BRCA1/2, PALB2 mutations and first-line CDK4/6 inhibitor efficacy in HR+ metastatic breast cancer.

1
Introduction
Approximately 20% and 77% of breast cancers occurring in germline BRCA1 and BRCA2 pathogenic variant carriers, are hormone receptor (HR)-positive HER2-negative respectively [1]. In the metastatic setting, endocrine therapy (ET) combined with CDK4/6 inhibition (CDK4/6i) is the standard-of-care front-line treatment in the absence of visceral crisis, irrespective of germline BRCA (gBRCA) status [2,3]. Real-world data suggest that patients with BRCA1/2 or PALB2 mutations may experience a reduced benefit from ET with or without CDK4/6i [[4], [5], [6], [7], [8]]. However, biases inherent to the retrospective assessment of germline alterations can significantly affect research outcomes and clinical interpretations. To address this and ensure accurate insights, we conducted a methodologically controlled study leveraging a large, real-world database. This study assessed the outcomes of HR-positive/HER2-negative, metastatic breast cancer (MBC) women with BRCA1/2 or PALB2 pathogenic variants who received a first-line combination of endocrine therapy and a CDK4/6 inhibitor.

2
Methods
This cohort study was conducted using the Epidemiological Strategy and Medical Economics program (ESME) MBC database (NCT03275311). We selected women who initiated a combination of ET and CDK4/6i as first-line treatment of their MBC between 2013 and 2023, with data last updated in March 2024, or until death or last contact for those lost to follow-up. The database was authorized by the French data protection authority (Initial authorization No. DE-2013-117 and subsequent authorization obtained in 2019 in accordance with General Data Protection Regulation (GDPR). The analysis was approved by an independent scientific committee. Formal informed consent was not required, but the patients consented to the reuse of their electronic data. This study was performed in accordance with the ESMO Guidance for Reporting Oncology real World (ESMO GROW) guidelines.
The primary objective was to compare progression-free survival (PFS) between women with BRCA1/2/PALB2 pathogenic alterations (BRCA/PALB2m, meaning a deleterious or likely deleterious alteration was present), and those with wild-type (BRCA/PALB2wt) or untested status. Secondary objectives included comparing overall survival (OS) across these groups. PFS was defined as the time from treatment initiation to disease progression or death, while OS was defined as the time from treatment initiation to death or last contact. Breast cancer was considered hormone receptor-positive if oestrogen or progesterone receptor expression was ≥10% by immunohistochemistry, per European guidelines. BRCA/PALB2m status was recorded in the ESME database as germline, tumoral, or undetermined. Three groups were defined by BRCA/PALB2 status: 1) BRCA/PALB2m (class 4 or 5 pathogenic variants), 2) BRCA/PALB2wt (no alterations), and 3) untested.
Germline BRCA1/2 and PALB2 testing was performed according to the French national oncogenetics framework. Patients were referred through genetic consultations at our comprehensive cancer center, and testing followed local implementation of national guidelines. Testing targeted high-risk patients based including early-onset breast cancer, triple-negative breast cancer before age 51, medullary carcinoma, male breast cancer, ovarian carcinoma, and relevant family history [9]. From 2014, multigene panels for hereditary breast and ovarian cancer (HBOC) genes were progressively implemented in accredited laboratories, with broader adoption by 2018 [10].
Comparisons were made using the Kruskal–Wallis test for continuous variables and Chi-square or Fisher exact test for categorical variables. Kaplan–Meier survival curves and log-rank tests were used to estimate and compare PFS and OS. Hazard ratios (HRs) and 95% confidence intervals (CIs) were calculated using Cox proportional hazards models were adjusted for age, number of metastatic sites, presence of visceral metastases, tumor grade, and de novo metastatic status. A propensity score matching analysis was performed, matching BRCA/PALB2m patients 1:1 to BRCA/PALB2wt and untested patients using a nearest neighbor method with a caliper of 0.01 and no replacement.
To address potential immortal-time bias related to the interval between first-line initiation and genetic testing, BRCA/PALB2 status was defined at baseline (mutations known at treatment initiation), and two sensitivity analyses were performed: 1) 6-month landmark analysis (mutations identified by 6 months, excluding patients who progressed or were censored earlier), 2) time-varying analysis (Cox-time model with BRCA/PALB2 status as a time-varying variable). Statistical significance was set at p < 0.05. All analyses were performed using Stata software, version 18.

