Musculoskeletal adverse events associated with CDK4/6 inhibitors: a real-world study using FDA Adverse Event Reporting System (FAERS) database

Abstract

Objective

Cyclin-dependent kinase (CDK)-4/6 inhibitors have significantly improved outcomes in several cancers but can also induce various organ system toxicities, including musculoskeletal disorders. This study aimed to comprehensively characterize the musculoskeletal adverse events (MSAEs) associated with CDK4/6 inhibitors based on real-world data.

Methods

Reports of MSAEs linked to CDK4/6 inhibitors from the first quarter (Q1) of 2015 and 2023 Q4 were extracted from the FAERS. Descriptive analyses evaluated report frequencies over time and patient characteristics. Disproportionality analyses using reporting odds ratios (RORs) identified signals for specific musculoskeletal preferred terms (PTs). Time-to-onset analyses examined the temporal patterns of MSAEs.

Results

A total of 10,095 MSAE reports associated with CDK4/6 inhibitors were identified, most involving Palbociclib (n = 7819). The median age of patients was 64 years (IQR: 55–72), predominantly female (97.73%). Most reports were submitted by consumers (47.62%) and the majority of reports were from the United States (71.53%). Disproportionality analyses revealed distinct signals, with Ribociclib showing prominent signals for bone pain and bone lesions, and Abemaciclib for osteonecrosis of the jaw and pathological fractures. Palbociclib demonstrated a consistent but less pronounced signal across musculoskeletal PTs. Time-to-onset analyses demonstrated a significantly longer onset of MSAEs for Palbociclib (median 82 days, IQR[14–311]) compared to Abemaciclib (32.5 days, IQR[12–119]) and Ribociclib (34 days, IQR[8-177]) using the nonparametric Kruskal-Wallis test (P-value = 3.048e-11).

Conclusion

Musculoskeletal toxicities is a significant adverse event that affects drug safety. Early identification and proper management of these events are crucial for patients receiving CDK4/6 inhibitors. Further research is warranted to elucidate the underlying mechanisms and improve risk mitigation strategies.

Introduction

Cyclin-dependent kinases (CDKs) are essential regulators of the cell cycle, specifically facilitating the transition from the G1 phase to the S phase through their interaction with D-type cyclins [1]. CDK4 and CDK6, when complexed with these cyclins, phosphorylate the retinoblastoma protein (Rb), releasing E2F transcription factors that drive DNA synthesis. This regulation ensures controlled cell division, which is essential for preventing unchecked cellular proliferation, a hallmark of cancer [2, 3]. Consequently, CDK4/6 inhibitors have emerged as pivotal agents in cancer therapy, particularly for hormone receptor-positive (HR+) breast cancer [4].

CDK4/6 inhibitors, including Palbociclib, Ribociclib, and Abemaciclib, have revolutionized the treatment of HR-positive, HER2-negative breast cancer with the significant extension of overall survival (OS) to 58.7 months [5]. Clinical trials such as PALOMA-3, MONALEESA-2 and MONARCH-3 have demonstrated their efficacy in combination with endocrine therapies, significantly improving progression-free survival (PFS) (9.5 vs. 4.6 months, 25.3 vs. 16.0 month, and 20.5 vs. 12.8 months, respectively) [6,7,8]. Despite their clinical benefits, these inhibitors are associated with various adverse events (AEs), such as musculoskeletal adverse events (MSAEs) [9,10,11]. These adverse events can severely impact patients’ quality of life and may lead to treatment discontinuation. Frequently observed events include arthralgia, back pain, muscle spasms, and bone pain, with lower incidences of serious outcomes like pathological fractures [11, 12].

