The evolving role of PARP inhibitors in advanced ovarian cancer

Olaparib is the first polyADP-ribose polymerase inhibitor approved for ovarian cancer both by EMA and FDA in 2014^[36]. It blockades in a reversible way PARP-1, −2 and −3 isoforms of the PARP family enzymes. The trapping of PARP1 is stronger and it constitutes the main mechanism of olaparib cytotoxic activity^[13]. Olaparib has low aqueous solubility characteristics and no lipophilic compound, which is translated in low membrane permeability. This may not affect the drug absorption but it impedes its penetration in the brain, so that the final tumor tissue exposure is about 60% less of the plasma concentration^{[37, 38]}. A fat or high calorie meal may cause delay in drug absorption and peak plasma concentration, but the final bioavailability remains intact, so that patients can receive the drug regardless of food^[48]. A steady state situation in plasma occurs within 3–5 days of drug initiation and it is almost 0.8 times higher when olaparib is taken as a tablet^[49]. It is metabolized in the liver, primarily mediated by CYP3A4 and CYP3A5, whilst its metabolites are excreted in feces (12–60%) and urine (35–60%). A dosage adjustment is necessary in severe liver and renal insufficiency, as well as with concomitant use of a CYP3A inhibitor^[48].

Olaparib is available for oral administration, initially only as capsule of 50mg and now as a tablet of 100mg and 150mg, as well. The new dose is 300mg twice daily. The need for another pharmaceutical form arose in an attempt to improve drug bioavailability. As capsules and tablets are not bioequivalent, the one formulation cannot be changed or replaced by the other one in any indication and time of treatment. Since 2017, based on SOLO-2 results, the tablet is the indicated pharmaceutical form for every clinical practice^{[36, 50]}. In fact, all phase III trials published, used the tablet formulation and FDA has permanently discontinued the capsule since 2018.

The efficacy and safety of olaparib in patients with germline BRCA1/2 mutations and recurrent ovarian cancer have been assessed in a number of trials, both in platinum resistant disease in comparison with chemotherapy, and in platinum sensitive disease as maintenance treatment^{[14, 51,52,53,54]}. Olaparib was initially granted FDA approval as monotherapy for patients with advanced OC and germline BRCA mutations, who had been previously treated with three or more chemotherapy regimens, following the results of a phase 2 study^{[55, 56]}. Benefit from olaparib monotherapy in terms of tumor objective response rate (ORR) (31.1%), duration of response (median, 225 days), PFS (median, 7 months) and OS (median, 16.6 months) was reported. Differences regarding the RR between BRCA1 and BRCA2 mutation carriers were not considered of note^{[55, 56]}.

The first indication of PARPi efficacy in OC patients without germline BRCA mutations was introduced after a multicenter, phase 2 trial, in patients with advanced, high-grade serous or poorly differentiated OC, as well as breast cancer^[57]. Objective responses were observed in all women with HGSOC, both with and without BRCA mutations, particularly in those with platinum-sensitive carcinomas, supporting the hypothesis that factors implicated in the HR repair mechanism, besides germline BRCA mutations, may predict sensitivity to PARPi treatment. Furthermore, the reported data of Study 19, confirmed PFS benefit in women with platinum-sensitive HGSOC and suggested that further investigation is required regarding the efficacy of PARPi in women with BRCAness phenotype^[14]. The prearranged analysis according to BRCA status indicated that the origin of BRCA mutations was not associated with the response to PARPi treatment, as women with somatic BRCA mutations gained similar benefit^[53]. A further updated analysis of these patients revealed statistically significant improvement regarding PFS (HR 0.23) and OS (HR 0.15)^[58].

Since then, more robust evidence for the efficacy of olaparib has been aggregated from phase 3 trials in various treatment settings. SOLO2 was designed to determine the efficacy of olaparib versus placebo, as maintenance therapy in platinum-sensitive, relapsed, BRCA-associated OC patients^[50]. Germline BRCA mutations were confirmed in all patients and deemed pathogenic or likely pathogenic in 97% of cases. Maintenance olaparib extended median PFS by 13.6 months with an HR of 0.30. Furthermore, improvement was observed regarding time to first subsequent therapy (median, 27.9 vs 7.1 months), time to second subsequent therapy (HR 0.37), as well as time to second progression or death (HR 0.50). As a result, the indications were expanded by both the FDA and the EMA in the maintenance therapy setting, to include relapsed OC patients, with response to previous platinum-based chemotherapy, irrespective of their BRCA status^{[45, 59]}. Very recently, final impressive overall survival data was announced, with a median OS improvement of nearly 13 months^[60]. Regarding first-line maintenance therapy, the results of the SOLO1, confirmed the efficacy of olaparib in this additional setting^[61]. The vast majority of participants was germline BRCA mutant (388/391 patients); two patients had somatic BRCA mutations and one had a BRCA variant of uncertain significance. Olaparib maintenance offered 36 months median PFS benefit, compared to placebo. Similar significant improvements were observed in time to first subsequent therapy and time to second disease progression, while OS did not appear to differ substantially at that point. Safety profile was as expected.

