Perifosine, a Bioavailable Alkylphospholipid Akt Inhibitor, Exhibits Antitumor Activity in Murine Models of Cancer Brain Metastasis Through Favorable Tumor Exposure

Abstract

Metastatic brain tumors are regarded as the most advanced stage of certain types of cancer; however, chemotherapy has played a limited role in the treatment of brain metastases. Here, we established murine models of brain metastasis using cell lines derived from human brain metastatic tumors, and aimed to explore the antitumor efficacy of perifosine, an orally active allosteric Akt inhibitor. We evaluated the effectiveness of perifosine by using it as a single agent in ectopic and orthotopic models created by injecting the DU 145 and NCI-H1915 cell lines into mice. Initially, the injected cells formed distant multifocal lesions in the brains of NCI-H1915 mice, making surgical resection impractical in clinical settings. We determined that perifosine could distribute into the brain and remain localized in that region for a long period. Perifosine significantly prolonged the survival of DU 145 and NCI-H1915 orthotopic brain tumor mice; additionally, complete tumor regression was observed in the NCI-H1915 model. Perifosine also elicited much stronger antitumor responses against subcutaneous NCI-H1915 growth; a similar trend of sensitivity to perifosine was also observed in the orthotopic models. Moreover, the degree of suppression of NCI-H1915 tumor growth was associated with long-term exposure to a high level of perifosine at the tumor site and the resultant blockage of the PI3K/Akt signaling pathway, a decrease in tumor cell proliferation, and increased apoptosis. The results presented here provide a promising approach for the future treatment of patients with metastatic brain cancers and emphasize the importance of enriching a patient population that has a higher probability of responding to perifosine.

Keywords: AKT; PI3K; apoptosis; brain cancer; metastasis; signaling pathway.

Figure 1

Pharmacokinetic profile of perifosine in plasma and brain tissue. Healthy animals were administered a single oral dose of perifosine, and at six intervals, mice from each group were then euthanized for plasma (A) and brain (B) extraction (n = 3). (C) The brain-to-plasma ratio of perifosine was calculated at each time point. (D) Relationship between the brain C_max and in vitro IC₅₀ for DU 145 and H1915 cells.

Figure 2

Representative morphological features of tumor cell lines following implantation. (A–D) Histological and (E) macroscopic images of the brains. Intracranial injection of DU 145 (A, B) and H1915 (C, D) cells into nude mice shows characteristic tumor growth patterns in brain on day 14. Representative H&E-stained (A, C) and cytokeratin (AE1/AE3)-stained (Ba, Da) sections are shown (scale bar = 500 μm). Boxed areas in panels (Ba) and (Da) show regions depicted in panels (Bb) and (Db-g), respectively. Cytokeratin (AE1/AE3) were used to distinguish brain tissues from metastatic DU 145 (Bb) and H1915 (Db-g) tumors; scale bar = 50 μm. (E) Gross features of representative brains excised from sham-operated (left; n = 5) and H1915-implanted (right; n = 4) mice on day 21. *p < 0.05 versus the sham-operated group.

Figure 3

Prolongation of life-span of orthotopic brain metastasis mice with perifosine. Mice were injected intracerebrally with DU 145 (A) or H1915 (B–E) cells on day 0, and the tumor was allowed to establish until day 3 (perifosine [D3]) or day 7 (perifosine [D7]). Mice were treated with either vehicle or perifosine under the conditions indicated in Tables 1 , 2 . (A, B) Survival curve analysis was performed for 64 days, and the death time point was defined when the animals appeared moribund. **p < 0.01; ***p < 0.001 versus the control group. (C, D) The H1915 orthotopic tumor mice were euthanized in a moribund state (C; up to day 63) or at the termination of the experiment (D; on day 64), and each brain was weighed. Although all 10 mice in the control group were euthanized in a moribund state until day 32, the brains were weighed only for seven mice (C). Age-matched normal (n = 3) and sham-operated (n = 5) mice were included for comparison and euthanized on day 21 (C). (E) Representative H&E-stained images of lesions in the H1915 model are shown; scale bar = 500 μm (a, d) or 50 μm (b, c, e, f). Low (a) and high (b, c) power views in the vehicle-treated mouse on day 14 and low (d) and high (e, f) power views in the perifosine (D3)-treated mouse on day 41. Boxed areas in panels (a) and (d) show regions depicted in panels (b, c, e, f), respectively. Arrowheads indicate tumor cells.

Figure 4

Induction of high antitumor efficacy against ectopic H1915 xenografts in association with blockade of the PI3K/Akt pathway by perifosine. Mice bearing subcutaneous DU 145 (A-E) and H1915 (F-J) tumors were randomized on day 1 and treated orally with either vehicle or perifosine using a 5-day-on/2-day-off × 3 cycle schedule (n = 5). On day 22, xenograft tumors were excised 4 h following the last dose. Changes in TV (A, F), tumor growth inhibition rate (TGI, %) at different days (B, G), tumor weight (C, H), and relative body weight (D, I). The expression level of key molecules was detected by western blot on Day 22 (E, J). Images are representative of at least two independent experiments with similar results. β-actin was used as a loading control. Protein levels were quantified and represented as the percentage of the control group. *p < 0.05; **p < 0.01; ***p < 0.001 versus the control group. (K) Representative H&E images of the H1915-tumor tissues treated with vehicle (a-c, e) and perifosine (d, f) on days 7 and 22 are shown; scale bar = 20 μm (a, b) or 50 μm (c-f). High-power views in the control group suggest that the tumor cells are large, round-to-oval, and possessed distinct multiple nucleoli in the clear oval nucleus (a, b).

Figure 5

Morphological changes in subcutaneous tumor xenografts evoked by oral administration of perifosine. The resected tumors in Figures 4C, H were analyzed by H&E staining, immunohistochemical staining for Ki-67 and cleaved caspase-3, and special staining for van Gieson. Representative images of the DU 145 (A–J) and H1915 (K–T)-tumor tissues treated with vehicle (A–E, K–O) and perifosine (F–J, P–T) are shown; scale bar = 500 μm (A, F, K, P) or 50 μm (B–E, G–J, L–O, Q–T). Granulomatous legion is delineated by a solid black line (P). Perifosine induced granuloma formation around the H1915 tumor lesion (P). Necrosis (asterisks), mitotic figures (dotted arrows), Ki-67-positive cells (arrows), and cleaved caspase-3-positive cells (arrowheads).

Figure 6

Plasma and tumor pharmacokinetic profiles of perifosine after single and repeated doses. Nude mice with DU 145 (A–C) and H1915 (D–F) subcutaneous xenograft tumors were randomized and treated with single doses of 45 and 180 mg/kg perifosine or a 5-day repeated dose of 180/45 mg/kg (180 mg/kg loading dose followed by maintenance doses of 45 mg/kg) perifosine. Plasma (A, D) and tumor (B, E) samples were collected at the indicated points. Chronological changes in tumor-to-plasma ratios of perifosine are shown for DU 145 (C) and H1915 (F) models (n = 3, except at 48-h time point for the 180 mg/kg group in (E, F), where n = 2 due to tumor sample processing error).