Effect of BRCA mutational status on survival outcome in advanced-stage high-grade serous ovarian cancer

At our institutional hospital, we recommend germline BRCA1/2 gene testing to all women with pathologically proven EOC in accordance with the position statements of the Korean Society of Gynecologic Oncology [20]. Previously, germline BRCA1/2 gene testing was performed using direct sequencing (Sanger sequencing), including whole exon and exon-intron boundaries of BRCA1 and BRCA2. However, since February 2016, next-generation sequencing (NGS) of BRCA1 and BRCA2 genes have been available and actively used. Germline mutations, discovered by NGS, were validated by Sanger sequencing. To determine whether the detected mutations were previously identified, we searched the Breast Cancer Information Core (BIC) database (https://​research.​nhgri.​nih.​gov/​projects/​bic/​), the National Institutes of Health open-access database of clinically observed variants and their classification (ClinVar) (https://​www.​ncbi.​nlm.​nih.​gov/​clinvar/​), and the previously published Korean germline BRCA1/2 mutations [21‐25]. In this study, patients with BRCA1 or BRCA2 gene sequence variants which were previously reported as deleterious mutations or classified as likely pathogenic or pathogenic according to the American College of Medical Genetics and Genomics and the Association for Molecular Pathology guidelines [26] were assigned to the BRCA mutation group. Otherwise, they were assigned to the wild-type BRCA group.

From our institution’s Ovarian Cancer Cohort Database, we searched relevant patients to identify those who met the following inclusion criteria: (1) diagnosed with EOC and treated between July 2008 and December 2017; (2) International Federation of Gynecology and Obstetrics (FIGO) stage III-IV disease; (3) High-grade serous histologic type; and (4) received germline BRCA1/2 gene testing at the time of or within 2 years from cancer diagnosis. Patients with insufficient clinical and/or pathologic data were excluded from the study. Figure 1 depicts the selection of the study population. Of the 128 patients with advanced-stage HGSOC who met these criteria, 51 (39.8%) and 77 (60.2%) were included in BRCA mutation group and wild-type BRCA group, respectively. Thereafter, we compared clinicopathological characteristics and clinical outcomes between the two groups.

Three representative domains of the BRCA1 protein are as follows: 1) the N-terminal Really Interesting New Gene (RING) domain (exons 2–7); 2) Exons 11–13, that covers over 65% of the sequence of BRCA1; and 3) the BRCA1 C-terminal (BRCT) domain (exons 16–24) [19]. Considering these three domains, 37 patients of the BRCA mutation group were divided into three subgroups depending on the position of BRCA1 mutations, and their survival outcomes were compared.

We reviewed medical records to collect information about clinicopathological characteristics, such as age and serum CA-125 levels at diagnosis, and FIGO stage, and primary treatment of EOC, such as use of NAC and extent of residual tumour after debulking surgery. Optimal debulking was defined when the maximal diameter of residual tumour was less than 1 cm. Whether the patients received NAC followed by interval debulking surgery or primary debulking surgery (PDS), all patients received post-operative adjuvant taxane- and platinum-based chemotherapy as the primary treatment. However, none of them received any PARP inhibitors. Patients’ personal and familial histories at the time of HGSOC diagnosis were retrieved from medical records. We collected the patients’ parity, menstruation status, personal and familial history of breast cancer and ovarian cancer, and the number of relatives (up to second-degree family members) who had a history of breast or ovarian cancers.

PFS was calculated as the time interval between the date of initial diagnosis and the date of disease progression using the Response Evaluation Criteria in Solid Tumours version 1.1 [27]. Treatment-free interval (TFI) was calculated as the time interval between the date of completion of primary treatment and the date of disease progression. Platinum-sensitive recurrence (PSR) was defined when the TFI was 6 months or longer, whereas platinum-resistant recurrence (PRR) was defined when the TFI was shorter than 6 months. Overall survival (OS) was calculated as the time interval between the date of initial diagnosis to the date of cancer-related death or end of the study.

Differences in clinicopathological characteristics were evaluated between the two groups: Student’s t-test and Mann-Whitney U-test were used to compare continuous variables, and Pearson’s chi-squared test and Fisher’s exact test were used to compare categorical variables. The Kaplan-Meier methods with log-rank test were used for survival analyses. Hazard ratios (HRs) and 95% CIs were calculated using Cox proportional hazards regression models. We used SPSS software (version 21.0; SPSS Inc., Chicago, IL, USA) for these analyses. All statistical tests were two-sided, and a P value < 0.05 was considered statistically significant.

The patients’ clinicopathological characteristics are presented in Table 1. No differences in age and serum CA-125 levels at the time of HGSOC diagnosis, use of NAC, and rates of optimal debulking surgery were observed between the BRCA mutation group and the wild-type BRCA group. The time interval between diagnosis and genetic test was not different either. Distributions of FIGO stage IIIA1 to IVB were not different between the two groups (P = 0.077). However, FIGO stage IV disease were less common in the patients who have BRCA mutation compared to those with BRCA wild-type (25.5% vs. 49.4%, P = 0.007). In the BRCA mutation group, 37 (72.5%) and 14 (27.5%) patients had mutation in BRCA1 or BRCA2, respectively; no patients had double mutations.