SOLO3 trial tested whether olaparib was more effective than physician-selected chemotherapy in women with germline BRCA-associated, platinum-sensitive, relapsed HGSOC treated with at least 2 prior platinum-based chemotherapy lines. ORR proved to be significantly higher (72.2% vs 51.4%) and PFS was further improved in favor of the olaparib group (median, 13.4 vs 9.2 months)^[62]. The clinical relevance of these findings remains debatable, since at present most of the patients receive olaparib as first-line maintenance therapy. Until new evidence surfaces regarding the role of olaparib in heavily pretreated patients, including PARPi as prior therapy, this question stands unanswered. On the other hand, olaparib monotherapy may be more relevant in patients with relapsed, platinum-resistant OC, as it is indicated by the preliminary results of the ongoing CLIO trial^[63]. ORR for olaparib monotherapy proved superior against standard-of-care chemotherapy (18% vs 6%, respectively), especially in patients with germline mutations. Further analyses of patients with somatic BRCA mutations are ongoing. Finally, the efficacy of olaparib has been tested in combination with other antitumor agents, which is discussed in another section.

Rucaparib is a small molecule inhibitor of PARP-1, −2 and −3 and of tankyrase-1 and −2. It is considered the less selective PARPi and its strongest trapping target is PARP-1. It's efficacy is increased by blockading both PARPs and hexose-6-phosphate dehydrogenase (H6PD)^[64]. Rucaparib's full name is rucaparib camsylate and it belongs to camphorsulfonate salts. Its water solubility is low, irrespective of the pH range, while its permeability is low- to moderate^[45]. Rucaparib was tested both intravenously and orally, whereas oral rucaparib achieved better results with regard to ORR and disease stabilization^[65]. It is available as an oral tablet and can be administered with or without food, though there is a moderate effect of food in the drug pharmacokinetics^[66]. Its average bioavailability is 36% and the maximum concentration is reached in 1.9 hours. Rucabarib succeeds a steady state in 1 week following the indicated dose of 600mg twice daily^{[39, 66]}. The drug is metabolized in liver via oxidation and glucuronidation, primarily by CYP2D6 and to a lesser extent by CYP1A2 and CYP3A4. The major metabolite is named M324 and it is inactive. The drug is excreted both intact and through its metabolites through feces (>72%) and urine (~17%). In cases of mild to moderate renal disease or in mild liver insufficiency no dose adjustments are required. There are no data for more severe impairments^{[39, 45]}. Rucaparib distribution in the brain is limited and this is potentially due to its interaction with p-glycoproteins, implying a drug resistance mechanism^{[67, 68]}. Regarding interactions with other drugs, two factors should be taken under consideration when rucaparib is administered: first that rucaparib inhibits p-glycoproteins and second its relationship with the CYP-enzymes.

Rucaparib, has shown acceptable toxicity profile and antitumor activity in patients with germline and somatic BRCA mutations and HRD^{[69,70,71,72]}. ARIEL2 was designed to determine whether rucaparib would maintain its efficacy across three subgroups of women with platinum-sensitive, relapsed HGSOC, based on BRCA mutation status and LOH; BRCA mutant, BRCA wild-type with high LOH and BRCA wild-type with low LOH^[71]. PFS was significantly improved in women with BRCA mutations as compared to the other subgroups, irrespective of mutation's origin (germline or somatic)^{[71, 72]}. Proceeding to the maintenance clinical setting, ARIEL3 recruited 564 patients with platinum-sensitive, recurrent HGSOC, with complete or partial response to their previous platinum-based therapy. PFS advantage was consistent across all prespecified subgroups, regardless of parameters such as LOH, BRCA status, disease at baseline or response to previous treatment. With regard to the HRD cohort, significant PFS prolongation was also noted for the patients on rucaparib and this overall benefit was consistent, regardless of the presence of BRCA mutations^[73]. However, the presence of mutations in other HR genes, both in LOH high and low subgroups, suggests that more sensitive biomarkers, predictive of the PARPis response, are required. Based on the results of Study 10, ARIEL2 and ARIEL3, rucaparib, initially approved by the FDA in 2016 for advanced high grade serous ovarian cancer with germline BRCA mutations as a third or later line treatment, was also approved in 2018 by the FDA as a maintenance treatment for the platinum sensitive disease irrespective of BRCA status^{[71, 73]}. With regard to maintenance treatment setting, its role is going to be elucidated by the MAMOC trial (NCT04227522), which is going to be initiated soon. Patients will be randomized to receive rucaparib vs. placebo, until disease progression and/or death, or unacceptable toxicity, after receiving bevacizumab maintenance for 12 to 15 months.

