Abstract
Background/Aim: Eribulin is an effective chemotherapeutic agent for the treatment of metastatic breast cancer and advanced or metastatic soft-tissue sarcomas. However, severe adverse effects (SAEs) occur in 30-40% of the patients, and significantly reduce the patients’ quality of life and disturb the recommended treatment schedules. Neutropenia is the main cause of treatment suspension, delay, and/or dose reductions, also leading to relative dose intensity reduction. This study aimed to examine the risk factors for SAE occurrence after eribulin treatment. Patients and Methods: Eighty patients with metastatic breast cancer or advanced or metastatic soft tissue sarcoma who received eribulin were retrospectively evaluated. Risk factors for SAE occurrence in the first cycle were primarily assessed. In addition, factors associated with SAE occurrence during all treatment cycles were evaluated. Results: SAEs in the first cycle occurred in 45% of patients. The primary SAE was neutropenia (91.7%). The incidence of SAEs during all treatment cycles was 61.3%. Multivariate analyses suggested that lower baseline neutrophil and hemoglobin levels were risk factors for SAE occurrence and severe neutropenia incidence in the first cycle. An independent factor associated with SAE occurrence during all cycles was age ≥65 years and a tendency was confirmed for baseline anemia. Conclusion: Baseline neutropenia and anemia were risk factors for SAE occurrence during the first eribulin treatment cycle. Age ≥65 years was also associated with SAE occurrence during all treatment cycles. Patients with these risk factors should be carefully monitored for assessment and prophylaxis.
Eribulin mesylate is a non-taxane inhibitor of microtubule dynamics and is classified as a halichondrin class of antineoplastic drugs (1). It has a different mode of action from other tubulin-targeting drugs, such as taxanes and vinca-alkaloids, and is considered effective in patients whose disease has progressed despite these treatments (2). In the EMBRACE trial, eribulin produced a significant clinically meaningful improvement in overall survival compared to physicians’ choice of treatment in metastatic breast cancer patients with heavy chemotherapeutic pretreatment, including anthracyclines and taxanes (3). Eribulin has also been suggested to have clinical benefits for overall survival in patients with advanced or metastatic soft tissue sarcoma (4, 5).
Neutropenia and chemotherapy-induced peripheral neuropathy (CIPN) are the main adverse effects caused by eribulin administration (3-6). In general, eribulin-induced CIPN is milder than taxane-based treatment (6) and occurs with cumulative dosing. However, attention is prudent as eribulin is generally administered after taxanes. Neutropenia appears during the early stages of treatment. A meta-analysis suggested that eribulin treatment has a higher incidence of neutropenia than non-eribulin regimens (7). Eribulin administration was suspended in 30%-40% of patients owing to neutropenia in previous studies (4-6), with dose-limiting toxicity described (8).
Serum albumin level is the most clinically important covariate predictive of inter-individual variability in the frequency of eribulin-induced neutropenia (9). In another report, creatinine clearance <70 ml/min was identified as a predictive factor in addition to a serum albumin level <3.9 mg/dl for severe neutropenia frequency in metastatic breast cancer (10). In contrast, Kobayashi et al. identified pretreatment neutrophil count and skeletal muscle gauge as risk factors for grade 4 neutropenia in 23 patients with advanced soft tissue sarcoma (11). However, smoking status or baseline platelet and hemoglobin levels, which are suggested risk factors for severe neutropenia frequency (12-21), and baseline liver dysfunction were not evaluated in these studies, suggesting that the evaluation may have been insufficient. Severe adverse effects (SAEs) significantly reduce the patients’ quality of life and disturb the recommended treatment schedules with frequent hospital visits. Treatment suspension, delays, and/or dose reduction due to SAEs also induce relative dose intensity (RDI) reduction, which can decrease treatment efficacy. Consequently, assessment of the factors associated with SAE occurrence is important for providing safer and less onerous treatments with appropriate management.
In this study, we aimed to further examine the risk factors for SAE occurrence in eribulin treatment in breast cancer and soft tissue sarcoma patients.
