Recurrent H3 Research Articles

NCOG-07. SAFETY OF ONC201 TREATMENT IN PATIENTS WITH PREVIOUSLY TREATED H3 K27M-MUTANT GLIOMA: RESULTS FROM ONC028, AN ONGOING COMPASSIONATE USE PROGRAM

Abstract PURPOSE H3 K27M-mutant diffuse midline glioma (DMG) has a poor prognosis, with a median overall survival of 1 year and no proven treatments beyond radiotherapy. ONC201 (dordaviprone), a first-in-class imipridone, previously demonstrated durable responses in patients with recurrent H3 K27M-mutant DMG. The present analysis reports safety data from ONC028, a compassionate use program (CUP) for patients with previously treated H3 K27M-mutant glioma. METHODS ONC028 is an expanded access program that provides ONC201 monotherapy to patients with H3 K27M-mutant glioma. Eligible patients are adults or pediatric patients with H3 K27M-mutant glioma, prior therapy including at least radiation, and adequate organ function. Patients must be ineligible for enrollment in other ongoing clinical studies evaluating ONC201. ONC201 is administered orally once weekly at 625 mg or dose-equivalent based on body weight for pediatric patients. MRIs are performed every two 28-day cycles for the initial 12 months, thereafter every 2-3 cycles. Treatment continues until disease progression, unacceptable adverse events (AEs), or withdrawal of consent. Treatment may continue post-progression at the discretion of the treating physician. Bevacizumab use for radiation necrosis or symptom management is acceptable. AEs are assessed by NCI CTCAE v5.0 criteria. RESULTS The 422 enrolled patients include 191 adults (45.2%) and 231 children (54.7%); 87.1% had a performance score ≥60 and the most common tumor location was the pons (57.3%). Any-causality Grade ≥3 AEs have been reported in 200 patients (47.4%). Grade ≥3 treatment-related AEs (TR-AEs) were reported in 19 (4.5%) patients; the commonest was decreased lymphocyte count, occurring in 3 (0.7%) patients. Any-causality serious AEs (SAE) were reported in 43.1% (n=182) patients. There were 10 (2.4%) patients with reports of treatment-related SAEs, the commonest were ascites, nausea, vomiting, and seizure, each occurring in 2 patients (0.2%). CONCLUSIONS ONC028 provides expanded access to ONC201 treatment for patients with previously treated H3 K27M-mutant glioma who are ineligible for enrollment in ongoing trials of ONC201. ONC201 treatment remains well-tolerated.

EXTH-37. ALLOSTERIC MITOCHONDRIAL CLPP AGONIST ONC206 ALTERS STRESS RESPONSE, METABOLIC AND EPIGENETIC PROFILES TO ELICIT ANTI-CANCER EFFICACY IN HIGH-GRADE GLIOMAS

Abstract Dordaviprone (ONC201) is a dopamine receptor D2/3 (DRD2/3) antagonist and allosteric ClpP agonist that has demonstrated durable responses in recurrent H3 K27M-mutant glioma patients. ONC206, a derivative of ONC201 with nanomolar potency that also targets DRD2/3 and ClpP, is in Phase I trials for central nervous system tumors. We characterized target relevance, human ClpP binding, in vitro efficacy, downstream signaling and response determinants associated with ONC206 in high-grade gliomas (HGG). Co-crystallization with ClpP that assembles as a tetradecameric complex revealed an allosteric ligand interaction with distinctions in the ONC206-ClpP overall conformation relative to the ONC201-bound or apo complex, including an increase in the resolved pore size and decreased complex height. ONC206 was more potent than ONC201 in cell-free human ClpP casein/peptide degradation assays). Accordingly, HGG cell lines and patient-derived cells exhibited increased sensitivity to ONC206 relative to ONC201 and generally required a shorter duration of exposure. Unlike DRD2, CRISPR-mediated ClpP knockout in HGG cell lines blunted ONC206 sensitivity. Metabolomics revealed elevated α-ketoglutaric acid, α-hydroxyglutaric acid, and glutaric acid. Proteomics indicated inhibition of multiple mitochondrial pathways including OXPHOS and TCA cycle. TCA cycle enzyme α-ketoglutarate dehydrogenase (KGDH) was inhibited in a ClpP-dependent manner. Western blotting showed downregulated ClpX, and upregulated ATF4/CHOP, H3 K27me3 and H3 K36me3. Next-generation sequencing of acquired resistance cell lines, which exhibited cross-resistance to ONC201 and ONC206, revealed diverse ClpP missense mutations distributed across various monomer interfaces. Both intrinsic and well as ONC206-stimulated ClpP proteolysis were mitigated by these mutations, which was congruent with an absence of the tetradecameric assembly. These ClpP mutations recapitulated resistance and, conversely, wildtype ClpP overexpression restored sensitivity. Thus, allosteric ClpP agonism is a key determinant of ONC206 activity in HGG by altering stress response, metabolic, and epigenetic profiles.

