Secondary Logo

Journal Logo

Mortality, cardiovascular risk, and androgen deprivation therapy for prostate cancer

A systematic review with direct and network meta-analyses of randomized controlled trials and observational studies

Scailteux, Lucie-Marie PharmD; Naudet, Florian MD, PhD; Alimi, Quentin; Vincendeau, Sébastien MD; Oger, Emmanuel MD, PhD

Section Editor(s): Lorenzo., Giuseppe Di

doi: 10.1097/MD.0000000000003873
Research Article: Systematic Review and Meta-Analysis

Androgen deprivation therapy (ADT) is a cornerstone therapy for advanced prostate cancer (PCa). We hypothesized that cardiovascular (CV) risk is different across the various ADT modalities to compare their effects on CV morbidity and mortality, and all-cause mortality in patients with PCa. To investigate more in depth potential CV risk heterogeneity focusing on coronary (main outcome) and cerebrovascular risk, CV, and overall mortality. We performed a Medline and Embase query, without language restriction, since 1950 up to July 2014. We included randomized controlled trials (RCTs) and observational studies providing that they compared at least 1 ADT modality to another one or to placebo and they gave data on CV event or all-cause mortality. Sixty-eight studies out of 3419 met our eligibility criteria. Eleven observational studies were analyzed. Direct meta-analyses showed that antiandrogen was associated with a 30% decrease risk for myocardial infarction (MI) compared to GnRH agonists (RR, 0.70 [0.54–0.91]); combined androgen blockade (CAB) was associated with a 10% increase risk for stroke when compared to antiandrogen (RR, 1.10 [1.02–1.19]). With regard to RCTs, 57 were included: direct meta-analyses suggested that CAB was associated with a 10% decrease of all-cause mortality when compared to GnRH agonist (RR, 0.90 [0.82–1.00]). Network analysis could only be performed for all-cause mortality and it remains difficult to disentangle benefit (positive impact on cancer survival) and risk (including CV risk). The impact of the ADT modalities on CV morbidity remains difficult to quantify and more detailed prospective collection is required. Registration: PROSPERO, CRD42014010598.

aPharmacovigilance, Pharmacoepidemiology and Drug Information Center, Rennes University Hospital, Rennes, France

bClinical Investigation Center, INSERM 1414, Rennes University Hospital and University of Rennes 1, Rennes, France

cUrology Department, Rennes University Hospital, Rennes, France.

Correspondence: Lucie-Marie Scailteux, Rennes Hospital University, Rennes, France (e-mail:

Abbreviations: AA = antiandrogen, ABIRA = abiraterone, ADT = androgen deprivation therapy, AS = androgen suppression, BT = brachytherapy, CAB = Combined androgen blockade = GnRH agonist + antiandrogen, CMD = chlormadinone acetate, CPT = cyproterone acetate, CV = cardiovascular, DES = diethylstilbestrol, ENZ = enzalutamide, GnRH = gonadotropin-releasing hormone, LHRH = luteinizing hormone-releasing hormone ( = GnRH), M0 = non metastatic disease, M1 = metastatic disease, MI = myocardial infarction, NA = not available, OT = orchiectomy, RCT = randomized controlled trials, RT = radiotherapy, SEER = Surveillance, Epidemiology, and End Results Program.

Authorship: LMS, FN, EO, and SV conceptualized and designed the study. LMS and QA selected articles for inclusion or exclusion, extracted data from included articles. LMS, FN, and EO analyzed the data and LMS drafted the initial manuscript. All authors contributed to data interpretation, critically reviewed and revised the manuscript, and approved the final manuscript for submission. All authors agree to be accountable for all aspects of the work by ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

Funding: This work was supported by the Rennes University Hospital.

Registration: In accordance with the guidelines, our systematic review protocol was registered with the International Prospective Register of Systematic Reviews (PROSPERO) on July 21, 2014 and was last updated on July 6, 2015 (registration number CRD42014010598).

The authors have no conflicts of interest to disclose.

Supplemental Digital Content is available for this article.

Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal's Website (

This is an open access article distributed under the Creative Commons Attribution License 4.0 (CCBY), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Received October 30, 2015

Received in revised form April 13, 2016

Accepted May 14, 2016

Back to Top | Article Outline

1 Introduction

Prostate cancer (PCa) is the most frequently diagnosed male cancer in the United States (US) and Europe.[1] It is the third leading cause of cancer-related death, yet an increasing number of men are living longer with PCa. Androgen deprivation therapy (ADT) is considered a cornerstone treatment for advanced symptomatic metastatic disease yet it is also used alone or in association with radiotherapy (RT) to treat less advanced tumors.[2]

Although data suggested that ADT, when associated with RT, can improve survival, its impact on cardiovascular (CV) event and CV risk is still controversial. Past[3] and more recent health-record-based studies[4,5] have reported positive association. The US FDA has recognized the potential adverse cardiometabolic profile of ADT[6] and has recommended that a patients’ CV risk be assessed prior to treatment.[7,8]

However, when focusing on type of ADT modality, results from observational studies suggested heterogeneity in the risk for myocardial infarction (MI) as well as for stroke.[5,9,10] However, indication bias remains a challenge, in particular in observational studies, the prescribers taking into account the individual benefic/risk balance and notably the comorbidities of their patients to choose the most suitable ADT modality. Nevertheless, risk heterogeneity across the different ADT modalities (GnRH agonist or antagonist, antiandrogens [AA], etc.) is plausible and may be explained by their different pharmacologic actions.

