Breast cancer is the most frequently diagnosed cancer in the United States (US) and worldwide with an increasing incidence rate of 0.5% each year [1, 2]. Hormone receptor-positive and human epidermal growth factor receptor type 2 negative (HR+/HER2–) is the most common subtype, with an approximate five-year survival rate of 90%. However, when diagnosed in late stages as advanced breast cancer (aBC), the five-year survival rate significantly decreases [3].

The cornerstone of treatment of aBC constitutes of endocrine therapy (ET): including aromatase inhibitors (such as letrozole, anastrozole, and exemestane), selective estrogen receptor (ER) modulators (such as tamoxifen), and ER down-regulators (such as fulvestrant). Unfortunately, the efficacy of these agents is limited by the presence of primary or acquired resistance [4, 5]. However, in recent years, it has been demonstrated that several solid tumors have direct modifications of genes that encode various proteins involved in cyclin-dependent kinases (CDK) and D-type cyclins activity and regulation [6]. As a result, small molecule inhibitors have been developed that target this mitogenic pathway, and three of them, known as cyclin-dependent kinases 4 and 6 (CDK4/6) inhibitors (palbociclib, ribociclib, and abemaciclib), are currently available for the treatment of aBC in combination with aromatase inhibitors or fulvestrant [6]. Three CDK4/6 inhibitors have been approved by the US Food and Drug Administration since 2015 as treatments in combination with ET for patients with HR+/HER– aBC, as informed by numerous randomized controlled trials (RCTs) [7, 8, 9, 10, 11, 12, 13]. In addition, several meta-analyses support this combined treatment which improves progression-free and overall survival of patients with aBC as compared to ET treatment alone [14, 15, 16, 17]. The American Society of Clinical Oncology guideline also recommends combined CDK4/6 and ET use [18, 19].

The use of CDK4/6 inhibitors, while providing clinical benefits, has raised concerns regarding cardiovascular events and cardiac toxicity [20, 21]. Significant QT prolongation was observed with ribociclib in the Mammary Oncology Assessment of LEE011’s (Ribociclib’s) Efficacy and Safety (MONALEESA) trials [9, 11, 13]. Furthermore, an updated review has reported that the use of palbociclib and ribociclib may be associated with an increased risk of thromboembolic events, based on data extracted from various clinical trials [22]. In addition, a prospective study has suggested that CDK4/6 inhibitors activate the inflammatory pathway, which may potentially lead to cardiovascular impairment [20].

Due to the heterogeneous safety profiles of different CDK4/6 inhibitors in relation to cardiovascular events, coupled with a lack of head-to-head trials comparing these drugs, we aimed to qualitatively synthesize the incidence of major adverse cardiovascular events (MACE) and hypertension associated with the use of CDK4/6 inhibitors reported in RCTs. Furthermore, this study conducted a network meta-analysis (NMA) to compare the MACE and hypertension risks associated with the use of CDK4/6 inhibitors.

This systematic review and network meta-analysis was reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) extension statement for systematic reviews, incorporating network meta-analyses for healthcare interventions.

The combination of CDK4/6 inhibitors and ET has been deemed an optimal treatment to prolong the survival outcomes of patients living with HR+/HER– aBC [15]. However, the negative impact of CDK4/6 inhibitors on the cardiovascular system has been investigated in some molecular studies [20, 31]. These studies have highlighted potential pathophysiological mechanisms such as increased vascular inflammation, left ventricular remodeling, and downregulation of the phosphatidylinositol 3‑kinase/protein kinase B (PI3/AKT) pathway [20, 31]. Despite the findings linking these effects to underlying cardiovascular disease development, a lack of sufficient evidence regarding the comparative cardiac safety of the three approved CDK4/6 inhibitors remains in the US. To address this gap, our systematic review and network meta-analysis confirms that the use of abemaciclib+ET could be associated with a lower risk of MACE, while ribociclib+ET had a less favorable safety profile regarding cardiovascular outcomes (including MACE and hypertension), based on mixed treatment comparisons. It is noteworthy that the sparse reporting of MACE events in the included phase III RCTs, may explain the lack of statistical significance in our study’s results. Nevertheless, both Bayesian and MH approaches indicated a similar trend of cardiac safety among the three CDK4/6 inhibitors in patients with HR+/HER– aBC.

Apart from ribociclib-induced QTc prolongation and non-cardiovascular related side effects, there is a lack of sufficient data available in other meta-analyses focusing on the incidence of cardiovascular events and outcomes associated with the use of these agents [21, 32, 33, 34]. The Palbociclib: Ongoing Trials in the Management of Breast Cancer (PALOMA) trials reported hematologic adverse events, primarily neutropenia, and leukopenia, as the most frequently observed adverse events with palbociclib [7, 34]. Conversely, neutropenia was the most commonly observed adverse event with ribociclib in the MONALEESA trials, where patients also showed QTc prolongation [9, 34]. Abemaciclib was associated with a higher incidence of gastrointestinal adverse events, which may be attributed to its distinct chemical structure and greater selectivity for CDK4 as compared to other CDK4/6 inhibitors [12, 35].

