Association of Androgen Deprivation Therapy with Thromboembolic Events in Patients with Prostate Cancer: a Systematic Review and Meta-analysis: Full-Text Article

Background: Whether androgen deprivation therapy (ADT) causes excess thromboembolic events (TEs) in men with prostate cancer (PCa) remains controversial and is the subject of the US Food and Drug Administration safety warning. This study aims to perform a systematic review and meta-analysis on previous studies to determine whether ADT is associated with TEs in men with PCa.

Methods: Medline, Embase, and Cochrane Library databases were searched for relevant studies. These studies comprised those that compared ADT versus control to treat PCa, reported TEs as an outcome and were published before January 2018. Multivariate-adjusted hazard ratios (HRs) and associated 95% confidence intervals (CIs) were calculated using random- or fixed-effects models.

Results: Five retrospective population-based cohort studies involving 170,851 ADT users and 256,704 non-ADT users were identified. Deep venous thrombosis (DVT) was found significantly associated with gonadotropin-releasing hormone (GnRH) agonists alone (HR = 1.47, 95% CI: 1.07–2.03; P = 0.017; I2 = 96.3%), GnRH agonists plus oral antiandrogen (AA) (HR = 2.55, 95% CI: 2.21–2.94; P < 0.001; I2 = 0.0%), and AA alone (HR = 1.49, 95% CI: 1.13–1.96; P = 0.004; I2 = 0.0%), but not with orchiectomy (HR = 1.80, 95% CI: 0.93–3.47; P = 0.079; I2 = 94.8%). In addition, pulmonary embolism (PE) was significantly associated with GnRH agonists alone (HR = 2.26, 95% CI: 1.78–2.86; P < 0.001; I2 was unavailable) and orchiectomy (HR = 2.12, 95% CI: 1.44–3.11; P < 0.001; I2 = 57.2%). This relationship was also supported with subgroup analyses based on different continents and races.

Conclusions: GnRH agonists alone, GnRH plus AA, and AA alone cause excess DVT in men with PCa after controlling the demographic and disease characteristics and other confounding factors, although statistically significant difference was not observed in orchiectomy group. Additionally, GnRH agonists alone and orchiectomy can increase the incidence of PE.

Prostate cancer (PCa), a leading cause of morbidity and mortality in men worldwide, has gradually become a major problem1–3. Androgen deprivation therapy (ADT) includes different types of treatments, such as oral antiandrogen (AA), gonadotropin-releasing hormone (GnRH) agonists, orchiectomy, and two or more types above combined as palliative therapy. ADT is the main treatment for PCa because the development and growth of PCa cells depend on androgen4–6. The use of ADT in the United States has increased by a factor of 26 over the past two decades, a trend that reflects a growing number of patients with local diseases whose benefits are less clear7–9.  However, this effect is unclear in terms of prolonging life expectancy or negative in a few clinical studies 10, 11. Several observational studies reported that ADT may increase the risk of cardiovascular diseases, including coronary heart disease and atherosclerotic plaque progression and instability12–14. Furthermore, one population-based cohort study demonstrated that ADT can significantly increase the risk of stroke15. Few studies suggested a positive association between ADT and thromboembolic events (TEs) in men with PCa16, 17. Nevertheless, conflicting results are also reported. Klil- Drori et al.18 demonstrated that orchiectomy and AA alone are not statistically associated with increased risk of deep venous thrombosis (DVT) incidence compared with those undergoing non-ADT. Nonetheless, no meta-analysis has explored the association of ADT and TEs in men with PCa globally. Hence, our study aimed to examine such a relation. A broad systematic review and meta-analysis of published studies were performed to evaluate the association of ADT and TEs in men with PCa precisely and help healthcare professionals make related clinical decisions.

Materials and methods:
The methods in this meta-analysis were conducted in accordance with the Cochrane Collaboration criterion 19. Furthermore, we followed the Meta-analysis of Observational Studies in Epidemiology approach 20. Thus, no ethical approval and patient consent are required.

