Conclusions. Keywords

Transcription

1 The impact of androgen-deprivation therapy (ADT) on the risk of cardiovascular (CV) events in patients with non-metastatic prostate cancer: a population-based study Giorgio Gandaglia*, Maxine Sun*, Ioana Popa*, Jonas Schiffmann*, Firas Abdollah, Quoc-Dien Trinh, Fred Saad, Markus Graefen, Alberto Briganti, Francesco Montorsi and Pierre I. Karakiewicz* *Cancer Prognostics and Health Outcomes Unit, University of Montreal Health Center, Montreal, Canada; Urological Research Institute, Vita-Salute San Raffaele University, San Raffaele Hospital, San Raffaele Scientific Institute, Milan, Italy; Department of Urology, University of Montreal Health Center, Montreal, Canada; Martini-clinic, Prostate Cancer Center Hamburg-Eppendorf, Hamburg, Germany; and Division of Urologic Surgery and Center for Surgery and Public Health, Brigham and Women's Hospital/Dana-Farber Cancer Institute, Harvard Medical School, Boston, MA, USA G.G. and MS contributed equally. Objective To examine and quantify the contemporary association between androgen-deprivation therapy (ADT) and three separate endpoints: coronary artery disease (CAD), acute myocardial infarction (AMI), and sudden cardiac death (SCD), in a large USA contemporary cohort of patients with prostate cancer. Patients and Methods In all, patients diagnosed with non-metastatic prostate cancer between 1995 and 2009 within the Surveillance, Epidemiology, and End Results (SEER)-Medicare linked database were abstracted. Patients treated with ADT and those not receiving ADT were matched using propensity score methodology. The 10-year CAD, AMI, and SCD rates were estimated. Competing-risks regression analyses tested the association between the type of ADT (GnRH agonists vs bilateral orchidectomy) and CAD, AMI, and SCD, after adjusting for the risk of dying during follow-up. Results Overall, the 10-year rates of CAD, AMI, and SCD were 25.9%, 15.6%, and 15.8%, respectively. After stratification according to ADT status (ADT-naïve vs GnRH agonists vs bilateral orchidectomy), the CAD rates were 25.1% vs 26.9% vs 23.2%, the AMI rates were 14.8% vs 16.6% vs 14.8%, and the SCD rates were 14.2% vs 17.7% vs 16.4%, respectively. In competing-risks multivariable regression analyses, the administration of GnRH agonists (all P < 0.001), but not bilateral orchidectomy (all P 0.7), was associated with higher risk of CAD, AMI, and SCD. Conclusions The administration of GnRH agonists, but not orchidectomy, is still associated with a significantly increased risk of CAD, AMI, and, especially, SCD in patients with non-metastatic prostate cancer. Alternative forms of ADT should be considered in patients at higher risk of CV events. Keywords prostate cancer, androgen-deprivation therapy, cardiovascular events, side-effects Introduction Prostate cancer represents the most common non-cutaneous malignancy for men, where an estimated new cases will be diagnosed in the year 2013 in the USA alone [1]. Androgen-deprivation therapy (ADT) represents a commonly used treatment option for patients with locally advanced or metastatic prostate cancer [2]. Additionally, an increasingly elevated proportion of men with localised disease receive this therapy, and as many as 45% of such patients receive some form of ADT in the course of their disease [3]. ADT improves survival in specific settings, e.g. combined with radiotherapy (RT) or after surgery in node-positive patients BJU Int 2014; 114: E82 E89 wileyonlinelibrary.com doi: /bju Published by John Wiley & Sons Ltd.

