Originally published as JCO Early Release 10.1200/JCO.2008.18.7393 on February 9 2009

This Article Full Text Full Text (PDF) Publisher's Note Erratum (v27,p2742) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Pintilie, M. Articles by Hill, R. P. Search for Related Content PubMed PubMed Citation Articles by Pintilie, M. Articles by Hill, R. P. Social Bookmarking                            What's this?

Heterogeneity and Power in Clinical Biomarker Studies

From the Departments of Biostatistics and Medical Oncology, Radiation Medicine Program, and the Division of Applied Molecular Oncology, Ontario Cancer Institute/Princess Margaret Hospital, University Health Network; Department of Laboratory Medicine, St Michael's Hospital; Dalla Lana School of Public Health and the Departments of Laboratory Medicine and Pathobiology, Medicine, University of Toronto; Department of Radiation Oncology, University of Toronto, Toronto, Canada.

Corresponding author: Melania Pintilie, MSc, 610 University Ave., Biostatistics Department, Ontario Cancer Institute/UHN, Toronto, M5G 2M9, Canada; e-mail: Pintilie{at}uhnres.utoronto.ca.

Purpose Many recent studies have suggested the possibility that a variety of different biomarkers may be associated with treatment outcome. However, it is also apparent that some of these biomarkers are heterogeneously distributed within a tumor. Due to this heterogeneous distribution of the biomarker, the association sought may appear weak or nonexistent. Thus, there is a wide range of conclusions in the literature on the association between a biomarker and an outcome.

Results This article presents how to quantify the heterogeneity and how it influences the observed effect size and the ability to detect it (power of the study). It can be shown that the estimated effect size and the power of the study are diminished when the biomarker is measured with error. The estimated effect of the association with outcome of the average of several replicates per patient is closer to the true effect size when the number of replicates increases.

Conclusion The first step in designing a study of association between a biomarker and outcome is to conduct a pilot study in which several measurements per patient are taken. Based on these data, the heterogeneity of the marker within and between individuals can be estimated and used in the process of designing an appropriate study of the association between the biomarker and outcome.

Supported by the Terry Fox Foundation.

Presented in part at the Biospecimen Research Network Symposium, March 13-14, 2008, Washington, DC; and Canadian Association of Radiation Oncology, September 10-13, 2008, Montreal, Quebec, Canada.

Authors' disclosures of potential conflicts of interest and author contributions are found at the end of this article.

CiteULike    Complore    Connotea    Del.icio.us    Digg    Facebook    Reddit    Technorati    Twitter    What's this?