Transplantation of haematopoietic stem cells derived from cord blood, bone marrow or peripheral blood: a single centre matched-pair analysis in a heterogeneous risk population.

BACKGROUND: Beside the transplantation of haematopoietic stem cells derived from bone marrow (BMT) and peripheral blood (PBSCT) in the absence of a well-matched donor, transplantation of cord blood (CBT) has been shown to be a valid alternative. To validate the efficacy of CBT in comparison to BMT and PBSCT we performed a single-centre based matched-pair analysis.

PATIENTS AND METHODS: Between 1996 and 2003, 15 paediatric patients with non-malignant and malignant diseases of heterogenous risk underwent CBT. 198 paediatric patients undergoing BMT or PBSCT during the same time and at the same centre were available for selection as appropriate controls for matched-pair analysis. Matching criteria in descending hierarchy were disease, risk status, type of donor, age at HSCT, gender and year of transplantation. 47 % of CB grafts were < or = 4/6 HLA-matched whereas close to 90 % of grafts in the BMT and PBSCT cohorts were completely matched.

RESULTS: Neutrophil engraftment was comparable in CB and BM recipients (p = 0.529) while engraftment following PBSCT occurs significantly earlier (p < 0.01). Median time to neutrophil recovery was 20 (range: 13-36), 19 (14-28) and 14 (9-24) days for the CBT, BMT and PBSCT cohort respectively. Of note contrary to the expectation, with regard to a reduced risk of Graft-versus-Host-Disease (GvHD) there was no clear advantage in the CBT cohort with a similar overall GvHD rate in all 3 groups. This observation can be attributed to the fact that in the CBT cohort the proportion of patients with an HLA-mismatched donor was higher than in the other cohorts. Rate of death of complications (DOC) was high in CB recipients (40 %), but not statistically different from BM (27 %) and PBSC recipients (13 %). In contrast to the CBT and BMT cohort with only 1 patient dead of disease (DOD), 4 PBSC recipients (31 %) died suffering from a relapse. 2-year event-free survival (EFS) in patients with malignant disease was 38.5 %, 69.2 % and 33.0 % for the CBT, BMT and PBSCT cohort respectively. 5-year overall survival (OS) was 53.3 % in the CBT, 66.4 % in the BMT and 50.9 % in the PBSCT cohort. There was no statistical difference between the cohorts transplanted with CB and BM or PBSC regarding EFS and OS (EFS: p = 0.24 and p = 0.72; OS: p = 0.53 and p = 0.64).

CONCLUSIONS: Transplantation of < or = 4/6 HLA-matched CB grafts seems to be associated with a higher risk of GvHD, graft rejection and lethal opportunistic infection. With an overall survival of 53 % in our 15 patients this analysis documents that even in high risk patients, CB may be a valid alternate HSC source in children who lack a well-matched donor. This is especially true, if a > 4/6 HLA-matched CB with > 2.0 x 10 (7) total nucleated cells/kg bodyweight is available. Thus, parallel to the search for a BM or PBSC donor, searching for an adequate CB unit should be initiated.