CTNI-54. RESPONSE BY RANO2.0 CRITERIA IN ONC201 (DORDAVIPRONE)-TREATED PATIENTS WITH RECURRENT H3 K27M-MUTANT DIFFUSE MIDLINE GLIOMA

Abstract PURPOSE Response assessment in neuro-oncology (RANO) 2.0 is a recently established response assessment for glial tumors that is agnostic to WHO grade. This updated response assessment includes quantitative evaluation of enhancing and non-enhancing disease and recommendations to account for pseudoprogression. This analysis evaluated response to ONC201 by RANO2.0 in recurrent, H3K27M-mutant diffuse midline glioma (DMG), which was previously assessed by RANO-HGG/RANO-LGG (Arrillaga-Romany, JCO, 2024). METHODS Fifty patients with recurrent H3 K27M-mutant DMG were selected based on criteria designed with input from regulators to evaluate objective response to ONC201 monotherapy. Patients had contrast-enhancing, measurable, progressive disease, and were ≥90 days from radiation, with adequate washout from other anti-cancer therapies. Diffuse intrinsic pontine glioma, primary spinal tumors, leptomeningeal spread, or CSF-disseminating tumors were ineligible. Response was assessed using RANO2.0 criteria by two-party blinded independent central review. RESULTS Best response by RANO2.0 included one complete response (CR, 2.0%), nine partial responses (PR, 18.0%), and four minor responses (MR, 8.0%). Overall response rate (CR+PR+MR) by RANO2.0 criteria was 28.0% (n=14; 95%CI, 16-42). Disease control rate (CR+PR+MR+SD) was 40.0% (n=20; 26-55). Median duration of response was 10.4 months (95%CI, 7.4-15.4). Median time to response was 4.6 months (range, 1.6-15.9). 6- and 12-month progression free survival rates were 32.0% (95%CI, 19-47) and 27.0% (15-41), respectively. RANO2.0 response as a time-varying covariate was significantly associated with overall survival in a multivariate analysis considering baseline performance score, multiple enhancing target lesions, enhancing tumor size, and number of prior recurrences (HR [95%CI], 0.22 [0.08–0.58]; p=0.0023). Patients with an objective response by RANO2.0 were more likely to be alive at 12 (100%) and 24 (86%) months, have reduced corticosteroid usage, and have improved performance status. Responses occurred at a higher rate among patients with a higher performance score and fewer target lesions. CONCLUSIONS RANO2.0 assessment of ONC201-treated patients suggests durable and clinically meaningful efficacy in recurrent H3K27M-mutant DMG.

Selective DRD2 antagonist and ClpP agonist ONC201 in a recurrent non-midline H3 K27M-mutant glioma cohort.

Diffuse midline glioma, H3 K27-altered (H3 K27M-altered DMG) are invariably lethal, disproportionately affecting the young and without effective treatment besides radiotherapy. The 2016 World Health Organization (WHO) Central Nervous System (CNS) Tumors Classification defined H3 K27M mutations as pathognomonic but restricted diagnosis to diffuse gliomas involving midline structures by 2018. Dordaviprone (ONC201) is an oral investigational small molecule, DRD2 antagonist, and ClpP agonist associated with durable responses in recurrent H3 K27M-mutant DMG. Activity of ONC201 in non-midline H3 K27M-mutant diffuse gliomas has not been reported. Patients with recurrent non-midline H3 K27M-mutant diffuse gliomas treated with ONC201 were enrolled in 5 trials. Eligibility included measurable disease by Response Assessment in Neuro-Oncology (RANO) high-grade glioma, Karnofsky/Lansky performance score ≥60, and ≥90 days from radiation. The primary endpoint was overall response rate (ORR). Five patients with cerebral gliomas (3 frontal, 1 temporal, and 1 parietal) met inclusion. One complete and one partial response were reported by investigators. Blinded independent central review confirmed ORR by RANO criteria for 2, however, 1 deemed nonmeasurable and another stable. A responding patient also noted improved mobility and alertness. H3 K27M-mutant diffuse gliomas occasionally occur in non-midline cerebrum. ONC201 exhibits activity in H3 K27M-mutant gliomas irrespective of CNS location.

Novel insights on genetics and epigenetics as clinical targets for paediatric astrocytoma.

