Adverse Cardiovascular Events Associated With Cyclin‐Dependent Kinase 4/6 Inhibitors in Patients With Metastatic Breast Cancer

Abstract

Background

Cyclin‐dependent kinase (CDK) 4 and 6 inhibitors have significantly improved survival in patients with hormone receptor–positive metastatic breast cancer. There are few data regarding the epidemiology of cardiovascular adverse events (CVAEs) with these therapies.

Methods and Results

Using the OneFlorida Data Trust, adult patients without prior cardiovascular disease who received at least 1 CDK4/6 inhibitor were included in the analysis. CVAEs identified from International Classification of Diseases, Ninth and Tenth Revisions (ICD‐9/10) codes included hypertension, atrial fibrillation(AF)/atrial flutter (AFL), heart failure/cardiomyopathy, ischemic heart disease, and pericardial disease. Competing risk analysis (Fine–Gray model) was used to determine the association between CDK4/6 inhibitor therapy and incident CVAEs. The effect of CVAEs on all‐cause death was studied using Cox proportional hazard models. Propensity‐weight analyses were performed to compare these patients to a cohort of patients treated with anthracyclines. A total of 1376 patients treated with CDK4/6 inhibitors were included in the analysis. CVAEs occurred in 24% (35.9 per 100 person‐years). CVAEs were slightly higher in patients who received CKD4/6 inhibitors compared with anthracyclines (P=0.063), with higher death rate associated with the development of AF/AFL or cardiomyopathy/heart failure in the CDK4/6 group. The development of cardiomyopathy/heart failure and AF/AFL was associated with increased all‐cause death (adjusted hazard ratio [HR], 4.89 [95% CI, 2.98–8.05]; and 5.88 [95% CI, 3.56–9.73], respectively).

Conclusions

CVAEs may be more common with CDK4/6 inhibitors than previously recognized, with increased death rates in these patients who develop AF/AFL or heart failure. Further research is needed to definitively determine cardiovascular risk associated with these novel anticancer treatments.

Nonstandard Abbreviations and Acronyms

AFL

atrial flutter

CDK

cyclin‐dependent kinase

CVAE

cardiovascular adverse event

HER2

human epidermal growth factor receptor 2

MONALEESA‐2

Study of Efficacy and Safety of LEE011 in Postmenopausal Women With Advanced Breast Cancer

Clinical Perspective.

What Is New?

• Our real‐world data suggest that cardiovascular adverse events, especially cardiomyopathy and atrial fibrillation, are more common among patients treated with cyclin‐dependent kinase 4/6 inhibitors than previously described in safety and efficacy studies and are not just limited to QT prolongation.

• Mortality is increased in patients treated with cyclin‐dependent kinase 4/6 inhibitors who develop atrial fibrillation/atrial flutter or heart failure.

What Are the Clinical Implications?

• These prospective real‐world data of patients treated with cyclin‐dependent kinase 4/6 inhibitors may indicate the need to develop class‐specific cardiotoxicity management and mitigation strategies.

The cell cycle has long been a mechanistic target for anticancer therapeutics. The G1‐S phase transition, regulated in part by cyclins and cyclin‐dependent kinases (CDKs), has become a major area of focus in breast cancer, given the increasing appreciation for the pivotal role CDKs play in cell‐cycle progression and, specifically, the aberrations in this pathway leading to increased dysregulated proliferation. ¹ Indeed, inhibition of CDKs 4 and 6 (CDK4/6), key cell cycle proteins that are regulated by cyclin D1 (which is overexpressed in up to 45% of breast cancers), ² has become a standard of care in combination with endocrine therapy for patients with hormone receptor–positive metastatic breast cancer, where this approach has led to improvements in progression‐free and overall survival. ³ , ⁴ , ⁵ CDK4/6 inhibition combined with endocrine therapy has also recently been approved in the adjuvant setting for high‐risk hormone receptor–positive early‐stage breast cancer, where it has shown to improve disease‐free survival. ⁶ Three CDK4/6 inhibitors have been approved by the US Food and Drug Administration—abemaciclib, palbociclib, and ribociclib, with palbociclib being the first CDK4/6 inhibitor approved in 2015.