3
Results
A total of 4283 patients in the ESME database met inclusion criteria. The flowchart is presented in eFig. 1. Baseline status was categorized as BRCA/PALB2m in 80 patients (1.9%), BRCA/PALB2wt in 467 patients (10.9%), and untested in 3736 patients (87.2%). Pathogenic variants at baseline were documented as germline, tumoral, or undetermined in 56 (70.0%), 6 (7.5%), and 18 (22.5%) patients, respectively. Overall BRCA/PALB2 status, considering all testing times, was BRCA/PALB2m for 152 patients (3.5%), BRCA/PALB2wt for 1051 patients (24.5%), and untested for 3080 patients (71.9%). The median time from first-line treatment initiation to genetic testing was 0.9 months (range: −312.4 to 78.8 months). Patient characteristics by baseline BRCA/PALB2 status are summarized in Table 1. In brief, BRCA/PALB2m carriers were younger, predominantly premenopausal, and more frequently had de novo MBC compared with BRCA/PALB2wt or untested. The proportions of visceral and bone-only diseases were similar among the groups. First-line treatments primarily included aromatase inhibitors (78.6%) and palbociclib (72.1%).
With a median follow-up of 43.7 months [95%CI: 42.6–44.9], progression-free survival under first-line ET combined with CDK4/6i was significantly shorter for baseline BRCA/PALB2m carriers than for BRCA/PALB2wt and untested patients (9.9 months [7.6–13.0], 15.4 months [13.9–17.3], and 18.3 months [17.6–19.3], respectively, Fig. 1). Baseline BRCA/PALB2m status was associated with worse PFS in the multivariable analyses (adjusted HR [95% CI], 1.61 [1.24–2.09]; p < 0.001) (Table 2). Exploratory analyses according to the nature of pathogenic variants at baseline (germline, somatic, or undetermined) are presented in eFig. 2. Sensitivity analysis to address immortal-time bias confirmed consistent findings using a 6-month landmark approach or a time-varying method (Table 2). Similarly, propensity score-matching analysis showed comparable results (HR [95% CI], 2.22 [1.53–3.22] and 1.59 [1.08–2.32] for BRCA/PALB2m and untested compared with BRCA/PALB2wt, respectively) (Fig. 2). Patient characteristics according to baseline BRCA/PALB2 status are summarized for the matched patients in eTable 1. The median OS was 49.0 months [95%CI: 41.9–not reached] for BRCA/PALB2m, 50.4 months [45.2–61.4] for BRCA/PALB2wt, and 52.8 [49.9-55.0] for untested patients, with no significant differences observed among the groups (Fig. 1).

