Using Laser Raman Spectroscopy to Reduce False Positives of Autofluorescence Bronchoscopies: A Pilot Study

Abstract

Preneoplastic lesions of the bronchial tree have a high probability of developing into malignant tumors. Currently, the best method for localizing them for further treatment is a combined white light bronchoscopy (WLB) and autofluorescence bronchoscopy (AFB) (WLB + AFB). The average specificity from large clinical trials for this combined detection method is approximately 60%, leading to many false positives. The object of this study is to determine whether adding point laser Raman spectroscopy (LRS) to a WLB + AFB has the potential to improve the specificity of preneoplastic lesion detection and what the implication is to the detection sensitivity. An LRS system was developed to collect real-time, in vivo lung spectra with a fiber optic catheter passed down the instrument channel of a bronchoscope. WLB + AFB imaging modalities were used to identify lesions from 26 subjects, from which 129 Raman spectra were measured. Multivariate statistical analyses were performed on the spectra with a leave-one-out crossvalidation. Clear in vivo Raman spectra were obtained in 1 second. The location of individual Raman peaks in the spectra correlated well with the known positions of Raman peaks generated by lipids, proteins, and water molecules. Preneoplastic lesions were detected with a sensitivity of 96% and a specificity of 91%. Adding point LRS analysis to WLB + AFB imaging has the ability to detect preneoplastic lesions in real time with high sensitivity and specificity. The use of LRS has great potential for substantially reducing the number of false-positive biopsies associated with WLB + AFB with very little reduction in the detection sensitivity.