In randomized controlled trials (RCTs), secondary safety outcomes such as MI, stroke, or CV death were less frequently evaluated than all-cause mortality, particularly in patients with comorbidities. Several direct meta-analyses did not detect a statistically significant difference between maximal androgen blockade and GnRH agonist monotherapy,[11] ADT (predominantly GnRH agonist) and no treatment,[12] GnRH agonist and GnRH antagonist in patients without CV disease and naïve of any cancer treatment.[13] Insufficient power and classification bias on such safety secondary outcomes prevent a definitive conclusion.

Therefore, to investigate more in depth potential CV risk heterogeneity, we performed direct and network meta-analyses comparing ADT modalities within each other (i.e., GnRH agonist versus complete androgen blockade [CAB], AA vs CAB, etc.). We focused on coronary and cerebrovascular risk, CV, and overall mortality. We included observational studies and RCTs in all PCa stage patients because their different designs (patients’ selection, main outcome studied, comparability, follow-up duration, etc.) lead to different biases and their results are thought to be complementary, thus they should be summarized but we subgrouped them by type of studies.

Back to Top | Article Outline

2 Evidence acquisition

2.1 Eligibility criteria

We undertook this study in accordance with the MOOSE and PRISMA statement.[14–16] We looked for RCTs and observational studies published up to July 28, 2014 without language restriction provided that they gave data on hormone sensitive PCa patients comparing 1 ADT modality to another or to either RT or total prostatectomy or placebo and that they considered MI, ischemic stroke, CV death, and all-cause mortality as primary or secondary (safety) outcomes.

Eligible ADT modalities, grouped in pharmacological classes each considered clinically homogeneous, were the following: GnRH agonists (buserelin, leuprorelin, goserelin, etc.), GnRH antagonists (degarelix, abarelix), antiandrogens (steroidal: cyproterone [CPT] or nonsteroidal: flutamide, bicalutamide, etc.), estrogen (diethylstilbestrol, polyestradiol phosphate, estradiol, etc.), and orchiectomy (OT). CAB was defined as an association of AA and GnRH agonist.

Because new drugs (abiraterone [ABIRA] or enzalutamide [ENZ]) were evaluated on top of ADT in castrate-resistant PCa patients, we decided to exclude those trials as our target population was hormone-sensitive PCa patients.

The primary outcome was MI. Secondary outcomes were ischemic stroke, CV death, and all-cause mortality. CV death included all patients who died by an ischemic process (coronary heart disease, ischemic heart disease, acute MI, stroke); we excluded death by congestive heart failure, arrhythmia, sudden cardiac death, deep vein thrombosis, and pulmonary or arterial embolism. For studies with insufficient detail on CV death, the outcome was extracted when it was reported (“cardiovascular death” or “cardiovascular mortality,” following the authors’ definition).

Back to Top | Article Outline

2.2 Search strategy

Literature search using Medline and Embase. We included MESH terms of all synonyms such as: PCa, prostatic neoplasm, targeted drug classes (gonadotropin releasing hormone agonist, luteinizing hormone releasing hormone agonist, etc.) and molecule name (flutamide, goserelin, etc.). The search formulated by LMS was reviewed by EO. For complete query see Appendix Text 1, We included grey literature such as letters and abstracts presented at relevant conference meeting. Title, abstracts, and full-text screening was performed in duplicate by LMS and QA. References list of obtained articles were hand searched. This review was registered in PROSPERO database (CRD42014010598).

Back to Top | Article Outline

2.3 Data extraction and study selection

LMS and QA independently extracted data from the selected studies into a standardized spreadsheet. Discrepancies were resolved by discussion until consensus was reached. When a publication was written in a language not fluently spoken by one of the 2 main reviewers, a translator did the extraction and the work was validated with an English-language extraction. The inclusion of data from multiple reports as separate studies (duplicate, overlapping, or companion studies) was allowed only when targeted outcomes were different. For observational studies pooling several ADT modalities, the author was contacted to obtain details on each ADT group. To avoid misestimating risk related to a specific ADT modality, studies not clearly defining drug exposure were excluded. When there were missing data on a specific outcome, we attempted to contact authors to obtain the relevant missing data. If data were not obtained, the study was discarded from the analysis on that specific outcome.

