Introduction

BRCA1 and BRCA2 genes were discovered three decades ago via analysis of breast-ovarian cancer pedigrees. Subsequent studies revealed that they render essentially similar elevation in breast cancer (BC) and ovarian cancer (OC) risk. Hence, both these genes are considered almost equivalent when discussing the surveillance for healthy carriers or the feasibility of prophylactic surgery. Inactivation of both BRCA1 and BRCA2 is associated with homologous recombination deficiency (HRD) and, consequently, tumor sensitivity to PARP inhibitors, platinum compounds, and other drugs capable of inducing DNA double-strand breaks. In practical terms, BRCA1/2 genes are commonly discussed as a single entity, and the differences between these two genes remain underappreciated.

Still, there are essential dissimilarities between BRCA1 and BRCA2 with regard to their biological functions, the spectrum of associated cancer types, and the nuances of tumor presentation. While both proteins are involved in DNA repair by homologous recombination, BRCA1 has a number of additional properties like ubiquitin-ligase activity and phosphorylation-dependent interaction with other proteins. It participates in the regulation of other than homologous recombination DNA repair pathways, cell cycle, transcription, R loop resolution, etc. [1]. Interestingly, BRCA1 appears to be essential for taxane-mediated apoptosis, whereas BRCA2 is not [2, 3]. Mutations in both BRCA1 and BRCA2 predispose to BC and OC, while a role in the development of prostate and pancreatic malignancies has been convincingly shown mainly for BRCA2 but not for BRCA1 [4–6]. With regard to BC, BRCA1 is generally associated with receptor triple-negative disease, while BRCA2-related tumors usually express steroid hormone receptors (HRs) [7]. Furthermore, several clinical studies revealed differences between BRCA1- and BRCA2-mutated hereditary tumors with regard to the efficacy of various drug regimens and overall disease outcomes [8–11].

Neoadjuvant chemotherapy (NACT) is a commonly used approach aimed at reducing tumor volume before surgery. NACT is highly informative for clinical studies, as it involves relatively uniform groups of treatment-naive patients [12]. Furthermore, the post-NACT tumor pathologic response is regarded as the most informative surrogate of the true efficacy of a given drug combination [13]. This study aimed to compare NACT outcomes in BRCA1- vs. BRCA2-associated BC and OC patients.

Materials and methods

The collection of hereditary high-grade serous ovarian carcinomas (HGSOCs) included 63 consecutive patients who carried BRCA1 or BRCA2 mutations and were subjected to NACT at the N.N. Petrov Institute of Oncology in the years 2019–2024. The majority of these patients were participants of a prospective clinical trial comparing the efficacy of mitomycin C plus cisplatin doublet vs. standard paclitaxel plus carboplatin regimen in BRCA1/2 mutation carriers (NCT04747717, https://clinicaltrials.gov/). There were 48 patients with germline BRCA1 mutations and 15 carriers of BRCA2 pathogenic alleles (Table 1 and Table S1). Of these, six patients (three cases with BRCA1 mutation and three cases with BRCA2 mutation) did not undergo debulking surgery. The remaining 57 patients underwent interval cytoreduction with omentectomy. The WT (wild-type) arm included 93 consecutive germline BRCA1/2 mutation-negative HGSOC patients eligible for NACT, who were treated with a standard paclitaxel plus carboplatin regimen within the years 2019–2024.

Pathologic responses were evaluated with the standard three-tier chemotherapy response score (CRS), which is routinely utilized for the assessment of NACT outcomes in women with HGSOC [14]. It is commonly accepted that CRS1 corresponds to no or minimal tumor response, CRS2 refers to multifocal or diffuse regression-associated changes with readily identifiable residual tumors, and CRS3 corresponds to a complete (no tumor cells) or near-complete response with minimal tumor foci up to a maximum of 2 mm [14].

BC study included consecutive patients who were treated with NACT within the years 2022–2024. Steroid HR status is a strong confounding factor for the response to NACT in BC, therefore, all comparisons considered both patients with BRCA1/2 mutations and women with sporadic carcinomas. ER, PgR, and HER2 receptor status were determined as a part of regular BC histopathological diagnosis using treatment-naive tumor tissue. HER2-positive cases were excluded from the analysis as they all received anti-HER2 therapy. In total, the study included 105 BRCA1 or BRCA2 mutation carriers and 150 consecutive cases without germline pathogenic BRCA1/2 variants. The majority of patients received standard anthracycline-taxane combinations. Carboplatin was added to the treatment scheme for some patients with BRCA1/2-associated and/or triple-negative disease, based on the physicians’ choice. 17 out of 255 patients received doxorubicin plus cyclophosphamide instead of taxane-containing regimens due to the preference of the doctor or the patient (Table 2).

BRCA1/2 full-length analysis covered coding regions of the genes and splice sites. DNA testing was performed using targeted NGS (next-generation sequencing) [15, 16].

The analysis of NACT efficacy in BC patients usually relies on the residual cancer burden (RCB) score, where RCB 0 indicates pathologic complete response (pCR), RCB I corresponds to minimal residual disease, RCB II reflects moderate residual disease, and RCB III refers to extensive residual tumor [17]. Detailed patient characteristics are presented in Table 2 and Table S2.

