AZD8835 is active in vitro and in vivo in ABC DLBCLs. (A) Downregulation of PI3Kα is required to induce cytotoxicity in HBL-1 and OCI-Ly10 cells treated with the PI3Kδ inhibitor idelalisib. In contrast, BJAB cells that were used as a negative control do not respond to PI3Kα knockdown and idelalisib treatment. Data are expressed as means ± standard deviation of at least 2 independent experiments. (B) PI3Kα/δ inhibition induces apoptosis. AZD8835 increases the rate of apoptotic cells in HBL-1 and TMD8 after 5 days of treatment. Data are expressed as means ± standard deviation of at least 2 independent experiments. (C) AZD8835 treatment reduces cell proliferation. CFSE dilutions after AZD8835 treatment are measured after 3 and 5 days. Cell proliferation is reduced in HBL-1 and TMD8 cells compared with DMSO-treated cells. Representative results of 3 independent experiments are shown. (D) Tumor growth curves of TMD8 and OCI-Ly10 xenograft mouse models following treatment with either AZD8835 (red) or vehicle control (blue). AZD8835 significantly reduces tumor growth in ABC DLBCL xenograft mouse models (P = .0091 for AZD8835 [n = 10] vs vehicle [n = 10] on day 22 in TMD8; P = .0003 for AZD8835 [n = 9] vs vehicle [n = 9] on day 11 in OCI-Ly10; 1-tailed 2-sample Student t tests). Treatment was initiated after animals developed macroscopic signs of tumors (day 6 after engraftment in TMD8; day 24 after engraftment in OCI-Ly10). Error bars indicate the standard error of the mean. (E) Tumor growth curves for KTC PDX mouse model. Treatment with AZD8835 (red) significantly reduces tumor growth compared with the vehicle control (blue) in the KTC mouse model (P = 5.36 × 10⁻⁵ for AZD8835 [n = 8] vs vehicle [n = 8] on day 15; 1-tailed 2-sample Student t test). Treatment was initiated after animals developed macroscopic signs of tumors (day 15 after engraftment). Error bars indicate the standard error of the mean. *P < .05; **P < .01; ***P < .001.