Open Collections

UBC Faculty Research and Publications

Automated synthesis of [18F]DCFPyL via direct radiofluorination and validation in preclinical prostate… Bouvet, Vincent; Wuest, Melinda; Jans, Hans-Soenke; Janzen, Nancy; Genady, Afaf R; Valliant, John F; Benard, Francois; Wuest, Frank May 4, 2016

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


52383-13550_2016_Article_195.pdf [ 1.9MB ]
JSON: 52383-1.0368604.json
JSON-LD: 52383-1.0368604-ld.json
RDF/XML (Pretty): 52383-1.0368604-rdf.xml
RDF/JSON: 52383-1.0368604-rdf.json
Turtle: 52383-1.0368604-turtle.txt
N-Triples: 52383-1.0368604-rdf-ntriples.txt
Original Record: 52383-1.0368604-source.json
Full Text

Full Text

ORIGINAL RESEARCH Open AccessAutomated synthesis of [18F]DCFPyL viadirect radiofluorination and validation inpreclinical prostate cancer modelsVincent Bouvet1, Melinda Wuest1, Hans-Soenke Jans1, Nancy Janzen2, Afaf R. Genady2, John F. Valliant2,Francois Benard3 and Frank Wuest1*AbstractBackground: Prostate-specific membrane antigen (PSMA) is frequently overexpressed and upregulated in prostatecancer. To date, various 18F- and 68Ga-labeled urea-based radiotracers for PET imaging of PSMA have beendeveloped and entered clinical trials. Here, we describe an automated synthesis of [18F]DCFPyL via directradiofluorination and validation in preclinical models of prostate cancer.Methods: [18F]DCFPyL was synthesized via direct nucleophilic heteroaromatic substitution reaction in a singlereactor TRACERlab FXFN automated synthesis unit. Radiopharmacological evaluation of [18F]DCFPyL involvedinternalization experiments, dynamic PET imaging in LNCaP (PSMA+) and PC3 (PSMA−) tumor-bearing BALB/cnude mice, biodistribution studies, and metabolic profiling. In addition, reversible two-tissue compartmentalmodel analysis was used to quantify pharmacokinetics of [18F]DCFPyL in LNCaP and PC3 tumor models.Results: Automated radiosynthesis afforded radiotracer [18F]DCFPyL in decay-corrected radiochemical yields of23 ± 5 % (n = 10) within 55 min, including HPLC purification. Dynamic PET analysis revealed rapid and highuptake of radioactivity (SUV5min 0.95) in LNCaP tumors which increased over time (SUV60min 1.1). Radioactivityuptake in LNCaP tumors was blocked in the presence of nonradioactive DCFPyL (SUV60min 0.22). The muscle asreference tissue showed rapid and continuous clearance over time (SUV60min 0.06). Fast blood clearance ofradioactivity resulted in tumor-blood ratios of 1.0 after 10 min and 8.3 after 60 min. PC3 tumors also showedcontinuous clearance of radioactivity over time (SUV60min 0.11). Kinetic analysis of PET data revealed the two-tissuecompartmental model as best fit with K1 = 0.12, k2 = 0.18, k3 = 0.08, and k4 = 0.004 min−1, confirming molecular trappingof [18F]DCFPyL in PSMA+ LNCaP cells.Conclusions: [18F]DCFPyL can be prepared for clinical applications simply and in good radiochemical yields via a directradiofluorination synthesis route in a single reactor automated synthesis unit. Radiopharmacological evaluation of[18F]DCFPyL confirmed high PSMA-mediated tumor uptake combined with superior clearance parameters.Compartmental model analysis points to a two-step molecular trapping mechanism based on PSMA bindingand subsequent internalization leading to retention of radioactivity in PSMA+ LNCaP tumors.Keywords: 18FDCFPyL, Positron emission tomography (PET), Automated radiosynthesis, Prostate-specificmembrane antigen (PSMA), Prostate cancer* Correspondence: wuest@ualberta.ca1Department of Oncology, University of Alberta, 11560 University Avenue,Edmonton, AB T6G 1Z2, CanadaFull list of author information is available at the end of the article© 2016 Bouvet et al. Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0International License (, which permits unrestricted use, distribution, andreproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link tothe Creative Commons license, and indicate if changes were made.Bouvet et al. EJNMMI Research  (2016) 6:40 DOI 10.1186/s13550-016-0195-6BackgroundProstate cancer is among the most common malignancyin men in Western countries and accounts for the fifthleading cause of cancer-related death in men [1]. Thissituation clearly underscores prostate cancer as a highlyfrequent but still unmet medical challenge which con-tinues to be the target of a significant proportion ofcurrent clinical research, including to a large extent thedevelopment of novel radiotracers for positron emissiontomography (PET) imaging of prostate cancer [2, 3].PET imaging of prostate cancer was successfully demon-strated with various 11C- and 18F-labeled choline analogs,as well as 11C-acetate for imaging cell membrane and fattyacid metabolism in prostate cancer [4, 5]. PET imaging ofprostate cancer with [18F]FDG, the most widely used ra-diopharmaceutical for cancer imaging, gave mixed results.[18F]FDG-PET is not useful in the detection of primaryorgan-confined prostate cancer and local recurrencesafter radical prostatectomy, as well as in differentiatingbetween post-operative scar and local recurrence. How-ever, [18F]FDG uptake in prostate cancer was reportedto correlate with prostate-specific antigen (PSA) levelas a biomarker to measure tumor aggressiveness [6].[18F]FDG-PET was also used to monitor therapeutic re-sponse of patients with androgen-independent disease [7, 8].The aforementioned limitations of [18F]FDG-PET inconnection with the importance of prostate cancer as asignificant unmet clinical need drive current develop-ments towards targeted molecular imaging of prostatecancer. One prominent class of agents that exemplifiesthe outcome of this effort includes radiotracers targetingthe trans-membrane protein, prostate-specific membraneantigen (PSMA) [9]. PSMA is a highly promising biomarkerfor targeted prostate cancer imaging due to its elevated ex-pression and up-regulation in poorly differentiated, meta-static, and androgen-independent carcinomas.PSMA is a particularly important molecular target inprostate cancer patients with negative bone scans whoare at high risk for metastatic disease. Moreover, PSMAis an ideal target for developing small-molecule radio-pharmaceuticals which typically show fast blood clear-ance and low background activity [9]. Over the pastdecade, several urea-based and phosphoramidate pepti-domimetic inhibitors of PSMA have been developedand labeled with positron emitters such as 11C, 18F,68Ga, 64Cu, 124I, and 86Y [9]. Among these small-moleculePSMA-imaging agents, especially 68Ga-labeled compoundshave been introduced into clinical applications for PSMAimaging in prostate cancer patients [10, 11]. Although68Ga-labeled small-molecule PSMA inhibitors exhibit favor-able clinical imaging properties, improved lesion detectioncould be further enhanced using 18F due to its lower posi-tron energy and longer half-life, or improved retention dueto the chemical nature of existing agents. This was recentlydemonstrated by the first comparative clinical study of18F-labeled PSMA inhibitor [18F]DCFPyL with [68Ga]Ga-PSMA-HBED-CC [12].Another challenge associated with 68Ga-labeled radio-pharmaceuticals for clinical imaging is the less than opti-mal availability of 68Ga generators, especially with respectto their regulatory