Abstract
Background/Aim: Abemaciclib is a cyclin-dependent kinase 4/6 inhibitor approved in combination with endocrine therapy for treating hormone receptor-positive and human epidermal growth factor receptor 2-negative early and advanced breast cancer patients. The safety profile of abemaciclib is characterized by frequent gastrointestinal toxicity, especially diarrhea. Therefore, we performed an exploratory analysis of clinical factors that may be potentially associated with diarrhea in patients treated with abemaciclib plus endocrine therapy. Patients and Methods: Factors potentially predisposing to diarrhea were selected, such as age ≥70 years, concomitant medications and diseases, diet, and use of laxatives. These variables were correlated with the onset of grade 2/3 diarrhea in a cohort of patients treated with abemaciclib from advanced breast cancer. Univariate and multivariate analysis was performed. Sensitivity and specificity were tested using the ROC curve. Results: Eighty women with advanced breast cancer were included in the study. The univariate analysis found a statistically significant correlation between grade 2/3 diarrhea and age ≥70 years, polypharmacy, and concomitant gastrointestinal diseases (p<0.05). In the multivariate analysis, the number of risk factors significantly correlated with the outcome of interest (p<0.0001). ROC analysis showed our model’s 82% sensitivity and 75% specificity. Conclusion: Taking into account specific pre-existing factors, it is possible to estimate the risk of diarrhea in hormone receptor-positive and human epidermal growth factor receptor 2-negative – advanced breast cancer patients, candidates for abemaciclib plus endocrine therapy. In these subjects, implementing proactive prevention and adopting a dose-escalation strategy may represent practical approaches to decrease the abemaciclib toxicity burden.
The cyclin-dependent kinase 4/6 inhibitors (CDK4/6i) abemaciclib, palbociclib, and ribociclib represent, in combination with endocrine therapy (ET), the standard treatment for patients with hormone receptor-positive (HR+) and human epidermal growth factor receptor 2-negative (HER2−) advanced or metastatic breast cancer (ABC) patients (1-4). After its initial approval as monotherapy for HR+/HER-ABC pretreated patients, abemaciclib entered the clinical practice combined with fulvestrant for the treatment of endocrine-resistant patients and in combination with a non-steroidal aromatase inhibitor (AI) for endocrine-sensitive women (5). Abemaciclib has been employed also in patients pretreated with palbociclib plus ET with good clinical results suggesting a possible lack of complete cross resistance between CDK4/6i especially in patients who had a significant response to first line palbociclib (6).
Abemaciclib is administered continuously by mouth at the starting dose of 150 mg bis in die (BID) (5). It is associated with a manageable safety profile, mainly characterized by diarrhea, while hematological toxicities are less pronounced (5). Palbociclib and ribociclib may also induce this adverse event (AE), but to a lesser extent and severity than abemaciclib (7, 8).
In the phase II MONARCH-1 trial, single-agent abemaciclib at the dose of 200 mg BID caused any-grade diarrhea in 90.2% of patients, with 19.7% grade 3 events according to NCCN-CTC (9). Diarrhea was the most frequent toxicity and showed an early onset - about one week after treatment initiation - but a short median duration of grade 2/3 events (7.5 and 4.5 days, respectively). Therefore, dose reduction was required in 20.5% of cases (9). In the phase III MONARCH-2 (abemaciclib versus placebo plus fulvestrant) and MONARCH-3 (abemaciclib versus placebo plus AI) trials, the most common any-grade abemaciclib-related AE was diarrhea (82.3% and 86.4%, respectively). Diarrhea was the second most frequent grade 3 toxicity (13.4% and 9.3%, respectively) after neutropenia (10, 11). In these trials, most patients experienced diarrhea after 6-8 days from treatment initiation, with a median duration of 6 days. The incidence of diarrhea tended to decrease after the fourth cycle of therapy. Dose interruptions were generally short and represented 1.7-3.8% of the total treatment duration (12). The incidence and severity of diarrhea determined dose reductions in up to 20% of patients. In the same trials, quality of life (QoL) evaluation favored the abemaciclib arm for any symptom except diarrhea, which was significantly better in the control arm (13). In the subsequent multinational phase III MONARCH-plus study, patients received abemaciclib plus an AI or fulvestrant, depending on the disease setting (14). Diarrhea occurred in 78-80% of cases but was generally mild, with grade 3 events only in 2-4% of cases. The lower occurrence of severe diarrhea in this trial can result from better prevention and treatment of the side effect. A study of abemaciclib plus tamoxifen reported similar results (15). Real-world evidence is in line with this observation, reporting an even lower incidence of any-grade diarrhea (43%-67%) with less than 10% severe events (16, 17).
