Efficacy and safety of PARP inhibitors for maintenance treatment of ovarian cancer, regardless of BRCA or HRD status: a comprehensive updated meta-analysis

Abstract

Without taking into account existing biomarkers like genetic mutations (BRCA mutation, Homologous recombination deficiency) with advanced ovarian cancer (OC), the overall survival (OS), progress-free survival (PFS) of the aggregate all groups that have been classified were hazard ratio (HR): 0.72, 95% confidence intervals (CI): 0.66–0.79 and HR: 0.48, 95%CI: 0.44–0.52, respectively. Meanwhile, the OS and PFS of the whole population (regardless of existing genetic mutation markers) were HR: 0.74, 95%CI: 0.64–0.87 and HR: 0.52, 95%CI: 0.42–0.65, separately. Furthermore, the OS and PFS of positive gene mutation markers were HR: 0.71, 95%CI: 0.61–0.83 (HRD and BRACm) and HR: 0.47, 95%CI: 0.42–0.52 (HRD and BRACm), individually. The poly ADP-ribose polymers (PARP) inhibitors have desired efficiency and security in the maintenance treatment of advanced OC patients with BRCAm or BRCAwt, HRD or HRP and unknown gene status.

Keywords:

• Ovarian cancer
• PARP inhibitor
• BRCA
• HRD
• OS
• PFS

Background

Ovarian cancer (OC) is the most lethal malignant tumor in gynecology. Type I epithelial ovarian cancer is considered to be a relatively inert and genetically stable tumor, usually caused by recognizable precursor lesions such as endometriosis or borderline tumors with low malignant potential. In contrast, type II epithelial ovarian cancer is considered from the outset to be an advanced bio-aggressive tumor with a tendency to metastasize from a small primary lesion (Pavlidis et al. 2021). Since there are no specific symptoms and signs in the early stage, OC is often diagnosed in the late stage (Siegel et al. 2015, Menon et al. 2021). Up to now, there has been no effective tool for general population screening. It also reflects research over the past decade into economic and cost-effective strategies for the early detection and prevention of ovarian cancer. The per capita cost of treating ovarian cancer patients remains the highest of all cancer types (Ghose et al. 2022a). Every year, there are about 23,900 new cases of OC and 152,000 deaths worldwide (Bray et al. 2018). The current standard of care for OC is to minimize the tumor burden through standardized radical surgery (Colombo et al. 2019). Generally speaking, paclitaxel plus carboplatin is the standard chemotherapy to eliminate residual tumor cells. The primary treatment strategy for ovarian cancer is cellular-reducing surgery, followed by six postoperative cycles of platinum-based chemotherapy. However, there is an alternative, which is three or more cycles of neoadjuvant chemotherapy (NACT) before surgery, and there is a lack of consensus on which treatment strategy is best and how to choose it. Importantly, NACT provides the opportunity to test drug sensitivity in advance and identify patients at higher risk of relapse (Moschetta et al. 2020). Meanwhile, maintenance methods delayed disease progression, prevented recurrence, prolonged the progression-free period to the maximum extent, and improved the clinical cure rate (LaFargue et al. 2019).

Up to now, poly ADP-ribose polymers (PARP) inhibitors have become highly promising molecular targeted therapies for precise and personalized treatment of OC (Barkauskaite et al. 2015). Additionally, Food and Drug Administration has approved PARP inhibitors in maintenance therapies for BRCA mutation (BRCAm), homologous recombination deficiency (HRD) and platinum-sensitive recurrent OC. However, a clinical application has demonstrated that advanced OC can be favorable over PARP irrespective of BRCA or HRD status (Lin et al. 2021). Therefore, a comprehensive meta-analysis of high-quality randomized controlled trials was included to investigate whether PARP inhibitors can improve the progression-free survival (PFS), overall survival (OS) and safety of patients with advanced OC with different gene mutation statuses, which were by enlisting different conditions, such as different BRCA, HRD state, all people, intention-to-treat population (ITT), investigation (INV) or Blind Independent Centre Assessment review (BICR), time to first subsequent therapy or death (TFST) or time to second subsequent therapy or end (TSST). Hence, this study explored whether PARP inhibitors, regardless of BRCA and HRD status, inclusion groups and different investigator evaluation methods, are accurate, personalized and targeted as maintenance treatment for patients with advanced OC, which can lengthen the OS and PFS, and have fewer adverse reactions.

