Preclinical and Clinical Trial Results Using Talazoparib and Low-Dose Chemotherapy

This article is featured in Highlights of This Issue, p. 1

There are currently four PARP inhibitors (PARPi) approved for clinical use with many others in clinical development (1). In the last 6 years, olaparib, rucaparib, and niraparib have been approved by the FDA as monotherapies in platinum-sensitive, high-grade serous ovarian carcinomas, and olaparib and talazoparib have been FDA-approved as monotherapies in individuals with HER2-normal breast cancers and deleterious germline BRCA1 or BRCA2 (BRCA1/2) mutations (2–12). While some of the ongoing trials are still focused on targeting tumors with suspected homologous recombination (HR) deficiencies; it is now apparent that many patients with wild-type BRCA1/2 ovarian cancers also derive clinical benefit from PARP inhibition, particularly in combination with chemotherapy (13, 14). This is not surprising, given that up to 50% of high-grade serous ovarian cancers are reported to have somatic defects in homologous recombination (15). Other recent trials have also demonstrated a role for PARPis in a subset of patients with metastatic castration-resistant prostate cancer and pancreatic cancer with biomarker-identified DNA damage repair defects (16–18).

Talazoparib is a potent PARPi as evidenced by its low nanomolar inhibition of PARP1 enzymatic activity (19, 20). Pommier and others have hypothesized that talazoparib may also work through “PARP-trapping” of PARP-1 onto sites of DNA damage (19, 21, 22). This mechanism involves the accumulation of PARP-1 at sites of DNA damage, physically blocking access to the DNA by proteins involved in DNA repair, transcription, and replication and ultimately leading to cell death (23). Moreover, in vitro studies demonstrate that PARP-trapping on DNA sensitizes cells to topoisomerase I inhibitors like irinotecan even in the absence of defects in DNA repair (24, 25). Although particular cancers may initially show sensitivity to PARPis, PARP-1, and PARP-2 are not essential for cell viability. Thus, development of resistance to PARPi monotherapy almost invariably occurs in cancers already prone to mutate and evolve in response to selection pressures. Currently, there are multiple mechanisms that have been described for augmenting the effects of PARPis and delaying the development of resistance (26, 27). This understanding has fueled several clinical trials of combinations of PARPi with chemotherapy (28). More importantly, if effective, this combination strategy may sensitize cancer cells typically considered unresponsive to PARPis into malignancies that may respond to these agents.

While chemotherapy combinations with virtually all PARPis have been studied in phase I clinical trials, it has been challenging to dose-escalate PARPis with standard or reduced doses of chemotherapy due to significant myelosuppressive toxicities (29–37). To address this issue, we first demonstrated that low doses of temozolomide and irinotecan (10%–15% of the usual dose) are sufficient to potentiate the efficacy of talazoparib both in vivo and in vitro (38). On the basis of our preclinical data, we launched an investigator-initiated phase I dose-escalation clinical trial combining talazoparib at its single agent maximum tolerated dose (MTD) with very low doses of chemotherapy that are below standard clinically therapeutic doses. We now report the preclinical data used to develop and test this hypothesis as well as the final results of an investigator-initiated phase I clinical study with biomarker correlates in patients with advanced malignancies.

All cell lines used in these studies are derived from ATCC with the exception of M296 (heterozygous for activating mutation NRAS Q61R), M238 (heterozygous for activating mutation BRAF V600E, CDKN2A -/-), M207 (heterozygous activating mutation NRAS Q61R, TP53 -/-), and M249 (activating mutation BRAF V600E, PTEN -/-, CDKN2A -/-), which are melanoma cell lines developed from tumor samples at UCLA. All media were supplemented with 10% heat-inactivated FBS (Omega Scientific, Inc, Tarzana, CA) and 1% penicillin and streptomycin (Irvine Scientific, Santa Ana, CA). All cell lines were kept in an atmosphere of 5% CO2 at 37°C.

Cells were plated in 6-well plates and treated with talazoparib 100 nmol/L, temozolomide 100 μmol/L, talazoparib 100 nmol/L plus temozolomide 100 μmol/L, or DMSO control. Cells were harvested 24 hours after treatment, resuspended in cold 70% ethanol, and kept at −20°C for up to a week before being resuspended in cold PERM buffer (PBS, 1% BSA and 0.2% Triton X-100) for staining with rabbit antibody to phosphorylated Ser139 on Histone H2AX (γH2AX; Millipore, catalog no. 16–202A). Cellular DNA was stained using propidium iodide. Samples were analyzed using FACS Scan flow cytometer (BD Biosciences) and analyzed on FlowJo software (Tree Star Inc., Ashland, OR).

Cells were harvested at 4°C in RIPA lysis buffer after 24 hours of treatment with talazoparib alone or in combination with temozolomide. Insoluble material was cleared by centrifugation at 10,000 × g for 10 minutes, and protein was quantitated using a Pierce Biochemicals BCA Protein Assay Kit (Thermo Fisher Scientific, catalog no. 23227). Protein was resolved by SDS-PAGE electrophoresis and transferred to nitrocellulose membranes (Invitrogen). Blots were incubated with mouse γH2AX antibody (Millipore, #05–636, MA), washed, incubated with goat anti-rabbit IgG horseradish peroxidase conjugate (Upstate, Bellerica, MA), developed using ECL Plus chemifluorescent reagent (Amersham Biosciences, Pistcataway, NJ), and imaged and quantified using Flour Chem Q (Alpha Innotach, San Leandro, CA).

Cells were plated at a density of 1.5×10⁶ cells with 10-mL growth medium and allowed to grow overnight. The following day, cells were treated with talazoparib 100 nmol/L, temozolomide 50 μmol/L, or talazoparib 100 nmol/L + temozolomide 50 μmol/L for 24 hours. Cytoplasmic, soluble nuclear, and chromatin bound proteins were isolated using Thermo Scientific Subcellular Protein Fractionation Kit (Catalog no. 78840) as per the manufacturer's instructions. Protein cell lysates concentrations were determined by BCA. Lysates were denatured at 95°C for 5 minutes prior to separation by protein electrophoresis on 4% to 20% Tris-Glycine Midi gels. Gels were then transferred to nitrocellulose membrane using iBlot Invitrogen transfer system. Western blots were performed with monoclonal rabbit Anti-PARP1 (Abcam, #32138) at 1:2,000 dilution.

Cells were seeded in complete medium onto coverslips in 24-well plates at a concentration of 30,000 to 50,000 cells/well. After recovery overnight, cells were treated with 100 nmol/L talazoparib, 50 to 100 μmol/L temozolomide, or combination for 24 hours. Cells were washed 3 times with PBS, fixed with ice-cold 100% methanol for 10 minutes, and rinsed with PBS. Slides were blocked in 1% BSA in 0.3% Triton X-100 at room temperature for 1 hour. Dual staining was performed using antibodies to PARP and γH2AX. Cells were incubated with anti-PARP antibody (Abcam mouse monoclonal, #110915) at 3 μg/mL in blocking buffer overnight at 4°C, washed with PBS, and labelled with anti-mouse Alexa Fluor 488 (Abcam, #150113) at 1:500 dilution in blocking buffer for 1 hour at room temperature. Cells were washed in PBS before incubation with γH2AX antibody (Abcam rabbit polyclonal, #11174) in 2 μg/mL in blocking buffer at room temperature for 1 hour. γH2AX was labelled with anti-rabbit Alexa Fluor 594 (Abcam, #150080) at 1:500 dilution at room temperature for 1 hour. Cells were washed with PBS and mounted to glass slides with ProLong Diamond Antifade Mountant containing DAPI (ThermoFisher P36971). Confocal laser scanning microscopy was performed at the CNSI Advanced Light Microscopy/Spectroscopy Shared Resource Facility at UCLA using a Leica Confocal SP2 MP-FLIM with 63x oil-immersion lens with fixed PMT setting.

