Abstract
Studies have demonstrated that the bivalent 111In-labeled leukotriene B4 (LTB4) antagonist DPC11870 reveals infectious and inflammatory lesions in various rabbit models. The radioactive tracer accumulates quickly at the site of infection and clears rapidly from the circulation, resulting in high-quality images. In this study, 2 new hydrazinonicotinamide (HYNIC)-conjugated compounds that are structurally related to DPC11870 were studied to further improve image quality. Methods: A bivalent HYNIC-conjugated LTB4 antagonist (MB81) and a monovalent one (MB88) were labeled with 99mTc. The radiolabeled compounds were intravenously injected into New Zealand White rabbits with E. coli infection in the left thigh muscle. The imaging characteristics of both compounds were compared with those of the bivalent 111In-labeled LTB4 antagonist. Results: Both 99mTc-labeled LTB4 antagonists revealed the abscess from 2 h after injection onward. Abscess uptake at 8 h after injection was similar for both compounds (0.22 ± 0.08 percentage injected dose per gram [%ID/g] and 0.36 ± 0.13 %ID/g for the bivalent and monovalent compounds, respectively). However, visualization of the abscess and the quality of the images were better after injection of MB88 than after injection of either of the bivalent LTB4 antagonists. The excellent delineation of the abscess by MB88 was mainly due to the more rapid clearance of this compound from nontarget organs. Conclusion: The 99mTc-labeled HYNIC conjugated LTB4 antagonists MB88 and MB81 revealed infectious foci in rabbits within a few hours after injection. Imaging characteristics of monovalent 99mTc-MB88 were superior to those of the bivalent LTB4 antagonists DPC11870 and MB81. Therefore, of the 3 LTB4 antagonists, the monovalent LTB4 antagonist MB88 is the most potent and promising agent for visualizing and evaluating infection and inflammation in patients.
Detection and localization of infectious and inflammatory lesions is important for appropriate treatment. One method of detecting infection and inflammation is scintigraphy. For scintigraphic detection of inflammatory and infectious lesions in patients, 67Ga-citrate and radiolabeled leukocytes are the most commonly applied radiopharmaceuticals (1). However, there is a continuous search for new radiopharmaceuticals that allow rapid and accurate visualization of the lesions and that lack the disadvantages of currently used compounds (e.g., laborious and time-consuming preparation and contamination risks) (2). Chemotactic and chemokinetic peptides and their antagonists are attractive candidates for this application (3). These compounds accumulate at the site of infection because of specific receptor interaction and clear rapidly from nontarget tissues. Compounds such as interleukin-2, interleukin-8, and platelet factor 4 accumulate at the site of infection because of specific interaction with receptors expressed on activated and infiltrated granulocytes (4–6). The leukotriene B4 (LTB4) receptor, expressed on polymorphonuclear granulocytes, is involved in leukocyte function during the inflammatory response (7,8). Several studies have been performed with a radiolabeled LTB4 antagonist to visualize infection and inflammation by targeting its specific receptors expressed on activated and infiltrated neutrophils (9,10). The LTB4 antagonist 111In-DPC11870 rapidly revealed infectious and inflammatory lesions in several rabbit models of acute inflammation (11,12). The compound cleared rapidly from the nontarget tissues, and physiologic uptake was seen only in the spleen, bone marrow, and kidneys. In an in vivo receptor-blocking experiment, we demonstrated that accumulation at the site of infection was receptor mediated (10).
Although the promising characteristics of 111In-DPC11870 warranted clinical evaluation in patients, for this application a 99mTc-labeled tracer is preferred. The use of 99mTc could improve image resolution and would reduce the radiation dose to the patient. Therefore, 2 analogs of DPC11870 were synthesized in which the diethylenetriaminepentaacetic acid (DTPA) chelator was replaced by the bifunctional chelator succinimidyl-hydrazinonicotinamide (S-HYNIC). One compound was structurally almost identical to DPC11870, differing only in metal ligand, and consisted of 2 LTB4 receptor-binding sites. The other LTB4 antagonist was a monomeric derivative of DPC11870 consisting of 1 LTB4 binding site and again conjugated with HYNIC. In this study, we labeled both LTB4 antagonists with 99mTc and compared the imaging characteristics of both compounds with those of 111In-DPC11870.
