The efficacy and safety of CDK4/6 inhibitors combined with endocrine therapy versus endocrine therapy alone in the adjuvant treatment of patients with high-risk invasive HR+/HER2-early breast cancer: A comprehensive updated meta-analysis of randomized clinical trials

Abstract

Background

This paper aimed to evaluate the effectiveness of incorporating CDK 4/6 inhibitors (CDK4/6i) into ET for the adjuvant treatment of HR + HER2-resected early-stage breast cancer (ESBC) patients, employing meta-analysis.

Methods

In this paper, we compiled randomized clinical trials focusing on CDK4/6i used in the adjuvant treatment of high-risk invasive HR-positive and HER2-ESBC patients. A meta-analysis was performed in line with the PRISMA guidelines.

Results

We identified four clinical trials that met our inclusion criteria and were published between 2020 and 2024. These trials involved a combined sample size of 17,749 patients diagnosed with breast cancer. The data obtained from the pooled analysis revealed a remarkable beneficial trend in terms of invasive disease-free survival (iDFS) for the use of ET in combination with CDK4/6i compared to the group receiving ET alone (HR = 0.81, 95 % CI: 0.67–0.98, p = 0.03). Of note, CDK4/6 inhibitors demonstrated a notably beneficial effect in both grade 2 (HR = 0.69, 95 % CI: 0.59–0.81, p < 0.001) and grade 3 (HR = 0.76, 95 % CI: 0.65–0.89, p < 0.001). Significant improvements were noted in terms of distant relapse-free survival (dRFS) in the groups treated with abemaciclib and ribociclib (HR = 0.65, 95 % CI: 0.56–0.76, p < 0.001; HR = 0.72, 95 % CI: 0.58–0.89, p = 0.003, respectively). CDK4/6i didn't yield a statistically significant difference in overall survival (OS) (HR = 0.96, 95 % CI: 0.77–1.19, p = 0.69). The use of CDK4/6i with ET was associated with an increased risk of adverse events, particularly anemia and neutropenia, compared with ET alone (OR = 9.12, 95 % CI = 5.04–16.48, p < 0.001).

Conclusion

The findings of this paper demonstrate a significant improvement in iDFS when ET is combined with CDK4/6i, compared to ET alone. Specifically, treatments with abemaciclib and ribociclib showed notable enhancements in dRFS.

Highlights (for review)

• •Conducting subgroup analyses.
• •Including safety analyses.
• •Expanding and deepening the discussion.
• •Checking spelling errors and language.

1. Introduction

Approximately 70 % of early-stage breast cancers are comprised of hormone receptor-positive (HR+) and human epidermal growth factor receptor 2-negative (HER2-) patients [1,2]. In this group of patients, treatment strategies including surgery, radiotherapy, adjuvant or neoadjuvant chemotherapy, and endocrine therapy (ET) are employed based on risk characteristics. Endocrine therapy constitutes the cornerstone of adjuvant systemic treatment in appropriately resected patients. In this context, aromatase inhibitors (AI), anti-estrogens, and ovarian suppression are widely used based on menopausal status and risk conditions [3,4]. Adjuvant ET provides a significant reduction in both disease recurrence and risk of death. Furthermore, approaches such as extending the duration of endocrine treatment aim to further reduce the risk [5]. Despite all these efforts, a remarkable risk of recurrence persists over the years, with approximately 20 % of patients experiencing recurrences, primarily in the form of distant metastases, within the first 10 years [6].

In recent years, the addition of cyclin-dependent kinase 4/6 inhibitors (CDK 4/6i) to endocrine therapy has shown significant benefits in terms of progression-free survival (PFS) and overall survival (OS) in patients with HR+/HER2-metastatic breast cancer, both in the first line and subsequent lines of treatment. These inhibitors have also exhibited a moderate and tolerable toxicity profile. In line with this, three CDK4/6i (palbociclib, ribociclib, and abemaciclib) have received approval and have become standard treatments for HR+/HER2-metastatic breast cancer [[7], [8], [9], [10], [11], [12], [13], [14]]. The significant contribution of CDK4/6i in metastatic disease has shifted attention toward the adjuvant treatment of early-stage disease. In this regard, the MonarchE trial was conducted with abemaciclib as the first step. The study conducted in high-risk, node-positive, HR+/HER2-early-stage breast cancer patients demonstrated that the addition of abemaciclib to endocrine therapy significantly contributed to invasive disease-free survival (iDFS) [15]. In the PALLAS study, which investigated the use of palbociclib in adjuvant therapy, the addition of palbociclib to endocrine therapy did not show a significant contribution in terms of iDFS in stage II-III HR+/HER2-patients [16]. Also, the addition of palbociclib to endocrine therapy in HR+/HER2-high-risk patients with residual disease after neoadjuvant chemotherapy and surgery has not demonstrated benefit [17]. Recently, preliminary results from the NATALEE study, which investigated the role of ribociclib in adjuvant therapy, have been published. In this study, which included stage IIA (including N0 patients), IIB, and III patients, and utilized a lower dose of ribociclib, it was demonstrated that the addition of ribociclib to endocrine therapy significantly contributed to iDFS and exhibited a reasonable toxicity profile even with longer duration of use [18,30].

