Donor lymphocyte infusions (DLI) are an well established treatment for relapsed chronic myeloid leukemia (CML) after allogeneic stem cell transplantation. In a previous analysis including mainly patients receiving DLI from HLA-identical siblings (Guglielmi et al., Blood 2002), it could be shown, that it was of major prognostic significance to start with a low cell dose (≤ 0.2 x 108 mononuclear cells/kg body weight).
In the present analysis we retrospectively analyzed patients receiving DLI from unrelated donors. In the data base of the Chronic Leukemia Working Party of the European Group for Blood and Marrow Transplantation 130 patients from 28 centers were identified, who received DLI from unrelated donors between 1987 and 2003 for treatment of relapse of Philadelphia chromosome positive CML and in whom data on the cell doses of DLI were available. According to the cell dose of the initial DLI patients were divided in 3 groups: 59 patients received DLI in a starting cell dose of ≤1x106 CD3 cells/kg body weight (group A), 39 patients received between 1 and 10x106 CD3 cells/kg (group B) and 32 patients received an initial cell dose of ≥10x106 CD3 cells/kg (group C). The three groups did not differ regarding age (median age 34 ys. in all groups), sex, sex mismatch and were comparable regarding the time interval from transplant to relapse (190d, 231d and 241d) and that from transplant to first DLI (453 d, 581d and 417d). In group C more patients were transplanted for CML beyond first chronic phase (41%), compared to 20% (group A) and 21% (group B), respectively. At first DLI relapse was molecular/cytogenetic in 60%, 64% and 38% of the patients, respectively. Relapse in accelerated phase/blast crisis at DLI was present in 3%, 13% and 22% of the patients. Multiple infusions were given in 80% of the patients in group A, 46% in group B and 6% in group C. In group A 61% of the patients received 2–3 DLI and 19% 4 or more DLI. The numbers were 33% and 13% in group B and in group C no patient received more than 2 DLI. The median total number of CD3 cells/kg given was 11 x 106 (0.5–311), 10 x 106 (3–860) and 75 x 106 (11–712) in the three groups.
A cytogenetic or molecular response was achieved in 68% of the patients in group A, in 65% in group B and in 63% in group C. Acute GvHD occurred in 27% of the patients in group A (I/IIo 17%, IIIo 9%), in 28% in group B (I/IIo 20%, IIIo 8%) and in 66% in group C (I/IIo 31%, IIIo 25%, IVo 10%). Myelosuppression occurred in 20%, 23% and 22% of the patients, respectively. Survival at 6 ys. was 68± 6% in group A, 67±12% in group B and 30±9% in group C. If patients treated for acc. phase or blast crisis were excluded, survival was 71±6% in group A, 72±13% in group B and 39±11% in group C.We conclude that treatment with DLI from unrelated donors with a starting dose of <10 x 106 CD3 cells/kg is efficient and results in a GvHD incidence of 27–28% with 8–9% GvHD>IIo. Starting with a lower cell dose (group A) did not compromise overall efficacy but did not reduce the incidence of GvHD or myelosuppression, probably due to the fact that total number of cells given was comparable to group B. An initial cell dose of > 10 x 106 CD3 cells/kg resulted in more GvHD and a poorer survival.