Radioiodinated versus Radiometal-Labeled Anti–Carcinoembryonic Antigen Single-Chain Fv-Fc Antibody Fragments: Optimal Pharmacokinetics for Therapy (original) (raw)

Skip Nav Destination

Experimental Therapeutics, Molecular Targets, and Chemical Biology| January 18 2007

Vania Kenanova;

1Division of Molecular Biology and

6Crump Institute for Molecular Imaging, Department of Molecular and Medical Pharmacology, David Geffen School of Medicine at University of California Los Angeles, Los Angeles, California

Search for other works by this author on:

Tove Olafsen;

6Crump Institute for Molecular Imaging, Department of Molecular and Medical Pharmacology, David Geffen School of Medicine at University of California Los Angeles, Los Angeles, California

Search for other works by this author on:

Nora H. Ruel;

5Department of Biostatistics, City of Hope National Medical Center, Duarte, California; and

Search for other works by this author on:

Jeffrey Longmate;

5Department of Biostatistics, City of Hope National Medical Center, Duarte, California; and

Search for other works by this author on:

Paul J. Yazaki;

3Department of Radioimmunotherapy,

Search for other works by this author on:

John E. Shively;

2Division of Immunology, Beckman Research Institute of the City of Hope;

Search for other works by this author on:

David Colcher;

3Department of Radioimmunotherapy,

Search for other works by this author on:

Anna M. Wu

1Division of Molecular Biology and

6Crump Institute for Molecular Imaging, Department of Molecular and Medical Pharmacology, David Geffen School of Medicine at University of California Los Angeles, Los Angeles, California

Search for other works by this author on:

Requests for reprints: Anna M. Wu, Crump Institute for Molecular Imaging, Department of Molecular and Medical Pharmacology, David Geffen School of Medicine at University of California Los Angeles, 700 Westwood Plaza, Los Angeles, CA 90095. E-mail: awu@mednet.ucla.edu.

Received: February 07 2006

Revision Received: July 22 2006

Accepted: November 07 2006

Online ISSN: 1538-7445

Print ISSN: 0008-5472

©2007 American Association for Cancer Research.

2007

Cancer Res (2007) 67 (2): 718–726.

Article history

Received:

February 07 2006

Revision Received:

July 22 2006

Accepted:

November 07 2006

• Split-Screen
• Views Icon Views
  • Article contents
  • Figures & tables
  • Video
  • Audio
  • Supplementary Data
  • Peer Review
• Open the PDF for in another window
• Tools Icon Tools
• Search Site
• Article Versions Icon Versions
  • Version of Record January 18 2007

Abstract

Antibody fragments with optimized pharmacokinetic profiles hold potential for detection and therapy of tumor malignancies. We studied the behavior of three anti–carcinoembryonic antigen (CEA) single-chain Fv-Fc (scFv-Fc) variants (I253A, H310A, and H310A/H435Q; Kabat numbering system) that exhibited differential serum persistence. Biodistribution studies done on CEA-positive tumor xenografted mice revealed that the 111In-labeled I253A fragment with the slowest clearance kinetics (_T_1/2β, 27.7 h) achieved the highest tumor uptake (44.6% ID/g at 24 h), whereas the radiometal-labeled H310A/H435Q fragment with the most rapid elimination (_T_1/2β, 7.05 h) reached a maximum of 28.0% ID/g at 12 h postinjection. The H310A protein was characterized by both intermediate serum half-life and tumor uptake. The 111In-based biodistribution studies showed that all three fragments were eliminated primarily through the liver, and hepatic radiometal activity correlated with the rate of fragment clearance. The 111In-labeled H310A/H435Q protein exhibited the highest liver uptake (23.5% ID/g at 24 h). Metabolism of the 125I-labeled scFv-Fc proteins resulted in low normal organ activity. Finally, the 125I/111In biodistribution data allowed for dose estimations, which suggest the 131I-labeled scFv-Fc H310A/H435Q as a promising candidate for radioimmunotherapy. [Cancer Res 2007;67(2):718–26]

Introduction

New approaches aiming to optimize radiolabeled anticancer antibodies for clinical usefulness are currently being pursued. Alterations in the native form of intact antibodies are made to enhance their desirable qualities while diminishing the suboptimal characteristics in imaging and therapeutic applications (reviewed in refs. 1–5). For example, engineered antibodies and their fragments can be designed with controlled targeting, distribution, and clearance kinetics. Both preclinical and clinical studies have confirmed the ability of radiolabeled antibody fragments to quickly target and accumulate at the tumor site while retaining relatively fast clearance kinetics. Single-chain Fv (scFv; 27 kDa) fragments have shown extremely rapid serum and tumor kinetics in clinical studies (6); however, their small molecular size and monovalency might be a limiting factor for their use. Larger, multivalent recombinant antibody fragments such as diabodies (dimers of scFvs, 55 kDa; ref. 7) or minibodies [(scFv-CH3)2 fusion proteins, 80 kDa; ref. 8] have shown usefulness as in vivo imaging agents in preclinical studies when radiolabeled with single-photon emitters such as 111In and 123I for single-photon emission computed tomography (8) or positron emitters such as 64Cu and 124I for positron emission tomography (PET; refs. 9, 10). Clinically, both the 123I-labeled diabody (11) and minibody (12) have been evaluated in patients and have shown promising results in detection of tumor lesions.

Manipulation of the receptor interactions controlling the extended serum persistence of intact antibodies offers an alternative approach in tailoring their pharmacokinetics for radionuclide delivery. The protective FcRn receptor is responsible for maintaining the levels of immunoglobulin Gs (IgG) in the circulation (13–15). Compromising the binding of antibodies to the FcRn leads to shortening of their serum half-life (16–18) whereas domain-deleted antibody fragments containing a complete Fc region show pharmacokinetic behavior similar to intact antibodies (19). We produced an anti–carcinoembryonic antigen (CEA) fragment, assembled VL-linker-VH-hinge-CH2-CH3 [(scFv-Fc)2, 105 kDa] along with five variants carrying specific mutations in the Fc region, where FcRn binding occurs (20). Biodistribution studies in BALB/c mice showed that the radioiodinated scFv-Fc antibody fragments possessed a spectrum of terminal half-lives ranging from 12 days to 8 h. The difference in clearance kinetics was also visualized by in vivo serial imaging of fragments labeled with 124I using a small animal PET scanner. Of the six scFv-Fc fragments, the three fastest-clearing variants (scFv-Fc I253A > H310A > H310A/H435Q) exhibited the most optimal serum kinetics and were considered as potentially suitable for therapy of CEA-expressing tumors.

