Clinical Characteristics and Prognosis of Gastric Cancer Patients with BRCA 1/2 Germline Mutations: Report of Ten Cases and a Literature Review.

BACKGROUND: The prognosis of gastric cancer (GC) is poor with a median overall survival (OS) of less than 12 months in advanced-stage disease. The search for distinct genetic subgroups of GC patients and predictive biomarkers is ongoing. While BRCA1 or BRCA2 germline mutations (gBRCAm) have potential therapeutic implications in ovarian, breast and pancreatic cancers, their significance in GC patients has not been established. PATIENTS AND METHODS: A retrospective multi-center data analysis of GC patients with gBRCAm was conducted, detailing the clinical characteristics and disease course in this unique subset of patients.
RESULTS: Ten GC patients with gBRCAm were identified, six of them with metastatic disease. The median OS of all ten GC patients was 47.5 (13-192) months. Median OS for patients diagnosed with operable disease was 55.5 (13-192) months and of the patients with metastatic disease (calculated from metastatic disease diagnosis) 32 (15-52) months with an exceptional 1-, 2- and 3-year survival rate of 100%, 83.3% and 50%, respectively.
CONCLUSION: These preliminary data suggest that gBRCAm in GC patients are associated with a favorable prognosis. Furthermore, gBRCAm might be a predictive biomarker to DNA-damaging agents response in GC patients, similarly to its established role in other malignancies. Further research is needed to confirm our findings.

Introduction

Gastric cancer (GC) is the fifth most frequently diagnosed malignancy worldwide and the third leading cause of cancer-related death.1 The prognosis of GC remains poor with a median overall survival (OS) rate of less than a year in the advanced setting despite great efforts to find new therapeutic agents and biomarkers.2

BRCA1 or BRCA2 germline mutation (gBRCAm) carriers with breast, ovarian, prostate and pancreatic cancer demonstrate durable responses to DNA-damaging agents (platinum agents and PARP (Poly ADP-ribose polymerase) inhibitors).3–7 Increased risk for GC in gBRCAm carriers has been previously suggested,8,9 yet GC is not considered a part of the cancer spectrum in gBRCAm carriers. Targeted sequencing of gastric cancer samples detected BRCA2 somatic mutations in 8% of the cases in a Chinese cohort, regardless of patients’ personal or familial background of malignancies.10 A similar number was reported in a different publication reporting mutations in homologous recombination deficiency (HRD) genes (ATM, BRCA1, BRCA2) in 7.5% of 400 GC patients undergoing both somatic and germline next generation sequencing (NGS).11 Figer et al detected the 6174delT BRCA2 Ashkenazi Jewish founder mutation in 5.7% of 35 GC patients of Ashkenazi origin. This prevalence is higher than the 1.16% expected prevalence rate of the mutation in the general Ashkenazi population.12 Alexandrov et al identified a unique somatic molecular signature found in various tumors of gBRCAm carriers. Interestingly, this distinct molecular signature was found also in tumors not harboring a BRCA1/2 mutation, suggesting that other mechanisms impairing DNA damage repair may exist.13

To the best of our knowledge, there are only a few reports describing the clinical course of GC patients harboring BRCA1/2 germline or somatic mutations.10,14 Chen et al report prolonged OS of 45 GC patients with BRCA2 mutations compared to BRCA2 wild type (WT) patients. The report does not specify whether patients had somatic or germline mutations, stages of GC diagnosis and treatments administrated.10 Ichikawa et al recently described three gBRCAm carriers with GC; two patients with operable T4N+ tumors are alive 2 years after surgery and adjuvant treatment. The third patient is alive and continuing treatment 5 years after metastatic disease diagnosis.14

In order to evaluate further the role of gBRCAm as a prognostic and/or predictive factor in GC patients, we describe the characteristics and clinical course of a cohort of GC patients harboring these mutations.