Paediatric and adult astrocytomas are notably different, where clinical treatments used for adults are not as effective on children with the same form of cancer and these treatments lead to adverse long-term health concerns. Integrative omics-based studies have shown the pathology and fundamental molecular characteristics differ significantly and cannot be extrapolated from the more widely studied adult disease. Recent clinical advances in our understanding of paediatric astrocytomas, with the aid of next-generation sequencing and epigenome-wide profiling, have led to the identification of key canonical mutations that vary based on the tumour location and age of onset. These driver mutations, in particular the identification of the recurrent histone H3 mutations in high-grade tumours, have confirmed the important role epigenetic dysregulations play in cancer progression. This review summarises the current updates of the classification, epidemiology, pathogenesis and clinical management of paediatric astrocytoma based on their grades and the ongoing clinical trials. It also provides novel insights on genetic and epigenetic alterations as diagnostic biomarkers, highlighting the potential of targeting these pathways as therapeutics for this devastating childhood cancer.

Epigenetic programming of pediatric high-grade glioma: Pushing beyond proof of concept to clinical benefit.

Pediatric high-grade gliomas (pHGG) are a molecularly diverse group of malignancies, each incredibly aggressive and in dire need of treatment advancements. Genomic analysis has revolutionized our understanding of these tumors, identifying biologically relevant subgroups with differing canonical mutational profiles that vary based on tumor location and age. In particular, the discovery of recurrent histone H3 mutations (H3K27M in diffuse midline glioma, H3G34R/V in hemispheric pediatric high-grade gliomas) as unique "oncohistone" drivers revealed epigenetic dysregulation as a hallmark of pediatric high-grade gliomas oncogenesis. While reversing this signature through epigenetic programming has proven effective in several pre-clinical survival models, early results from pediatric high-grade gliomas clinical trials suggest that epigenetic modifier monotherapy will likely not provide long-term disease control. In this review we summarize the genetic, epigenetic, and cellular heterogeneity of pediatric high-grade gliomas, and highlight potential paths forward for epigenetic programming in this devastating disease.

SYST-14 CLINICAL EFFICACY OF ONC201 IN RECURRENT H3 K27M-MUTANT DIFFUSE MIDLINE GLIOMA (DMG) PATIENTS

Abstract BACKGROUND H3 K27M-mutant DMG predominantly affects children and young adults; no effective therapy is known. ONC201 is a first-in-class, anti-cancer DRD2 antagonist and ClpP agonist. METHODS Fifty pediatric and adult patients with recurrent H3 K27M DMG who received oral ONC201 monotherapy in clinical trials and expanded access were selected for a planned efficacy analysis. Eligibility criteria included measurable contrast-enhancing disease by RANO-HGG criteria (excluding pontine and spinal cord tumors), KPS/LPS≥60, ≥90 days from prior radiation, and adequate washout from prior anti-cancer therapy. The primary endpoint was overall response rate (ORR) by RANO-HGG criteria. Secondary endpoints included duration of response, time to response, progression-free survival (PFS), overall survival (OS), corticosteroid response rate, performance status response rate, and ORR by RANO-LGG criteria. Radiographic endpoints were assessed by dual-reader blinded independent central review. Data cutoff was May 31, 2021. RESULTS ORR was 20.0% (95%CI, 10.0–33.7) by RANO-HGG criteria. Median duration of response was 11.2 months (95%CI, 3.8–not reached) and median time to response was 8.3 months (range, 1.9–15.9). PFS at 6 months was 35.1% (95%CI, 21.2–49.3). The ORR was 26.0% (95%CI, 14.6–40.3) by RANO-LGG criteria. Fifteen patients (30.0%; 95%CI, 17.9–44.6) achieved an objective response by RANO-HGG and/or RANO-LGG criteria. Of 15 patients receiving ≥4 mg daily dexamethasone at baseline, 7 (46.7%; 95%CI, 21.3–73.4) achieved ≥50% confirmed reduction in dose. Of 34 patients with baseline KPS/LPS <80, 7 (20.6%; 95%CI, 8.7–37.9) achieved a confirmed performance status improvement. With a median follow-up of 18.8 months, median OS was 13.7 months (95%CI, 8.0–20.3) and OS at 24 months was 34.7% (95%CI, 20.7–49.2). Twenty-five patients had serious adverse events with one possibly related to ONC201 by investigator assessment. CONCLUSIONS ONC201 monotherapy exhibits durable and clinically meaningful efficacy in recurrent H3 K27M-mutant DMG.