In general, CDK4/6 inhibitors have a relatively favorable toxicity profile, with myelosuppression, gastrointestinal, and skin toxicities most commonly reported in clinical trials. ³ , ⁴ , ⁵ Cardiovascular adverse events (CVAEs) are thought to be a rare occurrence, limited to QTc prolongation and venous thromboembolism. Interestingly, QT prolongation has not been demonstrated to be a class‐related effect and has been observed only with ribociclib. In the MONALEESA‐2 (Study of Efficacy and Safety of LEE011 in Postmenopausal Women With Advanced Breast Cancer) trial, 3.3% of the patients had an average QTc interval of >480 milliseconds when treated with the 600‐mg dose; however, there were no reported cases of torsades de pointes, and the QTc improved with dose reduction or cessation. ⁴ Thromboembolic events have been reported with palbociclib and ribociclib, with rates of 1.8% and 5%, respectively, in various clinical trials. ³ , ⁴ , ⁷ Nevertheless, there has yet to be a systematic evaluation of CVAEs in patients treated with CDK4/6 inhibitors. Using the OneFlorida database, this study aimed to evaluate the incidence of CVAEs in women with metastatic breast cancer treated with CDK4/6 inhibitors in real‐world clinical settings.

Methods

The OneFlorida data are Health Insurance Portability and Accountability Act limited data set and require a data use agreement with the OneFlorida Clinical Research Consortium. The data that support the findings of this study are available from the corresponding author upon reasonable request. Additional data access requests can be made through https://onefloridaconsortium.org/.

OneFlorida Clinical Research Consortium

This was a retrospective cohort study of patients in the OneFlorida Clinical Research Consortium. OneFlorida is a cooperative research team in the State of Florida that aims to improve statewide health research capacity and opportunities through community health care centers and local outpatient facilities. It is one of the original 13 clinical data research networks in the US Patient‐Centered Clinical Research Network. ⁸ OneFlorida partners provide health care to ≈15 million patients, which is >40% of Floridians, across all 67 Florida counties. ⁹ The OneFlorida Data Trust is the informatics infrastructure that includes deidentified information on electronic health records, administrative claims, and other individual‐level health‐related data on a broad‐based, unselected population of >15 million people in Florida. The protected health information in OneFlorida is limited to dates (eg, birthdates and dates of service) and locations (to zip code level), constituting a limited data set under the Health Insurance Portability and Accountability Act, to protect the health information of patients.

Cohort Development

For this analysis, electronic health record data of OneFlorida participants were queried to extract relevant information from January 1, 2013, until the latest data refresh before data analysis (April 30, 2020). Female patients diagnosed with metastatic breast cancer who received at least 1 cycle of a CDK4/6 inhibitor (palbociclib, ribociclib, abemaciclib) at standard dosing between January 1, 2013, and April 30, 2020, without exposure to either anthracyclines or human epidermal growth factor receptor 2 (HER2) inhibitor (eg, trastuzumab or pertuzumab) were included in the analysis. Age, race and ethnicity, obesity, diabetes, preexisting hypertension, and hyperlipidemia were baseline characteristics. Female patients with breast cancer treated with anthracyclines (but not CDK4/6 or HER2 inhibitors) were included as the comparator group, given its well‐established cardiotoxicity risk profile. The study was approved by the University of Florida Institutional Review Board (IRB201702876), and informed consent was not required, given the deidentified nature of the OneFlorida database.

Cardiovascular Adverse Events

Postchemotherapy CVAE was defined from International Classification of Diseases, Ninth and Tenth Revisions (ICD 9/10) codes (Table S1) for newly diagnosed hypertension, atrial fibrillation (AF)/atrial flutter (AFL), heart failure/cardiomyopathy, ischemic heart disease, and pericardial disease without preceding cardiovascular conditions documented in the 1 year before the first chemotherapy prescription. CVAE incidents within 12 months after the first prescription date of CDK4/6 inhibitor or anthracyclines were evaluated and calculated as 100 person‐years. The time to CVAE was calculated or censored on the basis of the date of April 30, 2020.

