Nuclear Medicine and Biology 40 (2013) 567–575

Contents lists available at SciVerse ScienceDirect

Nuclear Medicine and Biology

sev ie r .com/ locate /nucmedbio

j ourna l homepage: www.e l

5-Fluoro-[β-11C]-L-tryptophan is a functional analogue of5-hydroxy-[β-11C]-L-tryptophan in vitro but not in vivo

Olof Eriksson a,⁎, Ramkumar Selvaraju a, Beatrice Borg b, Veronika Asplund a,Sergio Estrada a, Gunnar Antoni a,b

a Department of Medicinal Chemistry, Preclinical PET Platform, Uppsala University, Uppsala, Swedenb Uppsala University Hospital, Uppsala, Sweden

⁎ Corresponding author. Preclinical PET Platform (PPPChemistry, Uppsala University, Dag Hammarskjölds vägSweden. Tel.: +46 707 903054 (cell), +46 18 4715304 (

E-mail address: olof.eriksson@pet.medchem.uu.se (O

0969-8051/$ – see front matter © 2013 Elsevier Inc. Alhttp://dx.doi.org/10.1016/j.nucmedbio.2013.02.005

a b s t r a c t

a r t i c l e i n f o

Article history:

Received 14 December 2012Received in revised form 21 January 2013Accepted 4 February 2013

Keywords:Neuroendocrine tumors5-HTPFTRP5-hydroxytryptophanFluoro-tryptophan

Introduction: 5-Hydroxy-[β-11C]-L-tryptophan ([11C]HTP) is an established positron emission tomography(PET) imaging agent for neuroendocrine tumors (NETs). It has also been used for other clinical researchpurposes in neurology and diabetes. However, its widespread use is limited by the short physical half-life ofthe radionuclide and a difficult radiosynthesis. Therefore, a Fluorine-18 labeled analogue, 5-[18F]Fluoro-L-tryptophan ([18F]FTRP), has been proposed as a functional analogue. There is no published method for thesynthesis of L-[18F]FTRP. We have therefore developed a synthesis of 5-fluoro-[β-11C]-L-tryptophan ([11C]FTRP), based on the existing chemo-enzymatic method for [11C]HTP and evaluated the potential usefulness ofradiolabeled FTRP as a substitute for [11C]HTP.Methods: The in vitro and in vivo behavior of [11C]FTRP, including the dependence of key enzymes in theserotonergic metabolic pathway, was investigated in NET cell lines, NET xenograft carrying immunodeficient

mice, normal rats and in non-human primate. [11C]HTP was used for direct comparison.Results: Uptake of [11C]FTRP in NET cell lines in vitro was mediated by enzymes involved in serotoninsynthesis and metabolism, similar to [11C]HTP. In vivo biodistribution, either in rodent or non-humanprimate, was not affected by selectively inhibiting enzymatic steps in the serotonergic metabolic pathway.Conclusion: [11C]FTRP has in vitro biological function similar to that of [11C]HTP. However, this function is notretained in vivo as shown by biodistribution and PET/CT studies. Radiolabeled FTRP is thus not likely toprovide an advantage over [11C]HTP in PET imaging in oncology, neurology or diabetes.

© 2013 Elsevier Inc. All rights reserved.

1. Introduction

5-Hydroxy-[β-11C]-L-tryptophan ([11C]HTP ) was originally de-veloped for studies of the decarboxylation step, mediated by DopaDecarboxylase (DDC), of monoamine neurotransmitter biosynthesisin vivo by positron emission tomography (PET) [1,2]. Since then, HTP-PET has found extensive use in the diagnosis and localization oftumors originating from cell types expressing the APUD (AminePrecursor Uptake and Decarboxylation) mechanism. [11C]HTP hasbeen established as a universal imaging agent for neuroendocrinetumors (NETs) [3]. The serotonergic metabolic pathway, expressed inthese tumors, contains many different enzymatic steps, frombiosynthesis of serotonin through decarboxylation of 5-HTP, totranslocation into secretory vesicles or degradation by monoamineoxidases (Fig. 1). More recently, [11C]HTP has been proposed as an

), Department of Medicinal14C, 3tr, SE-751 83, UppsalaOffice); fax: +46 18 4715307.. Eriksson).

l rights reserved.

imaging biomarker for native pancreatic islets of Langerhans [4–6], asthe serotonergic metabolic pathway including DDC is present in theendocrine pancreas [7].