In clinical practice, the management of diarrhea in patients receiving abemaciclib is well established. Therefore, the loperamide assumption is proactively recommended at the first sign of loose or watery stools. At the same time, dose interruption or reductions are indicated for grade ≥2 events according to their persistence or recurrence (12). Although abemaciclib-induced diarrhea is mild to moderate and has a rapid resolution in most cases, its occurrence can cause poor patient adherence, treatment delays, and dose omissions, reduce QoL, and eventually represent a barrier to prescription. In this context, identifying patients predisposed to this AE may be critical in the real-world setting. This issue is crucial since the combination of abemaciclib, and an AI to date represents the best adjuvant treatment for surgically excised, early breast cancer at high risk of recurrence, as shown by results of the MONARCH-e trial (18).
We conducted a prospective observational analysis on a cohort of HR+/HER2− ABC patients treated with abemaciclib to identify clinical factors potentially linked with diarrhea.
Patients and Methods
Study population and design. This multicenter prospective observational study was carried out in six oncology centers in Italy, including two academic hospitals, two private comprehensive cancer centers, one National Institute for cancer research, and one large peripheral hospital. Patients with HR+/HER2− ABC treated with abemaciclib plus an AI or fulvestrant were included in this analysis. Investigators entered in an electronic database patients’ age, medical history, physical examination, blood counts, serum chemistry tests, presence of previous or concurrent diseases, gastrointestinal illnesses, concomitant number and type medications, and dietary habits. Patients with incomplete toxicity or response evaluation, inadequate clinical data, or lost to follow-up were excluded from the analysis.
Cancer treatment. All patients were treated with abemaciclib at 150 mg BID as starting dose, according to the Italian label indication for the drug. Endocrine therapy consisted of letrozole 2.5 mg or anastrozole 1 mg once daily continuously or fulvestrant 500 mg intramuscular injection every two weeks for the first cycle, a loading dose, and then monthly. Premenopausal patients received luteinizing hormone-releasing hormone (LHRH) analog to induce menopause. Before prescribing abemaciclib, treating physicians evaluated the possible drug-drug interactions with a web-based drug checker (19). According to the best practice prescription flowchart, treating oncologists provided patients and their caregivers with detailed written instructions on a diet to be followed. Physicians also recommended maintaining adequate hydration, immediate use of loperamide from the first episode of diarrhea, and reporting side effects immediately to the cancer center. Dose interruption, reduction, or treatment discontinuation of abemaciclib were managed as per label indication (Table I) (20).
Grading and management of diarrhea.
Definition of events and toxicity. Investigators graded toxicity according to the National Cancer Institute Common Terminology Criteria for Adverse Events (NCI CTCAE) v5.0, as shown in Table I (20). Patients were monitored according to clinical practice, with the assessment of complete blood counts and liver function (transaminases, serum bilirubin) at baseline, every two weeks for the first two months, and then monthly. To limit patients’ access to the hospital during the COVID-19 pandemic, the first episode of diarrhea was referred to the treating physician via web-based tools, such as email or instant messaging. In the case of persistent diarrhea, the oncologist required a face-to-face visit to perform a clinical examination, provide recommendation reinforcement, and request additional tests, if needed. Hospitalization was required in case of grade >3 diarrhea.