Statistical analysis was performed using SPSS Statistics (version 26.0). Tumor response frequencies were compared by Fisher’s exact test and Mantel-Haenzsel method. All statistical comparisons were two-tailed, with a 5% threshold (P = 0.05) for statistical significance. Odds ratios (ORs) were calculated using the logistic regression function.

Results

Receptor status rather than BRCA1 or BRCA2 mutation is a strong predictor of BC sensitivity to NACT

We further analyzed the frequency of pCR in BCs with distinct BRCA status. The analyzed groups were well-balanced with regard to tumor stages (Table 2). pCRs were observed in 38/69 (55%) BRCA1 mutation carriers vs. 13/36 (36%) BRCA2-associated cancers (Fisher’s exact test, two-tailed, P = 0.1) or 43/150 (29%) BRCA-WT tumors (Fisher’s exact test, two-tailed, P = 0.0003). However, this comparison is of limited value because the above groups differed significantly from each other with regard to the proportion of patients with receptor triple-negative disease [BRCA1: 59/69 (86%); BRCA2: 10/36 (27%); WT: 76/150 (51%), Fisher’s exact test, P < 0.0001]. It is well established that triple-negative BCs are significantly more sensitive to NACT than tumors expressing steroid hormone receptors [11, 25]; therefore, we performed the comparison of the groups with different receptor statuses. The response rates in triple-negative patients were consistently higher than in HR+/HER2– patients across all analyzed subgroups [BRCA1: 35/59 (59%) vs. 3/10 (30%); BRCA2: 5/10 (50%) vs. 8/26 (31%); WT: 31/76 (41%) vs. 12/74 (16%); Mantel-Haenzsel test, P < 0.001].

It is of notice that the proportion of patients with unsatisfactory response to NACT (RCB III) was similarly high in both BRCA2-mutated [7/26 (27%)] and WT [21/74 (28%)] HR+ BCs (Figure 1 and Table S3). Overall, pCR was observed in 71/145 (49%) triple-negative vs. 23/110 (21%) HR+ BCs (Fisher’s exact test, two-tailed, P < 0.0001). The OR for achieving pCR was higher for receptor status (triple-negative vs. HR+: OR = 3.4, 95% CI 1.9–6.0, P < 0.001) than for BRCA status (any mutation vs. WT: OR = 2.1, 95% CI 1.2–3.6, P = 0.008).

Display full size

Figure 1. 

RCB distribution in triple-negative and hormone receptor-positive carcinomas. BC: breast cancer; pCR: pathologic complete response; RCB: residual cancer burden; WT: wild-type

The incorporation of carboplatin to the NACT is a common practice both in BRCA1/2-mutated patients and in women with triple-negative BC [11, 26–29]. In our dataset, 31/69 (45%) BRCA1-mutated, 11/36 (31%) BRCA2-mutated, and 26/150 (17%) WT patients received this drug. Strikingly, carboplatin did not improve pCR rates in BRCA1- or BRCA2-associated carcinomas. However, WT triple-negative BC patients receiving carboplatin showed a numerically higher frequency of pCR when compared to women treated by platinum-free regimens [14/26 (54%) vs. 15/44 (34%), Fisher’s exact test, two-tailed, P = 0.13].

Discussion

BRCA1/2 mutations account for approximately 25–30% of HGSOCs, with some studies providing even higher estimates [23, 30, 31]. BRCA1- and BRCA2-mutated HGSOCs are almost always discussed as a single entity, given that these categories of tumors do not have clear differences in clinical presentation, morphological and immunohistochemical appearance, or the pattern of chromosomal instability [32–34]. Only a few studies have compared treatment outcomes of BRCA1 vs. BRCA2 mutation carriers, and all these studies considered highly heterogeneous categories of HGSOC patients [8–10, 35]. This is the first study analyzing the immediate outcomes of NACT in BRCA1- vs. BRCA2-associated cancers. It demonstrates that BRCA2-associated HGSOCs have a higher rate of complete tumor cell elimination upon NACT when compared to BRCA1-mutated tumors. Although the size of this single-center study is small, the observed associations are in line with several other HGSOC investigations, which have demonstrated generally better survival for BRCA2 vs. BRCA1 mutation carriers [8, 9].

