Bladder tumor metabolic alterations in response to IFNα gene therapy.

Abstract

468 Background: Intravesical interferon-alpha (IFNα) gene therapy with Nadofaragene firadenovec has shown clinical efficacy in patients with non-muscle invasive bladder cancer (NMIBC) in a phase III clinical trial, highlighting the therapeutic potential of this approach in a disease with significant unmet clinical need. Optimizing the clinical efficacy of IFNα gene therapy requires an understanding of the underlying therapeutic mechanisms. Here, we investigate the impact of IFNα gene therapy on tumor metabolism using in vitro and orthotopic murine preclinical models and clinical trial data to elucidate mechanisms of tumor resistance and identify predictive biomarkers. Methods: In vitro murine bladder cancer cell lines treated with recombinant IFNα (rIFNα) and lentiviral IFNα (LV-IFNα) were analyzed by whole-transcriptome sequencing, glucose uptake, and lactate production. Preclinical murine bladder cancer models were treated with LV-IFNα (orthotopic tumor model) or Poly(I:C) (flank tumor model), a potent IFN inducer. Disease response was monitored by in vivo real-time luciferase imaging. Tumors were harvested and whole-transcriptome sequencing performed to assess effects of IFNα therapy on tumor metabolism and lipidomics. Lipidomic profiling was performed on patient urine samples from a phase II clinical trial of intravesical Nadofaragene firadenovec (7 clinical responders and 6 non-responders) to assess for clinically-relevant differences in lipid metabolism. Results: Following IFNα therapy in vitro and in murine orthotopic bladder cancer models, we identified downregulation of genes involved in fatty acid synthesis and upregulation of genes involved in glycolysis by whole-transcriptome sequencing. This was confirmed by higher glucose uptake and lactate production by IFNα-treated cells in vitro. These findings were recapitulated in whole-transcriptome sequencing data of human bladder tumors treated with intravesical Nadofaragene firadenovec. Lipidomics performed on murine MB49 tumors treated with poly(I:C) identified 79 upregulated lipids, including phosphotidyl choline, spingomyelin and phosphatidyl ethanolamine, and 12 downregulated lipids, notably the cardiolipin class. Lipidomics performed on patient urine samples collected pre- and post-treatment with intravesical Nadofaragene firadenovec detected >592 lipids with distinct expression profiles differentiating clinical responders and non-responders at both timepoints. Conclusions: We describe novel modulation of glucose and lipid metabolism by bladder tumor cells in response to IFNα gene therapy. These metabolic changes were reproducible across in vitro, in vivo and clinical trial studies and improve our mechanistic understanding of IFNα gene therapy, identify tumor escape pathways targetable with combination therapy regimens, and identify a new class of biomarkers for predicting clinical response of NMIBC to IFNα gene therapy.