Nuclear Medicine and Biology 39 (2012) 1026–1033

Contents lists available at SciVerse ScienceDirect

Nuclear Medicine and Biology

j ourna l homepage: www.e lsev ie r.com/ locate /nucmedb io

Preclinical characterization of 18 F-MAA, a novel PET surrogate of 99mTc-MAA

Shih-Yen Wu a,1, Jia-Wei Kuo a,b,1, Tien-Kuei Chang c, Ren-Shen Liu d, Rheun-Chuan Lee c, Shyh-Jen Wang d,Wuu-Jyh Lin b, Hsin-Ell Wang a,⁎a Department of Biomedical imaging and Radiological Sciences, National Yang-Ming University, Taipei, 11221, Taiwanb Institute of Nuclear Energy Research, Atomic Energy Council, Taoyuan, 32546, Taiwanc Department of Radiology, Taipei Veterans General Hospital, Taipei, 11217, Taiwand Department of Nuclear Medicine, Veterans General Hospital, Taipei, 11217, Taiwan

⁎ Corresponding author. Tel.: +886 2 28267215; fax:E-mail address: hewang@ym.edu.tw (H-E. Wang).

1 S.-Y. Wu and J.-W. Kuo contributed equally to this s

0969-8051/$ – see front matter © 2012 Elsevier Inc. Aldoi:10.1016/j.nucmedbio.2012.04.008

a b s t r a c t

a r t i c l e i n f o

Article history:

Received 28 February 2012Received in revised form 13 April 2012Accepted 23 April 2012

Keywords:99mTc-macroaggregated albuminPositron emission tomographyLung perfusion imaging90Y-labeled SIR-SpheresSelective internal radiation treatment

Introduction: 99mTc-labeled macroaggregated albumin (99mTc-MAA) scintigraphy scan is routinely performedfor lung perfusion imaging and for the assessment of in vivo distribution of 90Y-labeled SIR-Spheres prior toselective internal radiation treatment for hepatocellular carcinoma. Positron emission tomography (PET)imaging is superior to gamma scintigraphy in terms of sensitivity, spatial resolution and accuracy ofquantification. This study reported that 18 F-labeled macroaggregated albumin (18 F-MAA) is an ideal PETimaging surrogate for 99mTc-MAA.Methods: 18 F-MAA was prepared from the commercial MAA kit via a one-step conjugation with N-succinimidyl 4-18 F-fluorobenzoate (18 F-SFB). The biodistribution study and microPET/microSPECT imagingwere conducted in normal SD rats after intravenous injection of 18 F-MAA/99mTc-MAA. A comparison study ofthese two radiotracers was performed after co-injection via the intrahepatic arterial in a N1S1 hepatoma-

bearing SD rat model.Results: The optimal condition for 18 F-MAA preparation is coupling MAA (0.5 mg) with 18 F-SFB at 45 °C for5 min in a phosphate buffer of pH 8.5. 18 F-MAA was prepared in 60 min with high radiochemical yield (30%–35%) and high radiochemical purity (N95%). The in vivo distribution of 18 F-MAA after intravenous injectionmeets the specifications of MAA depicted in European Pharmacopeia. Our study demonstrated excellentcorrelation between 18 F-MAA and 99mTc-MAA in the regional distribution of tumor, liver and lungs(R2=0.965, 0.886 and 0.991, respectively), and also in the tumor-to-liver and tumor-to-lungs ratio(R2=0.965 and 0.987, respectively) in a N1S1 hepatoma-bearing SD rat model. The organ uptakes derivedfrom animal PET/CT and SPECT/CT imaging after administration of these two tracers were in accordance withthose obtained in the distribution studies.Conclusions: Starting from commercial MAA kit, an efficient preparation of 18 F-MAA was successfullyestablished. Highly correlated, almost parallel, regional distribution of 18 F-MAA and 99mTc-MAA in bothnormal rats and hepatoma-bearing rats was observed. The findings, taken together, demonstrate that 18 F-MAA is an ideal surrogate for 99mTc-MAA for clinical PET applications.

© 2012 Elsevier Inc. All rights reserved.

1. Introduction

Pulmonary embolism is a potentially fatal illness with an annualincidence rate of 0.2% inwestern countries [1,2]. Lungperfusion imagingwith 99mTc-labeled macroaggregated albumin (99mTc-MAA) is com-monly used for the detection of pulmonary embolism [3,4]. 99mTc-MAAhas a size rangeof 10–100 μmand is trapped in the capillary bedof lungs(~8 μm in diameter) during the first pass of the pulmonary circulationfollowing intravenous administration. Regional pulmonary perfusioncould be assessed by imaging the distribution of 99mTc-MAA in lungs.

+886 2 28201095.

tudy.

l rights reserved.

