Purpose: There is increasing evidence that the chimeric monoclonal antibody G250 (cG250) can be internalized by G250 antigen-expressing renal cell carcinoma (RCC) cells. Thus, accumulation in tumors of cG250 labeled with residualizing radionuclides might be higher than that of nonresidualizing 131I-cG250. Here, we present a study comparing intrapatiently the accumulation of 131I-cG250 and 111In-cG250 in RCC metastases.
Experimental Design: Five patients were i.v. injected with 222 MBq 111In-ITC-DTPA-cG250 and 222 MBq 131I-cG250 on days 0 and 4, respectively. Directly and 4 days after the injection of both antibody preparations, whole body gamma camera images were acquired. The scintigraphic images were analyzed visually and quantitatively. The radioactivity in tissues was calculated and expressed as percentage injected dose in organs or percentage injected dose/g in metastases. For the latter, tumor:blood ratios were also calculated. Twenty-five metastases were analyzed completely.
Results: At 4 days postinjection, the 111In-ITC-DTPA-cG250 images revealed more metastatic lesions (n = 47) than 131I-cG250 (n = 30). Quantitative analysis of the images showed higher activities of 111In-ITC-DTPA-cG250 than 131I-cG250 in 20 of 25 lesions. The mean overall half-life of both antibody preparations in plasma was similar.
Conclusions:111In-ITC-DTPA-cG250 outperformed 131I-cG250 for visualization of metastatic RCC lesions, not just because of the superior gamma camera characteristics of 111In, but more importantly, also because higher tumor:blood ratios were obtained. The higher activities of 111In-ITC-DTPA-cG250 in metastatic lesions might be caused by internalization and subsequent intracellular retention of the radiolabel, implying that in future radioimmunotherapy trials with cG250 in RCC patients, the use of a residualizing radionuclide should be considered.
For both experimental and human RIT3 studies, 131I labeled to MoAbs targeting a specific tumor-associated antigen has been frequently used as the radionuclide of choice (1, 2, 3). The radioiodination procedure of MoAbs is relatively easy, leading to high specific activities necessary for the administration of high-activity doses of radiolabeled MoAbs, and 131I is widely available at reasonable costs. Apart from the issues associated with radiation safety, another potential disadvantage of 131I is rapid excretion of the radioiodinated catabolite from the tumor cells when a radioiodinated MoAb, labeled with a conventional method, is internalized and metabolized by the target cell (4, 5). In contrast, MoAbs labeled with metallic radionuclides, such as 111In, 90Y, or 177Lu, result in radiolabeled catabolites that are trapped in the lysosomes after internalization by the target cells (4, 6, 7).
For radioimmunodetection studies in patients with primary RCC with MoAb G250, the radionuclide 131I has been used exclusively (8, 9). Visually, targeting of primary clear cell RCC tumors improved with time when the radiolabeled antibody cleared from the background (≤7 days p.i.). In subsequent RIT trials in one-sided nephrectomized patients, excellent visualization of metastases was noted ≤2 weeks p.i. of 131I-G250 (10, 11). In addition, in nude mice with RCC xenografts (NU-12 tumor model), accumulation of radioiodinated G250 antibody in the tumor was not significantly different from that of the 111In-labeled antibody. At 3 days p.i., accumulation in the tumor was equally high for both 131I-G250 and 111In-ITC-DTPA-G250 (12, 13). These observations led to the assumption that MoAb G250 was not substantially internalized and metabolized after binding to the G250 antigen on the tumor cells.
However, several recent observations led to the assumption that at least in some of the RCC tumors, internalization followed by processing of the radiolabeled antibody may play a role. Firstly, Dürrbach et al. (14) demonstrated in vitro internalization of MoAb G250 on binding to G250 antigen-expressing cells by fluorescence microscopy analysis. Furthermore, in the SK-RC-52 nude mice tumor model, we observed major differences in accumulated activities of G250 labeled with different radionuclides: the activity of MoAb G250 labeled with the residualizing radionuclides 88Y and 177Lu being approximately three to four times higher than 125I-G250 in the tumors at 7 days p.i.4
When these differences in handling of the radiolabeled G250 also play a role in humans, some RCC lesions in patients could be targeted more efficiently using MoAb G250 labeled with a residualizing radionuclide, such as 90Y or 177Lu. This would imply that in RIT studies, higher radiation doses could be guided to RCC metastases when MoAb G250 is radiolabeled with residualizing radionuclides (15, 16). Because 90Y does not emit γ-photons, 111In is generally used as a surrogate to determine the biodistribution of 90Y-labeled MoAbs in patients (17, 18, 19).
