Changes in the dielectric spectra of murine colon during neoplastic progression

Current colorectal cancer prevention and management rely on the endoscopic biopsy-based diagnosis of premalignant (dysplastic) and malignant (carcinomatous) neoplastic lesions. Due to substantial limitations in presently available endoscopic modalities, there is great demand for developing new diagnostic tools. Electrical impedance spectroscopy (EIS), which detects differences in dielectric properties between neoplastic and non-neoplastic tissues, may offer a promising alternative modality for endoscopic detection of neoplasia. In this study, we use EIS across the 10 kHz-10 MHz frequency range to characterize the dielectric properties of small intestinal and colonic dysplastic lesions (adenomas) in a well-established murine model of intestinal neoplasia, the APC^Min mouse. Impedance data from 22 adenomas (14 with low-grade dysplasia and 8 with high-grade dysplasia), as well as 11 different foci of normal mucosa adjacent to the adenomas, were obtained by direct placement of a specially-modified coaxial probe onto the mucosal surface of freshly-excised segments of murine intestine. Cole-Cole analysis was used to decompose the impedance data into a set of independent parameters, which were correlated with the results of histopathological evaluation of the measured tissue samples. Of the impedance parameters evaluated, the low-frequency tissue permittivity, dispersion broadening parameter, and DC conductivity exhibited significant changes as a function of the progression from normal to low and high-grade dysplasia (p < 0.05). Specifically, low-frequency permittivity and DC conductivity increased, and the dispersion broadening parameter decreased. The ability of EIS to discriminate between normal and dysplastic mucosa, as well as distinguish the degree of dysplasia (low versus high) is highly promising, as it may offer the ability to instantly detect and risk-stratify lesions during endoscopy, particularly in patients whose endoscopic surveillance is challenging using currently available diagnostic modalities.