Efficacy, safety, and predictive model of Palbociclib in the treatment of HR-positive and HER2-negative metastatic breast cancer

The research retrospectively examined data pertaining to patients with HR + /HER2- MBC who underwent Palbociclib therapy in conjunction with ET treatment at Zhejiang Cancer Hospital between August 2018 and August 2022. The inclusion criteria were as follows (1): individuals who were diagnosed with HR + /HER2- MBC within the time frame of August 2018 to August 2022, encompassing both de novo stage IV disease and relapsed disease, and who had undergone a minimum of one cycle of Palbociclib in combination with ET treatment. (2): complete medical information, including age upon diagnosis, estrogen receptor(ER) or progesterone receptor(PR) status, HER2 status, distant metastatic sites, local therapy, treatment approach, the Palbociclib combined with ET treatment plan, the initial Palbociclib dosage, and adverse reactions encountered during Palbociclib treatment. The exclusion criteria were as follows (1): lack of accessible medical records and presence of coexisting cancers, (2): positive HER2 status (3): both ER and PR status being negative, and (4): diagnosis of triple-negative breast cancer.

Two hundred twenty seven patients were eligible to participate in this research. After excluding patients with missing information (n = 9) or other primary malignant tumors (n = 4), 214 patients were finally included in the analysis. The flowchart of this retrospective study is displayed in Fig. 1. The Medical Ethics Committee of Zhejiang Cancer Hospital granted approval for this research. Informed consent was waived by the Medical Ethics Committee of Zhejiang Cancer Hospital due to the retrospective nature of this study.

Medical record data has been retrieved, comprising age, menstrual status, ER or PR status, HER2 status, history of adjuvant therapy, tumor metastasis organs, and number of metastatic sites, Palbociclib combined with ET treatment plan, past treatment plan, dose adjustment, efficacy, adverse reactions, etc. Progression-free survival (PFS) served as the primary endpoint in the study, while the objective response rate (ORR) and disease control rate (DCR) acted as the secondary endpoints. PFS was defined as the time interval (in months) from the initiation of CDK4/6i to the first recorded clinical and/or radiographic evidence of disease progression or death from any cause (whichever occurs earlier). Disease progression (PD) was defined as the beginning of radiographic progression, clinical progression, or the initiation of a new treatment line for MBC. A new treatment line was defined as a change in the treatment regimen or starting a new chemotherapy, endocrine therapy, or targeted therapy.

According to RECIST 1.1, the therapeutic effect after receiving Palbociclib was classified into four categories: Complete Response (CR), Partial Response (PR), Stable Disease (SD), and Progressive Disease (PD). ORR was defined as the percentage of patients achieving CR or PR, while DCR was defined as the proportion of patients achieving CR, PR, or SD. HR + /HER2- breast cancer cases were categorized into luminal A (PR ≥ 20% and Ki-67 < 20%) and luminal B (PR < 20% and Ki-67 ≥ 20%) [12]. Because guidelines of Chinese Society of Clinical Oncology(CSCO) breast cancer 2022 doesn’t give margin for Ki-67, the authors chose the median as the boundary value. All patients were followed up in October 2022.

Based on the treatment regimens involving Palbociclib, patients were categorized into three subgroups. These subgroups underwent Palbociclib administration in combination with ET, serving as the first-line, second-line, and later-line therapies for MBC. The PFS of each subgroup was separately analyzed to observe differences among each overall population. As the first CDK4/6i to be approved in China, Palbociclib has been widely employed in HR + /HER2- MBC patients. More specifically, it is usually administered in combination with aromatase inhibitors (AIs) or Fulvestrant. The typical initial dose of Palbociclib is 125 or 100 mg/day [13]. CSCO recommends an initial dose of 125 mg/day (Palbociclib is typically taken once daily for a duration of three weeks, followed by a one-week interval) [14]. However, doctors can adjust the dosage and timing of the medication based on the patient's age and physical condition.