Melanoma cell lines M207 and M238 and ovarian cell line RMG1 were grown as xenografts in mice. Six-week-old CD-1 athymic nude mice were purchased from Charles River Laboratories. A total of 5 × 10⁶ to 10 × 10⁶ cells in 50% Matrigel (BD Biosciences, San Jose, CA) were injected into the right flanks of 8 to 10 mice per treatment group for each cell line. Tumor xenografts were measured with calipers twice a week, and tumor volume in mm (3) was determined by multiplying height × width × length. When tumors reached an average size of 150 to 200 mm (3) for M207 and RMG1 or an average of 400 mm (3) for M238, mice were randomized into treatment groups. Talazoparib (0.33 mg/kg in sterile PBS) or sterile PBS (control) was given orally daily for duration of the experiments. 2.5 mg/kg or 5.0 mg/kg temozolomide (Sigma-Aldrich) was diluted in 10% DMSO and given intraperitoneally. Temozolomide dosing schedules varied: on days 1 to 5 only, for 5 days ×3 courses, or daily (continuously). Animals were euthanized per veterinary recommendations or on day 26 or 28 of treatment. Results are presented as mean volumes for each group. Error bars represent the SEM.

This single-center, open-label, phase I, investigator-initiated dose-escalation study enrolled heavily pretreated patients without deleterious germline BRCA1/2 mutations with unresectable, locally advanced or metastatic solid malignancies. The primary objectives of this trial were to determine safety and tolerability, dose-limiting toxicities (DLT), and MTD, for talazoparib in combination with temozolomide or irinotecan. Secondary endpoints included pharmacokinetics, antitumor activity, determination of a recommended phase II dose (RP2D), and biomarker analysis. Eligibility criteria included: ≥ 18 years of age, biopsy-confirmed advanced malignancy, measurable disease per Response Evaluation Criteria in Solid Tumors (RECIST) v1.1, Eastern Cooperative Oncology Group performance status (ECOG PS) of ≤ 1, adequate renal and liver function, and no known brain metastases. Patients with prior exposure to temozolomide or irinotecan were permitted, but not those with prior exposure to any PARPi. Additional exclusion criteria included an uncontrolled medical condition or hypersensitivity to temozolomide or irinotecan. All patients gave written informed consent per the Declaration of Helsinki. The clinical trial was approved by the UCLA Institutional Review Board. This study was monitored for safety by the UCLA Data and Safety Monitoring Board (DSMB). The study is registered with ClinicalTrials.gov with the identifier NCT02049593.

Patients were assigned at investigator discretion to either talazoparib plus temozolomide (Arm A) or talazoparib plus irinotecan (Arm B) in parallel cohorts. Treatment cycles were 28 days in length. In this design, talazoparib was dose-escalated from 500 mcg to the single agent MTD of 1 mg by mouth daily on days 1 to 28 before dose escalation of temozolomide or irinotecan. The starting dose of temozolomide was 25 mg/m²/day orally on days 1 to 5, and the irinotecan dose was 25 mg/m²/day intravenously on days 1 and 15. The sample sizes in the dose-escalation cohorts were derived from the standard modified Fibonacci 3+3 dose-escalation trial design (39). Dose-escalation decisions were based on data from all patients, including safety information, adverse events (AE), DLTs, and pharmacokinetics. For Common Terminology Criteria for Adverse Events (CTCAE) grade 3 and 4 treatment-related toxicities, daily dosing of talazoparib and chemotherapy was discontinued until toxicity resolved to grade 0 or 1, at which point therapy was resumed with a predefined dose reduction of talazoparib by 0.25 mg if resolution occurred within 10 days and 0.5 mg if resolution occurred within 11 to 20 days from the drug interruption start date. Talazoparib re-escalation was not allowed. Chemotherapy dose reductions were done at the investigator's discretion.

A DLT was defined as a CTCAE grade 3 or 4 AEs or abnormal lab value that could not otherwise be attributed to a patient's cancer, progression, unrelated illness, or non-study medications with the following exceptions: grade 3 leukopenia, grade 3 anemia, asymptomatic grade 3 thrombocytopenia that resolves within 7 days of treatment interruption, a non-hematologic grade 3 laboratory AE that was asymptomatic and returned to baseline or grade ≤ 1 within 7 days unless identified as relevant by the investigator, grade 3 fatigue that improves to grade ≤ 2 in ≤ 5 days, and grade 3 nausea, vomiting, or diarrhea that was medically managed to grade ≤ 2 within 24 hours. In addition, failure to receive at least 75% of planned doses in cycle 1 or delay to start cycle 2 by more than 14 days due to toxicity was considered a DLT. The MTD was defined as the highest combination drug dosage not causing medically unacceptable toxicity in 2 or more patients in a dosing cohort.

