Homologous recombination proficient subtypes of high-grade serous ovarian cancer: treatment options for a poor prognosis group

Approximately 50% of tubo-ovarian high-grade serous carcinomas (HGSCs) have functional homologous recombination-mediated (HR) DNA repair, so-called HR-proficient tumors, which are often associated with primary platinum resistance (relapse within six months after completion of first-line therapy), minimal benefit from poly(ADP‐ribose) polymerase (PARP) inhibitors, and shorter survival. HR-proficient tumors comprise multiple molecular subtypes including cases with CCNE1 amplification, AKT2 amplification or CDK12 alteration, and are often characterized as “cold” tumors with fewer infiltrating lymphocytes and decreased expression of PD-1/PD-L1. Several new treatment approaches aim to manipulate these negative prognostic features and render HR-proficient tumors more susceptible to treatment. Alterations in multiple different molecules and pathways in the DNA damage response are driving new drug development to target HR-proficient cancer cells, such as inhibitors of the CDK or P13K/AKT pathways, as well as ATR inhibitors. Treatment combinations with chemotherapy or PARP inhibitors and agents targeting DNA replication stress have shown promising preclinical and clinical results. New approaches in immunotherapy are also being explored, including vaccines or antibody drug conjugates. Many approaches are still in the early stages of development and further clinical trials will determine their clinical relevance. There is a need to include HR-proficient tumors in ovarian cancer trials and to analyze them in a more targeted manner to provide further evidence for their specific therapy, as this will be crucial in improving the overall prognosis of HGSC and ovarian cancer in general.

HRD refers to a loss of homologous recombination-mediated DNA repair (HRR), which is a pathway responsible for the highfidelity repair of double-stranded DNA breaks that restores the original DNA sequence at the site of damage.HRD contributes to genomic instability and consequently intact HRR plays a role in preventing malignant transformation (30,31).HRD is caused by inherited or somatic loss of function genetic alterations in wellknown driver genes such as BRCA1 and BRCA2, but also by mutations or methylation of other HRR related genes and potentially other currently undefined mechanisms (32).Patients with HRD HGSC are more likely to benefit from a favorable chemotherapy response, maintenance treatment with PARPis and consequently a longer OS (1,(33)(34)(35)(36)(37)(38).However, ~50% of HGSC are HR-proficient (HRP), an established poor prognostic marker associated with primary platinum and PARPi resistance and shorter survival times (36,39,40).Platinum-resistant ovarian cancer is defined as disease that relapses within six months of completing first-line treatment, and the probability of a response to platinum re-treatment is less than 10% (33,41,42).In fact, HGSC can also progress from HRD to become at least partially HRP by reversion of HR gene alterations through secondary genetic or epigenetic events (43)(44)(45).This acquired HR-proficiency is one of the most well described mechanisms of acquired treatment resistance and consequently a major clinical challenge.
HRD status in ovarian cancer is usually inferred by measurement of BRCA mutation status and/or the extent of cancer genome scarring associated with loss of HRR genes.Methodologies that assess HRD typically measure the extent of telomeric allelic imbalance, loss of heterozygosity, and large-scale transitions (31,46).However, these scores are based on permanent genomic scars, thus failing to reflect the current HRD status in the case of HRR restoration (47).An alternative is a dynamic assessment of HR status using functional assays in ex-vivo cultures (46).Immunofluorescence microscopy can be used to measure the presence of RAD51 formed molecular complexes which accumulate at sites of double-stranded DNA breaks in HRP cells.By contrast, HRD cells are unable to form RAD51 formed molecular complexes and their absence thus provides a functional indication of a defect in the HR pathway (48,49).However, such RAD51 assays are yet to be clinically validated.
Additionally, resistance to PARPis may be driven by RAD51independent mechanisms and consequently cannot be detected by RAD51 assays (50,51).Current HRD tests vary in the number and type of mutational features assessed, and the optimal thresholds to classify samples as HRD or HRP are not yet well defined.Variation in assays should be considered when evaluating the overall value of such assays in providing prognostic and predictive information.

Characteristics of patients with HR-proficient HGSC Clinicopathological
Variation in outcomes between patients with HGSC is in part determined by the molecular characteristics of the tumor, with HRstatus as one of the important determinants (Table 1).Patients with HRP tumors have an older median age at diagnosis compared to patients with HRD tumors (10,11,20,36).A retrospective analysis of 352 patients showed that HRP tumors required a higher number of cycles of NACT to be considered for interval cytoreductive surgery compared to those with germline BRCA mutations and other defects conferring HRD, and less complete gross resection (R0) could be achieved (11).While complete resection in primary and interval cytoreductive surgery remains one of the strongest prognostic features in ovarian cancer (2,3,62,63), the higher number of chemotherapy cycles and lower R0 rate also reflect an inherently resistant tumor (18, 22-24, 26, 33).

