Microscopic fluorescence spectral analysis of basal cell carcinomas

Abstract

Background and Objectives. Laser-induced autofluorescence (LIAF) is a promising tool for cancer diagnosis. This method is based on the differences in autofluorescence spectra between normal and cancerous tissues, but the underlined mechanisms are not well understood. The objective of this research is to study the microscopic origins and intrinsic fluorescence properties of basal cell carcinoma (BCC) for better understanding of the mechanism of in vivo fluorescence detection and margin delineation of BCCs on skin patients. A home-made micro- spectrophotometer (MSP) system was used to image the fluorophore distribution and to measure the fluorescence spectra of various microscopic structures and regions on frozen tissue sections. Materials and Methods. BCC tissue samples were obtained from 14 patients undergoing surgical resections. After surgical removal, each tissue sample was immediately embedded in OCT medium and snap-frozen in liquid nitrogen. The frozen tissue block was then cut into 16-&mgr;m thickness sections using a cryostat microtome and placed on microscopic glass slides. The sections for fluorescence study were kept unstained and unfixed, and then analyzed by the MSP system. The adjacent tissue sections were H&E stained for histopathological examination and also served to help identify various microstructures on the adjacent unstained sections. The MSP system has all the functions of a conventional microscope, plus the ability of performing spectral analysis on selected micro-areas of a microscopic sample. For tissue fluorescence analysis, 442nm He-Cd laser light is used to illuminate and excite the unstained tissue sections. A 473-nm long pass filter was inserted behind the microscope objective to block the transmitted laser light while passing longer wavelength fluorescence signal. The fluorescence image of the sample can be viewed through the eyepieces and also recorded by a CCD camera. An optical fiber is mounted onto the image plane of the photograph port of the microscope to collect light from a specific micro area of the sample. The collected light is transmitted via the fiber to a disperserve type CCD spectrometer for spectral analysis. Results. The measurement results showed significant spectral differences between normal and cancerous tissues. For normal tissue regions, the spectral results agreed with our previous findings on autofluorescence of normal skin sections. For the cancerous regions, the epidermis showed very weak fluorescence signal, while the stratum corneum exhibited fluorescence emissions peaking at about 510 nm. In the dermis, the basal cell island and a band of surrounding areas showed very weak fluorescence signal, while distal dermis above and below the basal cell island showed greater fluorescence signal but with different spectral shapes. The very weak autofluorescence from the basal cell island and its surrounding area may be attributed to their degenerative properties that limited the production of collagens. Conclusions. The obtained microscopic results very well explain the in vivo fluorescence properties of BCC lesions in that they have decreased fluorescence intensity compared to the surrounding normal skin. The intrinsic spectra of various microstructures and the microscopic fluorescence images (corresponding fluorophore distribution in tissue) obtained in this study will be used for further theoretical modeling of in vivo fluorescence spectroscopy and imaging of skin cancers.