Palbociclib and Trastuzumab in HER2-Positive Advanced Breast Cancer: Results from the Phase II SOLTI-1303 PATRICIA Trial

This article is featured in Highlights of This Issue, p. 5773

The availability of trastuzumab, pertuzumab, and T-DM1 has markedly improved the survival outcome of advanced HER2⁺ breast cancer (1–3). However, metastatic HER2⁺ breast cancer is still a life-threatening disease (4–8). Thus, new treatment strategies and molecular biomarkers to select patients better, are much needed.

The recent introduction of cyclin-dependent kinase 4 and 6 inhibitors (CDK4/6i) in combination with endocrine therapy has had a significant impact in the treatment landscape of advanced hormone receptor (HR)⁺/HER2⁻ disease (9–14). These drugs are well tolerated, orally administered, and have demonstrated an increase in overall survival (15–17). To date, however, the role of these drugs in HER2⁺ disease is less clear (18–21).

Preclinical data suggest that CDK4/6i are active in HER2⁺ cell lines (22, 23). In addition, Cyclin D1 expression mediates resistance to HER2-targeted therapies (24, 25) and this mechanism is overcome through CDK4/6i (26, 27). In transgenic mouse models of inducible HER2-driven mammary carcinomas, the combination of anti-HER2 therapy and CDK4/6 inhibition is more potent than either agent alone and can resensitize resistant HER2⁺ breast cancer to anti-HER2 therapies (26). Therefore, there is a strong rationale to evaluate the activity of CDK4/6i in patients with HER2-positive breast cancer.

The main purpose of the SOLTI-1303 PATRICIA study (NCT02448420) was to evaluate the efficacy and safety of palbociclib plus trastuzumab in postmenopausal pretreated patients with HER2⁺ metastatic breast cancer. Cohort A included patients with estrogen receptor (ER)-negative disease. Cohorts B1 and B2 included patients with ER⁺ disease treated without (B1) or with (B2) letrozole, respectively. We also performed gene expression profiling to characterize those patients who benefit the most from this new combination.

This is an open-label, multicenter, phase II study. Eligible patients were postmenopausal, had unresectable HER2⁺ metastatic breast cancer and had documented investigator-assessed progression after treatment with 2–4 regimens in the advanced setting, at least one including trastuzumab. Tumors (primary or metastatic) were required to have locally determined HER2 overexpression as defined by the current ASCO/CAP criteria. Patients with nonmeasurable or measurable disease according to Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1 were enrolled. Additional eligibility criteria included age of 18 years or older, a left ventricular ejection fraction (LVEF) of 50% or higher, an Eastern Cooperative Oncology Group performance status of 0–1 and an adequate organ function. All enrollment criteria are detailed in Supplementary Data (Supplementary Materials and Methods).

All patients provided written informed consent before the initiation of any study-related treatment or procedures, and the protocol was approved by the Ethics Committees from all participating institutions and Spanish Health Authorities. The study was conducted in accordance with Good Clinical Practice principles, the Declaration of Helsinki, and all local regulations.

All patients received palbociclib (200 mg per day orally for 2 weeks followed by 1 week off; refs. 28, 29) and trastuzumab (loading dose of 8 mg/kg followed by a maintenance dose of 6 mg/kg i.v. or 600 mg by subcutaneous injection every 21 days). The 2-week on/1-week off schedule was chosen to align the treatment with palbociclib with the administration of trastuzumab. Two dose reduction levels of palbociclib were permitted: 150 mg and 100 mg. Patients with ER⁺ disease were randomized to treatment without or with letrozole 2.5 mg/day (cohorts B1 and B2, respectively).

Patients had study visits and routine laboratory assessments on days 1, 8, and 15 for the first cycle; the day 8 and 15 visits were optional for subsequent cycles in the absence of significant toxicities. Tumor assessments were performed at baseline, within 28 days of randomization, and thereafter every 9 weeks until disease progression or discontinuation of the study treatment and at the end of the study. Efficacy assessments were performed locally using radiologic scans, including CT or MRI. Safety was monitored throughout the study.