Genomic characteristics
Extensive genomic and transcriptomic characterization has provided insight into HGSC with HRR pathway inactivation, most commonly caused by genetic or epigenetic alterations in the BRCA genes and alterations in other genes, including BRIP1, PALB2, RAD51C, or RAD51D, which encode proteins that are also involved in HR DNA repair (59).By contrast, the molecular drivers of HGSC that have no apparent defects in HR are less well defined (2).
HRP ovarian cancer cells are often characterized by genetic alterations in signaling pathways that contribute to cell cycle dysregulation, such as cyclin E1 (encoded by CCNE1) and cyclin dependent kinase (CDK) genes (44).Cyclin E1 is an important factor in the G1/S cell cycle transition through its activation of cyclin-dependent kinase 2 (CDK2), allowing the cell to enter the Sphase (64).Besides other cellular mechanisms, limiting the supply of cyclin E1 ensures that the cell remains in the G1 phase by keeping CDK2 inactive until mitogenic signals intervene (65).CCNE1 expression is dependent on E2F transcription factors that are bound to the retinoblastoma protein (Rb) in an inactivated state when cells are at rest.E2F is released through mitogenic stimuli such as c-MYC which increases the expression of D-type cyclins that in turn combine with CDK4 and CDK6 to phosphorylate and inactivate Rb (65).Furthermore, once activated, the cyclin E1/ CDK2 complex is able to phosphorylate Rb and thus upregulate its own expression in the form of a positive feedback loop through the continued release of E2F, independent of mitogenic stimuli (65).Additionally, the cyclin E1/CDK2 complex is an essential component of the chromatin remodeling process required for Adapted from (10-13, 17, 20, 36, 59-61).NACT, Neoadjuvant Chemotherapy; HRP, Homologous recombination proficient; HRD, Homologous recombination deficien; HR, Homologous recombination.
DNA replication.Overexpression of cyclin E1 increases the speed at which cancer cells transition from G1 to the S phase (66).This can lead to replicative stress, whole genome duplication, and further promote the dysregulation of genes responsible for proliferation and cell survival, which are also associated with resistance to cytotoxic and targeted therapies (67, 68).
CCNE1 amplification is currently the best characterized driver of HGSC with HR-proficiency.It is important to note, however, that cyclin E1 protein overexpression itself has not been shown to be a predictive biomarker for chemotherapy resistance in epithelial ovarian cancer (EOC), so methods to detect amplification of a gene (e.g.whole-genome sequencing, fluorescence in situ hybridization, polymerase chain reaction, single nucleotide polymorphism arrays) are required to identify the CCNE1 amplified subgroup (69).Approximately 40% of HRP HGSC show an CCNE1 amplification, which has been shown to be an early event in their development (43,64).HR pathway gene mutations and CCNE1 amplification have been shown to be mutually exclusive (44,60,65).This suggests that the pathogenesis of HGSC follows at least two distinct pathways, and that CCNE1-amplified tumors with cyclin E1 protein overexpression are more likely to be resistant to platinum-based chemotherapy and PARPi due to HR-proficiency (65).
AKT2 amplification is also a poor prognostic marker in EOC (34,70,71) and is associated with CCNE1 amplification (70).The co-amplification of the serine/threonine-protein kinase AKT2 and CCNE1 appears to be explained in part by their proximity on chromosome 19q.Pathway analysis indicates that CCNE1amplified cell lines are dependent on multiple genes within the CDK and AKT pathways, suggesting a specific dependence of CCNE1-amplified tumors on AKT activity (70).Consequently, combined CDK2 and AKT inhibition may have synergistic antitumor activity against CCNE1-amplified tumors and hold promise for clinical development (70).It should be noted that although CDK4/6 inhibitors have been investigated in ovarian cancer (72), it is the CDK2 inhibitor which is likely to be effective (73)(74)(75)(76).
CDK12-altered HGSC represent a unique subgroup that appear to be HR competent (36).Despite lacking the typical HRD genomic scarring, CDK12-altered tumors have a distinct tandem duplication signature and may be more susceptible to chemotherapy and PARPis than other HRP tumors (77).Aside from alterations in CCNE1, AKT2 and CDK12, the majority of HRP HGSC remain poorly defined, and integration of genomic, immune, proteomic and functional data is needed for their complete characterization (78)(79)(80)(81).

Immune profile
Tumor-infiltrating lymphocytes (TILs) are an established prognostic factor in ovarian cancer, regardless of the extent of surgical cytoreduction and chemotherapy (82-84).The presence of CD8+ TILs in the tumor microenvironment is associated with slower tumor progression, prolonged survival and may be essential for immunotherapy response (84)(85)(86).HRD tumors have a significantly increased CD8+/CD4+ ratio of TILs and a higher number of peritumoral T cells (44).This is likely due to HRD cells accumulating a high number of somatic mutations, which is predicted to result in the expression of more tumor neoantigens that elicit an adaptive immune response and cytotoxic T cell infiltration.These cells are capable of killing cancer cells (84), and in addition to a more favorable response to chemotherapy, explains the improved survival of patients with BRCA-mutated ovarian cancer.
By contrast, HRP tumors are characterized by a non-inflamed or "cold" immune phenotype, with fewer CD3+ and CD8+ TILs as well as decreased expression of PD-1 and PD-L1 (87)(88)(89).HRP tumors generally have a lower tumor mutational burden due to having intact DNA repair, which, together with a low TIL density, would predict a poor response to immune checkpoint blockade (84).Therefore, HRP tumors may be poor candidates for targeted immunotherapy with PD-1 and PD-L1 inhibitors as recently shown (90)(91)(92)(93)(94)(95)(96).Recent approaches to immunotherapy for cold tumors have focused on restoring inflammation by reprogramming myeloid cells, stromal cells, and vascular epithelial cells (97).Additionally, PARPi, low-dose radiotherapy, epigenetic drugs and anti-angiogenesis therapy may enhance T cell infiltration, suggesting their use in combination with vaccines and redirected T-cells using chimeric antigen receptors or bispecific antibodies (84,98).However, it should be noted that while T cell infiltration and the expression of PD-L1 and other immune checkpoint markers increases following chemotherapy, unlike primary disease, the extent of infiltration does not correlate with patient survival (99, 100).