In the present study, we performed an intrapatient comparison of the biodistribution and tumor targeting of metastatic RCC lesions of 131I-labeled cG250 and 111In-labeled cG250. Patients received consecutive injections with 131I-G250 and 111In-ITC-DTPA-G250. The scintigraphic images were analyzed visually and quantitatively by calculating the accumulated activity per metastasis. In addition, the biodistribution of the two radiolabeled forms of MoAb G250 in relevant organs was assessed.
Materials and Methods
Five patients (4 female, 1 male; mean age 57 years, range 47–71) who consented to take part in an ongoing Phase I/II RIT trial with 131I-labeled cG250 were included in this study. Patients with metastatic RCC who had undergone a tumor nephrectomy were eligible if the primary RCC tumor had a clear cell histology. The study was approved by the Institutional Review Board of the University Medical Center Nijmegen. Before study entry, written informed consent was obtained from all patients.
The isolation and immunohistochemical reactivity of MoAb G250 have been described elsewhere (9, 20). To reduce the immunogenicity of the antibody, a chimeric version has been developed (21). The reactivity of MoAb cG250 to normal human tissues is restricted to the (upper) gastrointestinal mucosa (stomach, ileum, proximal, and middle colon) and gastrointestinal related structures (intra and extrahepatic biliary system, pancreas) (20, 22). MoAb cG250 is reactive with the G250 antigen (Ka = 4 × 109 m−1), which is expressed on the cell surface of the majority of the clear cell type RCCs (9, 23). The G250 antigen has been identified as the tumor-associated isoenzyme carbonic anhydrase isotype IX (MN/CA IX) (24, 25).
Radiolabeling and Quality Control.
ITC-DTPA (Macrocyclics, Richardson, TX) was conjugated to MoAb cG250 in a 0.1 m NaHCO3 buffer (pH 8.2), using a 50-fold molar excess of ITC-DTPA as described by Ruegg et al. (26) with minor modifications (conjugation reaction during 1 h at room temperature). The number of ITC-DTPA ligands per cG250 molecule was determined as described by Hnatowich et al. (27). The DTPA-MoAb cG250 was purified by extensive dialysis against 0.15 m acetate buffer (pH 5.5), at 4°C, vialed in sterile vials, and stored at −20°C until use. The cG250-DTPA conjugate (5 mg, 1 mg/ml) was radiolabeled with 111InCl3 (Tyco Healthcare, Petten, the Netherlands). The specific activity of the final preparation was adjusted to 44.4 MBq 111In/mg MoAb cG250 by adding unlabeled MoAb cG250.
MoAb cG250 was radioiodinated with 131I (MDS Nordion, Fleurus, Belgium) according to the IodoGen method, using a remote system as described previously (9, 11). Radioiodinated cG250 was purified by AG1-XR resin filtration (Bio-Rad Laboratories, Hercules, CA) in PBS. Again, the specific activity of the final preparation was adjusted to 44.4 MBq 131I/mg MoAb cG250 by adding unlabeled antibody.
The radiochemical purity of each of the radiolabeled cG250 preparations was determined by ITLC using ITLC silica gel strips (Gelman Sciences, Inc., Ann Arbor, MI) using 0.15 m citrate buffer (pH 5.0) as the mobile phase (release criterion > 95% protein bound). The immunoreactive fraction at infinitive antigen excess of both radiolabeled cG250 preparations was determined on freshly trypsinized SK-RC-52 RCC cells essentially as described by Lindmo et al. (9, 28) with minor modifications.
