Abstract 1895: Identifying clonal T-cell receptor sequences and monitoring recurrent/persistent disease by T-cell receptor repertoire profiling in patients with mature T-cell neoplasms

This study demonstrates the broad potential of high-throughput sequencing of T-cell receptors to contribute to the post-therapeutic monitoring of T cell neoplasia in mature T-cell neoplasms. Identification of recurrent or persistent disease in T-cell neoplasms is important for individualized patient care. While patients with T-cell lineage lymphomas and leukemias are a small subset of all lymphoma and leukemia patients, the incidence of refractory disease in these patients can be higher than patients with B-cell lineage neoplasms. We recently developed a method to sequence the diversity of the TCR CDR3 rearrangements (Blood. 2009; 114(19): 4099-107) that exploits the capacity of high-throughput sequencing (HTS) to document the diverse repertoire of TCRB CDR3 chains simultaneously. These assays can describe both the breadth of the T-cell repertoire and quantify individual clones. This technology thus provides a potential opportunity to identify and then track the presence and frequency of clones in the context of an evolving, adaptive immune system, during the course of ongoing therapy. We amplified the TCR repertoire of 98 index samples to identify high-frequency TCRB or TCRG rearrangements. Clones were classified as neoplastic if occurring at a proportion greater than 7 standard deviations above the mean frequency of the most abundant rearranged TCRB in control samples of blood, bone marrow, or lymphoid tissues. Eight-four percent of index samples, based on these criteria, had a tractable TCR rearrangement. Samples that lacked a detectable TCRB clone were excluded. For 35 patients, at least one subsequent follow-up sample was available. For these 59 samples, we sequenced the TCRB repertoire to screen for the corresponding index clone. We find that for most samples, high-throughput sequencing concurs with currently available, routine clinical measures of disease, such as clinical flow cytometry or PCR-based evaluation of TCRG rearrangement. High-throughput sequencing of TCRB was concordant in 46 samples with identification of the index clones and in 7 additional samples without the identification of the index clones. However, 5 of 59 samples were only positive for recurrent disease based on HTS only, and 1 of the 59 samples was only positive for recurrent disease based on current diagnostic technology but not HTS. We find that for most disease diagnoses - high-throughput sequencing identifies a tractable clone. In addition, we find that for most samples, h