Efficacy of PARP Inhibitors in the Treatment of Ovarian Cancer: A Literature-Based Review

Importance of this Review

Recently, many new novel biological agents that act in multiple ways to conventional chemotherapy have been discovered. It is, therefore, very much needed to establish whether the addition of these new drugs to conventional chemotherapy is fruitful, in terms of survival and, if so, at what cost, in terms of additional AEs. As these compounds may be reported less toxic as compared with conventional chemotherapeutic agents, it may be feasible to advise these new therapeutic agents in patients who are not currently taking chemotherapy, to reduce the chance of, or delay, the recurrence of their OC.

PARPis and Synthetic Lethality

PARPs regulate a multiple biological mechanisms. PARP-1: Part of the PARP protein family; it is also found to be playing an important role in base excision repair mechanism (BER). DNA modifications induced exogenously or endogenously, can be repaired by BER. After eradication of the damaged base, single strand break (SSB) is produced.¹⁸ Usually, PARP-1 binds to the SSB and further attracts other proteins to start with the SSB repair.¹⁹ Previously, it was considered that the inhibition of PARP-1 would further lead to the stalling of the replication fork at the SSB,²⁰ further leading to the accumulation of double strand breaks (DSBs) in duplicating cells. Recent findings are also suggesting that some PARPis might also “trap” PARP1 on DNA and thereby further interfere with the catalytic cycle of PARP.²¹ The capability of PARPis to trap PARP1 on DNA has been found to add to the observed cytotoxicity.²² More recent data demonstrated that PARP also has a functional role in the repair mechanism of DSBs.²³ The efficiency of PARPis in BRCA1- and BRCA2-related tumors is dependent on the concept of synthetic lethality, whereas a deficiency in either one of two genes has no prominent effect on the viability of the cell but the combination of defects in both the genes will result in death of the cell.²⁴

Thus, especially in BRCA-deficient and homologous recombination-deficient (HRD) cells, inhibition of PARP enzyme will further lead to the death of the cell. Comparatively, normal cells have a sufficient homology directed repair (HDR) function and therefore they survive PARPi therapy. This further leads to a more precise and less toxic therapy as compared with chemotherapy.²² Therefore, PARPis are considered as a potent medication, especially in BRCA-mutated types of malignancies and other HDR-deficient cancers.²⁵

DNA Repair and BRCA

In past few years, dramatic advances have been reported in understanding about the mechanism and regulation of cellular components that are of vital importance in the repair of damaged DNA. DNA faces multiple types of assaults on its basic native structure and sequence throughout the life of a cell.²⁵ Human cells possess minimum five primary pathways of DNA repair, which are systems that serve to probe and identify defects protecting the genome. The major DNA repair pathways are direct repair pathways, mismatch repair (MMR), base excision repair (BER), nucleotide excision repair (NER), and double-strand break (DSB) recombinational repair, which includes both nonhomologous end-joining (NHEJ) and homologous recombinational repair.²⁶ Reduction, dysfunction, or absence of proteins committed to these pathways may further lead to disastrous cellular consequences, leading to the mutagenesis and toxicity.

Recently, BRCA-1 and BRCA-2 tumor suppressor genes have been corelated with a fundamental role in the response to cellular damage through the activation of specific DNA repair mechanism. Both BRCA1 and BRCA2 proteins are usually found in stable interaction, suggesting co-function of these proteins in pathways of tumor suppression. Both these genes have been proposed to function in DNA repair mechanism and as transcriptional regulators. BRCA-1 and BRCA-2 form a complex with Rad51, a protein with an established role in homologous recombination.²⁷

It has been found that BRCA-1 is also involved in forming a complex with and activation of p53.²⁸ The tumor suppressor protein p53 is involved in different types of human malignancies²⁹; the normal function of p53 is to signal the occurrence of DNA damage and temporarily arrest the cell cycle to either allow repair or trigger cell death. In-depth analysis of the effects of BRCA genes and their transcriptional functions may further result in a clearer understanding of their tissue-specific actions.

Olaparib (AstraZeneca)

Olaparib is one of the most focused and investigated PARPis. It was the first PARPis to investigate pharmacokinetics and pharmacodynamics in patients with, among others, breast and ovarian malignancies. Among the 60 patients enrolled, 22 had a BRCA-1 or BRCA-2 germ line mutation. Results showed a maximum tolerated dose (MTD) of olaparib 400 mg BID. Reported adverse events were mainly of grade 1 and 2 and included vomiting, nausea, altered taste, fatigue, and anorexia. Subsequent phase II clinical trials in patients with a gBRCAm showed an overall response rate (ORR) of 41% in patients with advanced breast malignancies⁴⁰ and of 33% in recurrent EOC^41,42 demonstrated a tumor response rate in metastatic breast cancer patients with ≥ 3 chemotherapy regimens of 12.9% (95% confidence interval [CI], 5.7–23.9). For platinum resistant relapsed EOC, the ORR was 31.1% (95% CI, 24.6–38.1). Depending on these results, a subsequent clinical trial (study 19) was started by Ledermann et al. This phase II clinical trial compared the efficiency of olaparib to placebo as maintenance therapy after response to platinum-based chemotherapy in 265 patients with platinum-sensitive recurrent serous EOC. Overall results demonstrated that olaparib increased median progression-free survival (PFS) as compared with placebo (8.4 vs. 4.8 months, respectively; hazard ratio (HR) 0.35; p < 000.1).⁴³ The subgroup analysis by overall BRCA mutation (BRCAm) status (germ line and somatic) demonstrated a considerable benefit in median PFS in the olaparib group as compared with the placebo group (11.2 vs. 4.3 months, respectively; HR 0.18; p < 0.0001).⁴⁴ Reported adverse events were mostly of grades 1 and 2. This study finally led to the EMA and FDA registration of olaparib for EOC.⁹

Veliparib (Abbvie)

Veliparib is another PARPi of which clinical studies reported promising results in recurrent EOC. The first phase I clinical trial by Puhalla et al⁴³ supported for a phase 2 dose (RP2D) of 400 mg BID for the management of platinum-resistant or -refractory EOC or basal-like breast cancer. Sixty out of 88 patients had a gBRCAm. Fatigue, nausea, and lymphopenia were the most commonly reported all-grade toxicities. In phase II clinical trial by Coleman et al,⁴⁰ veliparib monotherapy was given to 50 patients with persistent or recurrent EOC carrying a BRCAm. Results demonstrated an ORR of 26% with acceptable toxicity.

Niraparib (Tesaro)

Niraparib is another potent PARPi. This PARPi was first time clinically tested in phase I clinical trial by Sandhu et al.⁴⁵ The RP2D was 300 mg/day with reported adverse events of nausea, anorexia, anemia, fatigue, neutropenia, thrombocytopenia, constipation, and vomiting (grades 1 and 2). Efficiency evaluation showed that 40% of patients with BRCAm recurrent EOC and 50% of patients with metastatic breast cancer had a partial response. Recently, the placebo-controlled phase III clinical trial NOVA,⁴⁶ with niraparib as maintenance therapy after completing or near complete response to platinum-based chemotherapy in patients with platinum-sensitive recurrent EOC, showed a higher PFS in the niraparib group as compared with placebo group. Patients with a gBRCAm had the largest benefit with an increase in PFS of 21.0 versus 5.5 months (HR: 0.27), followed by a subgroup with HDR deficiency as defined by the My Choice HDR deficiency test (Myriad Genetics, Salt lake City, Utah, United States) with an increase in PFS of 12.9 versus 3.8 months (HR: 0.38). Even in the HDR-proficient group a PFS benefit of 6.9 versus 3.8 months was noted (HR: 0.58). OS data were not mature data. Finally it was concluded that niraparib is beneficial in all patients with platinum-sensitive recurrent EOC in response to platinum-based chemotherapy regardless of the BRCA mutation or HDR deficiency status. Furthermore,⁴⁷ it was concluded that patients managed with niraparib after complete or partial response can maintain their quality of life.

Rucaparib (Clovis Oncology)

One of the first phase I clinical trial focusing rucaparib was conducted by Kristeleit et al.⁴⁸ Results demonstrated a RP2D of 600 mg BID¹⁰ subsequently, efficacy of intravenous (IV) intermittent and oral continuous dosing schedules of rucaparib in gBRCAm recurrent ovarian and metastatic breast cancer was investigated. The IV intermittent dosing schedule resulted in an ORR of 2%, which was 15% for oral rucaparib. Forty-one percent of the patients on the IV intermittent dosing schedule achieved stable disease (SD) for ≥12 weeks. In the oral continuous dosing cohort 81% patients achieved RECIST (Response Evaluation Criteria in Solid Tumors) complete response, partial response or SD for ≥12 weeks. It was concluded that oral continuous rucaparib dosing is needed for getting better results. Furthermore, the ARIEL2 clinical trial studied the efficacy of rucaparib in 3 groups with relapsed EOC: patients with a BRCA mutation, patients with high loss of heterozygosity (LOH) as a definition for HDR deficiency, and patients with low LOH.⁴⁵ Results demonstrated a median PFS of 12.8, 5.7, and 5.2 months, respectively. PFS was found to be considerably on higher side in the BRCAm (HR: 0.27 p < 0.001) and LOH high group (HR: 0.62, p = 0.011) as compared with the LOH low group.

