Tumor Vaccines and Biotechnology Branch, Division of Cellular and Gene Therapies, Center for Biologics Evaluation and Research, Food and Drug Administration, Bethesda, Maryland
Tumor Vaccines and Biotechnology Branch, Division of Cellular and Gene Therapies, Food and Drug Administration, Center for Biologics Evaluation and Research, NIH Building 29B, Room 2NN20, HFM-710, 29 Lincoln Dr., Bethesda, MD 20892
This article is a US Government work and, as such, is in the public domain in the United States of America.
Epithelial ovarian cancer demonstrates high mortality due to diagnosis at an advanced stage. In the search for a biomarker for early diagnosis and a target for therapy, the issue of whether interleukin-13 receptor (IL-13R), shown to be expressed on a variety of human cancers, is expressed in ovarian tumor samples was explored. In addition, whether this receptor serves as a biomarker and can be targeted by IL-13 cytotoxin was examined.
IL-13R expression in 15 normal and 68 ovarian tumor tissue samples was determined by immunohistochemistry. Correlation between clinicopathologic features and IL-13R expression was analyzed. The efficacy of IL-13R-directed cytotoxin was determined in mice with subcutaneous, orthotopic, and peritoneal metastatic ovarian cancer.
Immunohistochemical analyses revealed that 83% of ovarian cancer specimens express IL-13Rα2, a high-affinity IL-13R subunit chain, whereas normal ovary samples expressed none or very low levels. The majority of clear cell ovarian carcinomas with the worst prognosis showed strong staining for IL-13Rα2. IL-13 cytotoxin was highly cytotoxic to the IGROV-1 ovarian cancer cell line in vitro, and it mediated significant antitumor activity against a xenografted tumor model. The antitumor effects were confirmed by treating orthotopically implanted or peritoneal metastatic ovarian tumors, which showed significant extension of survival in immunodeficient mice. IL-13 cytotoxin also prevented cachexia in treated mice. The soluble form of IL-13Rα2 was detected in the serum of mice with peritoneal metastasis, and the level decreased to baseline in the treated group.
IL-13Rα2 is a promising target for ovarian cancer therapy, and the soluble form of IL-13R may be a possible surrogate marker for disease monitoring. Cancer. Published 2006 by the American Cancer Society.
Ovarian carcinoma is estimated to be the fourth major cause of cancer-related deaths in females and the leading cause of death from gynecological malignancy in the U.S., with approximately 16,200 deaths anticipated for 2005.1 The majority of ovarian cancer patients (70%) are diagnosed at an advanced stage (Stage III or IV) of the disease, at which time the primary tumor has metastasized, leading to few therapeutic options for these patients. In addition, the incidence of distant metastasis at diagnosis in patients with ovarian cancer is highest among all cancer types.1 It is therefore extremely important to diagnose at an early stage. For diagnostic purposes, conventional methods such as computed tomography (CT), magnetic resonance imaging (MRI), ultrasonography, and nuclear medicine scans are generally utilized. In addition, CA 125, a tumor marker, has been tested in the plasma of patients with ovarian cancer. CA 125 serves as a biomarker for monitoring tumor recurrence and prognosis of patients.2 However, CA 125 is not a reliable predictor for diagnosis due to low sensitivity and there is currently no screening strategy that is proven to be effective.3 Therefore, for early diagnosis, novel approaches such as identifying cancer-specific cells as biomarkers in the blood or serum/urine by genomics or proteomics technologies are needed.4 Recently, there have been extensive studies and progress in the molecular pathogenesis analysis and gene profiling of ovarian cancers using serial analysis of gene expression and DNA microarray technologies.5 Although a number of novel genes that are involved in ovarian cancer have been identified by these studies, their role and significance as cancer biomarkers is still unknown. In a recent study, a serum proteomic pattern was reportedly linked with the early diagnosis of ovarian cancer.6 Whether these biomarkers will be confirmed for a large number of patient samples is not known.
