Quality control (QC) tissue array for the prospective rvaluation of translational research projects. A program of the EORTC GI and RGB Groups

Abstract

9600 Background: Translational research often correlates immunohistochemistry (IHC) on archival tissue with clinical information. However, IHC is not standardized and quality control is not yet widely established. Important parameters which can not be controlled are the fixation time and tissue heterogeneity. Therefore, a tissue array was designed to evaluate the suitability of a given antigen/antibody test before use in clinical trials. Methods: A 216 spot tissue array was designed using paraffin-embedded tissue from 14 colorectal cancers. Immediately after surgery, specimen from tumor rim or tumor center were formalin-fixed individually over night, for > 24 h, or for > 48 h, respectively. After evaluation by H&E staining, three cores from each specimen were assembled in a tissue array, i.e. 18 cores per tumor. Sections were distributed to 6 laboratories for IHC each using one or more of the commercial antibodies for p53, Ki67, CEA, CK20, Cox2 in routine staining protocols, or for 3 experimental antigens. Blinded reading of immunoreaction was performed by the investigators using a 3-step visual grading scale and by a central laboratory before unblinding of allocation. Results: In all laboratories, IHC was successful using the commercial antibodies. IHC with the experimental antigens failed due to high background or inconsistent staining. The staining pattern varied considerably between cores from individual specimen (e.g. Ki67: from 10–90% pos. nuclei; CEA: 50–100% pos. cells, CK20: 10–60% pos. cells) with moderate differences for staining intensity (variation from weak to moderate or moderate to strong). If average staining intensity or staining pattern for the three cores per location were calculated, there was no difference between tumor center and rim (p ≥0.62 in paired t-tests for all comparisons). Most importantly, fixation time had no impact on the expression levels of the five antigens (p values from 0.12 to 0.93). Conclusions: This QC tissue array is a useful tool to ensure the reliability of a given antibody and staining procedure before its use in translational research in multicenter clinical trials. No significant financial relationships to disclose.