Abstract
Objective:
Assess interaction of pazopanib, an oral antivascular endothelial growth factor inhibitor, with radiation in tumor xenograft models.
Methods:
Flank xenografts in female athymic nude mice of human lung cancer cell line, A549, and head and neck cancer cell line, UM-SCC-6, were allowed to grow to ∼5×5 mm. Groups were then treated with pazopanib and/or escalating doses of radiation and tumor measurements over time compared with untreated tumor-bearing controls. Pazopanib (100 mg/kg) began 7 days before radiation and continued for 28 days. Daily radiation was 0.5, 1, 2, or 3 Gy ×5 days.
Results:
Tumors in the A549 control group reached >4× the original size by day 36 postradiation. All treatment groups had less robust tumor growth (p<0.05) and the group receiving pazopanib+3 Gy radiation/day had tumor regression to less than baseline. In the UM-SCC-6-tumor-bearing animals, tumors in all treatment groups had less robust growth than untreated controls after day 23 post-treatment.
Conclusion:
The combination of pazopanib and radiation resulted in a trend of superior tumor growth inhibition compared with either agent alone. All treatment groups had impaired tumor progression compared with untreated controls.
Introduction
Pazopanib is an oral angiogenesis inhibitor targeting the tyrosine kinase activity associated with vascular endothelial growth factor (VEGF) receptors (VEGFR-1, -2 and -3), platelet-derived growth factor receptors-α and -β, and stem cell factor (c-KIT). Pazopanib is approved for the treatment of advanced renal cell carcinoma plus metastatic soft tissue sarcomas and is under study for other malignancies. 1 –4 Investigational studies involve its use in conjunction with other agents, including study in cooperative group trials. 5,6 Although published clinical trials and multiple preclinical studies have used pazopanib with chemotherapies/biologics, basic reports of its interaction with radiation are lacking. One clinical report testing the combination of radiation and pazopanib found it tolerable for skin toxicity, but did not assess for improved efficacy over radiation alone. 1 A positive interaction is anticipated as other antiangiogenesis VEGF inhibitors are known to be radiosensitizers. 7,8 This study examines the interaction of pazopanib with radiation in two in vivo models as a step to better understanding its clinical utility for cancer therapy.
Materials and Methods
The nonsmall cell lung cancer cell line, A549, and the head and neck cancer cell line, UM-SCC-6, were used to establish xenografts in female athymic nude mice. Separately for each cell line, 5×106 cells (UM-SCC-6) or 2×106 cells (A549) mixed with Matrigel were injected subcutaneously into the flank and tumors allowed to grow to ∼5×5 mm. Groups of 10 mice were then treated with pazopanib or vehicle (30% Captisol; CyDex Pharmaceuticals, Inc., Lenexa, KS) and/or escalating doses of radiation (IC-320; Kimtron, Inc., Woodbury, CT), and tumor measurements over time compared with untreated tumor-bearing controls (Captisol vehicle only) for 50 days. Pazopanib was given at 100 mg/kg by oral gavage, beginning 7 days before radiation and continuing for a total of 28 days. The choice of 100 mg/kg oral dose is based on preclinical studies where a dose/response of tumor inhibition was observed, with oral 100 mg/kg once or twice daily being the highest dose and comparing favorably to subcutaneous administration through continuous infusion. 9 These pharmacokinetic/pharmacodynamic studies showed that a pazopanib concentration of ∼40 μmol/L, which was achieved by 100 mg/kg, is required for maximum inhibition of VEGFR2 and inhibited multiple other kinases in mouse lung. 9 Results from these murine studies were applied to a Phase I clinical trial as a target steady-state concentration of ≥40 μmol/L. Near this concentration was achieved in the majority of patients receiving doses of ≥800 mg once daily or 300 mg twice daily (mean trough concentration 34 μmol/L). 10 Although dosing ranged up to 2000 mg/day, most patients reached acceptable steady-state levels at the 800 mg dose, which has become the recommended dose for FDA approved use in patients with renal cell carcinoma and is being studied in other malignancies. 11
Daily radiation dose was 0.5, 1, 2, or 3 Gy ×5 days. The repeat experiment used 2×106 UM-SCC-6 cells and the 3 Gy treatment group was replaced by a 0.5 Gy group. Data were combined from two A549 animal experiments (n=20) to evaluate if tumor growth was associated with treatment regimens and over time using a longitudinal analysis method with a mixed model procedure. Experiment days from the two data sets were combined for measures ±1 day. If an animal was euthanized due to the size of tumor, that tumor size was carried forward to subsequent study days. Univariate and multivariate analysis were used.
Biomarker studies
Thirty-one randomly selected animals were sacrificed on day 20 for A549 tumors and day 28 for UM-SCC-6 tumors. Tumors taken at the time of sacrifice were processed into blocks. Slides cut from each block were stained for Ki67 as a biomarker of proliferation, CD31 for angiogenesis, or with H&E to aid the pathologist in review for necrosis. Correlation analysis between biomarkers and tumor size was performed using the tumor size at day 17 since this was close to the time of sacrifice.
