Combining a PI3K Inhibitor with a PARP Inhibitor Provides an Effective Therapy for BRCA1-Related Breast Cancer

Pharmacodynamic effects of PI3K inhibitor NVP-BKM120 on breast carcinomas in MMTV-CreBrca1f/fTrp53^+/− mice. Female virgin mice developed spontaneous breast cancers at ages 8 to 12 months. A, representative ¹⁸FDG PET–CT scan images of a tumor-bearing mouse at baseline (image on the left) and within 48 hours after starting treatments with the PI3K inhibitor NVP-BKM120 (50 mg/kg/d by gavage, image on the right). This mouse had developed 4 simultaneous tumors. Arrows in orange, green, blue, and white are used to identify different tumors upon baseline (left) and posttreatment (right) imaging. The color palette for uptake ranges from dark blue to bright yellow with increasing count intensity. The changes in FDG uptake were determined as described in Methods. They decreased by 45% (tumor with orange arrow), 64% (green arrow), 64% (blue arrow), and 56% (white arrow). B, suppression of AKT phosphorylation on S473 as a result of treatments with NVP-BKM120 in vivo. Tumor tissue was obtained via core needle biopsy before and after 2 weeks of treatments with NVP-BKM120, fixed, and processed for IHC with anti-pAKT (S473) antibodies. For additional IHC images, see Supplementary Fig. S1. C, decrease in FDG uptake in 6 mammary carcinomas. Relative decrease in FDG uptake was determined by the ratio of uptake at 48 hours (blue bars) or 2 weeks (red bars) to baseline. Tumor-specific FDG uptake was determined as described in Methods. For additional PET–CT images, see Supplementary Fig. S2. D, concordance of decrease in FDG uptake and tumor shrinkage during a 2-week treatment with PI3K inhibitor NVP-BKM120. The tumor-bearing animal was imaged with FDG-PET (top, tumor indicated with a yellow arrow before and after treatment, decrease in uptake 93%) and concomitant CT scan (bottom) before (left) and on treatment (right). The tumor is again marked in the CT scan with a yellow arrow in the axial and sagittal plane. The red outline indicates the tumor circumference before treatment to visualize treatment effect on tumor size.

Antiangiogenic effects of PI3K inhibitor NVP-BKM120. A, gross pathologic images of an untreated tumor (left), a tumor treated for 2 weeks (middle), and a tumor treated for 6 weeks (right) with NVP-BKM120 at 50 mg/kg/d via gavage. IHC to detect CD31 in an untreated tumor (B), the center of a tumor treated for 6 weeks (C), and the tumor capsule of a mammary tumor treated for 6 weeks (D). E, determination of the Chalkley score to quantify CD31 staining. IHC with anti-CD31 antibodies was conducted in pretreatment biopsies and in tumor specimens from mice at the time of tumor progression.

PI3K inhibition increases PAR and H2AX phosphorylation. A, compensatory pathway activation induced by treatments with NVP-BKM120. HCC1937 or SUM149 cells were treated with NVP-BKM120, olaparib, or its combination as indicated for 72 hours, lysed, and subjected to immunoblotting with antibodies against total AKT, EGFR, ERK and their phospho-specific epitopes. B, in vivo increase of γ-H2AX–positive cells after treatment with NVP-BKM120 and proliferative activity at the “pushing margin.” Tumor-bearing mice were subjected to a pretreatment biopsy and then treated with NVP-BKM120 at 50 mg/kg/d. IHC of pretreatment biopsies and posttreatment tumor tissues was conducted with antibodies as indicated. C, BRCA1-mutant human HCC1937 or SUM149 cells were treated with vehicle control or NVP-BKM120 at the indicated concentrations for 24 hours, lysed, and subjected to immunoblotting with antibodies against PAR, phosphorylated AKT (p-AKT; S473), g-H2AX, CC3 as an apoptosis marker, and actin as a loading control. D, effects of combined PI3K and PARP inhibition on Brca1-mutant cells. Cells were treated with NVP-BKM120 at 1 μmol/L and olaparib at 10 μmol/L or their combination for 24 hours, lysed, and subjected to immunoblotting with antibodies against PAR, p-AKT, total AKT, g-H2AX, and actin as indicated. p-EGFR, phosphorylated ERK.

Effects of NVP-BKM120 and KU-55933 and their combination on the DNA damage response. A, HCC1937 cells were treated for 18 hours with NVP-BKM120 at 2.5 μmol/L, KU-55933 at 10 μmol/L, or their combination, subjected to ionizing radiation (IR) with 10 Gy or mock, lysed 6 hours later, and subjected to immunoblotting with antibodies as indicated. B–E, loss of Rad51 focus formation in response to ionizing radiation in the presence of NVP-BKM120. Breast cancer cells were isolated from primary tumors from MMTV-CreBrca1f/fTrp53^+/− mice and either treated with vehicle control (B and C) or NVP-BKM120 (D and E) for 18 hours, followed by irradiation with 10 Gy. Six hours later, cells were fixed and processed for immunofluorescence with antibodies against Rad51 and counterstained with 4′,6-diamidino-2-phenylindole (DAPI). F, induction of DNA-PK and H2AX phosphorylation and loss of RAD51 occur in response to PI3Kα, not PI3Kβ inhibition. SUM149 cells were transfected with siRNA pools depleting PI3Kα (left) or PI3Kβ (right). Cells were lysed after 48 hours and subjected to immunoblotting with antibodies as indicated.

Antitumor efficacy of PI3K inhibitor NVP-BKM120 alone and in combination with olaparib. A–D, tumor-bearing MMTV-CreBrca1f/fTrp53^+/− were treated with either vehicle control (A), NVP-BKM120 [B, 50 mg/kg/d (n = 11) or 30 mg/kg/d (n = 10)], olaparib [C, 50 mg/kg/d (n = 8)] or the combination of NVP-BKM120 and olaparib [D, NVP-BKM 50 mg/kg/d + olaparib 50 mg/kg/d (n = 8) or NVP-BKM 30 mg/kg/d + olaparib 50 mg/kg/d (n = 7)], and tumor volumes were measured every 2 to 3 days using calipers. Trend lines for vehicle control (red curve) and NVP-BKM120 treatments (green curve) were calculated using all data points to determine best fit. The functions of the best-fit curves were used to determine tumor-doubling times for all 3 treatment modalities and controls. E, stable body mass with PI3K inhibitor and PARP inhibitor treatments. Mice were weighed before and after treatments. F and G, target inhibition and pharmacokinetics in vivo. Tumor tissues harvested from animals treated with NVP-BKM120 (30 mg/kg/d) alone or in combination with olaparib (50 mg/kg/d) as indicated were harvested 3 hours after the last treatment and subjected to immunoblotting with antibodies against actin, p-AKT, and γ-H2AX (F) or lysed and subjected to mass spectrometry (G). For standards used, see Methods and Supplementary Fig. S5. H and I, responses of human BRCA1-related breast cancers implanted as xenotransplants into nude mice to NVP-BKM120, olaparib, or their combination. Breast cancer tissues from 2 patients, one with a 185delAG germline mutation (H) and the other one with a 2080delA germline mutation (I) were propagated as subcutaneous implants in nude mice. Tumors were allowed to grow to a size of 5 mm when mice were randomized to treatments with either vehicle control (black), NVP-BKM120 (red), olaparib (green), or their combination [blue (n = 6 for each cohort, same dosing as in F)]. Tumor assessment with electronic calipers was done as described in Methods.