Abstract
Background.
Preclinical studies demonstrate activity of poly(adenosine 5’-diphosphate-ribose) polymerase (PARP) inhibitors in isocitrate dehydrogenase (IDH) mutant gliomas. We investigated safety, tolerability, pharmacokinetics, and efficacy of the PARP inhibitor pamiparib in conjunction with metronomic low-dose temozolomide in patients with recurrent IDH mutant (IDHmt) gliomas in a multicenter Phase I/II/window of opportunity study.
Methods.
Patients received pamiparib in conjunction with daily temozolomide. Following Phase I determination of maximum tolerated dose (MTD), we enrolled 2 patient cohorts (Arm A, multiple prior chemotherapy regimens; Arm B, single prior regimen) in a 2-stage design. Exploratory cohorts examined grade 4 IDHmt patients and intratumoral pharmacokinetics of pamiparib. The primary endpoint was objective radiographic response (ORR) by RANO criteria.
Results.
Sixty-six subjects were enrolled. We established pamiparib 60 mg twice daily with temozolomide 20 mg daily as the phase II dose. In non-enhancing and enhancing tumor, pamiparib exhibited an unbound tumor/plasma ratio of 0.92 and 0.98, respectively. 0/15 Arm A and 1/24 Arm B patients achieved a centrally confirmed partial response. Median progression-free survival for Arm A was 5.9 months (95% CI, 1.2–14.8 months), and for Arm B was 9.7 months (95% CI, 5.7–21.7 months). Grade 3+ anemia and neutropenia affected 24% and 33% of patients, respectively. Twenty-two of 66 patients (33.3%) discontinued study treatment for reasons other than tumor progression.
Conclusions.
Pamiparib appeared to achieve sufficient pharmacologically active concentrations in both enhancing and non-enhancing tumors. While some patients achieved prolonged progression-free survival, combination with temozolomide did not produce a meaningful ORR in IDHmt recurrent gliomas. Cumulative hematologic toxicity was substantial and impacted long-term tolerability.
Keywords: clinical trial, glioma, IDH mutation, PARP inhibitors, temozolomide
Most non-glioblastoma adult gliomas harbor a mutation in the isocitrate dehydrogenase (IDH) gene.1 The associated widespread epigenetic and metabolic changes have led to multiple efforts to target the effects of this mutation.2 Recently, the INDIGO trial demonstrated the benefit of inhibiting the mutated IDH protein directly with vorasidenib in grade 2 IDHmt glioma treatment-naïve patients.3 Unfortunately, the activity of IDH inhibitors in patients experiencing tumor progression post radiation and chemotherapy, and particularly those with contrast-enhancing tumors, presently appears limited.3,4
At present, most patients with IDHmt grade 2 and 3 gliomas receive radiation and alkylator-based chemotherapy at diagnosis. When tumors recur or progress, patients are often treated with second-line alkylator drugs (nitrosourea-based chemotherapy if they previously received temozolomide, or vice versa).5 Occasionally patients are re-treated with temozolomide. However, none of these approaches has a high response rate, and subsequent treatment options with proven benefit are very limited. Additionally, a substantial minority of patients treated with temozolomide develop a hypermutator phenotype conveying further resistance to standard chemotherapy.6 Thus, there is a great unmet need for novel approaches that target these tumors.
IDH mutations result in the conversion of α-ketoglutarate to 2-hydroxyglutarate (2-HG), an oncometabolite with pleiotropic cellular effects. The Bindra laboratory and others recently demonstrated that IDH mutations induce a homologous recombination (HR) defect rendering tumor cells exquisitely sensitive to poly(adenosine 5’-diphosphate-ribose) polymerase (PARP) inhibitors, creating a “BRCAness” phenotype.7 This effect was recapitulated with treatment with 2-HG in the setting of intact IDH proteins and was reversible with use of small-molecule inhibitors of the mutated IDH protein. Poly(adenosine 5’-diphosphate-ribose) polymerase inhibitor sensitivity was confirmed in a range of clinically relevant models, including primary patient-derived glioma cells in culture and genetically matched tumor xenografts in vivo. The investigators localized the mechanism of action to two α-ketoglutarate dependent dioxygenase targeted by 2-HG and other oncometabolites, KDM4A and KDM4B, which mediate HR suppression.8,9 Several other groups since have independently reported a similar synthetic lethal interaction between IDH1/2 mutations and PARP inhibitors.10–12 These findings suggested a novel strategy of utilizing PARP inhibition to exploit the effect of 2-HG rather than trying to suppress 2-HG production.
Poly(adenosine 5’-diphosphate-ribose) polymerase inhibitors have been studied in gliomas, particularly in glioblastoma. They are commonly combined with DNA-damaging therapies, as they interfere with the base excision repair needed to fix the damage induced by temozolomide’s methylation at the N7 position of guanine. Additionally, temozolomide has also been shown to synergize with PARP inhibitors in homologous repair-defective cells via a separate mechanism of action, which likely can be attributed to an increase in collapsed replication forks and consequent one-ended double-strand breaks that require HR for resolution.13,14
Pamiparib (BGB-290) is a potent DNA-trapping PARP1 and PARP2 inhibitor with preclinical data suggesting good blood-brain barrier (BBB) penetration and greater potency than the FDA-approved PARP inhibitor olaparib.15 Additionally, we hypothesized that its superior PARP-trapping activity would be advantageous compared to the blood-brain barrier-penetrant PARP inhibitor veliparib. The observed synergy between temozolomide and PARP inhibition suggested exploring temozolomide as a “sensitizer” for pamiparib in IDHmt gliomas. Based upon prior studies of the combination of pamiparib with dose-dense temozolomide reporting hematologic toxicity,16,17 we chose to explore full dose pamiparib (60 mg twice daily) with a starting dose of temozolomide 20 mg daily.
