The OlympiAD trial enrolled 302 metastatic breast cancer patients with both triple negative (TN) (49.8%) and hormone receptor positive (HR+) HER2 negative (50.2%) tumors. Not more than two lines of chemotherapy for metastatic disease were permitted. Pre-treatment with platinum was allowed, but the last dose should have been administered at least 12 months before randomization. The patients were randomized in a 2:1 ratio to receive olaparib, 300 mg bid, monotherapy or treatment physician’s choice (TPC) among capecitabine, eribulin, or vinorelbine. The primary endpoint was PFS. The secondary endpoint included OS, overall response rate (ORR), and health-related quality of life (HRQoL). The primary analysis showed that PFS was significantly longer in the olaparib arm than the standard chemotherapy (7.0 vs. 4.2 months; HR: 0.58; 95% CI: 0.43 to 0.80; p < 0.001). ORR was also higher in the olaparib group than in the standard chemotherapy group (59.8 vs. 29.8%). OS, on the contrary, did not differ from the two arms (HR for death, 0.90; 95% CI, 0.63 to 1.29; P = 0.57), but the trial was not powered to assess OS differences. In the forest plot, HR was lower in the TN subgroup than in the HR+ subgroup (0.43 vs. 0.82). Finally, olaparib had a good safety profile: there were fewer grade 3 events and fewer discontinuations related to an adverse event in the olaparib arm than in the chemotherapy arm. The side effects reported were comparable to previously published phase I and II trials with anemia, nausea, vomiting, fatigue, headache, and cough occurring more frequently in the olaparib group than in the standard therapy group (37).

The planned study final analysis with OS update has been recently published (66). Overall, OS was not improved by olaparib treatment compared to standard chemotherapy (19.3 months with olaparib versus 17.1 months with TPC HR: 0.90, 95% CI: 0.66–1.23; p = 0.513); however, when patients were stratified according to pre-defined subgroups, an OS benefit was observed in patients who had not received prior chemotherapy for metastatic disease (first-line treatment, 22.6 versus 14.7 months; HR: 0.51; 95% CI: 0.29–0.90). Safety data were also updated: no new findings were reported. Overall, the incidence of grade 3 adverse events was 38%, while 5% of patients discontinued olaparib because of toxicity.

Another important secondary endpoint of the OlympiAD trial was the quality of life (QoL) of the patients. Investigators employed the European Organisation for Research and Treatment of Cancer Quality-of-Life Questionnaire Core 30-item module (EORTC QLQ-C30) to assess patient global health status/QoL. The final results (67) showed a significant QoL improvement in the olaparib arm compared to the TPC arm with a mean change of 3.9 (standard deviation 1.2) versus -3.6 (2.2), a difference of 7.5 points (95% confidence interval, CI: 2.48, 12.44; p = 0.0035). In addition, for EORTC QLQ-C30 symptoms and functioning subscales, only the nausea/vomiting symptom score was worse in the olaparib arm than in the TPC arm (across all visits compared with baseline) (68).

An extended follow-up exploratory analysis of the OlympiAD trial was presented at the San Antonio Breast Cancer Symposium in December 2019 (69). The median follow-up was 18.9 vs. 15.5 months in the olaparib vs. the TPC arms, respectively. Median study treatment duration was 8.3 months in the olaparib arm vs. 3.5 months in the TPC arm, and in the olaparib arm, 8.8% of patients received the treatment for more than 3 years, while no one did in the TPC arm. The results of the extended follow-up confirmed previously published results: no OS differences were registered between the two arms in the overall population (19.3 months for olaparib vs. 17.1 months in the TPC arm, HR: 0.84, 95% CI: 0.63–1.12), but an OS benefit was detected in the subgroup of patients treated with olaparib who had not received chemotherapy for metastatic setting (first-line treatment: OS 22.6 month in the olaparib arm vs. 14.7 months in the TPC arm, HR: 0.54, 95% CI: 0.32–0.92). In first-line subgroup, 40.8% of patients in the olaparib arm were alive at 3 years compared with 12.8% of patients in the TPC arm. No new safety data and no cumulative toxicity occurred at the extended follow-up analysis, confirming good olaparib tolerability even in long-term exposure.

The phase III EMBRACA trial was an open-label, randomized trial, comparing talazoparib versus choice of standard chemotherapy of the physician (capecitabine, eribulin, gemcitabine, or vinorelbine) in pretreated locally advanced (not amenable to curative treatment) or metastatic BRCA1/2 mutated breast cancer (36). A total of 431 patients were randomized in a 2:1 ratio to receive talazoparib at a dose of 1.0 mg daily (n = 287) vs. standard chemotherapy (n = 144). Forty percent of the enrolled patients were TN. No more than three previous chemotherapy regimens were admitted. Patients must have had previously received anthracyclines and taxanes, unless clinically contraindicated. Previous platinum-based adjuvant or neoadjuvant chemotherapy was admitted only if the patients had a disease-free interval of at least 6 months from the last platinum dose. The primary endpoint was PFS by blinded independent central review. The secondary endpoints were OS and ORR. Safety and patient-reported outcomes were also assessed.

