314
Bioconjugate Chem. 2010, 21, 314–318
Radioimmunotherapy of Human Colon Cancer Xenografts with Anti-CEA Monoclonal Antibody
131
I-Labeled
Zhaofei Liu,# Cunjing Jin,# Zilin Yu, Jing Zhang, Yan Liu, Huiyun Zhao, Bing Jia,* and Fan Wang* Medical Isotopes Research Center, Peking University, Beijing 100191, China. Received August 12, 2009; Revised Manuscript Received December 15, 2009
Radioimmunotherapy (RIT) is a promising approach for the treatment of a wide variety of malignancies. The aim of this study was to investigate the therapeutic efficacy of 131I-labeled anticarcinoembryonic antigen (CEA) monoclonal antibody CL58 in a human colon cancer mouse model. In vitro and in vivo characteristics of 125ICL58 were evaluated in LS180 human colon cancer cells and the nude mouse model. 131I-CL58 was prepared and its in vivo therapeutic efficacy was tested. 125I-CL58 showed high affinity to LS180 cells, as well as high tumor uptake and long tumor retention in LS180 tumor xenografts. 131I-CL58 exhibited dose-dependent inhibition of LS180 tumor growth. With the excellent in vitro and in vivo characteristics, and the effective therapy for colon cancer in animal model, 131I-CL58 is a promising agent for RIT of CEA-positive tumors including colon cancer.
INTRODUCTION Colorectal cancer, accounting for higher morbidity and mortality, is the second most lethal and third most common malignancy in western developed countries (1) and the third common gastrointestinal tumor in China (2, 3). Although multiple kinds of managements in prevention, screening, treatment, and surveillance of this disease have been developed systematically in the last decades, the major cause of cancer mortality is most likely due to dissemination and distant metastasis. In the exploration of methods to manage colorectal cancer, chemotherapy is the principal adjuvant therapy, and the addition of radiotherapy to chemotherapy has not been shown to improve outcomes (4). To reduce the systemic toxicity elicited from the chemotherapy, tactics involving biomarkers on cancer cells recognized by exogenous targeted molecules (e.g., peptides, oligonucleotides, aptamers, and antibodies) were conducted. Among these investigations, radioimmunitherapy (RIT), which typically consists of radionuclide and antibody, has been investigated extensively as a useful strategy for the treatment of a variety of neoplasms (5-7). For the agents of RIT, the antibody acts as a vehicle for selective delivery of cytotoxic radionuclides to destroy the tumor while sparing the normal organs and tissues. FDA approval of two radiolabeled antibodybased products for the treatment of non-Hodgkin’s lymphoma (NHL) was an initial success of hardworking investigation of the subject of RIT in the past decades (8, 9). Although the efficacy of RIT for solid tumors was less encouraging than that of hematologic tumors (10), it indeed gives hope for this subject; moreover, some lessons and experiences have already been acquired for further investigation. Many colorectal cancer markers, such as carcinoembryonic antigen (CEA), TAG-72, A33, Ep-CAM (17-1A), and CSAp, have been studied as the therapeutic targets in preclinical or * To whom correspondence should be addressed. Dr. Fan Wang, Medical Isotopes Research Center, Peking University, 38 Xueyuan Road, Beijing 100191, China. Phone and Fax: 86-10-82801145, E-mail:
[email protected]; or Dr. Bing Jia, Phone: 86-10-82802871, Fax: 86-10-82801145, E-mail:
[email protected]. # These authors contributed equally to this work.
clinical settings (11). Among all the colon cancer markers, CEA was first described as a gastrointestinal onco-fetal antigen, although it is now well-known to be overexpressed in a majority of carcinomas, including gastrointestinal cancers, respiratory and genitourinary cancers, and also breast cancers (12, 13). 131Ilabeled anti-CEA antibodies for RIT of colon cancer have shown encouraging results in preclinical and clinical trials (14, 15). Moreover, an adjuvant regimen of RIT showed exciting results in patients with colorectal cancer and liver metastases (16). CL58 is a murine anti-CEA monoclonal antibody (IgG1 subtype). Previous studies have shown the successful application of radiolabeled CL58 for CEA-positive tumor imaging and radioimmunoguided surgery (17, 18). In this study, in vitro and in vivo characteristics of iodinated CL58 were evaluated, and the RIT efficacy of 131I-labeled CL58 in nude mice bearing human colon cancer xenografts was investigated.
