Clinico-Immunological Effects of a Single-Agent CDK4/6 Inhibitor in Advanced HR+/HER2− Breast Cancer Based on a Window of Opportunity Study

Background: Cyclin-dependent kinase 4/6 inhibitors (CDK4/6 i), abemaciclib, palbociclib, and ribociclib, have been FDA-approved for the treatment of hormone receptor-positive (HR+), HER2−negative (HER2−) advanced breast cancer (aBC). This targeted therapy has revived hope in those aBC patients who did not respond to standard therapies. Interestingly, when administered as a single agent, CDK4/6 modulated several peripheral blood cells after a short-course treatment of 28 days. However, the impact of these immune effects has yet to be thoroughly investigated. Methods: We administered abemaciclib, palbociclib, and ribociclib monotherapy to 23 patients with HR+/HER2− metastatic breast cancer. The aim is to investigate the impact of on-treatment modifications on peripheral blood cells and their composite scores in patients after a 28-day course of CDK4/6 i alone. Results: In the current study, we observed a significant decrease in neutrophils (p-value < 0.001) for patients treated with abemaciclib, palbociclib, and ribociclib. An overall decrease of Tregs was observed and potentially linked to palbociclib treatment. The neutrophile to lymphocyte (N/L) ratio was also decreased overall and potentially linked to abemaciclib and palbociclib treatment. Platelets were decreased in patients administered with abemaciclib. Notably, the radiometabolic response was available only for those patients treated with ribociclib and abemaciclib, and only those lesions treated with ribociclib reached statistical relevance. Conclusions: Our study strongly supports the notion that CDK4/6 inhibitors induce tumour immune modulation. N/L ratio and platelet levels decreased due to treatment. Future studies should test whether patients would benefit from immunomodulators in association with CDK4/6 agents in a larger clinical trial. Moreover, the CDK4/6-induced immune modulation could also be considered a potential predictive clinical factor in HR+/HER2− advanced breast cancer.


Introduction
Breast cancer (BC) is the most diagnosed cancer in women worldwide, with over 2.2 million new cases in 2021 [1]. Two out of three cases of BC include hormone receptorpositive (HR+) and human epidermal growth factor receptor 2-negative (HER2−) histol-offer robust homogeneity compared to TILs. Only a few studies have observed that the neutrophil-to-lymphocyte ratio (NLR) and the platelet-to-lymphocyte ratio (PLR) before neoadjuvant treatment, mostly including taxanes and anthracyclines, are associated with responsiveness and prognosis [24][25][26].
The present study aimed to explore the biological effect of short abemaciclib, palbociclib, and ribociclib (28 days of continuous treatment for abemaciclib, 21 days of treatment and one week off for palbociclib and ribociclib) on circulating immune cells and platelets, in 23 advanced ER+/HER2− BC patients. Additionally, we investigated the radiometabolic effect of the aforementioned chemotherapeutic agents on 9 out of 23 patients (24 lesions in total). The data collected could provide a reference base for exploring other biomarkers to streamline treatment decisions in the future by enabling early identification of patients responding to therapy.

Study Design and Eligibility Criteria
From January 2019 to February 2020, 23 treatment-naïve patients diagnosed with metastatic Luminal A and B breast cancer (stage IV, according to the AJCC 8th edition) eligible for systemic therapy were enrolled at the ASST of Cremona Hospital. Only those patients with an Eastern Cooperative Oncology Group (ECOG) performance status of 1 or below, with good baseline cardiac, renal, hepatic, and haematological functions, were recruited for the study. The study was conducted following the Declaration of Helsinki, Good Clinical Practice principles, and all local regulations. The study (Mozart Trial EUDRACT 2018-002116-26) was granted ethical approval by the ethical committee of the ASST of Cremona Hospital (IRB Approval 27.12.18 n 433), and all patients involved provided written consent.

Study Objective and Endpoints
This study's main objective was to investigate the biological effects of the CDK4/6 i administered alone in a window of opportunity fashion in metastatic BC patients [17,27]. The primary endpoint was the exploratory evaluation of early changes in circulating immune cells for the overall cohort of patients (n = 23) induced by each agent under investigation. Secondary endpoints included the experimental assessment of radiometabolic (PET/TC) response to either abemaciclib, palbociclib, or ribociclib short course.

