Cardiovascular Events in Men with Prostate Cancer Receiving Hormone Therapy: An Analysis of the FDA Adverse Event Reporting System (FAERS)

Abstract

Purpose:

The comparative cardiovascular risk profiles of available hormone therapies for the treatment of prostate cancer is not known.

Materials and Methods:

We queried the U.S. Food and Drug Administration Adverse Event Reporting System, a retrospective, pharmacovigilance database, for cardiovascular adverse event reports in men with prostate cancer receiving gonadotropin releasing hormone (GnRH) agonists, GnRH antagonists, androgen receptor antagonists, and/or androgen synthesis inhibitors from January 2000 to April 2020.

Results:

Cardiovascular adverse events accounted for 6,231 reports (12.6%) on hormone monotherapy and 1,793 reports (26.1%) on combination therapy. Arterial vascular events were reported most commonly, followed by arrhythmias, heart failure, and venous thromboembolism. Compared to GnRH agonists, GnRH antagonists were associated with fewer cardiovascular adverse event reports as monotherapy (adjusted reporting odds ratio [ROR]=0.70 [95% CI 0.59–0.84], p <0.001) and as combination therapy (ROR=0.47 [0.34–0.67], p <0.0001), driven by reductions in arterial vascular events. Second generation androgen receptor antagonists and abiraterone were associated with more reports of hypertension requiring hospitalization (ROR=1.21 [1.03–1.41], p=0.02 and ROR=1.19 [1.01–1.40], p=0.03, respectively), and more heart failure events when used in combination with GnRH antagonists (ROR=2.79 [1.30–6.01], p=0.009 and ROR=2.57 [1.12–5.86], p=0.03).

Conclusions:

In this retrospective analysis of a pharmacovigilance database, arterial vascular events were the most commonly reported cardiovascular adverse events in men on hormone therapy for prostate cancer. GnRH antagonists were associated with fewer reports of overall cardiovascular events and arterial vascular events than GnRH agonists. Additional study is needed to identify optimal strategies to reduce cardiovascular morbidity among men with prostate cancer receiving hormone therapy.

Study Need and Importance:

Androgen deprivation therapy (ADT) is a mainstay of treatment for localized and metastatic prostate cancer. Observational studies have suggested an association of ADT with increased cardiovascular risk, though analyses of clinical trials have yielded conflicting results. In the recently published HERO trial, there was a 54% lower risk of adverse cardiovascular events on relugolix as compared to leuprolide. The comparative cardiovascular risk profile of currently available hormone therapies for prostate cancer is not known.

What We Found:

We queried the U.S. Food and Drug Administration Adverse Event Reporting System (FAERS), a pharmacovigilance database, for cardiovascular adverse event reports in men with prostate cancer receiving ADT from 2000 to 2020. Arterial vascular events were reported most commonly, followed by arrhythmias, heart failure, and venous thromboembolism. Compared to gonadotropin releasing hormone (GnRH) agonists, GnRH antagonists were associated with fewer cardiovascular adverse event reports as monotherapy and combination therapy, driven by reductions in arterial vascular events. Second generation androgen receptor antagonists and abiraterone were associated with more reports of hypertension requiring hospitalization and heart failure when used in combination with GnRH antagonists.

Limitations:

This study assessed adverse cardiovascular event reporting available in the FAERS database, which is subject to reporting bias due to the voluntary nature of reporting. Importantly, key clinical information such as prostate cancer characteristics, duration of anti-androgen therapy, cardiovascular comorbidities, and time to cardiovascular event were not available in the database. Additional study in data sets containing more granular clinical information is needed to extend our findings.

Interpretation For Patient Care:

Arterial vascular events are important cardiovascular adverse events on hormone therapy for men with prostate cancer. Additional study is needed to identify optimal strategies to reduce cardiovascular morbidity among men with prostate cancer receiving hormone therapy.

Graphical Abstract

Prostate cancer is the most common noncutaneous cancer in men, with more than 190,000 new cases diagnosed annually and 3.3 million survivors in the United States as of 2017.¹ Due to early detection and effective therapies, the 5-year prostate cancer survival rate approaches 100%, and the 10-year cancer-specific mortality rate is only 7.1% for localized prostate cancer.² However, men with prostate cancer are at increased risk of cardiovascular mortality compared to patients with other malignancies, with cardiovascular mortality risk that increases with age and time from cancer diagnosis.³ Effective strategies are needed to mitigate cardiovascular risk in men with prostate cancer.

