Presentation Authors: Jacob W. Greenberg, Stephen Proctor, M.D*, Ibifiri Wilcox, Jonathan L. Silberstein, M.D, FACS, L. Spencer Krane M.D, New Orleans, LA
Introduction: Micro RNAs (miRNA) are short non-coding RNA which are associated with post-transcriptional regulation of gene expression. As clear cell renal cell carcinoma (ccRCC) remains the most common primary renal malignancy and accounts for the majority of more than 14,000 deaths from renal malignancies each year, identification of miRNA signatures may portend prognostic significance could help subset patients with ccRCC eventually leading to therapeutic pathways. miRNAs have demonstrated importance in intratumor communication in ccRCC, however there are no established miRNA biomarkers. In this study, we sought to create a miRNA signature to predict risk overall mortality in patients with ccRCC.
Methods: Patient&[prime]s clinical data and level 3 miRNA expression profiles were obtained from the Cancer Genome Atlas (TCGA) repository. Clinical data was correlated with miRNA expression data. Regression analysis, Kaplan-Meier curves, and Heatmap clustering were performed using R packages ComplexHeatmap, Tidyverse, and Survival. Statistical analysis was performed using R Studio v3.4.4. Significant miRNAs were isolated and a diagnostic high, medium, and low score were created correlating to miRNA expression levels of each patient.
Results: We identified 4 miRNAs (mir-204, mir-181a-1, mir-29b-1, let-7d) that significantly affected survival and created a weighted score from these. (Score = (0.776Ã—hsa-mir-204)+(-0.507Ã—hsa-mir-181a-1)+(-0.459Ã—hsa-mir-29b-1)+(-0.42Ã—has-let-7d)) We were able to subset the cohort using hierarchical clustering into 3 survival categories: low, medium, and high. The low, medium, and high score groups had 278, 84, and 103 subjects respectively. Subjects with a medium and high miRNA score show a decreased significance, p < 0.05 and p < 0.00005 respectively (Figure 1). On multivariate analysis, medium and high score along with age, nodal status, metastatic deposit, and T3 or greater lesion were all independently associated with overall survival.
Conclusions: We have created a novel miRNA signature to predict survival in ccRCC. Prospective validation of these markers is ongoing along with further determination of let-7d&[prime]s role in aggressive tumor development.