The DNA damaging revolution

Abstract

Poly(ADP-ribose) polymerase (PARP) is a nuclear enzyme that plays a critical role in the repair of single-strand DNA damage via the base excision repair pathway. PARP inhibitors have substantial single-agent antitumor activity by inducing synthetic lethality. They have also emerged as promising anticancer targeted therapies, especially in tumors harboring deleterious germline or somatic breast cancer susceptibility gene (BRCA) mutations. PARP inhibition produces single-strand DNA breaks, which may be repaired by homologous recombination, a process partially dependent on BRCA1 and BRCA2. The PARP inhibitors olaparib, veliparib, talazoparib, niraparib, and rucaparib have predominantly been studied in women with breast or ovarian cancers associated with deleterious germline mutations in BRCA1 and BRCA2. Ongoing clinical trials are evaluating the role of PARP inhibitors alone and in combination with other therapies, including selective inhibitors against key targets involved in the DNA damage response. In this review we summarize the use of PARP inhibitors in various tumor types, as well as possible approaches for overcoming resistance to PARP inhibitors.

I. Introduction

We experience many types of damage to DNA in cells and repairing this DNA damage is crucial to cell survival. Many treatments, including some cytotoxic chemotherapies, function on the basis of DNA damage. Platinum agents, topoisomerase inhibitors, and alkylating agents, among others, have been cornerstones of therapy for many common tumors, all of which function by causing DNA damage. Cancer cells become resistant to these therapies by learning to repair the DNA damage, and so other drugs are needed to inhibit proteins involved in DNA repair pathways. Another group of effective drugs, including WEE1 inhibitors and CHK1 and CHK2 inhibitors, focus on cell cycle regulation. New studies are building on the recent success of poly adenosine diphosphate [ADP]–ribose polymerases (PARP) inhibitor monotherapy and will revolutionize treatment of different tumor and molecular subtypes.

II. The Biology of PARP Inhibitors: From Basics to Clinic

There are 2 main ways that DNA repair occurs. The first is through single-strand repair, and the second is through double-strand repair. DNA can either fix its damage 1 strand at a time or 2 strands at a time. There are 3 primary pathways involved in single-strand DNA repair. The first is the mismatch repair pathway, which is well known due to Lynch Syndrome and because these patients are sensitive to PD-1 inhibitors. The second is the nucleotide excision repair pathway, and the third is the base excision repair pathway. This third one is important, because PARP1 and 2 are the main enzymes involved in the base excision repair pathway (Table 1)[1, 2]. PARP is recruited to the site of the single-strand DNA damage, which leads to the activation of the base excision repair pathway and ultimately to cell survival. Once activated, it uses nicotinamide adenine dinucleotide (NAD) as substrate to add large, branched chains of poly (ADP-ribose) polymers (i.e PARPylation) to itself and to interacting partners. In the presence of PARP inhibitors, an alternate mechanism of DNA repair, mediated by the homologous recombination repair (HRR) pathway, gets activated, and the cell is still able to survive. The DNA repair mechanism remains intact. However, in the presence of PARP inhibitors, in underlying homologous recombination repair deficiency (HRD), the DNA is not able to repair, which leads to accumulation of double-strand DNA breaks, which leads to cell death. Some PARP inhibitors cause the PARP enzyme to bind to the DNA irreversibly, unable to disassociate. When PARP gets trapped on the DNA during S-phase (synthesis phase), it can cause collapse of the replication fork and a double-strand break ensues. Not only is PARP involved in enzymatic inhibition that leads to double-strand breaks, but too much irreversible PARP binding (trapping) can also lead to double-strand repair deficiency [3] (Figure 1). Preclinically, it has been suggested that both enzymatic inhibition and PARP trapping are important drug targets. The strongest PARP1 enzymatic inhibitor is talazoparib, which has the lowest half maximal inhibitory concentration (IC50), and the second best is rucaparib. The strongest PARP trapper is talazoparib, and veliparib is the weakest PARP trapper [4]. Mutations in DNA repair pathways lead to genomic instability.

Genomic instability in BRCA1-BRCA2-deficient cells impairs homolog recombination, and a loss of PARP-1 function in a BRCA1-or BRCA2 defective background could lead to cell cycle arrest and/or cell death [5, 6]. Bryant et al, show that BRCA2-deficient cells are sensitive to PARP inhibitors, because resultant collapsed replication forks are no longer repaired as a result of the cells’ deficiency in homolog recombination [7]. BRCAness can be defined as a defect in double-strand break repair by HRR [8]. DNA can be repaired 2 strands at a time via double-strand repair pathways, the most important mechanism of which is the homologous recombination repair process. Defects in one of the homologous recombination associated DNA repair pathways, such as ATM, ATR, and PALB2, causes sensitivity to PARP inhibitors in preclinical models[9, 10]. Genetic data revealed the presence of additional genes that are defective in cancer, such as CDK12, RAD51B, and RAD51C, which might also confer a BRCAness phenotype [11]. Synthetic lethality occurs when a mutation in either of two genes individually has no effect, but the combination of the loss of both genes is lethal [12]. PARP inhibitors may induce a synthetic lethal interaction between PARP and BRCA1 or BRCA2.