Direct radioiodination of residues on the surface of the scFv-Fc antibody fragments was practical and very useful in preliminary studies because small amounts of protein (micrograms) can be labeled with high labeling efficiency. However, the disadvantage of conventional radioiodination methods using chloramine-T or iodogen, which are based on modification of random surface tyrosine residues, is that the radioiodinated proteins are readily metabolized and dehalogenated in vivo (21). The free radioiodide and the metabolized iodinated peptide fragments are washed out of the tissues (including tumor) and excreted in the urine. In contrast, when radiometals are bound to antibodies via a bifunctional chelate, the radiolabeled antibody complex is relatively stable in vivo and the radiometal is generally trapped within cells, leading to accretion indicative of the overall organ and tumor distribution. Thus, in addition to the radioiodine label for biodistribution in tumor-bearing animals, studies with radiometal-conjugated scFv-Fcs were necessary to determine the therapeutic potential of scFv-Fc fragments conjugated to a therapeutic radiometal, as well as to provide information about the tumor uptake and routes of elimination.

In this work, we report the pharmacokinetics and biodistribution of three radioiodinated and radiometal-conjugated anti-CEA scFv-Fc antibody fragments in mice. We make several comparisons involving different antibody fragments, radionuclides, and animal systems to determine how these factors influence the kinetics of radioimmunoconjugate distribution. Finally, both radioiodine and radiometal distribution data were used to generate 131I and 90Y dose estimates to select the best agent with the appropriate radiolabel for therapy.

Materials and Methods

Production and purification of anti-CEA scFv-Fc antibody fragments. The recombinant anti-CEA scFv-Fc antibody fragments used in these studies were previously described (20). Briefly, the T84.66/GS18 scFv-Fc proteins were expressed in NS0 murine myeloma cells by using the pEE12 expression vector with a cytomegalovirus promoter and glutamine synthetase as the selectable marker (22). Proteins were purified by anion exchange chromatography (Poros HQ50, PerSeptive Biosystems, Framingham, MA), followed by hydroxyapatite chromatography (Macro-Prep Type 1, Bio-Rad Laboratories, Hercules, CA) and a final step of anion exchange chromatography (Source 15Q, GE Healthcare, Piscataway, NJ) using an AKTA Purifier (GE Healthcare).

Radioiodination of scFv-Fcs. Radioiodination of the T84.66 scFv-Fc proteins was done by the iodogen method as previously described (23). Briefly, ∼0.2 mg of purified protein was labeled with Na125I (Perkin-Elmer Life Sciences, Inc., Boston, MA) in 0.1 mL phosphate buffer (pH 7.5) using polypropylene tubes coated with 20 μg iodogen (Pierce, Rockford, IL) for 3 to 5 min at room temperature. The sample was purified by a Gilson high-performance liquid chromatography (HPLC) system on a Superdex 75 column (GE Healthcare). The fractions containing the radioactive peak were selected and diluted in saline/1% human serum albumin to prepare doses for injection. The radiolabeling efficiency was determined by integrating peak areas on the size exclusion HPLC trace and calculating the percentage of radioactivity associated with the 105-kDa peak of the scFv-Fc. Immunoreactivity and valency were determined by incubation of the labeled protein with a 20-fold molar excess of either CEA or the recombinant N-A3 fragment of CEA (24), followed by HPLC analysis on Superose 6 HR 10/30 columns (GE Healthcare) to assess the formation of antibody-antigen complexes.

Chelate conjugation and radiolabeling with 111In. Purified scFv-Fc proteins were conjugated to 1,4,7,10-tetraazacyclododecane-_N,N_′,_N_″,_N_‴-tetraacetic acid (DOTA; Macrocyclics, Dallas, TX) by using the water-soluble _N_-hydroxysuccinimide method as described (25, 26). Typically, 2 mg of protein were reacted with a 1,000:1 ratio of DOTA to protein for 18 to 24 h at room temperature, pH 7.0. After conjugation, the protein was dialyzed extensively in 0.2 mol/L NH4OAc (pH 7.2) and concentrated to >5 mg/mL. DOTA-scFv-Fc (∼400 μg protein) were incubated with 1.3 mCi of 111In-chloride (Mallinckrodt, Hazelwood, MO) in 0.25 mol/L NH4OAc (pH 5.0) for 1 h at 43°C. The reaction was terminated and conjugated monomeric proteins were purified by size-exclusion HPLC (25). Immunoreactivity was determined as described for radioiodination.

Biodistribution studies. All animal studies were conducted under protocols approved by the City of Hope Research Animal Care Committee or the Chancellor's Animal Research Committee at the University of California, Los Angeles. Xenografts were established in 7- to 8-week-old female nude mice (Charles River Laboratories, Wilmington, MA) by s.c. inoculation of 1 × 106 to 2 × 106 LS174T cells in the flank region. After 10 days, when tumor masses were in the range of 100 to 300 mg, a mixture of 3.8 to 4 μCi of 111In-DOTA-scFv-Fc and 10 μCi of the same scFv-Fc fragment radiolabeled with 125I (3–5 μg protein) was injected into the tail vein of each animal. Time points of analysis were 0, 2, 4, 6, 12, 24, 48, and 72 h, when groups of five mice at the selected time point were euthanized; necropsy was done; and organs were weighed and counted for radioactivity. Activities in liver, spleen, kidney, lung, tumor, carcass, and blood were determined; background, crossover, and decay corrections were done. Results were calculated as percentage of injected dose per gram of tissue (% ID/g).