PARPi and Chemotherapy

PARPis and Immunotherapy

There are numerous immune checkpoint inhibitors; pembrolizumab and nivolumab target the programmed death protein-1 (PD-1), avelumab, atezolizumab, and durvalumab target the programmed death ligand-1 (PD-L1) and ipilimumab and tremelimumab target cytotoxic T-lymphocyte–associated protein 4 (CTLA-4). These immune checkpoint inhibitors have been found to be beneficial in treating several types of malignancies and are currently under evaluation in female cancer patients. The combination of PARP inhibition and immune checkpoint inhibition showed promising results in patients with EOC and HDR deficiency.⁶¹ Tumors with a BRCAm typically harbor TP53 mutations and have a higher mutational load.⁵⁴ Hence, they also harbor a greater number of neoantigens that enhance the recruitment of tumor infiltrating lymphocytes (TILs); in BRCAm tumors a considerably increased expression of CD3+ and CD8+ TILs have been demonstrated,⁵⁵ as well as increased expression of PD-1 and PD-L1 as compared with wild type EOC.⁶² Furthermore, PARP inhibition can modulate immune signaling pathways through various mechanisms,⁶³ both activating and nonactivating. In vitro a CTLA-4 antibody, but not PD-1/PD-L1 blockade, synergized therapeutically with veliparib.⁶⁴ While the PARPi talazoparib increased the number of peritoneal CD8+ T-cells and natural killer cells and increased production of interferon (IFN)-γ and tumor necrosis factor-α in a BRCA1-mutated OC xenograft model.⁶⁵ The exact immune-modulating effects of checkpoint inhibitor plus PARPi combinations are currently under investigations and needs further research. Currently, multiple trials are, therefore, investigating these combinations.⁶⁶

Moreover, it is still unclear in which clinical setting the use of immune checkpoint inhibitors in EOC is most favorable. PARPis in combination with immune checkpoint inhibitors might be most beneficial in primary disease or early recurrence due to a lower tumor burden. Future clinical trials will test these combinations in first-line treatment of breast and EOC.

Antiangiogenic Therapy

Preclinical and clinical evidences suggest that there are existing interactions between the VEGF pathway and PARP inhibition. Multiple groups have demonstrated that PARP inhibition reduces VEGF-induced angiogenesis. A preclinical study conducted by Bindra and team⁶⁷ demonstrated that hypoxia is related with impaired HDR and, therefore, a state of BRCAness. They also postulated that in a hypoxic state the cells are pushed toward NHEJ because of impaired HDR and thus show increased genetic instability and cell death. Liu et al⁶⁸ designed a phase I clinical trial with olaparib bid and cediranib, a tyrosine kinase inhibitor directed against VEGF. The RP2D was cediranib 30 mg daily with olaparib 200 mg bid. Subsequently, Liu⁶⁹ investigated the efficiency of olaparib in combination with cediranib in a phase II clinical trial in 90 patients with recurrent platinum-sensitive EOC. Results demonstrated a PFS of 17.7 months for treatment with olaparib and cediranib against 9.0 months for olaparib monotherapy. A post hoc analysis demonstrated that median PFS was even greater for patients with a g BRCA-1/2m: 19.4 months in the combination arm as compared with 16.5 months in the olaparib monotherapy group, respectively. Most common grade-3 toxicities reported in the olaparib plus cediranib group were fatigue, hypertension, and diarrhea. Dose reductions were necessary in 77 and 24% of patients, respectively. Zimmer⁷⁰ conducted a phase I clinical trial investigating the RP2D for durvalumab + olaparib + cediranib in recurrent female cancers. They concluded a RP2D of 1,500 mg q28d durvalumab + 300 mg BID olaparib + 20 mg 5 days on/2 days off cediranib is tolerable and active. Overall, antiangiogenetic therapy in combination with PARP inhibition showed promising results due to its efficacy and potential synergism. At present many clinical trials are investigating the combination therapy of bevacizumab or cediranib with PARPis in ovarian or breast cancer.

Resistance to PARPis

Lord and Ashworth described four pathways leading to resistance to PARPis (Fig. 3).⁷¹ Most of the clinical studies conducted in vitro, with mice with different kind of knockout genes. The first mechanism is the incidence of a secondary mutation in the affected BRCA gene that would lead to the restoration of the BRCA open reading frame. Due to this restoration, the BRCA gene can be translated and it can further lead to (partial) functional protein to repair DSBs. Multiple studies have found this phenomenon in patients who had developed resistance to PARPis. The second mechanism depends on the (partial) restoration of HDR due to the somatic loss of expression of genes involved in the regulation of DSB repair pathway choice, like the tumor suppressor p53-binding protein 1 (53BP1) or REV7.⁷² This mechanism is shown in vivo in mice. A very recent publication describes the identification of shield in, a complex of REV7, RINN1, RINN2, and RINN3 proteins. This complex restrains DNA end resection and thereby promotes NHEJ. Deletion of one of the shield in components leads to resistance to PARPis in BRCA-1 depleted cells.⁷³ Third, the upregulation of the P-glycoprotein efflux pump, pumps PARPis out of the cell, resulting in a decreased inhibition of PARP. The fourth mechanism is the hypothesis that poses that acquired PARP1 loss-of-function mutations or down-regulation of transcription can result in PARPi resistance.