Part of the difficulty for early diagnosis and later therapy of epithelial ovarian cancer is the fact that these tumors are diverse in histologic types. The most common histologic type of epithelial ovarian cancer is papillary serous carcinoma, which is often associated with elevation of CA 125, followed by endometrioid, mucinous, and clear cell types.7 Among these pathologic types of ovarian cancer, the most difficult tumors to treat are clear cell carcinoma because of their chemoresistance.8 The literature is replete with references showing that inflammatory or immune reactions play a key role in the development and progression of cancer as a result of the natural host response to cancer, often along with normal tissue remodeling.9 It has been suggested that damage in the ovarian epithelium resulting from repeated ovulations may play a role in the development of cancer. These processes are coordinated by a number of growth factors, cytokines, chemokines, and their receptors. These factors may act in an autocrine or paracrine fashion influencing the outcome of tumor growth and metastasis, or tumor immunosurveillance.10 Among these cytokines, interleukin-13 (IL-13), a predominantly Th2-derived cytokine, is known to act as a proinflammatory cytokine. It is possible that IL-13 and its receptor may play a role in ovarian cancer oncogenesis, growth, or metastasis.
Recent studies have shown that IL-13 is expressed at elevated levels in both primary and metastatic ovarian tumors compared with normal ovary.11 In addition, IL-13 can inhibit apoptosis of tumors through the tumor necrosis factor (TNF)-α pathway.11, 12 IL-13 binds to 2 chains of IL-13R (IL-13Rα1 and IL-13Rα2). The IL-13Rα2 chain binds IL-13 with high affinity and has been shown to be overexpressed in brain tumor, head and neck cancer, renal cell carcinoma, and Kaposi sarcoma associated with the acquired immunodeficiency syndrome (AIDS).13 Recently, Wang et al.14 reported elevated expression of IL-13Rα1 in peritoneal and attached stroma samples in patients with ovarian cancer compared with samples derived from benign disease. To target IL-13R, IL-13 cytotoxins comprised of IL-13 and a truncated form of Pseudomonas exotoxin (IL13-PE38 or IL13-PE38QQR) were produced. These molecules mediated IL-13 receptor-specific cytotoxicity to tumor cells in vitro and in vivo.13
To determine the incidence and significance of IL-13R expression in human ovarian cancer specimens, we investigated whether normal and ovarian cancer tissue specimens express IL-13R. To determine whether IL-13R can provide a suitable target for cancer therapy, we investigated the antitumor activity of IL-13 cytotoxin against ovarian cancer cells in vitro and in animal models of human ovarian tumors. We also evaluated whether IL-13Rα2 is secreted in the serum of tumor-bearing mice that serves as a possible biomarker for disease monitoring.
MATERIALS AND METHODS
Cell Culture and Reagents
The human PA-1, SK-OV3, and OVCAR-3 ovarian cancer cell lines were purchased from the American Type Culture Collection (Manassas, VA) and cultured as described previously. The human IGROV-1 and OVCA429 ovarian cancer cell lines were kindly provided by Dr. K. Stromberg (Food and Drug Administration [FDA], Bethesda, MD). IL13-PE38 was produced and purified in our laboratory.15
Tissue Specimens, Tissue Array, and Immunohistochemistry
Human normal ovarian tissue sections (n = 15) and 22 primary ovarian tumor paraffin-embedded samples were obtained from the Cooperative Human Tissue Network (CHTN) after obtaining approval from the FDA Research Involving Human Subject Committee (RIHSC). In addition, a tissue array containing 46 different benign and malignant ovarian tumors was purchased from a commercial source (Imgenex, San Diego, CA). Immunohistochemistry was performed on all these sections using the Vector ABC peroxidase kit according to the manufacturer's instructions (Vector Laboratories, Burlingame, CA). Sections were incubated with antihuman IL-13Rα1 monoclonal antibody (Diaclone, Besancon, France), antihuman IL-13Rα2 monoclonal antibody (R&D Systems, Minneapolis, MN), or isotype control (immunoglobulin [Ig] G). Tissue sections were independently evaluated by 2 authors (M. Kioi and M. Kawakami) and by Dr. S. Takahashi (National Cancer Institute [NCI], Bethesda, MD). The results were scored on the basis of the density of staining: −, negative; −/+, slightly positive; +, moderately positive; and ++, strongly positive. The percentage of IL-13R-positive cells was estimated by counting IL-13R-positive cells divided by the total number of tumor cells in a given field. At least 1000 cells were counted in 3 different fields. Because of manual evaluation and counting, tissue array samples were excluded from assessment of percentage of IL-13R-positve cells.