Patients and Methods
Patients. Patients with metastatic breast cancer or advanced or metastatic soft tissue sarcoma who received eribulin between November 2011 and May 2021 at the Hokkaido University Hospital were enrolled in this retrospective study. All patients met the following baseline criteria: 1) age ≥20 years and 2) Eastern Cooperative Oncology Group performance status (ECOG-PS) of 0-2. Patients whose planned schedule differed from the original (± more than one day) in the first cycle, those who were not able to complete the first cycle of the treatment, and those without sufficient information from medical records were excluded.
For the primary endpoint of this study, the inclusion of 80 patients was necessary because the estimated SAE occurrence in the first cycle was approximately 40% from prior reports (4-6). We included approximately three to four covariates in the multivariate analysis.
The study was approved by the Ethical Review Board for Life Science and Medical Research of Hokkaido University Hospital (approval number: 021-0063) and performed in accordance with the Declaration of Helsinki. Due to the retrospective nature of the study, informed consent from the subjects was waived by the committee.
Treatment methods. Eribulin at a dose of 1.4 mg/m2 on days 1 and 8 was administered every three weeks. The dosage was reduced to 1.1 mg/m2 and 0.7 mg/m2 depending on adverse effect severity according to previous reports (4, 6). Trastuzumab (8 mg/kg at first administration, 6 mg/kg at subsequent administration) ± pertuzumab (840 mg at first administration, 420 mg at subsequent administration) was co-administered in cases of human epidermal growth factor receptor-2 (HER2) over-expressed breast cancer. Granisetron (3 mg) and dexamethasone (6.6 mg) were intravenously administered before chemotherapeutic agents according to the current Japanese antiemetic guidelines (22).
Evaluation of adverse effects. The required patient information was obtained from medical records. Adverse effects were assessed according to the Common Terminology Criteria for Adverse Events version 5.0. The primary endpoint was the detection of risk factors for SAE occurrence in the first cycle. The secondary endpoints were the assessment of the primary SAE in the first cycle and evaluation of the factors for SAE occurrence during all treatment cycles. We defined SAEs as grade 3 or higher hematotoxicities and grade 2 or higher non-hematotoxicities that induced further eribulin dose reduction and/or treatment suspension.
Statistical analysis. Univariate and multivariate analyses were performed using logistic analysis to identify independent risk factors related to SAE occurrence using the following covariates: sex, age, PS, clinical stage, number of distal metastases, treatment line, co-administration of HER2 inhibitors, body surface area (BSA), neutropenia, anemia, thrombocytopenia, liver dysfunction (grade 2 or higher aspartate aminotransferase, alanine aminotransferase, and total bilirubin elevation), renal dysfunction (creatinine clearance of less than 50 ml/min), hypoalbuminemia, C-reactive protein (CRP) elevation, and smoking history at baseline, and dose reduction from the initiation of chemotherapy, according to previous reports (9-11, 15). Variables that had potential associations, as suggested by univariate analysis (p<0.15), and hypoalbuminemia were considered while constructing the multivariable model.
All analyses were conducted using JMP statistical software (version 14.0; SAS Institute Inc.). Statistical significance was set at p<0.05.
Results
Patient characteristics. Eighty of the 86 patients were enrolled according to the eligibility criteria for this study (Figure 1). Baseline patient characteristics, including the frequency of neutrophils, hemoglobin, platelets, and serum albumin levels lower than the lower limit of normal (LLN), are shown in Table I. Seventy percent of the patients had metastatic breast cancer and 30% had advanced or metastatic soft tissue sarcoma. HER2 inhibitor co-administration was confirmed in 7.5% of patients. The proportions of patients with lower neutrophil, hemoglobin, and platelet levels at baseline were 11.3%, 57.5%, and 11.3%, respectively. Serum albumin levels lower than those in LLN, liver dysfunction, and renal dysfunction were evident in 61.3%, 11.3%, and 8.8% of the patients, respectively. The proportion of former or current smokers was 40.0%, and 18.8% of the patients received dose-reduced treatment from the beginning.
