Originally published as JCO Early Release 10.1200/JCO.2008.20.7175 on February 2 2009

This Article Full Text (PDF) 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 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 Schneider, B. P. Articles by Miller, K. D. Search for Related Content PubMed Articles by Schneider, B. P. Articles by Miller, K. D. Related Articles Related Article Social Bookmarking                            What's this?

In Reply

Indiana University School of Medicine, Indianapolis, IN

Molin Wang

Harvard School of Public Health; and Dana-Farber Cancer Institute, Boston, MA

George W. Sledge

Indiana University School of Medicine, Indianapolis, IN

Robert J. Gray

Harvard School of Public Health; and Dana-Farber Cancer Institute, Boston, MA

David A. Flockhart, Kathy D. Miller

Indiana University School of Medicine, Indianapolis, IN

We thank Min-Han et al for their insightful comments. The primary analysis in our correlative study¹ aimed to evaluate for a genotypic effect within each arm, and thus this is what was reported. When applying the Cox proportional hazard model for each of the VEGF-2578 and VEGF-1154 genotypes, including the genotype (three levels), treatment arm, and their interaction, the interaction between the VEGF-2578 genotype and treatment arm for overall survival (OS) was not significant (P = .11). The interaction between the VEGF-1154 genotype and treatment effect was significant (P = .020). In the model for VEGF-1154, based on the Wald test with a null hypothesis that the hazards for OS were equal for all three genotypes, the genotype effect was significant (P = .0017) in the experimental arm but not in the control arm (P = .70). Alternatively, when quantifying the VEGF-1154 genotype using the dose level of the VEGF-1154 A allele, the interaction between dose level of the VEGF-1154 A allele and treatment arm was significant (P = .0049). However, these statistical results should be interpreted with caution because they are exploratory.

We agree with the concern regarding the potential bias in the analysis of samples that were derived from only an unselected fraction within each treatment arm. We evaluated for the distribution of established clinical variables in the subgroups that we analyzed compared with those in the subgroups that we were unable to analyze (due to lack of samples) in each arm. We believe this was the best means to establish that there were no significant imbalances in clinical variables that might have explained the differences seen in OS (as opposed to the genetic variables that we reported), and those results are as follows. In the experimental arm, only the distribution of visceral disease was different in our analyzed subgroup compared with the unanalyzed subgroup, and this variable did not correlate with survival. In the control arm, only the presence of bone-only disease and visceral involvement were different. In multivariate analysis (Cox proportional hazard model) for the experimental arm, when evaluating the effect of the VEGF-2578 genotype on OS—with the potential clinical confounders included in the model—the genotype effect remained significant (P = .049). When performing the same analysis for VEGF-1154, the genotype effect remained significant (P = .0082). Thus, there are no apparent clinical imbalances that could have been responsible for the OS differences seen in our analyzed subgroups.

The concern regarding analysis based on ethnic admixture is reasonable. In this study, the percentages of white patients (84%) versus nonwhite patients (16%) in the samples that we tested were similar to the percentages (white patients [83%] v nonwhite patients [17%]) in the group as a whole. However, in the combination arm of E2100, the hazard ratio for OS was not affected by race. Although the distribution of genotype was indeed different among races, the outcome (ie, OS) was not different among races. Unfortunately, because of the small sample size in this trial, it would not be statistically feasible to evaluate for a genetic effect in the nonwhite population alone.

In summary, there are clear limitations to unplanned retrospective evaluations for biomarkers. We performed a number of exploratory evaluations to minimize the likelihood of nongenetic bias contributing to our results. Nonetheless, these results should be viewed with caution. Regardless of magnitude, validation in an independent data set is critical before drawing definitive conclusions regarding the role of vascular endothelial growth factor polymorphisms as biomarkers for bevacizumab.

AUTHORS' DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST

Although all authors completed the disclosure declaration, the following author(s) indicated a financial or other interest that is relevant to the subject matter under consideration in this article. Certain relationships marked with a "U" are those for which no compensation was received; those relationships marked with a "C" were compensated. For a detailed description of the disclosure categories, or for more information about ASCO's conflict of interest policy, please refer to the Author Disclosure Declaration and the Disclosures of Potential Conflicts of Interest section in Information for Contributors.

Employment or Leadership Position: None Consultant or Advisory Role: Bryan P. Schneider, Genentech (C); George W. Sledge, Genentech (C); Kathy D. Miller, Genentech (C), Roche (C), Entremed (C), Pfizer (C) Stock Ownership: None Honoraria: Kathy D. Miller, Roche Research Funding: Bryan P. Schneider, Genentech; Kathy D. Miller, Genentech, Roche Expert Testimony: None Other Remuneration: None

Acknowledgment

Supported by an American Society of Clinical Oncology Career Development Award, the Breast Cancer Research Foundation, and a grant from Genentech. D.A.F. is supported by funding for the Consortium on Breast Cancer Pharmacogenomics from the National Institute of General Medical Sciences Pharmacogenetics Research Network (Grant No. U01GM061373).

REFERENCE

1. Schneider BP, Wang M, Radovich M, et al: Association of vascular endothelial growth factor and vascular endothelial growth factor receptor-2 genetic polymorphisms with outcome in a trial of paclitaxel compared with paclitaxel plus bevacizumab in advanced breast cancer: ECOG 2100. J Clin Oncol 26:4672–4678, 2008.[Abstract/Free Full Text]

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

Related Article

• Race to Report: Are Vascular Endothelial Growth Factor Genetic Polymorphisms Associated With Outcome in Advanced Breast Cancer Patients Treated With Paclitaxel Plus Bevacizumab?
  Tan Min-Han, Tan Chuen Seng, and Lim Wei Yen
  JCO 2009 27: 1342 [Full Text]

This Article Full Text (PDF) 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 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 Schneider, B. P. Articles by Miller, K. D. Search for Related Content PubMed Articles by Schneider, B. P. Articles by Miller, K. D. Related Articles Related Article Social Bookmarking                            What's this?