A Phase I trial showed that olaparib could be administered as a single agent at a dose of 400 mg twice per day [13]. Patients with advanced solid tumors (n=60) were enrolled and screened for BRCA1 and BRCA2 mutations (n=22+1) to assess an objective anti-tumor response (15 patients with BRCA 1; 1 patient with BRCA2 mutated ovarian cancer). Nine (47%) BRCA carriers had a response according to the Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1, mostly in ovarian cancer. Twelve (63%) patients had a clinical benefit from treatment with olaparib. These observations are impressive because the study patients were heavily pretreated. Common dose limiting toxicities included reversible myelosuppression, central nervous system effects (somnolence, low mood), and fatigue.

Significant and durable responses were reported only in mutation carriers, all of whom had ovarian, breast, or prostate cancer. Two international multicenter phase 2 studies assessed olaparib in women with confirmed BRCA1 or BRCA2 mutations with advanced breast or ovarian cancer. Patients had been heavily pretreated with various chemotherapy regimens and most were resistant to a wide range of chemotherapies [14, 15]. The reported response rate was 41% in breast cancer at the tolerable dose of olaparib at 400 mg twice daily, and the clinical benefit rate was 52% (n=17) in advanced ovarian cancer with olaparib 400 mg twice daily.

Advanced ovarian cancer is a leading cause of cancer deaths in women. Up to 85% of patients recur after standard first-line platinum-based chemotherapy, thus outlining that standard first-line chemo and active surveillance are not sufficient to meet the needs of patients with newly diagnosed advanced ovarian cancer. When should PARP inhibitors be used? High-grade serous ovarian cancer is a clear indication for a PARP inhibitor, and perhaps all patients with this diagnosis should be exposed to one. Response to platinum therapy might be the best biomarker indicating sensitivity to a PARP inhibitor. The SOLO2, ARIEL3, and NOVA studies showed the benefit of maintenance treatment with PARP inhibitors after a platinum response in germline BRCA1 or BRCA2 mutant high-grade serous ovarian cancer. SOLO2 was published in 2017 and olaparib is now approved by the FDA as maintenance treatment after response to a platinum agent [16].

Around 50% of all high-grade serous ovarian carcinomas may have some defect in the homologous recombination pathway[17]. ARIEL3 and NOVA are the key studies here for maintenance treatment. These studies show that the improvement observed in the homologous recombination deficiency (HRD-positive) cohort was not driven solely by patients with a BRCA-mutant carcinoma. A benefit was still seen in patients with wild-type BRCA. The improvement in survival seen in patients with wild-type BRCA and high loss of heterozygosity (LOH) carcinoma was higher than those with wild-type BRCA and low LOH carcinoma [18]. In the NOVA study, platinum sensitive relapse in patients without germline BRCA mutations of the HRD positive cohort had a hazard ratio (HR) of 0.38, and for HRD negative subjects by the myChoice HRD test, the HR was 0.58 [19]. The ARIEL2 study, published in Lancet Oncology, assessed the ability of tumor genomic LOH to predict response to rucaparib, an oral PARP inhibitor, for high- grade serous or endometrioid ovarian cancer[20]. The results of ARIEL2 showed the usefulness of PARP inhibitors as maintenance therapy in platinum sensitive diseases.

Patients with mutant BRCA and wild-type BRCA with high LOH had longer progression-free survival (PFS) than patients with wild-type BRCA and low LOH. The 40 BRCA positive patients showed the best objective responses with rucaparib treatment. The response rate with rucaparib for patients with BRCA mutations was about 80% in ARIEL2. In Study 19, germline and somatic BRCA mutations produced the same effect in terms of PARP inhibitor sensitivity in platinum-sensitive high-grade serous ovarian cancer [21]. Twenty two percent (22%) of patients with wild-type BRCA had HRR mutations in Study 19 [22]. Study 19 patients had comparable PFS and hazard ratios to the overall BRCA wild-type populations (HR: 0.54) and BRCA-mutated populations (HR: 0.18). Subgroup analysis revealed that olaparib is associated with a greater PFS benefit in HRR-mutated patients without a BRCA mutation (hazard ratio 0.21). A recent study (ARIEL2 Part 1 trial) demonstrated that heterozygous and homozygous BRCA1 promoter methylation predicted response of platinum-sensitive recurrent high grade serous ovarian cancers to PARP inhibitors [23]. The QUADRA study for recurrent ovarian cancer with at least 3 prior therapies, showed that the overall response rate (ORR) to niraparib was 39% in patients with a BRCA mutation, the ORR was 26% in patients with HRD-positive disease, and the mean duration of response was 9.4 months [24]. According to a recent meta-analysis, PARP inhibitors benefited ovarian cancer patients regardless of their BRCA mutational status [25]. Platinum sensitivity is a good surrogate for PARP inhibitor sensitivity in relapsed disease, and HRD is equivalent to PARP inhibitor sensitivity. However, one of the best biomarkers is BRCA1 or BRCA2 mutation status. The U.S. Food and Drug Administration (FDA) approved the PARP inhibitors olaparib, rucaparib, niraparib, and now talazoparib.

Platinum sensitivity cannot be used as a marker in the first line setting because patients receiving PARP inhibitor therapy immediately after first line platinum therapy do not yet have a PFS interval. PRIMA/ENGOT-OV26/GOG-3012 was designed to evaluate the efficacy and safety of niraparib therapy after response to platinum-based chemotherapy in patients with newly diagnosed advanced ovarian cancer [26]. Niraparib provided similar clinical benefit in the HRD subgroups (BRCA mutated and BRCA wild-type). Niraparib provided clinically significant benefit in the HR-proficient subgroup with a 32% risk reduction for progression or death. At the 24-month interval, an interim analysis of overall survival (OS) showed an OS of 84% in the niraparib subgroup and an OS of 77% in the placebo group (hazard ratio (HR), 0.70; 95% CI, 0.44 to 1.11). Niraparib therapy in patients with advanced ovarian cancer provided a clinically significant improvement in PFS after response to first-line platinum- based chemotherapy in all patients (PFS for overall population: HR, 0.62; p<0.001; PFS for homologous recombination deficient: HR, 0.43; p<0.001; PFS for homologous recombination proficient: HR, 0.68; p=0.020).

The most common adverse events of grade 3 or higher were anemia (31 percent), thrombocytopenia (29 percent), and neutropenia (13 percent). Niraparib after first-line platinum-based therapy, is the first PARP inhibitor to demonstrate benefit in patients across biomarker subgroups, consistent with prior clinical studies of niraparib in recurrent ovarian cancer (NOVA and QUADRA). Patients with BRCA positive tumors showed an unprecedented PFS benefit from maintenance olaparib therapy in the phase III SOLO1 trial [27]. The risk of disease progression or death was lower with olaparib than with placebo (60 versus 27 percent; HR for disease progression or death 0.30, 95% CI 0.23-0.41). In PAOLA-1/ENGOT-ov25, women with advanced, high-grade, serous or endometrioid cancers who are receiving first-line standard of care treatment including bevacizumab, were randomly assigned to olaparib and bevacizumab maintenance versus placebo and bevacizumab maintenance [28]. Prespecified subgroup analyses showed that in patients with tumor BRCA mutations, the combination of olaparib plus bevacizumab improved PFS over bevacizumab alone (37 versus 22 months; HR 0.31, 95% CI 0.20-0.47). Patients lacking a BRCA mutation demonstrated PFS of 18.9 months with olaparib versus 16 months with placebo (HR, 0.71; 95% CI 0.58-0.88).

Veliparib potently inhibits PARP with minimal PARP trapping, avoiding myelosuppression due to PARP trapping activity [29]. This randomized, Phase III, placebo-controlled trial was conducted to assess the efficacy and safety of veliparib when added to carboplatin and paclitaxel and continued as maintenance therapy in patients with newly diagnosed high- grade serous epithelial ovarian cancer [30]. Reduction of hazard for recurrence or progression was 56% in patients with BRCA mutations, 43% in patients with HRD, and 32% in the intention-to-treat (ITT) population. Veliparib given only during the chemotherapy cycles did not demonstrate increases in PFS, though a numerically higher ORR was observed for both veliparib-containing arms. Veliparib was safely administered with carboplatin and paclitaxel but was associated with higher incidence of anemia and thrombocytopenia when combined with chemotherapy. Veliparib in combination with chemotherapy and continued as maintenance therapy should be considered a new treatment option for women with newly diagnosed, advanced-stage serous ovarian cancer, regardless of biomarker, type of surgery, or type of paclitaxel regimen. In high-grade serous ovarian cancer, we think the BRCA- related patient population should be offered a PARP inhibitor, possibly as early as maintenance therapy after platinum therapy. It is also a viable treatment with a good chance of response later in the course of treatment, in patients previously treated with platinum therapy. Even in the non-BRCA-mutated population with high-grade serous ovarian cancer, PARP inhibitor therapy can be considered for maintenance treatment after response to platinum therapy after either the first or second relapse.

There is ample evidence that there are problems with DNA repair in patients with breast cancer. This is a disease with a high degree of chromosomal instability as opposed to subtle sequence changes, just like ovarian cancer, and many structural changes within chromosomes are often driven by the presence of double-strand breaks. Whole genome sequencing analysis has revealed mutational signatures indicating DNA repair problems [31, 32]. Some of the most prevalent are abnormalities in homologous recombination pathways where BRCA-1 is active. HRD gets a lot of attention as it leaves a pathogenomic scar on the tumor genome [33]. Both OlympiAD with olaparib, published two years ago, and the EMBRACA study with talazoparib, are other investigations of PARP inhibitors in BCRA- related metastatic breast cancer (MBC) (Table 2) [34, 35]. Both of these studies provided a significant benefit over single agent chemotherapy of the physician's choice with respect to PFS. However, it is important to note that the analysis spanned three months, and no advantage in OS has been seen thus far. PARP inhibitor use did lead to improvements in quality of life. PARP inhibitors are FDA approved for patients with metastatic HER2-negative breast cancer and germline BRCA mutations (ER status does not matter).

The BROCADE-3 trial is the first phase III trial to evaluate a PARP inhibitor with highly- active platinum chemotherapy in patients with advanced breast cancer and a germline BRCA mutation [36]. Eligible patients had BRCA-mutated, HER2-negative MBC with ≤ 2 prior lines of chemotherapy for metastatic disease and ≤ 1 prior platinum-based regimen. Veliparib, when added to carboplatin and paclitaxel, provided a statistically significant and clinically meaningful benefit in patients with HER2 negative advanced breast cancer and a germline BRCA mutation. The hazard for disease progression was reduced by 29%. Median PFS for the veliparib + carboplatin/paclitaxel arm was 14.5 months, 1.9 months longer than that of the control arm. The observed benefit was durable, with 26% of patients on veliparib + carboplatin /paclitaxel alive and progression-free at three years, versus 11% of patients on placebo + carboplatin/paclitaxel alive and progression-free at three years. Addition of veliparib did not substantially alter the toxicity profile of carboplatin/paclitaxel. Platinum- based chemotherapy correlates with high levels of response in HRD and BRCA deficient cancers more than in non-mutated forms of triple negative breast cancer (TNBC) [37].

Neoadjuvant single-agent oral talazoparib once per day for 6 months without chemotherapy substantially reduced residual cancer burden rate with manageable toxicity. The rate of patients with residual cancer burden-0 (pathologic complete response) was 53% and the rate of patients with residual cancer burden-0/I was 63% [38]. GeparOLA (NCT02789332) randomized 102 patients to either paclitaxel 80 mg/m² weekly plus olaparib 100 mg twice daily for 12 weeks (n = 65) or paclitaxel 80 mg/m² weekly plus carboplatin (AUC2) weekly for 12 weeks (n = 37), both followed by epirubicin and cyclophosphamide [39]. Pathologic complete response (pCR) rate in the paclitaxel plus olaparib group was 55.1% compared with 48.6% in the paclitaxel plus carboplatin group. The Phase III BrighTNess trial included 634 patients with previously untreated histologically confirmed stage II–III TNBC who were candidates for potentially curative surgery [40]. The patients were randomly assigned to receive one of three regimens: weekly paclitaxel 80 mg/m2 for 12 weeks plus placebo; paclitaxel plus four cycles of carboplatin (AUC6) given every 3 weeks and placebo; or paclitaxel plus carboplatin and veliparib 50 mg twice daily. All patients were then given a second stage of treatment consisting of four cycles of doxorubicin and cyclophosphamide.

Overall, pCR was achieved by 31% of the paclitaxel arm and by 53% of the paclitaxel plus carboplatin and veliparib treatment arm. However, the frequency of patients achieving a pCR did not differ between those receiving paclitaxel plus carboplatin plus veliparib versus paclitaxel plus carboplatin ([53%] vs [58%]). In an international, phase 3, placebo-controlled trial, patients were randomly assigned in a 2:1 ratio to receive carboplatin/paclitaxel plus veliparib or carboplatin/paclitaxel plus placebo [36]. Patients had germline BRCA1/2 mutations and ≤2 prior lines of cytotoxic therapy for metastatic breast cancer. The most common (≥20%) adverse events up to grade 3 in the veliparib-combination arms were anemia (27% vs. 17%), neutropenia (52% vs. 50%), and thrombocytopenia (25% vs. 15%). Patients on the veliparib arm had a durable benefit compared to the control and demonstrated significant improvement in PFS. At this time, germline BRCA1/2 pathogenic variants are the best predictors of PARP inhibitor sensitivity in breast cancer.

The frequency of BRCAness is approximately 20% in metastatic castration-resistant prostate cancer (mCRPC) [41]. TRITON2 is an ongoing phase 2 study evaluating rucaparib in patients with mCRPC and a deleterious somatic or germline alteration in the DNA damage response (DDR) gene. All men included in the study progressed on AR-directed therapies such as abiraterone and enzalutamide and a prior taxane. Rucaparib demonstrates promising efficacy in patients with mCRPC and a germline or somatic BRCA or other DDR gene alteration: 43.9% had an objective response, and 52% had a confirmed PSA response [42]. GALAHAD is an ongoing open-label Phase II study assessing niraparib (300 mg daily) in patients with mCRPC and DNA repair defects, with disease progression on a taxane and an androgen receptor-targeted therapy.

