VPA/ATRA treatment efficacy in primary leukemia cells in vitro and in vivo. (A) VPA/ATRA-induced apoptosis in BCR/ABL positive leukemia is caspase-dependent. 32D-BCR/ABL cells (top row) and a Ph⁺-ALL patient sample no. 9 (bottom row) were treated with VPA/ATRA (A+V), VPA/ATRA plus the pan-caspase inhibitor z-VAD-fmk, (A+V+ z-VAD-fmk), or mock-treated (control). Forty-eight hours after start of exposure to these compounds, apoptosis was assessed using FITC-Annexin/propidium iodide staining and fluoresecence activated cell-scanning (FACS) analysis. The total percentage of apoptotic cells is shown in the right quadrants, respectively. A representative experiment of at least 3 independent experiments for the 32D-BCR/ABL cell line and the patient sample is shown. Box-and-whisker plots for the comparison of VPA/ATRA (A+V)–mediated apoptosis in Ph⁺ ALL versus AML blasts in vitro (B) and the effect of VPA/ATRA (A+V), rapamycin (Rapa), or both (A+V+rapa) on apoptosis induction in primary AML (C). Lines in boxes indicate the median; boxes display data points located in the middle 2 quartiles of all data points. Whiskers extend to the 2 extreme values of all data points. Significant differences of medians of apoptosis induction, as determined by Mann-Whitney U-test, are indicated (^**P < .01; ^*P < .03). (D) Western blotting of protein lysates generated from representative AML samples (patients no. 15 and no. 11) after in vitro treatment with indicated compounds for 48 hours. (E) A Ph⁺ ALL patient was treated with VPA/ATRA. Changes in peripheral white blood counts, percentage of granulocytes in the peripheral blood, and platelets are depicted during the first 45 days of treatment (i). Hypercellular bone marrow aspiration showing a homogenous ALL-blast infiltrate before treatment (ii, left panel), and bone marrow histology obtained after induction of remission with VPA/ATRA (ii, middle and right panels). Note the rare infiltration with CD34⁺/TdT⁺ ALL blasts in a hypocellular marrow (arrows in ii, middle and right panels).