Diagnosis of lung cancer based on direct-infusion electrospray mass spectrometry of blood plasma metabolites

Lung cancer is one of the most prevalent types of cancer in men and women, and is a leading cause of cancer-related death. Early detection of lung cancer may profoundly reduce cancer death rates. It is therefore extremely important to develop laboratory tests to detect human lung cancer, including at clinically asymptomatic stages. To this end, mass spectrometric metabolite analysis was performed on blood samples collected from patients with lung cancer (N = 100) and age-matched controls (N = 100). Proteins were extracted from blood plasma with methanol, and the remaining metabolite fractions were directly analyzed using electrospray ionization (ESI) mass spectrometry. Mass spectra obtained were converted into binary format, aligned, reduced to several variables by principal component analysis (PCA), and finally, classified as cancer cases versus controls by a support vector machine (SVM) algorithm. Repeated random sub-sampling validation revealed an accuracy of classification as high as 93.3% (sensitivity 94.1%, selectivity 92.4%), strongly indicating that direct-infusion ESI mass spectrometry of blood plasma metabolites offers great clinical potential in the diagnosis of early-stage human lung cancer.

Figure optionsDownload high-quality image (205 K)Download as PowerPoint slideHighlights
► Diagnostic power of blood plasma metabolites was investigated.
► Blood plasma metabolites were analyzed using direct-infusion ESI mass spectrometry.
► Mass spectra were used for samples classification.
► Accuracy of classification of lung cancer samples was 93.3% (sensitivity 94.1%, selectivity 92.4%).