| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Cancer Therapy: Preclinical |
Authors' Affiliations: 1 Division of Neurosurgery, Department of Surgery; Departments of 2 Molecular Genetics and Microbiology and 3 Pathology, Duke University Medical Center, Durham, North Carolina; and 4 Center of Biologics Evaluation and Research, Food and Drug Administration, Rockville, Maryland
Requests for reprints: John H. Sampson, Division of Neurosurgery, Duke University Medical Center, Box 3050, Durham, NC 27710. Phone: 919-684-9041; Fax: 919-684-9045; E-mail: john.sampson{at}duke.edu.
| Abstract |
|---|
|
|
|---|
Experimental Design: To evaluate for toxicity, PVS-RIPO was administered into the spinal cord of transgenic mice that express the human poliovirus receptor, CD155, and into the intrathecal space of athymic rats without tumor. To evaluate efficacy, two different doses of PVS-RIPO were administered intrathecally 3 days after athymic rats were inoculated intrathecally with an aggressive human glioblastoma multiforme xenograft.
Results: No clinical or histologic evidence of toxicity was found. In efficacy studies, median survival was increased by 174.47% from 8.5 days in the group treated with UV light-inactivated virus to 15 days in the rats treated with 1.0 x 107 plaque-forming units (pfu) of PVS-RIPO (P < 0.0001). A similar increase in median survival was seen in the group receiving 1.0 x 109 pfu PVS-RIPO (P < 0.0001); however, there was no statistically significant dose-response relationship (P = 0.345). In addition, 1 of 10 rats in lower-dose PVS-RIPOtreated group and 3 of 10 rats in higher-dose PVS-RIPOtreated group survived >60 days after tumor cell inoculation and had no evidence of residual tumor at autopsy.
Conclusion: These results suggest that intrathecal treatment with PVS-RIPO may be useful for treatment of neoplastic meningitis in patients with glioblastoma multiforme and provides a rationale for clinical trials in this area.
Poliovirus is a nonenveloped, plus-stranded RNA virus belonging to the Picornaviridae family and is the causative agent of paralytic poliomyelitis. In poliomyelitis selective targeting of motor neurons is determined by the distribution of the poliovirus cellular receptor (26), the immunoglobulin superfamily molecule, CD155 (27). We have previously shown that a number of tumors, including glioblastoma multiforme, also express this cellular receptor making it an ideal tumor associated target for oncolytic viral therapy (13, 28). Moreover, CD155 recently has been identified as a key determinant of tumor cell migration, invasion, and metastasis (29), and transcriptional activation of the CD155 gene has been linked to signaling pathways commonly deregulated in malignancy, including sonic hedgehog/gli (30), fibroblastic growth factor, and oncogenic ras (31).
To apply poliovirus to tumor therapy, we eliminated its inherent neuropathogenicity by manipulating an essential cis-acting regulatory element involved in gene expression control (32). Poliovirus, like all picornaviruses, uses an internal ribosomal entry site contained within its extensively structured 5' untranslated region to initiate translation in a cap-, 5' end-independent manner (33, 34). Replacing the cognate poliovirus internal ribosomal entry site with its counterpart from human rhinovirus type 2 eliminated the virus' ability to cause poliomyelitis in mice transgenic for CD155 and Cynomolgus macaques (32, 35). However, this alteration does not reduce the cytopathogenicity of the lytic virus for malignant cell types that express CD155, such as glioblastoma multiforme.
In this study, we evaluated the toxicity and efficacy of intrathecal administration of a novel clinical grade recombinant poliovirus, PVS-RIPO, developed through the NIH Rapid Access to Inventional Development program, in transgenic mice expressing the human poliovirus receptor, CD155, and in an athymic rat model of human glioblastoma multiforme neoplastic meningitis. Intraspinal and intrathecal administrations of PVS-RIPO were found to be nontoxic in both the transgenic mice and athymic rats. PVS-RIPO was also highly effective against human glioblastoma multiforme xenografts growing in the subarachnoid space of athymic rats.
