Similar response rates and survival with PARP inhibitors for patients with solid tumors harboring somatic versus Germline BRCA mutations: a Meta-analysis and systematic review

Since their discovery in 1994 and 1995 respectively, there has been significant clinical interest in the tumor suppressor genes BRCA1 and BRCA2, with particular focus in recent years on targeted therapies for patients with BRCA mutated cancers. The BRCA1-encoded protein is an important DNA damage response protein that interacts with multiple sensor and effector proteins in the DNA repair pathways, and is also involved directly in the repair of double-stranded DNA breaks, and BRCA2 is primarily involved in facilitating homologous recombination repair of DNA damage [1, 2]. Defects in function therefore lead to dysfunctional chromosomal rearrangement and cellular replication. Deleterious germline mutations in the BRCA1 protein have been found to significantly increase the risk of breast cancer and ovarian cancer, up to 72 and 44% respectively by age 80, as well as increased risk of many gastrointestinal, pancreatic and prostate cancers [3, 4]. BRCA2 deleterious mutations confer a similar risk of breast and ovarian cancer, as well as pancreatic cancer, prostate cancer, stomach cancer, and melanoma as well [3, 5]. Somatic BRCA mutations which are present only in the tumor cells, have been reported to be up to 15–30% of all BRCA1/2 mutations, and can be found in various malignancies, such as 3% of breast cancer cases, and over 12% of advanced prostate cancer patients [6‐10].

Poly(ADP-ribose) polymerase (PARP) is a nuclear protein that is activated by breaks in DNA single strands, which then further recruits DNA repair proteins with synthesis of poly(ADP-ribose) chains [11]. A class of medications, the PARP inhibitors (PARPi), were developed to specifically target the DNA repair pathways involved with PARP1, as well as the other PARP enzymes. Use of PARPi was found to increase apoptosis of BRCA1 and BRCA2 mutated cells, due to a combination of increased DNA lesions and inability for defective BRCA1 and BRCA2 products to repair the DNA, an example of the genetic concept called synthetic lethality [12, 13]. Multiple PARPi have since been FDA approved for advanced ovarian cancer and/or breast cancer, such as olaparib in 2014, rucaparib in 2016, niraparib in 2017, and talazoparib in 2018 [14]. The study of PARPi has been mostly limited to patients with germline BRCA mutations and hence there is a lack of data comparing response rates to PARPi in patients with somatic versus germline BRCA mutations. Furthermore, there is a concern that somatic BRCA mutations detected in tumor tissue, may be passenger mutations instead of driver mutations, and thus targeting it may not be of utility [15]. In order to better ascertain this, we performed a meta-analysis comparing overall response rate to PARPi in patients harboring a somatic versus germline BRCA mutation. We looked at all published studies including somatic and germline mutations in BRCA patients that received PARPi.

Figure 1 highlights our data collection process. 302 studies were identified from database searches. After removal of duplicates, 42 clinical trials and 57 chart reviews/corelative studies/basic science studies were assessed further. 18 studies met our criteria for including both somatic and germline BRCA mutations and were assessed further. Only 8 studies reported ORR for both somatic and germline BRCA mutations.

The PARPi are a novel drug group that inhibit the activity of PARP1, PARP2 and PARP3, a group of proteins closely involved in the repair of single strand DNA strand breaks [11]. When these molecules are inhibited, the BRCA-mutated cells are unable to undergo the significant homologous recombination or DNA strand repair needed to fix the errors, effectively leading to arrest during the cell replication cycle, and eventually to apoptosis [12, 13]. With further characterization of BRCA mutations and development of more PARPi, therapeutic options for individuals with advanced ovarian cancers as well as certain breast cancers are increasing, with expanded use of these drugs currently undergoing investigation [14, 38]. Olaparib subsequently gained approval in metastatic breast cancer with germline BRCA mutations based on the OlympiAD trial in 2017 [38], niraparib gained approval as maintenance treatment for patients with ovarian cancer who are responding to platinum-based chemotherapy in 2017 based on the NOVA trial [33], and talazoparib was also approved in 2018 for advanced breast cancer with germline BRCA mutations [39].