To quantitate the differences in blood clearance times, the ADAPTII software package (27) was used to estimate two rate constants characteristic of each engineered fragment. Biexponential functions were fitted, via the ID subroutine, to each blood clearance curve (% ID/g) after radiolabel decay correction. Significant differences in these values were examined by comparing the 95% confidence intervals for the variable estimates. When calculating the relative effectiveness of the proteins in a therapeutic application, radiation absorbed doses are proportional to % ID/g (28); thus, the time-activity curves for tumor and blood uptake were integrated. For each fragment, this pair of curves was fitted by a respective sum of two exponential functions, and integration was done over the interval from zero to infinity using ADAPTII to give the pharmacokinetics or radionuclide-associated area under the curve (AUC). Pharmacokinetic values refer to AUCs determined by first correcting blood, liver, kidney, and tumor uptake data for radiodecay. Such area values refer to a hypothetical situation where biodistributions could be conducted without a radiolabel. These areas are larger than those experimentally observed because of omitting the decay factor. The appropriate decay factor was added back to calculate the blood and tumor AUCs produced by the 131I and 90Y radioisotopes. Estimates of the maximum doses to tumors achieved by the three scFv-Fc fragments labeled with either 131I or 90Y were calculated based on the ratio of AUCs for the scFv-Fc proteins. These are the areas of %ID/g multiplied by time and should therefore represent dose. Edge effects and photon-induced doses were neglected.

Statistical analysis. To compare the effects of different mouse strains, the absence or presence of a tumor, and the radiolabel, two-way ANOVA test was done. All significance testing was done at the P < 0.01 level. The SAS/STAT software (SAS, Inc., Cary, NC) was used for all statistical models and analyses.

Results

125I/111In-DOTA-scFv-Fc Biodistributions in Tumor-Bearing Mice

Dual-label biodistribution and clearance kinetic studies of 125I– and 111In-DOTA–radiolabeled scFv-Fc I253A, H310A, and H310A/H435Q were conducted in athymic nude mice carrying CEA-positive LS174T xenografts. The labeling efficiency for the 111In radioisotope ranged from 89% to 96%, whereas the label incorporation for the 125I radioisotope by the H310A, H310A/H435Q, and I253A fragments was 100%, 75%, and 39%, respectively. The immunoreactivity was between 75% and 92% for the radioiodinated fragments and between 76% and 91% for the indium-labeled proteins. The specific activities were typically about 2.8 and 1.5 μCi/μg for 125I- and 111In-labeled proteins, respectively. Biodistribution results are shown in Table 1. The fastest-clearing fragment, H310A/H435Q double mutant, which has a terminal serum half-life (_T_1/2β) of 6.01 h (125I studies) and 7.05 h (111In studies), achieved a maximum tumor uptake more rapidly than the other two fragments (12 versus 24 h), although its tumor uptake was the lowest, 23.5% ID/g (Fig. 1B) and 28.0% ID/g (Fig. 1D) for the 125I- and 111In-labeled forms, respectively. The slower-clearing proteins, I253A (_T_1/2β, 28.4 and 27.7 h) and H310A (_T_1/2β, 21.9 and 23.8 h), achieved higher tumor uptake levels, 33.6% ID/g and 44.6% ID/g (I253A) and 18.0% ID/g and 33.4% ID/g (H310A) for the radioiodinated and radiometal-labeled forms at 24 h (Fig. 1B and D), respectively. In general, maximum tumor uptake was inversely related to serum clearance. For 111In-labeled proteins, liver showed the highest normal organ accumulation, with radionuclide activity elevating with the increase in the rate of clearance. Hence, the fastest-clearing fragment H310A/H435Q exhibited the highest liver uptake (23.5% ID/g; 24 h), followed by the H310A (19.8% ID/g; 48 h) and I253A (11.0% ID/g; 24 h) fragments (Table 1).

Table 1.

Biodistribution of 125I– and 111In-DOTA–labeled T84.66 scFv-Fc fragments in LS174T xenografted athymic nude mice

NOTE: Groups of five mice were analyzed at each time point. Organ uptake is expressed as percent injected dose per gram (% ID/g). Values are the means with corresponding SEs shown in brackets.

Figure 1.

Blood and tumor activity curves of 125I– and 111In-DOTA–conjugated scFv-Fc fragments. A, blood activity curves derived from biodistribution studies with 125I-labeled scFv-Fc fragments in LS174T xenografted athymic nude mice. H310A/H435Q*, blood curve of the radioiodinated fragment in non-tumor-bearing mice. B, tumor uptake curves of 125I-labeled scFv-Fc fragments in LS174T tumor–bearing mice. Bars, SE. C, blood activity curves derived from biodistribution studies with 111In-DOTA–labeled scFv-Fc fragments in LS174T xenografted athymic nude mice. H310A/H435Q*, blood activity curve of the radiometal-labeled fragment in non-tumor-bearing mice. D, tumor uptake curves of 111In-DOTA–conjugated scFv-Fc fragments in LS174T tumor–bearing mice.

Figure 1.

Blood and tumor activity curves of 125I– and 111In-DOTA–conjugated scFv-Fc fragments. A, blood activity curves derived from biodistribution studies with 125I-labeled scFv-Fc fragments in LS174T xenografted athymic nude mice. H310A/H435Q*, blood curve of the radioiodinated fragment in non-tumor-bearing mice. B, tumor uptake curves of 125I-labeled scFv-Fc fragments in LS174T tumor–bearing mice. Bars, SE. C, blood activity curves derived from biodistribution studies with 111In-DOTA–labeled scFv-Fc fragments in LS174T xenografted athymic nude mice. H310A/H435Q*, blood activity curve of the radiometal-labeled fragment in non-tumor-bearing mice. D, tumor uptake curves of 111In-DOTA–conjugated scFv-Fc fragments in LS174T tumor–bearing mice.

Close modal

125I/111In-DOTA-scFv-Fc Biodistribution in Non-Tumor-Bearing Mice

One dual-label study using the scFv-Fc H310A/H435Q was executed in non-tumor-bearing athymic nude mice (Table 2) to delineate the effect of tumor on the antibody distribution kinetics. The labeling efficiency and immunoreactivity to CEA for the 125I-labeled form were 80% and 95%, respectively, whereas the same variables measured for 111In-DOTA-scFv-Fc H10A/H435Q were 83% and 85%. The injected specific activities of radioiodinated and radiometal-labeled H310A/H435Q proteins were 3.3 and 2.4 μCi/μg, respectively. The data for the blood analyses in both tumor-bearing (Table 3A) and non-tumor-bearing (Table 3B) mice were fitted. In tumor-free mice, the terminal elimination phase half-life of the radioiodinated scFv-Fc H310A/H435Q fragment was 6.13 h, whereas the radioindium-conjugated form showed shorter persistence in the circulation with a half life of 4.36 h.

Table 2.