Patients and Methods

A retrospective review of databases from high-risk onco-genetic clinics in three hospitals – Sheba Medical Center, Rabin Medical Center and Hadassah Medical Center – was conducted in order to identify gBRCAm carriers with gastric adenocarcinomas. The first genetic registry was established in 1994, hence our data cover the period between 1994 until 2018. BRCA 1/2 Jewish founder mutations testing was carried out by targeted genotyping as described by Bernstein et al.15 In some cases, testing was done by single gene sanger sequencing or by next generation sequencing. Patients’ demographics, family history, as well as histopathologic characteristics, treatments administrated and disease outcomes were obtained from medical records.

The research was approved by the local ethics committees of the three participating centers: Sheba Medical Center, Rabin Medical Center and Hadassah Medical Center. Patient data access complied with relevant data protection and privacy regulations.

Results

We identified ten GC patients with a gBRCAm. Demographic and clinicopathological characteristics of these patients are described in Table 1 and summarized in Table 2. We did not identify any unique histopathological features in stage, location, grade and histology, neither could we identify differences between BRCA1 and BRCA2 carriers.

Table 1

BRCA1/2 Germline Mutation Carriers with GC

Case	Sex	Age at GC*	Ethnicity	BRCA Mutation	First Degree Relative with BRCA Associated Malignancy^	First Degree Relative with GC	Personal History of Other Malignancy	Stage of GC at Diagnosis	Histological Type	Tumor Location	HER2 Status	Adjuvant Treatment	Recurrent Metastatic Disease (Months from Surgery)	DNA Damaging Agents for Metastatic Disease	OS (mos)	OS from Diagnosis of Metastatic Disease	Clinical Status
1	F	71	Ashkenazi	BRCA1 185delAG	Yes	No	Colon, Ovary	IA	Unknown	Missing	Negative	No	No	NA	192	NA	DNED
2	F	56	Sepharadi	BRCA2 6024dupG	Yes	No	Breast	IIA	Unknown	Body	Unknown	Platinum based Chemoradiation	Yes (13)	Yes	28	15	DWD
3	F	74	Ashkenazi	BRCA2 6174delT	Yes	No	No	IB	Intestinal	Cardia	Positive	No	Yes (29)	No	56	27	DWD
4	M	65	Ashkenazi	BRCA2 6174delT	No	Yes	No	IV	Diffuse	Body	Negative	NA	NA	Yes	43	43	DWD
5	F	56	Yemenite	BRCA2 8765delAG	Yes	Yes	Breast- Bilateral	IV	Unknown	Body	Negative	NA	NA	Yes	52	52	DWD
6	M	64	Ashkenazi	BRCA2 6174delT	Yes	No	No	IV	Intestinal	GEJ	Negative	NA	NA	Yes	36	36	AWD
7	M	63	Yemenite	BRCA2 8765delAG	Yes	No	No	IA	Unknown	Body	Unknown	No	No	NA	81	NA	ANED
8	F	52	Ashkenazi	BRCA2 6174delT	No	No	No	IIA	Mixed	Antrum	Negative	Platinum based Chemotherapy	No	NA	55	NA	ANED
9	M	69	Ashkenazi	BRCA1 185delAG	No	No	No	IV	Unknown	Cardia	Negative	NA	NA	Yes	28	28	AWD
10	M	68	Ashkenazi	BRCA1 185delAG	Yes	No	No	IB	Unknown	GEJ	Unknown	No	No	NA	13	NA	ANED

Notes: *Age at GC diagnosis. ^Breast, ovarian or pancreatic cancer.

Abbreviations: GC, gastric cancer; GEJ, gastroesophageal junction; NA, not applicable; OS, overall survival; mos, months; DNED, died no evidence of disease; DWD, dead with disease; AWD, alive with disease; ANED, alive no evidence of disease.

Table 2

Demographic, Genetic and Clinicopathological Characteristics of BRCA1/2 Carriers with Gastric Cancer

Characteristics	N=10
Gender: Male	5
Female	5
Average age at GC diagnosis (range)	63.8 (52–74)
Ethnicity: Jewish Ashkenazi	7
Other	3
Mutation: BRCA1 185delAG	3
BRCA2 6174delT	4
BRCA2 8765delAG	2
BRCA2 6024dupG	1
Personal history of malignancy	3
First degree relative with BRCA associated malignancy*	7
First degree relative with gastric cancer	2
Stage of GC at diagnosis- Operable	6
Metastatic	4
Total number of patients with metastatic GC**	6
Her2: Positive	1
Negative	6
DNA-damaging treatment for metastatic GC	5

Notes: *Breast, ovary, pancreatic cancer. **Including patients presenting with metastatic GC or recurrence after operation.

Four patients with early-stage disease underwent surgery and did not receive any adjuvant chemotherapy; one of them had recurrent metastatic disease (29 months after surgery). Two patients received adjuvant platinum-based treatment; one had recurrent metastatic disease (13 months after surgery).

Four patients had metastatic disease at first diagnosis. Among the total six patients with metastatic disease (including patients presenting with metastatic disease and patients with recurring metastatic disease after surgery), all but one were treated with DNA-damaging agents including platinum agents and/or PARP inhibitors. Treatments administrated to metastatic patients included also other chemotherapy agents, Trastuzumab, Ramucirumab and immunotherapy.

The median OS of all ten GC patients in our cohort was 47.5 (13–192) months. Median OS for patients diagnosed with operable disease was 55.5 (13–192) months. The six patients with metastatic disease had a median OS rate (calculated from metastatic disease diagnosis) of 32 (15–52) months. The 1-, 2- and 3-year survival rates with metastatic disease were 100%, 83.3% and 50%, respectively. Two patients were alive and continuing treatment with an OS of 28 and 36 months at data cutoff (Figure 1).

Figure 1

Overall survival at data cutoff.

Abbreviations: DWD, dead with disease; AWD, alive with disease.

Discussion

Although GC in gBRCAm carriers has been previously described, it is not considered a BRCA-associated malignancy.6,8,9,12,14,16,17

Identifying GC as a possible BRCA associated malignancy may have both therapeutic and diagnostic implications: BRCA1/2 mutations may predict a better response to DNA-damaging agents. In addition, the possibility of screening and early detection of GC in BRCA carriers and their family members might be considered.

To our best knowledge, this is the largest detailed cohort of GC patients harboring a pathogenic gBRCAm. Data encompass patients’ genetics, personal and familial history along with clinicopathological characteristics and most important – systemic treatments administrated and disease outcome.

Most of our patients had a BRCA2 mutation. This is in concordance with previous reports of GC found mostly in BRCA2 families.8,9,12 Interestingly, only BRCA2 somatic mutations were detected in consecutive NGS of GC specimens, with no BRCA1 somatic mutations identified.10

All of the BRCA1/2 mutations in this cohort are well known pathogenic germline mutations described previously.15,18 Three are known founder or predominant Jewish mutations that are included in a targeted panel of 14 recurring Israeli BRCA1/2 mutations test.15 Most of the patients or families in our cohort were known to be gBRCAm carriers before GC diagnosis, with two patients, both BRCA2 carriers, having a first degree relative with GC. Unfortunately, none of them were screened for GC and both were diagnosed with metastatic disease.

Most patients in our cohort had a first-degree member with a BRCA associated malignancy, suggesting the phenotype of these families is not different from the classical breast and ovarian syndrome and that GC might be a part of the syndrome.

Treatment with DNA-damaging agents (platinum and PARP inhibitors) in BRCA1/2 associated cancers including breast, ovarian, pancreatic and prostate cancers has been proved to prolong progression-free survival, and the BRCA1/2 mutation serves as a predictive biomarker to these agents.3–7