Niraparib is the first PARPi to show PFS benefit as maintenance treatment in both BRCA mutant and wild type women with platinum sensitive recurrent OC. It equally and selectively blockades PARP-1 and −2, which is a common characteristic with veliparib. They are both carboxamides, so their carboxyl terminal domain interacts via hydrogen bonds with the α-helical domain of the PARP-1 and −2 enzymes, a subdomain which does not exist in other PARPis^[40]. Niraparib has a crystal structure and is classified according to BCS as class II. Its solubility is independent of the pH conditions. It is disposable only as capsule of 100mg and the standard dose of any clinically relevant indication is 300mg once daily. Niraparib is very rapidly absorbed irrespective of time or quality of meals intervals, having a bioavailability of 73% in humans, and reaching the peak plasma concentration in about 3 hours. Due to its high membrane permeability, it can cross the blood-cerebrospinal barrier, leading to brain concentration of 0.9-fold the concentration in plasma. Regarding tumor exposure to niraparib, both preclinical models and phase I trials have shown that tumor concentration ratio is 3-fold higher of that of the plasma^{[37, 38]}. It is metabolized in the liver primarily by carboxypeptidases (CEs) in a diphasic process^{[41, 74]}. The P450 enzymes have a minor role in niraparib metabolism, thus the risk of drug-to-drug interaction is uncommon, as are high aminotransferase levels. Niraparib is metabolized in more than ten non-active metabolites and it is excreted both in its initial formation and its metabolites, in feces (~39%) and in urine (~47%)^[74]. Its half life time is 36 hours. No dosage adjustments are needed in mild-to-moderate liver and renal impairment, while there are no pharmacokinetics data and consequently no accurate guidelines for its administration in severe impairments.

Niraparib was initially tested in HRD solid tumors in 2013 and gained approval by both the EMA and the FDA in 2017, based on the results of the ENGOT-OV16/NOVA trial^[75,76,77]. Among 553 patients with platinum-sensitive recurrent OC, 203 were germline BRCA mutation carriers, 138 of whom were assigned to 300 mg niraparib. This group performed significantly better in terms of PFS compared to those on placebo (median 21.0 vs 5.5 months), irrespective of their best achieved response to the last platinum-based therapy. Among the non-germline BRCA mutation cohort, further subgroups were created according to HRD status. Niraparib proved beneficial in almost every subgroup in terms of PFS, chemotherapy-free interval and time to first subsequent treatment. Women with HRD positive tumors experienced significantly prolonged PFS when treated with niraparib compared to placebo (median, 12.9 vs 3.8 months; HR 0.38). It is noteworthy that the favorable outcomes of niraparib were retained across subgroups with HRD positive and BRCA-wild type tumors, but also in HRD-negative patients^[75,76,77]. In 2019 FDA approved also its use as fourth line treatment in women with HRD ovarian cancer^[78].

PRIMA trial consolidated niraparib's pivotal position in patients with HRD^[80]. HRD status was defined as one of the stratification parameters and 373 women (50.9%) with response to first-line chemotherapy were categorized as deficient. In patients with HRD, significant PFS benefit of 21.9 months was reported with niraparib versus 10.4 months compared with placebo (HR:0.43). The risk of progression or death was generally lower in women regardless of HR deficiency or proficiency, but the greater effect was displayed in patients with BRCA-associated tumors^[79]. The antitumor activity of niraparib has been further illustrated in QUADRA trial, in heavily pretreated platinum-sensitive HRD-positive HGSC patients with no prior exposure to PARPi^[80]. The following analysis of the primary efficacy group yielded an ORR of 28% and a median PFS of 5.5 months. The strongest effect with regard to OS was 26 months observed in BRCA-mutant patients, followed by 19 months in the HRD-positive women. Despite the broad range of patient populations in this study, as no restrictions to mutational status or platinum-response were applied, the requirement for PARPi-naive patients and the subsequent per-protocol analysis of results raise concerns regarding the role of PARPi in this line of treatment. Finally, niraparib is about to be tested in the neoadjuvant treatment of ovarian cancer in the phase II NEOPRIMA trial.

Talazoparib targets PARP1, −2, −3 and tankyrases 1 and 2. It is unclear where talazoparib owes its unique properties regarding trapping. The theory of the allosteric effects in PARP proteins is questionable, whereas crystalized chiral-center structure could play a role^{[40, 46, 81]}. The phenomenon is even more obvious in populations with BRCA1/2 and PTEN dysfunctions. Although potency seems to be associated with bone marrow toxicity in both in vitro and in vivo data, it remains unclear if it is indicative of a better therapeutic result^[47].

