ORIGINAL 

Annals of Nuclear Medicine Vol. 6, No. 1, 21-27, 1992 

Pharmacokinetic analysis of antibody localization in human colon cancer : comparison with immunoscintigraphy 

Atsushi KUBO,* Kayoko NAKAMURA,* Michiaki KATAYAMA,* Shozo HASHIMOTO,* Tatsuo TERAMOTO** and Susumu KODAIRA** 

*Department of Radiology and **Surgery, School of Medicine, Keio University, Tokyo, Japan 


The biodistribution and imaging characteristics of the 111In-labeled anti CEA monoclonal antibody ZCE-025 were studied in five patients with suspicion of colorectal carcinoma. Evaluation included antibody pharmacokinetics and assessment of antibody distribution in surgical specimen, making a comparison with whole-body imaging with a gamma camera. ZCE-025 localization in tumors was demonstrated by gamma-camera imaging in 4 of the 5 patients, corresponding to surgical findings. Persistent accumulation of 111In in the lymph nodes was observed in one patient, whereas surgical exploration of these lymph nodes showed no gross or microscopic evidence of metastases of colon carcinoma. Analysis of individual plasma by size exclusion HPLC showed two radioactivity peaks, labeled antibody and free DTPA. No transchelation of 111In to circulating transferrin was observed. The blood clearance was fitted to a two-compartment equation and its half-lives were found to be 10.8+-8.7 h and 69.5+-21.8 h for t1/2a and t1/2b, respectively. Total urinary excretion averaged 0.3% of the injected dose/h with a small patient to patient variation. At 24hrs postadministration the predominant radiolabeled species in urine was free DTPA. Thereafter, radioactivity in urine was partly present as a low molecular weight catabolic product. No apparent correlation between CEA content and uptake of 111In-ZCE-025 in tumors resected by surgery could be found. How 111In-1abeled antibody is accumulated into tumors as well as into some nontumor tissues needs further study. 

Key words: monoclonal antibody, colon cancer, immunoscintigraphy, pharmacokinetics 


INTRODUCTION 

THE USE OF ANTIBODIES as targeting agents for cancer continues to attract much attention despite the many technical challenges that confront its successful clinical application. 
 ZCE-025 is one of the murine monoclonal antibodies (MoAbs) which have shown promise for the immunoscintigraphy and radioimmunotherapy of colon cancer. ZCE-025 is of the IgG1 subclass and is derived from the same clone as MAb-35.1 It reacts with approximately 95% colorectal tumors including 

 Received September 30, 1991, revision accepted November 20, 1991. 
 For reprints contact : Atsushi Kubo. Department of Radiology, School of Medicine, Keio University, 35 Shinanomachi, Shinjuku-ku, Tokyo 160 Japan 

cultured cell lines and primary and metastatic tissues obtained by biopsy. The antibody was provided by Hybritech, Inc. (San Diego) in the U.S.A., and the immunoscintigraphy of colorectal cancer with 111In-labeled ZCE-025 has been exclusively evaluated in the U.S.A.2-4 Few papers have, however, dealt with its pharmacokinetic characterizations.5 It is important to know how injected antibody is metabolized in humans from the point of antibody-guided targeting studies. 
 In this report we described the pharmacokinetic behavior of 111In-ZCE-025 injected into patients, comparing it with their immunoscintigraphic results. 

MATERIALS AND METHODS 

Radiolabeling of monoclonal antibody 

The anti-CEA monoclonal antibody (MoAb), ZCE-025 was of murine origin. ZCE-025-DTPA conjugates, prepared and characterized as described in a previous paper,6 were supplied by Hybritech, Inc. through Teijin Limited as a sterile, apyrogenic freeze-dried powder. The conjugated antibody was labeled with 75 MBq of 111In in 0.26 M citrate buffer solution (pH 1.5) followed by the addition of 0.13 M citrate buffer solution (pH 8.5) to neutralize the solution. Labeling efficiency higher than 90% was achieved without loss of immunoreactivity, which was confirmed by the cell binding assay. 
 The labeled antibody preparation was assayed by size exclusion HPLC using a single 7.5x300-mm Hitachi Gel Pack W-530 column with the UV 280-nm detector and 0.1 M phosphate buffer (pH 7.5) containing 0.2 M KCl as the eluant. Ten ul of the sample was analyzed with an automatic injector and fraction collector. Fractions were then counted in an automatic well gamma counter. 

