New pancreatic carcinoma model for studying oncolytic adenoviruses in the permissive Syrian hamster

Abstract

Syrian hamster is a practical animal model for studying the systemic effects of oncolytic vectors derived from adenovirus serotype 5 (Ad5). Ad5 replicates well in Syrian hamster tissues, and Syrian hamster cell lines are available that are known to support Ad5 replication. In this study, we established four new Syrian hamster cell lines from transplantable pancreatic, renal, hepatic and lung tumors. The pancreatic cell line (SHPC6) and the renal cell line were highly permissive for Ad5 replication. The SHPC6 cell line formed disseminated intraperitoneal tumors when cells were injected into the peritoneal cavity. INGN 007, an oncolytic Ad5-based vector, completely reversed the growth of disseminated intraperitoneal SHPC6 tumor nodules following intraperitoneal injection of the vector, leading to 100% survival of the treated animals. SHPC6 cells also formed subcutaneous tumors, whose growth was suppressed by INGN 007 following intratumoral injection. INGN 007 replicated in both the intraperitoneal and subcutaneous SHPC6 tumors. Following intraperitoneal injection, INGN 007 did not replicate in the livers of hamsters with intraperitoneal SHPC6 tumors, and was not hepatotoxic. These studies suggest that the SHPC6 cell line may be useful as a model for disseminated pancreatic cancer, and that INGN 007 may be a safe and effective vector to treat these tumors.

Introduction

Cancer is the second leading cause of death in the United States, and more advanced treatments are imperative in addressing this important public health issue. Gene therapy is one area in cancer research where large advances have been made. Oncolytic vectors, human adenovirus (Ad) type 5 (Ad5)-based vectors in particular, have been evaluated as promising cancer therapies. Several clinical trials have been conducted using replication-selective oncolytic Ads. These vectors have shown some efficacy, especially in combination with other therapies, and no dose-limiting toxicities were reported.^1–5

We have developed a family of oncolytic Ad vectors with increased antitumor potency due to overexpression of the adenovirus death protein (ADP).^6–10 ADP is an Ad5-encoded protein expressed at the final stage of infection that facilitates viral egress and cell-to-cell spread of infection.^11,12 One of our vectors, INGN 007^{8–10,13–18}, is about to enter phase I clinical trials. INGN 007 overexpresses ADP, has the immunomodulatory E3 genes deleted, but expresses all other Ad5 proteins at wild-type levels.⁸

Until recently, the study of oncolytic Ads as cancer therapeutics had been hindered by the lack of relevant animal models that are immunocompetent, permissive for Ad replication and for which tumor cell lines are available. Nude mice bearing human xenograft tumors have traditionally been the model for Ad vectors used as cancer therapies.^19–22 However, due to this model’s poor permissiveness for Ad replication, it cannot be used to evaluate the repercussions of systemic viral replication. ^{13,17,23–25} Further, as nude mice are immunocompromised, the model cannot address the immune response to these vectors. Ad5 replicates well in cotton rat tumors¹⁶ and tissues;^25,26 but cotton rats are costly, difficult to handle, and there are only a few cancer cell lines available for use in this model. Some large animal models that support Ad replication have been described,^27,28 but they are of limited use for the same reasons as the cotton rat. Furthermore, the housing requirements for these models make conducting large studies extremely difficult in most research facilities.

The Syrian hamster is one of the most useful and practical animal models for studying the therapeutic and systemic effects of human Ad5 and Ad5-based viral vectors.^{13,14,17,29–34} Ad5 replicates well in Syrian hamster tissues, and permissive cancer cell lines are available to be used for tumor-suppression studies.^14,29 Also, the model has been used to test the efficacy of antiviral drugs in preventing toxicity and mortality after lethal challenge of human Ad5.^32,35

