Abstract 4758: Use of highly multiplexed reference materials to facilitate validation of a clinical NGS tumor fusion RNA assay

Abstract

Abstract Introduction: Next Generation Sequencing assays for detection of tumor RNA fusions must undergo rigorous validation before clinical implementation. Validations include assessment of the assay's accuracy, precision, reproducibility and limits of detection or reportable range. Obtaining samples with all needed variants is difficult and time consuming. When found, they are often in limited quantities such that repeated testing for precision and reproducibility studies is not possible. Sample heterogeneity or lack of characterization further complicates interpretation of results. This study demonstrates how highly characterized, uniformly manufactured reference materials can facilitate clinical NGS assay validation. Methods: An RNA panel representing decreasing levels of 16 different RNA fusions and exon skipping events compared to the total cellular RNA was generated. GM24385 cell line was engineered to contain the fusion RNAs then formalin fixed, and the total RNA was extracted using the Maxwell RSC FFPE RNA kit. The fusion RNA was serially diluted into total RNA from similarly processed non-engineered GM24385 cells to create a panel with decreasing levels of each fusion. The panel was characterized by fusion-specific, TaqMan digital PCR assays to obtain a “truth set” on which to compare NGS results. Each panel member was tested in triplicate on three separate days as part of a validation study. Results: Precision and reproducibility were assessed quantitatively through analysis of the NGS unique reads. The intra-run precision varied among fusion targets with many targets having %CV of less than 15%, while a few targets had significantly higher variability between replicates. The linearity of the assay was good, with R-squared value greater than 0.95. Lower limits of detection for each fusion target were estimated using the digital PCR data and were different for different fusion targets; for example, detection of EML4-ALK fusion was five-fold more sensitive than detection of CD74-ROS1 fusion. Conclusions: Manufactured reference materials can supplement validation studies and their uniformity and digital PCR characterization allows greater insight into assay performance such as limits of detection than use of remnant specimens alone. The sensitivity of the NGS panel used was not the same for different fusion targets and, and underscores the need for highly multiplexed reference materials and for labs to test all clinically important fusions during validation Citation Format: Catherine Huang, Subit Barua, Deepika Philkana, Russell Garlick, Bharathi Anekella, Helen Fernandes. Use of highly multiplexed reference materials to facilitate validation of a clinical NGS tumor fusion RNA assay [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 4758.