Photodynamic imaging of a rat pancreatic cancer with pheophorbide a

Laser-induced fluorescence of pheophorbide a (Ph-a) was used for in vitro photodynamic imaging (PDI) of a rat pancreatic acinar tumor. A 400 nm excitation induced a 470 nm autofluorescence and a 678 nm dye fluorescence in tumors and their surrounding pancreas 24 h after a 9 mg kg-1 body weight Ph-a intravenous administration. With lower intensities in these blood-rich tumors than in pancreas, Ph-a fluorescence signals are unable to provide tumor images. A dimensionless function (the ratio of Ph-a fluorescence by autofluorescence, called Rt for the tumor and Rp for the pancreas) was used for fluorescence contrast calculation (C = Rt/Rp) between six tumors and their paired pancreas. Among five available laser excitation wave-lengths, only the 355 nm excitation gave a distinctive contrast (C = 1.5). The PDI of six intrapancreatic tumors and their intraperitoneal metastasis and of two control normal pancreas was thus performed ex vivo using a 355 nm excitation source delivered by a tripled Nd:YAG laser and a charged-coupled device camera. Fluorescence images were recorded at 680 nm (dye), 640 nm (background) and 470 nm (autofluorescence) through three corresponding 10 nm width bandpass filters. Computed division for each pixel of Ph-a fluorescence values by autofluorescence generated false color image. In this way, contrasted tumor images were obtained. But in five out of six animals false-positive images were present due to an autofluorescence decrease in some normal pancreatic areas. A 470 nm autofluorescence imaging on the same tumors gave in all cases false-positive image and false-negative in half of the cases. These observations suggest that autofluorescence alone is unable to achieve accurate PDI of pancreatic carcinoma and that using Ph-a as a PDI dye needs strong improvements.