Study Treatments and Procedures
Patients were assigned to orally receive either abemaciclib (n = 8, 34.8%), palbociclib (n = 6, 26.1%), or ribociclib (n = 9, 39.1%) at a dose of 150 mg/bid daily, 125 mg/daily, and 600 mg daily, respectively (28 days of continuous treatment for abemaciclib, 21 days of treatment and one week off for palbociclib and ribociclib) before starting with the standard combination with letrozole 2.5 mg/daily [17,27]. Blood sampling was performed at baseline and after 28 days of treatment. Treatment was stopped in cases of unacceptable toxicity or imaging confirmation of disease progression. Toxicities were classified according to the CT-CAE version 4.03. All patients enrolled in the study experienced at least one adverse event (AE), and all except 1 patient in the 600 mg dose group experienced at least one grade 3 AE. The most frequent AEs were hematologic, followed by nausea and increased laboratory values. The most frequent AEs of any grade were hematologic, a finding consistent with other studies of CDK4/6 i. The most frequent treatment-related ≥3 grade AEs were leukopenia (82%), neutropenia (72%), and lymphopenia (69%). However, these AEs were generally mild and none of them individually occurred in more than two patients enrolled in the trial. All events resolved without clinical action. None of the patients discontinued treatment due to an AE. FDG 18 -PET/CT scan was conducted at baseline and after 28 days of treatment ± 2 days. Patients were considered responsive to agents under investigation when a reduction of SUV max was significant after 28 days of treatment. Conversely, patients were considered nonresponsive when stability or increase in SUV max was observed according to PERCIST criteria. The same radiologist evaluated PET/CT responses at baseline and after 28 days of treatment.

Flow Cytometry Analysis
Flow cytometry analyses were used to investigate the circulating immune cells via whole blood samples, which were taken before and following either abemaciclib, palbociclib, or ribociclib administration. The use of Becton Dickinson (BD) FACSCanto TM and BD FACSCanto II TM with BD TM Cytometer Setup and Tracking (CS&T) control enabled signals to be reproducible and comparable in spite of any fluctuation in environmental conditions. BDFACSC Diva Software was used for the acquisition and assessment of at least 1.5 × 10 6 . The BD Multitest 6-color TBNK kit (BD) was used for the assessment of the following subpopulations: lymphocytes B, natural killer (NK), and T cells with CD4 and CD8 subpopulations.

Statistical Analysis
The distribution of continuous variables was inspected for normality using the Shapiro-Wilk test. Data were reported using the mean and standard deviation (SD) if normally distributed or median and range (minimum-maximum) if not. Categorical variables were described by frequency and proportion. Parameters of interest, including SUV, the overall count of lymphocytes (CD45+), count of lymphocytes B (CD19+), NK (CD16+CD56+), T (CD3+), T Helper (CD3+CD4+), T Suppressor (CD3+CD8+), T regs (CD3+CD4+CD25++CD127−), neutrophils, platelets, CD4/CD8, N/L (neutrophil to lymphocytes) and P/L (platelets to lymphocytes) ratios before and following 28-day treatment, were compared using the Student's t-test and the Wilcoxon matched-pairs test for continuous data. We used waterfall plots to provide an overview of how patients responded to the therapies. Statistical analysis was achieved by R software (version 4.1.21, R Core Team (2020). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria, URL https://www.R-project.org/ (accessed on 1 February 2022). A p-value < 0.05 was deemed to be significant, although formal comparisons were only exploratory.

Patients and Tumour Characteristics
A total of 23 patients were enrolled. The median patient age was 62.8, and surgery technique, namely, either conservative or mastectomy as initial treatment, was equally distributed. All patients were diagnosed with metastatic stage IV disease. In case of relapse, metastasis was biopsied for biomarker evaluation. In terms of molecular subtype, 16 (69.6%) patients carried the Luminal B subtype (ER-positive and/or PR-positive tumours with negative HER2 and low Ki67 proliferation (<14%)), while 7 (30.4%) patients carried the Luminal A subtype (ER-positive and/or PR-positive tumours with positive/negative HER2 and high Ki67 proliferation (≥14%)) [28]. All metastases were hormone-sensitive. Eight patients were administered with abemaciclib (34.8%), six patients received palbociclib (26.1%), and nine were administered with ribociclib (39.1%). At present, 14 patients (60.9%) are still alive, while 9 (39.1%) are deceased (median follow-up of 9.5 years). Patients' characteristics are summarised in Table 1.

Overall Treatment Effect on Immune Circulating Cells and Platelets
Immune circulating subpopulations and platelets were evaluated at baseline and after 28 days of treatment ( Figure 1 (from an average of 250.8 cells/µL to 228.7 cells/µL), CD4/CD8 ratio (from an average of 1.74 cells/µL to 1.56 cells/µL), and N/L ratio (neutrophil to lymphocytes) (from an average of 2.23 to 1.00). Contrastingly, the P/L ratio (platelets to lymphocytes) was increased (from an average of 167.6 cells/µL to 1784.2 cells/µL). Among these values, only those regarding T regs , neutrophils, and the N/L ratio were reported to be statistically significant, with p-values of 0.02, <0.001, and <0.001, respectively (Table 2).
Contrastingly, the P/L ratio (platelets to lymphocytes) was increased (from an average of 167.6 cells/μL to 1784.2 cells/μL). Among these values, only those regarding Tregs, neutrophils, and the N/L ratio were reported to be statistically significant, with p-values of 0.02, <0.001, and <0.001, respectively (Table 2).

Treatment Effect on Immune Circulating Cells and Platelets According to the Therapeutic Agent
Immune circulating subpopulations and platelets were further evaluated at baseline and after 28 days according to each specific treatment. In agreement with previously reported results, we observed an overall decrease-after 28-day therapy for each drug-in the number of T regs , neutrophils, platelets, and N/L ratio. However, T regs and platelets were found to be decreased at a statistically significant level in those patients administered with palbociclib (from an average of 59 cells/µL to 44 cells/µL, p = 0.04) and abemaciclib (from an average of 224.4 cells/µL to 187.8 cells/µL, p = 0.03, Figure 2). Interestingly, neutrophils were decreased at a statistically significant level in patients administered with all treatments under evaluation, namely, abemaciclib (from an average of 3.32 cells/µL to 1.32 cells/µL, p = 0.02), palbociclib (from an average of 3.31 cells/µL to 1.46 cells/µL, p = 0.03), and ribociclib (from an average of 3.45 cells/µL to 1.39 cells/µL, p = 0.008). Additionally, the N/L ratio was also reported to be significantly decreased in those patients administered with abemaciclib (from an average of 2.07 cells/µL to 0.84 cells/µL, p = 0.02) and palbociclib (from an average of 2.48 cells/µL to 1.11 cells/µL, p = 0.03) ( Table 3). No statistically significant difference was observed for other immune cells.

Radiometabolic Biomarkers According to Therapeutic Response
Radiometabolic response at baseline and after 28 days (28 days of cont treatment for abemaciclib, 21 days of treatment and one week off for palbocicl ribociclib) was available for only 9 out of 23 patients and a total of 24 malignant l Four of these patients were administered with abemaciclib (10 malignant lesions), a were issued with ribociclib (14 malignant lesions), as reported in Table 4. Ove significant decrease in median SUVmax was reported among all 24 lesions from ba The median SUVmax measure decreased from an average value of 4.8 at diagnosis to the end of treatment (p = 0.006). Regarding the specific treatment, only those treated with ribociclib reached statistical significance (p = 0.008), with a decrease in m

Radiometabolic Biomarkers According to Therapeutic Response
Radiometabolic response at baseline and after 28 days (28 days of continuous treatment for abemaciclib, 21 days of treatment and one week off for palbociclib and ribociclib) was available for only 9 out of 23 patients and a total of 24 malignant lesions. Four of these patients were administered with abemaciclib (10 malignant lesions), and five were issued with ribociclib (14 malignant lesions), as reported in Table 4. Overall, a significant decrease in median SUV max was reported among all 24 lesions from baseline. The median SUV max measure decreased from an average value of 4.8 at diagnosis to 2.9 at the end of treatment (p = 0.006). Regarding the specific treatment, only those lesions treated with ribociclib reached statistical significance (p = 0.008), with a decrease in median SUV max from an average value of 4.75 at diagnosis to 2.9 at the end of treatment. Those lesions treated with abemaciclib, although showing a decreasing trend for SUV max , did not achieve statistical relevance (Figure 3).