Statistical Analysis

The baseline characteristics of the 2 groups were compared using a t test for continuous variables and a chi‐square test for categorical variables as appropriate. To balance the differences in the baseline characteristics in the patients in the CDK4/6 inhibitor group and the anthracycline group, an overlap weighting propensity score method was used in the analysis. ¹⁰ , ¹¹ In this method, each patient's weight is the probability of that patient being in 1 of the 2 groups. ¹⁰ The properties of overlap weighting have been demonstrated theoretically and were found to improve in balance and precision relative to the inverse probability weighting. ¹⁰ The overlap weighting was used for adjusted model analysis, and the variables included in the propensity score calculation were age, race and ethnicity, obesity, diabetes, preexisting hypertension, and hyperlipidemia. Time to CVAE was compared between the 2 groups using competing risk survival analysis based on the Fine–Gray model, ¹² where death was considered as a competing risk for cardiotoxicity. Survival analysis modeling for all‐cause death using a Cox proportional hazards model was performed to compare the death between patients on CDK4/6 inhibitors and anthracyclines with a CVAE. The analysis was limited to a follow‐up of 3 years or censored on the basis of April 30, 2020. We also performed survival analysis using similar adjusted Cox regression hazard models for each CVAE or any CVAE in patients who received CDK4/6 inhibitors, wherein we compared them with those who did not experience any CVAE. Finally, a case–control sensitivity analysis of stage IV patients treated with anthracycline was compared with age‐ and race‐matched patients treated with CDK4/6 inhibitors. P values <0.05 were considered statistically significant. All analyses were performed in SAS version 9.4 (SAS Institute, Cary, NC) or R version 4.0.2 (R Foundation for Statistical Computing, Vienna, Austria).

Results

Demographics

A total of 1376 patients treated with CDK4/6 inhibitors with no exposure to anthracyclines or HER2 inhibitors were included in the analysis. Our cohort had a proportionately higher number of patients treated with palbociclib (n=1049) than ribociclib (n=156) or abemaciclib (n=74). Baseline demographics and clinical characteristics of the cohort are presented in Table 1. The mean age of women treated with CDK4/6 inhibitors was 62±18 years. Notably, 51% of patients included in this cohort self‐identified as non‐Hispanic White, 12% as non‐Hispanic Black, and 28% as Hispanic. Baseline cardiovascular risk factors included 14% having a preexisting diagnosis of hypertension, 11% categorized as obese (defined as body mass index ≥30), and 7% with diabetes or hyperlipidemia. In addition, we also included 1215 females with breast cancer treated with anthracyclines in this analysis to compare their CVAE profile to CDK4/6 inhibitors, given the known cardiotoxicity of anthracyclines. Compared with the women treated with CDK4/6 inhibitors, those who were treated with anthracyclines were younger (mean age, 53), more likely to be obese (16%), and less likely to have hypertension (11%) or hyperlipidemia (4%; Table 1).

Table 1.

Baseline Characteristics

Characteristic	CDK 4/6 inhibitors, N=1376	Anthracyclines, N=1215
Age	62±18	53±12
Race or ethnicity
White (non‐Hispanic)	706 (51)	467 (38)
Black (non‐Hispanic)	168 (12)	259 (21)
Hispanic	384 (28)	411 (34)
Other	118 (8.6)	78 (6.4)
Obesity	147 (11)	193 (16)
Diabetes	95 (6.9)	70 (5.8)
Hypertension	197 (14)	138 (11)
Hyperlipidemia	97 (7.0)	51 (4.2)

Continuous variables are presented as mean±standard deviation while categorical variables are presented as frequencies (percentages).

CDK indicates cyclin‐dependent kinase.