Based on medical literature and personal experience, VG, DP, and PV selected a list of parameters potentially correlated with the risk of diarrhea (21-24). The included risk factors were: age (≥70 years), concomitant medications and their number, type of concomitant medication, diabetes, fiber-rich diet, chronic use of laxatives, history of gastrointestinal diseases potentially predisposing to diarrhea, such as colitis, meteorism, irritable bowel, food intolerance, inflammatory bowel diseases, pancreatitis, previous pelvic radiotherapy, previous abdominal surgery. In addition, objective responses and survival parameters were defined as previously reported (25, 26).
Statistical analysis. Descriptive data were reported as absolute numbers and percentages rounded at the nearest unit with a 95% confidence interval (95%CI). An a-priori sample size calculator for multiple regression comprised four predictors of 4, not including the constant regression (27, 28). Therefore, an anticipated Cohen’s f2 effect size for an F-test of 0.15, with a power level b of 80% and a type I error rate of 0.80, required a sample size of at least 65 patients. Possible variables were evaluated at univariate analysis to reduce the sample size, and the number of predictors was minimized at multivariate analysis. Receiver operating characteristic (ROC) curve analysis was employed to assess predictivity for the occurrence of grade 2-3 diarrhea and its sensitivity and specificity rates. Pearson’s univariate correlation test, multivariate analysis, and received operator characteristics curve were carried out employing the GraphPad statistical package.
Ethical consideration. This study was performed in compliance with the Declaration of Helsinki. The Ethics Committee Palermo 1, University of Palermo, Italy, approved the study on July 17, 2019, protocol number 07/2019. Furthermore, the Ethics Committee waived informed consent collection, given the non-interventional nature of the research and the complete anonymization of the data.
Results
Patients’ population. Eighty patients were enrolled in the study (from July 2019 to June 2022). Population characteristics are reported in Table II. Thirty-five percent of patients were aged ≥70 years, 91% had metastatic disease, and 9% had locally advanced disease unsuitable for radical treatment. All but four patients had a performance status of <1, according to the Eastern Cooperative Oncology Group (ECOG). Most subjects (86%) had visceral metastases, while 11% had bone-only disease. Table III displays patients’ current or previous comorbidities and their related treatments. Cardiovascular diseases (e.g., hypertension) were the most common comorbidities (51.2%), followed by gastrointestinal conditions (47.5%). Consistently, drugs for cardiovascular disorders were frequent co-medications in our population, including antihypertensive (43.7%), antilipidemic (34.7%), and diuretics (17.5%). However, supplements (vitamin D, iron, and magnesium) were the most common co-medications taken by 91% of the patients included in the analysis.
Patient demographics and clinical characteristics.
Patients’ concomitant or previous diseases and concomitant medications.
Toxicity. Diarrhea was the most frequent toxicity, with 30 patients (37%) reporting grade 1 and 25 (31%) reporting G2/3 events, respectively. Diarrhea mainly occurred during the first treatment cycle with a median onset time of 7 days (range=5-11 days). The median duration of grade 2/3 diarrhea was eight days (range=4-10 days). In most cases, this AE was effectively managed with antidiarrheal medications and dose adjustments. Grade 2/3 diarrhea determined abemaciclib dose reduction to 100 mg BID in 23 cases (29%). No patient needed intravenous hydration or further de-escalation to 50 mg BID.
Incidence of grade 2/3 diarrhea was positively correlated with age (<70 or ≥70 years) in the univariate analysis (R2=0.1552; two-tailed p=0.0003). Although polypharmacy per sè was not related to the selected outcome, the number of different concomitant therapies significantly correlated with it (R2=0.1332; p two-tailed=0.0009). A positive correlation was also observed with previous intestinal diseases (R2=0.07878; p two-tailed=0.0117) but not with a fiber-rich diet, diabetes, or chronic use of laxatives. Additionally, the correlation between G2/3 diarrhea and the presence of one, two, or three risk factors was statistically significant (R2 the coefficient of determination=0.2687, p<0.0001). Figure 1 shows the receiver operating characteristic (ROC) curve analysis. The area under the ROC curve was 0.8485 (95% confidence interval=0.7873-0.9097; p<0.0001). The sensitivity and specificity of the risk score for grade 2/3 diarrhea were respectively 82.5% (95%CI=72.74%-89.28%) and 73.75% (95%CI=63.18-82.14%).