Close

Fig. 3 PARP inhibitor resistance mechanisms. BRCA, breast-related cancer antigens; PARP, poly (ADP-ribose) polymerase; PTEN, phosphatase and tensin homolog.

Export to PPT

Another mechanism explaining resistance to PARPis is replication fork stabilization. Deficiencies in PAX-interacting protein 1(PTIP), chromodomain-helicase-DNA-binding protein 4 (CHD4), and PARP-1 limit the action of MRE11 to single-strand DNA at stalled replication forks. MRE11 is involved in the degradation of stalled replication forks. When MRE11-dependent replication fork degradation is absent due to deficiencies in PTIP, CHD4, or PARP1, nascent DNA strands will be protected from degradation; therefore, the cell will be resistant to PARP inhibition.⁷⁴ BRCA-deficient cells become resistant to different types of DNA-damaging agents through the loss of PTIP, PARP1, and CHD4. Furthermore, survival analysis of patients with EOC with a BRCA2 mutation treated with platinum chemotherapy demonstrated that high PTIP expression has a correlation with a longer PFS.

Furthermore, increased phosphorylation of ribosomal protein S6 leading to upregulation of the mTOR pathway⁷⁵ and upregulation of NF-kB (nuclear factor kappa light chain enhancer of activated B cells) signaling can also lead to PARPi resistance. Depending on these mechanisms it was hypothesized that PARPi-resistant tumors should be treated with rapamycin (an mTOR [mammalian target of rapamycin] inhibitor) or with bortezomib (a proteasome inhibitor). The combination of PARP inhibition and rapamycin effectively suppressed tumor growth in mice⁶⁹ and the combination of PARP inhibition plus bortezomib led to increased cell death in PARPi–resistant cells.⁷⁶ Further investigations (both preclinical and clinical) of mechanisms of PARPi resistance will direct us to strategies that will optimally use PARPis in the clinic.

Biomarkers

As far as the application of PARPis is considered beyond BRCAm tumors, it is necessary to identify those classes of patients that get benefited mostly from PARPis to maximize the effects of treatment, further prevent futile therapy and toxicity, and limit the financial burden of health care. Multiple studies have already demonstrated that there is an additional group of tumors with HDR deficiency that also respond to PARPis.⁵⁹ These tumors have a so called “BRCAness” phenotype, a deficiency in HR in the absence of a BRCA mutation, which makes them responsive to PARPis.

At present, different approaches have been implemented to further identify HDR deficiency or BRCAness. The first way is to analyze tumors for loss of functional mutations in genes that are involved in HDR. These genes include ataxia telangiectasia mutated (ATM), ataxia telangiectasia and Rad3-related (ATR), CHEK2, BRIP1, PALB2, RAD51C, and RAD51D. The second way is to detect BRCAness is via transcriptional signatures. Larsen and team⁷⁷ analyzed 55 familial germ line BRCA-1 or BRCA-2 mutated breast cancer patients and 128 patients with sporadic breast cancer. They designed a transcriptional signature to distinguish BRCA-1 tumors from sporadic tumors with an accuracy of 83% and BRCA-2 tumors with 89%, which was validated in independent datasets. This type of transcriptional signature might also be implemented to identify BRCAness in non BRCA-related tumors. The third way to detect BRCAness is through the detection of a genomic signature in a tumor. These signatures represent a pattern of mutations or genomic alterations which are characteristic for the use of error-prone repair pathways in the absence of HDR.⁷⁸ These patterns consist of specific nucleotide substitutions (e.g., mutational signature 3) or sequence microhomology at breakpoints.⁷⁹ An algorithm dependent on the integration of six signatures associated with BRCA deficiency (including somatic nucleotide substitutions, insertion/deletion, and rearrangement patterns), termed HRDetect was developed by Davies.⁸⁰ Nowadays, SNP-based profiling has also been implemented to define a so-called HDR deficiency score, based on the combination of 3 DNA-based measures of genomic instability (i.e., loss of heterozygosity [LOH], telomeric allelic imbalance [TAI], and large-scale transitions [LST]).⁸¹ This is the myChoice HDR deficiency test (Myriad Genetic) which was used in the NOVA clinical trial as discussed earlier, but it could not discriminate between patients that demonstrated a benefit from treatment niraparib or not.⁸² A more recent approach to detect BRCAness is through functional biomarkers or a functional test for current HDR deficiency.⁵⁹ After the induction of DNA DSBs in fresh tumor tissue ex-vivo, RAD51 protein will accumulate at the sites of the breaks. This key step in the HDR pathway can be visualized by immunofluorescent staining as foci in the nucleus. The inability to form RAD51 foci after the induction of DSBs in replicating tumor cells is a biomarker for BRCAness. This biomarker technique requires fresh tumor tissue obtained before starting chemotherapy.⁸³ Finally, in the ARIEL3 clinical trial,⁴⁸ the percentage of genome-wide LOH quantification was used to compare effectiveness of rucaparib among groups with high or low genomic LOH levels. Tumors with high LOH appear to respond better to PARPi therapy and, therefore, genomic LOH quantification might be used to identify patients who might benefit from PARPis.⁸⁴

