Immuno-PET of Human Colon Xenograft– Bearing BALB/c Nude Mice Using ¹²⁴I-CDR–Grafted Humanized A33 Monoclonal Antibody

Antibodies and Cells

The CDR-grafted huA33 (IgG1) (6) was provided by the Ludwig Institute for Cancer Research (New York, NY) together with the A33-expressing colorectal cancer cell line SW1222. For controls, the colorectal cancer cell line LIM 2537 (non–A33 expressing) was provided by Dr. Robert Whitehead (Ludwig Institute for Cancer Research, Melbourne, Australia).

Radiolabeling

A 0.3-mL aliquot of 0.5 mol/L potassium phosphate buffer, pH 7.0, was added to 0.57 mL ¹²⁴I in 0.1N NaOH (13 MBq). An aliquot of 335 μL of this labeled mixture was added to 100 μL antibody (9 mg/mL) followed by 18.3 μL chloramine-T at a concentration of 0.46 mg/mL prepared in the same phosphate buffer. After 2 min, labeling was stopped by adding 18.3 μL of a solution containing 1.86 mg/mL sodium metabisulfite in phosphate buffer. The labeled mixture was purified by centrifugal desalting on columns of Biogel P6DG equilibrated in phosphate-buffered saline (PBS). The radiochemical purity of labeled antibody was analyzed by instant thin-layer chromatography silica gel (ITLC-SG) using 10% w/v trichloroacetic acid as solvent. Radioactivity bound to antibody remained at the origin, whereas free ¹²⁴I migrated with the solvent front.

Immunoreactivity Assays

The immunoreactivity of huA33 after radiolabeling was determined using the assay of Lindmo et al. (20). This cell-based assay consisted of incubating 20 ng radiolabeled antibody with increasing concentrations of antigen-expressing cells SW1222 ranging from 0 cells to 6.0 × 10⁶ cells in 1.0 mL cell culture medium and mixed continuously on a rotation mixer. After 45 min at room temperature, the cells were centrifuged and washed three times with medium before radioactive counting to determine the extent of binding compared with standards. The specificity of binding was shown by adding 20 μg unlabeled huA33 to the assays, and the extent of binding was similarly determined. Scatchard assays were also performed (20). Serial dilutions starting from 10 μg/mL unlabeled huA33 were added to 20 ng labeled antibody followed by the addition of 3 × 10⁶ SW1222 cells.

Serum Stability

The stability of radioconjugates in the blood circulation of mice was also determined. Serum was obtained at days 0, 3, and 7 after injection. An aliquot of serum was spotted on an ITLC-SG strip and developed using 10% w/v trichloroacetic acid as solvent. For immunoreactivity assay, concentrations of radiolabeled antibody in mouse serum were estimated from the blood time–activity curve of the biodistribution study and used in the cell binding assay.

Animals and Tumors

The A33 antigen–expressing melanoma cell line SW1222 was grown in RPMI 1640 with standard additives and 5% fetal calf serum. The control cell line LIM 2537 was cultured under similar conditions. Cells were harvested at the point of confluence using PBS and 0.05% w/v EDTA and resuspended in medium. Approximately 5.0 × 10⁶ cells of SW1222 and LIM 2537 lines in 0.1 mL PBS were injected intradermally contralaterally on the underside flanks of female BALB/c nude mice, 3–4 wk old. Both SW1222 and LIM 2537 tumors started to develop by 1–2 wk, and the mice were used 1–2 wk later when tumors weighed between 0.2 and 0.7 g. Tumors developed in all mice. An injected dose containing a sterile filtered mixture of 16 μg ¹²⁴I-huA33 (111 kBq) in 0.1 mL PBS was administered through retroorbital injection while the mice were under enflurane anesthesia (Ethrane; Baxter Pharmaceutical Products, New Providence, NJ). From 4 h after injection, groups of five mice were killed by cranial dislocation, and blood, tumors, and various tissues were removed for weighing and radioactive counting. Blood clearance kinetics were determined using a curve-fitting program, SAAM II, from the University of Washington (Seattle, WA) assuming a two-compartment model. Tumors were measured using the formula (length × width squared) × 0.5 (21), with length and width being two perpendicular measurements of diameter of palpable tumors.

PET Imaging

Mice were killed by overinhalation of enflurane and placed supine on an ECAT 951/31R PET scanner (Siemens Medical Systems, Inc., Hoffman Estates, IL/CTI, Knoxville, TN) linked to a SunSPARC 10 workstation (Sun Microsystems, Palo Alto, CA). The axial field of view of the PET scanner is 10.8 cm, and the in-plane resolution is 6.5 mm. The mice were imaged in two-dimensional mode (septa extended) for 30 min. Images were acquired in a 128 × 128 matrix and reconstructed with a Hanning filter at a cutoff of 0.4 cycles per pixel using the standard filtered backprojection algorithm. Single gamma emissions contribute significantly when using ¹²⁴I; however, the real-time sorter in the PET scanner eliminates the single emissions because these events are not true coincidences. A long imaging time of 30 min was chosen to ensure that sufficient counts were recorded for acceptable image quality. A transmission scan was not obtained.