Although the Syrian hamster is becoming a well-established model for studying Ad5 and Ad5-based vectors, more tumor cell lines are needed to maximize the relevance of this model in evaluating Ad therapies for cancer. To address this issue, we obtained four types of Syrian hamster transplantable tumors from the National Cancer Institute and used them to establish permanent tumor cell lines. It is advantageous to use cell lines for evaluating Ad vectors as opposed to the transplantable tumors themselves, because in vitro work can easily be done to study aspects of viral infection in each cell line that would be very difficult to address in vivo. The four cell lines, Syrian hamster pancreatic carcinoma (SHPC6), Syrian hamster renal carcinoma (SHRC), Syrian hamster lung carcinoma (SHLC) and Syrian hamster hepatic carcinoma (SHHC17) were evaluated for their ability to support Ad5 replication. The SHPC6 cell line was very permissive for Ad5 replication, and it was chosen for developing tumor models in hamsters. This cell line forms subcutaneous tumors as well as intraperitoneal (i.p.) pancreatic tumor nodules. Our oncolytic vector, INGN 007, completely reversed the growth of these nodules following i.p. injection of the vector. INGN 007 replicated in both the subcutaneous and i.p. SHPC6 tumors. Further, following i.p. injection, INGN 007 did not replicate in the livers of hamsters with i.p. SHPC6 tumors, and was not hepatotoxic. This suggests that Ad5-based oncolytic vectors may not be hepatotoxic when treatment is given via the i.p. route.

Materials and methods

Passaging transplantable tumors

Transplantable tumors 4671 (pancreatic duct carcinoma), 1382J (hepatic carcinoma), 8721R (renal carcinoma) and 11348P (pulmonary squamous cell carcinoma) were obtained from the Natural Products and Tumor Repositories at the National Cancer Institute (Frederick, MD). The tumors were implanted subcutaneously into recipient hamsters using a tumor implantation needle (Popper and Sons, New York, NY). When the tumors reached approximately 2ml size, they were surgically removed under aseptic conditions. The tumor tissue was placed into a tissue culture dish containing RPMI-1640 medium (without serum) and nontumor tissue and necrotic tumor tissue was removed. The viable tumor mass was cut into small pieces and washed in RPMI-1640 medium. The tumor pieces were either implanted into hamsters as described above, or frozen in RPMI-1640, 10% DMSO for storage at −80 °C.

Isolation of tumor cell lines

Tumor cell lines were isolated using the Panomics Cancer Cell Isolation Kit (Panomics Inc. Fremont, CA) according to the manufacturer’s instructions. Briefly, the tumor tissue was removed from the animals and processed as for passaging, then further dissociated by proteolytic digestion and strained to remove larger chunks and connective tissue. The resulting cell dispersion was fractionated according to sedimentation properties to get rid of cell debris and cells of blood origin. The fraction enriched in tumor cell aggregates was originally cultured in RPMI-1640 containing 10% fetal bovine serum (FBS) and 1% penicillin and streptomycin.

Cell lines and adenoviruses

All Syrian hamster and human cell lines were maintained in Dulbecco’s modified Eagle’s medium (DMEM; Sigma, St Louis, MO) containing 10% FBS (Thermo Scientific HyClone, Logan, UT), except the SHPC6 cell line. The SHPC6 cell line was maintained in DMEM containing 15% FBS and 1mM nonessential amino acids and 1mM sodium pyruvate (Invitrogen, Carlsbad, CA). HEK293 cells were purchased from Microbix (Toronto, Ontario), and cultured in DMEM containing 10% FBS. HaK and A549 cells were obtained from the American Type Culture Collection (Manassas, VA).

INGN 007 is an Ad5-based oncolytic Ad vector with an E3 region deletion replaced with the adp gene. Ad CMVpA is a replication-defective Ad5 vector with the E1 and parts of the E3 regions deleted. The INGN 007 and AdCMVpA viruses were obtained from Introgen Therapeutics Inc. (Houston, TX). Ad5 was grown and purified as previously described.³⁶ The Ad5, AdCMVpA and INGN 007 viruses were titered by TCID₅₀ assay on HEK293 cells.

Animal care

Four- to five-week-old female Golden Syrian hamsters (Harlan, Indianapolis, IN) were housed in polycarbonate microisolator caging (Allentown Inc., Allentown, NJ) and provided standard rodent care by the Saint Louis University Department of Comparative Medicine. All animal studies were approved by the Institutional Animal Care Committee and were conducted in the Department of Comparative Medicine.

