A Phase II Study of Talazoparib after Platinum or Cytotoxic Nonplatinum Regimens in Patients with Advanced Breast Cancer and Germline BRCA1/2 Mutations (ABRAZO)

Breast cancer susceptibility (BRCA) genes 1 and 2 [BRCA1 and BRCA2 (BRCA1/2)] are key components in the repair pathway for DNA double-strand breaks. Mutations in these genes account for 20%–25% of hereditary breast cancers and approximately 5% of all breast cancers (1).

With improved understanding of the functions of BRCA1/2 in DNA repair, the focus of developmental therapeutics has shifted in recent years from cytotoxic chemotherapy to molecularly targeted agents. Cancer cells with a BRCA1/2 mutation are deficient in homologous recombination DNA repair, and PARP inhibition causes synthetic lethality in these cells. This is at least, in part, because PARP inhibition traps PARP on sites of DNA damage, creating DNA damage that cannot be repaired in cancer cells with defective homologous recombination DNA repair mechanisms (2–4). This vulnerability in BRCA-mutant tumors has led to the clinical development of PARP inhibitors for a variety of cancers (2–9). Talazoparib is an oral PARP inhibitor that has recently emerged as a promising anticancer therapy (5). In nonclinical models, talazoparib has been shown to be the most potent PARP inhibitor in development, with the lowest concentrations required for inhibition of PARP enzymatic activity and PARP trapping at DNA breaks (9–11).

In the first-in-human phase I study, the MTD of talazoparib was defined as 1 mg once daily; the dose-limiting toxicity was reversible thrombocytopenia. Talazoparib treatment resulted in a 50% response rate and an 86% clinical benefit rate at 24 weeks in 18 patients with a germline BRCA1/2 mutation and advanced breast cancer (5). Treatment-emergent adverse events (AE) in the phase I trial were hematologic and gastrointestinal toxicity.

The purpose of this phase II study (ABRAZO, NCT02034916) was to evaluate the efficacy and safety of talazoparib as a single agent in patients with advanced breast cancer with a germline BRCA1/2 mutation. Cohort 1 included patients with platinum-sensitive disease to explore the impact of previous platinum therapy on the effectiveness of subsequent treatment with a PARP inhibitor, as it had been previously demonstrated that patients with ovarian cancer treated with other PARP inhibitors who had platinum-resistant disease (progression within 6 months after the last dose of platinum) or platinum-refractory disease (progression on or within 2 months following last dose of platinum) showed a lower observed response rate than patients with longer platinum-free intervals (6, 7). Cohort 2 was designed to investigate the response rate in a heavily pretreated population who had not received previous platinum therapy.

This open-label, two-cohort, phase II study evaluated talazoparib in patients with metastatic breast cancer with a deleterious or a suspected deleterious germline BRCA1/2 mutation using the BRACAnalysis CDx assay (Myriad Genetics). We enrolled patients from 33 locations throughout France, Germany, Spain, the United Kingdom, and the United States. Initially, a total of 35 patients were enrolled in stage I into each of two cohorts. If 5 or more patients had objective responses as determined by the local investigator in either cohort, an additional 35 patients were to be enrolled in stage II into that cohort, for a total of up to 140 patients in the study. Final assessment of objective response rate (ORR; all patients) was performed by an independent central review.

Cohort 1 included patients with a complete response or partial response to a previous platinum-containing regimen for metastatic disease, with no disease progression within 8 weeks of the last dose of platinum therapy. For patients treated with more than one previous platinum-containing regimen, platinum response and stability of disease from the last dose had to occur with the most recent regimen. Seven patients in cohort 1 were enrolled in violation of the eligibility criteria because of disease progression within 8 weeks of the last dose of platinum. These patients are included in all analyses.

Cohort 2 included patients who had received three or more previous cytotoxic chemotherapy regimens for metastatic disease and no previous platinum therapy for metastatic disease. Previous adjuvant or neoadjuvant therapy with platinum was allowed if the first disease recurrence was more than 6 months after the last dose of adjuvant or neoadjuvant platinum therapy. Patients with HER2-positive disease were eligible for either cohort, provided they were considered to be refractory to HER2-targeted therapy. Patients with HER2-positive breast cancer had to have discontinued HER2-directed therapies prior to day 1, although there might have been some overlap due to washout. Furthermore, HER2-directed combination therapy was not permitted on study.