Back to Top | Article Outline

2.4 Data collection

The following variables were recorded: details of study (year, design, name or registration number, country, financial support, total number of participants, follow-up duration, type of analysis in RCT); details of participants (median age, previous PCa treatment, cancer stage [T score, metastasis]); regimens (class, drugs, dose, timing of administration, length of treatment, number treated); outcomes measure (number of events for each treatment modality).

Back to Top | Article Outline

2.5 Quality assessment

LMS and EO independently assessed study quality using the Joanna Briggs reviewer's manual[17] for evaluating study biases using different tools for RCTs and for observational studies. Disagreements were resolved first by discussion and then by consulting a third author for arbitration.

Back to Top | Article Outline

2.6 Data analysis

Direct meta-analyses integrate the results of multiple independent studies addressing the same comparison. By extension, network meta-analyses allow inferences into the comparative effectiveness of those therapies that may or may not have been directly compared against each other, providing the network is connected.

For each outcome, adjusted risk estimates provided were chosen. If it was not given, we determined treatment effect along with 95% confidence interval (CI) from available raw data.

The estimate of overall effect (summary measure) was calculated with its 95% CI for each pair wise meta-analysis (head to head direct evidence) using random effects models separately for observational studies and RCTs through SAS macros.[18] To meta-analyze studies including no event in at least one arm, we used the statistical methods described by Kuss.[19] Statistical heterogeneity was documented with the I2 statistic (50–90%: may represent substantial heterogeneity) and investigated graphically by inspecting forest plots.[20] We then considered the variability in participant factors among trials and trial factors.

A stratification analysis on T stage could not be performed owing to the fact that these data were not always available; however, all T stages were homogeneously represented across studies without overrepresentation of a stage in particular.

Network meta-analysis was performed for RCTs. We used the graph–theoretical method[21] (see Appendix Text 2, for the script network meta-analysis). Results were reported in terms of OR and 95% CI. We used a design-based decomposition of Cochran Q for assessing the homogeneity in the whole network, the homogeneity within designs, and the homogeneity/consistency between designs. It allows also an assessment of the consistency assumption after detaching the effect of single designs. We used a net heat plot, a graphical tool for locating inconsistency.[22]

Analyses were run with R statistical package[23] and the netmeta library.[24]

Publication bias was investigated graphically using funnel plots for each meta-analysis when there were at least 4 studies. Funnel plot asymmetry was tested using the rank correlation test when there were at least 10 studies.[25]

Back to Top | Article Outline

2.7 Role of the funding source

This study received no funding.

Back to Top | Article Outline

2.8 Ethical review

Ethical approval was not necessary considering we used already published studies.

Back to Top | Article Outline

3 Results

Of the 3614 articles identified, 3419 left after deduplication (Fig. 1). After selection on the abstract then on full-text, 68 studies met our eligibility criteria: 11 observational studies and 57 RCTs. See Appendix eTable 1 which describes observational studies and Appendix eTable 2 which describes RCTs.

Figure 1

Figure 1

Back to Top | Article Outline

3.1 Observational studies

Studies pooling different ADT modalities were excluded, as well as CAPSURE or SEER[26] studies which did not distinguish if LHRH agonist was alone or associated with AA. In the Jespersen et al study on the Danish Cancer Registry,[10] the ICD-10 code BWHC covers LHRH agonist and AA modalities, and the isolated effect of LHRH agonists could not be assessed.

The 11 observational studies selected including 193,620 patients. Five studies[5,9,27–29] gave data on CV morbidity (coronary and/or cerebrovascular risk) but only 1 study[30] on CV death. The 6 studies given data on all-cause mortality (see Appendix Biblio, did not compare the same ADT modalities and could not be included in the meta-analysis.

Table 1 displays results from observational studies of the most frequently used ADT modalities. In synthesis, we observed an increased risk for stroke with CAB compared to AA (RR, 1.10 [1.02–1.19]); an increased risk for MI with GnRH agonist when compared to AA (RR, 1.43 [1.10–1.85]) and a consistent statistically nonsignificant association as regards stroke (RR, 1.22 [0.93–1.61]); thus AA appeared associated with a lower CV risk than GnRH alone or CAB. Appendix eTable 3 shows all comparisons with head to head direct comparison. When a relevant comparison (2 ADT compared in each other) was not available, we recalculated a relative risk from raw data. Data on CV death and all-cause mortality were sparse in observational studies.

Table 1

Table 1

See Appendix eTable 4 for quality assessment.

See Appendix eFigure 1 for funnel plot for publication bias. Of note, the limited number of studies (only 4 or 5 studies per head to head comparisons) hampered clearly ruling out publication bias.