status in North America. Given thesmall quantity of 68Ga eluted from 68Ga generators (typic-ally up to 1850 MBq), producing more than 1–3 doses perproduction batch is also challenging, which increases pro-duction costs.However, despite the inherent advantages of 18F likethe favorable physical half-life of 109.8 min and the con-venient availability as cyclotron-produced [18F]fluorideat high specific activity, the production of 18F-labeledsmall-molecule PSMA inhibitors is complicated andoften low yielding [13–20].In all cases, radiosyntheses were accomplished viachallenging prosthetic group-based multi-step synthesisroutes or aluminum-[18F]fluoride acceptor chemistry.Prosthetic group-based radiosyntheses included prepar-ation of 2,3,5,6-tetrafluorophenyl-6-[18F]fluoronicotinate([18F]FPy-TFP), 4-[18F]fluorobenzyl bromide [18F]FBB)and N-succinimidyl-4-[18F]fluorobenzoate ([18F]SFB) as18F-containing building blocks followed by bioconjuga-tion chemistry through N-acylation and S-alkylation re-actions, respectively (Fig. 1).First clinical results demonstrated that [18F]DCFPyL isa promising radiotracer for PET imaging of prostate can-cer and clear cell renal cell carcinoma. The promisingclinical results stimulated efforts for the development ofautomated radiosyntheses based on more desirable dir-ect radiofluorination reactions to simplify preparationand therefore to improve availability of [18F]DCFPyL asa radiopharmaceutical for PET imaging of PSMA.Here, we describe a novel automated radiosynthesis of[18F]DCFPyL via a direct radiofluorination procedure inan automated synthesis unit (ASU) suitable for use undergood manufacturing practice (GMP) guidelines. We alsopresent preclinical data on the radiopharmacological pro-file of [18F]DCFPyL in PSMA+ LNCaP and PSMA− PC3tumor xenografts which further confirmed the quality ofthe resulting radiotracer for targeted molecular imaging ofPSMA in prostate cancer.MethodsAll chemicals, reagents, and solvents for the synthesisand analysis were analytical grade. The amino acid pre-cursors H-Glu(OtBu)-OtBu.HCl and H-Lys(Z)-OtBu.HClwere purchased from EMD Millipore. The Bu4N+-HCO3−solution and Chromafix (30-PS-HCO3, Macherey-Nagel)cartridges were purchased from ABX. All other chemicalsand solvents were purchased from Sigma-Aldrich. Allsolvents were dried and/or distilled prior to utilization.Bouvet et al. EJNMMI Research  (2016) 6:40 Page 2 of 151H-NMR and 13C-NMR spectra were recorded on anAgilent/Varian Inova two-channel 400-MHz spectrom-eter, an Agilent/Varian Inova four-channel 500-MHzspectrometer, and an Agilent/Varian VNMRS three-channel 600-MHz spectrometer. Chemical shifts aregiven in parts per million (ppm) referenced to internalstandards (s = singlet, bs = broad singlet, d = doublet,dd = doublet of doublet, ddd = doublet of doublet ofdoublet, t = triplet, M =multiplet, m =massif ). Mass spec-tra were recorded using a Micromass ZABSpec HybridSector-TOF by positive mode electrospray ionization.High resolution mass spectra (HRMS) were carried out onan Agilent Technologies 6220 oaTOF. Crude reactionmixtures were analyzed by TLC and HPLC. Thin-layerchromatography (TLC) was monitored using HF254 silicagel. HPLC analyses were performed on a semi-preparativeLuna C18 column (100 Å, 10 μm, 250 × 10 mm) or JupiterC12 (100 Å, 10 μm, 250 × 10 mm).Both columns were connected to their correspondingguard columns (Phenomenex Nucleosil LUNA (II) RPC18 pre-column (5 μm, 50 × 10 mm) and Jupiter C12pre-column (5 μm, 50 × 10 mm)). UV detection was per-formed at 210 and 254 nm. Radioactivity detection wasachieved using a well-scintillation NaI (Tl) detector.[18F]Fluoride was produced by the 18O(p,n)18F nuclearreaction through proton irradiation of enriched (98 %)18O water (3.0 mL, ROTEM, Germany) using a TR19/9cyclotron (Advanced Cyclotron Systems, Inc., Richmond,BC, Canada).Chemical synthesis(S)-Di-tert.-butyl-(((2,5-dioxopyrrolidin-1-yl)oxy)carbonyl)-L-glutamate 2One hundred fifty milligrams (0.507 μmol) of Glu(OtBu)-OtBu·HCl 1 was dissolved in 2 mL of freshly distilledCH3CN. Seventy microliters of Et3N (0.507 μmol) and142 mg (0.554 μmol) of N,N′-disuccinimidyl carbonate wasadded to the solution. The reaction was stirred at 25 °C for12 h and concentrated under reduced pressure. After re-solubilization in 5 mL of EtOAc and successive washingwith 10 mL of 10 % citric acid and 10 mL of brine, the or-ganic layer was dried over Na2SO4 and concentrated underreduced pressure to afford 197 mg of a pale yellow powder.This unpurified powder contained 90 % of desired com-pound 2 (yield 78 %). TLC: (EtOAc/hexane, 3/1): Rf = 0.7.1H-NMR (400 MHz, CDCl3) δ: 1.46 (s, 9H), 1.50 (s, 9H),1.94–2.14 (m, 1H), 2.10–2.21 (m, 1H), 2.24–2.44 (m, 2H),2.83 (s, 4H), 4.23 (dt, 1H, J = 5.1 Hz, J = 7.7 Hz), 6.19 (d,1H, J = 7.7 Hz). 13C-NMR (100.5 MHz, CDCl3) δ: 25.47,27.59, 27.97, 28.00, 31.15, 54.78, 80.95, 83.09, 151.06,169.60, 169.76, 171.95.Fig. 1 Structure of 18F-labeled small-molecule PSMA inhibitorsBouvet et al. EJNMMI Research  (2016) 6:40 Page 3 of 15(S)-2-[3-((S)-1-tert.-Butylcarboxylate-(5-benzyloxycarbonylpentyl))ureido]-di-tert.-butylpentanedioate 4Compounds 4 and 5 and DCFPyL were previously syn-thesized [15, 16]. Herein, we describe an improvedmethodology. The yellow powder containing 90 % ofcompound 2 (197 mg, 90 % purity) was dissolvedwithout further purification in 3 mL of CH2Cl2 con-taining 100 μL of Et3N and 187 mg (501 μmol) of μs-benzyloxycarbonyl-L-lysine tert.-butyl ester hydrochloride(H-Lys(Z)-OtBu.HCl) 3. The reaction mixture was stirredfor 12 h at 25 °C, and progress of the reaction was moni-tored by TLC (CH2Cl2/MeOH 95/5). Upon completion,the reaction mixture was concentrated under reducedpressure and the residue was purified by column chroma-tography (CH2Cl2/MeOH 95/5) to afford 252 mg (92 %)of desired compound 4 as a clear gel. TLC: (CH2Cl2/MeOH 95/5): Rf = 0.2.1H-NMR (600 MHz, CDCl3) δ:1.16–1.24 (m, 1H), 1.24–1.30 (m, 1H), 1.35 (s, 9H), 1.37(s, 9H), 1.38 (s, 9H), 1.39–1.47 (m, 2H), 1.48–1.56 (m,1H), 1.61–1.69 (m, 1H), 1.69–1.77 (m, 1H), 1.93–2.01(m, 1H), 2.13–2.25 (m, 2H), 3.02–3.15 (m, 2H), 4.22–4.28(m, 1H), 4.28–4.33 (m, 1H), 4.99 (d, 1H, J = 12.5 Hz), 5.05(d, 1H, J = 12.5 Hz), 5.37 (d, 1H, J = 8 Hz), 5.44 (s, 1H),5.45 (d, 1H, J = 8.1 Hz), 7.20–7.24 (m, 1H), 7.24–7.30 (m,4H). 13C-NMR (150.9 MHz, CDCl3) δ: 21.36, 26.98, 27.00,27.05, 27.32, 28.33, 30.56, 31.61, 39.69, 51.84, 52.25, 65.46,79.44, 80.58, 81.14, 126.93, 127.03, 127.43, 135.76, 155.66,156.10, 171.33, 171.57, 171.89. m/z (ESI) C32H51N3O9([M +H+]) calcd. 622.4, found 622.4.(S)-2-[3-((S)-5-Amino-1-tert.-butoxycarbonylpentyl)ureido]pentanedioic acid di-tert.-butyl ester 5 [15, 16]A solution of compound 4 (220 mg, 0.354 μmol) in3 mL of MeOH was stirred with 10 mg of Pd/C underH2 atmosphere for 12 h. The reaction was filteredthrough celite and concentration under reduced pressureafforded 164 mg (95 %) of compound 5 as a clear gel.1H-NMR (600 MHz, CD3OD) δ: 1.45–1.47 (m,1H), 1.47(s, 9H), 1.47–1.49 (m, 1H) 1.49 (s, 9H) 1.50 (s, 9H),1.57–1.70 (m, 3H), 1.78–1.86 (m, 2H), 2.04–2.10 (m,1H), 2.28–2.40 (m, 2H), 2.79–2.88 (m, 2H), 4.18 (dd, 1H,J = 5.1 Hz, J = 9.0. Hz), 4.22 (dd, 1H, J = 5.0 Hz, J =8.9 Hz). m/z (ESI) C24H45N7O7 ([M +H+]) calcd. 