In addition to the key eligibility standards described for cohorts 1 and 2 above, additional criteria included measurable disease as defined by Response Evaluation Criteria In Solid Tumors version 1.1 (RECIST v 1.1), an Eastern Cooperative Oncology Group (ECOG) performance status of 0 or 1, and adequate organ and bone marrow function. A full list of exclusion criteria is provided in the Supplementary data (Supplementary Materials and Methods). The study was conducted in accordance with the protocol, good clinical practice standards, the Declaration of Helsinki, and the International Conference on Harmonisation. The appropriate Institutional Review Board or ethics committee at each participating institution approved the protocol. All enrolled patients provided written informed consent before undergoing study-specific procedures.

Patients received talazoparib 1 mg once daily by mouth as continuous therapy; patients were evaluated in repeated 21-day cycles. Talazoparib dosing could be interrupted for recovery from toxicity for up to 28 days. For interruptions longer than 28 days, treatment could be resumed at the same or a reduced dose at the discretion of the sponsor and investigator if clinical benefit was evident. Patients had study visits on days 1, 8, and 15 for the first two cycles; the day 8 and 15 visits were optional for subsequent cycles in the absence of significant toxicities. Treatment with talazoparib continued until radiographic disease progression, unacceptable toxicity, withdrawal of consent, or the investigator's decision to discontinue treatment. End-of-treatment assessments were performed 30 days after the last dose of the study drug or before initiation of a new antineoplastic therapy, whichever occurred first.

An independent radiology facility evaluated the imaging data and prior and on-study radiotherapy. Tumor assessments were performed by CT, MRI, or radiography. Tumor assessments were performed at screening or baseline, every 6 weeks for the first 24 weeks, and every 12 weeks thereafter until radiographic disease progression was confirmed by the independent radiology facility or until initiation of a new antineoplastic therapy. Tumor response could also be assessed as clinically indicated, at the time of clinical suspicion of disease progression, and to confirm complete or partial response at least 4 weeks after an initial response was observed.

All patients were monitored for survival and subsequent anticancer therapy every 60 days after the last dose of study drug for the first year, every 90 days thereafter, and at the sponsor's request. Survival follow-up continued until death or withdrawal of consent.

Safety and tolerability were evaluated by assessment of AEs, physical examinations, vital signs, laboratory evaluations, electrocardiograms, and use of concomitant medications during the treatment period through 30 days after the last dose of study drug or before initiation of a new antineoplastic therapy or investigational therapy, whichever occurred first.

Sparse pharmacokinetic sampling was performed on day 1 of cycles 1 through 4, consisting of a predose sample collected no more than 60 minutes prior to dosing and two postdose samples collected at least 30 minutes after dosing. The collection times of the two postdose samples were separated by at least 2 hours. Blood and tumor tissue samples were collected for pharmacodynamic and genomic research. These analyses will be reported separately.

The intention-to-treat analysis population was defined as all patients with an enrollment date (n = 84). The tumor-evaluable population was defined as all patients who had received at least one dose of talazoparib, had a baseline tumor assessment, and had at least one postbaseline tumor assessment (n = 83). This population was used for the efficacy analysis, with the exception of progression-free survival and overall survival, which used the intention-to-treat population.

The primary efficacy endpoint was ORR, defined as the proportion of patients in the tumor-evaluable population who had a confirmed objective response (best overall response of complete or partial response) confirmed by the independent radiology facility using RECIST v 1.1 at the time of data cutoff. Secondary efficacy endpoints were clinical benefit rate at 24 weeks and duration of response confirmed by the independent radiology facility, progression-free survival (PFS) confirmed by the investigator, and overall survival. Exploratory endpoints were ORR, clinical benefit rate at 24 weeks, and duration of response assessed by the investigator.

This study was designed to provide adequate power for the primary efficacy endpoint of ORR. We initially planned for a total enrollment of up to 140 patients, depending on the number of patients with an objective response in the first stage of each cohort using the two-stage design. If five or more responses were observed for 35 patients in the first stage of each cohort, enrollment of an additional 35 patients into each cohort was to occur in the second stage, for a total of 70 treated patients per cohort. Talazoparib would be considered effective as a single agent if at least 16 patients in a cohort had an objective response. This design provided 0.90 power to distinguish between an active drug with a 30% true response rate and a drug with a response rate of 15% or less with an alpha level of 0.05.

After the trial met the criteria for continuation to the second stage for each of the two cohorts, the sponsor stopped the enrollment in this study, after an amendment to the phase III EMBRACA trial (NCT01945775) of talazoparib versus physician's choice of chemotherapy in patients with advanced breast cancer and a germline BRCA1/2 mutation resulted in overlapping enrollment criteria with this phase II study.

Between May 2014 and February 2016, we enrolled 84 patients in the intention-to-treat population. Patient disposition for the intention-to-treat population is shown in Fig. 1. One patient in cohort 1 did not receive any study drug because of rapidly worsening liver function tests and was excluded, for a total of 83 patients in the tumor-evaluable population, including 48 patients in cohort 1 and 35 patients in cohort 2. All 83 patients were included in the safety population.

Baseline characteristics of patients are given in Table 1. BRCA1/2 status was assessed by a central laboratory for 79 patients; patients had to carry at least one risk factor for hereditary breast cancer to be screened by the central laboratory for this trial. BRCA1/2 status was determined by local assessment for 4 of the remaining 5 patients whose samples were not available for central assessment, and BRCA1/2 status was unknown for 1 patient (the local BRCA1/2 mutation report was not entered into the database prior to database lock. There was an insufficient amount of material to obtain a result so the patient was listed as unknown. However, the sample was then retested by a central laboratory and appeared to be a BRCA mutant). No patient carried mutations in both BRCA1 and BRCA2 genes.

The cut-off date for all efficacy analyses was September 1, 2016 (9 patients continued on treatment at this time), with the exception of overall survival for which the data cutoff was April 7, 2017. Talazoparib demonstrated efficacy in patients with a BRCA1/2 mutation and a response to previous platinum therapy (cohort 1), and in patients who had at least three previous chemotherapy regimens and no previous platinum therapy (cohort 2; Table 2). In an examination of subgroups in this study, the ORR [(95% confidence interval (CI)] in patients carrying a BRCA1 or BRCA2 mutation was 23% (11–39) and 33% (20–50; Fig. 2A and B), and the ORR (95% CI) in patients with triple-negative breast cancer (TNBC) and hormone receptor–positive breast cancer was 26% (13–43) and 29% (17–44; Fig. 2C and D). The study recruited 6 patients with HER2-positive breast cancer, who were considered refractory to previous HER2-targeting therapies. All HER2-positive breast cancers were also hormone receptor–positive. In these patients, an objective response was confirmed in 2 of 6 patients by the independent radiology facility.

Investigator-assessed median PFS was 4.0 months (95% CI: 2.8–5.4) in cohort 1 and 5.6 months (95% CI: 5.5–7.8) in cohort 2 (Supplementary Fig. S1). Based on reverse Kaplan–Meier estimates, the median follow-up time was 13.7 months for each cohort. The median overall survival was 12.7 months (95% CI: 9.6–15.8) in cohort 1 and 14.7 months (95% CI: 11.0–24.4) in cohort 2 (Supplementary Fig. S1), with 36 events (74%) in cohort 1 and 22 events (62.9%) in cohort 2.

The median time from the last platinum dose to progression in cohort 1 was 4.0 months (range, 0.03–49.15). Exploratory subgroup analyses of the effect of platinum-free interval on ORR and the median PFS suggested that a longer platinum-free interval following the last dose of platinum was associated with greater clinical activity (Fig. 3). Of the 7 patients who entered cohort 1 with disease progression within 8 weeks of the last dose of platinum, none responded to talazoparib.

Drug-related AEs that led to study drug discontinuation occurred in 3 patients overall, 2 patients in cohort 1, and 1 patient in cohort 2, and included anemia (1 patient) and liver function test abnormalities (2 patients). The most common reason for dose reduction was myelosuppression (anemia); 23 patients (28%) received at least one transfusion of red blood cells. The transfusion rate was higher in cohort 2 (37%), the group who had received more prior cytotoxic chemotherapy; 21% of patients in cohort 1 received at least one red blood cell transfusion. Few patients received an erythropoietin-stimulating agent (8% and 3% for cohorts 1 and 2, respectively). Time to first dose reduction was 10.9 weeks in cohort 1 and 13.3 weeks in cohort 2.

Hematologic treatment-emergent AEs and grade 3 or higher hematologic treatment-emergent AEs occurring in 15% or more and in 5% or more of patients, respectively, are shown in Table 3. In both cohorts, anemia was the most common hematologic AE (50% and 57% of patients in cohorts 1 and 2, respectively). Neutropenic sepsis occurred in 1 patient. No acute myeloid leukemia or myelodysplastic syndrome was observed. Nonhematologic treatment-emergent AEs and grade 3 or higher nonhematologic treatment-emergent AEs occurring in 20% or more and 5% or more of patients, respectively, are shown in Table 3. No grade 5 treatment-emergent AEs were observed. No clinically significant cardiovascular toxicity was observed.