Patient selection and administration of 111In-labeled antibody 

Patients in this study had suspected colorectal carcinoma. In each case, the patient had been scheduled for surgical procedures. Informed consent was obtained and the consent forms were approved by the Review Board of Keio University, School of Medicine. 
 An intradermal skin test with 0.1 ug of ZCE-025 was performed 15 minutes prior to antibody administration to detect immediate-type hypersensitivity reactions. No positive skin tests were observed. Each patient received 42 mg of ZCE-025 labeled with 74 MBq of 111In (2 mg of 111In-ZCE-025 mixed with 40 mg of unmodified MoAb) in 100 ml of normal saline containing 30% human serum albumin by intravenous injection over a period of 60 min. One and three days after the administration, planar and SPECT images were obtained. The gamma camera used was a Toshiba GCA-90B with a medium energy parallel collimator. 

Specimens resected, blood, and urine samples 

In all patients, blood was drawn just after and 1, 3, 24, 48, 72 and 120 minutes after the injection. The clearance and half-life of 111In in blood were calculated with a nonlinear regression analysis computer program. For all patients, plasma samples were analyzed by size-exclusion HPLC under the same conditions as for the labeled antibody. 
 In order to monitor 111In excretion, urine samples were collected as 24 hours aliquots for 72 hours following injection of the antibody. The 72 hour cumulative excretion radioactivity in urine was expressed as % of the total dose administered. Urine samples were analyzed by size-exclusion HPLC under the same conditions as for the labeled antibody. 
 Surgical procedures were planned for performance between 7 and 10 days after 111In-ZCE-025 injection. Tumor, the adjacent colon tissue and regional lymph nodes resected at exploration were examined so that the imaging results could be compared with the surgical findings. Excised tumor or non-tumor specimens were examined for the presence of CEA from snap-frozen samples with ZCE-025 by the avidinbiotin-immunoperoxidase procedure.7 In addition, the gamma camera images of specimens resected were obtained immediately after the surgery. 
 Aliquots of tissues in excess of that required for histopathological analysis were weighed and analyzed quantitatively for CEA and 111In content. Specimens resected were homogenized with PBS (9.6mM phosphate buffer solution (pH 7.3) plus saline) by the procedure described in our previous paper,8 and their CEA content and protein concentration were determined by means of the enzyme-linked immunoassay with anti-CEA monoclonal antibody9 and by the Lowry method,10 respectively. 


RESULTS 

Patient characteristics 

Characteristics of all 5 assessable patients are shown in Table 1. Two patients (No. 4 and 5) had liver metastases confirmed by surgery. Three patients (No. 1, 2, and 5) had well or moderately differentiated carcinoma with normal plasma CEA levels. One patient (No. 4) had moderately differentiated carcinoma with increased plasma CEA. Patient No. 4 had metastases in the liver and a paracolic lymph node. One patient (No. 3) had a history of rectosigmoid adenocarcinoma resected previously. No recurrence was confirmed by surgery, though a severe structure was seen in the rectum in a Baenema. 
 Following radiolabeling, in analysis by size exclusion HPLC a small amount of free 111In nonattached to antibody was detected, leading the labeling efficiency with higher than 90% (Fig. 1). 

Pharmacokinetics of 111In-ZCE-025 

In all patients documented with colorectal carcinoma, the mean peak blood level of 111In-ZCE-025 was 24.3+-4.1% of the injected dose per liter of blood, at the end of the 60-min injection. A representative pattern is shown in Figure 2. Clearance of 111In-ZCE-025 from the blood showed a half life, t1/2a, of 11.8+-8.4 hours and a t1/2b of 70.9+-24.2 hours (Table 2). In patient No. 3, the mean peak blood level of 111In was 19.6% ID/L, with blood clearance of 6.8 h for t1/2a and 64.2 h for t1/2b. 
 Size exclusion HPLC was performed on a total of 35 plasma samples from the five patients. The radiochromatograms most often showed a prominent peak corresponding to that of labeled antibody and species with longer retention times. Fig. 3 shows the most pronounced examples of the radiochromatograms, which were obtained by analysis of the 1, 3, 24, 48, 72, and 120-h plasma from patient No. 5. In this case, the first peak predominanted labeled antibody and was therefore at least largely due to the 111In-labeled antibody still present in the plasma. The second peak coeluted with radiolabeled DTPA. However, radiolabeled DTPA, administered as a small radiocontaminant in the injectate, may be expected to clear rapidly into the urine from plasma by glomerular filtration. Since this peak was observed in plasma collected after 120-hour post-administration, it was likely that the species responsible was not labeled DTPA but was a catabolic product with a similar retention time in size exclusion HPLC. 
 The radiochromatogram obtained by analysis of plasma from patient No. 4, who had increased plasma CEA, did not show a peak due to immunocomplexes resulting from high circulating CEA levels. Analysis by HPLC also indicated that there was no detectable transchelation of 111In from IgG to transferrin or to some other proteins. 