Receiver operatic curve (ROC) of diarrhea G2/3 versus the number of risk factors.
Clinical outcome. Table IV depicts the type and rates of objective response for the whole series and according to ET therapy (letrozole or fulvestrant), grade 2-3 diarrhea occurrence, and any reason for dose reductions of abemaciclib. In the whole population, the overall response rate (ORR) was 50%, with 3 (4%) complete responses (CR.), 37 (46%) partial responses (PR), 21 (26%) stable diseases (SD), and 19 patients (24%) progressive diseases (PD). In the 31 patients with grade 2/3 diarrhea, an objective response was observed in 15 cases (48%), with tumor stabilization in 8 cases (26%). Patients treated with abemaciclib plus AI or fulvestrant had 57% and 46% ORR, respectively. Among 23 patients who required abemaciclib dose reduction, 11 (48%) achieved PR and 7 (30%) SD. The differences between these cohorts were not statistically significant. Median progression-free survival was 9.8+ months and 14.6+ months for endocrine-sensitive and endocrine-resistant patients, respectively.
Objective response according to endocrine therapy and dose reduction.
Discussion
The occurrence of loose, watery, and frequent stools is the most common side effect of abemaciclib (10, 11, 13, 29, 30). The absolute risk for any grade diarrhea is far higher for abemaciclib [0.853 (95%CI=0.809-0.888, p<0.0001)] than for palbociclib [0.144 (95% CI=0.103-0.197, p<0.0001)], or ribociclib [0.258 (95%CI=0.181-0.355, p<0.0001)] (31).
The mechanisms underlying this AE still need to be fully elucidated (32-34). Besides CDK4/6 inhibition, abemaciclib also affects CDK9, which plays an essential role in intestinal cell proliferation. In rats, abemaciclib induced morphologic changes in the gut, such as crypt cell proliferation, loss of goblet cells, enterocyte degeneration, and mucosal inflammation (35). These effects can be related to abemaciclib inhibitory activity on CDK9, which plays an essential role in intestinal cell proliferation (35). In experimental preclinical models, CDK4/6 inhibition had a negligible antiproliferative effect on intestinal cells, while CDK9 and CDK1 knockdown significantly impacted intestinal cell-cycle progression. Abemaciclib also exerts inhibitory activity on glycogen synthase kinase-3 beta (GSK3β), part of a protein complex that phosphorylates β-catenin, preventing translocation into the nucleus. GSK3β also complexes with other transcription factors (TCFs) to form transcriptional activators of multiple genes, including MYC, CCND1, and AXIN2. The β-catenin/TCF-4 complex switches between proliferation and differentiation of intestinal epithelial cells, and inhibition of GSK3β induces Wnt pathway/β-catenin activation, favoring cellular proliferation (32-35). Moreover, abemaciclib profoundly inhibits Ca2+/calmodulin-dependent protein kinase CAMKII, which is involved in intestinal motility and can be linked to bowel movements (32-35).
Regardless of its pathogenesis, abemaciclib-induced diarrhea shows a typical temporal pattern, peaking during the first three cycles of treatment and then significantly decreasing as the result of adaption, antidiarrheal intervention, or dose reduction (31, 36, 37). In phase III trials, the occurrence of diarrhea caused abemaciclib dose reductions or omission in up to 15-19% of patients and withdrawal from therapy in up to 30% of subjects (10, 11, 13, 29, 30). Additionally, diarrhea was the only clinically meaningful symptom, more frequent in the abemaciclib than in the placebo arm, which affected patients’ QoL (36). In a study analyzing oncologists’ and patients’ preferences for dosing- and toxicity-related features of CDK4/6i, the risks of diarrhea and grade 3/4 neutropenia represented a crucial driver of choices (38). Overall, given its frequency and consequences, diarrhea may represent a potential barrier to abemaciclib prescription despite the proven activity of the drug.
