The Relevance of CD20 Targeting of Oncolytic Measles Virus in the Treatment of Adult Acute Lymphoblastic Leukemia

The Edmonston-B, vaccine strain of measles virus (MVNSe) has oncolytic activity in numerous preclinical tumour models with safety and some efficacy being demonstrated in phase 1 clinical trials for multiple myeloma and cutaneous non-Hodgkin lymphoma.

Adult acute lymphoblastic leukemia (ALL) is an aggressive malignancy where long term survival is achieved in only 50% of patients. The GRAALL study showed that anti-CD20 monoclonal antibody, rituximab, improved outcome in ALL and the results of the UKALL14 trial are eagerly anticipated in this regard. To determine whether targeting MV to the CD20 antigen can recapitulate some of these benefits we used a MV expressing a single chain Fv antibody against CD20 as a C-terminal extension of the H glycoprotein. We also used targeted MV 'blinded' to the native MV receptors previously cloned by site-directed mutagenesis. The mechanisms of MV oncolysis remain incompletely understood and our experiments were particularly designed to help clarify the role of the innate immune system.

First, we retrovirally transduced CD20 negative B-precursor ALL cell line NALM6 cells to express human CD20. The cells were flow sorted to give 'high' and 'low' expressing populations based on mean fluorescence intensity. Viral replication was measured by TCID₅₀ at consecutive 24 hourly time-points and MV-N mRNA/GADPH by RQ-PCR showing similar growth curves and MV-N mRNA levels for the targeted and parent virus. Oncolytic activity was measured by trypan blue exclusion at 24, 48, 72 and 96 hours post infection (hpi). Infection of CD20 negative NALM6 cells by MVHaCD20 virus showed similar levels of killing to the parent virus MVNSe, and there was significantly greater levels of killing by MVHaCD20 of the CD20 positive cells when compared to MVNSe ( p =0.03). Since it is also possible for the MVHaCD20 to enter cells via native MV receptors CD46 and CD150 (SLAM) we investigated the effect of 'blinding' the virus to these receptors. As expected, there was a small, non-significant reduction in efficacy in the CD46 blinded CD20 virus, MVHaCD20CD46blind, but no difference with the CD150 blinded virus, MVHaCD20SLAMblind, as CD150 is not expressed on NALM6 cells.

Our previous work has shown that neutrophils are important effectors in mediating MV oncolysis. To further investigate specific mechanisms, we carried out a FACS-based neutrophil phagocytosis assay in which target cells were labelled with PKH67 and the percentage phagocytosis determined by the frequency of dual expressing CD15+/PKH67+ cells relative to the total number of CD15+ cells. These assays demonstrated that in the presence of anti-MV antibody a similar level of phagocytosis was seen 24 hpi with MVNSe, at a multiplicity of infection of 1.0, to uninfected CD20 expressing NALM6 cells in the presence of rituximab (53% c/w 58%). The percentage phagocytosis strongly correlated with the anti-MV IgG concentration from individual donors' serum (Pearson's correlation, r =0.937; p <0.001) The level of phagocytosis was significantly increased when complement was active (student t-test, p =0.008). There was no increase in cell death as indicated by DAPI staining.

To further investigate the role of complement we carried out complement cytotoxicity (CDC) assays at 2hpi and 24hpi. As CD46 is a negative regulator to complement the CD46 blind viruses were included in the comparison. The results showed CDC to be a minor mechanism at 24hpi causing 5-10% cell death which was less effective than rituximab which resulted in 13-17% cell death. There was no additional effect with our 'blinded' viruses.

In conclusion, CD20 targeted MV has increased oncolytic efficacy in-vitro compared to the native MV when infecting CD20 expressing NALM6 cells. Our results suggest that ADCP is a mechanism of MV oncolysis however a recent study in chronic lymphocytic leukemia (Valgardsdottir et al Blood 2017) has indicated that this effect is 'trogocytosis' whereby the neutrophils take up small amounts of the cell membrane rather than engulfing the entire leukemic cell. In our study complement was implicated both in increasing the efficacy of this mechanism and also in CDC. Although this requires further exploration in ALL, it remains that neutrophils appear to be an important effector cell of MV oncolysis in the presence of anti-MV antibody. Overall our data suggest that targeting MV to CD20 could have some efficacy in the treatment of ALL.