Treatment options for patients with HRP HGSC Chemotherapy
Neoadjuvant chemotherapy with interval cytoreductive surgery is currently an alternative for patients with ovarian cancer who have a low chance of initial complete resection and chemosensitive histologic subtypes, or poor health status (1).However, there is a strong correlation between HR-status and response to platinumbased chemotherapy in HGSC; patients with HRP tumors have severely limited responses to chemotherapy, with reported median PFS ranging from 5.4 to 16.9 months (Table 1) (10-13).The chemoresistant nature of HRP tumors highlights the potential benefit of favoring the currently recommended option in HGSC (1) of primary debulking surgery followed by adjuvant platinum and taxane-based chemotherapy in these patients.
An ancillary data analysis of the VELIA/GOG-3005 trial focused on paclitaxel dosing schedule and BRCA mutation and HR-status (101).Dose-dense (weekly) paclitaxel was compared to a schedule of every three weeks showing an improved PFS with dosedense paclitaxel in HRP but not in BRCA-mutation or HRD tumors.Previous clinical trials of shorter versus longer paclitaxel intervals in ovarian cancer did not evaluate HR status and therefore further studies are needed to confirm this finding (102, 103).Interestingly, it has been shown that paclitaxel suppresses CDK1 expression via decreased BRCA1 phosphorylation, thereby reducing HR activity in response to DNA damage and increasing sensitivity to PARPis (104), so this combination represents a potential new treatment strategy that needs to be further investigated in HRP HGSC.
HGSC typically involves extensive peritoneal spread and therefore intraperitoneal chemotherapy and hyperthermic intraperitoneal chemotherapy (HIPEC) have been evaluated in multiple clinical trials.The goal of intraperitoneal chemotherapy is to increase local exposure to the chemotherapeutic agent, and in the case of HIPEC, heated chemotherapy has an additional cytotoxic effect and increases sensitivity to platinum compounds by inducing a transient state of HRD (105).Koole et al. analyzed the effect of HIPEC among patients with ovarian cancer previously enrolled in the phase III OVHIPEC1 trial (105) stratified by BRCAlike (HRD) versus non BRCA-like (HRP) (106) or BRCA mutation status.Although patients with HRD/BRCA-wildtype showed a strong benefit in terms of recurrence-free survival (RFS) and a promising trend in OS from HIPEC, this was non-significant in HRP/BRCA-wildtype patients and absent in patients with pathogenic BRCA mutations, both in terms of RFS and OS (58).It appears that HRP tumors remain resistant to chemotherapy despite hyperthermia.However, there is a lack of long-term survival data for HIPEC, and thus the benefit of this treatment modality remains unclear.The importance of tumor HR status in predicting response and survival following HIPEC may be addressed in ongoing studies (107).

Poly (ADP-ribose) polymerase inhibitors
Maintenance PARPi therapy after first-line treatment and in the platinum sensitive recurrent setting have become standard treatment options in patients with BRCA-mutated and HRD EOC (1,3).PARP is an enzyme that helps repair DNA damage and PARP inhibition causes an accumulation of single-and double-stranded DNA breaks (108).HRD cells are unable to effectively repair the DNA damage, resulting in an accumulation of chromosomal aberrations and cell death (109).As a maintenance therapy PARPi have led to improved PFS and shown a promising trend towards improved OS in EOC, particularly in patients with BRCA mutant and/or HRD tumors (17,18,20,21,23,24).While the greatest benefit is seen in HRD cancers, an exploratory analysis of the Phase III PRIMA trial showed improvements in PFS with niraparib versus placebo as first-line maintenance monotherapy, regardless of BRCA and HR-status (20).Patients with BRCAwildtype/HRP tumors treated with niraparib who responded to first-line chemotherapy had a median PFS of 8.1 months versus 5.4 months for placebo, with an estimated probability of survival at 24 months of 81% in the niraparib group versus 59% in the placebo group.Therefore, niraparib is clinically approved for use in patients with HRP HGSC, with beneficial effects and a manageable tolerability profile (110,111).
An exploratory analysis of the VELIA/GOG-3005 trial (27) showed that some patients with HRP ovarian cancer and also poor chemosensitivity may have gained a transient, but non-significant benefit from the addition of the PARPi veliparib to carboplatin-paclitaxel (median PFS 14.7 vs median 6.7 months, HR 0.62, 95% CI 0.37-1.05)(112).The authors of the study hypothesized that veliparib may have induced a chemosensitizing effect on HRP tumors (112,113).In addition, the Phase III ATHENA-MONO trial demonstrated improved PFS with rucaparib monotherapy compared to placebo in first-line maintenance in patients with newly diagnosed EOC without evidence of HRD (12.1 vs 9.1 months, HR 0.65, 95% CI 0.45-0.95)(13).As a result of such findings, the ESGO-ESMO-ESP consensus guidelines state that niraparib or rucaparib maintenance therapy may be used for patients with HRP HGSC if they have had a complete or partial response to first line chemotherapy or no evidence of disease (1).