To assess the stability of the 111In-labeled preparation in vitro, the preparation was diluted in plasma at an activity concentration of 50 μCi/ml. A final concentration of 0.5 mm EDTA was added to capture released radiometal, and the preparation was incubated at 37°C. Plasma samples were analyzed at 0 and 4 h and 1, 2, 4, 7, 12, 19, and 25 days by fast protein liquid chromatography using a Biosep Sec-S3000 gel filtration column (Phenomenex, Torrance, CA).
Study Design and Radioimmunoscintigraphy.
At day 0, patients received an i.v. injection of 222 MBq 111In-ITC-DTPA-cG250 in 0.9% NaCl (5 mg protein, total volume 10 ml) over a 5-min period. Directly p.i. and 4 days later, whole-body planar images, anterior and posterior view, were recorded using a double-headed gamma camera (MultiSpect 2; Siemens Inc., Hoffman Estates, IL) equipped with parallel-hole medium energy collimators (scan speed 8 and 4 cm/min, respectively). An aliquot of the injected dose was counted simultaneously. Symmetric 15% windows were used over both the 172 and 246 keV energy peaks. The data were stored digitally in a 256 × 1024 matrix. Directly after the recording of the 111In-ITC-DTPA-cG250 images at 4 days p.i., the patient was i.v. injected with 222 MBq 131I-cG250 (5 mg of protein, total volume 10 ml). Again, directly p.i. and 4 days later, whole-body planar images were recorded (high energy collimators, symmetric 15% window over 364 keV, scan speed 5 and 4 cm/min, respectively) and stored digitally in a 256 × 1024 matrix.
To block 131I uptake in the thyroid, patients received 100 mg of potassium iodide twice daily and 200 mg of potassium perchlorate four times daily, starting at the day of the administration of the radioiodinated MoAb. Vital signs were measured on the days of radiolabeled antibody administration until 3 h after injection. In addition, blood samples were drawn directly p.i. of each radiolabeled antibody preparation and at 4 days p.i. The activity in plasma samples was expressed as % ID/g. On the basis of these measurements, the overall half-life in the first 4 days after injection of both radiolabeled MoAbs in plasma was determined and expressed in hours.
Quantitative Analysis of the Scintigraphic Images.
The activity of the radiolabeled MoAb in the RCC metastases and relevant organs, showing accumulation of the radiolabeled antibody on the scintigraphic images, was derived from the four sets of images of each patient. ROIs were drawn manually over the total body; the organs heart, lung, liver, spleen, remaining kidney, and bowel; and metastatic lesions, as shown in Fig. 1. Adjacent background regions were drawn for organs in the chest, adjacent to heart and lung, respectively, whereas for the abdominal organs, one abdominal background region was drawn, not overlying any organ or metastasis.
The absolute activity in the organs and metastases was estimated using the conjugated view counting method by calculating the GM of the background corrected counts in an ROI on the anterior and posterior views (29, 30), using the formula GM =
Before calculation of the GM, both anterior and posterior measurements were corrected for background, using the partial background subtraction method as described previously (29, 30). In this method, ROI counts of the liver were not corrected for background activity because the liver occupies almost the entire thickness of the patient’s abdomen (F = 0). For the heart, this factor F was set at 0.5, and for lung, spleen, and kidney, F was set at 0.66. Furthermore, to correct for attenuation differences between the abdomen and pelvis as compared with the chest, the uptake of chest organs was multiplied by 0.85 (29). The uptake in organs and metastases at 4 days p.i. was corrected for physical decay. The activity in tissues was expressed as % ID, setting the number of counts in the total body image recorded directly after injection at 100%. The activity in the lungs was determined as twice the calculated activity in the right lung.
Only metastatic lesions that were identified and measurable on computed tomography and visualized on at least one of the scintigraphic images were analyzed quantitatively. The uptake in metastases was expressed as %ID/g, assuming that 1 ml of tissue equals 1 gram. The volume of metastatic lesions was derived from computed tomography measurements, using the formula π (length × width × height)/6. Since metastatic lesions were not visualized with either 131I- or 111In-labeled cG250 directly p.i., ROIs could not be drawn. Therefore, the %ID/g in the metastatic lesions at 4 days p.i. was calculated only. The activity in deeper located metastases was calculated using the GM method without background correction or attenuation correction. For s.c. metastases (n = 9), only the counts in the anterior or posterior ROI were used. Finally, the tumor:blood ratio was calculated for each metastasis.