Total RNA was isolated by Trizol reagent (Invitrogen, Carlsbad, CA) and reverse-transcriptase polymerase chain reaction (RT-PCR) was performed using specific primers as described previously.16
Radioreceptor Binding and Protein Synthesis Assays
Recombinant human IL-13 was labeled with 125I (Amersham, Arlington Heights, IL) by the IODO-GEN iodination reagent (Pierce, Rockford, IL) according to the manufacturer's instructions. The IL-13 equilibrium binding studies were performed by the method previously described.15 The in vitro cytotoxic activity of IL13-PE38 was measured by the inhibition of protein synthesis.15
Subcutaneous Xenografted Ovarian Tumor Model
OVCAR-3 and IGROV-1 cells were inoculated subcutaneously (s.c.) into the right dorsal flank of 6-week-old female BALB/c nude mice. Tumor growth was monitored and tumor size (mm2) was calculated by multiplying the length and width of tumor on a given day. Treatments were started 6 days after tumor cell inoculation, when tumors reached a mean size of 30 mm2 (5–6-mm dimensions). Mice were randomly divided into different therapeutic groups and 1 control group (6–7 mice per group) and injected with excipient (0.2% human serum albumin in phosphate-buffered saline [PBS]) or IL13-PE38 by either continuous intraperitoneal infusion (CIP) or intratumoral (IT) using a microinjection syringe.
Orthotopic Xenografted Ovarian Tumor Model
Tumor chunks obtained from s.c.-growing tumors were minced and approximately 3 × 3 mm tumor pieces were surgically sutured in the right ovary under anesthesia. Tumor volume (V) was calculated using the formula: V (mm3) = L × W2 × π/6, in which L equals the length of the tumor in mm and W equals the width of the tumor in mm.17 Treatments were initiated 5 days after tumor implantation. Mice were randomly divided into different therapeutic groups and 1 control group (5 mice per group). Animals were injected IP with excipient (0.2% human serum albumin in PBS) or various dosages of IL13-PE38 (25 μg/kg, 50 μg/kg, and 75 μg/kg) twice a day for 5 days.
Peritoneal Metastasis of Ovarian Tumor in an Animal Model
IGROV-1 cells (5 × 106) were inoculated into the peritoneal cavity of nude mice. After 5 days of tumor inoculation, IL13-PE38 was administered by either IP (70 μg/kg twice a day for 5 days) or by CIP infusion (140 μg/kg/day for 7 continuous days) using a continuous infusion pump (for CIP). For the advanced tumor model, 1 × 107 IGROV-1 cells were inoculated IP. After 10 days of tumor growth, IL13-PE38 treatment was given as described above for the 5 days tumor model. The survival of mice was monitored until all animals had died.
Detection of Soluble IL-13Rα2 in Serum
Serum was collected from tumor-bearing mice 4–5 weeks after IP inoculation of IGROV-1 tumor. Then 96-well immunoplates were coated with IL-13Rα2 monoclonal antibody (10 μg/mL), washed, and blocked with 5% milk in 0.1% Tween 20, 50 mM Tris-HCl (pH 7.5), and 0.15 M NaCl (TBST). The mouse serum (at 1:30 dilution) and serially diluted reference standard (recombinant human IL-13Rα2 extracellular domain) were then added. After 2 hours at 37°C, biotinylated goat antihuman IL-13Rα2 polyclonal antibody was added for 1 hour at room temperature. Finally, peroxidase-labeled streptavidin and TMB (3,3′, 5,5′tetra-methylbenzidine) peroxidase substrate were added and incubated to detect IL-13Rα2. The sensitivity of the assay was 1.56 ng/mL.