Antiangiogenic treatment
Vascular endothelial growth factor (VEGF) promotes increased vascularity and angiogenesis in response to hypoxic conditions and is a key promoter of tumor growth (114).The anti-angiogenic VEGF monoclonal antibody bevacizumab was the first targeted agent to be approved for use in stage III and IV EOC, showing an improved PFS when used in combination with chemotherapy and as maintenance therapy in the first-line setting, however without OS benefit (115,116).According to the ESGO-ESMO-ESP consensus guidelines, patients with HRP HGSC may receive platinum-based chemotherapy with bevacizumab followed by bevacizumab maintenance as an alternative to the option of maintenance with rucaparib or niraparib (1).Among other mechanisms of action, bevacizumab exposure may trigger HRD by inducing a hypoxic cellular state that can downregulate HR-related genes such as BRCA1/2 and RAD51 (117).In addition, the relative benefit of bevacizumab in EOC has been shown to increase as the disease becomes more platinum resistant (118).A retrospective analysis of 124 patients with platinum-sensitive recurrent ovarian cancer showed extended PFS with bevacizumab in patients with cyclin E1 overexpression (median 16.3 vs 7.1 months, P=0.010) (118).
Tumor VEGF secretion has been shown to be at least partially responsible for the development and maintenance of ascites, and the AURELIA trial demonstrated that the addition of bevacizumab to chemotherapy improved ascites control.This beneficial effect is certainly relevant for the HRP group as they are more frequently associated with suboptimal debulking, earlier recurrence and ascites (54).Furthermore, the combination of niraparib and bevacizumab evaluated in the pre-specified subgroup analysis of the AVANOVA trial showed a significant improvement in PFS compared to niraparib alone in the HRP population (HR 0.40, 95% CI 0.19-0.85)(119).The Phase III GOG-218 trial also showed prolonged PFS in patients with no HRR gene mutations who received bevacizumab in addition to standard chemotherapy with carboplatin and paclitaxel (HR 0.71, 95% CI 0.60-0.85,P = 0.0001).This benefit was not observed in patients with HRR gene mutations (HR 0.95; 95% CI 0.71-1.26)(120).Therefore, the ESMO guidelines recommend that the decision on bevacizumab versus niraparib maintenance in the HRP population should be based on the patient's disease and clinical characteristics, the toxicity profile of the two drug classes, the availability of each drug, and national guidelines (3,121).The ongoing Phase I/II MITO 25 trial (NCT03462212) may provide clearer evidence about potential therapy options by comparing whether the carboplatin-paclitaxelbevacizumab-rucaparib or carboplatin-paclitaxel-rucaparib arms improve PFS compared to standard carboplatin-paclitaxelbevacizumab in patients with HRP HGSC.
The inhibition of VEGF receptor-3 (VEGFR3) has been shown to decrease BRCA1 and BRCA2 expression in ovarian cancer cells and resulted in increased chemosensitivity (122).The randomized Phase II trial (NCT01116648) showed that the combination of olaparib plus cediranib, a VEGF receptor 1/2/3 inhibitor, significantly improved PFS in relapsed platinum-sensitive EOC compared to olaparib alone (median 17.7 months vs 9 months, P=0.005), with the greatest benefit in BRCA-wildtype patients (HR 0.32, P=0.008) (123).These results suggest that there may be greater synergism between the two agents in HRP tumors, with the response to olaparib in HRP tumors being enhanced by diminished HRR due to VEGFR3 inhibition.However, experimental in vivo efficacy data showed that the combination exhibited broad anti-tumor activity independent of HRR and that the combination effect was largely driven by influencing independent mechanisms affecting tumor cells and the tumor microenvironment (124).Clinically, the combination of cediranib and olaparib also showed some activity in the CONVERTO trial, a single-arm Phase IIb study of the two compounds in heavily pretreated, platinum-resistant, nongermline BRCA-mutated patients.However, the target objective response rate (ORR) of 20% was not reached (15.6%) and the overall benefit was unclear (OS 13.2 months, 95% CI 9.4-16.4;PFS 5.1 months, 95% CI 3.5-5.5)given it was a single arm study in a disease setting where most patients are expected to progress or die within 12 months (125).A Phase III trial [NCT02446600] in patients with relapsed platinum-sensitive ovarian cancer found that neither the combination of olaparib and cediranib nor olaparib monotherapy improved PFS compared to standard chemotherapy (126).An ongoing Phase II/III trial (NCT02502266) is evaluating cediranib plus olaparib compared to their monotherapies and standard chemotherapy.It remains to be determined if there is a clinical benefit of VEGF receptor inhibitors in treating EOC, particularly in chemoresistant HRP tumors.Future research efforts must focus on identifying other predictive biomarkers for antiangiogenic therapy, as not all observed responses can be explained by BRCA mutation or HR-status.

Secondary cytoreductive surgery
There have been significant advances in the surgical management of HGSC with improved PFS and OS due to intensification of surgical efforts (62,127,128).A multicenter, open-label, randomized, controlled Phase III trial SOC-1 (NCT01611766) demonstrated in 357 patients with platinumsensitive relapsed ovarian cancer that secondary cytoreductive surgery (SCS) followed by chemotherapy was associated with significantly longer PFS than with chemotherapy alone (median 17.4 vs 11.9 months, HR 0.58, 95% CI 0.45-0.74,P<0.0001) (129).Furthermore, the DESKTOP III trial (NCT01166737) analyzed 407 patients with platinum-sensitive recurrent ovarian cancer and showed that SCS followed by chemotherapy leads to a longer OS than chemotherapy alone (median 53.7 vs 46.0 months, HR for death 0.75, 95% CI 0.59-0.96,P=0.02).Patients with a complete resection had the most favorable outcome (130).In addition to these two positive studies, in the GOG-0213 trial, which also included patients with platinum-sensitive, recurrent ovarian cancer, SCS followed by chemotherapy did not result in a longer OS than chemotherapy alone (131).There are some differences between the trials that may explain the inconsistent results, such as the additional use of bevacizumab in the DESKTOP III trial (NCT01166737) or the process of selecting patients and centers (130).Therefore, it is important that patients are appropriately counseled about the option of SCS.
The role of surgery in patients with platinum-resistant disease has received increasing attention (132).In fact, patients with HRP tumors may benefit from SCS, similar to patients with low-grade serous ovarian cancer (133).To our knowledge, only three retrospective studies have been published analyzing the role of SCS in patients with platinum-resistant recurrent ovarian cancer.Both Petrillo et al. and Musella et al. showed a prolonged OS after recurrence when SCS was combined with chemotherapy instead of chemotherapy alone (median 32 months vs 8 months, P=0.002 and 67 months vs 24 months, P=0.035) (134,135).However, when evaluating these two studies, it is important to consider that they were carried out before the PARPi era and therefore their conclusions must be put into perspective with current treatment options.A recent multicenter retrospective series by Tuninetti et al. in 50 heavily pretreated platinum-resistant ovarian cancer patients showed a statistically significant longer OS in the group of patients who received complete cytoreduction after SCS compared to the very low survival of patients with residual disease (meidan 33 months vs 5 months, HR 4.21, 95% CI 2.07-8.60,P=0.001) (136).These retrospective studies did not include stratification by BRCA mutation or HR-status, and any discussion of the extent of surgical clearance should also consider how residual disease may be a marker of biology that drives outcome.However, in a recent multicenter retrospective study investigating platinum sensitive recurrent ovarian cancer, SCS was shown to be effective in BRCAwildtype patients, with an improvement in post-recurrence survival (PRS) when complete resection was performed (5-year PRS of 54% vs 42%, P=0.048), whereas in BRCA-mutated patients, prognosis appears to be related to molecular tumor characteristics rather than tumor resectability (137).A current prospective randomized controlled trial (NCT05633199) is now comparing SCS in platinum-resistant recurrent ovarian cancer and is expected to provide further information on whether and to what extent SCS can be used in the "platinum-resistant" HRP HGSC subgroup.Another advantage of SCS is to opportunistically obtain more comprehensive information on the pathological and molecular characteristics of HRP HGSC and how this may affect tumor evolution and clinical outcome (127).SCS in HGSC warrants further investigation in prospective trials, with particular attention paid to patient BRCA and HR-status.

Immunotherapy and antibodydrug conjugates
Immunotherapy for HGSC has fallen short of expectations, with immune checkpoint inhibitors so far showing limited benefit in ovarian cancer (138)(139)(140)(141)(142).However, there are new, potentially promising approaches, including ADCs that deliver a toxic 'payload' of chemotherapy directly to cancer cells via a linker attached to an antibody that binds to a specific surface antigen expressed on cancer cells (143).Mirvetuximab is a first-in-class ADC targeting folate receptor a (FRa), a cell surface protein that is commonly overexpressed on ovarian cancer (80-100%) and minimally expressed on normal tissue (144)(145)(146).This ADC incorporates the maytansinoid DM4 payload, a potent tubulintargeting antimitotic agent, and is the first novel agent to demonstrate an OS benefit when used as a single agent compared to chemotherapy alone in platinum-resistant ovarian cancer, as shown in the MIRASOL Phase III clinical trial (NCT04209855) (144).Patients with platinum-resistant, FRa-positive ovarian cancer treated with mirvetuximab (n=227) experienced an OS of 16.46 months (95% CI, 4.46-24.57)vs 12.75 months (95% CI, 10.91-14.36)for the chemotherapy arm (HR 0.67, 95% CI 0.50-0.89,P=0.005) and showed fewer Grade 3 or higher adverse events with mirvetuximab than with chemotherapy (41.7% vs 54.1%).
A n o t h e r p r o m i s i n g i m m u n o t h e r a p y a p p r o a c h i s Gemogenovatucel-T (Vigil, formerly known as FANG ® ), the first immunotherapy to demonstrate specific efficacy in the frontline maintenance setting for the HRP population.Vigil is a vaccine composed of autologous tumor cells derived from malignant tissue removed during cytoreductive surgery (147) (Figure 1).Tumor cells are transfected with a plasmid containing GM-CSF and bi-shRNA to reduce furin activity, which subsequently downregulates the expression of the immunosuppressive proteins TGF-b1 and TGF-b2 (transforming growth factor b).This is important because the expression of furin and the resulting immunosuppressive TGF-b isoforms are increased in ovarian tumors compared to normal ovarian tissue (148).Long-term safety of Vigil and evidence of patient benefit have been demonstrated in multiple solid tumors, including advanced ovarian cancer (149, 150).The ongoing Phase IIb VITAL trial (NCT02346747) evaluated the efficacy of Vigil in patients with stage III/IV ovarian cancer.RFS was 11.5 months for patients treated with Vigil versus 8.4 months for patients treated with placebo (HR 0.69, 90% CI 0.44-1.07,P=0.078) with an acceptable toxicity profile (151).Although the primary endpoint of RFS was not met, a small subgroup analysis (n=45) showed that RFS and OS was significantly improved with Vigil compared to placebo in HRP patients (HR 0.38 and 0.34, 90% CI 0.2-0.75 and 0.14-0.83,P=0.007 and P=0.019), while no difference was seen in patients with BRCA-mutated disease (151, 152).Vigil increases the expression of cancer-associated neoantigens by upregulating MHC-II and processing by dendritic cells, which enhances the afferent immune response, the initial phase of immune activation characterized by antigen presentation and recognition, resulting in a systemic anti-tumor immune response including CD3+/CD8+ T cell circulation (152).T cells showed to preferentially recognize clonal neoantigens over subclonal neoantigens to target the tumor in lung adenocarcinoma and melanoma (153).HRP tumors are associated with higher clonal neoantigen expression compared to HRD tumors, which therefore contain higher proportions of subclonal neoantigen subpopulations, which may explain why Vigil is more effective on HRP tumors (152).A Phase III trial is planned to validate the efficacy of Vigil compared to bevacizumab and niraparib in the HRP ovarian cancer population (152).It has been suggested that the increased expression of clonal tumor neoantigens and reduced tumor suppressive effect of TGF-b may synergistically enhance the activity of checkpoint inhibitor treatment (84,154,155).A prospective, randomized Phase I trial of Vigil plus the immune checkpoint inhibitor atezolizumab in patients with recurrent ovarian cancer explored this approach and showed that the combination was safe, supporting further investigation of this combination, particularly in BRCA-wildtype patients (155).
Adoptive cell therapy is another emerging personalized form of immunotherapy in which patients are treated with their ex vivo expanded natural TILs, genetically engineered T lymphocytes (CAR T cells) or T-cell receptor (TCR)-engineered T cells, which could offer a potential therapeutic option for patients with cold tumors.To date, CAR T cells that have been tested in clinical trials for HGSC have not yet demonstrated clear benefit (84,156).While this technology is promising, further development is required to investigate the full potential of T cell engineering and other novel immunotherapy approaches to address the problem of immunologically cold tumors (84).

Combined targeted therapies
Rational drug combinations are a potential strategy to prevent or delay the development of resistance and offer the opportunity to improve the therapeutic window by potentially reducing the required drug doses, resulting in fewer side effects (70).Several strategies to selectively disrupt HRR in cancer cells with drugs have been investigated both preclinically and in clinical trials in HGSC or EOC in general, including HRP tumors, and have provided the rationale for new potential therapeutic approaches (Figure 2, Table 2).Here we review the most promising approaches for HRP tumors that have been or are being investigated in ovarian cancer, including targeting the CDK, P13K/AKT or CHK pathways.