MATERIALS AND METHODS
Radiolabeling
The chemical structures of the 2 HYNIC-conjugated LTB4 antagonists are presented in Figure 1. Synthesis of the HYNIC-conjugated LTB4 antagonist MB81 (bivalent, BMS57868-81) and MB88 (monovalent, BMS57868-88) will be described in detail elsewhere. MB81 and MB88 (25 μg) were labeled with 555 MBq of Na-99mTcO4 (Tyco Healthcare Mallinckrodt) in the presence of 15 mg of tricine and 2 mg of isonicotinic acid in 350 μL of phosphate-buffered saline, pH 7.0, and 25 μL of SnSO4 (1 mg/mL) in 0.1N HCl. After incubation for 30 min at 100°C, quality control was performed using reverse-phase high-performance liquid chromatography (HPLC) (1100 series; Agilent Technologies) on a C18 column (Zorbax Rx-C18, 4.6 × 250 mm; Agilent Technologies) at a flow rate of 1 mL/min with a gradient mobile phase from 100% ammonium acetate (10 mmol/L; pH 7.2) to 100% acetonitrile in 20 min.
Chemical structures of MB81 (A) and MB88 (B).
DPC11870 was labeled with 111InCl3 (Tyco Healthcare Mallinckrodt) as described previously (10,12). The 20 μg of LTB4 antagonist DPC11870 were labeled with 74 MBq of 111In. Quality control of the DPC11870 compound was performed using the same HPLC protocols as described above.
Infection Model
Fifteen female New Zealand White rabbits weighing 2.3–2.8 kg were kept in cages (1 rabbit per cage) and fed standard laboratory chow and water ad libitum. The rabbits were anesthetized by subcutaneous injection of 0.7 mL of a mixture of fentanyl (0.315 mg/mL) and fluanisone (10 mg/mL) (Hypnorm; Janssen Pharmaceutica). Twenty minutes afterward, an E. coli infection was induced in the left thigh muscle by intramuscular injection of 4 × 109 colony-forming units of E. coli bacteria.
All animal experiments were approved by the local animal welfare committee in accordance with the guidelines of the Dutch legislation.
Imaging and Biodistribution
Twenty-four hours after induction of the abscess, when swelling of the infected muscle was apparent, tracer was injected via the lateral ear vein. Five rabbits received 37 MBq of 99mTc-MB81 (3.0 μg), 5 received 37 MBq of 99mTc-MB88 (3.0 μg), and 5 received 11 MBq of 111In-DPC11870 (3.0 μg). For scintigraphic imaging, the rabbits were immobilized in a mold and placed prone on a γ-camera (Orbiter; Siemens), using a low-energy parallel-hole collimator for the 99mTc-labeled compounds and a medium-energy collimator for the 111In-labeled compound. Images (300,000 counts per image) were obtained at several times, starting immediately after injection and continuing until 8 h after injection, and were stored digitally in a 256 × 256 matrix. All images were adjusted to the same background and intensity levels, allowing a fair comparison among them. The scintigraphic results were analyzed using regions of interest drawn over the heart, lungs, liver, spleen, kidneys, bone marrow, abscess, and contralateral muscle (background). Abscess-to-organ ratios were calculated. Blood samples were drawn at 1, 3, 5, 10, 30, 60, 120, 240, and 480 min after injection. The activity in the samples was determined in a γ-counter (Wizard; Canberra Packard) and expressed as a percentage of the injected dose (%ID), assuming that the total blood weight accounted for 6% of the total body weight (13). The distribution half-life (t1/2α) and clearance half-life (t1/2β) were calculated assuming a 2-phase linear model for blood clearance. At 8 h after injection, all rabbits were euthanized. A blood sample was taken by cardiac puncture. Tissues were dissected and weighed. The activity in tissues was measured in a shielded well-type γ-counter together with the injection standards and was expressed as %ID/g.
Statistical Analysis
All mean values are presented as mean ± SD. Statistical analysis was performed using 1-way ANOVA. Results were corrected for multiple datasets with the Bonferroni multiple-comparisons test. The level of significance was set at 0.05.
RESULTS
Radiolabeling
Results of the HPLC radioactivity analysis indicated that both the bivalent MB81 and the monovalent MB88 could be labeled with 99mTc with a specific activity of 37 MBq/μg (110 MBq of MB81 and 60 MBq of MB88 per nanomole). The labeling efficiency exceeded 95%. Reverse-phase HPLC analysis showed that the labeling efficiency of DPC11870 was more than 95%, with a specific activity of 3.7 MBq/μg (12 MBq/nmol).
