Combined delivery of engineered virus with immunotherapy is safe and improves outcomes in subset of patients with glioblastoma

Study validates safety of combination approach, offering possibilities to adapt treatment strategies to optimize outcomes

MD Anderson News Release May 15, 2023

Intratumoral delivery of an engineered oncolytic virus (DNX-2401) targeting glioblastoma (GBM) cells combined with subsequent immunotherapy was safe and improved survival outcomes in a subset of patients with recurrent GBM, according to results from a multi-institutional Phase I/II clinical trial co-led by researchers at The University of Texas MD Anderson Cancer Center and the University of Toronto.

The study, published today in Nature Medicine, met its primary safety endpoint and demonstrated the combination was well tolerated overall with no dose-limiting toxicities. The study did not meet its primary efficacy endpoint of objective response rate, but the combination achieved a 12-month overall survival (OS) rate of 52.7%, which is greater than the prespecified efficacy threshold of 20%. Three patients remained alive at 45, 48 and 60 months after treatment.

“This viral therapy is a different approach to the current standard of care,” said co-corresponding author Frederick Lang, M.D., chair of Neurosurgery. “Our previous trial demonstrated that not only does the virus act by killing cancer cells directly, it also effectively activates the innate immune system to convert these immunologically cold tumors into hot tumors. This led us to evaluate a combination with checkpoint inhibitors, which we now see can improve survival outcomes in a subset of patients.”

Glioblastoma is an aggressive brain cancer with a median OS of six months; patients typically experience recurrence with standard radiation and chemotherapy approaches. While immune checkpoint blockade has improved outcomes in other cancer types, the unique immunosuppressive tumor microenvironment in recurrent GBM shields it against immune cell infiltration, making it notoriously difficult to treat with immunotherapy.

Smart virus is efficient at eliminating GBM cells and activating immune response
Together with Lang, Juan Fueyo, M.D., and Candelaria Gomez-Manzano, M.D., both professors of Neuro-Oncology, are the co-inventors of DNX-2401, a cold virus engineered to selectively target and invade GBM cells while avoiding normal cells.

In previous Phase I trial results, DNX-2401 monotherapy effectively induced cancer cell death and changed the microenvironment to allow for increased T cell infiltration, resulting in an anti-tumor immune response. Twenty percent of patients with recurrent GBM remained alive for at least three years, and tumor reduction in complete responders continued for more than a year.

These results showed an increase in PD-1 checkpoint expression following treatment, suggesting that the immune system may be primed to respond to anti-PD-1 immunotherapy. Preclinical models supported this hypothesis, as treatment with pembrolizumab one week after DNX-2401 treatment improved survival outcomes compared to either treatment alone.

“Injecting a virus into a patient’s brain tumor is disruptive science, because this therapeutic strategy aims to awaken the patient’s immune system and trigger a healing from within,” Fueyo said. “After injection, patients that respond well develop inflammation inside the tumor, triggering an immune response that first kills the virus. Once the virus is wiped out, the continued immune reaction, stimulated by additional immunotherapy, destroys the cancer cells in a tightly regulated way without the side effects common to chemotherapy or radiation therapy.” 

Combination therapy prolongs survival and improves quality of life in subset of patients
The current trial was designed to evaluate the combination of intratumoral DNX-2401 followed by intravenous pembrolizumab. The study enrolled 49 patients with recurrent GBM from several institutions between September 28, 2016 to January 17, 2019. The median age of patients was 53 years and 41% were women.

Forty-eight of the 49 patients (98%) were treated with one dose of DNX-2401 after biopsy, followed by pembrolizumab given one week later. The majority of adverse events were grade 1 or 2, with the most common being brain edema (37%), headache (31%) and fatigue (29%).  

The combination achieved a clinical benefit, defined as stable disease or better, in more than half (56.2%) of the patients. Five patients had objective responses and two experienced tumor reduction of 80% or more at six months follow-up. By 18 months, these two patients had a complete response without evidence of disease progression.

Exploratory gene expression and immunophenotypic analysis also revealed that objective response occurred in patients with a moderately inflamed tumor microenvironment and modest PD-1 expression, meriting further investigation of which patient characteristics will determine who is more likely to benefit from this treatment.

While this study did not meet its primary efficacy endpoint, it did validate the use of DNX-2401 in combination with immune checkpoint inhibitors as a safe approach that opens the door to exploring other combinations. For instance, the researchers found that specimens from 10 patients showed elevated levels of several immune checkpoints after treatment including LAG3, TIGIT and B7-H3, highlighting these proteins as potential therapeutic targets.

“Our studies using this ‘smart virus’ are ongoing, but we are encouraged that we continue to see a small number of patients who have a very dramatic eradication of the tumor,” Gomez-Manzano said. “These results motivate us to keep searching for the best combination strategy that can optimize the use of this virus to improve patient outcomes.” 

Clinical trials currently are underway using mesenchymal stem cells to deliver more of the smart virus to the tumor and more widely through the tumor. Future clinical trials will evaluate alternate treatments, such as checkpoint inhibitors or CAR T cell therapy, in combination with DNX-2401. 

This study was supported by DNATrix, Inc. and Merck & Co. A full list of collaborating authors and their disclosures can be found with the full paper here.