| Materials and Methods |
|---|
|
|
|---|
EGFR, was obtained from W. Cavenee (University of California at San Diego; ref. 36). These cells are grown in zinc option medium with 0.23% HEPES, 0.22% NaCO3, and 10% fetal bovine serum. Preparation and inactivation of virus. PVS-RIPO was derived from a previous version of the recombinant virus, PV1(RIPO), by substituting the coding region of poliovirus type 1 (Mahoney) with its counterpart from the type 1 live attenuated Sabin [PV1(S)] vaccine strain. The synthesis of PVS-RIPO was reported previously (13). In brief, infectious PV1(S) cDNA [clone pS1 (T7)], kindly provided by A. Nomoto (University of Tokyo, Japan), was digested with AvaI. The resulting 7.0-kb restriction fragment was ligated with a PCR fragment amplified from PV1 (RIPOS) using primers (a) 5'-GGGTCGACTAATACGACTCCTATAGTAAAACAGCTCTGGGGTTGT-3' and (b) 5'-CCATTTCTCGGGCACTGGAGCG-3' and pBS vector DNA (New England Biolabs, Beverly, MA) digested with AvaI and SalI. PVS-RIPO cDNA was processed for rederivation of virus as described before (32).
Virus inactivation by UV irradiation was carried out in a Stratagene UV-linker according to previously established procedures. The absence of virus infectivity after UV treatment of samples was confirmed by plaque assay as described below.
Plaque assay. Cell monolayers, grown in DMEM with 10% fetal bovine serum, penicillin (10 units/mL), and streptomycin (100 µg/mL), were inoculated with a viral suspension at a multiplicity of infection of 10 and incubated for 30 minutes at room temperature. Culture dishes were thoroughly rinsed with five changes of growth medium to remove unbound particles and plated at 37°C thereafter. At the indicated intervals, culture dishes were removed and treated with three consecutive freeze/thaw cycles. Virus titers in infected monolayer cultures were quantified by plaque assay as described previously (24). The PVS-RIPO titers that we used in this study were determined based on our previous preclinical studies in experimental animals evaluating i.t. delivery of the agent (13, 24).
Antigen capture analysis of CD155 expression. The method of antigen capture for CD155 has been reported previously (13, 28). In brief, cells were grown in monolayer culture and homogenized in lysis buffer. The protein concentrations in cell lysates were determined using the Bradford method. ELISA plates were treated over night with monoclonal anti-CD155 antibody, D171, and suspended in carbonate buffer. Treated plates were rinsed thrice with PBS and individual wells were filled with cell lysate diluted with PBS to achieve the desired protein concentration. Plates were incubated for 4 hours at room temperature while gently rocking. Subsequently, the wells were thoroughly rinsed thrice with PBS and treated with SDS-PAGE loading buffer at 95°C to release the captured antigen. The recovered material was analyzed by SDS-PAGE, transferred to nitrocellulose membranes, and subjected to Western blot. Filters were blocked over night in TBST [10 mmol/L Tris (pH 8), 150 mmol/L NaCl, 0.05% Tween 20] containing 3% dry milk. After treatment with the polyclonal anti-CD155 antibody, D480 (diluted in TBST, 3% dry milk at 1:1,000) for 1 hour at room temperature, the blots were rinsed thrice in TBST and incubated for 1 hour at room temperature with secondary biotinylated anti-rabbit IgG antibody (diluted in TBST, 3% dry milk at 1:500; Vector Labs, Burlingame, CA). After three washes with TBST, the filters were treated with streptavidin-horseradish peroxidase complex (diluted in TBST, 3% dry milk at 1:2,000; Roche, Indianapolis, IN) and developed with a chemoluminescent substrate (enhanced chemiluminescence; Amersham, Arlington Heights, IL).
Toxicity study in CD155 transgenic mice. The evaluation of neurovirulence of poliovirus strains in CD155 transgenic mice has been described previously (37). Serotype 1 PV reference strain (WHO/I) and two clinical lots of PVS-RIPO (0402018, 2nd passage; 0404007, 5th passage in Vero cells) were administered intraspinally to CD155 transgenic mice (TgPVR21) and neurologic symptoms were evaluated using a standard clinical score as follows: grade 0, no neurologic symptoms; grade 1, paralysis of limb; grade 2, death. Evaluation of neurologic symptoms was done from days 0 to 14 after intraspinal PVS-RIPO inoculation. The rodent model for paralytic poliomyelitis in TgPVR21 mice was developed as a replacement for the traditional monkey neurovirulence assay in Cynomolgus macaques. It is a rigorous and standardized clinical assay. It has been established that clinical observation alone is sufficient to predict the neurovirulent potential of live attenuated poliovirus strains, as the sole neurologic manifestation of poliovirus infection in the CNS of TgPVR21 mice is motor neuron death (38). Therefore, and because of the large number of animals enrolled in our assay, we did not evaluate the spinal cords of infected TgPVR21 mice histopathologically.
