PARP inhibitors in breast and ovarian cancer with BRCA mutations: a meta-analysis of survival

Objective: To evaluate the efficacy of poly ADP ribose polymerase (PARP) inhibitors (PARPis) in breast and ovarian cancer with BRCA (BReast CAncer susceptibility gene) mutation (BRCAm). Methods: We conducted a meta-analysis of randomized controlled, phase II or III trials by searching of electronic databases from inception to September 1, 2020. The efficacy of PARPis measured by hazard ratios (HRs) and 95% confidence intervals (95% CIs) for progression free survival (PFS) and overall survival (OS) of patients. Results: By addition of PARPis to conventional therapy, breast or ovarian cancer patients carrying BRCAm significantly benefited PFS (breast cancer: HR 0.64, 95% CI=0.55-0.75, P<0.001; ovarian cancer: HR 0.33, 95% CI=0.27-0.42, P<0.001), but OS of patients did not increase significantly in these two cancer types (breast cancer: HR 0.87, 95% CI=0.76-1.01, P=0.065; ovarian cancer: HR 0.78, 95% CI=0.61-1.01, P=0.058). For ovarian cancer patients carrying BRCAm, the use of therapy with PARPis yielded longer PFS at the stage of newly diagnosed than the stage of recurrence (22.5 months vs 9.6 months). Conclusion: PARPis were beneficial to all with BRCAm, but they were "most" beneficial to the ovarian cancer subset when administered early after diagnosis, rather than after recurrence.


INTRODUCTION
Cancer is the leading cause of death in the world and has become a major public health problem that has persisted worldwide, for a long time. Whether in developed or developing countries, many cancer-related deaths occur every day. For example, based on human epidemiology data, 4,950 people die of cancer every day in the USA [1], this is even worse in China, where over 7,500 people die of cancer, daily [2]. Obviously, all countries will face great challenges in dealing with the huge and increasing burden of cancer at present and in the foreseeable future. To overcome these challenges, it is necessary to increase investment in basic and clinical research to further promote treatment options, which will undoubtedly accelerate the progress of fighting cancer. Patients suffering with cancer still face the challenge of high recurrence rate after surgery and the toxicity of conventional chemotherapy, so safer and more effective treatment schemes are needed.
Targeted molecular therapy holds great promise for the treatment of cancer and represents a revolutionary breakthrough in personalized medicine. Among breast AGING cancer and ovarian cancer patients with BRCA mutations, targeted therapy has always been a hot topic. In this regard, poly ADP ribose polymerase (PARP) inhibitor (PARPi), represent a novel cancer therapy targeting PARP, which have already achieved noteworthy therapeutic effects on cancer. Greater potency might be achieved by inhibiting PARP, because PARPi could sensitize cancer cells to conventional treatments including multiple chemotherapy or radiotherapy that cause DNA damage [3]. Using the genetic concept of synthetic lethality [3,4], PARP inhibitors are designed to target cancers harbouring specific DNA-repair defects, including those arising in carriers of BRCA1 or BRCA2 mutations [5]. More promisingly, tumors exhibiting BRCAness are sensitive to PARPis, and the concept of BRCAness can be described as a defect in DNA damage response by homologous recombination repair, regardless of the presence of germline BRCA1 or BRCA2 mutations.
Excitingly, numerous studies have demonstrated that PARPis achieve excellent anti-tumor efficacy as monotherapy or combination therapy with conventional treatments in various cancers patients, especially in breast cancer and ovarian cancer. Based on the results of clinical trials, the ASCO guidelines recommended PARP inhibitors for the treatment of ovarian cancer, breast cancer, and pancreatic cancer with BRCA1 / 2 mutations [6][7][8].
Enhanced understanding of the efficacy of PARPis in ovarian and breast cancer patients carrying BRCAm will provide more accurate treatment information and can improve clinical decision-making. The objective of this study was to perform a meta-analysis to provide a clinical reference by comprehensively evaluating the OS and PFS of ovarian and breast cancer with BRCA mutations.

Literature search
According to the search terms, a total of 2,829 related studies were identified from all searched databases. Because of duplications, 902 studies were removed. After eligibility screening of the titles, abstracts and full texts of the article, 1,895 studies were excluded because they have the characteristics of phase I clinical trial, basic research, preclinical trial, guideline, meeting, nontumor disease and insufficient data. After that, 32 trials related to evaluating the PFS or OS of PARPis in ovarian cancer, breast cancer, lung cancer, prostate cancer, pancreatic cancer, gastric cancer, glioblastoma, colorectal cancer and melanoma were retained. Finally, by analyzing breast and ovarian cancer with BRCA mutations, 15 randomized controlled trials were selected [9][10][11][12][13][14][15][16][17][18][19][20][21][22][23][24][25][26][27]. The flowchart of the trial selection process is shown in Figure 1.