The precise mechanisms by which CDK4/6 inhibitors cause MSAEs are not fully understood but may involve the disruption of normal cellular functions in muscle and bone tissues, inflammatory processes, mitochondrial dysfunction, and alterations in the extracellular matrix [10, 13]. These adverse reactions have a significant impact, reducing drug adherence by approximately 20% as reported in previous literature [14], it is crucial to understand their prevalence, onset, and severity through real-world evidence. The FDA Adverse Event Reporting System (FAERS) provides a valuable resource for this purpose, offering a large and diverse dataset for identifying potential safety signals that may not be evident in controlled clinical trials [15, 16]. This study aims to utilize the FAERS data to comprehensively characterize MSAEs associated with CDK4/6 inhibitors in the patients receiving therapy for ER+/HER2- breast cancer, focusing on identifying the types and frequencies of these events, analyzing their onset time, and assessing their severity. Moreover, the study will compare the MSAEs profiles of various CDK4/6 inhibitors to determine any differences or patterns. By leveraging real-world evidence, our study aims to enhance the understanding of the risks of MSAEs associated with CDK4/6 inhibitors and inform clinical practices to improve patient outcomes.

Methods

Data source and study design

This study utilized the FAERS database, which is a vital resource for monitoring the safety of pharmaceutical products post-marketing. The FAERS database collects spontaneous reports of adverse events from healthcare professionals, consumers, and manufacturers, covering a wide range of patient demographics, drug types, and adverse events [17]. Palbociclib was firstly approved in February 2015 for use in combination with letrozole for the treatment of postmenopausal women with estrogen receptor (ER)-positive, HER2-negative advanced breast cancer in the first line setting, while Ribociclib was firstly approved in March 2017 with an aromatase inhibitor for the adjuvant treatment of adults with hormone receptor (HR)-positive, HER2-negative breast cancer in the first line setting and Abemaciclib was approved in September 2017 in combination with in combination with fulvestrant for women with HR-positive, HER2-negative advanced or metastatic breast cancer with disease progression following endocrine therapy [18]. For this study, the data were extracted from the FAERS database starting from the first month of 2015 to the December of 2023 in order to cover all three drug adoption and monitoring of adverse events. The database includes seven data files: DEMO (demographic and administrative information), DRUG (drug information), REAC (adverse event information), OUTC (patient outcomes), RPSR (report sources), THER (therapy dates), and INDI (indications for drug use). The study was conducted in accordance with the Reporting of a Disproportionality Analysis for Drug Safety Signal Detection Using Individual Case Safety Reports in PharmacoVigilance (READUS-PV) guideline [19].

Data extraction

CDK4/6 inhibitors were identified in the FAERS database using their generic names, brand names, or any drug codes: Palbociclib (Ibrance, PD-0332991), Ribociclib (Kisqali, Kryxana, LEE011), and Abemaciclib (Verzenio, Ramiven, LY2835219). The data extraction process involved several steps to ensure a comprehensive and accurate collection of relevant reports. Initially, all reports mentioning these drugs were extracted. Next, we filtered these reports to include only those where CDK4/6 inhibitors were identified as the primary suspect (PS) drug to ensure that the adverse events were most likely attributable to the inhibitors. The Medical Dictionary for Regulatory Activities (MedDRA) terminology was used to code adverse events, focusing specifically on preferred terms (PTs) related to the musculoskeletal system. Data cleaning involved the removal of duplicate reports to ensure each case was unique following the FDA’s recommendations, selecting the higher PRIMARYID when the CASEID and FDA_DT were the same and selecting the latest FDA_DT when the CASEIDs were the same to ensure the most up-to-date and accurate record for each patient, reflecting the most recent adverse event reports.

Descriptive analysis

We performed a detailed descriptive analysis to compile the clinical characteristics of FAERS reports documenting MSAEs linked to CDK4/6 inhibitor treatment. The variables examined included gender, country, outcome, FDA receipt date, CDK4/6 inhibitors, report type, and other pertinent clinical details. The time-to-onset of MSAEs was calculated by subtracting the treatment start date (START_DT) from the event start date (EVENT_DT). Reports with incorrect or missing data, such as cases where EVENT_DT occurred before START_DT, were excluded. Cumulative distribution curves were employed to illustrate the time-to-onset among different comparison groups.