Athymic rat model of neoplastic meningitis. Subarachnoid catheters were implanted into female athymic rats using an established procedure described previously (39). Briefly, the rats (BIG:NIMRrnu; 190-260 g) were anesthetized by an i.p. injection of ketamine (55 mg/mL) and xylazine (9 mg/mL) and placed in a stereotactic frame (David Kopf Instruments, Tujunga, CA) with their necks flexed at a 70-degree angle using the tilt adapter. A midline sagittal incision was made from the inion to the laminar arch of C1, atlanto-occipital membrane was exposed and incised using a 20-gauge needle, and the underlying cisterna magna dura matter was opened. A PE-10 catheter (Intramedic, Franklin Lakes, NJ) with a 5-0 stainless steel wire stylet was then inserted into the subarachnoid space to the lumbar region (8.5 cm) by passing it along the posterior aspect of the spinal cord. The stylet was removed, a loose knot was tied in the catheter just above the opening of the dura, and the knot was secured in place with dental epoxy (Lang Dental Manufacturing Co., Chicago, IL). The exposed catheter end was then passed through the skin lateral to the incision. The incision was closed in three layers using 6-0 Ethilon (Ethicon, Somerville, NJ), and the catheter was temporarily occluded with 2-0 stainless steel wire. The rats were allowed to recover at least 7 days, and only rats showing normal weight and neurologic function and no evidence of infection were used in the experiments.
Neoplastic meningitis was induced by injecting tumor cells through the indwelling subarachnoid catheter. After anesthetizing rats with isoflurane, the wire stylet was removed, and 5.0 x 106 U87MG
EGFR cells were injected in 40 µL of PBS using a 1,000-µL Hamilton syringe and injector (Hamilton Co., Reno, NV). The catheter was then flushed with 20 µL of PBS and reoccluded with 2-0 wire.
Toxicity and efficacy studies in athymic rats. For intrathecal toxicity studies, athymic rats were injected with 1.0 x 109 pfu PVS-RIPO in 60 µL of PBS. The weight, neurologic status, survival, and histologic examination of the neural axis and internal organs were compared with a control group treated with PBS. For efficacy studies, rats in the treatment groups were injected with 1.0 x 107 or 1.0 x 109 pfu PVS-RIPO in 60 µL of PBS, and rats in the control groups were injected with 1.0 x 109 pfu UV-inactivated PVS-RIPO in the same volume of PBS or just a PBS. Clinical neurologic function included testing of the stepping and placing reflex and the ability to climb a 60-degree incline ramp. Histologic examination was conducted on six representative cross-sections of the CNS, including the forebrain at level of lateral ventricles, hindbrain at the level of occipital lobe, and four equally spaced spinal sections, including the cauda equina. These sections were evaluated microscopically for hemorrhage, necrosis, edema, demyelination, and arachnoid fibrosis.
Statistical analysis. Survival estimates and median survival was determined by using the method of Kaplan and Meier (40). Survival data were compared using the log-rank test.
| Results |
|---|
|
|
|---|
|
|
EGFRvIII expresses CD155 and is susceptible to PVS-RIPO. To evaluate the level of expression of the poliovirus receptor, CD155, in the human glioblastoma multiforme cell line U87MG
EGFR, we used an antigen capture/Western blot assay. This assay showed that CD155 was detected in the U87MG
EGFR cell line (Fig. 1). Next, we evaluated whether or not PVS-RIPO could be propagated in cultured U87MG
EGFR cells and found that PVS-RIPO propagated at levels comparable with cell lines commonly used for large-scale propagation of virus in the laboratory (i.e., HeLa cells; ref. 32; Fig. 2). These results indicated that U87MG
EGFR cells were appropriate targets for oncolytic PVS-RIPO therapy.
|
|
EGFR was investigated at two different doses, and survival was compared with groups of rats treated with PBS or UV light-inactivated PVS-RIPO. After inoculation of 5.0 x 106 U87MG
EGFR cells, median survival was increased by 187.5% from 8 days in the PBS-treated group to 15 days in the group treated with 1.0 x 107 pfu PVS-RIPO (P < 0.0001; Fig. 3). Similarly, the median survival time was increased by 174.47% when compared with the group treated with the UV light-inactivated PVS-RIPO. In the group receiving the higher dose of PVS-RIPO (1.0 x 109 pfu), the median survival was increased by 200% compared with the PBS-treated group and 188.23% compared with the group treated with UV-inactivated PVS-RIPO (P < 0.0001). There was, however, no statistically significant dose-response relationship (P = 0.345). In addition to the increase in median survival, 1 of 10 rats in lower-dose PVS-RIPOtreated group and 3 of 10 rats in higher-dose PVS-RIPOtreated group survived over 60 days after tumor cell inoculation. Histologic examination of their brains and spinal cords showed no evidence of residual tumor.