Study design.
Patient characteristics.
Frequency of SAE occurrence in eribulin treatment. The frequency of SAEs after eribulin treatment is shown in Table II. SAE occurrence in the first eribulin cycle was observed in 45.0% of patients, with treatment suspension on day 8 (32.5%) and delay in the second cycle (26.3%). The calculated RDI for the first cycle was 79.6%. The frequency of severe neutropenia in the first cycle accounted for 91.7% of all the SAEs (41.3% of the total patients). Other SAEs included thrombocytopenia, fatigue, and brain natriuretic peptide (BNP) elevation. In contrast, SAE occurrence during all treatment cycles was confirmed in 61.3% of patients, with an RDI of 77.3% (5 median treatment cycles, range=1-74).
Frequency and details of severe adverse effects (SAE) in eribulin treatment.
Univariate and multivariate analyses of risk factors related to SAE occurrence. Univariate and multivariate analyses suggested that neutropenia and anemia at baseline were risk factors for SAE occurrence during the first eribulin cycle (Table III). We also evaluated the factors for SAE occurrence during all treatment cycles and found that age ≥65 years was a significant factor, and patients with baseline anemia tended to experience SAEs, but this was not statistically significant (Table IV).
Univariate and multivariate analyses of the risk factors associated with severe adverse effects (SAE) occurrence in the first eribulin cycle.
Univariate and multivariate analyses of the risk factors associated with severe adverse effects (SAE) occurrence during all treatment cycles.
Discussion
Eribulin is an effective treatment strategy for metastatic breast cancer and advanced or metastatic soft tissue sarcoma (4-6). However, SAEs, especially severe neutropenia, occur in 30-40% of patients, leading to treatment suspension, delay, and/or dose reduction (4, 6). In addition, as most administrations are conducted in an outpatient setting, neutropenia poses the risk of infection and febrile neutropenia. Consequently, the evaluation of the risk factors for SAE occurrence is important for providing safe, less onerous, and effective chemotherapy. In this study, we first assessed factors associated with SAE occurrence during eribulin treatment.
The occurrence of SAEs in the first cycle was observed in 45% of patients, similar to previous reports (4-6). The primary SAE was severe neutropenia. Neutropenia and anemia at baseline were risk factors for SAE occurrence in the first cycle. Considering that the primary SAE was severe neutropenia, we additionally assessed factors related to its incidence, resulting in the same associated factors (data not shown). In addition, a trend was confirmed regarding the relationship between SAE occurrence during all treatment cycles and baseline anemia, although this was not statistically significant. Baseline neutropenia, which was considered to be caused by prior treatments, was suggested as a factor in a prior study evaluating eribulin (11) and could be a factor in drugs with strong hematotoxicity. In contrast, Suzuki et al. (10) suggested that this is not a factor. The authors set the cutoff baseline neutrophil count as 3,000/μl in their logistic analysis (10), which critically affected the results. Anemia has been reported to be a factor in severe neutropenia (16-19). The half-life of erythrocytes is approximately 120 days. In general, hemoglobin levels gradually decrease with each course of chemotherapy (23), reflecting longer-term hematotoxicity. As all patients in this study previously received hematotoxic treatment that included cytotoxic agents (taxanes) or cyclin-dependent kinases 4 and 6 inhibitors prior to eribulin administration, it is possible to speculate that bone marrow reserve reduction due to prior chemotherapy still affected the results. Furthermore, lower hemoglobin levels are negatively correlated with serum interleukin-6 (IL-6) levels, which is a proinflammatory factor that contributes to the worsening of neutropenia, although the mechanism underlying this correlation is unclear (16, 18, 24-26). We believe that baseline bone marrow reserve reduction and/or inflammation, which is not reflected by CRP elevation, might have affected the results. Further studies are required to elucidate this mechanism.
