Abstract
To the Editor: Pass et al. (Oct. 11 issue)1 report on fibulin-3 as a new soluble biomarker for malignant mesothelioma. The diagnostic accuracy of plasma levels of fibulin-3 in the Detroit and New York cohorts (area under the receiver-operating-characteristic curve [AUC], 0.99) exceeded that of any previously reported marker. In the external validation cohort, however, the diagnostic accuracy surprisingly decreased (AUC, 0.87). This marked decrease was probably due to differences in the study population, although the study showed no association between plasma fibulin-3 levels and common patient characteristics. To fully comprehend the potential and limitations of this new biomarker, researchers . . .

This publication has 7 references indexed in Scilit: