Background

Cyclin-dependent kinase 4/6 inhibitors (CDK4/6i) combined with endocrine therapy have shown significant clinical benefits in terms of progression-free survival (PFS) and overall survival (OS) for hormone receptor (HR)-positive advanced breast cancer. Randomized controlled trials (RCTs), including PALOMA, MONALEESA, and MONARCH series for palbociclib, ribociclib, and abemaciclib, have demonstrated that CDK4/6i combined with endocrine therapy is superior to endocrine therapy alone [1,2,3]. Since receiving approval from the Food and Drug Administration (FDA) in 2015, CDK4/6i treatment has become widespread worldwide and the standard of care for HR-positive advanced breast cancers.

While RCTs are considered as the gold standard for evaluating the efficacy and safety of a novel medication due to their strict study design, enrolled criteria, standardized evaluation, and specific patient populations, they have limitations in terms of patient diversity and may not fully represent actual clinical practice efficacy [4]. Real-world data (RWD) play an important role in providing additional evidence for clinical practice, especially in heterogeneous populations [5]. Recent RWD studies for CDK4/6i, such as P-REALITY X for palbociclib and RIBANNA and CompLEEment-1 study for ribociclib, have contributed valuable insights [6,7,8]. However, most RCTs have enrolled small numbers of Asian patients, and only a few RWD studies have focused on the Asian population [9]. Additionally, due to the timeline of FDA approval, many RWD studies have predominantly included palbociclib usage over ribociclib or abemaciclib when considering all types of CDK4/6i. In Taiwan, under the National Health Insurance (NHI), only palbociclib and ribociclib are reimbursed for HR-positive metastatic breast cancer (mBC) patients currently, resulting in the majority receiving palbociclib and ribociclib rather than abemaciclib.

Despite the known similarities in mechanism and structure, there is a lack of information on the efficacy, adverse effects, and treatment outcomes across distinct CDK4/6i. This study aims to investigate two different CDK4/6i (palbociclib and ribociclib) in patients with HR-positive mBC and relapsed breast cancer (rBC) and provide real-world evidence of safety and treatment strategies in the Asian population. Furthermore, according to the NHI policy in Taiwan, patients with luminal mBC can receive only 2 years of CDK4/6i. Consequently, some patients experienced economic burden and interrupted CDK4/6i treatment after 2 years of reimbursement. We also provided preliminary results regarding the efficacy of treatment after two years in this subpopulation.

Methods

Study population and treatments

This observational and retrospective study enrolled patients with HR-positive mBC or rBC patients who underwent treatment of either palbociclib or ribociclib between January 2019 and December 2023 at the Taipei Veterans General Hospital, a tertiary referral medical center in North Taiwan. Patients treated with abemaciclib as the first-line CDK4/6i, as well as those with other types of cancers or early-stage breast cancer, were excluded. Clinical treatment practices adhered to the guidelines established by the Comprehensive Breast Health Center at the Taipei Veterans General Hospital, which were based on recommendations from the National Comprehensive Cancer Network (NCCN), European Society for Medical Oncology (ESMO), and St. Gallen International Consensus guidelines [10,11,12]. All clinicopathological information, disease status, breast cancer-related prescriptions, and adverse effects were extracted from the Big Data Center and electronic medical records of the Taipei Veterans General Hospital. All potentially identifying patient information will be encrypted, anonymized, deleting names, addresses, and other identifiable items, to provide information security. The study protocols were approved by the Institutional Review Board (IRB-TPEVGH No.: 2024-04-014AC) with waivers of informed consent.

Data collections and outcomes

HR-positive (luminal) breast cancer was defined as those with estrogen receptor (ER)-positive, with ≥ 1% of tumor cells exhibiting nuclear staining according to immunohistochemistry (IHC) assay. Patients with human epidermal growth factor receptor 2 (HER2) test scores of IHC 3 + (positive) or 2 + (equivocal), and confirmed amplification by fluorescence in situ hybridization (FISH), were classified as HER2 positive; otherwise, they were considered HER2 negative [13]. The pathology results were determined from the most recent data, which could be from the original breast tumor, local recurrent tumor, or metastatic lesions. Adverse effects were documented by physicians using the Common Terminology Criteria for Adverse Events (CTCAE) version 5.0 score [14]. Dose reduction of CDK4/6i was defined as a continuous lower dose than standard dosage on more than two occasions. "Endocrine therapy resistance" was defined as patients who experienced disease relapse under adjuvant endocrine therapy or within 12 months of completing adjuvant endocrine therapy [15]. We also included CDK4/6i resistance, which was defined as the patients treated with CDK4/6i and experiencing disease progression within 12 months of treatment. The time of disease relapse and mortality were recorded and analyzed. Progression-free survival (PFS) was defined as the duration from the date of starting CDK4/6i to the local recurrence, distant metastasis, or death, whichever came first.

Statistical analysis

All clinicopathological details including age, body mass index (BMI), histology, IHC status, disease status, metastatic site, treatment strategies (endocrine partners, prior chemotherapy, and switching to another CDK4/6i), and safety profiles (dose reduction, regimen changes or discontinuation, adverse effects) for the comparison of the two CDK4/6i were assessed using 1:1 propensity score matching and compared using the χ2 test and Fisher's exact test.

PFS between the palbociclib and ribociclib groups was assessed using Kaplan–Meier survival curve analysis with the log-rank test. Additionally, the Cox proportional hazards model was employed to estimate the risk of relapse and mortality. Statistical significance was defined as P < 0.05 for all analyses. All statistical analyses were conducted using R statistical software, version 4.2.1 [16].

Results

Study population and patient characteristics

The cohort consisted of 340 patients, with 179 (52.6%) receiving palbociclib and 161 (47.4%) receiving ribociclib. The median follow-up duration was 27.8 months, with 29.4 in the palbociclib and 23.9 in the ribociclib group with an insignificant P-value of 0.114. Patient characteristics are detailed in Table 1, demonstrating no significant differences between the two CDK4/6i. After 1:1 propensity score matching by factors such as age, histology, PR status, visceral metastasis, bone-only metastasis, de novo stage IV status, and prior chemotherapy, the matched cohort consisted of 161 patients in both the palbociclib and ribociclib groups (Table 2). The mean age was 61 years, while the palbociclib group had a numerically higher prevalence of over 70 and the ribociclib group had more patients under 50. Most patients had a normal BMI, and the majority had invasive ductal carcinoma (IDC) histology. There were 34.78% and 41.61% of patients with de novo stage IV disease who received palbociclib and ribociclib, respectively. Additionally, 69.57% and 74.53% had visceral metastasis, while 6.83% and 4.35% had local recurrence only, respectively. No significant differences were observed in the distribution of disease statuses and distant metastatic sites between the two groups.

Table 1 Clinicopathological characteristics of patients receiving palbociclib or ribociclib before matching (n = 340)
Full size table
Table 2 Clinicopathological characteristics of patients receiving palbociclib or ribociclib after propensity score matching (n = 322)
Full size table

Treatment strategies and safety profiles

The combination of endocrine therapies, including aromatase inhibitors (letrozole, exemestane, and anastrozole), fulvestrant (a selective estrogen receptor degrader), and tamoxifen (a selective estrogen receptor modulator), did not differ significantly between the two groups. Aromatase inhibitors were the most frequently prescribed endocrine partner in both groups. The prevalence of endocrine resistance was approximately 30% in both groups, with no significant difference observed. Notably, the palbociclib group had more patients with prior chemotherapy compared to the ribociclib group before matching (16.7% vs. 8.70%, respectively, p < 0.05). The prevalence of prior chemotherapy yielded adequate balance after applied matching on the propensity score (9.94% vs. 8.70%, respectively, p = 0.848), as indicated in Table 3.

Table 3 The combination of endocrine treatment and the prevalence of endocrine resistance and prior chemotherapy for advanced disease after propensity score matching
Full size table

Table 4 presents the adverse effects and dose adjustment in the study. Hematotoxicity was the most common adverse effect in both groups. The palbociclib group had significantly higher rates of neutropenia (90.68%), anemia (36.65%), and thrombocytopenia (25.47%) compared to the ribociclib group (76.40%, 24.84%, 10.56%, respectively). In contrast, elevated liver enzymes, skin rash, and QT prolongation occurred more frequently in the ribociclib group (19.25%, 6.21%, 6.83%, respectively) compared to the palbociclib group (6.21%, 1.24%, and 0%). The dose reduction rate was similar in both groups, with 29.81% in the palbociclib group and 24.22% in the ribociclib group. However, patients with palbociclib had a significantly lower rate of regimen changes (10.56%) compared to those with ribociclib (21.12%), which included switching to another CDK4/6 inhibitor or changing the endocrine partner. Three patients switched from palbociclib to ribociclib, 10 patients switched from palbociclib to abemaciclib, 24 patients switched from ribociclib to palbociclib, 4 patients switched from ribociclib to abemaciclib, and only 14 patients changed to a third type of CDK4/6 inhibitor later on (Fig. 1). The most common reason for changing treatment in the palbociclib group was disease progression, while adverse effect was the primary reason in the ribociclib group. Moreover, the palbociclib group had a significantly higher rate of discontinuing CDK4/6i compared to the ribociclib group (70.19% vs. 59.01%, p < 0.05), primarily due to disease progression.

Table 4 Adverse effects of CDK4/6 inhibitors and causes of dose adjustment
Full size table
Fig. 1
figure 1

Sankey plot of changing regimens between different types of CDK4/6 inhibitors

Full size image

Treatment efficacy

The median PFS (mPFS) for the full cohort was 29.34 months. Analysis using a Cox regression model showed no significant difference in PFS between patients treated with palbociclib and those with ribociclib. The mPFS was 28.94 months in the palbociclib group and 29.86 in the ribociclib group, with a hazard ratio of 0.905 and a P-value of 0.542 (Fig. 2a). Given the significantly higher proportion of patients receiving prior chemotherapy in the palbociclib group compared to the ribociclib group, we conducted a subgroup analysis focusing on patients treated with CDK4/6i as the first-line treatment. The PFS did not differ significantly between palbociclib and ribociclib, with mPFS of 30.42 and 29.89 months, respectively. The hazard ratio was 0.868 with an insignificant P-value of 0.419 (Fig. 2b). Patients with prior chemotherapy or endocrine resistance experienced significantly decreased PFS in both groups (Figs. 3 and 4). Furthermore, dose reduction of CDK4/6i showed similar efficacy compared to standard-dose treatment in both groups (Fig. 5).

Fig. 2
figure 2

Kaplan–Meier survival curve analysis for progression-free survival (PFS) between patients treated with palbociclib and ribociclib. a Whole cohort patients and b patients treated in the first-line treatment setting

Full size image
Fig. 3
figure 3

Kaplan–Meier survival curve analysis for progression-free survival (PFS) between patients treated with prior chemotherapy or not. a Patients treated with palbociclib. Median PFS (mPFS) in non-prior chemotherapy (CT) and prior CT was 30.42 months versus 12.09 months. Hazard ratio = 2.023, P < 0.05. b Patients treated with ribociclib. mPFS was 29.89 months in non-prior CT and 7.06 months in prior CT, hazard ratio: 4.276, P < 0.05

Full size image
Fig. 4
figure 4

Kaplan–Meier survival curve analysis for PFS between patients treated with endocrine therapy resistance (ETR) or not. a Patients treated with palbociclib. mPFS in non-ETR and ETR was 30.91 months versus 21.85 months. Hazard ratio: 1.655, P < 0.05. b Patients treated with ribociclib. mPFS in non-ETR and ETR was 42.08 months and 19.78 months, hazard ratio: 1.842, P < 0.05. *mPFS: medium progression-free survival.

