Su1488 Laser-Induced Fluorescence to Distinguish Adenomatous From Non-Adenomatous Colorectal Polyps

Abstract

Laser-Induced Fluorescence to Distinguish Adenomatous From Non-Adenomatous Colorectal Polyps Carsten Schmidt*, Iver Petersen, Andreas Stallmach Department of Gastroenterology, Hepatology and Infectiology, University Clinic Jena, Jena, Germany; Institute of Pathology, University Clinic Jena, Jena, Germany Aim: Novel endoscopic techniques increase detection rates of polypoid lesions, but differentiation of neoplastic and non-neoplastic tissue (inflammation, pseudopolyps, e.g.) may be difficult based on the endoscopists interpretation alone. So far, no algorithm has been established in clinical practice to interpret these data on an objective basis. However, previous studies have shown the potential of laser-induced fluorescence (LIF) to distinguish between neoplastic and non-neoplastic polyps of the colon more objectively. Materials and Methods: The WavSTAT Optical Biopsy System employs a 337 nm pulsed nitrogen laser excitation source that produces light energy that is delivered to the targeted tissue via an optical fiber inserted through the biopsy/suction channel of an endoscope. The light energy produces the fluorescence signature of the tissue. The same optical fiber then returns the fluorescence signature for characterization, analysis and display of results (suspicious or non-suspicious). We examined 35 consecutive patients who presented with known or suspected colorectal polyps. Histopathological results of biopies and resected polyps served as the gold standard. Laser-induced fluorescence results was applied to normal mucosa and to polypoid structures, resp., and sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and accuracy were calculated. Results: Measurements of 48 neoplastic polyps and 51 non-neoplastic polyps (16) and normal mucosa (35) from 35 patients were performed. In 19 measurements no sufficient data were obtained to allow interpretation of the respective tissue as suspicious or non-suspicious for adenoma. 18 of these errors occurred in adenomatous polyps, one in a hyperplastic polyp. Excluding these data from the analysis LIF yielded the following sensitivity 72.7%, specificity 93.6%, PPV 88.9%, NPV 83.0%, accuracy 85.0%. Combining the endoscopists impression with LIF the following results were obtained: sensitivity 95.8%, specificity 76.5%, PPV 79.3%, NPV 95.1%, accuracy 85.9%. Please see table 1 for a summary of these data. Conclusions: LIF is a promising tool in order to interprete polypoid structures more objectively. However, the currently available algorithm yields uninterpretable data in a considerable percentage of measurements, probably due to an invasion of the probe in submucosal areas where an insufficient number of signals is obtained. Combining the endoscopists impression with data from LIF measurements can distinguish neoplastic from non-neoplastic tissues more accurately with a NPV of 95.1% and may therefore reduce the number of unnecessary polypectomies with potential side-effects.