Abstract 4154

T and B cell depletion of haploidentical peripheral stem cells with CD3/CD19 coated magnetic microbeads prevents GvHD and allows to coinfuse large numbers of donor NK cells and other accessory cells. The anti CD3 specific OKT3 antibody was routinely used as rejection prophylaxis without affecting CD3 negative cells. However, due to its restricted availability, the substance had to be substituted by polyclonal ATG preparations with a longer half life period and comprising a broader variety of antigen, with the CD56 antigen in particular. We present data with reduced ATG doses given at the beginning of the conditioning regimen in order not to impair cotransfused NK cells and immune recovery. A total of 27 pediatric patients (ALL/AML n=9, relapsed solid tumors n=13, nonmalignant diseases n=5) received either 3×5 mg/kg (n=7) or 3×10 mg/kg (n=20) ATG-F (Fresenius) starting at day -12, followed by fludarabine (160 mg/m2, d -8 to -5) thiotepa (10 mg/m2 d -4) and melphalan (140 mg/m2 d −3 to −2). A median number of 14.4×106 CD34/kg and 62×106 NK cells/kg with 37×103 T cells/kg were infused. Median time to ANC>500 was 9 days in both groups. Graft rejection occurred in 3/7 patients with 15 mg ATG (42%) and in 2/20 patients with 30 mg ATG (10%). After reconditioning with a TLI based regimen, final engraftment was achieved in all patients. Acute GvHD grade II-IV was observed in 1/4 (15 mg) and 1/18 (6%, 30 mg) patients without rejection. Extensive chronic GvHD occurred in 1/4 (15 mg group) and 1/18 (30 mg group) patients. Immune recovery was monitored in the 30 mg group and compared with a historical group receiving OKT3 and the same chemotherapy (n=34). Recovery of CD56+ NK cells was fast with a mean number of 473 vs. 230 cells/μl at day +14, 299 vs. 281 cells/μl at day +30 and 245 vs. 236 cells/μl at day +90. CD3+ T cells reached 12 vs. 16/μl at day +30 and 138 vs. 217/μl at day +90 (30 mg group vs. OKT3 group; no significant differences for all data pairs). ATG serum levels were measured in 9 patients by flow cytometry (amount of ATG binding to the Jurkat cell line, defined as T cell specific rabbit IgG). Median peak levels of 10.5 μg/ml (15 mg group) and 15.0 μg/ml (30 mg group) specific rabbit IgG were reached between day -8 and -6 and dropped to 1.2 and 2.6 μg/ml at day 0. In vitro incubation of NK cells from healthy donors with a comparable dosage of ATG-F (1 or 10 μg/ml) resulted in 26 (41)% apoptosis and 0.2 (0.2)% necrosis (70 (52)% vital cells) after 24 hours.


Our aim was to substitute OKT3 by ATG in patients who receive CD3/19 depleted haploidentical peripheral stem cells without hampering donor NK cells infused on day 0 and subsequent immune recovery. Administration of 15 or 30 mg/kg ATG-F was started at an early time point (day -12) of the regimen. Both doses resulted in low serum levels of specific ATG at day 0 and in a fast NK cell recovery. In vitro results suggested, that the majority of NK cells will not be damaged herewith. However, 15 mg/kg seemed to be not effective in preventing graft rejection and use of 30 mg/kg ATG-F has to be recommended. Immune recovery of T and NK cells was comparable to that of a historical control group who received OKT3. This approach will be also of interest for other transplantation strategies in which various components of the grafts and additionally given Tregs or specific T cells have to be preserved.


Martinius:Fresenius Biotech: Employment.

Author notes


Asterisk with author names denotes non-ASH members.