CDK pathway
Approximately 40% of HGSC with HR-proficiency have an amplification of CCNE1 (64).Cyclins are typically regulatory proteins that modulate the activity of CDKs (65).The CDK pathway offers attractive targets for the treatment of CCNE1amplified tumors due to its role as the kinase partner of cyclin E1 in the activated cyclin E1/CDK complex (65, 163) (Figure 3).Cyclin E1 is primarily regulated by CDK2 in CCNE1-amplified tumors, which are selectively dependent on CDK2 activity (73).Combination therapy with the multi-CDK inhibitor dinaciclib (targets CDK1/2/5/9) has shown positive preclinical responses in CCNE1-amplified HGSC (164-166), and there is currently an active but not recruiting Phase I trial [NCT01434316] evaluating dinaciclib in combination with the PARPi veliparib in advanced solid tumors.However, a disadvantage of broad-spectrum CDK inhibitors is their high toxicity (167).Recently, more selective CDK2 inhibitors have been investigated (74)(75)(76), including promising preclinical results using INX-315, a novel, potent and highly selective CDK2 inhibitor.INX-315 treatment resulted in tumor growth inhibition of CCNE1-amplified tumors by promoting retinoblastoma protein hypophosphorylation, inducing cell cycle arrest and delaying the onset of CDK4/6 inhibitor resistance in breast cancer (74).In addition, a recent first-in-human Phase I/IIa study (NCT04553133) of a novel and potent selective CDK2i (PF-07104091) found that it was well tolerated and showed antitumor activity in heavily pretreated metastatic breast cancer patients who had progressed on prior CDK4/6 inhibitors (75).Further development of selective CDK2 inhibitors in Phase I/II clinical trials are ongoing and may be of major importance for HRP HGSC.
Another strategy is to target Weel-like kinase (WEE1), which is highly upregulated in HGSC (108).Its inhibition causes activation of CDK1 and CDK2, resulting in cell cycle acceleration with an early mitotic entry and mitotic catastrophe leading to irreparable DNA damage (121).The multicenter Phase II IGNITE trial [ACTRN12619001185156P] is a non-comparative trial evaluating the WEE1-inhibitor adavosertib in two cohorts of platinum resistant recurrent HGSC (cyclin E1 overexpressed/CCNE1 Potential combination therapies to induce homologous recombination deficiency (HRD).VEGFR3i, Vascular endothelial growth factor receptor 3 inhibitor; EGFRi, Epidermal growth factor receptor inhibitor; CDKi, Cyclin-Dependent Kinases inhibitor; WEE1i, Weel-like kinase 1 inhibitor; PI3K, Phosphatidylinositol-3-Kinase and AKT, Serine/threonine protein kinases; ATRi, Ataxia teleangiectasia Rad3-related inhibitor; HDACi, Histone deacetylase inhibitor; HSP90i, Heat shock protein 90 inhibitor; BETi, Brodomain and extraterminal protein family inhibitor.amplified and cyclin E1 overexpressed/CCNE1 non-amplified) and demonstrated an ORR of 53% and a clinical benefit of 61% in an interim analysis of 32 patients in the cyclin E1 overexpressed/ CCNE1 non-amplified cohort (168).CDK1 is a key cell-cycle regulator and phosphorylates BRCA1, which is required for DNA damage-induced checkpoint control through the formation of BRCA1-containing foci (169); consequently, inhibition of CDK1 impairs the ability of cells to functionally repair DNA by HRR (165).Therefore, depletion or inhibition of CDK1 creates a state of "BRCAness" in transformed cells (170).Results from preclinical studies in other cancer modalities support the effect of WEE1 inhibition on HR, and thus the assumption that WEE1 inhibitors, in combination with a DNA damaging agent, specifically render HRP cell lines more susceptible to treatment (171,172).An ongoing Phase II trial (NCT03579316) in recurrent PARPiresistant EOC (including 98% HGSC) is evaluating the efficacy of the WEE1 inhibitor adavosertib with or without olaparib.The combination showed to have a greater clinical benefit rate than adavosertib alone (89% vs 63%), but the ORR was similar between the two arms (160).Interestingly, exploratory analyses showed a larger benefit of the combination in the BRCA-wildtype subgroup compared to the BRCA-mutated subgroup (39% vs 19% ORR).Translational analyses are underway to further explore potential predictive biomarkers (160).However, adavosertib use requires consideration of single agent toxicity as well as interactions when used as a drug combination.For example, the use of adavosertib in combination with carboplatin showed an increased incidence of bone marrow suppression, diarrhea, vomiting and fatigue (168,173).Additionally, adavosertib is metabolized via the enzyme cytochrome P450 3A4 (CYP3A4), which means that patients receiving any co-medications that are strong CYP3A4 inhibitors (for example, antibacterials such as clarithromycin and erythromycin, anticancer agents such as tamoxifen and irinotecan, anti-HIV agents such as ritonavir and delavirdine, or antihypertensives such as dihydro-dralazine and verapamil) (174) would be excluded from clinical trials.
Another promising therapeutic target of the CDK pathway specifically for CCNE1-amplified HGSC is PKMYT1 (68).PKMYT1 is a kinase encoding the pro-protein kinase Myt1, a negative regulator of CDK1, and was identified in a genetic screen of cellular dependencies in CCNE1 amplified HGSC (68).Inhibition of PKMYT1 results in activation of CDK1, causing unscheduled mitotic entry and genome instability.In contrast, the WEE1inhibitor showed no selectivity towards CCNE1-amplified cell lines (175,176).Ongoing first-in-human clinical trials are evaluating the PKMYT1 inhibitor lunresertib (RP-6306) as monotherapy or in combination with the ataxia-telangiectasia Rad3-related (ATR) inhibitor RP-3500 (NCT04855656) and in combination with gemcitabine (NCT05147272) in advanced solid tumors.