2 CV events after ADT for prostate cancer [4 6]. Despite this benefit, ADT is also characterised by a non-negligible risk of clinically relevant side-effects. These include decreased libido, erectile dysfunction, gynaecomastia, anaemia, hot flashes, osteopaenia and fractures, changes in cognition and mood, kidney failure, and metabolic alterations [2,7 9]. Interestingly, previous observational studies relying on historical populations hypothesised and proved that ADT, and in particular GnRH agonists are associated with an increased risk of cardiovascular (CV) events [8,10 12]. Consequently, the USA Food and Drug Administration (FDA) issued a safety warning requiring labelling on GnRH agonists about an increased risk of CV diseases. This measure may have resulted in more select use of ADT or in better preventive and follow-up measures aimed at treated patients. Additionally, other factors such as better nutrition, exercise, and overall more health conscious lifestyle of contemporary American men may also have resulted in improved overall health outcomes. This may, in turn, result in lower rates of CV events, such as coronary artery disease (CAD), acute myocardial infarction (AMI), and sudden cardiac death (SCD). Based on the hypothesis that the link between ADT and CV outcomes may have changed over time, we re-examined the effect of ADT on the rates of experiencing one of the three CV outcomes of interest, namely CAD, AMI, and SCD, in a large contemporary cohort of patients with prostate cancer representing the USA population. Additionally, as previous evidence suggests that surgical castration might be a safer alternative to GnRH agonists [10,13], we also assessed its effect on the rates of the same three endpoints. Patients and Methods The present study relied on the most recent version of the Surveillance, Epidemiology, and End Results (SEER)-Medicare insurance program-linked database. This database is 98% complete for case ascertainment. The SEER registries cover 28% of the USA population with Medicare administrative data. Medicare insurance includes 97% of Americans aged 65 years. In all, patients with histologically confirmed non-metastatic prostate cancer (International Classification of Disease for Oncology [ICD-O] site code 61.9, histological code 8140) aged 66 years diagnosed between January 1995 and December 2009 were identified. Exclusion criteria consisted of men with pre-existing CV diseases (ICD 9th edition [ICD-9], Current Procedural Terminology [CPT], and Healthcare Common Procedure Coding System [HCPCS] codes available upon request; n = ), unknown socioeconomic status (n = 6312), unknown stage (n = 6578), and unknown Gleason score (n = 6590). This resulted in a final population of assessable patients. For each patient, age, year of diagnosis, race, population density, marital status, Gleason score, clinical stage, primary treatment (radical prostatectomy [RP] vs RT vs brachytherapy) were assigned [14]. The Charlson comorbidity index (CCI) was derived from the Medicare claims 1 year before prostate cancer diagnosis using a previously validated algorithm [15]. The administration of GnRH agonists during follow-up was assessed, as previously reported [16]. The outcomes of interest were represented by the occurrence of CAD, AMI, and SCD, which were identified using the ICD-9, CPT, and HCPCS codes, as previously reported [10]. In patients receiving ADT, time to event was calculated from the date of ADT initiation to the first occurrence of the event. In ADT-naïve patients, time to event was calculated from the date of prostate cancer diagnosis to the first occurrence of the event. Means, medians, and interquartile ranges were reported for continuous variables. Frequencies and proportions were reported for categorical variables. The independent t-test and chi-square test were used to compare means and proportions, respectively. Our statistical analyses consisted of different steps. First, due to inherent differences between patients undergoing ADT and not receiving ADT, adjustment was performed using 0.5 to 0.5 propensity score matching. This methodology is commonly used in observational studies to select control subjects who are matched with treated subjects on the controlled background covariates, which, if uncontrolled for, might lead to biased estimates of treatment