The German quality assurance system for the molecular-pathological detection of KRAS-mutations in colorectal cancer

Abstract

4018 Background: In the beginning of 2008 the EMEA (European Medicines Agency) approved with panitumumab for the first time an EGFR (epidermal growth factor receptor) targeting therapy for patients with metastatic colorectal cancer overexpressing the EGFR and showing wildtypic sequences in the KRAS gene as a predictive biomarker. Thus, the need for assuring the quality of laboratories emerged. The German Society for Pathology in cooperation with the Federation of the German Pathologist supported by an unrestricted financial grant of Amgen Germany arranged a quality assurance system (QAS). In this context two round- robin tests were carried out which results are presented here. Methods: Collection of results from two round -robin tests and their statistical analysis applying binary classification tests. Results: Test sets of 4 histological sections from ten different cases of colorectal tumors with known mutational status of the KRAS gene were prepared for the round-robin tests. The method for the mutation detection was unrestricted. A total of 74 participants from universities (44 - 59.5 %) or other institutions (30 - 40.5 %) attended the tests. 11 participants (14.8 %) failed the test (6 universities: 13.6 %, 5 institutions: 16.6 %). For the analysis didesoxy-sequencing (DDS: 55 - 66.2 %), ARMS®-PCR (8 - 10.4 %), melting-point analysis (MPA: 7 - 9.1 %), pyrosequencing (PS: 6 - 7.8 %), hybridization (HYB: 4 - 5.2 %), or SSCP (1 - 1.3 %) were used, in which some participants (3) used more than one method. It turned out that all methods employed for the testing gave similar results when comparing the rate of correct or wrong hits or the rate of false positive detection: DDS (0.92, 0.07, 0.02), ARMS®-PCR (0.91, 0.08, 0.05), MPA (0.94, 0.06, 0.04), PS (0.95, 0.05, 0.03) or HYB (0.90, 0.10, 0.08). Conclusions: The quality of the molecular-pathological detection of KRAS mutations as precondition for an EGFR targeted therapy should be tested since about 15 % of laboratories did not meet a sufficient grade. For the mutation detection no method seemed superior. [Table: see text]