Imaging and Biodistribution
Analysis of the blood samples indicated that 99mTc-MB88 cleared rapidly from the circulation (Fig. 2). The t1/2α was 7 min for 99mTc-MB88 and 20 min for the bivalent 99mTc-labeled antagonist MB81. t1/2β also differed between the 99mTc-labeled LTB4 antagonists. The t1/2β of the monovalent MB81 was 16.5 h, whereas the t1/2β of 99mTc-MB88 was much longer (85.6 h).
Blood clearance of 111In-DPC11870, 99mTc-MB81, and 99mTc-MB88 in New Zealand White rabbits. Average amount of radioactivity of 3 rabbits is expressed as %ID/g present in blood. Curves represent best nonlinear exponential decay calculated by Prism software program (GraphPad Software, Inc.).
Scintigraphic images after injection of the bivalent 111In-DPC11870, the bivalent 99mTc-MB81, and the monovalent 99mTc-MB88 are shown in Figure 3. Immediately after injection, the whole-body distribution of all 3 radiolabeled LTB4 antagonists was similar. Accumulation of radioactivity was observed in the lungs, liver, kidneys, and bone marrow of all animals. At 2 h after injection of the 3 LTB4 antagonists, the abscesses could already be visualized in all animals. However, images acquired with the monovalent LTB4 antagonist at that time were qualitatively superior to images acquired with either bivalent antagonist. Images acquired at 2 h or later indicated that the monovalent MB88 cleared more rapidly from the circulation and nontarget tissues than did the bivalent LTB4 antagonists MB81 and DPC11870. Additionally, compared with the other 2 LTB4 compounds, MB88 showed a low radioactivity concentration in the kidneys. Because of the low radioactivity concentration in nontarget tissues, visualization of the abscess was better in animals injected with 99mTc-MB88 than in animals injected with either 111In-DPC11870 or 99mTc-MB81 at the same time point.
Anterior images of rabbits in which intramuscular infection was induced. Images were acquired immediately and 2, 4, 6, and 8 h after injection of 11 MBq of 111In-DPC11870 (A), 37 MBq of 99mTc-MB81 (B), or 37 MBq of 99mTc-MB88 (C).
Abscess-to-organ ratios calculated from region-of-interest analysis of images acquired 8 h after injection are shown in Figure 4. When the HYNIC-conjugated antagonists were compared, only a few differences between the monovalent and the bivalent compounds were observed. The abscess-to-kidney ratio was significantly higher for the monovalent 99mTc-MB88 than for the bivalent 99mTc-MB81 (P < 0.0001). Furthermore, of all the organs and tissues, the abscess had the highest radioactivity concentration of MB88, because all ratios of the monovalent LTB4 antagonist exceeded 1. When ratios of MB88 were compared with those of 111In-DPC11870, only the abscess-to-kidney ratio differed significantly (P < 0.01). The abscess-to-organ ratios of the 2 bivalent radiolabeled compounds (99mTc-MB81 and 111In-DPC11870) were similar.
Abscess-to-organ ratios calculated from region-of-interest data obtained from images acquired 8 h after injection of 111In-DPC11870, 99mTc-MB81, or 99mTc-MB88. *P < 0.05. **P < 0.01.
The biodistribution data derived from ex vivo counting of dissected tissues, as summarized in Figure 5, agreed with the scintigraphic images. Blood levels of both 99mTc-labeled compounds were low (0.10 ± 0.01 %ID/g and 0.09 ± 0.01 %ID/g for, respectively, the bivalent and the monovalent 99mTc-labeled compounds at 8 h after injection). Furthermore, abscess uptake did not differ significantly between the 2 HYNIC-conjugated antagonists (0.22 ± 0.08 %ID/g and 0.36 ± 0.13 %ID/g). However, the radioactivity concentration significantly differed between compounds in the liver and compounds in the kidneys. Kidney uptake of 99mTc-MB88 was significantly lower than that of the bivalent 99mTc-MB81 (0.58 ± 0.06 %ID/g) (P < 0.001). Also, the radioactivity concentration in the liver was higher after injection of 99mTc-MB81. When the results for the HYNIC-conjugated antagonists were compared with those for the initially synthesized reference compound (111In-DPC11870), more differences were found. First, the radioactivity concentration in the bone marrow was much higher with 111In-DPC11870 (0.57 ± 0.19 %ID/g) than with either of the 99mTc-labeled compounds (0.26 ± 0.1 %ID/g and 0.25 ± 0.06 %ID/g for MB81 and MB88, respectively). Second, kidney uptake of 111In-DPC11870 was low and significantly different from that of the bivalent 99mTc-MB81 (P < 0.001). However, kidney uptake of 111In-DPC11870 was comparable with that of the monovalent 99mTc-MB88. Third, significant differences were found among the radioactivity concentrations of the 3 compounds in liver and spleen. In both organs, the concentration of 99mTc-MB88 was lowest (0.07 ± 0.01 %ID/g and 0.25 ± 0.01 %ID/g for liver and spleen, respectively). Spleen and bone marrow were the organs that contained the highest radioactivity concentrations of 111In-DPC11870. For the bivalent 99mTc-MB81, the highest concentration was found in the kidneys, whereas for the monovalent 99mTc-MB88, the highest concentration was found in the abscess. As a result, the abscess was best visualized with 99mTc-MB88.
