Abstract
Background/Aim: Neutropenia and anemia are common and potentially serious adverse events associated with abemaciclib use in breast cancer, but their risk factors remain underexplored. This study aimed to identify the clinical predictors of severe neutropenia and anemia following the initiation of abemaciclib.
Patients and Methods: The medical records of 252 patients with breast cancer who received abemaciclib therapy between August 2011 and November 2024 were retrospectively reviewed. The risk factors for severe neutropenia and anemia were identified based on clinical characteristics and laboratory values. Moreover, the timing of dose reduction after the initiation of abemaciclib treatment was compared with progression-free survival and overall survival using log-rank analysis.
Results: Body mass index (BMI) <22.4 kg/m2 and pre-treatment neutrophil count <2,800×106 cells per liter were associated with an increased risk of grade ≥3 neutropenia [odds ratio (OR) (95% confidence interval, CI)=2.1 (1.1-4.0), p=0.03; OR (95%CI)=3.2 (1.5-6.7), p=0.003, respectively]. Also, BMI <22.4 kg/m2, hemoglobin level <8.9, and alanine aminotransferase (ALT) ≥60 U/l were significantly associated with an increased risk of grade ≥3 anemia [OR (95%CI)=5.0 (1.1-22.4), p=0.04; OR (95%CI)=1.6 (1.1-2.5), p=0.03; OR (95%CI)=9.6 (1.5-60.1), p=0.02, respectively].
Conclusion: BMI and neutrophil count before treatment were significantly associated with an increased risk of abemaciclib-induced grade ≥3 neutropenia. Moreover, BMI, ALT, and hemoglobin were significantly associated with grade ≥3 anemia following abemaciclib initiation. These factors should be considered in determining individual risk of neutropenia and anemia and risk-benefit ratio of abemaciclib.
Introduction
Breast cancer is diagnosed at an early stage in more than 90% of cases, and the mortality rate relative to incidence is approximately 15-20%, indicating a higher survival rate compared to other cancers (1, 2). However, recurrent breast cancer cases are often incurable, and patients who achieve surgical remission must complete postoperative pharmacotherapy to prevent recurrence (3). Furthermore, even in cases with metastasis, pharmacotherapy significantly impacts overall survival, with a 5-year survival rate of approximately 25% (4, 5). Hormone receptor-positive/human epidermal growth factor receptor 2-negative (HR+/HER2−) breast cancer accounts for nearly two-thirds of all breast cancer diagnoses (6, 7).
Abemaciclib is a novel, reversible cyclin-dependent kinase (CDK) 4/6 inhibitor used in the treatment of HR+/HER2 negative advanced breast cancer (7, 8). Current guidelines recommend the use of abemaciclib in combination with either a non-steroidal aromatase inhibitor or fulvestrant as a first-line therapy for patients with HR+/HER2 negative advanced breast cancer (9). Safety data from the MONARCH 2, 3, and E clinical trials have identified anemia and neutropenia as the primary hematological toxicities associated with abemaciclib use (10, 11). Neutropenia occurred in over 60% of patients receiving abemaciclib in the MONARCH 2 trial who had grade ≥3 adverse events. Among patients who experienced adverse events, 15.9% discontinued abemaciclib and 42.9% required dose reduction (12). Specifically, dose reduction due to neutropenia was implemented in 10% of the patients, while treatment interruption occurred in 16.3%. Similarly, in the MONARCH 3 trial, 41.3% of patients developed neutropenia, leading to dose reductions in 11.3% and treatment interruptions in 15.6% of the cases (12). In the MONARCH E trial, 45.8% of patients developed neutropenia, with 8.1% requiring dose reduction and 15.8% requiring treatment interruption (11). In addition to neutropenia, anemia also disrupts treatment schedules. The incidence of anemia was 29% in the MONARCH 2 trial (12), 28.1% in the MONARCH 3 trial (13), and 24.4% in the MONARCH E trial (11).