The patients’ personal and familial histories are presented in Additional file 1: Table S1. There were no differences in parity, menstruation status, and familial history of ovarian cancer between the two groups. However, patients with BRCA mutations had significantly higher personal history of breast cancer (23.5% vs. 7.8%, P = 0.012) and familial history of breast cancer (27.5% vs. 2.6%, P < 0.001) than those in the wild-type BRCA group. The number of relatives with breast cancer was also higher in the BRCA mutation group (P < 0.001).

Additional file 1: Table S2 depicts clinicopathological characteristics of patients according to the primary treatment strategy. In the BRCA mutation group, age and serum CA-125 levels at diagnosis, and residual tumour after surgery were similar between patients who received PDS and those with NAC. However, FIGO stage IV disease were more frequent in patients with NAC (47.4% vs. 12.5%, P = 0.009). In the wild-type BRCA group, patients with NAC had higher initial serum CA-125 levels (median, 1946.0 vs. 764.0, P = 0.011) and showed a trend towards more FIGO stage IV disease (63.3% vs. 40.4%, P = 0.050), compared to those with PDS. However, proportions of patients who achieved optimal debulking surgery were not different; 86.7 and 78.7% of patients who received NAC and PDS, respectively (P = 0.378).

The median observation period was 26.3 months (range, 8.1–94.4 months). During this time, 31 patients (60.8%) in the BRCA mutation group and 58 patients (75.3%) in the wild-type BRCA group experienced disease recurrence. Among them, the median TFI was longer in the patients with BRCA mutations (12.3 months vs. 9.0 months, P = 0.002). However, the proportions of those with PSRs among the recurred were similar between the two groups (80.6% vs. 63.8%, P = 0.099).

Survival outcomes for the BRCA mutation and wild-type BRCA groups are presented in Fig. 2. There were no significant differences in OS between the two groups (5-year survival rates, 75.1% vs. 66.4%, P = 0.257). By contrast, patients in the BRCA mutation group had significantly longer PFS than those in the wild-type BRCA group (median, 21.7 vs. 15.4 months, P = 0.001). In terms of specific BRCA gene type, the patients who had BRCA1 mutation and those who had BRCA2 mutation showed no differences in PFS (median, 21.7 vs. 26.7 months, P = 0.612).

Multivariate analyses adjusting age, FIGO stage, primary treatment strategy, residual tumour after debulking surgery, and BRCA mutational status revealed only NAC (compared to PDS) as an independent poor prognostic factor for OS (adjusted HR, 4.098; 95% CI, 1.478–11.359; P = 0.007) (Table 2). Meanwhile, the BRCA1/2 mutation was identified as an independent favorable prognostic factor for PFS (adjusted HR, 0.502; 95% CI, 0.318–0.795; P = 0.003), and NAC was also associated with poor PFS (adjusted HR, 2.103; 95% CI, 1.321–3.348; P = 0.002) (Table 3).

Next, we compared the survival outcomes in all patients according to the primary treatment strategy (Additional file 1: Figure S1). Patients who received NAC had significantly poorer survival outcomes than those who received PDS (OS, P = 0.003; and PFS, P < 0.001). We performed subgroup analyses considering BRCA mutational status. In the wild-type BRCA group, compared to PDS, patients who received NAC had poorer PFS (median, 14.2 vs. 16.9 months, P = 0.003), whereas no differences in OS were observed (Fig. 3 a, b). In the BRCA mutation group, patients who received NAC showed poorer OS (5-year survival rates, 57.9% vs. 82.8%; P = 0.040). However, PFS was not different between NAC and PDS treatments (median, 17.2 vs. 26.7 months; P = 0.082) (Fig. 3 c, d).

Lastly, we also performed subgroup analyses limited to the specific primary treatment strategy. In PDS group, multivariate analyses revealed that BRCA mutational status did not affect both OS and PFS (Additional file 1: Table S3). However, in NAC group, the BRCA1/2 mutation was identified as an independent favorable prognostic factor for PFS (adjusted HR, 0.433; 95% CI, 0.202–0.926; P = 0.031). Suboptimal deublking was a poor prognostic factor for PFS (adjusted HR, 3.753; 95% CI, 1.294–10.890; P = 0.015) (Additional file 1: Table S4).

Detailed deleterious BRCA1 gene mutations of 37 patients in the BRCA mutation group are displayed in Additional file 1: Table S5. The most frequently mutated domain was exons 11–13, in which 24 (64.9%) patients were included. Of the others, 6 (16.2%) and 7 (18.9%) patients had mutations in the N-terminal RING domain and BRCT domain, respectively. All 6 mutations in the N-terminal RING domain were the same nonsense mutation. In survival analysis, the three subgroups showed similar OS (P = 0.643) and PFS (P = 0.963) (Additional file 1: Figure S2).