Enhanced In vitro and In vivo Cytotoxicity of Combined Reovirus and Radiotherapy

Purpose: To test combination treatment schedules of reovirus and radiation in human and murine tumor cells in vitro and in vivo.

Experimental Design:In vitro cytotoxicity and cell cycle effects of reovirus given alone and combined with radiotherapy were assessed by colorimetric, tissue culture infectious dose 50, and fluorescence-activated cell sorting–based assays. Interactions between the agents were evaluated using combination index analysis. The effect of different schedules of reovirus and radiotherapy on viral replication and cytotoxicity was tested in vitro and the combination was assessed in three tumor models in vivo.

Results: Characterization of reovirus cytotoxicity in a panel of cell lines yielded a range of sensitivities. Combined reovirus and radiotherapy yielded statistically significantly increased cytotoxicity, particularly in cell lines with moderate susceptibility to reovirus alone. The enhanced cytotoxicity of the combination occurred independently of treatment sequence or schedule. Radiation did not affect viral replication and only reduced reoviral cytotoxicity after clinically irrelevant single doses (>50 Gy). Combination index analysis revealed synergy between radiation (3-10 Gy) and reovirus at multiplicities of infection between 0.001 and 1. Combination treatment significantly increased apoptosis in tumor cells relative to either single-agent treatment. In vivo studies using xenograft and syngeneic tumors showed enhanced activity of the combination relative to reovirus or radiation alone (P < 0.001).

Conclusions: Combining reovirus and radiotherapy synergistically enhances cytotoxicity in a variety of tumor cells in vitro and in vivo. These results offer strong support for translational clinical trials of reovirus plus radiotherapy that have been initiated in the clinic.

Reoviruses are nonenveloped icosahedral viruses with a segmented genome composed of 10 segments of double-stranded RNA. They are ubiquitous nonpathogenic viruses that can be isolated from the respiratory and gastrointestinal tracts of humans (1), although they are not associated with a specific clinical syndrome. Most healthy adults possess antireoviral antibodies, suggesting a high incidence of subclinical infection in early life (2). Three serotypes of reovirus (type 1 Lang, type 2 Jones, and type 3 Abney and type 3 Dearing) have been defined based on their antibody neutralization and hemagglutination-inhibitory activities (3).

Following the observation that normal and transformed cells manifest differential sensitivities to reovirus infection (4), it has become apparent that this agent is preferentially replication competent in cells with an activated Ras pathway through either Ras mutation or up-regulated epidermal growth factor receptor signaling (5–8). This effect is mediated, at least in part, through failure of activation of the double-stranded RNA-activated protein kinase in Ras-activated cells that are exposed to reovirus infection (8). In contrast, in normal cells, the presence of an intact double-stranded RNA-activated protein kinase system prevents the establishment of a productive infection. Reovirus has been shown to cause tumor regressions after intralesional injections in immunodeficient mice and after systemic administration in immunocompetent mice (5, 7, 8).

Reovirus represents a potential therapy for a range of cancers, given the fact that Ras activation is present in 30% to 40% of human tumors (9, 10). In particular, activating Ras mutations are present in 80% to 90% of pancreatic cancers (11), 40% to 50% of colorectal cancers (12), 50% of thyroid tumors (13), 30% of myeloid leukemias (14), and 24% of lung cancers (15). Ras can also be activated by upstream mitogenic signals, notably by receptor tyrosine kinases. Overexpression or mutation of receptor tyrosine kinases can result in elevated Ras activity and may render cells permissive to reovirus replication. Examples of this include overexpression of epidermal growth factor receptor in 90% of head and neck cancers (16), platelet-derived growth factor receptor and/or epidermal growth factor receptor in glioblastomas (17), and c-erbB2 in 25% to 30% of breast cancers (18). Recent studies have suggested that the activity of the Ras/RalGEF/p38 pathway is a major determinant of the susceptibility of tumor cells to reovirus-mediated oncolysis (6).

Ionizing radiation is an important treatment for a range of tumor types. For many tumors, radiotherapy is curative, either as a single modality or in combination with radiosensitizing chemotherapy (19–22). For other tumors, radiotherapy plays a critical role as part of adjuvant therapy following surgical resection. There is emerging evidence that the Ras signal transduction pathway plays a role in determining the radiosensitivity of tumors (23). Thus, overexpression of epidermal growth factor receptor, activating mutations of Ras, and phosphorylation of Akt (protein kinase B) and phosphoinositide-3-kinase have all been associated with increased radiation resistance in vitro and, in the case of epidermal growth factor receptor and Akt, to radiation treatment failure in cancer patients (23–26). In contrast, Ras pathway inhibitors, such as the farnesyltransferase inhibitors, sensitize cells to radiation (27) and similar results have been obtained with the use of adenoviral vectors encoding an inactivating single-chain antibody against Ras (28). Therefore, combining radiation with an oncolytic viral therapy that targets Ras-activated cells represents an attractive treatment option.

