Summary Human IgM monoclonal antibody (MAb) 16.88 recognizes an antigen strongly expressed by ... tumour deposits, a prerequisite for MAb-guided therapy.
40-43 (1990), 62, Suppl. X, X, 40-43
Br. J. Cancer Br. J. Cancer (1990), 62, Suppi.
Macmillan Press Ltd., 1990
Human IgM monoclonal antibody 16.88: pharmacokinetics and distribution in mouse and man H.J. Haisma', M.A.P. Kessell, C. Silva', M. R. McCabe' & E. Boven'
Muijen', J.C. Roos2, H. Bril3, H.J.M. Martens4,
Departments of 'Oncology, 2Nuclear Medicine, 3Pathology and 4Pharmacy, Free University Hospital, 1081 HV Amsterdam, Netherlands, and 5Bionetics Research Institute, Rockville MD 20850-4373, USA. Summary Human IgM monoclonal antibody (MAb) 16.88 recognizes an antigen strongly expressed by colon tissue. We used '3ll-labelled 16.88 for biodistribution and pharmacokinetic studies in nude mice bearing WiDr or NIH:OVCAR-3 xenografts. Serum half-life was 8 h. Maximum tumour uptake was between I and 8 h after administration and amounted to, respectively, 3% and 1% of the injected dose g ' for WiDr and NIH:OVCAR-3 tumours. Half-lives in these tumours were approximately 24 h. Tumour to normal colon uptake ratios increased from 2.3 at 24 h to 17 at 5 days after injection. Simultaneously, pharmacokinetic studies were performed in patients with advanced colon cancer reactive with 16.88. They were injected with 5 mCi '3'I-16.88 by intravenous infusion over 2 h. Serum half-life was 20 h with > 90% of the 'l'I bound to 16.88. Within 40 h 50% of the injected dose was excreted as free "'I in the urine. In one patient an accelerated clearance was found, possibly caused by pre-existing antibodies reacting with 16.88. None of the patients showed an immune response against 16.88 antibody. Immunoscintigraphy showed positive tumour localization in the majority of the patients, best visualized at later days. We conclude that 16.88 has tumour localization properties while its human origin accounts for the lack of immunogenicity. cancer
Clinical trial with murine monoclonal antibodies (MAb) have demonstrated the preferential localization of antibody in tumour deposits, a prerequisite for MAb-guided therapy. A major problem associated with the clinical use of such monoclonal antibodies is the development of anti-mouse antibodies in about 50% of the patients. Human MAb would alleviate these problems, making multiple injections of antibody a viable proposition. Two human IgM MAbs, 16.88 and 28A32 (Haspel et al., 1985), were developed from peripheral blood lymphocytes of colorectal cancer patients participating in an active immunotherapy trial for treatment of residual disease (Hoover et al., 1985). The antibodies were selected for their specificity, stability and growth properties. The purpose of the present study was to determine the pharmacokinetics and distribution of radioiodinated IgM MAb 16.88 in nude mice bearing human tumour xenografts and in patients with advanced colon cancer. Patients, materials and methods Cell lines The colon cancer cell line WiDr (Noguchi et al., 1979) and the ovarian cancer cell line NIH:OVCAR-3 (Hamilton et al., 1984) were grown as monolayers in Dulbecco's Modified Eagles Medium supplemented with 10% heat inactivated fetal calf serum (FCS). Cells were harvested with 0.2% ethylenediaminetetracetic acid (EDTA) in phosphate buffered saline (PBS). The xenografts were established by subcutaneous injection of 107 cells in female NMRI-nu/nu/Cpb/ Hsd mice and maintained by serial transplantation of fragments of 2-3 mm3. Immunoperoxidase studies of tumour tissue revealed a positive staining reaction with 16.88 in both WiDr and NIH:OVCAR-3.
Antibodies and radiolabelling Human IgM MAb 16.88 was obtained from Bionetics Research Institute, Rockville MD, USA. Labelling of 16.88
Correspondence: H.J. Haisma.
with '"'I and of control human IgM (Sigma, St. Louis, MO, USA) with '25I was performed according to the one vial method (Haisma et al., 1986). For clinical applications, 10 mg 16.88 was labelled with 15 mCi '3'I and the unbound iodine was removed with a Sephadex (Pharmacia, Uppsala, Sweden) column. Specific activities were approximately 1 mCi mg-' antibody. Precipitation with 10% trichloroacetic acid (TCA) indicated that 95% of the radioactivity was bound to protein in the final preparations.
