B. Delaloye

Tumor imaging with monoclonal antibodies

Immunoscintigraphy offers the possibility of specifically targeting human tumors, but the complexity of the human immune system, as well as tumor-related phenomena, prevent monoclonal antibodies from reaching a large number of tumor cells in which they can interact with the antigen. Possible ways to overcome these problems are the use of small fragments, in particular those of genetically engineered humanized antibodies including single immunoglobulin-variable domains, as well as techniques to label the antibody in vivo after a sufficient amount has been taken up by the tumor and the remainder has been eliminated. Despite the low absolute tumor uptake, results of European studies, presently available radiolabeled monoclonal antibodies in gastrointestinal and ovarian cancers yield an average sensitivity of more than 70% with an average specificity of more than 80%, even in otherwise occult tumors. Because of possible tracer uptake in normal liver, the detection rate of liver metastases varies from less than 10% to more than 90%. For the detection of local recurrence in the pelvis, immunoscintigraphy has been found to be more accurate than methods that are based on the imaging of structural changes. Fusion of morphological and functional images might improve the early detection of recurrent and metastatic disease. In melanoma, another tumor that has been extensively studied in Europe, similar results have been obtained, whereas only few data are presently available for other tumors (especially lung and breast cancer).

Immunoscintigraphy with antigranulocyte monoclonal antibodies for the diagnosis of septic loosening of hip prostheses

To determine the value of immunoscintigraphy (IS) with antigranulocyte monoclonal antibodies (Mab) in the diagnosis of subacute or chronic infection of hip prostheses, we prospectively studied 57 patients (23 women and 34 men; age 29–92 years, mean 72.7 years) sent to our institution in the past 6 years for clinical suspicion of septic loosening of a hip prosthesis. Nineteen patients had bilateral prostheses and one of them was studied twice. A total of 78 prostheses were examined. All patients had three-phase bone scans followed by IS with technetium-99m antigranulocyte Mab BW 250/183. Intervals between bone scans and IS varied from 2 days to 4 weeks. Final diagnosis was assessed by culture in 48 cases (articular puncture or intraoperative sampling) and by clinical follow-up of at least 8 months in 30 cases. Twelve prostheses were considered septic and 66 non-septic. The overall sensitivity and specificity were 92% and 64% respectively for bone scans, 67% and 75% for IS and 67% and 84% for both modalities together. In three cases, IS was doubtful and the final clinical diagnosis was negative for infection. False-positive results were observed in the presence of massive loosening of the prosthesis or in association with metaplastic peri-articular bone formation. In three of the four false-negative results, infection was proven only after enrichment of the culture, and the bacterium wasStaphylococcus epidermidis. In 12/33 (36%) positive bone scans IS allowed the diagnosis of infection to be excluded. Overall accuracy of both modalities together was 81% and the negative predictive value was 93%, which compares favourably with the results reported for other non-invasive methods.

Radiolabelled monoclonal antibodies in tumour imaging and therapy: out of fashion?

The initial enthusiasm for the development of diagnostic and therapeutic studies involving the use of monoclonal antibodies was replaced by scepticism as hopes remained unfulfilled. Against this background one needs to ask whether immunoscintigraphy (IS) serves clinical needs effectively and whether radioimmunotherapy (RIT) has a future. The current review considers these questions by reference to relevant studies. Taking colorectal cancer as an example, an appraisal is offered of the ability of IS to detect disease at an early stage and thereby to reduce mortality, and of the influence of the results of IS on patient management. It is concluded that in a limited number of cases of colorectal cancer and other solid tumors, IS will allow surgery to be performed at a stage where cure is still possible because of its ability to detect early recurrence. Turning to RIT, the results of studies in respect of various tumour types are reviewed, with due attention to reported toxicity. As regards colorectal cancer, no consistent therapeutic effects have been achieved, and myelotoxicity is typically the dose-limiting factor. Thus many questions remain to be answered, regarding antigens to be targeted, fractionation schedule, the use of “humanised” antibodies, choice of radionuclide and the use of intact immunoglobulins or fragments. These questions are considered. Overall it is concluded that the most promising application of RIT is as adjuvant therapy in patients with minimal residual disease, and a controlled multicentre trial is recommended. The development of more potent radioimmunoconjugates for therapeutic and ultimately diagnostic purposes will contribute to the improvement and development of IS by increasing its potential to influence prognosis.