Disproportionality analyses

Disproportionality analysis was conducted to identify signals of MSAEs associated with CDK4/6 inhibitors. This analysis compares the observed frequency of specific adverse events for a drug with the expected frequency in the FAERS database, helping to detect potential safety signals. In our study, the reporting odds ratio (ROR) were applied was used to identify significant MSAE signals and compare the association between each CDK4/6 inhibitor and MSAE. The ROR and its 95% confidence interval (CI) were calculated using Eq. 1:

$${\text{ROR}}=\frac{{{\text{a}}/{\text{c}}}}{{b/d}}$$

$$95\% CI={{\text{e}}^{\ln \left( {ROR} \right) \pm 1.96 \times \sqrt {\frac{1}{a}+\frac{1}{b}+\frac{1}{c}+\frac{1}{d}} }}$$

(1)

In this equation, a is the number of reports containing both the drug of interest and musculoskeletal adverse events; b is the number of reports containing musculoskeletal adverse events with other drugs; c is the number of reports containing the drug of interest with other adverse events; d is the number of reports containing other drugs and other adverse events. Signals were considered significant if the ROR lower limit of the 95% CI was greater than 1 and the number of cases was ≥ 3.

Statistical analysis

Descriptive statistics were used to summarize the clinical characteristics of the MSAEs reports, including patient demographics (age, gender), geographic distribution, and the types and frequencies of MSAEs. Median values and interquartile ranges (IQR) were calculated for continuous variables such as age. Distribution of categorical variables was summarized using frequencies and percentages. In our study, we employed the nonparametric Kruskal–Wallis test using R version 4.3.3 with statistical significance set at p < 0.05 to compare the median time-to-onset between groups among CDK4/6 inhibitors-associated adverse events. The Kruskal-Wallis test showed a statistically significant difference in median time-to-onset between groups (H = 12.5, df = 3, p = 0.006).

Results

Descriptive analysis of MSAE reports associated with CDK4/6 inhibitors

The selection process for MSAEs associated with CDK4/6 inhibitors in the FAERS database between 2015 Q1 and 2023 Q4 is depicted in Fig. 1. The FAERS database contained 14,101,548 records while 12,178,164 records were remained when the duplicate records removed. The drug-related (DRUG) records totaled 59,359,244, and reaction (REAC) records totaled 34,650,743. From these, 318,005 preferred terms (PTs) induced by CDK4/6 inhibitors were identified, and 95,313 adverse event reports were specifically associated with CDK4/6 inhibitors as the primary suspect (PS). Among these, 10,095 reports were identified as MSAEs linked to CDK4/6 inhibitors and the majority of musculoskeletal toxicity cases were reported with Palbociclib (n = 7819), followed by Ribociclib (n = 1895), and Abemaciclib (n = 381). The Fig. 2 illustrates the trend in the number of MSAE reports associated with CDK4/6 inhibitors in the FAERS database from 2015 to 2023. Overall, the number of reports increased significantly over this period, with the annual number reaching 2480 by 2023. Palbociclib had a steady increase in reports, peaking at 1596. Ribociclib reports also rose consistently, with a notable increase to 761 reports by 2023. In contrast, Abemaciclib had relatively fewer reports with 123 MSAEs reports in 2023.

Fig. 1

Flow chart showing the selection process of musculoskeletal adverse events for CDK4/6 inhibitors in the Food and Drug Administration Adverse Event Reporting System (FAERS). PTs preferred terms, PS primary suspected drugs

Fig. 2

The number of reported cases of three CDK4/6 inhibitors-induced musculoskeletal adverse events from 2015 to 2023

