Abstract
The aim of this study was to show that the multipinhole technique (high-resolution single-photon emission computed tomography [HiSPECT]) is suitable for dynamic imaging of both biodistribution and temporal uptake behavior of radiolabeled cationic liposomes in Balb/c-mice. HiSPECT uses multipinhole collimators adapted to a clinical SPECT scanner, together with a dedicated iterative reconstruction program. This technique provides both high spatial resolution and an improvement in sensitivity. Six male Balb/c mice received 9.8 ± 4.0 MBq of the In 111-labeled liposomes. The measurements started directly after the injection and tomographic data were acquired in steps of 5 minutes. The regional evaluation displayed a high initial uptake of liposomes in the lungs (45.4%), which decreased to 25.1% after 30 minutes and to below 2% after 48 hours. In contrast, liver uptake increased in the first 30 minutes from 13.1 to 21.2% and remained relatively stable at 24.4% (24 hours) and 18.8% (48 hours). The data are interpreted as a slow shift of liposomes from the lungs into the liver and later to other organs such as the spleen and bladder. This study shows that the HiSPECT technique is capable of dynamically visualizing the uptake behavior of radioactively labeled liposomes in vivo with high temporal and spatial resolution.
IN VIVO MOLECULAR IMAGING has revolutionized the way in which researchers study biologic processes. Such research calls for imaging systems with both high sensitivity and high-resolution. Singlephoton emission computed tomography (SPECT) and positron emission tomography (PET) are molecular imaging techniques that use radiolabeled molecules to investigate tumor growth, therapy efficacy, tissue pathology, and other physiologic parameters in vivo. PET typically offers greater sensitivity by two or more orders of magnitude beyond SPECT. In practice, the lack of sensitivity is often the limiting factor in reconstructed SPECT image quality and applications. In an attempt to address this sensitivity problem, in the last years, several systems employing multiple pinhole (MPH) collimations have been developed.1–6
SPECT is a major modality in the growing field of molecular imaging and is used in tracer development and preclinical in vivo testing of radiopharmaceuticals in small-animal studies. Hence, small objects require imaging systems with both high spatial resolution and high sensitivity. The first generation of animal imaging devices was equipped with parallel-hole ultra-high-resolution collimators, providing a moderate spatial resolution (2–3 mm in terms of full width at half maximum [FWHM]), but suffers from very low sensitivity (< 50 cps/MBq). Today, in small-animal SPECT imaging, high resolution is typically achieved with magnification geometries of single-pinhole collimation, resulting in limited sensitivity. Previously, we developed an imaging technique in which we improved the sensitivity of conventional single-pinhole imaging systems by providing the apertures with an array up to 20 pinholes.3,7,8–10
Our approach, high-resolution single-photon emission computed tomography (HiSPECT), consists of an imaging module made up of MPH SPECT add-ons that are compatible with all commercial gamma cameras. With HiSPECT, we have been able to increase sensitivity (> 1,000 cps/MBq) while maintaining the reconstructed resolutions of standard single-pinhole systems (<1 mm FWHM). This increase has been made possible by allowing the MPH projections to overlap on the detector, which, in turn, results in a more efficient use of the crystal surface. The sensitivity of HiSPECT is in the same order of magnitude as that of the present animal PET scanner, whereas the spatial resolution beats the theoretical limit (0.86 mm FWHM) of PET.
The science of liposomes as an extremely flexible delivery system for drugs and vaccines now has evolved through a variety of phases and culminated after interaction studies in vivo in the licensing of a great number of injectable liposome-based therapeutics.11–14 Primarily, liposomes approved by the US Food and Drug Administration are used to deliver cancer therapy drugs and antibodies. Homing molecules can be attached to liposome bilayers to make these structures target site specific. Parameters such as size, lamellarity, bilayer rigidity, bilayer surface modifications, and charge determine the fate of liposomes on the shelf and in vivo. 15 Cationic liposomes (or cationic lipoplexes) are structures that are made of positively charged lipids and are increasingly being researched for use in gene therapy owing to their favorable interactions with negatively charged deoxyribonucleic acid (DNA) and cell membranes. Here we present the first in vivo HiSPECT study to trace time-resolved biodistribution of In 111-labeled cationic liposomes in six male Balb/c mice. In this work, we measured baseline uptake values for an arthritic mouse model to determine the temporal behavior of liposomes in inflamed bony structures.