Materials and methods

This study was implemented strictly with the preferred reporting guidelines for meta-analysis. We systematically searched the relevant literature from Scope, PubMed, EMBASE, Cochrane Collaboration and China National Knowledge Infrastructure (CNKI) Database. The search keywords were OC, PAPR and randomized-controlled-trial (RCT) studies or clinical trials. The inclusion criteria were a complete or partial response to chemotherapy and relatively stable disease in patients with International Federation of Gynaecology and Obstetrics stage III and IV on advanced OC. The exclusion criteria were: (1) non-RCT; (2) no detailed HR and 95% confidence interval (CI) of PFS and OS; (3) PAPR inhibitors are not the only course of therapy; (4) conferences, case reports, reviews, etc.; (5) non-English and non-Chinese literature.

We included 19 studies with a total of 5316 participants, this included 3608 patients with PAPR inhibitors and 1708 patients on placebo (Ledermann et al. 2012, Ledermann et al. 2016, Matulonis et al. 2016, Mirza et al. 2016, Pujade-Lauraine et al. 2017, Moore et al. 2018, Gonzalez-Martin et al. 2019, Fabbro et al. 2020, Aghajanian et al. 2021, Banerjee et al. 2021, Oaknin et al. 2021, Poveda et al. 2021, Wu et al. 2021a, 2021b, Kristeleit et al. 2022, Monk et al. 2022, Swisher et al. 2022, Trillsch et al. 2022). The full text of a study was not available for relevant analysis, and as much help as possible was sought. Unfortunately, no response was received. One randomized controlled phase III clinical trial is currently underway and is expected to be completed in 2024. Information about the study has not been published.

Data extraction

In this analysis, patients with advanced OC who received PAPR inhibitors were enrolled in the experimental group, and those who received a placebo were in the control group. Two investigators identified the literature and evaluated the quality of the literature. The third investigator consulted for settlement if the two investigators had different opinions. In general, the main information about the literature was summarized, containing the principal investigator, registration number, year, indicators of clinical outcomes and related criteria.

Statistical analysis

The OS and PFS of this meta-analysis were calculated by HR and 95%CI. Data from the included literature were analyzed by Stata15 software, the total HR and 95%CI of the PFS and OS. p < 0.05 and I² > 50% indicates that there is considerable heterogeneity in the analysis of the data, the application of the random effects model can improve the confidence of the data on other statistics. Otherwise, the fixed effects model is implemented. Sensitivity analysis and subgroup analysis were performed to find the source of heterogeneity in the study. For example, HRD or HRP, TFST or TSST and INV or BICR. Begg’s test predicts publication bias. The safety of PARP inhibitors was examined with the OR and 95%CI.

Quality assessment

The Revman 5 software estimates the quality of the included literature from multiple aspects, including randomization of the included population (selective bias), allocation hiding (selective bias), double blinding of investigators and subjects (implementation bias), blinded evaluation of study results (measurement bias), data integrity of follow-up outcomes (reporting bias), selective reporting of study results (reporting bias), and bias from other news from. The literature bias assessment map and literature bias risk summary map were prepared by generalizing the information.

Results

Table 1 demonstrates the basic information about the involved literature. This study contained 5316 participants and one RCT of phase II in multiple national medical centers in North America, South America, Europe and Asia, including the US, Canada, France, the UK and China. Nine RCTs of phase III; The year of publication was one in 2014, four in 2016, one in 2017, two in 2019, eight in 2021 and three in 2022. All the included studies had registration numbers. Data from six studies involved OS in various BRCA states, 19 studies involved PFS in diverse BRCA types, two studies embodied OS in disparate HRD forms, and 21 studies embraced PFS in different HRD modes. 18 studies listed the outcome time point of the clinical study (TFST or TSST). Data from 17 studies were obtained from different investigators evaluating clinical outcomes (NIR or BICR). One study contained the outcome of RFS (Figure 1). All the blinded RCTs included in the study had a low-risk bias. The follow-up data were relatively complete (Figure 2). In one study, we defined risk as ambiguous because the process of the RCT was not described in detail.

Figure 1. Flow chart of included literature.