Talazoparib is available as an oral capsule and the indicated dose for its clinical use as monotherapy is 1g daily. After every dose administration it reaches its maximum plasma concentration in 2 hours while it needs 2 to 3 weeks from the starting dose to succeed a steady state situation. Food intake has a moderate effect in talazoparib bioavailability, which in general ranges between 55% – 65%. Thus, patients can take the drug regardless of meals. It has a low solubility, which is pH dependent, and moderate permeability, a fact that restricts the delivery of talazoparib in the central nervous system^[82]. Most of the drug (>65%) is eliminated unchanged through urine, while it is a potent p-glycoprotein inhibitor, so co-administration with similar drugs should be avoided^[44]. No dose adjustments are recommended for mild or moderate renal or hepatic dysfunction, while there are no data for its safety in more severe impairments.

Talazoparib has been mostly investigated in patients with BRCA1/2 mutations and PTEN mutated models, mainly in breast cancer populations, where it has an official treatment approval. As far as ovarian cancer is concerned, the field of investigation is limited in small pilot and phase I/II trials, which test talazoparib as monotherapy with optimistic results, but are still not adequate to forward its further investigation in phase III trials^{[44, 83]}. Of special interest was the attempt, that tried to find out if the drug could be active after a prior exposure to other PARP inhibitor, which was nevertheless, early terminated (NCT02326844)^[84].

Veliparib is the smallest PARPi and weakest, with regard to its trapping efficacy, whereas it is the most selective together with niraparib. It is orally available and the indicated dose is 400mg twice daily continuously. It is very rapidly absorbed and food only moderately affects the C_max and T_max time, so its administration is independent of meals^[42]. It is a very soluble molecule, categorized as class I in the BCS, having the ability to penetrate the central nervous system. Veliparib's bioavailability is high and reaches at least 73%^[85]. The majority of the drug is excreted unchanged in urine, while 20% is metabolized in the liver, mainly by CYP2DA. In renal or liver dysfunction, and potentially in CYP2DA insufficiency, drug exposure is increased, however it is not well-documented if a dose adjustment under these circumstances is necessary^{[43, 86]}. The ongoing trial NCT01366144 which combines veliparib with paclitaxel and carboplatin in solid tumors in patients with liver and renal impairment may elucidate this question. As it does not induce or inhibit the major Cytochromes P450 and it does not seem to significantly interact with P-gp, it is considered safe when co-administered with other drugs^[43].

Veliparib's efficacy was established in early phase trials, demonstrating promising antitumor activity^{[87, 88]}. In another phase 2 study, the results of 50 enrolled patients with germline BRCA mutations and recurrent OC were not as remarkable^[89]. Responses were reported in 35% of patients with platinum-sensitive, as compared to 20% of those with platinum-resistant disease; however, this difference did not reach statistical significance. Veliparib has still not managed to gain an approval in the treatment of ovarian cancer in any setting of disease, neither in USA nor in Europe. Combination treatment with veliparib will be discussed below.

Fluzoparib (SHR-3162) is a new PARP1 inhibitor, orally bioavailable, with favorable pharmacokinetic properties and tolerability in phase I trials. It has shown efficacy in HRD tumors, especially in platinum sensitive ovarian carcinomas^{[90, 91]}. Currently, ongoing phase III trials investigate the potent efficacy of fluzoparib as maintenance treatment in platinum-sensitive recurrent ovarian cancer (NCT03863860), and furthermore its combination with apatinib (VEGFR2 inhibitor) in the 1^st line maintenance setting (NCT04229615).

Pamiparib (BGB-290) is another new PARP 1/2 inhibitor with high selectivity and capability to penetrate the central nervous system^[92]. As all other PARPis, pamiparib is orally disposable. It is currently under investigation as maintenance treatment in platinum-sensitive ovarian cancer (NCT03519230), and in BRCA1/2 mutated ovarian cancer (NCT03933761)^{[93, 94]}.

E7449 (2X-121) is also an oral PARP 1/2 and tankyrases 1/2 inhibitor, which can enter the blood-brain barrier. Its action seems to be independent of p-glycoprotein expression. As many of the already known PARP inhibitors, it has been tested in preclinical BRCA 1/2 mutated models alone or with alkylating drugs, proving its antitumor activity and its capacity to potentiate DNA-damaging^[95]. At the moment, it is being tested within a phase II trial (NCT03878849) as a 3^rd line treatment in recurrent ovarian cancer, regardless of BRCA or HR status^[96].