Our analysis showed that the combined use of abemaciclib plus ET was associated with a relatively lower risk of MACE. The included RCTs pertaining to abemaciclib also supported that the use of abemaciclib was not associated with a risk of MACE [8, 12, 30]. These findings could be explained by the unique structure of abemaciclib, which provides additional potency as a CDK9 inhibitor, compared to the other two agents that only exhibit reversible binding to CDK4/6 [33]. CDK9 is a kinase involved in cardiac hypertrophy [36], which suggests that abemaciclib possesses cardioprotective properties. The use of both CDK4/6 inhibitors and hormonal treatment could lead to endothelial injury and hypertension through vascular inflammation [20]. However, no statistically significant differences were detected in our analysis of hypertension incidence between the three agents. It is still unclear whether such inflammation is due to malignancy or drug-induced cardiotoxicity in cancer patients taking CDK4/6 inhibitors [20, 37]. Specific biomarkers associated with vascular inflammation in patients receiving CDK4/6 inhibitors have not been identified [38].

Our NMA indicated that ribociclib had a relatively higher risk of MACE compared to other agents. This may be attributed to its significantly higher likelihood of inducing QTc prolongation compared to the other two agents [9, 39]. Prolongation of the QTc interval is associated with sudden cardiac death and other heart-related adverse events [40]. Hence, the use of ribociclib in cancer patients with concurrent cardiovascular dysfunction should be carefully monitored and ribociclib should be avoided in patients with a QTc prolongation prior to initiating treatment.

In our study, the MH method was used as an alternative approach to testing the robustness of our findings, and the results were consistent with the trends of cardiac safety identified through the Bayesian NMA. However, the Bayesian approach exhibited wider CI than the MH method, although the point estimates were similar. Notably, a growing number of recent studies have discussed the use of Bayesian framework in the analysis of sparse outcomes, such as mortality and serious adverse events [41, 42]. The Bayesian model might be a concern as the prior distributions for the models’ parameters can strongly influence the results. In our study, a non-informative prior distribution was used in the random-effects Bayesian NMA, as there was no other external information established for our outcomes of interest. However, the impact of prior distributions on the posterior estimates of the model was believed to be more significant when dealing with sparse data [27]. Therefore, the utilization of both the Bayesian and MH methods for our network in this study adds further credibility to our findings.

Unfortunately, the RCTs analyzed in this network meta-analysis were underpowered to detect statistically significant differences in MACE and hypertension events among patients with aBC. This might result from a limited follow-up period and infrequent reporting of outcomes. For example, three RCTs were excluded due to a lack of reported hypertension events. More RCTs with a longer follow-up period and real-world studies involving a larger population are needed to validate the findings of our study. Furthermore, a future examination of patient subgroups with different comorbidities and age ranges would be valuable in helping clinicians determine which CKD4/6 inhibitors offer the most cardiovascular protection for patients with aBC.

Our study demonstrates that the combination of abemaciclib and ET led to a reduced risk of MACE in patients with HR+/HER2– aBC, while ribociclib plus ET was associated with a higher risk of cardiovascular events. Furthermore, the use of palbociclib+ET appeared to have a favorable safety profile with respect to hypertension. While most of the results in our analyses did not reach statistical significance, both Bayesian and MH network meta-analyses in our study supported the trend of comparative cardiovascular safety among the three CDK4/6 inhibitors. Given the growing importance of cardio-oncology, our findings could serve as a foundation for future research to further explore direct comparisons and aid in determining the optimal treatment choice.

aBC, advanced breast cancer; CDK4/6, cyclin-dependent kinases 4 and 6; CIs, credible intervals; ET, endocrine therapy; HR+/HER2–, hormone receptor-positive and human epidermal growth factor receptor type negative; MACE, major adverse cardiovascular events; MH, Mantel–Haenszel; NMA, network meta-analysis; OR, odds ratios; RCTs, randomized controlled trials; SUCRA, the surface under the cumulative ranking curve.

YL and CP designed the study design. YL developed the literature search. YL and KD conducted the literature research and performed the study selection. YL and KD extracted the data while YL handled the statistical part. YL and KD interpreted the results and wrote the manuscript. CP provided help and advice on the manuscript write-up. All authors read and approved the final manuscript. All authors have participated sufficiently in the work and agreed to be accountable for all aspects of the work.

Not applicable.

Not applicable.

This research received no external funding.

Yi-Shao Liu: YL is a current student assistant (co-op) at Novo Nordisk Inc. and a graduate student at the University of Texas at Austin. The work was completed without any involvement with Novo Nordisk Inc.

The rest of two authors declare no conflicts of interest.