CM93 enters first-in-human testing in EGFR-altered recurrent glioblastoma
A new phase 1 study is set to test CM93 in adults with recurrent glioblastoma carrying EGFR mutation or amplification, a molecular subgroup that remains difficult to treat once standard therapy has failed. The trial combines dose escalation with a surgical window-of-opportunity design, giving investigators a chance to study both tolerability and whether the drug reaches active concentrations inside human tumor tissue. For a disease in which recurrent growth often outpaces available systemic options, the design is unusually direct: it asks not only whether the agent can be given safely, but whether it can reach the biologic target it is meant to inhibit.
The study is open-label and first-in-human. It is being run by NYU Langone Health and is scheduled to enroll an estimated 72 participants across three U.S. sites: NYU Langone Health in New York, Mass General Brigham in Boston, and Mayo Clinic in Rochester, Minnesota. Estimated study start is Dec. 1, 2026, with primary completion projected for June 1, 2031, and full completion for Dec. 1, 2032.
The scientific logic is straightforward. EGFR amplification and mutation are common drivers in glioblastoma biology, promoting proliferative signaling, invasive growth, and treatment resistance. In recurrent disease, those pathways can remain active even after surgery, radiation, and chemotherapy have already been used. A drug that can cross the blood-brain barrier and suppress EGFR signaling in tissue could offer a rational path forward, but the brain remains a formidable pharmacologic barrier, and recurrent glioblastoma often contains heterogeneous regions that may not be equally exposed to therapy.
Two-part design separates safety from tumor biology
The protocol is divided into two consecutive parts. Part 1 is a dose escalation and dose expansion phase designed to determine dose-limiting toxicities, maximum tolerated dose, and recommended phase 2 dose or doses. Investigators will use a Bayesian Optimal Interval dose escalation design, a framework intended to move dose selection forward while maintaining intensive safety monitoring. In early oncology studies, that balance matters: overcautious dose selection can stall development, but pushing dose too quickly can expose a fragile population to avoidable toxicity.
Part 2 is the more biologically ambitious section. Adults with EGFR-altered recurrent glioblastoma who are candidates for surgical debulking will be randomized 1:1 to receive CM93 before surgery or no pretreatment. The pretreated cohort will allow investigators to measure whether the drug crosses the blood-brain barrier and achieves therapeutic tumor concentrations. Comparison of pretreated and control tumor tissue will also be used to assess whether CM93 inhibits EGFR activation in the tumor microenvironment.
A drug that reaches plasma is not necessarily a drug that reaches glioblastoma.
That gap has frustrated neuro-oncology for years, because central nervous system delivery is often the limiting step, not target selection. The surgical window design is well suited to this problem because it links pharmacokinetics, pharmacodynamics, and tissue-based readouts in a single clinical framework. Tumor resection gives investigators a rare opportunity to confirm whether a candidate therapy is doing more than circulating in the bloodstream.
Eligibility reflects a heavily pretreated adult population
Eligibility criteria make clear that this is a study for adults with high-grade, molecularly defined disease who have already received standard first-line therapy. Participants must have histologically confirmed recurrent or progressive WHO grade 4 IDH wild type glioblastoma, documented EGFR mutation or amplification from archival tissue, prior radiotherapy with or without chemotherapy, and unequivocal progression on MRI with measurable disease by RANO criteria. The protocol also requires enough archival tumor tissue for repeat EGFR testing and additional sequencing, including either one FFPE tumor block or 20 unstained FFPE slides.
Functional and organ reserve requirements are substantial, which is typical for an early-phase oncology study. Patients must have a Karnofsky performance status of at least 60, adequate marrow and organ function, and a left ventricular ejection fraction of at least 50% on echocardiogram or MUGA scan. Those with known cardiac disease or prior exposure to cardiotoxic agents must meet New York Heart Association class II or better. Women of childbearing potential and male participants with partners of childbearing potential must use dual contraception during treatment and for four months afterward.
Age is restricted to 18 years and older. Children are excluded because no dosing or adverse event data are yet available for CM93 in pediatric patients, although the protocol leaves open the possibility of future pediatric trials.
Surgical tissue could answer the most important question first
The most revealing part of the study may be the tissue analysis in the surgical cohort. Patients entering Part 2 must be scheduled for clinically indicated surgery and must have enough resectable tumor so that both enhancing and non-enhancing components can be grossly or subtotally removed for correlative studies. That requirement is more than an operational detail. Glioblastoma is notoriously heterogeneous, and sampling both enhancing and non-enhancing regions may expose differences in drug penetration and pathway suppression that would be missed if only one compartment were examined.
If CM93 reaches the tumor and measurably inhibits EGFR activation, the study will have done more than establish a dose. It will have generated direct human evidence that the agent engages its intended target inside recurrent glioblastoma tissue, a critical threshold for a disease in which many candidates falter before biologic confirmation is ever possible.
The trial is currently not yet recruiting. J. Ricardo McFaline-Figueroa, MD, PhD, is listed as the study contact, and the protocol is registered under ClinicalTrials.gov identifier NCT07927491 and other study ID numbers 25-00065 and U19CA264504.
A phase 1 readout with no room for pharmacologic ambiguity
Early-phase glioblastoma trials rarely offer clean answers, but this design is built to minimize uncertainty where it matters most. Dose escalation should define whether CM93 can be administered safely in adults with recurrent EGFR-altered disease, while the surgical window portion should show whether systemic exposure translates into intratumoral delivery and pathway inhibition. Those are separate questions, and in neuro-oncology they often diverge.
What remains unresolved is whether those biologic signals, if they appear, will be strong enough to justify later-stage testing in a disease with relentless recurrence and limited therapeutic margin. The upcoming data will need to clarify not only tolerability and dosing, but whether EGFR-altered glioblastoma tissue is actually vulnerable to this approach once the drug is placed in the human brain environment.