There are several potentially positive theoretical interactions between reovirus and radiotherapy. First, cells with mutated Ras that are relatively radioresistant will be most sensitive to reoviral oncolysis. Therefore, combining radiation and reovirus potentially represents a form of spatial cooperation (29) in which the two different elements of the treatment combination target different cell populations. Second, the stress response in irradiated cells may enhance reoviral replication and oncolysis in an analogous fashion to the increased oncolysis that is seen in cells under osmotic stress.⁷

Third, radiation may increase the ability of reovirus to infect tumor cells by affecting the expression of cell surface receptors and a similar effect has been seen with adenovirus infection after therapeutic irradiation (30, 31). Fourth, radiation-induced changes in cell cycle distribution of tumor cells may render them more permissive to reovirus infection and/or replication. Fifth, in an in vivo system, radiation may reduce the high tumor interstitial pressure that represents a significant barrier to effective intratumoral spread of large molecules and biological particles (32).

In this report, we describe for the first time the effects of combining reovirus with ionizing radiation.

Cell lines. HN5, SIHN-5B, SIHN-11B, Detroit-562 (head and neck cancer; from Dr. S. Eccles, Institute of Cancer Research, London, United Kingdom; ref. 33), SK-Mel28, B16 [melanoma; from Professors C. Marshall (Institute of Cancer Research) and A.A. Melcher (St. James's University Hospital, Leeds, United Kingdom), respectively], HCT116, SW480 (colorectal cancer; from Professor C. Marshall), 293A (adenocarcinoma; Quantum Biotechnologies), and L929 (mouse fibroblast; Oncolytics Biotech, Inc.) were cultured in DMEM. MCF-7 (European Collection of Animal Cell Cultures) cells were cultured in RPMI 1640 with 0.01 mg/mL bovine insulin (Sigma-Aldrich). Both media contained 5% (v/v) FCS, 1% (v/v) glutamine, and 0.5% (v/v) penicillin/streptomycin. Cell lines were maintained at 37°C and 5% CO₂. Plating, irradiation, and reovirus infection were carried out in DMEM containing 2% (v/v) FCS, 1% (v/v) glutamine, and 0.5% (v/v) penicillin/streptomycin.

Reovirus stocks. Reovirus (Dearing type 3) stocks at 3.45 × 10¹⁰ tissue culture infectious dose 50 (TCID₅₀/mL) were obtained from Oncolytics Biotech and stored in the dark at neat and 1:50 concentrations either in PBS or DMEM containing 2% (v/v) FCS, 1% (v/v) glutamine, and 0.5% (v/v) penicillin/streptomycin at 4°C (maximum 3 months) or at −80°C (long-term storage).

Plaque assays. Viral stability was checked regularly using plaque assay. L929 cells, plated in six-well plates at a density of 1 × 10⁶ per well, were infected with dilutions of viral stocks. After incubation at 37°C for 3 h, the virus solution was removed and the wells were overlaid with a 1:1 mixture of 2% SeaPlaque agarose (BioWhittaker Molecular Applications) and 2× DMEM containing 4% (v/v) FCS, 2% (v/v) glutamine, and 1% (v/v) penicillin/streptomycin at 37°C. Wells were stained with 500 μL of 0.03% neutral red (Sigma-Aldrich) in PBS 72 to 96 h after infection and plaques were counted 3 to 4 h later.

Reovirus infection of cells. Reovirus stock was first diluted to the highest multiplicity of infection (MOI) that was to be used in the experiment (usually 10 or 100) and subsequently serially diluted to the various MOIs required for each individual study. Infections were done according to two protocols: for colorimetric cytotoxicity assays, cells were incubated with virus-containing medium or medium alone at 37°C, 5% CO₂ for 24 h after which time an equivalent volume of fresh medium was added to the well, and for fluorescence-activated cell sorting (FACS) assays, cells were incubated with virus for 2 to 4 h at 37°C, 5% CO₂ after which time the medium was aspirated and replaced with fresh medium.

Crystal violet survival assay. Cell viability was quantified using crystal violet stain (Sigma-Aldrich) at either 1% (w/v) for staining into medium or 0.2% (w/v) for direct cell staining, both in a 7% (v/v) solution of ethanol/PBS. After incubation at 37°C for 10 min, wells were washed twice with PBS. Images of the plates were captured on a Microtek ScanMaker 8700 (Microtek International, Inc.). Thereafter, crystals were solubilized in DMSO (VWR International Ltd.) and absorbance was measured at 550 nm on a SpectraMax 384 plate reader (Molecular Devices).

