Abstract 3566: Multimodal optical imaging for detection of oral neoplasia

Abstract

Abstract Despite advances in treatment methods the 5-year survival rate for patients with oral cancer in the United States is still only 61%, in part due to late stage diagnosis and subsequent difficulty of surgical treatment for advanced disease. To improve the ability of clinicians to detect early stage disease and to treat advanced cancers, we developed a multimodal optical imaging system (MMIS) to evaluate tissue structure in vivo at macroscopic and microscopic scales. The MMIS noninvasively measures changes associated with the development of precancer and cancer, including alterations in tissue autofluorescence properties (due to loss of collagen integrity, epithelial thickening, metabolic changes, and microvascularization) and alterations in tissue morphology (nuclear enlargement and crowding within the epithelium). The macroscopic imaging component of the MMIS acquires autofluorescence images of tissue with an excitation wavelength of 405 nm, a field of view of 45 mm, and spatial resolution of 0.1 mm. Loss of autofluorescence intensity is quantified by calculating the normalized red-to-green ratio in widefield autofluorescence images. Regions of tissue that display reduced blue-green autofluorescence are examined further using the high-resolution imaging component of the MMIS, which is a compact fluorescence microscope coupled to a fiber-optic imaging probe with a 720 μm diameter field-of-view and 4.4 μm spatial resolution. High-resolution imaging is performed at an excitation wavelength of 455 nm following topical application of proflavine, a fluorescent contrast agent which predominantly stains cell nuclei. Alterations in nuclear size and crowding are quantified by calculating the mean nuclear-to-cytoplasm ratio in high-resolution images. The MMIS was used to measure 100 sites in 30 patients undergoing surgery for oral cancer or precancer. Histopathology results from tissue specimens collected from the measured sites were used as the gold standard. Tissue specimens with a histopathologic diagnosis of dysplasia (mild, moderate, or severe) were further evaluated for molecular alterations by immunohistochemical staining for the markers Ki-67, p63 and PHH3. The MMIS correctly classified 98% of pathologically confirmed normal tissue sites and 95% of sites graded as moderate dysplasia, severe dysplasia, or cancer. MMIS measurements also correlated with molecular markers in 87% of sites with mild dysplasia. Following stratification of mild dysplasia sites by biomarker status, multimodal optical imaging classified all 100 measured sites with 93% sensitivity and 96% specificity. These findings support the ability of MMIS to guide the clinician in identifying neoplastic tissue, and potentially those sites with elevated risk of malignant transformation. This would have considerable impact on the detection and treatment of oral cancer and other epithelial malignancies. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 103rd Annual Meeting of the American Association for Cancer Research; 2012 Mar 31-Apr 4; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2012;72(8 Suppl):Abstract nr 3566. doi:1538-7445.AM2012-3566