Secondary Logo

An Early Naïve T Cell Population Lacking PD1 Expression at Day 100 As A Prognostic Biomarker of Chronic GVHD

Li, Amanda M.1,2; Drissler, Sybil3; Kariminia, Amina4; Subrt, Peter4; Brinkman, Ryan R.3,5; Schultz, Kirk1,4

doi: 10.1097/

1Division of Pediatric Hematology, Oncology, and BMT, British Columbia Children's Hospital, Vancouver, BC, Canada; 2Department of Pediatrics, University of British Columbia, Vancouver, BC, Canada; 3Terry Fox Laboratory, British Columbia Cancer Agency, Vancouver, BC, Canada; 4Michael Cuccione Childhood Cancer Research Program, Child and Family Research Institute, Vancouver, BC, Canada; 5Department of Medical Genetics, University of British Columbia, Vancouver, BC, Canada.

Chronic graft-versus-host disease (cGVHD) remains one of the leading causes of morbidity and non-relapse mortality in patients surviving allogeneic hematopoietic stem cell transplantation (HSCT) beyond 100 days. T cells have a well-recognized role in cGVHD, although distinct subpopulations of T cells likely play different roles in mediating and modulating cGVHD. There is a need to define more precise biomarkers of cGVHD that are predictive of clinical outcome.

Methods: Fifty-two peripheral blood mononuclear cell (PBMC) samples were collected from adults at Day +100 following HSCT. NIH clinical diagnostic consensus criteria were used to define patients with cGVHD (n = 23) and no cGVHD (n = 36) for two-group analyses, and then further divided into the following categories for four-group analyses: Tolerant (no GVHD; n = 11), Isolated (acute GVHD only, n = 20), De Novo (cGVHD only; n = 6), or Progressive (both acute and cGVHD; n = 15). PBMCs were sorted by flow cytometry using T cell markers CD45RA, CD3, CD4, CD5, CD25, CD31, CCR7, and PD1, using a gating strategy by flowDensity. The two-group analyses also included an unstructured gating technique using flowType pipeline to phenotype all T cell populations without an a priori gating sequence. Those with a minimum mean two-fold change using RchyOptimyx were reported.

Results: In a population of naïve CD4 + CD8-CD45RA+ cells T cells, (excluding CD25 + CD127- Treg cells), PD1- was predictive of cGVHD in both a two-group comparison of cGVHD+ versus cGVHD- (29% vs. 10%; p = 0.02, Table 1), and a four-group comparison of GVHD subgroups (p = 0.004), with an AUC of ROC curve of 0.76. Other markers which were significant in the cGVHD population included CD31+ (p = 0.01), and CCR7+ (p = 0.004), both with AUCs of 0.71. However, the addition of CD31+, CCR7+, or both, did not increase the AUC in unstructured analysis beyond the effect of the PD1- population alone.

Conclusion: At Day +100 following HSCT, patients with cGVHD had significantly larger populations of naïve CD4 T cell with CD31+, CCR7+, and PD1-, compared to patients without cGVHD. This was seen with both the structure and unstructured gating strategies. PD1- appeared to be the strongest factor, without significant additive effect of CD31+ and CCR7+ reflected in the AUC. Taken together, this early naïve T cell phenotype CD4 + CD8-CD45RA + PD1- may be an important prognostic biomarker to predict the later development of cGVHD.



Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved.