Abstract

Abstract 4192

Donor T cells are known to be the most powerful effectors for GVHD. Recently, the dose of invariant NKT and B cells in the allograft have been associated with incidence of aGVHD in sibling transplants (Chaidos et al, Blood 2012; Michonneau et al, Brit.J.Hem 2009). Murine models suggest a role for donor B cells and CD4+ T cells in development of cGVHD (Young et al, J.Immunol 2012). The influence of NK cells, NKT cells, B cells in allografts from matched unrelated donor (MUD) transplants on incidence of cGVHD is unknown. Here we present results on influence of innate and adaptive immune subsets in the allograft and the incidence of aGVHD and cGVHD in 108 patients (pts). We analyzed the absolute numbers of T, NK, NKT and B cells along with extensive immunophenotypic characterization of their activation status in consecutive allografts obtained from both sibling and MUD between 2010– 2011. All allografts were peripheral blood stem cells mobilized after 5 days of G-CSF administration. Of the 108 pts, 67 pts received grafts from MUD; 71 pts had received a reduced-intensity conditioning regimen. All pts receiving allografts from MUD received anti-thymocyte globulin. Median age of pts at transplant was 55 years (21–73 years). Median CD34+ dose was 6.51 x106/kg and median CD3+ dose was 2.33 x108/kg. The incidence of aGVHD and cGVHD was 58% and 38% respectively. We compared in an unsupervised approach whether absolute numbers of resting and activated NK, NKT, B, CD4+ and CD8+ T cells were different in pts with vs. without aGVHD/cGVHD. Univariate logistic regression models and Wilcoxon rank sum tests were used to assess differential expression of absolute and percent immune subset markers between those who did vs. did not develop GVHD, analyzing separately for aGVHD and cGVHD. In addition, some markers had a number of absolute values of zero, where the parameter of interest was whether or not a donor had a non-zero value and its association with GVHD in pts.

Results:

Median numbers of activated NK cells (CD3CD56+CD16+CD69+) cells were significantly higher in pts with vs. without aGVHD (586.4/μL vs. 270.5/μl, p=0.003). Higher median numbers of resting NK cells (CD3CD56+CD16+) and activated B cells (CD19+CD80+) were associated with cGVHD but not aGVHD. Non-zero levels of CD4+ T cells expressing prostaglandin D2 (PGD2) receptor (CD4+CD294+), chemokine receptor CCR3 (CD4+CD193+), chemokine receptor 4 CCR4 (CD4+CD194+) and CXCR3 (CD4+CD183+) in the allograft were associated with lower incidence of cGVHD. Also CD4+ T cells expressing both CCR4 and PGD2 receptors were associated with lower incidence of cGVHD. Significant differences were seen in median numbers (Table 1) and % of non-zero values (Table 2) of NK cells, B cells, and activated CD4+ T cells between those who did vs. did not develop cGVHD. Immunophenotypic analyses of pts receiving these allografts at days 30 and 100 post transplant were also performed. Updated results about correlation of these subsets between allografts and post-transplant patient samples will be presented. Multivariate analyses of these subsets in conjunction with clinical factors are being performed. In conclusion, activated B cells in the allograft may influence the incidence of cGVHD; in particular B cells expressing CD80, a co-stimulatory molecule for T cells and their association with cGVHD confirms results from murine models. Interestingly, CD4+ T cells associated with lower incidence of cGVHD expressed receptors for chemokines, CXCR3 and prostaglandin D2. These homing receptors are up regulated selectively in CD4+ T cells in the absence of chronic GVHD, suggesting a protective role for these receptors in the pathogenesis of cGVHD. Activated and resting NK cells were also associated with acute and chronic GVHD respectively. These results showing an association of specific immune subsets in the allograft and GVHD may provide opportunities for therapeutic interventions for graft engineering.

Table 1.

Median absolute numbers of NK and B cells by cGVHD status

SubsetcGVHDNo cGVHDp-value
CD3/CD56+/CD16+/CD117 6361 4892 0.036 
CD19+/CD80+ 109.6 71.2 0.012 
SubsetcGVHDNo cGVHDp-value
CD3/CD56+/CD16+/CD117 6361 4892 0.036 
CD19+/CD80+ 109.6 71.2 0.012 
Table 2.

Percent with non-zero absolute numbers of activated CD4+ T cells by cGVHD status

Subset% non-zero values in pts without cGVHD% non-zero values in pts with cGVHDp-value
CD4+/CD193+ 27.9% 5.1% 0.004 
CD4+/CD294+ 41.0% 10.3% 0.0013 
CD4+/CD183+ 41.0% 10.3% 0.0007 
CD4+/CD194+ 42.6% 10.3% 0.0006 
CD4+/CD294+/CD194+ 36.1% 7.7% 0.0017 
Subset% non-zero values in pts without cGVHD% non-zero values in pts with cGVHDp-value
CD4+/CD193+ 27.9% 5.1% 0.004 
CD4+/CD294+ 41.0% 10.3% 0.0013 
CD4+/CD183+ 41.0% 10.3% 0.0007 
CD4+/CD194+ 42.6% 10.3% 0.0006 
CD4+/CD294+/CD194+ 36.1% 7.7% 0.0017 
Disclosures:

Off Label Use: Lenalidomide in AML. Jaglowski:Pharmacyclics: Research Funding. Benson:Innate Pharma: Research Funding. Devine:Sanofi: Honoraria, Research Funding.

Author notes

*

Asterisk with author names denotes non-ASH members.