The major obstacle limiting more widespread clinical utilizationof [11C]HTP is the difficult radiosynthesis which requires a technicallydemanding 7-step chemo-enzymatic synthesis pathway [8]. Theavailability if further complicated by the requirement for individualtracer batches for each HTP-PET examination due to the short half-life of carbon-11 (20.3 min). For these reasons [11C]HTP is currentlyavailable at only a few PET sites worldwide while [18F]L-DOPA iscommonly used instead [9]. However, [18F]L-DOPA has lowersensitivity for imaging of the monoamine decarboxylation mecha-nism in vivo and is not a functional analog of [11C]HTP [10,11].Instead 5-fluoro-[18F]-L-tryptophan and 6-fluoro-[18F]-L-tryptophanhave been proposed to alleviate several of the challenges associatedwith production and use of [11C]HTP for quantification of DDCactivity [12,13]. Given the longer half-life of fluorine-18 (109.8 min),each batch would be sufficient for several patients in addition togenerating the opportunity of shipping tracer to remote locationsfrom PET facilities with advanced radiochemistry. In a previous study

Fig. 1. Schematic depiction of the serotonergic metabolic pathway which is expressedin entirety or in part in tissues such as the central nervous system, in neuroendocrinetumors and in the endocrine pancreas. 5-HTP is transported into the cell by LAT1, andconverted in 5-HT (serotonin) by DDC. The fate of 5-HT is either incorporation insecretory vesicles by VMAT2 for extracellular release, or degradation into HIAAby MAO-A.

568 O. Eriksson et al. / Nuclear Medicine and Biology 40 (2013) 567–575

by Atkins it was found that both 5-fluoro-[18F]-L-tryptophan and 6-[18F]fluoro-DL-tryptophan were promising as potential pancreasscanning agents [12].

There are some potential advantages with 5-fluoro-tryptophancompared to 6-fluoro-tryptophan (which is an irreversible inhibitor oftryptophan hydroxylase, in addition to DDC) in this study. Firstly wewould like to avoid irreversible tracer kinetics since the biomathe-matical handling of data for such a tracer demands on simultaneousmeasurements of blood flow. A reversible tracer will not be influencedby differences in regional blood flow. Secondly, a crucial step formeasurement of DDC activity is to avoid high uptake in tissues otherthan neuroendocrine. It could be speculated that the liver uptakewould be higher with a substrate for tryptophan hydroxylase, such as6-fluoro-tryptophan, which could cause problems in clinical detectionof hepatic metastases of neuroendocrine tumors, in addition tocausing specific ligand-receptor interactions other than betweentracer and DDC. Hepatic phenylalanine hydroxylase catalyses hydrox-ylation of tryptophan, and potentially also 5-fluoro-tryptophan.

The synthesis of 5-[18F]fluoro-L-tryptophan is associated withconsiderable radiochemical difficulties, and we therefore propose tolabel FTRP with carbon-11 at the β-position for initial preclinicalevaluation. Bjurling et al. have earlier shown that the chemo-enzymatic synthesis route to [11C]HTP also can be used for theproduction of [11C]FTRP by simply changing 5-hydroxyindole for 5-fluoroindole [14]. [11C]FTRP is structurally congruent with theproposed [18F]FTRP and thus the in vitro and in vivo evaluation of[11C]FTRPwill yield crucial insight in the biological function of FTRP inrelation to the serotonergic metabolic pathway.

In this study we evaluated [11C]FTRP in direct comparison with[11C]HTP in cell binding assays and xenograft tumor models usingseveral APUD cell lines, as well as the in vivo biodistribution in rats

Scheme 1. Synthesi

and non-human primate, targeting both neuroendocrine tumors andpancreatic islets of Langerhans.

2. Method and materials

2.1. Radiosynthesis

[11C]HTP was synthesized as described previously and [11C]FTRPwas produced by the same method modified by exchanging 5-hydroxyindole for 5-fluoroindole as shown in Scheme 1 [8]. Thetracers were purified by HPLC, filtered through a 0.2 μm sterile filterand delivered in 17 mM acetic acid/1 mM ascorbic acid.

2.2. In vitro cell binding

80,000–200,000 cells of the neuroendocrine cell lines INS-1, QGP1or H720 were seeded into borosilicate tubes containing 1 ml 50 mMTRIS buffer (pH 7.4). [11C]FTRP or [11C]HTP was added to the tubes(corresponding to 13–114 nM [11C]FTRP or 4–322 nM [11C]HTP). Toinvestigate the uptake dependence of key enzymes in the serotonergicmetabolic pathway (see Fig. 1), cells were co-incubated with non-radioactive 5-hydroxy-L-tryptophan (5-HTP; Sigma-Aldrich, St Louis,MO, USA), 5-[18F]Fluoro-L-tryptophan (FTRP; Sigma-Aldrich, St Louis,MO, USA), 2-Amino-2-norbornanecarboxylic acid (BCH; inhibitor ofLAT-1; Sigma, St Louis, MO, USA), carbidopa (inhibitor of DDC;Apoteket AB, Stockholm, Sweden), harmine (HAR; inhibitor of MAO-A; Sigma, St Louis, MO, USA) or tetrabenazine (TBZ; inhibitor ofVMAT2; BioTrend, Cologne, Germany). The concentrations andintended effect of these compounds on cellular metabolism areoutlined in Table 1.

All cell suspensions were incubated at 37 °C for 30 min and thenmoved onto a 1.2 μmWhatman filter (Brandel, Gaithersburg, MD, USA)by a C-48 cell harvester (Brandel, Gaithersburg, MD, USA). The filtercomponents associated to each incubation tube were measured in awell-counter (GE Healthcare, Uppsala, Sweden). Cell incubation tubes,references and filter binding controls were prepared in triplicates.