CVAEs Associated With CDK4/6 Inhibitor Therapy

We determined the incidence rate of 5 different CVAEs (Table 2). CVAEs occurred in 24% of the CDK4/6 cohort, calculated as 35.9 per 100 person‐years. The most common CVAE was hypertension (22.7 per 100 person‐years), cardiomyopathy/heart failure (7.8 per 100 person‐years), and AF/AFL (6.5 per 100 person‐years). The median time to event for all CVAEs was 2.3 months (interquartile range, 0.9–5.8). Among all patients who received CDK4/6 inhibitors for breast cancer treatment, those who had any CVAE had higher all‐cause death at 3‐year follow‐up (adjusted hazard ratio [aHR], 2.13 [95% CI, 1.46–3.11], Figure 1A). There was no difference in CVAE incidence among the different CDK4/6 inhibitors (Table S2).

Table 2.

Adverse Cardiovascular Events Associated With CDK4/6 Inhibitor or Anthracycline Exposure

Cardiovascular events, n (%)	CDK4/6 inhibitors, N=1376		Anthracyclines, N=1215
	Event, n (%)	Rate (per 100 person‐years)	Event, n (%)	Rate (per 100 person‐years)	P value*
Cardiomyopathy/heart failure	80 (5.8)	7.8	74 (6.1)	7.5	0.77
Atrial fibrillation/flutter	68 (4.9)	6.5	54 (4.4)	5.4	0.55
Pericardial disease	28 (2.0)	2.6	22 (1.8)	2.2	0.68
Hypertension	214 (16)	22.7	157 (13)	17	0.056
Ischemic heart disease	48 (3.5)	4.6	29 (2.4)	2.9	0.10
Any cardiotoxicity†	325 (24)	35.9	250 (21)	28.5	0.063

CDK indicates cyclin‐dependent kinase.

^*Statistical test performed: chi‐square test of independence.

^† Any cardiotoxicity is a composite of cardiomyopathy/heart failure, atrial fibrillation/flutter, pericardial disease, hypertension, or ischemic heart disease.

Figure 1. Kaplan–Meier curves of death in women treated with CDK4/6 inhibitors who developed cardiovascular adverse events compared with those who did not.*.

A, Any CVAE; B, hypertension; C, cardiomyopathy/heart failure; D, atrial fibrillation/flutter. *Adjusted P values were derived from Cox regression analysis adjusting for age, race and ethnicity, obesity, diabetes, preexisting hypertension, and hyperlipidemia. AFib/Aflutter indicates atrial fibrillation/atrial flutter; CDK, cyclin‐dependent kinase; CVAE, cardiovascular adverse event; and HR, hazard ratio.

Incident Hypertension

Hypertension was the most common CVAE noted after starting CDK4/6 inhibitors. Those who developed hypertension were generally older (66 years), had a higher burden of other cardiovascular disease risk factors (12% diabetes, 10% hyperlipidemia, 16% preexisting hypertension). The median time to event for the development of hypertension was 1.9 months (interquartile range, 0.5–4.7). The development of hypertension was not associated with all‐cause death at a 3‐year follow‐up, with an aHR of 1.44 (95% CI, 0.93–2.25; Figure 1B).

Cardiomyopathy/Heart Failure

Cardiomyopathy/heart failure was the second most common CVAE in women treated with CDK4/6 inhibitors. Those who developed cardiomyopathy/heart failure were generally older (63 years), had a higher burden of other cardiovascular disease risk factors (19% diabetes, 17% hyperlipidemia, 31% preexisting hypertension). The median time to event for the development of cardiomyopathy/heart failure was 2.4 months (interquartile range, 0.9–5.8). Development of cardiomyopathy/heart failure was associated with increased all‐cause death at a 3‐year follow‐up with an aHR of 4.89 (95% CI, 2.98–8.05; Figure 1C).

Atrial Fibrillation and Atrial Flutter

AF and AFL were the third most common CVAEs noted after starting CDK4/6 inhibitors. Those who developed AF/AFL were generally older (69 years) and had a higher burden of other cardiovascular disease risk factors (13% diabetes, 15% hyperlipidemia). The median time to event for the development of AF or AFL was 2.1 months (interquartile range, 0.95–6.44). Development of AF/AFL was associated with increased all‐cause death at a 3‐year follow‐up with an aHR of 5.88 (95% CI, 3.56–9.73; Figure 1D).

