Thirty-five children (19 males, 16 females; median (range) age 7.3 years, (0.2–16.2)) with various haematological and metabolic diseases underwent CD34+ selected peripheral blood stem cell transplantation (PBSCT) from mismatched family donors. Thirteen maternal, 19 paternal and 3 sibling donors who were mismatched for 2–3 HLA antigens. Twenty children underwent transplantation for refractory (n=2) or relapsed (14 patients in CR2 and 4 in CR3) acute leukaemia (ALL =15, AML =5). Other patients included Fanconi’s anaemia (n=3), Primary Haemophagocytic lympho-histiocytosis (n=3), Myelodysplastic syndrome (n=3), Juvenile myelomonocytic leukaemia (n=2), Congenital immunodeficiency (n=2) and Metabolic diseases (n=2). Fludarabine, cyclophosphamide and anti-thymocyte globulin (equine) were used for conditioning in all patients. In addition 21 patients received TBI (14.4 Gy) and 12 patients received oral busulphan (14 -16mg/kg). The mean ± s.d., CD34+ and CD3+ dose in the graft was 12.4 ± 7.4 X 106/kg (range 2.4 – 34.2) and 7.5± 10.4 X104/kg (range 0.1 – 54) respectively. All patients received prophylactic aciclovir, itraconazole, cotrimoxazole and intravenous immunoglobulin infusion (400mg/kg every 1–2 weeks). Ganciclovir 10mg/kg/day was used as pre-emptive prophylaxis for CMV reactivation. All patients were advised to follow a policy of semi isolation and safe diet. These measures were continued until the patient achieved a post-transplant CD3+ lymphocyte count of 0.3X109/l.

All patients achieved engraftment. Median time for neutrophil and platelet engraftment was 12 days (range 10–17 days) and 20 days (range 11–24 days) respectively. One patient with Fanconi’s anaemia had secondary graft failure but engrafted after second conditioning with OKT3. Sixteen patients received additional CD34+ stem cell infusions or donor lymphocyte infusions (DLI) for, progressive mixed chimerism (n=7), poor immune recovery (n=5), haematologic relapse (n=3) or persistent minimal residual disease (1). DLI either stabilised mixed chimaerism or reversed it to complete donor chimaerism in all seven patients. DLI was not beneficial in controlling haematologic relapse or improving lymphocyte recovery after transplantation though it was effective in eliminating persistent minimal residual disease in one patient.

Four (11.4%) patients developed acute GVHD ≥ grade 2 requiring treatment and fifteen patients (43%) developed opportunistic infections resulting in 4 (11.4%) deaths. Other causes of mortality included disease progression/relapse in 9 (25.78%), TBI toxicity in 1 (2.8%) and veno-occlusive disease in 1 (2.8%). Twenty (57%) children are alive and well after a median follow up of 54 months (range 16–70 months). Results from our single centre study of patients undergoing haploidentical PBSCT compare favourably with those of matched unrelated donor transplantation mainly due to low infection related mortality. Use of prophylactic antimicrobials, intravenous immunoglobulin infusions, a relatively high T cell dose and a policy of semi-isolation until lymphocyte subset recovery may be responsible for the low infection related mortality in our patient group.

Author notes

Corresponding author