EXPERIMENTAL PROCEDURES Radiolabeling of mAb. The murine anti-CEA monoclonal antibody CL58 was generated as previously reported (17, 18). An isotype-matched control murine IgG1 (mIgG) was purchased from Sigma. Iodination of mAb CL58 or mIgG was performed using Iodogen method as previously described (19, 20). Briefly, 2 mCi (74 MBq) of 125I or 131I (Beijing Atom High Tech, Beijing, China), 50 µg of mAb CL58 or mIgG, and 100 µL of 0.2 M phosphate buffer (pH 7.4) were added in the vial coated with 50 µg Iodogen (Sigma-Aldrich, St. Loius, MO). After reacting for 7-10 min at room temperature, the mixture was purified with PD-10 column (Amersham, Piscataway, NJ). Labeling yield and radiochemical purity of the products were measured by ITLC (85% methanol as the mobile phase) on a radio-thin layer scanner (Bioscan AR2000, Washington, DC). Cell Culture and Animal Model. The human colon carcinoma cell line LS180, with a cell surface CEA expression rate of 81% (17), was maintained in Dulbecco’s modified Eagle medium (DMEM) supplemented with 10% fetal bovine serum (FBS) at 37 °C under a humidified atmosphere containing 5% CO2. All animal experiments were performed in accordance with Guidelines of Peking University Health Science Center Animal Care and Use Committee. Five-week-old female BALB/c nude mice were purchased from the Department of Experimental
10.1021/bc9003603 2010 American Chemical Society Published on Web 01/15/2010
RIT of
131
I-CL58 for CEA-Positive Tumor
Bioconjugate Chem., Vol. 21, No. 2, 2010 315
Animal, Peking University Health Science Center (Beijing, China). The LS180 tumor model was generated by subcutaneous injection of 1 × 107 LS180 cells into the left flanks of the nude mice. Affinity of 125I-CL58. The equilibrium dissociation constant (Kd) of 125I-CL58 was determined by saturation binding study using the previously described method with some modifications (20). Briefly, LS180 cells seeded in 48-well plates were incubated in triplicate with increasing concentration of 125I-CL58 (from 20 ng/mL to 1.2 µg/mL, in 1% BSA binding buffer). The total volume of each well was adjusted to 200 µL. Nonspecific binding at each concentration was determined by adding an excess dose of unlabeled CL58 (∼100-fold). After 2 h of incubation at 4 °C, the reaction medium was removed and cells were washed with ice-cold PBS. The cells were lysed with 2 M NaOH, and the cell-associated radioactivity was measured in a gamma counter (Wallac 1470-002, Perkin-Elmer, Finland). Results were analyzed by nonlinear regression (one-site binding) using GraphPad Prism 4.0 (GraphPad Software, San Diego, CA) to obtain the best-fit Kd and Bmax values, and a Scatchard transformation was subsequently performed. Data were expressed as the average of triplicate samples.
Figure 1. Saturation binding of 125I-CL58 to EGFR on LS180 human colon cancer cells. Results were analyzed by nonlinear regression using GraphPad Prism 4.0 software, and a Scatchard transformation was subsequently performed. Data are expressed as the means ( standard deviation (SD) (n ) 3).
Biodistribution. LS180 tumor-bearing nude mice were randomly divided into groups (n ) 4/group) when the tumor size reached 0.5-0.8 cm in diameter (∼10 days after inoculation). Each mouse was injected intravenously with 100 µL of 125 I-CL58 (5 µCi, 5 µg). At selected time points (4 h; 1, 3, 5, and 7 d), mice were euthanized, and samples or organs of interest were removed, weighed, and counted in a gamma counter. The results were calculated as a percentage of the injected dose per gram tissue (%ID/g). To determine the specific accumulation of 125I-CL58 in LS180 tumors, 125I-mIgG (5 µCi, 5 µg) was injected intravenously into a group of 4 tumor-bearing mice, and the biodistribution (3 d postinjection) was determined as depicted above. γ-Imaging. For planar gamma imaging studies, two LS180 tumor-bearing mice were injected intravenously with 200 µCi of 125I-CL58 via tail vein after anesthetization. Mice were placed prone on a two-head γ-camera (Siemens, E.Cam) equipped with a parallel-hole, low-energy, and high-resolution collimator. Imaging was performed at 1, 2, and 6 d postinjection. After completion of imaging, mice were euthanized. Radioimmunotherapy of 131I-CL58. Therapy studies of 131ICL58 were investigated in LS180 tumor-bearing mice (the original tumor volume was 135 ( 58 mm3). Animals were randomly divided into six groups, each of which had 6 mice. The tumor-bearing mice in therapeutic groups were administrated intravenously with escalating doses (150 µCi, 300 µCi, and 450 µCi) of 131I-CL58 via tail vein. The mice in the control groups were administrated with normal saline, cold mAb CL58, and 150 µCi of 131I-mIgG via tail vein, respectively. The tumor volume was estimated, assuming the tumors were ellipsoid, using the formula/volume ) 4π/3 (1/2 length × 1/2 width × 1/2 height). Tumor size and animal weight were measured every 2-3 days, and the mouse was culled and euthanatized once the tumor volume reached >1500 mm3 or the body weight lost >20% of its original weight. Normal tissues including liver and kidney were isolated, fixed in 5% buffered formalin, embedded in paraffin, and sections were stained with hematoxylin and eosin (H&E) for histopathological analyses. Statistical Analysis. Quantitative data were expressed as mean ( SD. Statistical analysis was done using a one-way ANOVA for multiple groups and an unpaired Student’s t test.