The benefit of adding CDK 4/6i to ET in HR+ HER2-metastatic breast cancer patients has been documented in randomized phase III trials and supported by several meta-analyses [[19], [20], [21]]. While the addition of CDK 4/6i to ET has shown benefits in two out of three randomized trials in early-stage disease, one trial did not show a significant benefit. Furthermore, these three studies differ in terms of patient characteristics and the duration of CDK 4/6i use. Two meta-analyses evaluating the use of palbociclib and abemaciclib in adjuvant therapy for early-stage disease have shown a trend of CDK 4/6i improving iDFS, but with increased toxicity and treatment discontinuation rates [22,23]. It has been noted that the benefit reached statistical significance only in patients with N2/N3 disease [23]. On the other hand, there is only one meta-analysis that included ribociclib, and this analysis is based on congress presentation data of the NATALEE study. As a result of this analysis, it was shown that the addition of ribociclib to adjuvant ET significantly contributed to iDFS [31]. However, there is currently no meta-analysis available that includes the recently full-text published NATALEE trial involving ribociclib.

Therefore, based on the relevant recent developments, the purpose of this paper was to assess the contribution of adding CDK 4/6i to endocrine therapy in the adjuvant treatment of HR + HER2-resected early-stage breast cancer patients utilizing the meta-analysis method.

2. Methods

This study focused on compiling randomized clinical trials involving CDK4/6i, particularly abemaciclib, palbociclib, and ribociclib in patients with early-stage breast cancer. In adherence to the “Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA)” statement [24], we carried out a meta-analysis to scrutinize the assembled available data. Supp. Table S1 includes the PRISMA checklist, which summarizes the specific criteria and guidelines followed during the systematic review and meta-analysis process. The checklist served as a comprehensive reference, ensuring transparency and adherence to the PRISMA statement in our study.

2.1. Literature search and data sources

Data collection for the study was performed by two independent and qualified investigators (MEA and IO). The researchers executed an extensive search for published papers by utilizing databases involving PubMed/Medline, Web of Science, and Scopus. The search period was developed until March 23, 2024, ensuring the inclusion of the most up-to-date information available. In our research strategy, we incorporated several key factors to enhance the effectiveness of our search procedure. These included Medical Subject Headings (MeSH) or keywords, text terms, and Boolean operators (AND/OR). By utilizing these factors, we aimed to develop a comprehensive and focused search strategy to identify relevant studies and maximize the accuracy of our data collection process. The keywords used in the search strategy were carefully selected to target the specific research area. The keywords included in the search strategy were as follows: “breast cancer”, “HR-positive”, “HER2-negative”, “invasive”, “high risk”, “CDK4/6 inhibitor”, “clinical trial”. These keywords were used to optimize the search and ensure the identification of relevant studies related to the use of CDK4/6 inhibitors in early-stage breast cancer and the conduct of a meta-analysis following PRISMA guidelines. The literature searches were conducted exclusively in the English language, and publications in other languages were not included in our search strategy. To provide a clear and organized presentation of the search strategies used for each relevant database, a structured form is available in Supp. Table S2.

2.2. Study selection

During the initial search phase, the titles and abstracts of the identified papers were independently examined by two researchers (MEA and IO). The objective was to evaluate the relevance of each paper to the research topic. Any papers that were deemed inapplicable or not relevant to the study were quickly excluded at this stage. This rigorous screening process aimed to ensure that only the most relevant and pertinent papers were included for further evaluation and analysis. For further analysis, we selected clinical trials that investigated the impact of endocrine therapy and ET + CDK4/6i in two arms in patients with high-risk invasive HR-positive and HER2-negative early-stage breast cancer. These trials were deemed suitable for inclusion based on their relevance to our research objectives. In this study, we have excluded certain types of studies to maintain the focus and relevance of our analysis. Specifically, animal model experiments, cell culture studies, non-original publications (such as letters to the editor and case reports), and studies that did not evaluate the effect of CDK4/6i in combination with endocrine therapy were excluded from our study. The primary endpoint of interest for the study was identified as iDFS. This study also assessed secondary outcomes, which included distant relapse-free survival (dRFS), overall survival (OS) and any adverse events (aAEs) (all grades and grade 3-4-5).

To ensure data integrity, a systematic process was employed. Initially, the Mendeley data management program was utilized to identify and eliminate any duplicate articles that were identified across the relevant databases. This step helped to streamline the dataset and avoid duplication of results. Next, three independent authors, MEA, MK, and IO reviewed the results reported by each clinical trial individually. They carefully synthesized and extracted the relevant data from the selected studies. To maintain consistency and organization, the data were then recorded in a pre-prepared and structured Microsoft Excel® spreadsheet.

