PSCT is becoming increasingly more common, as large doses of PSCs may result in faster engraftment than bone marrow, and high PSC doses may be advantageous in decreasing graft rejection. PSCs contain several logs more T cells than bone marrow, and most studies have shown an increased risk of extensive chronic GVHD with unmodified PSCs. However, purified CD34+ cells may be associated with increased graft failure (

Bornhauser, et al.
Br J Haematol
) or relapse. To decrease the risks of both chronic GVHD and graft failure, and maintain potential graft vs. leukemia effect, we used CD34+ selection with a defined dose of CD3+ cells in the product at the time of infusion for URD or PMRD PSCT. We report on the first 23 patients (pts) on this IRB approved study. Conditioning was thiotepa 10mg/kg, cyclophosphamide 120 mg/kg and TBI 12 Gy/6 fractions. Lung shielding after 8 Gy and methylprednisolone 1 mg/kg at day+8 were initiated when it became apparent that engraftment syndrome was being observed in the first 10 patients. GVHD prophylaxis was cyclosporine, followed by oral tacrolimus, which was weaned by 120 days in the absence of GVHD. CD34+ selection was accomplished with the Isolex 300i device and resulted in an average T cell depletion of 4.2 logs. The positive fraction had an average purity of 94%, and T cells obtained from the negative fraction were added to the positively selected product at the time of infusion to achieve the defined T-cell dose of 5 x 105/kg CD3+. The remaining portion of the negative fraction was cryopreserved in multiple aliquots for DLI, if needed. There were 12 males; median age was10 yrs (range, 3–22). Diagnoses included: ALL: 6, AML/MDS:12, CML: 3, JMML: 2. URDs were used for 22 patients, and related donor for 1. This was the second transplant for one pt with MDS. Donors were matched by high resolution typing for13 pts; mismatches included A for 5 pts, B and C loci for 3 pts, DQB1 for 2 pts. The median CD34+ dose infused was 6.6 x 106/kg (1.3–12.4). Engraftment occurred in all pts, with ANC>500 at a median of 14 days (10–19), and platelets >20K in 15 pts at a median of 22 days (13–33). Eleven patients (48%) are alive in remission 7–25 months post SCT (med, 18). Grades I-II GVHD occurred in 15 pts, grade III in 1, and grade IV in association with HHV6 in 1 pt. Limited chronic GVHD occurred in 8 of 12 evaluable pts (67%), and has resolved in all but one. Two pts with AML relapsed at 3 and 18 months. Ten pts died at d 32–85 from CMV (2), ARDS (5), HHV6/infection (2), bacterial sepsis(1). Both pts who developed CMV pneumonitis were seropositive, with seronegative donors; one developed pneumonitis at day +20, with negative antigenemia. Pts with CML were negative for bcr-abl by PCR <6 months post SCT and have remained negative without GVHD. There was a trend toward higher mortality with mismatched donors (RR 2.6, p=0.06). This study demonstrates that PSC engineering with CD34+ positive selection with a defined dose of CD3+ cells results in prompt engraftment. With this approach, the risk of extensive chronic GVHD with URD or PMRD PSCs may be reduced.

Author notes

Corresponding author