A Tissue Section-Based Near-Infrared Spectroscopical Analysis of Salivary Gland Tumors

SIMPLE SUMMARY: Salivary gland tumors (SGTs) are a group of rare tumors that vary in clinical and histological behavior. Histological classification is difficult and requires an experienced pathologist. Based on recent research in other medical fields, this study investigated the value of near-infrared (NIR) spectroscopy in the diagnosis of SGTs supplementary to the histological investigation. The acquired spectra were analyzed with chemometric techniques. Enzymatic treatment (neuraminidase) resulted in spectral peak differences between healthy controls and different SGT types. Some malignant SGTs had higher spectral changes, suggesting bigger alterations in glycosylation of salivary mucins. Future biochemical research based on the further enzymatic dissection of SGTs and infrared spectroscopy could help pathologists to better understand the nature of these types of tumors. ABSTRACT: SGTs vary in histological behavior. Mucins, a major component in salivary glands, consist of a glycosylated and sialylated protein core. Rapid evolutions in glycobiology have demonstrated the important role of glycoproteins in cancer development. NIR spectroscopy is a method for the biochemical analysis of substrates. NIR spectra can be analyzed using specific chemometrics. Our aim was to explore the diagnostic possibilities of NIR spectroscopy in SGTs. 238 Hematoxylin and Eosine stained (H&E) SGT tissue sections were examined using NIR spectroscopy. 45 deparaffinized tissue sections were treated with neuraminidase to identify wavelengths in the NIR spectrum related to sialylation. NIR spectra were analyzed with chemometrics. NIR spectra could distinguish malignant SGTs from controls and benign SGTs. Prediction models based on the entire spectral range resulted in a 73.1% accurate classification of malignant SGTs and controls, while, based on neuraminidase experimental spectral peak differences (1436 nm; 1713 nm; 1783 nm; 1924 nm; 2032 nm; 2064 nm; 2178 nm; 2216 nm), an improved overall correct classification rate of 91.9% was obtained between healthy subjects and malignant tumors. H&E tissue section-based NIR spectroscopy can identify malignant SGTs from controls, promising an alternative method in the diagnosis of SGTs.