Subcutaneous tumors

Hamsters were injected subcutaneously with 1 × 10⁷ cells of each cell line in 50–100 μl of serum-free DMEM into the right flank. When tumors reached approximately 300 μl in size, treatment commenced. For intratumoral injections, each injection contained 100 μl of vehicle (20mM Tris HCl, pH 8.0) or 1 × 10¹⁰ TCID₅₀ of INGN 007 suspended in vehicle and was injected in multiple locations throughout the tumor using a fan-like injection pattern. Quantification of tumor growth was measured with electronic calipers and data were recorded with an in-house-developed tumor measurement and analysis software.

Intravenous disseminated SHPC6 lung tumors

The animals were injected into the jugular vein with 2 × 10⁶ cells in 200 μl of serum-free DMEM as previously published.³⁷

Intraperitoneal disseminated SHPC6 tumors

To establish tumors, 1 × 10⁷ cells in 500 μl of serum-free DMEM were injected into the lower right peritoneal cavity of hamsters. Tumors were formed by 4 days after injection.

Intraperitoneal injection of adenoviruses

For treatment of tumors, each injection contained 500 μl of vehicle or 1 × 10¹⁰ TCID₅₀ of INGN 007 suspended in vehicle. The viruses were injected in the same manner as the cells, either on the same day as the cell injection or 4 days after cell injection (for experiments with established i.p. tumors).

Hepatotoxicity evaluation

Six hamsters, three for each group, were injected i.p. with either vehicle or 1 × 10¹⁰ TCID₅₀ of INGN 007 for 3 consecutive days as described above. Blood was collected from hamsters from the retro-orbital sinus, and the serum was analyzed for transaminase levels (Advanced Veterinary Laboratory, St Louis, MO).

Histological preparation of tissues

Tissues fixed in 10% neutral buffered formalin were processed, embedded in paraffin, and sectioned onto glass slides. The tissue sections were then stained with hematoxylin and eosin.

Single step growth curve

Each cell line was plated into seven 35mm tissue culture dishes (TPP, Trasadingen, Switzerland) at 60–70% confluency. One dish from each cell line was counted for calculation purposes. The remaining six dishes were infected with Ad5 at 100 TCID₅₀ per cell in 0.5 ml of serum-free DMEM and incubated at 37 °C for 1 h. After incubation, the dishes were rinsed thrice with serum-free DMEM, and DMEM containing 5% FBS was added back. A dish from each cell line was frozen at 0, 1, 2, 4, 7 and 14 days postinfection (p.i.) The cells and media from the frozen dishes were collected into 5ml snap-cap culture tubes and freeze thawed 3 times. The cell lysates were clarified by centrifugation. TCID₅₀ assays to determine the virus titer from each cell line were performed with the supernatant.

TCID₅₀ assays of cell and tissue lysates

HEK293 cells were plated at approximately 1.3 × 10⁴ cells per well one day before infection in 96-well tissue culture plates (TPP). The cells were infected with serial dilutions of sample in serum-free DMEM and incubated for 1 h. In the case of tissue lysates, positive and negative controls were included in each plate.³⁸ Medium containing FBS was added back to the plate to a final serum concentration of 5%. The assay was read at 14 days p.i.

Cytopathic effect assay

HaK and SHPC6 cells were plated to be 60–70% confluent on the day of infection in 24-well plates (TPP). For the SHPC6 cell line, a 24-well Corning Cell Bind plate (Corning Life Sciences, Lowell, MA) was used. On the day of infection, one well from each plate of cells was counted. Wells were infected with serial dilutions of either INGN 007 or Ad5 from 100 to 0.01 TCID₅₀ per cell and observed daily for cytopathic effect (CPE). The cells were fixed with 3.7% formaldehyde (Fisher Scientific, Pittsburgh, PA), and stained with 1% crystal violet solution (Sigma).

Assessing viral yield in tissues

Tumor, liver and/or ascites fluid was collected from hamsters into sterile 1.7 ml screw cap tubes (Sarstedt, Newton, NC) and immediately frozen in dry ice. 300 μl of sterile phosphate-buffered saline without calcium and magnesium (Sigma) was added to liver and tumor samples. The samples were then homogenized in a bead-beater type homogenizer (TissueLyzer; Qiagen, Valencia, CA), freeze thawed thrice, and sonicated for 7 min.³⁸ A TCID₅₀ assay was performed to quantify virus content.

Statistics

Log rank test was performed to establish the significance of the differences between survival data. Differences with a P value smaller than 0.05 were considered significant.