Eligibility criteria
Studies that explored the association of ADT with TEs in patients with PCa were included without the geographical region, publication status, or language restrictions. The following inclusion criteria were adopted: (1) studies were published as original articles, and they contained original data (excluding reviews, editorials, and conference abstracts) and frequency data for TEs of any type or severity; (2) patients diagnosed with PCa only; (3) intervention groups must include ADT (either medical or surgical ADT); (4) treatments in the control groups were non-ADT (e.g., active surveillance, radical prostatectomy, and radiotherapy); and (5) studies must report comparative data of risk estimates with 95% confidence intervals (CIs). If more than one study was identified from the same population, then we used the most recent or complete report of that study. Studies were excluded if any of the following factors were identified: (1) secondary studies, (2) laboratory research, (3) animal studies, and (4) duplicated database and absence of detailed results.

Data sources and searches
To identify related studies, we performed a comprehensive literature search on the electronic databases of Medicine, Embase, and Cochrane Library from database inception until January 2018. Each database was searched using a combination of Medical Subject Headings (MeSH) and non-MeSH terms, such as “prostate cancer,” “prostate tumor” or “prostate carcinoma” and “androgen deprivation,” “androgen suppression,” “endocrine treatment,” “ADT” or “AST” and “thromboembolism,” and “embolism” or “thrombosis”. Manual searches of reference lists were also carried out in relevant original and review articles. The main search was completed by the senior author (ZLG). Any discrepancy was resolved by consulting another investigator (SSW) who was not involved in the initial procedure.

Study selection and data extraction
Two reviewers (ZLG and LLG) independently extracted data using a predefined data extraction form. Disagreements were resolved by discussion or consensus with a third reviewer (SG). The following data were extracted: first author’s name, study characteristics (i.e., year, duration, setting, and design), participant characteristics (i.e., mean age and sample size), study population, definition of TEs according to what was described in each included publication, covariates adjusted in the statistical analysis, and hazard ratios (HRs) with corresponding 95% CIs for TE risk in each comparison. Our meta-analysis involved different types of ADTs, including GnRH agonists, GnRH agonists plus AA, AA alone, and orchiectomy. With regard to studies with insufficient information, primary authors were contacted if possible to acquire and verify data.

Outcome and methodological quality assessment
Our analysis outcome aimed to evaluate the incidence of TEs in men with PCa with ADT compared with those with non-ADT. The assessment was separately conducted by two independent reviewers (YYH and FLC) in accordance with the modified version of the Newcastle–Ottawa Scale (NOS) for quality assessment of observational studies21. All disparities between the two reviewers were resolved with consultation with another field expert. The following domains were used to evaluate the methodological quality of the observational studies: selection and comparability of study groups and exposure/outcome of interest. Each numbered item within the categories of selection and exposure/outcome was awarded a maximum of one star. A maximum of two stars was given for comparability across studies. Furthermore, we considered studies that obtained 60% or more of the maximum number of stars as high-quality studies.

UroToday PCAN literature searches according to the Preferred Reporting Items for Systematic Reviews and Meta Analyses statement
Fig. 1 Flow diagram of literature searches according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses statement

Statistical analyses
Pooled analysis was conducted using STATA statistical package (version 14.0; serial number: 10699393; statacorp: wyb) to assess the association of ADT with TEs in patients with PCa. The reported risk estimates in each study were reanalyzed to work with consistent definitions. According to the Cochrane review guidelines, a severe heterogeneity of I2 ≥ 50% warrants the use of random-effects models. Otherwise, a fixed-effects model should be used 19. I2 was used to assess heterogeneity across studies, with I2 values of 0, 25, 50, and 75% representing no, low, moderate, and high heterogeneity, respectively, in compliance with the Cochrane review guidelines19. Statistical significance was set at P < 0.05. Sensitivity analysis was carried out by examining the exclusion of each study in a step-wise manner to evaluate the quality and consistency of results.  Subgroup analyses were performed according to different continents and race. A meta-regression analysis was conducted to investigate possible sources of heterogeneity on a certain variable. The restricted maximum likelihood method was used for analysis. Nevertheless, the use of Egger’s regression asymmetry test was limited because of the small number of studies evaluated22.