2.3. Data extraction and quality assessment

In this study, comprehensive data extraction was carried out from the selected trials. This involved meticulously gathering relevant information from each trial, including study characteristics, participant demographics, intervention details, outcome measures, and any other pertinent data. The data extraction process was performed in a systematic and standardized manner to ensure consistency and accuracy throughout the study. The investigators independently extracted varied key components of information from the trials, including: i) Baseline characteristics of the studies: This includes details such as authors, publication year, publication source, and research design; ii) Characteristics of study subjects: Important information about the study subjects, such as median age, sex distribution, and disease stage, was extracted to understand the patient population under investigation; iii) Intervention and control therapies: The specific details of the intervention and control therapies used in each trial were recorded to evaluate the treatment modalities being compared; iv) Follow-up time: The duration of follow-up for each study was noted to assess the long-term outcomes and observe any potential changes over time; v) Adverse events: Any adverse events or side effects observed during the study were carefully recorded to evaluate the safety profile of the interventions. Among the articles with more than one publication that were scrutinized, the article containing the most comprehensive information was selected. In situations where there was disagreement between the authors during the data extraction process, we prioritized open communication and dialogue to resolve any discrepancies. The authors engaged in discussions to reach a consensus on the contentious points, ensuring that the extracted data were accurate and aligned with the study objectives. In the quality evaluation of all clinical trials included in our study, we considered various important factors. These factors include allocation concealment, random sequence generation, blinding, missing information, reporting bias, and other biases. We conducted the evaluation of the clinical trials according to the guidelines provided in the Cochrane Handbook of Systematic Reviews of Interventions 5.1.0 [25]. This handbook serves as a comprehensive and recognized resource for conducting high-quality systematic reviews and evaluating the methodological quality of included studies.

PICOs.

• 1.Population: “Patients with breast cancer”
• 2.Intervention: “The effect of CDK4/6 inhibitors + endocrine therapy on patients with breast cancer”
• 3.Comparison: “Endocrine therapy alone”
• 4.Outcomes: i) “invasive disease-free survival”, ii) “distant relapse-free survival”, iii) overall survival and “any adverse events”.
• 5.Study: “Clinical trials”

2.4. Statistical analysis

To assess the heterogeneity of the included trials in this study, we employed two commonly used statistical methods: the χ2-based Cochran's Q test and I² statistics. The Cochran's Q test examines whether the observed differences between the study results are beyond what would be expected by chance alone. A p-value of less than 0.05 suggests significant heterogeneity among the included trials. The I² statistics quantifies the proportion of total variation across studies that is due to heterogeneity rather than chance. A p-value less than 0.05 in Cochran's Q test and an I² value greater than 50 % indicated substantial heterogeneity among the trials. The outcomes of iDFS, dRFS, OS, and any adverse events (stage III-IV-V alone) were calculated using the hazard ratio (HR) or risk ratio (RR) along with the corresponding 95 % confidence intervals (CIs). In the statistical calculations of both primary and secondary meta-analyses, we took into consideration the presence of heterogeneity among the included studies. If significant heterogeneity was detected among the studies, we employed the random effects model for the analysis. On the other hand, if there was no significant heterogeneity observed, we used the fixed effects model. To assess the potential for publication bias, we utilized schematic illustrations of funnel plots. A two-tailed p-value of less than 0.05 was employed as the threshold to discriminate statistical significance in all tests performed. Meta-analysis statistical computations were carried out via The Review Manager version (RevMan) 5.4 (The Nordic Cochran Centre, Copenhagen, Denmark) [26].

2.5. Subgroup analysis

In the meta-analysis, any adverse events were conducted based on both all grades and the grade 3-4-5 subgroup. Then, separate subgroup analyses were performed in the anemia, neutropenia, arthralgia, nausea, headache, fatigue, diarrhea, alanine aminotransferase (ALT) increased, aspartate aminotransferase (AST) increased, as well as hot flush subgroups. For iDFS, subgroup analyses were executed using available data based on age (≤50 vs > 50), grade, stage, menopausal status, neoadjuvant and adjuvant treatment group, and high and low Ki-67 situation.

3. Results

The initial database search yielded a total of 262 papers. Through the screening process, which involved excluding irrelevant papers and removing duplicates, we identified four clinical trials that met our inclusion criteria. These four clinical trials were published between 2020 and 2024, and they were represented by a total of 7 papers [[15], [16], [17], [18],27,28,30]. Fig. 1 illustrates the PRISMA flow diagram, which outlines the process of searching for literature and provides a visual representation of the inclusion and exclusion criteria. According to the Cochrane risk of bias tool, in general, the clinical trials included in the meta-analysis were observed to be of high quality. Three of the included studies, PALLAS, monarchE, and NATALEE were deemed to have the risk of bias in some concerns. This was primarily because these studies had an open-label design, which means that the participants and investigators were aware of the treatment assignment. This lack of blinding could potentially influence the outcome assessment and introduce bias into the results. However, without further information or clarification, it is uncertain how much this potential bias may have affected the study outcomes. The other trial, PENELOPE-B, was evaluated and determined to have a low risk of bias. Fig. 2 presents the results of the quality assessment conducted on the included clinical trials.

Fig. 1.

The PRISMA flow diagram illustrates the search strategy and identifies the studies that were included and excluded.

Fig. 2.

The summary and methodological quality assessment of the included clinical trials.

3.1. Baseline characteristics of clinical trials enclosed in the meta-analysis

A total of four phase III randomized clinical trials (comprising 7 papers) were included in the meta-analysis. These trials involved a combined sample size of 17,749 patients diagnosed with breast cancer. Among these patients, 8872 were assigned to the endocrine therapy + CDK4/6 inhibitor group, while 8877 were assigned to the endocrine therapy alone group. The median ages of the patients in the included trials ranged from 48 to 52 years. The median follow-up time varied across the studies, ranging from 15.5 months to 42 months. In the abemaciclib (monarchE trial), the administered dose was 150 mg twice daily for two years. For the PALLAS trial, the dose of palbociclib was 125 mg orally once daily on days 1–21 of a 28-day cycle for two years. In the PENELOPE-B trial, the dose of palbociclib was also 125 mg once daily for one year, compared to the placebo group. Unlike other studies, in the NATALEE study, the duration was longer for three years, and the ribociclib dose was 400 mg/day for three weeks on/one week off compared to the endocrine-only group. The relevant details regarding the included studies can be found in Table 1, which summarizes their baseline characteristics. Also, the basic features of the patients enrolled in the clinical trials were demonstrated in Table 2.

