Quantitative studies of the kinetics of 5-aminolaevulinic acid-induced fluorescence in bladder transitional cell carcinoma.

Photodynamic therapy is a potential treatment for superficial bladder cancer that utilizes photosensitizer drugs, which are activated by light to cause tissue destruction. However, first-generation photosensitizers cause prolonged phototoxicity, have poor tumour specificity and can accumulate within detrusor muscle, resulting in permanent loss of bladder capacity following treatment. A newer drug, called 5-aminolaevulinic acid (ALA), generates a sensitizer called protoporphyrin IX (PpIX) in situ and has been shown, qualitatively, to be more tumour specific. The fluorescence kinetics of ALA-induced PpIX was investigated in patient biopsies of bladder tumour, normal urothelium and detrusor muscle, both in vitro after incubation of specimens in ALA-rich culture medium for various times and in vivo after instillation of intravesical ALA before endoscopic resection. The fluorescence in tumour tissue was twice that of normal urothelium in vitro and up to tenfold in vivo. There was little ALA-induced fluorescence in detrusor muscle, both in vitro and in vivo. Most importantly, no patients experienced phototoxicity or other adverse events following intravesical instillation of ALA.

Photodvnamic therapx (PDT) is a method for the treatment of cancer based on the svstemic or topical administration of drugrs called photosensitizers. which are activated in the presence of light to cause cell death and tissue destruction. Ideally. photosensitizers should be selectively retained by tumour. leaving adjacent tissue undamaged following, light activation. although in practice this has been difficult to attain (Bown. 1989).
Transitional cell carcinomas (TCCs) of the bladder are ideal for this type of treatment because. with careful control of the lirht distribution. curative doses can reach all parts of the urothelial surface (D'Hallewin et al. 1992). This is of importance in problematic superficial tumours. which are often multifocal. Areas of poorly defined dysplasia and carcinoma in situ are therefore amenable to effective photodynamic therapy.
First-generation sensitizers. such as haematoporphyrin derivative. are usually administered systemically. This results in the inevitable distribution of drug to various organs such as the skin.
with skin photosensitivity often exceeding 6 weeks (Doughertv et al. 1990). The most significant morbidity from photodynamic treatment of the bladder arises. however. from the relatixve lack of selectivity of haematoporphynn-based sensitizers between mucosa and detrusor muscle. Furthermore. intravesical administration is unreliable with these photosensitizers. and this also means that the transitional cell layer cannot act as a natural barrier to the uptake of Received 7 October 1997Revised 3 March 1998Accepted 10 March 1998 Correspondence to: SN Datta. Department of Urology. University Hospital of Wales, Heath Park. Cardiff CF4 4XW. UK these sensitizers (Benson. 1988). Partly as a result of this. photodynamic therapy of the bladder using first-generation sensitizers has been associated with serious damage to. and therefore functional impairment of. the detrusor muscle (Nseyo et al. 1985).
The use of an endogenous substance. 5-aminolaev-ulinic acid (ALA). to generate the synthesis in situ of the pure porphyrin. protoporphyrin IX (PpLX) with minimal toxicity. represents a new strategy in the administration of photosensitizing drugs. ALA. a natural precursor of haem. is a low-molecular weight substance which is taken up by all nucleated cells. The immediate precursor of haem is PpIX and. because this final step is rate-limiting. exogenous ALA results in the accumulation of PpIX. which is an effective photosensitizer which can be exploited for photodynamic therapy. Intravenous administration of ALA results in rapid clearance from the body. with no detectable PpIX fluorescence in the skin or other organs after 24 h (Kennedy et al. 1991 ). It has been previously demonstrated (Steinbach et al. 1994). using qualitative measures. that intravesical ALA in human subjects can induce selective porphyrin accumulation within bladder tumour. compared with normal urothelium and detrusor muscle. Studies of intravesical ALA in the rat bladder confirm that selective accumulation of sensitizer was 11 times greater than in detrusor muscle (Chang et al. 1996a). The importance of mucosal selectivity was demonstrated by showing that subsequent photodynamic damage to the bladder wall was minimal (Chang et al. 1996b).
The purpose of this study was therefore to determine. quantitativ-ely. the fluorescence kinetics of ALA-induced PplX in human TCC. compared with normal urothelium and detrusor muscle in vitro. A further objective was to establish the kinetics of endooenous porphyrin induced by intravesical ALA in selected patients.
with a particular emphasis on recording adverse events. 1113