All values are expressed as mean ± SD. Statistical analysis was performed using the paired Student’s t test. Differences were considered significant when P < 0.05.
Radiolabeling and Quality Control.
The radiochemical purity of 111In-ITC-DTPA-cG250 was always >95%. The labeling efficiency of the radioiodination ranged between 71 and 90% (mean 83%). After purification, >98% of the 131I-activity in the antibody preparation was protein bound as determined by ITLC. The immunoreactive fraction of 111In-ITC-DTPA-cG250 and 131I-cG250 ranged between 77 and 100% (mean 90%) and between 87 and 100% (mean 96%), respectively. The fast protein liquid chromatography analysis of the preparation after incubation in plasma demonstrated that the 111In-ITC-DTPA-cG250 preparation was quite stable, because <3% of the radiolabel was released during 25 days of incubation.5
The injection of the radiolabeled antibody was tolerated well by all patients, and no side-effects were observed. No significant changes in blood pressure, pulse rate, and body temperature were detected.
Directly p.i. the 111In-ITC-DTPA-cG250 activity in plasma was 0.034 ± 0.006 %ID/g and 0.027 ± 0.003 %ID/g for 131I-cG250 (not significant). Four days later, the activity in plasma of 111In-ITC-DTPA-cG250 and 131I-cG250 was 0.0078 ± 0.0020 %ID/g and 0.0076 ± 0.0011 %ID/g, respectively (not significant). The mean overall half-life of the 111In-ITC-DTPA-cG250 and 131I-cG250 preparations in plasma in the first 4 days after injection were similar; 58 ± 18 h and 48 ± 6 h, respectively (not significant).
RCC metastases were detected with both 111In-ITC-DTPA-cG250 and 131I-cG250 at 4 days p.i. Visualization was superior with the 111In-labeled antibody as illustrated in Fig. 2,A, and therefore, the 111In-ITC-DTPA-cG250 images were much more informative (Fig. 2,B). With 111In-ITC-DTPA-cG250, more metastatic lesions could be detected in 4 patients (Fig. 2 C), whereas an adrenal metastasis could only be detected scintigraphically with 131I-cG250. In total, 111In-ITC-DTPA-cG250 detected 47 metastases in these 5 patients, whereas 131I-cG250 revealed 30 metastases.
Quantitative Analysis of the Scintigraphic Images.
A total of 25 RCC tumor lesions was evaluated quantitatively. The majority of the tumor lesions (n = 21) were both visualized with 111In-ITC-DTPA-cG250 and 131I-cG250. Of the 25 metastases selected for further qualitative assessment, two lung lesions in one patient and one s.c. lesion in another patient were only visualized scintigraphically with 111In-ITC-DTPA-cG250, whereas in yet another patient, one metastasis was only detected scintigraphically with 131I-cG250. The quantitatively calculated activities of both radiolabels in these lesions expressed as % ID/g are shown in Fig. 3 A. In 20 lesions, the calculated activity of 111In-ITC-DTPA-cG250 was higher than the calculated activity of 131I-cG250; in four metastases, the activities were equal; and in only one metastasis, the activity of 131I-cG250 was higher than the 111In-ITC-DTPA-cG250 activity.
Because different clearance rates of the radiolabeled antibody from the blood might affect tumor uptake, the T/B ratio for both radiolabels was calculated for each lesion (Fig. 3 B). The results were highly similar to the calculated % ID/g, showing substantially higher T/B ratios with 111In-ITC-DTPA-cG250 than with radioiodinated MoAb in 21 metastases. In three lesions, the ratios were equal, and in only one lesion, the ratio was higher for 131I than 111In.
