Localized Delivery of an IL-2–Expressing Oncolytic Vaccinia Virus Enhances PD-1 Pathway Inhibition in a De-Novo Esophageal Adenocarcinoma Model

Technical Summary: Localized Delivery of an IL-2–Expressing Oncolytic Vaccinia Virus Enhances PD-1 Pathway Inhibition in a De-Novo Esophageal Adenocarcinoma Model

Problem Statement
Esophageal adenocarcinoma (EAC) is an aggressive malignancy with poor survival rates and limited response to immune checkpoint blockade (ICB), particularly in patients with "immunologically cold" tumors characterized by low T-cell infiltration. While oncolytic viruses (OVs) offer a mechanism to convert cold tumors into "hot" environments via immunogenic cell death, systemic delivery faces barriers such as neutralization and toxicity. Furthermore, previous preclinical success with the oncolytic vaccinia virus expressing membrane-bound IL-2 (vvDD-IL-2-RG) was established in transplantable syngeneic models, which fail to recapitulate the complex tumor-stroma interactions and gradual immune editing of human EAC. There is a critical need to evaluate this combination therapy in a clinically relevant de novo EAC model and to assess the feasibility of localized delivery routes (oral gavage) versus systemic delivery (intravenous).

Methodology
The study utilized a modified Levrat rat model, an immunocompetent system that recapitulates the reflux-induced inflammation-to-carcinoma sequence of human EAC.

• Viral Construct: The researchers employed vvDD-IL-2-RG, a genetically engineered vaccinia virus (Western Reserve strain) with deletions in the vaccinia growth factor (VGF) and thymidine kinase (TK) genes. This virus expresses human IL-2 fused to a rigid linker and a glycosylphosphatidylinositol (GPI) anchor, designed to localize IL-2 signaling within the tumor microenvironment (TME) while minimizing systemic exposure. An unarmed control virus (vvDD-YFP) was used for in vitro validation.
• In Vitro Validation: Human EAC cell lines (FLO-1) and normal esophageal epithelial cells (Het-1A) were infected to assess viral infectivity, tumor selectivity (via RT-qPCR of viral gene A34R), and cytotoxicity (CCK-8 assay). IL-2 bioactivity was confirmed via mRNA expression.
• In Vivo Design: Rats with confirmed tumors (34 weeks post-surgery) were randomized into six groups: oral gavage virus (V(O)), intravenous virus (V(I)), PD-1 inhibitor alone (AUNP-12), V(O) + AUNP-12, V(I) + AUNP-12, and no treatment.
• Delivery: The virus was administered at 1×10⁹ PFU either intravenously or via oral gavage using a 0.75% sodium alginate gel vehicle to enhance mucosal adhesion and local retention. The PD-1 inhibitor (AUNP-12, 10 mg/kg) was administered intraperitoneally every two weeks for 13 weeks.
• Monitoring: Tumor burden was tracked via serial 7T MRI. Tumor microenvironment (TME) analysis was performed on subsets of animals euthanized at days 4 and 7 (for immune infiltration kinetics) and at the study endpoint (13 weeks) for gross tumor assessment and immunofluorescence (IF) staining of CD3 and CD8.

Key Results

• Viral Selectivity and Cytotoxicity: In vitro, vvDD-IL-2-RG demonstrated preferential replication in EAC cells (FLO-1) over normal epithelial cells (Het-1A), with A34R expression approximately 3.5-fold higher in tumor cells. The expression of the IL-2 transgene did not impair viral cytotoxicity.
• Tumor Response:
  • Monotherapy: Anti-PD-1 monotherapy (AUNP-12) was ineffective, resulting in a 96% increase in median tumor volume. Viral monotherapy showed modest effects (9.9% decrease for IV; 34.9% increase for oral).
  • Combination Therapy: The combination of vvDD-IL-2-RG and AUNP-12 resulted in substantial tumor regression. The oral gavage combination group showed a 63% decrease in median tumor volume, and the intravenous combination group showed a 58% decrease.
  • Statistical Note: While the combination groups showed dramatic volumetric reductions compared to controls, the overall Kruskal-Wallis test for tumor weight was not statistically significant (p = 0.072), and the group × time interaction in the mixed-effects model was not significant (p = 0.419). However, one pre-specified pairwise comparison (V(I) vs. V(I) + AUNP-12) was significant after correction (p = 0.035).
• Immune Microenvironment:
  • Kinetics: Rapid immune activation was observed. By day 4, oral viral administration led to a 2.3-fold increase in CD4+ and 3.6-fold increase in CD8+ T-cell infiltration compared to controls.
  • Sustained Response: By day 7, IL-2 expression returned to baseline (consistent with viral clearance), but CD4+ and CD8+ infiltration remained elevated.
  • Phenotype Shift: Immunofluorescence revealed a dramatic shift in the TME, with nearly 100% of infiltrating T cells being CD8+ at day 7 in treated groups, compared to ~30% in untreated controls.
• Safety: The treatment was well-tolerated. No animals exhibited systemic toxicity, significant weight loss, or organ abnormalities attributable to the viral or antibody treatment.

Significance and Claims
The paper claims to provide proof of principle that localized delivery of an IL-2-expressing oncolytic vaccinia virus enhances PD-1 pathway inhibition in a de novo EAC model. The authors assert that this approach addresses the limitations of current immunotherapies in "cold" EAC tumors by:

1. Converting the TME: Using the virus to induce immunogenic cell death and local IL-2 expression to recruit and activate cytotoxic T cells.
2. Overcoming Resistance: Combining this activation with PD-1 blockade to prevent the adaptive immune resistance (checkpoint upregulation) that typically limits viral monotherapy.
3. Feasibility of Localized Delivery: Demonstrating that oral gavage delivery in a viscous gel is a viable strategy for luminal GI malignancies, achieving high local viral concentrations and rapid immune activation comparable to intravenous delivery in terms of final tumor volume reduction.

The authors conclude that while the study has limitations regarding sample size and statistical power for certain endpoints, the observed biological effects (substantial tumor regression and CD8+ T-cell predominance) support further investigation into this strategy for clinical translation, particularly for patients with PD-L1-low or immunologically cold esophageal adenocarcinoma.