Raman Molecular Imaging Differentiates Epithelioid Mesothelioma from Metastatic-to-Pleura Bronchogenic Adenocarcinoma.
Andrew M Schreiner, John S Maier, Amy J Drauch, Subroto Paul. Weill Cornell Medical College, New York, NY; ChemImage Corporation, Pittsburgh, PA
Background: Raman molecular imaging (RMI), based on Raman spectroscopy, allows for reagentless, nondestructive molecular analysis of biologic samples. RMI produces a digital image with contrast resulting from laser light scatter. Epithelioid mesothelioma (EM) and metastatic-to-pleura bronchogenic adenocarcinoma (MAC) have similar morphology by light microscopy. We describe a pilot study using RMI to differentiate EM from MAC.
Design: Five cases of EM (3 women, 2 men; mean age 66 years) and 5 cases of MAC (2 women, 3 men; mean age 62 years) were selected based on classic clinicopathologic findings. Unstained thin sections from formalin-fixed paraffin-embedded tissue were prepared on aluminum-coated slides. RMI experiments were acquired on each sample using a Falcon II™ Raman microscope. Principal components analysis (PCA) was employed to reduce the data to a set of principal components (PCs) and to determine differences between the sample groups.
Results: Comparing the mean Raman spectra of EM and MAC demonstrated the greatest variance between 1200 and 1410 cm-1. Using this range, a PCA scatter plot showed a significant difference along PC2.
A leave-one-out cross validation for classification based on the weight in PC2 from a given case was done to estimate classification performance of RMI. Nine of 10 samples classified correctly, a sensitivity of 100% and a specificity of 80% for identifying MAC. Digital stains created with RMI data were mapped to corresponding H&E-stained images.
Conclusions: This pilot study shows a statistically significant difference in Raman scattering signatures between EM and MAC. A larger case study is warranted to validate our observations. RMI is a potentially useful diagnostic methodology and may be applicable in other organ systems.
Wednesday, March 2, 2011 9:30 AM
Poster Session V # 216, Wednesday Morning