Treatment with palbociclib with trastuzumab, with or without letrozole, continued until radiographic disease progression, unacceptable toxicity, withdrawal of consent, or investigator's decision to discontinue treatment. End-of-treatment assessments were performed 28 days after the last dose of the study drug or before initiation of a new antineoplastic therapy, whichever occurred first.

The modified intention-to-treat analysis population was defined as all patients who had at least one postbaseline tumor assessment after started the treatment. The primary objective was progression-free survival (PFS) rate at 6 months (PFS6) within each cohort, defined as the proportion of patients in the modified intention-to-treat population who were progression-free at 6 months according to RECIST version 1.1 assessed by the investigator. Secondary endpoints included toxicity according to the National Cancer Institute Common Terminology Criteria for Adverse Events version 4.0, PFS, overall response rate (ORR; complete plus partial response rate), and PFS and ORR according to the PAM50 subtypes (30).

Because of the lack of phase I data on the proposed regimen, the trial included a safety run-in phase with an intensified cardiac and hematologic safety monitoring of 6 patients treated with trastuzumab and palbociclib, and 6 patients treated with trastuzumab, palbociclib, and letrozole. This safety phase is fully described in Supplementary Materials and Methods. The preplanned safety interim analysis was evaluated by an independent data monitoring committee (independent medical oncologists, cardiologist, doctor in internal medicine, and statistician), which deemed the trial safe and supported its continuation to full recruitment.

A minimum of approximately 100 ng of total RNA was used to measure the expression of 55 breast cancer–related genes using the nCounter platform (NanoString Technologies), including the 50 genes of the PAM50 subtype predictor, androgen receptor, and 4 immune genes (CD4, CD8, PD1, and PDL1). Data were normalized using 5 housekeeping genes, and log₂ transformed. Intrinsic molecular subtypes were identified using the research based PAM50 predictor as described previously (31–33). Methods for RNA extraction, and quality assessment can be found in Supplementary Data (Supplementary Materials and Methods).

The study was based on a Simon two-stage design comprising 3 cohorts: ER⁻ (cohort A), ER⁺ without letrozole (cohort B1), and ER⁺ with letrozole (cohort B2), as illustrated by the schema in Supplementary Fig. S1. For each cohort, a Simon two-stage optimal design (34) was used to determine whether palbociclib in combination with trastuzumab had enough activity to warrant further development. Target accrual was a minimum of 15 patients (stage I) and a maximum of 46 patients (stage 1 and 2 combined) in each cohort. For a targeted agent where high benefit rates are expected, a true PFS6 rate of ≤ 30% would be considered unacceptable (null hypothesis), whereas a true PFS6 rate of ≥ 50% would merit further study (alternative hypothesis). In stage 1, if 5 or less patients out of 15 were progression-free at 6, the cohort should be closed. Otherwise, 31 additional patients would be accrued (stage 2) for a total of 46 patients per cohort. For the overall trial, the null hypothesis would be rejected for each cohort separately if at least 18 patients were progression-free at 6 months in each cohort. This design controls type I error rate at 5% in each cohort and generates 80% power for detecting active cohorts.

Analysis of the primary endpoint included the modified intention-to-treat population. PFS was summarized using Kaplan–Meier curves. Cox regression modeling using stepwise selection was used to determine the effect of various baseline covariates on PFS. OR with a 95% confidence interval (CI) were estimated using univariate and multivariable logistic regression analyses. Fisher exact tests were used for the difference in ORR between subtypes. All statistical tests were two-sided and considered significant when P ≤ 0.05. All statistical analyses were performed using the R v3.2.3 software.