Existing literature (14, 15) provides insufficient data on the risk factors for anemia and neutropenia in patients initiating abemaciclib therapy. Identifying risk factors using routine clinicopathological data after initiating abemaciclib therapy could provide clinicians with insights into specific toxicity risks. Accurately identifying risk factors would also improve the monitoring of adverse events and implementation of proactive strategies for toxicity management, ultimately allowing patients to continue beneficial treatment for a longer duration (14). Therefore, this study aimed to determine the predictive risk factors for anemia and neutropenia in patients with breast cancer following the initiation of abemaciclib therapy.
Patients and Methods
Study design and participants. Patients with early breast cancer (EBC) and metastatic breast cancer (MBC) who received abemaciclib at the Hokkaido Cancer Center Breast Department, Japan, between August 2011 and November 2024 were enrolled in this retrospective study. The inclusion criteria were (i) age ≥20 years, (ii) Eastern Cooperative Oncology Group performance status 0-2, (iii) treatment duration of at least 6 months, and (iv) availability of sufficient medical records. The exclusion criteria were as follows: no blood tests within 2 weeks before the start of abemaciclib treatment, transfer to another institution after the start of treatment, and lack of continuous clinical data. Figure 1 shows the study flowchart. This study was approved by the Hokkaido Cancer Center Ethics Committee (Approval No. R3-25) and was conducted in accordance with the Declaration of Helsinki. The ethics committee waived the requirement for informed consent from participants, owing to the retrospective nature of this study.
Patient flowchart.
Dose adjustment. The treatment plan comprised administration of abemaciclib at 150 mg once daily, and the dose was reduced to 100 mg and 50 mg once daily depending on the severity of the side effects. Omission of abemaciclib and/or dose reduction by up to two steps (i.e., 150 mg to 100 mg to 50 mg) was allowed based on the nature (hematological or non-hematological), severity, persistence, and recurrence of toxicity.
Predictors and outcomes. Patient data were retrieved from the medical records. Adverse events were assessed according to the Common Terminology Criteria for Adverse Events, version 5.0. The primary endpoint was the detection of risk factors associated with severe anemia and neutropenia within 1 year. Severe hematological toxicity was defined as grade ≥3 anemia and grade ≥3 neutropenia. The pre-treatment variables were selected based on availability, prior evidence, and biological plausibility and included age, neutrophil count, hemoglobin, BMI, and albumin. For continuous variables, the cut-off values were calculated using receiver operating characteristic analysis. The significance threshold was set at p<0.05. To assess multicollinearity, we calculated the variance inflation factor (VIF) for each variable, and a VIF value >10 indicated significant multicollinearity. We then performed multivariate analysis using logistic regression to identify independent risk factors associated with severe neutropenia and anemia.
Survival analysis. In the time-to-event analysis, progression-free survival and overall survival were estimated using the Kaplan–Meier method. The log-rank test was applied to compare survival curves between the groups. Patients who required further dose reduction were classified based on whether the dose reduction occurred within 3 months of the start of treatment or after 3 months. Statistical analysis was performed using R version 4.2.2 (R Development Core Team, Vienna, Austria) (16). This study was conducted in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines (17).
Results
Study population. Of the 274 patients who met the inclusion criteria, 22 were excluded. The characteristics of the patients are shown in Table I. The median (interquartile range) age of the 252 patients was 58 years (50-69 years). Only patients with HER2-negative (2+/fluorescence in situ hybridization or less) and hormone receptor-positive (estrogen or progesterone-positive) tumors were eligible for treatment with abemaciclib.
Demographic and clinical characteristics of patients.
Prediction of grade ≥3 neutropenia. In the multivariate analysis, BMI (<22.4 kg/m2) and neutrophil count before treatment (<2,800×106 cells per liter) were significantly associated with an increased risk of grade ≥3 neutropenia following initiation of abemaciclib therapy [OR (95% confidence interval (CI)=2.1 (1.1-4.0), p=0.03; OR (95%CI)=3.2 (1.5-6.7), p=0.003, respectively]. No significant association was observed between grade ≥3 neutropenia and age, albumin, AST, or ALT (Table II).