3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide assay. Cell viability was quantified using a 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay. Briefly, 20 μL MTT (thiazolyl blue; Sigma-Aldrich) at 5 mg/mL in PBS was added to treated cells in a 96-well plate. After 4-h incubation at 37°C, crystals were solubilized in DMSO and absorbance was measured at 550 nm on a SpectraMax 384 plate reader.

X-irradiation of cells or reovirus. All irradiations were done using a Pantak H.F. 320 kV X-ray machine (Ago X-ray Ltd.) with samples placed 27 cm beneath the X-ray source. Before irradiation of either cells or reovirus, the dose rate was determined using a Farmer Sub-Standard X-ray dosimeter Mk.2/S3 according to the manufacturer's instructions. Typically, the dose rate for irradiations was between 6.6 and 6.8 Gy/min at 240 kVp and 10 mA. Cells were irradiated in 24-well plates (BD Labware) or 96-well plates (Nunc) or in 75 cm² tissue culture flasks (BD Biosciences) depending on the experimental design. Exposure times were calculated to correspond to single fraction doses ranging from 1 to 10 Gy with a maximum exposure time of 91 s for 10 Gy. For irradiation of reovirus, viral stocks were diluted to a MOI of 100 (assuming subsequent infection of cells at a density of 1 × 10⁵ per well) in DMEM plus 2% FCS and exposed to single fraction doses of 0, 10, 25, 50, and 100 Gy over a maximum time period of 15 min and 9 s for 100 Gy.

UV-irradiation of reovirus. Viral stock at 3 × 10⁹ plaque-forming units/mL was divided into 125 μL aliquots in round-bottomed 96-well plates with the lid off and irradiated for 0, 10, and 30 min with an inverted 365-nm UV lamp.

Reovirus one-step growth curves. HCT116 cells were plated at 1 × 10⁵ per well in a 24-well plate. At 2 or 24 h before or after infection, cells received 8 Gy of radiation. Cells were infected with virus at a MOI of 5 (to ensure infection of all cells) for 2 h at 37°C and washed twice with medium and 0.5 mL of complete growth medium (DMEM) with 2% FCS was added to each well. At various times after infection, cells were scraped into the medium and the lysate was frozen at −70°C. Virus titer was determined by TCID₅₀ on L929 cells as described below.

TCID₅₀ assays. TCID₅₀ assays were done on either virus-infected cell lysates or diluted viral stocks. Virus-infected cells, which had or had not been irradiated, were subjected to three freeze-thaw cycles between −80°C and room temperature and lysates were collected. For assays of X- or UV-inactivation of reovirus, irradiated virus was diluted to 6 × 10⁷ plaque-forming units/mL in DMEM containing 2% (v/v) FCS, 1% (v/v) glutamine, and 0.5% (v/v) penicillin/streptomycin. The resulting lysates or viral suspensions were diluted in 10-fold series and used in quadruplicates to infect L929 cells plated in 96-well plates at 2 × 10⁴ per well. After incubation at 37°C for 3 h, the lysates or viral suspensions were removed and replaced with DMEM containing 2% (v/v) FCS, 1% (v/v) glutamine, and 0.5% (v/v) penicillin/streptomycin and incubated for a further 3 to 4 days at 37°C. Viral titer was calculated using the Kärber statistical method as follows. The proportion of cytopathic effect per dilution was scored (each of the four wells/dilution is given a score of 0.25 if it shows cytopathic effect and 0 if there is no cytopathic effect) and TCID₅₀ values for each condition were calculated using the following formula: titer = 10 × 10^{1 + 1(score − 0.5)}.

FACS analysis of cell survival. Infected and/or irradiated cells were harvested, washed thrice with PBS, and resuspended at 1 × 10⁶ in 500 μL PBS. Cells were stained with 5 μL propidium iodide (PI) at 1 mg/mL and 50 μL of fluorescein diacetate (FDA; Invitrogen-Molecular Probes) at 100 ng/mL and incubated for 10 min. Twenty thousand events were collected and analyzed on a FACSCalibur flow cytometer (Becton Dickinson) using CellQuest Pro version 5.2 (BD Biosciences).

FACS analysis of cell cycle distribution. Infected and/or irradiated cells were harvested, washed thrice with PBS, and resuspended in 200 μL PBS/0.1% FCS. While vortexing, ice-cold 70% ethanol was added dropwise to a final volume of 4 mL and cells were stored at 4°C for a minimum of 1 h and up to 1 week. Before FACS analysis, a further three PBS washes were carried out and cells were resuspended at 1 × 10⁶/mL in PBS. Cells were treated with RNase A (Sigma-Aldrich) at a final concentration of 100 μg/mL, stained with PI at a final concentration of 40 μg/mL, and incubated at 37°C for 20 min. Samples were analyzed on a FACSCalibur flow cytometer using CellQuest Pro version 5.2.