effects. When matching is performed, the covariates of the control and treatment groups are balanced in order to reduce possible biases to a minimum [17]. Propensity scores were computed by modelling a logistic regression with the dependent variable as the odds of receiving ADT, and the independent variable as age, race, marital status, CCI, population density, year of diagnosis, clinical stage, tumour grade, and the receipt of primary treatment (RP, RT, or brachytherapy). Subsequently, covariate balance between the matched groups were examined [18]. In the second part of the analyses, we examined crude rates of CAD, AMI, and SCD according to treatment type, namely men receiving GnRH agonists vs those treated with bilateral orchidectomy vs ADT-naïve patients. Subsequently, we relied on competing-risks regression analyses. Three separate endpoints were addressed in different competing-risks models, namely CAD, AMI, and SCD. In models focusing on CAD and AMI rates, we adjusted for all-cause mortality. In the model focusing on SCD, adjustment for other cause mortality was performed. All statistical tests were performed using the R statistical package (v2.15.2). All tests were two-sided with a significance level set at P < E83

3 Gandaglia et al. Results Baseline Characteristics In all, patients with non-metastatic prostate cancer were included in the study (Table 1). The mean (median) age at diagnosis was 73.6 (73) years. Overall, (44.7%) patients were treated with ADT over the study period. Among patients receiving ADT, (96.6%) and 2055 (3.4%) were treated with GnRH agonists and bilateral orchidectomy, respectively. During a mean (median) follow-up of 75.3 (74) months, (16.8%), (9.4%), and (8.9%) patients experienced CAD, AMI, and SCD, respectively. Additionally, 6229 (4.4%) and (20.5%) succumbed to prostate cancer and other cause mortality, respectively. Table 1 shows the characteristics of patients included in the study. When patients were stratified according to the administration of ADT, statistically significant differences were recorded for age at diagnosis, year of diagnosis, race, population density, marital status, Gleason score, clinical stage, and primary treatment (all P < 0.001). After propensity score matching, (50%) patients who received ADT and (50%) ADT-naïve individuals remained. The mean standardised differences of the matched covariates between the two groups were <10%, indicating a high degree of similarity in the distribution of both populations. All subsequent analyses were based on the post-propensity score-matched cohort. Cumulative Incidence of CAD Overall, the 10-year rates of CAD, AMI, and SCD were 25.9%, 15.6%, and 15.8%, respectively. After stratification according to ADT status (ADT-naïve vs GnRH agonists vs bilateral orchidectomy), the CAD rates were 25.1% vs 26.9% vs 23.2%, the AMI rates were 14.8% vs 16.6% vs 14.8%, and the SCD rates were 14.2% vs 17.7% vs 16.4%, respectively (Table 2). Competing-risks Regression Models In univariable competing-risks regression models, GnRH agonist exposure resulted in a CAD hazard ratio (HR) of 1.10 relative to ADT-naïve patients (P < 0.001; Table 3). Conversely, bilateral orchidectomy was not associated with increased risk of CAD (HR 1.00; P = 0.9). Additionally, GnRH agonists, but not bilateral orchidectomy, were significantly associated with the risk of AMI (HR 1.10; P < 0.001) and of SCD (HR 1.18; P < 0.001). For both endpoints (AMI and SCD), orchidectomy failed to show a statistically significant association. In multivariable competing-risks regression analyses, the exposure to GnRH agonists resulted in a 1.11-fold higher risk of CAD, after accounting for confounders (P < 0.001; Table 3). Similarly, GnRH agonists exposure was associated with increased risk of AMI (HR 1.09; P < 0.001) and SCD (HR 1.18; P < 0.001), after accounting for confounders. Conversely, bilateral orchidectomy was not significantly associated with any of the three examined outcomes (all P 0.1). Discussion The administration of ADT may detrimentally affect several health-related outcomes in patients with prostate cancer [7]. Previous retrospective studies reported an association between ADT, and in particular GnRH agonists, and the risk of CV morbidity and mortality [7,8,10,11,13,19,20]. Although these observations lead to a FDA warning about