Results
The growth pattern for most groups bearing A549 lung cancer xenografts initially treated (n=10) is shown in Figure 1. Multivariate analysis showed that the tumor size significantly increased over time (p<0.001) for all groups, and that changes of tumor size were different by treatment groups over time (p<0.0001). In the A549 untreated control group, tumors reached greater than 4× the original tumor size before day 26 and some animals were euthanized due to a large tumor burden. Among the treatment groups, only one animal in the 1 Gy group and one that received pazopanib alone had to be euthanized due to a large tumor burden.
Comparison of A549 mean tumor xenograft growth over time for untreated controls, and groups of mice treated with pazopanib and/or 1 or 3 Gy radiation/day ×5 days (n=10 mice/group). Error bars represent±SD.
The group receiving the most aggressive treatment of pazopanib+3 Gy radiation/day had tumor regression to less than baseline by day 21 postradiation treatment. Although 2 Gy groups are not shown in Figure 1 for simplicity, the growth patterns were relatively similar and slightly less favorable for 2 Gy versus 3 Gy±pazopanib and these were not statistically different (Table 1). The repeat experiment had similar patterns, showing a greater efficacy with increasing radiation and a trend of improved inhibition with the addition of pazopanib. Table 1 provides a comparison of most A459 xenografts, with data combined from the two experiments for the repeat groups. As shown in Table 1, where shaded rows indicate comparisons that were statistically different, the tumor size in all treatment groups was significantly less than untreated control tumors (p<0.0001).
This analysis includes combined data for the initial and repeat experiments (n=20) for the 1 and 2 Gy groups; 3 Gy groups were only in the initial experiment.
Estimate rounded to nearest whole number; standard error was 25.9–26.2 for all.
Similar to the results in the A549 tumor-bearing animals, there is more tumor inhibition with higher radiation doses and a positive interaction between radiation and pazopanib in the UM-SCC-6 groups. An example of the interaction is shown in Figure 2, where tumor inhibition after 0.5 Gy is improved with the addition of pazopanib. In the UM-SCC-6-tumor-bearing animals, tumors in all groups continued to grow during initial therapy and then declined by day 13 postradiation treatment. Those animals treated with pazopanib alone had a growth and regression pattern most similar to that of the untreated controls, whereas those receiving the combination of pazopanib and radiation had tumor regression below baseline. Confirming the radiation alone efficacy at the highest dose like the A459 groups, the UM-SCC-6 tumor-bearing animals treated with 3 Gy/day had the most vigorous regression pattern, and some animals had complete response to treatment.
Comparison of UM-SCC-6 tumor xenograft growth over time for untreated controls, and groups treated with pazopanib and/or 5 days of 0.5 Gy radiation/day (n=10 mice/group). Error bars represent±SD.
There are no significant associations between the biomarkers studied and treatment group. The percent of cells staining for Ki67 ranged from 5 to 85, but did not show any association with the tumor size using the Pearson correlation measurement with rho=0.1227, p=0.5109. There was a trend of association of angiogenesis marker CD31 with tumor size at day 17, but this was not statistically significant even at the 10% level (p=0.1044).
Categories of mild, moderate, and severe necrosis were used for the pathologist's assessment of necrosis. There was a significant difference in tumor size by category (p=0.0004) with mean tumor size of 147 mm2, median 140 mm2, and range 120–196 mm2 in the mild necrosis group, whereas the mean was 132 mm2, median 125 mm2, with a range of 100–168 mm2 in the moderate group, and the mean was 103 mm2, median 101 mm2, with a range 80–136 mm2 in the severe necrosis group.
Discussion
Although the pattern of growth and treatment effects were different for the lung versus the head and neck cancer xenografts, both showed an enhanced tumor inhibition with the addition of pazopanib over radiation alone. There was a positive trend of interaction at all radiation dose levels studied, despite some not being statistically significant. The lack of correlation of tumor growth outcome with biomarkers of proliferation and angiogenesis leaves open the need for additional study to determine the mechanisms of interaction. Since pazopanib is a potent inhibitor of multiple tyrosine kinases, the evaluation of kinase activities in tumor specimens represents a feasible pretreatment assessment for biomarker development, but this has not yet been accomplished. 12,13
Conclusions
The combination of pazopanib and limited radiation resulted in a trend of superior inhibition of tumor growth compared with either agent alone. All treatment groups had impaired tumor progression compared with untreated controls.
Footnotes
Acknowledgments
The authors thank Sheila Bright and Leisa Seay for technical assistance in animal model studies, and Sally Wakefield for article preparation. Supported by Glaxo Smith Kline.
Disclosure Statement
The author has no significant financial interests that are related to or would reasonably appear to be affected by the proposed article. All authors involved with the article have been informed of their obligations under federal regulations governing disclosure of significant financial interests and have no conflicts of interest or potential conflicts of interest that have not been disclosed.