High Numbers Of Memory T Cells Are Associated With Higher Risk Of Grade II-IV Acute Gvhd After Human Allogeneic Transplantation

Acute graft versus host disease is a frequent and life threatening complication following HSCT. Some predictive factors have been identified in the last decades. Experimental studies in mice suggest that the naïve cytotoxic T cells (CD3+/CD8+/CD45RA+/CD62L+) are the major mediators of acute GVHD and that removing this subset of the donor T cells, called ‘naïve T cells’, before transplant may reduce the frequency and intensity of GVHD. Detailed immunophenotyping of the graft including naïve and memory (CD3+/CD8+/CD45RA-/CD62L- ; CD3+/CD8+/CD45RA+/CD62L- ; CD3+/CD8+/CD45RA-/CD62L+ ) T-cell contents have never been explored in human GVHD. We studied the correlation between memory and naive T cell in bone marrow and peripheral blood grafts and development of acute GVHD after hematopoietic stem cell transplant.

We analyzed by detailed immunophenotyping, the grafts of a cohort of 210 patients among 402 patients who received an allogeneic stem cell transplantation from bone marrow and peripheral blood between January 2009 and June 2012 at a single center. There were no differences between the 210 studied patients and the other 192 in whom grafts were not studied. Characteristics of patients investigated for naïve and memory T cells were compared using Wilcoxon rank-sum tests and Fisher’s exact tests. The main outcome was occurrence of acute GVHD grade II – IV. Cumulating incidence of acute GVHD was estimated using usual methods and compared according to tertiles of T cytotoxic lymphocytes subpopulations using Gray’s test. Adjusted analyses were performed using Fine- Gray proportional hazards models. All tests were two - sided and p-values ≤ 0.05 were considered as indicating significant association. T cytotoxic lymphocytes were typed in all 210 grafts using CD3, CD8, CD45 RA and CD62L four colors immuphenotyping. Clinical and histological characteristics of patients were recorded. Including age, gender, ABO group and rhesus, viral serology of both the donor and the patient, characteristic of the grafts including HLA compatibility, bone marrow or peripheral blood, lymphocytes and nucleated cell and CD34 numeration, conditioning regimens, GVHD prophylaxis, characteristics of GVHD (date of onset, organs involved, stage and grade).

Median follow up from transplant was 18 months. Cumulative incidence of acute GVHD was 59% (95% CI range 45 to 59) overall, and 49% (95% CI 42 to 56) at 100 days. In univariate analysis increased absolute counts of memory T cell subtypes were significantly correlated with the onset of an acute GVHD grade II – IV. Risk factors for acute GVHD (multivariate analysis) were use of an unrelated donor, positive CMV donor for a negative recipient, and use of TBI 12Gy. In a multivariate analysis the subtype CD3+/CD8+/CD45 RA-/CD62L- was associated with the onset of acute GVHD grade II-IV (adjusted Hazard Ratio = 1.26 and 1.98, p=0.02). Adjusting analysis on the total number of total nucleated cells infused did not affect the results. Restricting analyses to patients receiving peripheral blood stem cells also provided same conclusions.

This first study on the relation between rate of memory T cell and GVHD revealed that CD3+/CD8+/CD45 RA-/CD62 L- T-cells numbers and percentage were associated with acute GVHD grade II – IV. In contrast to murine models we did not find evidence for a link between naïve T-cells and GVHD risk

Robin:novartis: Research Funding.