Figure 2. Risk bias assessment form.

Table 1. Basic clinical information table of the included literature.

Author	Number	Phase	Race	Year	Exp.	Con.	Exp.	Con.	Outcome	Status analysis
Antonio Gonzalez-Martin (Gonzalez-Martin et al. 2019)	NCT02655016	III	Americans and Europeans	2019	Niraparib	Placebo	487	246	PFS OS	HRD OP HRP BRCAm BRCAwt
Mansoor R. Mirza (Mirza et al. 2016)	NCT01847274	III	Americans and Europeans	2016	Niraparib	Placebo	372	181	PFS	gBRCA HRD HRP BRCAwt
Eric Pujade-Lauraine (Pujade-Lauraine et al. 2017)	NCT01874353	III	America, Europe, Asia	2017	Olaparib	Placebo	196	99	PFS	INV BICR TFST TSST
Andrés Poveda (Poveda et al. 2021)	NCT01874353	III	America, Europe, Asia	2021	Olaparib	Placebo	196	99	OS PFS	BRCAm TFST TSST
Susana Banerjee (Banerjee et al. 2021)	NCT01844986	III		2021	Olaparib	Placebo	260	131	PFS RFS	Higher /lower clinical risk BRCA1m BRCA2m
Jonathan Ledermann (Ledermann et al. 2012)	NCT00753545	II		2014	Olaparib	Placebo	131	123	PFS OS	OP BRCAm BRCAwt
Lingying Wu (Wu et al. 2021a)	NCT01844986	III	Entire world	2021	Olaparib	Placebo	260	131	PFS	INV BICR TSST TFST
Kathleen Moore (Moore et al. 2018)	NCT01847274	III	Entire world	2021	Olaparib	Placebo	260	131	PFS OS	TFST TSST BICR
Jonathan A. Ledermann (Ledermann et al. 2016)	NCT00753545	II		2016	Olaparib	Placebo	131	123	OS PFS	BRCAm BRCAwt TFST TSST BRCA1m BRCA2m
Rebecca Kristeleit (Kristeleit et al. 2022)	NCT02855944	III	America, Europe, Asia	2022	Rucaparib	Standard-of-care chemotherapy	233	116	PFS	Efficacy population ITT INV BICR HRD HRP
Michel Fabbro (Fabbro et al. 2020)	NCT01847274	III		2019	Niraparib	Placebo	372	181	PFS	BRCAm BRCAwt
Ursula A. Matulonis (Matulonis et al. 2016)	NCT00753545	II		2016	Olaparib	Placebo	136	129	OS	BRCAm PARP Inhibitor Sites
Elizabeth M. Swisher (Swisher et al. 2022)	NCT02470585	III		2021	Veliparib	Placebo	765	375	PFS	HRD HRP
Fabian Trillsch (Trillsch et al. 2022)	NCT01874353	III		2022	Olaparib	Placebo	196	99	PFS OS	＜65 year-old ≥65 year-old
Ursula A. Matulonis (Matulonis et al. 2016)	NCT00753545	III		2016	Olaparib	Placebo	131	102	OS	BRCAm PARP inhibitor sites
Bradley J. Monk (Monk et al. 2022)	NCT03522246	III	America, Europe, Asia	2022	Rucaparib	Placebo	612	160	PFS	HRD HRP ITT BICR BRCAm BRCAwt
Ana Oaknin (Oaknin et al. 2021)	NCT01968213	III		2021	Rucaparib	Placebo	375	189	PFS	BRCAm HRD ITT
X. H. Wu (Wu et al. 2021b)	NCT03705156	III	Asian	2021	Niraparib	Placebo	177	88	PFS	ITT BRCAm BRCAwt

Abbreviations. HRD: homologous recombination defect; OP: overall population; HRP: homologous recombination proficient; BRCAm: mutated BRCA; BRCAwt: wild-type BRCA; gBRCA: germline BRCA mutation; INV: investigator; BICR: blinded independent central review; TFST: time to first subsequent therapy; TSST: time to second subsequent therapy; ITT: intention-to-treat.

