A fluorescence imaging device applied to the detection of early cancer is described. The apparatus is based on the imaging of laser-induced fluorescence of a dye that localizes in a tumor with a higher concentration than in the surrounding normal tissue after iv injection. Tests carried out in the upper aerodigestive tract, the tracheobronchial tree, and the esophagus with Photofrin II (1 mg/kg of body weight) as the fluorescent agent are reported as examples. The fluorescence is induced by violet (410-nm) light from a continuous-wave (cw) krypton-ion laser. The fluorescence contrast between tumor and surrounding tissue is enhanced by real-time image processing. This is done by the simultaneous recording of the fluorescence image in two spectral domains (470-600 and 600-720 nm), after which these two images are digitized and manipulated with a mathematical operator (look-up table) at video frequency. Among the 7 photodetections performed in the tracheobronchial tree, 6 were successful, whereas it was the case for only 5 of the 15 lesions investigated in squamous mucosa (upper aerodigestive tract and esophagus). The sources of false positives and false negatives are evaluated in terms of the fluorescent dye, tissue optical properties, and illumination optics.

New results are presented on the pharmacokinetics of the fluorescing fraction of Photofrin II in patients with an early cancer in the oesophagus or the buccal cavity. The light-induced fluorescence signal shows a relatively high contrast between tumor and surrounding normal tissue at short times after i.v. injection. The magnitude of this contrast appears to correlate with the staging of the cancer, the more invasive tumors showing the highest contrast. Some early results on the ex vivo fluorescence analysis of biopsies taken on patients injected with the new sensitizer meta-tetra(hydroxyphenyl)chlorin (mTHPC) demonstrate the selectivity of this efficient second generation photosensitizer for advanced lung cancer. mTHPC appears to have good properties for photodetection and has a rather high rate of photobleaching. The significant of the latter is discussed in relation to simplifying light application in PDT.

The use of dyes has been helpful for the photodiagnosis of small cancers accessible to endoscopic examn. An important limiting factor of this technol. is that the presently used fluorescent dye mixt. has a relatively poor capacity to accumulate preferentially in malignant tissue and a low quantum yield of fluorescence. To improve these parameters, fluorescein was coupled to an anti-carcinoembryonic antigen (CEA) MAb and the biodistribution of several conjugates was studied in nude mice bearing a human colon carcinoma xenograft. In vitro, such conjugates with fluorescein to MAb molar ratios ranging 4-19, trace labeled with 125I, showed >82% binding to insolubilized CEA. However, since the aim of this work was the evaluation of MAb-dye conjugates designed for in vivo tumor localization, all newly prepd. MAb-fluorescein conjugates were tested in the exptl. model of nude mice bearing CEA expressing human colon carcinoma xenografts. Under these exptl. in vivo conditions, the conjugate contg. 10 fluorescein mols. per MAb mol. gave an excellent tumor localization (up to 30% of the injected dose per g tumor at 24 h), whereas a conjugate with 19 fluorescein mols. per MAb mol. gave almost no in vivo localization in the tumor, probably due to a very short half-life. Tumor-to-liver, -kidney, and -muscle ratios of 20, 30, and 72, resp., were obtained at 48 h after injection of the conjugate contg. 10 fluorescein mols. per MAb mol. In the spectrofluorometric anal., a high fluorescence intensity was obsd. in the tumor after injection of the anti-CEA MAb conjugate. To compare these results with a conventionally used dye, mice bearing the same xenografts received a purified form of hematoporphyrin, Photofrin II. The intensity of the fluorescence signal of the tumor after an injection of 0.44 mg fluorescein coupled to 20 mg of MAb was 8-fold higher than that obtained after injection of 60 mg of Photofrin II. These results illustrate the possibility of improving the specificity of in vivo tumor localization of dyes for laser-induced fluorescence photodetection and phototherapy by coupling them to MAb directed against tumor markers.