PI3K/AKT pathway
Phosphatidylinositol 3-kinase (PI3K) activity is stimulated by a wide range of oncogenes and growth factor receptors (177) and the activity of the PI3K pathway is important to the development of drug resistance in a variety of cancer types and treatment settings (178).Inhibition of the PI3K pathway also results in PI3K-mediated downregulation of BRCA, accompanied by extracellular signalregulated kinase (ERK) phosphorylation and subsequent abrogation of HRR (179).Preclinical work in ovarian cancer patient-derived xenograft models has shown that the PI3K inhibitor alpelisib (BYL719) inhibits HRR and consequently sensitizes ovarian cancer models with de novo or acquired HRproficiency to olaparib (180).A Phase I study in 28 patients with EOC (75% HGSC) provided preliminary clinical evidence of the efficacy of the combination of olaparib and alpelisib.An ORR of 33% was seen in patients with BRCA-wildtype platinum-resistant EOC compared to an ORR of 3-10% with olaparib or other PARPi monotherapy in the same setting, and with acceptable toxicity (121,181).Importantly, objective responses to this combination of agents occurred regardless of HR status, as measured by targeted DNA sequencing (181).Further evidence will be provided by the ongoing Phase III EPIK-O/ENGOT-OV61 trial (NCT04729387), which is evaluating the efficacy and safety of alpelisib/olaparib compared to single-agent cytotoxic chemotherapy in patients with platinumresistant or refractory BRCA-wildtype HGSC (161).The AKT serine/threonine protein kinases (AKT1, AKT2, AKT3) are key downstream mediators of PI3K signaling (182,183) and in particular, AKT2 has emerged as a poor prognostic marker and potential target in EOC (34,70,71).Drugs targeting AKT have shown activity in breast, endometrial, and ovarian cancer and are currently being investigated in Phase I/II/III trials (183,184).An active Phase Ib/II trial (NCT02208375) is evaluating the combination of olaparib and the AKT inhibitor capivasertib (AZD5363) in a heavily pretreated cohort of 159 patients, with encouraging clinical activity regardless of the presence of a BRCA mutation and despite platinum resistance (183).Further studies are needed to explore the potential of AKT and PI3K inhibitors in combination with PARPi or as monotherapy in HRP HGSC and ovarian cancer in general.

ATR inhibitors
ATR has a major role in the CHK1 (checkpoint kinase 1) pathway of DNA repair and is a regulator of several proteins in the HRR pathway, including activation of BRCA1, PALB2 and RAD51 (108).The potential for mechanistic synergism between ATR inhibitors (ATRi) and PARPis has been demonstrated in HRD and HRP ovarian cancer cells in preclinical models (108,185,186).Acquisition of PARPi resistance was shown to be associated with increased ATR-CHK1 activity, further supporting the potential benefit of combining of PARPis with ATR inhibitors (185).Patient-derived xenograft models of BRCA-wildtype and CCNE1amplified platinum-resistant ovarian cancer, which are associated with increased baseline activation of ATR/CHK1, demonstrated tumor reduction and a significant increase in OS when treated with the combination of PARPi and ATRi (185).Based on these preclinical data, an ongoing Phase II clinical trial of ceralasertib (AZD6738) in combination with olaparib was developed and initial results demonstrated the potential of ATRi to overcome PARPi resistance in an HRD HGSC patient population (187).
ATRis are also being investigated as potential monotherapy, and preliminary anti-tumor activity has been demonstrated in heavily pretreated tumors across a range of histologic types and gene alterations (188).Initial results from TRESR, a phase I trial of ATRi monotherapy with camonsertib, support preclinical findings that ATRi may be clinically active in other patient populations beyond those with loss of function of ataxia telangiectasia mutated (ATM) kinase, including those with other gene alterations (e.g., ARID1A, CCNE1, and MYC) or phenotypic (replication) markers (188, 189).The functional assessment of replication stress biomarkers is thought to be a better predictive biomarker for ATRi response than single aberrant genes in ovarian cancer (190).This statement can also be applied to the selective CHK1/2 inhibitor prexasertib, which showed an increased sensitivity to platinum and olaparib in mouse tumor transplantation models and monotherapy efficacy in BRCA-wildtype platinum-resistant ovarian cancer (191,192).To date, however, there is limited data on the safety and antitumor activity of CHK inhibitors, and a phase II trial of prexasertib was recently terminated prematurely due to COVID-19 and a shortage of investigational drug supplies (193).

HDAC inhibitors
The altered expression of HDACs (histone deacetylases) has been associated with resistance to platinum-based chemotherapy and poor prognosis (194) and HDAC inhibition leads to impaired HRR in cancer cells through reduced expression of critical genes such as BRCA1 and RAD51 (195,196).Konstantinopoulos et al. provided a preclinical rationale for the use of HDAC inhibitors (HDACi) to reduce HRR in HRP ovarian cancer, including CCNE1-amplified tumors, as a means to enhance PARPi activity (197).This approach has been confirmed by further preclinical studies showing that HDACi such as suberoylanilide hydroamic acid (SAHA), romidepsin, panobinostat and entinostat are synergistic with PARPi in HRP ovarian cancer cells (197,198).HDACi downregulate genes in the cyclin E/CDK and HR signaling pathways and thus show a synergistic cytotoxic effect in combination with a PARPi (198)(199)(200).Based on these preclinical results, there is an ongoing Phase I doseescalation trial (NCT04703920) of the combination of the PARPi talazoparib and the HDACi belinostat in metastatic ovarian, breast and prostate cancer.

HSP90
Another attempt to extend the benefit of PARPis to HRP patients is their combination with the heat shock protein 90 (HSP90) inhibitors.HSP90 mediates the maturation, stability and activation of several key proteins involved in DNA repair and HRR, such as CDK1, BRCA1 and BRCA2 (201).Due to its abundant expression, its dependence on adenosine ATP (adenosine triphosphate), and its massive protein interactome, it is an ideal target for pharmacological inhibition (201).Inhibition of HSP90 by ganetespib (STA-9090), a second-generation HSP90 inhibitor, sensitized HRP HGSC cells to talazoparib (201).HSP90 inhibition resulted in downregulation of BRCA1 and RAD51, HRR impairment and increased DNA damage (202).A recent Phase I dose-escalation study showed that the combination of the HSP90i onalespib and olaparib resulted in prolonged disease stabilization, without dose limiting toxicities, in a heavily pretreated patient population with advanced solid tumors (162).Due to limited efficacy as a monotherapy and in other combination studies, further development of onalespib was discontinued (162).However, preclinical and clinical data may support future evaluation of novel combinations of PARPis with other HSP90 inhibitors, such as pimitespib (203).While HSP90 inhibition has the potential to sensitize HRP HGSC to PARPi and other DNAdamaging agents, further clinical research is needed.