SYST-07 WINDOW-OF-OPPORTUNITY STUDY OF ONC201 IN PEDIATRIC PATIENTS WITH DIFFUSE INTRINSIC PONTINE GLIOMA (DIPG) AND THALAMIC GLIOMA

BACKGROUNDH3 K27M-mutant diffuse midline glioma is a universally fatal malignancy primarily affecting children and young adults; no effective systemic therapy is available. ONC201, a first-in-class imipridone, is an oral, blood-brain barrier penetrating, selective small molecule antagonist of dopamine receptor D2/3 and agonist of the mitochondrial protease ClpP. ONC201 monotherapy demonstrated durable objective responses in adults with recurrent H3 K27M-mutant glioma. This phase 1 trial will evaluate ONC201±radiotherapy (RT) in pediatric patients with H3 K27M-mutant midline glioma DIPG. METHODSThis multicenter, open-label, dose escalation and expansion phase 1 study of ONC201 is comprised of eight arms that will evaluate the recommended phase 2 dose (RP2D) of ONC201, biomarkers, and pharmacokinetics (PK) of ONC201±RT in various treatment settings (NCT03416530). Arm G previously defined the RP2D for twice-weekly ONC201 on consecutive days. Arm H, for which enrollment is ongoing, will estimate the influence of tumor location and blood-brain barrier integrity on PK and intratumoral ONC201 exposure in biopsy-eligible pediatric tumors (DIPG or contrast-enhancing thalamic glioma). Patients eligible for Arm H will be aged 2-≤19 years, ≥2 weeks from last RT administration, and have a KPS/LPS ≥50; prior confirmation of H3 K27M mutation is not required. In Arm H, single-agent ONC201 administration will occur twice-weekly on consecutive days during each 21-day cycle at the RP2D defined in Arm G. Arm H has a planned enrollment of 27 patients (DIPG, n=15; thalamic glioma, n=12), with three patients undergoing a single biopsy at each of the following time points: 1-3 h post-first dose, 22-26 h post-second dose, 1-3 h post-first dose, 6-10 h post-second dose, and 22-26 h post-second dose. The 22-26 h post-first dose biopsy in thalamic glioma was previously collected and will not be assessed in this treatment arm. Plasma for PK analysis will be collected from all patients.

Abstract 1671: Engineering new cellular models to decipher H3.3K27M mutation role in DIPGs' resistance to therapies

Abstract Among pediatric brain tumors, Diffuse Intrinsic Pontine Gliomas (DIPGs) display a particularly dismal prognosis, highlighted by their median survival lower than one year. Indeed, DIPGs’ location and infiltrative properties preclude their surgical resection. Moreover, DIPGs poorly respond to chemotherapeutic agents. In this context, the only treatment for these tumors remains palliative radiotherapy, systematically followed by tumor progression. In addition to their resistance to therapies, DIPGs are characterized by recurrent histone H3 mutations. The H3.3K27M mutation is the most frequent and results from a heterozygous single nucleotide variant in the H3F3A gene, inducing the lysine 27 substitution by a methionine. Although H3.3K27M’s driver role in DIPGs tumorigenesis is now established, its role in their chemo- and radioresistance remains unclear. Aiming to decipher the potential role of this mutation in pediatric gliomas’ resistance to therapies, we established isogenic cellular models of H3.3K27M induction and reversion.We first induced H3.3K27M mutation in three initially non-mutated supratentorial pediatric glioma cell lines. Thus, we generated models that stably expressed the dominant-negative H3.3K27M or the wild type H3.3 as controls. Complementarily, to study H3.3K27M roles in a DIPG cell context, we also developed H3.3K27M reversion models in two initially mutated DIPG cell lines by applying a gene-editing strategy based on the combinatorial use of the CRISPR/Cas9 technology and an insert.We showed that H3.3K27M induction in Res259 and KNS42 cells conferred a radioresistant phenotype to a fractionated radiotherapy schedule. Besides, we performed a screening of 80 anti-cancer drugs, which revealed a differential impact of the mutation on the drug sensitivity profiles of our three H3.3K27M-induced cell lines. These results indicate that H3.3K27M can control pediatric glioma cells’ resistance to therapies, but in a heterogeneous way depending on the cellular context. Along this line, we are currently characterizing the chemo- and radiotherapy response of our new DIPG H3.3K27M-reversed models. Altogether, our first results support a role for H3.3K27M in pediatric gliomas resistance to treatments, and our complementary models pave the way for identifying new H3.3K27M-dependent mechanisms and promising targets to sensitize DIPGs to therapies. Citation Format: Andria Rakotomalala, Paul Lewandowski, Quentin Bailleul, Clara Savary, Mélanie Arcicasa, Christine Bal, Maud Hamadou, Paul Huchedé, Audrey Restouin, Remy Castellano, Yves Collette, Audrey Vincent, Pierre-Olivier Angrand, Eric Adriaenssens, Xuefen Le Bourhis, Pierre Leblond, Marie Castets, Eddy Pasquier, Alessandro Furlan, Samuel Meignan. Engineering new cellular models to decipher H3.3K27M mutation role in DIPGs' resistance to therapies [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 1671.

Window-of-opportunity study of ONC201 in pediatric patients with diffuse intrinsic pontine glioma (DIPG) and thalamic glioma.