4
Discussion
This study enables a comprehensive analysis of the prognostic impact of BRCA/PALB2 alterations in patients with hormone receptor-positive, HER2-negative MBC. Our data strongly suggests an independent detrimental effect of BRCA/PALB2 mutations on PFS in HR-positive, HER2-negative MBC treated with first line ET + CDK4/6i. These conclusions are supported by a rigorous methodology including multiple sensitivity analyses to address potential immortality bias.
Our findings are consistent with five other retrospective studies involving 40 to 145 patients with BRCAm alterations, all of which reported poorer outcomes in this subgroup [[4], [5], [6], [7], [8]]. An analysis of 2968 patients from the US Flatiron Health database found a shorter time to first subsequent therapy or death among 85 gBRCAm HR+/HER2− MBC patients compared to 774 non-gBRCAm patients with a median (95% CI) OS was 26 (21-not estimated), 37 (31-51), and 33 (31-35) months, respectively. This study also included a Cox model which confirmed that the gBRCAm group had shorter time to first subsequent treatment (stratified hazard ratio [sHR] 1.24; 95% CI 0.96-1.59) and OS (sHR 1.50; 95% CI 1.06-2.14) than the gBRCAwt group [4]. Similarly, Safonov et al. reported a significantly shorter PFS in 145 BRCAm patients (germline and/or somatic) treated with first-line endocrine therapy plus CDK4/6 inhibitors (14.3 vs. 22.0 months; HR 1.95; p ≤ 0.0001) [5]. More recently, the SOLTI group presented retrospective data from 117 patients with pathogenic BRCA1/2 or other homologous recombination repair gene variants, showing significantly worse PFS for BRCA2 mutation carriers compared to matched controls (adjusted HR 2.18 [1.48–3.21]) [6]. Additional smaller cohorts have confirmed this trend [[11], [12], [13], [14], [15]].
Recent work from prospective series support these findings. The Young Pearl exploratory analysis evaluated biomarkers associated with response to palbociclib plus endocrine therapy in premenopausal women with HR+/HER2− metastatic breast cancer. Even though the number of patients with BRCA2 pathogenic mutation was small (N ¼ 3, 4%), these patients showed worse prognosis regardless PAM50 subtypes [16]. In a recent prospective study of 89 patients, comprehensive genomic and transcriptomic analyses of pre- and post-treatment tumors showed that a homologous recombination deficiency–high tumor cluster, including those with BRCA1/2 mutations, was associated with resistance to palbociclib and endocrine therapy [17].
Supporting this evidence, a recent meta-analysis of 14 studies including 618 gBRCAm patients reported significantly worse outcomes for gBRCAm patients compared to gBRCA wild type [18]. The pooled hazard ratio for progression-free survival was 1.68 (95% CI 1.37–2.05), and for overall survival was 1.73 (95% CI 1.12–2.67). Including patients with unknown gBRCA status yielded similar results, with a PFS HR of 2.02 and OS HR of 1.46. Most of the studies included in the meta-analysis were retrospective and frequently conducted at single centers, which introduces the possibility of selection bias. These studies differ in BRCA type, line of therapy, and cohort focus. Collins and Palazzo included only germline BRCA-mutated patients across various lines of CDK4/6 inhibitor therapy [4,7]. In contrast, SAFONOV and our study enrolled both germline and tumor BRCA1/2-mutated patients, focusing exclusively on first-line CDK4/6 treatment, the current standard of care [5]. Finally, the LAPENA study, also limited to first-line CDK4/6 therapy, specifically evaluated outcomes in BRCA2m carriers [6]. Additionally, the timing of BRCA testing was often not reported, raising concerns about immortal time bias. Performing genetic testing after treatment initiation may introduce survival bias, as it selectively includes patients who survive long enough to be tested, potentially confounding survival analyses. Notably, the Safonov study, which made a significant contribution to the meta-analysis, reported differing results in a secondary analysis by using a left truncation statistical method to adjust for the time gap between treatment initiation and BRCA testing. In contrast, our primary analysis included only patients with known BRCA/PALB2 status at the start of first-line treatment. To further reduce bias, we applied time-varying and landmark analyses to account for changes in mutation status over time. These approaches produced consistent findings, offering robust evidence despite the lack of prospective data.
The underlying mechanisms for this reduced benefit are not fully understood but may involve altered cell cycle regulation. gBRCA2-linked tumors are prone to developing RB1 biallelic loss as a resistance mechanism to CDK4/6i-based therapies [19]. This is driven by two key factors. First, gBRCA2-driven breast cancers are more likely to display RB1 loss of heterozygosity prior to CDK4/6i treatment, which increases susceptibility to losing the intact RB1 allele. Second, mutagenic processes in homologous recombination deficiency-driven tumors promote the emergence of RB1 loss-of-function mutations leading to biallelic inactivation via a second hit. Moreover, other explanations can be brought, such as the higher frequence of TP53 mutations in gBRCAm, contributing to genomic instability and heterogeneity that could lead to CDK4/6i resistance [20]. The loss of Erα expression in BRCA1m tumors is also associated with primary endocrine resistance [21]. Although these mechanisms of resistance mechanism may account for the shorter PFS, OS analysis showed no significant difference. This may reflect the higher number of treatment lines in gBRCA/PALB2m patients (median 3 vs. 2) and the greater prevalence of de novo disease among wild-type and untested patients.
These findings highlight the unique molecular characteristics of HR-positive HER2-negative MBC in gBRCA/PALB2 mutation carriers even when they are considered endocrine-sensitive. However, the combination of ET and CDK4/6i remains the standard first line treatment in the absence of better evidence. PARP inhibitors are recommended in the guidelines as monotherapy following first-line ET + CDK4/6i. In a recent cohort of 89 patients, those who received PARP inhibitors after progression on first-line CDK4/6i experienced a significantly longer median real-world progression-free survival (11.8 months) compared with patients treated with endocrine therapy, single-agent chemotherapy, or combination chemotherapy. This advantage was confirmed in a multivariable analysis, reinforcing PARP inhibitors as the preferred treatment option for eligible patients [22]. Promising results have emerged from combinations of ET and PARPi±immune checkpoint inhibitors [23]. The ongoing Young Paletta (NCT04819243) trial evaluates the combination of talazoparib and atezolizumab versus talazoparib alone in premenopausal women with HR+/HER2− metastatic breast cancer harboring a homologous recombination scars following progression on first-line palbociclib plus endocrine therapy. Meanwhile, the EvoPAR-Breast01 trial (NCT06380751) is investigating the combination of the PARP inhibitor saruparib with the novel SERD camizestrant compared with standard endocrine therapy plus CDK4/6 inhibition as a first-line treatment option.