Urine analysis 

A 72-hour cumulative urinary excretion was 20.4+-1.4% ID and approximately 78% of them was excreted in 24 hours after administration (Fig. 4). As in the case of blood, 111In was excreted in the urine with a similar pattern regardless of the tumor burden or plasma CEA level. 
 Size exclusion HPLC analysis showed a slight difference between early and late urine collection in the prominent peak of 18 min-retention time which differed from patient to patient. Radiochromatograms of urine sample from patient No. 4 at 0-24 h showed two distinct peaks (Fig. 5). The first peak with 10 min retention time was found only in the 0-24 h urine sample. The second (18 min-retention time) was exclusively present in the 48-h urine sample and all urine samples from patient No. 5 (Fig. 6). At 24 hrs, radioactivity in urine was primarily due to radiolabeled free DTPA excreted into urine following its introduction into plasma as a small radiocontaminant of the injectate. This 18 min-peak was still observed in the late urines, which might possibly mean a product of antibody catabolism. 

Imaging studies and surgical findings 

Gamma camera images were obtained 1 and 3 days after antibody injection. Specific antibody localization was assessed by correlating immunoscintigraphy with (1) Ba enema, (2) CT scan, (3) surgical findings, (4) pathological findings, and (5) radioactivity measurements in tumor and the adjacent colon tissue. 
 With progressive 111In-ZCE-025 clearance from the blood, metastatic foci in the liver in patient No. 4 as well as primary tumor sites in all four patients became obvious by day 3. The percentage of the injected dose per gram of primary tumors ranged from 1.2 to 2.8x10-2 (Table 1). There was no apparent relation of CEA content to the clarity of tumor images or the percentage of the injected dose of antibody taken up by the tumor. 
 A typical case with colon carcinoma is shown in Fig. 7. In some patients, the normal colon was faintly visualized in the 24 hr-image. As significant radioactivity could be removed from the bowel as time passed, this would suggest that some 111In activity is present in the feces. Furthermore, exvivo scan of species resected showed faint 111In uptake in the normal colon (Fig. 7B, C). Tsutsumi et al. reported that CEA was stained on the luminal surface of normal colonic epithelium,11 which was also indicated by our immunostaining findings in Fig. 7D. Two lymph nodes (12b1 and 12b2) in patient No. 4, which were clearly visualized in the SPECT image, were pathologically confirmed to have no tumor cells. A detailed description of the imaging results obtained in these patients has been included in other papers.12,13 


