Background
Despite recent advances in glioma treatments such as surgical resection, chemotherapies, and novel drug therapies, the survival rate of patients with advanced-grade glioma remains dismal [
1]. For that matter, the desperate clinical context of high-grade glioma warrants the present development of novel strategies with the potential for direct patient impact. An emerging, highly promising strategy for the treatment of glioma is the delivery of specific therapeutic agent(s) to tumor areas such as oncolytic viruses, aptly termed, oncolytic virotherapy [
1‐
4].
Among the oncolytic viruses, human adenovirus serotype 5 (Ad5) is the most commonly used due to its well-studied biology and flexibility for genetic modifications [
2,
3]. By taking advantage of these characteristics, a conditionally replicating oncolytic adenovirus, CRAd-S-pk7, has been generated [
5]. This viral vector infects cells by binding to anionic cell surface proteins through seven lysine residues (pk7) on the adenoviral fiber [
6,
7] and subsequently initiates replication by way of
E1 gene expression under the control of the tumor-specific promoter survivin (S) [
5]. This glioma selective oncolytic agent, CRAd-S-pk7, was shown to efficiently lyse glioma cells while sparing non-neoplastic cells [
8,
9]. However, the efficacy of CRAd-S-pk7 in the clinical setting can be nullified via both immune surveillance and pre-existing neutralizing antibodies against Ad5 [
8,
9]. Moreover, the biodistribution of oncolytic viruses (OVs) may be limited to the outer borders of a solid tumor mass with robust vascular supply, which often results in inconsistent viral penetration to the center of the tumor mass [
8‐
10].
To surmount these problems, we employed a cell carrier approach with neural stem cells (NSCs), which were shown to have a natural tropism toward tumor cells and the capability to penetrate the center of the tumor mass [
8‐
11]. In our previous studies, we showed that NSCs loaded with CRAd-S-pk7 could travel toward the tumor cells while simultaneously protecting CRAd-S-pk7 from immune surveillance as well as acting as host cells for CRAd-S-pk7 progeny production [
8‐
12]. Furthermore, NSCs loaded with CRAd-S-pk7 could be delivered into the center of the tumor mass and lyse tumor cells more efficiently than virus administration alone, which accounted for a significant survival benefit in xenograft murine models of GBM [
1,
8‐
12].
Given the fact that the murine model is not permissive to replication of human adenoviruses [
13‐
15], we herein investigated the biodistribution and toxicology of CRAd-S-pk7 after intracerebral administration of the therapeutic viral agent into Syrian Hamsters. This study was conducted under FDA guidelines, using Good Laboratory Practice (GLP) throughout, and will be used in filing of the final Investigational New Drug (IND). This Syrian Hamster model is permissive for human adenovirus replication and is fully immune competent [
9,
15], so the replication profiles and immune responses in different organs after the administration of this vehicle can recapitulate the replication kinetics and the toxicity of CRAd-S-pk7 in the human clinical setting. For this analysis, we administered single intracerebral doses of 2.5 × 10
7, 2.5 × 10
8, and 2.5 × 10
9 viral particles (vp) per animal. These doses were established based on our preclinical analysis using 50 vp/cell of CRAd-S-pk7 per 5 × 10
5 NSCs (2.5 × 10
7 vp/hamster in total) [
9].
In the immunological analysis, antibody response against E1 of the vector was observed in a dose-dependent manner as expected. Pathologically, minimal and transient reactive inflammatory changes were observed at all dose levels, which were not toxicologically significant. Furthermore, the viral vector was predominantly localized to the brain, with low levels of vector DNA observed in other tissues. Presence of vector in tissues was not correlated with any microscopic changes or vector-mediated toxicity. In summary, we demonstrated that the agreeable safety profile of CRAd-S-pk7 justifies proceeding with a planned Phase I clinical trial, utilizing this agent loaded onto FDA-approved NSCs for patients with GBM.