2.3. Ex vivo organ distribution

Sprague Dawley rats (n = 33, 373 ± 10 g) were kept understandard laboratory condition with unlimited access to chow andwater.The rodent experiments were approved by the local EthicsCommittee for Animal Research (C49/10, C242/11) and performedin accordance with local institutional and Swedish national rulesand regulations.

2.3.1. Baseline studiesThe animals were administered 4.7 ± 0.8 MBq [11C]FTRP or

14.4 ± 4.0 MBq [11C]HTP intravenously through the tail vein under

s of [11C]FTRP.

Table 1List of compounds co-incubated with [11C]FTRP and [11C]HTP in the cell binding assay.

Compound Final conc. (μM) Comment

5-HTP 340 Serotonin precursorFTRP 50 Fluorinated serotonin precursorBCH 100 Large Amine Transporter-1 inhibitorCarbidopa 10 Selective Dopa Decarboxylase (DDC) inhibitorHarmine 340 Selective Monoamine Oxidase-A (MAO-A)

inhibitorTetrabenazine 40 Selective Vesicular Monoamine Transporter 2

(VMAT2) inhibitor

Each of the compounds interacts with enzymes involved in the serotonin metabolicpathway in different ways and their effect on [11C]FTRP or [11C]5-HTP can be used togain insight into the biological mechanisms for cellular uptake.

569O. Eriksson et al. / Nuclear Medicine and Biology 40 (2013) 567–575

general anesthesia (isoflurane 3.0%). The animals were allowed towake up after tracer administration and were sacrificed after 30 or60 min by CO2.

2.3.2. Blocking studiesAnimals were pretreated by intravenous administration of 15 mg/

kg HTP, 15 mg/kg FTRP or 2 mg/kg Carbidopa. 10 min later, 3.1 ±0.2 MBq [11C]FTRP or 12.6 ± 3.1 MBq [11C]HTP was administered.The animals were then allowed to wake up andwere sacrificed by CO2

after 30 min.Organs were excised, weighed and measured for radioactivity in a

NaI well-counter (Uppsala Imanet AB, GE Healthcare, Uppsala,Sweden). The tissue uptake was corrected for tissue weight, totaladministered tracer dose and animal weight and expressed as aStandardised Uptake Value (SUV).

2.4. Animal PET xenograft imaging

Immuno-deficient Balb/c nu/nu mice (n = 4, 19.6 ± 2.5 g) wereinoculated by suspensions of INS-1 cells (14 × 106, n = 2) or QGP1cells (30 × 106, n = 2) at the frontal right flank. The tumor growthwas monitored visually and after approximately 4 weeks the animalswere examined in a dedicated small animal PET–CT camera, witheither [11C]FTRP (1.3–3 MBq) or [11C]HTP (3.75–8 MBq). One animalcarrying an INS-1 tumorwas examined by both [11C]HTP (day 28) and[11C]FTRP (day 31). Tracer was administered in a tail vein in amaximum volume of 100 μl under general isoflurane (1.5%–2.0%)anesthesia. The animals were examined in vivo for 10 min by an 8 cmField of View (FOV) Triumph Trimodality animal PET–CT–SPECTscanner (Gamma Medica Inc, Northridge, CA, USA) approximately20 min following tracer administration.

The PET data were reconstructed as static images using aMLEM 3Dalgorithm (10 iterations). μPET data were analyzed using PMOD v3.13(PMOD Technologies Ltd, Zurich, Switzerland). Volumes of Interest(VOIs) were drawn manually over the xenograft tumor and overmuscle tissue in the opposing flank.

The murine experiments were approved by the local EthicsCommittee for Animal Research (C51/10) and performed inaccordance with local institutional and Swedish national rulesand regulations.

2.5. Non-human primate PET–CT imaging

The non-human primate experiment was carried out in accor-dancewith the guidelines of the Uppsala University andwas approvedby the local ethics committee for animal research (C160/11). ACynomolgus monkey (n = 1, 4900 g) was sedated by intravenousKetamine (10 mg/kg) and Propofol before being intubated andconnected to a respirator. After intubation the animal was maintainedon 1.4%–4% sevoflurane inhalation anesthesia and artificial ventila-tion. Body temperature, heart rate, ECG, pCO2, pO2, SaO2 and blood

pressureweremonitored throughout the study. Two venous catheterswere applied, one for tracer and drug administration and one forsampling of blood samples for measurements of radioactivity.

The animal was positioned to include the abdomen in the 15 cmaxial field of view of a Discovery ST PET–CT scanner (GE Healthcare,Milwaukee, MI, USA) by assistance of a low dose CT scout view(140 kV, 10 mA). Attenuation correction was acquired by a 140 kV,Auto mA 10–80 mA CT examination.