Figure 2. (A) Biodistribution of 125I-CL58 (5 µCi per mouse) in LS180 tumor-bearing nude mice at 4 h, and 1, 3, 5, and 7 d postinjection. (B) Biodistribution comparison of 125I-CL58 and 125I-mIgG (isotypematched control) (5 µCi per mouse) in LS180 tumor-bearing nude mice on 3 d postinjection. Data are expressed as %ID/g ( SD (n ) 4 per group).
P values of 70% and the radiochemistry purity was >99% after purification with PD-10 column. The binding affinity of 125I-CL58 to LS180 cells was determined by a saturation binding assay. As shown in Figure 1, the equilibrium dissociation constant (Kd) was determined to be 9.73 ( 0.84 nM, and Bmax was calculated to be approximate 8.20 × 106 sites/cell. (means ( SD, n ) 3). Biodistribution Studies of 125I-CL58. Biodistribution studies of 125I-CL58 were performed in BALB/c nude mice bearing LS180 human colon carcinoma xenografts. As shown in Figure 2A, the tumor uptake of 125I-CL58 at 4 h, and 1, 3, 5, and 7 d
316 Bioconjugate Chem., Vol. 21, No. 2, 2010
Figure 3. Representative planar γ images of Arrow indicates the location of tumor.
125
Liu et al.
I-CL58 (200 µCi per mouse) in LS180 tumor-bearing nude mice on 1, 2, and 6 d postinjection.
postinjection were 13.34 ( 3.43%ID/g, 24.99 ( 2.00%ID/g, 24.11 ( 4.99%ID/g, 16.41 ( 5.40%ID/g, and 11.25 ( 3.82%ID/ g, respectively, which were significantly higher than that of other organs after 1 d postinjection (p < 0.001). The blood radioactivity eliminated from 23.91 ( 1.35%ID/g to 1.05 ( 0.66%ID/g from 4 h to 7 d postinjection. For most normal organs, there was a progressive washout of radioactivity with the time, which was concomitant with the clearance of radioactivity from blood. The uptake of 125I-CL58 in all normal organs was 1500 mm3 on day 12 posttreatment. Lower or moderate dose (150 µCi or 300 µCi) of 131 I-CL58 delayed tumor progression significantly compared with the control groups. The tumor size of mice receiving 150 µCi 131I-CL58 was significantly different to that in unlabeled mAb CL58 group on day 12 (P < 0.001). Mice receiving 300 µCi 131I-CL58 exhibited significant tumor growth inhibition until
day 19 before exponential growth of tumors resumed. For the group treated with the highest dose of 450 µCi, the tumor growth was significantly suppressed, and the average tumor size maintained in a similar level up to day 55 post-treatment. The tumor size in 450 µCi group was smaller than that in other groups from day 9 post-treatment. On day 30, the tumor volume of 450 µCi group was significantly smaller than that in 150 µCi and 300 µCi groups (p < 0.01). A dose-dependent effect was observed in this study as the tumor therapeutic efficacy of 131 I-CL58 followed the order of 450 µCi > 300 µCi > 150 µCi (Figure 4A). The CEA-targeted radioimmunotherapy with 131I-CL58 significantly prolonged the survival time of mice. As shown in Figure 4B, all mice in the control groups (saline, mAb CL58, and 150 µCi mIgG) died before day 23, whereas the median survival time for 150 µCi and 300 µCi therapeutic groups was prolonged to be 41 days and 48 days, respectively. All mice in the 450 µCi 131I-CL58 treatment group were still alive on day 55 post-treatment, and the median survival time was 84 days. All mice injected with 131I-CL58 did not show obvious signs of body weight loss (less than 20%), and histopathologic staining results revealed that no renal and hepatic toxicity was observed (Supporting Information Figure S1-S2), indicating the limited toxicity of 131I-CL58 in this therapeutic study.
