Enhanced tumour antiangiogenic effects when combining gefitinib with the antivascular agent ZD6126

Abstract

Current experimental and clinical knowledge supports the optimisation of endothelial cell targeting using a strategy combining anti-EGFR drugs with antivascular agents. The purpose of the present study was to examine the effects of the association of ZD6126, an antivascular microtubule-destabilising agent, with gefitinib and irradiation on the growth of six head and neck human cancer cell lines xenografted in nude mice and to study predictive and molecular factors responsible for antitumour effects. CAL33- and Hep-2-grafted cell lines were the most sensitive to ZD6126 treatment, with VEGF levels significantly higher (P=0.0336) in these tumour xenografts compared to Detroit 562- and CAL27-grafted cell lines with relatively low VEGF levels that were not sensitive to ZD6126. In contrast, neither IL8 levels nor EGFR expression was linked to the antitumour effects of ZD6126. ZD6126 in combination with gefitinib resulted in a synergistic cytotoxic interaction with greater antitumour effects than gefitinib alone. The synergistic interaction between ZD6126 and gefitinib was corroborated by a significant decrease in CD31 labelling. The present study may serve for future innovative clinical applications, as it suggests that VEGF tumour levels are possible predictors for ZD6126 antitumour efficacy. It also supports the notion of antitumour supra-additivity when combining gefitinib and ZD6126, and identifies neoangiogenesis as the main determinant of this synergistic combination.

Figure 1

Growth (log scale) of the different xenografted tumours in athymic nude mice. Bars represent standard error of the mean.

Figure 2

Distribution of VEGF (pg mg⁻¹ protein), IL8 (pg mg⁻¹ protein) and EGFR (fmol mg⁻¹ protein) levels in different xenografted tumours in athymic nude mice. Bars represent standard error of the mean.

Figure 3

Distribution of individual tumour VEGF levels as a function of tumour volume. A log–log scale was adopted due to the marked variations in the x- and y-axis.

Figure 4

Effects of ZD6126 against CAL33 human head and neck xenografts (filled triangles indicate the time of treatment). Bars represent standard error of the mean.

Figure 5

(A) Effects of ZD6126 and gefitinib or ZD6126+gefitinib on CAL33 tumour xenografts (filled triangles and arrows indicate the time of treatments). Bars represent standard error of the mean. (B) Effects of ZD6126+gefitinib+RT on CAL33 tumour xenografts (filled triangles and arrows indicate the time of treatments). Tumour growth curves of ZD6126+gefitinib and ZD6126+gefitinib+RT were superimposed. Bars represent standard error of the mean.

Figure 6

Effects of ZD6126, gefitinib, RT, ZD6126+gefitinib and ZD6126+gefitinib+RT on the labelling frequency and intensity of CD31 (endothelial cell marker) and Ki67 (proliferation marker) in CAL33 tumour xenografts. The final score was the result of the examination of three fields per tumour, and between four and eight tumours were investigated. Labelling intensities are shown as 0=no, 1=slight, 2=medium, 3=strong and 4=very strong, with the sum of all values in one histogram being 100%. Scoring was performed at the end of experiments on day 33.