We strive to innovate, simplify, and improve early diagnostic tools.
Research and Development
Lung Cancer Proteomics has spent 13 years in researching and developing a medical device that is non-invasive and can detect early stage lung cancer. We have evaluated over 110 commercially available protein biomarkers, 5,400 human plasma samples, and countless algorithms and machine learning methods in our search for a clinically useful product that detects early stage lung cancer with high accuracy, sensitivity and specificity. Below we present a summary of a blind study that validated the LCDT1 as a clinically useful, detection method for early stage non-small cell lung cancer.
Validation Study
A blind study was performed using a total of 228 Subjects processed in duplicate to yield 456 measurements (Table 1). Samples consisted of African-Americans, Caucasians, and Hispanics, and originated from the United States. Samples were randomized with the cohorts distributed evenly across the total plates of the study using R and Python. The assay service was performed at Eve Technologies in Canada.
Table 1. Blind Study Sample Size
By Pathology, Gender, Age
Table 2. Blind Test Results
For this test, only two classifications are predicted: Non-NSCLC or NSCLC.
The ROC Curve
The ROC (Receiver Operating Characteristic) Curve for the LCDT1 have an AUC (Area Under the Curve) of 0.96 which improved to 0.97 when other non-NSCLC cancers were removed from analysis.
Intra-assay precision was not performed. Inter-assay precision were between 1.5-13.1% and 3.8-10.3% for all analyte using MFI (Mean Fluorescence Intensity) and concentration CV (coefficient of variation), respectively.
The samples were evaluated for 19 proteins along with our locked proprietary algorithm. The results of the blind study was an accuracy of 95.6%, sensitivity of 89.1%, and specificity of 97.7% (Table 3). When other non-NSCLC cancers were removed from analysis, the specificity improved to 99.1%.
Summary
The blind test results indicate that the LCDT1 has great potential for clinical use in diagnosing early stage lung cancer with high accuracy, specificity, and sensitivity.