Treatment of Familial Hemophagocytic Lymphohistiocytosis with Bone Marrow Transplantation : A Single Center Experience of 48 Patients.

Familial Hemophagocytic Lymphohistiocytosis (FHLH) is a rare genetically heterogeneous autosomal recessive disease, curable only by bone marrow transplantation (BMT).We retrospectively analysed 48 children undergoing BMT between May 1982 and May 2004. Fourteen patients received BMT from matched sibling donor (MSD), 29 from family haploidentical donor and 5 from unrelated donor. The 48 patients had a total of 60 transplants. Genetic studies of perforin and Munc13, the two known genes, have been performed in 29 children: 7 had a perforin defect, 10 a Munc13 defect and 12 were not linked to perforin or Munc13. Median age at diagnosis was 3 months. Central nervous system involvement (CNS) was present at diagnosis in 30 patients (62%) and during the course prior to BMT in 38 patients (80%). Initial treatment before BMT was a chemotherapy regimen consisting of VP16, steroids, cyclosporin A and intrathecal methotrexate (IT) for 15 patients (31%). Thirtythree patients (69%) received an alternative treatment, based on the primary role of T cell activation in FHLH, including steroids, cyclosporin A, IT, and for 26 patients rabbit anti-thymocytes globuline (ATG). Twenty-seven children (56%) had a complete resolution before BMT, 14 (30%) a partial remission, 5 (10%) a unstable active disease and 2 a CNS involvement.Conditioning regimen consisted of Busulfan, Cyclophosphamide and ATG for 32 transplants and Busulfan, Cyclophosphamide and VP16 for 18 transplants. Marrow was T cell depleted for all transplants except MSD. Median age at BMT was 6 months and median time from diagnosis to BMT was 3,5 months. Donor cell engraftment was achieved in 37 children (77%). Seven patients (18%) had a acute GVHD of grade II-IV. The major regimen related toxicities were VOD in 14 children (29%) and severe infections (n=30). In the 19 deceased patients, death occured prior to day+100 in 15, caused by BMT complication (n=7) or reactivation of HLH (n=8). Four children died after day+100, 2 by GVHD and 2 by HLH. The median follow-up was 6 years. Twenty-nine children (60%) are alive at the time of evaluation, without recurrence of HLH. One child is alive with slight psychomotric retardation and 1 with a moderate spasticity. Survival with regard to donor was 100% in MUD, 57% in MSD, and 55% in haplo-identical donor. Among alive patients, 17 had a donor chimerism and 12 a mixed chimerism. Five patients had late disease relapse after 8 to 46 months of mixed chimerism. All these 5 children had mismatch BMT and relapse occured with less than 10% of donor cells, supporting the hypothesis that a certain amount of donor cells are needed to control T lymphocytes activation in HLH. Severe neurological involvement is associated with a lower survival (36%). Survival for patients with perforin, Munc13, no perforin or Munc defect were respectively 72%, 80% and 67%. With a pretreatment by immunotherapy 68% of patients survived, compared to 47% with chemotherapy. Survival for patients younger than 3 months was 61%, 66% for them between 3 months and 2 years, and 50% for older ones. Survival if BMT was performed early or late, using the median time to BMT as the cut-off time were 64% and 55%. We report the largest series of allogenic BMT in HLH from one center. Immunotherapy as pretreatment allows fast CR and short delay between diagnosis and BMT. Results of this non myeloablative approach are very encouraging with a great improvement of the survival.