Several trials have been conducted combining PARPis with chemotherapy in different types of cancer. Table 2 shows olaparib's combinations with chemotherapy agents. Regarding Olaparib's combination treatment with the classic chemotherapeutic drugs, namely carboplatin and paclitaxel in the platinum-sensitive setting and liposomic doxorubicin in the platinum-resistant setting, have been attempted. Both treatment strategies had promising clinical responses in terms of ORR and PFS benefit, especially for the subgroup of BRCA mutant patients^[98,99,101]. In spite of this, no phase III trials have been conducted to investigate if this is an effective treatment strategy, maybe due to overlapping myelosuppression, modifying treatment strategies to metronomic schedules. Moreover, paclitaxel and carboplatin combination with olaparib are now investigated in the neoadjuvant treatment of BRCA mutant ovarian cancer (germline and/or somatic) in the active, phase II NUVOLA trial.

Rucaparib is now under investigation as a combination treatment with many other therapeutic agents. Table 3 demonstrates all the currently ongoing rucaparib's combination trials. Combination treatment of rucaparib with chemotherapy has been reported only within phase 1 trials, and proved to be safe^[65]. According to our knowledge, there are no clinical trials which combine niraparib with alkylating agents. Talazoparib, is another PARPi which has been combined with other agents in early phase trials, especially with alkylating drugs (carboplatin and temozolamide) and the topoisomerase 1 inhibitor irinotecan, playing the role of the chemosensitizer^{[82, 102, 103]}. Limited data exist regarding its safety and efficacy in combination with chemotherapeutic agents in OC.

Veliparib's unique characteristic to act as a sensitizing factor for radiotherapy and chemotherapy increases its DNA-damaging efficacy^[92], making veliparib perhaps the ideal PARP inhibitor for potential combinations. It has been combined with many different chemotherapy agents such as alkylating agents, paclitaxel, liposomic doxorubicin, and topoisomerase inhibitors, proving to be tolerable in phase I trials and sometimes also effective (Table 4)^{[104,105,106,107,108]}. It has been also investigated as a radiosensitizer in peritoneal carcinomatosis, in a small phase I trial which included only 4 patients with ovarian cancer. The combination was tolerable, while only one gBRCA mutated patient with platinum sensitive disease achieved a response^[109]. Recently, results of VELIA trial, which tested the addition of veliparib to the standard neoadjuvant therapy and its continuation as maintenance therapy, were announced^[90]. Of the 1140 patients randomized, 214 had a germline BRCA mutation, while 84 had a BRCA mutation of somatic origin. Three nested cohorts of BRCA mutation, HRD and intention-to-treat population were prespecified. In general, substantial PFS benefit was demonstrated across all cohorts for the veliparib-throughout group. In the intention-to-treat cohort, median PFS was 23.5 versus 17.3 months for the veliparib-throughout and control group respectively (HR 0.68). Similarly, in the BRCA mutant population, a median PFS of 34.7 months was achieved in the veliparib-throughout group compared to 22.0 months in the control group (HR 0.44). The HRD cohort population achieved significantly longer PFS in comparison with the control group (median, 31.9 vs 20.5 months; HR 0.57). Women with BRCA mutations appeared to derive the greatest advantage from PARPi therapy, while those with HRD carcinomas had a borderline significant reduction in risk of progression or death. Interestingly, these differences were not observed during comparisons of the veliparib-combination-only with the control group, in either cohort^[91]. Therefore, the role of veliparib as an addition to induction chemotherapy, remains questionable.

The inhibition of PARP enzymes in parallel with the VEGF pathways seems also an effective strategy. The addition of cediranib, an anti-VEGFR inhibitor, to olaparib vs. olaparib alone demonstrated clinical activity not only in women with recurrent, platinum-sensitive OC, but also in platinum resistant disease^{[110,111,112]}. Surprisingly, subgroup analysis revealed that patients without germline BRCA mutations performed better with the combination, compared to BRCA mutant patients, in terms of both PFS and RR^[111]. Updated analysis revealed no OS benefit for the combination in the overall population, though OS was significantly increased by the cediranib/olaparib combination in the BRCA wild-type patients^[113]. The same combination when tested in the phase III GY004 trial, failed to meet its primary PFS endpoint in platinum-sensitive relapsed ovarian cancer, compared to standard of care chemotherapy^[114].

While most olaparib trials referred to patients with germline BRCA-associated OC, PAOLA-1 trial was designed to assess olaparib and bevacizumab as first-line maintenance treatment combination in high-grade OC patients, irrespective of BRCA status^[115]. Women were randomized to receive maintenance bevacizumab with either olaparib or placebo for a maximum of two years. The intention was to determine whether there would be synergistic benefit in combining bevacizumab with olaparib. In the intention-to-treat population there was a 6-month improvement in PFS with the combination (HR: 0.59). Impressively, in the HRD-positive patient population the HR was 0.33. However, the trial has been criticized for not including an olaparib monotherapy arm, meaning adding a third arm that would have been placebo plus olaparib, so that we can really define the relative contribution of the addition of bevacizumab to olaparib in the BRCA-mutated and HRD population.