The clinical characteristics of MSAE reports associated with CDK4/6 inhibitors in the FAERS database between Q1 2015 and Q4 2023 were analyzed. The median age of patients was 64 years (IQR: 55–72), predominantly female (97.73%). Most reports were submitted by consumers (47.62%), particularly for Ribociclib (66.39%), compared to Abemaciclib (58.53%) and Palbociclib (42.54%). Pharmacists contributed more reports for Palbociclib (32.15%) compared to Ribociclib (11.50%) and Abemaciclib (18.11%). The outcomes included death (6.38%), hospitalization (25.31%), disability (1.31%), life-threatening events (1.89%), and required intervention to prevent permanent impairment (0.11%). The outcomes showed that the overall 6.38% fatality proportion associated with musculoskeletal toxicities for these three CDK4/6 inhibitors and the death case was reported more frequently with Ribociclib (8.52%) compared to Palbociclib (5.60%) and Abemaciclib (2.12%). The majority of reports were from the United States (71.53%), with notable contributions from India (4.29%), Argentina (4.29%), Brazil (2.24%), and Colombia (1.64%). More detailed characteristics are presented in Table 1.

Table 1 Clinical characteristics of MSAE reports associated with CDK4/6 inhibitors in the FAERS database between 2015 Q1 and 2023 Q4

Disproportionality analysis of musculoskeletal PTs associated with different CDK4/6 inhibitors

The disproportionality analysis of musculoskeletal PTs (top 30 PTs) associated with different CDK4/6 inhibitors (Palbociclib, Ribociclib, and Abemaciclib), using the FAERS database from 2015Q1 to 2023Q4, revealed distinct differences exist among three CDK4/6 inhibitors (Table 2). Arthralgia showed significant RORs for Palbociclib (ROR 1.52, 95% CI 1.45–1.60) and Abemaciclib (ROR 1.59, 95% CI 1.26-2.00) but not for Ribociclib. Back pain had a significant signal for all three inhibitors, with the highest ROR for Ribociclib (ROR 2.12, 95% CI 1.93–2.34). Bone pain showed significant RORs across all inhibitors, notably higher for Ribociclib (ROR 5.14, 95% CI 4.55–5.81). Muscle spasms had a significant signal for Abemaciclib (ROR 2.12, 95% CI 1.57–2.86), while myalgia did not show significant signals across the inhibitors, especially for Ribociclib with a ROR of 0.59 (0.48, 0.72). It is lower than for other two CDK4/6 inhibitors and consistent with a reduced rate of myalgia for Ribociclib [9]. Significant signal for bone lesions was observed, particularly with high ROR for Ribociclib (ROR 28.08, 95% CI 22.55–34.97), highlighting a strong association. Pathological fractures also showed a high ROR for Abemaciclib (ROR 6.25, 95% CI 2.33–16.74). The ROR radar chart (Fig. 3) underscores these differences, with Ribociclib showing prominent signals for bone pain and bone lesions, and Abemaciclib for osteonecrosis of the jaw and pathological fractures. Palbociclib demonstrated a consistent but less pronounced signal across these events. These results highlight specific MSAEs with varying levels of association across the three CDK4/6 inhibitors.

Table 2 Signal detection for CDK4/6 inhibitors-associated musculoskeletal toxicities (top 30 PTs)

Fig. 3

Differences in reporting of 13 representative musculoskeletal adverse events among three different CDK4/6 inhibitors as a radar chart. The signal was normalized to the ROR of bone lesions for Ribociclib as 100%. The highest frequency top-13 PTs with positive signals were selected as the representative musculoskeletal adverse events for the radar chart

Analyses of time-to-onset profiles

The time-to-onset analysis for MSAEs associated with CDK4/6 inhibitors (Palbociclib, Ribociclib, Abemaciclib) revealed distinct temporal patterns through the Kaplan-Meier curves (Fig. 4). The median time-to-onset for Palbociclib was 82 days (IQR, 14–311), for Ribociclib was 34 days (IQR, 8-177), and for Abemaciclib was 32.5 days (IQR, 12–119). The cumulative event curves indicate that Palbociclib has a longer onset of MSAEs compared to Abemaciclib and Ribociclib. Statistical analysis using the nonparametric Kruskal-Wallis test yielded a highly significant difference among the three inhibitors (P-value = 3.048e-11), suggesting that the time-to-onset of MSAEs varies significantly among these drugs.