Back to Top | Article Outline

3.2 Randomized controlled trials

Fifty-seven RCTs were included with 31,037 patients. Seven studies contained data on MI[31–37] (101 events out of 2243 patients), 6 on stroke[31–37] (35 events out of 2008 patients), 18 on CV death[32–36,38–49] (1126 CV deaths out of 7787 patients), and 47 on all-cause mortality (see Appendix Biblio, (11,498 deaths out of 28,643 patients). Many ADT modalities were found and notably for CAB which required precise definition: short-term CAB corresponded to 3 or 4 months treatment; long-term CAB to only 6 to 8 months treatment; continuous CAB was a very long term (>1 year or permanent) treatment contrary to intermittent CAB which was also given on a very long term but episodically often because of progression or relapse of PCa disease.

Two publications were used to extract data on different outcomes in each, respectively the RTOG study 92-02[44,50] and the degarelix study.[37,51] As regards the 6 RCTs comparing GnRH antagonist to GnRH agonist included in Albertsen meta-analysis,[13] only 1 trial fulfilled our eligibility criteria (CS21[37,51]); for CS28, CS30, and CS31, no data concerning our outcomes were available; CS35 and CS37 were currently not published: some results are available on the website but no data on CV risk or mortality (CV or overall) are displayed for CS35, and no results nor publication are available for CS37. We did not use estimates reported by Albertsen because eligibility was supported by access to individual data studies published or not.

Table 2 displays results of the most frequently used ADT modalities from RCT. Appendix eTable 5 shows all comparisons with head-to-head direct comparison focusing on trials comparing active therapies. When a treatment effect for our chosen outcomes was available, we recalculated a relative risk from raw data. MI, stroke, and CV death were rarely reported and direct meta-analyses included only 2 studies; no difference was detected. All-cause mortality was reported as a main outcome or described in the safety data. As shown in Table 2, GnRH agonist was associated with a slight decrease in risk of all-cause mortality when compared to OT (OR, 0.93 [0.86–1.00]), as continuous CAB when compared to GnRH agonist (OR, 0.90 [0.82–1.00]).

Table 2

Table 2

See Appendix eFigure 1 for the funnel plots for publication bias. Once again, the limited number of RCTs hampered clearly ruling out publication bias. See also Appendix eTable 6 for quality assessment.

Back to Top | Article Outline

3.3 Network meta-analysis

A closed network could only be drawn for all-cause mortality because there were too scarce data for MI, stroke, and CV death outcomes.

Table 3 shows comparisons between the most frequently used ADT modalities. Appendix eTable 7 shows all comparisons between the different ADT modalities (see also Appendix eFigure 2 when using LHRH agonist as the reference). Appendix eFigure 3 shows the network model for all-cause mortality: there were a large number of ADT modalities and few RCTs comparing them.

Table 3

Table 3

In synthesis, we observed that AA had a 23% increased risk for all-cause mortality compared to continuous CAB (RR, 1.23 [1.01–1.49]) and a somewhat similar increase though not statistically significant when compared to the other main ADT modalities. Otherwise, we did not detect a substantial difference in overall survival between GnRH agonist, GnRH antagonist, continuous CAB, and OT.

Inconsistency and heterogeneity were identified across the network; inconsistency could have influence comparison “AA–OT” or the “long-term CAB–short-term CAB” (Appendix eTable 8

Back to Top | Article Outline

4 Discussion

4.1 Main findings

Our results support the hypothesis that the various ADT modalities have a different impact on CV risk. Focusing on MI and stroke, we observed through a comprehensive quantitative synthesis (direct meta-analysis) of observational studies that CAB differed from AA which differed from GnRH agonists.

Back to Top | Article Outline

4.2 Strengths

Our systematic review encompassed a large panel of observational studies and RCTs.

We excluded studies with pooled ADT modalities in their analyses or without clear definition of CV outcomes to avoid including studies with potential misclassification either on drug exposure or outcome that would have blurred relevant data from other studies more suited to our specific purpose.

Finally, indirect network meta-analysis gave the opportunity to estimate treatment effect between 2 ADT modalities without head-to-head data available.

Back to Top | Article Outline

4.3 Limits

Direct meta-analyses included very few studies (at most 5 for all-cause mortality). In addition, they suffered from substantial heterogeneity which could be related to population characteristics and methodological parameters. Analysis of publication bias could not be ruled out as funnel plots included only 4 or 5 studies. Network meta-analysis suffered from inconsistency in some comparisons. Eventually, no firm conclusion could be drawn from these data.