Biodistribution of 125I– and 111In-DOTA–labeled T84.66 scFv-Fc H310A/H435Q fragments in non-tumor-bearing athymic nude mice

NOTE: Values are the means with corresponding SEs shown in brackets.

Table 3.

Estimated values of blood half-times for the T84.66 scFv-Fc fragments in LS174T xenografted athymic nude mice and non-tumor-bearing athymic nude mice

*

The amplitudes of the two components are given by _A_α and _A_β, where the sum of _A_α and _A_β is the total amount of activity in the blood compartment at the time of injection.

†

AUC is a time integral of the various organ uptakes.

‡

Mean residence time (hours) is used to give a single variable for blood clearance.

The biodistribution study was also carried out to compare the liver uptake in tumor (Table 1) and non-tumor-bearing animals (Table 2), as higher liver uptake in the tumor-bearing animals could be indicative of trapping CEA-antibody immunocomplexes in the hepatocytes (29, 30). The 111In-labeled double mutant fragment showed significantly higher hepatic uptake in the tumor-free mice (max, 29.1% ID/g; 6 h) than in the tumor-bearing animals (max, 23.5% ID/g; 24 h) over the course of the entire study (P < 0.0001). The finding that there is more liver activity in the absence of a tumor suggests that there is no CEA antigen-antibody accumulation in the liver, although further experiments are needed to confirm this conclusion.

Biodistribution Comparisons

To understand the biodistribution properties of both radioiodinated and radiometal-labeled scFv-Fc fragments in different animal systems, several statistical comparisons were made.

Animal system effect. The radioiodinated scFv-Fc H310A/H435Q was used in biodistributions conducted in LS174T xenografted athymic nude mice (Table 1), as well as in tumor-free BALB/c (20) and athymic nude mice (Table 2). The fragment achieved significantly higher blood and tissue activity levels in the immunocompetent BALB/c animals compared with the immunocompromised athymic nude mice (tumor-bearing or not) in each of the collected organs (liver, kidneys, spleen, lung) and in blood. The same trend was also observed for the radioiodinated scFv-Fc I253A and H310A fragments.

Tumor effect. The effect of tumor on the biodistribution of 111In-DOTA-scFv-Fc H310A/H435Q was examined by comparing data obtained from tumor-bearing and non-tumor-bearing nude mice. The fragment behaved similarly in blood (Fig. 1C), lung, spleen, and carcass, whereas the liver and kidneys in tumor-bearing animals had significantly lower activity (P < 0.001) over the 72-h study (Table 1).

Radionuclide effect. In a two-way comparison of radioiodine versus radiometal label in non-tumor-bearing nude mice, the scFv-Fc H310A/H435Q protein yielded higher radioiodine levels in blood (P < 0.0001; Table 2; Fig. 1A and C) and increased levels of 111In-labeled fragment in liver, spleen, kidneys, and carcass (P < 0.0001; Table 2). In the lung, the radioiodine and radiometal activities were not significantly different (P = 0.0134). The same comparison of radioiodine versus radiometal-labeled scFv-Fc H310A/H435Q in tumor-bearing nude mice showed that the levels of the two radiolabels were similar in blood (P = 0.222; Fig. 1A and C), lung (P = 0.357, Table 1), and tumor (P = 0.0412; Fig. 1B and D), whereas the 111In-labeled fragment achieved higher levels in the kidney, liver, spleen, and carcass (P < 0.0001; Table 1). The radioiodinated and radiometal-labeled forms of the I253A and H310A fragments in the tumor-bearing animal model were also examined. In tumor and all organs, except lung (P = 0.284) and blood (P = 0.576), where the two labels were similarly present, the levels of the indium-labeled I253A were significantly higher than those of the radioiodinated I253A fragment (Table 1). In the case of the H310A protein, indium levels were higher in all organs but the blood (P = 0.0111), where the indium and radioiodine levels were similar. Overall, the indium-labeled proteins showed enhanced accretion in all organs but blood, where radioiodine and radiometal levels did not differ significantly.

Therapy Prospects

Estimation of the potential of the three scFv-Fc proteins in therapeutic applications required integration of the time-activity curves. Based on the 125I and 111In biodistribution data, Table 4 includes the calculated blood, tumor, liver, and kidney AUCs. In addition, the tumor-to-blood, tumor-to-liver, and tumor-to-kidney AUC ratios of unlabeled scFv-Fcs (pharmacokinetic case) as well as of the scFv-Fc fragments radiolabeled with potential therapeutic radionuclides, such as 131I or 90Y, were also determined. With the exception of pharmacokinetics derived from the H310A indium study, all tumor-to-blood AUC ratios were higher for the H310A/H435Q followed by the H310A and finally the I253A fragment (Fig. 2A). Enhancement was approximately a factor of 1.43 and 1.87 for the double mutant over the H310A and I253A fragments for pharmacokinetics derived from the iodine biodistribution data. For the indium pharmacokinetics, enhancement of the H310A fragment over the H310A/H435Q and I253A proteins was 1.29 and 2.41, respectively. When decay was included, the tumor-to-blood AUC ratio for the double mutant was better by 1.44- and 1.73-fold over the H310A and I253A fragments for 131I, and 1.17- and 1.60-fold over the H310A and I253A for the 90Y label. Thus, for radioimmunotherapy with 131I, the double mutant is expected to show an ∼44% improvement in tumor-to-marrow dose compared with the H310A fragment, whereas that dose would increase by 17% if the therapeutic label is 90Y (28).

Table 4.

AUC and ratio analyses of unlabeled (pharmacokinetic), 131I–, and 90Y-DOTA–labeled T84.66 scFv-Fc fragments

Abbreviations: T/L, tumor-to-liver; T/K, tumor-to-kidney; T/B, tumor-to-blood; PK, pharmacokinetic.

Figure 2.

AUC analysis and maximum tumor dose predictions of scFv-Fc fragments. A, tumor-to-blood AUC ratios of three scFv-Fc proteins, either not radiolabeled [pharmacokinetic (PK) data] or potentially labeled with 131I or 90Y therapeutic isotopes. B, tumor-to-liver AUC ratios. C, tumor-to-kidney AUC ratios. D, maximum tumor dose estimation of both 131I- and 90Y-labeled scFv-Fc fragments. The intact cT84.66 antibody is included as a reference. The limiting organ toxicity is indicated above the columns. M, marrow; L, liver.

Figure 2.