PARPis and Antiangiogenic Agents

Overexpression of PARP-1 has been noted to exert a clear proangiogenic effect in EOC by upregulation of VEGFA.⁸⁵ Antiangiogenic therapies are well known to induce a hypoxic cellular state, leading to down regulation of HR repair genes (BRCA1, BRCA2, and RAD51), with consequent enhancement of PARPis sensitivity.⁸⁶ Hypoxia is also found to be related to hypoxia inducible factor–1 α (HIF1α) upregulation, which is considered one of the most common mechanisms of resistance to angiogenesis inhibitors. Interestingly, PARP-1 may play a pivotal role in HIF-1α stabilization.⁸⁵ As a consequence, the use of PARPis may prevent its accumulation and signaling, thus overcoming this resistance mechanism, resulting in death of the targeted hypoxic cell. This is a suitable example of “contextual synthetic lethality” because of tumor microenvironment, in which hypoxia-induced repair-deficient tumor cells can be targeted by disrupting backup pathways. The combination of cediranib, an oral tyrosine kinase inhibitor of VEGF receptor (VEGFR) and olaparib versus olaparib alone was tested in a phase II clinical trial in platinum-sensitive ROC.⁵¹ Median PFS was considerably longer for cediranib/olaparib-treated patients than for patients on olaparib alone (17.7 vs. 9 months, HR 0.42; p = 0.005). A post hoc exploratory analysis showed an increased activity of cediranib plus olaparib versus olaparib alone in the subgroup of patients with wild type or unknown BRCA status, with an improvement in the median PFS from 5.7 to 16.5 months (HR = 0.32; p = 0.008) and in the ORR from 32 to 76% (p = 0.006). Among gBRCAm patients, there was a lesser trend toward increased activity for the combination arm, with a slighter gain of PFS (16.5–19.4 months) and ORR (benefit 63–84%). Notably, side effects (fatigue, diarrhea, and hypertension) of any grade occurred in the combination arm, resulting in dose reductions in more than 75% of patients. Three phase III clinical trials are currently ongoing to validate this combination in different settings. The aim of GY004 trial was to compare olaparib monotherapy versus olaparib and cediranib combination versus standard platinum-based chemotherapy in patients with platinum-sensitive recurrent ovarian/fallopian tube cancers. GY005 trials are evaluating the same therapeutic options in recurrent platinum-resistant or refractory OC. Finally, ICON 9 trial is examining maintenance therapy with cediranib and olaparib or maintenance olaparib alone after platinum-based chemotherapy in patients with recurrent platinum-sensitive high-grade OC. The combination of olaparib 400 mg BID with bevacizumab 10 mg/kg q2w was investigated in a phase I study in patients with advanced solid tumors; a good tolerability without serious AEs or dose-limiting toxicities were recorded.⁸⁷ On the basis of these results, a phase III trial, PAOLA1, was planned to determine the efficacy of olaparib or placebo combined with bevacizumab as maintenance treatment in patients with OC treated with standard first-line platinum-based chemotherapy plus bevacizumab. Finally, the results of a phase I clinical trial of bevacizumab (15 mg/kg q21 days) plus niraparib (300 mg oral QD) demonstrated an ORR of 45% in 12 patients and DCR of 91% in the combination treatment arm.⁵³ Class toxicities (hypertension, anemia, thrombocytopenia, fatigue, constipation, and nausea) were manageable and only one dose-limiting toxicity was reported: grade 3 thrombocytopenia that persisted for ≥5 days. Results from AVANOVA, a phase II ongoing clinical trial, comparing single-agent niraparib with niraparib plus bevacizumab in 94 women with platinum-sensitive OC, are still pending.