Methods
Study Design
This study had 5 arms: (1) a phase I study to determine the maximum tolerated dosing frequency of temozolomide 20 mg p.o. once daily in conjunction with pamiparib 60 mg p.o. twice daily; (2) a 2-stage phase II arm for subjects with initial grade 2 or 3 astrocytoma or oligodendroglioma who had tumor progression despite prior receipt of both temozolomide and another alkylating agent and thus were considered heavily pre-treated (Arm A); (3) a 2-stage phase II arm for subjects with initial grade 2 or 3 astrocytoma or oligodendroglioma who had tumor progression despite administration of a single prior alkylating agent >1 year previously (Arm B); (4) an exploratory arm for subjects with grade 4 astrocytoma with tumor progression despite prior alkylating agent therapy; (5) a phase 0 study to assess tumor tissue levels of pamiparib in subjects prior to tumor resection.
Eligibility
Eligible patients were age 18 or older with IDH1/2 mutant tumors as well as documented 1p/19q and MGMT methylation status. Given the mechanism of action targeting tumors with IDH mutation, both astrocytomas and oligodendrogliomas were included. Tumor classification was based upon WHO 2021 criteria. Requirements included MRI-measurable disease by FLAIR or contrast enhancement, KPS ≥ 60, and adequate organ and bone marrow function. Stable corticosteroid dose for ≥5 days was required before obtaining the baseline MRI scan. Ineligibility criteria included pregnancy, prior treatment with an inhibitor of IDH mutant protein, bevacizumab within the preceding 6 months, prior treatment with a PARP inhibitor, or use of an enzyme-inducing anti-epileptic drug. Phase I and II patients had to have available at least 3 prior MRI scans (not including screening), each separated by at least 2 months. Arm B patients were required to be ≥ 12 months since last treatment (chemotherapy or radiotherapy). Apart from Arm B subjects there was no limit on the number of prior therapies.
Treatment Plan
Phase I
The starting dose was pamiparib 60 mg twice daily with temozolomide 20 mg daily. Cycle length was 28 days. The standard 3 + 3 was used for dose finding, with a target Dose-Limiting-toxicity (DLT) rate of 33% and 6 patients treated at a putative maximum tolerated dose (MTD). Hematologic DLT was defined as NCI Common Terminology Criteria for Adverse Events (CTCAE) version 5.0. grade 4 neutropenia or thrombocytopenia, grade 3 or 4 febrile neutropenia, or hematologic toxicity preventing administration of at least 80% of planned temozolomide dose. Non-hematologic DLT included other grade 3 or 4 events that were symptomatic and not self-limited. Toxicity monitoring included weekly complete blood counts as well as serum chemistries every cycle. The toxicity evaluation period was the first 4 weeks of treatment. If 2–3 out of 3, or ≥3 out of 6 patients experienced DLT, another cohort would be enrolled with temozolomide 20 mg administered days 1–21 of each cycle. Additional potential dose level reduction cohorts incorporated temozolomide for 14 and 7 days of each cycle.
Phase II
Treatment comprised pamiparib 60 mg twice daily in conjunction with the temozolomide dose from the Phase I study. Dose reductions were permitted as protocol-specified, with dose reductions to begin with temozolomide; toxicity monitoring was as for phase I.
Surgical Arm
A subset of 10 patients with histologically confirmed grade 2–4 IDHmt gliomas progressive or recurrent following therapy, planning to undergo a biopsy or re-resection, were eligible for a correlative sub-study, in which pamiparib was administered for 7 days prior to surgery. This arm’s objective was to assess tumor tissue pamiparib levels. Patients were eligible to continue with pamiparib and temozolomide as per the Phase II treatment plan following recovery from surgery.
On the day of surgery, the patients received a single dose of pamiparib (60 mg). Blood and tumor (including enhancing and non-enhancing regions) samples were collected at 2–8 h post dosing during the surgery. Total and unbound drug concentrations of pamiparib in plasma and tumor samples were determined using a validated LC-MS/MS method.18
Grade 4 Arm
A subset of 10 patients with recurrent IDHmt grade 4 astrocytoma following radiation and chemotherapy were eligible for treatment with the Phase II treatment plan. The objective was to examine tumor response rates, PFS, and OS using descriptive statistics.
Statistics
The trial was an NIH-funded multi-institutional, open-label, nonrandomized Phase I/II study to evaluate safety of combination treatment of pamiparib at 60 mg BID, 28-day cycle with the standard temozolomide, and to test its efficacy in patients with recurrent IDH1/2 mutant glioma. The primary endpoint for Phase I was toxicity. The targeted DLT rate was 33% during the first cycle of treatment and among 6 patients. The dose was deemed safe and warranted for efficacy evaluation in a Phase II part of trial in patients with high-risk (multiple failures of prior alkylator therapies) or with lower-risk (single failure of prior alkylator therapy), independently.