Blood samples for plasma pharmacokinetic analysis of talazoparib, temozolomide, and irinotecan were collected from all patients at baseline, 0.5, 1, 1.5, 2, 4, 6, 8, and 24 hours during cycle 1, day 1 (single dose) and cycle 2, day 1 (multiple dose). These were used for comparison with the single agent phase I study of talazoparib (17).

Tumor response was evaluated by CT scans of the chest, abdomen, and pelvis with oral and intravenous contrast based on RECIST v1.1. Scans were performed at screening and every 2 cycles (8 weeks) after starting study treatment until disease progression. Safety was monitored at screening and throughout the treatment period by physical examination and collection of AE data with grading per NCI CTCAE v4.03.

Archival formalin-fixed, paraffin-embedded tumor tissue was collected (5 or more 10-μm sections/tumor sample) and analyzed for homologous recombination deficiency (HRD) status, gene mutations, and gene-specific loss of heterozygosity (LOH). Mutational status and LOH were determined for genes involved in DNA damage and repair was performed by Myriad Genetics and Foundation Medicine per previously described methods (40, 41). Next-generation sequencing was used to produce genome-wide SNP profiles to calculate an HRD score defined as the unweighted sum of LOH, telomeric-allelic imbalance, and large-scale state transitions scores by Myriad Genetics (42).

Descriptive statistics were used throughout the study. Progression-free survival, defined as the time from first dose of study treatment to first documentation of progression or death due to any cause, was analyzed by descriptive statistics. Pharmacokinetic parameters were calculated for each patient using non-compartmental analysis of concentration–time data.

Raw data for this study were generated at UCLA Translational Oncology Research Lab. Derived data supporting the findings of this study are available from the corresponding author upon request.

Because it has been suggested that PARP trapping results in collapse of the replication fork with formation of DNA double-strand breaks (DSB; refs. 43–45), phosphorylation of histone H2AX (γH2AX) was analyzed by flow cytometry in the presence of talazoparib monotherapy or combination talazoparib and temozolomide. Treatment with etoposide, a topoisomerase II inhibitor that induces DNA DSBs in actively replicating or transcribing cells, was used as a positive control for γH2AX. DNA DSB formation in cells treated with etoposide was not cell cycle phase-specific. γH2AX formation in cells treated with the combination of talazoparib plus temozolomide occurred in S-phase (M296), a line not sensitive to talazoparib or temozolomide monotherapy (Fig. 1A). Using Western blot, we measured protein expression of γH2AX in cells treated with dual talazoparib + temozolomide compared with monotherapy. The combination of talazoparib + temozolomide increased γH2AX compared with treatment with either talazoparib or temozolomide alone (Fig. 1B).

To further characterize the molecular differences between the combination, we used confocal microscopy to co-localize PARP and DNA DSBs in the presence of either monotherapy or the talazoparib/low-dose chemotherapy combinations. PARP is constitutively present in the nuclei in the absence of DNA damaging agents, but phosphorylation of histone H2AX is dynamic. Confocal data show that γH2AX formation in unsynchronized cells exposed to talazoparib, temozolomide, or combination therapy is increased when talazoparib is combined with temozolomide (Fig. 1C).

Temozolomide and irinotecan (SN-38) were chosen to study in combination with talazoparib as these cytotoxics have modest activity in these histologies and mechanistically are synergistic with PARPis (46, 47). We evaluated the effects of these combinations in ovarian (RMG1) and melanoma (M207 and M238) xenograft models. We treated the mice with talazoparib alone at a dose of 0.33 mg/kg daily or in combination with temozolomide 2.5 mg/kg in two different schedules (daily x 28 days or 5 days; Figure 1D). The combination of continuous talazoparib + temozolomide 2.5 mg/kg inhibited tumor growth greater than treatment with either single agent temozolomide 2.5 mg/kg or 0.33 mg/kg talazoparib and significantly (P = 0.02) when compared with vehicle control.