At a median follow-up of 11.2 months, the EMBRACA trial met its primary endpoint: the median PFS was significantly higher in the talazoparib arm (8.6 months; 95% CI, 7.2–9.3) than in the standard chemotherapy arm (5.6 months; CI, 4.2–6.7). The HR was 0.54 (95% CI, 0.41–0.71; p = 0.001), and it was confirmed by an independent radiologic review. The PFS HRs were consistent among subgroups, specifically, for HR+ and TN. The PFS HR was 0.47, 95% CI: = 0.32 to 0.71 for HR+/HER2− and 0.60, 95% CI: = 0.41 to 0.87 for TN (32, 33). The response rate by the investigators was 62.6% in the talazoparib arm compared with 27.2% in the chemotherapy arm.

At interim analysis, the median OS was longer in the talazoparib arm (22.3 months) than in the chemotherapy arm (19.5 months), but it did not reach statistical significance (HR: 0.76; 95% CI, 0.55–1.06, p = 0.11). Data about safety showed that the most common all-grade adverse events for talazoparib were anemia, fatigue, and nausea, while for chemotherapy nausea, fatigue, and neutropenia were more frequent. Grade 3 or 4 hematologic adverse events occurred in 55 vs. 36.1% of patients in the talazoparib and standard chemotherapy arms, respectively. Grade 3 non-hematologic adverse events occurred in 32 vs. 38% of patients in the talazoparib arm and in the standard chemotherapy arm, respectively. However, discontinuation rate due to an adverse event was low: 5.9% in the talazoparib group vs. 8.7% in the standard chemotherapy group. Moreover, talazoparib significantly delayed the onset of a clinically meaningful deterioration of global health status of QoL questionnaire, and it also significantly delayed deterioration according to breast symptom scale compared to chemotherapy (70).

In the final OS analysis, published after that 75% of the events occurred (324 patients), talazoparib showed no OS benefit compared to chemotherapy: median OS was 19.3 months (16.6–22.5 months) versus 19.5 months (17.4–22.4 months); HR: 0.848 (95% CI: 0.670–1.073; P = 0.17). A possible explanation for the lack of OS benefit relies on subsequent treatment that could have impaired the analysis: 32.6% of patients randomized to TPC received a PARP inhibitor in later lines of treatment (at the time of EMBRACA publication, olaparib had already been approved for metastatic breast cancer patients harboring gBRCA 1/2 mutation) (71).

More recently, a Cochrane metanalysis investigated the efficacy of PARPib in metastatic breast cancer patients with BRCA 1 or 2 mutations (72). The primary outcome was OS, while the secondary outcomes were PFS, tumor response rate, and safety. The authors included five trials involving 1,474 patients. PARPib showed a small OS benefit: HR: 0.87 (95% CI: 0.76 to 1.00; P = 0.05; high-certainty evidence), with no significant heterogeneity (I ² = 0%, P =0.81). Unfortunately, subgroup analysis could not be performed because data were not available for the included trials. On the contrary, PARPib significantly prolonged PFS with a HR of 0.63 (95% CI: 0.56 to 0.71; P < 0.00001; high-certainty evidence), with no significant heterogeneity (I ² = 2%, P = 0.39). For patients with TNBC (N = 664, four randomized controlled trials, RCTs), there was evidence of PFS benefit on pooling of studies (HR: 0.61, 95% CI: 0.47 to 0.80; P = 0.0003), with moderate heterogeneity (I ² = 44%, P = 0.15).

In addition to olaparib and talazoparib, other PARPib are currently under investigation in TNBC: rucaparib, niraparib, and veliparib. Of note is that the phase III Bravo trial, which investigated the role of niraparib versus TPC in BRCA mutated breast cancer, was prematurely closed because of high discontinuation rate in the control arm (the patients enrolled in the control arm did not continue the trial long enough to receive their first radiological scan, which is required to assess disease progression, resulting in an unusually high rate of censoring) (68). A complete list of other published trials (73) and monotherapy ongoing trials of PARPib in TNBC is summarized in Tables 1 , 2 (74, 75, 80, 81).

Despite the fact that the role of PARPib as a therapeutic milestone is now confirmed in the management of BRCA-mutant TNBC, approximately 50% of patients progressed during treatment (76). From preclinical studies, four principal mechanisms of resistance have been identified (77): (i) the influence of cellular availability of the inhibitor, mainly by overexpression of drug-efflux transporter genes; (ii) direct impact on the activity and abundance of PAR chains due to PARP1 mutations that diminish trapping of the protein on DNA or the loss of PAR glycohydrolase, which is responsible for the degradation of PAR chains; (iii) the occurrence of “reversion mutations” that lead to the reactivation of both BRCA1/2 function and of HR by the activation of a specific protein complex (53BP1–RIF1–Shieldin axis); and (iv) influence of replication fork protection, mainly due to the attack by MRE11 and MUS81 nucleases.

Although the clinical relevance of this issue needs to be proven, some new drugs are engineered to target the acquired vulnerabilities of resistant tumors, thus restoring PARPib sensitivity.

Overall, PARPib showed improved PFS and response rate compared with standard chemotherapy, but no difference in OS was observed in those studies (72). Thus, the development of new agents and/or combination strategies are urgently needed to overcome PARPib resistance and to better understand TNBC molecular aspects. Several ongoing clinical trials aiming at evaluating the safety and efficacy of PARPib in combination with immune checkpoint inhibitors (78, 79, 82–90), chemotherapy (91–94), or target agents (95–100) for advanced BC (including TNBC) are summarized in Tables 3 – 5 . Particularly promising are the data that emerged with combinations of PARPib and immunotherapy according to the durable response rates (101–103).

The combination between a PARPi and ICI is based on the evidence of the interaction between the abnormal presence of unrepaired DNA in the cytoplasm of TN tumor cells and the activation of the stimulator of interferon genes pathway which leads to the release of interferons and enhances T-cell infiltration inside the tumor (104). Thus, combining ICIs with a PARPib could be a great strategy to improve the antitumor immunity as well as response to treatment. Promising efficacy and safety findings have been reported in two single-armed phase 2 studies: TOPACIO and MEDIOLA for niraparib combined with pembrolizumab and for olaparib plus durvalumab, respectively (78, 79, 102, 103). The TOPACIO trial enrolled 55 patients of whom 15 were with BRCA mutations (103). Overall, an ORR of 21% (47% in patients with BRCA-mutated tumors) and a disease control rate (DCR) of 49% (80% in patients with BRCA mutated tumors) were reported. For the five patients harboring non-BRCA HRR pathway mutations, ORR was 20% (n = 1/5) and DCR was 80% (n = 4/5). In the overall population, ORR was higher in patients with PD-L1-positive TNBC (32%; n = 9/28) than in those with PD-L1-negative tumor (8%; n = 1/13). Despite the relatively small sample size (N = 47 for efficacy, N = 55 for safety), the combination of niraparib and pembrolizumab resulted to be active, regardless of BRCA mutation status, in patients with somatic or gBRCA-mutated and wild-type BRCA advanced/metastatic TNBC. Comparable results were obtained in the MEDIOLA trial where the combination of olaparib and durvalumab was associated with DCRs of 80 and 50% after 12 and 28 weeks, respectively, and a favorable tolerability in patients with gBRCA-mutated metastatic BC (101, 102).

PARPib are also being evaluated in combination with chemotherapeutic agents (91–94). In the phase 3 BROCADE3 trial (N = 509), addition of veliparib to carboplatin and paclitaxel resulted in a significant improvement in median PFS compared with placebo plus carboplatin and paclitaxel (14.5 vs. 12.6 months; HR: 0.71, 95% CI: 0.57–0.88; p = 0.002) in patients with gBRCA-mutated, HER2-negative, locally advanced or metastatic BC (105, 106). The PFS benefit was durable, and no additional toxicities were seen, although there was a high degree of toxicity in both treatment arms (105). However, veliparib appeared to be effective in terms of PFS benefit as monotherapy (HR: 0.49, 95% CI: 0.33–0.73) as well as in combination therapy (HR: 0.81, 95% CI: 0.62–1.06), regardless of the number of treatment cycles. Other ongoing phase II and phase III trials are reported in Table 4 .

Ongoing clinical trials are investigating PARPib in combination with new agents, including DDR molecules (ATR or Wee1 inhibitors). WEE1 is a kinase inhibitor which decreases kinases cyclin-dependent kinase1 (CDK1) expression, subsequently followed by activating replication firing and DSB repair (107). HR is scheduled but weakened by WEE1 inhibitor through phosphorylation of CDK1 in BRCA1/2-deficient tumor cells (108, 109). The combination of PARP and WEE1 inhibitors arrests G2 phase and results in chromosomal aberration and replication stress, which is proven to have an antitumor activity in numerous preclinical models (110). VIOLETTE is a global, multicenter, open-label, phase II study randomizing 1:1:1 450 patients with advanced TNBC to olaparib alone or in combination with AZD1775 (a WEE1 checkpoint inhibitor) or AZD6738 (an ataxia telangiectasia and Rad3-related protein inhibitor). Patients will be stratified in BRCA-mutated, non-BRCA HRR-mutated, and non-HRR mutated. The primary endpoint is PFS (95).