BET inhibitors
The BET (bromodomain and extraterminal) protein family includes BRD4, an epigenetic transcription modulator involved in the expression of proteins that regulate the cell cycle and DNA repair (204).BRD4 has been shown to be a necessary factor for the proliferation and survival of HGSC cells (205).In addition, BRD4 amplification is mutually exclusive with BRCA1 and BRCA2 mutations and tends to co-occur with CCNE1 amplification in HGSC, so BET inhibition may be particularly promising in the HRP group (38,(206)(207)(208).Preclinical studies have shown that BET inhibitors (BETis) suppress the expression of WEE1 and TOPBP1 (DNA Topoisomerase II Binding Protein 1) (209,210).WEE1 and TOPBP1 play critical roles in cellular processes related to DNA damage response and cell cycle regulation.WEE1 is a protein kinase that regulates the G2/M checkpoint in the cell cycle, controlling entry into mitosis and allowing time for DNA repair (173,176,211,212).TOPBP1 acts as a scaffold protein that coordinates the activation of ATR kinase in response to DNA damage, thereby initiating signaling cascades essential for DNA repair and cell cycle arrest (213).Dysfunction or dysregulation of these proteins can lead to genomic instability and contribute to the development of diseases such as cancer.Additionally, increased BRD4 expression has been identified as a factor contributing to PARPi resistance in HGSC (210).The specific BRD4 inhibitor INCB054329 was able to directly decrease the activity of both BRCA1 and RAD51 and induce an HRD phenotype (108,209).Consequently, in combination with PARPis, a synergistic effect is observed with decreased HR activity, increased DNA damage, and consequently increased tumor cytotoxicity (108,214).Unfortunately, initial clinical studies involving single agent use of BET inhibitors in various tumor types were disappointing, as preclinical results could not be replicated and resistance to therapy occurred rapidly in some cases (215).Specific evidence in ovarian cancer will be provided by an ongoing Phase II clinical trial (NCT05071937) of the BETi ZEN003694 in combination with the PARPi talazoparib in patients with recurrent ovarian cancer who have progressed on prior PARPi therapy.

Summary
The HRP HGSC subgroup exhibits complex molecular heterogeneity combined with an immune depleted microenvironment, and these are associated with therapy resistance and a poor prognosis.A subset of these cancers are driven by CCNE1 amplification and PI3K/ AKT alterations that contribute to cell cycle dysregulation and thus these pathways represent promising targets for novel therapeutic approaches.However, a significant subset of HRP HGSC lack CCNE1 amplification, and the molecular drivers of these cancers are still being defined.Additional studies, including the use of cell lines and potentially the use of existing data from systematic knockdown and knockout genetic screens (216,217) in the HRP non-CCNE1 amplified subgroup may define critical dependencies.
A large proportion of HRP HGSC are relatively immune depleted, likely in part due to a reduced mutational burden associated with intact DNA repair.The development of novel immunotherapies to boost the anti-tumor immune response remains a key area of focus for HRP tumors, including personalized approaches to enhance T-cell infiltration with therapeutic vaccines or adoptive cell therapy.Several new combination treatments are under investigation, which aim to sensitize HRP cancers to existing therapies, such as platinum and PARPis, by targeting the HRR pathway and impairing the ability of cells to functionally repair DNA.Antibody drug conjugates also represent a promising class of therapies to increase the potency and specificity of highly potent cytotoxic agents, while reducing toxicity.
These new approaches offer the opportunity to expand the otherwise very limited treatment options for patients with HRP HGSC.Importantly, explicit identification and enrollment of patients with HGSC tumors known to have intact HRR in clinical trials is crucial for the development of effective therapies for this medically underserved group.

FIGURE 1
FIGURE 1Immune profile of HRP vs. HRD tumors and effect of Vigil.HRP tumors show reduced immunophenotypic markers compared to HRD tumors.Gemogenovatucel-T (Vigil) is a vaccine composed of autologous tumor cells transfected with a plasmid containing GM-CSF and bi-shRNA resulting in a systemic anti-tumor immune response including CD8+ T cell circulation.HRP tumors have a higher proportion of clonal neoantigen expression, which explains the better effect of Vigil on HRP tumors compared to HRD.HRP, Homologous recombination proficient; HRD, Homologous recombination deficient; TILs, Tumor-infiltrating lymphocytes; Bi-shRNA, Bifunctional short hairpin RNA; GM-CSF, Granulocyte/Macrophage Colony Stimulating Factor; TGF, Transforming growth factor; APC, Antigen presenting cell.

FIGURE 3 CDK
FIGURE 3 CDK/cyclin E1 complex including targets for therapy.Activation of CDK2 by cyclin E1, allowing the cell to enter the S phase.Overexpression of cyclin E1 increases the rate at which cancer cells transition from G1 to S phase, leading to replicative stress and genomic instability.The WEE1 kinase is involved in regulating cell cycle progression by inhibiting CDK1 and CDK2 and WEE1 inhibition leads to cell cycle acceleration, with early mitotic entry and consequent mitotic catastrophe leading to irreparable DNA damage.CDK1 phosphorylates BRCA1 and CDK1 inhibition impairs the ability of cells to carry out functional DNA repair through homologous recombination.PKMYT1 encodes the protein kinase Myt1, a negative regulator of CDK1.

TABLE 2
Clinical trials of potential therapy options for HRP HGSC.