Figure 3. VLA-4, and not CD38, is essential for BM homing.

Human CLL cells from different patients were injected into NOD/SCID mice. After 180 min mice were sacrificed and the number of human cells detected by flow cytometry. (A) BM and spleen homing rates of CLL cells from patients separated into four groups according to their VLA-4 and CD38 risk status (VLA-4-/CD38-, n = 9; VLA-4+/CD38+, n = 10; VLA-4-/CD38+, n = 7; VLA-4+/CD38-, n = 10; Anova, Kruskal-Wallis test, BM: p<.0001, Spleen: p = 0.1637, and Dunn's Multiple Comparison test). (B) Correlation of percent VLA-4+ CLL cells of patients used in homing assays to corresponding BM homing rates (n = 36, Linear regression, R² = .5804, p<.0001) and correlation of percent CD38+ CLL cells of patients used in homing assays to corresponding BM homing rates (n = 36, Linear regression, R² = .0289, p = .3208). Homing rates were normalized as described[12]: number of human cells analyzed per 10⁶ mouse cells (total cells) per 10⁶ injected viable human target cells. *, P<.05; **, P<.01; ***, P<.001.