Sequential Acquisition of Inhibitory NK Cell Receptors After Allogeneic Hematopoietic Stem Cell Transplantation: The Role of Steroids and Other Transplantation-Related Factors.

Abstract 3543

Poster Board III-480

Donor-recipient incompatibility for ligands of killer immunoglobulin-like receptors (KIRs) was shown to result in NK cell alloractivity and graft-vs.-leukemia reaction after allogeneic hematopoietic stem cell transplantation (alloHSCT). However, clinical data are inconsistent suggesting the impact of various procedure-related factors on NK cell maturation and their possibility to display alloreactivity. In particular, only NK cells expressing KIRs but not CD94:NKG2A receptor are able to kill leukemic targets. Our goal was to prospectively evaluate reconstitution of NK cell inhibitory receptors after alloHSCT and to identify factors affecting this process.

Eighty-three adults with hematological malignancies treated with myeloablative T-replete alloHSCT from either HLA-matched sibling (n=32) or unrelated donor (n=51), were included. The expression of inhibitory receptors (KIR2DL1, KIR2DL2/DL3, KIR3DL1, NKG2A) on NK cells was analyzed by flow cytometry in donors as well as on days +28,+56,+100,+180,+365 after alloHSCT. In parallel, reconstitution of lymphocyte subpopulations (CD3+, CD3+CD4+, CD3+CD8+, CD19+) in peripheral blood was studied.

The median frequencies of KIR2DL1+ NK cells on days +28,+56,+100,+180,+365 equaled 4.7%, 5.6%, 5.9%, 9.1%, and 12.6%, respectively, and were significantly lower compared to 20.1% in donors. On day +28, frequency of KIR2DL1+ NK cells was higher for patients receiving peripheral blood compared to bone marrow transplants (p=0.02). Expression of KIR2DL1 correlated positively with the CD34+ cell dose on days +28, +56, +100 and +180 (p<0.05). Expression of KIR2DL2DL3 in subsequent study time-points was 14.4%, 20%, 19.6%, 20.9%, and 25.4%, compared to 24.5% in donors (p<0,05 up to day +100). On day +28 the frequency of KIR2DL2/DL3+ NK cells correlated positively with the absolute number of circulating CD3+ cells (p=0.007) and CD3+CD8+ cells (p=0.003). The frequencies of KIR3DL1 in subsequent time-points were 16.8%,13.8%,10.9%,10.1%, and 9.2% vs. 18.1% in donors. The expression of NKG2A on day +28 equaled 89.9% and decreased to 79%, 68.9%,64% and 51.9% on days +56, +100, +180, and +365, respectively (up to day +180, p<0.05 compared to 46.8% in donors). The frequency of NKG2A+ NK cells was lower for patients receiving PB compared to BM transplants on days +56, +100, and +180 (p<0.05). On day +365, median expression of NKG2A by NK cells was 60.5% for patients who were treated with steroids at any time after alloHSCT compared to 48.1% for the remaining subjects (p=0.04). On day +28, the frequency of NKG2A+ NK cells correlated negatively with the absolute number of circulating CD3+ cells (p=0.004) and CD3+CD8+ cells (p=0.04).

We conclude that: 1) the acquisition of KIRs by NK cells after alloHSCT is sequential and starts with KIR3DL1, followed by KIR2DL2/DL3 and KIR2DL1; 2) NKG2A is overexpressed within 6 months after transplantation and the expression remains elevated in case of the use of steroids; 4) T-cell recovery occurs in parallel to NK cell maturation. Our observations should be taken into account when trying to predict potential benefit from NK cell alloreactivity.