Methods
Animals
Syrian hamsters (Haslett, MI, USA) (n = 250) were randomized by body weight into one of four groups (groups 2–5) comprised of 120 male and 120 female hamsters distributed in equal numbers into core and satellite groups (60 male and 60 female hamsters each) (Table
1). The core groups (15 animals/sex/group) were used for toxicology evaluations including hematology and clinical chemistry, and the satellite groups (15 animals/sex/group) were used for assessment of the biodistribution and immunogenicity of the test article (CRAd-S-pk7) as well as the effect on coagulation parameters. Of note, 5 male and 5 female hamsters were arbitrarily placed into Group 1. They remained untreated and were used on day 1 for collection of baseline immunogenicity samples, after which they were removed from the study without further evaluation. All hamsters were approximately 5–6 weeks old on arrival at Southern Research (Birmingham, AL, USA) and the study was approved by the Institutional Animal Care and Use Committee. The hamsters in groups 2–5 were given an intracerebral injection of vehicle [GST150 Buffer (20 mM Tris, 150 mM NaCl, 2.5 % (w/v) glycerol, pH 8.0)] or CRAd-S-pk7 at 2.5 × 10
7, 2.5 × 10
8, or 2.5 × 10
9 viral particles (vp)/animal, respectively, at a fixed volume on day 1. Blood samples were drawn on days 6, 34, or 62 prior to necropsy from 5 male and 5 female hamsters from each of the core groups for the assessment of hematology and clinical chemistry parameters/analytes, and from another five male and five female hamsters in the satellite groups for the assessment of coagulation parameters.
Table 1
Assignment of hamsters to study groups
1 | Untreated | 0 | – | – | – | 5 M/5 F | – | – | – |
2 | Vehicle | 0 | 5 M/5 F | 5 M/5 F | 5 M/5 F | – | 5 M/5 F | 5 M/5 F | 5 M/5 F |
3 | Vector low | 2.5 × 107
| 5 M/5 F | 5 M/5 F | 5 M/5 F | – | 5 M/5 F | 5 M/5 F | 5 M/5 F |
4 | Vector mid | 2.5 × 108
| 5 M/5 F | 5 M/5 F | 5 M/5 F | – | 5 M/5 F | 5 M/5 F | 5 M/5 F |
5 | Vector high | 2.5 × 109
| 5 M/5 F | 5 M/5 F | 5 M/5 F | – | 5 M/5 F | 5 M/5 F | 5 M/5 F |
Clinical observations
All animals were observed at least twice daily during the pre-study and study periods for signs of mortality and moribundity.
Body weights and food consumption
All animals were weighed on day 1 and randomly assigned to each group. During the test periods, each animal was weighed weekly. Quantitative food consumption was measured weekly beginning on day 1 for each animal in the core group. Values were reported as an average consumption (grams/animal/day).
Tissue preservation and processing
Following injection of viral particles (2.5 × 107, 2.5 × 108, and 2.5 × 109, respectively), five male and female hamsters per group were sacrificed at 6, 34, and 62 days. Whole brains, as well as other organs, were collected for tissue bio-distribution analysis. Tissues (bone marrow (bilateral femurs; flushed), gonads, heart, incision site (skin, subcutis, muscle), kidneys, liver, lungs, mesenteric lymph nodes, and spleen) were snap frozen in liquid nitrogen and stored at −70 °C until analysis.
Viral DNA and transcript detection in the brain
Five male and five female animals per group were sacrificed at 6, 34, and 62 days. Whole brains were collected for detection and quantification of viral DNA at each time point. Tissues were snap frozen in liquid nitrogen and stored at −70 °C, or colder, until analysis. Brain tissue was analyzed for viral DNA via quantitative PCR (qPCR), using an assay previously evaluated and validated at Southern Research Institute. Tissue was homogenized and DNA was extracted using the Qiagen QIA amp 96 DNA QIAcube HT Kit (Qiagen, Venlo, The Netherlands). Samples were analyzed in triplicates, with one replicate spiked with 100 copies of adenovirus DNA to evaluate for possible inhibitors. The Qiagen QuantiTect Multiplex Master Mix was used for all qPCR reactions (Qiagen, Venlo, The Netherlands). Primers, probe, and control DNA were added to each reaction to serve as an internal amplification control and to rule out inhibition. DNA concentration was measured by spectrophotometer. The probes and primers used were validated specifically for human adenovirus serotype five hexon gene.
All brains were analyzed on days 6 and 34. Only those positive for vector DNA on day 34 were analyzed on day 62. PCR analysis was conducted using Applied Biosystems 7900HT Fast Real-Time PCR System (Applied Biosystems, Foster City, CA, USA) with software SDS 2.2.2. Settings used were: activation at 95 °C for 15 min, 45 cycles of denaturing at 95 °C for 10 s, and annealing/elongation at 60 °C for 45 s.