Forty patients were enrolled into two arms of a phase I dose-escalation study designed to assess the safety and efficacy of combinations of continuous oral talazoparib plus oral temozolomide days 1 to 5 (n = 18) or talazoparib plus intravenous irinotecan every 14 days (n = 22; Fig. 2). On the basis of the preclinical data, our strategy was to first dose-escalate talazoparib from 0.5 mg/day until it reached the full single agent dose of 1 mg/day, before increasing the chemotherapy doses (38). The chemotherapy dosing was initiated at substandard clinical doses because our in vivo preclinical data demonstrated that as little as 10% of the MTD of the cytotoxic agent was capable of achieving synergistic growth inhibition with talazoparib (Figure 2).

Patient characteristics in the two groups differed by histology with patients preferentially assigned to the temozolomide or irinotecan arm based on investigator preference (Table 1). All patients with melanoma or sarcoma were assigned to the temozolomide arm (n = 7), and all patients with colorectal cancer were assigned to the irinotecan arm (n = 13). The patient population was heavily pretreated with a median of 5 prior treatment regimens in the temozolomide arm (range, 1–16 regimens) and 4 in the irinotecan arm (range, 1–7 regimens). A majority (82%) of the patients in the irinotecan arm had received prior platinum therapy compared with 56% in the temozolomide arm. ECOG PS was somewhat better in the temozolomide arm compared with the irinotecan arm with an ECOG = 0 in 72% versus 55% patients, respectively. At the time of data cutoff in August 2019, treatment had been discontinued in 96% (17/18) patients in the talazoparib plus temozolomide arm due to disease progression and 1 patient remains on study for more than 2 years. Treatment was discontinued in all patients in the talazoparib plus irinotecan arm due to disease progression (n = 21) or withdrawal of consent (n = 1). No patients discontinued treatment due to AEs.

A total of 37 of 40 patients required a protocol-mandated dose reduction in talazoparib due to grade 3 and 4 hematologic toxicities, all of which resolved with treatment interruption. In 90% of patients, only one dose reduction was required (Supplementary Table S1). Three DLTs were identified in the talazoparib/temozolomide arm: grade 4 thrombocytopenia in 2 patients in cohort 4 (talazoparib 1 mg plus temozolomide 50 mg/m²) and grade 3 thrombocytopenia for > 7 days with concurrent grade 3 neutropenia in 1 patient in cohort 3.5 (talazoparib 1 mg plus temozolomide 37.5 mg/m²). Three DLTs were also identified in the talazoparib/irinotecan arm in 2 patients in cohort 4 (talazoparib 1 mg + irinotecan 50 mg/m²): one with grade 3 neutropenia for > 7 days and the second with grade 4 thrombocytopenia with concurrent grade 3 neutropenia, and 1 patient in cohort 3.5 (talazoparib 1 mg plus irinotecan 37.5 mg/m²) with grade 4 thrombocytopenia. The MTD for the talazoparib combinations were talazoparib 1 mg orally daily plus temozolomide 37.5 mg/m²/day on days 1 to 5 of a 28-day cycle and talazoparib 1 mg orally daily plus irinotecan 37.5 mg/m² intravenously every 14 days.

The overall safety profiles for both treatment arms is summarized in Table 2. AEs occurred in all patients in both arms, with the most common being anemia (94%), neutropenia (72%), thrombocytopenia (72%), and nausea (50%) in the talazoparib/temozolomide arm and anemia (68%), neutropenia (40%), and thrombocytopenia (45%) in the talazoparib/irinotecan arm. None of the thrombocytopenic events were associated with bleeding. All grade 3/4 AEs reported in this study were hematologic in nature (Table 2).