Biodistribution data obtained 8 h after injection of 111In-DPC11870, 99mTc-MB81, or 99mTc-MB88. Each bar represents mean value ± SD. Values were analyzed using ANOVA. *P < 0.05, **P < 0.01, and ***P < 0.001 for 111In-DPC11870 vs. 99mTc-MB81 or 99mTc-MB88. #P < 0.05, ##P < 0.01, and ###P < 0.001 for 99mTc-MB81 vs. 99mTc-MB88.
DISCUSSION
In previous studies, we demonstrated that the bivalent 111In-labeled LTB4 antagonist DPC11870 adequately revealed infectious and inflammatory foci in distinct rabbit models (10–12). Although the results of previous studies warranted clinical studies, we synthesized 2 HYNIC-conjugated analogs of this compound to allow radiolabeling with 99mTc. In this study, the imaging characteristics of 111In-DPC11870 were compared with those of the 99mTc-labeled bivalent analog MB81 and the 99mTc-labeled monovalent derivative MB88.
The results of the imaging and biodistribution experiments indicated that, besides 111In-DPC11870, both 99mTc-labeled antagonists revealed infectious foci within a few hours after injection. Furthermore, both MB81 and MB88 showed physiologic uptake in bone marrow, spleen, and, kidneys. The radioactivity concentration of the infected tissue at 8 h after injection was similar for both 99mTc-labeled compounds. The radioactivity concentration of the monovalent 99mTc-MB88 in the kidneys was much lower than that of the bivalent 99mTc-MB81. Comparison of the 99mTc-labeled bivalent and monovalent LTB4 antagonists revealed that images acquired after injection of the monovalent 99mTc-MB88 allowed superior abscess visualization because of fast clearance of 99mTc-MB88 from nontarget tissues and good accumulation of radioactivity in the abscess. Additionally, the abscess was delineated earlier with 99mTc-MB88 than with the bivalent analog. Delineation of the abscess was worse with the bivalent 99mTc-MB81 than with 99mTc-MB88 because images were scaled identically, based on the pixels with the highest radioactivity concentration. Because of the high kidney uptake after injection of 99mTc-MB81, the amount of radioactivity in the abscess seemed to be lower but was equal, as determined in the biodistribution experiment. When the characteristics of the best-performing 99mTc-labeled antagonist, MB88, were compared with those of 111In-DPC11870, the images acquired after injection of the 99mTc-compound were again superior. In this case, despite similar abscess and kidney uptake, better images were obtained because of lower radioactivity concentrations in the bone marrow of animals injected with the monovalent 99mTc-MB88.
In a previous study, we cautiously analyzed the pharmacodynamics of the 111In-labeled LTB4 antagonist DPC11870 (14). From the results of that study, we concluded that accumulation at the site of infection was based on initial targeting of receptor-positive cells in the bone marrow, followed by migration of radioactivity from the bone marrow to the site of infection. In addition we determined that, similarly to radiolabeled leukocytes, 111In-DPC11870 showed physiologic uptake in spleen, liver, and bone marrow (15,16). The radioactivity concentration in the bone marrow after injection of both newly synthesized HYNIC-conjugated LTB4 antagonists was lower than that after injection of 111In-DPC11870. The lower bone marrow concentration was an important factor determining the quality of abscess visualization after injection of the 99mTc-labeled compounds. The rapid blood clearance of the monomeric LTB4 antagonist may explain the low bone marrow uptake. However, because blood clearance was similar for both bivalent compounds (99mTc-labeled and 111In-labeled), a difference in blood clearance could not explain the lower bone marrow uptake of the bivalent 99mTc-MB81. Another explanation for the lower bone marrow uptake of both HYNIC compounds might be the introduction of the HYNIC chelator, which could affect the kinetics of the compound. The low radioactivity concentration of both HYNIC compounds in the spleen supports this explanation.