4 examinations were performed, consisting of both baseline andblocking studies of [11C]HTP or [11C]FTRP. The setup consisted of a [11C]HTP (7.6 MBq/kg) baseline scan at 10:00 and a [11C]FTRP (9.3 MBq/kg)baseline scan at 12:00. At 14:00, 3 mg/kg carbidopa (Karbidopa,Apoteket AB, Stockholm, Sweden) dissolved in 0.9% NaCl was admin-istered intravenously, to investigate to effect on tissue kinetics byinhibiting DDC. At 14:00, immediately following injection of carbidopa,[11C]FTRP (6.2 MBq/kg) was administered again for a blocking scan. At16:00 [11C]HTP (7.9 MBq/kg) was administered for a blocking scan.

For all examinations [11C]HTP or [11C]FTRP was administeredintravenously and the animal was examined by a dynamic PETprotocol for 90 min (33 frames; 12 × 10 s, 6 × 30 s, 5 × 120 s,5 × 300 s, 5 × 600 s), followed by a 3 bed static whole bodyexamination. Venous blood samples (0.2–0.5 ml) were collected at0.5, 1, 3, 5, 10, 15, 20, 30, 45, 60 and 90 min after each injection tomeasure the radioactivity concentration in whole blood and for selectsamples (5, 30 and 60 min) plasma.

2.6. Image analysis

Image acquisition is performed in 3D and reconstructed using aniterative OSEM VUEPOINT algorithm (2 iterations/21 subsets), ina128 × 128 matrix, zoom 50 cm diameter. Reconstructed data wereanalyzed by PMOD (PMOD Technologies Ltd., Zurich, Switzerland).Regions of Interest (ROIs) were drawn on co-registered CT images(pancreas and liver), or on early PET image summations (kidneycortex and aorta). Entire organs were measured as a Volume ofInterest (VOI) when possible.

Rate constants governing the radiotracer accumulation in tissues wasassessed by a 1-tissue compartmentmodel using venous plasma samplesas input (PKIN module; PMOD Technologies Ltd., Zurich, Switzerland).

2.7. Statistical considerations

Data on group level are reported as means ± SEM. Statisticalanalysis was performed in GraphPad Prism 5 (GraphPad, La Jolla, CA,USA) and differences between groups were assessed by the studentst-test using a significance level of p b 0.05.

3. Results

3.1. Radiosynthesis

[11C]FTRP (n = 8) was obtained with radiochemical purity higherthan 95%, in most batches exceeding 99%, and with a specificradioactivity of 29.2 ± 2.0 GBq/μmol. The total production time forthe 7-step chemo-enzymatic synthesis of [11C]FTRP, includingpurification, formulation and sterile filtration required 60–70 min.The non-optimised non-decay corrected radiochemical yield wasapproximately 1% giving 100–200 MBq of [11C]FTRP, ready forintravenous administration. The total mass of FTRP in each batchwas less than 2 μg and no chemical impurities exceeded the mass ofthe labelled compound.

3.2. In vitro cell binding

The cellular uptake of [11C]FTRP in INS-1, QGP1 and H720 wasstrongly inhibited by co-incubation with 5-HTP and to a lesser extent

570 O. Eriksson et al. / Nuclear Medicine and Biology 40 (2013) 567–575

carbidopa (Fig. 2). In INS-1 cells, unlabeled FTRP and BCH decreasedthe uptake similarly. Harmine and TBZ decreased the uptakemoderately (25%–50%) in all three cell lines.

5-HTP, harmine, TBZ and BCH had comparable inhibitory effectsfor [11C]HTP as compared to [11C]FTRP, but the effect of carbidopa wasincreased while the effect of FTRP was decreased.

The competition between [11C]FTRP and 5-HTP was assessed overtracer concentrations from 3 to 350 nM, using a direct comparisonwith [11C]HTP (Fig. 3). Saturation mechanisms were evident for both[11C]FTRP and [11C]HTP in INS-1, QGP1 and H720 cell lines. Theabsolute uptake of [11C]FTRP was markedly higher than for [11C]HTP,especially in INS-1 and QGP1 cell lines, as assessed by estimation of Kd

and Bmax.

3.3. Ex vivo organ distribution studies in rats

The absolute uptake (measured by SUV) of [11C]FTRP 30 minfollowing tracer administration was markedly increased compared to[11C]HTP, especially in pancreas (Fig. 4). After 60 min, the increasewas even more pronounced. However, the ratio tissue to blood wassimilar between [11C]HTP and [11C]FTRP and the higher uptake of[11C]FTRP in pancreas could be explained by a higher tracerconcentration in blood.

In vivo inhibition of DDC has been shown, somewhat counter-intuitively, to increase [11C]HTP uptake in tissues facilitating seroto-nin, such as NETs [15,16]. This can be explained by the substantialincrease in native [11C]HTP found in blood plasma, now availablefor tissue distribution, after inhibition of DDC in the kidney cortexwhere the major [11C]HTP degradation and excretion occur (the DDC-HTP effect).