TPS2082 Background: H3 K27M-mutant diffuse midline glioma is a universally fatal malignancy primarily affecting children and young adults; while radiotherapy (RT) provides transient benefit, no effective systemic therapy is currently available. ONC201, a first-in-class imipridone, is an oral, blood-brain barrier penetrating, selective small molecule antagonist of dopamine receptor D2/3 and agonist of the mitochondrial protease ClpP. Previously, ONC201 monotherapy demonstrated durable objective responses in adults with recurrent H3 K27M-mutant glioma. This phase 1 trial was designed to evaluate ONC201±RT in pediatric patients with H3 K27M-mutant midline glioma DIPG. Methods: This multicenter, open-label, dose escalation and expansion phase 1 study of ONC201 is comprised of eight arms that will evaluate the recommended phase 2 dose (RP2D) of ONC201, biomarkers, and pharmacokinetics (PK) of ONC201±RT in various treatment settings. Arm G previously defined the RP2D for ONC201 administered twice weekly on consecutive days in patients with H3 K27M-mutant glioma who had completed radiotherapy. Arm H, for which enrollment is ongoing, will estimate the influence of tumor location and blood-brain barrier integrity on PK and intratumoral ONC201 exposure in biopsy-eligible pediatric tumors (DIPG or contrast-enhancing thalamic glioma). Patients eligible for Arm H will be aged 2-≤19 years, ≥2 weeks from last RT administration, and have a Karnofsky/Lansky performance score ≥50; prior confirmation of H3 K27M mutation is not required. In Arm H, single-agent ONC201 administration will occur on two consecutive days each week during each 21-day cycle at the RP2D defined in Arm G. Evidence of disease progression is not required; as such, ONC201 may be administered in the maintenance setting or for recurrent disease. Arm H has a planned enrollment of 27 patients. Each patient will undergo biopsy at a single prespecified biopsy window, which will be assigned at enrollment (Table); plasma for PK analysis will be collected from all patients at all time points shown in the Table, with additional collection pre-dose and 0.5 h post first dose. Clinical trial information: NCT03416530. [Table: see text]

The diverse landscape of histone-mutant pediatric high-grade gliomas: A narrative review

Pediatric high-grade gliomas (pHGGs) are the leading cause of tumor-related death in children, with diffuse midline gliomas representing the worst prognoses. Despite decades of clinical trials, no effective treatment has been found, and we are in desperate need of novel therapeutics. The discovery of highly recurrent histone H3 mutations in pHGGs represents a major breakthrough in our understanding of tumor initiation and development. In this review, we summarize our current knowledge of the molecular pathology of these tumors, including their genomic/epigenetic alterations, mechanism of action, and partner mutations contributing to tumor progression.

Phase I dose escalation and expansion trial of single agent ONC201 in pediatric diffuse midline gliomas following radiotherapy.

ONC201, a dopamine receptor D2 (DRD2) antagonist and caseinolytic protease P (ClpP) agonist, has induced durable tumor regressions in adults with recurrent H3 K27M-mutant glioma. We report results from the first phase I pediatric clinical trial of ONC201. This open-label, multi-center clinical trial (NCT03416530) of ONC201 for pediatric H3 K27M-mutant diffuse midline glioma (DMG) or diffuse intrinsic pontine glioma (DIPG) employed a dose-escalation and dose-expansion design. The primary endpoint was the recommended phase II dose (RP2D). A standard 3 + 3 dose escalation design was implemented. The target dose was the previously established adult RP2D (625 mg), scaled by body weight. Twenty-two pediatric patients with DMG/DIPG were treated following radiation; prior lines of systemic therapy in addition to radiation were permitted providing sufficient time had elapsed prior to study treatment. The RP2D of orally administered ONC201 in this pediatric population was determined to be the adult RP2D (625 mg), scaled by body weight; no dose-limiting toxicities (DLT) occurred. The most frequent treatment-emergent Grade 1-2 AEs were headache, nausea, vomiting, dizziness and increase in alanine aminotransferase. Pharmacokinetics were determined following the first dose: T 1/2, 8.4 h; T max, 2.1 h; C max, 2.3 µg/mL; AUC0-tlast, 16.4 hµg/mL. Median duration of treatment was 20.6 weeks (range 5.1-129). Five (22.7%) patients, all of whom initiated ONC201 following radiation and prior to recurrence, were alive at 2 years from diagnosis. The adult 625 mg weekly RP2D of ONC201 scaled by body weight was well tolerated. Further investigation of ONC201 for DMG/DIPG is warranted.