A number of biological mechanisms may underlie these differences. Although BRCA1/2-associated HGSOCs almost always carry a somatic deletion of the remaining allele of the gene (loss-of-heterozygosity, LOH) detected in the gross tumor mass, some tumors have a small admixture of platinum-resistant cells with a preserved WT copy of the involved BRCA gene. This explains the nature of post-NACT residual tumor masses: while short-term platinum therapy does not cause restoration of BRCA1/2 open reading frame via a second mutation, the selection of pre-existing BRCA1/2-proficient tumor cells is observed at significant frequencies [36, 37]. Complete elimination of tumor cells upon NACT can be observed in HGSOCs, which do not have intratumoral heterogeneity with regard to LOH status. While intratumoral heterogeneity has already been demonstrated in BRCA1-related tumors, its involvement in the pathogenesis of BRCA2-associated cancers requires additional studies [38]. Some data suggest that BRCA2 deficiency results in complete inactivation of DNA repair, while cells with biallelic BRCA1 mutations still retain some DNA double-strand repair capacity due to the involvement of RAD51-driven pathways. Consequently, a higher level of sensitivity to platinum compounds and PARP inhibitors is generally observed in BRCA2- vs. BRCA1-defective cells [39]. Furthermore, taxanes are commonly included in NACT schemes for HGSOCs. Some laboratory and clinical studies suggest that taxanes are not effective against BRCA1-deficient cells, whereas BRCA2 inactivation does not preclude the action of taxanes [2, 3, 40]. Our data strongly support these observations: indeed, the TCbP doublet produced only a moderate CRS3 rate in BRCA1 mutation carriers, while excellent efficacy of this regimen was observed in BRCA2-associated HGSOCs.

Our study demonstrates that BRCA1-mutated BCs generally produce higher response rates to NACT as compared to BRCA2-related BCs. However, this difference is almost entirely attributed to the high frequency of triple-negative disease in BRCA1 vs. BRCA2 mutation carriers. When tumors were stratified according to receptor status, the NACT outcomes in the BRCA1 and BRCA2 groups appeared highly similar. Importantly, while BRCA1-mutated, BRCA2-mutated, and WT triple-negative tumors showed high rates of pCR, the frequency of NACT failure (i.e., RCB III) was similar between HR+ BRCA2-associated and WT cancers. It has to be acknowledged that pCR is a strong predictor of long-term survival only in triple-negative cancers, while the absence of pCR is not a potentially fatal indicator for HR+ disease [41, 42].

The high efficacy of single-agent platinum compounds has been convincingly demonstrated in a series of studies involving women with BRCA1/2 germ-line mutations [43–45]. These data are in strong agreement with preclinical experiments [46, 47]. However, the rationale for the addition of carboplatin to conventional drug regimens is more controversial. Indeed, virtually all standard BC NACT regimens include anthracyclines. In general, platinum drugs and anthracyclines are both DNA double-strand-inducing agents and have a similar mode of action toward tumors with HRD [48, 49]. A randomized comparison of neoadjuvant cisplatin vs. doxorubicin-cyclophosphamide in BRCA1/2 mutation carriers with BC did not reveal any differences between treatment arms [50]. Our study demonstrated that the addition of carboplatin did not improve the outcomes of NACT in BRCA1-mutated or BRCA2-mutated BCs, probably due to the redundancy of mechanisms of action of the involved drugs towards BRCA-deficient cells. This is in agreement with the results of the GeparSixto clinical trial, which involved triple-negative BC patients and subjected to separate analyses BRCA1/2-related and BRCA1/2-WT cases [26]. However, similar to the GeparSixto dataset, we observed a trend towards an improved pCR rate upon the addition of carboplatin in WT patients. This is an intriguing finding that deserves the analysis of the underlying biological mechanisms. BRCA1/2-WT BCs are a heterogeneous group of tumors, with some of them possibly having targets both for carboplatin and for conventional drug regimens.

Apparently, BRCA1 and BRCA2 mutations should not be regarded as equivalent factors when considering BC therapy. In addition to the role of HR status, other parameters, such as LOH for the remaining BRCA allele and the extent of HRD should be considered in future studies. Practical implementation of the upfront LOH and HRD testing may become complicated, as these techniques require a sufficient amount of tumor material and are not always compatible with BC biopsies.

NACT studies are highly informative because they involve therapy-naive patients, therefore, there are minimum factors potentially confounding treatment outcomes [12, 13]. Furthermore, they generally involve subjects with limited tumor spread, who are potentially amenable to surgery, i.e., variations in the tumor stages and patient comorbidities are usually small. This study suggests that ovarian carcinomas demonstrate better NACT outcomes in BRCA2- vs. BRCA1-mutated hereditary ovarian carcinomas, probably due to the higher sensitivity of the former to carboplatin-paclitaxel. The opposite trend is observed in BC, which is likely to be attributed to the high frequency of triple-negative disease in BRCA1- but not BRCA2-associated BCs.

The differences between the BRCA1 and BRCA2 mutations deserve further investigation. It is highly desirable to validate the above observations in large multi-center collections of BRCA1- and BRCA2-related tumors. At least two tendencies observed in the BC data set are potentially practice-changing and, therefore, require explicit clarification. First, HR status rather than BRCA1/2 status has to be considered while discussing the feasibility of NACT. Second, the incorporation of carboplatin into the NACT scheme provides an advantage to BRCA1/2 WT but not to BRCA1/2-mutated patients. Furthermore, the identification of biological mechanisms underlying clinical dissimilarities between BRCA1- and BRCA2-related tumors is of primary importance.

In conclusion, higher efficacy of NACT is observed in BRCA2- vs. BRCA1-associated OC. Steroid HR status but not BRCA1/2 status is the strongest predictor of NACT efficacy in breast carcinomas.