Format

Send to

Choose Destination
J Thorac Oncol. 2006 Jul;1(6):513-9.

Profiling tumor-associated antibodies for early detection of non-small cell lung cancer.

Author information

1
Department of Internal Medicine, Division of Pulmonary and Critical Care Medicine, University of Kentucky, Lexington, KY 40536, USA. lzhon2@uky.edu

Abstract

BACKGROUND:

A blood test for non-small cell lung cancer (NSCLC) may be a valuable tool for use in a comprehensive lung cancer screening strategy. Here we report the potential of autoantibody profiling to detect early-stage and occult NSCLC.

METHODS:

T7-phage NSCLC cDNA libraries were screened with patient plasma to identify phage-expressed proteins recognized by tumor-associated antibodies. Two hundred twelve immunogenic phage-expressed proteins, identified from 4000 clones, were statistically ranked for their individual reactivity with 23 stage I cancer patient and 23 risk-matched control samples. All 46 samples were used as a training set to define a combination of markers that were best able to distinguish patient from control samples; this set of classifiers was then examined using leave-one-out cross-validation. Markers were then used to predict probability of disease in 102 samples from the Mayo Clinic CT Screening Trial (six prevalence cancer samples, 40 drawn 1 to 5 years before diagnosis, and 56 risk-matched controls).

RESULTS:

Measurements of the five most predictive antibody markers in 46 cases and controls were combined in a logistic regression model that yielded area under the receiver operating characteristics curve of 0.99; leave-one-out validation achieved 91.3% sensitivity and 91.3% specificity. In testing this marker set with samples from the Mayo Clinic Lung Screening Trial, we correctly predicted six of six prevalence cancers, 32 of 40 cancers from samples drawn 1 to 5 years before radiographic detection on incidence screening, and 49 of 56 risk-matched controls.

CONCLUSIONS:

Antibody profiling may be a useful tool for early detection of NSCLC.

PMID:
17409910
[Indexed for MEDLINE]
Free full text

Supplemental Content

Full text links

Icon for Elsevier Science
Loading ...
Support Center