Talazoparib plasma concentration increased proportionally with dose increases and a peak concentration was achieved between 0.75 to 2 hours (Supplementary Table S2). Mean pharmacokinetic profiles of talazoparib in combination with either temozolomide or irinotecan were similar to that of talazoparib monotherapy seen in the phase I trial of talazoparib (Supplementary Fig. S1; ref. 20). Plasma concentrations after dosing talazoparib at 0.5 mg/day in the combination therapy arms were less than those of talazoparib 0.6 mg/day monotherapy and greater than talazoparib 0.4 mg/day monotherapy. Steady state plasma concentrations were reached by the end of the second week of dosing.

A total of 38/39 (97%) patients were evaluable for objective responses. One patient was not evaluable due to patient withdrawal on day 5. One patient was still receiving treatment at the data cutoff. Objective partial responses (PR) were confirmed in 2 patients with ovarian cancer and 1 with melanoma in the temozolomide arm. An additional PR was seen in a patient with Ewing sarcoma. A patient with isocitrate dehydrogenase wild-type intrahepatic cholangiocarcinoma remained on the combination with SD for 48 months. The median duration of response to talazoparib/temozolomide was 40.3 weeks (range, 4.4–55.1 weeks; Fig. 3A). In the talazoparib/irinotecan arm, confirmed objective PRs were observed in a total of 5 patients, two with ovarian cancer and 1 patient each with triple-negative breast cancer, small cell lung cancer (SCLC), and cervical adenocarcinoma (Fig. 3B). The median duration of response to the talazoparib/irinotecan was 28.2 weeks (range, 10.0–49.9). There were no complete responses (CR) in either arm. A sustained PR of over 4 months to talazoparib 1 mg/day + irinotecan 37.5 mg/m² every 2 weeks was observed in a patient with cervical adenocarcinoma with lung metastases who had progressed on four prior lines of combination chemotherapy, including three platinum-based regimens and a programmed cell death protein 1 monoclonal antibody (Fig. 3C).

After enrollment, 1 patient in each of the two arms was found to carry a germline deleterious BRCA1 or BRCA2 mutation. The patient in the talazoparib/temozolomide arm with BRCA1⁺ ovarian cancer had stable disease for 32 weeks. The patient in the talazoparib/irinotecan arm with BRCA2⁺ pancreatic cancer had progressive disease on the first imaging assessment. A genomic instability score was calculated using the Myriad Genetics myChoice test and specific genes from DNA damage repair pathways were interrogated on archived tissue in all patients with non-BRCA1/2⁺ ovarian cancer in both arms (n = 7) as well as in all observed PRs (n = 8; Table 3). Two patients with non-BRCA⁺ ovarian cancer with PRs in the talazoparib/temozolomide arm were found to have defects in homologous recombination repair genes with LOH, one with a deleterious mutation in PALB2 and one with defective Rad51D. Both patients also had high HRD scores. However, the patient with the highest HRD score did not respond to treatment while a patient with the lowest HRD score did respond. Of the partial responders in the talazoparib/irinotecan arm, 1 patient with breast cancer was found to have a somatic deleterious BRCA1 mutation with LOH and a high HRD score. None of the remaining patients (n = 4) were found to have defects in known DNA repair genes (Tables S3 and S4), and only one (an ovarian cancer patient) had a high HRD score. Taken together, these data showed that in 4/8 (50%) of patients without germline or somatic mutations in the BRCA1/2 genes or a high HRD score had confirmed objective responses using the strategy of causing DNA damage with low-dose chemotherapy in the presence of full-dose or near full-dose PARPi, suggesting that PARPis may be useful in histologies other than high-grade serous ovarian carcinomas and gBRCA1/2⁺-associated breast cancers.

