Near-IR imaging of fluorescent probes for cancer detection

Colorectal cancer is the third most common cancer in the US (not counting skin cancers). Colonoscopy with routine polypectomy has proved an effective screening tool for removing colorectal polyps before they turn into cancer. However, routine endoscopy still misses about 24 percent of lesions.

Dr. Umar Mahmood and his colleagues at Massachusetts General Hospital, Harvard Medical School, in Charlestown, Mass., wanted to see whether microfiber optic catheters would reveal colonic abnormalities more effectively after an optical “smart” probe was injected intravenously. The probe is quenched after initial injection, but then fluoresces brightly in areas of target overexpression.

As reported in the July issue of Radiology, the researchers tested their technique on 10 mice. Tumor cells were implanted into each mouse’s colon. Thirteen days later, a near-IR fluorescent probe (activated by proteases, particularly cathepsin B, and known to be overexpressed in abnormal growths in the colon) was injected into five of the mice. The probe, from VisEn Medical of Woburn, Mass., had a peak absorption of about 680 nm and a peak emission of 700 nm. The other five mice (the control group) were injected with an equivalent volume of saline.

The next day (day 14), an imaging microcatheter was inserted rectally into each mouse’s colon. Visible light was separated from near-IR light through a 670-nm dichroic mirror. White-light and near-IR images were recorded separately but simultaneously with white-light or near-IR video cameras.

After the microcatheter imaging, the mice were euthanized, and the colons were removed and imaged with an epifluorescence system from Siemens Medical Solutions of Malvern, Pa., to quantify the fluorescence intensity.

The excised colons also were evaluated with a histologic examination and with cathepsin B immunohistochemistry.

The fluorescence intensity of the tumors was more than 30 times greater in the mice that received an injection of the probe than in the control animals. The near-IR images after probe administration revealed a target-to-background ratio of 8.86, whereas white light images had a ratio of 1.14. The researchers easily identified all the tumors in the mice that had received an injection. Lesions — including smaller, flat ones — showed up as bright fluorescent signals. The tumors were not as easy to see in the control group, and some smaller ones were either missed or simply labeled as suspicious.

Histologic examination confirmed colon cancer in all 10 mice, and immunohistochemistry confirmed an overexpression of cathepsin B in the tumors as compared with surrounding tissue.

The researchers believe the results indicate that protease-activated probes used with microcatheters may help reveal tumors that might otherwise go undetected.