PATIENTS, MATERIALS AND METHODS
Before the instillation of intrax esical ALA into patients. an in vitro study of the kinetics of ALA-induced porphyrin fluorescence xxas performed on tissue samples taken from patients. This w as carried out to determine the approximate time for ALA-mediated PpIX fluorescence to reach a maximum and to determine the abilitv of different tissue samples to synthesize porphyrin photosensitizer.
In vitro study of ALA-mediated fluorescence kinetics Patients All studies involving patients had received local research ethics committee approval. Informed consent wxas obtained from all patients. Specimens for determining the fluorescence kinetics of ALA uptake were obtained from sexen patients with a prexious historof superficial transitional cell carcinoma of the bladder. Patients were admitted for endoscopic resection of their tumours after a recent diagnosis on flexible cystoscopx of tumour recurrence.

Collection of biopsy specimens
Before conxentional endoscopic resection. two or three cold-cup biopsies were taken from tumour. normal urothelium and detrusor muscle. These were dixided with a sterile scalpel such that there were four or fixe specimens of tumour. normal urothelium and detrusor muscle from each patient. The specimens were immediately transported to the laboratory in warm. sterile Hartmann's solution.
Incubation of specimens in ALA-rich culture medium Biopsy specimens were incubated in RPMI tissue culture medium (Imperial). supplemented with 10% fetal calf serum (FCS: Gibco BRL). 50 U ml penicillin/ 50 jgc ml-' streptomycin (Gibco BRL). 2 m-M L-glutarnine (Gibco BRL) and 1 nmL ALA (Signa) buffered to a pH of 7. In patients from whom five biopsy samples were available from each of the three tissue types. the fifth tissue sample was immediately snap frozen and used to determine tissue background fluorescence. The other four specimens were incubated for 2. 4. 6 or 24 h at 37 C in an atmosphere of 95% oxygen and 5%e carbon dioxide. Light exposure was minimiized to avoid bleachinc of photosensitizer. After incubation. biopsy material was embedded (Tissue Tek II embeddinc compound. BDH) and snap frozen in a bath of isopentane prechilled in liquid nitrogen. Paired frozen sections 10 jm thick wxere cut from each tissue block using a cryotome. One section from each pair was fixed and stained with haematoxylin and eosin. The other paired frozen section was left unstained and stored at -70 C until fluorescence microscopy and photometrvxwere carried out. All specimens were stored at -70cC until frozen sections were prepared and fluorescence microscopy and photometrv carried out. as described later.
In vivo clinical study of ALA-mediated fluorescence kinetics Patients Local research ethics committee approval was obtained. Permission was obtained from the UK Department of Health for instilling intraxesical ALA. under the prox isions of the Medicines (exemption from licences) (special cases and miscellaneous provisions) Order. 1972. Informed consent wxas obtained from ten patients with a previous known history of superficial transitional were taken for full blood count and baseline plasma porphyrin lexels and serum samples for urea and electrolyrtes and liver function tests. Total plasma porphvrins A ere determined spectrofluorometrically. In six patients. uroflowx parameters w-ere measured before instillation using a urine flowx meter. and the post-micturitional residual estimated using an ultrasound scanner (Bard Bladderscan).
Instillation of intravesical ALA Between 2 and 5 h preoperatively. 50 ml of 3%le ALA (Sigma) was instilled intravesical1v Xia a 12-F Lofric catheter. The solution w-as prepared by dissolvinc 1.5 g of ALA in 50 ml of physiological saline buffered to a pH of 6.5 with 8.4%sodium bicarbonate. The solution A as freshly prepared and sterilized by ultrafiltration at the Department of Pharmacy of the University Hospital. Cardiff. UK. Solutions of ALA were instilled w-ithin 3 h of preparation and refrigerated in the darkuntil used.