Full size image
Fig. 5
figure 5

Kaplan–Meier survival curve analysis for PFS between patients with dose reduction (DR) of CDK4/6 inhibitor or not. a Patients treated with palbociclib. mPFS in non-dose reduction and dose reduction was 28.94 months versus 24.51 months. Hazard ratio: 1.114, P = 0.641. b Patients treated with ribociclib. mPFS in non-dose reduction and dose reduction was 28.84 months and 57.59 months, hazard ratio: 0.722, P = 0.295. *mPFS: medium progression-free survival.

Full size image

Table 5 presents the risk of disease relapse as determined by the univariate and multivariate Cox proportional hazards model. In univariate analysis, elevated BMI (HR 0.943, 95% CI 0.907–0.980), bone-only metastasis (HR 0.556, 95% CI 0.350–0.884), and combination with letrozole (HR 0.566, 95% CI 0.408–0.734) were associated with a decreased risk of disease progression, all with p-values < 0.05. Conversely, there was a significantly increased risk of disease relapse in patients with elevated HER-2 IHC staining (HR 1.312, 95% CI 1.055–1.631), elevated Ki-67 level (HR 1.018, 95% CI 1.007–1.028), visceral metastasis (HR 1.462, 95% CI 1.009–2.118), liver metastasis (HR 2.663, 95% CI 1.900–3.732), endocrine therapy resistance (HR 1.619, 95% CI 1.166–2.249), and prior chemotherapy treatment (HR 2.713, 95% CI 1.687–4.361), all with p-values < 0.05. With multivariate analysis, only elevated Ki-67 level (HR 1.021, 95% CI 1.010–1.033), liver metastasis (HR 2.676, 95% CI 1.759–4.071), endocrine therapy resistance (HR 1.551, 95% CI 1.048–2.297), and prior chemotherapy treatment (HR 2.016, 95% CI 1.110–3.658) were associated with an increased risk of disease progression, all with p-values < 0.05.

Table 5 Risk factors of disease relapse in breast cancer patients treated with CDK4/6 inhibitor using the univariate and multivariate Cox proportional hazards model
Full size table

Patients treated with CDK4/6i who progressed within 12 months were considered CDK4/6i resistant. Among the 66 patients identified, 36 were treated with palbociclib, and 30 were treated with ribociclib (including 1 patient who switched from ribociclib to palbociclib due to adverse effects), showing resistance to CDK4/6i. Table 6 reveals that patient characteristics did not differ between those with or without CDK4/6i resistance, except for a higher proportion of liver metastasis (39.39% vs. 20.70%), prior chemotherapy treatment (19.70% vs. 6.64%), and endocrine resistance (40.91% vs. 26.95%) in CDK4/6i-resistant patients compared to those without (p < 0.05).

Table 6 The characteristics of patients with or without CDK4/6 inhibitor resistance
Full size table

Safety of discontinuation after 2 years of treatment

A subgroup analysis was conducted for 38 patients who discontinued CDK4/6i treatment after 2 years of NHI reimbursement without disease progression. Of these, 17 patients had been treated with palbociclib and 21 with ribociclib. Following discontinuation, 3 ribociclib patients switched to fulvestrant, letrozole, and exemestane, respectively, and experienced disease progression events 231, 276, and 81 days after discontinuing CDK4/6i. Two patients treated with palbociclib switched to exemestane and had disease progression 112 and 210 days after discontinuing CDK4/6i. One patient with ribociclib switched to fulvestrant and died 585 days later (not related to breast cancer), while another palbociclib patient switched to capecitabine and died after 308 days related to breast cancer.

