Development of Olaparib for BRCA-Deficient Recurrent Epithelial Ovarian Cancer

Ovarian cancer is the most lethal gynecologic cancer with 21,290 new cases and 14,180 deaths anticipated for 2015 in the United States (1). The lack of validated screening tools in the general population together with an absence of specific symptoms indicative of early disease accounts for the poor 5-year survival (1). The most common histologic subtype of epithelial ovarian (and fallopian tube) cancer is high-grade serous, which is characterized by genetic instability and almost universal p53 dysfunction (1).

Synthetic lethality occurs when a genetic defect or defective protein is compatible with cell viability, but is lethal when combined (i.e., synthesized) with another genetic/protein defect (1, 2). The most robust demonstration of the principle of harnessing synthetic lethality comes from the treatment of cancers resulting from loss of BRCA gene function. The discovery of BRCA1 and BRCA2 (BRCA1/2) is among the most important discoveries in human cancer genetics. The BRCA genes encode proteins involved in error-free repair of DNA double-strand breaks.

PARP1 is necessary for repair of DNA single-strand breaks (Fig. 1). PARP1 binds DNA and synthesizes PARP chains through PARylation (1,2). Posttranslational modification of substrate proteins leads to recruitment of DNA repair effectors such as XRCC1 to the site of the single-strand break. PARP1 is released from the damaged DNA through auto-PARylation (1,2).

Clinical trials of interest appear in Tables 1 (refs. 3–15) and 2 (phase II and randomized phase II; refs. 16–24). In the initial accelerated dose-escalation phase I study, Fong and colleagues (3) reported that the MTD of the oral PARP inhibitor olaparib was 400 mg two times a day (Table 1). Toxicities (myelosuppression and central nervous system effects) were mild and self-limited. Included in the 60-patient study were 19 BRCA1/2-deficient patients with breast, ovarian, or prostate cancer, for whom the observed objective response rate (ORR) was 47% and the disease control rate was 63% (3). A secondary analysis involving a planned expansion of this study enriched with BRCA-mutated cases only suggested that the most favorable responses to olaparib were among platinum-sensitive patients (ORR 69%) as compared with those who were platinum resistant (ORR 23%; ref. 4). The relationship between prior platinum response and sensitivity to PARP inhibition is explored further below and may result from the ability of both PARP inhibitors and platinum salts to stall replication forks and cause DNA damage that otherwise would be repaired through BRCA-mediated homologous recombination in the wild-type setting (2).

Phase II trials in both refractory breast and ovarian cancer studied olaparib at both the previously identified MTD dose of 400 mg two times a day, and a lower dose of 100 mg two times a day (Table 2; refs. 16, 17). Audeh and colleagues (17) reported the phase II study in women with BRCA-deficient ovarian cancer, for whom the median number of prior regimens was three. Two sequential cohorts were enrolled comprising 33 patients treated with the MTD, followed by 24 who received the lower dose. The higher dose was associated with a higher ORR (33% vs. 13%) and two cases (6%) of grade 3 and 4 nausea, one case (3%) of grade 3 and 4 fatigue, and one case (3%) of grade 3 and 4 anemia (17). These early studies served as a proof of concept for the clinical application of synthetic lethality.

Because of the molecular, histopathologic, and clinical similarities (i.e., BRCAness) shared between both high-grade serous ovarian cancer and triple-negative breast and germline BRCA-deficient cancers, these two subtypes were selected for further study with PARP inhibition (Table 2). Gelmon and colleagues (18) performed a phase II, nonrandomized study of olaparib (400 mg two times a day) in BRCA-positive and -negative patients with advanced high-grade serous and/or undifferentiated ovarian cancer (n = 65) and triple-negative breast cancer (n = 26). Among those with ovarian cancer, confirmed objective responses were seen in 7 [41%; 95% confidence interval (CI), 22–64] of 17 patients with BRCA1/2 mutations and 11 (24%; 95% CI, 14–38) of 46 without mutations (18). Responses correlated with prior platinum sensitivity. No confirmed objective responses were observed in the breast cancer cohort, although tumor-reduction effects and disease stabilization was higher in the mutants. Fatigue and nausea occurred in 50% or greater in both subpopulations (18).