Other Cardiovascular Events

Pericardial disease and ischemic heart disease were rare in our cohort, with an incidence of 2.6 and 4.6 per 100 person‐years of exposure to CDK4/6 inhibitors (Table 2). While there is no increase in deaths among those who developed ischemic heart disease, there is an increased death rate in those who developed pericardial disease (Figure S1).

Comparison With Anthracycline

In this study, we compared CDK4/6 inhibitors with anthracyclines due to their known cardiotoxicity profile. The overall CVAE in a group of female patients with breast cancer who had exposure to CDK4/6 inhibitors was slightly higher than for those treated with anthracyclines, with marginal significance (35.9 versus 28.5 per 100 person‐years; P=0.063; Table 2); however, there was no statistical difference between these 2 groups in the competing risk analysis (Figure 2A).

Figure 2. Competing risk analysis for the development of cardiovascular adverse events in women with breast cancer who were treated with CDK4/6 inhibitors vs those treated with anthracyclines.*.

A, Any CVAE; B, hypertension; C, cardiomyopathy/heart failure; D, atrial fibrillation/flutter. *Covariates in the adjusted analysis included age, race and ethnicity, obesity, diabetes, preexisting hypertension, and hyperlipidemia. AFib/Aflutter indicates atrial fibrillation/atrial flutter; CDK, cyclin‐dependent kinase; CVAE, cardiovascular adverse event; and HR, hazard ratio.

There was no difference in deaths among patients exposed to CDK4/6 inhibitors and anthracyclines after they experienced a CVAE, with an aHR of 1.35 (95% CI, 0.88–2.07; Figure 3A). There was no difference in risk of hypertension (Figure 2B) and death among patients exposed to CDK4/6 inhibitors and anthracyclines after they experienced hypertensive events (Figure 3B). Although there were no differences in the incidence of cardiomyopathy/heart failure and AF/AFL in the women treated with CDK4/6 inhibitors compared with those treated with anthracyclines (Figure 2C and 2D), those in the CDK4/6 inhibitor group had higher death rates compared with those in the anthracycline group, with an aHR of 2.20 (95% CI, 1.14–4.25) for cardiomyopathy/heart failure and 2.59 (95% CI, 1.20–5.61) for AF/AFL, respectively (Figure 3C and 3D). There was no difference in the incidence and death associated with pericardial disease and ischemic heart disease between the 2 groups (Figure S2).

Figure 3. Kaplan–Meier curves of death in women who developed CVAE after treatment with CKD4/6 inhibitors vs those treated with anthracyclines.

A, Any CVAE; B, hypertension; C, cardiomyopathy/heart failure; D, atrial fibrillation/flutter. AFib/Aflutter indicates atrial fibrillation/atrial flutter; CDK, cyclin‐dependent kinase; CVAE, cardiovascular adverse event; and HR, hazard ratio.

We performed a case–control sensitivity analysis of stage IV patients treated with anthracycline on the basis of data obtained from the OneFlorida tumor registry and compared them with age‐ and race‐matched patients treated with CDK4/6 inhibitors in an attempt to control for cofounding factors associated with more advanced malignancy. There was no difference in the rate of CVAEs between the 2 groups, and there was no significant difference in these results when compared with the larger cohort (Table S3).

Discussion

The present study demonstrates that a substantial proportion of women with breast cancer treated with CDK4/6 inhibitors (24% or 36 per 100 person‐years) developed CVAEs, most of which were hypertension, heart failure, and AF/AFL. Our real‐world data suggest that CVAEs are more common among patients treated with CDK4/6 inhibitors than previously described in safety and efficacy studies and are not just limited to QT prolongation. Although there were no differences in the incidence of cardiomyopathy/heart failure and AF/AFL in the women treated with CDK4/6 inhibitors compared with those treated with anthracyclines, there was a higher death rate when these CVAEs occurred in the setting of CDK4/6 inhibitor use. ¹³ , ¹⁴ Rates of cardiotoxicity were identical regardless of which CDK4/6 inhibitor was used. These prospective real‐world data of patients treated with CDK4/6 inhibitors are hypothesis generating and may indicate the need to develop class‐specific cardiotoxicity management and mitigation strategies.

