Image subtraction analysis with technetium-99M labeled monoclonal antibody and colloid for evaluation of liver lesions. Phantom measurements and patient studies.

The ideal radiolabeled monoclonal antibodies (MoAbs) against cancer antigens are taken up by liver metastases; the background activity of normal liver, however, causes problems for delineation and detectability. In order to study these phenomena, a liver phantom containing hot and cold lesions of different sizes (diameters 5-32 mm) was constructed. It was placed into an elliptic cylindrical container representing a cross section of the abdomen. The specific activities in hot lesions varied from 1.85 to 14.8 MBq/ml, whereas liver phantom and cylinder activities were kept constant during different measurements. Lesions of size 1.3 cm3 could be detected without any subtractions, if the signal to background ratio was larger than 1.2. Lesions larger than 5 mm in diameter could also be detected using subtraction, which gave additional information by a factor 2-9, when the lesion sizes varied from 0.3 to 5.3 cm3 and when the specific activity in the lesions was at least twice as high as in adjacent liver. This subtraction technique was applied in 32 breast and lung cancer patients after injecting about 1,000 MBq 99mTc-labeled anti-CEA MoAb; 24 h after the antibody injection 75 MBq 99mTc-phytate was injected. The phytate + residual MoAb image was subtracted from the original antibody image. Thirteen patients had liver metastases verified by (CT, US), but only four patients had clearly observable abnormal liver uptakes in planar MoAb images. In 9 cases, additional information concerning liver metastases was obtained by subtraction technique. To judge by our phantom measurements the enhanced detectability was not an artefact.

Radioimmunoimaging of non-small cell lung cancer with 111In- and 99mTc-labeled monoclonal anti-CEA-antibodies.

Radiolabeled monoclonal anti-CEA antibodies were used for radioimmunolocalization (RIL) of non-small cell lung cancer; in 30 patients with 111In labeled anti CEA F(ab')2 fragment (BW 431/31) and in 16 with 99mTc-labeled intact MoAb (BW 431/26). RIL results were compared with those of other imaging modalities. Paraffin sections from some patients were also studied immunohistochemically using anti-CEA antibody. Patients with 111In labeled MoAB were imaged twice 1-4 days after injection and for image enhancement pulmonary and liver/spleen subtraction were performed. Twenty-seven of 28 primary tumors were positive and metastases were detected in all patients. The total number of lesions was 78 of which 61 (78%) could be detected by RIL. For verification CT was applied to the study of 46 lesions detected by RIL. We found 6 unknown lesions subsequently verified histologically. Using subtraction techniques we detected 9 lesions in 4 patients, later verified as pulmonary metastases, not detected in unprocessed images. Pleural, mediastinal and pericardial lesions were also better delineated in subtracted images than in unprocessed images. Imaging of non-small cell lung cancer with 99mTc-labeled MoAB was performed twice 4-24 h after injection. RIL results were compared with other imaging methods; CT US, conventional radiography, and immunohistochemistry. Twelve out of 16 patients with suspected or known lung cancer had positive immunoscintigrams; 19 of 25 lesions could be detected by RIL. There were 5 false positive and 2 true negative findings. Immunoperoxidase (IP) stainings of paraffin sections of the tumours from 7 patients were performed using two different anti-CEA antibodies; BW 431/26 and ZCEA1. None of the seven tumors examined by immunohistochemistry were negative when stained by BW 431/26, which was the antibody used for immunoscintigraphy.

Imaging of soft-tissue sarcomas with indium-111-labeled monoclonal antimyosin Fab fragments.

Some soft-tissue sarcomas contain intracellular myosin. We therefore studied the possibility of localizing various soft-tissue sarcomas with 111In-labeled monoclonal antibody Fab fragments binding specifically to myosin, assuming that damage to the cell membrane could expose intracellular myosin. Nineteen patients with different types of soft-tissue sarcomas were studied. Eighteen patients were found to have abnormal antibody uptakes. Antibody uptake was not observed in an additional patient operated for a benign tumor (gastric leiomyoma). The immunoscintigraphy results were generally in good agreement with those of other radiologic findings (computed tomography, ultrasound, magnetic resonance imaging). Surprisingly, the immunohistochemistry results showed that tumors not stainable for myosin can also be imaged with antimyosin. Thus, the mechanism of antibody uptake does not seem to be related entirely to specific antigen recognition. Irrespective of the exact mechanism for the uptake of labeled antibody this method appears to be useful for localizing soft-tissue sarcomas.