P14.04 cfRNA from Liquid Biopsies Is More Abundant Than cfDNA, Informs Treatment Outcome and Is Concordant with Tissue

Abstract

The most analyzed nucleic acid in liquid biopsies is cfDNA due to its presumed greater stability compared with cfRNA. However, sometimes it is difficult to isolate sufficient tumor cfDNA from blood for analyses. In theory, cfRNA representing a particular gene expression should be present at much higher levels than the corresponding gene in cfDNA because a single gene is normally transcribed many times. cfRNA should contain only those mutations that are consequential for tumor development and also allows the measurement of tumor gene expressions. Here, we compared levels of cfDNA and cfRNA isolated from patients’ blood, to assess the stability of the cfRNA and its utility as a biomarker to follow therapy outcomes in non-small cell lung (NSCLC) cancer. 137 blood samples (20cc) from 54 pts with stage IIIB/IV NSCLC were drawn into both RNA and DNA BCT Tubes (Streck) pre-treatment and at three month intervals or progression. Levels of cfRNA, cfDNA, KRAS and PD-L1 gene expression were quantitated by qPCR and correlated with pts response (CR/PR/SD/PD), as determined by CT scans at the same time. Whole transcriptome RNA seq analysis was performed on 87 tissue-plasma paired samples in a separate study. There was a strong association of PD-L1 with Outcome to Immunotherapy in Tissue and Blood. 80% of pts positive for PD-L1 in tissue (9/11) were positive for PD-L1 in blood; 83% of these patients responded to immunotherapy with a PR. Also 46% of pts negative for PD-L1 in tissue (6/13) were also negative in blood. Pts negative for PD-L1 in blood only did not respond to immunotherapy with a PR. Changes in cfRNA and cfDNA as surrogates for monitoring disease status were seen when levels of cfDNA were significant. Levels of these nucleic acids exuded into the blood stream increased as patients progressed (PD) and decreased or remained stable as patients responded to therapy with a PR or SD. However, when cfDNA levels were at the limit of detection, changes in cfRNA levels still accurately predict patient outcome to disease. In 87% of treatment segments (55/63), changes in cfRNA levels were concordant with CT scans, whereas cfDNA levels were concordant with CT scans in 62% (24/63) of treatment segments. Finally our concordance analysis of NGS done in cfRNA in Tissue and Blood compare representative biomarker expressions (CD86, PD-L1, TIGIT, CD28, TIM3, CTLA4, LAG3 PDL2, PD1, FOXP3, IDO, OX40) in paired tissue and plasma samples using RNAseq whole transcriptome analysis (WTA). The qualitative and quantitative concordance between tissue and plasma gene expressions reached 84% and 76% respectively. CfRNA levels average 20 x higher than cfDNA levels and can track treatment outcome similarly to cfDNA even when cfDNA is limiting. cfRNA yields data on specific gene expression drug targets such as PD-L1 and is concordant with IHC. PD-L1 gene expression from cfRNA can accurately monitor response to immunotherapy in NSCLC. Levels of gene expressions in tissue and blood measured by RNAseq WTA are concordant.