Search

Background. The prognosis for colorectal cancer patients is related closely to the extent of tumor at the time of diagnosis, and early detection of metastatic or recurrent disease is an important prerequisite for successful treatment. Radioimmunodetection is a relatively new technique to image colorectal cancer using radiolabelled monoclonal antibodies (MoAb). The goal of this study was to evaluate the clinical use of radioimmunodetection with the anticarcinoembryonic antigen MoAb FO23C5, radiolabelled with 131 I, in patients submitted to surgery for colorectal cancer and with suspected local recurrences

CEA determination has attained an important role in the clinical management of patients with tumors of the colorectal tract. In this paper the experience of the Istituto Tumori of Milan is reviewed and the results are discussed. Three hundred and thirty-six patients were followed after curative resection of colorectal carcinoma. The follow-up period was 15 years, from January 1975 to December 1990 (global follow-up 1358 years). In the course of follow-up 136 patients developed recurrent disease. The number of CEA determinations for each patient ranged from 1 to 37 (mean 8, total 3330). CEA levels of presurgical patients were related to the clinical stage. Among patients who developed recurrences 61% showed an increase in CEA serum levels. In 200 patients with a negative follow-up we observed only 15 cases of false-positive results.

The monoclonal antibody (Mab) 131I-MOv18 was administered to 30 patients with ovarian carcinoma intravenously (n = 20) and intraperitoneally (n = 10). After intraperitoneal administration, higher tumour uptake (mean values 1.3% vs. 0.8%) and a better tumour/background ratio (mean values 2.8 vs. 1.9) than after intravenous injection were obtained. Moreover, after intraperitoneal administration the uptake in non-affected organs, such as liver and spleen, was lower. However, occasionally the favourable results of the intraperitoneal route were cancelled by persistent pelvic non-specific accumulations of 131I-MOv18. The possibility to change the biodistribution pattern in the latter cases with peritoneal washing was evaluated. 3 patients were submitted to this procedure and an improvement in the radiotracer biodistribution was obtained in 1 case. With regard to tumour detection, the average sensitivity (73%) showed a significant difference from the sensitivities for abdominal (61%) and pelvic lesions (90%). No false positive results were noted.

Twenty-five patients treated with debulking surgery and chemotherapy for ovarian cancer were prospectively studied to evaluate the efficacy of radioimmunoscintigraphy (RIS) in detecting residual tumor before second-look surgery. RIS was performed with the monoclonal antibody OC125 F(ab')2 labelled with 1-131 without knowledge of clinical data and compared with subsequent surgical results. Second look showed tumor persistence in 12 patients, mostly characterized by small lesions. The overall diagnostic sensitivity of RIS was 50% and the specificity was 85%. In particular, RIS showed better sensitivity for pelvic tumor localizations than for abdominal sites (73% vs 33%); this was due to the inability of RIS to detect upper abdominal lesions. Therefore, our conclusion is that, at present, RIS cannot substitute surgical second-look in the management of ovarian cancer, however, considering that also ultrasonography, computer tomography and magnetic resonance are not always able to give definite diagnostic evidence in the follow-up of ovarian carcinoma, RIS could be added to these procedures to balance the limitations of each method. In this regard, the best application of RIS could be in the follow-up of patients with marker elevation without clinical evidence of disease, especially in the case of pelvic fibrosis or adhesions due to previous therapy, where the other non-invasive tools can give doubtful diagnostic results.

Nineteen patients with clinical and mammographic diagnosis of breast cancer and 1 patient already mastectomized underwent radioimmunoscintigraphy (RIS) in order to evaluate the axillary node status before surgery. After histologic examination, ductal breast carcinomas were found in 17/19 patients and axillary metastases were found in 11 patients. RIS was performed with planar scintigraphies and SPET (Single Photon Emission Tomography) of mammary and axillary regions after intravenous injection of In-111 B72.3. Overall RIS sensitivity for breast tumors was 71% (12/17); according to the tumor site different RIS sensitivity was demonstrated (90% for lesions of external quadrants versus 43% for lesions of internal quadrants). RIS was negative in 2 patients with post-surgery diagnosis of mammary dysplasia. As regards axillary metastasis RIS sensitivity and specificity were 91% (10/11) and 78% (7/9), respectively. Human Anti-Murine Antibody (HAMA) production was shown in 16.7% (2/12) of the patients. The quantitation of In-111 B72.3 uptake (%I.D./g; mean+/-S.D.) was 0.0054+/-0.0021 in breast tumors, 0.0021+/-0.0011 in normal mammary tissue, 0.0053+/-0.0027 in axillary metastasis and 0.0032+/-0.008 in normal axillary nodes. Our data demonstrated that RIS can detect tumor spread to the axilla in patients with breast cancer. Larger pre-operative study is required to evaluate if RIS can alter the management of this disease.

An anti-CEA monoclonal antibody (FO23C5), belonging to the IgG1a class, was used to perform several preclinical and clinical studies on radioimmunoscintigraphy (RIS) in patients with colorectal carcinoma. Preliminary screening on different tissues showed high specificity of this antibody for gastric and colorectal carcinomas. A pilot study on 51 patients with 64 localizations of colorectal carcinoma was realized, followed by a multicenter validation study in which, among 509 patients bearing CEA secreting tumors, 254 patients had primary or relapsed colorectal tumors. High sensitivity and specificity values were obtained by these studies and many "unsuspected" localizations were recorded. In order to better define the clinical utility of this approach, a prospective trial was run on 59 patients previously submitted to surgery for colorectal carcinoma and with suspected local relapses. A comparative evaluation of RIS, CT scan, US and MRI was performed. RIS and MRI have the highest value of accuracy (86%) followed by CT scan (68%) and US (54%). No adverse reactions were noticed in any of the patients examined.

The studies on radioimmunodetection of malignant melanoma reported are based on the use of the 225.28S monoclonal antibody against a high molecular weight melanoma associated antigen, an IgG2a isolated and characterized by Ferrone et al. The strategy followed by us to check the characteristics and the behaviour of this antibody and to verify its possible use in clinical praxis, includes three groups of trials performed from 1982 to 1987. A pilot study was performed on 42 patients with 74 melanoma lesions, a multicenter study was realized, with the collaboration of 10 Italian nuclear medicine centers, on 254 patients with 539 lesions and prospective study performed on 44 patients with 54 suspected regional lymph nodes involvements. High sensitivity and specificity were obtained especially with the use of technetium compound which also allowed the attainment of very high predictive values of positive tests. No adverse reactions were noticed.

Preoperative detection of metastases in regional lymph nodes should constitute a fundamental step towards a better management of breast cancer. The National Cancer Institute of Milan has traditionally been engaed in this particular problem, that is, the search for a new imaging technique for detecting lymph nodal metastases from breast cancer. In 1984 axillary lymphoscintigraphy was performed on 26 patients with operable breast cancer by periareolar injection of 100 μCi of 99m Tc sulphur microcolloid. The regional lymph nodes were identified as spot areas and a comparison was made between the number of primary draining axillary lymph nodes and their pathological examination. In only 2 of the 26 patients have more than 3 spot areas been revealed by axillary scintiscan, but this finding did not correspond to metastatic invasion. The lymphoscintigraphy pattern of the other 24 patients did not suggest any pathological condition. From our experience the method does not appear to help either a precise diagnosis of breast disease or the definition of the extent of mammary cancer.

A prospective study to evaluate the clinical usefulness of radioimmunodetection of melanoma in clinical practice is ongoing at the National Cancer Institute of Milan, Italy. Technical conditions for the application of the method were previously reported. In this trial, 99mTc-labelled F(ab’) 2 fragments of the 225.28S monoclonal antibody were used against a high molecular weight melanoma associated antigen (HMW-MAA). Retrospective studies on radioimmunodetection of melanoma have already been made by our group and by other Centers in about 300 patients. This study concerns the evaluation of the regional extension of primary melanoma. 23 patients with 32 suspected lymphatic involvements of melanoma on the trunk and arms underwent immunoscintigraphy. No false positive results were observed; 3 false negatives, one corresponding to a micrometastasis, were noticed. Specificity corresponds to 100% and sensitivity to 78.6%.

In this paper we summarize the investigations performed by our group utilizing an anti-CEA monoclonal antibody (F023C5) labelled with different radionuclides in humans. Since 1983 radioimmunoscintigraphy (RIS) was performed on 51 patients with 64 localizations of colorectal carcinoma (pilot study). A multicenter clinical trial in a large number of patients (509 pts of which 284 with gastrointestinal cancer) was subsequently carried out in collaboration with ten nuclear medicine centres. High sensitivity and specificity values were obtained by these studies and many unsuspected lesions were recorded. In order to better define the clinical role of RIS, a prospective study was performed on 59 patients with suspected local relapses of colorectal cancer. A comparative evaluation of RIS, CT scan, US and MRI was done. RIS and MRI had the highest accuracy (86%) followed by CT scan (68%) and US (54%).