Vesicular Stomatitis Virus (VSV) Engineered to Express CD19 Stimulates Anti-CD19 Chimeric Antigen Receptor Modified T Cells and Promotes Their Anti-Tumor Effects

Although CD19-directed chimeric antigen receptor T cell (CART19) therapy is highly effective and was FDA approved for certain B-cell malignancies, most patients relapse after CART infusion within the first 1-2 years due to inadequate CART expansion in vivo. Vesicular stomatitis virus (VSV) has the ability to infect and lyse cancer cells. Clinical trials of VSV oncolytic therapy indicate that VSV efficiently infects cancer cells as well as innate immune cells. Therefore, we hypothesized that in patients who achieve suboptimal response to CART19, VSV engineered to express CD19 will augment anti-tumor activity through 1) direct lysis of cancer cells and 2) infecting cancer cells and innate immune cells with CD19 to further stimulate CART19.

To test our hypothesis, human CD19 or GFP (control) was engineered between the glycoprotein and large-protein (Fig.1A) in a modified VSV backbone. A matrix inactivating mutation (M51R) rendered it incapable of suppressing anti-viral reactions of infected targets, potentially promoting its immunogenicity.

First, we tested the anti-tumor activity of VSV-CD19 and VSV-GFP against the luciferase (luc)⁺CD19⁺ acute lymphoblastic leukemia cell line NALM6 and the luc⁺CD19^- acute myeloid leukemia cell line MOLM13. VSV-CD19 and VSV-GFP successfully lysed NALM6 (Fig.1B) or MOLM13, both in vitro and in vivo (data not shown). Next, we investigated the efficiency of VSV-CD19 in infecting tumor and immune cells. 24 hours after exposure to VSV-CD19 or VSV-GFP, we analyzed the surface expression of CD19 on MOLM13 and revealed efficient CD19 delivery (Fig.1C). Next, we assessed VSV infection of peripheral blood mononuclear cells (PBMCs) from healthy donors (HDs). Freshly isolated HD PBMCs were infected with VSV-CD19 for 6 hours and subsequently assessed for CD19 expression. Consistent with findings from clinical trials, VSV-CD19 selectively infected and induced CD19 expression on monocytes while other cells were not affected (Fig.1D). To exclude potential toxicities against CART19, we co-cultured CART19 with VSV-CD19 or VSV-GFP using second-generation 4-1BB costimulated CART19. Both VSV-CD19 and VSV-GFP did not infect CART19 as evident by preservation of CART19 viability and lack of CD19 or GFP expression (Fig.1E).

Having demonstrated that VSV-CD19 specifically delivered CD19 to monocytes, we next tested whether the infected monocytes stimulated CART19. VSV-CD19 infected monocytes induced potent antigen-specific proliferation of CART19 (Fig.1F) and resulted in enhanced anti-tumor activity against luc⁺NALM6 in vitro (Fig.1G).

Next, we aimed to confirm these findings in vivo. We generated luc⁺CART19 to track CART19 expansion in vivo. Freshly isolated HD monocytes were infected with VSV-CD19 ex vivo. After 4 hours, VSV-CD19 was washed away and immunocompromised NSG mice were intravenously injected with VSV-CD19 infected monocytes. After 24 hours, 3.5x10⁶ of luc⁺untransduced T cells (UTD) or luc⁺CART19 were injected intravenously. The T cell expansion was assessed by bioluminescence imaging (BLI). VSV-CD19 infected monocytes specifically stimulated and expanded CART19 (Fig.1H).

Finally, we tested whether VSV-CD19 can stimulate and rescue suboptimal anti-tumor effects of CART19 in vivo using a NALM6 relapsed model. Here, 1x10⁶ luc⁺NALM6 were injected intravenously into NSG mice on day -6. At day -1, mice were imaged and randomized according to tumor burden to receive 1x10⁶ UTD or CART19 on day 0. Subsequently, at day 4, mice were re-imaged and randomized. At day 5, HD monocytes were injected intravenously. Three hours after administering monocytes, mice received 1x10⁷ VSV-CD19 or VSV-GFP (Fig.1I). BLI revealed that CART19 plus VSV-CD19 showed better tumor control than CART19 monotherapy or CART19 plus VSV-GFP (Fig.1J-K). Furthermore, CART19 plus VSV-CD19 exhibited long-term survival (Fig.1L).

In summary, VSV-CD19 not only demonstrated direct anti-tumor effects but also specifically delivered CD19 to monocytes and tumor cells, thereby re-stimulating and enhancing the anti-tumor activity of CART19. This work provides a rationale to study VSV-CD19 in patients who demonstrate only suboptimal response to CART19. This approach could also be applied to augment CART therapy in other tumors.