5
Strengths and limitations
This study has several limitations. PFS assessment in real-world data studies can be challenging. However, in the ESME database, data are collected through independent patient file reviews ensuring consistent application of the strict PFS definition. Second, only a limited proportion of patients were tested for BRCA/PALB2 status, as the study period began before PARP inhibitors were approved in 2019, when routine testing was not yet clinically indicated. We could not completely rule out biases linked to the setting of BRCA/PALB2 testing and its potential association with specific tumor features and behaviors, as well as survivors’ biases.
This context helps explain why untested patients showed longer PFS compared with BRCA/PALB2 wild-type individuals, likely reflecting real-world selection bias, since those tested were often younger or had more aggressive disease. However, prognosis is assessed according to BRCA1/2 PALB2 status at baseline, which avoids most of these biases. Furthermore, time-varying analyses include both the onset of new mutants, wild type or untested status. Third, the majority of patients received first-line palbociclib, limiting the generalizability of the findings to other CDK4/6 inhibitors. Finally, differences in baseline characteristics, such as an imbalance in de novo status distribution across groups, may impact results. To mitigate this, we conducted several sensitivity analyses, all of which showed consistent results.

6
Conclusion
In this cohort, using careful methodology, carrier of BRCA1/2 and PALB2 pathogenic alterations showed worse PFS under standard first-line endocrine and CDK4/6 treatment than non-carriers. The potential benefit of more targeted strategies, such as first-line PARP inhibitor therapy in this population, is currently under investigation.

Access to data and data analysis
The datasets analysed for the current study were extracted from the ESME MBC database. In accordance with ethical and legal requirements related to the ESME Data Warehouse, individual data from the ESME databases cannot be made available. For any specific demand, please contact the corresponding author. Each demand will be examined on a case-by-case basis by the scientific committee.
JS Frenel and A Lusque had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

CRediT authorship contribution statement
Timothé Guinel: Writing – original draft, Project administration, Investigation, Formal analysis, Conceptualization. Amélie Lusque: Writing – original draft, Project administration, Investigation, Formal analysis, Conceptualization. Audrey Mailliez: Investigation. Vincent Massard: Investigation. William Jacot: Investigation. Severine Guiu: Investigation. Thibault de la Motte Rouge: Investigation. Etienne Brain: Investigation. Isabelle Desmoulins: Investigation. Monica Arnedos: Investigation. Caroline Bailleux: Investigation. Anthony Goncalves: Investigation. Christelle Levy: Investigation. Thomas Bachelot: Investigation. Lise Bosquet: Investigation. Suzette Delaloge: Writing – original draft, Project administration, Investigation, Formal analysis, Conceptualization. Jean-Sébastien Frénel: Writing – review & editing, Writing – original draft, Validation, Project administration, Methodology, Investigation, Formal analysis, Conceptualization.

Prior presentations
This work has been presented at the SABCS meeting as a poster in 2024.

Funding
This work was supported by 10.13039/501100021564Unicancer which independently ESME databases (i.e., data collection, analysis, and publication) and is the sole data controller for data processing The ESME MBC database receives financial support from industrial partners.

