Purpose: We validate the usefulness of laser-induced autofluorescence
for the detection of bladder carcinoma. Materials and Methods: We obta
ined and analyzed fluorescence spectra from 75 patients in whom bladde
r cancer was suspected. Tissue fluorescence was excited by a nitrogen
laser using a quartz optical fiber placed in gentle contact with the a
rea of interest. The laser-induced autofluorescence spectrum was recor
ded using an intensified optical multichannel analyzer system. Spectra
were corrected for the spectral response of the optical system, and t
he ratios of laser-induced autofluorescence intensities (I) at 385 and
455 nm. (I-385/I-455) were determined. We had previously established
this ratio as a diagnostic algorithm. We included only suspicious blad
der lesions (erythematous, edematous, raised and so forth) that were d
ifficult to diagnose by cystoscopy as well as areas from which random
biopsies were obtained. The fluorescence ratio algorithm was applied t
o 130 bladder areas. Results: Of the 130 biopsies obtained during rout
ine cystoscopy 107 (82%) were nonmalignant by histological classificat
ion. In contrast, because laser-induced autofluorescence effectively g
uides biopsies towards malignant lesions, only 30 biopsies (72% fewer)
would have been obtained from nonmalignant tissue if the fluorescence
ratio that identifies 95% of malignant lesions (95th percentile) had
been selected as the decision criterion during standard cystoscopy. Co
nclusions: By guiding the surgeon to suspicious lesions that are most
likely to be malignant, laser-induced autofluorescence substantially d
ecreases the number of biopsies obtained from nonmalignant tissue duri
ng cystoscopy to diagnose bladder carcinoma.