Immunoreactivity of radiolabelled antibody Binding assays for iodinated 16.88 and control IgM were set up according to Lindmo et al. (1984) using WiDr cells harvested from tissue culture and fixed with 0. 1% glutaraldehyde. Immunoreactivity of the '3'I-16.88 final preparation was between 80% and 90%, whereas '25I-IgM did not bind to WiDr or NIH:OVCAR-3 target cells. Nude mouse studies Radiolabelled 16.88 and control IgM (10 ,ug each) were diluted in 0.1 ml PBS and injected into the retro-orbital vein of mice bearing WiDr or NIH:OVCAR-3 xenografts with a tumour volume of 150-300 mm3. At 1, 4, 8, 24, 48, 168 and 196 h after injection the mice were sacrificed by ether anaesthesia. Three mice were used for each determination. Blood was collected and tissues were dissected from the mice and weighed, after which the radioactivity was meaured by gamma counting. The antibody uptake in the tumours and other tissues was calculated as the percentage of the injected dose g-' of tissue.
Clinical studies Formalin-fixed paraffin-embedded tissue sections from patients with advanced colon cancer were stained with 16.88 at 0.3 fg ml-' using the immunoperoxidase technique. Ten patients with positive staining of tumour sections were entered into the study after giving their informed consent. Patients received sodium perchlorate or potassium iodide prior to and continuing for 4 days after injection of the antibody. The antibody dose was 8 mg with approximately 5 mCi '3'I administered over a period of 2 h. Imaging was performed at the time of injection and at 2, 5 and 7 days
RADIOLABELLED HUMAN IgM 16.88 IN COLON CANCER
after injection using a gamma camera equipped with a medium energy collimator. Blood samples were obtained immediately before, within 15min after completion of the infusion and subsequently at various intervals for up to 2 weeks. Serial urine specimens were also collected from each patient. Radioactivity remaining in serum and urine was determined by counting 100 fil samples. The percentage injected dose was calculated by comparing the counts in the samples with an appropriate standard. The proportion of radioactivity associated with protein in serum and urine samples was determined by precipitation with 10% (w/v) trichloroacetic acid. Gel filtration analysis of serum and urine samples was performed using a Superose 6 column (Pharmacia, Uppsala, Sweden) with PBS as eluent. OD280 and radioactivity of samples were recorded. In an additional test, serum samples were assayed for ''I-16.88 immunoreactivity as described above, using 50 gl of serum and 5 x 106 glutaraldehyde-fixed WiDr cells.
Anti-16.88 assay For the detection of an antibody response against 16.88, a latex agglutination assay was used employing 16.88 or human IgM-coated beads. The sensitivity of this assay was 0.1 fg ml-' with goat anti-human IgM (Cappel, West Chester, PA, USA). An ELISA system, using 16.88 or human IgM-coated wells and peroxidase conjugated goat anti-human IgG (Cappel, West Chester, PA, USA) was used to detect IgG antibodies. Both the latex bead agglutination assay and ELISA were performed on patients' sera obtained pre-injection, and at 2 and 3 weeks thereafter. Results Nude mouse studies The pharmacokinetics and biodistribution of 3'I-16.88 were investigated in nude mice bearing WiDr or NIH:OVCAR-3 xenografts. The serum half-life of 16.88 and control IgM in both groups was 8 h (Figure 1). Biodistribution results are shown in Figure 2. 16.88 specifically accumulated in the tumours and had a slower clearance from tumours than from normal organs. This resulted in tumour to tissue ratios ranging from 7 (liver) to 36 (blood) at 10 days after injection. Tumours to normal colon uptake ratios increased from 2.3 at 24 h to 17 at 5 days after injection. Absolute uptake of antibody in the tumours was low, with maximum uptake between I and 8 h after injection which amounted to, respectively, 3% and 1% of the injected dose g' for WiDr and NIH:OVCAR-3 tumours. Uptake decreased to less than 0.01% of the doseg-' at day 7. Radiolabelled IgM showed similar kinetics and distribution as 16.88 but was not retained in tumours. The localization index (ratio of uptake of specific antibody and non-specific antibody, corrected for blood levels) was investigated in NIH:OVCAR-3 xenografts. 100
c0\ *cD0.01 0.001, 0
100 150 200 250 Hours after injection Figure 1 Pharmacokinetics of 3'I-labelled 16.88 in blood of nude mice bearing WiDr (+) or NIH:OVCAR-3 (*) xenografts. The difference in blood clearance was not significant.