the safety of GnRH agonists, the relationship between ADT and CV events is still highly controversial. For example, a recent meta-analysis of randomised trials failed to show an association between ADT and the risk of CV mortality [21]. On the other hand, clinical trials usually enrol younger and healthier patients, who might be considered at reduced risk of CV morbidity and mortality. Consequently, results obtained in this setting might not necessarily be applicable to the general population. Additionally, previous large retrospective investigations evaluating the USA population relied on historical cohorts [8,10,11]. Since then, the understanding and familiarity with ADT detriments has improved. Similarly, many individuals, including patients with prostate cancer who are ADT-candidates, have adopted more health-focused lifestyles. Consequently, the contemporary effects of ADT on CV endpoints may differ from those seen in previous investigations. Under this light, the present study was aimed at examining the association between the type of ADT (exposure to GnRH agonists or bilateral orchidectomy) and CAD, AMI, and SCD in a large contemporary cohort of patients with non-metastatic prostate cancer, representative of the USA population. Several results of the present study are noteworthy. First, relative to ADT-naïve patients, the rates of CAD, AMI, and SCD were increased in patients exposed to GnRH agonists, but not in those receiving bilateral orchidectomy. Particularly, at 10-year follow-up the proportion of patients experiencing CAD, AMI, and SCD was 27%, 16%, and 17% vs 25%, 15%, and 14% in patients exposed to GnRH agonists vs ADT-naïve individuals. This resulted in a corresponding 10-year number needed to harm (NNH) of 55, 56, and 28, respectively. On the other hand, surgical castration was unrelated to an increased risk of these CV endpoints. These observations were confirmed in competing-risks multivariable regression models, where the exposure to GnRH agonists, but not bilateral orchidectomy, was associated with an increased risk of CAD, AMI, and SCD, after accounting for the risk of dying from other causes. Of note, the magnitude in the increased risk of CAD, AMI, and SCD was similar to that previously reported by Keating et al. [10], and ranged between 1.11 and Additionally, the present analyses also corroborate the lack of statistically significant adverse effect of orchidectomy on CV endpoints. E84

4 CV events after ADT for prostate cancer Table 1 Descriptive statistics of patients diagnosed with non-metastatic prostate cancer between 1995 and 2009 within the SEER database stratified according to the type of treatment (observation vs ADT). Variable Before propensity score matching After propensity score matching Total ADT-naïve ADT Standardised difference Total No ADT ADT Standardised difference N (%) (100) (57.3) (42.7) (100) (50) (50) Age at diagnosis, years Mean (median) 73.6 (73) 72.6 (72) 75.1 (74) (74) 75.0 (74) 75.1 (74) 2.0 IQR Race, n (%) White (80.4) (80.9) (79.7) (80.1) (80.3) (79.8) 1.3 African-American (12.1) (12.0) (12.2) (12.1) (12.1) (12.2) Other (8.1) (7.1) (8.1) (7.8) (7.6) (8.0) Year of diagnosis Mean (median) (2003) (2004) (2003) (2003) (2003) (2003) 4.8 IQR Marital status, n (%) Married (69.0) (70.7) (66.6) (66.8) (67.0) (66.6) 0.9 Unmarried (31.0) (29.3) (33.4) (33.2) (33.0) (33.4) Population density (%) Metropolitan (82.9) (84.2) (81.3) (81.4) (81.6) (81.2) 1.3 Non-metropolitan (17.1) (15.8) (18.7) (18.6) (18.4) (18.8) CCI, n (%) (64.5) (68.0) (59.7) (61.4) (62.7) (60.0) (11.7) (10.7) (13.2) (11.9) (10.7) (13.0) (12.4) (11.2) (14.1) (13.5) (13.2) (13.9) (11.4) (10.2) (13.0) (13.2) (13.4) (13.1) Gleason score, n (%) (79.4) (88.0) (68.0) (71.4) (75.3) (67.5) (20.6) (12.0) (32.0) (28.6) (24.7) (32.5) Clinical stage, n (%) T (48.2) (53.1) (41.7) (42.4) (43.1) (41.7) 7.6 T (48.6) (45.5) (52.9) (53.3) (53.9) (52.7) T (2.5) (1.3) (4.2) (3.3) 673 (2.2) (4.3) T4 975 (0.7) 240 (0.3) 735 (1.2) 592 (1.0) 205 (0.7) 387 (1.3) RP, n (%) (18.9) (26.7) (8.4) (8.2) (8.0) (8.4) 1.4 RT, n (%) (40.5) (33.5) (49.9) (49.8) (49.8) (49.7) 0.3 Brachytherapy, n (%) (21.9) (20.1) (24.2) (24.1) (24.3) (23.9) 1.0 Type of ADT, n (%) GnRH agonists (41.2) (96.6) (48.3) (96.7) Bilateral orchidectomy (1.5) (3.4) 995 (1.7) 995 (3.3) IQR, interquartile range. E85