In clinical practice, the influence of diarrhea on abemaciclib effectiveness is unclear. Indeed, diarrhea poorly influences abemaciclib absorption since its effect on the fraction of the dose entering the portal vein from the intestinal wall is minimal, with an estimated reduction of 6% (39). Furthermore, abemaciclib dose reductions do not significantly influence its clinical efficacy (37). A study of abemaciclib in the neoadjuvant setting, i.e., the neo-Monarch, explored the benefit of prophylactic antidiarrheal therapy with loperamide (40). Although diarrhea represented the most common abemaciclib-related AE, most patients had easily manageable low-grade diarrhea, and the incidence of grade 3 diarrhea fell to 4% with the prophylactic administration of loperamide. Interestingly the incidence of grade 3 diarrhea and treatment withdrawals in patients without loperamide prophylaxis was lower than that reported in the Monarch 2-3 trials, which did not include prophylactic loperamide. These data are probably linked to increased awareness and improved toxicity management. In addition, the Monarch-plus trial in China did not report grade 3 diarrhea in patients treated with abemaciclib plus ET, and either IA or fulvestrant, suggesting possible metabolic differences in the Asian population (14).
Our study investigated the potential risk factors associated with diarrhea incidence and severity in ABC patients treated with abemaciclib plus ET to improve the management of this AE. According to our results, age ≥70 years, the number of concomitant medications, and the presence of concomitant or previous intestinal diseases significantly correlate with grade 2/3 diarrhea, while a fiber-rich diet, diabetes, or chronic use of laxatives did not show any association. Furthermore, the ROC curve analysis showed a statistically significant correlation between grade 2/3 diarrhea and the presence of one, two, or three risk factors (R2 the coefficient of determination=0.2687, p<0.0001), with a sensitivity and specificity of the risk score of 82.5% (95% CI=72.74-89.28%) and 73.75% (95%CI=63.18-82.14%), respectively. These results suggest the possibility of estimating the risk of diarrhea in HR+/HER− ABC patients, candidates for abemaciclib-based treatment.
Older age, comorbidities, and polypharmacy are strongly intertwined factors. Nearly 25% of those over 60 suffer from two or more long-term medical conditions, and polypharmacy is increasing with an aging population (41). Moreover, 30-50% of medicines prescribed for chronic conditions are not taken as programmed. Therefore, inadequate evaluation of side effects may expose patients to the so-called prescribing cascade. This phenomenon occurs when a medicine-induced adverse event is misinterpreted as a new medical condition, and a subsequent drug is prescribed to treat the previous drug-induced event (41). Therefore, prescribers and health professionals must consider drug interaction before prescribing abemaciclib. Indeed, potent CYP3A inhibitors increase the exposure of abemaciclib and its active metabolites M2 and M20, predisposing patients to toxicity. In this context, polypharmacy can be a significant driver of intestinal toxicity. Still, the polypharmacy definition remains controversial. This issue may represent a limitation of our study since our numerical definition of polypharmacy does not account for specific comorbidities and makes it difficult to assess the safety and suitability of therapies (42).
A retrospective study explored risk factors linked to abemaciclib -induced liver injury. Hepatic steatosis and age >65 years showed a weak correlation with liver injury while concomitant use of an AI was an independent risk factor, suggesting careful evaluation of liver function in patients prior to prescription of abemaciclib plus ET (43).