CDK4/6 inhibitors have been generally regarded as tolerable and safe from a cardiovascular standpoint. ¹⁵ , ¹⁶ , ¹⁷ With the exception of ribociclib‐induced QTc prolongation, ⁴ there is otherwise a paucity of data regarding the incidence of CVAEs and outcomes with the use of these agents. Nevertheless, several basic and translational studies have suggested that CDK4/6 inhibitors may directly affect the cardiovascular system, including alteration of potassium and sodium channel activity, increased vascular inflammation and left ventricular remodeling, and downregulation of the PI3/AKT pathway, all of which have been associated with the development of cardiovascular disease. ¹⁸ , ¹⁹ , ²⁰ To our knowledge, this is the first observational study investigating the cardiovascular toxicity profile of CDK4/6 inhibitor therapy in a real‐world cohort of patients with advanced breast cancer. The sample size is considerably larger than previous studies detailing more common treatment‐related adverse effects. ²¹ , ²² Furthermore, our cohort comprises a diverse racial and ethnic population, with 40% of patients treated with CDK4/6 inhibitor identifying as Black or Hispanic; this proportion of historically underrepresented patients is significantly higher than the randomized controlled trials conducted to gain Food and Drug Administration approval. ³ , ⁴ , ⁵ , ²² Nevertheless, the lack of prospectively defined cardiovascular end points in clinical trials for these agents remains a significant gap in our understanding of their true potential for cardiotoxicity.

To further investigate the outcomes of CDK4/6 inhibitor CVAEs, a group of female patients with breast cancer treated with anthracyclines was used as a comparator group. Anthracyclines were chosen due to their established CVAE risk profile in patients with breast cancer with significant associated morbidity and death, most notably heart failure. ²³ , ²⁴ CVAEs were slightly higher in patients who received CKD4/6 inhibitors compared with anthracyclines (P=0.063). It is important to note that indications for treatment with anthracyclines and CDK4/6 inhibitors are not necessarily the same, and due to limitations of the OneFlorida database, we could not establish the stage or pathology of the patients treated with anthracycline. Moreover, patients treated with anthracyclines were younger, which could have impacted the likelihood of developing adverse cardiovascular events. While fewer events were observed with the anthracycline group, this may be due to the typically lower doses (240 mg/m²) used in breast cancer treatment protocols, though we cannot confirm exact dosing due to constraints of the database. Thus, confounding effects related to disease severity, prior chemotherapy exposure, and baseline patient characteristics might limit detecting CVAEs in a standard comparison. It must be emphasized that while these data demonstrate an association between CVAEs and exposure to CDK4/6 inhibitors, they do not prove causation.

Survival analyses suggested poor outcomes for patients treated with CDK4/6 inhibitors who developed CVAE. Notably, there was a statistically significant decrease in survival during the 3‐year follow‐up period among patients exposed to CDK4/6 inhibitors who experienced a CVAE compared with patients who did not experience a CVAE (aHR, 2.13 [95% CI, 1.46–3.11]). Subgroup analysis revealed that this finding was primarily driven by the development of AF/AFL and heart failure/cardiomyopathy. Moreover, patients treated with a CDK4/6 inhibitor who developed AF/AFL or heart failure/cardiomyopathy had worse overall survival compared with patients treated anthracycline who developed the same CVAE. Similar results were observed with a matched case–control sensitivity analysis of patients with stage IV breast cancer treated with anthracyclines; however, the sample size was small, and further studies should be conducted with a larger group of patients with advanced cancer to confirm these results. Although this result is noteworthy, given the well‐recognized cardiotoxicity of anthracyclines and the understudied CVAE profile of CDK4/6 inhibitors, it must be emphasized that the broader range of breast cancer phenotypes included in the anthracycline comparator group (including those with earlier‐stage breast cancer) may have impacted the outcomes reported in this study.

