Hypericin-based fluorescence diagnosis of bladder carcinoma M.-A. D’HALLEWIN, A.R. KAMUHABWA*, T. ROSKAMS†, P.A.M. DE WITTE* and L. BAERT Department of Urology, UZ Gasthuisberg, *Laboratory of Pharmaceutical Biology and Phytopharmacology, and †Department of Pathology, UZ Gasthuisberg, KU Leuven, Belgium Objective To determine the use of hypericin instillation for the fluorescent detection of papillary bladder cancer and carcinoma in situ. Patients and methods Eighty-seven patients with papil- lary bladder cancer and/or carcinoma in situ received instillations with 40 mL of an 8 mmol/L hypericin solution for at least 2 h. Fluorescent excitation with blue light was effective for up to 16 h, and biopsies were examined by fluorescence microscopy. Results There were no side-effects reported, no photo- bleaching and all papillary lesions fluoresced red. The sensitivity and specificity for detecting carcinoma in situ was 94% and 95%, respectively. An interval of 4 months is recommended after BCG instilla- tions before using this test. Fluorescence microscopy showed that hypericin was selectively localized in the epithelium. Conclusions Hypericin-induced fluorescence has a high sensitivity and specificity for detecting bladder cancer. After 4 months there are few false-positive results in patients treated with BCG. Keywords bladder carcinoma, hypericin, fluorescence detection Introduction It is now commonly accepted that there are two separate pathways of carcinogenesis in bladder cancer, with a different morphological outcome from the start, resulting in papillary noninvasive tumours that rarely progress and flat carcinoma in situ (CIS) that will progress to invasive disease [1]. TUR of superficial papillary tumours will only cure 20–50% of the patients; the remainder will develop one or more recurrences within a year [1,2]. This is partly attributed to the natural history of the disease, but small tumours can be overlooked at the initial resection [3]. Hudson and Herr [4] recently stated that CIS may have a long- or short-term in situ phase before progression to invasion, the duration of which may be several years. However, CIS is inexorably pro- gressive and will become invasive if uncontrolled and given enough time. The early recognition of superficial bladder cancer is therefore mandatory, to offer the patients the highest possible cure rate. The first reports by Kriegmair et al. [5] of fluorescence-guided endoscopy showed the superiority of this technique over white-light endoscopy. However, with this technique a valid inter- pretation and reliable diagnosis is hampered because there are many false-positive results, showing its relative lack of specificity. In this study we evaluated the use of hypericin, a potent photosensitizer (a hydroxylated phenantro- perylenequinone) present in several plants of the genus Hypericum, the most common of which is H. perforatum [6], as a selective tumour marker for bladder cancer. The oral administration of Hypericum extract (daily dose 1–2.5 mg) is used clinically as an antidepressive agent, with no side-effects [7]. Patients and methods Hypericin was synthesized as previously described by Falk et al. [8]; hypericin is insoluble in water and there- fore plasma protein was used as an effective hypericin transporter/carrier. The solution for bladder instilla- tion was prepared as follows: 16 mmol/L hypericin in absolute ethanol was diluted 1000-fold in a 1% plasma protein solution in buffered saline and sterilized by membrane filtration. Aliquots of 20 mL were kept frozen until use and 20 mL saline added to obtain a 40-mL instillation solution containing 8 mmol/L hypericin. The study comprised 87 patients, including 52 with visible papillary disease on white-light endoscopy, 12 with positive urinary cytology but no abnormality on urography and white-light endoscopy, six treated 2 months after BCG instillation for CIS and/or T1G3 disease, 15 treated after o 4 months and two treated after radiotherapy. The instillation procedure was described previously [9]. The system used for fluorescence diagnostics was the D Light system (developed by Storz Company, Germany) and consists of a xenon-arc lamp with a band-pass filter Accepted for publication 10 January 2002 BJU International (2002), 89, 760–763 # 2002 BJU International 760