4′,6-Diamidino-2-phenylindole staining for apoptosis. Cells were plated in six-well plates at 1 × 10⁵ per well and infected with reovirus at MOIs of 0, 0.1, or 1. Plates were irradiated to 5 Gy (or mock irradiated) 16 h later. At 48 h after infection, cells were fixed in 3.8% formaldehyde (BDH Laboratory Supplies) in PBS and stored at 4°C. At the time of assay, cells were washed twice in PBS and stained with 2 μL of a 1 mg/mL solution of 4′,6-diamidino-2-phenylindole (Invitrogen-Molecular Probes) in 1 mL of PBS. After 20 min, a further 1 mL PBS was added and cells were viewed under a Nikon Eclipse E500 confocal microscope. Cells were scored as apoptotic if they showed the typical microscopic features of DNA condensation and fragmentation. The proportion of apoptotic cells in four microscopic fields was counted.

Tumor irradiation. Dosimetry of the in vivo irradiation system was checked using thermoluminescent dosimetry in wax phantom mice (25 g) bearing right flank “tumors” made from tissue equivalent material. Data were calibrated against thermoluminescent dosimetry from the same batch irradiated to 1 to 10 Gy on a 6 MV linear accelerator (Varian) in the Radiotherapy Department, Royal Marsden Hospital NHS Trust that were read on a Toledo 654 thermoluminescent dosimetry reader (data not shown; D.A. Pitman). Before therapeutic tumor irradiation, mice received an i.p. injection of 100 μL of a 1:1:4 mixture of Hypnorm (0.315 mg/mL fentanyl citrate and 10 mg/mL fluanisone; Janssen-Cilag Ltd.), Hypnovel (5 mg/mL midazolam; Roche Products Ltd.), and water for injection BP (Fresenius Health Care Group). Control animals were also anesthetized in the same way. This well-established regimen has been shown to provide effective short-duration anesthesia and has the advantage of maintaining better tissue perfusion than barbiturate anesthesia (34). Furthermore, tumor blood flow is only slightly reduced with this combination of anesthetic agents (35). Anesthetized animals were positioned in an irradiation jig with the s.c. tumors exposed under an aperture in a 3-mm lead sheet that shielded the rest of the body. Following irradiation, to limit hypothermia to a minimum, animals were wrapped in toweling jackets until they recovered consciousness (approximately 30-60 min).

In vivo studies. Xenograft tumors (SW480 and HCT116) were established in female MF1 nude mice (Charles Rivers plc) and syngeneic tumors (B16) were established in immunocompetent C57BL6 mice (Charles Rivers plc). Animals were maintained under specific pathogen-free conditions in sterile filter-top cages on sterile bedding and fed an irradiated diet and autoclaved, acidified water (pH 2.8) ad libitum. In all cases, tumors were established by s.c. injection of 1 × 10⁶ cells suspended in 100 μL PBS in the right flank using a 25-gauge needle. None of the cell lines required coinjection with Matrigel. Once tumors had reached 5 to 8 mm in diameter, the mice were randomly allocated into treatment groups that were stratified by tumor size. In all experiments, the following four treatment groups were used: (a) control (no radiotherapy, intratumoral PBS), (b) radiotherapy alone [12 Gy in four fractions (for SW480 and HCT116) or 20 Gy in five fractions (for B16), intratumoral PBS], (c) reovirus alone [no radiotherapy, one (SW480), two (HCT116), or three (B16) doses of intratumoral reovirus], and (d) radiotherapy plus reovirus [12 Gy in four fractions (for SW480 and HCT116) or 20 Gy in five fractions (for B16), one (SW480), two (HCT116), or three (B16) doses of intratumoral reovirus]. The mean (SD) tumor volumes (mm³) on the first day of treatment for the various study groups were as follows: SW480: control, 136.6 (60.1); radiotherapy alone, 70.3 (29.8); reovirus alone, 108.2 (43.1); reovirus plus radiotherapy, 73.6 (28.2); HCT116: control, 47.0 (31.1); radiotherapy alone, 41.4 (17); reovirus alone, 46.4 (24.9); reovirus plus radiotherapy, 43.5 (17.3); B16: control, 65.5 (28.6); radiotherapy alone, 82.8 (54.6); reovirus alone, 64.4 (24.8); reovirus plus radiotherapy, 82.4 (51.2). In all cases, reovirus was administered slowly by direct intratumoral injection in 100 μL PBS. Controls received 100 μL PBS alone without reovirus. A single cutaneous puncture site was used, and once in a s.c. location, the 25-gauge needle was redirected along multiple tracks within the tumor to achieve maximal dispersal of the reovirus. By careful application of this technique, it was possible to achieve direct intratumoral injection without backflow of the injectate. This route of administration was selected for these studies to provide essential background data in support of a phase I clinical trial (that is currently ongoing) in which patients receive intratumoral injections of reovirus during a course of radiotherapy. Future work will involve detailed analysis of the i.v. route of viral administration as a prelude to subsequent clinical evaluation of radiotherapy combined with i.v. reovirus. Three orthogonal tumor diameters (d₁, d₂, and d₃) were measured by Vernier calipers thrice weekly and tumor volume was calculated from the formula V = π/6 d₁·d₂·d₃. The tumor volume on the first day of radiotherapy was defined as V_o and animals were studied until the tumor volume had increased to 3V_o. Use of this measure was designed to spare the animals from the physical distress of unnecessarily large tumor burdens and to comply with the Medical Research Council guidelines (Responsibility in the Use of Animals for Medical Research, 1993).