Many ongoing trials investigate tolerability and efficacy of different niraparib combinations with other agents (Table 5). Thus far, niraparib has been examined as definitive treatment with bevacizumab, compared to niraparib alone, in women with platinum-sensitive, relapsed HGSOC in AVANOVA2 phase 2 study^[116]. Stratification was based on HRD status and chemotherapy-free interval following previous therapy. In general, PFS was significantly prolonged for the combination group, regardless of HRD status (median, 11.9 months vs 5.5; HR 0.35) and origin of BRCA mutation (germline or somatic).

Based on preclinical studies that suggested PARP inhibition may create a more immunogenic tumor milieu in OC, combination of PARPis with immune checkpoint inhibitors proceeded to early phase clinical trials. Olaparib and durvalumab, initially tested in germline-mutated (gBRCAm) platinum-sensitive OC, showed modest clinical activity in recurrent OC^{[117, 118]}. Combination of pembrolizumab with niraparib has been tested in women with relapsed, platinum-resistant, BRCA-mutant and HRD OC, in the phase 1/2 TOPACIO/KEYNOTE-162 trial^[119]. ORR was estimated at 18% for somatic BRCA-mutant and 14% for the HRD-positive women, respectively. Although antitumor effect was reported in all subgroups regardless of their biomarker status, intriguingly, long-term responses appeared to especially favor women with platinum-refractory or resistant disease and BRCA wild type or HRD-negative tumors. Another combination trial which did not meet its primary endpoint of improving PFS in patients with OC and has since been discontinued, is JAVELIN Ovarian PARP 100 trial, in which talazoparib was tested together with avelumab, an anti-PDL1 agent^[120]. The decision upon discontinuation, also made due to the modest benefit observed with avelumab in frontline OC in JAVELIN Ovarian 100. Translational research studies indicate that VEGF/VEGFR pathway blockade would be necessary for improved efficacy of this combination. As a consequence, triplet combinations are already being tested in early phase clinical trials.

Future is focused on combinations with the before mentioned effective therapies in phase II and III trials, trying to overcome resistance and improve patient outcomes. At the same time, into play come new target agents, such as the MEK and PI3K inhibitors, most of the times in the recurrence phase of ovarian cancer. Selumetinib, a novel MEKi, having shown antitumor activity in recurrent low grade serous ovarian cancer and ER positive HGSOC^{[121, 122]}, it is now tested in early phase clinical trials in combination with olaparib (NCT03162627). On the other hand, preliminary clinical evidence of synergism between olaparib and alpelisib has paved the way to further investigation^[123]. Furthermore, targeting more than one DNA repair pathway, including ATR, ATM, Chk1, Wee1 and DNA-PK, has the potential to enhance responses to PARP inhibitors. Other agents having shown activity in combination with PARPis, particularly in preclinical studies and early phase trials, are HDAC inhibitors and hypomethylating agents^{[124, 125]}. Table 6 shows olaparib's combination trials with immunotherapy, HDAC inhibitors and hypomethylating agents.

Analysis of BRCA1 missense mutations suggests that the conserved N- and C-terminal domains are most important for the response to HR-deficiency targeted therapies. Specifically, tumors carrying the BRCA1-C61G mutation which disrupts the N-terminal RING domain respond poorly to PARPis, and rapidly develop resistance^[127]. Interestingly, mutations in the BRCAC-terminal (BRCT) domain of BRCA1 commonly create protein products that are subject to protease-mediated the BRCT domain of these mutant BRCA1 proteins under PARP inhibitor selection pressure^[128]. The HSP90-stabilized mutant BRCA1 proteins can efficiently interact with PALB2-BRCA2-RAD51, form RAD51 foci, and confer PARP inhibitor and cisplatin resistance. Treatment of resistant cells with an HSP90 inhibitor may reduce the mutant BRCA1 protein level and restore their sensitivity to PARP inhibition^[128].

In BRCA1/2-mutated tumors, the most common acquired mechanism of resistance to PARPis is secondary intragenic mutations restoring the BRCA1 or BRCA2 protein functionality^{[129, 130]}. Restoration of BRCA1/2 function occurs either by genetic events that cancel the frameshift caused by the original mutation and restore ORF leading to expression of a functional nearly full length protein which can be similar to the full length wild-type protein. These genetic events were originally observed in BRCA1- and BRCA2-mutated cancer cells under in vitro selective pressure, due to exposure to cisplatin or PARPis and as a result would eventually cause platinum and PARPi resistance^[126]. This mechanism of resistance is highly clinically relevant for patients with BRCA-mutated cancers who are treated with platinum-based therapy; 46% of platinum resistant BRCA-mutated HGSOC exhibit tumor-specific secondary mutations that restore the ORF of either BRCA1 or BRCA2^[126]. Lastly, multiple somatically-reverted BRCA1/2 alleles have been detected from patients undergoing PARPi or platinum therapy through analysis of circulating free DNA (cfDNA)^[131].