Androgen deprivation therapy (ADT), a mainstay of treatment for localized and metastatic prostate cancer, has been associated with increased cardiovascular risk particularly among men with more medical comorbidities.^4–8 The observed association of ADT with increased cardiovascular risk led to the 2010 publication of a joint scientific statement from the American Heart Association and the American Urological Association regarding this finding,⁹ as well as the issuance of cautionary labeling for gonadotropin-releasing hormone (GnRH) agonists by the U.S. Food and Drug Administration (FDA).¹⁰

Recent clinical trial data suggest that cardiovascular risk may vary between different ADT drug classes, with lower risk on GnRH antagonists compared to GnRH agonists.^8,11,12 Potential differences in cardiovascular risk profile between ADT drug types in clinical practice are currently unknown. Furthermore, novel anti-androgen therapies such as second generation androgen receptor antagonists and androgen synthesis inhibitors are now standard of care for metastatic hormone sensitive and metastatic castrate-resistant prostate cancer. The cardiovascular safety of these newer hormonal therapies as compared to traditional ADT is not well understood.

In this study, we aimed to characterize the cardiovascular risk profile of hormone therapies used to treat prostate cancer, including GnRH agonists, GnRH antagonists, androgen receptor antagonists and androgen synthesis inhibitors, using the FDA Adverse Event Reporting System (FAERS), a pharmacovigilance database.

MATERIALS AND METHODS

Study Design and Data Source

This was a retrospective, observational, cross-sectional pharmacovigilance study using adverse drug reactions reported in the FAERS database. FAERS is a publicly available database created in 1968 containing adverse event reports, medication error reports, and product quality complaints resulting in adverse events submitted by health care professionals, consumers and manufacturers worldwide. Clinical reviewers evaluate reports submitted to FAERS in the Center for Drug Evaluation and Research and the Center for Biologics Evaluation and Research to monitor products’ safety after receiving FDA approval.

Data in each event report include an individualized case identification number, event date, the suspected pharmaceutical agent, clinical indication for use, observed adverse reaction(s), reaction severity, reaction outcome(s) (ie hospitalization, death), basic patient demographics (gender, age, weight), source of event report (ie health care professional, consumer, manufacturer), concomitant medications, country where the event occurred and manufacturer control number for the pharmaceutical agent.

Procedures

The FAERS database was queried between January 1, 2000 and August 8, 2020 for cardiovascular adverse events in men aged ≥18 years receiving hormone therapy for prostate cancer. Specific pharmaceutical agents queried included GnRH agonists: leuprolide, buserelin, goserelin, histrelin, triptorelin; GnRH antagonists: degarelix; first generation androgen receptor antagonists (ARAs): flutamide, nilutamide, bicalutamide; second generation ARAs: enzalutamide, apalutamide, darolutamide; and the androgen synthesis inhibitor abiraterone acetate. Monotherapy-related adverse event reports were those associated with only 1 queried therapeutic class. Combination therapy-related adverse event reports were those associated with either a GnRH agonist in combination with an ARA or abiraterone, or a GnRH antagonist in combination with an ARA or abiraterone.

Several categories of cardiovascular adverse events were queried; specific search terms for each category of adverse events are listed in supplementary table 1 (https://www.jurology.com). FAERS adheres to the international safety reporting guidance issued by the International Conference on Harmonization (ICH E2B), where all adverse events are coded using terms in the Medical Dictionary for Regulatory Activities.¹³ These included arterial vascular events (coronary artery disease [CAD], myocardial infarction [MI], ischemic stroke, peripheral vascular disease [PVD], and hypertension requiring hospitalization), heart failure, venous thromboembolism (VTE; deep venous thrombosis and pulmonary embolism), and arrhythmias (atrial fibrillation and QT prolongation).

Date of last FAERS access was August 8, 2020.