Literature search
A flow chart depicting the search process and study selection is shown in Fig. 1. A total of 598 studies were identified through our comprehensive search. After removing 203 duplicates, only 395 studies were retrieved. After reading the titles and abstracts, 377 studies were excluded because they did not meet inclusion criteria, and the reasons are as follows: secondary studies (259 studies), laboratory search (67 studies), and not the related topic (51 studies). Finally, we carefully read the full text of each of the remaining 18 trials and 13 trials were excluded for the following reasons: TEs were not the endpoint (6 studies), ADT was used in the control groups (3 studies), duplicated database (2 studies), and endocrine therapy was not ADT (2 studies). Study characteristics and quality of evidence Five retrospective population-based cohort studies16–18, 23, 24 were included in the meta-analysis. The basic characteristics of the included studies are described in Table 1. These studies were published between 2010 and 2017, and they involved 170,851 ADT users and 256,704 non-ADT users. One study was conducted in America16, one in the United Kingdom18, one in Denmark17, one in Sweden23, and one in Taiwan24. In the included clinical studies, the sample sizes varied in the range of 7181–182,757 participants. All of the included studies reported adjusted covariates in the statistical analysis16–18, 23, 24. The association between ADT and DVT risk was reported in four studies16–18, 23, and pulmonary embolism (PE) risk was investigated in two studies23, 24. Moreover, the methodological quality of four studies16–18, 23 was considered of high quality, whereas one study24 was regarded as poor quality on the basis of the NOS. The main deficiency was reporting bias related to the insufficient information on tumor stage and follow-up.

UroToday PCAN Characteristics of the included studies 1

AA oral antiandrogens, ADT androgen deprivation therapy, CPRD UK Clinical Practice Research Datalink, DNRP Danish National Registry of Patients, DVT deep venous thrombosis, GnRH gonadotropin-releasing hormone, IQR interquartile range, NA not available, NHI National Health Insurance system of Taiwan, NPCR National Prostate Cancer Register of Sweden, PE pulmonary embolism, PSA prostate-specific antigen, SEER Surveillance Epidemiology and End Results Medicare database, y years

Association of GnRH agonists alone with DVT risk
Figure 2a shows the effects of GnRH agonists alone versus non-ADT in the endpoint of DVT among 163,670 ADT users compared with those in 256,704 non-ADT users16–18, 23. Pooled HR showed that the incidence of DVTm in the GnRH agonists alone group was 47% higher than that in non-ADT users (HR = 1.47, 95% CI: 1.07–2.03; P = 0.017), and significant heterogeneity was observed (I2 = 96.3%, P < 0.001). Thus, a random-effects model was used for pooled analysis. Sensitivity analysis indicated that finding stability exhibited no significant changes by omitting any single study sequentially (Table 4). With regard to subgroup analyses of different continents (Table 2), findings on the association between GnRH agonists alone and DVT were consistent among studies conducted in Europe and North America. When the studies were stratified by different races, the association among black men was significant compared with that among white men (HR = 1.27, 95% CI: 1.18–1.36; P < 0.001). However, the subgroups showed no considerable contributions to heterogeneity. Meta-regression analysis results indicated that none of the covariates (P = 0.434) resulted in heterogeneity among four studies. Additionally, the adjusted R2 of −2.38% indicated that the regressor is contributing little to the explanation of the response variable 25, 26 (Table 3).

Association of orchiectomy with DVT risk
The effects of orchiectomy versus non-ADT on the endpoint +of DVT among 142,430 ADT users compared with those in 228,411 non-ADT users are shown in Fig. 2b16,18, 23. Results revealed that the incidence of DVT in the orchiectomy group was 80% higher than that in the non-ADT users, although statistically significant difference was not observed (HR = 1.80, 95% CI: 0.93–3.47; P = 0.079) with significant heterogeneity (I2 = 94.8%, P < 0.001) through a random-effects model. Sensitivity analysis results showed that finding stability exhibited no significant changes by omitting any single study sequentially (Table 4).  In the subgroup analyses (Table 2), a significant association was observed among the studies performed in North America and Europe. Nevertheless, no considerable contributions to heterogeneity were detected based on subgroup analyses. Furthermore, in the meta-regression analysis, none of the covariates (P = 0.309) resulted in heterogeneity among three studies, and the adjusted R2 of 6.80% in between-study variance indicated that the regressor is also contributing little to the explanation of the response variable25, 26 (Table 3).