Selective internal radiation therapy (SIRT) is an emergingtreatment for hepatocellular carcinoma (HCC), which relies onselective accumulation of 90Y-labeled SIR-Spheres in tumor followingintrahepatic arterial (i.a.) injection [5]. Specific targeting of 90Y-labeled SIR-Spheres to tumor lesion is due to the unique pattern ofhepatic blood flow. The vast majority of the tumor blood supplycomes from the hepatic artery, whereas hepatic blood flow is largelyderived from the portal vein. The best prognostic indicator for tumorprogression and patient response would be the tumor-to-normaltissue (T/N) ratio of 90Y-labeled SIR-Spheres [6]. Arteriovenousshunting of micro-particulates from liver to lungs is a dose-limitingfactor for HCC treatment with particle emission radionuclide-labeledmicrospheres [7]. Radioembolization to lungs can cause acutedamage, resulting in radiation pneumonitis. Previous studies with

1027S-Y. Wu et al. / Nuclear Medicine and Biology 39 (2012) 1026–1033

90Y-labeled SIR-Spheres have reported that patients with shuntingfraction of N20% or the potential lung absorbed radiation dose of morethan 30 Gy were at high risk of developing radiation pneumonitis andshould be excluded from 90Y-labeled SIR-Spheres treatment [5,7].99mTc-MAA has a particle size comparable to that of 90Y-labeled SIR-Spheres. Gamma imagingwith 99mTc-MAA has been adopted to assessthe distribution of 90Y-labeled SIR-Spheres within HCC and normalliver, and also to predict the extrahepatic shunting prior to SIRT [5–7].

It is well known that positron emission tomography (PET) is apromising imaging modality with excellent sensitivity, favorablespatial resolution and quantification capability compared withgamma imaging and single photon emission computed tomography(SPECT) [8–10]. Thus, PET imagingwith positron emitter-labeledMAAmay provide benefits for the diagnosis of pulmonary perfusion, theprediction of in vivo distribution, and the more precise quantitativeassessment of radiation dose to hepatoma before treatment with 90Y-labeled SIR-Spheres. F-18 is the most widely used positron emitterwith a favorable half-life of 110 min and low positron energy of0.64 MeV. To the best of our knowledge, 18 F-labeledMAAhas not beenreported yet. N-succinimidyl 4-18 F-fluorobenzoate (18 F-SFB) was acommonly used agent for labeling F-18 radionuclide to a wide varietyof biomolecules, including Annexin-V, octreotide and RGD peptide[11–13]. In this study, various conditions have been applied to coupleMAAwith 18 F-SFB, and 18 F-MAAcanbepreparedwith a high couplingefficiency. The distribution pattern and pharmacokinetics of 18 F-MAAand 99mTc-MAAwere studied and compared to tell whether 18 F-MAAis a potential PET surrogate of 99mTc-MAA for clinical applications.

2. Materials and methods

2.1. General

Unless otherwise stated, all reagents and solvents were obtainedfrom commercial sources and used without further purification. Sep-Pak QMA and Sep-Pak C18 cartridges were purchased from Waters(Milford, MA). Lichrolut SCX cartridge was obtained from Waters(Darmstadt, Germany). Radioactivity was measured with a dosecalibrator (CRC-15R, Capintec, Bioscan, Washington, DC, USA) or agamma counter (1470 Wizard automatic gamma counter, Perkin-Elmer, Waltham, MA, USA). Radioisotope distribution on the thin-layer chromatography (TLC) plate was analyzed using a Bioscansystem AR-2000 imaging scanner (Bioscan, Washington, USA).Pulmolite MAA kits were purchased from CIS-US Inc., Bedford, MA,USA. 99mTc-MAAwas prepared by using commercial CIS-US PulmoliteMAA kit, following the manufacturer's instruction. The radiolabelingefficiency and the radiochemical purity of 99mTc-MAAwere bothmorethan 95%. Nylon syringe filters (0.22 μm) were purchased from PallCorporation, Ann Arbor, MI, USA.

2.2. Synthesis of 18 F-MAA

2.2.1. Preparation of MAA suspensionMAA particles from a commercial Pulmolite MAA kit were

suspended in 2 mL of sterile saline and centrifuged at 3000-g for2 min. The supernatant containing SnCl2 and human serum albumin(HSA) was discarded and MAA particles were re-dissolved in 2 mL ofsaline. The washing procedure was repeated twice and MAA particleswere finally suspended in phosphate buffer (PB, 0.1 M, pH 7.5–9.5).

2.3. Synthesis of 18 F-SFB

No-carrier-added 18 F-SFB was synthesized according to literaturewithminormodifications [14]. H18F was produced by the 18O(p,n)18 Fnuclear reaction by irradiation of 95% 18O-enriched water with a 17-MeV proton beam at 18 μA for 30 min. At the end of irradiation, H18Fwas transferred through a QMA Sep-Pak cartridge, which has been

previously conditioned with 0.5 M K2CO3 (10 mL) and water (20 mL).The 18O-enriched water was then recovered. The 18 F- trapped on thecartridgewas desorbed by elution with 0.8 mL of 40 mM tetra-n-butylammonium (TBA) bicarbonate solution in acetonitrile/water (4/1, vol/vol) under helium gas and was sent to the reaction vessel. The TBA18Fsolution in the reaction vessel was heated at 100 °C under reducedpressure for 10 min. The residue was dried again with 1.6 mLanhydrous acetonitrile at 100 °C (azeotropic evaporation). To thisdry residue was added a solution of tert-butyl 4-N,N,N-trimethylam-moniumbenzoate triflate (2 mg) in anhydrous acetonitrile (0.8 mL),and the mixture was heated at 100 °C for 10 min to give tert-butyl[18 F]fluorobenzoate. The tert-butyl ester was subsequently hydro-lyzed using 20 μL of tetramethylammonium hydroxide (TMAOH, 25%in MeOH) at 100 °C for 10 min, and then the mixture was azeotropi-cally dried using acetonitrile (1.6 mL). Subsequently, a solution ofO-(N-Succinimidyl)-1,1,3,3-tetramethyluronium tetrafluoroborate(TSTU) (22 mg) in acetonitrile (0.8 mL) was added and the solutionheated at 100 °C for 2 min. After cooling, 5% aqueous acetic acid(10 mL) was added. The reaction mixture was passed through a C18Sep-Pak cartridge and a Lichrolut SCX cartridge in series. Thecartridges were washed with 10% aqueous acetonitrile (10 mL) andthen the product 18 F-SFB was eluted with acetonitrile (1.5 mL). Thesolventwas then removedunder a streamof nitrogen at 80 °C and 18 F-SFB was re-dissolved in 50 μL of dimethyl sulfoxide. 18 F-SFB wasprepared in 45 minwith high radiochemical yield (55%–60%) and highradiochemical purity (N95%).