PARPis and Immune Checkpoint Inhibitors

Achievement of tumor antigenicity sensitizes cancers to checkpoint blockade therapy. Multiple clinical studies have demonstrated that BRCA-1/2m and wt-BRCA1–2 HRD ovarian tumors display a higher neoantigen load than HR-proficient cancers.⁵⁴ PARPis have been considered to enhance the response to immunotherapy in HRD OC by yielding a greater mutational burden, thereby expanding neoantigen expression. In fact, the presence of danger signals, followed by DNA damage, activates the stimulator of interferon genes (STING) pathway which plays a vital role in the innate immunity by inducing type I interferon and proinflammatory cytokine production.⁵⁵ Moreover, PARPis administration upregulates PD-L1 expression on tumor cells, which in turn attenuates PARPis efficacy via cancer-associated immunosuppression. As a consequence, the targeted blockade of PD-L1 pathway can restore antitumor immunity and potentiate the antitumor activity of PARPis.⁵⁶ Altogether, these observations provide a scientific rationale for evaluating the combination of immune checkpoint blockade with PARPis in OC clinical trials. The phase I/II TOPACIO clinical trial³ demonstrated niraparib in combination with pembrolizumab, an anti–PD-1 antibody, to be a promising therapeutic option for the management of platinum-resistant OC. Following dose finding in phase I, the RP2D of niraparib and pembrolizumab was assessed at 200 mg orally QD and 200 mg IV every 21 days, respectively. During phase II, among 60 out of 62 enrolled patients who were considered evaluable for initial response assessment, 64% of them were platinum-resistant, 19% had platinum-refractory disease, and 17% had platinum-sensitive OC ineligible for further platinum. In the whole population, ORR and DCR were 25 and 68%, respectively, while in the BRCA1/2m cohort, ORR and DCR were 45 and 73%, respectively.

In regard to olaparib, phase I/II basket MEDIOLA clinical trial evaluated the combination with durvalumab, an anti-PDL1 antibody, for the management of gBRCA 1/2m platinum-sensitive relapsed OC.⁵⁷ Thirty-two patients were given olaparib 300 mg (tablets) BID for the period of 4 weeks, followed by a combination of olaparib 300 mg BID and durvalumab 1.5 g intravenous (IV) every 4 weeks, until disease progression. DCR at 12 weeks was 81% while ORR was 63%. This combination was well tolerated, with a low incidence of grade 3 toxicities and all-grade immune-related AEs. The results of TOPACIO clinical trial and MEDIOLA study were presented at European Society of Gynaecological Oncology (ESGO) congress 2018. In another phase II clinical trial the combination of durvalumab and olaparib was tested in 34 ROC patients with platinum-sensitive or -resistant disease. The preliminary efficacy results, which were presented during the ESMO congress 2018, showed a response rate and DCR of 15 and 53%, respectively. The combination of olaparib at a dose of 300 mg BID with tremelimumab, a CTLA-4 antibody, at a dose of 10 mg/kg monthly, was tested in another phase I clinical trial⁴ for the treatment of BRCA-1/2m ROC and demonstrated good safety.

Multiple clinical studies assessing the efficacy of PARPis in combination with immunotherapy are ongoing. A phase I/II trial is evaluating side effects and optimal dose of olaparib in combination with durvalumab and tremelimumab for the treatment of ROC. The phase III FIRST clinical trial was designed to evaluate platinum and TSR-042 (PD-L1 inhibitor) followed by niraparib and TSR-042 maintenance therapy versus adaptive standard platinum-based treatment for newly diagnosed advanced OC patients. In addition, ENGOT-ov46/AGO/DUO-O is a phase III clinical trial whose aim is to assess the efficacy and safety of standard of care platinum-based chemotherapy and bevacizumab followed by maintenance bevacizumab either as monotherapy, or in combination with durvalumab, or in combination with durvalumab and olaparib for newly diagnosed advanced OC patients. Regarding rucaparib, phase III ATHENA study is investigating the association with nivolumab as maintenance therapy in OC patients, after response to frontline platinum-based chemotherapy.

PARPis and Other Therapeutic Agents

As PARPis emerged as an efficient therapeutic strategy in the management of HR-deficient tumors, the real question is to establish whether PARPis can also be effective in the treatment of HR-proficient carcinomas. The condition of HR proficiency, which plays a major role in PARPis resistance, can occur de novo or be acquired. In the latter case, epigenetic or genetic events occurring under selective pressure due to PARPis exposure are responsible for reversing the original HR alterations, hence leading to HR restoration. The most commonly acquired mechanism of resistance to PARPis is a somatic genetic reversion of the original truncating BRCA mutation that restores functional protein expression.⁵⁸ Alternatively, an acquired epigenetic reversion of BRCA-1, like promoter hypermethylation, has been noted to restore normal BRCA-1 protein expression levels.⁸⁸ Of note, tumors carrying a specific BRCA-1 mutation, which further disrupts the N-terminal RING domain, respond poorly to platinum drugs and PARPis and rapidly develop resistance.⁵⁹ Another possible mechanism of PARPis resistance may be the decreased expression of PARP enzymes due to epigenetic silencing or accelerated/high protein turnover.⁶⁰ Furthermore, intracellular concentrations of PARPis could be reduced by increased P-glycoprotein–mediated efflux, thus resulting in decreased antitumor effect.⁶¹ The loss of 53BP1 in BRCA1m tumors, partially restoring the error-free repair mechanism mediated by HR, enhances DNA-damage tolerance and induces PARPis resistance.⁵⁴