5 Gandaglia et al. Table 2 The 10-year CAD, AMI, and SCD rates in the overall population, and after stratifying patients according to the type of ADT (no ADT vs GnRH agonists vs bilateral orchidectomy). Type of ADT CAD AMI SCD 10-year rates (95% CI), % P 10-year rates (95% CI), % P 10-year rates (95% CI), % P No ADT 25.1 ( ) Ref ( ) Ref ( ) Ref. GnRH agonists 26.9 ( ) < ( ) < ( ) <0.001 Bilateral orchidectomy 23.2 ( ) ( ) ( ) 0.4 Taken together, these observations suggest that, when ADT is recommended, orchidectomy might represent a safer option, especially if CAD avoidance is paramount. Conversely, one patient would experience CAD, AMI, and SCD at 10-year follow-up every 55, 56, and 28 men exposed to GnRH agonists. These considerations should be included at treatment decision-making process and at informed consent. This is particularly important in the context of previous studies that failed to show a survival benefit associated with the exposure to ADT in patients with clinically localised prostate cancer [22]. Consequently, hormonal manipulation should be ideally performed exclusively in selected patients with non-metastatic disease [4 6,23,24]. Finally, novel findings pertaining to GnRH blockers relating to reduction of life-threatening CV events also warrant consideration [12]. Such agents may provide the desired testosterone suppression at a faster and more complete rate [25], with fewer CV detriments compared with GnRH agonists [12], and without the stigma of surgical castration. It is also noteworthy that the availability of data about a possible association between ADT and CV morbidity and mortality from the landmark study by Keating et al. [10] and other retrospective investigations [7,8,10,11,13,20] did not result in a decline in the rates of CAD, AMI, and SCD over the last decade in patients with prostate cancer exposed to medical castration. One might speculate that healthier lifestyles might have resulted in a reduction in the rates of life-threatening complications related to GnRH agonists exposure. Similarly, it might have been postulated that better clinical follow-up and/or better prevention might have been implemented. This does not appear to be the case. Consequently, the present results suggest that sustained efforts aimed at risk modification, including smoking cessation, adoption of healthier lifestyles, and better management of existing comorbidities, are still very much needed to reduce the risk of CV morbidities and mortality in patients with non-metastatic prostate cancer exposed to ADT [8,26]. Additionally, accurate patient selection is necessary to spare the substantial risk of CV-related side-effects in men at higher risk of CAD, AMI, and SCD. Last but not least, stringent selection criteria for use of ADT in non-metastatic setting should also be encouraged to limit or shorten the exposure. From a pathophysiological perspective, the present observations suggest potential explanations about the mechanisms underlying the detrimental effects of GnRH agonists on CV outcomes. Historically, the association between ADT and CV events has been attributed to the low testosterone levels induced by medical or surgical castration, which result in metabolic changes, e.g. increase in body weight, insulin resistance, and dyslipidaemia [7,27]. Additionally, testosterone was postulated to directly affect cardiac contractility and to lead to systemic vasodilatation [28]. Consequently, ADT might result in increased arterial wall thickness and/or endothelial dysfunction. This in turn might promote formation of atherosclerotic plaques [29,30]. The non-specific association of ADT in either surgical or medical forms is questioned by orchidectomy-specific data. Specifically, the lack of association between surgical castration and the endpoints of CAD, AMI, and SCD questions the validity of the non-specific pathophysiological mechanisms of testosterone suppression. Indeed, our observations suggest that CV morbidities might be a drug class effect of GnRH agonists, but not a detriment of castration and hypogonadism itself. In this context, it has been proposed that GnRH agonists might directly regulate cardiac contractility via GnRH receptors expressed by cardiomyocytes. Additionally, T cells express GnRH receptors, and their activation by GnRH agonists might represent another trigger for