Differential diagnosis of human lung tumors using surface desorption atmospheric pressure chemical ionization imaging mass spectrometry

Abstract

Distinguishing tumors from normal tissue is a key component in lung cancer-conserving surgery. In this study, accurate diagnosis of human squamous cell carcinoma lung cancer in untreated tissue sections is achieved by ambient mass spectrometry imaging using liquid-assisted surface desorption atmospheric pressure chemical ionization mass spectrometry (DAPCI-MS) combined with multivariate statistical analysis. DAPCI-MS imaging shows great promise as a molecular pathology technique that uses the phosphatidylcholine (PC) and sphingomyelin (SM) profiles of tissues to visualize and differentiate lung cancer from normal tissue, and the use of multivariate statistical analysis significantly increases the confidence of this diagnosis through data interpretation. Multivariate statistical analysis has played an important role in biomarker discovery and lung cancer diagnosis, highlighting the need for the use of multivariate statistical analyses to reduce the high-dimensional mass spectral data. Partial least-squares linear discriminate analysis (PLS-LDA) was successfully used for visualization and classification of 14 tissue pairs (28 tissue samples) using the full scan mass spectra data, only with a misclassification rate of 2.16% determined from the validation set. Multiple distinctive PC and SM species between the tumor and non-tumor samples derived from massive full scan mass spectral data using PLS-LDA were tentatively identified by individual ion images, compared with the pathological examination of the hematoxylin and eosin (H&E) stained tissue sections. A significant increase in multiple phosphatidylcholines (PCs) and a decrease in several specific sphingomyelins (SMs), particularly as well as increased levels of choline (C5H14NO+) were uncommonly observed in tumor regions with respect to adjacent noncancerous areas. These could be the signature compounds and have the largest possibility as potential biomarker compounds for identifying and differentiating tumor regions from adjacent normal tissue sections. Overall, the DAPCI-MSI combined with multivariate statistical analysis provides an effective tool for direct ambient analysis of such complex heterogeneous lung tissue samples, which has great potential in the application of intraoperative tumor assessment.