Here we show that this effect is pronounced in rat pancreas, where[11C]HTP uptake in pancreas is increased by pretreatment using either

Fig. 2. Comparison of uptake of [11C]FTRP (left) and [11C]5-HTP (right) inneuroendocrine cell lines INS-1 (top panels), QGP-1 (middle) and H720 (bottom).The bars show uptake of tracer alone, or the effect on tracer uptake after co-incubationwith unlabeled 5-HTP, carbidopa, harmine or tetrabenazine (TBZ). Stars indicatesignificant difference from baseline.

5-HTP or carbidopa (Fig. 5). As expected, the concentration of tracer inblood is increased in parallel, and the escalated uptake in pancreas isless pronounced when normalizing for this fact. Pretreatment by15 mg/kg FTRP, on the other hand, has no discernible effect on either[11C]HTP or [11C]FTRP biodistribution measured as SUV.

Conversely, pretreatment by 5-HTP or carbidopa had no effecton [11C]FTRP tissue distribution, measured as SUV or tissue-to-blood ratio.

3.4. Animal PET–CT xenograft imaging

Both INS-1 and QGP-1 xenograft tumors were readily visualized invivo by [11C]HTP and PET–CT (Fig. 6). The absolute uptake of [11C]FTRP in tumors was comparable to [11C]HTP, but the backgrounduptake in reference tissues was increased, resulting in reduced tumor-to-background contrast for especially INS-1 but also QGP1.

3.5. Cynomolgus PET/CT imaging

Baseline [11C]FTRP biodistribution in abdominal tissues such aspancreas, liver and kidney followed a distinctly different patternthan from [11C]HTP (Fig. 7). [11C]HTP is characterized by asignificant accumulation in the pancreas during the initial 10–20 min followed by washout and renal excretion (Fig. 8). [11C]FTRPon the other hand does not accumulate in any abdominal tissuesduring the first 20 min, except for the early perfusion phase. Afterthis point, uptake in blood and liver is constant, while the uptake inpancreas increases continuously and irreversibly for the duration ofthe examination. Additionally, renal excretion is markedly reducedfor [11C]FTRP.

Pretreatment by carbidopa has previously been shown to inhibitthe conversion of [11C]HTP into [11C]serotonin in endocrinepancreas in both non-human primates [6] and humans [16] in vivo.Here, we show that inhibition of DDC by carbidopa has a negligibleeffect on the pancreatic uptake of [11C]FTRP (Figs. 7 and 8). As adirect comparison in the same animal, [11C]HTP pancreaticuptake was almost completely abolished following intravenouscarbidopa administration.

Volume of distribution (vT) of [11C]FTRP in pancreas, as assessedby a 1-tissue compartment model, was similarly not decreased byinhibition of DDC (Table 2). By contrast, vT of [11C]HTP was more thanhalved following carbidopa administration. Baseline vT was higher for[11C]FTRP compared to [11C]HTP, which can be mainly attributed tothe lower estimation of k2 (tissue elimination of radiotracer) ratherthan increase in K1 (tissue accumulation).

4. Discussion

Here, we present evidence that radiolabeled FTRP is a functionalanalogue of the neuroendocrine biomarker [11C]HTP in vitro.However, baseline and displacement studies in rats, xenograftcarrying mice and non-human primate suggest that in vivo [11C]FTRP uptake is mediated by mechanisms other than the serotonergicmetabolic pathway.

4.1. Radiochemistry

The synthesis method for [11C]HTP was used as such without anymodifications apart from exchanging 5-hydroxyindole for 5-fluor-oindole and the amount of [11C]FTRP produced was sufficient for theintended purpose. The key enzymatic step in the synthesis of [11C]HTPand [11C]TRP has been shown to proceed stereospecifically [8] andthere is no reason to believe that a fluoro substituent in the 5-position,not actively participating in the enzymatic coupling between 11C-pyruvic acid and the indol, should have any impact on thestereoselectivity. This is further strengthened by the observation

Fig. 3. Uptake of [11C]FTRP (left panels) and [11C]5-HTP (right panels) in different neuroendocrine cell lines at increasing concentrations of radiotracer. Non-specific (or uptakeunrelated to the serotonergic system) was assessed by co-incubation with 5-HTP in excess. The magnitude of uptake was higher for [11C]FTRP compared to [11C]5-HTP, especially inINS-1 and QGP1.

571O. Eriksson et al. / Nuclear Medicine and Biology 40 (2013) 567–575

that enantiomeric purity of 11C-4-methyl-tryptophan, using 4-methylindole and the same enzymatic synthesis route, was shownto proceed stereo-specifically [14]. Since the aim was to evaluate theusefulness of FTRP as a substitute for HTP and not develop it as a tracerfor clinical use no further optimisations were performed.

4.2. In vitro studies

Cell binding studies in three different NET cell lines indicate that[11C]FTRP has affinity for serotonergic metabolism in the same rangeas [11C]HTP, as assessed by competition with unlabeled 5-HTP inexcess. In especially INS-1 and QGP1 cells, [11C]FTRP had asignificantly increased Bmax(9 and 6 times, respectively) comparedto [11C]HTP. Additionally, both tracers responded similarly to co-incubation with inhibitors of LAT-1, DDC or VMAT2. These results allpoint to that [11C]FTRP is facilitated in a similar manner to [11C]HTP inneuroendocrine cell lines in vitro.