Germline mutations of BRCA1 and BRCA2 are significantly associated with the development of multiple neoplasms, including breast, ovarian, stomach, pancreas, colon, and melanoma [3‐5]. However somatic BRCA mutations are under-recognized and represent a missed opportunity for further targeted therapy [6]. Many of the existing clinical trials for the PARPi don’t specifically include somatic BRCA patients, leading to underrepresentation in the data. The initial clinical trial, Study 42, leading to FDA approval for olaparib in advanced ovarian cancer, only examined germline BRCA mutations [40]. The OlympiAD phase III trial for olaparib in breast cancer by Robson et al. required a deleterious or suspected deleterious germline BRCA mutation for eligibility [38]. The phase III trial for talazoparib in advanced breast cancer by Litton et al. included only germline mutation in BRCA1/2 as well [39]. The recent POLO trial for pancreatic cancer which studied olaparib as maintenance therapy also did not include any somatic BRCA patients [41].

Exclusion of such patients from clinical trials has thus led to a lack of data of efficacy of PARPi in this subgroup. Furthermore, given concerns over whether somatic BRCA mutations are passenger or driver mutations, further data is needed to establish efficacy in this setting [15]. A potential expansion of the indication of approved medications for germline BRCA to somatic BRCA can provide care to a greater number of patients [6, 8‐10].

Our data demonstrates a comparable ORR of PARPi in somatic and germline BRCA patients. A subgroup analyses accounting for different malignancy types and different PARPi also did not reveal any significant difference in response rates. Although we were unable to perform a meta-analysis of PFS data, the reviewed literature indicates similar efficacy of PARPi between somatic versus germline BRCA in terms of PFS. Our data indicates that broader testing for somatic BRCA mutations should be considered.

Our study was limited by inconsistent reporting of PFS and ORR. The heterogeneity of the setting of use for PARPi in the studies included may limit generalizability as well, as in some studies the PARPi was used for maintenance therapy and in others for later lines of treatment.

Furthermore, we could not include several significant publications in our analysis due to exclusion of either somatic of germline BRCA patients, or no clear reporting of outcome specifically for each subgroup. For example, the POLO trial studying olaparib for pancreatic cancer did not include somatic BRCA patients and could not be included [41]. Conversely, the PROFOUND trial in prostate cancer which is looking at the efficacy of PARPi in patients with an array of somatic mutations involved in homologous recombination repair pathways does not include any germline mutation patients, and could not be included in our study [42]. Some studies are known to have included both somatic and germline patients, but do to a lack of subset analysis reported specifically for somatic BRCA patients, they could not be included for analysis, such as the QUADRA study [32].

Our study is the first to review all currently published data to compare ORR between somatic BRCA mutations and germline BRCA mutations when treated with a PARPi. Our analyses revealed a similar ORR between both cohorts. Our data would suggest that broader indications for PARPi therapy may be considered. Furthermore, our data supports the inclusion of somatic BRCA patients in future clinical trials for PARPi therapy. An example of such a study is the NIRAPANC trial enrolling metastatic pancreatic cancer patients with somatic and germline BRCA mutations, as well as other DNA repair deficiencies [43]. Notably in the COVID-19 era, PARP inhibitors, due to its per oral administration, has the added advantage of limiting patient in person healthcare visits and associated exposure, for instance compared to intravenously administered anti-cancer agents, especially with the availability of telemedicine follow-up.

The use of PARPi in BRCA-related malignancy have largely been limited to BRCA germline mutations, and thus a precise estimate of efficacy of PARPi in somatic BRCA mutation in lacking. Our meta-analysis and systematic review of the literature indicates similar response rates of PARPi therapy in patients with somatic and germline BRCA mutations. Investigation of use of PARPi therapy in a broader patient population, and the inclusion of patients with somatic BRCA mutations in further clinical trials is paramount in improving therapeutic options for our patients.