The primary efficacy endpoint for Phase II was tumor response by Response Assessment in Neuro-Oncology (RANO) criteria with a minimum of 4 weeks, including PR and CR. Response rate was chosen because at study inception there were no established criteria for progression-free survival at a fixed time point (eg 6 or 12 months) with other therapies in patients with recurrent IDHmt tumors previously treated with radiation and chemotherapy. Patients with measurable enhancing disease were assessed by the high grade glioma (HGG) RANO criteria,19 while patients with only T2/FLAIR abnormalities were assessed by the low grade glioma (LGG) RANO criteria.20 Simon’s MiniMax 2-stage single-armed design was used to plan Phase II study in both high and lower-risk groups. A true response rate of 25% was considered clinically promising for Arm A patients to compare a null of 5%. A total of 25 patients were required to have 90% statistical power detecting a 20% improvement from a null of 5% with a false positive rate at 5%. The same statistical method was used to design the Arm B group with a true response rate of 50% considering clinically promising compared to a null of 30%. A total of 53 patients were required in the lower-risk group to have 90% statistical power detecting a 20% improvement with a false positive rate at 5%.
Two interim analyses were planned after patients were treated during the first stage of Phase II. Fifteen high-risk patients were enrolled onto stage 1, the study would be stopped for the null if no response was observed among the first 15 patients. The probability of early stopping for futility is 0.46 when the null was true and 0.01 when the true response was 25%. Twenty-four lower-risk patients were enrolled to stage 1, the study would be stopped for the null if 7 or fewer responses were observed among the first 24 patients. The probability of early stopping for futility is 0.57 when the null was true and 0.03 when the true response was 50%. A measurement on percent change in 6-month tumor growth rate was built in for additional consideration during the interim analysis if no response was observed in either high or lower risk group.
Secondary Phase II objectives included determining progression-free and overall survival in each arm as well as the duration of response. Exploratory objectives included evaluation of changes in tumor growth rate in subjects with non-enhancing glioma based on fluid attenuated inverse recovery (FLAIR) tumor volume measurements of serial MRI exams and assessing whether such changes were associated with progression by RANO or overall survival. Volumetric analyses of tumor growth rates are incorporated into a separate manuscript in preparation.
Baseline patient and disease characteristics were summarized using descriptive statistics. Safety data were evaluated based on NCI CTCAE v.5, and the incidence rate of adverse events was presented using a proportion. Tumor responses were initially classified by investigator assessment and followed by a central review. If imaging was not done, the missing imaging was considered a non-response. There was no imputation performed on the missing imaging data at any given time point. The response rate was estimate using a binomial distribution along with 95% CI. The duration of the response was calculated from the date of initial response being observed to the date of disease progression. Progression-free survival and overall survival were measured from date of trial registration to date that event occurred or censored at time of database locking. Survival probability was estimated using the Kaplan-Meier method.21 The CI of median survival time was constructed by the method of Brookmeyer-Crowley.22 The exploratory studies were exploratory and mechanistic in nature, and the data were presented descriptively. The analysis of clinical data was performed using SAS, version 9.4 (SAS Institute, Cary, NC).
Results
Between February 2020 and March 2022, 67 subjects were enrolled; one withdrew prior to treatment. The CONSORT diagram is shown in Figure 1. Table 2 summarizes baseline patient and tumor characteristics. Due to the expiration of ABTC funding, the database was locked in June 2023. Patients still on treatment were allowed to transition to compassionate use of pamiparib.
Figure 1.
CONSORT diagram.