2.4. Synthesis of 18 F-MAA

By varying the amount of MAA, the reaction temperature, bufferpH and the reaction time, the optimal condition for 18 F-MAApreparation was assessed. 18 F-SFB was mixed with MAA suspension(0.25–1.0 mg) in 0.25 mL of PB (pH 7.5–9.5). The reaction mixturewas incubated at 25–55 °C for 1–20 min. The resulting solution wascentrifuged at 3000-g for 2 min and the supernatant containingunconjugated 18 F-SFB was discarded. The 18 F-MAA pellet was re-suspended in 2 mL of sterile saline, the washing procedure wasrepeated twice to afford 18 F-MAA final product in sterile saline.

2.5. Quality assurance

Radiochemical purity was determined by filtering 18 F-MAA finalproduct through a 0.22 μm nylon syringe filter, which selectivelyretained 18 F-MAA. The radioactivity of the filtrate and filter wascounted in a dose calibrator. The radiochemical purity was calculatedto be the filter radioactivity as a percentage of the total radioactivity.The radiochemical purity was also determined by thin layer chroma-tography (radio-TLC) method. The radio-TLC was performed on SilicaRP18 F256S plate (MERCK, Darmstadt, Germany)withMeCN/H2O (1/1,v/v) as the mobile phase. The retention factor (Rf) was 0.0 for18 F-MAA, 0.6 for 18 F-SFB, 0.8 for 18 F-FBA in radio-TLC determination.

2.6. Particle size distribution

The particle size distribution of 99mTc-MAA and 18 F-MAA wasdetermined using a Coulter LS 230 particle size analyzer (BeckmanCoulter, Fullerton, CA, USA). The results were processed using LS230software, based on the general Fraunhofer optical model, with wateras solvent, to plot particle size distribution curves and calculate meanparticle size and standard deviation.

2.7. Serum stability

Aliquots of 18 F-MAA (0.74 MBq, 50 μL) were added to 0.5 mL offetal bovine serum (FBS). Samples were incubated at 37 °C through astudy period of 4 h. In vitro stability of 18 F-MAA was determined

1028 S-Y. Wu et al. / Nuclear Medicine and Biology 39 (2012) 1026–1033

using a filtration method, as previously described, at different timepoints (15, 30, 60, 120, and 240 min) post incubation (n=3).

2.8. Tumor cell line and N1-S1 hepatoma-bearing rat model

Animal experiments were approved by the Institutional AnimalCare and Use Committee of National Yang-Ming University. N1-S1hepatoma cells were routinely cultured in Dulbecco's Modified EagleMedium (containing 5% FBS, 1% L-glutamine and 20% horse serum,100 U/mL penicillin and 100 μg/mL streptomycin) at 37 °C in ahumidified atmosphere of 5% CO2.

4×106 N1-S1 hepatoma cells (in 0.1 mL of phosphate buffer saline)were implanted into the hepatic lobe of male Sprague–Dawley rats (SDrats, 200–250 g), using aprotocol as described inourprevious report [15].

2.9. Distribution pattern in normal rats after intravenous injection of18 F-MAA and 99mTc-MAA

To evaluate the potential of 18 F-MAA as a lung perfusion PETimaging agent, the distribution pattern of 18 F-MAA and 99mTc-MAA innormal rats was conducted in parallel and compared in this study.

2.10. Biodistribution study

Each normal SD rat (200–250 g) was anesthetized with isoflurane(Abbott Laboratories, Queensborough, Kent, England) using a vaporizersystem (A.M. Bickford, Wales Center, NY, USA). At 15 and 120 min afterintravenous injection with 18 F-MAA (1.85 MBq) or 99mTc-MAA(1.85 MBq), three rats at each time point were sacrificed with CO2.Blood samples were obtained by cardiac puncture. The organs wereexcised, weighed and assayed for radioactivity with a gamma counter.The tissues' uptakewasexpressed in counts perminute (cpm) correctedwith decay and normalized as the percentage injected dose per organ(%ID) or the percentage injected dose per gram of tissue (%ID/g).

2.11. Imaging study

Multimodality dynamic imaging of SD rats was conducted afterintravenous injection of 18 F-MAA (~3.7 MBq) or 99mTc-MAA(~18.5 MBq) using a microPET scanner (Concorde R4, Siemens Medical

Fig. 1. The optimal condition for 18 F-MAA preparation is coupling 18 F-SFB with 0.5 m

Solutions, Malvern, Pa USA) or a microPET/SPECT/CT scanner (FLEXTriumph Regular FLEX X-OCT, SPECT CZT 3 Head System, LabPET4 Tri-modality system, GEHealthcare, Northridge, CA, USA). The sensitivity ofmicroPET R4 scanner and FLEX Triumph PET/SPECT/CT scanner was2.1% and 1.1%, respectively. The spatial resolution of microPET R4scanner and FLEX Triumph PET/SPECT/CT scanner was 1.6 mm and0.91 mm, respectively. Animals were anesthetized by inhalation of 2%isoflurane in 2 L/min oxygen in the prone position. Dynamic scan for 2 hwith lungs on the center of field of viewwas acquired immediately afterinjection of radiotracers. MicroPET images were reconstructed usingfiltered backprojection with a 128×128 pixel image matrix, 16 subsets,4 iterations and use of a Gaussian filter. The microSPECT images wereacquired using a N5F65Amultipinhole collimator. The radius of rotation(ROR) for rat multipinhole was 60 mm with an FOV of 37.23 mm2. Atotal of 32projectionswere acquired ina60×60acquisitionmatrixwithaminimumof 8000 counts per projection. DynamicmicroSPECT imageswere reconstructed using an ordered-subset expectationmaximizationalgorithm (5 iterations and 8 subsets).

2.12. Distribution pattern in N1-S1 hepatoma-bearing rats afterintrahepatic arterial (i.a.) co-injection of 18 F-MAA and 99mTc-MAA

To evaluate 18 F-MAA as a potential PET imaging surrogate of99mTc-MAA, the biodistribution and pharmacokinetics of 18 F-MAAand 99mTc-MAA in N1-S1 hepatoma-bearing rat after intrahepaticarterial (i.a.) co-injection of these two radiotracers were investigated(n=9). The protocol for the i.a. injection of radiotracers was detailedin the previous study [15].

2.13. Biodistribution study

The biodistribution study of 18 F-MAA and 99mTc-MAA in rats wasconducted on the 10th day after tumor implantation. Tumor-bearingrats were anesthetized with isoflurane and injected with a combineddose of 18 F-MAA (1.85 MBq) and 99mTc-MAA (1.85 MBq) through thehepatic artery. At 15 and 120 min after injection, rats were sacrificedwith CO2. Blood samples were obtained by cardiac puncture. Theorgans were excised, weighed and assayed for radioactivity witha gamma counter using a dual-isotope counting programs for 18 Fand 99mTc, enabling comparison between 18 F-MAA and 99mTc-MAA

g MAA (A) at 45 °C (B) for 5 min (C) in a phosphate buffer of pH 8.5 (D) (n=3).

Fig. 2. Typical size distribution of 18 F-MAA (A) and 99mTc-MAA (B) injections.

1029S-Y. Wu et al. / Nuclear Medicine and Biology 39 (2012) 1026–1033

distribution. The tissue activity was expressed as percentage injecteddose per gram of tissue (%ID/g).

Table 1Radioactivity distribution in normal SD rats at 15 min and 120 min after intravenousinjection of 18 F-MAA.

%ID %ID/g

15 min 120 min 15 min 120 min

blood 2.95±0.76 4.72±0.80 0.35±0.09 0.54±0.09bone 0.04±0.01 0.04±0.01 0.35±0.31 0.47±0.07heart 0.09±0.01 0.09±0.01 0.27±0.06 0.29±0.04kidneys 0.35±0.03 1.87±0.17 0.38±0.06 2.13±0.12large intestine 0.04±0.01 0.04±0.01 0.20±0.08 0.29±0.10lungs 91.82±3.93 77.09±3.43 95.43±2.74 74.68±5.71muscle 1.12±0.39 1.16±0.22 0.01±0.01 0.01±0.01pancreas 0.04±0.01 0.05±0.01 0.05±0.02 0.32±0.01S.I. 0.04±0.01 0.04±0.01 0.30±0.07 0.39±0.10spleen 0.04±0.01 0.08±0.01 0.15±0.04 0.27±0.03stomach 0.04±0.01 0.06±0.01 0.10±0.01 0.14±0.02liver 1.04±0.18 3.00±0.24 0.16±0.20 0.36±0.25

Rats, anesthetized with 0.2% isoflurane, were injected with 1.85 MBq of 18 F-MAAthrough lateral tail vein. Results were expressed as the percentage of injection dose perorgan (%ID) and the percentage of injection dose per gram of organ (%ID/g). (n=3;mean±S.D.)

