Sequential therapy with JX-594, a targeted oncolytic poxvirus, followed by sorafenib in hepatocellular carcinoma: preclinical and clinical demonstration of combination efficacy

Abstract

JX-594 is a targeted and granulocyte-macrophage colony stimulating factor (GM-CSF) expressing oncolytic poxvirus designed to selectively replicate in and destroy cancer cells through viral oncolysis and tumor-specific immunity. In a phase 1 trial, JX-594 injection into hepatocellular carcinoma (HCC) was well-tolerated and associated with viral replication, decreased tumor perfusion, and tumor necrosis. We hypothesized that JX-594 and sorafenib, a small molecule inhibitor of B-raf and vascular endothelial growth factor receptor (VEGFR) approved for HCC, would have clinical benefit in combination given their demonstrated efficacy in HCC patients and their complementary mechanisms-of-action. HCC cell lines were uniformly sensitive to JX-594. Anti-raf kinase effects of concurrent sorafenib inhibited JX-594 replication in vitro, whereas sequential therapy was superior to either agent alone in murine tumor models. We therefore explored pilot safety and efficacy of JX-594 followed by sorafenib in three HCC patients. In all three patients, sequential treatment was (i) well-tolerated, (ii) associated with significantly decreased tumor perfusion, and (iii) associated with objective tumor responses (Choi criteria; up to 100% necrosis). HCC historical control patients on sorafenib alone at the same institutions had no objective tumor responses (0 of 15). Treatment of HCC with JX-594 followed by sorafenib has antitumoral activity, and JX-594 may sensitize tumors to subsequent therapy with VEGF/VEGFR inhibitors.

Figure 1

JX-594 replication in liver cancer lines is inhibited in the presence of sorafenib in vitro. (a) Infectability of PLC/PRF/5 (human hepatoma) cells (parental) and sorafenib-adapted PLC/PRF/5 cells was determined by addition of JX-594 expressing green fluorescent protein at multiplicities of infection (MOI) of 0, 0.1 and 1. Images were taken with a Zeiss Axiovision microscope fluorescent microscope and analyzed using Axiovision acquisition and image storage software 24 hours postinfection. (b,c) JX-594 replication in the presence of sorafenib was measured by plaque formation and burst assay after incubation with sorafenib for 3 days (plaque formation assay) or 24 hours (burst assay). Results are expressed as percent of no-sorafenib control (replicate mean + SD). (d) The ability of JX-594 to form plaques in the absence or presence of increasing concentrations of sorafenib on monolayers of hepatocellular carcinoma (HCC) cell lines. Results are expressed as percent of no sorafenib control. (e) JX-594 replication was tested by burst assay using a panel of HCC cell lines infected at MOI of 0.1 for 2 hours, followed by change to media with 4 µmol/l sorafenib. After 48 hours, cells and supernatants were collected for titration by plaque assay on A2780 cells. Results are expressed as percent of no-sorafenib control (replicate mean + SD). (f) Cell viability in the presence of sorafenib was determined 24 hours after addition of sorafenib to PLC/PRF/5 and HepG2 cells plated at 60% and 100% densities. Results are expressed as percent of no-sorafenib control (replicate mean + SD).

Figure 2

Combination therapy with sorafenib enhances JX-594 efficacy in two murine tumor models. (a,b) SCID mice with subcutaneous HepG2 tumors of 12–14 mm diameter were randomized to five groups and treated with PBS, daily sorafenib, weekly JX 594, simultaneous sorafenib and JX-594, or sequentially treated with sorafenib for 2 weeks followed by JX-594, or with 2 weekly dose of JX-594 followed by sorafenib. Analysis of time-to-tumor-progression (TTP) was performed. Kaplan–Meier curves show TTP, with 5,000 mm³ tumor volume considered as progression. (c) B16 model of lung metastases: C57BL/6 mice were injected with 3 × 10⁵ B16-F10-LacZ cells intravenously and 24 hours later were treated with phosphate buffered saline (PBS), sorafenib alone (50 µg/kg per oral dosing daily for 2 weeks) JX594 alone (10⁷ plaque-forming units (pfu) intravenously three times per week for 1 week) or in combination (n = 5 per group). Three weeks after treatment initiation, mice were killed and lungs were fixed and stained to detect surface B16 tumor nodules.

Figure 3

JX-594 treatment of patients with advanced hepatocellular carcinoma cell may sensitize to subsequent sorafenib therapy. (a) Patient 1705 was treated with JX-594 at a dose level of 10⁸ plaque-forming units (pfu) intratumorally for three treatments every 2 weeks (week 0, week 2, week 4). Sorafenib therapy was initiated at week 10.5. Antitumor response was evaluated by dynamic contrast-enhanced magnetic resonance imaging (MRI) at baseline, after treatment with JX-594 alone and after sorafenib initiation. Red circles indicate target tumors. The darker areas within the target tumors at week 13 represent significant increasing necrosis within the target tumors is seen at week 11, manifested by nonenhancement within these tumors. (b) Patient 1702 was treated with JX-594 at a dose level of 10⁹ pfu intratumorally for three treatments every 2 weeks (week 0, week 2, week 4). Sorafenib therapy was initiated at week 13. Response was evaluated by dynamic MRI imaging at baseline, after treatment with JX-594 alone and after sorafenib initiation. Red circles indicate target tumors. Mild amount of necrosis is seen in the larger tumor at baseline. The necrosis increases somewhat following JX-594 alone, but significant necrosis is identified following sorafenib therapy. (c) Patient 1712 was treated with JX-594 at a dose level of 10⁹ pfu intratumorally for three treatments every 2 weeks (week 0, week 2, week 4). Sorafenib therapy was initiated at Week 8. Response was evaluated by dynamic MRI imaging at baseline, after treatment with JX-594 alone and after sorafenib initiation. Red ovals indicate areas of the liver not injected with JX-594. These images demonstrate antitumor activity in noninjected lesions similar to that seen in injected lesions from (a) and (b). Progressively developing necrosis within these tumors is seen following JX-594 therapy alone and then sorafenib.

Figure 4

Three-dimensional segmentation analysis reveals significant tumor necrosis induction following JX-594 and sorafenib combination treatment. (a) Patient 1705 dynamic magnetic resonance imaging scans collected at baseline, week 8, and week 13 were evaluated using 3D segmentation analysis to assess the volumes of viable and necrotic tumor at each timepoint. Dark orange, green, and yellow indicate viable tumor tissue. Lighter colors within each tumor indicate geographic areas of necrosis. The liver margin is displayed as a wire frame. Note that on the week 13 image, the viable tumor in the largest lesion has been nearly completely replaced by necrosis. (b) Percent tumor necrosis (compared to total tumor volume) at baseline, after JX-594 therapy alone and after subsequent sorafenib initiation by tumor.

Figure 5

Tumor destruction in patient with advanced, poor prognosis renal cell cancer following JX-594 and sunitinib combination treatment. (a) Patient 301 was treated with JX-594 at a dose level of 10⁹ plaque-forming units (pfu) intratumorally in liver metastases for four treatments every 3 weeks (week 0, week 3, week 6, and week 9). Sunitinib therapy was initiated at ~week 17. Contrast-enhanced computed tomography scans of patient 301 at baseline and 9 months after JX-594 and sunitinib treatment. A noninjected, large hypervascular mass in the left abdomen shows essentially complete resolution 9 months following JX-594 and sunitinib therapy. (b) Whole body positron emission tomography scan of patient at baseline and 9 months after JX-594 and sunitinib treatment shows complete disappearance of fludeoxyglucose-avid tumor following JX-594 and sunitinib.