AUC analysis and maximum tumor dose predictions of scFv-Fc fragments. A, tumor-to-blood AUC ratios of three scFv-Fc proteins, either not radiolabeled [pharmacokinetic (PK) data] or potentially labeled with 131I or 90Y therapeutic isotopes. B, tumor-to-liver AUC ratios. C, tumor-to-kidney AUC ratios. D, maximum tumor dose estimation of both 131I- and 90Y-labeled scFv-Fc fragments. The intact cT84.66 antibody is included as a reference. The limiting organ toxicity is indicated above the columns. M, marrow; L, liver.

Close modal

Tumor-to-liver and tumor-to-kidney AUC ratios (Table 4) were also calculated because these clearance organs could be dose limiting, especially in the case of 90Y therapy. The double mutant exhibited the lowest tumor-to-liver ratio (1:1; Fig. 2B) in the 90Y scenario, which indicates a relatively high liver exposure when an objective therapeutic response is sought with a high tumor dose. The tumor-to-kidney AUC ratio (Fig. 2C) was about the same for all three 90Y-labeled scFv-Fcs and was ∼4-fold higher than the tumor-to-liver AUC ratio. Labeled with the therapeutic 131I isotope, the H310A/H435Q double mutant shows superior tumor-to-liver and tumor-to-kidney ratios (Table 4; Fig. 2B and C) over the other two fragments. Thus, based solely on the results from the AUC analyses, the H310A/H435Q double mutant fragment seems to be the best candidate for radioimmunotherapy, armed with the 131I radioisotope. This prediction was also strengthened by the estimates of maximum tumor dose achieved by the three scFv-Fcs labeled with either 131I or 90Y (Fig. 2D). The toxicity limits were set to 150 cGy for marrow, 1,500 cGy for kidney, and 3,000 cGy for liver (31). The parental intact cT84.66 antibody was also included in the comparison as a control (32). Based on the dosimetry simulation, the 131I-labeled scFv-Fc H310A/H435Q is expected to deliver the highest tumor dose of >7,000 cGy, limited by red marrow toxicity. From the scFv-Fc fragments, the I253A protein is best suited for 90Y delivery, although it cannot match the intact cT84.66 antibody.

Discussion

Many of the limitations associated with the use of intact antibodies as radionuclide delivery vehicles in radioimmunotherapy have been addressed by employing advances made in antibody engineering. The approach we have used in this work is based on manipulating antibody-receptor interactions to modulate the immunoconjugate pharmacokinetics in vivo. Previously, we engineered six anti-CEA scFv-Fc recombinant antibody fragments exhibiting a wide range of serum half-lives (20). After extensive in vitro and in vivo characterization, the three fastest-clearing fragments, I253A, H310A, and H310A/H435Q double mutant, were selected for further analysis aiming to extend their clinical usefulness.

In the present work, the biodistribution properties of 125I– and 111In-DOTA–labeled scFv-Fc I253A, H310A, and H310A/H435Q fragments were investigated in tumor-bearing animals. In general, the data suggested that there is an inverse relationship between blood clearance and tumor uptake. Specifically, as seen from the radiometal-derived data, higher clearance rate was associated with lower tumor uptake. This conclusion is not surprising considering that longer serum residence time allows for better exposure of the antibody fragment to the tumor mass. Nevertheless, even the fastest-clearing H310A/H435Q double mutant achieved relatively high tumor-to-blood ratio at 72 h (73:1) and a maximum tumor uptake (28.0% ID/g at 12 h), which is substantial for an antibody fragment with a serum half-life of 7.05 h (25). The 111In-labeled H310A variant exhibited intermediate clearance profile and tumor uptake between the I253A and H310A/H435Q mutant fragments.

The order of 125I-labeled fragment blood clearances was the same as the one observed with the radiometal-labeled scFv-Fcs. However, the tumor uptake achieved by the 125I-labeled H310A fragment was not intermediate between the I253A and H435Q/H310A fragments. It was overall lower than that of the double mutant. Tumor heterogeneity and size should not have been a factor because the radioiodine and radiometal studies were done in the same animals. The behavior of the 125I-labeled H310A fragment could be explained by a greater degree of dehalogenation and a loss of signal, resulting in apparently lower uptake. This possibility was suggested when comparing the blood curves produced by the radioiodinated and radiometal-labeled H310A fragment. The 125I-labeled H310A fragment blood curve showed overall lower activity, whereas the same comparison within the I253A and H310A/H435Q fragments yielded almost overlapping radiometal and radioiodine blood activity data. The same observation was made for the lung, where the radioiodinated I253A and H310A/H435Q fragment uptakes were very similar, yet the H310A was lower. The inconsistencies in the radioiodine data emphasize the need for radiometal biodistributions. Specifically, significant dissociation of 111In radiometal from DOTA chelate has not been reported (33); thus, a better approximation of the fragments' overall body distribution was obtained through radiometal conjugation.

The scFv-Fc fragments (105 kDa) were engineered to exceed the molecular weight threshold (∼60 kDa) for first-pass renal clearance and thus led to elimination through the liver (34). Thus, elevated hepatic uptake of the 111In-DOTA–labeled fragments in mice was expected. However, activity retained in the liver could have also been due to circulating CEA antigen (shed from the tumor), bound by the anti-CEA scFv-Fc fragment and taken up by the hepatic immunoglobulin receptor system (30). This possibility was considered based on the observed hepatic uptake of the parental intact 111In-labeled cT84.66 antibody (9.8% ID/g at 96 h) in mice, which had been attributed to circulating CEA-antibody immunocomplexes retained by the hepatocytes (29). The radiometal-labeled scFv-Fc H310A/H435Q protein, which showed the highest liver uptake in tumor-bearing mice, was therefore injected in non-tumor-bearing mice. The noted increase of hepatic accumulation in non-tumor-bearing mice compared with LS174T xenografted mice contradicted the hypothesis of circulating CEA-scFv-Fc complexes being the cause of elevated hepatic activity. The same conclusion was also reached with the 111In-labeled anti-CEA T84.66 minibody, which produced similar hepatic activity in both tumor-bearing and non-tumor-bearing mice (25). Thus, the presence of liver activity in our study is most likely due to a direct accumulation of the 111In-DOTA-scFv-Fc fragments and/or 111In-labeled metabolites. A high-performance liquid chromatography analysis of liver homogenates would be useful for revealing the nature of the radiometal-labeled scFv-Fc metabolites.

