Abstract
Tumor antigen-driven selection may expand T cells having T cell receptors (TCRs) of shared antigen specificity but different amino acid or nucleotide sequence in a process known as TCR convergence. Substitution sequencing errors introduced by TCRβ (TCRB) repertoire sequencing may create artifacts resembling TCR convergence. Given the anticipated differences in substitution error rates across different next-generation sequencing platforms, the choice of platform could be consequential. To test this, we performed TCRB sequencing on the same peripheral blood mononuclear cells (PBMC) from individuals with cancer receiving anti-CTLA-4 or anti-PD-1 using an Illumina-based approach (Sequenta) and an Ion Torrent-based approach (Oncomine TCRB-LR). While both approaches found similar TCR diversity, clonality, and clonal overlap, we found that Illumina-based sequencing resulted in higher TCR convergence than with the Ion Torrent approach. To build upon this initial observation we conducted a systematic comparison of Illumina-based TCRB sequencing assays, including those employing molecular barcodes, with the Oncomine assay, revealing differences in the frequency of convergent events, purportedly artifactual rearrangements, and sensitivity of detection. Finally, we applied the Ion Torrent-based approach to evaluate clonality and convergence in a cohort of individuals receiving anti-CTLA-4 blockade for cancer. We found that clonality and convergence independently predicted response and could be combined to improve the accuracy of a logistic regression classifier. These results demonstrate the importance of the sequencing platform in assessing TCRB convergence.