HR repair can be restored by reversal of BRCA1 promoter methylation. The primary sensitive sample may show extensive promoter methylation and low BRCA1 expression, while the sample from the relapsed disease has lost BRCA1 methylation and the BRCA1 gene is expressed at comparable levels to homologous recombination proficient tumors^[126]. The loss of BRCA1 promoter methylation may result from an active demethylation event but more likely from a heterogeneous tumor from which the tumor cells with less promoter methylation undergo positive selection in PARPi^[126].

Since PARPis function by blocking the enzymatic action of PARP enzymes, another possible mechanism of PARPi resistance may be decreased expression of PARP enzymes. This mechanism of resistance maybe particularly relevant to the PARP-trapping mechanism of action of PARPis. Interestingly, a recent study identified a PARP1 mutation (1771C > T) in an ovarian cancer patient who demonstrated de novo resistance to PARP inhibitor^[132].

Several mechanisms of resistance involving reacquisition of DNA end resection capacities have also been described. There are several proteins which are known to working a pathway which suppresses end resection and thereby inhibits HR repair. These proteins include 53BP1, REV7, PTIP, and RIF1. Discovery of this mechanism came from the observation that the requirement of BRCA1 for HR can be alleviated by concomitant loss of 53BP1. 53BP1 blocks CtIP-mediated DNA end resection via downstream effectors like Rif1 and PTIP and thus commits DNA repair to CNHEJ^[134]. Loss of 53BP1 partially restores the HR defect and their hypersensitivity to DNA-damaging agents in knockout mice for BRCA-1^[134].

In BRCA1-mutant cells, loss of 53BP1 confers resistance to PARPi. More recently, a set of new proteins which bind to REV7, including RINN1, −2, −3 have been identified^[135]. Knockdown of these proteins also results in PARP inhibitor resistance.

Pharmacological effects that alter the cellular response to PARPis may also be relevant. Several studies have shown that PARPi responses maybe modified by ATP-binding cassette (ABC) transporters^[136]. Increased expression in tumor cells of ABC transporters, such as the P-glycoprotein (PgP) efflux pump (also known as multidrug resistance protein 1) have been implicated in reducing the efficacy of many compounds by enhancing their extracellular translocation.

Importantly, BRCA1/2 deficient tumor cells can acquire resistance to PARPi by finding independent mechanisms for protecting their replication forks. Research has demonstrated how the study of drug resistance can uncover novel principles of DNA repair and replication fork biochemistry. Multiple studies have previously demonstrated the role of BRCA1, BRCA2, and other FA proteins in the protection of RFs^{[137, 138]}. It has been demonstrated that different nucleases are necessary for fork protection. The first study showed that BRCA1 cells were shown to become resistant to PARPi by reducing the recruitment of the nuclease, MRE11 (required for the RF stability), to the stalled fork, thereby resulting in fork protection through reduced expression of the protein, PTIP^[139]. In a second study it was demonstrated that BRCA2 deficient tumors cells can become resistant to PARPi by reducing the recruitment of a different nuclease, MUS81, to the stalled fork^[140]. As a mechanism of PARPi resistance, these tumors downregulate EZH2 activity at the fork and downregulate MUS81 recruitment^[140]. Interestingly, through both of these mechanisms, the tumors achieve PARPi resistance without restoring HR repair.

Recently it has been suggested from in vitro trials that inhibition of nuclear protein PolyADP-ribosylation reaction catalyzed by PARP1 somehow triggers the activation of cytoplasmically localized Akt. PARP1 activated by DNA strand breaks PolyADP-ribosylation itself, thereby creating a scaffold that recruits a complex of proteins^[141]. If excess PARP activation is prevented by a pharmacological inhibitor or genetic manipulation, one of the complex proteins does not undergo ribosylation it translocates to the cytoplasm and in part associates with the mitochondrial outer membrane^[142]. There, the complex recruits Akt and mTOR, and together with them forms a signalosome that activates Akt, leading to activation of downstream pro-survival pathways^[143].

Fatigue is the most common AE, presented in 50% to70% of patients receiving PARPis. It is mainly grade 1–2 and most common with rucaparib, niraparib and olaparib^[145]. According to Coleman et al, the incidence of fatigue with veliparib is about 26%^[88]. When not related to hematological toxicity as anemia or thrombocytopenia, activity and physical exercise may be helpful to manage this adverse event. Another class related AE is anorexia and decrease in appetite. This appears to be a common side effect observed in about 20–25% of patients, to whom nutritional consultation may be necessary^[145].