Association of GnRH agonists plus AA and AA alone with DVT risk Meta-analysis on two studies18, 23 indicated that GnRH agonists plus AA significantly increased the DVT risk by 2.55 times (HR = 2.55, 95% CI: 2.21–2.94; P < 0.001; I2 = 0.0%). In addition, the incidence of DVT in the AA group was 49% higher than that in non-ADT users (HR = 1.49, 95% CI: 1.13–1.96; P = 0.004; I2 = 0.0%) (Fig. 2c). Association of GnRH agonists alone and orchiectomy with PE risk Meta-analysis on two studies23, 24 revealed that PE was significantly associated with GnRH agonists alone (HR = 2.12, 95% CI: 1.44–3.11; P < 0.001; I2 = 57.2%) and orchiectomy (HR = 2.26, 95% CI: 1.78–2.86; P < 0.001; I2 was unavailable). Thus, a random-effects model was used for pooled analysis (Fig. 3).

UroToday PCAN Fig. 2 a HRs of DVT related to GnRH agonists alone
UroToday PCAN b HRsofDVTrelated toorchiectomy
UroToday PCAN HRsofDVTrelated toGnRHagonists plusAAandAAalone

Fig. 2 a HRs of DVT related to GnRH agonists alone. b HRs of DVT related to orchiectomy. c HRs of DVT related to GnRH agonists plus AA and AA alone. AA oral antiandrogens, CI confidence interval, DVT deep venous thrombosis, GnRH gonadotropin-releasing hormone, HRs hazard ratios.

UroToday PCAN Results of subgroup analyses
UroToday PCAN Resultsofmetaregression

The association between ADT and TE risk is still controversial, although the incidence of TEs in men undergoing ADT with PCa has been an emerging problem in recent years. This meta-analysis systematically provides the first available summary of five observational studies to generate the current most remarkable estimates of TE risk in men undergoing ADT with PCa. According to the main result, GnRH agonists alone, GnRH plus AA, and AA alone caused excess DVT compared with those of non-ADT in men with PCa, although statistically significant difference was not observed in the orchiectomy group. Moreover, GnRH agonists alone and orchiectomy can increase the incidence of PE. This significant association was observed among studies performed in North America and Europe. In particular, we found that the risk of DVT was higher in black men than that in white men.

Most of the included studies suggested a significant association between ADT and TEs, and one study reported conflicting results18.. Klil-Drori et al. 18 demonstrated that orchiectomy and AA alone are not statistically associated with increased risk of DVT incidence, and the multivariate HR is 1.56 (95% CI: 0.63–3.81) and 1.43 (95% CI: 0.98–2.10), respectively, compared with those undergoing non-ADT. One study was considered of low quality on account of the reporting bias related to insufficient information about residual confounding factors (e.g., missing information on tumor grade and stage)24. To reduce this bias, a sensitivity analysis was performed to compare orchiectomy monotherapy with non-ADT. When this study was excluded, the finding stability exhibited no significant changes. Notably, significant heterogeneity was detected in GnRH agonists alone and the orchiectomy group with DVT risk. Therefore, we carried out a meta-regression analysis based on certain variable and subgroup analyses stratified by different races to investigate possible sources of heterogeneity.  Results showed no considerable contributions to heterogeneity.

Mentioning that two retrospective population-based cohort studies on the topic have been excluded from the study because they no longer meet the inclusion criteria is important. We found that Ehdaie et al.27 and O’Farrell et al.28 used the same database as that of Hu et al.16 and Van Hemelrijck et al.23 to collect information, respectively. Nevertheless, Ehdaie et al.27 indicated that GnRH agonists alone and orchiectomy are statistically associated with increased risk of DVT incidence, in which the multivariate HR is 1.54 (95% CI: 1.49–1.60) and 1.97 (95% CI: 1.72–2.26), respectively, compared with those undergoing non-ADT. Moreover, O’Farrell et al.28 demonstrated that DVT is significantly associated with GnRH agonists alone (HR=1.67, 95% CI: 1.40–1.98) and orchiectomy (HR=1.61, 95% CI: 1.15–2.28). Additionally, they also found that GnRH agonists alone (HR=1.59, 95% CI: 1.22–2.06) and orchiectomy (HR=1.61, 95% CI: 1.42–1.82) can significantly result in PE. However, our findings remain reliable and rigorous based on the results of the sensitivity analyses and meta-regression.