Table 1.

Baseline characteristics of the clinical trials incorporated into the meta-analysis.

First author/study name	Study design	Year	Sample size (n) (ET + CDK4/6i/ET alone)	CDK4/6 inhibitors	Prior chemotherapy (Neoadjuvant) (n %) (ET + CDK4/6i/ET alone)	Prior chemotherapy (Adjuvant) (n %) (ET + CDK4/6i/ET alone)	Follow-up time (median)	Primary and secondary endpoint(s)
Johnston et al./monarchE trial	Open-label, randomized, phase III	2020–2022	2808/2829	Abemaciclib, 150 mg twice daily	1039 (37.0)/1048 (37.0)	1642 (58.5)/1647 (58.2)	15,5 months	iDFS, DRFS, aAEs, OS
Mayer et al., Gnant et al./PALLAS trial	Open-label, randomized, phase III	2021	2884/2877	Palbociclib, 125 mg orally once daily on days 1–21 of a 28-day cycle	965 (33.5)/974 (33.9)	1448 (50.2)/1427 (49.6)	31 months	iDFS, DRFS, LCFS, iBCFS, OS aAEs
Loibl et al./PENELOPE-B trial	Double-blind, placebo-controlled, phase III	2021	631/619	Palbociclib, 125 mg once daily vs placebo	N/A	N/A	42 months	iDFS, DRFS, LRRFI, OS, aAEs
Slamon et al./NATALEE trial	Open-label, randomized, phase III	2024	2549/2552	Ribociclib, dose of 400 mg per day for 3 weeks, followed by week off, for 3 years	1085 (42.6)/1095 (42.9)	1223 (48.0)/1220 (47.8)	34 months	iDFS, DRFS, DDFS, OS, aAEs

ET: endocrine therapy, CDK4/6i: CDK4/6 inhibitor, iDFS: invasive disease-free survival, DRFS: distant relapse–free survival, LCFS: locoregional cancer-free survival, iBCFS: invasive breast cancer–free survival, DDFS: distant disease–free survival, OS: overall survival, LRRFI: locoregional relapse-free interval, N/A: not available, aAEs: any adverse events, SD: standard deviation, IQR: interquartile range.

Table 2.

Baseline characteristics of the patients that were included in the clinical trials.

First author/study name	Age, (median (range or min-max)) (ET + CDK4/6i/ET alone)	Sex, female, (n %) (ET + CDK4/6i/ET alone)	Menopausal status (premenopausal; postmenopausal) (n %) (ET + CDK4/6i/ET alone)	Histopathological grade at diagnosis (grade 1; grade 2; grade 3) (n %) (ET + CDK4/6i/ET alone)	Prior radiotherapy (Neoadjuvant; Adjuvant) (n %) (ET + CDK4/6i/ET alone)	Initial adjuvant endocrine therapy (Aromatase inhibitors; tamoxifen; toremifene or other(s)) (n %) (ET + CDK4/6i/ET alone)
Johnston et al./monarchE	51 (23–89)/51 (22–86)	2787 (99.3)/2814 (99.5)	1221 (43.5)/1232 (43.5); 1587 (56.5)/1597 (56.5)	209 (7.4)/215 (7.6); 1373 (48.9)/1395 (49.3); 1090 (38.8)/1066 (37.7)	71 (2.5)/82 (2.9); 2620 (93.3)/2628 (92.9)	1928 (69.1)/1891 (67.5); 857 (30.7)/898 (32.1); 6 (0.2)/11 (0.4)
Mayer et al., Gnant et al./PALLAS trial	52.0 (25–90)/52.0 (22–85)	2867 (99.4)/2858 (99.3)	1303 (45.2)/1323 (46.0); 1562 (54.2)/1534 (53.3)	300 (10.4)/313 (10.9); 1622 (56.3)/1658 (57.6); 836 (29.0)/767 (26.7)	N/A; 2558 (88.7)/2560 (89.0)	1955 (67.8)/1917 (66.6); 922 (32.0)/950 (33.0); N/A
Loibl et al./PENELOPE-B trial	49.0 (22.0–76.0)/48.0 (19.0–79.0)	631 (100)/619 (100)	300 (47.5)/316 (51.1); 331 (52.5)/303 (48.9)	31 (5.0)/36 (5.9); 355 (57.0)/330 (54.0); 237 (38.0)/245 (40.1)	N/A	317 (50.2)/311 (50.2); 314 (49.8)/308 (49.8); 108 (17.1)/113 (18.3)
Slamon et al./NATALEE trial	52.0 (24.0–90.0)/52.0 (24.0–89.0)	2538 (96)/2453 (96)	1115 (43.7)/1123(44.0); 1423 (55.8)/1420 (55.6)	218 (8.6)/240 (9.4); 1458 (57.2)/1451 (56.9); 521 (20.4)/549 (21.5)	1085 (42.6)/1095 (42.9); 1223 (48.0)/1220 (47.8)	1601 (62.8)/1592 (62.4); N/A; 4 (0.2)/13 (0.5)

ET: endocrine therapy, CDK4/6i: CDK4/6 inhibitor, IDFS: invasive disease-free survival, DRFS: distant relapse–free survival, N/A: not available, SAE: serious adverse events, SD: standard deviation, IQR = interquartile range.