Systemic biodistribution of virally encoded transcripts after CRAd-S-pk7 intracerebral injection
On the planned day of necropsy, five male and five female animals in the satellite groups 2–5 were anesthetized and blood samples (~0.5 mL) were collected from the retro-orbital sinus into anticoagulant tubes (EDTA). Samples were mixed by gentle inversion, then snap frozen on dry ice and stored at −70 °C or below until analysis. The animals were then euthanized and the following whole tissues were collected: bone marrow, femur (both; flushed), brain, gonads (ovary/testis), heart, incision site (skin, subcutis, muscle), kidney, liver, lungs, lymph nodes, mesentery, and spleen. Tissue samples were snap frozen in liquid nitrogen and stored at −70 °C in the same fashion as blood samples. Blood and tissues collected were assayed for the presence of vector DNA using a qPCR protocol previously evaluated and validated at Southern Research Institute. Tissue was homogenized and DNA was extracted using the Qiagen QIA amp 96 DNA QIAcube HT Kit (Qiagen, Venlo, The Netherlands). Samples were analyzed in triplicates, with one replicate spiked with 100 copies of adenovirus DNA to evaluate for possible inhibitors. The Qiagen QuantiTect Multiplex Master Mix was used for all qPCR reactions (Qiagen, Venlo, The Netherlands). Primers, probe and control DNA were added to each reaction to serve as an internal amplification control and to rule out inhibition. DNA concentration was measured by spectrophotometer. The probes and primers used were validated specifically for human adenovirus serotype five hexon gene.
Since brain was the only tissue positive for vector DNA at day 34, blood and other tissue samples were not assayed at day 62. PCR analysis was conducted using Applied Biosystems 7900HT Fast Real-Time PCR System (Applied Biosystems, Foster City, CA) with software SDS 2.2.2. Settings used were: activation at 95 °C for 15 min, 45 cycles of denaturing at 95 °C for 10 s and annealing/elongation at 60 °C for 45 s.
Blood collection
On days 6, 34, and 62 post-injection, blood samples were collected prior the euthanasia for hematology and clinical chemistry evaluation from each hamster scheduled for necropsy. On these days blood was also collected from each hamster in satellite groups for evaluation of coagulation prior to euthanasia. Blood was collected from each hamster from the retro-orbital sinus using tubes containing either EDTA as an anticoagulant for hematology, or no coagulant for clinical chemistry, or sodium citrate as an anticoagulant for coagulation evaluation. Tubes were mixed by gentle inversion upon collection. Clinical pathology samples were analyzed on the same day that the samples were obtained. The following hematology parameters were assayed: total leukocyte count, erythrocyte count, hemoglobin, hematocrit, mean corpuscular volume, mean corpuscular hemoglobin, mean corpuscular hemoglobin concentration, reticulocyte count, platelet count, differential leukocyte counts, RBC morphology, and nucleated red blood cell count. The following clinical chemistry parameters were assayed: urea nitrogen, aspartate aminotransferase, alanine aminotransferase, alkaline phosphatase, glucose, creatinine, creatine kinase, total protein, albumin, globulin (calculated), albumin/globulin ratio (calculated), sodium, potassium, chloride, cholesterol, total bilirubin, calcium, phosphorus, blood urea nitrogen (BUN)/creatinine ratio (calculated). For coagulation analysis, prothrombin time and fibrinogen were assayed.
Immunogenicity
Control blood samples for immunogenicity analysis (~0.5 mL) were collected on day 1 from five animals per sex in the untreated group. Terminal blood samples were collected from satellite group animals on days 6, 34, and 62 immediately after blood collection for biodistribution. Samples were collected into tubes without anticoagulant. Serum was frozen and stored at −70 °C until analysis. One aliquot of each sample was assayed for antibodies against the vector using an ELISA assay developed and validated for GLP use at Southern Research.
Histopathology and toxicity
In the pathologic evaluation of CRAd-S-pk7 when administered as a single intracerebral dose in hamsters, viral toxicity was assessed through both gross and microscopic evaluations. Half of the animals were used for tissue collection for histopathology and the remaining satellite animals were used for biodistribution studies. The doses of vectors administered to groups 2 through 5, respectively, were: 0 (vehicle), 2.5 × 107 (vector, low), 2.5 × 108 (vector, mid), or 2.5 × 109 (vector, high). Animals were then euthanized on days 6, 34, and 62 and examined post-mortem.