Results

The new Syrian hamster tumor cell lines are easily grown in tissue culture

We obtained samples of spontaneous Syrian hamster pancreatic, renal, pulmonary and liver transplantable tumors from the National Cancer Institute, implanted them subcutaneously in Syrian hamsters and generated cell lines from the resulting tumors. All four cell lines grew well and were easily maintained as adherent tissue culture, with cell doubling times of 24.6, 19.5, 18.9 and 22.5 h for the SHPC6, SHRC, SHLC and SHHC17 cells, respectively (not shown).

All four cell lines are tumorigenic in the Syrian hamster, and the tumors are histologically similar to the original transplantable tumor

Each cell line was tested to ensure that it remained tumorigenic in hamsters. After injecting the cells subcutaneously, the tumors were palpable within several days and most reached 10mm diameter in a week. Once tumors reached approximately 1ml or more in size, a necropsy was performed. On gross examination, the only animals that had metastatic sites in the lungs were those with subcutaneous SHPC6 tumors. For all four cell lines, tumors generated from both the transplantable tumors and the tumor cell lines were examined histopathologically. Overall, tumors from the cell lines were histologically similar to the original transplantable tumor. The transplantable pancreatic tumors were solid to trabecular carcinomas (Figure 1a); the tumors formed with the SHPC6 cells were tubulopapillary carcinomas with occasional trabeculae (Figure 1b), which is compatible with their ductal origin.³⁹ The SHLC cells formed a more solid carcinoma (Figure 1d) compared to the original transplantable pulmonary tumors (Figure 1c), which were solid to trabecular histologically. The transplantable hepatic tumors (Figure 1e) and the SHHC17 tumors (Figure 1f) were both solid carcinomas, except that the SHHC17 tumors were often divided into small packets by fine fibrovascular stroma. The transplantable renal tumors (Figure 1g) and the SHRC tumors (Figure 1h) were trabecular carcinomas and contained extensive necrotic and hemorrhagic areas.

Figure 1.

Subcutaneous tumors formed from the established tumor cell lines are morphologically similar to the original transplantable tumors. Tumor sections were stained with hematoxylin-eosin and photomicrographs were taken. Size bars are 500 μm in the larger image and 50 μm in the inset. Labels: T, trabecular morphology; N, necrosis; H, hemorrhage; P, packets; TP, tubulopapillary morphology.

The SHPC6 and SHRC cell lines are permissive for Ad5 replication

The ability of Ad5 to replicate in and produce progeny in each cell line was assessed by a single-step growth curve and compared to HaK cells, a hamster kidney cell line previously shown to support Ad5 replication,¹⁴ and the highly permissive human A549 lung adenocarcinoma cell line. The final Ad5 burst sizes in the SHPC6 and SHRC cell lines were comparable to the A549 and HaK cell lines (Figure 2a). At 4 days p.i., the burst size in the SHPC6 and SHRC cells was approximately 2 logs less than that in A549 cells and approximately fivefold more than that in HaK cells. By 7 days, the burst size in both SHRC and SHPC6 cells had increased to within a log of the burst size in A549 cells, and was consistently greater than that in HaK cells. Replication in the SHRC cells was similar to that in HaK cells at 14 days p.i. Very little replication was evident in the SHLC and SHHC17 cell lines. From these data, we chose the SHPC6 cell line for further study.

Figure 2.

The SHRC and SHPC6 cell lines are permissive for adenovirus 5 (Ad5) replication. (a) The ability of each cell line to support the replication of Ad5 was established by measuring the burst-size of the virus in a single-step growth curve. (b) INGN 007 and Ad5 lyse SHPC6 cells more efficiently than the HaK cells. The SHPC6 plate was fixed and stained 6 days postinfection (p.i.), and the HaK plate was stained 8 days p.i.

Ad5 and INGN 007 lyse SHPC6 cells

The ability of the oncolytic Ad vector INGN 007 and Ad5 to lyse SHPC6 and HaK cells was assessed by infecting the cells at multiplicities of infection ranging from 0.01 to 100 TCID₅₀ per cell. With the SHPC6 cells, CPE was grossly evident at 1 TCID₅₀ per cell (Figure 2b), and microscopically evident in wells down to 0.1 TCID₅₀ per cell for both viruses at 6 days p.i. Wells infected with INGN 007 had visibly higher amounts of CPE compared to Ad5; this is because INGN 007 expresses more ADP than Ad5.⁸ Both viruses lysed SHPC6 cells more efficiently at 6 days p.i. than HaK cells at 8 days p.i. (Figure 2b).