Curiously, uptake of [11C]FTRP or [11C]HTP was increased, ratherthan decreased, following inhibition of MAO-A, the expected mainroute of degradation. This indicates either that other modes ofdegradation of 5-HT and F-HTmay be present in these cell lines or thatthe majority of tracer rapidly is stored in secretory vesicles, protectedfrom degradation by MAO-A.

4.3. In vivo biodistribution

Absolute [11C]FTRP uptake accumulated in most tissues to a higherdegree than compared to [11C]HTP, both after 30 and 60 min. Thislikely reflects increased free and non-specific uptake of tracer in bloodand tissue, as in vivo inhibition of conversion of 5-HTP to serotonin byDDC had no effect on [11C]FTRP.

As mentioned above, pretreatment of either carbidopa orunlabeled 5-HTP in excess seems to increase, rather than decrease,uptake of [11C]HTP in some endocrine tissues, in this case pancreas.This is likely due to depletion of [11C]HTP due to excessive urinaryexcretion during baseline conditions. Studies in non-human primate[6] and in humans [16] show that [11C]HTP is significantly affected bycarbidopa pretreatment, resulting in markedly increased NET uptakeand subsequently improved tumor-to-background contrast. Carbi-dopa is today administered orally prior to clinical examination by[11C]HTP-PET/CT due to known or suspected NETs. The fact that [11C]FTRP biodistribution is unaffected by carbidopa thus indicatesdiminished interaction with DDC.

However, the first alternative seems more likely, as FTRP in excessdid not influence [11C]HTP biodistribution in a manner comparable to5-HTP or carbidopa, which further indicates that no or only modestcompetition between HTP and FTRP occurs in vivo.

Fig. 5. Effect of carbidopa, 5-HTP or FTRP pre-treatment on [11C]FTRP and [11C]5-HTP biodistribution. [11C]FTRP was affected by neither of the compounds (top right panel), asopposed to [11C]5-HTP which exhibited increased uptake in most tissues including pancreas following administration of carbidopa or 5-HTP (top right panel). However, FTRP had noeffect on [11C]5-HTP, indicating negligible competition for enzymatic processes. When correcting for radioactivity in blood (bottom panels), the effect of carbidopa and 5-HTP on[11C]5-HTP diminished. Stars indicate significant difference in treatment group compared to baseline.

Fig. 4. Comparison between ex vivo organ distribution of [11C]FTRP and [11C]5-HTP. The absolute tissue uptake in pancreas, liver and heart of [11C]FTRP was higher than [11C]5-HTPafter both 30 and especially 60 min post tracer administration (top panels). However, the difference between the tracers is diminished when correcting for tracer concentration inblood at respective time-point (bottom panels). Stars indicate significant difference between [11C]FTRP and [11C]5-HTP within groups.

572 O. Eriksson et al. / Nuclear Medicine and Biology 40 (2013) 567–575

Fig. 6. (A) Uptake in neuroendocrine tumor xenografts of [11C]FTRP and [11C]HTP. The sensitivity for visual detection of both INS-1 (top panels) and QGP1 (bottom panels) cell linetumors is slightly decreased for [11C]FTRP due to increased background uptake compared to [11C]HTP. (B) Contrast as measured by tumor-to-muscle ratio in PET images.

573O. Eriksson et al. / Nuclear Medicine and Biology 40 (2013) 567–575

To investigate whether [11C]FTRP accumulation in pancreaswas anactive process mediated by LAT-1 (as shown by cell binding studies)we pretreated rats by 1–10 mg/kg BCH, a selective LAT-1 inhibitor(data not shown). This had no effect on biodistribution of either [11C]FTRP or [11C]HTP, but this likely reflects that LAT-1 cannot besaturated at the administered doses rather than that LAT-1 is notinvolved in active uptake of [11C]FTRP or[11C]HTP. Further studiesexplaining the role of LAT-1 in endocrine tissue uptake of [11C]FTRP invivo are therefore required.

4.4. In vivo xenograft imaging

Despite having an increased Bmax in INS-1 and QGP1 cell linescompared to [11C]HTP, [11C]FTRP did not have improved tumor-to-

Fig. 7. Cynomolgus biodistribution in the abdominal tissues of [11C]FTRP and [11C]HTP as meblocking uptake in endocrine pancreas by inhibiting the enzyme DDC by carbidopa (top anuptake was not mediated by DDC (bottom left) and therefore not related to the decarboxylaspecific background uptake (for example hepatic and splenic uptake) was higher than fonormalized to SUV = 6.

background contrast in xenografts originating from such cells. Thismay partially be explained by the increased retention of [11C]FTRP inblood as well as non-serotonin facilitating reference tissues seen inthe organ distribution studies, combined with modest or no increasein tumor uptake.