The purpose of this phase I clinical trial was to evaluate the safety and efficacy of talazoparib in combination with low-dose temozolomide or irinotecan, with prioritization of talazoparib dose escalation to reach the MTD before beginning dose escalation of chemotherapy. The current approach represents a departure from what has been tried in numerous other PARPi clinical trials where a PARPi was dose-escalated in the setting of standard-of-care chemotherapy doses (31, 32, 36). This commonly investigated approach is nearly universally characterized by dose-limiting hematologic AEs even with minimal doses of PARPis. Using our preclinical findings that small doses of chemotherapy potentiate the effects of talazoparib and appreciating the limitations of combining chemotherapy with PARPis, we used ∼10% of what would be considered clinically therapeutic doses of temozolomide and irinotecan: 200 mg/m² every day × 5 days and 180 mg/m² every 2 weeks, respectively (38). The combination MTD was determined to be talazoparib 1 mg with temozolomide 37.5 mg/m² or irinotecan 37.5 mg/m². However, given the number of patients with grade 3/4 hematologic AEs (many occurring after the 4-week DLT evaluation period), and the fact that the majority of patients sustained clinical benefit at 25 mg/m² of either cytotoxic chemotherapy at these combined doses, our RP2D was 25 mg/m² of either temozolomide or irinotecan with 1 mg of talazoparib. Our study did not recommend decreasing dosing of both talazoparib and chemo in parallel, and the priority was to maintain the talazoparib at 1 mg. However, it is possible that lower doses of talazoparib in combination with chemotherapy may lead to more sustainable treatments. In this study, as with virtually all PARPi trials, AEs were primarily hematologic with the majority of patients requiring talazoparib dose reductions (as per clinical trial protocol) and all of which resolved with treatment interruption.

While there were no objective CRs in this heavily pretreated patient population, there were a total of 9/38 (24%) PRs by RECIST. In addition, a total of 23/38 (60%) of patients had clinical benefit defined as a PR or SD for > 6 months. There were 13/18 (72%) of patients with either SD or PR in the talazoparib/temozolomide arm and 10/20 (50%) in the talazoparib/irinotecan arm. Biomarker analysis in patients with ovarian cancer on the talazoparib/temozolomide arm identified HR defects in tumors of 2 patients with PRs, 1 showing a PALB2 mutation and 1 with a mutation in Rad51D. Of note, the patient with ovarian cancer with the highest HRD score in cohort 1 did not respond to talazoparib/temozolomide. Of the patients with PRs in the talazoparib/irinotecan arm, 1 was found to have somatic loss of BRCA1. HR defects related to LOH for specific HR genes were not identified in the remaining 4 patients, although 1 had a high HRD score. This would suggest that features other than HR deficiency may have contributed to the responses seen in most of these patients, but given the small numbers in this trial, it is difficult to conclude any definitive HRD biomarker correlation with clinical outcomes.

To our knowledge, this is the first trial to examine the combination of full-dose PARP inhibition with low doses of chemotherapy. We have demonstrated that the single agent recommended dose of talazoparib combined with low-dose temozolomide or irinotecan can result in promising clinical activity and tolerability in a heavily pretreated patient population with a diverse range of malignancies. While defects in DNA DSB repair pathways may predict sensitivity to combination therapies of low-dose chemotherapy and talazoparib, we have shown both in vitro, in vivo, and now in patients that low-dose chemotherapy combined with full-dose or near full-dose talazoparib has efficacy in malignancies lacking known HR deficiencies. Taken together, these data now suggest that BRCA1/2 wild-type cancers from a diverse range of histologies may be sensitive to and successfully treated with the combination of talazoparib with chemotherapy in a setting for which either talazoparib or low-dose cytotoxic monotherapy would not have been predicted to be clinically beneficial. Given the low doses of chemotherapy used in this trial and the absence of clinical responses in many of these cancers as seen in previous trials with monotherapy PARP inhibition (48), it is possible that the DNA damage induced by the low doses of temozolomide or irinotecan used here elicits a picture similar to “PARP trapping” in patients who otherwise would not respond to single agent PARPi. This approach opens new therapeutic possibilities for an effective combination use of PARPis with DNA-damaging agents, but larger randomized clinical trials will be needed to address this question more conclusively.

Our study has several limitations which include both the heavily pretreated patient population (which may reflect favorable biology) and the lack of corresponding biomarker information in the entire trial population. Many of the specimens used to assess biomarker information were archived tissue from years prior to study enrollment. Because pretreatment biopsies were not mandated, this may not reflect evolutionary changes in tumor biology which may explain sensitivity or resistance to PARPis. However, while our trial was not randomized, our clinical results are supported by a recent trial in SCLC demonstrating the successful combination of olaparib and temozolomide (49). In fact, based on the phase I trial reported herein, two additional trials are ongoing: NCT03672773 and NCT03830918. However, given the findings of continued hematologic toxicities, several trials are also being performed with intermittent dosing strategies. Should these data be confirmed in a larger study and this new approach prove effective and tolerable in additional datasets, it could open the use of PARP inhibition therapy to a much broader spectrum of malignancies beyond their proven roles in BRCA1/2⁺ and HRD-associated cancers.

