Related Experiment Videos
Intermediate markers as surrogate endpoints in cancer research
1Nutritional Epidemiology Branch, National Cancer Institute, Bethesda, Maryland, USA.
Hematology/Oncology Clinics of North America
|August 19, 2000
Summary
Surrogate endpoints in cancer research offer faster, cheaper studies but carry uncertainty. Validating these markers requires large, expensive trials, highlighting the need for robust evidence despite the cost.
Area of Science:
- Oncology
- Biostatistics
- Epidemiology
Background:
- Surrogate endpoints in cancer research offer advantages like smaller, faster, and less expensive studies compared to those using frank cancer outcomes.
- Advancements in cell and molecular biology are generating new therapies and potential surrogate markers, increasing the attractiveness of surrogate endpoint studies.
- Surrogate endpoint studies can be suggestive and play a role in Phase II trials, but their validity for inferring cancer outcomes is often uncertain.
Purpose of the Study:
- To address the uncertainty associated with surrogate endpoints in cancer research.
- To advocate for rigorous investigations to evaluate the validity of potential cancer surrogates.
- To emphasize the need for studies that generalize surrogate endpoint findings to actual cancer outcomes.
Main Methods:
- The article discusses the critical evaluation of surrogate markers through surrogate-cancer studies and mediation analyses.
- It highlights the importance of considering the totality of causal connections for surrogate validity, not just placement on the causal pathway.
- The text examines the inferential strength gained by using 'downstream' markers and the associated increase in study size and cost.
Main Results:
- Many potential surrogate markers carry uncertainty, and plausible alternatives can make inferences about cancer problematic.
- There is a history of surrogate markers providing incorrect answers for chronic disease therapies, suggesting cancer surrogacy is not immune.
- Establishing surrogate validity for one intervention or exposure does not guarantee validity for another due to exposure dependence.
Conclusions:
- The full evaluation of surrogate endpoints often requires large, long, and expensive studies, ironically negating their initial cost-saving advantage.
- Inferential certainty with surrogate endpoints is directly associated with study cost; 'you get what you pay for'.
- The limitations of surrogacy underscore the complexity of cancer causation and affirm the continued importance of large clinical trials and observational studies with explicit cancer endpoints.