or even existing cancer cells) or in tissues surrounding the area of interest (such as evidence of neo-vascularization or inflammation surrounding a cancer). Biomarkers may be in exfoliated cells, or they may be soluble or suspended molecules (for example proteins, DNA, microRNA) in circulation or in secretions. Finally, inherited germline biomarkers can be evaluated from circulating leukocytes or exfoliated cells from easily accessible tissues, such as from a cheek swab.


Biomarkers have many important potential roles in clinical research and in clinical practice (IOM, 2010; see also Tables 1-1 and 1-2 in Chapter 1). These include prognosis, prediction of response to therapy (effect modifiers), prediction of clinical outcome (surrogate endpoints), risk assessment, screening, diagnosis, pharmacogenetics, and patient monitoring during and after treatment. Although these uses are applicable to most if not all disease processes, this appendix refers to oncologic examples because most of the case studies for this report arose from the field of oncology.

Prognostic Factors and Effect Modifiers1

Prognostic Factors

Prognostic factors, used to estimate the risk of or time to clinical outcomes such as disease recurrence or progression, may be useful even though these biomarkers are simply correlated with the causal mechanisms of the disease process (Fleming, 2005). In oncology, biomarkers may have roles as both prognostic factors and effect modifiers, and, in fact, these may be mixed. For example, overexpression and/or amplification of the human epidermal growth factor receptor 2 (HER2) portend a poor prognosis in breast cancer patients who do not receive any adjuvant systemic therapy. Moreover, this same biomarker appears to be associated with poorer response to endocrine therapy in patients with estrogen receptor (ER)-positive breast cancer (compared to those who have ER-positive, HER2-negative cancers), but it has been associated with higher response rates to various chemotherapies (e.g., anthracyclines, taxanes) and it is especially related to benefit from anti-HER2 therapies, such as trastuzumab and lapatinib (Wolff et al., 2007). Therefore, any study addressing the prognostic role of a biomarker needs to take into account the specific, intended clinical use of the biomarker and the potential confounding effects of how patients in the study cohort are treated. Perhaps


1Also see the discussion in Chapter 1.

The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement