Graft-versus-host disease in children: The AIEOP-BMT group experience with cyclosporin A

We retrospectively analyzed the data base of the Italian Association of Pediatric Hematology/Oncology BMT Group on the incidence and severity of GVHD in children given allogeneic BMT from HLA-identical sibling and receiving cyclosporin A (CsA) alone as GVHD prophylaxis. The study population included 145 patients for acute GVHD and 114 children at risk for chronic GVHD. Twelve patients had non-malignant diseases and 133 patients were affected by malignant disorders. Among the 145 patients (50 females, 95 males), 107 (74%) presented acute GVHD and 38 (26%) had no sign of disease. In the group of patients with acute GVHD, 38 children (26% of the whole study population) were found to have grade II disease, 9 (6% of the whole) grade III, 4 (3%) grade IV. Donor-recipient sex pairs had no significant influence on incidence of acute GVHD neither did donor-recipient age class stratification. Of the 114 patients evaluated for chronic GVHD, 86 (76%) developed no disease while 23 patients (20%) presented secondary chronic GVHD and 5 (4%) had de novo chronic GVHD. The incidence of chronic GVHD was higher in F-M than in M-M donor-recipient sex pairs (33% vs 11%, p < 0.05), with no difference between F-F and M-F. In patients of > 10 years, a higher incidence of chronic GVHD was observed in both female donors and recipients compared with male donors and recipients (48% vs 20% and 47% vs 19%, respectively, p < 0.05). Disease status at BMT was the most important factor determining outcome of patients with acute leukemia, disease-free survival (DFS) being 63%, 28% and 7%, respectively, in children transplanted at an early phase, at a more advanced stage and in relapse (p < 0.005). Patients with acute leukemia presenting moderate-severe acute GVHD had a better although not statistically significant DFS compared with both patients presenting grade I or no acute GVHD and children developing de novo or secondary chronic GVHD (62% vs 49% vs 44%, respectively). The anti-leukemic effect of acute GVHD was clearly evident only in patients at a less advanced stage of the disease. In fact, in the group of children with early leukemia, patients with grade II-IV acute GVHD showed a better disease free interval (DFI) than patients without or with grade I acute GVHD or chronic GVHD (100% vs 70% vs 64%, respectively, p < 0.05). We conclude that in pediatric patients CsA is effective in preventing and controlling severe acute and chronic GVHD without substantially eliminating the protective role played by donor immunocompetent cells against leukemia relapse.