Niraparib has high clinical activity in patients with BRCA1/2 mutant-positive disease with an objective response rate of 41% and a median radiographic PFS (rPFS) of 8.2 months [43]. PROfound was a randomized, open-label, phase III trial evaluating the efficacy and safety of olaparib vs physician’s choice in patients with pretreated mCRPC with HRR alterations whose disease had progressed on a prior second-generation hormonal therapy (Table 2) [44]. Patients were randomized 2:1 to olaparib (300 mg bid) or to physician’s choice of enzalutamide (160 mg/day) or abiraterone (1000 mg/day + prednisone 5 mg BID). In patients with mCRPC with disease progression on prior second-generation hormonal therapy, olaparib provided a statistically significant and clinically meaningful improvement in rPFS compared with physician’s choice of enzalutamide or abiraterone + prednisone in patients with BRCA1, BRCA2 and ATM mutations. Olaparib improved multiple clinical and patient-reported endpoints (rPFS, ORR, time to pain progression). Despite >80% cross-over, at an interim analysis, olaparib had a favorable trend in OS for patients with alterations in BRCA1, BRCA2 and/or ATM (HR=0.64), as well as in the overall population (HR=0.67). Adverse effects were very similar to what we see in non- prostate cancer patients: anemia, thrombocytopenia, neutropenia, fatigue, nausea and vomiting. PARP inhibitors as monotherapy have high levels of anti-tumor activity in mCRPC patients with identified alterations in BRCA2 and select other DNA repair genes. The National Comprehensive Cancer Network (NCCN) guidelines now recommend germline genetic testing for all patients with high-risk, lymph node-positive or metastatic prostate cancer. Somatic genomic testing is indicated for all patients with metastatic prostate cancer.

Olaparib can be considered for maintenance treatment following first-line platinum-based chemotherapy in patients with a germline BRCA mutation and metastatic pancreatic cancer [45]. In the POLO trial, patients were randomly assigned to olaparib (300 mg twice daily) or placebo (92 patients to receive olaparib and 62 to receive placebo). The median PFS was significantly longer in the maintenance olaparib group than in the placebo group (7.4 months vs. 3.8 months, with a hazard ratio for disease progression or death of 0.53; [CI] 0.35 to 0.82; P=0.004). OS was similarly high in both groups (median 18.9 versus 18.1 months) (Table 2). A meta-analysis of the efficacy of PARP inhibitors in various solid tumors showed that the PARP inhibitor was superior for long-term survival in terms of OS and PFS compared with the control group [46].

III. Partners With PARP?

Combination approaches are also being investigated to allow further patients to benefit from PARP inhibitor therapy. Single agent treatments are rarely effective for solid tumors. Ideal combinations are those that inhibit tumor growth and show synergistic interactions, and both tumor and patient characteristics need to be considered in the selection of the optimal treatment. Is one drug sufficient or do we need rational combinations? Not all patients with BRCA1/2 mutant cancers will respond, and drug resistance is nearly inevitable in all patients. With combination therapy, there is the potential to deepen responses, to increase the durability of these responses, and to explore wider applications of PARP inhibitors beyond BRCA and beyond BRCA-ness. Is it possible to raise the tail on a survival curve with rational PARP inhibitor combinations? In the next five to ten years, our prediction is that due to overlapping toxicities, there will be less emphasis on concurrent chemotherapy combinations and a greater emphasis on immunotherapy (IO), targeted agents including DDR, and also on radiotherapy.

The goal is to build on the success of PARP inhibitor monotherapy. What does this approach entail? It is well known that a homologous recombination-deficient tumor may be sensitive to DDR agents. The sensitivity to PARP inhibitors can be enhanced by inducing HRD phenotype in HR proficient tumors with molecularly targeted agents [47]. In this way, combination treatment strategies can expand patient populations that might benefit from DDR inhibitors. There are now several examples of this strategy. Hypoxia leads to impaired HR by down-regulating HR genes [48]. Cediranib suppresses homology-directed DNA repair by down-regulating BRCA1/2 and RAD51 [49]. Cediranib/olaparib compared with olaparib alone significantly improves PFS (23.7 versus 5.7 months, P = 0.002) and OS (37.8 versus 23.0 months, P = 0.047) in germline BRCA wild-type/unknown patients in a randomized phase 2 trial [50]. However, diarrhea, fatigue, and hypertension are common with cediranib and olaparib. Other potential targeted agents might include inhibitors against MEK, BET, or PI3K/AKT and also inhibitors against the androgen receptor pathway, such as the CYP17 inhibitor, abiraterone. MEK (MAP kinase/ERK kinase) inhibition in combination with PARP inhibitor is being investigated in RAS mutant cancers [51-55]. The PARP inhibitor olaparib, given in combination with the PI3K inhibitor alpelisib, is clinically feasible with no unexpected toxic effects. The combination showed an ORR of 33% in patients with advanced ovarian cancer who were largely platinum resistant [56]. BET bromodomain (BRD4) protein inhibition decreases homologous recombination competency by depletion of the DNA double-stand break resection protein CtIP (C-terminal binding protein interacting protein) [57]. BRD4 inhibition resensitizes the cell with acquired PARP inhibitor resistance to the PARP inhibitor. BET inhibitors in combination with PARP inhibitors may have wide applications in the future.