OS changes in maintenance therapy with PRPA inhibitors in OC patients

Figure 3(a) manifests the meta-analysis of 5 RCTs, and the results suggested that PARP inhibitors effectively prolonged the OS in patients with advanced OC, and the HR of OS was 0.72 (95%CI: 0.66–0.79, n = 1673, I² = 0.00%, p = 0.885; p < 0.001), there was no heterogeneity in the merged studies, and the prolongation of the OS was statistically significant. The three pooled studies were a meta-analysis of a randomized population without any type of grouping. Compared with the placebo, the application of PARP improved the OS of the population (HR: 0.74, 95%CI: 0.64–0.87, I² = 0.00%, p heterogeneity = 0.67; p < 0.001, Figure 3(b)). A meta-analysis of three studies referred BRCAm or HRD and BRCAwt or HRP. PARP inhibitors administration protracted the OS in the population, contrasted with the placebo (HR: 0.71, 95%CI: 0.61–0.83, I² = 0.00%, p heterogeneity = 0.85; p < 0.001, Figure 3(c)). The OS of the positive mutated biomarker gene positive (BRCAm or HRD) (HR: 0.67, 95%CI: 0.56–0.80), and the OS of the negative mutated biomarker (BRCAwt or HRP) (HR: 0.81, 95%CI: 0.62–1.06).

Figure 3. Forest plot of the OS for meta-analysis of RCTs in PARP maintenance therapy. (a) All enrolled populations; (b) ungrouped population; (c) BRCA mutation or HRD and BRCA wild-type or HRP.

The PFS of OC patients treated with PRPA inhibitors during maintenance therapy

Meta-analyses of 15 RCTs were done on all people. The results imply that PARP inhibitors extend PFS in patients with advanced OC (HR: 0.48, 95%CI: 0.44–0.52, I² = 74.1%; p heterogeneity < 0.001, p < 0.001, Figure 4(a)). Meta-analysis of three RCTs with randomly selected populations, the PARP inhibitors improved the PFS in the population by contrasting placebo (HR: 0.52, 95%CI: 0.42–0.65, I² = 66.9%; p heterogeneity = 0.002, p < 0.001, Figure 4(b)). Compared with placebo-controlled, the meta-analysis of pooled three PARP inhibitors RCT studies with a statistically significant prolongation of the PFS by different BRCA mutations or BRCAwt and HRD or HRP status (HR: 0.47; 95%CI: 0.42–0.52; I² = 71.3%; p heterogeneity < 0.001, p < 0.001, Figure 4(c)). The meta-analysis includes three studies with different homologous recombination situations. Contrasted placebo, PARP increased the PFS of HRD (HR: 0.48, 95%CI: 0.40–0.59; p < 0.001, Figure 4(d)) and HRP (HR: 0.70, 95%CI: 0.61–0.81; p < 0.001, Figure 4(d)), there is heterogeneity among them.

Figure 4. Forest plot of the PFS from the meta-analysis of RCTs in maintenance therapy with PARP inhibitors. (a) All registered populations; (b) the included ungrouped population; (c) the BRCA mutation or HRD versus BRCA wild-type or HRP; (d) the HRD and HRP.

A meta-analysis of the BRCAm and BRCAwt-related studies was conducted. PARP inhibitors protracted the PFS in patients regardless of the BRAC status (n = 10, HR: 0.41; 95%CI: 0.35–0.49; p < 0.001, Figure 5(a)). The PFS in the BRCAm group (HR: 0.33, 95%CI: 0.29–0.38; p < 0.001) and the PFS in the BRCAwt group (HR: 0.57, 95%CI: 0.48–0.67; p < 0.001) showed little heterogeneity in the subgroup analysis (BRCAm I² = 22.3%, p heterogeneity = 0.206; BRCAwt I² = 48.3%, p heterogeneity = 0.036). Meta-analysis of RCTs of the OC with negative biomarker (BRCAwt HR: 0.61, 95%CI: 0.54–0.68; p < 0.001 and HRP HR: 0.70, 95%CI: 0.61–0.81; p < 0.001) expression represents that maintenance treatment with PARP inhibitors can affect the PFS (HR: 0.64, 95%CI: 0.59–0.70, I² = 33.7%; p heterogeneity = 0.085, p < 0.001, Figure 5(b)). Meta-analysis was performed according to different trial evaluators and methods, and we discovered that RCTs of both INV (HR: 0.51, 95%CI: 0.36–0.72; p < 0.001) and BICR (HR: 0.46, 95%CI: 0.35–0.60; p < 0.001) to evaluate the efficacy of PARP inhibitor maintaining therapy in the PFS was satisfactory (HR: 0.48, 95%CI: 0.39–0.59, I² = 85.1%; p heterogeneity < 0.001, p < 0.001, Figure 5(c)). In the meta-analysis of some RCTs, maintaining treatment with PARP inhibitors significantly prolonged advanced OC in the TFST (HR: 0.34, 95%CI: 0.30–0.38; p < 0.001) and TSST (HR: 0.49, 95%CI: 0.43–0.55; p < 0.001) in the total (HR: 0.41, 95%CI: 0.37–0.44, I² = 48.3%; p heterogeneity = 0.012, p < 0.001, Figure 5(d)).