Kaufman and colleagues (19) studied olaparib (400 mg two times a day) in a spectrum of BRCA1/2-mutated cancers in Study 42. Enrollment was restricted to germline BRCA1/2-positive patients with (i) platinum-resistant recurrent ovarian cancer, (ii) metastatic breast cancer following three or more chemotherapy regimens, (iii) pancreatic cancer with prior gemcitabine treatment, or (iv) prostate cancer with progression on hormonal therapy and one systemic therapy. In the entire cohort of 298 evaluable patients, the ORR was 26.2% overall (19). The highest responses were reported in patients with prostate (50.0%, n = 4 of 8) and ovarian cancer (31.1%, n = 60 of 193; 95% CI, 24.6–38.1; ref. 19). Only 8 (12.9%; 95% CI, 5.7–23.9) of the 62 women with breast cancer experienced objective tumor response by RECIST criteria. Grade 3 or higher adverse events (AE) were reported for 54% of patients with anemia (17%) being most common (19).

The first randomized phase II study of olaparib was reported by Kaye and colleagues (20), and included nearly 100 women with BRCA1/2-positive, recurrent ovarian cancer who were randomized to two different dosages of olaparib or to two different dosages of pegylated liposomal doxorubicin (PLD; Table 2). Equivalence of all three arms for progression-free survival (PFS) and ORR was reported, with the activity and tolerability of olaparib consistent with previous studies.

Ledermann and colleagues (21) performed a randomized, double-blind, placebo-controlled, phase II study using olaparib as maintenance therapy (400 mg two times a day) among patients with platinum-sensitive, relapsed, high-grade serous ovarian cancer who had received two or more platinum-based regimens and had a partial or complete response to their most recent platinum-based regimen. This trial is known as Study 19 and its primary endpoint, PFS, was significantly longer with olaparib than with placebo (median 8.4 vs. 4.8 months; HR for progression or death, 0.35; 95% CI, 0.25–0.49; P < 0.001; Table 2; ref. 21). The majority of AEs were grade 1 and 2 and those occurring by more than 10% in the olaparib group compared with placebo included, nausea, fatigue, vomiting, and anemia.

Of those patients in the trial with known BRCA status, 74 (56%) in the olaparib group and 62 (50%) in the placebo group had a deleterious or suspected germline BRCA mutation (22). In a protocol-specified retrospective analysis stratified by mutational status, Ledermann and colleagues (22) reported that among patients with BRCA-deficient cancers, those treated with olaparib had a significantly longer median PFS as compared with the placebo group (11.2 vs. 4.3 months; HR, 0.18; P < 0.0001). Thus far, two interim analyses suggest that overall survival (OS) is not significantly different between the olaparib and placebo cohorts, whether stratified by mutation status or not.

Clinical development of olaparib has included multiple attempts to combine the PARP inhibitor with chemotherapy. In the phase I trial reported by Samol and colleagues (Table 1; ref. 7), the combination of olaparib plus topotecan was associated with significant dose-limiting hematologic AEs resulting in a subtherapeutic MTD, essentially halting further development of this combination. The combination of olaparib with cisplatin plus gemcitabine was piloted by Rajan and colleagues (8) who reported significant myelosuppression even at relatively low dosages (Table 1). Similarly, an MTD could not be determined for the combination of olaparib plus weekly paclitaxel, which was complicated by significant clinical interaction and greater than expected rates of neutropenia despite secondary prophylaxis (9). In phase I studies evaluating continuous and intermittent dosing of olaparib with cisplatin (11) or with gemcitabine (12), only chemotherapy combined with intermittent dosing of olaparib at 50 mg (days 1–5) or 100 mg (days 1–14) was tolerable.

At the recent 2015 Annual Meeting of the American Society of Clinical Oncology, Chiou and colleagues (13) reported results from their phase I/Ib trial of olaparib with escalating dosages of carboplatin in heavily pretreated patients with high-grade serous ovarian cancer with low genetic risk. The median number of prior therapies was six. Among evaluable subjects, platinum-sensitive patients (n = 11) experienced a higher ORR than those with platinum-resistant disease (n = 14; 36% vs. 7%).

In an open-label, phase II study (Study 41), Oza and colleagues (23) randomized 162 women with platinum-sensitive high-grade serous ovarian cancer who had remained progression free for a minimum of 6 months before study entry to carboplatin (AUC 4) plus paclitaxel (175 mg/m²) with and without olaparib, 200 mg two times a day d1–10, followed by 400 mg two times a day monotherapy continuously. The arm administering olaparib was associated with significantly improved PFS (Table 2), with the greatest benefit conferred to those with known BRCA mutations (n = 20; HR, 0.21; 95% CI, 0.08–0.55; P = 00015; ref. 23). Serious AEs were reported in 15% of patients in the olaparib-plus-chemotherapy arm as compared with 21% in the chemotherapy-alone group. Interestingly, when comparing the hazard of progression for the platinum-sensitive BRCA-mutated patients in this study (23) with that of the platinum-sensitive BRCA mutation–positive patients in Study 19 (HR, 0.21 vs. 0.18, respectively; ref. 21), it would appear that there is no additional benefit to adding olaparib to chemotherapy.

Phase I studies of olaparib (400 mg two times a day) combined with drugs that target the VEGF axis have been informative, with no dose-limiting toxicities (DLT) reported in the olaparib plus bevacizumab trial (Table 1; ref. 14). Provocative data from the phase I combining olaparib with cediranib identified the MTD (Table 1; ref. 15) and led directly to a randomized phase II study of 46 women with either measureable platinum-sensitive, recurrent, high-grade serous, or endometrioid ovarian carcinoma or those with deleterious germline BRCA1/2 mutations (Table 2; ref. 24). Compared with olaparib monotherapy, the combination was associated with significantly improved PFS (17.7 vs. 9.0 months; HR, 0.42; 95% CI, 0.23–0.76; P = 0.005; ref. 24). Grade 3 and 4 fatigue, diarrhea, and hypertension were more common with combination therapy (24). The olaparib-plus-cediranib combination may constitute a chemotherapy-free alternative for select patients with recurrent disease.

During February 2014, AstraZeneca filed a U.S. regulatory submission for olaparib as a maintenance therapy in platinum-sensitive recurrent disease based on Study 19 (21). On June 25, 2014, the Oncology Drugs Advisory Committee (ODAC) panel members voted 11 to 2 against regulatory approval of olaparib and in response to a request by the FDA for additional data, AstraZeneca submitted a major amendment to their New Drug Application on July 24, 2014, highlighting Study 42 (19).

These data prompted the FDA to extend the original October 3, 2014, Prescription Drug User Fee Act action date to January 3, 2015. On December 19, 2014, the FDA granted accelerated approval to olaparib as fourth-line therapy for women with BRCA-deficient (germline only) ovarian carcinoma. Approval was based on the analysis of 137 patients from Study 42 with measureable BRCA-deficient recurrent disease treated with a median of 3.4 prior lines of chemotherapy. The ORR for patients in this cohort with measureable disease was 34% (95% CI, 26%–42%) and the median duration of response was 7.9 months (95% CI, 5.6–9.6 months).

The phase III randomized, placebo-controlled olaparib monotherapy study, SOLO 1 (NCT01844986), will attempt to validate the findings of Study 19, while SOLO 2 (NCT01874353), the phase III randomized, placebo-controlled olaparib-without-chemotherapy trial in platinum-sensitive disease was designed on the heels of Study 41. Both SOLO 2 and its successor, SOLO 3 (NCT02282929), are integral to the phase IV commitment following accelerated approval. FDA approval was done in conjunction with regulatory approval of a companion diagnostic genetic test (BRCAnalysis CDx) that will screen serum from ovarian cancer patients for mutations in the BRCA genes (gBRCAm). Interestingly, one day before regulatory approval in the United States, AstraZeneca announced on December 18, 2014, that the European Commission granted marketing authorization for olaparib as a first-line maintenance therapy for adult patients with platinum-sensitive, relapsed BRCA-mutated (germline and/or somatic) high-grade serous epithelial ovarian, tubal, or peritoneal carcinoma.

The clinical implications of approving a drug for fourth-line therapy need to be framed in the context of toxicity assessment. This is particularly important in the setting of recurrent disease, where quality of life/disease control is the goal and not cure. Contemporary regimens for recurrent platinum-sensitive disease include carboplatin (or cisplatin) plus either paclitaxel or gemcitabine, and carboplatin plus PLD. Women with partially platinum-sensitive and platinum-resistant disease may receive bevacizumab plus either PLD, topotecan, or weekly paclitaxel or monotherapy using PLD, topotecan, weekly albumin-stabilized nanoparticle-formulation paclitaxel, trabectedin, or pemetrexed (the latter two not having a label in the United States). Olaparib as monotherapy may have a much more favorable safety profile compared with conventional chemotherapy used in the recurrent setting.

All patients with recurrent ovarian cancer will ultimately progress on olaparib. On the basis of the preclinical data, early essayists were concerned that treatment with PARP inhibitors would result in acquisition of secondary BRCA1/2 mutations that would limit the efficacy of subsequent chemotherapy, specifically, platinum-based agents. Ang and colleagues (25) collected data from 89 patients with BRCA-deficient ovarian cancer who had been previously treated with olaparib at daily dosages of 200 mg and higher. The ORR by RECIST to post-olaparib chemotherapy was 36% (n = 24 of 67 patients). For patients treated post-olaparib with platinum-based chemotherapy, the ORR was 40% (n = 19 of 48 patients; ref. 25). The corresponding ORRs when incorporating CA-125 were 45% and 49%, respectively (25). The length of the platinum-to-platinum interval with intervening olaparib was associated with an increased likelihood of response (25). Tumor was available from 6 cases and subjected to massively parallel sequencing, which demonstrated no evidence of secondary BRCA mutations (25). These data suggest that not only does PARP inhibitor-resistant ovarian cancer retain the potential to respond to chemotherapy, but that treatment with PARP inhibitors may enhance the response to subsequent platinum.

Using targeted capture and massively parallel genomic sequencing, Pennington and colleagues (26) recently reported that germline (24%) and somatic (9%) mutations were detected in one or more of 13 homologous recombination genes, including BRCA1, BRCA2, ATM, BARD1, BRIP1, CHEK1, CHEK2, FAM175A, MRE11A, NBN, PALB2, RAD51C, and RAD 51D. Homologous recombination deficiency was found in both serous and nonserous ovarian carcinomas, including clear cell, endometrioid, and ovarian carcinosarcoma. Both germline and somatic homologous recombination mutations were highly predictive of platinum sensitivity and improved OS, with OS durations of 66 months (germline), 59 months (somatic), and 41 months (no homologous recombination deficiency) noted (26). These data support interrogating ovarian tumors for genomic scars indicative of homologous recombination deficiency to assess candidacy for treatment with olaparib and other PARP inhibitors.

Moving forward, the prevalence of homologous recombination deficiency in tumors other than ovarian and breast requires assessment and predictive biomarkers need to be discovered. In addition, indications for combining olaparib with chemotherapy and additional chemotherapy-free olaparib-based regimens should be explored.

Regulatory approval of the first-in-class PARP inhibitor, olaparib, as a monotherapy for patients with BRCA-deficient or suspected BRCA-deficient recurrent ovarian cancer is accompanied by discrete consequences for drug development in the fourth-line space. At the very minimum, newer agents will need to deliver ORR in excess of 30% with the lower limit of a 95% CI settling in at 26% to 28%.

R.N. Eskander reports receiving speakers bureau honoraria from AstraZeneca. B.J. Monk reports receiving speakers bureau honoraria from AstraZeneca, and is a consultant/advisory board member for AstraZeneca and TESARO. No potential conflicts of interest were disclosed by the other author.

Conception and design: K.S. Tewari, R.N. Eskander, B.J. Monk

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): B.J. Monk

Writing, review, and/or revision of the manuscript: K.S. Tewari, R.N. Eskander, B.J. Monk

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): B.J. Monk

The authors thank Daniele A. Sumner, BA, for her assistance in editing the article.