Fluorescence cystoscopy
Under spinal or general anaesthesia. all instilled ALA w as drained and subsequent cystoscopy carried out using sterile 1.5%e glNcine. Conventional cystoscopy A-ith w-hite light w-as briefly carried out. followed by fluorescence cystoscopy. This has previously been descnrbed to determine qualitatively the preferential nature of ALA fluorescence kinetics within tumour and for possible early detection of bladder cancer (Kriegmair et al. 1994(Kriegmair et al. . 1996a. Collection of biopsy specimens Using conventional cystoscopy. cold-cup biopsies W ere tak-en from each of tumour or abnormal urothelium. normal urothelium and detrusor muscle. These specimens A-ere immediately transported to the laboraton-. where tissue blocks were embedded (Tissue Tek II embedding compound. BDH) and snap frozen in a bath of isopentane prechilled in liquid nitrogen. Paired frozen sections 10 gm thick wxere cut from each tissue block using a cryotome.
One section from each pair A-as fixed and stained w-ith haematoxvlin and eosin. The other paired frozen section was left unstained and stored at -70-C until fluorescence microscopy and photometrx were carried out.
Post-operative care Approximately 1-5 h after instillation of ALA. a blood sample for plasma porphy-rin determination w-as collected from each patient. via a 14-G enous cannula inserted perioperatively. Approximately 24 h after instillation. further samples were taken from the venous cannula for full blood count. creatinine. electrolyrtes. liver function tests and further plasma porphyrin levels. Patients were usually discharged on the second post-operative day. Patients w-ere advised prospectively of possible phototoxicity and asked to record all adverse exents. Patients were reviewed in the outpatient clinic w ithin 2-3 w-eeks and adverse events noted. Further blood samples wxere taken for full blood count. creatinine. electrolytes and liver function tests. Urofloxx parameters were recorded in those patients w ith preoperative flowrate and post-micturition residual measurements. Bladder symptom scores (IPSS were recorded in all patients at this stace.

Fluorescence microscopy and photometry
Unstained frozen section slides wxere transported to the laboratory in drn ice and only allow ed to thaxx before fluorescence microscopy. Fluorescence microscopy was performed as described prexiouslI (Bedwxell et al. 1992: Loh et al. 1992l. An inverted microscope (Oix mpus IM-2) w ith epifluorescence and phasecontrast attachments vvas used. A 10 x objectixe w-as used to giVe images of 880 x 550 gm dimensions. Fluorescence A-as excited using an 8-mW helium neon laser operating at 632.8 nm. with the output directed onto a dichroic mirror in the epifluorecence microscope through a liquid light guide and xia a 10-nm bandpass filter to remove extraneous light. The advantage of using the helium-neon laser for excitation is its spectral puritx and the induction of less tissue autofluorescence. Exposure time w-as set at 25 s. using an excitation fluence of <1 J cm-'. Fluorescence w-as detected in the range 660-710 nm using a combination of bandpass (Omeca Optical) and longpass (Schott RG655) filters. Under these conditions. porphyrin photodegradation was negligyible. The fluorescence sianal was detected by a highlx sensitixe crvogenicallv cooled slow--scan charge-coupled dexice (CCD) camera of resolution 400 x 600 pixels (Wright Instruments) attached to the microscope. The signal w-as processed via an IBM personal computer into a colour-coded digital image of the section depicting mean signal counts per pixel. Fluorescence in terms of counts per pixel (four photoelectrons per count: quantum efficiencv 0.5 at this A-avelength wAas quantified digitally oxer at least three equal areas of interest for each section. Areas of interest w-ere chosen to be representatix e of the entire histological section. avoiding cold biopscrush artefacts. Both unstained and stained pairs of tissue sections were checked histologicallx using light microscopy. Autofluorescence of control specimens of bladder tumour. normal urothelium and detrusor muscle. which A-ere not exposed to ALA. A-as measured and this data w as subtracted from the relatixe fluorescence -alues of corresponding ALA-exposed tissue.

Results of in vitro fluorescence kinetics of ALAmediated accumulation of PpIX
Incubation of tissue samples in ALA resulted in an initial timedependent increase in tissue fluorescence for all three tissue ty pes over the first 2-6 h. Howexer. the increase in detrusor muscle tissue fluorescence w-as relatively small compared w-ith that of tumour and normal urothelial tissue. The peakfluorescence in tumour tissue was double that of normal urothelium and nearly six times that of detrusor muscle. Tumour fluorescence appeared to peak between 2 and 6 h of incubation in ALA. but. by 24 h. this had diminished to fluorescence levels similar to those of normal urothelium and detrusor muscle (Ficure 1).

Results of in vivo study
Results of fluorescence cystoscopy Of ten patients in whom intravesical ALA A-as administered. eight had evaluable superficial bladder tumours. Cystoscopx of these patients using the violet excitation light source resulted in bright red fluorescence of all tumours identified using conventional A hite light. In one patient. conventional cystoscopy revealed a flat. red area of bladder which did not fluoresce when the light source was sw itched. This area A-as biopsied and subsequent histological examination confirmed features consistent with chronic inflammation. In a second patient with a previous history of carcinoma in situ. however. w-hite light cystoscopy A-as unremark-able although an area near the trigone fluoresced brightly. This area w as biopsied