4.5. Cynomolgus studies

The results seen in rodents were confirmed also in a non-humanprimate when examined by PET–CT. As previously shown, [11C]HTPuptake in endocrine pancreas was mediated by DDC, as shown byselective inhibition of carbidopa. [11C]FTRP uptake in pancreas, on theother hand, was not affected by inhibition by DDC. Still substantialaccumulation in pancreas was found in both baseline and blocking

asured by PET–CT. [11C]HTP acts as a precursor for [11C]serotonin, which was shown byd bottom right panels). [11C]FTRP also accumulated in the pancreas (top left), but thetion step in the serotonin metabolic pathway, contrary to [11C]HTP. Generally, the non-r [11C]HTP. Images are summations 10–90 min following tracer administration and

Fig. 8. [11C]FTRP and [11C]5-HTP kinetics in cynomolgus monkey. The uptake of [11C]HTP in tissues expressing DDC (endocrine pancreas and kidney cortex), but not DDC-negativeliver, can be modulated by inhibition of DDC (top and bottom right panels). [11C]FTRP uptake in pancreas follows a distinctly different time-dependant uptake in pancreas, includinga perfusion phase followed by steady accumulation from 20 min and onward (top left panel). This accumulation was not mediated by DDC (bottom left). [11C]FTRP is not excretedinto kidneys and the urinary tract to the same extent as [11C]HTP, and is instead excreted into the biliary ducts resulting in a higher hepatic background signal. The FTRP uptake inpancreas is clearly different from liver and kidney cortex and the accumulation in the pancreas thus indicates an active process.

574 O. Eriksson et al. / Nuclear Medicine and Biology 40 (2013) 567–575

studies.This could potentially be explained by a higher bloodconcentration of FTRP as compared with HTP due to a lower excretionrate and slower metabolism.

The increased uptake in non-serotonin facilitating tissues such asliver is possibly non-specific due to the higher lipophilicity of [11C]FTRP. For the same reason, it is likely that [11C]FTRP tends to beexcreted through the biliary rather than the renal system, which is theroute of excretion for [11C]HTP. The increased background uptake intissues where NETs metastasize (e.g. liver) significantly reduces thesensitivity for detection of small metastasized lesions with [11C]FTRP.

To further analyse and explain the differences between [11C]FTRPand [11C]HTP uptake in especially pancreas the per cent intact tracerin plasma needs to be determined. However, in this study we are notaiming for a full kinetic model describing the FTRP tissue uptake, andthe main objective is instead to evaluate the usefulness of [11C]FTRPas a substitute for [11C]HTP in neurology, oncology and diabetes. TheFTRP uptake in pancreas is clearly different from liver and kidneycortex and during the course of the study, apparent irreversiblekinetics of accumulation in the pancreas indicate an active process.This is in line with earlier studies with racemic 5- and 6-[18F]fluorotryptophan [12]. The mechanism for this accumulation is to

Table 2Rate constants and compound parameter vT for [11C]FTRP and [11C]5-HTP eitheradministered alone or following pretreatment of DDC-inhibitor carbidopa, as assessedby a 1-tissue compartment model.

Unit [11C]FTRP [11C]HTP

Baseline Block Baseline Block

K1 ml/ccm/min 0.05 0.07 1.21 0.04k2 1/min 0.002 0.001 0.19 0.014vT ml/ccm 32.1 86.4 6.4 2.7Chi Sq 22.3 26 9.9 11.2

our best knowledge not known, but it can be hypothesized to berelated to cellular accumulation through LAT-1, given the structuralsimilarities between FTRP and HTP. Furthermore, a simple 1-tissue compartment model demonstrates slower tissue eliminationof [11C]FTRP than [11C]HTP (lower k2), indicating some form ofmetabolic trapping.

5. Conclusion

[11C]FTRP retains in vitro biological function similar to that of 5-HTP. However, this functionality is lost in vivo. The in vivo results,especially the non-human primate study, are based on few animals,and generalization should therefore be approached with caution.Radiolabeled FTRP tracers are based on these results not likely toprovide an advantage over [11C]HTP in PET imaging of the serotoner-gic metabolic pathway in oncology, neurology or diabetes. Instead, itmay possibly be used for imaging of LAT-1.

Acknowledgments

INS-1 cells were obtained as a kind gift fromHans Hohmeier (DukeUniversity Medical Center). QGP1 and H720 cell lines were obtainedas a kind gift from Prof. Barbro Eriksson (Uppsala University). Theauthors wish to thank Dr. Marie Karlsson for excellent technicalassistance. OE is supported by EXODIAB (Excellence of DiabetesResearch in Sweden) and Barndiabetesfonden, and RS’s position issupported by the VINNOVA foundation.

References

[1] Reibring L, Agren H, Hartvig P, Tedroff J, Lundqvist H, Bjurling P, et al. Uptake andutilization of [beta-11C]5-hydroxytryptophan in human brain studied by positronemission tomography. Psychiatry Res 1992;45(4):215–25.

