Date of Award

12-2026

Document Type

Thesis

Degree Name

Master of Science in Biology

Department

Biology

First Reader/Committee Chair

Nicole Bournias-Vardiabasis

Abstract

Epithelial ovarian cancer remains the most lethal gynecologic malignancy in the United States, with a five-year survival rate below 30% for advanced-stage disease. Most patients present with Stage III or IV disease at diagnosis, characterized by widespread peritoneal dissemination and malignant ascites. Despite initial responsiveness to platinum-taxane chemotherapy, over 70% of patients relapse within 18 months, and the immunosuppressive peritoneal tumor microenvironment remains a fundamental barrier to durable therapeutic responses. Neural stem cell-delivered oncolytic adenovirus represents a novel immunotherapeutic platform with the potential to exploit the natural migratory tropism of neural stem cells toward tumor deposits, replicating selectively within malignant cells while simultaneously engaging innate and adaptive immune effectors in situ. Prior work in glioblastoma models demonstrated that HB1.F3.CD21 neural stem cells loaded with CRAd-S-pk7, a survivin promoter-driven oncolytic adenovirus with pk7 fiber modification, can selectively home to and replicate within tumor tissue following systemic delivery. Whether this platform could be translated to intraperitoneal ovarian carcinomatosis, and what dosing schedule would be required to maximize therapeutic efficacy, remained unknown.

We tested the central hypothesis that intraperitoneal delivery of NSC.CRAd-S-pk7 would extend survival and modulate the peritoneal immune microenvironment in an immunocompetent syngeneic murine model of ovarian cancer, and that the dosing schedule would be a critical determinant of efficacy. Three prospective studies were conducted in C57BL/6 mice bearing intraperitoneally established ID8-EGFP-ffLuc tumors. Study CLIN1-24-02 compared four dosing schedules: single dose, three weekly doses, three biweekly doses, and three triweekly doses against an untreated tumor-only control to identify the optimal dosing interval. Study CLIN1-25-03 tested whether escalating from three to six biweekly doses would provide additional benefit beyond the optimal schedule identified in CLIN1-24-02. Study CLIN1-25-06 administered a single late dose at Day 48 post-tumor induction in animals with advanced disease and characterized the immune kinetic response and disease burden over 15 days using multicolor flow cytometry, bioluminescence imaging, and a murine adaptation of the Peritoneal Carcinomatosis Index.

Three biweekly doses of NSC.CRAd-S-pk7 on Days 7, 21, and 35 emerged as the optimal regimen across both dosing studies. In CLIN1-24-02, biweekly treatment conferred a median overall survival of 90 days versus 58 days in untreated controls (hazard ratio 4.85, 95% CI 1.95-9.73; p < 0.0001) and delayed mean ascites onset to 84.0 ± 8.6 days compared to 56.4 ± 4.4 days in controls. Contrary to the hypothesis that more doses would improve outcomes, CLIN1-25-03 demonstrated that six biweekly doses significantly reduced survival compared to three (median 79.5 vs. 92.5 days; HR 2.56, p = 0.0084), with post-Day 42 bioluminescence imaging revealing accelerated tumor growth in animals receiving continued dosing. In the advanced-disease immune kinetics study CLIN1-25-06, a single dose of NSC.CRAd-S-pk7 elevated tumor-infiltrating CD8⁺ cytotoxic T cells 2.2- to 2.6-fold above controls from Day 5 through Day 15 (p < 0.0001), reduced peritoneal ascites volume by 50-60% at Days 7 through 15, arrested omental tumor growth (treated slope: 3.8 mg/day, p = 0.30, R² = 0.07; control slope: 23.5 mg/day, p < 0.0001), and significantly reduced total Peritoneal Carcinomatosis Index across all timepoints (treatment effect p < 0.0001; pooled Hedges' g = 1.78). Macrophage infiltration was substantially reduced in treated tumors from Day 7 onward, primarily driven by depletion of the CD206⁺ M2-like immunosuppressive subset, while no treatment-induced expansion of myeloid-derived suppressor cell-like populations was observed at any time point.

These findings reveal a dosing-ceiling effect in which extended NSC.CRAd-S-pk7 treatment cycles impair rather than enhance therapeutic efficacy. Notably, BLI tumor kinetics did not track survival outcomes across dosing schedules, indicating that BLI alone is insufficient as a surrogate endpoint in this model. Immune profiling from CLIN1-25-06 documented an innate-to-adaptive response sequence consistent with active immune engagement. Together, these data suggest that biweekly dosing aligns with intrinsic immune activation windows, translating into delayed ascites onset and improved long-term survival. This work provides a data-driven framework for optimizing oncolytic immunotherapy schedules. It highlights the importance of treatment timing in maximizing the therapeutic potential of cell-delivered viral agents against aggressive gynecological cancers.

Included in

Biology Commons

Share

COinS