3.2. Outcomes of efficacy analysis

A total of four eligible clinical trials [[15], [16], [17], [18],27,28,30] were assessed for the association between CDK4/6i + ET versus ET alone and iDFS. As illustrated in Fig. 3, the data obtained from the pooled analysis, which included a total of 17,749 participants from four trials, revealed a remarkable beneficial trend in terms of iDFS for the use of ET in combination with CDK4/6i compared to the group receiving ET alone (HR = 0.81, 95 % CI: 0.67–0.98, p = 0.03). Based on the statistical analyses performed, a significant level of heterogeneity was observed among all the groups (Tau² = 0.03, I² = 78.0 %, p = 0.003). Established on this, the analyses were conducted using a random-effects model, and the results were quantified. When CDK4/6i were evaluated in separate arms, significant and higher benefits were observed in the abemaciclib and ribociclib groups (HR = 0.66, 95 % CI: 0.58–0.76, p < 0.001; HR = 0.75, 95 % CI: 0.62–0.91, p = 0.003, respectively) (Fig. 3). Of note, according to the results of trials that used the random-effects model, palbociclib did not significantly change the iDFS (HR = 0.95, 95 % CI: 0.83–1.09, p = 0.46) either without considerable heterogeneity (I² = 0.0 %, p = 0.98) (Fig. 3). The schematic illustration of the funnel plot indicated no evidence of publication bias among the included trials (Fig. S1).

Fig. 3.

The forest plot of the impact of CDK4/6 inhibitors with endocrine therapy compared to endocrine therapy alone on invasive disease-free survival. CI = confidence interval.

In the pooled subgroup analysis for iDFS of two eligible studies for grade 2 and three eligible studies for grade 3, the use of ET in combination with CDK4/6 inhibitors was associated with a significant beneficial trend in both grade 2 (HR = 0.69, 95 % CI: 0.59–0.81, p < 0.001) and grade 3 (HR = 0.76, 95 % CI: 0.65–0.89, p < 0.001) compared to the ET alone group (Fig. 7). Again, in the pooled subgroup analysis for iDFS of two eligible studies, a similar trend was found in stage II (HR = 0.74, 95 % CI: 0.58–0.94, p = 0.01) and stage III (HR = 0.67, 95 % CI: 0.58–0.76, p < 0.001) without significant heterogeneity. The outcomes of other subgroup analyses fir iDFS are demonstrated in Fig. 7.

Fig. 7.

The forest plot of the subgroup analysis of the impact of CDK4/6 inhibitors with endocrine therapy compared to endocrine therapy alone on invasive disease-free survival. CI = confidence interval.

Four relevant clinical trials [[15], [16], [17], [18],27,28,30] were evaluated to determine the relationship between the combination of CDK4/6i and ET versus solely ET concerning dRFS. In the meta-analysis conducted with the random effect model (Tau² = 0.06, I² = 85 %, p = 0.0002), as shown in Fig. 4, data from the pooled analysis revealed that ET in combination with CDK4/6i showed no significant difference in dRFS compared to the group receiving ET alone (HR = 0.84, 95 % CI: 0.65–1.08, p = 0.18). Significant improvements were noted in terms of dRFS in the groups treated with abemaciclib and ribociclib when CDK4/6i were analyzed in distinct arms (HR = 0.65, 95 % CI: 0.56–0.76, p < 0.001; HR = 0.72, 95 % CI: 0.58–0.89, p = 0.003, respectively) (Fig. 4). Based on the findings from trials employing the random-effects model, palbociclib did not significantly impact the dRFS (HR = 1.04, 95 % CI: 0.90–1.21, p = 0.58), and this was observed without substantial heterogeneity (I² = 0.0 %, p = 0.86) (Fig. 4). A schematic representation of the funnel plot is provided in Fig. S2.

Fig. 4.

The forest plot of the impact of CDK4/6 inhibitors with endocrine therapy compared to endocrine therapy alone on distant relapse-free survival. CI = confidence interval.

An evaluation of four suitable clinical trials [[15], [16], [17], [18],27,28,30] was conducted to assess the impact of combining CDK4/6i with ET compared to ET alone on OS. Using the random effect model for the meta-analysis, the pooled data analysis indicated that combining ET with CDK4/6i did not yield a statistically significant difference in OS when compared to the group treated with ET alone (HR = 0.96, 95 % CI: 0.77–1.19, p = 0.69) (Fig. 5). When CDK4/6i were pooled in different arms, there were no significant improvements in OS in all groups treated with abemaciclib, palbociclib, and ribociclib. Fig. S2 offers a schematic representation of the funnel plot.

Fig. 5.

The forest plot of the impact of CDK4/6 inhibitors with endocrine therapy compared to endocrine therapy alone on overall survival. CI = confidence interval.