488.3,found 488.3.(S)-2-[3-((S)-1-Carboxy-5-[3-(6-fluoropyridine)carbonyl)amino)pentyl)ureido]-pentanedioicacid (DCFPyL) [15]In a flame-dried flask, HBTU (156 mg, 410 μmol),DIPEA (72 μL, 410 μmol), and 6-fluoronicotinic acid 6(58 mg, 410 μmol) were added to a solution of 100 mg(0.205 μmol) of compound 5 in 4 mL of distilled CH2Cl2.The reaction was stirred at 25 °C for 3 h under nitrogen,and progress of the reaction was monitored by TLC (hex-ane/ethyl acetate 1/1 v/v). Upon completion, the reactionmixture was concentrated under reduced pressure and theresidue was purified by column chromatography (CH2Cl2/MeOH 97/3) to afford 97 mg (78 %) of tert.-butyl esterintermediate as a white powder. TLC: (EtOAc/hexane1/1): Rf = 0.45.1H-NMR (600 MHz, CDCl3) δ: 1.39 (s, 9H),1.41–1.46 (m, 19H), 1.51–1.63 (m, 2H), 1.64–1.71 (m, 1H),1.73–1.80 (m, 1H), 1.80–1.86 (m, 1H), 1.98–2.08 (m, 1H),2.25–2.39 (m, 2H), 3.33–3.42 (m, 1H), 3.50–3.59 (m, 1H),3.69–3.77 (m, 1H), 4.45–4.24 (m, 2H), 5.59 (t, 1H, J =9.5 Hz), 5.88 (t, 1H, J = 9.8 Hz), 7.00 (d, 1H, J = 8.1Hz),7.81–7.90 (m, 1H), 8.41 ( t, 1H, J = 7.9 Hz), 8.81 (s, 1H).13C-NMR (150.9 MHz, CDCl3) δ: 23.36, 27.81, 27.88, 27.95,28.04, 28.66, 31.51, 32.43, 39.93, 53.12, 53.69, 80.72, 81.64,82.54, 109.09 (d, J = 37.6 Hz), 128.56 (d, J = 4.3 Hz) 140.96(d, J = 8.7 Hz), 147.95 (d, J = 13.1 Hz), 157.60, 164.86164.87 (d, J = 244 Hz), 172.16, 172.31, 173.49. m/z (ESI)C30H47FN4O8 ([M + H+]) calcd. 611.3, found 611.3.Tert.-butyl ester intermediate was treated with 6 mLof CH2Cl2/TFA (1/1 v/v) for 8 h at 25 °C, concentratedunder reduced pressure, and purified by HPLC. HPLCpurification was performed on a semi-preparative JupiterC12 column (100 Å, 10 μm, 250 × 10 mm). The elutingsolvent started with a 10/90 CH3CN/(water 0.5 % TFA)gradient for 5 min at a flow rate of 2 mL min−1 followedby a gradient from 10/90 to 70/30, v/v, for 25 min. Com-pound DCFPyL eluted at 14.5 min, and after solvent re-moval, 54 mg (67 %) of DCFPyL was isolated as a whitepowder. 1H-NMR (400 MHz, D2O) δ: 1.42–1.56 (m, 2H),1.60–1.73 (m, 2H), 1.73–1.83 (m, 1H), 1.85–1.95 (m, 1H),1.95–2.05 (m, 1H), 2.13–2.24 (dt, 1H, Jd = 20.6 Hz, Jt =7.5 Hz), 2.51 (t, 2H, J = 7.2 Hz), 3.43 (t, 2H, 6.5 Hz), 4.25(ddd, 2H, J = 9.4 Hz, J = 5.1 Hz, J = 9.2 Hz), 7.23 (d, 1H,J = 8.2 Hz), 8.31 (dt, 1H Jt = 8.2 Hz, Jd = 2.2 Hz), 8.57 (d,1H, J = 2.2 Hz).13C-NMR δ (125.7 MHz, D2O): 23.26,27.11, 28.56, 31.00, 31.55, 40.68, 53.51, 54.07, 109.05 (d,J = 37.8 Hz), 128.47 (d, J = 4.4 Hz) 140.87 (d, J = 8.1 Hz),147.98 (d, J = 12.9 Hz), 160.26, 164.62 (d, J = 240.1 Hz),168.16, 177.12, 178.05, 178.19. m/z (HRMS) C18H22FN4O8([M −H+]) calcd. 441.1427, found 441.1430.6-Trimethylammonium-nicotinic acid 2,3,5,6-tetrafluorophenyl ester triflate salt 8 (adapted fromreference [21])One gram (6.34 mmol) of 6-chloronicotinic acid 7; 1.1 g(6.5 mmol) of 2,3,5,6 tetrafluorophenol; and 1.31 g(6.34 mmol) of N,N′-dicyclohexylcarbodiimide (DCC) werestirred in dioxane (40 mL) for 5 h at 25 °C. Progress ofthe reaction was monitored by TLC (EtOAc/hexane 4/1).Upon completion, the reaction mixture was filtered andBouvet et al. EJNMMI Research  (2016) 6:40 Page 4 of 15concentrated under reduced pressure and the residue waspurified by recrystallization in hot hexane to afford 1.35 g(70 %) of the 6-chloronicotinic acid active ester inter-mediate as a white powder [18]. TLC: (EtOAc/hexane4/1): Rf = 0.28.1H-NMR (600 MHz, CDCl3) δ 7.11 (tt, 1H,J = 7.1 Hz, J = 9.9 Hz), 7.57 (d, 1H, J = 8.3 Hz), 8.43 (dd,1H, J = 8.3 Hz, J = 2.7 Hz), 9.21 (d, 1H, J = 2.7Hz). m/z(ESI) C12H4ClF4NO2 ([M +H+]) calcd. 305.0, found 304.9.6-Chloronicotinic acid active ester intermediate (130 mg)[18] was dissolved in 3 mL of a 1 M Me3N solution in THFand stirred 2 h at 25 °C. After 5 min, a white precipitatewas formed. After completion of the reaction, the precipi-tate was collected by filtration and washed with diethylether and cold CH2Cl2. The obtained white powder wassuspended in 5 mL of CH2Cl2 containing 2 % TMSOTf andsonicated for 10 min. The reaction mixture was concen-trated under reduced pressure and washed with diethylether to afford 140 mg (68 % over two steps) of a graypowder after drying. 1H-NMR (600 MHz, CD3CN) δ7.43 (tt, 1H, J = 7.4 Hz, J = 10.5 Hz), 8.07 (dd, 1H, J =8.6 Hz, J = 0.8 Hz), 8.85 (dd, 1H, J = 8.6 Hz, J = 2.3 Hz), 9.34(dd, 1H, J = 2.3 Hz, J = 0.8 Hz). m/z (ESI) C15H13F4N2O2([M+]) calcd. 330.1, found 330.0.(S)-2-[3-((S)1-Carboxy-5-((6-trimethylammonium-pyridine-3-carbonyl)-amino)-pentyl)-ureido]-pentanedioic acidtrifluoroacetate salt 9To a solution of 80 mg (164 μmol) of compound 5 inCH2Cl2 (4 mL) was added compound 8 (100 mg, 209 μmol)and 100 μL of DIPEA (572 μmol). The reaction was stirredfor 2 h at 25 °C and then concentrated under reducedpressure.Progress of the reaction was monitored by TLC:(CH2Cl2/MeOH 4/1): Rf = 0.26. HPLC purification wasperformed on a semi-preparative Jupiter C12 column(100 Å, 10 μm, 250 × 10 mm). The eluting solvent startedwith a gradient from 5/95 to 70/30 acetonitrile/(water 0.5 %TFA) for 20 min at a flow rate of 2 mL min−1. Then theeluent was kept at 70/30 acetonitrile/(water 0.5 % TFA) for10 min to elute the desired compound at 25.8 min. Afterremoval of the solvent under reduced pressure gave 96 mg(77 %) of desired compound 9 as a white powder. 1H-NMR(600 MHz, D2O) δ: 1.32 (s, 9H), 1.34 (s, 9H), 1.35(s, 9H),1.35–1.39 (m, 2H) 1.55–1.66 (m, 3H), 1.70–1.82 (m, 2H),1.95–2.03 (m, 1H), 2.30 (M, 2H), 3.36 (t, 2H, J = 6.8 Hz),3.57 (s, 9H), 4.02 (ddd, 2H, J = 9.5 Hz, J = 8.7 Hz, J =5.1 Hz), 7.94 (d, 1H, J = 8.8 Hz), 8.35 (dd, 1H Jt = 8.8 Hz, Jd= 2.3 Hz), 8.57 (d, 1H, J = 2.3 Hz). 13C-NMR (125.7 MHz,D2O) δ: 23.35, 27.63, 28.19, 28.20, 28.31, 28.78, 31.92,32.58, 40.80, 54.34, 54.99, 56.06, 83.47, 84.18, 84.36, 115.48,118.47, 133.37, 141.14, 148.90, 160.16, 167.46, 174.55,175.14, 175.33. m/z (HRMS) C33H56NO8 ([M+]) calcd.650.4123, found 650.4116. Mp = 56 °C.Radiosynthesis and quality control of [18F]DCFPyLRadiosynthesis of [18F]DCFPyL was performed on aGE TRACERlabTM FX (GE Healthcare, Mississauga,ON, Canada). The synthesis module was modified interms of program and hardware (see Fig. 3). The synthesisunit was installed and operated in a shielded hot cell.Analytical HPLC was carried out using a Gilson HPLC(Mandel Scientific Company Inc.; Guelph, Ontario,Canada) by injection of HPLC-purified [18F]DCFPyLonto a Phenomenex Nucleosil Luna C18 column (10 μm,250 × 10 mm) and elution with 20 % CH3CN/0.2 % TFAfor 5 min at 2 mL min−1, followed by gradient elutionfrom 20 % to 38 % CH3CN for 5 min and from 38 % to70 % CH3CN for 15 min with isocratic elution at 70 %CH3CN for 15 min. Radio-TLC analysis on silica gel platesgave a Rf value of 0.6 in 95 % CH3CN/H2O (Additional file1: Figure S4).Automated synthesis of [18F]DCFPyLRadiosynthesis of [18F]DCFPyL was performed on a GETRACERlabTM FX (GE Healthcare, Mississauga, ON,Canada). The synthesis module was modified in terms ofprogram and hardware (Fig. 3). The synthesis unit wasinstalled and operated in a shielded hot cell.In vivo tumor modelsAll animal experiments were carried out in accordancewith the guidelines of the Canadian Council on AnimalCare (CCAC) and approved by the local animal care com-mittee (Cross Cancer Institute, University of Alberta).PET