Impact of CD3+ and CD34+ Doses in the Mobilized Peripheral Blood Stem Cell Grafts on the Outcome of Reduced-Intensity Conditioning Allogenic Transplantation from Matched Unrelated Donors for Acute Myeloid Leukemia – an Analysis from the Acute Leukemia Working Party of the EBMT

Introduction:

The impact of peripheral blood stem cell (PBSC) grafts composition on the outcome of reduced-intensity conditioning (RIC) allogenic transplantations (allo-SCT) is still controversial. Inconsistent results have been reported regarding the influence of CD3+ and CD34+ cells dose on incidence of GVHD, disease control and survival. These discrepancies may be at least in part explained by the differences in disease categories, disease status at transplant, donor type and conditioning.

Aim and methods:

This retrospective EBMT registry study aimed to analyze the impact of CD3+ and CD34+ cells dose in PBSC grafts on the outcome of RIC allo-SCT in patients with acute myeloid leukemia (AML) allografted from matched unrelated donors (MUD). Two hundred and three adults with AML in first complete remission (CR1) treated with RIC allo-PBSCT from MUD (10 of 10 match) between 2000 and 2012 (median 2011) were included. Ex-vivo T-cell depletion was an exclusion criterion. Median age was 58 (range, 21-73) years; gender: male-116, female-87; 142 had intermediate, 42 unfavorable, 4 favorable, and 15 unknown cytogenetic features; median time to achieve CR1 was 51 (range, 21-350) days, whereas time from CR1 to allo-SCT was 115 days (15-351). Median donor age was 34 (range, 19-61) years. The preparative regimen was based on chemotherapy in 143 (70%) of the RIC allo-CST (Flu-Bu:110; Flu-Mel:12; Flu-Treosulfan:14, other:7), whereas in 60 (30%) low dose TBI was applied (2 Gy:57; 4 Gy:3). In-vivo T-cell depletion (ATG or Campath) was used in 166 (82%) of the transplants. Median transplanted CD3+ and CD34+ doses were 250 (50-885) x 10^6 and 6.53 (1.34-413) per kg of recipient b.w., respectively. Follow-up was 22 months (3-105).

Results:

The engraftment rate equaled 99% (202 pts) and was not affected by CD3+ nor CD34+ counts. In univariate analysis, patients transplanted with the highest CD3+ doses (above the third quartile cut-off point value, >347 x 10^6/kg) had an increased incidence of acute (a) GVHD grade >2 (45% vs. 26%, p=0.007) and grade 3-4 (20% vs. 6%, p=0.003), respectively. In the quartile of pts transplanted with the highest CD34+ graft content (>8.25 x 10^6 /kg) increased incidence of grade 3-4 aGVHD (18% vs. 7%, p=0.02) was observed. Other risk factors for aGVHD were transplantation from CMV seropositive donors (grade >2; 44% vs. 24%, p=0.005) and from females to males (grade 3-4; 19% vs. 8%, p=0.04). There was no association between cellular composition of grafts and non-relapse mortality (NRM), AML relapse, incidence of chronic (c) GVHD and survival. There was also no significant correlation between CD3+ and CD34+ cells infused. In multivariate analysis, CD3+ dose was the only adverse predicting factor for aGVHD grade >2 (HR= 2.1; 95%CI: 1.25-3.55, p=0.005) and together with CD34+ dose for aGVHD grade 3-4 (CD3+, HR=3.6; 95%CI: 1.45-9.96, p=0.006; CD34+, HR=2.65; 95%CI: 1.07-6.57, p=0.04). Other factors that independently affected unfavorable outcome were: time from diagnosis to CR1 above median for NRM, transplantation without in-vivo T-cell depletion for cGVHD, time from CR1 to allo-SCT above median and unfavorable karyotype for leukemia-free and overall survival and unfavorable karyotype for relapse incidence.

Conclusions:

These results suggest that the incidence of severe acute GVHD post RIC allo-SCT, still a major cause of morbidity and mortality, is associated with the composition of the PBSC grafts, specifically higher numbers of infused CD3+ and CD34+ cells. As graft composition can be manipulated, careful assessing the CD3+ and CD34+ graft content and the cell dose infused may help in reducing severe aGVHD risk and improving transplantation outcome.