Tracer development for detection and characterization of neuroendocrine tumors with PET Neels, Olivier Christiaan

Transcription

1 University of Groningen Tracer development for detection and characterization of neuroendocrine tumors with PET Neels, Olivier Christiaan IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below. Document Version Publisher's PDF, also known as Version of record Publication date: 2008 Link to publication in University of Groningen/UMCG research database Citation for published version (APA): Neels, O. C. (2008). Tracer development for detection and characterization of neuroendocrine tumors with PET s.n. Copyright Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons). Take-down policy If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim. Downloaded from the University of Groningen/UMCG research database (Pure): For technical reasons the number of authors shown on this cover page is limited to 10 maximum. Download date:

2 Tracer development for detection and characterization of neuroendocrine tumors with PET Oliver Christian Neels

3 This thesis was supported by grant of the Dutch Cancer Society. Financial support for printing this thesis was provided by: Von Gahlen B.V. Siemens Nederland N.V. Veenstra Instrumenten B.V. ISBN: Cover and lay-out designed by Henriette Kiss Printed by Ponsen & Looijen, Wageningen 2007 by Oliver Christian Neels. All rights are reserved. No parts of this book may be reproduced or transmitted in any form by any means without the permission of the author.

4

5

6 RIJKSUNIVERSITEIT GRONINGEN Tracer development for detection and characterization of neuroendocrine tumors with PET Proefschrift ter verkrijging van het doctoraat in de Medische Wetenschappen aan de Rijksuniversiteit Groningen op gezag van de Rector Magnificus, dr. F. Zwarts, in het openbaar te verdedigen op maandag 21 januari 2008 om uur door Oliver Christian Neels geboren op 22 november 1975 te Nordenham, Duitsland

7 Promotores: Prof. dr. E.G.E. de Vries Prof. dr. P.L. Jager Prof. dr. R.A.J.O. Dierckx Copromotores: Dr. P.H. Elsinga Dr. I.P. Kema Beoordelingscommissie: Prof. dr. ir. M. de Jong Prof. dr. J. Martens Prof. dr. P.H.B. Willemse ISBN:

8

9 Paranimfen: Bram Maas Klaas Pieter Koopmans

10

11

12 Contents Chapter 1 Introduction and outline of this thesis Molecular imaging in neuroendocrine tumours: molecular uptake mechanisms and clinical results. Klaas P. Koopmans, Oliver C. Neels, Ido P. Kema, Philip H. Elsinga, Thera P Links, Elisabeth G.E. de Vries, Pieter L. Jager. Submitted for publication 1 7 Chapter 2 Development of a reliable remote-controlled synthesis of β-[ 11 C]-5-hydroxy-L-tryptophan on a Zymark robotic system. Oliver C. Neels, Pieter L. Jager, Klaas P. Koopmans, Elisabeth Eriks, Elisabeth G.E. de Vries, IP Kema, Philip H. Elsinga. J Label Compd Radiopharm 2006; 49: Chapter 3 Manipulation of [ 11 C]HTP and [ 18 F]FDOPA accumulation in neuroendocrine tumor cells. Oliver C. Neels, Klaas P. Koopmans, Pieter L. Jager, Laya Vercauteren, Aren van Waarde, Janine Doorduin, Hetty Timmer-Bosscha, Adrienne H. Brouwers, Elisabeth G.E. de Vries, Rudi A. Dierckx, Ido P. Kema, Philip H. Elsinga. Submitted for publication Chapter 4 Staging of carcinoid tumours with 18 F-DOPA PET: a prospective, diagnostic accuracy study. Klaas P. Koopmans, Elisabeth G.E. de Vries, Ido P. Kema, Philip H. Elsinga, Oliver C. Neels, Wim J. Sluiter, Anouk N.A. van der Horst-Schrivers, Pieter L. Jager. The Lancet Oncol 2006; 7:

13 Chapter 5 Chapter 6 Improved staging and characterization of lesions in patients with carcinoid and islet cell tumors with 18F-DOPA and 11C-5-HTP positron emission tomography. Klaas P. Koopmans, Oliver C. Neels, Ido P. Kema, Philip H. Elsinga, Wim J. Sluiter, Koen Vanghillewe, Adrienne H. Brouwers, Elisabeth G.E. de Vries, Pieter L. Jager. In press for J Clin Oncol 5-Fluorotryptophan as potential PET tracer for neuroendocrine tumors. Oliver C. Neels, Klaas P. Koopmans, Pieter L. Jager, Laya Vercauteren, Hetty Timmer-Bosscha, Adrienne H. Brouwers, Elisabeth G.E. de Vries, Rudi A. Dierckx, Ido P. Kema, Philip H. Elsinga. Submitted for publication Summary and Future Perspectives Samenvatting / Zusammenfassung Dankwoord Appendix full-color figures

14

15

16 Introduction and outline of this thesis 1

17 Introduction Neuroendocrine tumors are rare and often slowly growing tumors originating from neuroendocrine cells. These tumors frequently have the ability to secrete peptides and hormones such as serotonin and catecholamines, which can cause symptoms. Best known is the overproduction of serotonin resulting in symptoms such as flushing, diarrhea and rightsided heart disease. These amines are produced in neuroendocrine cells by uptake and subsequent decarboxylation of amine precursors (APUD) such as 5-hydroxytryptophan (5- HTP) or levodopa (L-DOPA) in neuroendocrine cells. Several diagnostic methods for imaging neuroendocrine tumors have been developed in the past years. Therapy can be monitored by morphological techniques such as computed tomography (CT) or magnetic resonance imaging (MRI). However, given specific characteristics of neuroendocrine tumors functional imaging techniques such as somatostatin receptor scintigraphy (SRS) and positron emission tomography (PET) are also of interest as they image respectively the somatostatin receptor expression and intracellular metabolism. In oncology, [ 18 F]fluorodeoxyglucose (FDG) is used as very sensitive but non-specific PET tracer. However, FDG rarely shows sufficient uptake in neuroendocrine tumors. Based on the APUD principle, PET tracers like 6-[ 18 F]fluoro-levodopa ([ 18 F]FDOPA) and [ 11 C]-5- hydroxytrytophan ([ 11 C]HTP) have been developed because of the capacity of neuroendocrine tumor cells to have a high accumulation of these tracers. Only a few clinical studies with these tracers have been performed and little is known about the precise mechanisms involved in the accumulation of these tracers. Because of their slow growth and non-specific symptoms, neuroendocrine tumors are often already metastasized at diagnosis. Proper staging is important to make the right treatment decisions. PET tracers [ 11 C]HTP and [ 18 F]FDOPA have the potential to improve the yield of diagnostic imaging as they are precursors for respectively the serotonin and catecholamine pathways and can therefore be used to obtain knowledge on the biochemical behavior of neuroendocrine tumors. Following uptake of L-DOPA and 5-HTP by large amino transporters (LAT) into tumor cells they are decarboxylated by amino acid decarboxylase (AADC) to their corresponding amines. These amines are then stored into cellular vesicles via the vesicular monoamine transporter (VMAT). After release they are metabolized by the enzyme monoamine oxidase (MAO) to 5-hydroxyindole acetic acid, respectively homovanillic acid and subsequently excreted. Exploring the metabolic pathways with tracers visualizing these pathways could lead to a better understanding of biochemical mechanisms in neuroendocrine tumors. Furthermore the tumor image quality and the sensitivity of tumor detection might be improved by the use of these PET tracers. PET allows visualization of metabolic pathways by incorporating a positron emitting radionuclide into a molecule taking part in biochemical and physiological processes. These radionuclides can be produced by irradiation of target material with highly accelerated protons or deuterons using a cyclotron. Commonly used radionuclides in PET imaging are carbon-11, nitrogen-13, oxygen-15 and fluorine-18 with relative short half-lives varying from 2 to 110 minutes. Radioactive decay takes place through emission of a positron. A positron has the same mass as an electron but carries the opposite charge. Radiochemical synthesis is used to incorporate the radionuclide into a molecule which is then defined as tracer. Important aspects for radiopharmaceutical preparations are: 1) radiochemical yield, 2) reliability of the production, 3) reaction time and 4) choice of the radionuclide. 2

18 Introduction and outline of this thesis Following administration to a patient, the positron interacts after a short distance with an electron from tissue resulting in annihilation. During the annihilation the masses of the positron and electron are converted into energy according to Einstein s formula E=mc 2. The energy appears in the form of two photons with an energy of each 511 kev emitted under an angle of 180 from each other. The PET camera consists of a full ring of multiple detectors. Two detectors that are exactly opposite to each other will detect these photons. This allows coincidence detection of two simultaneously emitted photons within 10 nanoseconds. After injection of the radiopharmaceutical, the distribution can be followed in time in a part of the body (dynamic scan) or by moving the bed into several positions (static whole-body scan). PET shows a higher resolution than conventional gamma cameras. While human PET cameras currently have a resolution of 4-5 mm, so-called micropet cameras used for small animal imaging reach a resolution of 1-2 mm. Aim and outline of this thesis The aim of this thesis is to study the development, biochemical behavior and value of new PET tracers for imaging of neuroendocrine tumors. In chapter 1, a literature overview is presented on uptake mechanisms of tracers used in nuclear medicine to detect neuroendocrine tumors. Different detection methods and their diagnostic values for several subtypes of neuroendocrine tumors are reviewed. In order to visualize the serotonin pathway the availability of [ 11 C]HTP was required. In chapter 2 the enzymatic synthesis of L-[ 11 C]HTP on a Zymark robotic system is described. The aim was to optimize the synthesis of enantiomerically pure L-[ 11 C]HTP and to obtain radiochemical yields reliable for patient studies. The origin of used enzymes was analyzed for safer human use. For routine production the radiation exposure for the radiochemist had to be minimized. Thereafter [ 11 C]HTP was applied to study its diagnostic value in patients with neuroendocrine tumors and to get more insights in the biochemical behavior of neuroendocrine tumor cells. In chapter 3 in vitro and small animal studies were performed using the PET tracers [ 11 C]HTP and [ 18 F]FDOPA to analyze tracer metabolism and accumulation. Several inhibitors of LAT, AADC and MAO were administered to a human neuroendocrine tumor cell line to obtain knowledge on the biochemical behavior. The influence of the decarboxylase inhibitor carbidopa on tumor uptake was studied in an animal model using a micropet camera to get a better understanding of the in vivo metabolism. The aim of the study described in chapter 4 was to determine the diagnostic value of [ 18 F]FDOPA for imaging patients with carcinoid tumors. Fifty-three patients with a metastatic carcinoid underwent a [ 18 F]FDOPA PET scan. The diagnostic sensitivity was compared with combined SRS and CT. Biochemical parameters of the catecholamine and serotonin pathways were compared with the used imaging methods. As no data are available with a head to head comparison of [ 11 C]HTP and [ 18 F]FDOPA PET imaging we performed the study described in chapter 5. The diagnostic value of [ 11 C]HTP and [ 18 F]FDOPA PET in patients with neuroendocrine tumors was evaluated. 24 patients with carcinoid tumors and 23 patients with islet cell tumors were studied with both PET tracers and CT and SRS scan. Currently several PET tracers for imaging of the catecholamine pathway are available. The serotonin pathway can be visualized with [ 11 C]HTP only. Due to the short half-life of this 3