imaging experiments were carried out in LNCaPand PC3 tumor-bearing Balb/c nude mice (Charles RiverLaboratories, Quebec, Canada). Male Balb/c nude micewere housed under standard conditions with free accessto standard food and tap water. LNCaP and PC3 cells(5 × 106 cells in 100 μL of PBS) were injected into theupper left flank of the mice (20–24 g). Before injectingLNCaP cells, the mice received a 1.0-mg/pellet containingtestosterone in a 60-day release preparation (InnovativeResearch of America, Sarasota, FL, USA).The pellet was implanted subcutaneously into theupper right flank in order to provide a constant level oftestosterone needed by the androgen receptor positiveLNCaP cells. Tumors reached sizes of approximately1 cm3 which were suitable for all in vivo experiments.Radiometabolite analysisNormal BALB/c mice were injected with 10–20 MBq of[18F]DCFPyL. Venous blood samples were collected at 5,15, 30, and 60 min p.i. via the mouse tail vein and furtherprocessed. Blood cells were separated by centrifugation(13,000 rpm × 5 min). Precipitation of proteins in thesupernatant was achieved by the addition of 2 volumeparts of MeOH, and the samples were centrifuged againBouvet et al. EJNMMI Research  (2016) 6:40 Page 5 of 15(13,000 rpm × 5 min). Fractions of blood cells, proteins,and plasma were measured in a Wizard gamma counterto determine radioactivity per sample. The clear plasmasupernatant was injected onto a Shimadzu HPLC system.The samples were analyzed using a Phenomenex Luna10u C18 [2] 100 A, 250 × 4.6 mm column at a constantflow rate of 1 mL min−1 and the following gradient withwater/0.2 % TFA as solvent A and CH3CN as solvent B:0–7.5 min 20 % B, 7.5–15 min gradient to 90 % B, 15–20 min 90 % B.Dynamic PET imagingGeneral anesthesia of tumor-bearing mice was inducedwith inhalation of isoflurane in 40 % oxygen/60 % nitro-gen (gas flow = 1 mL min−1), and the mice were subse-quently fixed in prone position. The body temperaturewas kept constant at 37 °C for the entire experiment.The mice were positioned in a prone position into thecenter of the field of view. A transmission scan for at-tenuation correction was not acquired. The mice wereinjected with 2–10 MBq of [18F]DCFPyL (60–150 ng) in100–200 μL of isotonic NaCl solution (0.9 %) through atail vein catheter. For blocking studies, the animals werepre-dosed with 300 μg of DCFPyL in 50 μL saline about10 min prior to radiotracer injection. Data acquisitionwas performed over 60 min in a 3D list mode. The dy-namic list mode data were sorted into sinograms with 53time frames (10 × 2, 8 × 5, 6 × 10, 6 × 20, 8 × 60, 10 × 120,5 × 300 s). The frames were reconstructed using max-imum a posteriori (MAP) as reconstruction mode. Thepixel size was 0.085 × 0.085 × 0.121 mm3 (256 × 256 ×63), and the resolution in the center of the field of viewwas 1.8 mm. No correction for partial volume effectswas applied. The image files were processed using theROVER v 2.0.51 software (ABX GmbH, Radeberg,Germany). Masks defining 3D regions of interest (ROI)were set and the ROIs were defined by thresholding.Mean standardized uptake values [SUV]mean = (activity/mL tissue)/(injected activity/body weight), mL/g, were cal-culated for each ROI. Time-activity curves (TACs) weregenerated for the dynamic scans only. All semi-quantifiedPET data are presented as means ± SEM. Statistical differ-ence for the blocking study was tested by unpaired Stu-dent’s t test and was considered significant for P < 0.05.Internalization experiments105 LNCaP or PC3 cells were seeded in poly-D-lysine-coated 12-well plates 24–48 h before the assay so thatcells could reach 95 % confluency. The medium was re-moved 1 h before the assay, and the cells were rinsedtwice with PBS. After the addition of Krebs buffer(1 mL) to each well, the cells were incubated at 37 °C.Krebs buffer was aspirated, and the cells were incubatedwith 300 μL of [18F]DCFPyL in 0.9 % NaCl (0.1 MBq )for 60 min at 37 °C. Cellular uptake was stopped by re-moving incubation media from the cells and washing thewells twice with ice-cold PBS buffer (1 mL). Surface-bound radioactivity was removed from the cells throughincubating the cells twice with 0.5 mL glycine-HCl inPBS (50 mM, pH 2.5) for 5 min at 37 °C. Cells werewashed again with ice-cold PBS before the addition ofradio-immunoprecipitation assay (RIPA) buffer (400 μL)to lyse the cells. Cells were returned into the incubatorfor 10 min, and cell lysates were collected for counting.Radioactivity of surface-bound and internalized fractionwas measured in a WIZARD2 Automatic gamma coun-ter (Perkin Elmer, Waltham, MA, USA). Total proteinconcentration in the samples was determined by thebicinchoninic acid method (BCA; Pierce, Thermo Scien-tific 23227) using bovine serum albumin (800, 600, 400,300, 200, 100, 50 μg/mL, blank) as protein standard.Data are expressed as percent of total uptake per 1 mgprotein (% of total uptake/mg protein).Tracer kinetic analysisTracer kinetic analysis was performed using a two-tissuecompartmental model using dynamically acquired PETimaging data. Full details on tracer kinetic analysis aregiven in the Additional file 1.ResultsChemistry and radiochemistrySynthesis of lysine-urea-glutamate peptidomimetics ashighly potent PSMA-binding motifs is given in Fig. 2.Compounds 4, 5, and DCFPyL were previously describedin the literature [15, 16]. Herein, we present an improvedsynthesis route [15, 16].Synthesis commenced with the activation of tert-butylester protected amino acid H-Glu(OtBu)-OtBu.HCl 1with N,N′-disuccinimidyl carbonate to give correspond-ing NHS ester (S)-(1,5-di-tert-butoxy-1,5-dioxopentan-2-yl)carbamic acid 2 in 87 % yield. Coupling of active ester2 with μ-benzyloxycarbonyl-L-lysine tert-butyl ester (H-Lys(Z)-OtBu.HCl) 3 afforded 2-(3-{1-tert-butylcarboxy-late-5-[(carboxybenzyl)-amino]-pentyl}-ureido)-di-tert-butyl pentanedioate 4 in 92 % yield. Removal of the Zprotecting group in compound 4 using hydrogenationon Pd/C gave 2-{3-[1-tert-butylcarboxylate-(5-aminopen-tyl)-ureido}-di-tert-butyl pentanedioate 5 in high chemicalyields of 95 %.The free amine in 5 was acylated with 6-fluoronicotinicacid 6 in the presence of HBTU followed by acidic cleavageof tert-butyl ester groups using trifluoroacetic acid (TFA) togive reference compound 2-(3-{1-carboxy-5-[(6-fluoropyri-dine-3-carbonyl)-amino]-pentyl}-ureido)-pentanedioic acid(DCFPyL) in 53 %. The total yield for the four step synthe-sis of DCFPyL was 40 %.Bouvet et al. EJNMMI Research  (2016) 6:40 Page 6 of 15Trimethylammonium salt 9, which was the precursorfor direct radiofluorination, was prepared by acylationreaction of compound 5 with active ester N,N,N-tri-methyl-5-((2,3,5,6-tetrafluorophenoxy)carbonyl)-pyridin-2-aminium triflate 8. Active ester 8 was prepared withslight modifications based on a published procedure [21]in 68 % yield in two steps starting from 6-chloronicotinic6 acid. Activation with 2,3,5,6-tetrafluorophenol gavecompound 7 which was treated with trimethylaminefollowed by TMSOTf to afford compound 8. Compound 9was obtained through reaction of compound 5 with activeester 8 in 77 % yield after purification by HPLC.Automated radiosynthesis of [18F]DCFPyL was per-formed on a GE TRACERlabTM FXFN automated synthesisunit (ASU) equipped with an integrated HPLC systemusing a Luna semi-prep HPLC column (Fig. 3).The overall radiosynthesis of [18F]DCFPyL is displayedin Fig. 4, and the optimized HPLC purification is given inthe supplementary materials (Additional file 1: Figure S3).In the first reaction step (R1), cyclotron-produced no-carrier added (n.c.a.) [18F]fluoride was captured from[18O]H2O target solution onto a Chromafix cartridge. Asolution (0.6 mL) containing Bu4N+-HCO3− (0.3 mL,0.075 M) and CH3CN (0.3 mL) was used to elute n.c.a.