Data on MI, stroke, and CV death were limited especially in RCTs. As regards all-cause mortality, it remains difficult to disentangle benefit with better survival through a positive impact on cancer progression and risk including CV risk. Cancer staging such as presence of metastasis and CV history are major issues and can induce a shorter survival duration compared with nonmetastatic patients who could have time to develop CV disorders and in whom CV death can be anticipated. The negative prognostic impact of severe comorbidity could also be due to cancer therapy adapted to comorbidity making it difficult to discern whether worse survival is due to comorbidity or less efficacious treatment. This is notably claimed by a study on the importance of comorbidity in cancer patients.[52] RCT included in our meta-analysis were rarely stratified on CV comorbidity including coronary heart disease or cerebrovascular disease, and we did not explore the risk of CV death nor all-cause mortality across CV comorbidity.

Back to Top | Article Outline

4.4 Comparison to other studies

Previous meta-analyses have been published but did not precisely address our hypothesis because of an analysis that pooled several ADT modalities,[11,12,53,54] a no treatment comparison group,[54,55] restrictive criteria, or different objectives. The first meta-analysis[11] including 27 trials which focused on metastatic (88%) and locally advanced (12%) PCa patients concluded that maximal androgen blockade [MAB] (OT + AA or CAB or OT + CPT) improved the 5-year survival by about 2% or 3% compared to androgen suppression alone [AS] (OT or LHRH agonist). The second meta-analysis[12] analyzed data from 8 studies enrolling nonmetastatic and nonhormone-refractory PCa patients and did not detect any evidence that immediate ADT (pooling several ADT modalities) increased CV death compared to no immediate ADT. The third study[53] reanalyzed data from the previous meta-analysis[12] and showed a nonsignificant association between CV death and ADT use. The fourth meta-analysis[55] included 8 observational studies assessing the risk of fatal and nonfatal CV event with different ADT modalities compared to no treatment, irrespective of PCa stage. A fifth and more recent meta-analysis identified 6 population-based observational studies comparing ADT modality versus watchful waiting or active surveillance.[54] Some studies used data from SEER database[26] or some national cancer registries[10] where extracting codes included many modalities, such as LHRH agonist and AA without clearly distinguishing them. Results indicated that LHRH agonists were associated with an increased risk for stroke and MI (fatal or not) and that AA were associated with an increased risk for any nonfatal CV disease compared to no treatment. We excluded these studies to avoid any mixing between ADT modalities. Nevertheless, we also found an increased risk of MI and stroke with LHRH agonist versus no endocrine treatment, as well as OT when compared with no endocrine treatment.

Another meta-analysis[13] focused on 6 trials comparing LHRH agonist to LHRH antagonist in metastatic and nonmetastatic, locally or not advanced PCa patients, naïve of ADT treatment. Safety data including CV morbidity, CV mortality, and all-cause mortality were scarce due to the small follow-up duration (12 months); in 2 trials, data were not available or published. Results indicated, in patients with CV history, a decrease in cardiac event with LHRH antagonist patients compared to LHRH agonist. The last meta-analysis focused on 8 RCTs comparing intermittent androgen deprivation to continuous androgen deprivation and did not detect any difference in overall survival.[56]

From a pharmacological point of view, a differential impact of the various ADT modalities on CV risk might be explained. Studies have established that ADT could increase weight gain, body fat percentage, triglycerides rate, and decrease lean body mass, and insulin sensitivity.[57–59] The link with diabetes, metabolic syndrome, and ADT is claimed by some authors[58,59]; those metabolic changes may increase CV risk but the underlying mechanism remains unclear. ADT suppress androgen activity by central or peripheral mechanisms: by decreasing testicular and/or extra gonadal androgen production with GnRH agonist and antagonist or by blocking androgen receptor activation using AA. GnRH receptors are synthesized in several extrapituitary tissues as well as the reproductive system,[60] bladder,[61] or heart.[62] Dong et al[63] observed an impact of the GnRH agonist on cardiomyocytes contractile function in mice. Other authors[64,65] suggested, through studies on human mononuclear cells, that GnRH receptors located on T-lymphocytes could indirectly explain a modification of the stability of the atheromatous plaques due to their activation and proliferation after administration of GnRH agonist (T-lymphocytes are the main immune cells infiltrating the atheromatous plaques). This hypothesis could explain the increase of cerebrovascular and coronary heart diseases observed with GnRH agonists.

Back to Top | Article Outline

5 Conclusion

Our results support the hypothesis that the various ADT modalities have different impact as regards CV risk.

However, we should be cautious and consider that the question is currently not totally resolved. RCT does not seem adapted to this issue and we are currently conducting a large nationwide population-based study (ADTCR) using the French medico-administrative database.

