In order to study the clinical impact of T-cell receptor (TCR) diversity in the setting of umbilical cord blood transplantation (UCBT), we retrospectively analyzed samples from 76 patients in 2 independent study cohorts at separate institutions. At Fred Hutchinson Cancer Research Center (FHCRC), we followed 34 patients with hematological malignancies who underwent myeloablation and primarily double UCBT (2 single cord). This cohort was composed of 11 pediatric and 23 adult patients (median age, 26.5yrs), primarily with acute leukemia (n=26), 50% of whom had evidence of MRD at UCBT. They received fludarabine, cytoxan or treosulfan, and total body irradiation (TBI) with cyclosporine and mycophenylate mofetil as GvHD prophylaxis. At Dana-Farber Cancer Institute (DFCI), we followed 42 adult patients (median age, 52.3 yrs), 19 with acute leukemia, 18 of whom had evidence of disease at UCBT. All patients received reduced-intensity conditioning with fludarabine, melphalan, and anti-thymocyte globulin (ATG) followed by double UCBT with tacrolimus and sirolimus as GvHD prophylaxis. DFCI participants were selected to have predominantly cord T cell chimerism. We analyzed peripheral blood pre-transplant, and at 1, 2, 3, 6 and 12 months after UCBT, based on sample availability.

We performed high-throughput sequencing of rearranged (TCR) loci to track presence and frequency of individual T cell clones in each patient across time-points, as well as to calculate estimated diversity of the TCR repertoire as a whole. We correlated our measure of TCR repertoire diversity with clinical outcome, using one-year survival as our primary endpoint.

Diversity of the T-cell repertoire can provide a measure of immune competence. We hypothesized that restoration of TCR repertoire diversity would be associated with overall survival after UCBT. For the combined group of 63 patients with samples available 3 months after UCBT, 19 subsequently died within 1 year of transplant. These patients had significantly lower TCR repertoire diversity at 3 months than patients who survived at least one year (data are given as percentiles within each cohort due to differing amounts of blood available for analysis; median 28th percentile in patients who died vs. median 56th percentile for patients who survived 1 year, p-value = 0.003 by permutation test). When analyzed as 2 independent cohorts, the 29 patients at FHCRC with samples at 3 months after UCBT demonstrated significantly lower TCR repertoire diversity in those who died within 1 year of UCBT versus those who survived beyond 1 year (median TCR diversity 12,000 vs. 27,000, p-value = 0.015), as did the 34 patients with samples available 3 months after transplant at DFCI (median TCR diversity 1,200 vs. 6,400, p-value = 0.027). Between 100 days and 1 year after dUCBT 6 patients died of non-relapse causes (NRM) and 4 with relapse at FHCRC and 6 patients died of NRM and 3 with relapse at DFCI. Correlation of TCR diversity with clinical outcomes such as NRM, relapse, GvHD and correlation with clinical outcomes beyond 1 year are ongoing.

In this study, we demonstrate that measurement of TCR repertoire diversity generated using high-throughput sequencing genes at 3 months after UCBT is significantly correlated with subsequent mortality within the first year. This correlation was demonstrated separately in two unrelated cohorts of UCBT recipients at different transplant centers of differing ages and conditioning regimens and in combined analysis of all recipients. Low T cell diversity in the peripheral blood as soon as 3 months may therefore be an early indicator of inadequate immune reconstitution and could potentially be used to tailor monitoring and therapy after UCBT in several clinical contexts.


Emerson:Adaptive Biotechnologies: Employment, Equity Ownership. Sherwood:Adaptive Biotechnologies: Employment, Equity Ownership. Carlson:Adaptive Biotechnologies: Consultancy, Equity Ownership, Patents & Royalties. Robins:Adaptive Biotechnologies: Consultancy, Equity Ownership, Patents & Royalties.

Author notes


Asterisk with author names denotes non-ASH members.