Decreased liver as well as kidney uptake of the H310A/H435Q double mutant fragment in tumor-bearing mice (compared with non-tumor-bearing animals) could be explained by the LS174T xenograft acting as an antigen sink. It has been estimated that the expression of CEA in the LS174T tumor xenografts is ∼0.56 nmol/g of tissue or 5.0 (±2.5) × 105 binding sites per cell (35). We also measured that there are ∼1.6 × 108 LS174T cells/g. The LS174T tumors weighed an average of 100 mg, which translates to ∼8 × 1012 CEA binding sites in a tumor (based on the assumption that the tumor is composed exclusively of malignant LS174T cells). Each animal received ∼5 μg of radiolabeled scFv-Fc double mutant protein, which consisted of ∼2.9 × 1013 bivalent scFv-Fc molecules (able to bind either one or two CEA moieties). From these rough calculations, it seems that the tumor has the capacity of capturing almost all injected scFv-Fc species. With a slower off-rate, these scFv-Fc molecules, which exhibit nanomolar relative binding affinity for the CEA antigen (20), should be withdrawn from the circulation and become less available to the clearance organs such as liver and kidneys.

The liver uptake of the fast-clearing 111In-labeled H310A/H435Q fragment was comparable to the level of tumor accumulation. With increasing serum half-life, the overall radioactivity in the liver measured during the course of the study (72 h) decreased. Hence, the slow-clearing 111In-labeled I253A fragment had the lowest liver activity, followed by the H310A and the H310A/H435Q mutants. This observation could possibly be due to a combination of factors, such as the antibody fragment kinetics (including binding affinity for the FcRn) and tumor size. A physiologically based model that describes the concentration-time profile on a per organ basis could be useful in providing a more detailed picture of the interaction between the scFv-Fc fragments and liver (36). This interaction needs consideration because it might have implications in radiometal-based therapy using the scFv-Fc fragments, where liver could become a dose-limiting organ.

The radioiodinated scFv-Fc H310A/H435Q double mutant was injected in immunocompetent BALB/c mice, as well as in immunocompromised non-tumor-bearing and LS174T tumor–bearing athymic nude mice, enabling comparison of the behavior of this fragment in the three animal systems. We found that the serum clearance of the scFv-Fc fragment in BALB/c mice was slower than that in both non-tumor-bearing and tumor-bearing nude mice. This finding is consistent with reports suggesting rapid elimination of injected mouse IgG2a, human IgG1, and chimeric (mouse/human) monoclonal antibodies in nude mouse strains (37, 38). Our scFv-Fc fragments were designed to be chimeric molecules, incorporating human IgG1 Fc region and mouse scFv. The reports reveal that rapid clearance of the injected antibody molecules occurred in nude mice exhibiting decreased level of endogenous IgG2a and IgG2b. CD64, also known as FcγRI receptor and expressed in high numbers on macrophages in the liver and spleen, may be responsible for the noted rapid blood clearance. When endogenous IgG2a levels are low, FcγRI binding sites are not occupied and the receptor would bind the injected antibody molecules, facilitating their degradation. Thus, when biodistribution studies are carried out in nude mice, the phenomenon of faster serum clearance of injected antibodies and some antibody fragments needs to be taken into consideration, especially when one is trying to extrapolate data and predict the pharmacokinetic characteristics of these molecules in humans.

The 131I and 90Y radionuclides are currently the primary choices for radioimmunotherapy (39). More recently, 177Lu has emerged as a promising radiopharmaceutical due to its low β energy, which reduces the negative effect of radiation on normal tissues surrounding the tumor. Due to its low tissue-penetration range, 177Lu has been projected to be appropriate for smaller tumors. In addition, 177Lu also emits γ rays; thus, it can be used for biodistribution and imaging studies. Because 90Y does not emit γ-photons, 111In-labeled antibodies are generally used as chemical and biological surrogates (40–42) to study biodistribution and estimate radiation dosimetry of 90Y-labeled antibodies. If the biodistribution of 111In-conjugated fragments is similar to the biodistribution of the same fragments bound to 177Lu (using the same chelate DOTA), then the calculated dose ratios of 177Lu are expected to be similar to those for 90Y. This speculation remains to be confirmed experimentally.

The ultimate purpose of this work was to estimate the therapeutic potential of the three anti-CEA scFv-Fc fragments. To achieve this aim, both 111In- and 125I-derived biodistribution data were used to make dosimetry predictions for 90Y- and 131I-based therapy studies. In the case of antibody molecules conjugated to radionuclides, the cumulative exposure of the tumor and normal tissues provides a first estimate of both the therapeutic efficacy and toxicity. The potential advantage of the scFv-Fc double mutant over the other two fragments is suggested by the AUC ratios between tumor and blood, which are 4.7:1 for the 131I-labeled H310A/H435Q and 5.6:1 for the 90Y-labeled H310A/H435Q. From this analysis, it seems that the 90Y-labeled scFv-Fc H310A/H435Q double mutant would be the best candidate for therapy. However, examining the tumor-to-organ ratios reveals that the scFv-Fc H310A/H435Q mutant labeled with radioyttrium has ∼1:1 AUC ratio between tumor and liver, which implies that the dose to the tumor will be equal to the liver dose. The radioiodinated double mutant fragment, on the other hand, shows more favorable tumor-to-liver ratio of about 29:1. The tumor-to-kidney ratios are again in favor of the 131I label coupled to the H310A/H435Q double mutant fragment. Thus, based on AUC ratios alone, the scFv-Fc H310A/H435Q is the protein of choice when labeled with the 131I radioisotope. Thus, the 131I-labeled scFv-Fc H310A/H435Q is predicted to deliver >7,000 cGy to the tumor. This provides a significant advantage of selecting the 131I-labeled scFv-Fc H310A/H435Q fragment over the use of radioiodinated intact cT84.66 antibody for therapy of CEA-expressing malignancies, whereas for the 90Y and possibly the 177Lu labels, the intact cT84.66 antibody may be the protein of choice.