Fig. 4

The onset time of musculoskeletal adverse events after treatment with different CDK4/6 inhibitors. Statistical tests were conducted using the nonparametric Kruskal-Wallis test. The curve was generated by the R (version 4.3.3) with the ggplot2 package

Discussion

This real-world pharmacovigilance study, based on over 10,000 cases from FAERS, is the largest evaluation of musculoskeletal toxicities with CDK4/6 inhibitor therapy to date. The results suggest a strong association between this drug class and musculoskeletal adverse events, supporting the safety signals observed in clinical trials. In this large pharmacovigilance study, we comprehensively characterized the diverse musculoskeletal toxicity profile associated with CDK4/6 inhibitors. Our findings suggest that in addition to commonly reported musculoskeletal events like arthralgia and bone pain, CDK4/6 inhibitors may increase the risk of more serious toxicities such as osteonecrosis of the jaw, arthritis, and spinal pain. These events often necessitated drug discontinuation affecting approximately 1/3 patients receiving CDK4/6 inhibitors, which highlights their potential clinical impact [10].

The MSAEs we identified are consistent with those reported in the pivotal trials of CDK4/6 inhibitors, providing reassurance on the validity of our findings. For example, arthralgia rates of up to 30.8% were seen with Palbociclib in the PALOMA-2 trial [20], while the MONALEESA trials documented arthralgia in up to 40.7% of patients receiving Ribociclib plus letrozole [21]. Muscle spasms and musculoskeletal pain were also commonly reported adverse events across these studies with the incidence of 10.1% and 7.2–12.5%, respectively. Our results extend the clinical trial data by suggesting that more serious events like osteonecrosis of the jaw, arthritis, and spinal pain may be part of the musculoskeletal toxicity spectrum. Notably, apart from the well-recognized musculoskeletal AEs, this analysis uncovered several novel safety signals like pathological fractures, osteonecrosis, bone lesion, and musculoskeletal chest pain across different CDK4/6 inhibitors. While rare, such events can have serious consequences impacting patients’ quality of life and treatment compliance. Clinicians should remain vigilant for these toxicities when prescribing CDK4/6 inhibitor regimens.

The median time-to-onset of musculoskeletal AEs was around 4–9 weeks, which aligns with previous observations that many events manifest within the first few months of treatment initiation [10]. Palbociclib had a relatively longer median onset compared to Ribociclib and Abemaciclib, potentially related to differences in pharmacokinetics or mechanism of action. For example, Ribociclib possesses half-lives of more than 24 h and is administered on a daily basis. In contrast, Abemaciclib has a relatively shorter half-life and is dosed twice a day. Given their myelosuppressive properties, Abemaciclib is administered continuously without any break and is linked to a higher incidence and more severe bone marrow suppression when compared to Ribociclib [22]. Nonetheless, vigilant monitoring in the early treatment phase is prudent regardless of the specific CDK4/6 inhibitor used. The overall 6.38% fatality proportion associated with musculoskeletal toxicities, while low, is not negligible. Ribociclib exhibited the highest fatality rate (8.52%) based on the results from the FAERS database, which could be attributed to its enhanced potency and continuous dosing schedule compared to Palbociclib and Abemaciclib. However, most fatal cases likely resulted from disease progression rather than the musculoskeletal AE per se. Larger studies are needed to definitively establish any causal relationships.