SHPC6 cells generate disseminated pancreatic tumors when injected intravenously and intraperitoneally

As the subcutaneous SHPC6 tumors metastasized to the lung, we explored the ability of the SHPC6 cell line to be used as a disseminated tumor model. SHPC6 cells were injected intravenously (i.v.) into the jugular vein of three hamsters. Two hamsters were necropsied at 28 days after injection and one at 40 days after injection. In all three hamsters, tumor nodules were observed in the lungs. Histologically, these tumors (not shown) were very much like the subcutaneous SHPC6 tumors described previously. Tumor nodules were not found in any other organ.

In addition to this experiment, we injected SHPC6 cells into the peritoneal cavity of two hamsters. At 6 days after injection, necropsy was performed. In both animals, the abdominal cavity contained a large amount of serosanguinous ascites fluid, but more interestingly, multiple oval shaped nodules had localized and grown in the mesentery adjacent to the pancreas and spleen. Histology was performed on the nodules and adjacent tissues, as well as on an ascites fluid smear. The nodules were undifferentiated solid carcinomas (not shown). The ascites fluid contained large amounts of neoplastic cells and inflammatory cells. Mild to moderate peritonitis was also noted.

Intraperitoneal injection of INGN 007 prevents SHPC6 disseminated tumor formation

We tested whether INGN 007 could suppress the growth of SHPC6 tumors when both cells and virus are injected i.p. Vehicle or 3 × 10¹⁰ TCID₅₀ of INGN 007 were administered i.p. fractionated into three daily doses, beginning at 4 h after SHPC6 cell injection. At 7 days after cell injection, the study was terminated and the animals were necropsied.

Remarkably, no tumors or ascites were found in any of the INGN 007-treated animals (n = 12), although varying amounts of adhesions throughout the abdominal cavity were noted. On histological examination of tissues from the INGN 007-treated animals, no evidence of neoplasia was found in the liver, lung, pancreas, spleen, mesentery, kidney, diaphragm or abdominal lymph nodes.

In the vehicle-treated group, the abdomens of 10 animals contained ascites fluid ranging in volume from 1.5 to 17.5 ml (Figure 3a), and multiple tumor nodules were found in the splenic/pancreatic mesentery and on the diaphragm, liver and parietal peritoneum. Increased amounts of ascites fluid coincided with the extent of tumor formation; hamsters with small volumes of ascites fluid (less than 2 ml) only had tumors formed on the splenic/pancreatic mesentery, whereas animals with large volumes of ascites also had tumor nodules on the diaphragm, liver, peritoneum and floating in the ascites fluid. Two vehicle-treated animals did not have any tumors or evidence of ascites. Histology of the tumors in the splenic/pancreatic mesentery and the ascitic fluid from the vehicle-treated hamsters were as previously described. The nodules on the diaphragm were solid carcinomas with marked inflammation containing macrophages and lymphocytes. The small nodules found on the liver, peritoneum and free within the peritoneal cavity were aggregates of tumor cells. Although the tumor nodules and tumor cell aggregates were attached to these organs, histologically, these nodules were confined to the serosal surface of the organs and were not present within the organ parenchyma. In addition, no neoplastic lesions were found in the lung, pancreas, spleen, kidney or abdominal lymph nodes in the vehicle-treated animals. All abdominal organs examined in both groups had evidence of subacute, chronic, nonsuppurative inflammation affecting all of the serosal membrane surfaces.

Figure 3.

Intraperitoneal SHPC6 tumors were effectively treated with INGN 007. (a) The ability of SHPC6 cells to form tumors and ascites in the peritoneal cavity was eliminated when INGN 007 treatment was initiated at 4 h after cell injection. The volume of ascites indicates the progression of the neoplasia (n = 12 per group). (b) At 4 days after intraperitoneal injection, INGN 007 replicated in established intraperitoneal SHPC6 tumors, but not in the liver (n = 6 per group). The dashed line represents the limit of quantifiability. Symbols above this line represent values obtained from individual samples. Symbols below this line represent samples in which no virus was detected. (c) Treatment with INGN 007 eliminated intraperitoneal SHPC6 tumors in the majority of animals resulting in 100% survival (n = 13 per group).