Statistical analysis. Comparisons between groups were done using the t test. Survival curves were estimated using the Kaplan-Meier method, and significance was assessed using the log-rank or the χ² test. Statistical analysis was done using the Statistical Package for the Social Sciences version 14.0 (SPSS, Inc.) for Figs. 5D and 6A–C. GraphPad Prism version 4.0 was used for Figs. 1A and C, 2B, 3A, and 4B and C.

The effect of the combination of reovirus and radiation on cell proliferation was assessed by calculating combination index (CI) values using CalcuSyn software (Biosoft). Derived from the median-effect principle of Chou and Talalay (36), the CI provides a quantitative measure of the degree of interaction between two or more agents. A CI of 1 denotes an additive interaction, >1 denotes antagonism, and <1 denotes synergy. Experiments were done using a range of doses in a nonconstant ratio checkerboard design.

Reovirus cytotoxicity in tumor cell lines. The effect of reovirus infection in a range of tumor cell lines was assessed. HN5, SIHN-5B, SIHN-11B, Detroit-562 (head and neck cancer), SK-Mel28, B16 (melanoma), HCT116, SW480 (colorectal cancer), MCF-7 (breast), 293A (adenocarcinoma), and L929 (mouse fibroblast) were infected with reovirus at MOIs from 0.01 to 10 and survival was measured by MTT assay (Fig. 1A). These data show differing sensitivities of the various cell lines to reovirus-induced cytotoxicity. L929 and 293A cells showed dramatic sensitivity to reovirus (controls versus virus-infected cells, P < 0.001 for all MOI), whereas Detroit-562, MCF-7, SIHN-5B, SK-Mel28, HCT116, and HN5 cells were moderately sensitive (controls versus virus-infected cells, P < 0.001 for MOI of 1, 5, and 10). B16 cells were sensitive to reovirus at the highest MOI of 5 and 10 (P < 0.01 and 0.001, respectively). SW480 and SIHN-11B cells seemed to be relatively resistant to reovirus cytotoxicity.

Reovirus is not inactivated by clinically relevant doses of X-irradiation. Before proceeding to studies in which reovirus was combined with ionizing radiation, the effect of X- and UV-irradiation on the cytotoxicity of reovirus on HCT116 cells was measured. Clearly, if therapeutic doses of ionizing radiation are able to inactivate reovirus in vitro, this agent will not have clinical utility in combination with radiotherapy. Therefore, reovirus was irradiated to single-fraction doses of 0, 10, 25, 50, and 100 Gy at a dose rate of 6.8 Gy/min in 50 mL tubes (BD Biosciences Falcon) after the stock solution (3.45 × 10¹⁰ TCID₅₀/mL) had been diluted in DMEM to a nominal MOI of 100. These doses of radiation are greatly in excess of doses that are used as part of daily fractionated radiotherapy in the clinic. HCT116 cells at a density of 1 × 10⁵ per well in 24-well plates were subsequently infected with the irradiated virus stocks at MOI between 0.0001 and 10. Cell survival was assessed 96 h later by crystal violet assay (Fig. 1B and C). These data showed no evidence of reovirus inactivation at radiation doses of ≤25 Gy, such that cytotoxicity was significantly greater (P < 0.001) for virus-infected cells compared with controls for MOI as low as 0.0001. However, there was loss of reovirus cytotoxicity at the higher radiation doses (50 and 100 Gy) at MOI lower than 0.1 and 1, respectively. In separate experiments, the effect of X- and UV-irradiation on viral titer was measured using a TCID₅₀ assay. Stock solutions of reovirus received X-irradiation to 0, 10, or 50 Gy or UV-irradiation for 0, 10, or 30 min. Once again, X-irradiation up to 50 Gy was seen to cause a minor reduction (<1 log) in reovirus titer. In contrast, 10 and 30 min of UV-irradiation caused virtually 100% loss of virus titer (>7 logs; Fig. 1D).