3.3. Outcomes of safety analysis

For safety analysis, an evaluation of four appropriate clinical trials [[15], [16], [17], [18],27,28,30] was conducted to assess the impact of combining CDK4/6 inhibitors with endocrine therapy compared to endocrine therapy alone on adverse events. Analyses were performed separately for all grades and only for grades 3-4-5. In a pooled analysis of 17,552 participants for all grades, the use of CDK4/6 inhibitors with endocrine therapy was associated with an increased risk of any adverse events compared with endocrine therapy alone (OR = 9.12, 95 % CI = 5.04–16.48, p < 0.001) (Fig. 6). As seen in Fig. S4, a high level of heterogeneity was detected among the pooled clinical studies (Tau² = 0.25; I² = 83.0 %, p < 0.001). A similar trend was observed in the meta-analysis for grades 3-4-5 (OR = 9.50, 95 % CI = 5.35–16.87, p < 0.001; Tau² = 0.34, I² = 98.0 %, p < 0.001).

Fig. 6.

The forest plot of the impact of CDK4/6 inhibitors with endocrine therapy compared to endocrine therapy alone on adverse event for all grade (A) and grade 3-4-5 (B). CI = confidence interval.

Adverse events across all grades were also pooled in subgroup analyses. In particular, the combination of CDK4/6 inhibitors and endocrine therapy has been associated with a significant increase in anemia (OR = 6.58, 95 % CI: 5.08–8.53, p < 0.001) and neutropenia (OR = 43.92, 95 % CI: 16.84–114.56, p < 0.001) (Fig. S5). Combination use of CDK4/6 inhibitors and endocrine therapy has been associated with a reduction in arthralgia (OR = 0.72, 95 % CI: 0.63–0.84, p < 0.001) and hot flush (OR = 0.77, 95 % CI: 0.64–0.93, p = 0.006) compared to endocrine therapy alone. Forest plots for other subgroup analyses are presented in Fig. S5.

4. Discussion

In this meta-analysis, the contribution of adding CDK 4/6 inhibitors to ET in the adjuvant treatment of patients with HR + HER2-resected early stage breast cancer was evaluated. As a result of this analysis, including the recently published NATALEE study, it was shown that the combination of ET and CDK 4/6 inhibitor provided a significant improvement in iDFS compared to ET alone, but brought with it a significant level of toxicity.

Conflicting results have been reported in three previous meta-analyses regarding the role of CDK 4/6i in the adjuvant treatment of early-stage breast cancer. In the first two of these [22,23], MonarchE, Pallas and Penelope-B studies were included in the analysis. Agostinetto E et al. [22] reported that the addition of CDK 4/6 inhibitors to ET tended to improve iDFS (HR: 0.85, p = 0.07), but there was no difference in dRFS (HR: 0.83, p = 0.31).

Gao H.F. et al. [23] observed an improvement in iDFS in the second analysis (HR:0.87, p = 0.03; however, “None of the cumulative z-curves crossed the trial monitoring boundaries in TSA, and no reliable conclusion could be drawn”) in the disease group with N2 and N3 it was stated that the benefit reached statistical significance (HR: 0.83, p = 0.02). It was reported that there was no significant difference in the other groups (TNM stage, T stage, histological grade, prior neoadjuvant chemotherapy). The third analysis included the congress presentation data of the NATALEE study in addition to the three studies above, and in this study, it was shown that the addition of ribociclib to adjuvant ET significantly contributed to iDFS (HR: 0.81). In the NATALEE study, both node+ and node-patients were included, and it was reported that the iDFS benefit was similar in stage 2 and stage 3 (absolute iDFS benefit in the 3rd year was 3 % and 3.2 %, respectively). In the subgroup analysis, it was stated that the contribution in question was independent of menopausal status, tumor grade, Ki-67 index and previous neoadjuvant chemotherapy, but there was a difference in terms of stage, and while the contribution in stage 3 was significant (HR: 0.67), in stage 2 It has been stated that the one in is in the direction of recovery (HR: 0.74) [31]. In our study, a pooled analysis was performed with the current full-text published results of the NATALEE study, as well as the MonarchE, Pallas and Penelope-B studies. As a result, unlike the above three meta-analyses, it was determined that there was a significant improvement in both iDFS in the whole group (HR: 0.81, p = 0.03) and in dRFS (HR: 0.65, p < 0.001) in the group treated with abemaciclib and ribociclib. However, no significant difference was detected in OS. While the two analyzes above, in which ribociclib was not included, reported an improvement in iDFS and no difference in dRFS, in our analysis, the contribution to both iDFS and dRFS was found to be statistically significant. In the subgroup analysis, it was observed that this contribution was independent of menopausal status, tumor grade, Ki-67 index, stage and prior neoadjuvant chemotherapy. We think that in addition to the MonarchE study, the results of the current NATALEE study may also contribute to these positive results.