[18F]fluoride from the resin into the reactor. The aqueous[18F]fluoride solution was dried azeotropically throughconsecutive addition and removal of anhydrous CH3CN.The first drying cycle was performed at 55 °C for 1 min at0.1 bar. After the addition of CH3CN (2 mL), solventswere evaporated at 55 °C for 1 min at 0.1 bar followed byan evaporation step at 95 °C for 3 min under high vacuum(0.01 bar). The end of the process yielded activated n.c.a.[18F]fluoride suitable for subsequent nucleophilic radio-fluorination. This three-step azeotropic distillation allowsfor minimum dispersion of [18F]fluoride on the side of thereactor vial and concentrates most of the radioactivity atthe bottom of the reactor. This is important for the subse-quent radiolabeling step which uses only a small volumeof solvent (0.5 mL).The second reaction (R2) involved combining n.c.a.[18F]fluoride and 9 to form compound 10 via a nucleo-philic heteroaromatic substitution reaction. Radiofluori-nation was carried out with 5 mg of labeling precursor 9in anhydrous CH3CN (0.5 mL) at 60 °C for 10 min.Final reaction step (R3) included removal of the tert.-butyl ester protecting groups in compound 10 via acidiccleavage using 1.5 mL of HCl (10 N) in 1 mL CH3CN at40 °C for 5 min to give crude [18F]DCFPyL.The pH of the crude reaction mixture was adjusted bythe addition of 2 mL of 0.1 M NaOAc (pH 5.8), and thereaction mixture was transferred into a 5-mL injectionloop and injected onto a Luna C-18 column (10 μm,250 × 10 mm) for purification employing isocratic elu-tion with 18 % EtOH containing 0.2 % H3PO4 at a flowrate of 2 mL min−1. Product peak was collected between22 and 23 min (Additional file 1: Figure S3). The collectedpeak (~2 mL) was transferred, through a sterile filter, intoa sterile vial containing 12 mL of 0.1 M NaOAc (pH 5.8).The total synthesis time was 55 min, including HPLCpurification. The overall isolated radiochemical yield of[18F]DCFPyL was 23 ± 5 % (n = 10, decay-corrected).Quality control of [18F]DCFPyLIdentity of [18F]DCFPyL was confirmed by radio-HPLCand radio-TLC using co-injection and co-spotting withreference compound [19F]DCFPyL, respectively. Repre-sentative HPLC traces are given in Fig. 5a. Quality con-trol revealed high radiochemical purity of 98 %, and theFig. 2 Synthesis of lysine-urea-glutamate peptidomimeticsBouvet et al. EJNMMI Research  (2016) 6:40 Page 7 of 15specific activity was determined to be 80–100 GBq/μmol(Fig. 5b, Additional file 1: Figures S5 and S6) starting with15 to 26 GBq of n.c.a. [18F]fluoride. Radio-TLC analysis onsilica gel plates gave a Rf value of 0.6 in 95 % CH3CN/H2O(Additional file 1: Figure S4), and the product was stable(>95 %) in saline of up to 4 h.Cell-specific internalization experimentsRadiotracer [18F]DCFPyL revealed substantial higher cellmembrane accumulation and internalization in PSMA+LNCaP cells compared to PSMA− PC3 cells (Fig. 6). Afterincubation for 60 min, cellular uptake of [18F]DCFPyL wasstopped and cells were washed with glycine-HCl to re-move radioactivity bound to the membrane. Internalizedfraction of radioactivity was determined after lysing thecells with RIPA buffer. After 60 min, radioactivity uptakein PSMA+ LNCaP cells reached 45.47 ± 1.01 % of total up-take/mg protein for the membrane-bound fraction and35.05 ± 0.69 % of total uptake/mg protein for the internal-ized fraction. This reflects a 44 % fraction of internalizedradioactivity in PSMA+ LNCaP cells after incubation withradiotracer [18F]DCFPyL for 60 min. In contrast, only4.30 ± 0.40 % of total uptake/mg protein and 0.76 ± 0.05 %of total uptake/mg protein were found in the membrane-bound and internalized fraction, respectively, in the caseof PSMA− PC3 cells.Dynamic PET imaging experimentsRadiotracer [18F]DCFPyL was injected into PSMA+ LNCaPand PSMA− PC3 tumor-bearing BALB/c nude mice. Asshown in Fig. 7, high radioactivity uptake and retention wasobserved in the PSMA+ LNCaP tumor, whereas only veryFig. 3 Automated synthesis unit for the radiosynthesis of [18F]DCFPyLBouvet et al. EJNMMI Research  (2016) 6:40 Page 8 of 15Fig. 4 Radiosynthesis of [18F]DCFPyLFig. 5 a HPLC traces for confirmation of identity after co-injection of [18F]DCFPyL with reference compound DCFPyL and b quality control of thefinal product. Detectors are connected in seriesBouvet et al. EJNMMI Research  (2016) 6:40 Page 9 of 15low uptake and no retention was seen in the PSMA− PC3tumor. Analyzed TACs describe a continuous increase ofradioactivity accumulation and retention in the PSMA+LNCaP tumor over 60 min reaching a standardized uptakevalue (SUV60min) of 1.1 ± 0.1 (n = 5).In contrast, PSMA− PC3 tumors showed some radio-activity uptake during the perfusion phase followed byrapid washout of radioactivity reaching a SUV of 0.1(n = 2) at 60 min p.i..In both prostate cancer models, radioactivity is rapidlycleared from blood and muscle. Radiotracer [18F]DCFPyLis eliminated through the kidneys with somewhat more re-tention of radioactivity in the kidneys in the PC3 model.Only very little radioactivity is found in all other organsresulting in a low background signal. Rapid clearance fromblood and muscle tissue led to high tumor-to-blood andtumor-to-muscle ratios of 8.9 ± 1.9 and 17.0 ± 3.4, respect-ively, after 60 min p.i. in LNCaP-bearing mice.Specificity of radiotracer [18F]DCFPyL for PSMA wasdemonstrated through blocking experiments in theLNCaP model. LNCaP PSMA+ model was injected with300 μg of nonradioactive DCFPyL 5 min prior to the ad-ministration of radiotracer [18F]DCFPyL. Control andblocking experiments were carried out in the same ani-mal on 2 consecutive days.Acquired PET images confirmed substantial decreaseof radioactivity uptake in the LNCaP tumor, which wasstatistical significant (P = 0.0069). Radioactivity uptake wasreduced by 80 % after 60 min p.i. upon pretreatment withDCFPyL (SUV60min control: 1.1 ± 0.1, n = 5; SUV60minblocked: 0.2 ± 0.05, n = 4). Uptake in the kidneys was alsoreduced (Fig. 8).In vivo metabolic stability of [18F]DCFPyLIn vivo metabolic stability of [18F]DCFPyL was studiedby analyzing murine blood samples at different timepoints and urine after 60 p.i. using radio-HPLC. Figure 9summarizes the results reflecting the distribution patternof radioactivity in blood cells, plasma proteins, and plasmaover time. The overall distribution of radioactivity in thedifferent blood compartments remained mainly unchangedover time, suggesting rapid equilibration of radioactivitydistribution between analyzed blood compartments. Themajority of radioactivity is found in the plasma fraction in-dicating high bioavailability of radiotracer [18F]DCFPyL.Radio-HPLC analysis of the plasma samples at differenttime points confirmed