Transcriptome-based inference of immune cell subsets and pathway deregulation in rectal cancer reveals biological processes related to therapy response

SENDOYA J; ISEAS S; GOLUBICKI; ROBBIO J; CORAGLIO F; ABBA MC; PODHAJCER O; LLERA A

Congreso; AACR Annual Meeting 2018; 2018

Purpose: Colorectal cancer is the second most frequent cancer and the second and third cause of cancer-related mortality in Argentina for men and women respectively. For locally advanced rectal cancer (LARC), preoperative chemoradiotherapy (CRT) ± induction chemotherapy (iCT) comprises the standard treatment. However, only 15-27% of patients achieve a pathological complete response, while 40-60% respond variably. This scenario poses a strong need for biomarkers able to predict response to CRT and provide guidance towards optimal treatment for each patient. Here we explore LARC gene expression data to identify baseline immune populations and pathways that might be associated with response to CRT.Methods: Using Agilent Human Genome Microarrays, we analyzed gene expression in pre-treatment tumor biopsies of 26 LARC patients as part of an ongoing prospective translational study in Argentina. After treatment with CRT (capecitabine + radiotherapy) ± iCT (capecitabine + oxaliplatin) and surgery, pathological response was assessed according to College of American Pathologists (CAP) guidelines. We grouped CAP scores 0-1 as responders and 2-3 as nonresponders. We estimated cellular heterogeneity of the microenvironment from transcriptomic data by applying two independent computational methods: GSVA and xCell, and inferred patient-specific pathway deregulation with PARADIGM. To elucidate differential features between both patient groups, we performed Two-sample t-Tests for GSVA and xCell, and Rank Product test for PARADIGM. All tests were adjusted for FDR<0.01.Results: Of 26 analyzed patients, we focused on 16 with currently available response data. T-test yielded 98 differentially expressed genes. From the cell type estimators, among the responder group we observed upmodulation of genes related to B-cells (p-GSVA: 0.0001; p-xCell: 0.004), mast cells (p-GSVA: 0.038; p-xCell: 0.029), activated dendritic cells (p-GSVA: 0.035) and effector memory CD8+cells (p-xCell: 0.034). PARADIGM revealed upmodulated activities in responders such as ?DNA damage', miR34A (regulation of tumor suppression), and TP73 (apoptotic response to DNA damage). Downmodulated pathways included E2F1/3 (transcriptional/cell cycle activation), PARP1 (DNA repair) and IL8 (tumorigenic/proangiogenic chemokine). For all cases, FDR-adjusted p < 0.0001.Conclusions: Considering the caveats regarding our current sample size, we could detect higher involvement of adaptive immunity cell populations, upmodulation of proapoptotic features, and downmodulation of DNA repair and proliferation activities in baseline biopsies of responders when compared to nonresponders. Overall, a cell cycle-decelerated tumor with poor DNA repair abilities immersed in a proapoptotic environment seems to be the one in which CRT would be more effective. With necessary validation of these analyses underway, we envision this effort as the basis on which we will evaluate genomic biomarkers for response prediction in our LARC population.Citation Format: Juan M. Sendoya, Soledad Iseas, Mariano Golubicki, Juan Robbio, Mariana F. Coraglio, Ubaldo Gualdrini, Ana Cabanne, Mirta Kujaruk, Vanesa Mikolaitis, Mariana Rizzolo, Ruben Salanova, Cecilia Rotondaro, Martin Abba, Osvaldo Podhajcer, Enrique Roca, Andrea Llera. Transcriptome-based inference of immune cell subsets and pathway deregulation in rectal cancer reveals biological processes related to therapy response [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 LB-374.