Prospective two center study of CD38 bright CD8+ effector memory T-cells as a predictor of acute GVHD

We conducted a prospective validation study at Cincinnati Children's Hospital Medical Center and Ohio State University Medical Center to test if absolute CD38 bright CD8+TEM cells > 30/µL would predict acute GVHD, similar to our pilot data. Blood was collected twice weekly following HSCT. If CD38 bright CD8+ TEM ≥ 30 cells/µL, Epstein-Barr virus and cytomegalovirus specificity was determined by tetramer staining, granzyme B content was assessed, Ki-67 staining performed to assess T-cell proliferation. Cells were incubated with alemtuzumab, daratumumab and cyclophosphamide in vitro to determine susceptibility to depletion. Of the 182 enrolled patients, 83 (45.6%) developed acute GVHD by day+100 but 171 patients were evaluable (acute GVHD n = 77 and no GVHD n = 94). There was no difference in the maximum absolute CD38 bright CD8+TEM cells prior to clinical symptoms and also after CMV and EBV tetramer positive patients were excluded from both cohorts. Ki-67 or Granzyme B expression in patients were comparable between patients with and without acute GVHD. Lastly CD38 bright CD8+ T-cells were effectively depleted with alemtuzumab and cyclophosphamide in vitro. Absolute peripheral blood CD38 bright CD8+TEM cells ≥30 do not predict acute GVHD in a large validation cohort of adult and pediatric HSCT recipients.