An international study to increase concordance in Ki67 scoring

Area of Science:

• Oncology
• Pathology
• Biomarker Standardization

Background:

• Ki67 is a crucial breast cancer biomarker, but its clinical utility is hindered by a lack of standardized scoring methods.
• Previous studies highlighted significant variability in Ki67 scoring among different laboratories.

Purpose of the Study:

• To assess the effectiveness of a standardized Ki67 scoring method using a web-based calibration tool.
• To improve inter-laboratory reproducibility for Ki67 assessment in breast cancer tissue microarray slides.

Main Methods:

• 16 laboratories from eight countries calibrated to a specific Ki67 scoring protocol via web-based training and test images.
• Laboratories then scored 50 centrally stained MIB-1 tissue microarray cases using the standardized method.
• Performance was evaluated using Root Mean Square Error, Maximum Absolute Deviation, and intraclass correlation (ICC).

Main Results:

• Calibration showed promising trends in reducing scoring errors (RMSE decreased from 0.6 to 0.4, P=0.07).
• Post-calibration, tissue microarray scoring achieved a high ICC of 0.94 (95% CI: 0.90-0.97), exceeding the target of >0.70.
• Despite improvements, some discrepancies persisted, particularly around clinically relevant cutoffs.

Conclusions:

• A standardized Ki67 scoring method, facilitated by a web-based tool, can achieve high inter-laboratory reproducibility on tissue microarray slides.
• The findings suggest standardization is feasible but emphasize the need for further research.
• Future studies should address staining variability, extend to biopsies/whole sections, and correlate with clinical outcomes before widespread clinical adoption.