Defective Triggering of NK Cells Results in Primary CLL Cells Resistance to Cytotoxicity,

Abstract 3876

Chronic lymphocytic leukemia (CLL) remains an incurable disease except after allogenic transplantation. Natural killer (NK) cells are one of the main effectors of immune surveillance involved in tumor control. Alterations of NK cells functions have been well characterized in myeloid malignancies. However the role of NK cells in immune escape of CLL in less known and controversial. Here we describe extensive phenotypic and functional characterization of NK cells and primary CLL cells and their interactions in vitro and in vivo.

Twenty eight untreated CLL patients, twenty four age-matched healthy donors and ten AML patients were enrolled in the study.

We have previously shown that expression and function of NK cell-triggering receptors is defective in AML. We then assessed the phenotypic and functional properties of NK cells from CLL patients. Unlike the results found in AML, no significant differences were observed in term of activating receptors, NKp46, DNAM-1, NKG2D, 2B4 and CD16. Only the natural cytotoxicity receptor (NCR) NKp30 was weakly decreased compared to healthy donors (p=0.0107). There wasn't any difference in the expression of inhibitory receptors CD158a, b, e, ILT2 and NKG2A. Looking at the spontaneous NK-mediated cytotoxicity, CLL NK cells displayed a cytolytic activity similar to that of healthy donors against K562 cell line. To further evaluate the functional consequences of the decreased expression of NKp30, mAb redirected killing assays was performed against P815 cell lines. The NK cells killing was slightly lower in CLL patients compared to healthy donors when anti-NKp30 was used although no difference could be observed with anti-NKp46 and anti-CD16. All these results supported that NK cells cytotoxicity should be effective in CLL.

We then studied the susceptibility of CLL B cells to allogenic NK killing both in vitro and in vivo. Unlike AML cells and K562 cells, CLL cells were resistant to NK cytotoxicity mediated by resting cells. Exogenous stimulation of allogenic NK cells with IL2 and IL15 restored partially CLL killing, which was nevertheless still lower than AML blasts and K562 cells killing (p=0.0288 and <0.0001 respectively). Murine xenotransplantation model using NOD/SCID g null (NSG) mice allowed us to study the anti-leukemic capacity of purified NK cells after activation with IL2. We didn't observe any clearance of CLL cells after allogenic NK cell injection while CLL and NK cells were checked to be present in blood, bone marrow, spleen and liver. These experiments confirmed the CLL resistance to NK-mediated killing.

To investigate the potential mechanisms of this resistance, we analyzed the surface expression of ligands for activating and inhibitory NK receptors on CLL cells. CLL cells displayed poor expression of ligands for activating NK receptors MICA/B, ULBP1-3, PVR, nectin-2 and CD54. Interestingly, this profile of surface expression was similar to that of normal B cells except a slight increase of ULBP3 expression on CLL cells. Regarding ligands for inhibitory NK receptors, HLA-class I molecules were significantly down-regulated while HLA-E tended to be up-regulated on CLL cells compared to normal B cells.

Finally, we tested ADCC in order to overcome the resistance of CLL cells to NK killing: the presence of rituximab increased significantly CLL lysis. Of note, priming of NK cells with IL2+IL15 still increased CLL cytotoxicity (p<0.0001).

Our findings demonstrate that primary CLL cells are resistant to NK mediated killing. This defect is mainly due to the lack of ligands for NK receptors on CLL cells surface leading to deficient triggering of NK cells. However NK cells of CLL patients are fully competent. Attempts to optimize NK cell therapy for treatment of CLL will require overcoming the low immunogenicity of B-CLL cells. Our xenograft model provides the tools for such preclinical development.