Statistical Analysis

Descriptive statistics were computed using the chi-square test to differentiate the demographic and clinical characteristics of those who received combination therapy vs monotherapy. Multivariable polytomous logistic regression modeling (adjusted for age, source of event report, and hospitalization requirement) was performed to estimate the reporting odds ratio for each adverse event by class of anti-androgen therapy as compared to GnRH agonists (monotherapy) or GnRH agonists combined with any other agent (combination therapy).

RESULTS

Total Adverse Event and Cardiovascular Adverse Event Reporting

In total, 20,292,362 adverse events were reported in the complete FAERS database. During the study period (January 1, 2000 to April 30, 2020), there were 18,389,339 total adverse events reported (cardiovascular and noncardiovascular) in FAERS, of which 49,634 were due to the selected therapies used as monotherapy and 6,881 were due to combination therapy. Cardiovascular adverse events accounted for 6,231 reports (12.6% of all reports) on hormone monotherapy and 1,793 reports (26.1% of all reports) on combination therapy. Arterial vascular events were the most common cardiovascular adverse events reported (3,235 on monotherapy, 1,004 on combination therapy), followed by arrhythmias (900 on monotherapy, 301 on combination therapy), heart failure (907 on monotherapy, 271 on combination therapy), and VTE (417 on monotherapy, 159 on combination therapy).

Characteristics of cardiovascular adverse event reports on hormone monotherapy and combination therapy are shown in table 1. Cardiovascular adverse event reporting varied significantly by age (p <0.001), with the vast majority of reports, including men ≥age 60 years. Most events were categorized as “serious” and required hospitalization, with “serious” and hospitalized events, reported more frequently on combination therapy as compared to monotherapy (p <0.001). Mortality outcome occurred more frequently with events reported on monotherapy than combination therapy (p <0.001). Over 70% of cardiovascular adverse events were reported after 2013. There were no significant differences in these characteristics among the 4 cardiovascular event categories (supplementary table 2, https://www.jurology.com).

Table 1.

Characteristics of cardiovascular adverse event reporting on androgen deprivation monotherapy (6,231) and combination therapy (1,793)

	No. Monotherapy (%)	No. Combination Therapy (%)	p Value
Yrs age:			<0.001
18–39	91 (1.5)	47 (2.6)
40–59	208 (3.3)	71 (4.0)
60–79	3,603 (57.8)	1,189 (66.3)
≥80	2,329 (37.4)	486 (27.1)
Year of reporting:			0.09
2000–2013	1,821 (29.2)	487 (27.2)
2014–2020	4,410 (70.8)	1,306 (72.8)
Reporting source:			<0.001
Consumer	7 (0.1)	2 (0.1)
Pharmaceutical company	3,856 (61.9)	1,358 (75.7)
Health care professional	2,167 (34.8)	229 (12.8)
Other	201 (3.2)	204 (11.4)
Reaction type:			<0.001
Serious	5,526 (88.7)	1,736 (96.8)
Nonserious	705 (11.3)	57 (3.2)
Hospitalized	3,336 (53.5)	1,192 (66.5)	<0.001
Died	1,292 (20.7)	221 (12.3)	<0.001

Statistical testing was performed using Wilcoxon test for continuous variables and chi-squared test for categorical variables.

Hormone Monotherapy

Cardiovascular adverse event reporting by hormone monotherapy is shown in table 2. Total and cardiovascular adverse event reporting was an order of magnitude lower for GnRH antagonists and first generation ARAs than other therapeutic categories. Compared to GnRH agonist monotherapy, GnRH antagonists were associated with fewer cardiovascular adverse events reports (adjusted reporting odds ratio [ROR]=0.70 [95% CI 0.59–0.84], p <0.001), second generation ARA (ROR=0.80 [0.74–0.86], p <0.001), and abiraterone (ROR=0.87 [0.80–0.94], p=0.0003). Conversely, more cardiovascular events were reported on first generation ARA monotherapy (ROR=1.23 [1.09–1.38], p=0.0009).

Table 2.

