Prodrugs are harmless in their native state, as they are not targeted by human enzymes. But they can be converted into highly toxic compounds (the “drug”) by viral or bacterial enzymes. Thus, provided that the delivery of the gene encoding the activating enzyme is confined to the cancer, DDEPTs hold the promise of avoiding the general and severe toxicity of conventional chemotherapeutic approaches, most of which lack specificity and besides killing the cancer cells, also damage normal tissues and organs. The prodrug regimen Herpes simplex virus type 1 (TK1) thymidine kinase and ganciclovir (HSV-tk/GCV) was tested in phase III clinical trial; the TK1 gene from a retroviral vector was injected by craniotomy. No benefits resulted. The reasons included low-level and short-lived gene expression, lack of enzyme potency, insufficient attention to generate a strong bystander effect; also, there are concerns about the use of viruses: immune recognition, insertional mutagenesis, and inflammatory toxicity. Similar reasons account for the failure to proceed beyond phase I/II stages of clinical trials of CB1954 (tretazicar) + Escherichia coli nitroreductase (NTR) GDEPT for treating prostate cancer – the NTR encoding adenovirus was injected directly into the localized cancer.  Our GDEPT employs a new prodrug, CNOB, and a new bacterial enzyme, which we discovered, improved and humanized (“HChrR6”). This enzyme converts CNOB into the drug MCHB. The latter is not only highly lethal, it is strongly fluorescent and can be visualized in living mice. This greatly helped in ensuring that the activation of CNOB was confined to orthotopic implanted HER2 xenografts in mice. We utilized extracellular vesicles (EVs) as vehicles for gene delivery, which, being nature’s own antigen delivery system, circumvent the issues of viral use. Further, we utilized mRNA for gene delivery instead of DNA used in the above trials; mRNA is superior to DNA for gene delivery and for rapid expression. We also quantified the degree of transfection with the gene (as mRNA) of tumor cells for an effective BE. mRNA was loaded in HEK293 generated EVs by a novel plasmid that we generated; this was the first time that functionally competent exogenous mRNA loading into the EVs was accomplished. The EVs displayed a novel protein (EVHB) that binds to the EV surface and displays a high affinity anti-HER2 scFv, thus targeting the EVs specifically to the HER2 receptor; these loaded and directed EVs are termed “EXO-DEPTs. Ip injection of EXO-DEPTs into mice with implanted HER2+ breast cancer xenografts, along with iv CNOB resulted in near-complete arrest of tumor growth. Thus, GDEPT was accomplished using systemic delivery, eliminating the limitation imposed by the current approaches that necessitate gene injection directly into the tumor site. To enhance translatability of this approach we have accomplished this with the prodrug CB1954 (paper in submission), for which a safe dose in humans has been established, and with 220-fold less EVs than before. Further, we have used in vitro synthesized mRNA to load the EVs; this eliminates the potential toxicity that inclusion in the EVs of the plasmid genetic material might cause. No notable off-target toxicity was seen. Work is in progress to test this regimen in immune-competent mice, and so are IND-enabling studies. We hope to conduct first-in-human, phase 1 clinical trials using EVs obtained from patients own dendritic cells; we have successfully done this before.

Learning Objectives: 

1. Use of extracellular vesicles in cancer treatment
2. How to make cancer treatment directed to the tumor so as to avoid off-target toxicity of conventional chemotherapy.