From July 2015 to November 2018, 72 patients were recruited across 17 centers in Spain (Fig. 1). A data cut-off date of July 11, 2019 was used for all reported analyses. Two patients with ER⁻ were included in Cohort B1 and B2 by mistake and were therefore included in these cohorts for the analyses. Patient characteristics are listed in Table 1. The median number of prior lines of systemic therapy for metastatic disease was 3 (Supplementary Table S1 details the prior oncologic treatments for metastatic or locally advanced disease), most patients (95.8%) had received prior therapy in metastatic setting with another anti-HER2 in addition to trastuzumab, including pertuzumab (50.7%), lapatinib (50.7%), T-DM1 (60.7%), or neratinib (8.5%).

In stage 1 (n = 45), the PFS6 rate was 33.3% (5/15) in cohort A, 40.0% (6/15) in cohort B1 and 53.3% (8/15) in cohort B2. Per protocol, cohorts B1 and B2, but not cohort A, were successful (i.e., achieved a PFS6 rate of ≥ 40% in the first 15 patients). For this reason, recruitment proceeded to stage II for cohorts B1 and B2, and was halted for cohort A. Following the analysis of the stage I results, especially those related to the PFS benefit in the patients with PAM50 Luminal tumors, recruitment of the stage II part was halted, as there was a significant protocol amendment (see below). The PFS6 rate in cohorts B1 and B2 after recruiting 56 of the 92 (61%) expected patients was 42.9% (12/28) and 46.4% (13/28), respectively (Table 2; Fig. 2).

The median PFS was 4.2 months (95% CI, 0.7–20.2) in cohort A, 6.0 months (95% CI, 4.1–27.1) in cohort B1, and 5.1 months (95% CI, 3.7–9.1) in cohort B2 (Table 2). No statistically significant differences in PFS was observed among the 3 cohorts (Supplementary Fig. S2). Finally, the ORR in 59 patients with measurable disease was 15.5% (95% CI, 8.0–26.0) in all patients: 6.6% (95% CI 0.2–31.8) in cohort A, 14.3% (95% CI 4.0–32.7) in cohort B1, and 21.4% (95% CI 8.3–41.0) in cohort B2 (Table 2).

Neutropenia was the most common hematologic treatment-related adverse event (AE) across the three cohorts. Grade 3 or higher hematologic treatment-related AEs occurred in 64.8% of patients, and grade 3 or higher neutropenia and thrombocytopenia in 63.4% and 11.3% of patients, respectively. Febrile neutropenia occurred in 3 (4.2%) patients. Grade 3 or higher nonhematologic treatment-related AEs occurred in 7% of patients, most commonly G3 asthenia in two patients. No clinically significant cardiovascular toxicity was observed (Table 3).

Thirty-nine (54.9%) patients had one or more dose reductions (Supplementary Table S2), mostly due to neutropenia (22.6%). Drug-related AEs that led to study drug discontinuation occurred in 2 (2.8%) patients (1 patient with neutropenia and 1 patient with thrombocytopenia). No deaths were observed during the study.

Gene expression was performed in 59 (83.1%) tumor samples. These samples came from primary tumors (54.2%), metastatic tumors (42.4%), or unknown site (3.4%). Only 4 samples were collected after the last treatment before the patient was included in the trial. Baseline characteristics of this subset of patients were like those of the overall study cohort (Supplementary Table S3). The 49.2% of the tumor samples were identified as HER2-enriched (HER2-E; Supplementary Fig. S3A), followed by Luminal B (22.0%), Luminal A (16.9%), normal-like (10.2%), and Basal-like (1.7%). As expected, the distribution of the intrinsic subtypes differed on the basis of HR status (Supplementary Fig. S3B and S3C; P < 0.001). Of note, no luminal subtype was identified in ER⁻ disease, and the HER2-E subtype was identified in 36.4% (n = 16) and 86.7% (n = 13) of ER⁺ and ER⁻ disease, respectively.

