Quality assurance trials for Ki67 assessment in pathology

Ki67 is a broadly used proliferation marker in surgical pathology with an obvious need for standardization to improve reproducibility of assessment. Here, we present results of the so far only existing round robin tests on Ki67, organized annually in Germany, Austria, and Switzerland from 2010 to 2015 with up to 160 participating laboratories (QuIP). In each quality assessment trial, eight probes from each breast cancer, neuroendocrine tumor, and malignant lymphoma were compiled on a tissue microarray (TMA). TMAs were stained in the participants' laboratories with antibodies and procedures also applied in their daily routine. Participating pathologists were expected to assign Ki67 values to one of four different categories for each tumor type. All local stainings and evaluations were reassessed by the organizing panel and compared to a preset standard. On average, 95% of participants reached the benchmark of over 80% concordance rates with the Ki67 category pre-established by the panel. Automatization and type of antibody did not affect the success rate. Concordance rates differed between tumor entities being highest in each tumor type with either very high or very low labeling indices. Lower rates were seen for intermediate Ki67 levels. Staining quality improved during the observation period as did inter-observer concordance with 85% of participants achieving excellent agreement (kappa > 0.8) in the first year and over 95% in 2015. In conclusion, regular external quality assurance trials have been established as a tool to improve the reproducibility and reliability of the prognostic and predictive proliferation marker Ki67.