Back to Top | Article Outline


[1]. Droz J-P, Aapro M, Balducci L, et al. Management of prostate cancer in older patients: updated recommendations of a working group of the International Society of Geriatric Oncology. Lancet Oncol 2014;15:e404–14.
[2]. Allan CA, Collins VR, Frydenberg M, et al. Androgen deprivation therapy complications. Endocr Relat Cancer 2014;21:T119–29.
[3]. Byar DP, Corle DK. Hormone therapy for prostate cancer: results of the Veterans Administration Cooperative Urological Research Group studies. NCI Monogr 1988;7:165–70.
[4]. Saigal CS, Gore JL, Krupski TL, et al. Androgen deprivation therapy increases cardiovascular morbidity in men with prostate cancer. Cancer 2007;110:1493–500.
[5]. Van Hemelrijck M, Garmo H, Holmberg L, et al. Absolute and relative risk of cardiovascular disease in men with prostate cancer: results from the population-based PCBaSe Sweden. J Clin Oncol 2010;28:3448–56.
[6]. FDA. FDA Drug Safety Communication: Update to Ongoing Safety Review of GnRH Agonists and Notification to Manufacturers of GnRH Agonists to Add New Safety Information to Labeling Regarding Increased Risk of Diabetes and Certain Cardiovascular Diseases [10-20-2010]. Accessed May 24, 2015.
[7]. Levine GN, D’Amico AV, Berger P, et al. Androgen-deprivation therapy in prostate cancer and cardiovascular risk: a science advisory from the American Heart Association, American Cancer Society, and American Urological Association: endorsed by the American Society for Radiation Oncology. Circulation 2010;121:833–40.
[8]. Heidenreich A, Bastian PJ, Bellmunt J, et al. EAU guidelines on prostate cancer. Part II: treatment of advanced, relapsing, and castration-resistant prostate cancer. Eur Urol 2014;65:467–79.
[9]. 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.
[10]. Jespersen CG, N⊘rgaard M, Borre M. Androgen-deprivation therapy in treatment of prostate cancer and risk of myocardial infarction and stroke: a nationwide Danish population-based cohort study. Eur Urol 2014;65:704–9.
[11]. Prostate Cancer Trialists Collaborative Group. Maximum androgen blockade in advanced prostate cancer: an overview of the randomised trials. Prostate Cancer Trialists’ Collaborative Group. Lancet 2000;355:1491–8.
[12]. Nguyen PL, Je Y, Schutz FAB, et al. Association of androgen deprivation therapy with cardiovascular death in patients with prostate cancer: a meta-analysis of randomized trials. JAMA 2011;306:2359–66.
[13]. Albertsen PC, Klotz L, Tombal B, et al. Cardiovascular morbidity associated with gonadotropin releasing hormone agonists and an antagonist. Eur Urol 2014;65:565–73.
[14]. Liberati A, Altman DG, Tetzlaff J, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. Ann Intern Med 2009;151:W65–94.
[15]. Shamseer L, Moher D, Clarke M, et al. Preferred reporting items for systematic review and meta-analysis protocols (PRISMA-P) 2015: elaboration and explanation. BMJ 2015;349:g7647.
[16]. Stroup DF, Berlin JA, Morton SC, et al. Meta-analysis of observational studies in epidemiology: a proposal for reporting. JAMA 2000;283:2008–12.
[17]. The Joanna Briggs Institute, The University of Adelaide. Joanna Briggs Institute. Reviewer's Manual—2011 Edition; 2011.
[18]. Senn R, Weir J, Hua T, et al. Creating a suite of macros for meta-analysis in SAS: a case study in collaboration. Stat Probab Lett 2011;81:842–51.
[19]. Kuss O. Statistical methods for meta-analyses including information from studies without any events-add nothing to nothing and succeed nevertheless. Stat Med 2015;34:1097–116.
[20]. Hedges LV, Olkin I. Statistical Method for Meta-Analysis. London: Academic Press; 1985.
[21]. Rücker G. Network meta-analysis, electrical networks and graph theory. Res Synth Methods 2012;3:312–24.
[22]. Krahn U, Binder H, König J. A graphical tool for locating inconsistency in network meta-analyses. BMC Med Res Methodol 2013;13:35doi:10.1186/1471-2288-13-35.
[23]. R Development Core Team. R: A Language and Environment for Statistical Computing. Vienna, Austria: R Foundation for Statistical Computing; 2009.
[24]. Rücker G, Schwarzer G, Krahn U, et al.. Package “netmeta”. Network Meta-Analysis using Frequentist Methods (Version 0.7-0). February 2015.
[25]. Sterne JAC, Sutton AJ, Ioannidis JPA, et al. Recommendations for examining and interpreting funnel plot asymmetry in meta-analyses of randomised controlled trials. BMJ 2011;343:d4002.
[26]. Keating NL, O’Malley AJ, Smith MR. Diabetes and cardiovascular disease during androgen deprivation therapy for prostate cancer. J Clin Oncol 2006;24:4448–56.
[27]. Robinson D, Garmo H, Lindahl B, et al. Ischemic heart disease and stroke before and during endocrine treatment for prostate cancer in PCBaSe Sweden. Int J Cancer 2012;130:478–87.
[28]. 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.
[29]. Martín-Merino E, Johansson S, Morris T, et al. Androgen deprivation therapy and the risk of coronary heart disease and heart failure in patients with prostate cancer: a nested case-control study in UK primary care. Drug Saf 2011;34:1061–77.
[30]. Bittner N, Merrick GS, Galbreath RW, et al. Primary causes of death after permanent prostate brachytherapy. Int J Radiat Oncol Biol Phys 2008;72:433–40.
[31]. Waymont B, Lynch TH, Dunn JA, et al. Phase III randomised study of zoladex versus stilboestrol in the treatment of advanced prostate cancer. Br J Urol 1992;69:614–20.
[32]. Lukkarinen O, Kontturi M. Comparison of a long-acting LHRH agonist and polyoestradiol phosphate in the treatment of advanced prostatic carcinoma. An open prospective, randomized multicentre study. Scand J Urol Nephrol 1994;28:171–8.
[33]. Robinson MR, Smith PH, Richards B, et al. The final analysis of the EORTC Genito-Urinary Tract Cancer Co-Operative Group phase III clinical trial (protocol 30805) comparing orchidectomy, orchidectomy plus cyproterone acetate and low dose stilboestrol in the management of metastatic carcinoma of the prostate. Eur Urol 1995;28:273–83.
[34]. Chang A, Yeap B, Davis T, et al. Double-blind, randomized study of primary hormonal treatment of stage D2 prostate carcinoma: flutamide versus diethylstilbestrol. J Clin Oncol Off J Am Soc Clin Oncol 1996;14:2250–7.
[35]. Mikkola AK, Ruutu ML, Aro JL, et al. Parenteral polyoestradiol phosphate vs orchidectomy in the treatment of advanced prostatic cancer. Efficacy and cardiovascular complications: a 2-year follow-up report of a national, prospective prostatic cancer study. Finnprostate Group. Br J Urol 1998;82:63–8.
[36]. Schröder FH, Whelan P, De Reijke TM, et al. Metastatic prostate cancer treated by Flutamide versus Cyproterone acetate: final analysis of the “European Organization for Research and Treatment of Cancer” (EORTC) protocol 30892. Eur Urol 2004;45:457–64.
[37]. Smith MR, Klotz L, Persson B-E, et al. Cardiovascular safety of degarelix: results from a 12-month, comparative, randomized, open label, parallel group phase III trial in patients with prostate cancer. J Urol 2010;184:2313–9.
[38]. Aro J, Ruutu M, Juusela H, et al. Polyestradiol phosphate (160 mg/month) or LHRH analog (buserelin depot) in the treatment of locally advanced or metastasized prostatic cancer. The Finnprostate Group. Ann Chir Gynaecol Suppl 1993;206:5–8.
[39]. Iversen P, Rasmussen F, Klarskov P, et al. Long-term results of Danish Prostatic Cancer Group trial 86. Goserelin acetate plus flutamide versus orchiectomy in advanced prostate cancer. Cancer 1993;72(12 suppl):3851–4.
[40]. Vogelzang NJ, Chodak GW, Soloway MS, et al. Goserelin versus orchiectomy in the treatment of advanced prostate cancer: final results of a randomized trial. Zoladex Prostate Study Group. Urology 1995;46:220–6.
[41]. Zalcberg JR, Raghaven D, Marshall V, et al. Bilateral orchidectomy and flutamide versus orchidectomy alone in newly diagnosed patients with metastatic carcinoma of the prostate—an Australian multicentre trial. Br J Urol 1996;77:865–9.
[42]. Bono AV, DiSilverio F, Robustelli della Cuna G, et al. Complete androgen blockade versus chemical castration in advanced prostatic cancer: analysis of an Italian multicentre study. Italian Leuprorelin Group. Urol Int 1998;60(suppl 1):18–24.
[43]. De Voogt HJ, Studer U, Schroder FH, et al. Maximum androgen blockade using LHRH agonist buserelin in combination with short-term (two weeks) or long-term (continuous) cyproterone acetate is not superior to standard androgen deprivation in the treatment of advanced prostate cancer. Final analysis of EORTC GU group trial 30843. Eur Urol 1998;33:152–8.
[44]. Efstathiou JA, Bae K, Shipley WU, et al. Cardiovascular mortality and duration of androgen deprivation for locally advanced prostate cancer: analysis of RTOG 92-02. Eur Urol 2008;54:816–23.
[45]. Roach M III, Bae K, Speight J, et al. Short-term neoadjuvant androgen deprivation therapy and external-beam radiotherapy for locally advanced prostate cancer: long-term results of RTOG 8610. J Clin Oncol 2008;26:585–91.