Discussion
Despite the number of studies already published, no clear indication of a role for peripheral systemic immunity has been observed so far in patients administered with CDK4/6 i, partly due to the lack of standardized cut-off values and the limited number of patients enrolled.
In our study, we first confirmed that CDK4/6 i-based therapy resulted in a considerable decrease in peripheral blood white cells (primarily neutrophils), as previously observed in phase II and phase III studies investigating CDK4/6 i agents [15,[29][30][31][32][33]. The overall decrease in total white blood cell and neutrophil counts is a wellestablished and common adverse event (AE) of palbociclib, ribociclib, and abemaciclibbased therapy [10,15,[29][30][31][32]. This side effect is likely due to the CDK6 inhibition of transcription modulation of the EGR1 gene, which results in the impairment of hematopoietic stem cell proliferation and differentiation in bone marrow [34]. In a pooled safety analysis (including the PALOMA 1, 2, and 3 trials) 1% of patients reported grade 3-4 lymphopenia and the level of any grade 5-year lymphopenia was 2.8% [10,11,35]. In the MONALEESA 2 trial, grade 3 and 4 lymphopenia rates were 7% and 0.9%, respectively [12]. However, the overall impact of CDK4/6 i-based therapy on peripheral blood counts in PALOMA, MONARCH, and MONALEESA trials has not been accurately discussed,