The plasma concentration of talazoparib at approximately 2 hours following the first dose was 3,650 pg/mL [coefficient of variation (CV%) 93.6]. On cycle 2 day 1 (approximately 21 days later), the plasma concentration was 10,700 pg/mL (CV% 46.2) in patients without a dose modification. Talazoparib plasma trough concentrations were similar for cycles 2, 3, and 4 for patients without any dose modification (between 3,300 and 3,900 pg/mL), suggesting that talazoparib reached steady state by cycle 2. In general, trough concentrations had high interpatient variability, between 46% and 60% for patients without dose modification, which was consistent with previous observations (5). Talazoparib showed accumulation at steady state where postdose concentrations (two postdose samples, >2 hours between) in cycles 2 to 4 were 2.7- to 3-fold higher than observed in cycle 1, consistent with historical data.

In this open-label, phase II study of patients with metastatic breast cancer carrying a germline BRCA1/2 mutation, talazoparib had clinical activity both in patients who had a previous response to platinum chemotherapy and in those who received at least three previous cytotoxic regimens for advanced breast cancer without prior platinum exposure. In the group who previously received platinum, an exploratory analysis demonstrated an association between higher ORR and longer median PFS with a longer platinum-free interval.

Platinum chemotherapy and PARP inhibitors target defective homologous recombination DNA repair in BRCA1/2-mutant breast cancer. Although the mechanisms of resistance to platinum chemotherapy in this setting are incompletely understood, reversion mutations that restore DNA repair proficiency appear to play a role in at least some patients. Such mutations have been identified in preclinical models and in the circulating tumor DNA of heavily pretreated breast cancers (12, 13). In an exploratory analysis, response to talazoparib was less likely in patients progressing with a short platinum-free interval than in patients with a long platinum-free interval (Fig. 3), suggesting that patients who progress on platinum chemotherapy or shortly after stopping may show reduced sensitivity to PARP inhibitors. Previous studies of PARP inhibitors in breast cancer have included relatively few patients with prior platinum chemotherapy. In a phase II study with olaparib, the ORR was 9.5% in 42 patients who had received platinum and was 20% in 20 patients without prior platinum exposure (14). In the OlympiAD phase III study of olaparib, prior platinum chemotherapy as a whole did not affect the relative benefit of olaparib (HR for PFS was 0.67 and 0.60 for patients with and without prior platinum), although olaparib efficacy in relation to the timing of platinum chemotherapy was not reported (15). Our results emphasize the importance of robustly assessing platinum-free interval in future studies.

The safety profile of PARP inhibitors is generally consistent across the different agents, with the primary toxicity being clinically manageable myelosuppression. Additional toxicities include mild-to-moderate gastrointestinal toxicity (i.e., nausea, vomiting, diarrhea, and abdominal pain) and fatigue (16). Three of the 83 evaluable patients from our study discontinued therapy for a drug-related toxicity, suggesting that most toxicity could be managed with supportive care. Although myelosuppression was common, the clinical sequelae (i.e., hemorrhage and infection) were rare and included 1 patient with a severe but transient episode of epistaxis and 1 with neutropenic sepsis successfully treated for infection who later had a confirmed partial response. Grade 3/4 thrombocytopenia occurred in 29% (14/48) of patients in cohort 1 and 23% (8/35) of patients in cohort 2. Hemorrhage was rare, occurring in only 1 patient, suggesting that thrombocytopenia only rarely had clinical sequelae. No acute myeloid leukemia or myelodysplastic syndrome was noted during the median follow-up time of 13.7 months. Acute myeloid leukemia or myelodysplastic syndrome has been noted to occur at an incidence of 1%–3% among patients following a few months up to several years of treatment with PARP inhibitors (3, 17–19), although this has been exclusively reported in patients with ovarian cancer.

There are several limitations to our study. The open-label design may have led to investigator bias in assessment of response and progression; therefore, the primary endpoint was ORR by independent imaging facility assessment. In addition, the termination of enrollment prior to completion of the targeted 70 patients per cohort limits the conclusions that may be drawn from this study. We recruited few patients with HER2-positive breast cancer. Although antitumor activity was observed in these patients, the data are preliminary and further studies would be required to assess the activity of talazoparib in patients with HER2-positive disease. In addition, the two cohorts differed in tumor subtype, which may have had an influence on PFS and overall survival (Table 1), with triple-negative cancers in general having adverse PFS and overall survival compared with estrogen receptor–positive cancers. Of note, a phase III clinical trial (EMBRACA, NCT01945775) recently reported improved PFS for talazoparib compared with treatment of physician's choice, with a doubling of ORR and prolonged duration of response (20).