The pathophysiologic mechanisms underlying the musculoskeletal toxicities of CDK4/6 inhibitors are still being elucidated. Preclinical studies indicate that CDK4/6 plays an important role in muscle development, regeneration, and homeostasis [23]. Inhibition of these kinases may lead to muscular dysfunction and wasting. Other hypotheses include mitochondrial toxicity, alterations in calcium homeostasis, and inflammatory effects [24, 25]. The inhibition of CDK4/6 may disrupt the balance between pro-inflammatory and anti-inflammatory signaling pathways, leading to an exaggerated inflammatory response that manifests as musculoskeletal symptoms [26]. Further research into the biological basis is needed to develop preventive strategies and interventions. Our study raises the possibility that there may be differences in musculoskeletal toxicity risks among the individual CDK4/6 inhibitors. Abemaciclib appeared to have higher reporting odds for several events like osteonecrosis of jaw and pathological fracture compared to Palbociclib and Ribociclib. Significant signal for bone lesions was observed, particularly with high ROR for Ribociclib. Conversely, Palbociclib had the relatively lower reporting odds for 13 representative musculoskeletal toxicities versus the other agents as demonstrating in the radar chart (Fig. 3). While these disproportionality signals require prospective validation, they suggest that varying selectivity profiles and pharmacologic/pharmacokinetics properties among the three CDK4/6 inhibitors could translate into distinct musculoskeletal adverse event patterns. Such differences have been noted for other adverse effects, including hematologic events like neutropenia and respiratory events like respiratory injury [27].

From a clinical perspective, our findings underscore the importance of proactive monitoring and management of musculoskeletal complaints in patients prescribed CDK4/6 inhibitors. Arthralgia and myalgia were among the most common events and often occurred weeks to months after treatment initiation. Therefore, patient education on these potential delayed effects is crucial. Regular symptom assessment using standardized tools may allow earlier intervention before symptoms escalate and impact function or quality of life. Established guidelines recommend conservative management strategies like anti-inflammatory medications, muscle relaxants, physical therapy, and optimizing calcium/vitamin D intake [28]. However, our data suggest that more serious musculoskeletal toxicities like osteonecrosis of the jaw, arthritis, and spinal pain can develop, which may warrant the earlier involvement of rheumatologists as well as other specialists like neurologists, oral and maxillofacial surgeons experienced in managing these conditions.

However, some limitations are inherent to spontaneous reporting databases like the FAERS. These include potential under-reporting, lack of detailed clinical data, inability to confirm causality, and absence of medication adherence or dosing information. Furthermore, confounding factors like concomitant medications, comorbidities, and cancer types could influence the risk of musculoskeletal events. For example, the combination of CDK4/6 inhibitors with anti-hormonal therapies such as aromatase inhibitors (AIs) or fulvestrant introduces an additional layer of complexity in interpreting the mechanism of bone lesions. AIs are well-known to be associated with MSAEs, including bone loss and joint pain, due to estrogen depletion, which leads to reduced bone density and increased fracture risk. Furthermore, tamoxifen has a mixed estrogen agonist/antagonist effect depending on the target tissue, which may also influence bone health. Additionally, lack of denominators and underreporting in the FAERS database prevent incidence calculations. Despite these limitations, pharmacovigilance analysis of FAERS provides a real-world complementary perspective on the musculoskeletal toxicity profile of CDK4/6 inhibitors beyond data from highly-controlled registration trials. The large sample size, ability to evaluate individual agents, and capture of serious/rare events represent key strengths. Our findings can help raise awareness among clinicians and support the need for prospective studies to validate potential risk factors and optimal preventive/treatment algorithms.

Conclusion

This comprehensive pharmacovigilance analysis highlights musculoskeletal toxicity as an important safety consideration with CDK4/6 inhibitors in real-world clinical practice. A wide range of MSAEs, including some novel and potentially serious toxicities, were associated with different CDK4/6 inhibitors. While most events occur within a few months, their early identification and proper management are crucial to prevent treatment discontinuation and preserve quality of life. Ongoing surveillance, coupled with translational research exploring the underlying mechanisms, could guide strategies to mitigate the musculoskeletal risks of this promising drug class. Clinicians should carefully weigh the potential benefits and risks when selecting CDK4/6 inhibitor regimens for individual patients.