One of the several possible explanations for the observed differences in outcome observed in adjuvant CDK4/6i studies is the difference in efficacy and potency of CDK 4/6 inhibitors preclinically [27]. For example, ribociclib and abemaciclib have greater potency against CDK4 than CDK6, whereas palbociclib has similar potency against cyclin D1-CDK4 and cyclin D2-CDK6 complexes [27]. Some differences in efficacy have also been reported in CDK4/6i studies in the metastatic setting, with palbociclib appearing more effective only in patients with bone disease, and a significant proportion of high-risk patients enrolled in PALLAS in the adjuvant setting did not benefit more from palbociclib than the overall study population, consistent with PENELOPE-B [17,27]. Abemaciclib, on the other hand, appears to be more effective in patients with aggressive and poor prognosis disease in metastatic disease, which is consistent with the benefit reported in the monarchE study on high-risk patients with adjuvant abemaciclib [27,29]. Ribociclib is effective even in patients with clinically aggressive statement in metastatic disease and has shown PFS contribution [32]. The effectiveness of ribociclib in adjuvant treatment in both node-negative and node-positive patients was also consistent with the effectiveness in metastatic disease [30].

The four studies included in the analysis differ in terms of patient recruitment characteristics and treatment durations. Although all of them were performed in patients with HR + HER2-early stage breast cancer, the MonarchE study included patients with HR + HER2-node + early stage breast cancer, while the Pallas and Penelope-B studies included stage II-III patients, the majority of whom were node positive. While the treatment duration was two years in the MonarchE and Pallas studies, it was one year in the Penelope-E study. In the NATALEE study, unlike the other three studies, both node+ and node-patients were included and ribociclib was administered at a lower dose for three years. In addition, there are differences in terms of risk characteristics; while patients with high risk of recurrence were included in the MonarchE and Penelope-B studies, only 58.7 % patients with clinically high risk characteristics were included in the Pallas study.

Before the studies of CDK 4/6i in adjuvant therapy, as a result of studies and meta-analyses on the optimal adjuvant hormonal therapy treatment in early stage HR + HER2-breast cancer, in low-risk (pT ≤ 2 cm, G1, ki 67< %20, pN0) premenopausal patients, only tamoxifen is recommended for 5 years, and in postmenopausal patients, AI or TMX is recommended for 5 years or 2–3 years TMX->2–3 years AI; in high-risk (pT > 2 cm, G3, ki 67 > 20 %, pN+) patients, TMX or AI and extended hormonotherapy with OFS for 5 years in premenopausal patients, for postmenopausal patients, extended hormonotherapy (provided that AI was used for at least 5 years) was recommended for 7–8 years. In high-risk patients with germline BRCA 1/2 mutant, olaparib was recommended for 1 year, in addition to hormonal therapy [33].

After studies showing the positive contribution of abemaciclib and ribociclib, which are CDK 4/6i, on DFS, it was recommended to use concomitant in addition to endocrine treatment for 2 years for abemaciclib and 3 years for ribociclib in available patients [30,33]. Additionally, for BRCA 1/2 mutant patients, patients were included in the study within the first 16 months after surgery for abemaciclib, within 18 months from the time of diagnosis for ribociclib, or within 12 months for patients starting standard (neo)adjuvant ET, it is primarily recommended to start abemaciclib or ribociclib followed by Olaparib [33]. In meta-analyses, HR + HER2-is defined as high risk in breast cancer, it has been shown that factors such as histological grade, histological type, lymph node status, tumor size, Ki 67 status, hormone receptor (estrogen and progesterone) expression level, Residual Cancer Burden (RCB) after neoadjuvant treatment, status, genomic signatures status (Oncotype DX®, MammaPrint® (MP), Prosigna®, and Endopredict®) are used primarily [34]. However, in the definition of high-risk patients determined to add CDK 4/6i treatment to adjuvant therapy, lymph node status, ki 67 level, histological grade and tumor size are used for abemaciclib treatment, while for ribociclib, in addition to these factors, genomic risk profile is also used as a factor prognostic for the treatment decision [30,33]. Future studies will show whether some resistance-related mutations (PI3K, ESR1 mutation, etc.) and ctDNA levels will be included in addition to the genomic risk profile when defining high-risk patients in the future.

When the results are evaluated in terms of toxicity, it is reported that there is a significant increase in toxicities and treatment discontinuation rates with combined treatment in all four analyses. Agostinetto E et al. [22] stated that there was a significant increase in all grade toxicities (RR: 9.36) and treatment discontinuation rates (RR: 22.11), while Gao HF et al. [23] reported a significant increase in grade ¾ toxicities (RR: 4.14) and treatment discontinuation rates (RR: 19.16) in the second analysis. Similarly, in our study, it was found that combined use significantly increased the risk of adverse events (RR: 3.71).

When the distribution of adverse effects is examined, in the meta-analysis conducted by Agostinetto E. et al., neutropenia (OR: 38.14 (6.19–235.09)), diarrhea (OR: 14.51 (1.02–205.84)), thrombocytopenia (OR: 12.43 (7.58–20.38)), anemia (OR: 6.84 (4.16–11.24)) increased in favor of the combination of CDK 4/6 inhibitor and endocrine therapy, while arthralgia (OR: 0.66(0.50–0.86)) and hot flushes (OR: 0.70 (0.54–0.91)) increased in favor of those receiving only endocrine therapy [22]. In the meta-analysis performed by Gao HF et al. including three adjuvant studies (PALLAS, PENOLOPE-B, monarchE), leukopenia (RR: 86.69 (95 % CI 20.52–366.23, p = 0.0005)), neutropenia (RR: 67.39 (95 % CI 19.56–232.17, p < 0.0001)), lymphopenia (RR: 11.12 (95 % CI 7.20–17.18, p < 0.00001)), thrombocytopenia (RR: 8.26 (95 % CI 2.33–29.34, p = 0.001)) and anemia (RR: 4.17 (95 % CI 2.37–7.32, p < 0.0001)). 0.00001)) increased in favor of the combination of CDK 4/6 inhibitor and endocrine therapy [23]. Similar to these meta-analyses, in our study, neutropenia (OR = 43.92, 95 % CI: 16.84–114.56, p < 0.001) and anemia (OR = 6.58, 95 % CI: 5.08–8.53, p < 0.001) increased in favor of CDK 4/6 inhibitors, while arthralgia (OR = 0.72, 95 % CI: 0.63–0.84, p < 0.001) and hot flushes (OR = 0.77, 95 % CI: 0.64–0.93, p = 0.006) increased only in favor of endocrine therapy.