CTNI-36. SAFETY OF ONC201 ADMINISTERED TWO CONSECUTIVE DAYS PER WEEK IN PEDIATRIC H3 K27M-MUTANT GLIOMA PATIENTS

Abstract ONC201, an anti-cancer DRD2 antagonist and ClpP agonist, is in Phase II trials for adult H3 K27M-mutant diffuse midline gliomas. The recommended phase 2 dose (RP2D) of 625mg ONC201 once a week has been established as a biologically active dose that is well tolerated in adult and pediatric populations. Radiographic regressions with single agent ONC201 have been reported in recurrent H3 K27M-mutant glioma patients. In another study, twice/week dosing was explored in adult patients and deemed to be safe (no DLTs observed). This warranted exploration of twice/week dosing in pediatric patients and will be discussed in this presentation. This multi-arm, dose-escalation and dose-expansion trial (ONC014; NCT03416530) determined the pediatric RP2D of ONC201 administered once per week and twice per week on two consecutive days. ONC201 was orally administered and scaled by body weight. Dose escalation was performed by a 3 + 3 design beginning with two 125mg capsules less than the adult RP2D equivalent. Twelve children (8 females; 4 males) with H3 K27M-mutant gliomas (pons: 8; thalamus: 2; spinal cord: 2) aged 4-19 years have been treated post-radiation: 3 at dose level -1; 3 at dose level 1; 6 as part of the dose expansion cohort on dose level 2. Median KPS was 90 (range 70-100). One treatment cycle was 21 days (6 doses), which also defined the DLT window. Patients were on-treatment for a median length of 4 cycles (range: 2-11). Twice weekly dosing of ONC201 was tolerated well, as observed with weekly dosing, with no instance of DLT. A total of 4 SAEs were reported, none of which were related to the study drug. The most common AEs (regardless of relatedness) included headache, facial nerve disorder, abducens nerve disorder, nausea, fatigue and ataxia. Additional safety data, PK, and clinical outcomes from this arm will be reported.

DIPG-52. PHASE I CLINICAL TRIAL OF ONC201 IN PEDIATRIC H3 K27M-MUTANT GLIOMA OR NEWLY DIAGNOSED DIPG

H3 K27M-mutant gliomas often manifest as midline gliomas, have a dismal prognosis, and have no effective treatments. ONC201 efficacy has been shown in high-grade glioma preclinical models and durable responses with single agent ONC201 have been reported in adults with recurrent H3 K27M-mutant gliomas. These observations led to a Phase I pediatric clinical trial of ONC201 dosed by body weight. This multi-center, open-label, 3 + 3 dose-escalation and dose-expansion clinical trial (NCT03416530) for H3 K27M-mutant glioma or non-biopsied DIPG has 6 arms: arms A and E determine the RP2D in pediatric post-radiation (recurrent or not-recurrent) H3 K27M-mutant glioma patients with ONC201 administered as an oral capsule as well as a liquid formulation, respectively. Both arms have completed accrual. The study is currently enrolling newly diagnosed DIPG patients to determine the RP2D for ONC201 in combination with radiation (arm B). Dedicated assessment of intratumoral ONC201 concentrations in midline gliomas patients (arm C) and the effects of ONC201 in H3K27M DNA levels in circulating CSF (arm D) are currently enrolling patients. ONC201 as a single agent in patients with progressive H3K27M mutant tumors following irradiation (excluding DIPG/spinal cord tumors) was recently opened (arm F). Once the RP2D is confirmed, there is a dose-expansion cohort to confirm the safety, radiographic efficacy and survival with ONC201. The primary endpoints of arms A, B, and E have been established with the RP2D of 625mg scaled by body weight as a capsule or liquid formulation administered alone or in combination with radiation without incidence of dose-limiting toxicity.

Advances in the classification and treatment of pediatric brain tumors.

Recent genetic and molecular findings have impacted the diagnosis, prognosis, and in some instances, treatment strategies for children with pediatric central nervous system tumors. Herein, we review the most up-to-date molecular findings and how they have impacted tumor classification and clinical care. It is now recognized that aberrations of the mitogen-activated protein kinase pathway are present in the majority of pediatric low-grade glioma. Also, there has been the identification of recurrent histone H3 K27M mutations in diffuse intrinsic pontine and other midline gliomas. Medulloblastoma is now divided into four molecular subgroups with distinct characteristics and prognoses. The classification of other unique embryonal tumors is also highlighted. Finally, we present the newest classification of ependymoma; supratentorial ependymomas comprise two subtypes based on expression of the chromosome 11 Open Reading Frame 95-reticuloendotheliosis Viral Oncogene Homolog A or yes-associated protein 1 fusion, whereas posterior fossa ependymomas are divided into two distinct molecular subgroups, posterior fossa-A and posterior fossa-B. These advances in the molecular classification of pediatric central nervous system tumors have not only assisted in diagnoses, but they have led to a new era of tumor classification and prognostication. They also have served as drivers for the evaluation of new targeted therapies based upon molecular aberrations with the hope for improved survival outcomes for our patients.