We further analyzed a total of 89 patients who continued CDK4/6i for over 2 years with out-of-pocket medication and 38 patients who interrupted CDK4/6i due to economic reasons, to evaluate the impact of early discontinuation. The median follow-up times were 40.8 months and 39.0 months, respectively, with p-value < 0.05. Figure 6 illustrates the results of a Kaplan–Meier survival curve, showing no significant difference in PFS between patients who continued treatment and those who interrupted treatment after 2 years (hazard ratio: 1.439, 95% confidence interval: 0.624–3.321, P = 0.393).

Fig. 6
figure 6

Survival analysis for patients who persisted with treatment and who interrupted treatment after 2 years

Full size image

Discussion

CDK4/6 inhibitors have emerged as a cornerstone in the treatment of HR-positive metastatic breast cancer over this decade. Three CDK4/6i—palbociclib, ribociclib, and abemaciclib—share a common mechanism of action, inhibiting CDK4/6 to halt the phosphorylation of retinoblastoma protein and arrest the G1 cell cycle phase in tumor cells [17, 18]. These agents have demonstrated similar efficacy in multiple clinical trials [1,2,3]. Palbociclib, in combination with an aromatase inhibitor, was the first to receive FDA approval for HR-positive/HER2-negative metastatic breast cancer since February 2015, followed by ribociclib and abemaciclib in March 2017 and September 2017, respectively. In Taiwan, National Health Insurance began reimbursing  palbociclib and ribociclib for HR-positive metastatic breast cancer in 2019. Consequently, palbociclib has been the most widely used CDK4/6i, followed by ribociclib, primarily due to the timeline of FDA approvals, marketing, and insurance coverage. In our study, we found no significant difference in patient characteristics between palbociclib and ribociclib. However, there was a numerical predominance of patients aged younger than 50 years treated with ribociclib than palbociclib, while patients older than 70 tended to receive palbociclib. The MONALEESA-7 trial enrolled approximately 27.7% of premenopausal and around 30% of Asian patients, which may have influenced physicians’ choices [19].

Regarding the adverse effects of CDK4/6 inhibitors, hematologic toxicity was the most prevalent in both groups in our cohort. The palbociclib group exhibited significantly higher rates of neutropenia, anemia, and thrombocytopenia compared to the ribociclib group. Conversely, the ribociclib group experienced a higher incidence of nausea, vomiting, elevated liver enzymes, skin rash, and QT prolongation. The adverse events in the palbociclib group were consistent with findings from the PALOMA trial [1]. However, patients in the ribociclib group exhibited significantly lower rates of gastrointestinal symptoms and skin rash compared to those reported in the MONALEESA and CompLEEment-1 trials [20, 21]. These discrepancies may be attributed to potential biases in patient-reported outcomes and medical record documentation. Moreover, palbociclib demonstrated no age-related differences between adverse effect in the PALOMA-2/3 trials in the elderly population. However, the MONALEESA series showed a higher incidence of anemia and gastrointestinal symptoms in patients over 65 years of age treated with ribociclib compared to younger patients. The risk of QT interval prolongation from ribociclib is also a consideration [22]. Despite the safety profiles of both CDK4/6i being established in clinical studies, physicians may lean toward prescribing palbociclib in elderly patients due to these observed differences. Moreover, patients treated with ribociclib, which is associated with a higher prevalence of digestive symptoms, skin rash, and elevated liver enzymes compared to palbociclib, may switch to other CDK4/6 inhibitors due to intolerance of adverse effects or physicians’ and patients’ concerns. In the previous studies, the PALOMA-2 trial failed to demonstrate an overall survival benefit, whereas the MONALEESA-2 trial reported a significant benefit, with an overall survival of 63.9 months [2, 23]. Meanwhile, the further clinical trial of palbociclib, PARSIFAL-Long, determined that the mPFS was 33.2 months, and the OS was 65.4 months, consistent with the OS reported in MONALEESA-2 [24]. This variation underscores the importance of continuously monitoring the evolving patterns of CDK4/6 inhibitor usage in RWD.