There are multiple nonclinical and translational studies which suggest the potential for synergism when PARP inhibitors are combined with androgen receptor (AR) signaling inhibitors. Inhibition of AR signaling suppresses the expression of genes associated with DNA damage response [58]. Also, PARP activity has been shown to support the function of the androgen receptor, suggesting that co-blockade of PARP with androgen receptor- directed therapy may be synergistic [59]. In addition, clinical resistance to AR blockade has been shown to associate with codeletion of RB1 and inactivation of BRCA2, which means inactivation of BRCA2 may lead to PARP inhibitor sensitivity [60]. In a phase II trial, 140 patients were randomized to the combination of abiraterone plus olaparib versus placebo plus abiraterone. Patients must have had prior docetaxel chemotherapy for mCRPC but no previous second-generation antihormonal agents [61]. Median rPFS was 13.8 months with olaparib and abiraterone and 8.2 months with placebo and abiraterone ([HR] 0.65, 95% CI 0.44-0.97, p=0.034).

The phase II TALAPRO-1 trial enrolled approximately 100 patients with progressive metastatic castration-resistant prostate cancer (mCRPC) who have DNA damage repair mutations [62]. The overall objective response was 25%, the objective response rate for patients with BRCA1/2 mutations was 50.0%, and the rate was 7.1% for those with ATM mutations. This was novel, because so far, no responses have been seen with PARP inhibitors in patients that harbor ATM mutations in the tumor. In patients who had BRCA1 and BRCA2 mutations, the PSA responses were 64%, and they decreased to 33% in patients who had a PALB2 mutation. For all 53 patients, regardless of their DNA repair defect, the median rPFS was 5.6 months. Interestingly, the median rPFS at 8.2 months was much higher in patients who harbored BRCA1 and BRCA2 mutations. The rPFS was lower in other mutant types. The aforementioned Phase III trials are investigating whether these combinations of a PARP inhibitor with a novel hormonal therapy are only efficacious in patients who have an HRR mutation, or whether these combinations may be effective in patients who do not harbor these DNA repair defects.

PARP inhibitors can also elicit antitumor immunity. Recent studies demonstrated increasing evidence that links PARP inhibitors and IO. The initial hypothesis was that PARP inhibition would lead to S phase DNA damage, leading to increased new antigen expression, and thus creating a more antigenic immune microenvironment [63]. S phase-specific DNA damage leads to accumulation of cytosolic DNA, which activates the c-GAS-STING innate immune response, stimulates type 1 IFN, and upregulates PD-L1 expression, therefore sensitizing to IO agents [64]. PARP inhibition inactivates GSK3β, leading to PD-L1 upregulation, which also demonstrates in vivo synergy between PARP inhibitors and PD-1 inhibitors.

Combining DDR and PD-1/PD-L1 inhibitors is a rational antitumor strategy. The MEDIOLA study investigated the combination of olaparib plus avelumab for platinum sensitive recurrent BRCA1 and 2 mutant ovarian cancers (ClinicalTrials.gov Identifier: NCT02734004). The result was an impressive disease control rate of 81%, and notable super responders who, at more than a year, are still on treatment. On the other hand, there were also very refractory patients. We were impressed by the complete response rates of about 20% in this patient population and were pleased to see that there is currently an ongoing global phase 3 trial of this combination in first line ovarian cancer. The TOPACIO study evaluated the PARP inhibitor niraparib combined with pembrolizumab in patients with platinum resistant ovarian carcinoma [66].

It compared BRCA mutant and HRD-positive patients versus BRCA wild type and HRD-negative patients. HRD status did not correlate with response to this combination in platinum resistant/refractory disease. The addition of pembrolizumab to niraparib in BRCA wild type tumors with HDR negative status, led to a similar ORR as PARP inhibitor monotherapy in a tumor mutant BRCA population. We also see myelosuppression with DDR and DDR inhibitor combinations. It seems like the only class of drugs used in combination that does not have overlapping toxicities with PARP inhibitors is PD-1/ PD-L1. Induction of DNA damage can induce an immune response that is enabled by checkpoint inhibition in recurrent TNBC [67]. 67 patients with metastatic TNBC were randomized to nivolumab with or without induction, including irradiation (3 × 8 Gy), cyclophosphamide, cisplatin, or doxorubicin, all followed by nivolumab. This study indicates that short-term doxorubicin and cisplatin may induce a more favorable tumor microenvironment and increase the likelihood of response to PD-1 blockade in TNBC.

In a phase II study, patients who had progressive mCRPC and who had progressed on enzalutamide and/or abiraterone were enrolled. This was a single-arm study, and patients received durvalumab with olaparib. The primary endpoint was clinical efficacy, and secondary endpoints were response rates, safety, duration of response, and prostate-specific antigen (PSA) response [68]. Interestingly, there was no limitation on previous standard therapies for the mCRPC population, so the patients could have received any number of prior therapies in the castration-sensitive or castration-resistant setting. Median rPFS for all patients is 16.1 months, although the PFS was longer in patients who harbored DNA damage repair defects. Cybrexa has also developed CBX-11, a conjugate of talazoparib with an alphalex peptide. CBX-11 targets these PARP inhibitors to acidic cells, like tumor cells, and therefore spares normal tissue from the toxic effects of PARP inhibitors. We need to determine the optimal order of these rational PARP inhibitors combination therapies. Rational combinations will hopefully widen the breadth of application of PARP inhibitors to different tumor and molecular subtypes beyond the current approved indications. Concurrent chemotherapy is challenging, and thus approaches with IO, DDR agents, and other targeted agents are encouraging. Combination efficacy needs to be balanced against the synergistic toxicities that have been observed. Ultimately, further studies and better trial designs are needed to examine these combinations.
IV. Novel DNA Damage Response Modulators in the Clinic

