Ki67 Quantitative Interpretation: Insights using Image Analysis

Area of Science:

• Digital Pathology
• Oncology
• Quantitative Analysis

Background:

• Ki67 is a crucial proliferation marker for cancer diagnosis and prognosis.
• Current methods for Ki67 quantification are debated, impacting clinical decisions.

Purpose of the Study:

• To evaluate an automated image analysis tool for Ki67 quantification in neuroendocrine tumors.
• To compare the accuracy and efficiency of digital analysis against manual counting.

Main Methods:

• A Leica digital pathology platform with a nuclear algorithm was used.
• Tumor regions of interest (ROIs) were analyzed by the algorithm and manually by pathologists.
• The impact of ROI size and number on Ki67 counts was assessed.

Main Results:

• The algorithm provided rapid and reproducible Ki67 results, significantly faster than manual counting.
• Users reported improved accuracy, time savings, and ease of use with the digital tool.
• Variations in ROI selection and cell counts influenced Ki67 results around critical cutoffs for tumor grading.

Conclusions:

• Digital image analysis offers a faster, more accurate method for Ki67 quantification.
• Standardization of cell number and ROI selection is crucial to avoid clinically significant variations in grading.
• This highlights the risks of relying on fixed cutoffs for continuous variables like Ki67.