More than 45% of patients will die outside of the PCa, and TEs are the second common cause of death in all patients with cancer29, 30. Although our findings are related to the clinical decisions of individual patients, they also affect public health and the economic burden of PCa. Moreover, PCa accounts for a large proportion of cancer cases and deaths annually, and most men are diagnosed early and shown a favorable prognosis. ADT accounts for half of all patients with PCa in 1990, and it accounts for one-third of Medicare expenditures in PCa treatment31.  Furthermore, ADT usage is sensitive to the reimbursement level of clinicians 32, 33. The precise mechanism by which ADT affects the veins and central and peripheral arteries remains unclear; ADT can increase fat mass and decrease insulin sensitivity, which may result in arterial stiffness and atherosclerosis34, 35.

UroToday PCAN Results of sensitivity analyses
CI confidence interval, DVT deep venous thrombosis, gonadotropin-releasing hormone, HRs hazard ratios

Although the ADT benefits may outweigh its risks, the optimal treatment duration is uncertain. A randomized trial found that the overall survival of 36 months is higher in patients with high-risk PCa receiving radiation therapy than that of 6 months36. However, another randomized trial reported that a long duration of ADT is unassociated with improved survival compared with that of short duration37. Our understanding of the age of patients and duration of ADT is insufficient because some of these factors have been rarely investigated in the included studies. Moreover, the association between ADT and TEs in men with different tumor grades and stages remains unclear because of the lack of studies that examined such a relationship from included studies. Thus, further research is warranted to verify the findings of this meta-analysis in terms of the possible influence of age, tumor stage, and duration of ADT on extensive consequences for preventive measures.

Strengths and limitations
In general, our study exhibited strengths in several aspects. First, this study was the first to explore a potential association between ADT and TE risk in men with PCa through a comprehensive and rigorous meta-analysis. Second, the overall combined estimates were based on the validity of the codes analyzed. Additionally, the included studies have collected primary data from authoritative Medicare databases and were approved by the Institutional Review Board.  Thus, the rationality and reliability of our meta-analysis results were evidently improved. Third, confounding factors that may influence TE risk were minimized because multivariable-adjusted risk estimates were applied.

UroToday PCAN HRsofPErelatedto GnRHagonists
This study also displayed limitations that must be acknowledged before accepting the findings. First, studies varied in their degree of control for confounding factors, such as age, ethnic, clinical tumor stage, Gleason score, prostate-specific antigen, and lymph node involvement. Hence, various studies were difficult to reconcile. Second, heterogeneity is another critical issue that may have been due to insufficient information about tumor grade and stage, which is distinctive in the study of Yii et al.24. Nonetheless, when we removed the outlier study from the meta-analysis, no strong evidence of heterogeneity was observed in the remaining studies.  Third, selection bias may have influenced our results. To minimize this bias, we carried out a predesigned search strategy with the independent selection, and data were extracted by two reviewers (ZLG and LLG). Moreover, potential reporting bias is likely to exist, and we could not get any important additional information from the primary authors.

Fourth, all eligible reports were based on retrospective population-based studies, which may introduce recall limitation. Hence, the integrity of records may weaken the reliability of the results to a certain extent. However, as an adverse effect of ADT, TEs may not be the main end-points that randomized controlled trials focus on, and strict inclusion in randomized controlled trials might lead to limitations in external validity, as a result38. Finally, we can only relatively evaluate the association between different types of ADT and TE (DVT and PE) incidence. Therefore, to investigate adverse drug reactions, carrying out large-scale randomized studies with long follow-up duration and high quality of design and implementation is further credible.

ADT can increase the risk of TEs. Men exposed to ADT, which was defined as GnRH agonists alone, GnRH plus AA, and AA alone, displayed a high risk of DVT after controlling the demographic and disease characteristics and other confounding factors, although statistically significant difference was not observed in orchiectomy group. Additionally, GnRH agonists alone and orchiectomy can significantly result in PE. These findings may contribute to the clinician’s awareness regarding the potential risk of ADT and ensure clinical management when prescribing this treatment. Additional randomized studies should also focus on other TEs and possible variations by age and tumor stage because they require different approaches to treatment.