Study characteristics
The main characteristics of the eligible trials are showed in Table 1. 15 trials were involved in the analysis,   Supplementary Figures 1, 2).

The efficacy of each PARPi in BRCAm cancer by pooling data from breast and ovarian cancer patients
Compared with the control groups, all PARPis included in the analysis were statistically significant in improving    Figure 4).  Table 2). Therefore, the patients of BRCAm ovarian cancer can get better survival when PARPis are administered earlier in cancer progression.

Publication bias
As visually assessed, substantial asymmetry was not identified in the Begg funnel plot (see Supplementary Figure 3). Moreover, no significant publication bias was found by the Begg rank correlation test and Egger linear regression test.

DISCUSSION
Rapid death is the inevitable outcome of patients with advanced or metastatic breast and ovarian cancer. It is urgent that an effective solution emerges to manage patients of advanced or metastatic breast and ovarian cancer.
Multiple meta-analysis of patients with ovarian cancer or breast cancer had proved that addition of PARPis to therapy was beneficial [28][29][30][31]. Moreover, Gu, L., et al. included 12 clinical trials containing six types of cancer patients for meta-analysis, and then concluded that with the acceptable and controllable toxicity profile, PARPis improved survival of cancer patients, and were more beneficial to ovarian cancer patients with BRCAm [32].
In order to evaluate PARPis in-depth and in detail, this meta-analysis comprehensively focused on evaluating the efficacy of each PARPi, with respect to different intervention strategies in patients with BRCAm-positive breast and ovarian cancer, and also differences in the therapeutic effectiveness of PARPis when administered at the time of relapse versus at the time of ovarian cancer diagnosis. In a subgroup of BRCAm cancers, our pooled analysis showed that compared with the control, the PARPis treatment group showed a statistically significant reduction in disease risk progression of patients with breast (36%) or ovarian (67%) cancer. From the perspective of OS data from meta-analysis, more RCTs are needed to confirm whether PARPis effectively improve the OS of BRCA mutant patients with breast or ovarian cancer.
For ovarian cancer, despite standard therapy which includes cytoreductive surgery and conventional chemotherapy, about 70% of patients with newly diagnosed advanced ovarian cancer will face to relapse within the subsequent 3 years [33]. Promisingly, Ibrahim, E.M., et al. demonstrated that in newly diagnosed patients with advanced high-grade ovarian cancer, PARPis significantly decreased the risk of PFS by 46% when compared with placebo [34]. As firstline maintenance therapy, PARPis greatly benefit PFS of patients with newly diagnosed advanced ovarian cancer [35]. Consistent with these results, our study found that the addition of PARPis to standard therapy at the beginning of diagnosis for advanced ovarian cancer patients with BRCAm lead to an additional PFS benefit of 12.9 months, when compared at the time of relapse.

Use PARPis in patients of cancer with homologous recombination deficiency (HRD)
Importantly, by using novel biomarkers of homologous recombination repair deficiency, the benefits of PARPis were extended to wider populations of patients beyond breast or ovarian cancer, even beyond BRCA mutant cancer [36]. For other cancers, in patients with BRCAm and metastatic pancreatic or prostate cancer, olaparid group had a better PFS than control group [37,38]. Furthermore, for patients with HRD, PARPi used in the populations of ERCC1 or BRCA deficient cancers, might potentiate their therapeutic effects by regulating the signal pathway related to antitumor immunity [39,40]. ATM (ataxia telangiectasia mutated) promoted survival by decreasing sensitivity to PARP inhibition AGING and playing a role upstream of homologous recombination repair in the repair of certain types of double-strand breaks [41]. PARP inhibitors had a significant killing effect on many cancers with ATM deficient [42][43][44][45]. Results from a phases 3 trial enrolled prostate cancer patients with qualifying deleterious or suspected deleterious alterations in at least 1 of 15 prespecified genes in homologous recombination repair: BRCA1, BRCA2, ATM, BRIP1, BARD1, CDK12, CHEK1, CHEK2, FANCL, PALB2, PPP2R2A, RAD51B, RAD51C, RAD51D, and RAD54L, indicated that olaparib improved progression-free survival when compared with enzalutamide or abiraterone [38]. However, because there is a lack of reliable clinical data to confirm PARPis are suitable for treating all types of cancer, patients should have genetic testing for defects in BRCA1/2 and other genes related to homologous recombination deficiency, before PARPis are safely applied [46].

Strengths and limitations of study
Strengths: Firstly, 1,426 breast cancer patients and 2,300 ovarian cancer patients with BRCA mutations were included in this analysis. Secondly, through reasonable stratification and grouping, the therapeutic characteristics of PARPis in patients with breast or ovarian cancer carrying BRCAm, BRCA1m and BRCA2m were analyzed. Thirdly, after considering the pooled, satisfactory results of median PFS in newly diagnosed patients with BRCA mutant ovarian cancer, we recommended PARPis as the first-line maintenance therapy for BRCAm ovarian cancer.
Limitations: We acknowledge that this evidence-based medical report has some limitations, based on heterogeneity from cancer type, PARPi type and therapeutic schedule, which may reduce the accuracy of our results. Firstly, two cancer types were included in this study, and no further detailed analysis was made in pathological types in each cancer. Secondly, diversification existed in the phase of the treatment for patients, as well as the treatment options of PARPis. Thirdly, data on the OS of several studies was not mature enough or published. Fourthly, although stratification analysis was conducted, a few trials were included in some subgroup analysis.