Cardiovascular adverse event reporting by hormone monotherapy

Monotherapy	No. Cardiovascular/Total Event Reports	% Reported Events	Adjusted ROR	95% CI	p Value
GnRH agonist	1,752/11,809	14.8	—	—	—
GnRH antagonist	173/1,250	13.8	0.70	0.59, 0.84	<0.001
1st generation ARA	474/2,389	19.8	1.23	1.09, 1.38	0.0009
2nd generation ARA	2,311/23,181	10.0	0.80	0.74, 0.86	<0.001
Abiraterone	1,521/11,005	13.8	0.87	0.80, 0.94	0.0003

Total and cardiovascular adverse event reporting on hormone monotherapy are shown by therapeutic category. Odds ratio for cardiovascular adverse event reporting (adjusted for age, reporting source, and need for hospitalization) is shown as compared to GnRH agonists for every other therapeutic category.

Adverse event reporting by cardiovascular event type is shown in figure 1, part A. Arterial vascular events accounted for the vast majority of cardiovascular adverse event reports in every therapeutic category (range 49%–61%). GnRH antagonists were associated with fewer reports of arterial vascular events (ROR=0.69 [0.56–0.87], p=0.001), VTE (ROR=0.40 [0.20–0.79], p=0.009), and arrhythmias (ROR=0.45 [0.27–0.77], p=0.004) as compared to GnRH agonists without significant difference in heart failure (supplementary table 3, https://www.jurology.com). Similarly, second generation ARA and abiraterone monotherapy were associated with fewer reports of arterial vascular events (ROR=0.78 [0.71–0.86], p <0.001 and ROR=0.76 [0.68–0.84], p <0.001, respectively) and VTE (ROR=0.53 [0.41–0.69], p <0.001 and ROR=0.66 [0.51–0.85], p=0.002). Conversely, first generation ARAs were associated with more reports of heart failure (ROR=1.38 [1.07–1.78], p=0.02), VTE (ROR=1.40 [1.01–1.95], p=0.04), and arrhythmias (ROR=1.94 [1.52–2.48], p <0.001) as compared to GnRH agonists.

Figure 1.

Specific cardiovascular adverse event reporting by hormone monotherapy. Adverse event reporting by cardiovascular event type (A) and arterial vascular event type (B) for hormone monotherapy. RORs performed for each category of adverse event reporting as percentage of total adverse event reporting (adjusted for age, reporting source, and need for hospitalization) as shown in supplementary table 3 (https://www.jurology.com).

Adverse event reporting by arterial event type is shown in figure 1, part B. PVD represented the minority of reported arterial vascular events (range 2%–5%). While GnRH antagonists were associated with fewer reports of CAD/MI (ROR=0.48 [0.32–0.71], p=0.0002) and hypertension requiring hospitalization (ROR=0.66 [0.46–0.94], p=0.02) as compared to GnRH agonists, there was no difference in reporting of ischemic stroke or PVD (supplementary table 3, https://www.jurology.com). Second generation ARA and abiraterone monotherapy were associated with fewer reports of CAD/MI, ischemic stroke, and PVD but increased reports of hypertension requiring hospitalization as compared to GnRH agonists. There was no significant difference in arterial event type reporting with first generation ARA as compared to GnRH agonists.

Combination Hormone Therapy

Cardiovascular event reporting by combination hormone therapy is shown in table 3. Total and cardiovascular event reporting was higher in combination with GnRH agonists than GnRH antagonists. Compared to any combination therapy with GnRH agonist, any combination therapy with GnRH antagonists was associated with fewer cardiovascular adverse event reports (ROR=0.64 [0.50–0.81], p=0.0003). Similarly, GnRH antagonists in combination with either first generation ARA (ROR=0.45 [0.25–0.79], p=0.006) or second generation ARA (ROR 0.65 [0.46–0.92], p=0.02) were associated with fewer cardiovascular adverse event reports than GnRH agonists. There was no difference in cardiovascular adverse event reporting between GnRH agonists and GnRH antagonists in combination with abiraterone.

Table 3.