The organ uptake (%ID) of the 111In-labeled and radioiodinated cG250 directly after injection and 4 days later is summarized in Fig. 4. Directly after injection, there were no significant differences between the two preparations, except for the kidney; activity of 131I-cG250 in the kidney being significantly higher than activity of 111In-ITC-DTPA-cG250 (Fig. 4,A). At 4 days p.i., the activity of 111In-ITC-DTPA-cG250 was slightly higher in all organs compared with the activity of 131I-cG250 and markedly higher in the liver [12.7 ± 3.5 %ID versus 4.5 ± 1.1 %ID (P < 0.05)] and also in the bowel [3.2 ± 0.3 %ID versus 1.1 ± 0.3 %ID (P < 0.05; Fig. 4 B)].
In this study, the quantitatively calculated activities in RCC metastases of a radioiodinated form of MoAb cG250 were directly compared with the calculated activities of cG250 labeled with 111In, within the same patient. Because both 131I- and 111In-labeled cG250 are highly stable in serum, we assumed that the observed activity distribution is representative for the distribution of MoAb cG250, labeled with either 131I or 111In. Although a visual comparison of activity in metastases after i.v. injection of a radioiodinated MoAb and an 111In-labeled MoAb has been performed by other investigators, quantitation of the activity in metastatic lesions was not attempted in earlier studies (31, 32, 33). In the current study not only more metastatic lesions were visualized in 4 of the 5 patients studied, which is in line with these earlier observations, but also quantitative analysis of the activity in 25 metastases demonstrated higher accumulation levels of 111In-ITC-DTPA-cG250 compared with 131I-cG250 in 21 lesions. The activity in the metastases was quantitated to provide objective measurements of labeled antibody accumulation in tumor and organs, since visually apparent differences between the 131I-cG250 and 111In-DTPA-cG250 images might have been attributable to the inferior resolution of the 131I images due to a difference in radioisotope characteristics and collimators. These higher levels were not caused by differences in clearance rates of the radiolabeled antibody from the blood, because the mean overall half-life in plasma did not differ significantly, and the T/Bs showed a highly similar pattern as the quantitatively calculated activity in the metastases.
Because of the shorter half-life, but more importantly the lower energy γ-radiation of 111In, the 111In-labeled MoAb was injected first, and the 131I-labeled cG250 was injected 4 days later. In a previous study, we have shown that the accumulation of two radioiodinated cG250 preparations in primary tumors was identical when the two separate injections were given 4 days apart (34). Moreover, we also demonstrated in patients that at antibody protein doses between 2 and 10 mg accumulation of the radiolabel in the tumor was not affected by antibody protein dose (9). Therefore, it is unlikely that the accumulation of 131I-cG250 in RCC metastases in this study was reduced because of occupation of G250 antigens on the tumor lesions by the 111In-labeled MoAb, and this does not explain our observations.
The absolute values of the calculated activities in the metastases in the present study have to be interpreted with some caution. In the dosimetric calculations applied in this study, various assumptions could have especially affected the activity values in small metastases. Therefore, no background or attenuation corrections were applied in the analysis of the metastatic lesions. However, the activity of 111In-ITC-DTPA-cG250 and 131I-cG250 in RCC lesions was measured in the same patient, and the 111In and 131I images, although not recorded in an identical manner, were subsequently analyzed using the same method. Although the true activity in vivo of both radiolabeled preparations may not be reflected completely by the calculated %ID/g, their relative magnitude does allow direct comparison of each metastasis.
The difference in %ID in the organs and total body between 111In-ITC-DTPA-cG250 and 131I-cG250 at 4 days p.i. is in line with earlier observations (32, 33). In general, 111In-labeled MoAbs have a higher liver accumulation than radioiodinated MoAbs (32, 33). This may hamper the visualization of metastases in the liver. However, in this study, metastases in the hepatic region were at least as effectively visualized with 111In-ITC-DTPA-cG250 than with the radioiodinated antibody, despite a ∼3-fold higher accumulation in the liver of 111In-labeled G250 compared with 131I-cG250.