Figure 5. Forest plot of meta-analysis of the efficacy of maintenance treatment with PARP inhibitors. (a) Forest map of PFS in the BRCAm state and BRCAwt state; (b) forest map of the PFS with the HRP type and BRACwt type; (c) forest maps of the PFS evaluated by the INV and BICR; (d) forest diagram for the TFST and TSST.

Sensitivity analysis

To explore the sources of heterogeneity in the meta-analysis, the sensitivity analysis was operated to evaluate the stability and credibility of HR: values (PFS) of meta-analysis. The horizontal boxplots with the leave-one-out method confirmed that none of the studies had a qualitative impact on HR: value for the PFS (Figure 6).

Figure 6. The plot of sensitivity analysis. (a) Meta-analysis of the PFS for survival outcomes in the total population; (b) sensitivity analysis of the PFS for the HRD and BRCAm; (c) sensitivity analysis chart of the PFS on the HRD; (d) sensitivity analysis of the PFS assessed by INV; (e) analysis of the PFS evaluated by BICR.

Publication bias

Begg’s test funnel plot was implemented to evaluate whether there was a publication bias in the survival outcomes (OS and PFS) of the RCTs included in the meta-analysis. First of all, the visual evaluation denoted that the funnel plot was symmetric and there was no obvious publication bias. Begg’s test was adopted to estimate publication bias in the OS (p = 0.115) and PFS (p = 0.007). There was no obvious change in the funnel plot after pruning, indicating that the results were relatively stable (Figure 7).

Figure 7. Funnel plot of Begg’s test of the meta-analysis. (a) Funnel diagram of the Begg’s test on the OS; (b) funnel plot of the Begg’s test on the PFS.

Adverse reactions

The OR: was the outcome indicator of meta-analysis for PARP inhibitors on adverse reactions (grade 3 or above) (Figure 8). The overall risk of major adverse events was assessed, which were anemia (incidence = 19.47%, OR: 13.7, 95%CI: 8.19–20.87; p < 0.001), nausea (incidence = 1.97%, OR: 5.49, 95%CI: 1.97–15.25; p < 0.001), fatigue (incidence = 4.91%, OR: 3.91, 95%CI: 2.10–7.30; p < 0.001), abdominal pain (incidence = 1.29%, OR: 1.23, 95%CI: 0.64–2.35; p = 0.54, p = 0.018), diarrhea (incidence = 1.00%, OR: 2.97, 95%CI: 1.20–7.33; p = 0.018), and vomiting (incidence = 1.89%, OR: 2.55, 95%CI: 1.19–5.49; p = 0.017) in a fixed-effect model, as well as neutropenia (incidence = 13.00%, OR: 3.27, 95%CI: 1.78–6.03; p < 0.001) and thrombocytopenia(incidence = 18.67%, OR: 5.38, 95%CI: 0.90–32.20, p = 0.066) in a random-effect model. The results of the meta-analysis indicate that the adverse reactions of PAPR inhibitors are mainly hemotoxicity and gastrointestinal reactions, which are safe after symptomatic treatment.

Figure 8. Adverse reactions after maintenance treatment with PARP inhibitors. (a) Forest map of anemia; (b) forest map of neutropenia; (c) forest map of nausea; (d) forest diagram of fatigue; (e) forest map of diarrhea; (f) forest map of abdominal pain; (g) forest map of thrombocytopenia; (h) forest map of vomiting.

Discussion

The outcome of our meta-study demonstrates that maintenance treatment with PARP inhibitors can effectively prolong the OS and PFS with advanced OC compared with placebo, with relatively fewer controllable adverse effects (Valabrega et al. 2021). In addition, the INV or BICR assessed that patients with OC could benefit from PAPR inhibitors despite their BRCA and HRD status, extending the TSST or TFST (Huang et al. 2020). Importantly, PARP suppression may even represent a therapeutic strategy for sporadic cancers with BRCA-like properties, known as ‘BRCAness’. Thus, mutations in genes other than BRCA in the HR: pathway – ATM, CHEK2, BARD1, BRIP1, MRE11, RAD50, NBS1, RAD51C, RAD51D, RAD52, PALB2 and DNA-PK – expand the fitness of PARP inhibitors. Broader applications of synthetic lethality targeting the HR: pathway are still being investigated (Boussios et al. 2022). Therefore, the results of these pooled analyses of RCTs support the possibility that PAPR inhibitors could be expanded to the entire population as an emerging target therapy for the maintenance treatment with advanced OC.

It has been proposed that advanced OC requires to be scientifically managed like a chronic disease. Recurrence and drug resistance of advanced OC after chemotherapy are still tricky problems to be solved urgently in clinical practice (Salani et al. 2011, Mullen et al. 2019). It is worth adding that proteomic techniques such as mass spectrometry and protein array analysis have facilitated the anatomical and proteomic characterization of the underlying molecular signaling events in several cancers. In this case, proteomic analysis of ovarian cancers, as well as their adaptive response to treatment, can uncover new treatment options, which can reduce the emergence of drug resistance and potentially improve patient outcomes (Ghose et al. 2022b). However, this study shows that (1) The PAPR inhibitors can significantly prolong the OS and PFS with OC (BRCAm, BRCAwt, HRD or HRP) (Cancer Genome Atlas Research Network 2011); (2) the PAPR inhibitors can be clinically beneficial when it was in a randomized population or patients with unknown gene status in OC (Mirza et al. 2016); (3) a significant survival advantage was observed for advanced OC with PAPR inhibitors, both at the time of first follow-up treatment or death from disease progression and from the time of second follow-up treatment or death (Pennington et al. 2014, Gonzalez-Martin et al. 2019); (4) the assessment of imaging data, disease changes and drug response by the INV or BICR, which presents confidently that this meta-analysis of PARP inhibitors will have a favorable effect on the entire population of patients with advanced OC (Stone et al. 2019, Aghajanian et al. 2021). In conclusion, the application of PARP inhibitors in the maintenance treatment of advanced OC can apparently lengthen the OS, PFS, and delay the progression, whether with newly diagnosed OC or after secondary treatment (Ray-Coquard et al. 2019).

At the same time, this meta-analysis has some limitations: (1) despite the inclusion of high-quality RCTs, the existence of bias cannot be ruled out, such as the design and implementation of open-arm trials, the maturity of the included RCTs and the funding of drug manufacturers (Dizon, 2017); (2) the PAPR inhibitors in the BRCAm and HRD with advanced OC has been reported in relevant meta-analysis literature on the extension of the PFS (Lin et al. 2021). (3) the PAPR inhibitors in this study are not single drugs and are tablets or capsules produced in different countries, and the therapeutic effects may be different (Domchek et al. 2016, Perego et al. 2020). Nevertheless, this meta-analysis provides a comprehensive explanation. It supports that the PAPR inhibitors for maintenance treatment in patients with advanced OC can effectively on extending the OS and PFS (Smith and Pothuri 2022, Vergote et al. 2022). Recently, the approved indications for PARP inhibitors are shrinking due to the increased incidence of myelodysplastic syndrome and AML.

Conclusion

The results of this meta-analysis manifest that PAPR inhibitors can achieve excellent OS and PFS with advanced OC regardless of BRCA, BRCAwt, HRD or HRP and unknown gene status. Although grade 3 or above adverse reactions may occur, close monitoring and timely treatment are relatively safe.

Ethical approval

This study did not involve human or animal experiments, so ethical approval was not required.

Acknowledgment

The author thanks the participants for their cooperation, at the same time, the author is particularly grateful to the Obstetrics and Gynecology Department of Shandong University Qilu Hospital for providing support.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

The author(s) reported there is no funding associated with the work featured in this article.