Related Experiment Video
Updated: May 21, 2026
06:19
Constructing and Visualizing Models using Mime-based Machine-learning Framework
Published on: July 22, 2025
Deriving benefit of early detection from biomarker-based prognostic models
1Department of Biostatistics, University of Washington, Seattle, WA 98195, USA. linoue@u.washington.edu
Biostatistics (Oxford, England)
|June 26, 2012
Summary
Lowering cancer biomarker thresholds for early detection may not improve survival due to lead time bias. Survival benefits depend on biomarker change rates and lead time, not just detection timing.
Area of Science:
- Oncology
- Biostatistics
- Epidemiology
Background:
- Prognostic models for cancer often utilize biomarkers for early detection and survival prediction.
- Higher biomarker levels typically correlate with poorer prognosis, suggesting early detection might improve survival.
Purpose of the Study:
- To investigate whether prognostic models imply a true survival benefit under early detection after accounting for lead time.
- To analyze the influence of lead time and biomarker dynamics on survival benefit in cancer prognostic models.
Main Methods:
- Examined prognostic models using age and/or biomarker levels at disease detection.
- Assessed the impact of biomarker change rate, lead time, and baseline biomarker levels on survival outcomes.
Main Results:
- The implied survival benefit from early detection is contingent upon the rate of biomarker change and the lead time.
- Lowering biomarker detection thresholds does not automatically guarantee an extended life expectancy.
Conclusions:
- Survival benefits suggested by prognostic models under early detection may be artifacts of lead time bias.
- Careful consideration of lead time and biomarker dynamics is crucial when interpreting survival benefits from early cancer detection strategies.