V30O 0 a
Figure 2 Biodistribution of '3'I-labelled 16.88 in blood, tumours and other organs of nude mice bearing a WiDr or b NIH:OVCAR-3 xenografts. _ 4h; M 24h; M 240h. It increased from 1.0 at 24 h to 10.9 at 10 days after injection.
Clinical studies Ten patients with 16.88 positive staining of their tumour tissue received 8 mg '3'I-!abelled 16.88 (5 mCi). The pharmacokinetics of the radionuclide in the serum and urine are shown in Figure 3a and b. As all patients received the same dose, the serum clearance values were expressed per surface area in order to correct for blood volume. The average half-life in the serum was 20 h with more than 90% of the radioactivity TCA-precipitable. Immediately after injection 90% of the radioactivity in the serum could be shown to be immunoreactive, while lower immunoreactive fractions were found at 24 (85%) and 48 h (80%) after injection. Elimination of the radionuclide was via the urine, with 50% of the dose excreted in 40 h. We observed no severe adverse reactions after the administration of '3'I-labelled 16.88 MAb. Immediately after completion of the infusion an average peak concentration of 14.6% dose litre-' serum m2 was found in nine patients. In one patient only 2.5% of the dose litre-' serum m-2 remained at the end of infusion. Urinary excretion in this patient was similar to that of the other patients. Apparent antibodies reactive with 16.88 could be detected in sera obtained pre-injection and 2 weeks thereafter, using either 16.88-coated latex beads or the ELISA system. In both assays reactions could be blocked by the addition of excess cold 16.88 antibody. The apparent antibodies were not specific for 16.88 because reactivity was also observed with control human IgM. In all patients gel filtration analysis of serum samples revealed a major peak (>90%) of radiolabelled IgM and a smaller peak of free '3'I. Urine samples contained only free In six patients tumour deposits could be localized. In general, lesions smaller than 4 cm in diameter could not be visualized. In four of six patients with liver metastases cold areas were detected in the liver, corresponding to the tumour locations. These areas were filled in with radioactivity at 5 days after injection.
H.J. HAISMA et al.
Figure 3 a Pharmacokinetics of I3i in serum of patients after injection of "'I1-labelled 16.88, corrected for surface area (s.e.m. < 15%). b Urinary clearance of '3II in patients after injection of i"IIlabelled 16.88 (s.e.m* < 5%);
Discussion Murine MAbs are extremely valuable for in vitro diagnostic conjugation with radionuclides they are being investigated for tumour localization in clinical trals (Boven and Pinedo, 1985; Haisma et at., 1987). Antibodies conjugated to radionuclides drugs or toxins are being studied as new anti-cancer agents in patients (Carrasquillo et atl. 1984; Pimm et at., 1987; Spitler et at. 1987). A major problem associated with the clinical use of munine MAbs is the development of anti- mouse antibodies which may result in neutralization of the injected antibody or allergic reactions (Herlyn et al., 19858 Shawler et a, 1985 Moseley et a., 1988). Human MAbs or mouse/human chimeric MAbs would alleviate these problems, making multiple injections of antibody a viable proposition In this study we determined the distribution and pharmacokinetics of one of the first human MAbs, 16.88 which is being used in clinical trials (Steis et at., 1989). The human igM MAb 16d88 was demonstrated to preferentially localize in WiDr colon cancer and NIH:OVCAR-3 ovarian cancer xenograftst Control IgM was not retained in tumours in contrast to 1688i The localization index in tumours increased from 1 at 24 h to 10.9 at 10 days after injection. This clearly showed the tumour localizing properties of 16.88 antibody regardless of its direction against a cytoplasmic antigen. In this respect the antibody could have been bound to locally released antigen from dead or dying tumour cells Similarly Welt et at (1987) and tests. After
