Glioblastoma is one of oncology's hardest problems: an aggressive brain tumor that almost always recurs after surgery, radiation, and chemotherapy, and that has resisted decades of new approaches. A Phase 1/2 trial recently posted to the U.S. registry attacks it with a gene-therapy design that tries to do two jobs at once from a single shot — killing dividing tumor cells directly while simultaneously recruiting the immune system to finish the work.
The study, NCT07346144, is sponsored by Trogenix and is recruiting patients with newly diagnosed or recurrent high-grade glioma, including glioblastoma. The drug, TGX-007, is delivered by an adeno-associated virus vector injected directly into the tumor, and what makes it notable is that the vector carries not one therapeutic gene but two. The registry spells out the dual mechanism plainly.
"TGX-007 is a gene therapy drug delivered by a harmless adeno-associated virus (AAV) vector which delivers two combined therapeutic payloads to enable killing of proliferative cells and activation of an anti-tumour immune response. One is herpes simplex virus thymidine kinase (HSV-tk), which converts the pro-drug valaciclovir into an active drug that can kill tumour cells and the other is interleukin 12 (IL-12), which activates the body's immune system to recognise and fight the tumour."— ClinicalTrials.gov, source
The two payloads work on complementary principles. The first, herpes simplex virus thymidine kinase, is a so-called suicide-gene strategy with a long pedigree: the enzyme is harmless on its own, but it converts the antiviral prodrug valaciclovir into a toxic molecule inside the cells that took up the gene. Because the toxic product preferentially damages dividing cells, the system aims to kill the proliferating tumor while sparing the brain's largely non-dividing healthy tissue. That is why the trial pairs the single intratumoral injection of TGX-007 with oral valaciclovir taken three times daily for 14 to 21 days — the prodrug is the trigger that makes the enzyme lethal.
The second payload, interleukin-12, is a potent immune-activating cytokine. Delivered locally by the same vector, it is intended to wake up an anti-tumor immune response right where the tumor sits. The rationale for combining them is that the suicide-gene arm not only kills cells but releases tumor debris, and a local cytokine signal can help the immune system recognize that debris and mount a broader, potentially durable response. Packaging both into one AAV vector is the engineering bet at the center of this program: a single injection that delivers a direct cytotoxic mechanism and an immune-priming mechanism together.
What the trial is structured to learn first
The trial is split into two phases, and the endpoint design is appropriately cautious. The first primary objective is safety and tolerability and the identification of an optimal biological dose — measured through the incidence of adverse events and serious adverse events, dose-limiting toxicities within 28 days of administration at each dose level, and confirmation of HSV-tk expression. Only the second primary objective addresses overall survival, assessed at 18 months for newly diagnosed patients and six months for recurrent patients. The sequencing tells the reader what a near-term success looks like: a tolerable dose with confirmed transgene expression, not yet a survival win.
That HSV-tk expression endpoint deserves emphasis, because it is the kind of mechanistic checkpoint that separates a credible gene-therapy program from a hopeful one. Confirming that the injected vector actually drives expression of the enzyme is the proof that the delivery worked at all; without it, neither the prodrug-killing arm nor any survival signal could be interpreted. Building that confirmation into the primary objectives is a sign the trial is designed to learn whether the biology is doing what the design intends.
The brain is also an unusually demanding setting for any of this to work. The blood-brain barrier that frustrates systemic drugs is precisely why TGX-007 is injected directly into the tumor rather than infused — local delivery is the only practical way to get a viral vector and its payloads where they need to be. That same anatomy raises the stakes on safety: interleukin-12 is a powerful cytokine whose systemic use has historically been limited by toxicity, so confining its expression to the tumor by local AAV delivery is part of the design's safety logic, and the dose-limiting-toxicity window over the first 28 days is where the success of that confinement will first show up. The trial's two distinct populations — newly diagnosed patients heading into standard surgery and chemoradiotherapy, and recurrent patients with fewer options — also carry different survival benchmarks, which is why the protocol sets separate overall-survival timepoints for each. Reading the eventual data will require keeping those two groups, and their very different baselines, carefully apart.
Reading the record honestly
This is a non-randomized, open-label, sequential study planning to enroll 68 participants, and it sits in one of the most treatment-resistant settings in oncology. The dual-payload, locally injected design is genuinely novel in its combination, but novelty is a hypothesis, not a result. The disciplined read of NCT07346144 is that a single AAV vector carrying a suicide gene and an immune-stimulating cytokine, activated by oral valaciclovir, is being tested first for safety and an optimal biological dose in high-grade glioma — with survival a downstream question the later phase is built to probe. Whether two payloads in one shot translate into longer survival for glioblastoma patients is exactly what this trial exists to find out, and exactly what its early data cannot yet claim.