Table 2.
Patient and tumor characteristics
| Baseline Characteristics | Phase I n = 7 | Arm A n = 15 | Arm B n = 24 | Grade 4 n = 10 | Surgical n = 10 | Total n = 66 |
|---|---|---|---|---|---|---|
|
| ||||||
| Age median (range) | 45.2 (25.4–54.6) | 42.7 (35.9–58.4) | 45.5 (22.2–68.6) | 35.9 (26.5–72.9) | 46.7 (32.8–66.9) | 44.5 (22.2–72.9) |
| KPS median (range) | 90 (70–100) | 80 (70–100) | 90 (60–100) | 90 (80–100) | 95 (70–100) | 90 (60–100) |
| Race n (%) White | 6 (85.7%) | 12 (80%) | 21 (87.5%) | 7 (70%) | 8 (80%) | 54 (81.8%) |
| Gender n (%) male | 4 (57.1%) | 10 (66.7%) | 12 (50%) | 5 (50%) | 6 (60%) | 37 (56.1%) |
| No. of Relapses, median (range) | 3 (1–4) | 2 (2–3) | 1 (1–3) | 2 (1–7) | 1 (1–2) | 1 (1–7) |
| Measurable disease, n (%) yes | 7 (100%) | 15 (100%) | 24 (100%) | 9 (90%) | 9 (90%) | 64 (97%) |
| IDH Mutant n (%) yes | 7 (100%) | 15 (100%) | 24 (100%) | 10 (100%) | 10 (100%) | 66 (100%) |
| Initial Histology | ||||||
| Astro gr 2 | 0 (0%) | 4 (26.7%) | 5 (20.8%) | 0 (0%) | 2 (20%) | 11 (16.7%) |
| Astro gr 3 | 4 (57.1%) | 2 (13.3%) | 1 (4.2%) | 2 (20%) | 2 (20%) | 11 (16.7%) |
| Oligo gr 2 | 0 (0%) | 6 (40%) | 12 (50%) | 0 (0%) | 3 (30%) | 21 (31.8%) |
| Oligo gr 3 | 2 (28.6%) | 3 (20%) | 6 (25%) | 0 (0%) | 0 (0%) | 11 (16.7%) |
| Astro gr 4 | 1 (14.3%) | 0 (0%) | 0 (0%) | 8 (80%) | 3 (30%) | 12 (18.2%) |
| Enrolling Histology | ||||||
| Astro gr 2 | 0 (0%) | 1 (6.7%) | 4 (16.7%) | 0 (0%) | 2 (20%) | 7 (10.6%) |
| Astro gr 3 | 4 (57.1%) | 4 (26.7%) | 2 (8.3%) | 0 (0%) | 3 (30%) | 13 (19.7%) |
| Oligo gr 2 | 0 (0%) | 3 (20%) | 8 (33.3%) | 0 (0%) | 1 (10%) | 12 (18.2%) |
| Oligo gr 3 | 2 (28.6%) | 6 (40%) | 9 (37.5%) | 0 (0%) | 1 (10%) | 18 (27.3%) |
| Astro gr 4 | 1 (14.3%) | 1 (6.7%) | 1 (4.2%) | 10 (100%) | 3 (30%) | 16 (24.2%) |
| MGMT methylated n (%) | 1 (14.3%) | 9 (60%) | 19 (79.2%) | 6 (60%) | 5 (50%) | 40 (60.6%) |
| Number of prior surgeries median (range) | 3 (1–3) | 2 (1–4) | 2 (1–3) | 2 (1–3) | 1 (1–3) | 2 (1–4) |
| Procedure, n (%) | ||||||
| GTR | 4 (57.1%) | 9 (60%) | 9 (37.5%) | 7 (70%) | 5 (50%) | 34 (52%) |
| STR | 2 (28.6%) | 6 (40%) | 14 (58.3%) | 2 (20%) | 4 (40%) | 42 (42%) |
| Biopsy | 1 (14.3%) | 0 (0%) | 1 (4.2%) | 1 (10%) | 1 (10%) | 4 (6%) |
| Prior RT, n (%) yes | 7 (100%) | 15 (100%) | 21 (87.5%) | 10 (100%) | 10 (100%) | 63 (95.5%) |
| ASM usage, n (%) yes | 4 (57.1%) | 9 (60%) | 16 (66.7%) | 7 (70%) | 6 (60%) | 42 (63.6%) |
| Steroid usage, n (%) yes | 2 (28.6%) | 5 (33.3%) | 2 (8.3%) | 0 (0%) | 1 (10%) | 10 (15%) |
| Number of prior chemo regimens median (range) | 3 (1–7) | 3 (2–5) | 1 (1–3) | 3 (1–6) | 1 (1–2) | 1 (1–7) |
| Prior chemo regimens | ||||||
| TMZ | 7 (100%) | 15 (100%) | 23 (95.8%) | 10 (100%) | 10 (100%) | 65 (98.5%) |
| CCNU/BCNU/PCV | 3 (42.8%) | 15 (100%) | 1 (4.2%) | 2 (20%) | 0 (0%) | 21 (31.8%) |
| Immunotherapy | 0 (0%) | 1 (6.7%) | 0 (0%) | 3 (30%) | 0 (0%) | 4 (6.1%) |
Abbreviations: ASM, anti-seizure medication; BCNU, Carmustine; CCNU, Lomustine; KPS, Karnofsky Performance Score; PCV, Procarbazine, Lomustine, and Vincristine; RT, radiation therapy; TMZ, temozolomide.
Phase I Results
The phase I sub-study enrolled 7 subjects at dose level 1 (temozolomide 20 mg daily); one was replaced for non-compliance. One patient experienced DLT with grade 4 neutropenia and thrombocytopenia. Consequently, the phase II dose was determined to be temozolomide 20 mg daily with pamiparib 60 mg twice daily. The median number of 4-week cycles was 2 (range, 1–14). The median PFS was 14.6 months (95% CI, 0.9–41 months) and the median OS was 31.2 months (95%CI: 1.2–41.3 months).
Pharmacokinetic Analysis
Total and unbound concentrations of pamiparib were quantitated in non-enhancing and enhancing tumor samples in 7 and 5 patients, respectively (Figure 2A). Median unbound drug concentrations in non-enhancing and enhancing tumors were 0.188 (range, 0.135–1.133) and 0.209 (0.117–0.731) μmol/L (or μmol/kg). The in vitro IC50 of pamiparib for PARP1/2 inhibition is ~1 nM.23 Target engagement ratio (TER),24 defined as the ratio of unbound drug tumor concentration to the in vitro (IC50) for target inhibition, serves as a crude indicator for potential intracranial activity. Median TER was estimated to be 188 (range, 135–1133) and 209 (range, 117–731) in the non-enhancing and enhancing tumors, respectively.
Figure 2.