Virus replication is evident in INGN 007-treated established intraperitoneal SHPC6 tumors, but not in tumors treated with a nonreplicating Ad5 control

We examined the replication of INGN 007 in the tumors and livers of 12 hamsters with established i.p. SHPC6 tumors when the vector was administered by i.p. injection. This experiment differs from the previous one in that INGN 007 was administered 4 days following the injection of the SHPC6 cells, a time when the disseminated tumors were well established. The hamsters were treated with one injection of either 1 × 10¹⁰ TCID₅₀ of INGN 007 or AdCMVpA, a replication-defective Ad5 vector with no transgene. Tumors, ascites fluid and livers were collected at 4 days after virus injection and the infectious virus content was quantified by TCID₅₀ assay. In the AdCMVpA group, four of six animals had tumors with average weight of 113mg in the splenic/pancreatic mesentery, and all tumor-positive animals had large volumes of ascites fluid. In the INGN 007 group, three of six animals had similar tumors, but they were smaller (average weight 52 mg), and only one animal had ascites.

Substantial virus yields were recovered from the tumors of the three INGN 007-treated hamsters that had tumors (4.7 × 10³ to 4.2 × 10⁵ TCID₅₀ per gram of tumor tissue; Figure 3b). The titer of INGN 007 in the ascites fluid from the one INGN 007-treated animal with ascites was 4.1 × 10⁵ TCID₅₀ per gram (not shown); this was also the animal with the highest tumor virus yield (Figure 3b). With the AdCMVpA-treated animals with ascites, we could detect infectious virus in the ascites fluid of only one of the four animals, but the amount was below the limit of quantification (not shown). No virus was recovered from the tumors in the AdCMVpA-treated group (Figure 3b).

As the liver is the primary target for Ad toxicity in hamsters,^13,14,32 we assayed the liver for infectious Ads as well. The livers of hamsters in both the AdCMVpA- and INGN 007-treated groups were negative (Figure 3b).

Treatment with INGN 007 results in 100% survival in hamsters with established intraperitoneal SHPC6 tumors

To assess the antitumor efficacy of INGN 007 in established i.p. SHPC6 tumors, we injected cells 4 days before treatment, and then administered either 1 × 10¹⁰ TCID₅₀ of INGN 007 or vehicle i.p. daily for 3 consecutive days. The animals were killed as they became moribund and a survival curve was generated. INGN 007 treatment resulted in a dramatically significant difference in survival rates (P = 0.0023; Figure 3c). None of the INGN 007-treated animals (n = 13) became moribund by 29 days after cell injection (Table 1). In the vehicle-treated group (n = 13), the median survival time was 16 days (Figure 3c); these animals had tumors and ascites fluid that were grossly and histologically identical to the vehicle-treated tumors described in previous studies (Figure 4a and c).

Table 1.

Summary of findings for established i.p. tumors treated with INGN 007

Group	Animals surviving at day 29	Animals developing tumors	Tumor size at killing (n)	Animals developing ascites	Tumor histology (n)
Vehicle	5/13	10/13	2–3 mm (8) 5–10 mm (2)	8/13	Carcinoma (10)
INGN 007	13/13	3/13	2–3 mm (3)	2/13	Carcinoma (2) Granulomaa (1)

^aNo viable tumor cells were found.

Figure 4.

The ability of INGN 007 to treat established SHPC6 tumors is grossly and histologically evident. (a) SHPC6 tumors formed in the mesenteric connective tissue between the spleen and pancreas. (b) SHPC6 tumors treated with INGN 007 were much smaller and appeared necrotic. Untreated tumors are compact, solid carcinomas when examined histologically (c), whereas a large area of necrosis is evident, and the overall structure has become granulomatous in the treated tumor (d). Labels: PI, pancreatic islets; PA, pancreatic acini; T, tumor; S, spleen; P, pancreas; N, necrosis. The arrows point to tumor nodules. Size bar = 500 μm in the larger image and 50 μm in the inset.

Surviving animals were killed at 29 days after cell injection. At necropsy, two of the five surviving vehicle-treated hamsters had multiple large tumors, ranging from 5 to 10mm in size, in and around the duodenal lobe of the pancreas and in the splenic/pancreatic mesentery. No tumors were found in the other three vehicle-treated hamsters. With the INGN 007-treated hamsters, 10 of the 13 hamsters did not have any tumors in the abdominal cavity; two had small, 2–3mm size tumors in the splenic/pancreatic mesentery and ascites and one had a single, small, tan to yellow nodule in the splenic/pancreatic mesentery (Figure 4b) and no ascites (Table 1). The appearance of this latter nodule was quite different from the tumor nodules described previously. Histologically, it was a granuloma consisting of macrophages, lymphocytes and plasma cells surrounding a large area of central necrosis; no viable tumor cells were definitively identified (Figure 4d). This histopathologic finding is consistent with effects of treatment observed in other cell lines with INGN 007.¹⁸ Two animals in the INGN 007-treated group had adhesions in multiple locations throughout the abdomen.

INGN 007 was not hepatotoxic after intraperitoneal injection

To ascertain the safety of INGN 007, we evaluated serum chemistry parameters of liver function after i.p. injection of the virus. Vehicle or 3 × 10¹⁰ TCID₅₀ of INGN 007, fractionated into three daily doses, was injected i.p. into hamsters. Liver damage was assessed by measuring serum aspartate aminotransferase (AST) and alanine aminotransferase (ALT) levels 4 and 7 days after injection. No elevation in AST or ALT levels was found for either group at either time point (not shown). Even though hepatotoxicity was not observed, the i.p. injection of INGN 007 in this study was not without consequence: Frequent adhesions had formed in the abdominal cavity in the INGN 007 group by 29 days after virus injection (not shown).

Intratumoral injection of INGN 007 increased the survival of animals with subcutaneous SHPC6 tumors

Although disseminated tumor models may better mimic the challenges of treating human pancreatic cancer, subcutaneous tumor models are routinely used in research to evaluate virus-based cancer therapeutics. To evaluate the antitumor efficacy of INGN 007 on subcutaneous SHPC6 tumors in hamsters, tumors averaging 332 μl were injected intratumorally with either 1 × 10¹⁰ TCID₅₀ of INGN 007 or vehicle for 6 days, and then the tumors were measured. When tumor volumes reached 5000 μl, the animals were killed and a survival curve was generated. The median survival time with the INGN 007 treatment group increased to 32.5 days from 21 days with the vehicle group (P < 0.01; Figure 5a). Importantly, the tumors completely regressed in three hamsters from the INGN 007 treatment group, whereas all tumors grew in the vehicle-treated group. These three animals from the INGN 007 treatment group were observed for 18 days after the vehicle and other virus-treated animals were killed, and no evidence of tumor recurrence, metastasis or other pathology was grossly evident at necropsy.

Figure 5.

INGN 007 replicated in established subcutaneous SHPC6 tumors after intratumoral injection and increased the survival time of tumor-bearing hamsters. (a) INGN 007 treatment of established subcutaneous SHPC6 tumors significantly increased animal survival (n = 10/group). (b) At 3 days after i.t. injection, higher amounts of infectious INGN 007 were recovered from tumors compared to AdCMVpA. A symbol below the dashed line represents an unquantifiable positive sample.

INGN 007 replicates in subcutaneous SHPC6 tumors after intratumoral injection

The replication of INGN 007 in subcutaneous SHPC6 tumors was assayed in hamsters after a single intratumoral injection of 1 × 10¹⁰ TCID₅₀ of either INGN 007 or AdCMVpA. Tumors were harvested at 3 days after injection, and infectious virus content was assessed by TCID₅₀ assay. Overall, the viral yield from the INGN 007-treated tumors was much higher than those treated with nonreplicating vector, indicating replication of INGN 007 (Figure 5b). This is in accordance with previously published data.^14,18

Discussion

We have successfully isolated four new hamster carcinoma cell lines from solid tumors. We identified the SHPC6 pancreatic carcinoma cell line as a highly permissive tumor model that can be used to study Ad5-based oncolytic vectors. Although the SHRC cell line was also highly permissive for Ad5 replication, it was not ideal for development as a tumor model to study oncolytic Ad vectors because of its aggressive nature. Animals were killed by 10 days after subcutaneous cell injection due to excessive tumor burden.

With the SHPC6 tumor models that we developed, we showed that INGN 007, an oncolytic Ad5-based vector, replicated in both i.p. and subcutaneous SHPC6 tumors. Intratumoral injection of INGN 007 into subcutaneous SHPC6 tumors extended the survival of the hamsters, as was shown previously with other hamster tumor cell lines.^14,18

The SHPC6 cell line is likely to be useful as a model of disseminated pancreatic cancer. When this cell line is injected i.p., the neoplastic progression is similar to end-stage human pancreatic cancer with malignant ascites, in that the ascites can develop with relatively low tumor burden, and that the presence of ascites is strongly associated with mortality.^40,41 The SHPC6 cell line also behaves comparably to other disseminated pancreatic cancer models in mice and hamsters.^42–44 Another interesting observation in our studies is that two hamsters developed large tumors in the duodenal lobe of the pancreas after i.p. cell injection. We speculate that this was the result of inadvertent introduction of cells directly into the pancreas. Given this, we plan to surgically implant SHPC6 cells into the pancreas and further explore using this cell line as an orthotopic tumor model.

Oncolytic Ads are being studied as treatment for pancreatic cancer.^45–47 The i.p. SHPC6 tumor model can be used as a preliminary evaluation method for these vectors. Not only is the SHPC6 cell line as permissive or more permissive for Ad5 replication than other hamster cell lines,^14,29 but studies can be conducted very quickly due to the rapid tumor formation observed with this cell line. Most untreated animals were moribund by 16 days after i.p. injection of cells.

Another interesting finding is the remarkable efficacy with which INGN 007 eliminated i.p. SHPC6 tumors. The treatment resulted in 100% survival, and 77% of these animals were disease free. This is at stark contrast to the 39% survival and 15% disease-free status of the untreated animals at the conclusion of the study. One possible explanation as to why INGN 007 performed better in the i.p. model than in the subcutaneous one is that in the former the vector may have better access to the multiple smaller tumor nodules than to all areas of the single, but heavily compartmentalized subcutaneous tumor. This result might indicate that the dispersed i.p. tumors are a preferable indication for treatment with oncolytic Ads.

Our results also suggest that the i.p. injection of replication-competent oncolytic Ads may be safe. No infectious virus was recovered from the livers following a single i.p. injection of 1 × 10¹⁰ TCID₅₀ of INGN 007 (Figure 3b), and no detectable elevation in serum liver enzymes was observed in any animals after i.p. injection of 1 × 10¹⁰ TCID₅₀ of INGN 007 daily for 3 consecutive days. In earlier studies, we injected INGN 007 or Ad5 i.v. rather than i.p. as in this study, and we observed replication in the liver as well as elevations in hepatic transaminases. In one study,³² in which we injected 1.6 × 10¹⁰ TCID₅₀ of Ad5, we found a mean of 2.4 × 10⁸ TCID₅₀/g of virus in the liver and a ~30-fold increase in serum AST at 3 days after injection. In another study,17 in which we injected 1.9 × 10¹² virus particles (equivalent to 3 × 10¹⁰ TCID₅₀) of INGN 007 or Ad5, we found about 10⁷ TCID₅₀/g of INGN 007 and Ad5 in the liver at 1 day after i.v. injection and, in half of the hamsters, between 2 × 10³ and 1 × 10⁶ TCID₅₀/g of INGN 007 and Ad5 at 6 days after injection. In a parallel study,¹³ using the same dose of i.v. INGN 007 and Ad5, we saw significant elevation in AST and ALT at 6 days after injection. Although these three i.v. studies used slightly more virus per injection (1.5- to 3-fold more virus) than in this study in which we injected 1 × 10¹⁰ TCID₅₀ of INGN 007 i.p., it appears that i.p. injection of INGN 007 is less toxic to the liver than i.v. injection. Our results in this study mirror human clinical experience: only mild toxicities were observed in human patients receiving i.p. injection of a replication-selective oncolytic Ad.⁴⁸

The Syrian hamster is already used extensively in pancreatic cancer research.^49,50 Our findings illustrate the utility of the i.p. SHPC6 tumor model in evaluating oncolytic Ad’s in the permissive, immunocompetent Syrian hamster. More work is needed to identify whether this cell line could be useful in evaluating other treatments for pancreatic cancer.