Material and Methods
SPECT Camera and Collimators
The HiSPECT camera is based on a commercial two-headed human SPECT scanner (Prism 2000S, Philips, Eindhoven, the Netherlands) that employs a NaI(Tl) crystal of 8 mm thickness and has a rectangular field of view (FOV) of 390 × 510 mm 2 . The MPH collimators are based on an original pinhole frame for the Prism 2000S and extended by a 12 mm pyramidal lead shielding and a MPH aperture, which is made of a 10 mm thick tungsten alloy (HPM1850). The height of each collimator is 150 mm. We used an aperture for the mice studies with 10 holes each measuring 1.5 mm in diameter, leading to a reconstructed resolution of 1.2 mm FWHM.
Labeling of the Liposomes
For our experiments, we used a representative cationic liposome type of the most recent development with the internal code name XRAU44. This type carries diethylenetriamine pentaacetic acid (DTPA) as chelator for In 111. From Covidien Deutschland GmbH (Neustadt/Donau, Germany), we received 200 MBq In 111-Cl3 dissolved in 300 μL NaCl; 175 μL of In 111-Cl3 was mixed with the same amount of citrate buffer (pH 4.5), and 320 μL of this mixture was incubated for 30 minutes with 320 μL of the liposome solution. The total available amount of labeled liposomes for injection was 634 μL with an activity of ~80 MBq. Quality control was executed using thin-layer chromatography, which resulted in a > 98% share of labeled liposomes in the injectable solution.
Measurement Protocol
After anesthesia (60 μL of ketamine-xylazine, 5:1, intraperitoneally), six male Balb/c mice (6 weeks; 21.4 ± 1.1 g) were placed into a semicylindrical plastic holder and received 9.8 ± 4.0 MBq of the labeled liposome type XRAU44 via the tail vein. The HiSPECT measurement at an object-collimator distance of 37 mm started directly after the injection and tomographic data were acquired in steps of 5 minutes (six frames). To include both In 111 energy lines (171 and 243 keV), the window was set to 209 keV ± 24%. Two further measurements, each 10 minutes in length, were executed after 24 and 48 hours p.i. 1 To facilitate orientation, additional images of bone metabolism were simultaneously acquired as previously described 11 with Tc 99m diphosphono-1,2-propandicarbonacid as a radiotracer. Here the energy window was set at 140 keV ± 7.5% (Figure 1).
Evaluation
Image reconstructions were carried out using a dedicated MPH algorithm (HiSPECT, SciVis GmbH, Göttingen). This iterative reconstruction is based on the maximum likelihood approach that solves the set of linear equations relating the unknown activity distribution to the measured projections. The reconstructed volume had a voxel size of 0.6 × 0.6 × 0.6 mm 3 with a matrix size of 100 × 100 × 240. The regional evaluation of the whole body, lungs, and liver of the mice was executed using the imaging tool AMIDE, version 0.9.1 (A. Loening, <http://amide.sourceforge.net>). To receive a temporal description of the liposome biodistribution, we defined a 40% isocontour region of interest (ROI) around the lungs, liver, and spleen and determined the values for each time step. Furthermore, a cylindrical ROI of 40 mm in diameter over the complete FOV was drawn. This value was considered to be 100% of the injected liposomes and was defined as whole body value (WBV). Owing to the long half-time of In 111 (67.92 hours), the WBVs had been half-time corrected only for 24- (78.3%) and 48-hour (61.3%) data.
Comparison of the statistics of the liposome uptake behavior in the lungs and liver in Balb/c mice. Data are presentede as mean ± SD of five animals. 1 p.i. = post-***injection.
Results and Discussion
One mouse died after 30 minutes; the other five mice survived all three examinations.
