Data and metadata supporting the article: Systematic evaluation of scoring methods for Ki67 as a surrogate for 21-gene recurrence score

Although Ki67 labeling index is a potential predictive marker for chemotherapy benefit, its clinical utility has been limited by the lack of a standard scoring method resulting in poor inter-observer reproducibility. Especially, there is no consensus on the use of average versus hotspot score for reporting.In order to determine the best method for Ki67 scoring and validate manual scoring method proposed by the International Ki67 Working Group (IKWG), the authors of this paper systematically compared average versus hotspot score in 245 cases with a public domain image analysis program QuPath. They used OncotypeDx Recurrence Score (RS) as a benchmark to compare the potential clinical utility of each scoring methods.
Data access: The Ki67 image analysis data supporting the findings of this study, are publicly available in the figshare repository, as part of this data record. The scanned microscopic images are not publicly available as informed consent for image publication was not obtained for individual images. The images can be obtained from the corresponding author upon reasonable request. For data access requests, please contact Soonmyung Paik, MD, Yonsei University College of Medicine, email address: [e-mail] (or [e-mail]).
Study approval and patient consent: This study was approved by the institutional review board of Yonsei University College of Medicine (1-2014-0067). All patients signed consent to allow routine surgical pathology specimens for research.
Study aims and methodology: The meta-analysis of gene expression profiles in breast cancer suggested that the prognostic ability of gene expression based tests including the OncotypeDx is mainly driven by the expression levels of proliferation genes. Thus, Ki67 labeling index (LI) based on immunohistochemistry has a potential to serve as a surrogate for OncotypeDx.However, utilization of Ki67 LI has been hampered by poor inter-laboratory reproducibility mainly due to the lack of a standard scoring method as demonstrated by a multicenter reproducibility study conducted by the International Ki67 Working Group (IKWG). In order to determine the best scoring method for Ki67 LI and to evaluate the feasibility of utilizing an open source image analysis software for objective scoring, the authors utilized QuPath for unbiased analysis of whole sections stained with Ki67 from a cohort of 245 cases previously tested for OncotypeDx.
A total of 245 subjects with estrogen receptor positive breast cancer were analyzed in the study. All patients were diagnosed with estrogen receptor positive breast cancer at the Korean National Cancer Center and were tested for OncotypeDx between 2011 and 2015.
As part of a routine immunohistochemistry panel, ER, PR, and Ki67 staining (1:200, clone MIB-1, DAKO) was performed at National Cancer Center using the Ventana platform (Ventana Medical Systems, Inc, Tucson, AZ, USA). Slides were scanned with Aperio ScanScope with 20x objective lens (Leica Biosystems, Seoul, South Korea).QuPath (version 0.2.0.), an open source image analysis software was used. Square grids were created to fill the entire invasive tumor areas drawn by a pathologist. Grid sizes of 300, 400, and 500 micrometers were used. Average 651 (range 7 to 2920), 423 (range 6-2021), and 290 (range 4-1146) grids were generated per case for each grid size respectively. Grids containing ductal carcinoma in situ or lymphoid aggregates, as well as those containing staining artifacts were manually excluded upon visual inspection of the grids.
Data supporting the findings of the study: Dataset countdata.xlsx, is in .xlsx file format, and contains Ki67 image analysis data from 240 cases of breast cancer with known OncotypeDx Recurrence Score (RS).Scanned whole slide image files of 240 cases of breast cancer stained with Ki67, are in .SVS file format.
The scanned microscopic images will be made available on request as described in the data access section above.