For each post-mortem pathologic examination at the respective time points, all tissues were examined both grossly and microscopically. Hematoxylin and eosin (H&E) staining was performed on formalin-fixed paraffin-embedded (FFPE) tissue for histopathological evaluation. In addition to the standard tissue evaluation, target organs (brain and cervical spinal cord) were evaluated on day 34 and any gross lesions were documented, examined, and histologically graded on day 62. Lesions were graded according to their degree of involvement.
Microscopic lesions observed in this study were graded using a numerical scoring system in which 1 = minimal, 2 = mild, 3 = moderate, and 4 = marked. In general, lesions that affected less than 10 % of the tissue were considered as minimal, lesions that affected 11–50 % of tissue were considered as mild, lesions that affected 51–75 % of tissue were considered to as moderate, and lesions that affected greater than 75 % of tissue were considered as marked. All graded pathological evaluation was blinded to treatment.
Statistical analyses
For CRAd-S-pk7 dose effect analyses, cytotoxicity and clinical pathology data among four groups (groups 2, 3, 4, and 5) were compared. When the equality of variances among the groups assumption holds, One-way analysis of variance (ANOVA) with post hoc Bonferroni correction for multiple comparison were used. When the variances among the groups were not equal, Kruskal–Wallis test with post hoc Mann–Whitney U tests with the Bonferroni correction were performed. Linear regression models were conducted to assess the effect of the four groups on viral copies with log10 transformed. Generalized Estimating Equation approach was used to estimate the parameters in the regression models taking into account the longitudinal structure of the data. For the hematology, clinical chemistry, and coagulation parameters/analytes, the individual’s percent change (%Δ) on each day of analysis relative to the vehicle group’s (i.e., group 2) mean for the time point by sex was calculated with the following formula: %Δ in parameter/analyte for individual on day X = [(individual’s day X value − Y)/Y] × 100, where X = days 6, 34, or 62 and Y = vehicle group mean value for day X by sex and time point. Statistical analyses were performed with the statistical analysis system (SAS) software Version 9.4 (SAS Institute Inc., Cary, NC, USA).
Discussion
We have previously demonstrated the effectiveness of CRAd-S-pk7 as a single or combined therapy against high-grade gliomas [
4,
11,
17‐
19]. As compared with other oncolytic viruses, CRAd-S-pk7 showed the highest level of viral replication and tumor oncolysis in GBM cell lines, with minimal toxicity to normal brain tissue [
4,
11,
18]. Intracranial injection of CRAd-S-pk7 alone reduced tumor growth by 300 and 67 % of treated animals were long-term survivors [
4,
11,
18]. Synergistic effect of CRAd-S-pk7 with the current standard of care for anti-glioma therapy, radiotherapy and temozolomide, was also positively validated in animal models [
17,
19]. Taken together, these findings provide the rationale for further application of this tumor-specific oncolytic virus in a phase 1 clinical trial in patients with recurrent high-grade glioma. In preparation for this, herein we evaluated the biodistribution, toxicology, and anti-viral immune response of CRAd-S-pk7 in Syrian Hamsters, an immunocompetent model that is permissive for the replication of human adenoviruses [
11]. We observed that following intracranial administration of vehicle or low, mid, or high doses of the test article formulation of CRAd-S-pk7 (2.5 × 10
7, 2.5 × 10
8, or 2.5 × 10
9 viral particles/animal, respectively), viral particles were largely confined to the brain. Low levels of vector DNA were detected in other tissues in a few animals suggesting that systemic circulation of the virus can occur. Microscopic changes and virus-related toxicity were considered minor and CRAd-S-pk7 intracerebral injection was not associated with lethality. In addition, CRAd-S-pk7 was able to elicit measurable dose-dependent IgG immune response by day 34 after intracranial injection, which was considered to be marginal in the lowest dose group. In sum, the above results provide safety and toxicology data justifying the clinical application of CRAd-S-pk7 in humans with recurrent malignant glioma.