Age ≥65 years was a risk factor for SAE occurrence during all treatment cycles. The most common SAEs were neutropenia (75.5%), fatigue (10.2%), CIPN (6.1%), PS reduction (4.1%), and gastrointestinal symptoms (4.1%). Aging has been previously suggested as a risk factor for severe hematotoxicity and related complications, and CIPN (27-29). Hurria et al. (30) summarized the features of elderly patients. The authors described that, in general, hepatic mass and content of cytochrome P450 (CYP), which is the main metabolic enzyme, decrease in liver biopsies with increasing age, although the impact on liver function remains controversial. Furthermore, renal mass and renal blood flow decrease with age. Moreover, body fat increases and total body water decreases with age, increasing the volume distribution of lipid-soluble medicines such as eribulin (30). In addition, there is an increase in bone marrow fat and a decrease in bone marrow reserve with age, placing older adults at higher risk for chemotherapy-induced hematotoxicity, particularly after cumulative administration (31). Consequently, we consider that elderly patients have cumulative chemotherapy-induced damage due to physiological reduction, which might have affected the results in all treatment cycles in our data. Interestingly, the factors in the first cycle and in all cycles were different. We speculate that prior treatment affected baseline neutropenia and that cumulative damage was caused by age-related physiological reduction, resulting in the difference. Baseline anemia may be a good indicator of SAE occurrence following eribulin treatment.
In contrast, hypoalbuminemia was not detected. In a previous pooled data evaluation study, it was described that some studies adopted different administration methods from the usual eribulin regimen (9). We consider that the difference in dosage and administration, and patient population from the real world affected the results. In addition, liver dysfunction was not detected as a risk factor in this study. Eribulin is mainly excreted in its unchanged form, and its principal elimination route is biliary excretion (32), although it is partially metabolized by CYP3A4 (33). Devriese et al. reported higher area under the blood concentration-time curve in patients with hepatic impairment than in those with normal hepatic function (34). Therefore, patients with Child-Pugh class A and B were supposed to receive a reduced dosage (34). In addition, Macpherson et al. suggested that elevated bilirubin levels are associated with increased toxicity and a greater requirement for dose modifications in patients with metastatic breast cancer (35). There were no patients with abnormal bilirubin levels, and two out of nine patients with ≥ grade 2 liver dysfunction, who were evaluated as Child-Pugh A, received dose reduction according to the guidelines from treatment initiation in this study. We believe that these factors have significantly affected our results. Considering these results, a decreased dose should be administered to patients with elevated bilirubin or Child-Pugh A or B levels to prevent SAEs. Moreover, renal dysfunction, which was also a suggested risk factor for severe neutropenia in a previous study (10), was not detected in our analysis.
This study had some limitations. First, it was retrospective in nature and included a relatively small patient population from a single institution. Second, we did not evaluate the treatment efficacy. As baseline absolute lymphocyte count and neutrophil-to-lymphocyte ratio and dynamic changes in absolute lymphocyte count were significantly associated with survival benefit in eribulin treatment (36), evaluation of relationship between SAE occurrence and treatment efficacy will provide further comprehension. Third, CYP3A4 has been reported to have polymorphisms (37, 38), although it remains unclear whether it affects metabolism and the adverse effects of eribulin. As we did not assess the genetic background of such patients, further studies may provide new findings. Fourth, we did not evaluate proinflammatory cytokine levels such as IL-6, IL-2, and tumor necrosis factor-alpha.
In conclusion, baseline neutropenia and anemia were risk factors for SAE occurrence, particularly severe neutropenia, during the first eribulin treatment cycle. Age ≥65 years was also a factor for its occurrence during all treatment cycles. Therefore, adverse effects should be cautiously monitored in patients with these risk factors.
Footnotes
Authors’ Contributions
Designed study: Y.S.; Performed research: Y.S.; Analyzed data: Y.S.; Contributed new methods or models: Y.S.; Wrote the paper: Y.S.; All Authors read and approved the final manuscript.
Conflicts of Interest
The Authors have no relevant financial or non-financial interests to disclose in relation to this study.
- Received May 24, 2022.
- Revision received June 11, 2022.
- Accepted June 14, 2022.
- Copyright © 2022 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.