BIOM-47. LONGITUDINAL MONITORING OF PLASMA H3K27M IN DIFFUSE MIDLINE GLIOMA PATIENTS TREATED WITH ONC201

Abstract Diffuse midline glioma, H3 K27M-mutant (DMG) is a 2016 WHO Grade IV glioma that has no established treatment beyond first-line radiation. ONC201 is an investigational small molecule that has been shown to be clinically active in recurrent DMG clinical trials. While biopsies of DMG are sometimes feasible, many patients defer secondary to complication risk. MR scans have many limitations in monitoring DMG progression, including distinguishing pseudoprogression and pseudoresponse and measuring diffuse lesions that often do not contrast enhance. Digital droplet PCR (ddPCR) is capable of sensitively detecting and quantifying the allelic frequency of circulating-tumor DNA (ctDNA) fragments against a backdrop of non-tumor DNA. Using sequence-specific probes for H3F3A (H3.3 K27M) and HIST1H3B (H3.1 K27M) ddPCR detects very low frequency variants and provides an assessment of mutational burden. A pilot cohort of 5 patients treated with ONC201 who had a range of outcomes were assessed with serial ctDNA analyses. Two patients with immediately progressive disease had a concordant H3 K27M ctDNA increase that precedes radiographic detection by 4 weeks. Two patients with >50% tumor regressions while on ONC201 had concordant H3 K27M ctDNA burden at the onset of response and subsequent radiographic progression was preceded by increases in ctDNA 8–16 weeks prior. One patient who had prolonged stable disease had decreased H3 K27M ctDNA burden over time. Upon radiographic progression, the addition of bevacizumab with ONC201 caused a radiographic pseudoresponse, however H3 K27M ctDNA remained stable. These pilot results suggest H3 K27M ctDNA may be a sensitive and accurate biomarker of disease burden. Longitudinal evaluation of H3 K27M ctDNA in a cohort of 34 recurrent contrast-enhancing H3 K27M-mutant glioma patients while on ONC201 will be reported. Primary tumor locations range across the thalamus, cerebellum, basal ganglia, temporal lobe, and midbrain; median age is 31 years old (range 20–70).

CTNI-37. EFFICACY OF ONC201 IN PATIENTS WITH ONC201 FOR RECURRENT H3 K27M-MUTANT DIFFUSE MIDLINE GLIOMA

Abstract H3 K27M-mutant diffuse midline gliomas (DMG) have a dismal prognosis. We report an integrated analysis for ONC201, a DRD2 antagonist and ClpP agonist, in patients with recurrent H3 K27M DMG administered as monotherapy in 3 clinical studies (NCT03295396, n=20; NCT02525692, n=6; or expanded access, n=4) by independent, central radiology review among 30 patients. At baseline, patients had measurable contrast-enhancing disease by RANO criteria, KPS>60, and were >90 days from prior radiation. Patients with primary lesions that involved the pons or spinal cord or had evidence of leptomeningeal or cerebrospinal fluid dissemination were excluded. ONC201 was orally administered at 625 mg (scaled by body weight for patients < 18 years old) weekly in 29 patients and every 3 weeks in 1 patient. Median age was 31 years (range 8–70; two patients < 18). There were 16 women and 14 men. Tumors were predominantly thalamic (73%), with other locations including the cerebellum (5%), brainstem (non-pontine) (3%), basal ganglia (3%), and midbrain (3%). Median time from prior radiation was 7.5 months. The most frequent drug-related adverse events were low grade nausea (10%) and fatigue (10%). Nine patients (30%) had >50% regression of T1 contrast enhancement and 11 (36.7%) patients had regression of T2/FLAIR. Objective response rate and progression-free survival by RANO criteria, as well overall survival will be reported. Among 8 patients with sustained radiographic regressions, 6 were tapered off dexamethasone and 4 had improvement in KPS. Five patients (16.7%) remain on treatment with a median of 9.5 months (range 7.3–12). In conclusion, single agent ONC201 is well tolerated and clinically active in recurrent H3 K27M DMG patients.