Cardiovascular adverse event reporting by combination hormone therapy

Combination Therapy	No. Cardiovascular/Total Event Reports	% Reported Events	Adjusted ROR	95% CI	p Value
GnRH agonist, any combination	1,395/5,241	26.6	—	—	—
GnRH antagonist, any combination	92/450	20.4	0.64	0.50, 0.81	0.0003
GnRH agonist+1st generation ARA	948/3,232	29.3	—	—	—
GnRH agonist+1st generation ARA	15/93	16.1	0.45	0.25, 0.79	0.0055
GnRH agonist+2nd generation ARA	305/1,417	21.5	—	—	—
GnRH antagonist+2nd generation ARA	50/242	20.7	0.65	0.46, 0.92	0.02
GnRH agonist+abiraterone	142/592	24.0	—	—	—
GnRH antagonist+abiraterone	27/115	23.5	0.98	0.61, 1.59	0.94

Total and cardiovascular adverse event reporting on combination hormone therapy are shown. Odds ratio for cardiovascular adverse event reporting (adjusted for age, reporting source, and need for hospitalization) is shown as compared to GnRH agonist combination therapy.

Adverse event reporting by cardiovascular event type is shown in figure 2. Any combination therapy with GnRH antagonists was associated with fewer reports of arterial vascular events (ROR=0.47 [0.34–0.67], p <0.0001) and arrhythmias (ROR=0.55 [0.31–0.97], p=0.04; supplementary table 4, https://www.jurology.com). In combination with GnRH antagonists, first generation ARAs (ROR=0.11 [0.03–0.47], p=0.003) and second generation ARAs (ROR=0.45 [0.28–0.73], p=0.001) were associated with fewer reports of arterial vascular events as compared to combination therapy with GnRH agonists. Similarly, arrhythmias were reported less frequently on second generation ARAs (ROR=0.19 [0.05–0.78], p=0.02) in combination with GnRH antagonists as compared to GnRH agonists. Conversely, there was increased reporting of heart failure events on second generation ARA (ROR=2.79 [1.30–6.01], p=0.009) and abiraterone (ROR=2.57 [1.12–5.86], p=0.03) in combination with GnRH antagonists. There were no significant differences in VTE reporting between GnRH agonist vs GnRH antagonist combination therapy.

Figure 2.

Specific cardiovascular adverse event reporting by combination hormone therapy. Adverse event reporting by cardiovascular event type (A) and arterial vascular event type (B) for GnRH agonist and GnRH antagonist combination therapy is shown. RORs performed for each category of adverse event reporting as a percentage of total adverse event reporting (adjusted for age, reporting source, and need for hospitalization) as shown in supplementary table 4 (https://www.jurology.com).

Adverse event reporting by arterial event type is shown in supplementary table 4 (https://www.jurology.com). Compared to any combination therapy with GnRH agonist, any combination therapy with GnRH antagonist was associated with fewer reports of CAD/MI (ROR=0.19 [0.08–0.43], p <0.0001) and hypertension requiring hospitalization (ROR=0.43 [0.25–0.75], p=0.003). This was driven by lower reporting of CAD/MI, and hypertension requiring hospitalization in patients receiving second generation ARAs combined with GnRH antagonists compared to GnRH agonists. There were no differences in reporting of ischemic stroke or peripheral vascular disease between GnRH agonist and GnRH antagonist combination therapy.

To evaluate for potential reporting bias related to increased awareness of adverse cardiovascular effects of GnRH agonists after the FDA’s warning label was issued, cardiovascular event reporting was analyzed separately for the time periods of 2000–2013 and 2014–2020 (supplementary table 5, A and B, https://www.jurology.com). We did not find a consistent pattern of increased event reporting on GnRH agonists in the latter time period, as would be expected with this type of reporting bias.

DISCUSSION

In this retrospective pharmacovigilance analysis of the FAERS database, GnRH antagonists were associated with fewer reports of cardiovascular adverse events than GnRH agonists, both as monotherapy and as combination therapy. Arterial vascular events accounted for most reported cardiovascular adverse events and were reported more commonly with GnRH agonist monotherapy and combination therapy. Second generation ARAs and abiraterone were associated with fewer reports of CAD/MI, ischemic stroke, and PVD than GnRH agonists, but more reports of hypertension requiring hospitalization and heart failure.

