Can biomarkers help us understand the nutritional and lifestyle factors important in cancer prognosis?

Abstract

In attempting to discover the causes of cancer, investigators have recognized that biomarkers can confirm biological plausibility, enhance relative risks, and serve as surrogate endpoints in observational and intervention studies. In the arena of cancer survival, the potential value of biomarkers is increasingly appreciated. A broad range of histological, cellular, and molecular markers have been identified among persons diagnosed with cancer. Molecular and cellular markers are being used to stage disease, predict prognosis, and target therapeutic interventions. Biomarkers in survivors can also help us to understand factors that influence prognosis by both elucidating pertinent biological pathways and sharpening risk estimates. However, as in the case of incident cancer, the use of biomarkers as surrogate endpoints postdiagnosis is problematic because of the potential existence of alternative pathways to recurrence and death that bypass the surrogate endpoint. In evaluating potential surrogates, an understanding of the causal structure underlying the interrelations of exposures, surrogate, and recurrence or death is essential. Three questions can help to shed light on this structure: 1) What is the relation of the surrogate endpoint to recurrence or death? 2) What is the relation of the intervention (or exposure) to the surrogate? 3) To what extent does the surrogate endpoint mediate the relation between intervention (exposure) and recurrence or death? To address these questions, it is imperative to integrate biomarker studies into ongoing pharmacotherapeutic and lifestyle intervention studies with recurrence or mortality as explicit endpoints. © 2007 American Society for Nutrition.