Multivariable analysis for occurrence of grade ≥3 neutropenia during 365 days.
Prediction of grade ≥3 anemia. Multivariate analysis revealed that BMI (<22.4 kg/m2), hemoglobin level (<8.9), and ALT (≥60 U/l) were significantly associated with an increased risk of grade ≥3 anemia following the initiation of abemaciclib treatment [OR (95%CI)=5.0 (1.1-22.4), p=0.04; OR (95%CI)=1.6 (1.1-2.5), p=0.03; OR (95%CI)=9.6 (1.5-60.1), p=0.02, respectively]. No significant associations were found between grade ≥3 anemia and age, albumin level, or AST (Table III).
Multivariable analysis for occurrence of grade ≥3 anemia during 365 days.
Reasons for discontinuation of abemaciclib. Of the 252 patients, 190 discontinued abemaciclib during the follow-up period. The reasons for discontinuation were adverse events in 74 cases (29%) and disease progression in 88 cases (35%). The adverse events that led to treatment discontinuation included pneumonitis in 21 cases (8%), elevated liver enzymes in 13 cases (5%), and fatigue in six cases (2%) (Table IV).
Reasons for dose reduction and treatment discontinuation.
Reasons for dose reduction due to adverse events. Among the 252 patients, 152 (60.3%) experienced a one-step dose reduction and 49 (19.4%) experienced a two-step dose reduction due to adverse events. The top three adverse events leading to a one-step dose reduction were neutropenia in 53 cases, fatigue in 34 cases, and elevated liver enzymes in 15 cases. The top three adverse events leading to a two-step dose reduction were neutropenia in 15 cases, fatigue in nine cases, and anemia in six cases (Table IV).
Survival analysis results: dose reduction time cut-off (365 days). The median overall survival (mOS) for patients who experienced a dose reduction within 365 days was 60.4 months (95%CI=42.5 months – not reached), while the mOS for patients who received the full dose (no dose reduction) was not reached (95%CI=40.1 months – not reached). No significant difference was observed between the two groups (p=0.17) (Figure 2A). In addition, the median progression-free survival (mPFS) for patients who experienced a dose reduction within 365 days was 26.0 months (95%CI=18.8-33.0 months), while it was 9.8 months (95%CI=6.7-18.0 months) for patients who received the full dose (no dose reduction). A significant difference was observed between the two groups (p< 0.0001) (Figure 2B).
Kaplan–Meier survival analysis by dose reduction status (365-day cut-off). A) Overall survival (OS) curves comparing patients with abemaciclib dose reduction within 365 days and those without dose reduction. The bold line represents the no dose-reduction group; the regular line represents the dose-reduction group. B) Progression-free survival (PFS) curves for the same patient groups. The bold line represents the no dose-reduction group; the regular line represents the dose-reduction group.
Survival analysis results: dose reduction time cut-off (50 days). The patients were divided into three groups: the group that underwent dose reduction for ≥50 days, the group that underwent dose reduction for <50 days, and the group that did not undergo dose reduction (no dose reduction group). The mOS for patients who experienced a dose reduction for <50 days was 49.8 months (95%CI=39.0 months – 60.7 months), the mOS for patients who experienced a dose reduction for ≥50 days was not reached (95%CI=42.8 months – not reached), and the mOS for patients who received the full dose (no dose reduction) was not reached (95%CI=40.1 months – not reached). A significant difference was observed between the three groups (dose reduction for ≥50 days versus the other two groups, p=0.02) (Figure 3A).
Kaplan–Meier survival analysis by timing of dose reduction (50-day cut-off). A) Overall survival (OS) curves for three groups: patients with dose reduction ≥50 days after treatment initiation (dotted line), <50 days (regular line), and those with no dose reduction (bold line). B) Progression-free survival (PFS) curves for the same three groups. The bold line indicates no dose reduction; the regular line indicates dose reduction <50 days; the dotted line indicates dose reduction ≥50 days.
In addition, the mPFS was 20.7 months (95%CI=10.6-26.0 months) for patients who experienced a dose reduction for <50 days, 36.4 months (95%CI=20.7-60.0 months) for those who experienced a dose reduction for ≥50 days, and 9.8 months (95%CI=6.7-18.0 months) for patients who received the full dose (no dose reduction). A significant difference was observed between the three groups (dose reduction ≥50 days versus other the two groups, p<0.00001) (Figure 3B).
Survival analysis results: by reason for dose reduction (hematotoxicity or non-hematotoxicity). The patients were categorized into three groups: those who underwent dose reduction due to hematotoxicity after treatment initiation (hematotoxicity group), those who underwent dose reduction for non-hematotoxicity reasons after treatment initiation (non-hematotoxicity group), and those who did not undergo dose reduction (no dose reduction group). The mPFS and mOS were compared.
The mOS was 50.9 months (95%CI=41.6 months – not reached) for the hematotoxicity group, not reached (95%CI=40.1 months – not reached) for the non-hematotoxicity group, and not reached (95%CI=40.1 months – not reached) for the no dose reduction group (p= 0.31) (Figure 4A).
Kaplan–Meier survival analysis by reason for dose reduction. A) Overall survival (OS) curves comparing patients who underwent dose reduction due to hematotoxicity (regular line), non-hematotoxicity (dotted line), and those with no dose reduction (bold line). B) Progression-free survival (PFS) curves for the same groups. The bold line represents the no dose-reduction group; the regular line represents the hematotoxicity group; the dotted line represents the non-hematotoxicity group.
In addition, the mPFS was 25.3 months (95%CI=15.1-37.1 months) for the hematotoxicity group, 26.0 months (95%CI=13.6-34.8 months) for the non-hematotoxicity group, and 9.8 months (95%CI=6.7-17.9 months) for the no dose reduction group (p=0.0003) (Figure 4B).
Safety information. The most common hematological toxicity was neutropenia (22%), followed by anemia (18%) and hepatic dysfunction (18%). Among these, grade ≥3 neutropenia was particularly common (14%). The median cumulative incidences of neutropenia, anemia, and hepatic dysfunction were 25, 50, and 29 days, respectively.
Meanwhile, the most common non-hematological toxicities were diarrhea, nausea, and fatigue in 85%, 62%, and 51% of the patients, respectively. The incidence of grade ≥3 non-hematological toxicities tended to be lower than that of hematological toxicities, at less than 5%. The median cumulative incidences of diarrhea, nausea, and fatigue were 14, 15, and 29 days, respectively (Table V).
Safety information.
Discussion
This study investigated the pre-treatment risk factors for hematological toxicity that can lead to the interruption or dose reduction of abemaciclib in patients with breast cancer. We identified neutrophil count (<2,800×106 cells per liter) and BMI <24.3 kg/m2 as risk factors for grade ≥3 neutropenia and hemoglobin level (<8.9) and BMI <24.3 kg/m2 as risk factors for grade ≥3 anemia. This report identified a potential association between serious hematological toxicity and low BMI.
Neutropenia and anemia are common adverse events of CDK 4/6 inhibitors due to their effect on the hematopoietic bone marrow. Avelumab-induced grade ≥3 neutropenia is usually managed by dose reduction and interruption (14). Therefore, patients at high risk of developing grade ≥3 neutropenia should be identified before starting treatment to prevent occurrence of neutropenic sepsis (18). Consistent with our results, a previous report observed a significantly higher risk of grade ≥3 neutropenia in underweight patients (19). In other studies that compared clinical efficacy and treatment-related toxicity according to BMI using data from the MONARCH 2 and 3 trials, patients with low and/or normal body weight had a higher incidence of neutropenia and better overall response rates (19). One possible reason for this variation in the risk of neutropenia is differences in blood concentration of abemaciclib between obese and non- obese patients, since the dose of abemaciclib is standardized to the same dose and does not consider body weight or height differences. In the MONARCH 2 and 3 trials, the pharmacokinetics of the overall population and Japanese patients were compared, but no difference was reported between the two populations. However, this pharmacokinetic comparison only assessed abemaciclib itself and did not consider its active metabolites (20, 21). Abemaciclib is primarily metabolized into the active metabolites M2, M20, and M18, which have the same efficacy as the parent drug. Abemaciclib and its active metabolites are thought to be transported by P-glycoprotein (22, 23). Among the genes encoding P-glycoprotein, patients with the ABCB1 2677G>T/A homozygous genotype are reported to have increased concentrations of abemaciclib and its active metabolites, and are thus likely to discontinue treatment or reduce the dose (22, 23). Thus, studies investigating the pharmacokinetics of abemaciclib should measure the concentrations of both the parent drug and its active metabolites.