Finally, HR proficient OC with concurrent amplification of cyclin-E genes has been demonstrated to be resistant to PARPis.⁵⁵ As a consequence, the rationale to combine PARPis with molecularly targeted agents capable of inhibiting HRR may represent a promising effective strategy to expand their use in HR-proficient OC.⁸⁴ Among all the combinations that are currently under assessment, the association with PI3K inhibitors appears the most reasonable approach, as its phase I evaluation in ovarian and triple-negative breast cancers has just been completed.⁶² The rationale for this trial was based on two preclinical studies published in 2012,⁶³ which showed that PI3K inhibition significantly decreased BRCA1/2 expression, thus resulting in acquired HRD underlying the antitumor effects of PARPis. In the phase I study, combined exposure to olaparib and BMK120 showed an ORR of 29% in 46 advanced OC patients, irrespective of platinum-sensitivity status.⁶⁴ The maximum tolerated dose was 50 mg QD of BKM120 and 300 mg BID of olaparib. Randomized phase II studies are warranted to further define the efficacy of PI3K/PARP-inhibitor combinations compared with PARPis alone in different settings of ROC.⁶⁶ Finally, the use of PARPis after PARPis represents an interesting field of investigation. In this direction, the OReO/ENGOTOv-38 is ongoing trial that will evaluate efficacy and safety of maintenance retreatment with olaparib in patients that retain platinum sensitivity despite progression on olaparib maintenance therapy.

Related Clinical Trials of PARPis

Clinical studies of PARPis, including olaparib, rucaparib, and niraparib, in OC are summarized in Table 2. To begin with, the FDA permitted olaparib as the fourth-line treatment for advanced OC with gBRCAm, based on the results from a phase II study representing an objective response rate (ORR) of 31% and a median overall survival (OS) time of 16.6 months with olaparib treatment in 193 OC patients [NCT01078662]. Patients with platinum-resistant disease or those unsuited for further platinum therapy due to noteworthy toxicity or hypersensitivity to platinum, were also included in this study. This level of activity appreciably exceeded that of conventional third-/fourth-line therapy; hence, the FDA approved olaparib for this indication.¹⁰⁸ A pooled analysis of 6 phase I/II clinical trials [NCT00516373 (Study 2), NCT00777582 (Study 24), NCT00494442 (Study 9), NCT00628251 (Study 12), NCT00679783 (Study 20), and NCT01078662 (Study 42)] recognized the ORR as 36% and the median duration of response as 7.4 months with olaparib treatment among patients with gBRCAm and advanced relapsed OC. The ORR among patients who had received three or more lines of prior chemotherapy was found to be 31%, with duration of response of 7.8 months, showing that a sustained response to olaparib could be achieved in heavily pretreated, relapsed, gBRCAm-associated OCs.¹⁰⁹ The highest tolerated dose (MTD) of olaparib was recognized as 400 mg twice daily in a phase I trial [NCT00516373].⁴⁴

A dose–response association between diverse olaparib dose levels was studied in two phase II clinical trials. In the first phase II study [NCT00628251], the ORR was observed to be higher in the 400 mg olaparib group (31%) than in the 200 mg olaparib group (25%).¹¹⁰ In an additional phase II study [NCT00494442], a variation of 3.9 months in the median PFS was reported in the 400 mg olaparib group as compared with the 100 mg olaparib group, in favor of the 400 mg olaparib group. And the ORRs were 33% and 13% in the 400 mg olaparib and 100 mg olaparib groups, respectively.⁴³