very high metabolic stability of theradiotracer.High stability of [18F]DCFPyL was also found in urinesamples after 60 min p.i. In all cases, no radiometaboliteswere detected in plasma and urine samples over the studiedtime course of 60 min p.i. cells (3) PC3 cells (3) membrane-bound internalizednietorpgm/e katpul atotfo%Fig. 6 Cell-specific internalization experiments of [18F]DCFPyL inPSMA+ LNCaP and PSMA− PC3 cells 60 min after incubation withthe radiotracer at 37 °CFig. 7 PET images of [18F]DCFPyL after 60 min p.i. into LNCaP (left) and PC3 (right) tumor-bearing BALB/c nude mice. Middle: Time-activity curvesfor radioactivity uptake in LNCaP and PC3 tumorBouvet et al. EJNMMI Research  (2016) 6:40 Page 10 of 15Tracer kinetic analysisCellular uptake of [18F]DCFPyL is presumed to follow atwo-step process, consisting of first binding to PSMA onthe cell membrane and then transport of the [18F]DCFPyL-PSMA complex into the cytoplasm. We have thereforemodeled the tracer kinetics of [18F]DCFPyL using a revers-ible two-tissue compartmental model. Tracer kineticanalysis was carried out with PET imaging data acquireddynamically from PSMA+ LNCaP and PSMA− PC3tumor-bearing mice (see Fig. 7). Four kinetic parameters(K1, k2, k3, k4) describing the flow of the radiotracer fromthe blood to the tissue (K1) and from the tissue into theblood (k2), binding of the radiotracer to the tissue (k3) anddissociation from the tissue (k4), were calculated.The analysis was performed by fitting the measuredTACs with a two-exponential model of the general formas described, e.g., by van den Hoff (Additional file 1:Figure S2) [22]. Results for the four kinetic parametersare given in Fig. 10.The net delivery (K1/k2) of the radiotracer to the tissuewas comparable for LNCaP (K1/k2 = 0.62) and PC3 (K1/k2 = 0.49) tumors. However, radiotracer [18F]DCFPyLaccumulated significantly (10.25 times) more in PSMA+LNCaP tumors compared to PSMA− PC3 tumors (k3(LNCaP) 0.082 vs. k3 (PC3) 0.008). The coefficient k3/(k3 + k2) describes the fraction of radiotracer enteringthe second compartment (i.e., being bound as opposedto being released back into the blood pool): the valuefor LNCaP (0.314) dominates that of PC3 (0.058) bymore than a factor 5. Significant differences were alsofound in the dissociation rate (k4) of the radiotracerfrom the tumor (k4 (LNCaP) 0.004 vs. k4 (PC3) 0.061).Overall, this led to high retention of the radiotracer inLNCaP tumors (k3/k4 = 20.5) and negligible retentionin PC3 tumors (k3/k4 = 0.13).DiscussionThe present study described the automated radiosynthesisand preclinical validation of (S)-2-[3-((S)-1-carboxy-5-[3-(6-fluoropyridine)carbonyl)amino)pentyl)ureido]-pentanedioicacid ([18F]DCFPyL) as a radiotracer for molecular imagingof prostate-specific membrane antigen (PSMA) in prostatecancer. We performed the radiosynthesis of [18F]DCFPyLemploying direct nucleophilic heteroaromatic substitutionas novel synthesis route in an ASU. The radiopharmacolo-gical evaluation of [18F]DCFPyL included dynamic PET im-aging, metabolic profiling, and tracer kinetic analysis.The following important results emerged from thisstudy: [1] radiotracer [18F]DCFPyL can be prepared ingood radiochemical yields suitable for clinical applica-tions via a direct radiofluorination synthesis route in anautomated GE TRACERlabTM FXFN synthesis unit; [2]radiopharmacological profile of [18F]DCFPyL preparedin an ASU via direct radiofluorination agrees as expectedwith previously published work such as high specificuptake and retention in PSMA+ tumors, very highmetabolic stability, and high bioavailability in vivo.The first part of this study was focused on the devel-opment of a novel and automated radiosynthesis of[18F]DCFPyL using a direct radiolabeling approach withcyclotron-produced n.c.a. [18F]fluoride. Radiochemistrywith prosthetic groups for 18F labeling represents a specialchallenge for robust and reliable radiotracer synthesis, es-pecially when automated processes according to GMPguidelines for clinical applications are required. Literaturereports on syntheses of 18F-labeled small-molecule PSMAFig. 8 PET images of [18F]DCFPyL after 60 min p.i. into a LNCaP tumor-bearing BALB/c nude mouse. Left: control; Middle: time-activity curves foruptake and blocking in LNCaP tumors in the absence and presence of nonradioactive DCFPyL; Right: blockedBouvet et al. EJNMMI Research  (2016) 6:40 Page 11 of 15inhibitors like [18F]DCFPyL mainly involved multiple stepreactions using manual synthesis procedures which pro-vided modest to good radiochemical yields [15, 17–19].Reported decay-corrected radiochemical yields for thepreparation of [18F]DCFPyL using prosthetic group 2,3,5,6-tetrafluorophenyl-6-[18F]fluoronicotinate ([18F]FPy-TFP)vary greatly from 5 to 53 % [12, 13, 15]. Low radiochemicalyields (decay-corrected) in the range of 5–12 % are reportedfor [18F]DCFPyL when the radiotracer was produced forclinical trials [12, 13]. In one case, the radiosynthesis wascarried out in a modified dual-run FDG synthesis module[13]. Higher decay-corrected radiochemical yields of36–53 % were reported for the original manual synthesisof [18F]DCFPyL when modest amounts (1.6–2.2 GBq) of[18F]fluoride were used [15]. Various literature reportsdocumented excellent radiochemical yields for directnucleophilic heteroaromatic substitution reactionswith n.c.a. [18F]fluoride at the ortho-position to preparevarious 2-[18F]fluoropyridines, including the prostheticgroup [18F]FPy-TFP, under mild conditions [21]. RecentFig. 9 Distribution of [18F]DCFPyL in the blood (top) revealing only minimal binding to plasma proteins and therefore good bioavailability andevaluation of the metabolic stability of [18F]DCFPyL (bottom) following injection in BALB/c mice (n = 3)Bouvet et al. EJNMMI Research  (2016) 6:40 Page 12 of 15applications and advantages of this methodology for thedesign and synthesis of numerous 18F-labeled radiotracershave extensively been reviewed [23]. Building on thiswork, we envisaged a direct radiofluorination synthesisroute based on 2-[18F]fluoropyridine chemistry for thepreparation of [18F]DCFPyL using the trimethylammo-nium salt 9 as the labeling precursor. Attempts to prepare[18F]DCFPyL in satisfying radiochemical yields from re-lated 6-chloronicotinic acid compound were not success-ful (data not shown).The use of trimethylammonium salt 9 as labeling pre-cursor offers several advantages. Notably as a solid, com-pound 9 allows for convenient handling and storage.Compound 9 was stored at 4 °C without significant de-composition for at least 3 months. Moreover, excellentleaving group properties of the Me3N group in hetero-aromatic nucleophilic substitution reactions with n.c.a.[18F]fluoride at the ortho-position of various pyridinesenables radiosynthesis under mild reaction conditionscompatible with technology and equipment of ASU likethe GE TRACERlabTM FXFN synthesis module.Automated radiosynthesis of [18F]DCFPyL was accom-plished in three reaction steps using a one reactor set-up in-volving drying and activation of cyclotron-produced n.c.a.