In clinical practice, several approaches may be helpful to prevent or limit the occurrence of diarrhea, improve adherence, avoid abrupt patient-initiated drug withdrawal and improve abemaciclib compliance (44, 45). The National Community Oncology Dispensing Association recently reported an exhaustive guide for managing CDK4/6i-related diarrhea based on scientific evidence (46). Suggestions include increasing fluid intake, eating frequent and small meals, avoiding lactose, alcohol, sorbitol, spicy, fried, and fatty foods, and adopting the BRAT diet (banana, rice, apple puree, toast). Proactive education of patients and caregivers, i.e., recommending loperamide at the first sign of loose stool, is also pivotal for properly managing abemaciclib-induced diarrhea. Additionally, oncologists should provide detailed information about diet and adequate hydration.
Another strategy to prevent diarrhea may be dose escalation. The National Comprehensive Cancer Network and network guidelines recently included a dosing strategy in patients at risk of toxicity, including tailored dosing strategies (4). The step-up dosing approach has been successfully studied for some anticancer medications at high risk of inducing diarrhea, such as regorafenib and neratinib (47, 48). A similar strategy may also be used with abemaciclib, starting the treatment at 100 mg BID and increasing the dose to 150 mg after one of two cycles if the drug is well tolerated. Dose escalation might be helpful, especially in patients with risk factors for diarrhea, such as those identified in our study. Since data about abemaciclib dose-escalation are lacking, studies addressing its feasibility are needed. Based on our results, we plan to conduct a prospective trial to explore the impact of the abemaciclib dose-escalation strategy on the incidence and severity of diarrhea, dose reductions, and patient outcomes, including QoL. Personalizing abemaciclib dosing may eventually represent the key to minimizing adverse events and maximizing treatment efficacy, aiming to improve HR+/HER2− ABC patient management.
Conclusion
The peculiar tolerability profile of abemaciclib with a high rate of gastrointestinal side effects can be attributed to the principal involvement of CDK9, compared to CDK6, which is remarkably implicated in hematopoiesis, which in turn explains the lower frequency of myelosuppression compared to other CDK4/6i (32, 33). Moreover, the favorable toxicity profile of abemaciclib may be linked to its lower inhibitory activity against CDK1, CDK7, and CDK9 (32, 33).
Patients and their caregivers should closely monitor diarrhea since it may sometimes cause dangerous dehydration and electrolyte waste. In the case of diarrhea, serum chemistry tests can help identify alterations in electrolyte levels. In clinical trials, diarrhea was often managed using antidiarrheal agents, sparing the need for dosage reductions or interruptions in most populations. Clinical evidence suggests proactively antidiarrheal medications, such as loperamide and diphenoxylate/atropine, to prevent complications. Besides providing loperamide supply to the patient, the proactive education of patients and caregivers has a pivotal role. Oncologists must provide detailed information on diet and suggest using loperamide given prophylactically or at the first episode of diarrhea, adequate hydration, and seeking medical advice or preplanned precocious second visits to provide advice reinforcement. In addition, personalized dosing strategies, proactive symptom management techniques, patient education, and appropriately timed patient follow-ups are necessary to ensure that patients are in the best position to succeed in treatment.
Footnotes
Authors’ Contributions
Vittorio Gebbia and Paolo Vigneri were involved in the conception and design of the study. All Authors reported data relevant to the study. Vittorio Gebbia, Federica Martorana, and Dario Piazza analyzed the data. Vittorio Gebbia, Maria Vita Sanò, and Paolo Vigneri interpreted the data. All Authors critically reviewed the manuscript, revised the paper for intellectual content, provided detailed feedback, read, and approved the final manuscript, and agreed to be accountable for all aspects of the work.
Conflicts of Interest
All Authors disclose they received honoraria as expert testimony or scientific board members from Eli Lilly, Novartis, and Pfizer.
Funding
This work was partially supported by the GSTU Foundation for Cancer Research, Palermo, Italy.
- Received December 30, 2022.
- Revision received January 11, 2023.
- Accepted January 12, 2023.
- Copyright © 2023 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.