Estimating Tumor Proportion in Smear Slides for Reliable Molecular Analysis

Area of Science:

• Cytopathology
• Molecular Pathology
• Oncology Diagnostics

Background:

• Molecular pathology is crucial for cancer diagnostics and theranostics.
• Cytological smears are increasingly used for molecular testing.
• Accurate tumor cell percentage estimation is vital for reliable molecular test results.

Purpose of the Study:

• To evaluate the reliability of tumor cell percentage estimations among different pathology specialists.
• To compare the consistency of estimations by an expert cytopathologist, a molecular cytopathologist, and a molecular pathologist.

Main Methods:

• Digital images of May-Grünwald-Giemsa stained mediastinal lymph node smear slides were used.
• Three pathologists independently estimated tumor cell percentages in predefined categories.
• Manual cell counting served as the gold standard for comparison.

Main Results:

• The molecular cytopathologist demonstrated the highest consistency (Kappa = 0.69), followed by the expert cytopathologist (Kappa = 0.64).
• The molecular pathologist showed moderate consistency (Kappa = 0.52).
• Agreement was highest in the 71-100% category, but tumor proportions were often overestimated in the 11-20% category.

Conclusions:

• Variability in tumor percentage estimation necessitates standardized protocols and training.
• Substantial agreement was achieved in certain categories, but discrepancies in borderline cases require attention.
• Further research is needed to enhance tumor cell estimation methods for molecular pathology.