Spectroscopic Measurements and Fluorescence Imaging for Treatment and Diagnosis of Skin Cancer

This thesis deals with the use of light for treatment and diagnosis of skin cancer. The occurrence of skin cancer is increasing worldwide, among the most common types being basal cell carcinoma, squamous cell carcinoma and actinic keratosis, a skin cancer precursor. These types of lesions today constitute a heavy burden to the medical system, and hence the need for effective treatments and accurate diagnostic tools is great. Photodynamic therapy (PDT) is based on an initial photosensitisation of the tumour, followed by irradiation with visible light, inducing a tissue-toxic photochemical reaction killing the cells. By application of delta-5-aminolaevulinic acid (ALA), the photosensitisation is obtained by elevated production of endogenous protoporphyrin IX (Pp IX). In addition to the therapy, the fluorescence from Pp IX has proved to be useful for diagnostic purposes in so-called fluorescence imaging. The full mechanism involved in PDT is not yet clear. The work in this thesis deals with some of the questions remaining. The photobleaching has been studied in solution as well as clinically. It was found that lower fluence rates resulted in higher bleaching rates and preferable treatment outcome. This seems related to the consumption of oxygen in tissue during irradiation. Moreover, the pain experienced by the patients was observed to be related to the size of treatment area, location of lesion and the photodynamic reaction itself. As for fluorescence imaging, the technique has hitherto not gained clinical acceptance. The selectivity of the Pp IX fluorescence has been shown to be time-dependent. In a clinical investigation the highest contrast between tumour and surrounding normal skin was found after 3 hours of ALA application. Furthermore, preferential results were obtained when using multispectral fluorescence imaging combining the ALA-induced fluorescence with endogenous tissue fluorescence, so-called autofluorescence, controlled by histopathological mapping. The discriminability between tumour and normal skin was further improved by implementing texture analysis and linear discrimination