Median PFS was 10.6 months (95% CI, 4.1–14.8) in Luminal B, 8.2 months (95% CI, 2.2-0.24.1) in Luminal A, 4.3 months (95% CI, 2.1–27.1) in HER2-E, and 3.7 months (95% CI, 1.7–11.2) in normal-like tumors (Fig. 3A). Only 1 patient had ER-negative/Basal-like disease and experienced disease progression at 21 days following palbociclib and trastuzumab (cohort A). Significant differences in PFS curves were observed between intrinsic subtypes (P < 0.001 by log-rank test). The previous analysis suggested that PAM50 Luminal versus non-luminal disease had distinct PFS outcomes, and this was confirmed pooling patients with Luminal A/B tumors versus patients with non-Luminal tumors: median PFS 10.6 versus 4.2 months, respectively (P = 0.009 by log-rank test; Fig. 3B).

In the multivariate analysis, PAM50 Luminal subtype vs. non (adjusted hazard ratio = 0.35; 95% CI, 0.16–0.76; P = 0.008) and number of prior lines of systemic therapy for metastatic [2–3 lines vs. 4 lines; adjusted hazard ratio = 0.48; 95% CI, 0.25–0.95; P = 0.035) were found associated with PFS] after adjusting for age, ECOG, HR status, and visceral disease (Supplementary Table S4).

Like the overall population, PAM50 Luminal intrinsic subtype remained significantly associated with PFS in patients with ER⁺ disease (P = 0.001 by log-rank test; Fig. 3C and D; Supplementary Table S5). Median PFS was 10.6 months (95% CI, 4.1–14.8) in Luminal B, 8.2 months (95% CI, 2.2–24.1) in Luminal A, 3.8 months (95% CI, 2.1–10.9) in HER2-E, and 6.0 months (95% CI, 1.7–11.2) in normal-like. No patients with Basal-like disease were identified in cohorts B1 and B2. In the univariate analysis, patients with luminal disease had better PFS than patients with non-luminal tumors (median PFS 10.6 vs. 4.2 months; hazard ratio = 0.32; 95% CI, 0.25–0.98; P = 0.002; Fig. 3D). Age, ECOG, visceral disease, cohort, and previous hormone therapy in metastatic setting were not found significantly associated with PFS (Supplementary Table S5).

The highest ORR was observed in the Luminal A (20.0%) and Luminal B (23.1%) subtypes, followed by normal-like (16.6%) and HER2-enriched (10.3%) subtypes (P = 0.707; Supplementary Table S6). No statistically significant differences were observed between luminal and non-luminal tumors (21.7% vs. 12.9%; OR 2.15; 95% CI, 0.51–9.06, P = 0.296). Within ER⁺ disease, the ORR of luminal and non-luminal tumors were 21.7% and 14.3%, respectively (OR = 1.66; 95% CI, 0.35–8.03, P = 0.525).

In November 2018, after reviewing the preliminary results of the stage I of the trial (i.e., the marked differences in PFS between patients with Luminal and non-Luminal disease), the Steering Committee of the trial decided to stop enrollment and open a new cohort including only patients with ER⁺, PAM50 Luminal A or B tumors. The cohort C is now randomizing 232 patients with pretreated (minimum 1 and up to 4 prior lines) HER2⁺/ER⁺ and PAM50 Luminal A or B disease, to receive palbociclib, trastuzumab, and endocrine therapy versus physician's choice of treatment (i.e., T-DM1 or single agent chemotherapy plus trastuzumab) (Supplementary Fig. S4). Patients previously recruited in cohort B2 and with a PAM50 Luminal subtype will not be considered for the final analysis of cohort C. The primary objective of this cohort is to demonstrate the superiority of palbociclib, trastuzumab, and endocrine therapy in terms of PFS, assuming a PFS in palbociclib arm of 10 months versus 6.2 months in control arm (hazard ratio = 0.62 for the experimental group). This cohort begun recruitment in August 2019.