Competing risk analyses were performed to account for potential differences in cancer‐related death between the groups, where death was treated as a competing risk for CVAE development; this methodology helps address the possibility of biased risk assessments. ¹² , ²⁵ In our adjusted Fine–Gray model, arrhythmia occurred more frequently in CDK4/6‐exposed patients compared with anthracycline‐exposed patients during the 1‐year follow‐up period when accounting for the competing risk of death. Additionally, there was an increased risk of ischemic heart disease that reached statistical significance in the CDK4/6 cohort; however, the CIs were wide, and there was no significant difference in the risk of developing hypertension, pericardial disease, or heart failure/cardiomyopathy between the CDK4/6 inhibitor–exposed group and the anthracycline‐exposed group. While CVAEs frequently occur with CDK4/6 exposure, the cardiotoxicity risk profile is similar to that of anthracyclines with the exception of AF/AFL, which were associated with worse outcomes in the CDK4/6 group.

Our study has several limitations. It is retrospective and therefore has the inherent biases of this study design. Moreover, reliance on ICD codes and not direct patient‐level data can lead to inaccuracies in the diagnosis of CVAEs. In addition, the OneFlorida registry lacks the granularity of certain patient‐level cancer characteristics, including hormone receptor status, previous systemic chemotherapy exposure, and prior surgery or radiation therapy. Because these characteristics help define disease severity, which can impact overall death rates, their absence introduces the risk of confounding. Moreover, without the aforementioned disease‐classifying data, it is possible that there could be differences in death outcomes between the CDK4/6 inhibitor and anthracycline arms on the basis of more aggressive histologic/molecular subtypes or progression through multiple lines of therapy. Similarly, more aggressive screening of patients treated with anthracyclines compared with CDK4/6 inhibitors may have allowed for the earlier identification of CVAEs, thereby impacting death rates. We did perform a matched case–control sensitivity analysis comparing stage IV patients treated with anthracycline with those treated with CDK4/6 inhibitors (albeit small numbers due to the small number of patients with documented stage IV diseases in the tumor registry), and there were no significant differences when compared with the larger data set. Finally, we were not able to assess patient adherence or dose reductions on the basis of toxicity. Future prospective studies in a population with additional detailed cancer‐specific data and a more narrowly matched comparator group are necessary.

Conclusions

Cardiac events were common in patients with metastatic breast cancer treated with CDK4/6 inhibitors, occurring in 24% of exposed individuals enrolled in the OneFlorida database. Additionally, patients treated with a CDK4/6 inhibitor who developed AF/AFL or heart failure/cardiomyopathy had worse overall survival compared with patients treated with anthracycline who developed the same CVAE. These findings suggest that patients taking CDK4/6 inhibitors may warrant closer monitoring for CVAE to minimize treatment disruption and ensure patient safety. Further prospective studies are necessary to fully characterize the cardiovascular toxicity profile and potential associated morbidity and death of this novel class of medications.

Sources of Funding

Research efforts of Dr Gong, including those reported in this study, are supported in part by the University of Florida Clinical and Translational Science Institute, which is supported in part by the National Institutes of Health National Center for Advancing Translational Sciences under award number UL1TR001427. The content is solely the authors' responsibility and does not necessarily represent the official views of the National Institutes of Health. In addition, Dr Fradley is supported by NIH R01HL151659, Dr Gong is supported by National Institutes of Health R01HL151659, R56DE030538, R01HG011800, R01HL149752, and R01HD071779, and Dr Guha is supported by American Heart Association‐Strategically Focused Research Network Grant in Disparities in Cardio‐Oncology (Nos. 847740 and 863620); however, these funders did not have a role in supporting this research.

Disclosures

Dr Fradley reports a research grant from Medtronic and personal fees from AbbVie, AstraZeneca, Johnson and Johnson, Myovant, Pfizer, and Zoll. The remaining authors have no disclosures to report.

Supporting information

Tables S1–S3

Figures S1–S2