[18F]fluoride (R1), incorporation of activated [18F]fluorideinto compound 10 via nucleophilic heteroaromatic substi-tution (R2), and removal of the tert-butyl protecting groupsthrough treatment with HCl and subsequent HPLCpurification to give final product [18F]DCFPyL (R3).The fully automated synthesis provided [18F]DCFPyL indecay-corrected radiochemical yields of 23 ± 5 % (n = 10)within 55 min, including final HPLC purification. In a typ-ical synthesis, 3 GBq of [18F]DCFPyL ready for injectionwas obtained starting from 20 GBq of n.c.a. [18F]fluoride.This is sufficient for several patient doses assuming a sin-gle patient dose of 300–400 MBq of [18F]DCFPyL. The ob-tained good and highly reproducible radiochemical yieldof 23 ± 5 % clearly demonstrates the favorable features ofthe direct radiofluorination synthesis route compared tothe existing method [12, 13, 15] and should help ensurethe availability of the agent. The specific activity was deter-mined to be 80–100 GBq/μM. Radiotracer [18F]DCFPyLpassed standard quality control test, making module-prepared [18F]DCFPyL suitable for human studies.The second part of this study dealt with the evaluationof module-prepared [18F]DCFPyL in preclinical prostatecancer models to further confirm pharmaceutical qualityand suitability of the radiotracer for PET imaging ofPSMA in vivo.Dynamic PET imaging of [18F]DCFPyL has not yetbeen reported using PSMA+ LNCaP and PSMA− PC3tumors. Previously reported studies were using PSMA+PC3-PIP and PSMA− PC3-flu xenografts [15].Prostate cancer cell lines LNCaP and PC3 have well-characterized PSMA expression levels to evaluate PSMAradiotracers and are therefore suitable for validation of[18F]DCFPyL prepared from the new method in com-parison with the original method [15, 17, 18].As shown in Fig. 6, radiotracer [18F]DCFPyL showedhigh uptake in PSMA+ LNCaP tumors with favorableclearance pattern from blood and muscles. The high up-take and retention of radioactivity in PSMA+ LNCaP tu-mors correlates with the reported high inhibitorypotency of [18F]DCFPyL (Ki = 1.1 nM) towards PSMA[15]. The fast clearance of the radiotracer from the blood,muscle, and most tissues and organs is in agreement withthe hydrophilic nature of compound [18F]DCFPyL as typ-ical for peptidomimetic PSMA inhibitors. Uptake of[18F]DCFPyL in PSMA+ LNCaP tumors (SUV60min = 1.1)was about 11 times higher compared to PSMA− PC3 tu-mors (SUV60min = 0.1). After 60 min p.i., only some radio-activity was also found in the kidney with only veryminimal background radioactivity in the liver. Most of theinjected radioactivity was accumulated in the bladder after60 min p.i..This biodistribution profile is consistent with previouslyreported studies using [18F]DCFPyL for PET imaging ofPSMA in other prostate cancer models [15]. As expectedfor PSMA− PC3 tumors, only very little radioactivity wasfound in the tumor at 60 min p.i.. The highest levels ofradioactivity in the PC3 model were also observed in thekidneys and bladder (Fig. 6).Renal clearance pathway as seen for urea-based pepti-domimetic PSMA inhibitor [18F]DCFPyL was also re-ported for 18F-labeled phosphoramidate peptidomimeticas small-molecule PSMA-imaging agent in LNCaP- andPC3-bearing mice [18]. Uptake of radiolabeled phos-phoramidate in PSMA+ LNCaP tumors was about fourtimes higher compared to the uptake in PC3 tumors asdetermined by ex vivo biodistribution studies.Favorable radiopharmacological profile and specificityof [18F]DCFPyL for PSMA imaging in vivo was furtherK1 k2 k3 k40. 1Fig. 10 Tracer kinetic analysis of [18F]DCFPyL in PSMA+ LNCaP andPSMA− PC3 tumor-bearing Balb/c miceBouvet et al. EJNMMI Research  (2016) 6:40 Page 13 of 15confirmed by specific blocking studies with nonradioactiveDCFPyL using the same animal on 2 consecutive days.The observed 80 % reduction of [18F]DCFPyL uptake inPSMA+ LNCaP tumors upon blocking with DCFPyL is in-dicative of specific targeting of [18F]DCFPyL to PSMA.Metabolic profiling of [18F]DCFPyL in mice revealedlow binding of the radiotracer to blood cells and plasmaproteins leading to high bioavailability. Analyzed plasmasamples further confirmed remarkable high metabolicstability of [18F]DCFPyL. No radiometabolites were de-tected by radio-HPLC analysis over 60 min p.i. Highmetabolic stability was further confirmed by analysis ofurine samples at 60 min p.i., which contained only par-ent compound [18F]DCFPyL.In the last part of this study, we applied tracer kineticanalysis to quantify and compare pharmacokinetics of[18F]DCFPyL for PSMA imaging in preclinical LNCaPand PC3 prostate cancer models. Delivery of the radio-tracer (K1) is dependent on blood flow, and our data in-dicate that radiotracer delivery was elevated for LNCaPtumors. However, net delivery as expressed by the K1/k2ratio was comparable for both tumor models (LNCaP:0.62; PC3: 0.49).The observed high uptake and retention of [18F]DCFPyLin LNCaP tumors was also reflected by the calculated k3values for both tumors. This was further supported by thecoefficient k3/(k3 + k2), which was five times higher forLNCaP tumors compared to PC3 tumors.Kinetic parameter k3 reflects the binding of the radio-tracer to the tissue, and the k3 value for LNCaP is about10 times higher than the k3 value for PC3. Tracer bind-ing is favored in PSMA+ LNCaP tumors, as expressedby the ratio k3/k4, and is more than 150 times higherthan for PC3 tumors. This led to an 11 times higher up-take of the radiotracer in LNCaP tumors, based on SUVvalues at 60 min p.i.. The high retention of radioactivityin the LNCaP model can be explained by high affinitybinding of the radiotracer to PSMA and internalizationof the PSMA-ligand complex. This is supported by thecell-specific internalization experiments in PSMA+ LNCaPand PSMA− PC3 cells. High retention of radioactivity inLNCaP tumors and rapid clearance of radioactivity frommost tissues and organs support findings from earlier workon [18F]DCFPyL as an ideal radiotracer for PSMA imagingin vivo [15].ConclusionsWe have developed an automated synthesis for radiotracer[18F]DCFPyL based on a direct radiofluorination synthesisroute. We demonstrated that [18F]DCFPyL can be preparedin good radiochemical yields and pharmaceutical qualitysuitable for clinical applications for PSMA imaging inhumans. The automated synthesis based on direct radio-fluorination should improve availability of [18F]DCFPyL forPSMA imaging in humans. Dynamic PET imaging of[18F]DCFPyL in PSMA+ LNCaP and PSMA− PC3tumor-bearing mice confirmed the previously reporteddata describing high PSMA-mediated tumor uptake andfavorable clearance profile of the radiotracer. Compart-mental model analysis points to a two-step moleculartrapping process based on PSMA binding and subsequentinternalization leading to retention of radioactivity inPSMA+ LNCaP tumors. These findings further under-line the excellent characteristics of [18F]DCFPyL for PETimaging of PSMA in vivo.Additional fileAdditional file: Supplementary information. Automated synthesis of[18F]DCFPyL via direct radiofluorination and radiopharmacologicalevaluation in preclinical prostate cancer models. (DOCX 446 kb)AbbreviationsMAP: maximum a posteriori; PET: positron emission tomography;PSMA: prostate-specific membrane antigen; ROI: regions of interest;SUV: standardized uptake value; SUV: standardized uptake value;TAC: time-activity curve; TLC: thin-layer chromatography.Competing