The present results indicate that the chelator (HYNIC or DTPA) obviously affected the in vivo behavior and biodistribution of the radiolabeled LTB4 antagonist. Several studies have focused on the in vivo effect of a chelator (17,18). These studies demonstrated that a chelator might affect the in vivo characteristics of a compound. In a comparative study in which the biodistribution of 99mTc-labeled and 111In-labeled octreotide was evaluated, kidney retention of the 99mTc-labeled compound was found to be less than that of the 111In-labeled compound (19). The characteristics of MB81 were less favorable than those of MB88 because of the higher kidney retention and slower t1/2α of the compound. Because the t1/2α of 111In-DPC11870 (30 min) is comparable with that of MB81, the longer t1/2α of MB81 probably is caused by its bivalency. The higher radioactivity concentration of 99mTc-MB81 in the kidney is difficult to explain. Because kidney retention of both 111In-DPC11870 (bivalent, DTPA) and 99mTc-MB88 (monovalent, HYNIC) was lower than that of 99mTc-MB81, it is not clear whether the kidney retention was caused by the bivalency or the HYNIC conjugation of the LTB4 antagonist. The high-quality images obtained with 99mTc-MB88 resulted from use of the high-resolution radiolabel 99mTc, rapid clearance of MB88 from nontarget tissue, and preserved accumulation of this LTB4 antagonist at the site of infection.
The ideal imaging agent for detection of infectious and inflammatory foci needs to satisfy several requirements. First, the agent should accumulate specifically and rapidly in the foci and clear quickly from the nontarget tissues to allow visualization of the lesion shortly after injection. Second, the labeling procedure should be relatively simple, and labeling should preferably be with 99mTc instead of other radionuclides. Finally, the agent should not provoke side effects when administered to patients (20). The results of the present study indicate that 99mTc-MB88 meets the requirements for a potential infection-imaging agent. When these results are compared with those for other currently studied compounds, 99mTc-MB88 remains a promising candidate. Other agents currently under investigation include other radiolabeled cytokines and chemotactic peptides (interleukin-2, interleukin-8, and platelet factor 4) (4,21,22). These compounds, however, exhibit biologic activity and have the disadvantage of provoking side effects in patients. Furthermore, production of these recombinant peptides is complicated and expensive. Antimicrobial agents (99mTc-ciprofloxacin [Infecton; Draximage Inc.] and human neutrophil peptides) (23,24) have an advantage over 99mTc-MB88 in that they may be targeted specifically at bacteria. Therefore, these agents can potentially differentiate between infectious lesions and inflammatory foci. Infecton has been evaluated in several clinical studies and yields a high overall sensitivity and specificity (85.4% and 81.7%, respectively) (23). Another advantage of Infecton over MB88 is lack of bone marrow uptake, possibly suiting Infecton for detection of infection after orthopedic interventions (e.g., osteomyelitis).
In summary, clinical evaluation of 99mTc-MB88 is warranted to assess its value as an infection-imaging agent and to allow a fair comparison with the other currently studied agents.
CONCLUSION
In the present study, we compared the imaging characteristics of 3 radiolabeled LTB4 antagonists for visualizing infection and inflammation. Visualization of infectious foci on images obtained with the monovalent LTB4 antagonist 99mTc-MB88 was superior to that on images obtained with the other 2 compounds. We consider 99mTc-MB88 to be a potent and promising agent for visualizing infection and inflammation in patients.
Acknowledgments
We thank Gerry Grutters (Central Animal Laboratory, Radboud University of Nijmegen) for technical assistance.
Footnotes
Received Apr. 21, 2005; revision accepted May 26, 2005.
For correspondence or reprints contact: Julliëtte E.M. van Eerd, MSc, Department of Nuclear Medicine, Radboud University Nijmegen Medical Centre, P.O. Box 9101, 6500 HB Nijmegen, The Netherlands.
E-mail: J.vaneerd{at}nucmed.umcn.nl