DISCUSSION 

All five patients who enrolled in the investigation were suspected to have colorectal carcinoma from the Ba-enema and/or CT-scan, and all four primary tumors could be visualized by immunoscintigraphy. Detectability cannot be discussed here since the number of cases studied was too small. It should be highly appreciated that the tumors that produced CEA accumulated 111In-labeled anti-CEA MoAb. Furthermore, it is noteworthy that the lesion, as in patient No. 3, which was confirmed to be pathologically non-malignant (a benign structure caused by previous surgery), was determined to be negative in the immunoscintigraphy, since it is very difficult to differenciate tumor recurrence from postoperative granuloma by other modalities. 
 The injectate contained a small amount of free 111In-DTPA throughout the investigation. The free 111In-DTPA was not detected in the first labeling vial before the addition of cold MoAb and the 100-ml saline, suggesting that 111In attached to MoAb partly dissociated from MoAb during dilution with MoAb and/or saline. The presence of free 111In-DTPA in the injectate would not affect the image qualities because of rapid excretion of 111In-DTPA in the urine. 111In was still attached to MoAb in the plasma, meaning that 111In-labeled MoAb was quite stable in the blood. Some reports described the presence of the immunocomplex between the antibody injected and circulating antigen.14,15 We experienced one patient (No. 4) with slightly increased plasma CEA, and observed no immunocomplex in the plasma in size-exclusion HPLC. 
 Few reports have described 111In-MoAb characterization in cumulative urine excretion. Patt et al. reported that 48 h-cumulative urine excretion from patients with colorectal cancer under a similar regime to ours was 9.9+-2.2%, which was lower than in our results (18.0+-1.1%).3 The clearance rate of 111In in the urine did not vary among patients, meaning that it did not depend on the tumor burden or on the plasma CEA level. What was bound with 111In in the urine was different from patient to patient, suggesting that it depended on several factors including renal function. In general, 111In was cleared into the urine primarily as 111In-DTPA. 
 The uptake of 111In in tumor tissues was not high, as already pointed out in other papers.16,17 When 111In-ZCE-025 was injected into nude mice bearing human gastric cancer MKN45 producing CEA, 111In-ZCE-025 was localized in the tumor with 11.5% ID/g 2 days after the injection. The reason for the discrepancy between clinical and animal experiment results remains unclear. 
 The uptake of 111In in tissues was not correlated with their CEA content, which disagrees with Duda's report.17 The reason might be the different accessibility of each monoclonal antibody. 
 One regional lymph node resected by surgery, which was confirmed to be pathologically malignant, was not visualized in the immunoscintigraphy because it was located very near the primary colon tumor. On the other hand, some lyrnph nodes (in patient No. 4, 12b1 and 12b2), which were visualized in both planar and SPECT images with intensive 111In uptake (0.41 and 0.03 % ID/g, respectively), did not contain high CEA (86 and 37 ng CEA/mg prot., respectively), compared with primary tumor (771 ng CEA/mg prot. and 0.01% ID/g of 111In uptake). Furthermore, they were histopathologically confirmed to have no tumor cells. As shown in Fig. 7, some spots showed intensive 111In activity in exvivo scans of the specimens resected. They were also confirmed to be pathologically non-malignant. False-positive cases in the immunoscintigraphy were reported in several papers.4,19,20 They pointed out the following factors as the reasons for the false positive ; (1) involvement of macrophages, (2) the presence of a blood pool, and (3) presence of shedding CEA or antigen. Actually 111In-ZCE-025 was localized in the abscess containing macrophages, which were formed in nude mice with 6.03% ID/g, which was lower than that in tumor MKN45 (11.5% ID/g) grown on the same mice. The lymph nodes we discussed in patient No. 4 did not contain high shedding CEA (10.5 ng/ml) and were not stained positively with anti-CEA MoAb immunohistologically. The lymph nodes did not contain many macrophages. Our false positive result, therefore, could not be explained by any factors already described. As shown in Table 1, immunohistological findings indicated that CEA was present in some normal colon, but 111In-ZCE-025 did not concentrate at this site, possibly becauseIof poor accessibility to the antigen. The cross reactivity of ZCE-025 or antigen accessibility in the lymph nodes might be other factors as the reasons for the false positive result. 
 In conclusion, immunoscintigraphy with 111In-labeled ZCE-025 shows promise in detecting CEA-producing tumor and appears to be useful in distinguishing recurrent tumor from glanuloma which other imaging modalities failed to depict. The quality of our images was almost the same as those already reported, although the pharmacokinetics of 111In-ZCE-025 was slightly different from them. 111In are present in the blood or in the urine simply in the form of 111In-MoAb or 111In-DTPA. The accumulation mechanism of 111In-labeled MoAb in tumors as well as in some normal tissues needs further study in order to develop MoAb targeting for the radioimmunotherapy of cancer. 


ACKNOWLEDGMENTS 

This work was supported in part by Grant 63480254 from the Ministry of Educations and a grant from the Ichiro Kanehara-Foundation. The authors wish to thank Teijin Limited for its cooperation in making ZCE-025 available to us and to Toshikazu Sammiya of Keio University for his help in obtaining the images. 


REFERENCES 

1. Buchegger F, Haskell CM, Schreyer M, et al: Radiolabeled fragments of monoclonal antibodies against carcinoembryonic antigen for localization of human colon carcinoma grafted into nude mice. J Exp Med 158: 413-427, 1983 
2. Abdel-Nabi HH, Schwartz AN, Higano CS, et al: Colorectal Carcinoma: Detection with indium-111 Anticarcinoembryonic-Antigen Monoclonal Antibody ZCE-025. Radiology 164: 617-621, 1987 
3. Patt YZ, Lamki LM, Haynie TP, et al : Improved Tumor Localization with Increasing Dose of Indium-111-Labeled Anti-Carcinoembryonic Antigen Monoclonal Antibody ZCE-025 in Metastatic Colorectal Cancer. J Clin Oncol 6: 1220-1230, 1988 
4. Duda RB, Zimmer AM, Rosen ST, et al: Radioimmune Localization of Occult Carcinoma. Arch Surg 125 : 866-870, 1990 
5. Halpern SE, Haindl W, Beauregard J, et al: Scintigraphy with In-111-labeled Monoclonal Antitumor Antibodies: Kinetics. Biodistribution, and Tumor Detection. Radiology 168 : 529-536, 1988 
6. Krejcarek GE, Tucker KL: Covalent attachment of chelating groups to macromolecules. Biochem Biophys Res Commun 77: 581-585, 1977 
7. Schreyer M, Haskell CM, Girardet C, et al: Localization of tumor-associated antigens on section of human colon carcinoma grafted into nude mice with mouse monoclonal antibodies using the avidin-bioth-immunoperoxidase reaction. In Proceedings of the Fourth International Workshop on Immune-deficient Animals in Experimental Research, Basel, S. Karger A.G., pp 287-290, 1984 
8. Nakamura K, Kubo A. Hashimoto S, et al: Radioimmunodetection of human pancreatic tumor xenografts using DU-PAN II monoclonal antibody. Ann Nucl Med 2: 1-6, 1988 
9. Gold P : Specific carcino embryonic antigens of the human digestive system. J Exp Med 122: 467-469, 1965 
10. Lowry OH, Rosebrough NJ, Farr AL, et al: Protein measurement with the Folin Phenol reagent. J Biol Chem 193 : 265-275, 1951 
11. Tsutsumi Y, Onoda N, Misawa M, et al: Immunohistochemical demonstration of nonspecific crossreacting antigen in normal and neoplastic tumor tissue using a monoclonal antibody comparison with carcinoembryonic antigen localization. Acta Pathol Jpn 40: 85-97, 1990 
12. Katayama M, Nakamura K, Kubo A, et al: Immunoscintigraphy of colorectal cancer with In-111-labeled anti CEA monoclonal antibody (ZCE-025). Jpn J Nucl Med 28 : 701-109, 1991 
13. Kubo A, Nakamura K, Kodaira S: False positive in Immunoscintigraphy. J Nucl Med 32: 2194-2195, 1991 
14. Mach JP, Buchegger F, Forrni M, et al: Use of radiolabeled monoclonal anti CEA antibodies for the detection of human carcinomas by external photoscanning and tomoscintigraphy. Immunol Today 2 : 239-249, 1981 
15. Hnatowich DJ, Rusckowski M, Brill AB, et al : Pharmacokinetics in Patients of an Anti-Carcinoem-bryonic Antigen Antibody Radiolabeled with Indium-111 Using a Novel Diethyleneamine Pentaacetic Acid Chelator. Cancer Res 50 : 7272-7278, 1990 
16. Sfakianakis GN, Garty II, Serafini AN : Radioantibodies for the Diagnosis and Treatment of Cancer; Radioimmunoimaging (RAI) and Radioimmunotherapy (RAT). Cancer Investigation 8 : 381-405, 1990 
17. Lamki LM: Radioimmunoscintigraphy of Cancer. Problems, Pitfalls, and Prospects. In Nuclear Medicine Annual 1990, Freeman LM, (ed.), New York, Raven Press, pp 113-150, 1990 
18. Duda RB, Beatty JD, Sheibani K, et al : Imaging of Human Colorectal Adenocarcinoma with Indium-Labeled Anticarcinoembryonic Antigen Monoclonal Antibody. Arch Surg 121: 1315-1319, 1986 
19. Patt YY, Lamki LM, Shanken J, et al: Imaging with indium-111-Labeled Anticarcinoembryonic Antigen Monoclonal Antibody ZCE-025 of Recurrent Colorectal or Carcinoembryonic Antigen-producing Cancer in Patients with Rising Serum Carcinoembryonic Antigen Levels and Occult Metastases. J Clin Ocnol 8 : 1246-1254, 1990 
20. Abdel-Nabi HH, Chan HW, Doerr RJ: Indium-Labeled Anti-Colorectal Carcinoma Monoclonal Antibody Accumulation in Non-Tumored Tissue in Pstients with Colorectal Carcinoma. J Nucl Med 31 : 1975-1979, 1990