2.14. MicroPET/SPECT/CT imaging

MicroPET/SPECT/CT imaging of N1-S1 hepatoma-bearing SD ratswas performed at 1 h after intravenous co-injection of 18 F-MAA(9.3 MBq) and 99mTc-MAA (18.5 MBq). Animals were anesthetizedby inhalation of 2% isoflurane in 2 L/min oxygen in the proneposition. After in vivo microPET/SPECT/CT imaging, animals weresacrificed and hepatoma with surrounding liver was excised. Ex vivomicroPET/SPECT/CT images of hepatoma with surrounding liver wereobtained by using the microPET/SPECT/CT scanner. The SPECT imageswere acquired using a multipinhole collimator. A total of 32projections were acquired in a 60×60 acquisition matrix with aminimum of 8000 counts per projection for SPECT imaging. SPECTimages were reconstructed using an ordered-subset expectationmaximization algorithm (5 iterations and 8 subsets). The acquisitionof PET/SPECT images was followed by CT images acquisition (X-raysource: 50 kVp, 0.28 mA; 512 projections). The co-registration ofmicroPET/SPECT /CT images was performed using VIVID (VolumetricImage Visualization, Identification and Display) software (based onAmira 4.1 platform). The microPET images were reconstructed toproduce an image volume of 240×240×31 with an image resolutionof 1.0 mm. The CT images were also reconstructed using theFeldkamp cone-beam algorithm for filtered backprojection in an

image volume of 512×512×512 with an image resolution of0.08 mm. The CT data were not corrected for scatter or beamhardening. VIVID software was also used for the image fusion ofmicroPET images, microSPECT images and microCT images. Afterregistration, the image of SPECT/PET/CT had 256×256×256 voxels inan isotropic 0.24 mm voxel size.

Table 2Radioactivity distribution in normal SD rats at 15 min and 120 min after intravenousinjection of 99mTc-MAA.

%ID %ID/g

15 min 120 min 15 min 120 min

blood 3.60±0.57 2.62±0.77 0.26±0.04 0.19±0.05bone 0.01±0.01 0.01±0.01 0.03±0.06 0.11±0.11heart 0.04±0.02 0.05±0.02 0.08±0.03 0.10±0.04kidneys 0.84±0.04 3.14±1.81 0.59±0.07 2.30±1.56large intestine 0.01±0.01 0.01±0.01 0.02±0.02 0.03±0.01lungs 36.90±5.08 75.40±5.32 106.67±25.29 60.54±16.89muscle 1.79±0.19 2.51±0.36 0.02±0.01 0.03±0.01pancreas 0.01±0.01 0.02±0.01 0.04±0.01 0.05±0.02S.I. 0.01±0.01 0.02±0.01 0.04±0.01 0.07±0.03spleen 0.02±0.01 0.07±0.04 0.06±0.01 0.19±0.12stomach 0.04±0.02 0.17±0.11 0.05±0.02 0.26±0.18liver 1.95±0.19 2.49±0.81 0.16±0.05 0.23±0.11

Rats, anesthetized with 0.2% isoflurane, were injected with 1.85 MBq of 99mTc-MAAthrough lateral tail vein. Results were expressed as the percentage of injection dose perorgan (%ID) and the percentage of injection dose per gram of organ (%ID/g). (n=3;mean±S.D.)

1030 S-Y. Wu et al. / Nuclear Medicine and Biology 39 (2012) 1026–1033

2.15. Statistical methods

Results are expressed as means±SD. Statistical analysis wasperformed using the Student t test for unpaired data. A 95%confidence level was chosen to determine the significance ofdifferences between groups, with a P value of less than 0.05 indicatinga significant difference. Linear regression was used to analyze therelationship between the radioactivity distribution of both tracers.

3. Results

3.1. Synthesis of 18 F-MAA

18 F-SFB was synthesized using a three-step procedure. Nucleo-philic 18 F-fluorination of tert-butyl 4-N,N,N-trimethylammonium-

Fig. 3.MicroPET/microSPECT coronal images of normal SD rats after intravenous injection ofanesthesia and the image acquisition time was 10 min.

benzoate triflate precursor gave a intermediate product, which wasthen hydrolyzed with TMAOH and activated by using TSTU to afford18 F-SFB. 18 F-MAA was prepared by coupling MAA with 18 F-SFB in aone-step reaction. The coupling efficiency was highly dependent onthe quantity of MAA, the reaction temperature, the pH of the reactionsolution and the reaction time (Fig. 1). A small amount of MAA(0.5 mg) was sufficient to afford a high coupling efficiency with 18 F-SFB (58% yield, based on the radioactivity of 18 F-SFB). Raising theamount to 1.0 mg increased the yield slightly. Prolonging the couplingtime from 1 to 5 min significantly improved the completeness ofreaction, while longer reaction time (N5 min) reduced the reactionyield. The pH of the reaction medium and the reaction temperaturewere also found to be critical factors that affected the conjugation ofMAA with 18 F-SFB. The optimal condition with a conjugation yield of58% could be achieved when MAA and 18 F-SFB were incubated at45 °C for 5 min in a phosphate buffer of pH 8.5. Increasing the reactiontemperature (N45 °C) and the pH of the reaction medium (N8.5)diminished the coupling efficiency, probably due to the enhanceddecomposition of 18 F-SFB [16]. In all, 18 F-MAA can be prepared in60 min (from the end of bombardment) with an overall radiochemicalyield of 30%–35% and high radiochemical purity (N95%).

