Vector Integration and Efficacy of CD19-Directed CAR T Cell Therapy in Acute Lymphoblastic Leukemia (ALL) and Chronic Lymphocytic Leukemia (CLL)

CD19-specific Chimeric Antigen Receptor (CTL019)-engineered T-cells provide a breakthrough for personalized cancer therapy. An anti-CD19 CAR gene with 41BB costimulatory domain is delivered into patient T-cells ex vivo using a lentiviral vector, expanded in culture and then reinfused into patients. While dramatically successful for some treatment-refractory cancers, a significant proportion of patients do not experience therapeutic levels of CAR T cell expansion - thus it is important to investigate factors driving successful expansion in responders in more detail. Here we have analyzed sites of lentiviral vector integration in CAR T cells from trials to ALL and CLL, comparing successful and unsuccessful therapy in longitudinal data sets for 40 subjects. The location of each integrated vector marks a cell lineage uniquely allowing the fate mapping of individual CAR-engineered T cells in the infusion product and after adoptive transfer. We found that 81.4% of integrations had occurred in annotated transcription units which is consistent with previous reports for lentiviral vector integration sites. Relatively larger and more diverse populations of CAR-modified T-cells were associated with improved outcome (Chao1 index, p=0.043). Population sizes were also significantly more diverse in the infusion product compared with day 28 post-infusion, and more diverse at this time point when comparing responders with non-responders, or even partial responders with non-responders (p<0.05). Thus, the population size measured in the infusion product and one month after infusion forecasts the patient's response to CTL019. Vector integration can also modify activity of nearby genes, as we recently reported for an integration event in the DNA methylcytosine dioxygenase gene TET2, where clonal CTL019 expansion was associated with successful therapy (Fraietta et. al., Nature, 2018). Insertional mutagenesis was evaluated here over five criteria, including i) clonal expansion after infusion, ii) increasing frequency of unique integration sites per gene after infusion, iii) development of orientation bias, iv) long-term persistence, and v) accumulation of integration site clusters. Our analysis disclosed genes and cell pathways, including apoptosis and epigenetics, associated with superior cell proliferation and persistence. These data thus provide multiple approaches for improvement of the anti-leukemia activity of CAR T cells.