Comparison of standard immunohistochemistry and automated quantitative immunohistochemistry for quantification of signaling proteins in archival Ewing sarcoma samples.

Abstract

10045 Background: Targeted therapies are increasingly being evaluated for patients with Ewing sarcoma (EWS). Optimal strategies for quantifying key signaling proteins from paraffin-embedded material remain unclear. We sought to quantify tumor expression of IGF-1R, EGFR, and PTEN in EWS using two different methodologies. Methods: Decalcified paraffin-embedded tissue was obtained from 40 patients with EWS. 28 samples were obtained at time of diagnosis, 6 after initial chemotherapy, and 6 at time of relapse. Tumor was evaluated for the expression of IGF-1R, PTEN, and EGFR using standard immunohistochemistry (IHC) and automated quantitative analysis (AQUA) immunohistochemistry. Standard IHC results were categorized as low, medium and high expression based on a pathologist’s review. Expression by AQUA was measured as a continuous variable in arbitrary units. One sided ANOVA analysis was used to compare the mean AQUA expression for each analyte between categories of expression by IHC. Results: The mean age of patients was 15 years (range 1-49 yrs) at time of diagnosis. 68% were male and 66% had localized disease. Samples displayed a wide range of expression by AQUA: mean IGF-1R = 10702 (range 393 - 14424); EGFR = 2750 (range 672 – 9798); and PTEN = 2250 (range 251 – 6557). Mean IGF-1R expression by AQUA did not differ significantly between standard IHC expression categories (mean IGF-1R expression by AQUA for low IHC = 11255, medium IHC = 11070, high IHC = 11023; p = 0.98). Mean PTEN expression by AQUA was higher in the medium and high IHC categories (mean PTEN expression by AQUA for low IHC = 1229, medium IHC = 2715, high IHC = 2940; p = 0.064). Only two samples expressed EGFR by standard IHC. These samples qualitatively had higher AQUA expression levels, but there were too few samples for a reliable determination of statistical significance. Conclusions: AQUA provides a more dynamic range for each marker, though there was poor correlation with standard IHC results. Larger sample sets are needed to determine the optimal approach for quantifying signaling proteins in Ewing sarcoma.