Discussion
Despite the number of studies already published, no clear indication of a role for peripheral systemic immunity has been observed so far in patients administered with CDK4/6 i, partly due to the lack of standardized cut-off values and the limited number of patients enrolled.
In our study, we first confirmed that CDK4/6 i-based therapy resulted in a considerable decrease in peripheral blood white cells (primarily neutrophils), as previously observed in phase II and phase III studies investigating CDK4/6 i agents [15,[29][30][31][32][33]. The overall decrease in total white blood cell and neutrophil counts is a well-established and common adverse event (AE) of palbociclib, ribociclib, and abemaciclib-based therapy [10,15,[29][30][31][32]. This side effect is likely due to the CDK6 inhibition of transcription modulation of the EGR1 gene, which results in the impairment of hematopoietic stem cell proliferation and differentiation in bone marrow [34]. In a pooled safety analysis (including the PALOMA 1, 2, and 3 trials) 1% of patients reported grade 3-4 lymphopenia and the level of any grade 5-year lymphopenia was 2.8% [10,11,35]. In the MONALEESA 2 trial, grade 3 and 4 lymphopenia rates were 7% and 0.9%, respectively [12]. However, the overall impact of CDK4/6 i-based therapy on peripheral blood counts in PALOMA, MONARCH, and MONALEESA trials has not been accurately discussed, and only a single case study describes a patient who developed neutropenia, thrombocytopenia, and anaemia following a palbociclib regimen [36]. It is worth mentioning that lymphopenia is a robust predictor of chemotherapy-induced toxicity and is also a predictive factor of treatment efficacy in lung, breast, and colon cancer [37][38][39]. It has been observed that palbociclib and ribociclib have higher selectivity towards CDK4 and CDK6 than abemaciclib, although the latter demonstrated higher offtarget activity and approximately 13 times more specificity for CDK4 than CDK6 [40,41]. For this reason, the incidence of neutropenia is more commonly observed in patients treated with abemaciclib. Unlike chemotherapy-induced neutropenia, CDK4/6 i-induced neutropenia is not correlated with an increased risk of developing severe infections, possibly because CDK4/6 i does not result in irreversible damage and apoptotic cell death in white blood cell precursors in the bone marrow [42]. On the contrary, CDK4/6 i can enable cytotoxic activity by mitigation of myeloid-cell-mediated immunosuppression [43]. In light of the above, our analysis is an emerging study describing the significant reduction of peripheral blood counts following single-agent abemaciclib, palbociclib, and ribociclib administration.
A decline in neutrophil-to-lymphocytes (N/L) ratio was also observed during this study, mainly in the abemaciclib and palbociclib groups. No statistically significant dif-ferences were observed for the ribociclib group. It was recently observed that peripheral immunity and inflammation balance indicators such as the N/L ratio might be a predictive response marker and correlate with good outcomes [24]. Vernieri et al. recently showed that the N/L ratio was associated with response to platinum agents in metastatic triplenegative breast cancer patients [44]. Further studies have investigated the N/L ratio in breast cancer patients following neoadjuvant chemotherapy [45][46][47]. Overall, a low N/L ratio potentially suggests a systemic background of decreased immune activation and resulting inflammation, leading to better treatment response [24].
A decrease in the overall count of peripheral regulatory T cells (T regs ) was also observed. To the best of our knowledge, this is one of the first studies observing an in vivo correlation between CDK4/6 i administration and a decrease in peripheral T regs in patients diagnosed with ER+/HER2− breast cancer. The ability of CDK4/6 i to decrease T reg count has been described only in preclinical studies [48]. In this study, we observed a significant decrease in T reg levels, particularly in patients administered palbociclib. Overall cohort data reported a reduction of T regs cells, particularly the effector T regs subset (CD3+CD4+CD25++CD127−), which is involved in immunosuppressive functions.
Ultimately, we would conclude that ribociclib showed the most significant effect on tumour shrinkage-according to the radiometabolic response, whilst no statistically significant reduction was reached in those lesions treated with abemaciclib. We cannot exclude that tumour shrinkage induced by abemaciclib treatment did not reach statistical relevance due to the limited number of lesions under investigation. However, abemaciclib appeared to be the CDK4/6 with the strongest effect, according to its ability to reduce neutrophil, N/L ratio, and platelet levels; this is likely due to its higher selectivity for CDK4 and CDK9 [49].
With regards to thrombocytopenia, although we observed decreased platelet counts among the entire cohort of patients, this was most apparent in patients treated with abemaciclib. Thrombocytopenia is another emerging side effect of CDK4/6 i treatment reported in PALOMA−2, PALOMA−3, MONALEESA, and MONARCH−3 trials. More specifically, any-grade thrombocytopenia was observed in 15-25% of patients in PALOMA−2 and PALOMA−3 trials, with grade 3 and grade 4 thrombocytopenia observed in 1 to 2% and 0.2 to 1% of patients, respectively [29,30]. The incidence of any-grade thrombocytopenia was 9.3 to 11% in MONALEESA trials, with grade 3 and grade 4 thrombocytopenia observed in 1% of patients [33,50]. As for abemaciclib, any grade thrombocytopenia occurred in 36.2 to 53.2% of patients in MONARCH−2 and MONARCH−3 trials, with grade 3 and grade 4 thrombocytopenia reported in 1.3 to 2% and 0.6 to 1.4% of patients, respectively [13,15].

Conclusions
Collectively, these data strongly support the idea that CDK4/6 inhibitors induce tumour immune modulation. Moreover, when considering the critical need for the identification of patients with either a good treatment response or resistance to CDK4/6 agents, CDK4/6 immune modulation should be evaluated as a novel and robust predictive factor. Furthermore, modifications of peripheral immune subsets and tumour-infiltrating lymphocytes have been found to be closely related to one another, and immune profiling might become a crucial tool in guiding treatment choice.
Despite some limitations, including the limited sample size, long follow-up, and data concerning LDH as an indicator of tumour burden, the study brings some important knowledge to the field, showing that there are immunological effects correlated to the use of CDK4/6 inhibitor as a single agent. Furthermore, to overcome the aforementioned limitations, these results must be investigated in large-scale studies with extensive follow-up and assessment of other inflammatory markers.

Institutional Review Board Statement:
The studies involving human participants were reviewed and approved by the Ethical Committee of the ASST of Cremona Hospital. The patients/participants provided their written informed consent to participate in this study.
Informed Consent Statement: Informed consent was obtained from all subjects involved in the study.

Data Availability Statement:
The original contributions presented in the study are included in the article. Further inquiries can be directed to the corresponding authors.