PARP Inhibition in Epithelial Ovarian Cancer: High Hopes Undergo a Reality Check

ABSTRACT: The treatment strategy of poly(ADP-ribose) polymerase (PARP) inhibition capitalizes on the inherent defect in homologous recombination that occurs in BRCA-deficient tumors by inhibiting the alternative DNA repair pathway involving base excision repair. Although PARP inhibitors were initially considered a potential treatment specifically for tumors with germline BRCA mutations, evidence of frequent somatic deficiency in the BRCA pathway has caused reconsideration of that approach. PARP inhibitors have been shown to have activity in epithelial ovarian, fallopian tube, and primary peritoneal cancer in phase I and II clinical trials. Responses have been seen in both BRCA-deficient and sporadic tumors, and they do not appear to require platinum sensitivity. Although PARP inhibitors are well tolerated as monotherapy, additional study is required to determine their efficacy and toxicity in combination with chemotherapy and other targeted agents. Many hurdles remain along the pathway to drug registration, but the motivation of the community of ovarian cancer patients, researchers, and clinicians to find new treatments may speed the process.

The identification of BRCA1 and BRCA2 in the early 1990s as genes that, when mutated, are associated with an increased risk of breast, epithelial ovarian, fallopian tube, primary peritoneal, and other cancers opened a new frontier in managing hereditary cancer risk. The ability to delineate a high-risk subset of patients was of particular benefit to research into epithelial ovarian cancer (EOC), given its poor prognosis, frequent late stage at diagnosis, and relative lack of effective screening techniques. Further study has documented that, depending on the population evaluated, 5% to 15% of EOC cases are hereditary in origin. Mutations in BRCA1/2 or the mismatch repair mutations of Lynch syndrome account for almost all of that risk, although the great majority of EOC cases are sporadic in origin.

Groups at high risk for a disease provide ideal populations for research given their relatively high event rate, which makes studies shorter, smaller, and less expensive to conduct than when the general population is evaluated. These patients also tend to be highly motivated to participate in research, since the results may benefit not only their own treatment but also the care of their family members. In light of the desperate need for improvement in the treatment of EOC, development of poly(ADP-ribose) polymerase (PARP) inhibitors in the treatment of BRCA-deficient cancer has been closely monitored by patients, as well as by clinicians and researchers. The great hope is that lessons learned from the high-risk population will lead to insights into sporadic EOC.

This article will review the trials that have been conducted with PARP inhibitors in EOC, fallopian tube cancer (FTC), and primary peritoneal cancer (PPC). The impact of those results will then be placed in the larger context of PARP inhibitor development.

PARP Inhibitor Trials in Ovarian Cancer

The early history of PARP inhibition has been chronicled thoroughly, including in reviews by Comen et al[1] and Rios et al,[2] and will not be discussed in depth here. The strategy capitalizes on the inherent defect in homologous recombination that occurs in BRCA-deficient tumors by inhibiting the alternative DNA repair pathway involving base excision repair. Base excision repair relies on PARP. If that pathway is blocked via PARP inhibition, the loss of both repair mechanisms leads to accumulation of DNA breaks and, ultimately, cell death. The concept of synthetic lethality, originally described in 1946, has gained new life in the setting of these two DNA repair defects, which individually are not lethal but become so when combined.[3]

TABLE 1
Completed Clinical Trials of PARP Inhibitors in Epithelial Ovarian Cancer

Tables 1 and 2 include the trials listed on www.ClinicalTrials.gov that evaluate PARP inhibition in EOC and other gynecologic cancers. The list was synthesized by searching for “PARP inhibitors AND ovarian cancer” and conducting separate searches for each of the PARP inhibitors currently in development (AG014699 [PF-01367338], olaparib [KU59436, AZD2281], veliparib [ABT-888], iniparib [BSI-201], INO-1001, GP121016, CEP-9722, MK4827, and BMN-673). Of these agents, results for trials in EOC were found for AG014699, olaparib, veliparib, iniparib, and MK4827.

Although PARP inhibitors were initially considered a potential treatment specifically for tumors with germline BRCA mutations, evidence of frequent somatic deficiency in the BRCA pathway has led to reconsideration of that approach. In an analysis of 235 EOC cases, Hennessy et al noted that among 44 BRCA1/2 mutations detected, at least 43% of the BRCA1 mutations and 29% of the BRCA2 mutations were somatic in nature. (Not all patients had germline DNA available for testing.)[4] In addition, the Cancer Genome Atlas analysis of serous EOC recently documented that a combination of germline and somatic events led to mutations in BRCA1/2 in 22% of cases, but that the frequency of other mutations in EMSY, PTEN, RAD51C, ATM, ATR, and Fanconi anemia genes suggests that approximately 50% of serous EOC cases have disruption of the homologous recombination pathway and may be susceptible to PARP inhibitor therapy.[5] The shift towards evaluating PARP inhibitor activity in broader clinical populations than BRCA-deficient EOC, FTC, and PPC is illustrated in Tables 1 and 2; the completed trials shown in Table 1 all focused on populations with germline mutations, whereas 21 of 26 ongoing trials listed in Table 2 have a sporadic component.

Completed trials

TABLE 2

Ongoing Clinical Trials of PARP Inhibitors in Epithelial Ovarian Cancer From ClinicalTrials.gov

The trials included in Table 1 are listed as “completed” on www.ClinicalTrials.gov and/or have results published in manuscript format. Fong et al described the earliest experience with PARP inhibitor treatment in EOC in a phase I trial with dose escalation of olaparib from 10 mg daily to 600 mg twice daily in a population of patients with recurrent solid tumors.[6] The maximum tolerated dose was 400 mg twice daily. Interestingly, PARP inhibition as measured in peripheral mononuclear cells exceeded 90% after a dose of 60 mg twice daily, although clinical responses were not seen until the 100-mg twice-daily dose level. Patients with BRCA-deficient cancer (eight with EOC and one with breast cancer) accounted for all of the partial or complete radiologic responses seen. This observation led to an expansion cohort in the BRCA-deficient population.[7] A dose of 200 mg twice daily was chosen due to concerns about possible increased toxicity in mutation carriers. However, no increase in side effects was seen; olaparib was well tolerated, with the most common complaints being fatigue and gastrointestinal symptoms. The clinical benefit rate was 46% in the expansion cohort, with a median response duration of 28 weeks. A statistically significant increase in response rate was seen among platinum-sensitive vs platinum-resistant EOC patients, leading to questions about whether platinum sensitivity is a surrogate for predicting response to PARP inhibition and whether patients with platinum-resistant tumors should be excluded in future studies.

A subsequent phase II trial of olaparib was performed as an international collaboration known as ICEBERG (International Collaborative Expertise for BRCA Education and Research through Genetics) 2.[8] The study population included women with recurrent, BRCA-deficient EOC, FTC, or PPC. Two dose cohorts were included: a 100-mg twice-daily group (the dose at which responses were first seen in the phase I trial) and a 400-mg twice-daily group (the maximum tolerated dose in the phase I trial). The primary objective of overall response rate varied from 13% in the low-dose group to 33% in the high-dose group. As in the phase I trial, olaparib was well tolerated, with fatigue, nausea, and anemia as the most common toxicities.

Another phase II trial with olaparib was performed in Canada and included women with recurrent, high-grade serous or undifferentiated sporadic EOC; BRCA-deficient EOC, FTC, or PPC; or triple-negative breast cancer.[9] Among 91 patients enrolled, 64 with EOC were treated. As expected, most of these patients had sporadic disease (47 of 64); among the 17 mutation carriers, 65% had BRCA1 mutations, while 29% had BRCA2 mutations and 1 patient had both. The primary endpoint of objective response rate was 41% in mutation carriers and 24% in sporadic EOC patients, a marked difference from the dose-escalation portion of the phase I trial, where no responses were seen in non–mutation carriers. Interestingly, no responses were seen in 26 breast cancer patients, even among the 10 patients with germline mutations.

The only other completed trial of a PARP inhibitor including EOC patients is a phase I study combining veliparib with temozolomide(Drug information on temozolomide) (Temodar). The combination was of interest due to preclinical evidence of synergy.[10,11] No final results for the phase I study have been reported to date, although a follow-up phase II trial has completed accrual.

Ongoing trials

Table 2 lists the 26 trials that were identified in ClinicalTrials.gov as ongoing, with 14 in phase I, 2 in phase I/II, and 10 in phase II. No phase III trials are in progress currently. Nine trials are active but have completed accrual; two others either have been suspended or terminated but were included in this table because final results have not been published in manuscript format. Preliminary results presented as abstracts appear in the last column of Table 2 and are summarized below.

Monotherapy trials of olaparib, veliparib, iniparib, MK4827, and AG014699 are included, of which only the veliparib study continues to enroll patients. Preliminary results from a phase I study of MK4827 show that 10 of 12 partial responses seen were in EOC patients, with a mix of BRCA-deficient (7) and sporadic (3) cases.[12] In addition, four of eight patients with stable disease had EOC, evenly split between BRCA-deficient and sporadic tumors. Another monotherapy trial of AG014699 in a phase II setting included women with BRCA-deficient EOC and/or breast cancer.[13] Of 41 patients enrolled, 24 had EOC. The overall response rate for the entire cohort was quite low at 5%, but stable disease contributed to a clinical benefit rate of 32%.

Finally, two randomized phase II trials incorporating olaparib monotherapy have been reported. In the first, women with recurrent, BRCA-deficient EOC were randomized between olaparib at 200 mg twice daily, olaparib at 400 mg twice daily, and pegylated liposomal doxorubicin(Drug information on doxorubicin) (PLD).[14] Initial results show median progression-free survival times of 6.5 months, 8.8 months, and 7.1 months, respectively, for the three treatment groups. The highest rate of confirmed response was in the high-dose olaparib group (31%). In the second randomized phase II trial, olaparib at 400 mg twice daily was compared with placebo in a cohort of women with recurrent, platinum-sensitive, serous EOC as maintenance therapy after partial or complete response to platinum therapy.[15] Preliminary findings include an improvement in progression-free survival from 4.8 to 8.4 months, which was statistically significant. The data were not mature enough to allow an analysis of overall survival.

It is evident from the remaining trials listed that combinations of a PARP inhibitor with traditional chemotherapy and/or other targeted agents are of interest. Chemotherapy agents under study in combination with various PARP inhibitors include carboplatin(Drug information on carboplatin), paclitaxel(Drug information on paclitaxel), irinotecan(Drug information on irinotecan), carboplatin/pac-litaxel, carboplatin/gemcitabine, PLD, and topotecan(Drug information on topotecan). The fundamental question of drug sequence is being investigated in a phase I study of olaparib and carboplatin.[16] Preclinical evidence suggests that administering olaparib before carboplatin diminishes DNA damage, possibly because PARP inhibition preceding cytotoxic chemotherapy enhances alternative DNA repair pathways and recovery from chemotherapy.[17] Preliminary results from another phase I study of olaparib and carboplatin with maintenance olaparib until progression include a recommended phase II dose of olaparib 400 mg twice daily and carboplatin AUC5, as well as an impressive 83% clinical benefit rate in patients with recurrent EOC.[18]

A phase I study of veliparib and irinotecan in recurrent malignancies is evaluating escalating doses of veliparib given twice daily for 15 days out of a 21-day cycle while irinotecan is held at a fixed dose of 100 mg/m² on days 1 and 8.[19] Preliminary results include a clinical benefit rate of 61% in 32 patients, 7 of whom have EOC. Dose-limiting toxicities include fatigue, diarrhea, febrile neutropenia, and bone-marrow suppression. The recommended phase II dose is 40 mg of veliparib twice daily for 15 days on/6 days off, with irinotecan at a dose of 100 mg/m² intravenously on days 1 and 8.

Single-arm phase II trials of iniparib with gemcitabine(Drug information on gemcitabine) and carboplatin in platinum-sensitive and platinum-resistant recurrent EOC yielded preliminary results recently.[20,21] In the platinum-sensitive group, 12 of 17 women had a confirmed response, while 6 of 19 had a confirmed response in the platinum-resistant group.

PARP inhibitors are also being studied in combination with the antiangiogenic agents bevacizumab(Drug information on bevacizumab) (Avastin) and cediranib (AZD2171). Veliparib is being evaluated along with carboplatin, paclitaxel, and bevacizumab in the Gynecologic Oncology Group phase I trial 9923. This is the only trial of frontline treatment of EOC, FTC, or PPC now in progress. Cediranib is being evaluated with olaparib in a phase I/II trial enrolling patients with recurrent EOC, FTC, or PPC; preliminary results from the phase I component include a 56% unconfirmed response rate, without unexpected toxicity.[22]