The mechanism that caused the differences in accumulation in RCC metastases of the 111In-labeled and radioiodinated form of MoAb cG250 can be explained by internalization and subsequent metabolization of the radiolabeled MoAb by the tumor cells. Although in vitro cG250 is not internalized very rapidly, exemplified by a relatively low in vitro processing rate, the internalization rate appears to be highly dependent on the characteristics of a particular cell line (5). On the other hand, in recent biodistribution experiments in nude mice with SK-RC-52 xenografts, a marked difference between tumor accumulation of coinjected radioiodinated and 111In-labeled cG250 was observed caused by differences in tumor retention after internalization of both radiolabeled antibody preparations (5). These paradoxical results between in vitro and in vivo data are in line with studies showing predominantly different internalization routes of radiolabeled MoAbs in cells in vitro and in vivo (35). After incubation of lymphoma cells with a fluorescent anti-CD20 antibody, Mattes and Michel observed internalization and translocation of the antibody into a nonlysosomal compartment, identified as the endocytic recycling compartment without subsequent excretion from the cell. In contrast, in vivo the MoAb was presumably transported to the lysosomes with subsequently rapid excretion of the nonresidualizing radiolabel (35). Thus, the translocation of MoAbs to the endocytic recycling compartment may in fact be an in vitro artifact. Internalization and subsequent translocation of MoAb G250 to the endocytic recycling compartment in vitro has been recently described by Dürrbach et al. (14).
The mechanism of internalization and subsequent excretion may also explain why radioimmunoscintigraphy with 131I-cG250 was inferior in detecting RCC metastases in a direct comparison to FDG-PET (36). Because evidently more metastases can be visualized with 111In-ITC-DTPA-cG250 than 131I-cG250, more metastatic lesions might have been detected with 111In-ITC-DTPA-cG250.
Finally, one might argue that the superior retention of 111In-ITC-DTPA-cG250 in RCC metastases could also be caused by deiodination of the radioiodinated antibody in the body. There is no evidence that the latter process plays a significant role for radioiodinated antibodies in vivo (4). Regardless of the mechanism, this study shows a marked difference in tumor accumulation between a residualizing and a nonresidualizing radiolabel. Because higher activities in metastases will lead to higher radiation doses to the tumor if residualizing radionuclides, such as 177Lu and 90Y, are used (15, 16), these radionuclides should be considered for a potentially more effective RIT in patients with metastatic RCC using MoAb cG250.
In conclusion, we have demonstrated that evidently higher activities in RCC metastases can be achieved after i.v. injection of 111In-ITC-DTPA-cG250 as compared with 131I-cG250. This difference is attributable to different handling of the radiolabeled MoAb, most likely internalization and subsequent metabolization. After processing in the tumor cells, the radiolabeled catabolite of 131I-cG250 is rapidly excreted from the cells, whereas the catabolite of 111In-ITC-DTPA-cG250 is retained in the tumor cells. The higher activity levels obtained with 111In-ITC-DTPA-cG250 implies that in RIT trials, potentially higher radiation doses could be guided to RCC metastases when MoAb G250 is radiolabeled with residualizing radionuclides, such as 177Lu or 90Y. Therefore, in future RIT trials with cG250 in RCC patients, the use of a residualizing radionuclide should be considered.
Presented at the “Ninth Conference on Cancer Therapy with Antibodies and Immunoconjugates,” October 24–26, 2002, Princeton, NJ. Supported by the Dutch Cancer Society (Koningin Wilhelmina Fonds), Grant KUN99-1973, and sponsored by the Ludwig Institute for Cancer Research (New York, NY), and Wilex A. G., Munich, Germany. E. O. was supported by the Ludwig Institute for Cancer Research (New York, NY).
The abbreviations used are: RIT, radioimmunotherapy; MoAb, monoclonal antibody; cG250, chimeric G250; RCC, renal cell carcinoma; %ID/g, percentage injected dose per gram; p.i., postinjection; ITC-DTPA, isothiocyanatobenzyl-diethylenetriaminepentaacetic acid; ITLC, instant thin layer chromatography; ROI, region of interest; GM, geometric mean; T/B, tumor:blood ratio.
Manuscript submitted for publication.
Data on file.
We thank W. J. M. van den Broek and E. B. Koenders for labeling of cG250 with 131I and 111In, respectively, and J. van Eerd and C. Frielink for excellent technical assistance in the stability and animal studies. We also thank L. A. Kiemeney, Department of Epidemiology and Biostatistics and Department of Urology, University Medical Center Nijmegen, for his advice on the statistical analysis.