Z.A.Wainberg reports grants from AbbVie and other support from Amgen, Astellas, AstraZeneca, Daiichi, Bayer, Genentech, Merck, BMS, Novartis, Ipsen, Incyte, Seagen, Pfizer, and Lilly outside the submitted work. A.S. Singh reports grants and personal fees from Rain Therapeutics and Deciphera, grants from Tracon, and personal fees from Epizyme outside the submitted work. J.R. Hecht reports grants and nonfinancial support from Amgen; grants from Tesaro, A2, Tizona, Merck, Bold Therapeutics, NGM Bio, Mirati, and ARMO BioSciences/Lilly; personal fees from Cancer Panels, Notch Therapeutics, Research to Practice, Scripps, Istari Oncology, Corcept Therapeutics, AstraZeneca, and Regnancy; and grants and personal fees from Astellas, Gritstone, and Exelixis outside the submitted work. J. Goldman reports grants from Pfizer outside the submitted work. B. Chmielowski reports personal fees from Iovance Biotherapeutics, IDEAYA Biosciences, Deciphera, Sanofi Genzyme, OncoSec, Nektar, Instil Bio, Delcath, and Novartis outside the submitted work. R.S. Finn reports grants, personal fees, and nonfinancial support from Pfizer during the conduct of the study as well as personal fees from AstraZeneca, CStone, Exelixis, and Hengrui; grants, personal fees, and nonfinancial support from Bayer, Eli Lilly, Merck, and Roche/Genentech; grants and nonfinancial support from Bristol-Myers Squibb; and grants and personal fees from Novartis and Adaptimmune outside the submitted work. M. Brennan reports other support from 1200 Pharma and TORL Biotherapeutics outside the submitted work. D.J. Slamon reports nonfinancial support and other support from BioMarin; grants, nonfinancial support, and other support from Pfizer; grants, personal fees, and nonfinancial support from Novartis; personal fees from Eli Lilly; and other support from Amgen, Seattle Genetics, 1200 Pharma, and TORL Biotherapeutics outside the submitted work. No disclosures were reported by the other authors.

Z.A. Wainberg: Conceptualization, resources, data curation, software, formal analysis, supervision, funding acquisition, validation, investigation, visualization, methodology, writing–original draft, project administration, writing–review and editing. A.S. Singh: Investigation, visualization, methodology, writing–original draft, writing–review and editing. G.E. Konecny: Investigation, project administration, writing–review and editing. K.E. McCann: Data curation, formal analysis, supervision, validation, investigation, writing–original draft. J.R. Hecht: Investigation, writing–review and editing. J. Goldman: Investigation, writing–review and editing. B. Chmielowski: Investigation. R.S. Finn: Investigation. N. O'Brien: Data curation, formal analysis, methodology. E. Von Euw: Data curation, methodology. M.M. Price: Investigation. D. Martinez: Methodology, project administration. L. Yonemoto: Investigation. M. Brennan: Investigation. J.A. Glaspy: Investigation. D.J. Slamon: Conceptualization, resources, supervision, funding acquisition, investigation, writing–original draft, writing-review and editing.

We would like to thank the UCLA DSMB for monitoring this trial. We would also like to thank Evelyn Wang, Gilles Gallant, and Joshua Henshaw of BioMarin Pharmaceuticals for their help with this clinical trial.

Talazoparib was previously owned by BioMarin, followed by Medivation, and is now owned by Pfizer.

The trial was sponsored by the Jonsson Comprehensive Cancer Center at University of California Los Angeles (UCLA) with partial support provided by both BioMarin Pharmaceuticals and Medivation.

The publication costs of this article were defrayed in part by the payment of publication fees. Therefore, and solely to indicate this fact, this article is hereby marked “advertisement” in accordance with 18 USC section 1734.