Low hemoglobin levels were identified as a risk factor for grade ≥3 anemia. Some studies have reported that low hemoglobin concentration before treatment impairs anti-proliferative activity due to restricted tumor oxygenation and reduces the efficacy of chemotherapy in breast cancer (24). However, it is unknown whether the same mechanism applies to molecular targeted therapies, such as abemaciclib.
In this study, 88 out of 252 patients with breast cancer (35%) discontinued abemaciclib treatment due to disease progression, followed by 74 patients (29%) who discontinued treatment due to adverse events. The main adverse events that led to discontinuation of abemaciclib were interstitial pneumonia in 21 patients (8%), increased liver enzymes in 13 patients (5%), fatigue in six patients (2%), gastrointestinal symptoms such as nausea and anorexia in five cases (2%), and skin rash in eight cases (3%). In the MONARCH 2 and MONARCH 3 studies, 70 out of 441 patients (15.9%) and 64 out of 328 patients (19.6%), respectively, discontinued treatment due to adverse events (12, 13). Discontinuation of treatment due to drug-induced lung disorders was observed in two of 441 patients (0.5%) in the MONARCH 2 study compared to 21 of 252 patients (8%) in our study (Asian patients). The incidence of drug-induced lung disorders in our study (8%) was still higher than the 2.7% among the Japanese sub-cohort of the MONARCH 2 study. Interstitial lung disease is a well-recognized, potentially serious complication of many cancer agents (25).
Additionally, regarding hepatic dysfunction (increase in ALT and/or AST), in our study (for Asians), treatment discontinuation was observed in 13 out of 252 cases (5%), whereas in the MONARCH 2 and 3 trials, treatment discontinuation due to hepatic dysfunction was observed in 0.5-2.1% of the patients. This discrepancy may be due to ethnic differences, as only approximately 30% of patients in the MONARCH 2 and 3 trials were Asian. Indeed, when considering only the Japanese population in the MONARCH 3 trial, the discontinuation rate due to hepatic dysfunction was 5.3-10.5%, which was similar to our results (20, 21). Further research is needed to investigate the influence of ethnicity on the occurrence of hepatic dysfunction following abemaciclib treatment for breast cancer.
With regard to dose reduction, 60.3% of patients in the present study had a one-level dose reduction, and 19.4% had a two-level dose reduction. The top three adverse events in patients with a one-level dose reduction were neutropenia, fatigue, and hepatic dysfunction, and the top three adverse events for patients who had a two-level dose reduction were neutropenia, fatigue, and anemia. In cases where the dose was reduced by two levels, 60% of patients with neutropenia also experienced one-level reduction for the same reason. In the MONARCH 2 and MONARCH 3 studies, 189 out of 441 patients (42.9%) and 142 out of 328 patients (43.4%), respectively, had dose reductions due to adverse events (12, 13). In the Japanese subcohort of the MONARCH 2 and 3 trials, dose reductions due to neutropenia occurred in 12.7-13.2% of patients, which was higher than the overall rate (10.0% to 12.8%) (19, 20) and lower than our observed rate of 21% (20, 21). The Japanese pharmacokinetics, safety, and health-related quality of life profiles were consistent across various studies (11-13, 20, 21, 26, 27). Furthermore, in the Japanese sub-cohort of the MONARCH 2 and 3, the proportion of grade ≥3 hematological events and hepatic dysfunction (ALT and/or AST increase) was higher than those in the overall population (11, 12, 27), confirming clinically important toxicities in Japanese patients treated with abemaciclib. Similarly, a higher frequency of hematological toxicities was previously observed in the Japanese sub-cohort of the global phase 3 study of another CDK4/6 inhibitor, palbociclib, in women with HR+/HER2− advanced breast cancer, with a higher incidence of grade ≥3 neutropenia in Japanese patients treated with palbociclib compared to the overall study population (28). Although the mechanism underlying this finding is unknown, it may indicate a drug-class effect. Despite the higher incidence of neutropenia in the Japanese sub-cohort of the MONARCH 2 study, neutropenia was not associated with an increase in severe infections or febrile illness, and only few patients discontinued treatment due to neutropenia.
