News Release

Other highlights in the February 16 JNCI

Peer-Reviewed Publication

Journal of the National Cancer Institute

Trio of Papers Address Issues in Using Serum Proteomics for Detecting Ovarian Cancer

In 2002, a study reported that an analysis of protein patterns in a blood sample could be used to detect ovarian cancer. However, questions have been raised about whether the technology produces results that are reproducible and reliable enough for use as a screening test. A new study and two responding commentaries discuss these issues.

Keith A. Baggerly, Ph.D., of the University of Texas M. D. Anderson Cancer Center in Houston, and colleagues examined the reproducibility of the proteomic pattern found in a previous study from a researcher at the State University of New York, Stony Brook, that used publicly available ovarian cancer datasets produced by Lance A. Liotta, M.D., Ph.D., of the National Cancer Institute, and Emanuel F. Petricoin III, Ph.D., of the U.S. Food and Drug Administration. Baggerly and colleagues maintain that the patterns, which were derived from blood samples to determine whether or not women had ovarian cancer, were not biologically plausible. In addition, they found that the method used in the Stony Brook study to classify patterns performed no better than chance. They conclude that the reproducibility of the proteomic profiling approach remains an open question.

In a commentary, Liotta, Petricoin, and colleagues respond to the study by Baggerly et al. They note that, because the parameters of their experiments changed for each of the publicly available datasets, they were not surprised that the patterns were different. They write that this discrepancy highlights the need for improved communication among scientists who produce and analyze data and for the establishment of a common platform and standardized operating procedures for this new technology so such studies can be compared across different laboratories.

A second commentary, from David F. Ransohoff, M.D., of the University of North Carolina at Chapel Hill, describes how chance and bias can affect the validity of experiments in many of the "-omics" fields and may cause flawed results and inflated expectations. "To address such threats and to realize the potential of new -omics technology will require application of appropriate rules of evidence in the design, conduct, and interpretation of clinical research about molecular markers," he writes.

Contacts:

  • Baggerly et al.: Nancy Jensen, M. D. Anderson Cancer Center, 713-792-0655, nwjensen@mdanderson.org
  • Liotta et al.: National Cancer Institute Press Office, 301-496-6641, NCIPressOfficers@mail.nih.gov, or Lenore Gelb, U.S. Food and Drug Administration Press Office, 301-827-6242, lgelb@oc.fda.gov
  • David Ransohoff, University of North Carolina at Chapel Hill, ransohof@med.unc.edu

    Study Explores Potential of Gene Methylation Detection for Cervical Cancer Screening

    Detection of changes in DNA methylation patterns has potential for use in cervical cancer screening, according to a new study.

    Changes in DNA methylation are an early event in carcinogenesis and are often present in the precursor lesions of various cancers, including cervical cancer. To examine whether DNA methylation changes might be used as markers of cervical intraepithelial neoplasia (CIN) and invasive cervical cancer, Nancy B. Kiviat, M.D., of the University of Washington in Seattle, and colleagues studied these changes in 20 genes in cervical cell samples from Sengalese women with increasingly severe CIN and invasive cervical cancer.

    The frequency of hypermethylation for at least one of three genes (DAPK1, RARB, or TWIST1) increased with increasing severity of neoplasia but was rare in samples with less severe neoplasia, those with CIN-1 or less. The authors estimated that, among the population studied, detection of the DAPK1, RARB, or TWIST1 hypermethylated gene would reveal histologically confirmed CIN-3 or worse with a sensitivity (proportion of true-positive results) of 60% and a specificity (proportion of true-negative results) of 95%, which are comparable to values achievable with thin-layer Pap smear screening. The authors conclude that methylation analysis is a potential diagnostic tool for cervical cancer screening.

    Contact: Susan Gregg-Hanson, Harborview, 206-731-4097, sghanson@u.washington.edu

    Higher Tissue Zinc Concentration Associated With Reduced Risk of Esophageal Cancer

    Higher tissue concentrations of zinc are associated with a reduced risk of esophageal squamous cell carcinoma, according to a new study.

    Animal studies have found an association between zinc deficiency and the risk of esophageal cancer. However, human studies have been complicated by the need to measure zinc concentration in esophageal tissue directly because the mechanisms by which zinc deficiency increases the risk of esophageal carcinogenesis are thought to occur locally in the target tissue.

    To determine whether concentrations of zinc and other elements influence esophageal cancer risk, Christian C. Abnet, Ph.D., M.P.H., of the National Cancer Institute, and colleagues used X-ray fluorescence spectroscopy to directly measure zinc, copper, iron, nickel, and sulfur concentrations in esophageal tissue sections from residents of Linzhou, China. They found that the risk of developing esophageal cancer over 16 years of follow-up was much lower for subjects with higher concentrations of zinc in esophageal tissue than for those with lower concentrations.

    Contact: National Cancer Institute Press Office, 301-496-6641, NCIPressOfficers@mail.nih.gov

    Also in the February 16 JNCI:

  • Study Examines Effect of Managed Care on Quality of Care for Cancer Patients: http://www.eurekalert.org/emb_releases/2005-02/jotn-see021005.php
  • Hepatitis B Viral Load, Genotype Affect Liver Cancer Risk, Study Finds: http://www.eurekalert.org/emb_releases/2005-02/jotn-hbv021005.php
  • Studies Examine Coffee Drinking and Risk of Liver and Colorectal Cancers: http://www.eurekalert.org/emb_releases/2005-02/jotn-sec021005.php

    ###

    Note: The Journal of the National Cancer Institute is published by Oxford University Press and is not affiliated with the National Cancer Institute. Attribution to the Journal of the National Cancer Institute is requested in all news coverage. Visit the Journal online at http://jncicancerspectrum.oupjournals.org/.


  • Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.