In conclusion, talazoparib is a novel and highly potent PARP inhibitor with a manageable safety profile, pharmacokinetics that support once-daily dosing, and encouraging single-agent efficacy in germline BRCA1/2 mutation carriers with advanced breast cancer previously treated with platinum and in patients treated with multiple lines of prior nonplatinum-based chemotherapy. The impact of prior platinum exposure and platinum-free interval on sensitivity to PARP inhibitors should be further investigated.

N.C. Turner reports receiving commercial research grants from and is a consultant/advisory board member for Pfizer. M.L. Telli is a consultant/advisory board member for Tesaro, Pfizer, Merck, and AstraZeneca. H.S. Rugo reports receiving commercial research grants from Pfizer, Novartis, Lilly, Genentech, Macrogenics, Merck, Daichi, Immunomedics, and Eisai, and reports receiving speakers bureau honoraria from Celltrion. J. Ettl reports receiving speakers bureau honoraria from, is a consultant/advisory board member for, and reports other remuneration from Pfizer. J. Balmaña reports receiving commercial research grants from Pharmamar, and is a consultant/advisory board member for AstraZeneca, Clovis Oncology, Tesaro, and Bristol-Myers Squibb. P.A. Fasching reports receiving commercial research grants from Novartis to his institution, reports receiving speakers bureau honoraria from Novartis, Roche, Pfizer, Celgene, Daiichi-Sankyo, Puma, Teva, and Amgen, and is a consultant/advisory board member for Novartis, Roche, Pfizer, Celgene, Daiichi-Sankyo, Puma, Teva, and Celgene. S.A. Hurvitz reports receiving commercial research grants from Ambryx, Amgen, Bayer, Bi Pharma, Biomarin, Cascadian, Daiichi-Sankyo, Dignitana, Genentech, GlaxoSmithKline, Lilly, Macrogenics, Medivation, Merrimack, Novartis, Obi Pharma, Pfizer, Pieris, Puma, Roche, and Seattle Genetics. A.M. Wardley is an employee of, reports receiving speakers bureau honoraria from, and is a consultant/advisory board member for Pfizer. A.L. Hannah is a consultant/advisory board member for Pfizer. M.E. Robson reports receiving commercial research grants from AstraZeneca, Abbvie, Tesaro, and Medivation, and is a consultant/advisory board member for AstraZeneca, Pfizer, and McKesson. No potential conflicts of interest were disclosed.

Conception and design: N.C. Turner, H.S. Rugo, S.A. Hurvitz, A.L. Hannah, M.E. Robson

Development of methodology: N.C. Turner, S.A. Hurvitz, A.L. Hannah, M.E. Robson

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): N.C. Turner, M.L. Telli, H.S. Rugo, A. Mailliez, J. Ettl, E.-M. Grischke, L.A. Mina, J. Balmaña, P.A. Fasching, S.A. Hurvitz, A.M. Wardley, M.E. Robson

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): N.C. Turner, M.L. Telli, H.S. Rugo, J.Ettl, E.-M. Grischke, L.A.Mina, P.A. Fasching, S.A. Hurvitz, A.M. Wardley, C. Chappey, A.L. Hannah

Writing, review, and/or revision of the manuscript: N.C. Turner, M.L. Telli, H.S. Rugo, A. Mailliez, J. Ettl, E.-M. Grischke, L.A. Mina, J. Balmaña, P.A. Fasching, S.A. Hurvitz, A.M.Wardley, C.Chappey, A.L.Hannah, M.E. Robson

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): C. Chappey, A.L. Hannah

Study supervision: N.C. Turner, E.-M. Grischke, L.A. Mina, S.A. Hurvitz, A.L. Hannah

Other: All authors had full access to the study data, contributed to the revision and approval of the manuscript, and participated in the decision to submit the manuscript for publication.

We would like to thank the patients and their families/caregivers for their participation. This study was sponsored by Medivation, which was acquired by Pfizer in September 2016. In Manchester, this study was carried out in the NIHR Manchester Clinical Research Facility at The Christie NHS Foundation Trust. Editorial and medical writing support was provided by Edwin Thrower, PhD, and Mary Kacillas at Ashfield Healthcare Communications, and was funded by Pfizer.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.