PARP inhibitors induce increased DNA adducts which stall replication forks, and ATR is required for the repair of such stalled replication forks (Figure 2). In the setting of ATM deficiency, PARP will lead to increased DNA damage and cell death [69] ATR inhibitors are important in DNA damage repair. Double-stranded DNA breaks induce activated ATR to phosphorylate CHK1, resulting in cell cycle arrest, so that the DNA can repair itself and ensure cellular survival. Inhibition of ATR could prevent cell cycle arrest, and hopefully lead to cell death [70]. It has been shown that ATR plus CHK1 inhibition leads to replication fork arrest, DNA SSB accumulation, replication collapse, and synergistic cell death in cancer cells [71]. On the other hand, BRCA1/2 deficient cells can use ATR to protect forks and enable HR repair. PARP inhibitor-resistant, BRCA1-deficient cells are increasingly dependent on ATR for survival. Thus, ATR inhibition is a unique strategy to overcome the PARP inhibitor resistance in BRCA-deficient cancers[72]. AZD6738 is a potent selective oral ATR inhibitor.

45 patients with advanced cancer were treated with olaparib plus AZD6738 [73]. Of 39 evaluable patients, 1 RECIST complete response (CR), 5 partial responses (PR) and 1 unconfirmed response were observed in patients with advanced breast (3 patients), ovarian, prostate, pancreatic and ampullary cancer and BRCA1/2 mutations independent of ATM status. Thrombocytopenia and neutropenia were dose- and schedule limiting toxicities. VX- 970 is a potent, selective ATR inhibitor that has demonstrated marked preclinical antitumor activity in combination with chemotherapy [74, 75]. In a study of twenty-six patients, one patient with ATM deficient colorectal cancer had a complete response and four patients had stable disease. This potentially suggests a biomarker-selected population to consider for treatment with VX-970. PFS was also improved when combined with a platinum. VX-970 was generally well tolerated with mainly G1-2 toxicities. In another study, the same drug under a different name showed that a combination of selective ATR inhibitor M6620 (previously VX- 970) and topotecan, a selective TOP1 inhibitor, would be tolerable and active, particularly in tumors with high replicative stress [76]. Three out of the five patients with small cell lung cancer were responders, all of whom were previously rapidly progressing.

The most common treatment-related grade 3 or 4 toxicities were anemia, leukopenia, and neutropenia (19% each); lymphopenia (14%); and thrombocytopenia (10%). There are a couple ATM inhibitors in current use in clinic, as well. Two key checkpoint kinases, CHK1 and CHK2 may be regulated by ATM and ATR. As mentioned before, CHK1 and CHK2 are important proteins that play a critical role in DNA damage-induced cellular responses[77]. However, these drugs have been inordinately difficult to develop. For 15 years, they have been available to clinicians, but their use has been significantly limited by toxicity and by fears of limited activity. AZD7762 is a checkpoint kinase 1 (Chk1) inhibitor that increases sensitivity to DNA- damaging agents, including gemcitabine. Sausville et al. evaluated the safety of AZD7762 in monotherapy and with gemcitabine in the setting of advanced solid tumors [78]. First, patients received a cycle of monotherapy, revealing minimal toxicity at lower dose levels. At higher dose levels, even as a single agent, significant cardiac toxicity and hyperglycemia was observed, which has been dose-limiting. Adding gemcitabine even at lower doses significantly increased toxicities, which made this an intolerable regimen. In terms of activity, only two patients had responses out of 42 patients that were treated, and both of those were gemcitabine naive non-small cell lung cancer patients, i.e. not showing any added benefit.

V. What is New in DNA Damage Response

PARP inhibitors work very well as monotherapy in HRD cancers. Many approaches arrest or affect repair of DNA lesions at the replication apparatus. Deoxyuridine triphosphatase (duTPase) generates deoxyuridine monophosphate (dUMP), which is the substrate for thymidylate synthase, and finally produces thymidine. Thymidylate synthase is a target for 5- fluorouracil. dUTPase inhibitors have no effect on replication stress but increase 5-FU induced replication defects and DNA damage, and potentiate 5-fluorouracil treatment [100]. PARP inhibitors have seen use in the clinic in combination with 5-fluorouracil, but 5- fluorouracil alone has minimal efficacy. The DNA polymerase θ (Polθ also known as POLQ, encoded by POLQ) is involved in alternative end joining and suppresses HR [101]. POLQ overexpression is a promising genetic instability and a poor prognostic marker for breast cancer [102]. Removing POL theta can enable effective targeting of HRD defective cancers [103]. Polθ depletion significantly impaired tumor growth after PARP inhibitor treatment. POLQ is a polymerase that is very important for maintaining the health of the DNA replication fork in the setting of homologous recombination deficiency. A process called microhomology- mediated end-joining is a druggable target, so PolQ inhibitors are in development [101].

