Strategies for developing Ki67 as a useful biomarker in breast cancer

Increased proliferation is a hallmark of malignant tumors. The proliferation marker Ki67 has been investigated as a breast cancer biomarker, but despite 32 years of research the best cutpoints and the best methods for determination are still under debate. This review is based on an overview on the efforts to standardize Ki67 and to optimize its performance that was presented at the St. Gallen oncology conference 2015. The clinical validity of Ki67 as a prognostic marker as well as a predictive marker (in the neoadjuvant setting) has been shown in several meta-analyses. Depending on cohort characteristics, molecular subtype and clinical setting, Ki67 is a prognostic marker, a predictive marker, or both. Many different cutpoints for Ki67 have been reported, but it is has not been possible to determine an evidence-based "optimal" cutpoint. This supports the view that Ki67 is continuous marker, reflecting the continuous variation of the proliferation rate in different tumors. We should probably stop looking for an "optimal" cutpoint for Ki67 because it simply does not exist. It is evident from the results of several ring trials that intermediate levels of Ki67 are particularly difficult for standardization. Due to the low analytical validity in the intermediate range as well as intratumoral heterogeneity, the clinical utility of intermediate Ki67 levels is limited. Clinical decisions should not be based on small differences in the intermediate range and additional molecular tests might be necessary for tumors with intermediate Ki67 levels. For the two groups of tumors with a very low or a very high Ki67 a clinical interpretation could be straightforward. Despite these limitations, the assessment of proliferation is a central parameter for tumor characterization and an important element of the pathological assessment.