interestsThe authors declare that they have no competing interests.Authors’ contributionsVB was responsible for chemical syntheses and radiosynthesis of theradiotracer. He also supported writing of the manuscript. MW wasresponsible for the radiopharmacological evaluation of the radiotracer,including analysis of the PET data. HSJ was responsible for the kineticanalysis experiments. NJ and ARG supported the chemical synthesis andcharacterization of the PSMA ligands. JFV, FB, and FW were responsible forthe design of the study and critically revised the manuscript. FW wrote themanuscript. All authors read and approved the final manuscript.AcknowledgementsThe authors would like to thank Dr. John Wilson, David Clendening, andBlake Lazurko from the Edmonton PET Center for providing 18F produced ona biomedical cyclotron. This work was supported by Medical Imaging ClinicalTrials Network (MITNEC) from the Canadian Institutes of Health Research(CIHR) and the Movember Foundation.Author details1Department of Oncology, University of Alberta, 11560 University Avenue,Edmonton, AB T6G 1Z2, Canada. 2Department of Chemistry and ChemicalBiology, McMaster University, Hamilton, Canada. 3Department of Radiology,University of British Columbia, Vancouver, Canada.Received: 10 February 2016 Accepted: 26 April 2016References1. Siegel R, Naishadham D, Jemal A. Cancer statistics, 2012. CA Cancer J Clin.2012;62:10–29.2. Kiess AP, Cho SY, Pomper MG. Translational molecular imaging of prostatecancer. Curr Radiol Rep. 2013;1:216–26.3. Hong H, Zhang Y, Sun J, Cai W. Positron emission tomography imaging ofprostate cancer. Amino Acids. 2010;39:11–27.4. Brogsitter C, Zöphel K, Kotzerke J. 18F-Choline, 11C-choline and 11C-acetatePET/CT: comparative analysis for imaging prostate cancer patients. Eur JNucl Med Mol Imaging. 2013;40 Suppl 1:S18–27.5. Vāvere AL, Kridel SJ, Wheeler FB, Lewis JS. 1-11C-acetate as a PETradiopharmaceutical for imaging fatty acid synthase expression in prostatecancer. J Nucl Med. 2008;49:327–34.Bouvet et al. EJNMMI Research  (2016) 6:40 Page 14 of 156. Seltzer MA, Barbaric Z, Belldegrun A, Naitoh J, Dorey F, Phelps ME, GambhirSS, Hoh CK. Comparison of helical computerized tomography, positronemission tomography and monoclonal antibody scans for evaluation oflymph node metastases in patients with prostate specific antigen relapseafter treatment for localized prostate cancer. J Urol. 1999;162:1322–8.7. Morris MJ, Akhurst T, Osman I, Nunez R, Macapinlac H, Siedlecki K, Verbel D,Schwartz L, Larson SM, Scher HI. Fluorinated deoxyglucose positronemission tomography imaging in progressive metastatic prostate cancer.Urology. 2002;59:913–8.8. Morris MJ, Akhurst T, Larson SM, Ditullio M, Chu E, Siedlecki K, Verbel D,Heller G, Kelly WK, Slovin S, Schwartz L, Scher HI. Fluorodeoxyglucosepositron emission tomography as an outcome measure for castratemetastatic prostate cancer treated with antimicrotubule chemotherapy.Clin Cancer Res. 2005;11:3210–6.9. Mease RC, Foss CA, Pomper MG. PET imaging in prostate cancer: focus onprostate-specific membrane antigen. Curr Top Med Chem. 2013;13:951–62.10. Afshar-Oromieh A, Hetzheim H, Kratochwil C, Benesova M, Eder M, NeelsOC, Eisenhut M, Kübler W, Holland-Letz T, Giesel FL, Mier W, Kopka K,Haberkorn U. The theranostic PSMA-ligand PSMA-617 in the diagnosis ofprostate cancer by PET/CT: biodistribution in humans, radiation dosimetryand first evaluation of tumor lesions. J Nucl Med. 2015;56:1697–705.11. Haberkorn U, Kopka K, Hadaschik B. Positron emission tomography-computedtomography with prostate-specific membrane antigen ligands as a promisingtool for imaging of prostate cancer. Eur Urol. 2015;S0302-2838(15):00855–6.12. Dietlein M, Kobe C, Kuhnert G, Stockter S, Fischer T, Schomäcker K, SchmidtM, Dietlein F, Zlatopolskiy BD, Krapf P, Richarz R, Neubauer S, Drzezga A,Neumaier B. Comparison of [(18)F]DCFPyL and [(68)Ga]Ga-PSMA-HBED-CCfor PSMA-PET imaging in patients with relapsed prostate cancer. MolImaging Biol. 2015;17:575–84.13. Szabo Z, Mena E, Rowe SP, Plyku D, Nidal R, Eisenberger MA, Antonarakis ES,Fan H, Dannals RF, Chen Y, Mease RC, Vranesic M, Bhatnagar A, Sgouros G,Cho SY, Pomper MG. Initial evaluation of [(18)F]DCFPyL for prostate-specificmembrane antigen (PSMA)-targeted PET imaging of prostate cancer. MolImaging Biol. 2015;17:565–74.14. Rowe SP, Gage KL, Faraj SF, Macura KJ, Cornish TC, Gonzalez-Roibon N,Guner G, Munari E, Partin AW, Pavlovich CP, Han M, Carter HB, BivalacquaTJ, Blackford A, Holt D, Dannals RF, Netto GJ, Lodge MA, Mease RC, PomperMG, Cho SY. 18F-DCFBC PET/CT for PSMA-based detection andcharacterization of primary prostate cancer. J Nucl Med. 2015;56:1003–10.15. Chen Y, Pullambhatla M, Foss CA, Byun Y, Nimmagadda S, SenthamizhchelvanS, Sgouros G, Mease RC, Pomper MG. 2-(3-{1-Carboxy-5-[(6-[18F]fluoro-pyridine-3-carbonyl)-amino]-pentyl}-ureido)-pentanedioic acid, [18F]DCFPyL, a PSMA-based PET imaging agent for prostate cancer. Clin Cancer Res. 2011;17:7645–53.16. Maresca KP, Hillier SM, Femia FJ, Keith D, Barone C, Joyal JL, ZimmermanCN, Kozikowski AP, Barret, JA, Eckelman WC and Babich JW. A series ofhalogenated heterodimeric inhibitors of prostate specific membraneantigen (PSMA) as radiolabeled probes for targeting prostate cancer. J MedChem. 2009;52:347–57.17. Mease RC, Dusich CL, Foss CA, Ravert HT, Dannals RF, Seidel J, Prideaux A,Fox JJ, Sgouros G, Kozikowski AP, Pomper MG. N-[N-[(S)-1,3-Dicarboxypropyl]carbamoyl]-4-[18F]fluorobenzyl-L-cysteine, [18F]DCFBC: anew imaging probe for prostate cancer. Clin Cancer Res. 2008;14:3036–43.18. Lapi SE, Wahnishe H, Pham D, Wu LY, Nedrow-Byers JR, Liu T, Vejdani K,VanBrocklin HF, Berkman CE, Jones EF. Assessment of an 18F-labeledphosphoramidate peptidomimetic as a new prostate-specific membraneantigen-targeted imaging agent for prostate cancer. J Nucl Med. 2009;50:2042–8.19. Malik N, Machulla HJ, Solbach C, Winter G, Reske SN, Zlatopolskiy B.Radiosynthesis of a new PSMA targeting ligand ([18F]FPy-DUPA-Pep). ApplRadiat Isot. 2011;69:1014–8.20. Malik N, Baur B, Winter G, Reske SN, Beer AJ, Solbach C. Radiofluorination ofPSMA-HBED via Al(18)F(2+) chelation and biological evaluations in vitro. MolImaging Biol. 2015;17:777–85.21. van den Hoff J. Principles of quantitative positron emission tomography.Amino Acids. 2005;29:341–53.22. Olberg DE, Arukwe JM, Grace D, Hjelstuen OK, Solbakken M, Kindberg GM,Cuthbertson A. One step radiosynthesis of 6-[(18)F]fluoronicotinic acid2,3,5,6-tetrafluorophenyl ester ([(18)F]F-Py-TFP): a new prosthetic group forefficient labeling of biomolecules with fluorine-18. J Med Chem. 2010;53:1732–40.23. Dollé F. [18F]fluoropyridines: from conventional radiotracers to the labelingof macromolecules such as proteins and oligonucleotides. Ernst ScheringRes Found Workshop. 2007;64(62):113–157.Submit your manuscript to a journal and benefi t from:7 Convenient online submission7 Rigorous peer review7 Immediate publication on acceptance7 Open access: articles freely available online7 High visibility within the fi eld7 Retaining the copyright to your article    Submit your next manuscript at 7 springeropen.comBouvet et al. EJNMMI Research  (2016) 6:40 Page 15 of 15


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items