The selective targeting of tumors by enzymes conjugated to monoclonal antibodies (mAb) may be an ideal approach to convert relatively nontoxic prodrugs into active agents at the tumour site. We used the anti-carcinoembryonic antigen mAb BW431/26 conjugated to alkaline phosphatase (AP) and phosphorylated etoposide (etoposide-P) as a prodrug to study the feasibility of this concept. Etoposide was phosphorylated with POCl3. Quantitative hydrolysis of etoposide-P to etoposide occurred within 10 min in the presence of AP. BW431/26 and AP were conjugated using a thioether bond. The AP conjugate retained 93% of its calculated activity. 125I-labelled AP conjugate did not show a reduction of immunoreactivity as determined by a cell-binding assay. SW1398 colon cancer cells were used to analyse the cytotoxicity of etoposide and etoposide-P. Etoposide (IC50 22 microM) was 100 times more toxic than etoposide-P (20% growth inhibition at 200 microM). Pretreatment of the cells with BW431/26-AP prior to etoposide-P exposure resulted in a dramatic increase in cytotoxicity (IC50 70 microM). The pharmacokinetics and tumour-localizing properties of BW431/27 and the AP conjugate were assessed in nude mice bearing SW1398 tumours. BW431/26 showed excellent tumour localization (10% of the injected dose/g tissue retained from 8 h to 120 h), whereas the AP conjugate showed a reduced tumour uptake (3%-0.3% of the injected dose/g tissue at 8-120 h), a faster clearance from the circulation and a high liver uptake. Radiolabelled AP showed a similar pharmacokinetic profile to the AP conjugate. Gel filtration analysis of blood, liver, and tumour samples indicated good stability of the conjugate.

Twenty colorectal cancer patients were given an intravenous injection of human IgM monoclonal antibody (MAb) 16.88 (8 mg) conjugated to 131I for tumor localization. After a 2-week interval, a second injection with 200, 500, or 1000 mg of unlabeled antibody added was given to groups of five patients each. at the end of the 2-hour infusion, 66% of the radioactivity remained in the circulation. Blood clearance of the 131I-labeled MAb 16.88 was biphasic with a mean half-life (T1/2 alpha) of 12 hours and T1/2 beta of 45 hours. Clearance rate was 0.09 L/hour. More than 90% of the 131I in serum was protein bound, with an immunoreactive fraction of 80% in the first 48 hours. Size exclusion chromatography indicated no degradation products other than 131I in serum and urine. The urinary excretion rate of 131I increased to 1.5% of the dose per hour at 24 hours, with 50% of the dose excreted in 34 hours. The pharmacokinetic profile of 131I-labeled MAb 16.88 was neither influenced by the total protein dose of antibody administered nor affected by specific uptake in tumor tissue in individual patients, as determined on early immunoscintigrams. The larger antibody doses showed a slightly slower excretion of 131I. The assays applied to determine immunogenicity were enzyme-linked immunosorbent assay, radioimmunoassay, and the dot-blot assay. They had sensitivities ranging from 5 ng/mL to 0.5 micrograms/mL for goat or rabbit antihuman IgM. The assays did not reveal antihuman antibody responses.