Concerning virus biodistribution, 6 days after intracranial administration, viral genomic DNA was present in the brains of all animals at all time-points, with the exception of one animal. The levels of vector DNA in the brains were dose-dependent and decreased over time, with a maximum decrease observed between days 6 and 34 after intracranial injection. The persistence of virus DNA 62 days after local administration was consistent with low levels of inflammation observed at the viral distribution sites. The significant time-dependent decrease of vector DNA in the brains of the studied animals was consistent with previous reports showing that these conditionally replicative adenovirus vectors are tumor-specific and do not replicate in non-malignant brain tissues [
3,
4,
9,
11,
18,
20,
21]. Systemic dissemination of intracranially administered CRAd-S-Pk7 was assessed in immunocompetent hamsters by qPCR. Blood was collected and various organs were harvested from animals sacrificed 6, 34, and 62 days after virus injection. On day 6, vector DNA was present at low levels at the incision site of many animals as well as in the blood and other tissues of a few animals in the mid and high-dose groups. The presence of vector DNA in these non-brain/incision site tissues indicated that for a few animals the vector was able to enter the systemic circulation. However, there were no microscopic lesions in these tissues from core group animals on day 6, suggesting that the presence of vector DNA in extraneural tissues was of no toxicological significance. Moreover, vector DNA was not detected in tissues outside the brain after day 6. These results are consistent with previously published data and confirm the safety and limited biodistribution of CRAd-S-pk7 after intracranial administration [
11].
The toxicological profile of CRAd-S-pk7 after intracranial administration was assessed via local and peripheral inflammatory response, measured mainly by immune cell count, fibrinogen and albumin levels, and macro- and microscopic pathological changes in harvested tissues at the above described time points. At all dose levels there was an increase in total leukocyte, neutrophil, and/or monocyte counts and fibrinogen levels, and a significant decrease in mean albumin level for male hamsters in the high dose group on day 6. We also observed microscopic inflammatory lesions in the brain, spinal cord, and meninges for hamsters of both sexes at all three time points (days 6, 34, and 62). However, these observed changes in clinical pathology parameters were for the most part present at relatively low levels. Moreover, there were no clinically significant test article-related inter-group differences for male hamsters on days 34 or 62, suggesting recovery from the adverse effects seen on day 6. There were no apparent test article-related inter-group differences for female hamsters at any time point. Therefore, the observed changes in clinical pathology parameters were considered to be of little toxicological significance other than as possible indicators of the underlying inflammation occurring in the nervous system. Considering additional toxicological parameters, we did not observe any changes in body weight between the studied groups. Changes in hematology, clinical chemistry, and coagulation parameters were minor and transient, and were consistent with the inflammatory changes that were observed microscopically. These changes were considered of little toxicological significance.
In terms of immune reactions, CRAd-S-pk7 was able to elicit a measurable IgG immune response at all dose levels by day 34. The magnitude of this response was related to the dose, with the response in the low dose group considered to be marginal. Because the levels of anti-CRAd-S-pk7 antibodies were higher in the high dose group than in the mid dose on both days 34 and 62, it was not possible from the data in this study to say whether a maximal response had been reached at the high dose. Based on previous reports and the above-described data, it is expected that CRAd-S-pk7 may induce anticancer immunity through viral replication followed by tumor cell lysis in glioma-bearing subjects [
22‐
24]. However, modulating the immune response can be tricky and a fine balance between antiviral and antitumor immune responses must be achieved so that the virus will be able to first infect and lyse the tumor cells and then induce an antitumor immune response against neighboring malignant tissues.
In summary, administration of CRAd-S-pk7 at different doses produced a positive antibody response against the virus. The vector remained localized primarily in the brain and to some degree in the tissues at the incision site, although the presence of low levels of virus DNA in other tissues indicated that the vector was able to enter the systemic circulation in a few animals. However, the presence of virus in those tissues from the satellite group animals was not correlated with any microscopic changes in the same tissues of core group animals that would suggest toxicity of the vector. Test article-related microscopic changes were observed that were consistent with viral disease affecting the central nervous system; these changes appeared to decrease in incidence and severity over time, indicating that recovery was in progress. Because of the microscopic changes seen in the 2.5 × 107 vp/animal group, a no observed adverse effect level (NOAEL) could not be identified for the CRAd-S-pk7 vector under the conditions of this study. These results provide feasibility data related to the biodistribution, toxicology, and immune response of CRAd-S-pk7 in preparation for a phase 1 clinical trial.