Pamiparib penetration into brain tumors in patients with gliomas. (A) Total and unbound concentrations of pamiparib in the contrast-enhancing and non-enhancing tumor regions. (B) The extent of pamiparib tumor penetration, assessed by the total or unbound tumor-to-plasma concentration ratio (Kp or Kp, uu) in the contrast-enhancing and non-enhancing tumor regions. (C) Fraction unbound (Fu) of pamiparib in the plasma, contrast-enhancing and non-enhancing tumor regions. Contrast-enhancing tumor samples were obtained from 5 patients and non-enhancing tumor samples were obtained from 7 patients. The symbols (○▲×) represent data from individual patient samples, and the lines present the median values.
The extent of drug tumor penetration is assessed as the total or unbound drug tumor-to-plasma partition ratio (Kp or Kp, uu). Pamiparib exhibited similar extent of drug penetration into the non-enhancing and enhancing tumor regions (Figure 2B), with the respective median Kp of 0.23 (range, 0.17–0.49) and 0.39 (range, 0.19–0.74), and the respective median Kp, uu of 0.92 (range, 0.38–1.52) and 0.98 (range, 0.42–1.77). Interestingly, pamiparib showed a higher binding to plasma proteins (median unbound fraction in plasma [Fu, plasma], 0.031) than to tumor tissue (median unbound fraction [Fu, tumor] of 0.083 and 0.076 in non-enhancing and enhancing tumors, respectively) (Figure 2C), explaining pamiparib’s higher Kp, uu than Kp, where Kp, uu = Kp × Fu, tumor/Fu, plasma.
Phase II Results
In Arm A, (multiple prior lines of alkylator chemotherapy), none of the 15 patients had a confirmed partial or complete response. Arm B (a single chemotherapy regimen) had a single centrally confirmed partial response among 24 patients. The responder was a 45 year old woman 5 years out from radiation and temozolomide for an incompletely resected grade 3 oligodendroglioma. She had a partial response after 2 cycles of study treatment but progressed after 12 cycles. Consequently, neither arm met the threshold for advancing to the second stage of accrual. Table 3 summarizes the best response for each arm; any locally determined partial or complete response was centrally reviewed, and the only confirmed partial response was a single Arm B subject. The best radiographic response in Arm A was 1 unconfirmed complete response, 8 subjects with stable disease, and 5 subjects with progressive disease. For Arm B, these figures were 1 unconfirmed complete response, 2 partial responses (1 centrally confirmed), and 17 patients with stable disease. The duration of the response was 41 weeks. The median number of cycles for Arm A and B, respectively, were 2 (range, 1–20) and 7 (range, 1–32). The median progression-free survival (PFS) for Arm A was 5.9 months (95% CI, 1.2–14.8 months), and for Arm B was 9.7 months (95% CI, 5.7–21.7 months). The median overall survival in Arm A was 26.1 months (95% CI, 6.0 months-not reached); in Arm B the median OS was not reached (95% CI, 22.7 months—not reached). Figure 3 shows the swimmer plot for subjects on Arms A and B. Figure S1 shows OS and PFS for Arm A (Figure S1a) and Arm B (Figure S1b).
Table 3.
Best response by treatment arm
| Best Treatment Response | Phase I n = 7 | Arm A n = 15 | Arm B n = 24 | Grade 4 n = 10 | Surgical n = 10 | Central Review Confirmation | Overall Response n = 66 (%) |
|---|---|---|---|---|---|---|---|
|
| |||||||
| Complete Response | 1 | 1 | 0 | 0 (0%) | |||
| Partial Response | 1 | 2 | 1 | 1 (1.5%) | |||
| Stable Disease | 2 | 8 | 17 | 3 | 3 | 37 (56%) | |
| Progressive Disease | 2 | 5 | 1 | 5 | 2 | 15 (22.7%) | |
| Unassessable | 2 | 1 | 3 | 2 | 5 | 13 (19.7%) | |
Figure 3.
Swimmer Plot for Phase II Arms A and B.
Grade 4 Results
In the exploratory grade 4 arm, the best response among 10 patients was stable disease (3 patients). The median number of cycles was 2 (range, 1–36) and the median PFS was 2.9 months (95% CI, 1.6 months—not reached). The median OS was 15.5 months (95% CI, 5.5—not reached).
While PFS-6 was not a study endpoint, 30 of the 66 subjects (45%) were free of progression at 6 months.
Toxicity
Common terminology criteria for adverse events grade 3 and higher toxicities at least possibly attributable to treatment are noted in Table 4, and those of any grade occurring in ≥10% in Table 5. Myelosuppression and gastrointestinal distress were anticipated side effects; nausea was common (45%) but invariably ≤grade 2, though 1 patient had grade 3 vomiting. There was substantial cumulative hematologic toxicity with subsequent cycles. Grade 3 or higher anemia and neutropenia ultimately affected 24% and 33% of patients, respectively. The median time to development of any grade hematologic toxicity was 1 cycle (range 1–2.5). The median time to grade 2 anemia was 2 cycles. Of the 15 patients developing grade 3 anemia, 2 developed after the first cycle, 8 after the second cycle, 4 after the third cycle, and 1 after the 16th cycle. Thirteen of 66 patients (19.7%) came off study for adverse events (Table S1). Three patients (4.6%) each went off treatment for patient preference, physician decision, or non-compliance. Slightly more than half (37 patients, 56.0%) discontinued treatment due to progressive disease.