Author contributions SSW conceived the study idea. ZLG, YYH, LLG, FLC, SG, and STX performed a literature search, study selection, and data extraction. ZLG, YYH, CMG, STX, and SSW performed statistical analyses and interpretation of corresponding results. ZLG drafted the initial manuscript. SSW modified the initial manuscript. SSW had primary responsibility for final content. All authors made a critical comment for the initial manuscript.

Compliance with ethical standards
Conflict of interest The authors declare that they have no conflict of interest.
Acknowledgments This study was supported by grants from the National Natural Science Foundation of China (Grant No. 61301294). The supporting institution had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Zhenlang Guo1, Yiyu Huang1, Leiliang Gong2, Shu Gan3, Franky Leung Chan4, Chiming Gu3, Songtao Xiang3, Shusheng Wang3
Received: 4 February 2018 / Revised: 22 March 2018 / Accepted: 27 March 2018 / Published online: 9 July 2018 © Macmillan Publishers Limited, part of Springer Nature 2018

1. Ferlay J, Shin HR, Bray F, Forman D, Mathers C, Parkin DM. Estimates of worldwide burden of cancer in 2008: GLOBOCAN 2008. Int J Cancer. 2010;127:2893–917.
2. Buja A, Lange JH, Perissinotto E, Rausa G, Grigoletto F, Canova C, Mastrangelo G. Cancer incidence among male military and civil pilots and flight attendants: an analysis on published data. Toxicol Ind Health. 2005;21:273–82.
3. FDA Drug Safety Communication. FDA requests label changes and single-use packaging for some over-the-counter topical antiseptic products to decrease risk of infection. Clin Infect Dis. 2014;58:i–ii.
4. Falchook AD, Basak R, Mohiuddin JJ, et al. Use of androgen deprivation therapy with radiotherapy for intermediate-and highrisk prostate cancer across the United States. JAMA Oncol. 2016;2:1236–8.
5. Klotz L, Higano CS. Intermittent androgen deprivation therapy-an important treatment option for Prostate Cancer. JAMA Oncol. 2016;2:1531–2.
6. Bekelman JE, Mitra N, Handorf EA, et al. Effectiveness of androgen-deprivation therapy and radiotherapy for older men with locally advanced prostate cancer. J Clin Oncol. 2015;33:716–22.
7. Bruce JY, Lang JM, McNeel DG, et al. Current controversies in the management of biochemical failure in prostate cancer. Clin Adv Hematol Oncol. 2012;10:716–22.
8. Sartor O, Silberstein J. Prostate cancer: primary ADT monotherapy not suitable for localized disease. Nat Rev Urol. 2014;11:309–10.
9. Gilbert SM, Kuo YF, Shahinian VB. Prevalent and incident use of androgen deprivation therapy among men with prostate cancer in the United States. Urol Oncol. 2011;29:647–53.
10. Sammon JD, Abdollah F, Reznor G, et al. Patterns of declining use and the adverse effect of primary androgen deprivation on all cause mortality in elderly men with prostate cancer. Eur Urol. 2015;68:32–9.
11. Heidenreich A, Aus G, Bolla M, et al. EAU guidelines on prostate cancer. Eur Urol. 2008;53:68–80.
12. Keating NL, O’ Malley AJ, Freedland SJ, et al. Diabetes and cardiovascular disease during androgen deprivation therapy: observational study of veterans with prostate cancer. J Natl Cancer Inst. 2010;102:39–46.
13. Li S, Li X, Li J, et al. Experimental arterial thrombosis regulated by androgen and its receptor via modulation of platelet activation. Thromb Res. 2007;121:127–34.
14. Zhao J, Zhu S, Sun L, et al. Androgen deprivation therapy for prostate cancer is associated with cardiovascular morbidity and mortality: a meta-analysis of population-based observational studies. PLoS One. 2014;9:e107516.
15. Azoulay L, Yin H, Benayoun S, et al. Androgen-deprivation therapy and the risk of stroke in patients with prostate cancer. Eur Urol. 2011;60:1244–50.
16. Hu JC, Williams SB, O’Malley AJ, et al. Androgen-deprivation therapy for nonmetastatic prostate cancer is associated with an increased risk of peripheral arterial disease and venous thromboembolism. Eur Urol. 2012;61:1119–28.
17. Nguyen-Nielsen M, Borre M, Horváth-Puhó E, et al. Risk of venous thromboembolism among prostate cancer patients treated with androgen deprivation therapy in Denmark: a population-based cohort study, 1997-2011. Eur Urol Suppl.2014;13:e976.