The statistical significance of data was calculated using the Student t-test. All statistical tests were 2-sided. Survival rates were calculated by the Kaplan-Meier method. For analysis of statistical significance between the therapeutic group and the control group in survival assays, the log-rank test was used. The difference in IL-13Rα2 expression between malignant and normal ovarian tissues or cancer tissues and benign tumor tissues was calculated by the chi-square test. The chi-square test was used for statistical analysis to evaluate the relation between the expression of IL-13Rα2 chain and patient age, clinical stages of disease, grade, or histology of disease.
In Situ Expression of IL-13R Chains
Fifteen normal ovary and 68 ovarian tumor tissue sections including tissue array samples were analyzed by immunohistochemistry for the expression of IL-13Rα1 and IL-13Rα2 chains. Approximately 83% of malignant ovarian tumor specimens (44 of 53 specimens) demonstrated positive staining (cell surface and intracytoplasmic) for the IL-13Rα2 chain (Fig. 1 shows staining of IL-13Rα2 in 2 samples, TA10 and TA57) (Table 1). In contrast, 87% of normal ovary tissue specimens showed slight staining (+/−) or no staining (Fig. 1, Sample numbers 1 and 3). Approximately 34% of ovarian cancer tissues expressed the IL-13Rα2 chain at high density (++) and 49% at moderate density (+) (Table 1). Seventeen percent of malignant samples showed weak or no staining for the IL-13Rα2 chain. When IL-13Rα2-positive cells in tissue sections were analyzed, approximately 67 ± 22% cells were positive in the high IL-13Rα2-expressing sections, whereas 49 ± 17% were positive in the moderate density-expressing sections. Some background staining was noted in weakly positive samples. It is interesting to note that clear cell carcinoma showed the highest percentage of IL-13Rα2 positive cells (data not shown). In contrast, all normal ovary and ovarian cancer tissue specimens showed similar staining for IL-13Rα1 chain.
Table 1. Immunohistochemical Analysis of IL-13Rα2 in Normal and Malignant Ovarian Tissues
We also performed in situ hybridization assays to further confirm the immunohistochemistry results by examining IL-13Rα2 mRNA expression. Similar to immunohistochemistry, ovarian cancer tissue sections showed intense positivity for IL-13Rα2 mRNA, whereas normal ovary showed weak positivity (data not shown).
Analysis of Correlation between IL-13R Expression in Ovarian Tumor Samples and Clinicopathologic Parameters
To investigate the significance of IL-13Rα2 expression in ovarian cancer, we compared IL-13Rα2 expression with clinicopathologic parameters. As summarized in Table 2, the expression level of the IL-13Rα2 chain was significantly higher in ovarian tumors compared with normal ovary tissues (P < .0001). Similarly, malignant ovarian tumors expressed a significantly higher level of IL-13Rα2 compared with benign tumors (P < .0001). There was a significant correlation between histology of ovarian cancer and the expression level of IL-13Rα2 chain (P < .05). IL-13Rα2 chain-overexpressing cells were much higher in clear cell carcinoma sections compared with other histologic types of ovarian cancers (75% of samples were ++ positive in clear cell carcinoma compared with 27% strongly positive in other types of ovarian cancers). There was no significant correlation between IL-13Rα2 expression and the distribution of ages, tumor grade, or International Federation of Gynecology and Obstetrics (FIGO) stage classification.