Two hundred and fourteen HR + /HER2- MBC patients were randomly allocated in a 3:1 ratio to either the training set (n = 161) or validation set (n = 53) by using R version 4.1.2 (http://​www.​r-project.​org/​). Then, the independent samples t-test or chi-square test was used to evaluate differences in variables between the training and validation sets. In the training set, univariate and multivariate analyses were conducted using SPSS version 25.0 to determine variables significantly associated with PFS and to calculate the hazard ratio (HR) and 95% confidence interval(CI) for each variable. Utilizing the findings from a comprehensive multivariate analysis, a predictive nomogram model was subsequently established. The discriminative ability of the established model was assessed using the receiver operating characteristic (ROC) curve and the area under the ROC curve (AUC). Afterward, the performance of the nomogram was evaluated in the validation set by using 200-fold cross-validation in R. Furthermore, R was utilized to generate calibration plots for the training and validation sets, which can visually measure the degree of closeness between the actual conditions and the predicted conditions of the nomogram. At the same time, The programming language R was employed to generate decision curves, thereby assessing the predictive efficacy of the model within the clinical context. A two-tailed P-value below 0.05 was deemed to have statistical significance.

The research's median follow-up duration spanned 28.6 months (95% CI, 22.0–35.2 months). Among the 214 patients, 62.62% (134/214) were under 60 years old, with a mean age of 57 years. 82.24% (176/214) of HR + /HER2- MBC patients were already postmenopausal prior to the administration of Palbociclib in combination with ET. 37/214 patients (17.29%) were diagnosed as stage IV at their first diagnosis. In the latest immunohistochemistry analysis prior to using Palbociclib, more than half of the patients had PR < 20% (113/214, 52.80%) and Ki-67 ≥ 20% (119/214, 55.61%). Notably, the most common histological type among these patients was invasive ductal carcinoma (163/214, 76.17%), and the most common molecular subtype was luminal B (135/214, 63.08%). 85.51% (183/214) of patients underwent surgical intervention. At the same time, 57.01% (122/214) of the patients received adjuvant radiotherapy, whereas 78.04% (167/214) of the patients received adjuvant endocrine therapy. Furthermore, 42.52% (91/214) of patients received previous systemic chemotherapy before using Palbociclib. Patients were classified into primary resistance and acquired resistance based on whether they received adjuvant endocrine therapy for more than 2 years or first-line endocrine therapy for more than 6 months. Patients who experienced recurrence after 1 year after completion of adjuvant endocrine therapy or who had not received any prior endocrine therapy were considered sensitive to endocrine therapy [15]. In this study, 69.16% (148/214) of patients developed primary resistance to ET, while only 30.84% (66/214) were sensitive to ET or developed secondary resistance. The median PFS of Palbociclib was 7.17 months (95% CI, 7.61 months to 10.05 months). Prior to the use of Palbociclib in combination with ET, 63.55% (136/214) of patients had already developed visceral metastasis, and 55.61% (119/214) had ≥ 2 metastatic sites. Fortunately, only 34.58% (74/214) of patients manifested liver metastasis. Prior studies have reported that patients with liver metastasis who were given Palbociclib in combination with ET have a lower treatment efficacy than those without. Finally, 64.95% (139/214) of patients had a disease-free survival (DFS) period of more than 24 months.

The number of patients who received treatment involving the combination of Palbociclib and ET as first-line, second-line, and later-line therapies were 103/214 (48.13%), 63/214 (29.44%), and 48/214 (22.43%), respectively. Moreover, 97.20% (208/214) of patients received an initial dose of 125 mg/day of Palbociclib, while only 2.80% (6/214) of patients were initiated on a dose of 100 mg/day. Lastly, 42.06% (90/214) of patients had their dosage adjusted owing to intolerable side effects, with the most common side effect leading to dosage adjustment being grade 3/4 neutropenia. The most commonly co-administered ET drug with Palbociclib was AIs (131/214, 61.21%), followed by Fulvestrant (83/214, 38.79%). 152/214 (71.03%) patients discontinued combination treatment during the study period. Among them, 92.11% (140/214) switched to another regimen due to disease progression, 5/214 (3.29%) patients discontinued treatment due to financial reasons, 6/214 (3.95%) due to intolerable adverse reactions, and 1/214 (0.66%) due to the COVID-19 pandemic. Detailed clinical characteristics are presented in Table 1.

