Towards improved bladder cancer diagnosis using fluorescence imaging and Raman spectroscopy

Abstract

Bladder cancer is the fourth most common type of cancer worldwide. Its high recurrence rate makes bladder cancer one of the most prevalent types of cancer in the western world and the most costly type of cancer over the patient’s lifetime. In the Netherlands, each year 5,400 new patients with bladder cancer are diagnosed. Approximately 75% of these suffer from non-muscle invasive bladder cancer. These patients have a good prognosis regarding life expectancy. However, 40-80% will develop at least one tumor recurrence within 12 months after primary treatment. Disease progression, though relatively rare (10-15%), to muscle invasive or disseminated disease will severely affect their prognosis. Evaluation of the bladder wall and biopsy of suspect lesions of the bladder are performed during transurethral resection of bladder tumor (TURBT). Histopathology of the acquired biopsy samples is used to establish the final diagnosis of bladder cancer. Frequent reexamination of the bladder and repeated TURBT of suspicious lesions place a significant burden on the public healthcare system. Optimization of this procedure by enhancing early detection will directly affect patient outcome as well as the costs associated with bladder cancer. Fluorescence cystoscopy, as an extension to ordinary visual bladder inspection to improve the detection, has slowly been introduced in urology. It considerably enhances the sensitivity of bladder cancer detection, but suffers from a mediocre specificity. Raman spectroscopy is an optical technique that examines the interaction of light with molecular bonds, which has many applications in the identification of substances and processes because of its inherent specificity. The aim of this thesis is to optimize the clinical application of fluorescence imaging and to assess the place for optical diagnosis by means of Raman spectroscopy. The ultimate goal is to enhance bladder cancer diagnosis, which in the long run should result in the possibility to perform an optical biopsy without the need for confirmation by histopathology. Clinical observations in fluorescence imaging led to an evaluation of the differences between the initial and subsequent fluorescence cystoscopy procedures particularly with respect to its specificity. Raman spectroscopy proved not only able to distinguish between benign and malignant tissues of the bladder, but also allowed prediction of the grade and stage of the lesion in vitro. An evaluation of prerequisites for in vivo fluorescence guided Raman spectroscopy showed this a promising tool to enhance the specificity of bladder cancer detection.