DDRE-06. CELLULAR STRESS RESPONSE IN DIPG THERAPY

Abstract Diffuse Intrinsic Pontine Glioma (DIPG) is an incurable pediatric brain tumor of the pons and brainstem. Therefore, there is a desperate need for new therapeutics. Genomic profiling of tumors identified a highly prevalent dominant negative somatic mutation at lysine (K)-27 in histone genes HIST1H3B and H3F3A. Clonal evolution modeling suggests these mutations are truncal, and studies have demonstrated their contribution to tumorigenesis. ONC201, a first-in-class DRD2 antagonist and ClpP agonist is an anticancer drug developed by Oncoceutics, which targets the unfolded protein response (UPR) and integrated stress response (ISR) signaling and is actively being investigated in patients with recurrent H3 K27M-mutant gliomas. In adults with recurrent glioma, single agent studies showed benign-safety, no dose-limiting toxicities and a durable objective response when administered orally. In addition, intra-tumoral drug levels exceeded therapeutic thresholds, and induced tumor cell apoptosis. Based on this and response seen in a pediatric patient with DIPG for whom compassionate use of ONC201 was approved, a multi-arm, non-randomized multi-institutional Phase I clinical trial (NCT03416530) is actively accruing patients. However, the strength of UPR and ISR in DIPGs and their effect on DIPG response to ONC201 is not known. Our group employed bulk/single cell transcriptomic and single cell proteomic approaches to demonstrate substantial heterogeneity in UPR and ISR signaling in human DIPG samples. Consistent with this, DIPG cell lines exhibited considerable variability in sensitivity to ONC201. Single cell profiling identified tumor sub-populations with significant proliferative capacity even after ONC201 exposure. Incomplete response promotes recurrence. To target these cells, we performed a synthetic lethality screen with a library of 360 FDA-approved CNS penetrant compounds, which identified HDAC inhibitors and DNA damage-inducing chemotherapy as having synergy with ONC201. Thus, we suggest that tumor heterogeneity impacts sensitivity to ONC201 and that this can be reduced by combination treatments.

CTNI-15. CLINICAL EFFICACY OF ONC201 IN NEWLY DIAGNOSED DIPG AND IN PREVIOUSLY IRRADIATED PEDIATRIC H3 K27M-MUTANT GLIOMAS

Abstract ONC201, an anti-cancer DRD2 antagonist and ClpP agonist, is in Phase II trials for adult H3 K27M-mutant diffuse midline gliomas. In adults, the recommended phase 2 dose (RP2D) of 625mg ONC201 once a week has been established as a biologically active dose that is well tolerated. Radiographic regressions with single agent ONC201 have been reported in adult recurrent H3 K27M-mutant glioma patients. This multi-arm, dose-escalation and dose-expansion trial determined the pediatric RP2D of ONC201 administered as an oral capsule (Arm A) or liquid formulation (Arm E) in post-radiation H3 K27M-mutant glioma (Arm A) or in newly diagnosed DIPG (Arm B) patients. Molecular assessments include intratumoral ONC201 concentrations (Arm C) and CSF H3 K27M DNA levels (Arm D). Enrollment as of April 30, 2020 is complete in Arm A (22) and Arm E (26) and continues in Arm B (18/24), Arm C (5/12), and Arm D (22/24). The RP2D of weekly 625mg ONC201 scaled by body weight was confirmed when administered as a capsule or a liquid formulation as a single agent or in combination with radiation without dose-limiting toxicity. The most frequent adverse events regardless of attribution to the drug were predominantly low grade: ONC201 capsule alone was headache (54.5%), nausea (36.4%), and fatigue (36.4%); ONC201 liquid formulation was vomiting (31.8%), headache (22.7%), VIth nerve disorder (22.7%); ONC201 capsules in combination with radiation (Arm B) was headache (47.1%), vomiting (52.9%), nausea (41.2%). Pharmacokinetic analysis in plasma of Arm A patients revealed T1/2: 8.4h; Tmax: 2.1h; Cmax: 2.3ug/mL; AUC0-tlast: 16.4ug/mL, with similar exposure across body weights. In conclusion, when scaled by body weight the ONC201 capsule or liquid formulation alone or in combination with radiation were associated with safety and pharmacokinetic profiles in pediatric H3 K27M-mutant diffuse midline glioma patients that are similar to the experience in adults.

H3 K27M and EZHIP Impede H3K27-Methylation Spreading by Inhibiting Allosterically Stimulated PRC2

Diffuse midline gliomas and posterior fossa type A ependymomas contain the recurrent histone H3 lysine 27 (H3 K27M) mutation and express the H3 K27M-mimic EZHIP (CXorf67), respectively. H3 K27M and EZHIP are competitive inhibitors of Polycomb Repressive Complex 2 (PRC2) lysine methyltransferase activity. Invivo, these proteins reduce overall H3 lysine 27 trimethylation (H3K27me3) levels; however, residual peaks of H3K27me3 remain at CpG islands (CGIs) through an unknown mechanism. Here, we report that EZHIP and H3 K27M preferentially interact with PRC2 that is allosterically activated by H3K27me3 at CGIs and impede its spreading. Moreover, H3 K27M oncohistones reduce H3K27me3 in trans, independent of their incorporation into the chromatin. Although EZHIP is not found outside placental mammals, expression of human EZHIP reduces H3K27me3 in Drosophila melanogaster through a conserved mechanism. Our results provide mechanistic insights for the retention of residual H3K27me3 in tumors driven by H3 K27M and EZHIP.