Table 2. Correlation between IL-13Rα2 Expression in Normal Ovary and Ovarian Cancer Sections and Various Clinicopathologic Features
We also determined the expression of IL-13Rα1 and IL-13Rα2 mRNA by semiquantitative RT-PCR in 5 ovarian cancer cell lines. The mRNA expression for neither IL-13Rα1 nor IL-13Rα2 was detected in the OVCA-429 cell line, but all other ovarian cancer cell lines tested showed a similar expression level of IL-13Rα1 mRNA (Fig. 2A). In contrast, expression of IL-13Rα2 mRNA was detected in only the IGROV-1 and SK-OV3 cell lines, whereas other cell lines (PA-1 and OVCAR-3) did not show any expression. To further confirm expression of IL-13R on the cell surface, we performed radiolabeled IL-13 binding assays (Fig. 2B). IGROV-1 cells clearly showed high binding of radiolabeled IL-13, which was inhibited by unlabeled IL-13 but not by IL-4, suggesting that the binding was IL-13 specific. The SK-OV3, PA-1, and OVCAR-3 cell lines showed low binding, which did not appear to be inhibited by IL-13 or IL-4.
Sensitivity of Ovarian Cancer Cell Lines to IL13-PE38
Expression of cell surface receptors may provide potential targets for selective delivery of therapeutic agents. Because many ovarian tumors and 2 ovarian cancer cell lines expressed high to moderate levels of IL-13Rα2, we determined the sensitivity of ovarian cancer cell lines to IL-13 cytotoxin (IL13-PE38), which is composed of IL-13 and a truncated form of Pseudomonas exotoxin. The cytotoxicity of IL-13 cytotoxin was determined by protein synthesis inhibition assays. As shown in Figure 2C, the IGROV-1 cell line was the most sensitive to IL13-PE38 (IC50: 90 ng/mL). Conversely, consistent with the lack of specific IL-13 binding, the SK-OV3, OVCAR-3, and PA-1 cell lines were not sensitive to IL-13-PE38 (IC50: >1000 ng/mL). These results are also consistent with the IL-13Rα2 mRNA expression.
Intratumoral Administration of IL13-PE38 Eradicates Ovarian Tumors
To determine the antitumor activity, IL13-PE38 was injected in 2 different s.c. xenografted ovarian tumor models. After tumor nodules were established in immunodeficient animals, IL13-PE38 was injected intratumorally for 5 consecutive days. As shown in Figure 3A, IL13-PE38 treatment (100 or 250 μg/kg doses) inhibited IGROV-1 s.c. tumor growth in a dose-dependent manner. At the highest dose, 5 of 7 mice showed complete eradication of their tumors. The remaining small tumors in 2 mice did not show growth as long as we monitored these animals (>95 days). However, the tumors in control mice continued to grow exponentially and showed ulceration, and therefore mice were sacrificed on Day 95. Although IL13-PE38 treatment did not cause complete regression of IGROV-1 s.c. tumor at a lower dose, the overall reduction in tumor size was 55% at the 100 μg/kg dose (P < .002) and 97% at the 250 μg/kg dose (P < .0001). In contrast, OVCAR-3 tumor-xenografted mice were resistant to IL13-PE38 treatment and the average tumor size in both treatment groups was not different from the control group. These results were consistent with in vitro cytotoxicity data indicating that the expression of IL-13Rα2 is necessary for the antitumor effect of IL13-PE38 in vivo.