The regional evaluation displayed an average uptake of liposomes (Figure 2) in the lungs of 45.4% WBV (± 4.7% SD), which decreased to 25.1% WBV (± 10.1%) after 27.5 minutes. The 24- and 48-hour values of the lungs do not differ from each other (2.4 ± 1%, 1.7 ± 10.7% WBV). The liver uptake increased in the first 30 minutes from 13.1 ± 7.4% to 21.2 ± 7.6% WBV. After 24 and 48 hours, the uptake was increased to 24.4 ± 10.6% and 18.9 ± 1.6% WBV, respectively. The average sum of the liposomal concentration in the liver and lungs only decreased slowly from 56% (t = 0) to 49% after 30 minutes. The uptake values in the spleen at the first 30 minutes remained below 1% (0.56 ± 0.49%) and increased slowly to 1.55 ± 0.77% WBV after 24 hours and 2.12 ± 0.27% WBV after 48 hours (data not shown).
After 24 and 48 hours, 69.3 ± 21.6% and 49.7 ± 8.3%, respectively, of the injected activity was measured in the cylindrical whole-body ROI, and good retention of the In 111-labeled XRA144 in the animals was demonstrated. Comparison between the half-time-corrected WBVs for 24 and 48 hours with the measured values resulted in liposome excretion of about 20%/24 hours.
These data can be interpreted as a slow shift of liposomes from the lungs to the liver and other organs such as the spleen and bladder. In the first 30 minutes, the decrease in the liposome concentration in the lungs was 45%; the uptake growth in the liver was calculated to be 62%. After 24 and 48 hours, a diffuse and negligible accumulation in the bones in the range of < 0.1% WBV was observed.
Direct application methods, such as intratumoral or intraperitonal application, 13 , 16 appear to be an impressive alternative to increase local concentration of the liposomes and the encapsulated therapeutic agents. Thus, the therapeutic effects of β-emitters (eg, 186 Re, 188 Re) could be drastically improved. In this study, the XSPECT camera was equipped with a parallel-hole collimator to achieve a higher temporal resolution and the liposomal uptake was determined dynamically in steps of 1 minute in planar views. Recently, the slow uptake behavior of similar labeled liposomes justified the longer SPECT frames of 1 hour after 4, 16, 24, 48, 72, and 96 hours p.i. 16 In our study, we focused on the early and three-dimensional clearance of the liver and lungs and the determination of the liposome half-life in the laboratory animal.
Sagittal views of high-resolution single-photon emission computed tomographic data showing the uptake behavior of the In 111-labeled liposomes in one male Balb/c mouse after 7.5, 12.5, and 27.5 minutes and 24 and 48 hours postinjection (p.i.). Whereas in the first 30 minutes, most of the liposomes were trapped in the lungs, after 24 and 48 hours, the highest uptake could be observed in the liver, spleen, and bladder. The last image on the right represents the distribution of Tc 99m-labeled DPD, a clinical bone metabolism marker, which was injected in the same mouse 2 hours before the 48-hour measurement of the liposomes. The gray values are normalized to the maximum value of 7.5 minutes p.i.
For tiny anatomic regions such as inflamed mice ankles and other bony structures, the systemic route seems to be more practical. The in vivo stability of the liposome-DTPA-In 111 complex was not tested; the ex vivo stability in the injection solution was > 85% after 48 hours. The high percentage of the remaining and half-time-corrected activity in the animals after 24 and 48 hours, as well as the low bladder activity in the first 30 minutes, is a good but not sufficient indicator 17 of the in vivo stability of the complex.
Our results provided baseline data for comparing where the specific accumulation of liposomes in inflamed ankles occurred for arthritis mouse models.
The results show that the HiSPECT technique is capable of dynamically visualizing the uptake behavior of radioactive labeled liposomes in vivo in a high temporal and spatial resolution.
Compared with previous procedures, HiSPECT offers a unique way to visualize the temporal biodistribution of liposomes. Additionally, it provides three-dimensional information about organ interactions in vivo; in our study, the loss of activity concentration in the lungs favors uptake in the liver and spleen.
In our next study, the quantitative considerations of the ROI data will include correction factors for self-absorption (~10% for mice using In 111), partial-volume effect (object size dependent), and serum stability of the liposome-DTPA-In 111 complex.
Furthermore, the number of animals and the experimental effort can be reduced since it is no longer needed to sacrifice animals at each time step to measure each organ separately. Thus, follow-up measurements become possible, avoiding the interindividual differences of the used laboratory animals.