3.2. Characterization of 18 F-MAA

The determined particle size of 18 F-MAA showed a typical sizedistribution as that of 99mTc-MAA, ranging from 0.4 μm to 100 μm,with no particle above 100 μm and 13% of particles below 10 μm(Fig. 2A). 18 F-MAA had a similar size distribution to 99Tc-MAA (34.9±23.59 μm and 26.67±18.21 μm, respectively) (Fig. 2). The adoptedreaction condition, especially heating at 45 °C, did not change thephysical appearance of MAA. This is of vital importance for radi-olabeled MAA products, as any change in size and shape maysignificantly affect their in vivo distribution. 18 F-MAA exhibited agood in vitro stability while incubating with FBS. The radiochemicalpurity was 99.01%±0.85%, 97.65%±0.92%, 96.24%±0.11%,95.72%±1.36% and 94.10%±0.35% after 15, 30, 60, 120 and 240 min

18 F-MAA (~3.7 MBq) and 99mTc-MAA (~18.5 MBq). The animals were under isoflurane

Table 3Radioactivity distribution in N1-S1 hepatoma-bearing SD rats at 1 h after intrahepaticarterial co-injection of 18 F-MAA and 99mTc-MAA.

Uptake (%ID/g)

18 F-MAA 99mTc-MAA

blood 0.03±0.01 0.05±0.02bone 0.01±0.01 0.05±0.02heart 0.01±0.01 0.02±0.01kidneys 0.07±0.04 0.38±0.17L.I. 0.01±0.01 0.02±0.01lung 0.73±0.85 0.73±0.79muscle 0.01±0.01 0.01±0.01pancreas 0.01±0.01 0.02±0.01S.I. 0.03±0.03 0.06±0.04spleen 0.01±0.01 0.04±0.01stomach 0.06±0.06 0.11±0.16tumor 46.42±23.32 44.71±24.97whole liver 4.96±1.38 4.97±1.84

Tumor-bearing rats were anesthetized with isoflurane and injected with a combineddose of 18 F-MAA (1.85 MBq) and 99mTc-MAA (1.85 MBq) via the hepatic artery.Results were expressed as the percentage of injection dose per organ (%ID) and thepercentage of injection dose per gram of organ (%ID/g). (n=9; mean±S.D.)

1031S-Y. Wu et al. / Nuclear Medicine and Biology 39 (2012) 1026–1033

of incubation, respectively. Defluorination of 18 F-MAA to release free18 F-fluoride was not observed through the study period of 4 h.

3.3. Biodistribution of 18 F-MAA and 99mTc-MAA in normal rats

The biodistribution of 18 F-MAA and 99mTc-MAA in normal ratswasshown in Tables 1–2. More than 80% of administered 18 F-MAA (91.82±3.93 %ID) and 99mTc-MAA (86.90±5.08 %ID) were retained in lungs

Fig. 4. Correlation between tissue accumulation of 18 F-MAA and 99mTc-MAA in tumor (A), liEach point represents one rat. Radioactivity of tumor and tissues and tumor-to-tissue ratios w18 F-MAA and 99mTc-MAA. (n=9).

at 15 min after intravenous injection, and then excreted very slowlyfrom lungs (77.09±3.43 and 75.40±5.32 %ID at 2 h postinjection,respectively). There was no significant difference in the percentadministered activity in the lungs between 18 F-MAA and 99mTc-MAAfor each time point (P N0.05). The distribution of 18 F-MAA meets thespecifications of MAA in European Pharmacopeia, which states thatN80% of radioactivity is found in lungs and b5% of radioactivity isobserved in liver and spleen at 15 min after intravenous injection. Theradioactivity accumulation in liver and kidneys was slightly increasedat 2 h postinjection (3.00±0.24 and 1.87±0.17 %ID, respectively),probably due to the biodegradation of 18 F-MAA in lungs.

3.4. Dynamic imaging in normal rats

Dynamic microPET/microSPECT imaging exhibited highly inten-sive radioactivity accumulation in lungs soon after injection of 18 F-MAA and 99mTc-MAA in normal rats (Fig. 3), while other organsshowed only low level of radioactivity during the study period. Thetime–activity curve (TAC) derived from dynamic microPET imagesrevealed that, after injection of 18 F-MAA, the radioactivity in lungsreached maximum at 1–2 min, then kept nearly steady in thefollowing 2 h (Figure 1S). The results implied that 18 F-MAA wasstable in vivo during the experiment period.

3.5. Biodistribution study of 18 F-MAA and 99mTc MAA in N1-S1hepatoma-bearing rats

Hepatic tumor derives blood supply mainly from hepatic artery,while non-tumor liver receives blood via portal vein. Thus,

ver (B), and lungs (C). (D) and (E) shows the tumor-to-liver and tumor-to-lungs ratios.ere obtained from biodistribution study at 1 h after intrahepatic arterial co-injection of

Fig. 5. Typical microPET/SPECT/CT coronal images of N1-S1 hepatoma-bearing SD rats (A), and ex vivo microPET/SPECT/CT images of hepatoma and whole liver (B) at 1 h afterintrahepatic arterial co-injection of 18 F-MAA (9.3 MBq) and 99mTc-MAA (18.5 MBq). The animals were under isoflurane anesthesia and the image acquisition time was 10 min.Arrow indicated the location of tumor. The tumor was clearly visible with significant radioactivity accumulation. Because normal liver tissue also derives blood flow from the hepaticartery, a portion of radioactivity was also entrapped in normal liver tissue.