IL13-PE38 Eradicates Large Subcutaneous Ovarian Tumors
Because both dosages (100 and 250 μg/kg) of IL13-PE38 showed strong antitumor activity against small IGROV-1 s.c. tumors, we next examined whether larger ovarian tumors can also be treated with IL13-PE38. For this study, IGROV-1 tumors were grown to a size of 75 mm2 and on Day 70 of tumor establishment IL13-PE38 was given by the IT or CIP routes. For comparison, the same total amount of IL13-PE38 (2 mg/kg in 2 courses) was given by 2 routes. To achieve this, IL13-PE38 was given by CIP for 7 days from Day 70 to 77 and Day 85 to 92 or IT for 5 days from Day 70 to 75 and Day 85 to 90. The tumors began to regress after completion of IL13-PE38 infusion in both groups of treatment and no regrowth of tumors was observed until the end of the experiment (Day 120, Fig. 3B). Overall, tumor size in the CIP- and IT-treated groups was significantly smaller compared with the control group (P < .007 control vs. CIP and P < .001 control vs. IT). It is interesting to note that animals receiving IL-13 cytotoxin by the IT route continued to show tumor regression even after the treatment cycles had been completed, and 4 of 5 mice showed complete eradication of their tumors. In contrast, tumors in control mice continued to grow, resulting in sacrifice of animals on Day 107 for ethical reasons.
Intraperitoneal Treatment with IL13-PE38 Prolongs Survival and Prevents Cachexia of Mice Implanted with Orthotopic Ovarian Tumors
We also evaluated overall survival of ovarian tumor-bearing mice when treated with increasing doses of IL13-PE38. Mice orthotopically implanted with IGROV-1 tumors were treated with excipient only (control) or IL13-PE38 (25, 50, or 75 μg/kg twice daily for 5 days from Days 5 to 9) and their survival was monitored. As shown in Figure 4A, survival of mice was extended in a dose-dependent manner. At the 50 μg/kg IL13-PE38 dose, survival of the tumor-bearing host was significantly enhanced compared with control animals (P < .05). At the highest IL13-PE38 dose (75 μg/kg), 60% of IGROV-1 tumor-bearing mice survived for as long as we followed these animals (>250 days), whereas all untreated mice died in <100 days.
Orthotopically implanted IGROV-1 ovarian tumor-bearing mice showed substantial body weight loss with tumor progression in the excipient treatment group, whereas IL13-PE38-treated mice showed weight gain (Fig. 4B). After 4 weeks of tumor implantation, the body weight of control mice did not increase, gradually stabilized, and after 6 weeks a loss in body weight occurred despite growing tumor size. In contrast, tumor-bearing mice treated with the 50 or 75 μg/kg dose of IL-13 cytotoxin showed a linear increase in body weight for up to the 12 weeks we followed these animals. It is interesting to note that the 25 μg/kg IL13-PE38 dose did not significantly improve survival (Fig. 4A); nevertheless, these animals showed an increase in body weight, followed by a decline (Fig. 4B). Overall, the body weight of the 25 μg/kg IL13-PE38-treated animals remained significantly higher compared with untreated control animals (P < .05).
Intraperitoneal Treatment with IL13-PE38 Prolongs Survival of Mice with Peritoneal Metastasis
We also evaluated the survival of mice with peritoneal spread of ovarian tumor after IP treatment with several dosages of IL13-PE38. IGROV-1 cells were inoculated IP in mice and after 5 days of tumor growth animals were treated with excipient (control) or IL13-PE38 (70 μg/kg, twice a day for 5 days, IP group; 140 μg/kg/day for 7 days, CIP group) and their survival was monitored. As shown in Figure 5A, control mice showed a peritoneal spread of ovarian tumors to multiple organs including liver, kidney, mesentery, and peritoneum. In addition, these animals developed large amounts of ascitis within 5–6 weeks after inoculation of tumor cells. These metastatic tumor-bearing mice died within 60 days at the 5 × 106 dose and < 40 days with the 107 tumor cells dose. In contrast, the survival of mice was significantly extended in the IL13-PE38 treatment groups when given by both IP and CIP routes (P < .007 and P < .002 vs. control, respectively) (Fig. 5B). There was no statistically significant difference in survival between IL-13 cytotoxin administered by the IP and CIP routes (P = .81). It is interesting to note that both IP and CIP treatment with IL13-PE38 was effective (P < .0006 and P < .0006 vs. control, respectively) in extending the survival of mice even in the advanced IP metastasis model. These results indicate that IL13-PE38 may be a useful agent even in patients with IP metastasis of ovarian tumor.
