Method validation of circulating tumour cell enumeration at low cell counts

Area of Science:

• Oncology
• Biomarker Analysis
• Clinical Diagnostics

Background:

• Circulating tumour cells (CTCs) are vital prognostic and predictive biomarkers in cancer care.
• Accurate enumeration of CTCs, especially at low cell counts (<10 cells/7.5 ml), is critical for clinical decision-making.
• The CellSearch™ system is a key technology for CTC detection, necessitating rigorous method validation.

Purpose of the Study:

• To validate the CellSearch™ system's performance for enumerating circulating tumour cells (CTCs) in clinical samples.
• To assess and quantify inter-operator variability in CTC detection, particularly in the clinically relevant low cell count range.
• To identify factors contributing to variability and propose methods for improvement.

Main Methods:

• Analytical accuracy was evaluated using quality controls (QCs) and spiked blood samples from healthy volunteers.
• Results were analyzed using β-expectation tolerance intervals (BETI) and β-content γ-confidence tolerance intervals (BCTI).
• Inter-operator error was characterized across 6 readers analyzing 38 patient samples, with a focus on samples containing ≤5 CTCs.

Main Results:

• Imprecision was 3-4 times higher at low QC levels (48 cells) compared to high QC levels (987 cells).
• Individual analysts introduced systematic errors in CTC image interpretation, correlated with training and experience.
• Selecting readers based on BETI and BCTI analysis reduced inter-operator error at ≤5 CTCs from 170% to <30%.

Conclusions:

• Significant inter-operator variability exists in enumerating CTCs at low cell counts.
• This variability is not inherent to the method but can be substantially reduced.
• Implementing simple guidance and ensuring adequate training can improve the reliability of CTC enumeration.