DISCUSSION
Photodvnamic therapy of bladder cancer was first described in 1976. although results were limited because of the lack of reliable light dosimetry (Kellv and Snell. 1976). The spheroid shape of the bladder also enables reasonably uniform light delivers and the abilitv for the whole urothelial surface to be treated simultaneouslv (Pope and Bow-n. 1991). This is of importance in resistant superficial tumours that are often multifocal. Areas of poorly defined dysplasia and carcinoma in situ do not. therefore. necessarilv need to be precisely defined for effective photodynamic therapy to take place. The results described in this paper are complementary to the pioneering work of Kriegmair and 5 6 7 8 colleagues (Kriegmair et al. 1994. 1996a  PpIX accumulation within the X arious tissues has been well described and A-as carried out using, fluorescence microscopy. utilizing a highly sensitive CCD camera (Bedwell et al. 1992: Regula et al. 1995). Loh et al (1993 were able to show. using in patients with carcinoma in situ (D'Hallewin and Marijnissen. 1995). with 100% initial complete response rates reported. The overall success of photodynamic therapy of the bladder using haematoporphyrin derivative and Photofrin is. therefore. of a similar order of magnitude to intra-esical chemotherap-. w-ith verencouraging results for carcinoma in situ. although no comparative studies exist in this respect. Furthermore. there are no long-term studies looking at lon2-term disease progression or survival.
Despite the encouraging results of limited early clinical studies. haematoporphyrin deri ative and Photofrin-mediated photodynamic therapy of bladder have remained experimental treatment modalities within the field of uro-oncologv. Much of this has been due to adverse events which. while usually not danaerous. often result in symptoms which have a profound effect on the quality of life. The use of an endogenous substance. 5-aminolaevulinic acid (ALA). to generate the synthesis in situ of the pure porphyrin. protoporphyrin IX (PpIX). therefore represents a newstrategy.
Intravenous administration of ALA results in rapid clearance from the body uwith no detectable PpIX fluorescence in the skin or other organs after 24 h (Kennedy and Pottier. 1992).
Despite the promising nature of the role of ALA-mediated photodynamic therapy of bladder tumours. little work has as vet been reported in this field. One group was able to induce necrosis of the urothelial layer of rat bladder using 50 J of laser light. but with minimal detrusor damage (Chang et al. 1996b). The opportunity of administering ALA intravesically not onlv allows the urothelium to act as a potential barrier to sensitizer accumulation within detrusor. but also theoreticallv reduces the risk of skin photosensitivity. assuming that there is minimal absorption of ALA by the transitional cell laver. It has been demonstrated qualitativelv that intravesical ALA in human subjects induces photosensitizer uptake within normal and neoplastic urothelium. with minimal detrusor accumulation (Steinbach et al. 1994). Using a krypton laser to induce fluorescence rather than tumour destruction. the same group has exploited the preferential accumulation of sensitizer within tumour as a diagnostic tool to enhance cystoscopic visualization of poorly defined dysplasias and carcinomas in situ (Kriegmair et al. 1994. 1 996a). However. these studies were confined to establishina the qualitative differences between ALAinduced fluorescence in tumour and non-tumour tissue. W'e have been able to reproduce the technique of fluorescence cystoscopy. and also determine. quantitatively. that ALA-mediated fluorescence is up to 11 times greater in transitional cell tumour compared with normal urothelium. The fluorescence intensitywithin the Various tumour specimens is summarized in Table 1. There does not appear to be a clear difference between patients with moderatelv differentiated or well-differentiated tumours.
To date. there is only one report of ALA-mediated photodynamic therapy of bladder cancer in patients ( Kriegmair et al. 1996b). This study is an early report of the treatment of ten patients with bladder cancer refractorto other treatment modalities. Different light doses and laser wavelengths were used in different patients after the instillation of 10`%1 ALA. but complete or partial responses were reported in six of these subjects. The ability of intravesical ALA to induce the preferential accumulation of photosensitizer within tumour tissue. with minimal adverse e-ents. is an important consideration in further development of photodynamic therapy of superficial bladder cancers. Superficial bladder cancers have long been shown to be a suitable target for photodynamic treatment. Early carcinoma in situ is a clear example of a potentially dangerous bladder cancer. with si5nificant recurrence rates despite intravesical chemotherapy. Often. radical cystectomy is the only solution for this scenario. but. with close to 100'% response rates. photodynamic therapy is clearly a promising solution for this particular problem. However. serious adverse events associated with first-generation sensitizers has meant that this treatment modality has remained experimental within the urological community. Intravesical ALA appears to generate the preferential accumulation of photosensitizer w-ithin bladder tumour with minimal toxicitv. thus offerin2 the prospect of photodvnamic therapy occupying a potentially important niche in the urologist's armamentarium.