Serum IL-13Rα2 Levels in Ovarian Tumor-Bearing Mice with Peritoneal Metastasis
Because ovarian tumors express high to moderate levels of IL-13Rα2 and the soluble form of IL-13Rα2 can be detected in the serum of mice with certain disease states, we determined whether immunodeficient animals with human ovarian tumor metastasis also secrete the soluble form of IL-13Rα2. Nude mice were injected with IGROV-1 tumor cells and peritoneal metastasis was established. These mice were divided into 2 groups: control and treatment with IL13-PE38. Serum was collected 1 to 5 weeks after tumor inoculation. Untreated control mice with ovarian tumor showed elevation of serum IL-13Rα2 (Fig. 5C) at Week 5 after tumor inoculation, whereas IL13-PE38 treatment by both the IP and CIP routes showed significantly lower levels of serum IL-13Rα2 (P < .001 control vs. IP and P < .004 control vs. CIP). The serum in naive mice showed an undetectable level of soluble IL-13Rα2.
In the current study, we demonstrate that the IL-13Rα2 chain is expressed in 34% of malignant ovarian tumor tissue samples at high levels and 49% at intermediate levels. In contrast, normal ovary and benign ovarian tumor tissue samples expressed no or low levels of IL-13Rα2. There was a correlation between IL-13Rα2 expression and histology of ovarian cancer. Clear cell carcinoma seemed to overexpress IL-13Rα2 in 75% of specimens, which was the highest among all histologic types of ovarian carcinoma. Several other genes or gene products have been shown to be highly expressed in clear cell carcinoma. These include p21, hepatocyte nuclear factor-1β, and cyclin E.18, 19 IL-13Rα2 joins these overexpressed molecules in clear cell ovarian carcinoma, and perhaps may serve as a biomarker for advanced disease.
It has been demonstrated that IL-13 cytotoxin is extremely cytotoxic to IL-13Rα2-expressing human tumor cells, including those derived from glioblastoma, head and neck carcinoma, renal cell carcinoma, and AIDS-associated Kaposi sarcoma.13 In addition, IL-13 cytotoxin is highly efficacious in mediating antitumor activity in various human tumor animal models.13, 20 In contrast, IL-13 cytotoxin is not cytotoxic to normal cells lacking expression or with low expression of this receptor chain. Based on these studies, several Phase I/II clinical trials using IL-13 cytotoxin in patients with recurrent malignant glioma were initiated. Some of these clinical trials have been completed and show that IL-13 cytotoxin up to a concentration of 0.5 μg/mL is extremely well tolerated, without any evidence of toxicity to normal brain cells, as normal brain cells do not express IL-13Rα2. A Phase III clinical trial involving infusion of IL-13 cytotoxin in normal brain surrounding a tumor cavity after resection has been recently completed at numerous U.S. and European clinical centers (PRECISE Study). Patients are currently being followed for the safety and efficacy of IL13-PE as a treatment of glioblastoma multiforme.23
Consistent with these preclinical and clinical activities, we show that IL-13 cytotoxin mediates antitumor activity in s.c., orthotopic, and peritoneal metastatic ovarian xenograft tumor models. It is interesting to note that when ovarian tumors were grafted orthotopically onto the ovaries of immunodeficient mice, IL-13 cytotoxin mediated remarkable antitumor effects when injected IP. Not only were tumors completely regressed, the treatment also enhanced the survival of these animals. The highest dose of IL-13 cytotoxin by the IP route showed a curative effect in 60% of mice. This is a remarkable observation very rarely seen with other therapeutic approaches. Administration of IL-13 cytotoxin directly into the peritoneal cavity allowed targeting of ovarian tumors and maintained a high concentration of the agent at the tumor site without increasing the risk of systemic toxicity.