1032 S-Y. Wu et al. / Nuclear Medicine and Biology 39 (2012) 1026–1033

intrahepatic arterial administration of 18 F-MAA exhibited a significant-ly higher hepatic tumor retention than that of normal liver(46.42±23.32 %ID/g compared to 4.96±1.38 %ID/g, Table 3). Theaccumulation of 18 F-MAA and 99mTc-MAA in hepatoma showedexcellent correlation at 1 h postinjection (R2=0.965, Pb0.05, Fig. 4A).Besides tumor retention, agreement of these two radiotracers on lungsuptake, tumor-to-liver ratio and tumor-to-lungs ratio (R2=0.991, 0965and 0.987, respectively, Pb0.05, Fig. 4C–E)was observed. The uptake of18 F-MAA in liver was moderately correlated with that of 99mTc-MAA(R2=0.886, Pb0.05, Fig. 4B). The results of biodistribution studyindicated a similar distribution pattern of 18 F-MAA to that of 99mTc-MAA in the hepatoma-bearing rat model (Table 3).

3.6. MicroPET/SPECT/CT imaging of N1-S1 hepatoma-bearing rats

The microPET/SPECT/CT imaging of N1-S1 hepatoma-bearing ratsafter co-injection of 18 F-MAA and 99mTc-MAA via the hepatic arteryexhibited highly intensive regional accumulation in the liver andN1-S1 hepatoma (Fig. 5). The distribution patterns of these two MAAtracers were almost identical, consistent with those observed in thebiodistribution studies.

4. Discussion

18 F-MAA can be readily prepared by conjugating 18 F-SFB to MAA,which is obtained from a commercial kit for 99mTc labeling. This kitcontains not only MAA particulates but also albumin human and SnCl2for 99mTc labeling. Additional manipulation was needed to separateMAA from albumin human and SnCl2. In the future, the widespreadapplication of 18 F-MAA may urge the manufacturer to supply pureMAA for 18 F-SFB conjugation, and then the separation process couldbe omitted. The best conjugation condition was reacting at 45 °C inslightly basic solution. No significant aggregation or breakdown ofMAA in the product solution was observed. The entire synthesis time

of 18 F-MAA was 60 min with an overall radiochemical yield of 30%–35 % (decay corrected, based on the radioactivity of 18 F-fluoride) andhigh radiochemical purity (N95%). The preparation of 18 F-MAArequires only commercially available chemicals and no sophisticatedequipment or manipulation is needed. Compared with the one-steppreparation of 99mTc-MAA, however, the overall preparation of 18 F-MAA is less straightforward. Size and shape play key roles in thedistribution of lung perfusion agent, any change in morphology mightaffect the in vivo distribution ofMAA [17]. The high in vitro stability of18 F-MAA and similar particle size distribution compared to 99mTc-MAA may account for the similar distribution pattern of these twoMAA tracers in vivo.

The distribution study of 18 F-MAA in normal SD rats showedhighly intensive retention in lungs at 15 to 120 min post intravenousadministration, while the radioactivity accumulation in the otherorgans was at a much lower level. The uptake and retention of 18 F-MAA in lung fulfill the European Pharmacopeia criteria for MAA andwere comparable to those of 99mTc-MAA [18]. In the microPET imagesof normal rat, lungs can be clearly delineated with high lung-to-background ratio, consistent with those observed in biodistributionstudies. The time–activity curve extracted from microPET imagesrevealed persistent radioactivity retention in specific organs of normalrats and also suggested high stability of 18 F-MAA in vivo.

In this study, a dual tracer of 18 F-MAA and 99mTc-MAA was co-injected into N1-S1 hepatoma-bearing rats to compare their distri-bution profiles. Both tracers showed remarkable tumor retentioncompared with surrounding normal liver tissue. The tumor-to-liverand tumor-to-lungs ratios of 18 F-MAA were highly correlated withthose of 99mTc-MAA (r2=0.961 for tumor-to-liver ratio and 0.990 fortumor-to-lungs ratio). Negligible uptake in bone after injection of18 F-MAA indicated its good stability in vivo. MicroPET/SPECT/CTimages exhibited that the biodistribution pattern of 18 F-MAAcorrelated well with that of 99mTc-MAA. It was reported that thenon-tumorous part of the liver also obtains 20% of its blood flow from

1033S-Y. Wu et al. / Nuclear Medicine and Biology 39 (2012) 1026–1033

the hepatic artery [19]. Thus, a significant portion of radioactivitysubstance injected through hepatic artery is also delivered to normalliver tissue. After18F-MAA and 99mTc-MAA were co-injected throughhepatic artery, microPET/SPECT/CT images revealed high radioactivityretention in the tumor and also a portion of activity accumulation inthe normal liver. The results found in the microPET/SPECT/CT imagingof hepatoma-bearing animals were consistent with those observed inbiodistribution studies. This study demonstrated that 18 F-MAA PETwould be promising to predict the in vivo distribution andextrahepatic shunting fraction of 90Y-labeled SIR-Spheres as areplacement of 99mTc-MAA scan. PET imaging with 18 F-MAA wouldallow more accurate dosimetry estimation and more reliablepretreatment prediction for SIRT with 90Y-labeled SIR-Spheres.However, the expense and limited availability of the PET scannerand the cyclotron, and also the higher absorbed radiation dose due tothe positron, may prevent the widespread acceptance of 18 F-MAA forclinical use. On the other hand, the half-life of fluorine-18 is longenough that 18 F-MAA can be shipped from the site of manufacture toimaging centers lacking an on-site cyclotron.

