Production and efficacy of CLL-XNK cells. (A) NK cells were purified and expanded as described above. Portions of each culture were frozen at various points during expansion, and the graph shows projected growth based on the cells left in expansion culture. Doubling times were calculated based on an exponential fit of growth curves. Both graphs show means ± standard deviation of 9 ND-XNK, 5 CLL-XNK, and 4 Ibru-CLL-XNK samples. (B) NNKs (n = 6 experiments), ND-XNKs (n = 13), CLL-XNKs (n = 15), and Ibru-CLL-XNKs (n = 8) were tested for cytotoxicity against primary allogeneic CLL cells using 4-hour calcein release assay, as described above. Graph depicts mean ± standard deviation for each NK cell type. (C) Patient-derived XNK cells (n = 4 CLL-XNK, 2 Ibru-CLL-XNK) were cocultured with autologous (n = 6) or allogeneic (n = 18) CLL cells. Cytotoxicity was measured by calcein release and depicted as mean ± standard deviation. (D) ND-XNKs (n = 5) and CLL-XNKs (n = 8) were characterized by flow cytometry. Median fluorescent intensity (MFI) was calculated for total live CD3⁻CD19⁻CD56⁺ cells. ∗P < .05 by ANOVA (A), mixed effect modeling (B,C), or Student t test (D).