High number of memory t cells is associated with higher risk of acute graft-versus-host disease after allogeneic stem cell transplantation.

The pathophysiology of acute graft-versus-host disease (GVHD) remains poorly understood in humans. Although T cell subsets have been identified to play a major role in disease initiation in rodents, clinical data on the effect of these different subsets are scarce and conflicting. To address this question, immunophenotyping analyses were performed on the graft in 210 patients. The onset of acute GVHD was retrospectively correlated with these subpopulations. In an adjusted analysis, only the absolute count of CD45lo/CD62Llo CD8(+) T cells (effector memory T cells) was significantly associated with the onset of grade 2 to 4 acute GVHD. Thus, in contrast to experimental data, we found that the number of effector memory but not of naïve T cells was associated with the onset of GVHD. These results should be kept in mind while clinical trials, which aim to deplete naïve T cells, are underway in several institutions.

Long-term immune reconstitution and infection burden after mismatched hematopoietic stem cell transplantation.

Mismatched unrelated donor (MMUD) or umbilical cord blood (UCB) can be chosen as alternative donors for allogeneic stem cell transplantation but might be associated with long-lasting immune deficiency. Sixty-six patients who underwent a first transplantation from either UCB (n = 30) or 9/10 MMUD (n = 36) and who survived beyond 3 months were evaluated. Immune reconstitution was prospectively assessed at sequential time points after transplantation. NK, B, CD4(+), and CD8(+) T cells and their naïve and memory subsets, as well as regulatory T cells (Treg), were studied. Detailed analyses on infections occurring after 3 months were also assessed. The 18-month cumulative incidences of infection-related death were 8% and 3%, and of infections were 72% and 57% after MMUD and UCB transplantation, respectively. Rates of infection per 12 patient-month were roughly 2 overall (1 for bacterial, .9 for viral, and .3 for fungal infections). Memory, naïve CD4(+) and CD8(+)T cells, naïve B cells, and Treg cells reconstitution between the 2 sources were roughly similar. Absolute CD4(+)T cells hardly reached 500 per µL by 1 year after transplantation and most B cells were of naïve phenotype. Correlations between immune reconstitution and infection were then performed by multivariate analyses. Low CD4(+) and high CD8(+)T cells absolute counts at 3 months were linked to increased risks of overall and viral (but not bacterial) infections. When assessing for the naïve/memory phenotypes at 3 months among the CD4(+) T cell compartment, higher percentages of memory subsets were protective against late infections. Central memory CD4(+)T cells protected against overall and bacterial infections; late effector memory CD4(+)T cells protected against overall, bacterial, and viral infections. To the contrary, high percentage of effector- and late effector-memory subsets at 3 months among the CD8(+) T cell compartment predicted higher risks for viral infections. Patients who underwent transplantation from alternative donors represent a population with very high risk of infection. Detailed phenotypic analysis of immune reconstitution may help to evaluate infection risk and to adjust infection prophylaxis.