Abstract Targeting relapsed acute myeloid leukemia (AML) with engineered T cells holds exceptional promise, but obstacles to efficacy exist. Understanding resistance mechanisms is essential to improve therapy and may broadly inform cancer immune escape. We and others have developed transgenic T cell receptor (TCR) therapies targeting the intracellular WT1 oncoprotein that promotes growth of most high-risk AMLs. TCR therapies generally target epitopes processed by the proteasome for Class I presentation. As mutation or loss of WT1 or loss of HLA expression is infrequent in AML, we hypothesized resistant clones may have defective antigen processing. A 25-year-old man with AML recurrent after a 2nd allogeneic stem cell transplant received reinduction chemotherapy and transgenic donor-derived T cells recognizing the HLA-A2 restricted WT1126-134 epitope (NCT01640301). The patient experienced a 12-month remission followed by relapse despite robust T cell persistence (>6% of CD8+ T cells). The relapsed leukemia expressed WT1 and HLA-A2 but was refractory to retreatment with additional WT1126-134-restricted T cells. Single cell RNA sequencing (scRNAseq) was performed on cryopreserved PBMCs from remission (n=4976 cells) and relapsed (n=2780) time points (10x Genomics platform). During apparent remission, transgenic T cells had a transcriptome distinct from native CD8s, expressing activation markers including HLA-DR, IL32, CD69 and CTLA4, suggesting recent TCR triggering from antigen encounter. However, at the time AML progression was observed, transgenic cells clustered with native quiescent CD8s, implying they were no longer encountering the targeted epitope. Transcriptome analysis of relapsed AML demonstrated downregulation of a single subunit of the immune proteasome (β1i), which we recently demonstrated in cell lines to be critical for processing the targeted WT1126-134 epitope. This patient's refractory AML was killed by T cells restricted to an alternate HLA-A2 epitope of WT1 processed independently of β1i, confirming the functional impact of immune proteasome downregulation. At an early responding timepoint, scRNAseq detected clinically unrecognized rare circulating AML cells already β1i-low, suggesting β1i loss represented immunoediting/outgrowth of a pre-existing clone as a consequence of T cell immunotherapy pressure. ScRNAseq is poised to become an important diagnostic in immunotherapy, with sufficient depth to profile both transgenic T cells and low burden/minimal residual AML, and, in our case, successfully identified a potentially targetable mechanism of immune escape. Our data suggest T cell therapy in AML and other tumors may be improved by in depth analysis of individual leukemic cells, and from selecting target epitopes processed by both standard and immunoproteasomes and/or early combination therapy combining TCRs recognizing differentially processed epitopes. Citation Format: Kelly G. Paulson, Thomas M. Schmitt, Daniel Egan, Valentin Voillet, Miranda C. Lahman, Felecia D. Wagener, Daniel S. Hunter, Petri R. Muhlhauser, Paul C. Hendrie, Cecilia Yeung, Nathalie Vigneron, Benoit Van den Eynde, Jason H. Bielas, Merav Bar, Raphael Gottardo, Aude G. Chapuis, Philip D. Greenberg. Single cell RNA sequencing reveals AML immunoediting under pressure from engineered T cell therapy [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 2980.