Co-Adaptive Escape-Resistant Antineoplastics

Abstract public domain (NIH RePORTER)

Project Summary/Abstract Modern oncology has delivered transformative successes, but most patients with advanced cancers ultimately develop treatment resistance and succumb to progressive disease. The problem is that state-of-the-art cancer therapies target cancer antigens or enzymes and require that tumors harbor either druggable mutations or have a high-antigenic burden. However, many tumors lack these enzymatic or antigenic targets, and even when they are present, the targets can become dispensable for tumor growth, or the cells can ‘rewire’ – leading to therapy resistance and treatment failure. Consequently, even with the most intensive immuno- and polychemo-therapies humans can tolerate, many cancers are minimally treatable, incurable, or resistance-prone. A case in point is metastatic pancreatic ductal adenocarcinoma (PDAC), the 2nd leading cancer killer—it rapidly develops resistance to existing therapies, and is almost universally lethal. Fundamentally, the tumor-cell phenotype is defined by the cell’s transcriptome, which is exceptionally challenging to drug and can dynamically rewire to create drug-resistant cells. To overcome the technical barriers of drugging a shifting target, we propose to engineer molecular parasites that exploit the neoplastic cell’s transcriptome and dynamically adapt to it. Beginning with PDAC and glioblastoma (GBM), we will harness the adaptive potential and transcriptional malleability of lentiviruses to develop a safe, effective transcriptome-targeted virotherapy that we term a Neoplasm Emergent Oncolytic virus (or ‘neovirus’). The neovirus will ablate tumor cells via viral cytopathic effect and employ molecular ‘gates’ to ensure specificity. The key departure from previous oncolytic virus approaches is the chassis endowing co-evolution in concert with transcriptionally shifting tumor cells, enabling establishment of a co-evolutionary arms race to circumvent resistance. The concept will be tested in PDX mice which are superior to cell lines in recapitulating tumor heterogeneity and are a top choice to study cancer progression in vivo. These models have been shown to be predictive of clinical outcomes and are widely used for preclinical drug evaluation, biomarker identification, biological studies, and personalized medicine strategies. Clearly, such an evolving ‘living’ therapy – even one for otherwise untreatable late-stage metastatic cancers –carries unique safety risks. However, metastatic disease (e.g., PDAC and GBM) kills in months, whereas untreated lentivirus infections progress to disease slowly, over ~10 years. Moreover, engineering a neovirus from a lentivirus chassis has a unique safety advantage: a broad spectrum of > 45 approved antiretroviral drugs could provide a replicative ‘kill switch’. Finally, in contrast to current drug development which inexorably generates lead candidates with ‘locked in’ off-target toxicities, neovirus selection in ‘evolutionary reactors’ would enable therapeutic efficacy and off-target toxicity to co-adapt, enabling optimization of the ‘therapeutic window’. This new therapeutic paradigm would offer a level of controllability unmatched by existing therapies

Public health relevance public domain (NIH RePORTER)

PROJECT NARRATIVE Despite decades of drug development, most cancers become resistant to therapy and, once metastatic, are almost universally lethal. To overcome this critical problem, this project proposes a fundamental shift in therapeutic paradigm—a new class of medicines with the unique potential to co-adapt with tumors; these treatments would constitute resistance-proof therapies that are tumor agnostic.