Combined reovirus and radiation treatment enhances cytotoxicity in tumor cell lines. A panel of six tumor cell lines was infected with reovirus at MOI ranging from 0.001 to 10 in 24-well plates. Sixteen hours after infection, cells received either mock irradiation or a single 5 Gy fraction of radiation. Cell survival was measured at 96 h by crystal violet (Fig. 2A) and MTT (Fig. 2B; Supplementary Fig. S1) assay. As seen in Fig. 2B, significant increases in cytotoxicity were observed for the combination of reovirus and radiation in many of the cell lines at MOI as low as 0.001. These differences remained statistically significant when the survival data were normalized either to the respective irradiated controls (i.e., unirradiated cells treated with reovirus normalized to the 0 Gy, no reovirus control and irradiated cells treated with reovirus normalized to the 5 Gy, no reovirus control; Fig. 2B) or when all data were normalized to the 0 Gy, no reovirus control (Supplementary Fig. S1). The effect of the combined treatment was most marked at low MOI and in those cell lines that showed only moderate susceptibility to reovirus cytotoxicity in the absence of radiation [HCT116 (P < 0.001 for MOI of ≥0.01), SIHN-5B (P < 0.001 for MOI of 0.001-0.1), MCF-7 (P < 0.001 for MOI of 0.001 and 5, P < 0.01 for MOI of 0.1 and 1), and Detroit-562 (P < 0.001 for MOI of 0.001-0.01)]. A statistically significant benefit from the combination of reovirus and radiation was not seen in 293A cells that were particularly sensitive to reovirus infection.

In these analyses, colorimetric assays were used because formal clonogenic assays were shown to be unreliable. Rates of 100% cell kill with reovirus alone (and reovirus plus radiation) were consistently seen in clonogenic assays even with very low MOI of input reovirus (Supplementary Fig. S2). Photomicrographs of tumor cell lines treated with reovirus confirmed that treatment was associated with cell death rather than a purely cytostatic action (Supplementary Fig. S3).

Enhanced cytotoxicity is radiation dose dependent and most marked at low MOI. HCT116 cells at a density of 1 × 10⁵ per well in 24-well plates were infected with reovirus at MOI between 0.0001 and 50 and irradiated 16 h later at doses of 0, 1, 3, 5, and 10 Gy. Cell survival was determined after 96 h by crystal violet assay. These data showed that the enhancement of cell killing was particularly evident at radiation doses of ≥3 Gy and that this effect seemed to operate for cells treated at MOI between 0.001 and 1. In contrast, the effect was less evident when MOI of 10 or 50 was used. Therefore, in unirradiated cells, there was a stepwise incremental effect of increasing MOI, whereas in cells irradiated to 10 Gy the level of cytotoxicity was equivalent across a range of MOI from 0.001 to 50 (Fig. 3A; Supplementary Fig. S4). The interaction between radiation and reovirus was assessed using the CalcuSyn program based on the principle of Chou and Talalay (36). CIs were calculated and are shown in Supplementary Table S1 and Fig. 3B. Using this methodology, a CI of <0.7 is indicative of synergy, with a CI of <0.3 considered strongly synergistic. Strong synergism or synergism was shown for cells exposed to 3, 5, and 10 Gy radiation and reovirus at MOI of 0.001 to 1. Within these treatment variables, the CI values were between 0.26 and 0.36 (Supplementary Table S1). Outside of these variables, the combination proved to be moderately synergistic or additive and at the extremes of dose (high radiation and very low reovirus or low radiation and very high reovirus) antagonistic. At these extreme doses, the agent at the low dose had little additional effect on cell survival compared with the agent at the high dose alone.

The effect of the combination of reovirus and radiation was further assessed in HCT116 cells by means of FACS analysis using dual staining with PI and FDA. In this case, following treatment with reovirus, radiation, or reovirus plus radiation, the percentage of cells with low PI and high FDA (PI-ve/FDA+ve) staining (equivalent to live cells) was compared with that with high PI and low FDA (PI+ve/FDA-ve) staining (equivalent to dead cells). For the combined treatment, there was a significant increase in the percentage of PI+ve/FDA-ve–stained cells compared with either single-agent treatment or controls (Fig. 3C and D).

Cytotoxicity is not affected by the sequence of treatment combination. To evaluate the effect of radiation schedule on the cytotoxicity of reovirus, a series of experiments was done in which the interval between the two treatments was varied. In the first schedule, HCT116 cells were infected with reovirus at MOI between 0.0001 and 100 and radiation doses of 0, 1, 3, and 5 Gy were administered 1, 4, 8, 16, and 24 h later. In the second schedule, the order of treatments was reversed such that the same radiation doses were delivered 1, 4, 8, 16, and 24 h before cells were infected with reovirus at MOI between 0.0001 and 100. In all cases, cell survival was assessed by crystal violet assay at 96 h after infection (Fig. 4A and B). Both schedules showed enhanced cytotoxicity in a radiation dose-dependent fashion that was most marked at lower MOI. For each schedule, cytotoxicity was independent of the timing of radiation. These findings were confirmed by assessing cytotoxicity by means of FACS analysis of cell survival by PI/FDA staining. In this case, cells were irradiated with 5 Gy either 24 h before or after reoviral infection at a MOI of 1. The cytotoxicity seen with the combined treatment was equivalent (86.7% versus 87.7%) irrespective of the sequence in which the treatments were delivered (Fig. 4C).