Dairkee & Hackett (1988) have shown tumour localization of monoclonal antibodies to intracellular antigens in, respectively, human melanoma and breast cancer. The amount of antibody taken up by the tumours was highest at 1-8 h after injection with a maximum of 3% of the injected doseg'I tumour tissue. Serum half-life of the radionuclide was 8 h for mice bearing either xenograft. This clearance rate is in good agreement with other studies in which IgM antibodies were injected into tumour-bearing mice (Ballou et al., 1985; Halpern et al., 1988). In a similar approach, McCabe et al. (1988) have studied 16.88 in another xenograft model (THO) and found a 6-8 h serum half-life and a tumour uptake of antibody of 1% of the injected dose g' at 3 days after injection. Including our study, these data demonstrate that IgM MAbs are cleared more rapidly from the circulation than gG MAbs. Uptake of IgM antibodies in tumours was approximately ten times less than uptake of relevant IgG antibodies, possibly due to the rapid blood clearance and slow tissue penetration of the large IgM molecules (Jain & Baxter, 1988). Ten patients were injected with 8 mg '31I-labelled 16.88 (5 mCi). Analysis of serum showed that '3'I remained associated with immunoreactive 16.88. Serum half-life of '31I-labelled 16.88 in patients was 20 h. This differs from a half-life of around 30 h for IgG, 25 h for F(ab')2 fragments and 2 h for Fab (Zuckier et al., 1989). Elimination of the radioactivity was via the urine with 50% of the dose excreted as free 'l'I within 40 h. The biological half-life of native IgM in man is 5.1 days (Barth et al., 1964). The reason for the short halflife of 16.88 in serum is not yet clear. It may be because of clonal variation in carbohydrate composition, or due to production and purification methods. The difference in kinetics between IgM MAbs and IgG MAbs may be explained by the carbohydrate composition of IgM and IgG molecules. It has been suggested that IgM with a carbohydrate content of 7-11% can be cleared faster by the liver than IgG which contains less than 5% of carbohydrate (Halpern et al., 1988). Immunoscintigraphy showed positive localization of 16.88 in six of 10 patients. The best images were produced 5 and 7 days after injection. The percentage of positive images (60%) in our small series of patients is similar to that of other studies using 31I-labelled IgG derived antibodies (Goldenberg et al., 1980; Mach et al., 1983). Thus far we have found one patient with a rapid clearance who apparently had antibodies reactive with 16.88 and human IgM. Sera of this patient, obtained before the injection and at 2 and 3 weeks, showed a positive reaction in an assay using 16.88-coated latex beads, as well as in an ELISA system with immobilized 16.88 and anti-human IgG conjugate. In both systems the reaction could be blocked with excess 16.88 antibody, but non-specific reactions with IgM antibody were also noted. Analysis for the presence of rheuma factors in this serum was negative. None of the other patients' sera showed a positive reaction in the ELISA or latex bead assay, indicating the lack of immunogenicity of human IgM 16.88 after a single injection of 8 mg of antibody. In conclusion, human IgM MAb 16.88 can specifically localize in antigen-positive tumours in patients whereas immunogenicity has not yet been detected. Our results with 16.88 may help to design further studies to improve human MAb-guided immunotherapy. This work was supported in part by a grant from the 'Praeventiefonds'.
References BALLOU, B., REILAND, J., LEVINE, G., KNOWLES, B. & HAKALA, T.R.
(1985). Tumour localization using F(ab')2u from a monoclonal IgM antibody: pharmacokinetics. J. Nucl. Med., 26, 283. BARTH, W.F., WOCHNER, R.D. & WALDMAN, T.A. (1964). Metabolism of gamma macroglobulins. J. Clin. Invest., 43, 1036.
BOVEN, E. & PINEDO, H.M. (1985). Monoclonal antibodies in cancer treatment: where do we stand after ten years? Radiother. Oncol., 5, 107.
RADIOLABELLED HUMAN IgM 16.88 IN COLON CANCER CARRASQUILLO, J.A., KROHN, K.A., BEAMIER, P. & 5 others (1984). Diagnosis of and therapy for solid tumours with radiolabelled antibodies and immune fragments. Cancer Treat. Rep., 68, 317. DAIRKEE, S.H. & HACKETT, A.J. (1988). Internal antigens accessible in breast cancer: implications for tumour targeting. J. Natl. Cancer Inst., 80, 1216. GOLDENBERG, D.M., KIM, E.E. & DELAND, F.H. (1980). Radioimmunodetection of cancer with radioactive antibodies to carcinoembryonic antigen. Cancer Res., 40, 2984. HAISMA, H.J., HILGERS, J. & ZURAWSKI, V.R. (1986). lodination of monoclonal antibodies for diagnosis and therapy using a convenient one vial method. J. Nucl. Med., 27, 1890. HAISMA, H.J., GOEDEMANS, W. & HILGERS, J. (1987). Tumour imaging with monoclonal antibodies. Res. Monogr. Immunol., 11, 171. HALPERN, S.E., HAGAN, P.L., CHEN, A. & 7 others (1988). Distribution of radiolabelled human and mouse IgM antibodies in murine models. J. Nucl. Med., 29, 1688. HAMILTON, T.C., YOUNG, R.C., LOUIE, K.G. & 4 others (1984). Characterization of a xenograft model of human ovarian carcinoma which produces ascites and intraabdominal carcinomatosis in mice. Cancer Res., 44, 5286. HASPEL, M.V., MCCABE, R.P., POMATO, N., HOOVER, H.C. & HANNA,
M.G. (1985). Human monoclonal antibodies: generation of tumourcell reactive monoclonal antibodies using peripheral blood lymphocytes from actively immunized patients with colorectal carcinoma. In Monoclonal Antibodies in Cancer Therapy, Reisfeld, R.A. & Sell, P. (eds), p. 505. New York. HERLYN, D., LUBECK, M., SEARS, H. & KOPROWSKI, H. (1985). Specific detection of anti-idiotype immune responses in cancer patients treated with murine monoclonal antibodies. J. Immunol. Methods, 85, 27. HOOVER, H.C, SURDYKE, M.G., DANGLE, R.B., PETERS, L.C. &
HANNA, M.G. (1985). Prospective randomized trial of adjuvant active specific immunotherapy for human colorectal cancer. Cancer, 55, 1235. JAIN, R. & BAXTER, L. (1988). Mechanism of heterogeneous distribution of monoclonal antibodies and other macromolecules in tumours: significance of elevated interstitial pressure. Cancer Res., 48, 7022.
LINDMO, T., BOVEN, E., CUTTITTA, F., FEDORKO, J. & BUNN, P.A. Jr
(1984). Determination of the immunoreactive fraction of radiolabelled monoclonal antibodies by linear extrapolation to binding at infinite antigen excess. J. Immunol. Methods, 72, 77. MACH, J.P., CHATAL, J.F. & LUMBROSO, J.D. (1983). Tumour localization in patients by radiolabelled monoclonal antibodies against colon carcinoma. Cancer Res., 43, 5593. MCCABE, R.P., PETERS, L.C., HASPEL, M.V., POMATO, N., CARRASQUILLO, J.A. & HANNA, M.G. (1988). Preclinical studies on the pharmacokinetic properties of human monoclonal antibodies to
colorectal cancer and their use for detection of tumours. Cancer Res., 48, 4348. MOSELEY, K.R., KNAPP, R.C., & HAISMA, H.J. (1988). An assay for the detection of human anti-murine immunoglobulins in the presence of CA125 antigen. J. Immunol. Methods, 106, 1. NOGUCHI, P., WALLACE, R., JOHNSON, J. & 8 others (1979). Characterization of WiDr: a human colon carcinoma cell line. In Vitro, 15, 401. PIMM, M.U., ARMITAGE, N.C. & BALLANTYNE, K. (1987). Imaging of primary and metastatic colorectal carcinoma with monoclonal antibody 791 T/36 and the therapeutic potential of antibody-drug conjugates. Cancer Detect. Prev., 1, 249. SHAWLER, D., BARTHOLOMWE, R., SMITH, L. & DILLMAN, R. (1985).
Human immune response to multiple injections of murine monoclonal IgG. J. Immunol., 135, 1530. SPITLER, L.E., DER RIO, M. & KHENTIGAN, A. (1987). Therapy of patients with malignant melanoma using a monoclonal antimelanoma antibody-ricin A chain immunotoxin. Cancer Res., 47, 1717. STEIS, R.G., CARRASQUILLO, J.A., MCCABE, R. & 15 others (1989). An evaluation of the toxicity, immunogenicity, and tumour radioimmunodetection ability of two human monoclonal antibodies in patients with metastatic colorectal carcinoma. J. Clin. Oncol., (in press). WELT, S., MATTES, M.J., GRANDO, R. & 7 others (1987). Monoclonal antibody to an intracellular antigen images human melanoma transplant in nu/nu mice. Proc. Nat!. Acad. Sci. USA, 84, 4200. ZUCKIER, L.S., RODRIGUEZ, L.D. & SCHARFF, M.D. (1989).
Immunologic and pharmacologic concepts of monoclonal antibodies. Semin. Nucl. Med., 19, 166.