Please note that we have stopped the regular imports of Gene Expression Omnibus (GEO) data into ArrayExpress. This may not be the latest version of this experiment.
E-GEOD-63850 - Tumor infiltrating IL-33 producing effector-memory CD8+ T cells in resected hepatocellular carcinoma prolong patient survival
Released on 8 December 2014, last updated on 2 January 2015
Background: IL-33, a cytokine with pleiotropic functions, is elevated in serum of patients with hepatocellular carcinoma (HCC). This study investigated effects of local IL-33 expression in resected HCC on patient survival and on immunological and molecular tumor microenvironment. Methods: Tissue of resected HCCs was stained for H&E, masson trichrome, alpha smooth muscle actin, IL-33, CD8 and IL-13 and analysed by flow cytometry. Besides histomorphologic evaluation, the immunohistochemical stainings were analysed for the respective cell numbers separately for tumor area, infiltrative margin and distant liver stroma. These findings were correlated with clinical data and patient outcome. Further, gene expression of different HCC risk groups was compared using Micro Arrays. Results: In multivariable analysis, infiltration of HCCs by IL-33+ cells (P=0.032) and CD8+ cells (P=0.014) both independently were associated with prolonged patient survival. Flow cytometry demonstrated that cytotoxically active CD8+CD62L-KLRG1+CD107a+ effectory-memory cells are the main producers of IL-33 in these HCC patients. Using infiltration by IL-33+ and CD8+ cells as two separate factors, a HCC immune score (HCCIS) was designed and evaluated that stratified patient survival (P=0.0004). This HCCIS identified high and low risk patients who differ in gene expression profiles (P<0.001). Conclusion: Infiltration of HCCs by IL-33+ and CD8+ cells is independently associated with prolonged patient survival. We suggest that this is due to an induction of highly effective cytotoxically active CD8+CD62L-KLRG1+CD107a+ effector-memory cells producing IL-33. Based on these two independent factors we established a HCC immune score that provides risk stratification for HCC patients and can be used in the clinical setting. To investigate if HCCIS 0 high risk and HCCIS 2 low risk tumors exhibit a distinct molecular environment and gene expression pattern, RNA from fresh tumor tissue was isolated and analyzed by whole genome Microarray 4 patients with low risk tumors (HCCIS 2) were compared to 4 patients with high risk tumors (HCCIS 0).
transcription profiling by array
Stefan Brunner <email@example.com>, Andreas Teufel, Christoph Rubner, Hans J Schlitt, Maria Martin, Petra Ruemmele, Rebecca Kesselring, Stefan Fichtner-Feigl, Stefan M Brunner, Thomas Stempfl