3D-AML drug screening in vivo. (A) Histologic assessment of ex vivo 3D scaffolds 3 weeks after implantation: representative image of hematoxylin and eosin (H&E, ×10 magnification, scale bar = 100 µm), human CD45 (×100 magnification; scale bar, 10 µm), murine CD31 (×20 magnification; scale bar, 50 µm), Alizarin red, von Kossa, and human-osteopontin (×100 magnification, scale bar = 10 µm) staining. (B) 3D scaffolds were seeded with stromal cells (at t = 0) and cultured in vitro for 24 hours before adding the luciferase-transduced SHI-1 AML cell line. After 24 hours, scaffolds were subcutaneously implanted in the back of previously irradiated NSG mice. Luciferase activity was measured before (at day 10) and after (at day 32) treatment. RLI, relative luminescence intensity; total flux = photons/s. Animals were treated daily at the following doses: venetoclax 100 mg/kg (orally), Ara-C 50 mg/kg (intraperitoneally), I-BET151 30 mg/kg (intraperitoneally), 5-azacytidine 5 mg/kg (intraperitoneally); n = 3 mice per group. (C) Tumor growth of luciferase-transduced primary AML cells in 3D scaffold in NSG mice, measured by luminescence activity during treatment with Ara-C 12.5 mg/kg (intraperitoneally) and lercanidipine 3 mg/kg (intraperitoneally), as single agents or in combination (dual targeting). Combination index (CI) = 0.1 at the end of treatment (day 44). CI = 0.5 after treatment withdrawal (day 79). CI <1.0 = synergistic effect. (i) Representative images of bioluminescence in differently treated mice before (at day 7) and after treatment (at day 44) and at treatment withdrawal (at day 79). n = 5 mice per group. (D) Representative images of H&E (×10 magnification; scale bar, 100 µm). Scale bar of images at higher magnitude, 2 µm), human osteopontin immunohistochemical analysis (×10 magnification; scale bar, 100 µm), and CD73 immunofluorescence (×60 magnification; scale bar, 40 µm) staining of scaffolds harvested from mice at the end of treatment. *P < .05, **P < .01, ****P < .0001.