Based on previous experiments in nude mice, showing that fluoresceinated monoclonal antibodies against carcinoembryonic antigen localized specifically in human carcinoma xenografts and could be detected by laser-induced fluorescence, we performed a feasibility study to determine whether this immunophotodiagnosis method could be applied in the clinic. Six patients, with known primary colorectal carcinoma, received an i.v. injection of 4.5 or 9 mg of mouse-human chimeric anti-carcinoembryonic antigen monoclonal antibody coupled with 0.10-0.28 mg of fluorescein (molar ratio 1/10 to 1/14). The monoclonal antibody was also labeled with 0.2-0.4 mCi of 125I (1 Ci = 37 GBq). Photodetection of the tumor was done ex vivo on surgically resected tissues for the six patients and in vivo by fluorescence rectosigmoidoscopy for the sixth patient. Upon laser irradiation, clearly detectable heterogeneous green fluorescence from the dye-antibody conjugate was visually observed on all six tumors; almost no such fluorescence was detectable on normal mucosa. The yellowish tissue autofluorescence, which was emitted from both tumor and normal mucosa, could be subtracted by real-time image processing. Radioactivity measurements confirmed the specificity of tumor localization by the conjugate; tissue concentrations of up to 0.059% injected dose per g of tumor and 10 times less (0.006%) per g of normal mucosa were found. The overall results demonstrate the feasibility of tumor immunophotodiagnosis at the clinical level.

A single multimode optical fiber is used to excite and collect tissue autofluorescence as well as the fluorescence of an IV-injected fluorescent tumor marker. Measurements of the relative fluorescence intensity of a tumor marker as a function of the time after IV injection permit measurement of the kinetics of this substance in tumor, normal tissue, and skin. The authors believe that these are the first measurements of this kind in patients. Furthermore, the autofluorescence spectrum generated at several excitation wavelengths in different tissues is compared, for instance in the oesophagus, the bronchi, and the tongue. The measuring system is based on an optical multichannel analyzer which measures the fluorescence excited by monochromatic radiation from a spectrally filtered Xe lamp. A correlation between the observed pharmacokinetics and tumor properties like the degree of vascularization is of fundamental importance for each selected tumor marker. Also, the results of these measurements are used for the optical detection of tumors.

Two methods for clinical optical light dosimetry are developed. In the first method, which is invasive, a fluorescent probe attached to an optical fiber is inserted by means of a thin hypodermic needle and measures light transmitted through the cheek as a function of the penetration depth. In the second noninvasive method, the diffusely reflected light intensity, from a small illuminated spot on the surface of the tissue to be investigated, is measured as a function of the radial distance along the surface. Preliminary results with both methods are presented. Simulations of the second measurements, which allow for a simplified extraction procedure of the relevant optical data from such measurements, are also shown.

The performance of a fluorescence endoscope for the detection of early cancer is clinically evaluated. the apparatus is based on the imaging of the laser-induced fluorescence (LIF) of a dye which localizes in the tumor after IV injection with a higher concentration than in the surrounding normal tissue. The tests are carried out in several of the hollow organs, such as the upper aerodigestive tract, the bronchi, and the colon. In the two former cases the dye used is photofrin II, whereas in the latter case conjugates between monoclonal antibodies (Mab) directed against carcinoembrionic antigen (CEA) and fluorescein molecules are injected. The fluorescence contrast between tumor and surrounding tissue is enhanced by real-time image processing which eliminates most of the tissue autofluorescence as well as the fluorescence due to the relatively small amount of dye localized in the normal tissue. This is done by recording the fluorescence image in two spectral domains, after which these two images are digitized and manipulated with a mathematical operator (lookup-table). The sources of false positives and false negatives are evaluated in terms of the fluorescent dye and tissue optical properties.

An apparatus is designed and realized to detect "early" cancer at the surface of the hollow organs in the human body by endoscopic means. The tumor is localized by the laser induced fluorescence of a dye (HPD) which concentrates selectively in the neoplastic tissue after intravenous injection. Fluorescence contrast between the tumor and its normal surroundings is enhanced by subtracting the background autofluorescence which occurs in both types of tissue. This is done by means of 2-color digital images manipulation in real-time. Preliminary clinical tests of the apparatus demonstrated the detection of carcinoma in situ in the esophagus.

The efficacy of photodynamic therapy (PDT) alone was evaluated on 41 ‘early’ squamous cell carcinomas of the pharynx (10), oesophagus (15) and tracheo-bronchial tree (16). All lesions but two were synchronous second primaries in ENT-patients suffering from a more extensive cancer, governing the overall oncological prognosis.