During a median follow-up duration of 28.6 months (95% CI, 22.0–35.2 months), 140 patients experienced disease progression. The median PFS of patients on Palbociclib combined with ET was 7.17 months (95% CI, 7.61–10.05 months). All lesions were measurable, with no CR cases, 6 PR cases, 68 cases of SD, and 140 cases of PD. The DCR was 34.58%, and the ORR was 2.80%.

According to the RECIST1.1 criteria for tumor assessment, all lesions in this retrospective study were measurable. Among all patients, 2.80% (6/214) had a tumor assessment of PR, 31.78% (68/214) achieved SD, and 65.42% (140/214) achieved PD. Among them, the ORR and DCR of HR + /HER2- MBC patients using Palbociclib in combination with ET as first-line treatment were 5.83% and 48.54%, respectively. Interestingly, the ORR and DCR of second-line and later-line treatments were 0.0% and 28.57%, and 0.0% and 12.50%, respectively. Detailed tumor evaluations for the different treatment groups are listed in Table 2.

The most prevalent grades 3–4 adverse reaction was neutropenia (38.79%), which had a lower incidence compared with that of the Paloma series studies but comparable with previous studies (Table 3). After using Palbociclib combined with ET for a period of time, 65.42% (140/214) of HR + /HER2- MBC patients experienced disease progression. Herein, electronic medical records were reviewed, and telephone follow-ups were conducted to gain a better understanding of their treatment regime following disease progression. After excluding lost-to-follow-up and death cases, 42.97% (55/128) were observed to choose chemotherapy as the next treatment modality, with Abraxane being the most commonly selected chemotherapeutic drug. Additionally, 32.03% (41/128) chose to continue with endocrine therapy, while 17.19% (22/128) chose targeted therapy. The patients' survival time was also extended to varying degrees (Fig. 1).

After conducting a univariate analysis on the training dataset, the following variables were noted to be statistically significant (P < 0.05): age, PR, Ki-67, molecular subtype, adjuvant radiotherapy or endocrine therapy, late-stage chemotherapy, sensitivity to endocrine therapy, number of visceral metastases, presence of liver metastases, and number of Palbociclib treatment lines. After incorporating the aforementioned parameters into a multivariate Cox proportional hazards regression model, patients aged < 60 years old, PR < 20%, Ki-67 ≥ 20%, luminal B molecular subtype, primary resistance to ET, receipt of late-stage chemotherapy, and presence of liver metastasis or ≥ 2 visceral metastases were identified as independent prognostic factors associated with poor PFS (P < 0.05) (Table 4, Supplementary Fig. 1).

A nomogram model was established based on the independent prognostic factors identified from multivariate analysis. As displayed in Fig. 2, the specific scores for the eight variables were summed to obtain a total score, which was then used to determine the corresponding PFS rates of 6 months, 12 months, and 18 months. In addition, the study examined the discriminative performance of the nomogram model via AUC. On the one hand, the AUC values for PFS at 6, 12, and 18 months were 0.771, 0.783, and 0.790 in the training set, respectively (Fig. 3A), indicating the strong predictive ability of the model. On the other hand, the AUC values for PFS at 6, 12, and 18 months were 0.720, 0.766, and 0.754 in the validation set, respectively (Fig. 3B), implying favorable discriminative ability. The calibration curve also delineated a goodness-of-fit with the ideal curve (Figs. 4A-F). Moreover, decision curve analysis (DCA) was employed to compare the clinical effectiveness of various predictive models. The x-axis illustrates a range of probability thresholds for PFS representing varying levels of risk potential, while the y-axis denotes the net benefit derived from the model. The DCA denoted that the model brought clinical benefits to patients. In the training set, the DCA at 6, 12, and 18 months depicted clinical benefit within the risk threshold of 0.15–0.65, 0.30–0.90, and 0.55–0.90, respectively. In the validation set, the DCA at 6, 12, and 18 months showed clinical benefit within the risk threshold of 0.30–0.65, 0.25–0.85, and 0.30–0.95, respectively (Figs. 5A-B).