Despite great efforts to identify new treatments, the standard systemic treatment for advanced GC remains chemotherapy, adding Trastuzumab in HER2-positive patients. Median OS rate is less than a year in HER2 negative and up to 16 months in HER2-positive patients.2,19 Triple chemotherapy regimens increases the response rate, but median OS remains less than a year. The one- and two-year survival rate with triple chemotherapy is 40% and 18%, respectively.20 Second-line chemotherapy achieves a median OS of about 4 months.2 The addition of the biologic agent Ramucirumab to second-line chemotherapy extends median OS of pretreated GC patients by another 2 months.21 Immunotherapy in GC patients gives hope for prolonged responses for a subset of patients, with the predictive biomarkers of High Microsatellite Instability (MSI-H) and possibly PDL-1 expression.22–24 The addition of a PARP inhibitor to chemotherapy in metastatic GC patients was tested in the GOLD trial. The trial recruited metastatic GC patients and stratified them by the Ataxia-Telangiectasia (ATM) protein expression. The trial failed to prove benefit either in unselected patients or in the ATM negative subgroup. The publication does not include data regarding germline or somatic BRCA mutations of the patients.25 Several ongoing clinical trials are evaluating the effect of PARP inhibitors in GC (Table 3). In addition, various strategies to increase the efficacy of PARP inhibitors in GC are being investigated. Koutas et al found that c-MET inhibition increases the anti-tumor activity of PARP inhibitors in cell lines.26

Table 3

Clinical Trials with PARP Inhibitors in GC

Trial	Phase	Treatment	Patient Population	Genomic Alteration	Primary End Point	Trial Status
NCT02033551	1	Veliparib	Previously treated	BRCA mut	Safety	Completed
NCT01123876	1	Veliparib + FOLFIRI	Previously treated	Non required	Safety	Completed
NCT03026881	1	Fluzoparib + Apatinib + Paclitaxel	Previously treated	Non required	Safety	Unknown
MEDIOLA NCT02734004	1/2	MEDI4736 + Olaparib	Previously treated	Non required	Safety, DCR, ORR	Active, not recruiting
NCT03008278	1/2	Olaparib + Ramucirumab	Previously treated	Non required	Safety, ORR	Recruiting
NCT03008278	1/2	Olaparib + Ramucirumab	Previously treated	Non required	ORR	Recruiting
NCT04209686	2	Paclitaxel + Pembrolizumab + Olaparib	Previously treated	Non required	OS	Recruiting
LODESTARNCT04171700	2	Rucaparib	Previously treated	BRCA or deleterious HRR mut	ORR	Recruiting
NCT03427814	2 randomized	Pamiparib/placebo	Previously treated and responded to first-line platinum	Non required	PFS	Active, not recruiting
NCT04550494	2 randomized	Pactiltaxel + Olaparib/placebo	Progressed following first line-therapy	Known ATM status	PFS	Completed

Abbreviations: PARP, poly ADP-ribose polymerase; GC, gastric cancer; Mut, mutation; HRR, homologous recombination repair; ATM, ataxia telangiectasia; DCR, disease control rate; ORR, overall response rate; OS, overall survival; PFS, progression free survival.

The median OS of our metastatic GC patients was strikingly better than the reported OS of the general population of metastatic GC patients and is consistent with the few previous publications of better outcome of BRCA1/2 carriers with GC.2,10,14 This outcome is also superior to data from the Israeli cancer registry, that reports a median OS survival of 13.6 months for all-stage GC patients between the years 2000–2014.27

Our findings suggest that gBRCAm can be a predictive biomarker for response to DNA-damaging agents in GC patients.

Our study has several limitations that should be acknowledged: it is retrospective and includes a small number of patients. GC patients are not routinely tested for BRCA1/2 mutations. Most patients are tested only for the common Jewish mutations and do not undergo full BRCA1/2 sequencing. Patients in this cohort were treated according to the treating physicians’ preferences hence diverse protocols were employed.

The strength of our study is the detailed information including genetic and family history along with clinicopathological information and, most importantly, treatment outcomes and long term follow up.

Conclusions

This retrospective case series describes a specific subgroup of GC patients: gBRCAm carriers. We report a favorable course with a prolonged OS of metastatic GC patients harboring these germline mutations. Our findings buttress the significance of BRCA1/2 germline mutations as a tumor agnostic biomarker.

Further research is needed to understand the incidence, disease characteristics and response to various treatments in gBRCAm carriers with GC.