[46]. Calais da Silva FEC, Bono AV, Whelan P, et al. Intermittent androgen deprivation for locally advanced and metastatic prostate cancer: results from a randomised phase 3 study of the South European Uroncological Group. Eur Urol 2009;55:1269–77.
[47]. Crook J, Ludgate C, Malone S, et al. Final report of multicenter Canadian Phase III randomized trial of 3 versus 8 months of neoadjuvant androgen deprivation therapy before conventional-dose radiotherapy for clinically localized prostate cancer. Int J Radiat Oncol Biol Phys 2009;73:327–33.
[48]. Bolla M, Van Tienhoven G, Warde P, et al. External irradiation with or without long-term androgen suppression for prostate cancer with high metastatic risk: 10-year results of an EORTC randomised study. Lancet Oncol 2010;11:1066–73.
[49]. Denham JW, Steigler A, Lamb DS, et al. Short-term neoadjuvant androgen deprivation and radiotherapy for locally advanced prostate cancer: 10-year data from the TROG 96.01 randomised trial. Lancet Oncol 2011;12:451–9.
[50]. Horwitz EM, Bae K, Hanks GE, et al. Ten-year follow-up of radiation therapy oncology group protocol 92-02: a phase III trial of the duration of elective androgen deprivation in locally advanced prostate cancer. J Clin Oncol 2008;26:2497–504.
[51]. Klotz L, Boccon-Gibod L, Shore ND, et al. The efficacy and safety of degarelix: a 12-month, comparative, randomized, open-label, parallel-group phase III study in patients with prostate cancer. BJU Int 2008;102:1531–8.
[52]. Piccirillo JF, Tierney RM, Costas I, et al. Prognostic importance of comorbidity in a hospital-based cancer registry. JAMA 2004;291:2441–7.
[53]. Bourke L, Kirkbride P, Hooper R, et al. Endocrine therapy in prostate cancer: time for reappraisal of risks, benefits and cost-effectiveness? Br J Cancer 2013;108:9–13.
[54]. 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.
[55]. Bosco C, Bosnyak Z, Malmberg A, et al. Quantifying observational evidence for risk of fatal and nonfatal cardiovascular disease following androgen deprivation therapy for prostate cancer: a meta-analysis. Eur Urol 2015;68:386–96.
[56]. Brungs D, Chen J, Masson P, et al. Intermittent androgen deprivation is a rational standard-of-care treatment for all stages of progressive prostate cancer: results from a systematic review and meta-analysis. Prostate Cancer Prostatic Dis 2014;17:105–11.
[57]. Smith MR, Lee H, Nathan DM. Insulin sensitivity during combined androgen blockade for prostate cancer. J Clin Endocrinol Metab 2006;91:1305–8.
[58]. Smith MR, O’Malley AJ, Keating NL. Gonadotrophin-releasing hormone agonists, diabetes and cardiovascular disease in men with prostate cancer: which metabolic syndrome? BJU Int 2008;101:1335–6.
[59]. Nguyen PL, Alibhai SMH, Basaria S, et al. Adverse effects of androgen deprivation therapy and strategies to mitigate them. Eur Urol 2015;67:825–36.
[60]. Khan MAH, Ferro VA, Stimson WH. Use of a highly specific monoclonal antibody against the central variable amino acid sequence of mammalian gonadotropin releasing hormone to evaluate GnRH-I tissue distribution compared with GnRH-I binding sites in adult male rats. Am J Reprod Immunol 2003;49:239–48.
[61]. Reichler IM, Jöchle W, Piché CA, et al. Effect of a long acting GnRH analogue or placebo on plasma LH/FSH, urethral pressure profiles and clinical signs of urinary incontinence due to Sphincter mechanism incompetence in bitches. Theriogenology 2006;66:1227–36.
[62]. Kakar SS, Jennes L. Expression of gonadotropin-releasing hormone and gonadotropin-releasing hormone receptor mRNAs in various non-reproductive human tissues. Cancer Lett 1995;98:57–62.
[63]. Dong F, Skinner DC, Wu TJ, et al. The heart: a novel gonadotrophin-releasing hormone target. J Neuroendocrinol 2011;23:456–63.
[64]. Chen HF, Jeung EB, Stephenson M, et al. Human peripheral blood mononuclear cells express gonadotropin-releasing hormone (GnRH), GnRH receptor, and interleukin-2 receptor gamma-chain messenger ribonucleic acids that are regulated by GnRH in vitro. J Clin Endocrinol Metab 1999;84:743–50.
[65]. Tanriverdi F, Gonzalez-Martinez D, Hu Y, et al. GnRH-I and GnRH-II have differential modulatory effects on human peripheral blood mononuclear cell proliferation and interleukin-2 receptor γ-chain mRNA expression in healthy males. Clin Exp Immunol 2005;142:103–10.

androgen deprivation therapy; cardiovascular morbidity; cardiovascular mortality; network meta-analysis; prostate cancer

Supplemental Digital Content

Back to Top | Article Outline
Copyright © 2016 The Authors. Published by Wolters Kluwer Health, Inc. All rights reserved.