Table 4.
Summary of AEs Grades 3–4 with attribution of possible, probable, and definite to the intervention treatment for all patients and all treatment arms combined
| Adverse Events with All patients combined: n (% of patients): | Grade 3–4 n = 66 |
|---|---|
|
| |
| Agitation | 1 (2%) |
| Alanine aminotransferase increased | 2 (3%) |
| Anemia | 16 (24%) |
| Anorexia | 1 (2%) |
| Aspartate aminotransferase increased | 1 (2%) |
| Fatigue | 1 (2%) |
| Headache | 1 (2%) |
| Leukocytosis | 1 (2%) |
| Lymphocyte count decreased | 10 (15%) |
| Neutrophil count decreased | 22 (33%) |
| Platelet count decreased | 11 (17%) |
| Renal calculi | 1 (2%) |
| Vomiting | 1 (2%) |
| White blood cell decreased | 19 (29%) |
Table 5.
Summary of all AEs among 10% or more patients
| Adverse Events with All patients combined: n (% of patients): | Any Grade n = 66 |
|---|---|
|
| |
| Alanine aminotransferase increased | 14 (21%) |
| Anemia | 49 (74%) |
| Anorexia | 8 (12%) |
| Aspartate aminotransferase increased | 9 (14%) |
| Constipation | 9 (14%) |
| Dizziness | 8 (12%) |
| Fatigue | 34 (52%) |
| Hypophosphatemia | 8 (12%) |
| Lymphocyte count decreased | 29 (44%) |
| Nausea | 30 (45%) |
| Neutrophil count decreased | 40 (61%) |
| Platelet count decreased | 40 (61%) |
| Vomiting | 8 (12%) |
| White blood cell decreased | 44 (67%) |
The AEs were with attribution of possible, probable, and definite to the intervention treatment for all patients and all treatment arms combined.
Discussion
Preclinical promise has led to several studies of PARP inhibition in IDHmt gliomas. A phase II study of single agent olaparib in recurrent IDHmt gliomas, while failing to meet its primary endpoint, reported 1 CR and 1 PR.25 The authors hypothesized poor blood-brain barrier penetration26 likely contributed to the negative results. A second study in this population showed no responses.27 Ongoing PARP inhibitor studies in this population include NCT05188508 (olaparib, temozolomide, and pembrolizumab), NCT05406700 (niraparib), and NCT03749187 (pamiparib and temozolomide in the pediatric population). This topic was recently and comprehensively reviewed elsewhere.28
Poor penetration of the blood-brain barrier is a contributor to the generally low chemosensitivity of gliomas.29 Our study demonstrated that pamiparib penetrated the blood-tumor barrier well with similar concentrations achieved in both non-enhancing and enhancing glioma. As a TER ≥1 theoretically results in ≥50% target inhibition, the median TER of approximately 200 suggests sufficient PARP inhibition.
Based upon prior experience with the pamiparib/temozolomide combination, we anticipated hematologic toxicity to represent the DLT of our combination.16,17 In this study, based upon DLT during the !rst cycle, we identified temozolomide 20 mg daily with pamiparib 60 mg b.i.d. as the MTD for the phase II components. It is possible that the low concentration of temozolomide this dose achieved could be insufficient to induce synergistic tumor cell-killing with PARP inhibitors.
For the phase II sub-studies, we selected radiographic response as the primary endpoint given the lack of published norms for PFS in recurrent IDHmt gliomas. The paucity of confirmed responses led to accrual stop following the first stage. Interestingly, however, the median PFS of 5.9 months in the heavily pretreated Arm A and 9.7 months in Arm B suggest a subset of patients may have benefited from protocol therapy.
Our phase I design failed to predict the cumulative hematologic toxicity experienced by a substantial minority of study participants. Even with the option of dose reduction, one-third of patients without known disease progression discontinued treatment for adverse events, patient or physician preference, or non-compliance. While we lack additional details regarding patient or physician reasoning for study discontinuation, fatigue (seen in 52%) and nausea (45%) may have contributed. Grade 3 anemia, uncommon with most glioma therapies, affected 24% of patients. This unanticipated outcome underlines the missed opportunity of incorporation of quality of life measures into this trial.
In conclusion, while pamiparib effectively reached glioma tissue, in combination with metronomic low-dose temozolomide it failed to produce RANO imaging responses. There are no established benchmarks for PFS for salvage therapy in patients with IDHmt gliomas that have progressed following radiation and alkylating chemotherapy. Some patients achieved prolonged disease stabilization, suggesting possible activity, but cumulative hematotoxicity limited this regimen’s tolerability.
Why was activity with this combination limited? Virtually all patients on ABTC 1801 had previously received temozolomide, predisposing to a hypermutated phenotype that may have contributed to therapeutic resistance. Given pamiparib’s excellent penetration into tumor, in contrast to olaparib, insufficient drug delivery does not explain the lack of response. Conceivably a lower dose of pamiparib with a higher temozolomide dose would have improved tolerability and efficacy. Alternatively, it is possible intermittent administration of a higher dose of TMZ would have yielded better results. An ongoing study (NCT05188508) explores whether the addition of the immune checkpoint inhibitor pembrolizumab to olaparib and temozolomide benefits patients with recurrent IDHmt gliomas. Future studies testing PARP1-selective PARP inhibitors that convey less hematologic toxicity, especially those with enhanced CNS penetration,30 including molecules devoid of trapping activity, may demonstrate more robust efficacy in combination with TMZ against IDH mutant gliomas.31
Supplementary Material
Supplementary material is available online at Neuro-Oncology (https://academic.oup.com/neuro-oncology).
