Performance Evaluation of a Next-Generation Sequencing-Based T-Cell Receptor Gene Rearrangement Assay

Abstract

T-cell receptor (TCR) gene rearrangement clonality studies help resolve atypical T-cell proliferations in the context of suspected malignancy. However, the interpretation criteria for this assay using a next-generation sequencing (NGS) platform have not been extensively explored and standardized. Thus, this project assessed the current performance of the Stanford Health Care in-house NGS-based TCR clonality diagnostic testing with the goal of optimizing the interpretation criteria and identifying recurrent analytical challenges. The current assay identifies a predominant clonotype when its sequence comprises at least 2.5% of the total reads with at least 5× fold change from the background. Using concurrent pathology reports as the clinical truth, this project analyzed 619 cases and determined that the current assay performs at 74% sensitivity and 85% specificity. Receiver operating characteristic analysis identified an optimized interpretation criterion that only improved the diagnostic yield marginally compared with the preexisting algorithm. Further clinicopathologic evaluation of discordant cases revealed that discrepancies mostly arose from technical limitations or the underlying nuanced biology of the diagnosis. Overall, this study provides an objective approach in establishing the interpretation criteria of the current NGS-based TCR clonality test and offers a roadmap for other laboratories considering implementing a similar assay.