Abstract
Abstract
Abstract 4174
The first-in-human clinical trial using Sleeping Beauty (SB) transposition for T-cell immunotherapy is under way at the University of Texas MD Anderson Cancer Center. To understand potential genotoxic risks involved in our non-viral SB transposition-based immunotherapy, we employed high throughput sequencing (Illumina) and profiling (BlueBioU supercomputer) of the SB insertion sites in T cells. Our SB clinical vector inserted 99.999% of the time at expected TA dinucleotide sites, with 44% of the insertions localizing to intragenic loci and the remainder 56% intergenic. The vast majority (96.5%) of intragenic insertions are intronic while the majority (>60%) of intergenic transpositions fall within non-coding repeat regions. By linking microarray gene expression profile to the vector insertion profile, we find that the number of transcriptional start sites (TSS) hit is proportional to the expected ‘open' conformational loci for the starting population of quiescent T cells. Compared to immunotherapy infusing T cells genetically modified with retrovirus, the integration profiles for SB-modified T cells are favored as: 1) there are more insertions within intergenic regions (56% SB vs. 44% in retroviral), 2) insertions concentrated at TA sites which have a different genomic distribution profile than retroviral insertion sites, 3) TSS associated with quiescent T cells were favored reflecting the electro-transfer of SB DNA plasmids into non-proliferating T cells vs. retroviral protocols that require transduction of activated T cells. Scanning for potential danger loci (e.g. oncogene, tumor suppressor, miRNA etc.), SB integration profiles compare favorably with retroviral insertions. In addition, we identified potential “safe harbor” genetic loci for future targeted modifications. Opportunities in immunotherapy and other personalized medicine modalities will be fully realized when the actual and perceived risks are understood and perhaps mitigated. In the absence of targeted insertions by clinical vectors, the next best option is random integration with an evenly distributed insertion profile across the whole genome to minimize hot spot and clonal expansion of therapeutic T cells. We now seek to proactively profile the T-cell products prior to infusion and follow up sampling to help clinicians in data-driven decision making in a time-sensitive manner. Our methods and findings are consistent with the human application of SB system and the harnessing high throughput supercomputing resources to help clinicians mitigate potential risks.
Disclosures:
No relevant conflicts of interest to declare.