Radiation does not increase reoviral replication in a one-step growth curve assay. The effect of in vitro irradiation of HCT116 cells on their ability to support viral replication was measured using a one-step growth curve assay. Early (2 h) and late (24 h) time points were used and cells were irradiated either before or after viral infection. There was no evidence of any effect of a single 8 Gy fraction of radiation administered before, or after, infection on the titer of reovirus recovered from infected HCT116 cells (Fig. 4D).

Reovirus and radiation individually induce cell cycle arrest and together increase apoptosis. The effect of either irradiation or reovirus infection on cell cycle distribution was investigated in HCT116 cells by means of FACS analysis with PI staining of permeabilized cells. Radiation at doses of 1, 3, 5, and 10 Gy resulted in a radiation dose-dependent accumulation of cells in the G₂-M phase of the cell cycle, with significant reduction in cells in G₁ and S. This blockade increased between 4 and 16 h after radiation and persisted at the higher radiation doses (3-10 Gy) at 24 h (Fig. 5A). Infection with reovirus also caused cell cycle perturbation with a virus dose-dependent accumulation of cells in S and G₂-M phases of the cell cycle. This was most marked at the 72-h time point (Fig. 5B). The effect of the combined reovirus and radiation treatment on cell cycle distribution is shown in Fig. 5C. In this case, the redistribution of cells to G₂-M (with radiation) and to G₂-M and S (with reovirus) was still evident, but when the two agents were combined, it was overshadowed by the presence of a significant sub-G₁ population. This suggests that the combined treatment caused increased levels of apoptosis in excess of that seen with either treatment modality alone. This mechanism of cell death was studied further by direct quantitation of apoptosis using 4′,6-diamidino-2-phenylindole staining and counting of treated cells (Fig. 5D). This study showed statistically significant enhancement of apoptosis when reovirus was combined with radiation at a dose of 5 Gy (P = 0.006 for MOI of 0.1; P < 0.001 for MOI of 1).

Combined reovirus and radiation treatment enhances tumor growth delay in vivo in immunodeficient and immunocompetent mice. The in vivo effects of the combination of reovirus and radiation were tested in three different tumor cell lines (Fig. 6A-C). Both SW480 and HCT116 tumor xenografts showed increased tumor growth delay in nude mice in response to the combined treatment and syngeneic B16 tumors showed the same effect in C57BL6 mice. For SW480 tumors, the median survival was increased from 7.1 days in the control group to 25.7 days with radiation, 13.3 days with reovirus alone, and was not reached in animals treated with a combination of reovirus and radiation (P < 0.0001; Fig. 6A). For HCT116 tumors, a similar pattern of results was seen with the median survival increased from 5.7 and 5.3 days in controls and reovirus alone groups, respectively, to 15.2 days with radiation and 43.6 days with reovirus and radiation (P < 0.0001; Fig. 6B). The results in the B16 tumors also showed a highly statistically significant increase in median survival with reovirus and radiation compared with the other groups. Median survivals were 5.1, 14.2, 10.8, and 31.5 days in control, radiation alone, reovirus alone, and reovirus plus radiation groups, respectively (P < 0.0001; Fig. 6C). There was no evidence of exacerbation of cutaneous toxicity in the mice treated with the combination therapy compared with radiotherapy alone (data not shown). Similarly, there was no difference in the weights of the mice that received the combination therapy or radiotherapy alone (data not shown).

Reovirus type 3 (Dearing) represents a promising new anticancer agent that is currently in phase I clinical trials. Thus far, it is safe and well tolerated by intratumoral and i.v. routes. There have been some preliminary suggestions of antitumor activity in these studies (37), but it is unlikely that reovirus possesses sufficient single-agent activity to represent a stand-alone therapy. Rather, it is probable that this agent will be developed as part of a combination treatment strategy with radiotherapy, chemotherapy, or chemo-radiotherapy. The potential advantages of this approach include exploiting additive or synergistic interactions between reovirus and other treatment modalities, the ability of reovirus to target radiation- or drug-resistant cell populations in the tumor, the potential to use radiation to enhance favorably the biodistribution/penetration of reovirus in tumors, and the prospect of radiation- or drug-induced local or systemic immune suppression reducing the humoral and cellular immune response against the virus.