|
| Discussion |
|---|
|
|
|---|
In our previously published work, we showed that the poliovirus receptor, CD155, is stably expressed in established glioblastoma multiforme cell lines as well as primary explant cultures (28). We have also previously shown that the oncolytic recombinant poliovirus, PVS-RIPO, is efficacious in an athymic rat model of intracerebral breast cancer metastasis (13). We now present new data based on our investigations with a clinical grade recombinant poliovirus, PVS-RIPO, prepared for human use by the NIH Rapid Access to Inventional Development program, that this oncolytic viral approach may also have efficacy in glioblastoma multiforme neoplastic meningitis after intrathecal delivery.
Although we show here that intrathecal PVS-RIPO therapy may be effective against human glioblastoma multiforme neoplastic meningitis in an athymic rat model, there may still be significant limitations to its success in humans that will need to be evaluated. For example, in our preclinical study, a two log increase in virus dose had no effect on the efficacy of intrathecal viral treatment. Although the lower dose of virus may be "saturating," it remains possible that treatment failures are due to the existence of resistant subpopulations of cells, perhaps those not expressing CD155, or tumor locales not accessible to inoculated virus, notwithstanding the dose.
Another potential obstacle is the fact that all patients that might receive this therapy will have been previously vaccinated against poliovirus. However, available evidence suggests that neutralizing antibodies do not exist in the CNS of vaccinated individuals, and only patients who experience neurologic symptoms of poliovirus infection, indicating intraspinal replication of virus, produce neutralizing antibodies in the cerebrospinal fluid. Still, given the disruption of the blood-brain barrier in patients with glioblastoma multiforme, conclusive evidence for or against existing neutralizing antibodies within the CNS will need to await human studies.
Similarly, monkeys have been shown to be able to contract poliomyelitis after intracerebral administration of pathogenic poliovirus, even if they had been previously vaccinated, suggesting that prior vaccination does not limit viral spread within the CNS (41). Although this is concerning for toxicity, with PVS-RIPO, our preclinical work has shown it to be devoid of neuropathogenic potential even in the complete absence of a preexisting immune response. This is partially due to the heterologous HRV2 internal ribosomal entry site that selectively abrogates viral replication in neuronal cells. Perhaps more important then, is the role that preexisting immunity may play in thwarting i.t. propagation of PVS-RIPO, and whether such a response will limit the efficacy of this approach.
Another potential concern with regard to human application is the potential for deleterious adaptation of the virus by mutation because all RNA viruses are inherently genetically unstable. All of our previous experiments have shown that adaptation to neurovirulence does not occur under the circumstances proposed for our clinical trials (13, 24). Nonetheless, such events will need to be monitored carefully during clinical trials as they can never be categorically excluded with replicating RNA viruses.
In conclusion, intrathecal delivery of PVS-RIPO seems to have promise as a safe and efficacious agent for human glioblastoma multiforme neoplastic meningitis and should be investigated in phase I clinical trials.
| Footnotes |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Received 7/25/05; revised 10/19/05; accepted 12/ 5/05.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
L. M. Hales, N. J. Knowles, P. S. Reddy, L. Xu, C. Hay, and P. L. Hallenbeck Complete genome sequence analysis of Seneca Valley virus-001, a novel oncolytic picornavirus J. Gen. Virol., May 1, 2008; 89(5): 1265 - 1275. [Abstract] [Full Text] [PDF] |
||||
![]() |
H. Toyoda, J. Yin, S. Mueller, E. Wimmer, and J. Cello Oncolytic Treatment and Cure of Neuroblastoma by a Novel Attenuated Poliovirus in a Novel Poliovirus-Susceptible Animal Model Cancer Res., March 15, 2007; 67(6): 2857 - 2864. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. A. Campbell, M. Mulvey, I. Mohr, and M. Gromeier Attenuation of Herpes Simplex Virus Neurovirulence with Picornavirus cis-Acting Genetic Elements J. Virol., January 15, 2007; 81(2): 791 - 799. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cancer Prevention Research |
| Cancer Prevention Journals Portal | Cancer Reviews Online |
| Annual Meeting Education Book | Meeting Abstracts Online |