Prospective technical validation and assessment of intra-tumour heterogeneity of a low density array hypoxia gene profile in head and neck squamous cell carcinoma

Abstract Background and purpose Tumour hypoxia is associated with a poor prognosis in head and neck squamous cell carcinoma (HNSCC), however there is no accepted method for assessing hypoxia clinically. We aimed to conduct a technical validation of a hypoxia gene expression signature using the TaqMan Low Density Array (TLDA) platform to investigate if this approach reliably identified hypoxic tumours. Materials and methods Tumour samples ( n = 201) from 80 HNSCC patients were collected prospectively from two centres. Fifty-three patients received pimonidazole prior to surgery. TaqMan Low Density Array-Hypoxia Scores (TLDA-HS) were obtained by quantitative real-time PCR (qPCR) using a 25-gene signature and customised TLDA cards. Assay performance was assessed as coefficient of variation (CoV). Results The assay was sensitive with linear reaction efficiencies across a 4log10 range of inputted cDNA (0.001–10 ng/μl). Intra- (CoV = 6.9%) and inter- (CoV = 2.0%) assay reproducibility were excellent. Intra-tumour heterogeneity was lower for TLDA-HS (23.2%) than for pimonidazole (67.2%) or single gene measurements of CA9 (62.2%), VEGFA (45.0%) or HIG2 (39.4%). TLDA-HS in HNSCC cell lines increased with decreasing pO2 . TLDA-HS correlated with Affymetrix U133 Plus 2.0 microarray HS ( p < 0.01) and positive pimonidazole scores ( p = 0.005). Conclusions Gene expression measurements of hypoxia using a 25-gene signature and TLDA cards are sensitive, reproducible and associated with lower intra-tumour heterogeneity than assaying individual genes or pimonidazole binding. The approach is suitable for further assessment of prognostic and predictive capability in clinical trial material.