Our analysis of the mPFS in patients receiving the first-line CDK4/6i was 30.42 and 29.89 months for the palbociclib and ribociclib groups, respectively. These findings are slightly longer than the results of the PALOMA-2, which reported a mPFS of 27.6, and the MONALEESA-2, which reported a mPFS of 25.3 months. In the aspect of RWD, the P-REALITY X study demonstrated that HR-positive/HER2-negative metastatic breast cancer patients receiving palbociclib plus an aromatase inhibitor (AI) as first-line treatment had a mPFS of 32.6 months in a retrospective comparative analysis from routine practice in the USA [25]. On the other hand, the CompLEEment-1 study evaluated ribociclib plus letrozole in patients with HR-positive/HER2-negative advanced breast cancer, revealing a mPFS of 27.1 months [21]. A single-center study from the UK, focusing on the first-line therapy for HR-positive metastatic breast cancers, also showed similar results, with a mPFS of 27.9 for palbociclib and 29 months for ribociclib [26]. Additionally, a multi-center study in Germany reported a mPFS of 23 months for CDK4/6i in the first-line setting [27]. This study also highlights that patients with a history of extensive prior therapy had a shorter mPFS. Our analysis further demonstrated that patients with prior chemotherapy or endocrine resistance experienced shorter mPFS regardless of whether they received palbociclib or ribociclib. This subset of patients may exhibit aggressive tumor behaviors or have been heavily pretreated, potentially leading to resistance to CDK4/6i.

Regarding the diversity of ethnicity, 14.6% of the patients in PALOMA-2 were Asian, and only 23.9% were from races other than Black (8.1%) and White (68.1%) in the P-REALITY X study [1, 6]. In ribociclib studies, only 8.4% of patients were Asian in MONALEESA-2 and 7% in CompLEEment-1 [20, 21]. A RWD on the use of palbociclib in advanced/metastatic breast cancer in the Asian population from Singapore showed a mPFS of 28.17 months in first-line treatment [9]. Our study cohort consisted entirely of an Asian population, and the outcomes were consistent with previous clinical trials and RWD. These results suggest that there is no significant difference in treatment efficacy between Asian and Western populations. However, there is still a need for more comprehensive RWD on the Asian population to date.

According to previous studies, an increase in BMI is associated with a higher risk of metastasis in obese breast cancer patients. However, they primarily focused on early breast cancers [28, 29]. In our study, we found that a higher BMI was an independent prognostic factor for better outcomes in metastatic HR-positive breast cancer. The mean BMI in our cohort was 23.5 kg/m2, which falls within the normal range. In a multi-center national-wide retrospective database study of metastatic breast cancers, it was concluded that overweight and obesity were not associated with poorer outcomes, while being underweighted emerged as an independent adverse prognostic factor [30]. Underweighted patients, which are more common in the Asian region, may be associated with poor health conditions, aggressive disease behavior, or a higher risk of treatment intolerance. Another study focusing on metastatic breast cancer patients treated with CDK4/6i found no difference in mPFS across BMI categories [31]. The correlation between BMI and breast cancer prognosis remains a topic of debate.

Regarding the multivariate Cox regression analysis in our study, only an elevated Ki-67 level, liver metastasis, endocrine resistance, and prior chemotherapy treatment showed a significantly increased risk of disease progression. Additionally, liver metastasis, endocrine resistance, and prior chemotherapy were more prevalent in patients with CDK4/6i resistance compared to those without, suggesting that these patients may have a more aggressive disease status or may be more resistant to treatment due to prior therapies. However, the use of letrozole decreased risk of disease progression in the univariate analysis, which could introduce bias into the study. In clinical practice, patients considered to have endocrine resistance or those receiving higher-line therapy tend to favor fulvestrant over aromatase inhibitors or tamoxifen, based on the findings from the PALOMA-3 and MONALEESA-3 trials [32, 33].