In the SOLTI1303-PATRICIA study, palbociclib in combination with trastuzumab showed clinical activity in HER2⁺ metastatic breast cancer. Of note, the study population had received a median of 3 previous regimens, and patients were required to have had disease progression on trastuzumab in the advanced setting. Interestingly, a secondary analysis showed an association of longer median PFS in the luminal subtypes compared with the non-luminal disease, including durable disease control for up to 12 months in 39% of these patients. These findings, together with the importance of developing well-tolerated non-chemotherapy-based regimens, are important therapeutic progress in the context of heavily pretreated HER2⁺ patients.

These results, together with the strong preclinical and early clinical data support dual inhibition of CDK4/6 and HER2 in ER⁺/HER2⁺ breast cancer (19, 20, 27, 35). The phase III PATINA trial (36) is testing the value of adding palbociclib to trastuzumab (± pertuzumab) and endocrine therapy maintenance after induction therapy in the first-line setting. The phase II monarcHER trial (21) compared abemaciclib and trastuzumab, with or without fulvestrant, to chemotherapy plus trastuzumab. The triplet combination showed a statistically significant improvement in PFS compared with chemotherapy and trastuzumab, with a median PFS of 8.3 versus 5.7 months, respectively. Consistently with the SOLTI-1303 PATRICIA trial, results from monarcHER suggest that the addition of endocrine therapy to CDK4/6 inhibition plus trastuzumab might improve outcome in these patients.

The safety profile of CDK4/6i is generally consistent, with the primary toxicity being clinically manageable myelosuppression. The rate of drug-related discontinuation in PATRICIA was low (3%), suggesting an overall good toxicity profile. Most of toxicities induced by palbociclib can be managed with drug hold and reduction. Although myelosuppression was common, the clinical sequelae were rare, including 3 patients with uncomplicated febrile neutropenia and no grade 3/4 hemorrhage episodes were reported. Of note, we used a 2-week on/1-week off schedule in PATRICIA to align palbociclib administration with trastuzumab infusion. However, a more standard 3-week on/1-week off schedule of palbociclib will be used moving forward.

All the four main intrinsic subtypes of breast cancer can be found in HER2⁺ tumors (37–40), but to date this has no impact on patient clinical management. In SOLTI 1303-PATRICIA study, more than half of HR⁺/HER2⁺ tumors were found to be Luminal A or B, and these patients had a statistically significant longer PFS than patients with non-luminal disease (median PFS 10.6 vs. 4.2 months).

In HER2⁺ disease, evidence for the ability of the intrinsic subtype to predict prognosis and/or treatment benefit is mostly confined to early breast cancer (37–47). Albeit, similar evidence in advanced breast cancer was previously limited to a single retrospective PAM50 analysis of 157 HR⁺/HER2⁺ tumor samples from the first-line EGF30008 study (32). This trial evaluated the effect of adding lapatinib to letrozole as first-line treatment. In this study, all intrinsic subtypes seemed to benefit to some degree from lapatinib therapy based on the estimate of the hazard ratio. However, the non-luminal subtype was significantly associated with poor PFS in patients treated with or without lapatinib. This result, together with that observed in the PAM50 subpopulation of SOLTI 1303-PATRICIA, confirms the non-luminal subtype to be a consistent and poor prognostic biomarker for HR⁺/HER2⁺ advanced breast cancer.

CDK4/6i efficacy observed depending on intrinsic subtype is aligned with other previous reports in HER2⁻ disease. Finn and colleagues (48) have recently published the exploratory results of PFS according to the intrinsic subtype in 455 patients recruited in PALOMA-2. As expected, patients with Luminal subtypes tumors benefited substantially from palbociclib plus letrozole versus placebo plus letrozole. In contrast, the patients with non-luminal subtypes tumors had small absolute benefit from palbociclib. Similar results were seen in the PEARL trial (49), which randomly assigned 601 patients with HR⁺/HER2⁻ advanced breast cancer to palbociclib plus endocrine therapy versus capecitabine. In this trial, patients with non-luminal tumors that received chemotherapy had a significantly longer PFS when compared with palbociclib plus endocrine therapy (13.7 months vs. 2.7 months). In contrast, patients with luminal tumors had a median PFS of 9.3 and 11.0 months in the palbociclib and the capecitabine arms, respectively.