A two-part, open-label, non-randomized, phase 2, ongoing trial is testing the combination of ZEN003694 (a bromodomain inhibitor) with talazoparib in patients with TNBC without BRCA 1/2 germline mutations. The part 1 of this trial is a dose escalation study, with primary outcome incidence of treatment-related adverse events and treatment-related serious adverse events. The part 2 is a Simon 2-stage design, with primary outcome ORR (97).

The other group of agents that are interesting are the AKT inhibitors: previous research has shown that PI3K inhibitors (PI3Kib) lower nucleotide pools required for DNA synthesis and S-phase progression. Additionally, inhibition of PI3K/mTOR could inhibit PI3K interaction with the homologous recombination complex, increasing the dependency on PARP enzymes for DNA repair (111). Based on this data, the combination of PI3Kib and PARPib could potentially lead to a new, chemotherapy-free treatment option for BRCA wild-type TNBC as well as to improve the modest PFS/OS seen with the PARPib as single agents in BRCA1/2 mutant advanced setting. At the ASCO 2020, two randomized phase 2 studies, LOTUS and PAKT, reported the role of AKT inhibitors in combination with taxanes. Both trials demonstrated some improvement in PFS, with hints toward improvement in OS, in advanced TNBC. The results also showed some suggestions that PTEN loss or a PI3K-altered pathway could be a biomarker to predict who is going to benefit the most AKT inhibitors. Thus, a dual mTOR/PI3K inhibitor (gedatolisib) for metastatic or recurrent/unresectable TNBC could be a promising strategy in combination with talazoparib (98). Finally, the combination of olaparib plus trastuzumab for HER2-positive BC (OPHELIA trial) and a phase I trial with veliparib plus lapatinib are also under evaluation (99, 100). A complete list of clinical trials evaluating the combination of PARPib with other targeted therapies is summarized in Table 5 .

Along with BRCA1 and BRCA2, multiple HRR genes, including ATM, BARD1, BRIP1, CHEK2 (encodes CHK2), MRE11A, PALB2, RAD50, RAD51C, and RAD51D, are also implicated in hereditary cancer risk and are recently considered new potential biomarkers in patients with non-gBRCA HRR gene mutations (112).

Clinical studies that showed positive findings for PARPib in settings other than gBRCA-mutated BC include single-arm phase 2 studies of olaparib (113), rucaparib (114), and talazoparib (75) monotherapy ( Table 6 ).

In the olaparib expanded study, in 54 patients with metastatic BC and germline mutations in various non-BRCA DDR genes (cohort 1) or somatic mutations in DDR genes including BRCA (cohort 2), ORR was 33 and 31%, respectively (115). Overall, antitumor activity was reported in patients with somatic BRCA or gPALB2 mutations, but not in those with ATM or CHEK2 mutations. At the ASCO 2020 symposium, a study investigating the role of olaparib in women with HER2-negative breast cancer and a germline alteration in DDR pathway, such as PALB2, CHEK2, and ATM, or a somatic tumor mutation without a germline BRCA1/2 mutation was presented (115). Of the two cohorts, the first included patients with germline mutations other than BRCA. Olaparib demonstrated a high response rate, and the trial met its primary endpoint. Specifically, patients with a germline PALB2 mutation had 80% ORR, whereas in the somatic mutation cohort, patients with a somatic BRCA1/2 mutation reported 50% ORR.

In the RUBY trial, rucaparib monotherapy was investigated in 41 patients with HRD, including four patients harboring somatic BRCA mutations. Five patients (13.5%) demonstrated clinical benefit, comprising three patients with high loss of heterozygosity, one with a somatic BRCA1 mutation, and another patient with a somatic BRCA2 mutation (116).

In the phase 2 study of single-agent talazoparib, patients with BRCA wild-type, HER2-negative, advanced BC and non-BRCA HRR pathway mutations were enrolled. Based on 12 evaluable patients, the ORR and the clinical benefit rate were 25 and 50% after 6 months of treatment, respectively (117). In detail, two-thirds of the responders had gPALB2 mutations; the others had gCHEK2, gFANCA, and somatic PTEN mutations.

According to the above-mentioned reported data, PARPib demonstrated to have a role beyond BRCA2 germline mutation carriers, although the responses seem to be gene specific: the ATM and CHEK2 cohorts seemed not to respond, but the sample size was small.