AI-04 Immunotherapy for lupus in a mouse model to define pathogenesis and therapeutic targeting

Background Recent advances in immunotherapy using genetically modified chimeric antigen receptor (CAR) T cells have made it possible to selectively and completely eliminate cells expressing specific cell surface targets. Such a system could potentially be applied to lupus and other autoimmune diseases for identifying new targets in model systems and for therapy. To test the possible utility of such an approach, we therefore conducted a preliminary study to determine the efficacy of anti-CD19 CAR T cell in eliminating B cells and disease development in a spontaneous mouse model of lupus. Methods In these experiments, a modulatable CAR T cell system was used consisting of CAR T cells recognizing an antigenic site on a soluble anti-CD19 Fab ‘switch.’ Thus, activation and killing activity of CAR T cells was dependent on the presence of an independently injected anti-CD19 switch. To study CAR T cell efficacy, lupus-prone male BXSB mice at 3 months of age were initially conditioned with cyclophosphamide i.p., then 24 hour later given CAR T cells and either the anti-CD19 switch or PBS alone every other day (3 mice/group). Mice were followed for mortality, autoantibodies, proteinuria, and peripheral blood and spleen cells populations up to 9 weeks after CAR T cell transfer. Immunoglobulins and autoantibodies in sera were measured by ELISA. Flow cytometry was used to analyze immune cell populations and included a FITC-conjugated switch peptide to identify CAR-expressing T cells. Proteinuria was determined by dipstick and kidney sections were PAS-stained and scored for glomerulonephritis on a 0–4 scale. Results An initial experiment documented that a single i.p. injection of 100 mg/kg of cyclophosphamide at 3 months of age did not reduce the development of lupus in BXSB male mice compared to PBS controls (4 and 3 mice/group). When mice treated with conditioning and CAR T cells were analyzed, the group given anti-CD19 switch but not PBS, had low levels of circulating B cells by one week after CART T cell transfer and for the duration of the experiment. Immunoglobulin and autoantibody levels were present in the PBS group, but undetectable in the anti-CD19 switch group at the end of the 9 week study. All PBS-, but none of the anti-CD19 switch-treated mice group developed severe glomerulonephritis (glomerulonephritis scores: 3.1±0.07 vs 0.43±0.23, p Conclusions In a small study, anti-CD19 CAR T cell treatment of lupus was highly effective in preventing the development of severe lupus glomerulonephritis. Strikingly, at 9 weeks after transfer, there was complete deficiency of circulating immunoglobulins suggesting that long-lived plasma cells either express sufficient levels of CD19 to be targeted by CAR T cells or less likely that the plasma cell population in lupus requires replenishment from newly generated B cells. These findings support the possibility of using the switchable CAR T cell approach to define the role of immune cell subsets in lupus and treatment of severe lupus. Acknowledgements This work was supported by grants from the NHLBI, NIAMS, and NCI.