate cancer. Ann Rheum Dis 2015;193:e756.
Osborn VW, Chen SC, Weiner J, Schwartz D, Schreiber D. Impact of aspirin on clinical outcomes for African American men with prostate cancer undergoing radiation. Tumori 2016;102:65-70.
Ratnasinghe LD, Graubard BI, Kahle L, Tangrea JA, Taylor PR, Hawk E, et al.
Aspirin use and mortality from cancer in a prospective cohort study. Anticancer Res 2004;24:3177-84.
Dhillon PK, Kenfield SA, Stampfer MJ, Giovannucci EL, Chan JM. Aspirin use after a prostate cancer diagnosis and cancer survival in a prospective cohort. Cancer Prev Res (Phila) 2012;5:1223-8.
Stock DC, Groome PA, Siemens DR, Rohland SL, Song Z. Effects of non-selective non-steroidal anti-inflammatory drugs on the aggressiveness of prostate cancer. Prostate 2008;68:1655-65.
Cardwell CR, Flahavan EM, Hughes CM, Coleman HG, O'Sullivan JM, Powe DG, et al.
Low-dose aspirin and survival in men with prostate cancer: A study using the UK Clinical Practice Research Datalink. Cancer Causes Control 2014;25:33-43.
Goluboff ET, Prager D, Rukstalis D, Giantonio B, Madorsky M, Barken I, et al.
Safety and efficacy of exisulind for treatment of recurrent prostate cancer after radical prostatectomy. J Urol 2001;166:882-6.
D'Amico AV, Chen MH, Renshaw AA, Loffredo B, Kantoff PW. Risk of prostate cancer recurrence in men treated with radiation alone or in conjunction with combined or less than combined androgen suppression therapy. J Clin Oncol 2008;26:2979-83.
Daugherty S, Pfeiffer R, Ghazarian A, Izmirlian G, Prorok P, McGlynn K. Frequency of aspirin use and prostate cancer mortality among prostate cancer cases in the control arm of the Prostate, Lung, Colorectal, and Ovarian (PLCO) cancer screening trial. Cancer Res 2013;73:LB-15.
Veitonmäki T, Murtola TJ, Talala K, Taari K, Tammela T, Auvinen A, et al.
Non-steroidal anti-inflammatory drugs and cancer death in the Finnish prostate cancer screening trial. PLoS One 2016;11:e0153413.
D'Amico AV, Kantoff PW, Chen MH. Aspirin and hormone therapy for prostate cancer. N
Engl J Med 2007;357:2737-8.
[Figure 1], [Figure 2], [Figure 3], [Figure 4]
[Table 1], [Table 2], [Table 3]