Phase I Study of DNX-2401 (Delta-24-RGD) Oncolytic Adenovirus: Replication and Immunotherapeutic Effects in Recurrent Malignant Glioma

Abstract

Purpose DNX-2401 (Delta-24-RGD; tasadenoturev) is a tumor-selective, replication-competent oncolytic adenovirus. Preclinical studies demonstrated antiglioma efficacy, but the effects and mechanisms of action have not been evaluated in patients. Methods A phase I, dose-escalation, biologic-end-point clinical trial of DNX-2401 was conducted in 37 patients with recurrent malignant glioma. Patients received a single intratumoral injection of DNX-2401 into biopsy-confirmed recurrent tumor to evaluate safety and response across eight dose levels (group A). To investigate the mechanism of action, a second group of patients (group B) underwent intratumoral injection through a permanently implanted catheter, followed 14 days later by en bloc resection to acquire post-treatment specimens. Results In group A (n = 25), 20% of patients survived > 3 years from treatment, and three patients had a ≥ 95% reduction in the enhancing tumor (12%), with all three of these dramatic responses resulting in > 3 years of progression-free survival from the time of treatment. Analyses of post-treatment surgical specimens (group B, n = 12) showed that DNX-2401 replicates and spreads within the tumor, documenting direct virus-induced oncolysis in patients. In addition to radiographic signs of inflammation, histopathologic examination of immune markers in post-treatment specimens showed tumor infiltration by CD8+ and T-bet+ cells, and transmembrane immunoglobulin mucin-3 downregulation after treatment. Analyses of patient-derived cell lines for damage-associated molecular patterns revealed induction of immunogenic cell death in tumor cells after DNX-2401 administration. Conclusion Treatment with DNX-2401 resulted in dramatic responses with long-term survival in recurrent high-grade gliomas that are probably due to direct oncolytic effects of the virus followed by elicitation of an immune-mediated antiglioma response.

Trial registration: ClinicalTrials.gov NCT00805376.

Figures

Fig 1.

In situ replication and clinical activity of DNX-2401 in recurrent malignant glioma. (A) Schema outlining treatment strategies for group A (single injection of DNX-2401 into recurrent tumor without other therapeutic interventions to evaluate safety and efficacy), and group B (two-stage surgical design to provide post-treatment specimens for analysis). (B) Photomicrograph of en bloc resected post-treatment surgical specimen (left) obtained at open craniotomy from a patient in group B who underwent intratumoral injection of DNX-2401 14 days before this specimen was obtained. The arrows indicate the location of the proximal end of the injection catheter whose distal end extends into the center of the tumor and marks the site of DNX-2401 administration. After fixation, the en bloc specimen was blocked perpendicular to the catheter, and these blocks (right) were then sectioned and analyzed for viral replication; the arrows indicate the location of the catheter in each block, which allowed for precise identification of the site of DNX-2401 injection. (C) Photomicrographs of sections from en bloc resection specimens taken 14 days after virus injection showing immunohistochemical staining for viral E1A protein (left, ×100), which is a marker of viral infection, and for viral hexon protein (middle, ×20; right, ×40), which is a marker of replication. E1A immunostaining (left) is primarily intranuclear, as would be expected for actively infecting virus. The middle photomicrograph shows three distinct concentric zones indicative of active viral spread: a central zone of virus-induced necrosis without pseudopalisading, where the virus has infected and lysed the tumor cells leaving necrosis (N; zone 1), an intermediate zone of active viral replication as demonstrated by high hexon protein expression, which surrounds the central area of necrosis consistent with the centrifugal spread of the virus (V; zone 2), and a peripheral zone of yet-to-be-infected tumor cells (T; zone 3). The right panel shows a high-power view (×40) of tumor cells immunostaining positively for hexon protein; note the viral inclusions in the cytoplasm and the viral blebbing from the cell membrane consistent with active replication. (D) Waterfall plot showing maximal change in tumor size for all patients enrolled in group A after a single DNX-2401 treatment (n = 25). Bars represent the maximal tumor change from baseline on the basis of contrast-enhanced magnetic resonance imaging (complete responses, gray; partial responses, gold). Numbers at the bottom correspond to each patient’s identification number. (E) Bar plots showing the survival for each patient in group A by DNX-2401 dose. The patient numbers and bar colors correspond to Figure 1D. The length of the bar represents survival. The arrow indicates ongoing survival. vp, viral particles.

Fig 2.