CONCLUSIONS
In this meta-analysis, application of therapy with PARPis provided a substantial PFS benefit among breast and ovarian cancer with BRCA mutations; and among patients with BRCAm ovarian cancer, PARPis provided longer PFS benefit at the stage of newly diagnosed than at the stage of recurrence.

MATERIALS AND METHODS
We performed a meta-analysis of PARPi efficacy in BRCA mutant cancers according to the recommendations of the Cochrane Handbook and PRISMA statement guidelines [47].

Search strategy
The search strategy and selection criteria are similar to the previous study we published [29]. We conducted a comprehensive systematic search of PubMed, Embase and Cochrane from inception to September 1, 2020 for all RCTs, and then checked the trial registration number and more relevant information in the https://www.clinicaltrials.gov/ and international clinical trials registry platform. For database searches we used the "parp OR poly adp ribose polymerase OR poly adenosine diphosphate ribose polymerase OR olaparib OR veliparib OR iniparib OR rucaparib OR niraparib OR talazoparib" as the search terms in all fields.

Selection criteria
Exclusion criteria and inclusion criteria were prespecified and used in literature search and screening. To be eligible, the selected randomized controlled trials not only met the condition of researching the clinical efficacy of PARPis in patients, but also met the following conditions. Firstly, the population was patients with breast and ovarian cancers carrying BRCA mutations, irrespective of cancer stage or grade, surgery, recurrence, drug resistance, histology. Secondly, intervention: treated with PARPis (olaparib, veliparib, rucaparib, iniparib, talazoparib, niraparib) as monotherapy or combination with conventional chemotherapy or molecular target therapy regardless of dosage and duration. Thirdly, main outcome: the primary outcome was OS or PFS measured as HR. Studies were excluded if they were nonrandomized control trials, phase I clinical trials, literature reviews and meta-analysis, case reports, retrospective or prospective observational cohort studies, basic science papers, commentaries, quality of life studies, and cost effectiveness analyses. In addition, we excluded those studies that did not explore PFS and OS or whose data have not yet been published. Moreover, updated and published follow-up data meeting the inclusion criteria which may have appeared in multiple articles or different publications were considered for one trial to analyze.

Data extraction and risk of bias assessment
The key purpose of this meta-analysis was to evaluate the efficacy of PARPis measured by HRs of OS or PFS. Using a pilot-tested data extraction sheet, two investigators independently reviewed the eligible AGING literature, and extracted the data including: cancer type and clinical stage or grade, first author, year of publication, phase of clinical trial, number of patients enrolled, intervention method, hazard ratios (HR) and their 95% confidence intervals (CIs) for OS and PFS stratified by BRCA status. The risk of bias was evaluated by using the Cochrane Risk of Bias Tool that consists of random sequence generation; allocation concealment; blinding of participants and personnel to the study protocol; blinding of outcome assessment; incomplete outcome data; and selective reporting [48]. The risk of bias was divided into three different levels: high, low, or unclear. Two investigators completed the review independently and in the event of any differences, resolved them through discussion and consultation.

Statistical analysis
The primary objectives of the meta-analysis were OS and PFS in breast and ovarian cancers patients with BRCAm. Subgroup analysis was conducted to explore the efficacy of PARPis in breast or ovarian cancer by stratification BRCA1m or BRCA2m, and also to evaluate the efficacy of each PARPi and different intervention methods by integrating data from these two cancer types. Cochrane's Q-test and I2 statistics were used to assess heterogeneity across the different studies. P ≤ 0.10 or I 2 ≥ 50% indicated significant heterogeneity.
The random-effect model was used to increase reliability because of the obvious heterogeneity attributed to differences by cancer type, PARPi type and therapeutic schedule in this meta-analysis [32]. Therefore, it was necessary to perform subgroup analysis to reduce heterogeneity and improve reliability. Potential publication bias was assessed by the Begg's and Egger's test [49,50]. P < 0.05 was refer to indicate statistical significance. All analysis was carried out using Stata version 15.1 (StataCorp, College Station, TX, USA).

Data availability statement
All data relevant to the study are included in the article or uploaded as supplementary information. All data generated or analysed during this study are included in this published article (and its supplementary information files).

CONFLICTS OF INTEREST
The authors declare that they have no conflicts of interest.

FUNDING
This study and the article processing charges were supported by the National Natural Sciences Foundation of China [grant number 81772764].