Statistical considerations for enumeration of circulating tumor cells.

Area of Science:

• Oncology
• Biomedical Engineering
• Medical Statistics

Background:

• Circulating tumor cells (CTCs) are rare in carcinomas.
• In metastatic breast cancer, ≥5 CTCs/7.5ml blood correlates with poor survival.
• Understanding CTC detection limitations is crucial due to clinical implications.

Purpose of the Study:

• To develop a statistical model for CTC detection and enumeration.
• To identify sources of error in CTC analysis.
• To evaluate the impact of reducing errors on prognostic thresholds.

Main Methods:

• Statistical analysis of multi-center clinical trial data.
• Utilized the CellSearch® System for CTC isolation and enumeration.
• Modeled statistical issues from blood collection to image analysis.

Main Results:

• Developed a statistical model for CTC isolation and detection.
• Employed Poisson and empirical distributions for process steps.
• Reader variability identified as a key error source affecting the 5 CTC threshold.

Conclusions:

• Error reduction in CTC identification could lower the poor prognosis threshold.
• Potential to reduce threshold from 5 to 1 CTC/7.5ml blood.
• Improved patient stratification for metastatic breast cancer.