Diagnostic applications and therapeutic approaches with different preparations of anti-CEA antibodies.

Various preparations of anti-CEA antibodies have shown to detect very sensitively CEA producing tumors. The development of human anti-mouse antibodies (HAMA) prevents from the widespread use of immunoscintigraphy (IS) in the follow-up of patients with colorectal carcinoma. It is, however, not yet clear if genetically reshaped antibodies will solve this problem and it is even less clear if this will change prognosis of these patients. There is certainly room for new therapeutic approaches in colorectal carcinoma. Radioimmunotherapy in combination with other techniques might advantageously complete surgery, chemotherapy and radiotherapy, but needs to be further developed.

Progress in Diagnostic Immunoscintigraphy and First Approach to Radioimmunotherapy of Colon Carcinoma

More than thirty years ago, Pressman and Korngold (1953) and Bale and Spar (1957) reported the first experiments indicating that radiolabeled antibodies injected into a tumor bearing animal can localize preferentially into the tumor, which could be detected by external scanning. These early studies, however, were performed with poorly characterized antibodies obtained by absorption and elution from tumor cells or crude tumor fractions. Thus, this type of antibodies could not be used for clinical applications.

Immunoscintigraphy of CEA-Producing Tumors with Special Emphasis on the Use of Mab Fragments and ECT

The first reports that carcinoma could be detected in patients by external scanning following injection of purified I-131-labeled anti-CEA antibodies were made by Goldenberg et al. [1, 2]. They claimed that almost all the CEA-producing tumors could be detected by this method and that there were no false-positive results. However, our experience with highly purified goat anti-CEA antibodies [3] and the same blood pool subtraction technique used by Goldenberg was that only 42% of CEA-producing tumors (22 out of 53 tested) could be detected by this method [4]. Furthermore, we found that in several patients the labeled anti-CEA antibodies localized nonspecifically in the reticuloendothelium, particularly in the liver. Despite the use of the subtraction technique, it was difficult to differentiate this nonspecific uptake from the specific uptake in liver metastases. The discrepancy in the results obtained by the group of Goldenberg and our own is unlikely to be due to a difference in the quality of the anti-CEA antibodies used, since a direct measurement of the radioactivity in tumors resected after injection showed that our antibodies were capable of excellent tumor localization [4]. Furthermore, in a few patients scheduled for tumor resection, we injected simultaneously 1 mg goat anti-CEA antibodies labeled with 1 mCi I-131 and 1 mg control normal goat IgG labeled with 0.2 mCi I-125. With this paired labeled method adapted to the patient situation, we demonstrated that the antibody uptake was four times higher than that of control normal IgG [4].

Biodistribution of anti-CEA F(ab')2 fragments after intra-arterial and intravenous injection in patients with liver metastases due to colorectal carcinoma.

SummaryThe biodistribution of simultaneous intra-arterial and intravenous injections of a radiolabelled anti-CEA MAb F(ab′)2 fragment was studied in three patients with liver metastases from colorectal cancer. Identical MAb fragments, labelled with either 125I or 131I, were injected over a period of 30 min into the hepatic artery and into a peripheral vein. After 1 or 2 days, biodistribution was measured in the surgically removed metastases, normal tissue samples and blood. By tissue radioactivity counting, tumour uptake in the range 6.3–9.1% of injected dose per gram was found. Superimposable metastasis-to-blood and metastasis-to-normal liver ratios were obtained for both iodine isotopes in all three patients. The results indicate that the intra-arterial injection of MAb F(ab′)2 fragments gives no measurable advantage over more convenient injections into a peripheral vein.

Immunoscintigraphy in the Early Diagnosis of Tumours of the Abdomen

Radiolabelled monoclonal antibodies (MoAbs) have been used to detect various tumours by scintigraphy (Goldenberg et al. 1978; Mach et al. 1980; Larson 1985; Baldwin and Byers 1985; Delaloye et al. 1986; Chatal 1989; Nunz and Emrich 1990), but the specific role of antibody imaging in the clinical evaluation of patients still remains to be defined. In the following, the possible clinical usefulness of immunoscintigraphy in malignant tumours of the abdomen, especially bladder, ovarian and colorectal carcinoma, will be analyzed. From this perspective it would be fastidious to enumerate and describe in detail all the antibodies, radiolabels, labelling and imaging methods which might be used for immunoscintigraphy. A comprehensive review on the technical aspects of immunoscintigraphy has been published recently (Britton et al. 1991).