575O. Eriksson et al. / Nuclear Medicine and Biology 40 (2013) 567–575

[2] Lundquist P, Blomquist G, Hartvig P, Hagberg GE, Torstenson R, Hammarlund-Udenaes M, et al. Validation studies on the 5-hydroxy-L-[beta-11C]-tryptophan/-PET method for probing the decarboxylase step in serotonin synthesis. Synapse2006;59(8):521–31, http://dx.doi.org/10.1002/syn.20268.

[3] Orlefors H, Sundin A, Garske U, Juhlin C, Oberg K, Skogseid B, et al. Whole-body(11)C-5-hydroxytryptophan positron emission tomography as a universalimaging technique for neuroendocrine tumors: comparison with somatostatinreceptor scintigraphy and computed tomography. J Clin Endocrinol Metab2005;90(6):3392–400, http://dx.doi.org/10.1210/jc.2004-1938 [pii].

[4] Di Gialleonardo V, Signore A, Scheerstra EA, Visser AK, van Waarde A, Dierckx RA,et al. 11C-Hydroxytryptophan uptake and metabolism in endocrine and exocrinepancreas. J Nucl Med 2012, http://dx.doi.org/10.2967/jnumed.112.104117 [pii].

[5] de Heide LJ, Glaudemans AW, Oomen PH, Apers JA, Totte ER, van Beek AP.Functional imaging in hyperinsulinemic hypoglycemia after gastric bypasssurgery for morbid obesity. J Clin Endocrinol Metab 2012;97(6):E963–7, http://dx.doi.org/10.1210/jc.2011-3063 [pii].

[6] Eriksson O, Espes D, Selvaraju R, Sörensen J, Lubberink M, Antoni G, et al. In vivoimaging of beta cell mass by [11C]5-hydroxy-L-tryptophan PET Presented at 48thEASD Annual Meeting; 2012. p. OP261.

[7] Ohta Y, Kosaka Y, Kishimoto N, Wang J, Smith SB, Honig G, et al. Convergence ofthe insulin and serotonin programs in the pancreatic beta-cell. Diabetes2011;60(12):3208–16, http://dx.doi.org/10.2337/db10-1192 [pii].

[8] Bjurling P, Watanabe Y, Tokushige M, Oda T, Långström B. Synthesis of B-11C-labelled L-tryptophan and 5-hydroxy-L-tryptophan using a multi-enzymaticreaction route. J Chem Soc Perkin Trans 1989;1(1):1331–4.

[9] Jager PL, Chirakal R, Marriott CJ, Brouwers AH, Koopmans KP, Gulenchyn KY.6-L-18F-fluorodihydroxyphenylalanine PET in neuroendocrine tumors:

basic aspects and emerging clinical applications. J Nucl Med 2008;49(4):573–86, http://dx.doi.org/10.2967/jnumed.107.045708 [pii].

[10] Hartvig P, Tedroff J, Lindner KJ, Bjurling P, Chang CW, Tsukada H, et al. Positronemission tomographic studies on aromatic L-amino acid decarboxylase activity invivo for L-dopa and 5-hydroxy-L-tryptophan in the monkey brain. J NeuralTransm Gen Sect 1993;94(2):127–35.

[11] Eriksson B, Bergstrom M, Sundin A, Juhlin C, Orlefors H, Oberg K, et al. The role ofPET in localization of neuroendocrine and adrenocortical tumors. Ann N Y Acad Sci2002;970:159–69.

[12] Atkins HL, Christman DR, Fowler JS, Hauser W, Hoyte RM, Klopper JF, et al. Organicradiopharmaceuticals labeled with isotopes of short half-life. V. 18 F-labeled 5-and 6-fluorotryptophan. J Nucl Med 1972;13(10):713–9.

[13] Kramer SD, Mu L, Muller A, Keller C, Kuznetsova OF, Schweinsberg C, et al. 5-(2-18F-fluoroethoxy)-L-tryptophan as a substrate of system L transport for tumorimaging by PET. J Nucl Med 2012;53(3):434–42, http://dx.doi.org/10.2967/j-numed.111.096289 [pii].

[14] Bjurling P. Synthesis of 11C-labelled compounds using multi-enzymatic catalysis(doctoral thesis). Acta Universitatis Upsaliensis 1989;225.

[15] Neels OC, Koopmans KP, Jager PL, Vercauteren L, van Waarde A, Doorduin J, et al.Manipulation of [11C]-5-hydroxytryptophan and 6-[18F]fluoro-3,4-dihydroxy-L-phenylalanine accumulation in neuroendocrine tumor cells. Cancer Res2008;68(17):7183–90, http://dx.doi.org/10.1158/0008-5472.CAN-08-0095 [pii].

[16] Orlefors H, Sundin A, Lu L, Oberg K, Langstrom B, Eriksson B, et al. Carbidopapretreatment improves image interpretation and visualisation of carcinoidtumours with 11C-5-hydroxytryptophan positron emission tomography. Eur JNucl Med Mol Imaging 2006;33(1):60–5, http://dx.doi.org/10.1007/s00259-005-1891-z.