Lung cancer diagnosis and classification is typically accomplished by light microscopy and current classifications are of very little utility in predicting outcome or individualizing therapy. We have previously shown the utility of matrix-assisted laser desorption/ionization mass spectrometry (MALDI-MS) in classifying non-small cell lung cancer (NSCLC) from surgical frozen tissue sections (Yanagisawa, et al., 2003). However, frequently the only diagnostic samples available in clinical situations are brushings or fine needle aspirates, and these are usually heavily contaminated by normal cells making blind proteomic analysis impossible. In order to further develop mass spectrometry for use in the clinic to quickly and accurately diagnose and classify clinical samples, we have developed a method to acquire mass spectrometric data selectively from tumor cells in these clinical samples. We accomplish this by cytocentrifugation of clinical specimens onto transparent, metal-coated glass slides, followed by staining with a nuclear stain, allowing us to directly visualize the cancer cells in the sample. Application of a matrix solution directly to the cancer cells followed by targeted laser desorption for MALDI-MS results in the production of high quality protein profiles from these samples. Using this technique we can quickly analyze multiple clinical specimens including pleural effusions, brushes, washes, and fine needle aspirates. This technique, while developed for lung cancer in our laboratory, could be applied to similar biopsies taken from individuals for many other types of cancer and will be very useful in the clinical diagnosis, classification, and potentially, the individualized treatment of cancer patients.
[Proc Amer Assoc Cancer Res, Volume 46, 2005]