One of the characteristics of ovarian cancer is their IP spread, leading to an advanced stage of the disease invading other organs. Thus, the IP route of administration of therapeutic agents seems to be an appropriate route for disseminated ovarian cancer therapy. Therefore, we investigated the efficacy of IL-13 cytotoxin administrated IP using either a continuous infusion pump or bolus administration. By these routes of administration, IL-13 cytotoxin mediated dramatic antitumor effects in very large established s.c. xenografted tumors. Therefore, the IP route of administration of IL-13 cytotoxin may be a viable option. Although we observed remarkable antitumor effects of IL-13 cytotoxin in ovarian cancer models, no visible toxicities were seen. It is reported that IL-13 binds to IL-13R in the mouse system and also mediates its signaling.21 Therefore, testing the therapeutic efficacy and systemic toxicity of IL-13 cytotoxin in the mouse model is clinically relevant. In previous studies, IL-13 cytotoxin was administrated intravenously (i.v.) to CD2F1 mice or cynomolgus monkeys, and only reversible hepatic toxicity was observed at dosages up to 50 μg/kg given every alternate day for 3 days.13 In our study, we also did not observe any visible toxicity of IL-13 cytotoxin when given by the IP or IT routes. In addition, when IL-13 cytotoxin was given by CIP at 143 μg/kg/day for 7 days, no abnormal visible changes were observed in animals. This observation is consistent with our previous observation that higher doses of IL-13 cytotoxin can be administrated to mice by continuous infusion compared with bolus injection.20
It is interesting to note that the presence of peritoneal tumor metastasis resulted in secretion of the extracellular domain of IL-13Rα2 in the serum of mice, whereas normal mice did not show a detectable level of soluble IL-13Rα2. Because treatment of mice with IL-13 cytotoxin resulted in a statistically significant decline in serum soluble IL-13Rα2 levels, our results indicate for the first time that soluble IL-13Rα2 may serve as a biomarker for the diagnosis of ovarian cancer and perhaps other cancers expressing high levels of IL-13Rα2, or a biomarker of treatment response in patients with ovarian cancer. Additional animal models of ovarian cancer are needed to confirm this observation. As a majority of ovarian cancers are diagnosed at an advanced stage, it would be highly desirable to develop novel diagnostic tools and markers that can increase the frequency of diagnosis at an early stage of ovarian cancer.
Because IL-13 binds with high affinity to IL-13Rα2, the soluble form of this receptor may inhibit the cytotoxicity of IL13-PE. Consistent with this hypothesis, the efficacy of IL-13 cytotoxin will be compromised or somewhat lower when high levels of soluble receptor is cleaved and detected in the blood. It is of interest to note that the serum level of soluble receptor in the current study was relatively low compared with the >3-fold higher concentration of IL-13 cytotoxin detected even after a single i.v. injection of IL-13 cytotoxin in mice (100 μg/kg).22 Thus, it is possible that the excess serum concentration of IL-13 cytotoxin will overcome the inhibitory effects of soluble receptor and enough drug will be available for antitumor effects. As seen in the current study, even in the presence of soluble serum receptor levels IL-13 cytotoxin was able to significantly mediate antitumor effects and increase survival of tumor-bearing mice. Thus, the IL-13Rα2 chain not only serves as a target for ovarian cancer therapy, but soluble IL-13R may serve as an important biomarker for tumor response.
In conclusion, IL-13Rα2 is a potential diagnostic marker to identify ovarian cancer, monitor tumor response, and a potential target for therapy. Therefore, a large number of ovarian cancer specimens should be examined for the expression of IL-13Rα2 to confirm this conclusion. Finally, because IP administration of IL-13 cytotoxin has revealed remarkable antitumor activity in animal models, it is expected that IL-13 cytotoxin may be a useful agent in ovarian cancer therapy, and additional preclinical studies should be performed to support future clinical studies in this disease.
We thank Dr. Andrew Byrnes at CBER/FDA for helpful suggestions and critical reading of the article, Dr. Satoru Takahashi at NCI for the assessment of immunohistochemistry, and Ms. Pamela Dover for help and technical support.