atherosclerotic plaque destabilisation and rupture, which may result in CV events [31,32]. These GnRH agonists-specific pathways explaining the link with CV morbidities await further validation. Despite several strengths, the present study is not devoid of limitations. First, it is in part limited by its retrospective design. Particularly, we cannot exclude that a treatment selection bias might partially affect the validity of our results. To minimise this potential bias, we performed propensity score matching. Second, the lack of data on certain risk factors, e.g. smoking status, body mass index, and family history of CAD, precluded us from adjusting our analyses for these important confounders. However, our model accounted for CCI, which might be considered a reliable proxy of the general patient health status. Third, the SEER-Medicare does not provide data on oral medications. However, treatment E86

6 CV events after ADT for prostate cancer Table 3 Multivariable competing-risks regression analysis predicting CAD, AMI, and SCD in patients with non-metastatic prostate cancer included in the analyses after propensity score matching. Variable CAD AMI SCD Univariable analyses Multivariable analyses Univariable analyses Multivariable analyses Univariable analyses Multivariable analyses HR (95% CI) P HR (95% CI) P HR (95% CI) P HR (95% CI) P HR (95% CI) P HR (95% CI) P Age at diagnosis 1.00 ( ) ( ) ( ) < ( ) < ( ) < ( ) <0.001 Year of diagnosis 0.98 ( ) < ( ) < ( ) ( ) ( ) < ( ) 0.01 Race: White 1 (ref.) 1 (ref.) 1 (ref.) 1 (ref.) 1 (ref.) 1 (ref.) African-American 0.99 ( ) ( ) ( ) ( ) ( ) < ( ) <0.001 Other 0.78 ( ) < ( ) < ( ) ( ) ( ) < ( ) <0.001 Marital status: Married 1 (ref.) (ref.) (ref.) (ref.) (ref.) < (ref.) 0.8 Unmarried 0.97 ( ) 0.96 ( ) 1.08 ( ) 1.04 ( ) 1.09 ( ) 1.00 ( ) Population density: Metropolitan 1 (ref.) (ref.) (ref.) (ref.) (ref.) < (ref.) <0.001 Non-metropolitan 0.93 ( ) 0.92 ( ) 1.03 ( ) 1.02 ( ) 0.88 ( ) 0.87 ( ) CCI: 0 1 (ref.) 1 (ref.) 1 (ref.) 1 (ref.) 1 (ref.) 1 (ref.) ( ) < ( ) < ( ) < ( ) < ( ) < ( ) < ( ) < ( ) < ( ) < ( ) < ( ) < ( ) < ( ) < ( ) < ( ) < ( ) < ( ) < ( ) <0.001 Gleason score: 7 1 (ref.) < (ref.) < (ref.) < (ref.) < (ref.) (ref.) ( ) 1.11 ( ) 0.91 ( ) 0.90 ( ) 0.97 ( ) 0.96 ( ) Clinical stage: T1 1 (ref.) 1 (ref.) 1 (ref.) 1 (ref.) 1 (ref.) 1 (ref.) T ( ) ( ) ( ) ( ) ( ) ( ) 0.4 T ( ) ( ) ( ) ( ) ( ) ( ) 0.2 T ( ) ( ) ( ) ( ) ( ) ( ) 0.2 RP 0.92 ( ) ( ) ( ) < ( ) ( ) < ( ) 0.3 RT 1.09 ( ) < ( ) ( ) ( ) ( ) ( ) 0.05 Brachytherapy 1.07 ( ) ( ) ( ) ( ) ( ) ( ) 0.01 ADT: No 1 (ref.) 1 (ref.) 1 (ref.) 1 (ref.) 1 (ref.) 1 (ref.) GnRH agonist 1.10 ( ) < ( ) < ( ) < ( ) < ( ) < ( ) <0.001 Orchidectomy 1.00 ( ) ( ) ( ) ( ) ( ) ( ) 0.1 E87

7 Gandaglia et al. with anti-androgen monotherapy is not approved in the USA. Finally, as the present findings were obtained in a cohort of patients aged 66 years representative of the North American population, external validation is necessary to confirm their validity in younger patients treated in other settings. Ideally, this should be performed in the context of prospective multi-institutional studies. In conclusion, the exposure to GnRH agonists, but not bilateral orchidectomy, increases the risk of CAD, AMI, and SCD. Therefore, when ADT is contemplated, the detriments of GnRH agonists relative to orchidectomy warrant strong consideration. Substantial efforts aimed at improving patient selection, lifestyles modifications, as well as more stringent patient selection should be emphasised in patients with non-metastatic prostate cancer receiving ADT. Conflict of Interest None declared. References 1 SiegelR,NaishadhamD,JemalA.Cancer statistics, CA Cancer J Clin 2013; 63: Sharifi N, Gulley JL, Dahut WL. Androgen deprivation therapy for prostate cancer. JAMA 2005; 294: 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: 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: 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: Messing EM, Manola J, Yao J et al. Immediate versus deferred androgen deprivation treatment in patients with node-positive prostate cancer after radical prostatectomy and pelvic lymphadenectomy. Lancet Oncol 2006; 7: Conteduca V, Di Lorenzo G, Tartarone A, Aieta M. The cardiovascular risk of gonadotropin releasing hormone agonists in men with prostate cancer: an unresolved controversy. Crit