To our knowledge, this study is the first to explore differences in cardiovascular risk profiles among hormone therapies for prostate cancer in a pharmacovigilance database. Our findings are concordant with the favorable cardiovascular toxicity profile of GnRH antagonists as compared to GnRH agonists that have previously been described in the clinical trial setting. In 934 patients with advanced prostate cancer enrolled in the phase III HERO trial, the novel GnRH antagonist relugolix was associated with a 54% lower risk of nonfatal myocardial infarction, nonfatal stroke, and all-cause mortality as compared to leuprolide (2.9% vs 6.2%, hazard ratio 0.46 [95% CI 0.24–0.88]) after 48 weeks of treatment.¹² A smaller study of 80 men with prostate cancer and pre-existing cardiovascular disease found fewer major cardiovascular and cerebrovascular events on degarelix (3%) as compared to GnRH agonists (20%; p=0.013).¹¹ The ongoing PRONOUNCE clinical trial (NCT02663908) aims to determine prospectively the extent of cardiovascular risk reduction with GnRH antagonist therapy in patients with pre-existing cardiovascular disease.¹⁴ The mechanistic basis of cardiovascular risk reduction with GnRH antagonists requires further study.

Hypertension has been reported as a prominent cardiovascular toxicity of second generation ARAs and abiraterone. However, the risk of hypertension on these agents compared to GnRH agonists has not been well studied. A meta-analysis of 5 randomized controlled trials comparing second generation ARAs or abiraterone with placebo found an increased risk of all-grade (HR 1.84 [95% CI 1.37–2.46]) and high-grade (HR 1.77 [95% CI 1.13–2.77]) hypertension on these agents.¹⁵ Other meta-analyses of enzalutamide and abiraterone have found similar associations with hypertension in the clinical trial setting.^16,17 In addition to hypertension requiring hospitalization, we found higher reporting of heart failure on second generation ARAs and abiraterone, which frequently accompanies severe hypertension in the hospital setting. Heart failure has been reported as an adverse effect of enzalutamide in randomized clinical trials¹⁸ though the relevance of this toxicity in real-world clinical practice has not until now been well described.

Our findings also highlight an important potential role for the cardiovascular community to improve overall clinical outcomes for men with prostate cancer. As prostate cancer outcomes have improved, cardiovascular morbidity and mortality have risen^3,19 in this population, particularly arterial vascular events as supported by our data. Aggressive cardiovascular risk factor modification and primary prevention therapy with aspirin and/or statins are cornerstones of therapy for patients at increased risk for arterial vascular events, with newer agents such as low-dose anticoagulants also showing clinical benefit in this setting.²⁰ Although men with prostate cancer have been shown to be at increased cardiovascular risk,²¹ targeted clinical guidelines for cardiovascular risk reduction have yet to be produced. Additional study is needed to identify optimal strategies to reduce cardiovascular morbidity among men with prostate cancer receiving hormone therapy.

There are several important limitations to our study. First, this was a retrospective analysis of adverse cardiovascular event reporting available in the FAERS database. Key clinical information, such as prostate cancer characteristics, duration of anti-androgen therapy, cardiovascular comorbidities, and time to the cardiovascular event, is not available in this database. Additional study in databases containing more granular clinical information would allow multivariable and/or propensity score matched analyses to address potential confounders. Second, entries in the FAERS database are subject to potential reporting bias due to the voluntary nature of adverse event reporting. Third, we found many total and cardiovascular events reported on second generation ARAs and abiraterone monotherapy in the FAERS database; however, these agents are not routinely used as single-agent therapies as part of the clinical standard of care. Without access to the associated clinical data for these reported monotherapy uses, the clinical context is unclear. Our findings are therefore hypothesis-generating, and do not establish causality for the reporting associations identified.

CONCLUSIONS

Cardiovascular adverse event reporting among prostate cancer patients was higher with GnRH agonists than nonGnRH agonists in FAERS, a pharmacovigilance database. Arterial vascular events accounted for most cardiovascular adverse events reported and occurred more frequently on GnRH agonist monotherapy and combination therapy. Additional study is needed to identify optimal strategies to reduce cardiovascular morbidity among men with prostate cancer receiving hormone therapy.

Abbreviations and Acronyms

ADT

androgen deprivation therapy

ARA

androgen receptor antagonist

CAD

coronary artery disease

FAERS

FDA Adverse Event Reporting System

FDA

U.S. Food and Drug Administration

GnRH

gonadotropin-releasing hormone

MI

myocardial infarction

PVD

peripheral vascular disease

ROR

reported odds ratio

VTE

venous thromboembolism