18. Klil-Drori AJ, Yin H, Tagalakis V, et al. Androgen deprivation therapy for prostate cancer and the risk of venous thromboembolism. Eur Urol. 2016;70:56–61.
19. Higgins JPT, Green S. Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0. The Cochrane Collaboration. 2011.
20. Stroup DF, Berlin JA, Morton SC, Olkin I, Williamson GD, Rennie D, Moher D, Becker BJ, Sipe TA, Thacker SB. Meta-analysis of observational studies in epidemiology: a proposal for reporting. JAMA. 2000;283:2008e12.
21. Stang A. Critical evaluation of the Newcastle-Ottawa scale for the assessment of the quality of non-randomized studies in meta-analyses. Eur J Epidemiol. 2010;25:603–5.
22. Egger M, Smith GD, Schneider M, et al. Bias in meta-analysis detected by a simple, graphical test. BMJ. 1997;315:629.
23. Van Hemelrijck M, Adolfsson J, Garmo H, et al. Risk of thromboembolic diseases in men with prostate cancer: results from the population-based PCBaSe Sweden. Lancet Oncol. 2010;11:450–8.
24. Yii SC, Chung SD, Huang CY. The association between androgen deprivation therapy and pulmonary embolism: a population-based study. J Urol. 2017;24:36–7.
25. Higgins JPT, Thompson SG. Controlling the risk of spurious findings from meta-regression. Stat Med. 2004;23:1663–82.
26. Knapp G, Hartung J. Improved tests for random effects meta-regression with a single covariate. Stat Med. 2003;22: 2693–710.
27. Ehdaie B, Atoria CL, Gupta A, et al. Androgen deprivation and thromboembolic events in men with prostate cancer. Cancer. 2012;118:3397–406.
28. O’ Farrell S, Sandström K, Garmo H, et al. Risk of thromboembolic disease in men with prostate cancer undergoing androgen deprivation therapy. BJU Int. 2016;118:391–8.
29. Levitan N, Dowlati A, Remick SC, et al. Rates of initial and recurrent thromboembolic disease among patients with malignancy versus those without malignancy. Risk analysis using Medicare claims data. Medicine (Baltimore). 1999;78:285–91. 30. Brown BW, Brauner C, Minnotte MC. Noncancer deaths in white adult cancer patients. J Natl Cancer Inst. 1993;85:979–87.
31. Shahinian VB, Kuo YF, Freeman JL, et al. Increasing use of gonadotropin-releasing hormone agonists for the treatment of localized prostate carcinoma. Cancer. 2005;103:1615–24.
32. Shahinian VB, Kuo YF, Gilbert SM. Reimbursement policy and androgen-deprivation therapy for prostate cancer. N Engl J Med. 2010;363:1822–32.
33. Weight CJ, Klein EA, Jones JS. Androgen deprivation falls as orchiectomy rates rise after changes in reimbursement in the U.S. Medicare population. Cancer. 2008;112:2195–201.
34. Smith MR, Lee H, Nathan DM. Insulin sensitivity during combined androgen blockade for prostate cancer. J Clin Endocrinol Metab. 2006;91:1305–8.
35. Dockery F, Bulpitt CJ, Agarwal S, et al. Testosterone suppression in men with prostate cancer leads to an increase in arterial stiffness and hyperinsulinemia. Clin Sci (Lond). 2003;104:195–201.
36. Bolla M, de Reijke TM, Van Tienhoven G, et al. Duration of androgen suppression in the treatment of prostate cancer. N Engl J Med. 2009;360:2516–27.
37. D’ Amico AV, Denham JW, Bolla M, et al. Short- vs long-term androgen suppression plus external beam radiation therapy and survival in men of advanced age with node-negative high-risk adenocarcinoma of the prostate. Cancer. 2007;109:2004–10. 38. Zumsteg ZS, Zelefsky MJ. Short-term androgen deprivation therapy for patients with intermediate-risk prostate cancer undergoing dose-escalated radiotherapy: the standard of care? Lancet Oncol. 2012;13:e259–e269.

Read More: A Commentary from the Editor of PCAN - Andrew J. Armstrong, MD