POLQ is over expressed in cancer and is required for survival of HRD cancer. The antibiotic novobiocin directly binds to the POLθ ATPase domain, inhibits its ATPase activity, and phenocopies POLθ depletion. BRCA-deficient tumor cells and those with acquired PARP inhibitor resistance are sensitive to novobiocin in vitro and in vivo. Zhou et al. showed that novobiocin may be useful alone or in combination with a PARP inhibitor in treating HR- deficient tumors [104]. Another new approach is to target G-quadruplex (G4s) DNA structures. G-quadruplexes (G4s) occur across the genome and are enriched at telomeres. G-quadruplexes have proven to have important roles in the regulation of gene expression and transcription, DNA repair, protein translation and proteolysis, and other cancer-related cell functions. Currently, there are small molecule drugs which can bind and stabilize them. This can arrest the DNA replication fork in a way that requires functional homologous recombination [105]. CX-5461 is a promising therapy in combination with a PARP inhibitor in HR-deficient high grade serous ovarian cancer-patient-derived xenograft (PDX) in vivo [106].

CX-5461 is a G-quadruplex stabilizer, now in an advanced phase I clinical trial for patients with BRCA1/2 deficient tumors (Canadian trial, NCT02719977). Groove binding agents are another selective inhibitor of active transcription. This agent causes irreversible stalling of RNA polymerase II on DNA and may lead to R loop accumulation and induces HR for repair of arrested broken forks [107]. In patients with BRCA2 mutations, ORR was 61%, PFS was observed to be 5.9 months and OS was observed to be 26.6 months. In patients with a BRCA1 mutation, ORR was 26 %, PFS was observed to be 3 months, and OS was observed to be 15.9 months. The overall response rates are particularly impressive, but in BRCA2 subsets the study is not powered to tell the difference between genotypes. However, there is an interesting level of activity in a heavily pretreated population.

MTHFD2, a mitochondrial methylenetetrahydrofolate dehydrogenase and cyclohydrolase involved in one-carbon metabolism, is one of the most highly overexpressed enzymes during neoplastic transformation. MTHFD2 mRNA expression is high in proliferating cells and is cancer specific. MTHFD2 is an NAD(P)-dependent enzyme and plays an essential role in mitochondrial one-carbon folate metabolism [108]. It may play critical roles in cellular detoxification. Targeting MTHFD2 causes cancer-selective replication stress. MTHFD2 is normally expressed during embryogenesis, and not in adult tissues, but it can be re- expressed in cancer. It negatively correlates with survival in breast cancer patients [109]. MTHFD2 knockdown impairs proliferation of cancer cells independent of tissue of origin [109, 110]. MTHFD2 suppression decreased leukemia burden and prolonged survival. [111]. MTHFD2 inhibitors are not toxic to non-transformed cells, but toxic only to the cancer cells. This is one way that we can avoid the normal tissue toxicity that is often limiting for DNA repair inhibitors. Gustafsson et al. identified the first MTHFD2 inhibitor LY345899 [112]. The folate analog LY345899 acting as the MTHFD2 inhibitor, displays therapeutic activity against colorectal cancer [113]. Synergy between MTHFHD2 inhibitors and ATR inhibitors cause replication stress in cancer cells. We need to study how these approved agents change the tumor landscape and DNA repair mechanisms, and how other drugs may help delay or tackle resistance.

V. Conclusion

We have discussed various tumor types and DNA damaging agents in this review, as well as some recently published studies that provide a different perspective. Germline BRCA mutations in ovarian, breast and prostate cancers are a good example of DDR dysfunction that can be targeted with PARP inhibitors. Ovarian cancer treatment paradigm is rapidly improving. In the maintenance therapy, PARP inhibitors have led to a substantial, unprecedented improvement in PFS in patients with ovarian cancer. BRCA as a mutation is the best for single agent activity, but not all BRCA-related cancers are the same (ovary>prostate>breast). Further studies need to build on PARP inhibitor monotherapy efficacy, and rational combinations will widen the breadth of application of PARP inhibitors.

Concurrent chemotherapy combinations are challenging, and we should really be switching our attention to rational approaches with IO, with DDR agents, and with other targeted therapies. Combination approaches that target DNA damage, such as DDR targeting agents with immunotherapy have a sound rationale, but much more information is needed to maximize the combination’s effects. Combination efficacy needs to be balanced against synergistic toxicities. The main issue to date has been the inability to select patients who do not benefit from PARP inhibition (NOVA, ARIEL3, PRIMA). MK-4827 More research is required to improve our ability to predict PARP inhibitor sensitivity, particularly in patients who do not have evidence of HRD. A functional dynamic biomarker may be the best to recapitulate the HR status at any timepoint during tumor evolution, especially in patients in whom resistance to treatment develops. A key area of focus going forward will be key mechanisms of resistance and how to abrogate these. ATR, CHK1 or Wee1 inhibitors are in clinical trials and may target frequent replication stress and reduce emergence of PARP inhibitor resistance. The optimization of treatment selection and sequencing strategies of current agents deserve further research attention in order to make good use of these promising agents.