Multistep approaches, such as streptavidin-biotin–based pretargeting, amplification pretargeting, and bispecific antibody pretargeting (reviewed in ref. 1), hold the potential to greatly reduce systemic toxicity associated with conventional radioimmunotherapy. Recombinant antibody fragments with controlled pharmacokinetics, such as our anti-CEA scFv-Fc fragments, could be good candidates for a multistep study because the time between the administration of the unlabeled antibody fragment (streptavidin-conjugated antibody; step 1) and the administration of the radiolabeled protein/peptide (radiolabeled biotin; step 2), for example, can be reduced. This will possibly lead to diminished organ toxicity associated with radiation damage.

Although the dosimetry predictions are quite helpful in selecting the most optimal antibody fragment-therapeutic radioisotope combination, actual therapy studies need to be conducted. Specifically, factors such as the particle energy range, energy deposition, organ-to-organ crossfire, and the γ emission of the 131I isotope were not taken into account when toxicity levels were estimated. For the 90Y estimates, subtle differences in radionuclide retention of the metal chelating agent used for 111In and 90Y labeling can result in a differential release between the two radiometals. Overall, this work provides a scheme for evaluation of the therapeutic potential of antibody fragments using biodistributions with a variety of radionuclides.

Summary

The in vivo pharmacokinetic behavior of three scFv-Fc variants exhibiting differential distribution profiles in mice has been evaluated in different mouse systems and radioisotope combinations leading to conclusions affecting their therapeutic potential. Biodistribution studies provided a platform for dosimetry predictions, yielding the fastest-clearing scFv-Fc fragment labeled with the therapeutic 131I radioisotope as the species able to deliver the highest tumor dose. An actual therapy study using the 131I-labeled scFv-Fc H310A/H435Q variant is needed to confirm these predictions.

Note: L.E. Williams, J.E. Shively, A.A. Raubitschek, and A.M. Wu are members of the City of Hope Comprehensive Cancer Center. A.M. Wu is a member of the University of California Los Angeles-Jonsson Comprehensive Cancer Center.

Acknowledgments

Grant support: NIH grants CA 43904 and CA 86306, Department of the Army grants DAMD 17-00-1-203 and DAMD 17-00-1-0150, and NIH grants CA 33572 (L.E. Williams, J.E. Shively, A.A. Raubitschek, and A.M. Wu) and CA 16042 (A.M. Wu).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

We thank Lin Li for DOTA conjugation of the three scFv-Fc fragments, Anne-Line Anderson and Randall Woo for 125I/111In labeling of the engineered proteins, Desiree Crow for assistance in the biodistribution studies, and Militza Bocic for the analysis of immunoreactivity.

References

1

Kenanova V, Wu AM. Tailoring antibodies for radionuclide delivery.

Expert Opin Drug Deliv

2006

;

3

:

53

–70.

2

Wu AM, Senter PD. Arming antibodies: prospects and challenges for immunoconjugates.

Nat Biotechnol

2005

;

23

:

1137

–46.

3

Huhalov A, Chester KA. Engineered single chain antibody fragments for radioimmunotherapy.

Q J Nucl Med Mol Imaging

2004

;

48

:

279

–88.

4

Todorovska A, Roovers RC, Dolezal O, Kortt AA, Hoogenboom HR, Hudson PJ. Design and application of diabodies, triabodies and tetrabodies for cancer targeting.

J Immunol Methods

2001

;

248

:

47

–66.

5

Wu AM, Yazaki PJ. Designer genes: recombinant antibody fragments for biological imaging.

Q J Nucl Med

2000

;

44

:

268

–83.

6

Begent RH, Verhaar MJ, Chester KA, et al. Clinical evidence of efficient tumor targeting based on single-chain Fv antibody selected from a combinatorial library.

Nat Med

1996

;

2

:

979

–84.

7

Holliger P, Prospero T, Winter G. “Diabodies”: small bivalent and bispecific antibody fragments.

Proc Natl Acad Sci U S A

1993

;

90

:

6444

–8.

8

Hu S, Shively L, Raubitschek A, et al. Minibody: a novel engineered anti-carcinoembryonic antigen antibody fragment (single-chain Fv-CH3) which exhibits rapid, high-level targeting of xenografts.

Cancer Res

1996

;

56

:

3055

–61.

9

Wu AM, Yazaki PJ, Tsai S, et al. High-resolution microPET imaging of carcinoembryonic antigen-positive xenografts by using a copper-64-labeled engineered antibody fragment.

Proc Natl Acad Sci U S A

2000

;

97

:

8495

–500.

10

Sundaresan G, Yazaki PJ, Shively JE, et al. 124I-labeled engineered anti-CEA minibodies and diabodies allow high-contrast, antigen-specific small-animal PET imaging of xenografts in athymic mice.

J Nucl Med

2003

;

44

:

1962

–9.

11

Santimaria M, Moscatelli G, Viale GL, et al. Immunoscintigraphic detection of the ED-B domain of fibronectin, a marker of angiogenesis, in patients with cancer.

Clin Cancer Res

2003

;

9

:

571

–9.

12

Wong JY, Chu DZ, Williams LE, et al. Pilot trial evaluating an 123I-labeled 80-kilodalton engineered anticarcinoembryonic antigen antibody fragment (cT84.66 minibody) in patients with colorectal cancer.

Clin Cancer Res

2004

;

10

:

5014

–21.

13

Ghetie V, Hubbard JG, Kim JK, Tsen MF, Lee Y, Ward ES. Abnormally short serum half-lives of IgG in β2-microglobulin-deficient mice.

Eur J Immunol

1996

;

26

:

690

–6.

14

Junghans RP, Anderson CL. The protection receptor for IgG catabolism is the β2-microglobulin-containing neonatal intestinal transport receptor.

Proc Natl Acad Sci U S A

1996

;

93

:

5512

–6.

15

Israel EJ, Wilsker DF, Hayes KC, Schoenfeld D, Simister NE. Increased clearance of IgG in mice that lack β2-microglobulin: possible protective role of FcRn.

Immunology

1996

;

89

:

573

–8.

16

Kim JK, Firan M, Radu CG, Kim CH, Ghetie V, Ward ES. Mapping the site on human IgG for binding of the MHC class I-related receptor, FcRn.

Eur J Immunol

1999

;

29

:

2819

–25.

17

West AP, Jr., Bjorkman PJ. Crystal structure and immunoglobulin G binding properties of the human major histocompatibility complex-related Fc receptor(,).

Biochemistry

2000

;

39

:

9698

–708.