Declaration of competing interest
The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: FRENEL reports a relationship with AstraZeneca Pharmaceuticals LP that includes: consulting or advisory, speaking and lecture fees, and travel reimbursement. FRENEL reports a relationship with Novartis that includes: consulting or advisory, speaking and lecture fees, and travel reimbursement. FRENEL reports a relationship with Eli Lilly and Company that includes: consulting or advisory, speaking and lecture fees, and travel reimbursement. FRENEL reports a relationship with Pfizer that includes: consulting or advisory, speaking and lecture fees, and travel reimbursement. MAILLIEZ reports a relationship with AstraZeneca Pharmaceuticals LP that includes: consulting or advisory, speaking and lecture fees, and travel reimbursement. MAILLIEZ reports a relationship with Novartis that includes: consulting or advisory, speaking and lecture fees, and travel reimbursement. MAILLIEZ reports a relationship with Eli Lilly and Company that includes: consulting or advisory, speaking and lecture fees, and travel reimbursement. MAILLIEZ reports a relationship with Pfizer that includes:. MAILLIEZ reports a relationship with Pfizer that includes: consulting or advisory, speaking and lecture fees, and travel reimbursement. MASSARD reports a relationship with Novartis, that includes: consulting or advisory, speaking and lecture fees, and travel reimbursement. MASSARD reports a relationship with AstraZeneca Pharmaceuticals LP that includes: consulting or advisory, speaking and lecture fees, and travel reimbursement. Massard reports a relationship with Eli Lilly and Company that includes: consulting or advisory, speaking and lecture fees, and travel reimbursement. Massard reports a relationship with Pfizer that includes: consulting or advisory, speaking and lecture fees, and travel reimbursement. JACOT reports a relationship with Novartis Pharma SAS that includes: consulting or advisory, speaking and lecture fees, and travel reimbursement. JACOT reports a relationship with AstraZeneca Pharmaceuticals LP that includes: consulting or advisory, speaking and lecture fees, and travel reimbursement. JACOT reports a relationship with Eli Lilly and Company that includes: consulting or advisory, speaking and lecture fees, and travel reimbursement. JACOT reports a relationship with Pfizer Inc that includes: consulting or advisory, speaking and lecture fees, and travel reimbursement. GUIU reports a relationship with Novartis Pharma SAS that includes: travel reimbursement. DE LA MOTTE ROUGE reports a relationship with AstraZeneca Pharmaceuticals LP that includes: consulting or advisory, speaking and lecture fees, and travel reimbursement. DE LA MOTTE ROUGE reports a relationship with Novartis that includes: consulting or advisory, speaking and lecture fees, and travel reimbursement. DE LA MOTTE ROUGE reports a relationship with Eli Lilly and Company that includes: consulting or advisory, speaking and lecture fees, and travel reimbursement. DE LA MOTTE ROUGE reports a relationship with Pfizer Inc that includes: consulting or advisory, speaking and lecture fees, and travel reimbursement. BRAIN reports a relationship with AstraZeneca PLC that includes: consulting or advisory, speaking and lecture fees, and travel reimbursement. BRAIN reports a relationship with Novartis Pharma SAS that includes: consulting or advisory, speaking and lecture fees, and travel reimbursement. BRAIN reports a relationship with Eli Lilly and Company that includes: consulting or advisory, speaking and lecture fees, and travel reimbursement. BRAIN reports a relationship with Pfizer that includes: consulting or advisory, speaking and lecture fees, and travel reimbursement. ARNEDOS reports a relationship with AstraZeneca Pharmaceuticals LP that includes: consulting or advisory, speaking and lecture fees, and travel reimbursement. ARNEDOS reports a relationship with Novartis that includes: consulting or advisory, speaking and lecture fees, and travel reimbursement. ARNEDOS reports a relationship with Eli Lilly and Company that includes: consulting or advisory, speaking and lecture fees, and travel reimbursement. ARNEDOS reports a relationship with Pfizer that includes: consulting or advisory, speaking and lecture fees, and travel reimbursement. BAILLIEUX reports a relationship with AstraZeneca Pharmaceuticals LP that includes: consulting or advisory, speaking and lecture fees, and travel reimbursement. BAILLIEUX reports a relationship with Novartis that includes: consulting or advisory, speaking and lecture fees, and travel reimbursement. BAILLIEUX reports a relationship with Eli Lilly and Company that includes: consulting or advisory, speaking and lecture fees, and travel reimbursement. BAILLIEUX reports a relationship with Pfizer that includes: consulting or advisory, speaking and lecture fees, and travel reimbursement. GONCALVES reports a relationship with AstraZeneca PLC that includes: travel reimbursement. GONCALVES reports a relationship with Novartis Pharmaceuticals Corporation that includes: travel reimbursement. GONCALVES reports a relationship with Pfizer that includes: travel reimbursement. BACHELOT reports a relationship with AstraZeneca PLC that includes: consulting or advisory, speaking and lecture fees, and travel reimbursement. BACHELOT reports a relationship with Novartis Pharma SAS that includes: consulting or advisory, speaking and lecture fees, and travel reimbursement. BACHELOT reports a relationship with Eli Lilly and Company that includes: consulting or advisory, speaking and lecture fees, and travel reimbursement. BACHELOT reports a relationship with Pfizer Inc that includes: consulting or advisory, speaking and lecture fees, and travel reimbursement. DELALOGE reports a relationship with Novartis Pharma SAS that includes: funding grants and non-financial support. DELALOGE reports a relationship with AstraZeneca Pharmaceuticals LP that includes: funding grants and non-financial support. Given her role as editor, Suzette Delaloge has not been involved in the peer-review of this article and has no involvement in the peer review of this article and has no access to information regarding its peer review. Given his role as Editorial Board Member, Etienne Brain had no involvement in the peer-review of this article and has no involvement in the peer review of this article and has no access to information regarding its peer review If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.