In Vitro Properties of ¹²⁴I-HuA33

The radiolabeling efficiency of huA33 using ¹²⁴I and chloramine-T was 40%. After chromatographic purification, >98% of ¹²⁴I was bound to protein as determined by thin-layer chromatography. The specific activity of ¹²⁴I-huA33 obtained was 7.4 MBq/mg. Labeled antibody retained the ability to bind A33 antigen–expressing cells as shown using the assay of Lindmo et al. (20) (Fig. 1A). Double-reciprocal plotting of values from the same data (Fig. 1A) and extrapolation to the ordinate provided a quantitative measure of immunoreactivity in the presence of infinite antigen excess. This immunoreactive fraction of the labeled antibody was found to be 0.78. Alternatively, the immunoreactivity is expressed as 78% binding. Scatchard analysis of binding of ¹²⁴I-huA33 to SW1222 colorectal cancer cells is shown in Figure 1B. The intercept at the abscissa represents the binding capacity of the cell. This value was determined to be 371,000 antibody molecules bound per cell. The apparent association constant can be derived from the slope of the line (Fig. 1B), giving a value of 2.2 × 10⁹ M⁻¹.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1.

Binding assay for determination of immunoreactive fraction of anti–colorectal cancer ¹²⁴I-huA33. (A) Plot of specific binding over total applied radioactivity against increasing SW1222 cell concentration. (B) Scatchard plot of binding of ¹²⁴I-huA33 mAb to colorectal cancer cells. From intercept value at abscissa and slope, binding capacity and apparent association constant can be determined. Ka = association constant.

Serum Stability

The stability of radioconjugate was also evaluated by analyzing serum taken from mice injected with antibodies on days 0, 3, and 7 after injection. Thin-layer chromatographic analysis indicated that 98.7% and 85.3%, respectively, of ¹²⁴I was bound to protein on days 0 and 3. By day 7, the radioactivity in serum was too low to be analyzed by this method. The immunoreactivity of ¹²⁴I-huA33 in mouse serum at these times after injection was 72%, 69.8%, and 51.3%, respectively. From these results, the radioconjugate was shown to remain stable and to retain an ability to bind to target A33–expressing cells for several days while remaining in the circulation.

Biodistribution and Pharmacokinetics

The relative levels of ¹²⁴I-huA33 in colon xenografts and blood at times after injection are shown in Figure 2. Assuming a two-compartment model with a four-parameter fit of mean blood levels, the T_1/2α was 1.5 h and the T_1/2β was 38.2 h.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2.

Radioactivity concentration of ¹²⁴I-huA33 in blood (•), SW1222 human colon tumor (▴), and LIM 2537 control tumor (▪) in BALB/c nude mice bearing xenografts, as function of time after infusion. %ID = percentage injected dose.

The mean uptake of ¹²⁴I-huA33 by antigen-expressing cells SW1222 reached a maximum percentage injected dose (%ID) per gram of 50 ± 7.0 by 4 d after injection, with a gradual reduction over the next 6 d. In contrast, the maximal uptake in the control colon xenograft LIM 2537 was 6.27 ± 0.80 %ID/g at 4 h. The tumor-to-blood ratio for SW1222 tumor increased from 5:1 to 35:1 from day 2 to day 7 after injection.

The tissue distribution of ¹²⁴I-huA33 in xenografted mice is shown in Table 1. Radioactivity was found mainly in blood and several organs by 4 h after injection as a result of blood-pool activity. Activity in the tissues continued to decrease with time, whereas tumor (SW1222) uptake of radioconjugate was substantial and specific. Uptake of radioconjugate in control tumor (LIM 2537) was low.

PET Imaging

At 4 h after injection (Fig. 3A), blood-pool activity was evident in the heart, with some tumor visualization. Tumor uptake of ¹²⁴I-huA33 was distinct by 24 h after injection (Fig. 3B) in SW1222 tumors. No radioconjugate uptake was observed in LIM 2537 control tumor located in the contralateral flank. Blood-pool activity in the heart continued to decrease by 48 h after injection (Fig. 3C), and delayed imaging up to 10 d after injection was possible (Fig. 3D). At that late time, other than radioactivity in the tumor and thyroid, the whole-body activity is essentially clear. Tumor sizes ranged from 0.5 to 0.81 g in mass and from 530 to 730 mm³ in volume by 7 d, with growth of tumor up to 1.0 g by day 10.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 3.

Immuno-PET images of ¹²⁴I-huA33 in groups of two BALB/c nude mice bearing SW1222 colorectal cancer xenografts of masses ranging from 0.2 to 0.7 g. Images were acquired 4 h (A), 24 h (B), 48 h (C), and 10 d (D) after injection of ¹²⁴I-huA33. Location of SW1222 tumors in left flank is indicated in (A) by arrow. Control LIM 2537 tumors are in contralateral flank.

In another experiment, the effect of tumor size on PET imaging was examined with mice bearing tumors weighing 0.21 g (292 mm³), 0.68 g (615 mm³), and 1.2 g (1,079 mm³) (Fig. 4). The mice were imaged 96 h after injection. All tumors could be imaged clearly, and control tumors again did not have any detectable radioconjugate. Larger tumors showed reduced blood-pool activity, consistent with the increased uptake of radioconjugate in tumor mass (Fig. 4).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 4.

Effect of tumor mass on immuno-PET detection of ¹²⁴I-huA33 uptake in SW1222 colorectal cancer xenografts in BALB/c nude mice 96 h after infusion. SW1222 tumors are in left flank, as in Figure 3. Tumor masses were 0.21, 0.68, and 1.2 g.