CDK6 plays a crucial role in regulating the proliferation of hematopoietic precursors, leading to hematotoxicity as a major adverse effect of CDK4/6i. Evidence suggests that Asian patients may experience a higher rate of neutropenia and require dose reductions compared to other ethnic populations in the PALOMA series and RWD from Asian countries [34, 35]. Our study also revealed a numerically higher rate of all-grade neutropenia in the palbociclib group (90.7%) compared to 79.5% in the PALOMA-2 study, and in the ribociclib group (76.4%) compared to 74.3% in MONALEESA-2 trial. These findings may be attributed to lower baseline absolute neutrophil counts (ANC) in Asian compared to non-Asian patients [36]. However, the dose reduction rate in the palbociclib group (29.8%) and the ribociclib group (24.2%) in our study was lower than that reported in the PALOMA trial (33.8%), P-REALITY X (41%), MONALEESA-2 (57.5%), and CompLEEment-1 (68.9%) [1, 20, 21, 25] This discrepancy may be due to the inconvenience of changing the dose within the constraints of NHI reimbursement policies, which leads physicians to interrupt treatment rather than to adjust the dose. Previous studies have suggested that neutropenia may be associated with better treatment outcomes, as it is related to the "on-target effect," which may predict efficacy [37,38,39]. Despite that, this association was not observed in our study, as there was no significant difference in PFS between patients who required dose reductions and those who did not, regardless of being treated with palbociclib or ribociclib. While the safety was confirmed, the impact of dose modification on treatment outcomes remains unclear. Therefore, further evaluation of patients requiring dose reductions is warranted.

Being a novel treatment modality, the economic toxicity of CDK4/6i becomes a significant factor influencing physicians' and patients' decisions. The Taiwan NHI system balances CDK4/6i expenses and patient benefits. Consequently, a decision was made to reimburse two years of CDK4/6i for the 1st and 2nd lines of HR-positive mBC based on the mPFS results from clinical trials. However, there is currently no evidence regarding the prognosis after discontinuation of CDK4/6i treatment after 2 years of reimbursement. In our study, the median follow-up time for the patients who discontinued CDK4/6i due to economic burden was significantly shorter than for patients with continuous treatment of CDK4/6i. This limited follow-up time was insufficient to assess the outcomes of patients treated with interrupted CDK4/6i for up to two years under the NHI reimbursement policy. Understanding these outcomes is crucial for ensuring patients’ safety and evaluating the efficacy of the NHI resource allocation.

This study has several limitations. Firstly, it is based on a retrospective database, which may introduce biases and limitations in data collection. There may be incomplete or inaccurate reporting of adverse effects, especially symptoms relevant to the patient's subjective well-being, such as gastrointestinal symptoms and fatigue without a formal quality-of-life assessment instrument. This might cause an underestimate of adverse effects, compared to the corresponding clinical trials. Also missing data on other facilities' treatments outside our institute, such as previous medical history and self-paid medications, cannot be retrieved from the electrical medical records or the Big Data Center. For some patients, the lack of pathology for primary or metastatic tumors may limit the accuracy of the analysis. Secondly, the sample size in this study was limited, which may affect the generalizability of the findings and subgroup analyses. Thirdly, the follow-up time in this study was limited, which makes it challenging to determine overall survival and assess the long-term effects of CDK4/6i treatment. Lastly, the study did not account for the sequential treatment effects of different CDK4/6 inhibitors and endocrine therapy shifting, which may impact treatment outcomes. Future studies should consider these factors to provide a more comprehensive understanding of the efficacy and safety of CDK4/6i treatments, especially for the Asian population.

Conclusion

Treatment outcomes for distinct CDK4/6i did not differ significantly, indicating the safety and efficacy of CDK4/6i in the Asian population. Ribociclib and palbociclib showed similar efficacy in terms of PFS in RWD of the Asian population. Patients with prior chemotherapy and endocrine therapy resistance had a negative impact on PFS, yet dose reduction showed no significant difference in PFS between palbociclib and ribociclib.