The observation that patients with Luminal/HER2⁺ tumors benefit from palbociclib is intriguing. Prior work (50) has revealed that Luminal/HER2⁻ tumors have genomic and genetic alterations similar to those of Luminal/HER2⁺ tumors except for the HER2 amplicon, which is only overexpressed/amplified in those Luminal tumors that are HER2⁺. Overall, these data suggest that the non-luminal subtypes do not benefit much to CDK4/6 inhibition within both HR⁺/HER2⁻ and HR⁺/HER2⁺ disease, and Luminal A or B subtypes benefit substantially from CDK4/6 inhibitors.

This study has several limitations. First, it is a single-arm, open label study and the difference in PFS between luminal and non-luminal subtypes could be due to inherent prognosis. However, given the differences in PFS that were observed, if there is a benefit of palbociclib in non-luminal tumors, this is likely to be rather small from an absolute perspective. Second, the termination of enrollment prior to completion of the targeted accrual limits the final conclusions. Third, the dosing schedule for palbociclib was not standard, which could have led to higher rates of hematologic toxicity compared with current recommended dosing schedule. For this reason, we have re-defined dosing schedule for current cohort C with the 125 mg dose. Fourth, the genomic correlative analyses did not include all specimens of the study (16.9% of samples were not evaluable). Moreover, 54% of the samples profiled were from primary tumors, rather than metastatic specimens, and only 4 samples were collected right before the patient was included in the PATRICIA trial. Although we cannot predict how many samples would have shifted subtype and what the findings would have been if the analysis had been done on baseline samples only, the reality is that metastatic tissues are not always available in clinical practice and it is currently unknown if the prognostic and/or predictive associations would have been improved if metastatic tumors had been profiled. Finally, a small number of genes was evaluated in the gene expression analysis. Therefore, we were limited regarding the ability to derive new gene signatures and identify new biological processes associated with treatment response.

In conclusion, SOLTI-1303 PATRICIA is the first trial to prospectively show benefit of palbociclib plus anti-HER2 therapy with trastuzumab in HER2-positive advanced breast cancer. On the basis of these findings, a new cohort of 232 patients is being enrolled to test the hypothesis that in patients with HER2⁺/ER⁺ and PAM50 luminal disease, palbociclib, trastuzumab, and endocrine therapy is superior to physician's treatment of choice.