18

Martin WL, West AP, Jr., Gan L, Bjorkman PJ. Crystal structure at 2.8 A of an FcRn/heterodimeric Fc complex: mechanism of pH-dependent binding.

Mol Cell

2001

;

7

:

867

–77.

19

Slavin-Chiorini DC, Kashmiri SV, Schlom J, et al. Biological properties of chimeric domain-deleted anticarcinoma immunoglobulins.

Cancer Res

1995

;

55

:

5957

–67s.

20

Kenanova V, Olafsen T, Crow DM, et al. Tailoring the pharmacokinetics and positron emission tomography imaging properties of anti-carcinoembryonic antigen single-chain Fv-Fc antibody fragments.

Cancer Res

2005

;

65

:

622

–31.

21

Eary JF, Krohn KA, Kishore R, Nelp WB. Radiochemistry of halogenated antibodies. In: Zalutsky MR, editor. Antibodies in radiodiagnosis and therapy. Boca Raton (FL): CRC Press, Inc.; 1989. p. 83–102.

22

Bebbington CR, Renner G, Thomson S, King D, Abrams D, Yarranton GT. High-level expression of a recombinant antibody from myeloma cells using a glutamine synthetase gene as an amplifiable selectable marker.

Biotechnology (N Y)

1992

;

10

:

169

–75.

23

Wu AM, Williams LE, Zieran L, et al. Anti-carcinoembryonic antigen (CEA) diabody for rapid tumor targeting and imaging.

Tumor Targeting

1999

;

4

:

47

–58.

24

You YH, Hefta LJ, Yazaki PJ, Wu AM, Shively JE. Expression, purification, and characterization of a two domain carcinoembryonic antigen minigene (N-A3) in Pichia pastoris. The essential role of the N-domain.

Anticancer Res

1998

;

18

:

3193

–201.

25

Yazaki PJ, Wu AM, Tsai SW, et al. Tumor targeting of radiometal labeled anti-CEA recombinant T84.66 diabody and t84.66 minibody: comparison to radioiodinated fragments.

Bioconjug Chem

2001

;

12

:

220

–8.

26

Lewis MR, Kao JY, Anderson AL, Shively JE, Raubitschek A. An improved method for conjugating monoclonal antibodies with _N_-hydroxysulfosuccinimidyl DOTA.

Bioconjug Chem

2001

;

12

:

320

–4.

27

D'Argenio DZ, Schumitzky A. A program package for simulation and parameter estimation in pharmacokinetic systems.

Comput Programs Biomed

1979

;

9

:

115

–34.

28

Siegel J, Wessels B, Watson E, et al. Bone marrow dosimetry and toxicity for radioimmunotherapy.

Antibodies Immunoconjugates Radioimmunopharm

1990

;

3

:

213

–33.

29

Beatty BG, Beatty JD, Williams LE, Paxton RJ, Shively JE, O'Connor-Tressel M. Effect of specific antibody pretreatment on liver uptake of 111In-labeled anticarcinoembryonic antigen monoclonal antibody in nude mice bearing human colon cancer xenografts.

Cancer Res

1989

;

49

:

1587

–94.

30

Beatty JD, Beatty BG, O'Conner-Tressel M, Do T, Paxton RJ. Mechanisms of tissue uptake and metabolism of radiolabeled antibody-role of antigen: antibody complex formation.

Cancer Res

1990

;

50

:

840

–5s.

31

Emami B, Lyman J, Brown A, et al. Tolerance of normal tissue to therapeutic irradiation.

Int J Radiat Oncol Biol Phys

1991

;

21

:

109

–22.

32

Williams LE, Wu AM, Yazaki PJ, et al. Numerical selection of optimal tumor imaging agents with application to engineered antibodies.

Cancer Biother Radiopharm

2001

;

16

:

25

–35.

33

Anderson CJ. Metabolism of radiometal-labeled proteins and peptides: what are the real radiopharmaceuticals in vivo?

Cancer Biother Radiopharm

2001

;

16

:

451

–5.

34

Behr TM, Goldenberg DM, Becker W. Reducing the renal uptake of radiolabeled antibody fragments and peptides for diagnosis and therapy: present status, future prospects and limitations.

Eur J Nucl Med

1998

;

25

:

201

–12.

35

Berk DA, Yuan F, Leunig M, Jain RK. Direct in vivo measurement of targeted binding in a human tumor xenograft.

Proc Natl Acad Sci U S A

1997

;

94

:

1785

–90.

36

Ferl GZ, Kenanova V, Wu AM, Distefano JJ III. A two-tiered physiologically based model for dually labeled single-chain Fv-Fc antibody fragments.

Mol Cancer Ther

2006

;

5

:

1550

–8.

37

van Gog FB, Brakenhoff RH, Snow GB, van Dongen GA. Rapid elimination of mouse/human chimeric monoclonal antibodies in nude mice.

Cancer Immunol Immunother

1997

;

44

:

103

–11.

38

Reddy N, Ong GL, Behr TM, Sharkey RM, Goldenberg DM, Mattes MJ. Rapid blood clearance of mouse IgG2a and human IgG1 in many nude and nu/+ mouse strains is due to low IgG2a serum concentrations.

Cancer Immunol Immunother

1998

;

46

:

25

–33.

39

Milenic DE, Brady ED, Brechbiel MW. Antibody-targeted radiation cancer therapy.

Nat Rev Drug Discov

2004

;

3

:

488

–99.

40

DeNardo SJ, Kramer EL, O'Donnell RT, et al. Radioimmunotherapy for breast cancer using indium-111/yttrium-90 BrE-3: results of a phase I clinical trial.

J Nucl Med

1997

;

38

:

1180

–5.

41

Lovqvist A, Humm JL, Sheikh A, et al. PET imaging of (86)Y-labeled anti-Lewis Y monoclonal antibodies in a nude mouse model: comparison between (86)Y and (111)In radiolabels.

J Nucl Med

2001

;

42

:

1281

–7.

42

Vallabhajosula S, Kuji I, Hamacher KA, et al. Pharmacokinetics and biodistribution of 111In- and 177Lu-labeled J591 antibody specific for prostate-specific membrane antigen: prediction of 90Y-J591 radiation dosimetry based on 111In or 177Lu?

J Nucl Med

2005

;

46

:

634

–41.

©2007 American Association for Cancer Research.

2007

599 Views

68 Web of Science

78 Crossref

Citing articles via

Email alerts