GLUT1 and CAIX as intrinsic markers of hypoxia in bladder cancer: relationship with vascularity and proliferation as predictors of outcome of ARCON

Abstract

Glucose transporter-1 protein (GLUT1) and carbonic anhydrase IX (CAIX) are regulated by hypoxia inducible factor-1 (HIF-1) and have been studied as putative intrinsic cellular markers for hypoxia. This study directly compares CAIX and GLUT1 with pimonidazole binding in a prospective series of bladder cancer patients and also studies the prognostic significance of the markers, in combination with vascularity and proliferation, in a retrospective series of bladder cancer patients treated in a phase II trial of radical radiotherapy with carbogen and nicotinamide (ARCON). A total of 21 patients with a diagnosis of transitional cell carcinoma of the bladder received 0.5 g m(-2) pimonidazole. Serial tumour sections were stained for pimonidazole, GLUT1 and CAIX and compared. Tissue sections obtained from a series of 64 patients previously treated for invasive bladder cancer using ARCON were stained for GLUT1 and CAIX together with Ki-67 and CD31/34. There was a good geographical colocalisation of both intrinsic markers with pimonidazole and a highly significant agreement in individual patients; correlation coefficients were 0.82 (P=0.0001) for GLUT1 and 0.74 (P<0.0001) for CAIX. In both series of patients, the intrinsic hypoxia markers were highly correlated with each other and a correlation with proliferation was also evident in the retrospective study. In univariate and multivariate analyses, GLUT1 and CAIX were independent predictors for overall and cause specific survival. The hypoxia markers did not predict for local control or metastases-free survival although higher Ki-67 indices showed a trend towards local failure. The data suggest that both hypoxia modification and accelerated treatment may be valid treatment options in bladder cancer.

Figure 1

Serial sections stained for pimonidazole (top), GLUT1 (middle) and CAIX (bottom).

Figure 2

Comparison between intrinsic and extrinsic markers of hypoxia. The data are presented as the intrinsic markers subtracted from the pimonidazole score for each tumour in the prospective study. Box 1 indicates good absolute agreement, while boxes 2 and 3 represent disparity between the markers.

Figure 3

Correlation between GLUT1 and CAIX in the prospective (•) and retrospective (▵) studies.

Figure 4

Examples of double staining. Serial sections showing similar areas stained for CAIX (brown) vs CD31/34 (red) (upper), CAIX (red) vs Ki-67 (brown) (middle), and GLUT1 (brown) vs CD31/34 (red) (lower).

Figure 5

Relationship between the proliferation and expression of CAIX and GLUT1. The data show the individual values for Ki-67 positivity and the mean and s.e.m. for the three categories of intrinsic marker expression.

Figure 6

Influence of GLUT1 (upper panel) and CAIX (lower panel) on the overall survival of patients treated with radical radiotherapy, carbogen and nicotinamide. The data were classified into tumours with no expression of the proteins and using a cutoff of 10% for those that were positive.