E. Ciruelos reports personal fees from Pfizer and nonfinancial support from Pfizer during the conduct of the study; personal fees from Roche, Lilly, AstraZeneca, Novartis, and MSD outside the submitted work. P. Villagrasa reports personal fees from NanoString and grants from Pfizer during the conduct of the study. M. Oliveira reports grants from Pfizer (for the institution) during the conduct of the study; grants, personal fees, and nonfinancial support from Roche, Novartis; grants and personal fees from Genentech, Seattle Genetics, AstraZeneca, PUMA Biotechnology; grants from Immunomedics, Boehringer-Ingelheim; nonfinancial support from Eisai, Grunenthal, GP Pharma, and Pierre Fabre outside the submitted work. S. Pernas reports nonfinancial support from Novartis (travel grant); personal fees and nonfinancial support from Roche (consulting and travel grant); personal fees from Daiichi Sankyo (consulting), AstraZeneca (lecture), and Polyphor (consulting) outside the submitted work. S. Escrivá-de-Romaní reports grants from Pfizer (for the institution) during the conduct of the study; grants, personal fees, and nonfinancial support from Roche (for the institution); personal fees and nonfinancial support from Pierre-Fabre; nonfinancial support from Daiichi Sankyo; personal fees from Eisai, Kyowa Kirin, Esteve; and grants from Synthon (for the institution) outside the submitted work. J. Cortés reports personal fees and other from Roche (advisor/honoraria); personal fees from Celgene (advisor/honoraria), AstraZeneca (advisor), Cellestia (advisor), Biothera (advisor), Merus (advisor), Seattle Genetics (advisor), Daiichi Sankyo (advisor/honoraria), Erytech (advisor), Athenex (advisor), Polyphor (advisor), Lilly (advisor/honoraria), Servier (advisor), Merck Sharp & Dohme (advisor/honoraria), GSK (advisor), Leuko (advisor), Bioasis (advisor), Clovis Oncology (advisor), Boehringer (advisor), Novartis (honoraria), Eisai (honoraria), Pfizer (honoraria), Samsung Bioepis; and other from MedSIR (stock) outside the submitted work (travel, accommodation, expenses: Roche, Novartis, Eisai, Pfizer, Daiichi Sankyo). N. Martínez reports personal fees from Novartis, Eisai, Imegen, Roche, and Pfizer during the conduct of the study. B. Bermejo reports other from Pfizer (congres travels); personal fees from Roche (advisory), MSD (advisory), and Eisai (advisory) outside the submitted work. E. Vega reports grants from HM Sanchinarro during the conduct of the study. P. Nuciforo reports other from SOLTI [institution received funding for the conduct of the trial (central lab)] during the conduct of the study. B. González-Farré reports personal fees from SOLTI (medical monitoring of the trial) during the conduct of the study. A. Prat reports grants and personal fees from Pfizer and NanoString Technologies during the conduct of the study; personal fees from Roche, Novartis, AstraZeneca, Daiichi Sankyo, Seattle Genetics, Guardant Health, Foundation Medicine, BMS, MSD, AbbVie, and Oncolytics Biotech outside the submitted work. No potential conflicts of interest were disclosed by the other authors.

E. Ciruelos: Conceptualization, resources, formal analysis, supervision, funding acquisition, investigation, writing-review and editing. P. Villagrasa: Conceptualization, resources, data curation, formal analysis, writing-review and editing. T. Pascual: Resources, data curation, investigation, visualization, methodology, writing-original draft, writing-review and editing. M. Oliveira: Conceptualization, resources, investigation, writing-review and editing. S. Pernas: Conceptualization, resources, validation, writing-review and editing. L. Paré: Resources, data curation, software, visualization, writing-review and editing. S. Escrivá-de-Romaní: Resources, investigation, writing-review and editing. L. Manso: Resources, investigation, writing-review and editing. B. Adamo: Resources, investigation, writing-review and editing. E. Martínez: Resources, investigation, writing-review and editing. J. Cortés: Supervision. S. Vazquez: Investigation, writing-review and editing. A. Perelló: Investigation, writing-review and editing. I. Garau: Investigation, writing-review and editing. M. Melé: Investigation, writing-review and editing. N. Martínez: Investigation, writing-review and editing. A. Montaño: Investigation, writing-review and editing. B. Bermejo: Investigation, writing-review and editing. S. Morales: Investigation, writing-review and editing. M.J. Echarri: Investigation, writing-review and editing. E. Vega: Investigation, writing-review and editing. B. González-Farré: Investigation, visualization, methodology, writing-review and editing. D. Martínez: Validation, investigation. P. Galván: Investigation, visualization. J. Canes: Data curation, project administration. P. Nuciforo: Investigation, writing-review and editing. X. Gonzalez: Conceptualization, investigation, writing-review and editing. A. Prat: Conceptualization, resources, data curation, supervision, funding acquisition, investigation, visualization, writing-review and editing.

The authors thank Pfizer for their provision of palbociclib and their financial contribution to this clinical study. They also thank the patients and their families/caregivers for their participation. This work was supported by grants from Pas a Pas and Save the Mama to A.P. Fundación SEOM (SEOM 2018 grant: Fellowship for Training in Research in Reference Centers; to T. Pascual).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.