Ki-67 Testing in Breast Cancer: Assessing Variability With Scoring Methods and Specimen Types and the Potential Subsequent Impact on Therapy Eligibility

Therese Bocklage ¹, Virgilius Cornea ¹, Caylin Hickey ¹

⁰Departments of Pathology and Laboratory Medicine.

Applied Immunohistochemistry & Molecular Morphology : Aimm +

|

March 7, 2024

View abstract on PubMed

Summary

This study highlights significant variation in Ki-67 scoring for abemaciclib eligibility. Refined Ki-67 assessment methods are crucial to reduce variability and ensure accurate patient selection for ER-positive/HER2-negative breast cancer treatment.

Area Of Science

Oncology
Pathology
Biomarker Analysis

Background

Abemaciclib is approved for ER-positive/HER2-negative breast cancer with Ki-67 ≥20%.
Lack of standardized guidelines for Ki-67 testing and scoring methods creates ambiguity.
Patient eligibility for abemaciclib therapy may be impacted by scoring variability.

Purpose Of The Study

To evaluate the variability of Ki-67 expression across different specimen types and scoring methods.
To compare manual Ki-67 scoring techniques (estimated, unweighted, weighted) with quantitative image analysis (QIA).
To assess the impact of scoring variations on determining eligibility for abemaciclib therapy.

Main Methods

Comprehensive analysis of breast specimens from 50 patients eligible for abemaciclib.
Three pathologists performed blinded reviews using estimated (EST), unweighted (UNW), and weighted (WT) manual scoring methods.
Quantitative image analysis (QIA) using the HALO platform was employed for comparison.
Evaluation of Ki-67 expression in core needle biopsy (CNB), resection (RES), and axillary lymph node metastasis (ALN) specimens.

Main Results

Average Ki-67 expression varied by scoring method: EST (14.68%), UNW (14.46%), WT (14.15%), and QIA (11.15%).
The weighted (WT) method demonstrated lower interobserver variability (ICC1=0.959) compared to EST (ICC1=0.922) and UNW (ICC1=0.949).
Treatment eligibility based on Ki-67 ≥20% ranged from 24% (QIA) to 40% (EST), indicating significant discrepancies.
No statistically significant difference in Ki-67 expression was found among CNB, RES, and ALN specimens.

Conclusions

Significant clinical variation exists in Ki-67 scoring for abemaciclib eligibility.
The weighted scoring method and QIA show potential for reducing interobserver variability.
Further refinement of Ki-67 scoring methodologies is essential to ensure accurate patient selection for targeted breast cancer therapies.