Abstract #5418: Serum based immunoassay for lung cancer (LC) detection

LC is a leading cause of cancer death, accounting for 29% of cancer mortality. At present, there is no recommended screening method, and symptomatic patients are referred to chest X ray or CT to detect pulmonary nodules that need further confirmation via CT, PET scan or biopsies. CT is costly and has a high rate of incidental nodule detection. An easy-to-use immunodiagnostic assay would have clinical applications in LC diagnosis and screening. We have developed a proprietary approach for the discovery of cancer biomarkers (Bm) based on: i) an in-house generated library of monoclonal antibodies (mAb), ii) a large collection of clinical samples, and iii) our multiplex screening platform technology (Matrix Protein Array Technology, MPAT). We have applied this approach to the discovery and selection of mAb and Bm with utility for LC detection in serum. The MPAT enables us to screen a large number of clinical samples with a large number of antibodies. Briefly, a protein matrix comprising samples from cancer patients, benign and normal controls, is replica printed on the MPAT solid support. Each sample matrix is simultaneously interrogated with a different mAb. mAb-Bm reaction is detected using fluorescence-based Odyssey imaging system (Li-Cor), followed by spot quantification and statistical analysis to select Bm that are significantly and differentially expressed in cases versus controls. We first examined by MPAT mAb in tissue protein extracts derived from patients with LC (156 early and 43 late stage), benign (15), and normal controls (208), in addition to breast, colon, and ovary cases and controls, amounting to 1329 samples. A number of mAb-based Bm show excellent discriminatory power between cancer and control, with up to 90% sensitivity at 90% specificity, as determined by ROC curves using statistical package GB-STAT (Dynamic Microsystems). Immunohistochemistry on in-house and commercial tissue microarrays (75 cases) confirmed cancer specificity, and mAb specifically stained LC tissues, with membrane, nuclear and cytoplasmic localization. Some candidate Bm appeared to be secreted and overexpressed in LC patient serum, when tested by Western blot. Selected LC specific mAbs were further tested by antibody capture immunoassay with 250 clinical samples comprising 64 early stage non small cell lung carcinoma (24 adenocarcinoma, 24 squamous cell, and 16 large cell carcinoma) and 186 non-LC patient serum samples (41 normal including high risk smokers, 25 colon and 120 prostate cancers). mAb-Bm reactivity was detected by chemiluminescence using a CCD camera, and quantified using Scanalyze software. When comparing LC to non-LC group based on ROC curve analysis, most mAb sensitivity ranged from 60-80% at 80% specificity. We are also developing two-antibody ELISA sandwich assays using optimal capture and detection mAb pairs. Such immunodiagnostic assays could complement CT diagnosis and possibly aid in CT screening of high risk individuals.

Citation Information: In: Proc Am Assoc Cancer Res; 2009 Apr 18-22; Denver, CO. Philadelphia (PA): AACR; 2009. Abstract nr 5418.