DISCUSSION In this study, we evaluated the in vitro and in vivo characteristics of radioiodinated anti-CEA monoclonal antibody CL58 and investigated whether 131I-labeled CL58 could be used for radioimmunotherapy of colon cancers. RIT can specifically deliver radiation to tumors with minimal toxicity to normal organs. The radiation energy given off by the radionuclides would also kill the adjacent tumor cells, which do not express the target antigen (so-called crossfire). Herein, we have chosen carcinoembryonic antigen (CEA) as the target for RIT of colon cancer xenografts because CEA is expressed at high levels in approximately 95% of colorectal carcinomas but is restrictedly expressed in normal tissues. CEA is located
RIT of
131
I-CL58 for CEA-Positive Tumor
Figure 4. (A) Radioimmunotherapy of established LS180 tumors in nude mice with saline, unlabeled mAb CL58, 131I-mIgG (150 µCi), or different doses (150, 300, and 450 µCi) of 131I-CL58. Volumes of tumors in each treatment group were measured and expressed as a function of time (means ( SD, n ) 6 per group). (B) Kaplan-Meier plots of individual treatment groups receiving saline, unlabeled CL58, 131 I-mIgG (150 µCi), or different doses (150, 300, and 450 µCi) of 131 I-CL58.
on the cell surface of tumor cells, and it has been one of the first tumor-associated antigens (TAA) used for RIT. A series of anti-CEA antibodies have been investigated for colon cancer targeted RIT, some of which have been used for clinical trials (see review, ref 21). In this study, we used a murine anti-CEA monoclonal antibody CL58 as the vehicle for radiation delivery. The mAb CL58, which has been successfully used in clinical trials for tumor imaging and radioimmunoguided surgery (17, 18), exhibited a moderate affinity (Kd ) 9.73 ( 0.84 nM) to CEApositive LS180 cells (Figure 1). The low antigen binding affinity of an antibody will certainly affect the in vivo radionuclide delivery in RIT trials; however, it has also been hypothesized that very high affinity interactions between antibodies and tumor antigens may impair efficient tumor penetration of the antibodies and thus diminish effective in vivo targeting (22). Consequently, antibodies with moderate affinity may be advantageous, as they can effectively penetrate into tumors and exhibit longer retention on tumor cells (23). LS180 cells had an extremely high CEA expression level (Figure 1). In the clinical situation, the CEA expression of other types of colon cancers may not reach such a high level as LS180 tumors, but the highly restrictive expression of CEA in colon cancers as compared with normal organs would still allow the specific delivery of radiation by the anti-CEA mAb CL58. The in vivo behavior of 125I-CL58 was tested in nude mice bearing CL180 colon cancer xenografts by both biodistribution and gamma imaging studies. 125I-CL58 showed efficient tumor targeting properties, with rapid normal organ clearance and high tumor-to-nontumor ratios (Figures 2 and 3). The longer tumor retention of radioactivity was also found by the biodistribution study, and the tumor uptake of 125I-CL58 was still as high as 11.25 ( 3.82%ID/g on 7 d postinjection. We tested the internalization property of 125I-CL58 by the method reported previously (20), and the result demonstrated that 125I-CL58 exhibited an undetectable internalization after binding to CL180
Bioconjugate Chem., Vol. 21, No. 2, 2010 317
cells (data not shown). The negligible internalization property of 125I-CL58 would effectively avoid the rapid diffusion of iodotyrosine in the target cells after internalization and catabolism of the radioiodinated mAb, leading to the high tumor uptake and long tumor retention of 125I-CL58 in LS180 tumor xenografts. Importantly, the liver and spleen uptake of 125I-CL58 was relatively low compared with that of metal isotope (e.g., 64 Cu, 111In) labeled antibodies (24, 25), which possibly makes radioiodinated CL58 less toxic. The significantly higher tumor uptake of 125I-CL58 as compared with 125I-mIgG clearly demonstrated the in vivo CEA-targeting specificity of mAb CL58 (Figure 2B). 131 I and 90Y are the most commonly used isotopes for radioimmunotherapy. Compared with the pure β emitter 90Y, 131 I is much easier for radiolabeling, lower cost, and can also be used for imaging, which makes it possible for monitoring the therapeutic efficacy during the period of RIT. The experimental RIT study using 131I-CL58 showed promising results on the therapy of colon cancer xenografts. In the 450 µCi 131ICL58 group, the tumor growth was efficaciously inhibited up to the end of therapy study (Figure 4A). The overall body weight loss of mice was
