Patient-Level Data Show Extended OS Advantage With DCVax-L vs External Controls in Glioblastoma

The autologous dendritic cell treatment DCVax-L extended overall survival (OS) vs external controls in patients with glioblastoma (GBM), according to updated individual patient–level data from multiple independent analyses of an ongoing phase 3 trial (NCT00045968).¹

Findings presented at the 2026 Annual Meeting of the British Neuro-Oncology Society showed that the study’s report of OS extension with DCVax-L vs controls was likely underestimated in the original analysis. The updated dataset included findings from several statistical analyses conducted by independent statisticians and using individual patient–level data.

In the individual patient–level data–based analysis, propensity score matching analyses of known prognostic factors that included data from 2 large randomized controlled trials in newly diagnosed GBM showed median OS advantages with DCVax-L ranging from 4.9 to 6.3 months. Additionally, a propensity score matching analysis of a third randomized controlled trial in this setting revealed median OS advantages ranging from 3.4 to 3.7 months with DCVax-L. The HRs range was 0.69 to 0.77 (P range = .004-.027). The investigators noted that these findings were greater than what was seen in the original trial analysis using cohort-level data, which showed a median OS advantage with DCVax-L of 2.8 months.

An inverse probability weighting analysis of known prognostic factors showed similar findings, including median OS advantages with DCVax-L ranging from 3.4 to 4.3 months (P range = .008-.03).

The investigators also performed 2 types of sensitivity analyses to account for potential unknown factors that may have confounded or added bias to the results: Rosenbaum’s Gamma measures, and E values. These sensitivity analyses showed that the observed efficacy of DCVax-L was not attributable to any hidden imbalance between patients in the DCVax-L vs control populations.

“These new propensity score matching and inverse probability weighting analyses, together with the E Value and Rosenbaum’s Gamma sensitivity analyses, provide strong evidence and reinforcement of the survival benefit with DCVax-L treatment, based on my review of the statisticians’ analyses,” Marnix Bosch, PhD, chief technical officer of Northwest Biotherapeutics, stated in a news release. “The key is that all these analyses, which have taken multiple different approaches and have applied them to multiple newly diagnosed GBM randomized controlled trials, have produced results that are all directionally consistent and all in the same range of magnitude. The statisticians’ findings indicate that the median survival benefit reported from the original analysis of the DCVax-L phase 3 trial, based on cohort-level data, was likely substantially underestimated.”

What is the design of the phase 3 trial of DCVax-L in GBM?

The original trial design used cohort-level data to compare outcomes between patients in the DCVax-L arm of this phase 3 trial and those in the control arms of comparator randomized controlled trials. At the time, Northwest Biotherapeutics, the developer of DCVax-L, only had access to cohort-level data from the control population. However, the company recently gained access to individual patient data from 3 randomized controlled trials in patients with newly diagnosed GBM, which allowed for more precise matching between the DCVax-L and control populations. Notably, this adoption of individual patient data aligned with the trial’s pre-specified statistical analysis plan.

What findings have been previously reported with DCVax-L in GBM?

The efficacy of DCVax-L in this phase 3 trial was first compared with cohort-level data from external controls in randomized controlled trials (5 in newly diagnosed GBM, and 10 in recurrent GBM).² Among patients with newly diagnosed GBM, in the DCVax-L cohort (n = 232), the median OS was 19.3 months from random assignment and 22.4 months from surgery; in the control population (n = 1366), the median OS was 16.5 months from random assignment (HR for DCVax-L vs external control, 0.80; P < .001). Among patients with recurrent GBM, in the DCVax-L cohort (n = 64), the median OS was 13.2 months from recurrence vs 7.8 months from recurrence in the control population (n = 640; HR for DCVax-L vs external control, 0.58; P < .001).

Longer-term data showed that DCVax-L more than doubled the 5-year OS rate vs external controls in patients with newly diagnosed disease, with 5-year rates of 13.0% vs 5.7%, respectively (relative rate, 228%). In patients with recurrent disease, DCVax-L more than doubled the OS rate at 30 months post-recurrence, at 11.1% vs 5.1%, respectively (relative rate, 217%).

References

1. Northwest Biotherapeutics presents updated survival data from phase III trial of DCVax-L for glioblastoma using individual patient level data in multiple independent analyses. News release. Northwest Biotherapeutics. July 8, 2026. Accessed July 9, 2026. https://nwbio.com/northwest-biotherapeutics-presents-updated-survival-data-from-phase-iii-trial-of-dcvax-l-for-glioblastoma-using-individual-patient-level-data-in-multiple-independent-analyses/
2. DCVax-L-associated survival extension assessed through propensity score matching analysis. Presented at: 2026 Annual Meeting of the British Neuro-Oncology Society. July 1-2, 2026; Birmingham, United Kingdom.