SOLO2 and SOLO1 are all randomized, double-blind, placebo-controlled phase II/III clinical trials of olaparib monotherapy that are highly noteworthy showed that olaparib maintenance monotherapy considerably improved PFS (median 8.4 vs. 4.8 months; HR 0.35; 95% CI 0.25–0.49; p < 0.001) as compared with placebo in patients with platinum-sensitive, recurrent HGSOC who had received two or more prior lines of platinum-based chemotherapy and established a CR/PR to the most recent platinum-based chemotherapy.⁸⁹ Retrospective germ line and somatic BRCA mutation testing was conducted on all patients with an supplementary 2 years of follow-up. A total of 51% of the HGSOC population showed germ line or somatic BRCA mutation, and patients with or without g/sBRCA mutations both obtained the PFS benefit from olaparib maintenance therapy versus placebo, with a greater PFS benefit in the g/sBRCA1/2-mutated group as compared with the wild-type BRCA group. The first, second and third interim OS analyses from study 19 were conducted after 38, 58, and 77% of patients had died, respectively. The final OS analysis was performed after 210 deaths (79% data maturity), after a median follow-up of 6.5 years.¹¹¹ Neither of the first¹¹² or second interim OS analyses demonstrated a benefit for olaparib versus placebo for either the BRCA1/2-mutated or BRCA wild-type groups in the overall population.¹¹³ The third interim OS analysis and the final OS analysis both demonstrated an OS advantage of olaparib against placebo in all patients and inpatients with BRCA mutations. However, the predefined threshold for statistical significance (p = 0.0095) was not met. In the final OS analysis, 32 patients (24%) had received olaparib maintenance therapy for over 2 years, and 15 (11%) had received olaparib maintenance therapy for over 6 years, which showed the long-term safety and tolerability of olaparib maintenance therapy.¹¹⁴ Therefore, olaparib maintenance appreciably improved PFS in patients with platinum-sensitive recurrent HGSOC treated with two or more previous lines of platinum-based chemotherapy, and patients with a g/sBRCA mutation reported the greatest benefit from olaparib. The analyses for time to first subsequent therapy or death, time to second progression, and time to second subsequent therapy or death reported that the PFS benefit was continued until succeeding treatment and that a long-term benefit was achieved irrespective of BRCA1/2 mutation status. The long follow-up time necessary to obtain sufficient OS data increases the chance that postprogression PARP inhibitor therapy and patient crossover will affect the OS data. When excluding the patients from places where placebo patients were treated with postprogression PARP inhibitors, the OS hazard ratio was found to be extensively improved, demonstrating that in study 19, postprogression PARP inhibitor therapy had a confounding effect on the interim OS analysis for patients with BRCA mutations. SOLO2 [NCT01874353] intended to look into the effectiveness and safety of olaparib in platinum-sensitive, recurrent OC patients with a g/sBRCA1/2 mutation who had received two or more lines of previous chemotherapy and confirmed a CR/PR to the most recent platinum-based chemotherapy. The median PFS was considerably longer with olaparib (19.1 months) than with placebo (5.5 months). The PFS benefit from olaparib maintenance as compared with that from placebo in SOLO2 to a large extent exceeded that was reported in study 19, which is not surprising as SOLO2 included only patients with g/sBRCA1/2-mutated tumors.⁴⁶ Furthermore, heavily pretreated patients with BRCA1/2-mutated OC whose disease progressed following SOLO 1 was the first clinical trial to further investigate the efficacy of olaparib as a first-line maintenance therapy for primary advanced OCs. Patients who had no evidence of disease at the completion of 2 years stopped receiving the trial intervention. Maintenance olaparib led to an extensive development in the PFS of patients with newly diagnosed advanced OC and BRCA mutation, with a difference of ∼3 years in the median PFS for olaparib as compared with that of placebo. The Kaplan–Meier curves for the olaparib group did not appreciably change after 2 years, which suggests an enduring treatment benefit after treatment stop. The second PFS showed a statistically significant upgrading, suggesting that olaparib did not reduce patients’ ability to benefit from subsequent therapy. Considering the significant PFS advantage in favor of first-line maintenance therapy with PARPis, patients would have to undergo g/sBRCA testing immediately after OC diagnosis and adopt the PARPi first-line maintenance therapy if they were found to be positive for g/sBRCA mutation. Moreover, the PFS benefit from maintenance olaparib as compared with that from placebo in SOLO1 also substantially exceeded that in SOLO2, showing that olaparib is more beneficial to BRCA mutation carriers as a first-line maintenance therapy than as third-line treatment.

Latest FDA Guidelines: Major Updates in Maintenance Therapy

The FDA guidelines make quite a few updates in maintenance therapy in stage 2, 3, and 4 disease: Olaparib is recommended as first-choice of maintenance therapy for patients with BRCA1/2 mutations in absolute clinical remission or partial remission (Fig. 4). The recommendation is category 1 for germ line mutations and category 2B for somatic mutations; O’Malley and team noted that this occurred because there were only few patients with somatic mutations studied. The recommendation applies whether or not the patient was previously treated with bevacizumab.

Close

Fig. 4 Summary of FDA-approved PARP inhibitors. BRCA, breast-related cancer antigens; FDA, Food and Drug Administration; PARP, poly (ADP-ribose) polymerase.

Export to PPT

The recommendation for olaparib is dependent on results from the SOLO-1 clinical trial, which studied the PFS depending on RECIST criteria and noted that as compared with placebo, median PFS was not reached in the olaparib arm and was 13.8 months in the placebo arm.⁴⁶ Bevacizumab is also suggested for maintenance therapy post remission for patients with partial or whole responses who received it in primary treatment, or for patients with stable disease. Updates for bevacizumab were based on the GOG-218¹¹⁵ and the ICON7 clinical trials, which O’Malley reviewed. GOG-218 was cited in the June 13, 2018, FDA approval for bevacizumab in combination with paclitaxel or carboplatin, followed by bevacizumab as a single agent, for stage 3 or 4 epithelial ovarian, fallopian tube, or primary peritoneal cancer after initial resection.¹¹⁶

In reviewing the ICON7 study results, O’Malley noted that even though the overall results for did not reach statistical significance, bevacizumab was found to be very effective for the highest-risk patients; published results showed that the estimated median PFS was 10.5 months with standard therapy, versus 15.9 months with bevacizumab (HR 0.68; 95% CI 0.55–0.85; P < 0.001).