When the mortality rates in the CDK 4/6 inhibitor arms of the studies were evaluated, fatal serious side effects occurred in 8 patients in the palbociclib adjuvant study Penelope-B, while non-drug related (cardiogenic shock and influenza) deaths occurred in two patients in the palbociclib arm [17]. In the other study of palbociclib, the PALLAS study, the death rate in the palbociclib and ET-only arms was almost similar (0.5 % vs 0.3 %, respectively), and no deaths were related to the drug [27]. In the MonarchE study, it was lower in the abemaciclib arm compared to the ET arm only ((7.4 % vs 8.3 %) [35]. Lastly, in the NATALEE study, the rate of death from any cause was lower in the ribociclib arm (2.4 % vs 3 %), but the treatment related mortality rate was slightly higher in the ribociclib arm (0.5 % vs 0.2 %) [30]. These results show that it is useful to consider toxicity properties when using CDK 4/6 inhibitors in an area such as adjuvant therapy, where quality of life and treatment compliance are important. In terms of treatment-related quality of life, the patient-reported outcomes (PRO) findings for abemaciclib in the MonarchE study confirmed a tolerable and reversible toxicity profile for abemaciclib. QoL was preserved with the addition of abemaciclib as adjuvant to endocrine therapy, supporting its use in patients with HR+, HER2-, high-risk early breast cancer [36]. For ribociclib PRO analysis results of the studies in question have not yet been published, and considering the differences in treatment durations and toxicities, it is thought that the results of the PROM analyzes of the studies will be effective in the decision on the cost-effective and preferable CDK 4/6 inhibitor. In this respect, PROM analysis results are also eagerly awaited.

Also, in the future, identifying biomarkers that will help distinguish which patients will benefit from adding a CDK 4/6 inhibitor to ET and in which patients only ET will be sufficient will provide guidance in an important area such as adjuvant therapy and will protect some patients from unnecessary toxicity.

One of the most prominent features of our study is that it is an analysis that also includes current data from the NATALEE study. It is a study that includes more patients and includes a more comprehensive analysis of CDK 4/6i, and also analyzes both efficacy and toxicity data.

The limitations of our study are that the studies included in the analysis had different patient recruitment criteria, different risk group definitions, different treatment durations (1, 2 and 3 years), and that it was not an “individual patient data meta-analysis” but a “pooled data meta-analysis”. Therefore, in this meta-analysis, the studies published so far on the effect of CDK 4/6 inhibitors in adjuvant treatment in HR + HER2-early stage breast cancer was evaluated what the optimal treatment duration should be, the effect of prolonged treatment duration on toxicity and its contribution to OS and DFS, the survival effect when changing the CDK 4/6i agent (palbociclib, abemaciclib, ribociclib), the appropriate patient population in which these treatments are effective (lymph node positive, high Ki67, high degree, young age, etc.), PRO analysis, costeffectiveness etc. haven't been able to clarified to questions that are currently unknown.

In conclusion, current findings revealed that the application of CDK 4/6 inhibitors in combination with ET in the adjuvant treatment of HR + HER2-resected early stage breast cancer patients provided a significant improvement in iDFS, while specifically abemaciclib and ribociclib also provided a significant increase in dRFS. These results are consistent with previous meta-analysis results and strongly support the role of CDK 4/6 inhibitors in adjuvant therapy. In the future, the results of studies investigating molecular profile-based approaches as well as clinicopathological factors will better illuminate our path in this field.

CRediT authorship contribution statement

Merve Keskinkilic: Writing – review & editing, Writing – original draft, Supervision, Software, Formal analysis, Data curation. Mehmet Emin Arayici: Writing – review & editing, Writing – original draft, Software, Methodology, Formal analysis. Yasemin Basbinar: Writing – review & editing, Writing – original draft, Software, Methodology, Data curation. Hulya Ellidokuz: Writing – review & editing, Writing – original draft, Software, Formal analysis, Data curation. Tugba Yavuzsen: Writing – review & editing, Writing – original draft, Supervision, Software, Formal analysis, Data curation. Ilhan Oztop: Writing – review & editing, Writing – original draft, Supervision, Software, Formal analysis, Data curation.

Data availability statement

The datasets used and/or analyzed in this study are available upon reasonable request from the corresponding author.

Ethics approval statement

Ethical approval is not required for this study, and all techniques followed the Declaration of Helsinki.

Funding statement

None.

Declaration of competing interest

All authors declare that they have no conflicts of interest to disclose.

Appendix ASupplementary data

The following is the Supplementary data to this article:

Multimedia component 1