In these studies, we have shown a marked increase in cytotoxicity when reovirus and radiotherapy are used in combination, especially at low input MOIs of reovirus. Interestingly, the combination effect was statistically significant in those cell lines that exhibit relative resistance to reoviral oncolysis, suggesting that this may represent more than a simple additive effect. Further detailed analysis of the two treatments using the CI analysis confirmed that there was synergy between them across a broad range of clinically relevant doses. Therefore, at clinically deliverable fractional radiation doses of 3 and 5 Gy, there was strong synergy between the two agents between MOI of 0.001 and 1. Such MOIs are likely to occur at or around tumor cells in clinical trials. For instance, after intratumoral injection of large viral titers, reoviral leakage, sequestration in necrotic areas, and immune clearance are likely to mean that the effective MOI in certain regions of the tumor will be within the synergistic window (0.001-1). Similarly, after i.v. administration, it is likely that tumor colonization by reovirus will occur at relatively low levels (due to immune clearance in the circulation, nonspecific binding to nontarget tissues, and difficulties in accessing tumor cells through the process of extravasation). Therefore, it seems that clinical studies are well placed to exploit the synergistic interaction between reovirus and radiation. For example, our current phase I dose escalation trial is designed to use radiation doses of 3 to 4 Gy per fraction in combination with intratumoral injections of reovirus.

Despite the observation of a synergistic interaction between the two treatments, the precise nature of the interaction between the two agents remains to be determined. The data from the one-step growth curve clearly showed that the enhanced cytotoxicity of the combined treatment was not due to increased viral infectivity of irradiated cells because there was no difference in viral output when the cells were irradiated before or after infection. Furthermore, the data from the one-step growth curve exclude a radiation-mediated increase in viral replication. Instead, the increased cytotoxicity seemed to be due to a marked increase in apoptosis with the combination treatment. These findings are in keeping with data showing abrogation of reovirus-induced apoptosis (but not infection or replication) in C26 colorectal cancer cells in which expression of mutant KrasD12 gene has been repressed (38). Current work in our laboratory is focussing on elucidating the role of the Ras signaling pathway in determining the apoptotic response to combined reovirus and radiation treatment.

The observation that the combination effect was most marked at relatively low input MOIs, when a minority of tumor cells were infected, raises the possibility that the synergistic effect of the combined treatment was due, at least in part, to a bystander effect that may have been mediated by a secreted substance. Reovirus infection has been shown to induce tumor necrosis factor-α and tumor necrosis factor–related apoptosis-inducing ligand expression in tumor cells (39–41) and these agents have been shown to sensitize to the effects of ionizing radiation in several in vitro and in vivo systems (42–45). We are currently investigating the potential role of tumor necrosis factor-α and tumor necrosis factor–related apoptosis-inducing ligand signaling as mediators of the enhanced cytotoxicity seen with combined radiation and reovirus treatment.

In these in vitro and in vivo studies, we also attempted to address several important issues relating to the clinical implementation of combined reovirus and radiation. We have shown that, in contrast to its sensitivity to UV, reovirus is relatively resistant to ionizing radiation even at high, clinically irrelevant single fraction doses of 100 Gy. Therefore, we can confidently conclude that, if the two treatments are combined in the clinic, there will be no significant risk of reovirus being inactivated by the radiation that is delivered to the tumor. In addition, these studies showed that the enhanced cytotoxicity of the combination was not schedule or sequence dependent. This is an important consideration for clinical studies that use fractionated radiotherapy schedules because it is likely that, if multiple virus injections are used, reovirus administration will follow some fractions of radiotherapy but precede others. The in vivo data from three different tumor models delivering radiotherapy doses over four or five fractions provided further evidence of the fact that reovirus is stable in irradiated tissues and acts independently of treatment schedule. Current investigations in the laboratory are assessing the effects of combining reovirus with cytotoxic chemotherapy, with and without radiotherapy, with a view to designing appropriate combination schedules for in vivo testing.

Therefore, in these studies, we have shown that combining reovirus with ionizing radiation results in enhanced cytotoxicity in a range of tumor cell types in in vitro and in vivo experiments. CI analysis confirms that, within a range of clinically deliverable fractional radiation doses and low reovirus MOI, there is evidence of synergism between the two agents. Consequently, we are currently conducting a phase I dose escalation study of reovirus in combination with two different schedules of short-course palliative radiotherapy in patients with advanced cancers. In the first part of this study, patients received 20 Gy in five fractions with two reovirus injections (on days 2 and 4) according to a dose escalation schedule between 10⁸ and 10¹⁰ TCID₅₀ per injection. In the second part of the study, patients have received 36 Gy in 12 fractions with increasing numbers of injections of reovirus (from two to six) at 10¹⁰ TCID₅₀. This clinical trial aims to assess the safety and efficacy of the combined approach in patients with cancer and includes exploratory biological end points of antitumor activity and immune response. In addition, further clinical trials of reovirus in combination with cytotoxic chemotherapy have begun to recruit. Separate protocols involve i.v. reovirus in combination with paclitaxel and carboplatin (in patients with ovarian, head and neck cancer, and melanoma), gemcitabine (in patients with advanced cancer), and docetaxel (in prostate cancer).