Phase I clinical trial of peposertib plus radiation in adults with newly diagnosed MGMT-unmethylated glioblastoma.

Background: DNA-dependent protein kinase (DNA-PK) is a pivotal component of DNA damage repair (DDR) pathways and is an attractive target in glioblastoma because its expression renders tumors less vulnerable to radiotherapy. Peposertib is a small molecule DNA-PK inhibitor that has been pre-clinically shown to potentiate radiotherapy and regress glioblastoma tumors. We conducted a two-stage phase I trial of peposertib plus radiation in patients with newly-diagnosed MGMT-unmethylated glioblastoma (NCT04555577). Methods: In stage 1, patients received concurrent peposertib plus standard-of-care radiotherapy to determine the maximum-tolerated dose (MTD) based on the Bayesian Optimal Interval design (maximum n=24). Stage 2 is a window-of-opportunity expansion cohort and will include 5 surgical patients to evaluate intratumoral drug concentration. Both groups receive 6 cycles of adjuvant temozolomide. Results: Eighteen patients completed the 10-week dose-limiting toxicity (DLT) period; 3@50mg, 3@100mg, 3@200mg, 9@300mg. One DLT (G3 radiation necrosis [RN] at 300mg) was observed. Enrollment of the last three patients into the 300mg dose level began in December 2023 and will complete stage 1. To date, most notable toxicity was transient G3 dermatitis of the scalp (2@200mg, 1@300mg; not a DLT per protocol). Therefore, radiation dose constraints to the skin were incorporated and subsequent patients did not experience this toxicity. After a median follow up of 14.3 months (9.3-18.4), the median OS was 22.9 months [95% CI (16-NR)] and median PFS was 12.7 months [95% CI (9-NR)]. Next-generation sequencing (NGS) was available for 16 archival tumor tissue specimens from patients treated on this trial. Two patients had pathology-proven RN (one within and one outside of the DLT period). Both tumors had baseline mutations in DDR genes (i.e. ATRX, DICER1). Recurrent tissue was available for 4 patients after treatment with peposertib, 2 of which demonstrated gain of DDR gene mutations (2 ATM mutations in one patient and gain of MAD2L2 mutation in another patient). Conclusions: The initial safety data of peposertib plus radiation in patients with newly-diagnosed MGMT unmethylated glioblastoma is favorable. The cases with RN in tumors with DDR mutations, the gain of DDR mutations in recurrent tumors, and the cases of scalp dermatitis suggest peposertib activity may correlate with DDR function, and possibly potentiates the effect of radiation. Stage I is an independent stage of the protocol that will enable MTD determination. Complete safety and survival data and correlations with genomics and spatial transcriptomics for Stage I patients will be presented at the meeting. The study was supported by EMD Serono (CrossRef Funder ID: 10.13039/100004755).