Rev Oncol Hematol 2013; 86: Keating NL, O Malley AJ, Freedland SJ, Smith MR. Does comorbidity influence the risk of myocardial infarction or diabetes during androgen-deprivation therapy for prostate cancer? Eur Urol 2013; 64: GandagliaG,SunM,HuJCetal.Gonadotropin-releasing hormone agonists and acute kidney injury in patients with prostate cancer. Eur Urol 2014 [Epub ahead of print]. doi: /j.eururo Keating NL, O Malley AJ, Smith MR. Diabetes and cardiovascular disease during androgen deprivation therapy for prostate cancer. J Clin Oncol 2006; 24: Keating NL, O Malley AJ, Freedland SJ, Smith MR. Diabetes and cardiovascular disease during androgen deprivation therapy: observational study of veterans with prostate cancer. JNatlCancerInst2010; 102: Albertsen PC, Klotz L, Tombal B, Grady J, Olesen TK, Nilsson J. Cardiovascular morbidity associated with gonadotropin releasing hormone agonists and an antagonist. Eur Urol 2014; 65: Jespersen CG, Norgaard 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: Wong YN, Mitra N, Hudes G et al. Survival associated with treatment vs observation of localized prostate cancer in elderly men. JAMA 2006; 296: Klabunde CN, Potosky AL, Legler JM, Warren JL. Development of a comorbidity index using physician claims data. J Clin Epidemiol 2000; 53: Hu JC, Gu X, Lipsitz SR et al. Comparative effectiveness of minimally invasive vs open radical prostatectomy. JAMA 2009; 302: D Agostino RB Jr. Propensity score methods for bias reduction in the comparison of a treatment to a non-randomized control group. Stat Med 1998; 17: Ho DE, Imai K, King G, Stuart E. MatchIt: nonparametric preprocessing for parametric causal inference. J Stat Softw 2011; Rosario DJ, Bourke L, Keating NL. Androgen deprivation therapy and cardiovascular harm: are all men created equal? Eur Urol 2014; 65: 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: Nguyen PL, Je Y, Schutz FA et al. Association of androgen deprivation therapy with cardiovascular death in patients with prostate cancer: a meta-analysis of randomized trials. JAMA 2011; 306: Lu-Yao GL, Albertsen PC, Moore DF et al. Survival following primary androgen deprivation therapy among men with localized prostate cancer. JAMA 2008; 300: Bolla M, Gonzalez D, Warde P et al. Improved survival in patients with locally advanced prostate cancer treated with radiotherapy and goserelin. NEnglJMed1997; 337: 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 randomised trial. Lancet Oncol 2011; 12: 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: Saylor PJ, Keating NL, Smith MR. Prostate cancer survivorship: prevention and treatment of the adverse effects of androgen deprivation therapy. J Gen Intern Med 2009; 24 (Suppl. 2): S Smith MR, Finkelstein JS, McGovern FJ et al. Changesinbody composition during androgen deprivation therapy for prostate cancer. J Clin Endocrinol Metab 2002; 87: Dockery F, Bulpitt CJ, Agarwal S, Donaldson M, Rajkumar C. Testosterone suppression in men with prostate cancer leads to an increase in arterial stiffness and hyperinsulinaemia. Clin Sci (Lond) 2003; 104: Fahed AC, Gholmieh JM, Azar ST. Connecting the lines between hypogonadism and atherosclerosis. Int J Endocrinol 2012; Liu PY, Death AK, Handelsman DJ. Androgens and cardiovascular disease. Endocr Rev 2003; 24: Chen HF, Jeung EB, Stephenson M, Leung PC. 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: E88

8 CV events after ADT for prostate cancer 32 Tanriverdi F, Gonzalez-Martinez D, Hu Y, Kelestimur F, Bouloux PM. GnRH-I and GnRH-II have differential modulatory effects on human peripheral blood mononuclear cell proliferation and interleukin-2 receptor gamma-chain mrna expression in healthy males. Clin Exp Immunol 2005; 142: Correspondence: Giorgio Gandaglia, University of Montreal Health Center, 1058, rue St-Denis, Montreal, QC, H2X 3J4, Canada. Abbreviations: ADT, androgen-deprivation therapy; AMI, acute myocardial infarction; CAD, coronary artery disease; CCI, Charlson comorbidity index; CPT, Current Procedural Terminology; CV, cardiovascular; FDA, USA Food and Drug Administration; HCPCS, Healthcare Common Procedure Coding System; HR, hazard ratio; ICD(-O), International Classification of Diseases (for Oncology); RP, radical prostatectomy; RT, radiotherapy; SCD, sudden cardiac death; SEER, Surveillance, Epidemiology, and End Results. E89