Importance of the study.
This study translated the laboratory discovery that 2-HG, the product of the IDH mutation, creates a BRCA-like defect in homologous recombination that renders IDHmt cells sensitive to PARP inhibition. We demonstrated that the PARP inhibitor pamiparib achieves excellent penetration of the blood-brain barrier in both enhancing and non-enhancing tumor. We found that the combination of pamiparib and very low dose temozolomide was ineffective in shrinking tumors though it was associated with prolonged disease stabilization in some patients. The cumulative hematologic toxicity limited tolerability, suggesting non-PARP-trapping PARP inhibitors may be worthy of further study.
Key Points.
Pamiparib demonstrates good penetration of the blood-brain barrier.
Although some patients achieved prolonged disease stabilization, radiographic responses were rare.
Cumulative hematologic toxicity not anticipated from our phase I study, particularly anemia, limited long-term tolerability for many patients.
Acknowledgements
The Phase I results were presented at the 2021 Society for Neuro-Oncology Annual Meeting, and the Phase II results at the 2023 Society for Neuro-Oncology Annual Meeting.
Funding
This study was supported by the Gateway for Cancer Research and Adult Brain Tumor Consortium UM1CA137443 grants.
Footnotes
Conflict of Interest Statement
None (ICJME forms available upon request).
Contributor Information
David Schiff, Division of Neuro-Oncology, University of Virginia School of Medicine, Charlottesville.
Xiaobu Ye, Department of Neurosurgery, Johns Hopkins School of Medicine, Baltimore.
Jing Li, Department of Oncology, Wayne State University, Detroit.
Benjamin M. Ellingson, Department of Radiological Sciences, David Geffen School of Medicine, UCLA, Los Angeles
Patrick Y. Wen, Neuro-Oncology Center, Dana-Farber Cancer Institute, Boston
Tobias Walbert, Hermelin Brain Tumor Center, Henry Ford Hospital, Detroit.
Jian Campian, Department of Neurology, Washington University School of Medicine, St Louis.
Louis B Nabors, Department of Neurology, University of Alabama at Birmingham, Birmingham.
Byram H. Ozer, Department of Neurology, Johns Hopkins School of Medicine, Baltimore
Arati Desai, Department of Medical Oncology, University of Pennsylvania Perelman School of Medicine, Philadelphia.
Antonio Omuro, Department of Neurology, Yale School of Medicine, New Haven.
Serena Desideri, Department of Oncology, Johns Hopkins School of Medicine, Baltimore.
Neeraja Danda, Department of Oncology, Johns Hopkins School of Medicine, Baltimore.
Stuart Grossman, Department of Oncology, Johns Hopkins School of Medicine, Baltimore.
Ranjit S. Bindra, Department of Therapeutic Radiology, Yale School of Medicine, New Haven
Data Availability
Data will be made available upon reasonable request.
References
- 1.Yan H, Parsons DW, Jin G, et al. IDH1 and IDH2 mutations in gliomas. N Engl J Med. 2009;360:765–773. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Schiff D. Low-Grade gliomas: a new mutation, new targeted therapy, and many questions. Neurology. 2024;103:e209688. [DOI] [PubMed] [Google Scholar]
- 3.Mellinghoff IK, van den Bent MJ, Blumenthal DT, et al. ; INDIGO Trial Investigators. Vorasidenib in IDH1- or IDH2-mutant low-grade glioma. N Engl J Med. 2023;389:589–601. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Mellinghoff IK, Ellingson BM, Touat M, et al. Ivosidenib in isocitrate dehydrogenase 1-Mutated advanced glioma. J Clin Oncol. 2020;38:3398–3406. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Miller JJ, Gonzalez Castro LN, McBrayer S, et al. Isocitrate dehydrogenase (IDH) mutant gliomas: a society for Neuro-Oncology (SNO) consensus review on diagnosis, management, and future directions. Neuro Oncol. 2023;25:4–25. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Yu Y, Villanueva-Meyer J, Grimmer MR, et al. Temozolomide-induced hypermutation is associated with distant recurrence and reduced survival after high-grade transformation of low-grade IDH-mutant gliomas. Neuro Oncol. 2021;23:1872–1884. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Sulkowski PL, Corso CD, Robinson ND, et al. 2-Hydroxyglutarate produced by neomorphic IDH mutations suppresses homologous recombination and induces PARP inhibitor sensitivity. Sci Transl Med. 2017;9:eaal2463. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Sulkowski PL, Oeck S, Dow J, et al. Oncometabolites suppress DNA repair by disrupting local chromatin signalling. Nature. 2020;582:586–591. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Sulkowski PL, Sundaram RK, Oeck S, et al. Krebs-cycle-deficient hereditary cancer syndromes are defined by defects in homologous-recombination DNA repair. Nat Genet. 2018;50:1086–1092. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Lu Y, Kwintkiewicz J, Liu Y, et al. Chemosensitivity of IDH1-Mutated gliomas due to an impairment in PARP1-mediated DNA repair. Cancer Res. 2017;77:1709–1718. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Molenaar RJ, Radivoyevitch T, Nagata Y, et al. IDH1/2 mutations sensitize acute myeloid leukemia to PARP inhibition and this is reversed by IDH1/2-Mutant inhibitors. Clin Cancer Res. 2018;24:1705–1715. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Schvartzman JM, Forsyth G, Walch H, et al. Oncogenic IDH mutations increase heterochromatin-related replication stress without impacting homologous recombination. Mol Cell. 2023;83:2347–2356.e8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Petropoulos M, Karamichali A, Rossetti GG, et al. Transcription-replication conflicts underlie sensitivity to PARP inhibitors. Nature. 2024;628:433–441. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Murai J, Zhang Y, Morris J, et al. Rationale for poly(ADP-ribose) polymerase (PARP) inhibitors in combination therapy with camptothecins or temozolomide based on PARP trapping versus catalytic inhibition. J Pharmacol Exp Ther. 2014;349:408–416. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Xiong Y, Guo Y, Liu Y, et al. Pamiparib is a potent and selective PARP inhibitor with unique potential for the treatment of brain tumor. Neoplasia. 2020;22:431–440. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Stradella A, Johnson M, Goel S, et al. Phase 1b study to assess the safety, tolerability, and clinical activity of pamiparib in combination with temozolomide in patients with locally advanced or metastatic solid tumors. Cancer Med. 2024;13:e7385. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Shih KS,D, Kim L, Battiste J, et al. ACTR-30. Phase 1b/2 study to assess the clinical effects of pamiparib (BGB-290) in combination with radiation therapy (RT) and/or temozolomide (TMZ) in patients with newly diagnosed or recurrent/refractory glioblastoma (GBM). Society for Neuro-Oncology Annual Meeting. 2018;20:vi17–vi18. [Google Scholar]
- 18.Jiang J, Bao X, Yue Y, Schiff D, Bindra R, Li J. Development and validation of a liquid chromatography coupled with tandem mass spectrometry method for determining total and unbound pamiparib in human plasma and brain tumors. Biomed Chromatogr. 2022;36:e5478. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Wen PY, Macdonald DR, Reardon DA, et al. Updated response assessment criteria for high-grade gliomas: response assessment in neuro-oncology working group. J Clin Oncol. 2010;28:1963–1972. [DOI] [PubMed] [Google Scholar]
- 20.van den Bent M, Wefel J, Schiff D, et al. Response assessment in neuro-oncology (a report of the RANO group): assessment of outcome in trials of diffuse low-grade gliomas. Lancet Oncol. 2011;12:583–593. [DOI] [PubMed] [Google Scholar]
- 21.Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc. 1958;53:457–481. [Google Scholar]
- 22.Brookmeyer R, Crowley J. A confidence interval for the median survival time. Biometrics. 1982;38:29–41. [Google Scholar]
- 23.Wang H, Ren B, Liu Y, et al. Discovery of pamiparib (BGB-290), a potent and selective poly (ADP-ribose) polymerase (PARP) inhibitor in clinical development. J Med Chem. 2020;63:15541–15563. [DOI] [PubMed] [Google Scholar]
- 24.Li J, Jiang J, Bao X, et al. Mechanistic modeling of Central Nervous System pharmacokinetics and target engagement of HER2 tyrosine kinase inhibitors to inform treatment of breast cancer brain metastases. Clin Cancer Res. 2022;28:3329–3341. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Esparragosa Vazquez I, Sanson M, Chinot OL, et al. Olaparib in recurrent isocitrate dehydrogenase mutant high-grade glioma: a phase 2 multicenter study of the POLA network. Neurooncol Adv. 2024;6:vdae078. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Hanna C, Kurian KM, William K, et al. Pharmacokinetics, safety, and tolerability of olaparib and temozolomide for recurrent glioblastoma: results of the phase I OPARATIC trial. Neuro Oncol. 2020;22:1840–1850. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Fanucci K, Pilat MJ, Shyr D, et al. Multicenter phase II trial of the PARP inhibitor olaparib in recurrent IDH1- and IDH2-mutant glioma. Cancer Res Commun. 2023;3:192–201. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Cella E, Bosio A, Persico P, et al. PARP inhibitors in gliomas: mechanisms of action, current trends and future perspectives. Cancer Treat Rev. 2024;131:102850. [DOI] [PubMed] [Google Scholar]
- 29.Sarkaria JN, Hu LS, Parney IF, et al. Is the blood-brain barrier really disrupted in all glioblastomas? A critical assessment of existing clinical data. Neuro Oncol. 2018;20:184–191. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Staniszewska AD, Pilger D, Gill SJ, et al. Preclinical characterization of AZD9574, a blood-brain barrier penetrant inhibitor of PARP1. Clin Cancer Res. 2024;30:1338–1351. [DOI] [PubMed] [Google Scholar]
- 31.Geurts M, Gramatzki D, Merler S, et al. Abstract LB_A12: initial results from 2 phase I studies of NMS-03305293, a selective PARP1 inhibitor. Mol Cancer Ther. 2023;22:LB_A12. [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Data will be made available upon reasonable request.