Recipient T-Cell Repertoire Diversity after Double Umbilical Cord Transplantation Correlates with Non-Relapse Mortality and Survival

Complications of allogeneic stem cell transplant, such as infection, relapse and GvHD, are linked to failure of balanced immune reconstitution.  Early quantification of immunologic function is crucial to identify those at risk and institute pre-emptive therapies.  We hypothesize that T-cell receptor (TCR) diversity after double umbilical cord blood transplantation (dUCBT) will predict clinical outcomes.  Methods: 42 adult dUCBT recipients received fludarabine, melphalan, ATG with tacrolimus/sirolimus as GvHD prophylaxis.  TCRb CDR3 regions were amplified from unsorted PBMCs and sequenced using the ImmunoSeq platform (Adaptive Biotechnologies) at 3, 6 and 12 months after dUCBT.  Samples analyzed had predominantly donor T-cell chimerism and no morphologic relapse.  We analyzed 34 samples at 3 mos, 26 at 6 mos and 23 at 12 mos.  Entropy/Shannon's diversity index and richness/Fisher's alpha diversity were calculated from number and frequency of TCRb sequences and normalized for sequencing depth.  Higher entropy and richness, measures of uncertainty and unseen species within a complex population, reflect increased diversity.  Results: Median age was 52 years (19-67).  57.1% were in remission at dUCBT.  Primary indications for dUCBT were AML, NHL and MDS. Median follow-up among survivors was 47 months (18-91).  4-year OS was 41%; 4-year PFS was 26%.  As a whole, dUCBT recipients increased entropy and richness of their T cell repertoire in the first year.  We categorized subjects into 3 groups: relapse, NRM, or alive with neither event.  In the NRM cohort, median entropy and richness values rose slightly from 3 to 6 mos then fell at 12 mos.  In comparison, median entropy and richness were higher in the no event cohort between 3 and 12 mos and rose substantially to diversity levels found in healthy donors.  (p=0.01, 0.09, 0.004 for entropy at 3,6,12 mos; p=0.06, 0.11, 0.006 for richness at 3,6,12 mos). (Figure)  Entropies of subjects who relapsed did not differ from those of the no event cohort (not depicted).  In a similar analysis, entropy values for subjects with chronic GvHD fell between 3 and 12 mos and were significantly lower than in subjects without any cGvHD, relapse, or NRM at 12 mos (median entropy 5.9 vs 12.3, p=0.04).  In multivariable analyses controlling for age, patient sex, disease status, and diagnosis, high entropy was associated with lower NRM (HR 0.68, p<0.001), PFS (HR 0.79, p=0.009) and OS (HR 0.73, p=0.005).  High richness (log10 transformed) was associated with lower NRM (HR 0.3, p<0.001) and OS (HR 0.30, p=0.0003).  Conclusion: T-cell repertoire diversity as early as 3 months after dUCBT is significantly lower in subjects who eventually experience NRM or cGvHD and does not expand with time after transplant.  These quantitative predictive indices may be used to identify dUCBT recipients who merit more intensive monitoring or prophylactic strategies against infection and GvHD.