Summary of the three complete tumor responses of recurrent glioblastomas after a single dose of DNX-2401. (A) Magnetic resonance imaging (MRI) scans and graph of tumor size for Patient No. 12 demonstrating a complete response of a recurrent left parietal glioblastoma treated with 1 × 108 viral particles (vp) DNX-2401. Contrast enhancement was observed until 4 months post-treatment, followed by complete regression, which lasted until 41.5 months after DNX-2401 treatment. MRI scans at baseline and months 1, 4, 8, 12, and 31 after treatment are shown. Circle indicates tumor mass. The graph shows the percent change in tumor size (contrast enhancement) over time; note the initial increase in enhancement followed by a progressive decrease. (B) MRI scans and graph of tumor size for Patient No. 33 demonstrating a recurrent glioblastoma of the left superior temporal gyrus treated with 1 × 1010 vp DNX-2401. MRI scans at baseline and months 2, 3, 6, 18, and 37 after treatment are shown. Circle indicates tumor mass. Complete regression was concurrent with a decrease in peritumoral hyperintensity on fluid-attenuated inversion recovery (FLAIR) images (bottom left). As in Figure 2A, the graph of percent change in tumor size (contrast enhancement) over time shows an initial increase in enhancement followed by a progressive decrease. (C) MRI scans and graph of tumor size for Patient No. 37 with a recurrent glioblastoma of his right parietal lobe injected with 3 × 1010 vp DNX-2401. Complete regression was observed 1 year after treatment, as shown on MRIs from baseline and months 1, 2, 7, and 14. The graph of percent change in tumor size (contrast enhancement) over time shows a staggering initial increase in enhancement followed by a progressive decrease. A contrast-enhanced nodule (bottom right) arose away from the initial recurrence 29 months after DNX-2401 treatment (arrows note location of new lesion); surgical resection of this lesion revealed necrosis and inflammation and no evidence of tumor.

Fig 3.

Immune-mediated response to DNX-2401. (A and B) Magnetic resonance imaging (MRI) scans and pathologic specimens of left temporal glioblastoma from Patient No. 20. (A) Baseline MRI (far left) and the exact site of DNX-2401 injection into recurrent glioblastoma of the left temporal lobe (second from left shows enlarged image with yellow line and red dot showing trajectory of injection needle and site of injection of DNX-2401, respectively). MRI 1.5 months after treatment (third and fourth from left) shows necrosis near injection site (star) and worsening of enhancement. This entire tumor was resected, and postoperative MRI shows complete removal of the mass (far right). (B) The pathologic examination of the tumor resected in Figure 3A, including hematoxylin and eosin–stained section (far left) demonstrating large numbers of small hematoxylin-stained inflammatory cells around blood vessels. Immunostaining shows infiltration of macrophages (CD68; second from left), T cells (CD-3; middle), few CD4+ T cells (second from right), and large numbers of CD8+ T cells (far right), all consistent with a cytotoxic immune response. The patient died of a pulmonary embolism 1.5 months after the surgery; therefore, his outcome related to the tumor could not be assessed. (C) Representative cases of post-treated specimens from patients from group B taken 14 days after treatment with DNX-2401 and immunostained for macrophages (CD68), T cells (CD3), cytotoxic T cells (CD8), and T-bet+ T cells; note the prominent perivascular cuffing of all these cell types. Each dot represents a different patient. Different colors and shapes were used in the plotted points because many overlapped. The CD8 and T-bet specimens are serial sections from the same patient. T-bet expression correlated with increased CD8+ T cells as expected of a Th1 immune response (far right). Graphs showing the density of (D) CD3+, CD4+, and CD8+ cells and (E) TIM-3+ cells on the basis of quantitative analyses of pretreatment (pre; n = 5) and post-treatment (post; n = 10) tumor specimens. Post-treatment specimens were from patients in group B treated with DNX-2401 14 days previously, and five pretreatment specimens were the matched pretreatment biopsies of three patients in group B and two unmatched specimens (one from group B and one from group A). The mean values are noted by horizontal bars; although CD3+ changes were not evident, both CD4+ and CD8+ cells increased after treatment, with increases in CD4+ cells reaching statistical significance. A decrease in the exhaustion marker TIM-3 was seen after treatment with DNX-2401. (F) Graphs showing analyses of damage-associated molecular patterns (DAMPs) in glioma stem cells cultured from the surgical specimens of two patients enrolled in group B. Glioma sphere-forming cells (GSC) 327 and GSC308 were cultured and treated with either mock infection (blue bars) or with DNX-2401 (dose of 20 active viral particles per cell; gold bars). After 72 hours, the supernatant was collected and analyzed by enzyme-linked immunosorbent assay for the following damage-associated molecular patterns: high-mobility group proteins B1 (HMGB1), heat-shock protein (HSP)90α, HSP70, and ATP. Data represent mean ± standard deviation. *P ≤ .05; ***P ≤ .001.