Background: The assessment of the proliferation fraction is becoming more and more important; however, there is no consensus concerning optimal validation. Depending on the institute the proliferation fraction is determined either from a core biopsy (SB) or resection specimen (OP).
Patients and methods: The interobserver variability and the results of SB and OP were investigated whereby two pathologists independently estimated the proliferation fraction of 90 cases of invasive breast cancer. The results (Ki-67) were quantified, categorized, and compared.
Results: Identical (accuracy of 5% steps) results between the 2 pathologists were achieved in 43% (n=39) for SB, 47% (n=42) for OP and 60% (n=54) for SB versus OP. When categorizing the proliferation fraction (low ≤ 15%, moderate 20-30% and high ≥ 35%) the following results were achieved: 76% (n=68) for SB, 82% (n=74) for OP and 81% (n=73) for SB versus OP.
Conclusions: There was a clear interobserver variability (SB: kappa=0.32, OP: kappa=0.34) but this could be dramatically improved by forming proliferation categories (SB: kappa=0.62, OP: kappa=0.72, 60 versus 81%). In SB with a low proliferation fraction a repeated analysis in OPs can be advisable as a higher proliferation fraction is observed in up to 12% of OPs.