5. Conclusion

An efficient and convenient preparation of 18 F-MAA via a one-stepconjugation of 18 F-SFB to MAA has been developed. The distributionpattern and the pharmacokinetics of 18 F-MAA correlated well withthose of 99mTc-MAA in both normal rat and hepatoma-bearing ratmodels. The findings, taken together, demonstrated that 18 F-MAA canserve as an ideal PET surrogate for 99mTc-MAA in clinical use.

Supplementary data to this article can be found online at http://dx.doi.org/10.1016/j.nucmedbio.2012.04.008.

Acknowledgments

This study was supported by the grant from National ScienceCouncil, Taiwan (NSC 98 -2314-B-010-009-MY2). We thank the staffof National PET and Cyclotron Center in Taipei Veterans GeneralHospital, who kindly provided F-18 and excellent technical assistance.This work was also supported by the Molecular and Genetic ImagingCore /National Research Program for Genomic Medicine at NationalYang-Ming University. Instrumentation Center of National TaiwanUniversity was also gratefully acknowledged.

References

[1] Bosson JL, Pouchain D, Bergmann JF. A prospective observational study of a cohortof outpatients with an acute medical event and reduced mobility: incidence ofsymptomatic thromboembolism and description of thromboprophylaxis practices.J Intern Med 2006;260:168–76.

[2] Heit JA. The epidemiology of venous thromboembolism in the community.Arterioscler Thromb Vasc Biol 2008;28:370–2.

[3] Parker JA, Coleman RE, Hilson AJW, Royal HD, Siegel BA, Sostman HD. Society ofNuclear Medicine procedure guideline for lung scintigraphy. Version 3.0, 2004.http://interactive.snm.org/docs/Lung%20scintigraphy_v3.0.pdf.

[4] Worsley DF, Alavi A. Radionuclide imaging of acute pulmonary embolism. SeminNucl Med 2003;33:259–78.

[5] Murthy R, Nunez R, Szklaruk J, Erwin W, Madoff DC, Gupta S, et al. Yttrium-90microsphere therapy for hepatic malignancy: devices, indications, technicalconsiderations, and potential complications. Radiographics 2005;25(Suppl. 1):S41–55.

[6] Ho S, Lau WY, Leung TW, Chan M, Chan KW, Lee WY, et al. Tumour-to-normaluptake ratio of 90Y microspheres in hepatic cancer assessed with 99mTcmacroaggregated albumin. Br J Radiol 1997;70:823–8.

[7] Pöpperl G, Helmberger T, Münzing W, Schmid R, Jacobs TF, Tatsch K. Selectiveinternal radiation therapy with SIR-Spheres in patients with nonresectable livertumors. Cancer Biother Radiopharm 2005;20:200–8.

[8] McBride WJ, Sharkey RM, Karacay H, D'Souza CA, Rossi EA, Laverman P, et al.A novel method of 18F radiolabeling for PET. J Nucl Med 2009;50:991–8.

[9] Urakami T, Akai S, Katayama Y, Harada N, Tsukada H, Oku N. Novel amphiphilicprobes for [18F]-radiolabeling preformed liposomes and determination ofliposomal trafficking by positron emission tomography. J Med Chem 2007;50:6454–7.

[10] Carli MF, Dorbala S, Meserve J, El Fakhri G, Sitek A, Moore SC. Clinical myocardialperfusion PET/CT. J Nucl Med 2007;48:783–93.

[11] Guenther KJ, Yoganathan S, Garofalo R, Kawabata T, Strack T, Labiris R, et al.Synthesis and in vitro evaluation of 18F- and 19F-labeled insulin: a new radiotracerfor PET-based molecular imaging studies. J Med Chem 2006;49:1466–74.

[12] Yagle KJ, Eary JF, Tait JF, Grierson JR, Link JM, Lewellen B, et al. Evaluation of 18F-annexin V as a PET imaging agent in an animal model of apoptosis. J Nucl Med2005;46:658–66.

[13] Okarvi SM. Recent progress in fluorine-18 labelled peptide radiopharmaceuticals.Eur J Nucl Med 2001;28:929–38.

[14] Tang G, Zeng W, Yu M, Kabalka G. Facile synthesis of N-succinimidyl 4-[18F]fluorobenzoate ([18F]SFB) for protein labeling. J Labelled Compd Radiopharm2008;51:68–71.

[15] Wang SJ, Lin WY, Lui WY, Chen MN, Tsai ZT, Ting G. Hepatic artery injection ofyttrium-90-lipiodol: biodistribution in rats with hepatoma. J Nucl Med 1996;37:332–5.

[16] Chen X, Park R, Shahinian AH, Tohme M, Khankaldyyan V, Bozorgzadeh MH, et al.18F-labeled RGD peptide: initial evaluation for imaging brain tumor angiogenesis.Nucl Med Biol 2004;31:179–89.

[17] Lacoeuille F, Hindré F, Denizot B, Bouchet F, Legras P, Couturier O, et al. Newstarch-based radiotracer for lung perfusion scintigraphy. Eur J Nucl Med MolImaging 2010;37:146–55.

[18] Council of Europe (COE) – European Directorate for the Quality of Medicines(EDQM). European pharmacopeia. 5th ed. Strasbourg: European Directorate forthe Quality of Medicines; 2005.

[19] Ho S, Lau WY, Leung TW, Johnson PJ. Internal radiation therapy for patients withprimary or metastatic hepatic cancer: a review. Cancer 1998;83:1894–907.