A recent meta-analysis has shown that the risk of high grade gastrointestinal toxicities associated with PARPis is significantly increased in ovarian cancer patients, with the exception of constipation^[146]. Of these, nausea and vomiting seem to appear early on treatment with a tendency to decline later on. According to the SOLO-2 trial, the incidence of all – grade nausea with olaparib was 76% and of vomiting 37%. Lower incidence rates were reported with diarrhea and stomatitis, 33% and 20% respectively^[50]. Nausea and vomiting related to PARP inhibition usually subside with medication and treatment often involves drugs as metoclopramide, benzodiazepines or 5-HT3 receptor antagonists. Last but not least, dyspepsia (10–20%) and dysgeusia (10–40%) have frequently occurred in patients treated with PARP inhibitor maintenance, the later more commonly noted with rucaparib. Prescription of PPI's and certain dietary and oral hygiene measures may be helpful to control these AEs^[145].

Hematologic AEs are very common during treatment with PARPis and require complete blood counts regularly to exclude enhancing blood toxicity. According to a meta-analysis from Zhou et al including 2479 patients from 12 randomized controlled trials treated with PARPis, the incidence of severe neutropenia, thrombocytopenia and anemia were 32.9%, 15.9% and 9.1% respectively^[147]. Hematologic toxicity is more common with niraparib and talazoparib and less with rucaparib, olaparib and veliparib^[144]. According to one trial, the incidence of neutropenia of any grade in patients on treatment with PARPis is about 20%^[145]. On the other hand, febrile neutropenia is extremely rare when the PARP inhibitor is used a single agent^{[60, 73, 79]}. In general, the use of granulocyte growth factors during treatment with PARP inhibitors is not recommended. MDS/AML has been reported as an AE related to PARPi therapy although in a very small percent (approximately 1%). Exposure to DNA-damaging agents is demonstrated to be a significant risk factor for developing MDS/AML especially in the presence of BRCA mutations^[78].

Anemia with hemoglobin levels <8g/dL generally require interruption of treatment until hemoglobin recovers to 9 g/dL. Blood transfusions are often required and dose modifications with lower doses are necessary after persistent occurrence of anemia with treatment. PARPis should be withheld when platelet counts are <100×10⁹ cells/L. Treatment adjustment is required when platelet count reduces <75×10⁹ cells/L or at a second occurrence with the exception of talazoparib, where treatment may be continued if platelet counts are >50×10⁹ cells/L. Severe thrombocytopenia with platelet counts <10×109 cells/L require emergency platelet transfusions^[145].

The most common respiratory symptoms noted on treatment with PARP inhibitors are dyspnea, cough and nasopharyngitis. Usually mild, they affect 10–20% of patients and may relate to infectious conditions. Clinical and radiographic assessment may be necessary to exclude olaparib related pneumonitis which is reported in <1%^[145].

Insomnia and headache are frequently related with PARP inhibition. Headache is common with both olaparib and niraparib at a rate of 26% while insomnia is observed in 27% of patients who are treated with niraparib. Dizziness is related to niraparib too with an incidence rate of 18%. Pharmacological measures may appear necessary to control refractory cases of drug related neurological adverse events^[145].

Transient transaminase elevations have been described mainly with rucaparib. All grade incidences have been reported to be approximately 34% and usually transaminase normalization occurs after a few weeks from initiation of treatment. Olaparib and niraparib can cause transaminase elevations too, but to a lesser extent than rucaparib^[145]. Elevations in creatinine have been described in approximately 10–12% of patients across different PARPis^[145]. A retrospective trial by Zibetti et al. showed that elevation of blood creatinine levels in patients receiving PARPis was not related to a decrease in the GFR measured by renal scan^[148]. In the event of persistent blood creatinine rise, other causes should be ruled out before withholding treatment.

Among cardiac adverse events, hypertension and palpitations occurrence have been related mainly to niraparib. In the NOVA trial, all grade hypertension was noted in 20% of patients, while palpitations in 10% respectively^[77]. This potential may be due to the drug's off-target pharmacologic activity on the dopamine, the norepinephrine and the serotonin transporters^[149]. Accordingly, thorough investigation is needed to exclude other causes of tachycardia and palpitations, and medication to be used when necessary.

Skin disorders related to treatment with PARP inhibitors include rash, pruritus and photosensitivity. In the ARIEL3 clinical trial, 17% of the patients receiving rucaparib reported photosensitivity, 13% pruritus and 12% rash^[73]. Sun protection and Dermatologic consultation may be indicated.