The risk factors reported in the present study can guide clinicians in identifying patients at high risk of abemaciclib-induced neutropenia and anemia. Clinicians can then consider preventive strategies (e.g., prophylactic granulocyte colony-stimulating factor, dose reduction of abemaciclib, or more rigorous monitoring of white blood cell count, neutrophil count, and hemoglobin levels) to facilitate effective and safe long-term abemaciclib treatment without the need for dose reductions. In addition, predicting adverse events can minimize the need for continuous intervention by medical staff, which can help reduce patients’ anxiety about their treatment (29).
Study limitations. First, as this was a retrospective observational study, we could not adjust for confounding factors other than those included in the analysis. Second, this study was conducted in a single center and had a relatively small sample size, which limited the number of variables that could be analyzed in the multivariate analysis. Third, this study only included Japanese patients. As mentioned earlier, racial differences may affect the risk of adverse events. In the future, it will be necessary to accumulate more cases and conduct more detailed studies that consider factors such as genetic mutations and racial differences. The MONARCH 2 and 3 trials showed progression-free survival with abemaciclib, regardless of dose reduction or early onset of diarrhea and neutropenia (11). Therefore, if a side effect that is difficult to manage occurs despite appropriate management, treatment with abemaciclib can be continued, even if early dose reduction is necessary, rather than risk having to stop abemaciclib due to worsening of the side effect. By identifying the risk factors for neutropenia and anemia, the occurrence of adverse events can be monitored for early detection and appropriate management.
Conclusion
This study identified pre-treatment risk factors for hematological toxicity associated with abemaciclib in patients with breast cancer, including low neutrophil count and BMI (<24.3 kg/m2) for grade ≥3 neutropenia and low hemoglobin level and BMI for grade ≥3 anemia. Our findings highlight the potential link between low BMI and severe hematological toxicity.
Given the high incidence of neutropenia and anemia observed in Japanese patients, clinicians should assess these risk factors before initiating treatment. Early identification of high-risk patients may help implement preventive strategies such as dose adjustments, proactive monitoring, or supportive care to reduce toxicity while maintaining therapeutic efficacy.
Despite the study’s limitations, including its retrospective design, single-center setting, and limited sample size, our results suggest that individualized management based on patient characteristics may improve the safety and tolerability of abemaciclib treatment. Further research is needed to explore the impact of genetic and racial factors on treatment outcomes.
Acknowledgements
The Authors thank all patients who participated in this study and their families, as well as the staff of the pharmacy and Breast Surgery of the Hokkaido Cancer Center, Sapporo, Japan. The Authors would like to thank Editage (www.editage.com) for English language editing.
Footnotes
Authors’ Contributions
ST, KU, TK, MY, NT, KW, and HH confirmed the medical assessment and designed this study. KU and TK provided advice regarding statistical analyses. ST and KU performed the statistical analyses. ST, KW, and HH edited the manuscript. All Authors have discussed the results and commented on the manuscript.
Conflicts of Interest
The Authors declare no conflicts of interest in association with the present study.
Funding
No funding to declare.
- Received March 24, 2025.
- Revision received April 7, 2025.
- Accepted April 8, 2025.
- Copyright © 2025 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.