19 tracer, its use is restricted to centers with cyclotron facilities. Therefore we aimed to develop a tracer with a longer half-life and started a synthesis route for 18 F labeled tryptophan. First, the affinity of 5-fluorotryptophan to accumulate in a neuroendocrine cell line was investigated. The synthesis of a precursor towards an enzymatic route of [ 18 F]-5- fluorotryptophan and the development of an enzymatic synthesis route towards 5- fluorotryptophan using fluorodestannylation are described in chapter 6. Finally, a summary of the studies performed in this thesis and future perspectives are given in English, Dutch and German. 4

20 5

21 6

22 Chapter 1 Molecular imaging in neuroendocrine tumors: molecular uptake mechanisms and clinical results Klaas P. Koopmans 1, Oliver C. Neels 1, Ido P. Kema 2, Philip H. Elsinga 1, Thera P. Links 3, Elisabeth G.E. de Vries 4, Pieter L. Jager 1 Departments of Nuclear Medicine and Molecular Imaging 1, Clinical Chemistry 2, Endocrinology 3 and Medical Oncology 4, University of Groningen and University Medical Center Groningen, The Netherlands Submitted for publication 7

23 Chapter 1 Summary Neuroendocrine tumors can originate almost everywhere in the body and consist of a great variety of subtypes. This paper focuses on molecular imaging methods using nuclear medicine techniques in neuroendocrine tumors, coupling molecular uptake mechanisms of radiotracers with clinical results. A non-systematic review is presented on receptor based and metabolic imaging methods. Receptor-based imaging covers the molecular backgrounds of somatostatin, vaso-intestinal peptide (VIP), bombesin and cholecystokinin (CCK) receptors and their link with nuclear imaging. Imaging methods based on specific metabolic properties include metaiodobenzylguanidine (MIBG) and dimercaptosuccinic acid (DMSA-V) scintigraphy as well as more modern positron emission tomography (PET) based methods using radiolabeled analogues of amino acids, glucose, dihydroxyphenylalanine (DOPA), dopamine and tryptophan. Diagnostic sensitivities are presented for each imaging method and for each neuroendocrine tumor subtype. Finally, a Forest plot analysis of diagnostic performance is presented for each tumor type in order to provide a comprehensive overview for clinical use. Introduction Neuroendocrine tumors are unique and rare tumors originating from neuroendocrine cells. These neuroendocrine cells are postulated to arise from common precursor cells of the embryologic neural crest and are dispersed throughout the human body. Characteristic is a common phenotype consisting of the simultaneous expression of general protein markers of neuroendocrine cells and hormonal products specific to each cell type 1. The main function of neuroendocrine cells is to regulate a large variety of body functions through paracrine action with dedicated amines and peptides, of which the biogenic amines, such as serotonin and catecholamines, are most prominent. In order to be able to synthesize these amines, neuroendocrine cells have the ability to take-up and decarboxylate amine precursors (APUD; amine precursor uptake and decarboxylation). Other biogenic amines, substances such as adrenocorticotrophic hormone, growth hormone, neuropeptide K, substance P, bradykinin, kallikrein and prostaglandins can also be secreted 2,3. Neuroendocrine tumors arise in nearly every organ but primary sites in gastrointestinal (56%) and bronchopulmonary (12%) tracts are most frequent 4. Due to the slow growth of most neuroendocrine tumors and the long time span between the onset of symptoms, many patients present with metastases (figure 1). Even with advanced disease, patients may survive for many years. However, there are also subtypes, which behave more aggressively. The diagnosis is based on histology and can be considered when specific symptoms induced by tumor products are present, such as diarrhea or flushing. Apart from clinical symptoms, determination of tumor secretory products using biochemical assays can assist in obtaining a diagnosis. A division based on the embryological origin of the organs in which neuroendocrine tumors arise has been made in the past. This division classifies these tumors as foregut, midgut and hindgut tumors. However, the recent World Health Organization (WHO) classification of the different subtypes of neuroendocrine tumors of gastro-intestinal and pancreatic origin is based on histopathologic characteristics consisting of cellular grading, primary tumor size, primary tumor localization, proliferation markers, degree of invasiveness and the production of biologically active substances. The main categories defined by this 8

24 Molecular imaging in neuroendocrine tumors: molecular uptake mechanisms and clinical results classification are the well differentiated endocrine tumors with a low grade of malignancy, well differentiated, more aggressive carcinomas, poorly differentiated endocrine carcinomas with a high grade of malignancy and a poor prognosis and finally mixed exocrine endocrine tumors. In this WHO classification the term carcinoid is abandoned and replaced by neuroendocrine tumor 5. Currently, there is a broad variety of treatment options for patients with neuroendocrine tumors. Surgery is the only curative option. When cure is not possible, palliative treatment aims to control symptoms, maintain local tumor control and prolong life. Palliative procedures include surgical debulking, correction of bowel obstruction, chemoembolizations, or systemic treatment using interferon-α or somatostatin analogues 2. Other treatment options are chemotherapy and radionuclide therapy. New targeted therapies such as anti-angiogenic therapies are currently being explored 4,6,7. Figure 1. Full-color in appendix. Metastasized carcinoid. 18 F-DOPA PET scan (A), Octreotide scan (B) and 18 F-DOPA PET-CT fusion image (C) of a female patient presenting with a metastasized carcinoid. This patient illustrates the intra-individual heterogeneity in the uptake of different tracers by tumor metastases. Accurate localization of tumor lesions can guide treatment decisions. In general, radiological techniques such as CT, ultrasound or MRI are applied in staging and restaging, but also nuclear medicine techniques, such as somatostatin receptor scintigraphy (SRS), have proven to be of great value. The presence of somatostatin receptors on many neuroendocrine tumors was the driving force that enabled the development of SRS. Besides receptors, the remarkable metabolic activity of specific biochemical pathways used in substance synthesis provides possibilities for molecular nuclear medicine imaging. The unique characteristics of neuroendocrine tumors have led to the development of interesting new diagnostic methods for these tumors over the last years, both in receptor imaging and metabolic imaging. Especially the increase in positron emission tomography (PET) facilities allows developments of tracers for new molecular targets with a highresolution method for imaging. In addition new insights in the genetic, biochemical and metabolic aspects of subtypes of the neuroendocrine tumor family have arisen over the last years. It is therefore important to understand the receptor and metabolic targets for neuroendocrine tumors, as the molecular mechanisms that drive tracer uptake, translate into 9

25 Chapter 1 images and determine the final clinical applicability. The unique characteristics of neuroendocrine tumors are increasingly exploited to successfully enhance our imaging and therapeutical options for these tumors. Three directions can be seen in the development of new tracers for imaging use. The first uses tumor receptor expression, the second uses the metabolic properties and the last method uses antibodies. Therefore, the purpose of this review is to describe the receptor and metabolic imaging methods of neuroendocrine tumors and translate the molecular uptake mechanisms into clinical parameters such as sensitivity and specificity. For this purpose, a review of current literature is presented, both for imaging methods and for tumor types. Search strategy and selection criteria For this non-systematic review, a Medline and PubMed search was performed. Due to the many subtypes of neuroendocrine tumor and imaging methods a multitude of different search terms was necessary. A detailed list is available on request. Only papers with an English abstract published over the last 10 years (1995) were included. Material from review articles also referring to older studies was evaluated and was used if relevant. Reference lists of individual papers were also analyzed for study selection. Only studies from which a clear description of sensitivity or specificity for individual tumor subgroups could be derived were included. Data from all sources were sorted and divided over subcategories of tracer and tumor type. In general, studies with fewer than 10 subjects were excluded, however, due to the rarity of a number of tumor types, some of these reports or small studies were included. The number of patients included in the studies was used as a weight factor in summarizing results for different tracers and is represented by a square in the Forest plot analysis. The size of the square represents the weight that the studies exert in the analysis. Sensitivity values given in the figures denote a lesion-based sensitivity for the detection of all types of tumor deposits. Nuclear imaging methods The methods for nuclear imaging of neuroendocrine tumors can be divided in three main categories, namely tracers based on a) the selective expression of different receptors, b) metabolic properties of tumors and c) tracers which exploit antigens expressed by the tumors. For each category, currently available tracers will be described and their clinical impact. A schematic overview of the uptake mechanisms of these tracers can be found in figure 2. A general principle in nuclear medicine is that detectability of lesions primarily depends on the amount of tracer localized in a lesion, and only indirectly on the size of that lesion. In theory a 1 mm lesion can be detected as long as there is enough tracer uptake. On the images, such a lesion will result in a hot-spot with larger dimensions and vague borders. This is especially true for imaging with gamma cameras, such as SPECT, which have a considerable lower spatial resolution than methods based on PET. A prerequisite for molecular uptake mechanisms therefore is that the magnitude of tracer uptake is high and the background uptake low. In daily practice of commonly used tracer methods, these principles translate into a detection limit of 1-2 cm for conventional gamma camera imaging and cm for PET imaging. 10

26 Molecular imaging in neuroendocrine tumors: molecular uptake mechanisms and clinical results 99m Tc-V-DMSA Passive diffusion Active transport 11 C-5- HTP 18 F-DOPA Phosphate metabolism Serotonin Catecholamine pathway IMT Lysosome Nucleus Secretory vesicle Internalisation of receptorligand complex LAT 1 4F2HC complex transporter (family of LAT) Noradrenalin transporter NaPi co-transporter Glucose transporter VMAT transporter, located on secretory vesicle L-3-[ 123 I]-iodo-alphamethyltyrosine (IMT) Glucose metabolism Somatostatin receptor with (labelled) ligand Bombesin receptor with (labelled ligand) 18 F-Dopamine MIBG 18 FDG CCK receptor with (labelled) ligand VIP receptor with (labelled) ligand Antigen with (labelled) antibody Figure 2. Full-color in appendix. Metabolic pathways. In this figure the different metabolic pathways by which neuroendocrine tumors can be visualized using nuclear medicine imaging techniques are schematically depicted. Three major routes can be identified: receptor based techniques, techniques which use the metabolic properties of these tumors and labeled antibody based techniques. Receptor based imaging methods Somatostatin receptor imaging Currently somatostatin receptor imaging is the method of choice for the staging of neuroendocrine tumors 8. Somatostatin is a small regulatory peptide, which is widely distributed in the human body. Besides its function as a neurotransmitter in the hypothalamus, it has an inhibitory effect on the production of several exocrine hormones in the gastrointestinal tract and anti-proliferative effects 9. Somatostatin receptors are G protein coupled receptors on the cell membrane which recognize the ligand and generate a transmembrane signal. The resulting hormone-receptor complexes have the ability to be internalized. Once internalized, these vesicles fuse with lysosomes, resulting in hormone degradation or receptor recycling 10. Thus far, 6 different somatostatin receptors have been cloned, namely sst 1 -sst 5 (sst 2 can be alternately spliced to yield two products, sst 2A and sst 2B ). Neuroendocrine tumors frequently express a high density of somatostatin receptors, which is exploited by imaging techniques using somatostatin analogues. These analogues have been developed because somatostatin itself has a very short plasma half life (~ 3 min). For most somatostatin analogues internalization of the 111 In-octreotide complex with residualization of the 111 In label is the most likely mechanism accounting for the good scintigraphic tumor to background ratio observed 24 h after injection 11. Currently radiolabeled analogues of somatostatin such as octreotide, vapreotide and MK678, are in clinical use for imaging. All octreotide analogues bind with high affinity to sst2 and sst5 and with 11

27 Chapter 1 varying affinity to the sst3 and sst4 receptors. When chelators such as DTPA or DOTA are coupled to somatostatin analogues, these molecules can thereafter be labeled with for instance 111 In or 99m Tc for scintigraphic purposes. When labeled with positron emitting isotopes, such as 18 F, 64 Cu or 68 Ga the somatostatin analogues can be used for PET imaging 12,13. There are many new developments in newer better chelators (i.e. DOTA, EDDA or HYNIC) and analogues with more rapid internalization properties 14. Based on the high receptor expression, somatostatin receptor imaging using 111 In-octreotide provides important information on tumor localizations of many neuroendocrine tumors. SRS is now widely available, and yields the best results in paragangliomas and neuroendocrine gastrointestinal tumors (87 and 88% sensitivity). SRS is least suitable for medullary thyroid carcinoma (44% sensitivity). In general the SRS whole body scan information is complementary to the more focal information given by CT or MRI. In figure 3 literature results are summarized in a Forest plot. The three imaging methods described below, namely VIP, bombesin and CCK receptor imaging are still experimental and are not yet available for routine clinical use. In figure 4 results for VIP, CCK and antibody imaging are presented. Vasoactive intestinal peptide receptor imaging (VIP) VIP and pituitary adenylate cyclase activating peptide (PACAP), both member of the secretin like peptides, are neuropeptides which regulate a broad spectrum of biological activities, including vasodilatation, stimulation of secretion of various hormones, immunomodulation and promotion of cell proliferation. There are two groups of receptors, namely VPAC 1 and VPAC 2, which are receptors with high affinity for VIP, PACAP and PAC 1, which is characterized, by a high affinity for PACAP but a low affinity for VIP. These receptors function through two distinct G-protein-coupled receptor subtypes that can also be internalized 15. VPAC 1 is expressed in most epithelial tissues and the brain, while VPAC 2 is only present in smooth muscle. Subsequently, these receptors are expressed in tumors derived from these tissues. The PAC 1 receptor is the only VIP/PACAP receptor found on catecholamine producing tumor cells of neuroendocrine origin and neuroblastomas. Interestingly VIP/PACAP receptors are absent in medullary thyroid cancers 16. Proteolytic degradation of VIP in vivo as well as high VPAC 1 receptor expression in normal epithelial tissues and VPAC 2 expression in smooth muscle hampers the applicability of this target for neuroendocrine tumor imaging. Recently a 64 Cu labeled VIP analogue has been developed, which is more stable and has been shown to have a higher tumor uptake then the 99m Tc labeled VIP analogue. Therefore this analogue might become of interest to study over expression of VIP receptors 17. Clinical application is however still limited. Bombesin receptor imaging Bombesin and gastrin releasing peptides (GRP) are members of the brain-gut peptides present in the nervous system, gastrointestinal tract and the pulmonary tract 18. GRP regulates several physiologic processes in the central and enteric nerve systems. An autocrine feedback mechanism involving the expression of bombesin, GRP receptors and the production of peptides in tumor cells (i.e. small cell lung cancer or neuroblastoma), can stimulate growth of neighboring tumor cells

28 Molecular imaging in neuroendocrine tumors: molecular uptake mechanisms and clinical results Abdominal carcinoids 111 In-octreotide 29 58% (44, 71) 111 In-octreotide 54 95% (82, 100) 111 In-octreotide 55 86% (81, 91) 111 In-octreotide 28 57% (46, 67) 111 In-octreotide % (97, 100) 111 In-octreotide 30 46% (43, 50) 111 In-octreotide 56 93% (78, 99) 111 In-octreotide 63 81% (54, 96) Pheochromocytoma Gastric carcinoids Merkel cell tumour Medullary thyroid carcinoma Neuroblastoma Pancreatic neuroendorine tumours Paraganglioma Small cell lung cancer Bronchial carcinoid 123 I-Tyr 3 -octreotide In-octreotide In-octreotide In-DTPA-D-Phe1-Oct In-DOTA-lanreotide In-DOTA-TOC 62 99m Tc-EDDA/HYNIC-TOC 63 99m Tc-EDDA/HYNIC-TOC In-octreotide In-octreotide In-octreotide In-octreotide In-octreotide In-octreotide In-octreotide In-octreotide In-octreotide In-octreotide In-octreotide In-octreotide In-octreotide In-lanreotide In-DOTA-TOC In-octreotide In-octreotide In-octreotide In-octreotide In-octreotide In-octreotide In-octreotide In-octreotide In-octreotide In-octreotide In-octreotide Sensitivity (%) 85% (73, 94) 52% (42, 62) 92% (89, 95) 87% (77, 93) 64% (46,, 78) 87% (72, 96) 81% (77, 85) 100% (92, 100) 63% (51, 73) 75% (68, 82) 100% (68, 100) 90% (84, 95) 78% (39, 98) 78% (40, 95) 29% (17, 42) 44% (29, 60) 75% (51, 92) 41% (23, 61) 25% (16, 35) 52% (32, 71) 71% (41, 92) 100% (79, 100) 95% (76, 100) 61% (35, 83) 64% (53, 74) 83% (34, 100) 93% (73, 99) 77% (56, 91) 54% (32, 71) 100% (91, 100) 83% (74, 91) 26% (15, 39) 86% (74, 94) 71% (52, 86) Figure 3. Forest plot analysis of SRS imaging per tumor type. Results of SRS imaging sorted per tumor type in a Forest plot analysis. On the left side of this plot the tumor type is given with the tracers used in the presented studies. On the right side sensitivities from literature data are given with their calculated confidence interval. The size of the solid black square represents the weight the corresponding study exerts in this analysis and is calculated using the number of patients included (Mantel-Haenszel weight). 13

29 Chapter 1 Abdominal carcinoids 123 I-VIP 73 0% (0, 28) 111 In-minigastrin 22 91% (70, 99) Pancreatic neuroendocrine tumours 111 In-Cg-3S-ISG % (68, 100) Medullary thyroid carcinoma 111 In-Cg-3S-ISG 50 96% (81, 100) 123 I-VIP 85 70% (65, 76) Sensitivity(%) Figure 4. Forest plot analysis of VIP, CCK and anti-body imaging per tumor type. Results of VIP, CCK and anti-body imaging sorted per tumor type in a Forest plot analysis. On the left side of this plot the tumor type is given with the tracers used in the presented studies. On the right side sensitivities from literature data are given with their calculated confidence interval. The size of the solid black square represents the weight the corresponding study exerts in this analysis and is calculated using the number of patients included (Mantel-Haenszel weight). The bombesin receptor family consists of four receptor subtypes; gastrin releasing peptide receptor (BB 2 ), neuromedin B receptors (BB 1 or NMB), and bombesin receptors BB 3 and BB 4. GRP receptor proteins are over expressed in tumors such as prostate, breast, renal cell and small cell lung cancer. The bombesin and GRP receptors are G protein coupled and can internalize after a receptor-agonist complex has been formed. In neuroendocrine tumors GRP receptors are preferentially expressed by gastrinomas. Ileal carcinoids express NMB receptors whereas small cell lung carcinomas and bronchial carcinoids express the BB 3 receptors 19. Bombesin analogues for the GRP receptor have been successfully labeled and seem to be interesting candidates for tumor imaging, especially in prostate cancer imaging to possibly improve lymph node staging and recurrence detection 20. CCK receptor imaging Cholecystokinin and gastrin, both members of the cholecystokinin peptide family, play an important role in the neurotransmission in the central nervous system as well as in the gastrointestinal physiology. In the gastrointestinal tract they play a role as a growth factor for physiologic processes, but also for several neoplasms, i.e. colon and brain tumors 16. Three G protein-coupled CCK receptors have been identified, namely CCK 1, CCK 2 and CCK C. These CCK receptors have the same internalization mechanisms as somatostatin receptors. This internalization is limited to receptor ligands with agonistic activity 21. CCK 1 receptors have been detected in i.e. meningioma and neuroblastoma. The CCK 2 receptor was identified in medullary thyroid carcinoma, astrocytomas, and some neuroendocrine gastroenteropancreatic tumors (especially insulinomas) and several soft tissue tumors. In pheochromocytomas and paragangliomas CCK 2 receptors are rarely expressed. This CCK 2 receptor can be targeted by radiolabeled CCK octapeptides 21. The last family member, the CCK C receptor, seems to be mainly involved in gastrin mediated proliferation in tumors of the nervous system. 14

30 Molecular imaging in neuroendocrine tumors: molecular uptake mechanisms and clinical results Only a few clinical studies with CCK imaging have been published. CCK-2 imaging is feasible with 111 In-DTPA-D-GLU 1 -minigastrin in patients with metastatic medullary thyroid cancer. In 32 patients, the sensitivity for this tracer was 91% 22. Metabolic imaging The production of different peptides distinguishes neuroendocrine tumors of other malignancies. Therefore the metabolic pathways by which neuroendocrine tumors synthesize these peptides and the intracellular processes which are essential to be able to sustain production of these peptides are ideal candidates for the development of tracers specific for neuroendocrine tumors. These pathways can be targeted at different levels. Tracers can be developed as a marker for the transporter proteins, which are necessary to supply the substrate into the cell (or intracellular vesicle), as a true substrate for the involved pathway, as a substrate which irreversibly binds to an enzyme involved in the pathway or as a marker for (re-)uptake of the end product. The thus far targeted metabolic pathways and the large amino acid transporter system (LAT, which is responsible for the uptake of precursors for both the serotonin and catecholamine pathway) will be described below in order of clinical importance. In figure 5 literature results of metabolic imaging methods for PET are presented. Abdominal carcinoids 18 FDOPA 29 86% (74, 94) 18 FDOPA 28 65% (55, 74) 18 FDOPA 30 96% (95, 98) 11 C-HTP % (80, 100) 11 C-HTP 59 89% (70, 89) 18 FDG 54 65% (38, 86) 18 FDG 28 29% (20, 39) 123 I-IMT 47 43% (35, 52) Pheochromocytoma 18 FDOPA % (76, 100) 18 F-dopamine 36 93% (88, 97) 18 F-dopamine % (87, 100) Medullary thyroid carcinoma 18 FDG 89 72% (53, 87) 18 FDOPA 28 63% (42, 81) 18 FDG 42 96% (91, 99) 18 FDG 43 78% (72, 84) 18 FDG 53 44% (25, 65) Small cell lung cancer 18 FDG 90 83% (80, 100) 18 FDG % (92, 100) 18 FDG 92 96% (90, 99) Paraganglioma 18 FDOPA % (76, 100) Sensitivity (%) Figure 5. Forest plot analysis of metabolic PET tracer imaging per tumor type. Results of 18 F-FDG PET, 18 F-DOPA PET, 18 F- Dopamine PET and 11 C-5-HTP PET sorted per tumor type in a Forest plot analysis. On the left side of this plot the tumor type is given with the tracers used in the presented studies. On the right side sensitivities from literature data are given with their calculated confidence interval. The size of the solid black square represents the weight the corresponding study exerts in this analysis and is calculated using the number of patients included (Mantel-Haenszel weight). 15

31 Chapter 1 Catecholamine pathway In the catecholamine pathway phenylalanine and intermediate products such as L-3,4 - dihydroxyphenylalanine ( L-DOPA) are taken up via the LAT system into the cytoplasm of the cell 23. Here these precursors can be metabolized to dopamine, which is transported into secretory vesicles via the vesicular monoamine transporter (VMAT) system 24. In these vesicles dopamine can be further metabolized to noradrenalin and adrenalin. The secretory vesicles are responsible for the secretion of end products. Finally, these end products can be transported back via i.e. dopamine and noradrenalin transporters. Tracers developed for this pathway are the precursor 6-18 F-L-3,4-dihydroxyphenylalanine ( 18 F-DOPA), the end product 6-18 F-dopamine and Metaiodobenzylguanidine (MIBG), which is a substrate for the noradrenalin transporter F-DOPA PET F-DOPA is an F labeled variant of L-DOPA used for PET imaging (figure 6). Although the presence of the 18 F atom in 18 F -DOPA influences the metabolism, it has no or little effect for the transport into the intracellular environment via the cell membrane bound LAT2 transporter F -DOPA is decarboxylated to 18 F-dopamine via the enzyme aromatic acid decarboxylase (AADC) at a faster rate then L-DOPA. The thus formed dopamine is then probably transported into secretory vesicles by VMAT transporters. Although the precise uptake mechanism is not fully understood, it appears that the high 18 F-DOPA uptake in neuroendocrine tumors is the result of increased LAT2 transporter activity to satisfy a high precursor turnover due to an increased catecholamine pathway or at least increased AADC activity 26. Somewhat paradoxically, the AADC inhibitor carbidopa is sometimes used in conjunction with 18 F-DOPA PET imaging. In the proximal tubuli of the kidney 18 F-DOPA is rapidly converted by AADC to 18 F-dopamine, which is then excreted. This leads to rapid loss of 18 F-DOPA and may generate renal, ureter or bladder artifacts in 18 F-DOPA PET imaging. Figure 6. Patient with metastatic pancreatic islet cell tumor. 18 F-DOPA PET (left) and 11 C-5-HTP PET (right) in a patient with metastatic pancreatic islet cell tumor. 16

32 Molecular imaging in neuroendocrine tumors: molecular uptake mechanisms and clinical results The use of oral carbidopa as pre-treatment for 18 F-DOPA PET studies improves image quality by reducing the conversion of 18 F-DOPA and the excretion of 18 F-dopamine in the kidney in urine. The use of carbidopa also lowers physiological 18 F-DOPA uptake in the pancreas. The combined effect is a higher availability of 18 F-DOPA for tumor uptake 27. Only a few studies with 18 F-DOPA PET in neuroendocrine tumors have thus far been published. 18 F-DOPA PET yields a very high sensitivity in the detection of carcinoid tumors, paragangliomas and pheochromocytomas. For example, in the detection of gastrointestinal neuroendocrine tumors, the average sensitivity of 18 F-DOPA PET is very high (89%, figure 5), whereas in these studies the currently used standard methods (CT and SRS imaging) performed not nearly as good (sensitivities of 56% for CT versus 47% for SRS imaging) 28,29,30. When comparing the thus far published results, 18 F-DOPA PET enabled best localization of primary tumors and lymph node staging 28,29,30. There is a very small risk that the use of a catecholamine precursor in patients with a carcinoid syndrome can lead to the development of a carcinoid crisis. Only one case has thus far been published. This complication can be prevented by a slow tracer injection and if necessary an intravenous injection of octreotide I and 131 I- Metaiodobenzylguanidine (MIBG) - The precise uptake and retention mechanism in neuroendocrine tumors for MIBG has not been clarified, but the noradrenalin transporter seems to play an important role for MIBG uptake (figure 3). Reports indicate that MIBG acts as an intracellular substrate for the vesicular monoamine transporters VMAT 1 and VMAT These transporters are located on the membrane of secretory vesicles of neuroendocrine cells. It seems likely that, once MIBG has passed the cell membrane, the VMAT transporters transport MIBG into secretory chromaffin granules 32. MIBG can be labeled with 123 I and 131 I. 123 I labeled MIBG yields the best image quality, due to the superior physical qualities for imaging. It has high photon energy of 159 KeV, a half life of 13 hours and it can be administered in a higher dose then 131 I-MIBG. These properties enable the use of 123 I-MIBG for SPECT 33. Figure 7. Double side pheochromocytoma. 18 F-DOPA PET (left) and 123 I-MIBG (right) of a patient with double-sided pheochromocytoma. 17

33 Chapter 1 MIBG scintigraphy has become the imaging method of choice for neuroblastoma and pheochromocytomas (for an example see figure 7). MIBG scintigraphy has a lower sensitivity for the detection of other neuroendocrine tumors, such as carcinoid (averaging 50%) (figure 8). Its specificity in detecting pheochromocytoma and neuroblastoma is superior to other imaging modalities. There is however no explanation for the variation in uptake of MIBG by the different neuroendocrine tumors (figure 4). Most studies report specificities for MIBG ranging from % for the detection of pheochromocytoma, but less then 80% for the detection its malignant variant. Specificity for neuroblastomas is 84%. But, since other neuroendocrine tumors in childhood are rare, a positive MIBG scan is nearly diagnostic for a neuroblastoma 34. Abdominal carcinoids Pheochromocytoma 131 I-MIBG 94 73% (61, 83) 123 I-MIBG or 131 I-MIBG 58 51% (37, 65) 131 I-MIBG 95 88% (68, 98) 123 I-MIBG 96 92% (74, 98) I-MIBG 90% (61, 99) 131 I-MIBG 33 86% (68, 96) 123 I-MIBG 87 92% (74, 98) 131 I-MIBG 36 72% (64, 80) 123 I-MIBG or 131 I-MIBG 89 86% (68, 96) 123 I-MIBG or 131 I-MIBG 65 81% (70, 89) 123 I-MIBG 97 35% (20, 53) Neuroblastoma 123 I-MIBG 98 89% (66, 99) 123 I-MIBG 99 76% (62, 87) 123 I-MIBG 75 78% (52, 94) 123 I-MIBG % (70, 82) 123 I-MIBG 76 94% (87, 98) Paraganglioma 123 I-MIBG % (55, 86) 123 I-MIBG 80 50% (15, 85) 123 I-MIBG % (85, 100) Sensitivity 0(%) Figure 8. Forest plot analysis of MIBG imaging per tumor type. Results of MIBG imaging sorted per tumor type in a Forest plot analysis. On the left side of this plot the tumor type is given with the tracers used in the presented studies. On the right side sensitivities from literature data are given with their calculated confidence interval. The size of the solid black square represents the weight the corresponding study exerts in this analysis and is calculated using the number of patients included (Mantel-Haenszel weight) F-Dopamine PET F-Dopamine is a substrate for the monoamine transporters DAT (dopamine transporter) and the norepinephrine transporter. After this trans-membrane transport, 18 F-dopamine is stored in cytoplasmatic secretory vesicles through the VMAT system. However, 18 F- dopamine plasma levels decline rapidly after injection due to metabolization The PET tracer 6- F-dopamine was developed to visualize sympathiconeuronal innervation. This tracer is actively taken up, stored and metabolized by cells from organs with a sympathetic innervation

34 Molecular imaging in neuroendocrine tumors: molecular uptake mechanisms and clinical results Organs which have a high 18 F-dopamine uptake are the heart, liver, spleen, salivary glands and chest wall. 18 F-dopamine does not cross the blood-brain barrier, and therefore virtually no uptake is seen in the brain. 18 F-dopamine PET is a useful imaging method for the detection of pheochromocytomas 36. These tumors have a high expression of monoamine transporters, which makes these tumors ideal candidates for 18 F-dopamine imaging. Due to a low specific activity and the pharmacological activity of both labeled and unlabelled dopamine, the maximum injectable dose of 18 F-dopamine is limited. Serotonin pathway The serotonin and catecholamine pathway have many common features. Precursors for the serotonin pathway (tryptophan and 5-hydroxytryptophan, 5-HTP) are taken up via the same LAT transport system as utilized by the catecholamine pathway. The conversion from 5- HTP to serotonin is performed by the same enzyme, which decarboxylizes L-DOPA to dopamine, namely the AADC enzyme. The end product, serotonin, is also transported via the same VMAT transporter system into secretory vesicles C-5-Hydroxytryptophan ( 11 C-5-HTP) PET - Thus far, 11 C-5-HTP PET is the only tracer for this pathway, which has reached clinical application (figure 6). In neuroendocrine tumors, the high uptake via the over expressed system L transporter, the rapid decarboxylation by AADC and the subsequent storage in secretory granules allow for an excellent discrimination of these tumors with 11 C-5-HTP and 18 F-DOPA as compared with normal tissue 37. Due to low organ uptake of 11 C-5-HTP, scans are characterized by a low background activity. 11 C-5-HTP uptake by the kidneys and metabolization to serotonin by the AADC enzyme followed by subsequent urinary excretion, result in an intense signal in these organ systems. There is also some physiologic pancreatic uptake noticeable. This can lead to difficulties in the interpretation of lesions in the direct vicinity of these organs. However, the oral pre-medication with carbidopa lowers the metabolization to serotonin thereby reducing the signal intensity in this area, thus improving image quality and interpretability 38. There are however two draw-backs for this tracer. The tracer synthesis is very complex since it relies on two complex multi-enzyme steps. Also the short half life of 20 min for 11 Carbon limits the use of this tracer to specialized centers with their own cyclotron facilities. Nevertheless, the published results with this tracer justify the use of this tracer. Phosphate metabolism Inorganic phosphate (Pi) molecules and Na + ions are taken up by cells via the Na + /Pi cotransporters 39. Three different families of Na + /Pi co-transporters have been reported, type I, II and III. These transporters are involved in the inorganic phosphate transport in cells. Type II and III are regulated by extra cellular ph whereas type I is indifferent to ph. Type II activity is decreased by acidic ph and increased with an alkaline ph. Type III functions in the opposite way, i.e. acidic ph increases its activity. The physiologic function of the Na + /Pi transporters is still not entirely clear. These transporters are predominantly expressed in kidney, liver and brain. The three families of this transporter are expressed differently in these organs and within the tissues of these organs 39,40. 19

35 Chapter 1 99m Tc -(V)-Dimercaptosuccinic acid ( 99m Tc-(V)-DMSA) 99m The uptake mechanism for Tc-(V)-DMSA is based on the resemblance between the TcO complex in labeled DMSA and the phosphate molecule PO Due to the over expression of type III Na + /Pi transporters, lack of Type II Na + /Pi transporters and a more acidic extra cellular ph in tumor cells than normal tissue, tumor cells have a higher phosphate uptake. As 99m Tc (V)-DMSA resembles the phosphate molecule, it is also actively taken up. 99m Tc-(V)-DMSA is therefore a marker for the phosphate metabolism 41. This tracer has thus far mainly been used for imaging medullary thyroid tumors. For the detection of this tumor type, 99m Tc-(V)-DMSA is the routinely used tracer with the best results (figure 3). It has an average sensitivity of 76 % for medullary thyroid tumors (figure 9). In these tumors, morphological imaging (CT/MRI) performs equally (sensitivity ranging from 67% to 87%) 42, Tc-(V)-DMSA % (49, 96) 99 Tc-(V)-DMSA 71 69% (55, 80) 99 Tc-(V)-DMSA 72 30% (17, 43) 99 Tc-(V)-DMSA 42 57% (41, 71) 99 Tc-(V)-DMSA 43 33% (21, 45) 99 Tc-(V)-DMSA 74 50% (27, 73) 99 Tc-(V)-DMSA % (84, 100) Sensitivity (%) Figure 9. Forest plot analysis of 99m Tc-(V)-DMSA imaging for medullary thyroid tumors. Results of DMSA imaging for medullary thyroid carcinoma in a Forest plot analysis. On the left side of this plot the tumor type is given with the tracers used in the presented studies. On the right side sensitivities from literature data are given with their calculated confidence interval. The size of the solid black square represents the weight the corresponding study exerts in this analysis and is calculated using the number of patients included (Mantel-Haenszel weight). Glucose metabolism Cells rely mostly on their glucose metabolism to obtain ATP as their energy source. After uptake by glucose transporters, glucose is phosphorylated by the hexokinase enzyme to a phosphorylated intermediate, which is eventually metabolized to pyruvate and lactate. This step does not require oxygen (anaerobic glycolysis). The next step, the citric acid cycle, requires oxygen (aerobic) and produces most of the ATP molecules. In the citric cycle, pyruvate is eventually metabolized to CO 2. Tumor cells rely mainly on the anaerobic glycolysis with its relatively low energy yield, and therefore require much more glucose. 20

36 Molecular imaging in neuroendocrine tumors: molecular uptake mechanisms and clinical results L I-alpha-methyl-tyrosine ( 123 I-IMT) 123 I-IMT, an artificial amino acid derived from tyrosine, was initially developed as a functional imaging agent for neutral amino acid transport in brain tumors. 123 I-IMT accumulates fast in neuroendocrine tumor cells due to uptake via LAT1, but is not further metabolized intra-cellularly 48. Therefore, it can be regarded as a true marker for transport capabilities of neuroendocrine tumor cells. A study in 22 carcinoid patients showed an overall lesion detection of 43% with a lower lesion contrast and image quality than 111 In- octreotide I-IMT is not generally available. 18 F-Fluor-2-deoxy-D-Glucose PET 18 F-2-deoxy-D-glucose (FDG) is transported into the cells and then phosphorylated by hexokinase. This results in a polar intermediate (FDG-6p), which crosses the cell membrane poorly. The increased expression of the glucose transporter molecules and the hexokinase enzyme result in an increased uptake and retention of FDG in tumor cells compared to normal cells 44. In general, FDG uptake increases when tumors behave more aggressively 45. However, neuroendocrine tumors do not have a high glycolysis rate. Results of FDG imaging in neuroendocrine tumors are inferior to these obtained for metabolically active tumors. Neuroendocrine tumors which have a high uptake, such as small cell lung cancer, are characterized by a more aggressive behavior 46. The use of the FDG PET scan in neuroendocrine tumors is therefore more or less limited to imaging small cell lung cancer (with approximately 93% sensitivity) and medullary thyroid carcinoma (76% sensitivity). Large amino acid transport system To import large branched and aromatic neutral amino acids cells rely on the plasma membrane bound system L transporters. This system consists of two heterodimers composed of a large glycoprotein part and a variable light chain, thus forming the LAT 1 to 5. System L transporters, which are amino acid transporters, are obligatory exchange transporters which can only function by exchanging an intracellular amino acid for an extra-cellular one. In combination with other unidirectional transporters with overlapping amino acid sensitivities, cells can control the activity of the system L transport. Over expression of amino acid transporters helps to satisfy the metabolic needs, but tumor cells can consume more nutrients than required for the metabolic needs. In neuroendocrine tumors the LAT2 transporters play an important role, due to their ability to take up large neutral amino acids such as phenylalanine and tryptophan. Thus far only L I-alphamethyl-tyrosine ( 123 I-IMT) has been developed to exploit the over expression of the LAT system for imaging purposes in neuroendocrine tumors 46,47. Radiolabeled monoclonal antibodies Only a few reports of patient studies with radio labeled monoclonal antibodies against antigens expressed on neuroendocrine tumors used for nuclear medicine imaging are available. Described applications are anti-cea for paragangliomas, anti CgA for medullary thyroid carcinoma and anti- UJ13A and anti GD2 for neuroblastoma These reports should be seen as experimental, due to the limited data available and the fact that these methods have not yet been adapted for clinical use. 21

37 Chapter 1 Abdominal carcinoids 111 In-Octreotide 79% (76, 82) 111 In-DOTA-TOC 87% (72, 96) 99m Tc EDDA 90% (85, 94) 123 I-VIP 70% (65, 76) 18 F-DOPA 89% (81, 94,) 11 C-5-HTP 95% (78, 99) 123 I and 131 I-MIBG 63% (54, 72) 18 FDG 35% (26, 45) 123 IMT 43% (35, 52) Pancreatic neuro-endocrine tumours Pheochromocytoma 111 In-Octreotide 74% (63, 83) 111 In-Cg-3S-ISG 96% (81, 100) 111 In-Octreotide 63% (51, 73) 18 F-DOPA 100% (76, 100) 18 F-dopamine 98% (88, 100) 123 I and 131 I-MIBG 79% (68, 82) 18 FDG 72% (53, 87) Merkel cell tumours Medullary thyroid carcinoma 111 In-Octreotide 78% (51, 92) 111 In-Octreotide 49% (42, 57) 111 In-DOTA TOC 95% (76, 100) 123 I-VIP 0% (0, 28) 111 In Minigastrin 91% (70, 99) 111 In-Cg-3S-ISG 100% (68, 100) 18 F-DOPA 63% (42, 81) 99m Tc-V-DMSA 55% (47, 64) 18 FDG 79% (70, 86) Neuroblastoma Paraganglioma 111 In-Octreotide 64% (53, 72) 123 I and 131 I-MIBG 84% (79, 89) 111 In-Octreotide 97% (86, 100) 18 F-DOPA 100% (76, 100) 123 I and 131 I-MIBG 81% (69, 91) Bronchial carcinoids Small cell lung cancer 111 In-Octreotide 71% (52, 86) 111 In-Octreotide 56% (38, 71) 18 FDG 95% (88, 98) Sensitivity (%) Figure 10. Summary of results. Results of all described imaging methods sorted per tumor type in a Forest plot analysis. On the left side of this plot the tumor type is given with the tracers used in the presented studies. On the right side sensitivities from literature data are given with their calculated confidence interval. The size of the solid black square represents the weight the corresponding study exerts in this analysis and is calculated using the number of patients included (Mantel-Haenszel weight). Conclusion Most receptor-based tracers have been developed for scintigraphic use, and changing to positron emitting labels (i.e. 18 F, 64 Cu, 68 Ga) could make these tracers suitable for PET imaging. For SRS, different somatostatin analogues are investigated which are more stable and bind more receptor subtypes with a higher affinity. Most neuroendocrine tumors share common metabolic pathways, such as the catecholamine 22

38 Molecular imaging in neuroendocrine tumors: molecular uptake mechanisms and clinical results and serotonin pathways. Different strategies in the development of tracers suitable to visualize the metabolic pathways are possible. For example, three different tracers for the catecholamine pathway have been developed: the tracer 18 FM-6-FmT is an aromatic L- AADC inhibitor, 18 F-DOPA is a substrate in the catecholamine synthesis, 18 F-dopamine is stored in vesicles where it is metabolized to 18 F-norepinephrin and 18 F-epinephrin. Several other tracers, which exploit the metabolic characteristics, are in development. The use of antibodies specific for neuroendocrine tumors has thus far not been as successful as the receptor and metabolic based imaging methods. It has been troubled by the high background uptake. The developments in the area of image fusion are also of great interest for neuroendocrine tumors. Nuclear medicine techniques lack anatomical information, whereas morphological imaging lacks functional information. Co-registration of these modalities, either by software or by hardware, assists in tumor localization. The combined functional and anatomic images give surgeons essential information to guide surgical decision-making. An enormous amount of expertise and knowledge has been gathered in the past years with somatostatin analogues, both for diagnostic and treatment purposes. It is more or less clear which tumor types can be visualized to what extend with this tracer. It can be expected that combining somatostatin analogues with positron emitting labels will secure its position within both diagnostic and therapeutic procedures in nuclear medicine in the near future. However, somatostatin receptors are not expressed evenly both in receptor subtypes and quantity of expressed receptors on the cellular membrane. Therefore other techniques will improve diagnostic imaging. Due to their common capability for the uptake of large amino acids for the incorporation in metabolic pathways, precursors (i.e. 18 F-DOPA and 11 C-5- HTP) for these pathways are interesting candidates for diagnostic and therapeutic purposes. Although results are promising, experience with these PET tracers is still limited. It is foreseeable that in the future the first step in neuroendocrine tumor imaging will be with the use of PET tracers, such as 18 F-DOPA on a PET-CT machine. When this combination yields negative results, other techniques can be used for further analysis. References 1. Reubi JC. Neuropeptide receptors in health and disease: the molecular basis for in vivo imaging. J Nucl Med 1995; 36: Schnirer LL, Yao JC, Ajani JA. Carcinoid - a comprehensive review. Acta Oncol 2003; 42: Kloppel G, Heitz PU, Capella C, Solcia E. Pathology and nomenclature of human gastrointestinal neuroendocrine (carcinoid) tumours and related lesions. World J Surg 1996; 20: Modlin IM, Lye KD, Kidd M. A 5-decade analysis of 13,715 carcinoid tumours. Cancer 2003; 97: Solcia E, Kloppel G, Sobin LH et al. Histologic typing of endocrine tumours. WHO International Histological Classification of Tumours. 2nd ed. Heidelberg: Springer Verlag; Öberg K. Chemotherapy and biotherapy in the treatment of neuroendocrine tumours. Ann Oncol 2001; 12 Suppl 2: S111-S

39 Chapter 1 7. Warner RR, O'dorisio TM. Radiolabeled peptides in diagnosis and tumour imaging: clinical overview. Semin Nucl Med 2002; 32: Plockinger U, Rindi G, Arnold R et al. Guidelines for the diagnosis and treatment of neuroendocrine gastrointestinal tumours. A consensus statement on behalf of the European Neuroendocrine Tumour Society (ENETS). Neuroendocrinology 2004; 80: Lamberts SW, de Herder WW, Hofland LJ. Somatostatin analogs in the diagnosis and treatment of cancer. Trends Endocrinol Metab 2002; 13: Hofland LJ, Lamberts SW. The pathophysiological consequences of somatostatin receptor internalization and resistance. Endocr Rev 2003; 24: Krenning EP, Kwekkeboom DJ, Bakker WH et al. Somatostatin receptor scintigraphy with [ 111 In-DTPA-D-Phe1]- and [ 123 I-Tyr3]-octreotide: the Rotterdam experience with more than 1000 patients. Eur J Nucl Med 1993; 20: Schottelius M, Poethko T, Herz M et al. First 18 F-labeled tracer suitable for routine clinical imaging of sst receptor-expressing tumours using positron emission tomography. Clin Cancer Res 2004; 10: Lewis JS, Srinivasan A, Schmidt MA, Anderson CJ. In vitro and in vivo evaluation of 64 Cu-TETA-Tyr3-octreotate. A new somatostatin analog with improved target tissue uptake. Nucl Med Biol 1999; 26: Ginj M, Chen J, Walter MA, Eltschinger V, Reubi JC, Maecke HR. Preclinical evaluation of new and highly potent analogues of octreotide for predictive imaging and targeted radiotherapy. Clin Cancer Res 2005; 11: Shetzline MA, Walker JK, Valenzano KJ, Premont RT. Vasoactive intestinal polypeptide type-1 receptor regulation. Desensitization, phosphorylation, and sequestration. J Biol Chem 2002; 277: Reubi JC. Peptide receptors as molecular targets for cancer diagnosis and therapy. Endocr Rev 2003; 24: Thakur ML, Aruva MR, Gariepy J et al. PET imaging of oncogene overexpression using 64 Cu-vasoactive intestinal peptide (VIP) analog: comparison with 99mTc- VIP analog. J Nucl Med 2004; 45: Seretis E, Gavrill A, Agnantis N, Golematis V, Voloudakis-Baltatzis IE. Comparative study of serotonin and bombesin in adenocarcinomas and neuroendocrine tumours of the colon. Ultrastruct Pathol 2001; 25: Reubi JC, Wenger S, Schmuckli-Maurer J, Schaer JC, Gugger M. Bombesin receptor subtypes in human cancers: detection with the universal radioligand 125 I- [D-TYR(6)], beta-ala(11), PHE(13), NLE(14)] bombesin(6-14). Clin Cancer Res 2002; 8: Reubi JC, Waser B. Concomitant expression of several peptide receptors in neuroendocrine tumours: molecular basis for in vivo multireceptor tumour targeting. Eur J Nucl Med Mol Imaging 2003; 30: Behe M, Behr TM. Cholecystokinin-B (CCK-B)/gastrin receptor targeting peptides for staging and therapy of medullary thyroid cancer and other CCK-B receptor expressing malignancies. Biopolymers 2002; 66:

40 Molecular imaging in neuroendocrine tumors: molecular uptake mechanisms and clinical results 22. Behr TM, Behe MP. Cholecystokinin-B/Gastrin receptor-targeting peptides for staging and therapy of medullary thyroid cancer and other cholecystokinin-b receptor-expressing malignancies. Semin Nucl Med 2002; 32: Soares-da-Silva P, Serrao MP. High- and low-affinity transport of L-leucine and L-DOPA by the hetero amino acid exchangers LAT1 and LAT2 in LLC-PK1 renal cells. Am J Physiol Renal Physiol 2004; 2: F Ericksson JD, Schafer MK, Bonner TI, Eiden LE, Weihe E. Distinct pharmacological properties and distribution in neurons and endocrine cells of two isoforms of the human vesicular monoamine transporter. Proc Natl Acad Sci U.S.A 1996; 93: Uchino H, Kanai Y, Kim DK et al. Transport of amino acid-related compounds mediated by L-type amino acid transporter 1 (LAT1): insights into the mechanisms of substrate recognition. Mol Pharmacol 2002; 61: Firnau G, Sood S, Chirakal R, Nahmias C, Garnett ES. Metabolites of 6- [ 18 F]fluoro-L-dopa in human blood. J Nucl Med 1988; 29: Hoffman JM, Melega WP, Hawk TC et al. The effects of carbidopa administration on 6-[ 18 F]fluoro-L-dopa kinetics in positron emission tomography. J Nucl Med 1992; 33: Hoegerle S, Altehoefer C, Ghanem N et al. Whole-body 18 F dopa PET for detection of gastrointestinal carcinoid tumours. Radiology 2001; 220: Becherer A, Szabo M, Karanikas G et al. Imaging of advanced neuroendocrine tumours with 18 F-FDOPA PET. J Nucl Med 2004; 45: Koopmans KP, de Vries EG, Kema IP et al. Staging of carcinoid tumours using 18 F-DOPA positron emission tomography: a diagnostic accuracy study. Lancet Oncol 2006; 7: Koopmans KP, Brouwers AH, De Hooge MN et al. Carcinoid crisis after injection of 6-18 F-fluorodihydroxyphenylalanine in a patient with metastatic carcinoid. J Nucl Med 2005; 46: Kolby L, Bernhardt P, Levin-Jakobsen AM et al. Uptake of meta- in neuroendocrine tumours is mediated by vesicular iodobenzylguanidine monoamine transporters. Br J Cancer 2003; 89: Furuta N, Kiyota H, Yoshigoe F, Hasegawa N, Ohishi Y. Diagnosis of pheochromocytoma using [ 123 I]-compared with [ 131 I]-metaiodobenzylguanidine scintigraphy. Int J Urol 1999; 6: Hiorns MP, Owens CM. Radiology of neuroblastoma in children. Eur Radiol 2001; 11: Goldstein DS, Holmes C. Metabolic fate of the sympathoneural imaging agent 6- [ 18 F]fluorodopamine in humans. Clin Exp Hypertens 1997; 19: Ilias I, Yu J, Carrasquillo JA et al. Superiority of 6-[ F]-fluorodopamine positron emission tomography versus [ 131 I]-metaiodobenzylguanidine scintigraphy in the localization of metastatic pheochromocytoma. J Clin Endocrinol Metab 2003; 88: Bergström M, Lu L, Eriksson B et al. Modulation of organ uptake of 11 C-labelled 5-hydroxytryptophan. Biog Amines 1996; 12:

41 Chapter Örlefors H, Sundin A, Lu L et al. Carbidopa pretreatment improves image 11 interpretation and visualisation of carcinoid tumours with C-5- hydroxytryptophan positron emission tomography. Eur J Nucl Med Mol Imaging 2006; 33: Werner A, Dehmelt L, Nalbant P. Na + -dependent phosphate cotransporters: the NaPi protein families. J Exp Biol 1998; 201: Lam AS, Puncher MR, Blower PJ. In vitro and in vivo studies with pentavalent technetium-99m dimercaptosuccinic acid. Eur J Nucl Med 1996; 23: Denoyer D, Perek N, Le Jeune N, Frere D, Dubois F. Evidence that 99m Tc-(V)- DMSA uptake is mediated by NaPi cotransporter type III in tumour cell lines. Eur J Nucl Med Mol Imaging 2004; 31: De Groot JW, Links TP, Jager PL, Kahraman T, Plukker JT. Impact of 18 F-fluoro- 2-deoxy-D-glucose positron emission tomography (FDG-PET) in patients with biochemical evidence of recurrent or residual medullary thyroid cancer. Ann Surg Oncol 2004; 11: Diehl M, Risse JH, Brandt-Mainz K et al. Fluorine-18 fluorodeoxyglucose positron emission tomography in medullary thyroid cancer: results of a multicentre study. Eur J Nucl Med 2001; 28: Bar-Shalom R, Valdivia AY, Blaufox MD. PET imaging in oncology. Semin Nucl Med 2000; 30: Adams S, Baum R, Rink T, Schumm-Drager PM, Usadel KH, Hor G. Limited value of fluorine-18 fluorodeoxyglucose positron emission tomography for the imaging of neuroendocrine tumours. Eur J Nucl Med 1998; 25: Meier C, Ristic Z, KLauser S, Verrey F. Activation of system L heterodimeric amino acid exchangers by intracellular substrates. Embo J 2002; 21: Jager PL, Meijer WG, Kema IP, Willemse PH, Piers DA, de Vries EG. L-3- [ 123 I]Iodo-alpha-methyltyrosine scintigraphy in carcinoid tumours: correlation with biochemical activity and comparison with [ 111 In-DTPA-D-Phe1]-octreotide imaging. J Nucl Med 2000; 41: Shikano N, Kanai Y, Kawai K et al. Isoform selectivity of 3- I-iodo-alpha- tympanicum tumours by In-111 labeled monoclonal antibody using single photon methyl-l-tyrosine membrane transport in human L-type amino acid transporters. J Nucl Med 2003; 44: Kairemo KJ, Himi T, Hopsu EV, Ramsay HA. Radioimmunoimaging of glomus emission computed tomography. Am J Otol 1997; 18: Siccardi AG, Paganelli G, Pontiroli AE et al. In vivo imaging of chromogranin A- positive endocrine tumours by three-step monoclonal antibody targeting. Eur J Nucl Med 1996; 23: Goldman A, Vivian G, Gordon I, Pritchard J, Kemshead J. Immunolocalization of neuroblastoma using radiolabeled monoclonal antibody UJ13A. J Pediatr 1984; 105: Larson SM, Pentlow KS, Volkow ND et al. PET scanning of iodine-124-3f9 as an approach to tumour dosimetry during treatment planning for radioimmunotherapy in a child with neuroblastoma. J Nucl Med 1992; 33:

42 Molecular imaging in neuroendocrine tumors: molecular uptake mechanisms and clinical results 53. Hoegerle S, Altehoefer C, Ghanem N, Brink I, Moser E, Nitzsche E. 18 F-DOPA positron emission tomography for tumour detection in patients with medullary thyroid carcinoma and elevated calcitonin levels. Eur J Nucl Med 2001; 28: Belhocine T, Foidart J, Rigo P et al. Fluorodeoxyglucose positron emission tomography and somatostatin receptor scintigraphy for diagnosing and staging carcinoid tumours: correlations with the pathological indexes p53 and Ki-67. Nucl Med Commun 2002; 23: Chiti A, Fanti S, Savelli G et al. Comparison of somatostatin receptor imaging, computed tomography and ultrasound in the clinical management of neuroendocrine gastro-entero-pancreatic tumours. Eur J Nucl Med 1998; 25: Krausz Y, Bar-Ziv J, de Jong RB et al. Somatostatin-receptor scintigraphy in the management of gastroenteropancreatic tumours. Am J Gastroenterol 1998; 93: Montravers F, Grahek D, Kerrou K. Can fluorodihydroxyphenylalanine PET replace somatostatin receptor scintigraphy in patients with digestive endocrine tumors? J Nucl Med 2006; 47: Nocaudie-Calzada M, Huglo D, Carnaille B, Proye C, Marchandise X. Comparison of somatostatin analogue and metaiodobenzylguanidine scintigraphy for the detection of carcinoid tumours. Eur J Nucl Med 1996; 23: Örlefors H, Sundin A, Ahlstrom H et al. Positron emission tomography with 5- hydroxytryprophan in neuroendocrine tumours. J Clin Oncol 1998; 16: Raderer M, Kurtaran A, Leimer M et al. Value of peptide receptor scintigraphy using 123 I-vasoactive intestinal peptide and 111 In-DTPA-D-Phe1-octreotide in 194 carcinoid patients: Vienna University Experience, 1993 to J Clin Oncol 2000; 18: Shi W, Johnston CF, Buchanan KD et al. Localization of neuroendocrine tumours with [ 111 In] DTPA-octreotide scintigraphy (Octreoscan): a comparative study with CT and MR imaging. QJM 1998; 91: Virgolini I, Patri P, Novotny C et al. Comparative somatostatin receptor scintigraphy using In-111-DOTA-lanreotide and In-111-DOTA-Tyr3-octreotide versus F-18-FDG-PET for evaluation of somatostatin receptor-mediated radionuclide therapy. Ann Oncol 2001;12 Suppl 2: S41-S45. 99m 63. Gabriel M, Muehllechner P, Decristoforo C et al. Tc-EDDA/HYNIC-Tyr(3)- octreotide for staging and follow-up of patients with neuroendocrine gastro-enteropancreatic tumours. J Nucl Med Mol Imaging 2005; 49: Hubalewska-Dydejczyk A, Fross-Baron K, Mikolajczak R et al. 99m Tc- EDDA/HYNIC-octreotate scintigraphy, an efficient method for the detection and staging of carcinoid tumours: results of 3 years' experience. Eur J Nucl Med Mol Imaging 2006; 33: Tenenbaum F, Lumbroso J, Schlumberger M et al. Comparison of radiolabeled octreotide and meta-iodobenzylguanidine (MIBG) scintigraphy in malignant pheochromocytoma. J Nucl Med 1995; 36:

43 Chapter Gibril F, Reynolds JC, Lubensky IA et al. Ability of somatostatin receptor scintigraphy to identify patients with gastric carcinoids: a prospective study. J Nucl Med 2000; 41: Briganti V, Matteini M, Ferri P, Vaggelli L, Castagnoli A, Pieroni C. Octreoscan SPECT evaluation in the diagnosis of pancreas neuroendocrine tumours. Cancer Biother Radiopharm 2001; 16: Schillaci O, Spanu A, Scopinaro F et al. Somatostatin receptor scintigraphy with 111 In-pentetreotide in non-functioning gastroenteropancreatic neuroendocrine tumours. Int J Oncol 2003; 23: Durani BK, Klein A, Henze M, Haberkorn U, Hartschuh W. Somatostatin analogue scintigraphy in Merkel cell tumours. Br J Dermatol 2003;148: Guitera-Rovel P, Lumbroso J, Gautier-Gougis MS, et al. Indium-111 octreotide scintigraphy of Merkel cell carcinomas and their metastases. Ann Oncol 2001; 12: Adams S, Baum RP, Hertel A, Schumm-Draeger PM, Usadel KH, Hor G. Comparison of metabolic and receptor imaging in recurrent medullary thyroid carcinoma with histopathological findings. Eur J Nucl Med 1998; 25: Arslan N, Ilgan S, Yuksel D et al. Comparison of In-111 octreotide and Tc-99m (V) DMSA scintigraphy in the detection of medullary thyroid tumour foci in patients with elevated levels of tumour markers after surgery. Clin Nucl Med 2001; 26: Berna L, Chico A, Matias-Guiu X et al. Use of somatostatin analogue scintigraphy Nucl Med in the localization of recurrent medullary thyroid carcinoma. Eur J 1998; 25: Kurtaran A, Scheuba C, Kaserer K, Schima W, Czerny C, Angelberger P et al. Indium-111-DTPA-D-Phe-1-octreotide and technetium-99m-(v)- dimercaptosuccinic acid scanning in the preoperative staging of medullary thyroid carcinoma. J Nucl Med 1998; 39: Kropp J, Hofmann M, Bihl H. Comparison of MIBG and pentetreotide scintigraphy in children with neuroblastoma. Is the expression of somatostatin receptors a prognostic factor? Anticancer Res 1997; 17: Schilling FH, Bihl H, Jacobsson H et al. Combined In-pentetreotide scintigraphy and 123 I-MIBG scintigraphy in neuroblastoma provides prognostic information. Med Pediatr Oncol 2000; 35: Corleto VD, Scopinaro F, Angeletti S et al. Somatostatin receptor localization of pancreatic endocrine tumours. World J Surg 1996; 20: Rickes S, Unkrodt K, Ocran K, Neye H, Wermke W. Differentiation of neuroendocrine tumours from other pancreatic lesions by echo-enhanced power Doppler sonography and somatostatin receptor scintigraphy. Pancreas 2003; 26: Duet M, Sauvaget E, Petelle B et al. Clinical impact of somatostatin receptor scintigraphy in the management of paragangliomas of the head and neck. J Nucl Med 2003; 44:

44 Molecular imaging in neuroendocrine tumors: molecular uptake mechanisms and clinical results 80. Muros MA, Llamas-Elvira JM, Rodriguez A et al. 111 In-pentetreotide scintigraphy is superior to 123 I-MIBG scintigraphy in the diagnosis and location of chemodectoma. Nucl Med Commun 1998; 19: Bohuslavizki KH, Brenner W, Gunther M et al. Somatostatin receptor scintigraphy in the staging of small cell lung cancer. Nucl Med Commun 1996; 17: Bombardieri E, Crippa F, Cataldo I et al. Somatostatin receptor imaging of small cell lung cancer (SCLC) by means of 111In-DTPA octreotide scintigraphy. Eur J Cancer 1995; 31A: Fanti S, Farsad M, Battista G et al. Somatostatin receptor scintigraphy for bronchial carcinoid follow-up. Clin Nucl Med 2003; 28: Kurtaran A, Scheuba C, Kaserer K et al. Indium-111-DTPA-D-Phe-1-octreotide and technetium-99m-(v)-dimercaptosuccinic acid scanning in the preoperative staging of medullary thyroid carcinoma. J Nucl Med 1998; 39: Raderer M, Kurtaran A, Leimer M et al. Value of peptide receptor scintigraphy using 123 I-vasoactive intestinal peptide and 111 In-DTPA-D-Phe1-octreotide in 194 carcinoid patients: Vienna University Experience, 1993 to J Clin Oncol 2000; 18: Örlefors H, Sundin A, Ahlstrom H et al. Positron emission tomography with 5- hydroxytryprophan in neuroendocrine tumours. J Clin Oncol 1998; 16: Hoegerle S, Nitzsche E, Altehoefer C et al. Pheochromocytomas: detection with 18 F DOPA whole body PET--initial results. Radiology 2002; 222: Pacak K, Eisenhofer G, Carrasquillo JA, Chen CC, Li ST, Goldstein DS. 6- [ 18 F]fluorodopamine positron emission tomographic (PET) scanning for diagnostic localization of pheochromocytoma. Hypertension 2001; 38: Shulkin BL, Thompson NW, Shapiro B, Francis IR, Sisson JC. Pheochromocytomas: imaging with 2-[fluorine-18]fluoro-2-deoxy-D-glucose PET. Radiology 1999; 212: Chin R Jr, McCain TW, Miller AA et al. Whole body FDG-PET for the evaluation and staging of small cell lung cancer: a preliminary study. Lung Cancer 2002; 37: Pandit N, Gonen M, Krug L, Larson SM. Prognostic value of [ 18 F]FDG-PET imaging in small cell lung cancer. Eur J Nucl Med Mol Imaging 2003; 30: Schumacher T, Brink I, Mix M et al. FDG-PET imaging for the staging and follow-up of small cell lung cancer. Eur J Nucl Med 2001; 28: Hoegerle S, Ghanem N, Altehoefer C et al. 18 F-DOPA positron emission tomography for the detection of glomus tumours. Eur J Nucl Med Mol Imaging 2003; 30: Hoefnagel CA, Taal BG, Valdes Olmos RA. Role of [ 131 I]metaiodobenzylguanidine therapy in carcinoids. J Nucl Biol Med 1991; 35: Berglund AS, Hulthen UL, Manhem P, Thorsson O, Wollmer P, Tornquist C. Metaiodobenzylguanidine (MIBG) scintigraphy and computed tomography (CT) in clinical practice. Primary and secondary evaluation for localization of phaeochromocytomas. J Intern Med 2001; 249:

45 Chapter De Graaf JS, Dullaart RP, Kok T, Piers DA, Zwierstra RP. Limited role of metaiodobenzylguanidine scintigraphy in imaging phaeochromocytoma in patients with multiple endocrine neoplasia type II. Eur J Surg 2000; 166: van der Harst E, de Herder WW, Bruining HA et al. [ 123 I]metaiodobenzylguanidine and [ 111 In]octreotide uptake in benign and malignant pheochromocytomas. J Clin Endocrinol Metab 2001; 86: Hadj-Djilani NL, Lebtahi NE, Delaloye AB, Laurini R, Beck D. Diagnosis and follow-up of neuroblastoma by means of iodine-123 metaiodobenzylguanidine scintigraphy and bone scan, and the influence of histology. Eur J Nucl Med 1995; 22: Hashimoto T, Koizumi K, Nishina T, Abe K. Clinical usefulness of iodine-123- MIBG scintigraphy for patients with neuroblastoma detected by a mass screening survey. Ann Nucl Med 2003; 17: Pfluger T, Schmied C, Porn U et al. Integrated imaging using MRI and I metaiodobenzylguanidine scintigraphy to improve sensitivity and specificity in the diagnosis of pediatric neuroblastoma. AJR Am J Roentgenol 2003; 181: Erickson D, Kudva YC, Ebersold MJ et al. Benign paragangliomas: clinical presentation and treatment outcomes in 236 patients. J Clin Endocrinol Metab 2001; 86: Virotta G, Medolago G, Zappone C et al. Meta-[ I]iodobenzylguanidine single photon emission computed tomography in chemodectomas. J Nucl Med 1995; 39: Adalet I, Kocak M, Oguz H, Alagol F, Cantez S. Determination of medullary thyroid carcinoma metastases by 201 Tl, 99m Tc (V)DMSA, 99m Tc-MIBI and 99m Tctetrofosmin. Nucl Med Commun 1999; 20: Ugur O, Kostakglu L, Guler N et al. Comparison of 99m Tc(V)-DMSA, 201 Tl and 99m Tc-MIBI imaging in the follow-up of patients with medullary carcinoma of the thyroid. Eur J Nucl Med 1996; 23:

46 31

47 32

48 Chapter 2 Development of a reliable remote-controlled synthesis of β-[ 11 C]-5-Hydroxy-L-tryptophan on a Zymark robotic system Oliver C. Neels 1, Pieter L. Jager 1, Klaas P. Koopmans 1, Elisabeth Eriks 1, Elisabeth G.E. de Vries 2, Ido P. Kema 3, Philip H. Elsinga 1. Departments of Nuclear Medicine and Molecular Imaging 1, Medical Oncology 2 and Pathology and Laboratory Medicine 3, University of Groningen and University Medical Center Groningen, The Netherlands J Label Compd Radiopharm 2006; 49:

49 Chapter 2 Summary Precise staging of neuroendocrine tumors (NET) using positron emission tomography (PET) tracers visualizing their specific metabolic activity is of interest. Besides [ 18 F]FDOPA, staging NET with carbon-11 labeled 5-hydroxytryptophan (5-HTP) is reported in recent literature. We implemented the multi-enzymatic synthesis of enantiomerically pure [ 11 C]-L-5-HTP on a Zymark robotic system to compare both tracers in patient studies. [ 11 C]-5-HTP can be synthesized in up to 24% radiochemical yields (EOB). Average specific activity is 44,000 GBq/mmol in ca. 50 min from [ 11 C]methyl iodide in radiochemical purities > 99 %. The synthesis of 5-HTP is difficult due to its multi-enzymatic reaction steps but typical yields can be achieved of ca. 400 MBq. [ 11 C]-5- HTP is now reliably used in ongoing studies for staging NET. Introduction Neuroendocrine tumors (NETs) are slowly growing malignant tumors with the capacity of uptake and decarboxylation of amine precursors (APUD) such as 5-hydroxytryptophan (5- HTP) and L-dihydroxyphenylalanine (L-DOPA) 1. DOPA and 5-HTP are transported into the cell via LAT1 transporters 2. Fluorine 18 labeled L-DOPA 3 is in use for staging neuroendocrine tumors in several PET centers to measure the DOPA-pathway. 11 C labeled 5-HTP is also of special interest because it is a direct precursor for serotonin. As 5-HTP is decarboxylated by aromatic amino acid decarboxylase (AADC) the carbon-11 atom in carboxyl-position will be lost immediately and tumor detection will be impossible 4. Therefore, 5-HTP has to be labeled in an other than the carboxyl position, e.g. in the β- position which has been described by Bjurling et al. and applied in staging of NET 1,4,5,6. Radiosynthesis of β-[ 11 C]-5-hydroxy-L-tryptophan has been described earlier. First, a glycine derivate is labeled with [ 11 C]methyl iodide, hydrolyzed to racemic [ 11 C]alanine and converted by a multi-enzymatic reaction into enantiomerically pure [ 11 C]-L-5-HTP. Fully automated synthesis 7 is reported using immobilized 8,9 as well as free enzymes 10,11. We now report the implementation of this production on a Zymark robotic system. Results and Discussion [ 11 C]-5-HTP was obtained in decay corrected yields of 15 ± 12 % (figure 1) calculated from the time of release of [ 11 C]methyl iodide. This is in the range of earlier published results. Over time these yields increased up to 24%. In a typical run approximately 400 MBq of [ 11 C]-5-HTP was synthesized in approximately 50 min calculated from trapping of [ 11 C]methyl iodide (mean 9 GBq). Radiochemical purity was > 99 % and average specific activity was 44,000 GBq/mmol. The minimum amount of radioactive 5-HTP for a patient study was determined as 60 MBq. Average reliability after 36 syntheses was > 95 %. Enantiopurity of the obtained radiolabeled amino acid was assessed on a chiral HPLC column in > 99 %. The most critical part during the synthesis of [ 11 C]-5-HTP is the enzymatic step where in total 4 enzymes plus co-enzymes are used in an equilibrium reaction described by Bjurling et al. 10 and Watanabe et al. 12. The concentration of the used solutions is crucial for reliable yields. Racemic [ 11 C]alanine has to be converted into [ 11 C]pyruvic acid by the use of D- amino acid oxidase (DAO), catalase (CAT) and glutamic pyruvic transaminase (GPT) 34

50 Development of a reliable remote-controlled synthesis of β-[ 11 C]-5-Hydroxy-L-tryptophan on a Zymark robotic system (figure 2) and then into [ 11 C]-5-HTP by the reversed tryptophanase reaction (figure 3). The enzymatic reaction resulted only in good yields if the ph is adjusted in a range of Measuring and correction of ph is therefore important and can lead to dramatic decrease in yield if omitted ields in percentage Y Number of syntheses Figure 1. Yields [ 11 C]-5-HTP after trapping [ 11 C]methyl iodide corrected for decay Another issue for safe human use is the origin of enzymes. For example, catalase as reported by Bjurling et al was from the bovine liver. We used catalase from Aspergillus niger to avoid Bovine spongiform encephalopathy risk. Ph Ph N N-(Diphenylmethylene) glycine tert-butyl ester CO 2 C(CH 3 ) 3 1. [ 11 C]CH 3 I, KOH 60 C, 90 sec. 2. HCl 160 C, 10 min. H 2 N CO 2 H 11 CH 3 DL-[ 11 C]Alanine DAO CAT GPT α-kg O CO 2 H 11 CH 3 [ 11 C]Pyruvic acid Figure 2. Synthesis of DL-[ C]Alanine / [ C]Pyruvic acid NH 2 O CO 2 H 11 CH 3 [ 11 C]Pyruvic acid HO Tryptophanase (NH 4 ) 2 SO 4 N H HO H 2 11 C N H β-[ 11 C]-5-Hydroxy-L-tryptophan CO 2 H Figure 3. Synthesis of β-[ 11 C]-5-Hydroxy-L-tryptophan from [ 11 C]Pyruvic acid 35

51 Chapter 2 During the synthesis of [ 11 C]-5-HTP different parameters are playing important roles. The volume of the minivials used in the Zymark robot system is limited to 4 ml and the volume of the needle-hand has a maximum of 2.5 ml. Therefore the volume of liquids used in the SPE cleaning step had to be scaled down as compared to reported methods. All reaction vials are newly prepared for each synthesis. Reduction of radiation exposure for nuclear workers is a critical point in the synthesis of [ 11 C]-5-HTP. While replacing the SPE column, measuring and adjusting ph before and adding enzymes in the last step of the synthesis, the radiochemist is exposed to high levels of radiation for a period of about 1.5 min. Radiation dose per synthesis could be restricted to an average minimum of 260 μsv for the skin and 40 μsv for the whole body by use of a fast to open and close manual sliding door from lead. Experimental General For the synthesis of [ 11 C]-5-HTP we used a Zymark robotic system controlled by Easylab software (Hopkinton, MA, USA). Easylab software could easily be configured for [ 11 C]-5- HTP syntheses by adapting existing steps and changing parameters as reaction times and temperature. The robotic arm is centered in the hot cell and surrounded by workstations e.g. two ovens for heating (ambient temperature to 160 C) and cooling (-20 to +60 C), racks to hold septa-sealed (mini)-vials, air pressured solid phase extraction (SPE) purification and an injection unit for HPLC. Both ovens are combined with compatible needle devices that can shift up and down pneumatically and is controlled by Easylab software. Via the needle devices [ 11 C]methyl iodide as used in this synthesis can be trapped. Also helium can be streamed through the reaction vial by gas flow for evaporation. For standard procedures a finger hand and a ml needle hand are used. The sterile vial containing synthesized [ 11 C]-5-HTP is placed in a lead container situated near the sliding door of the hot cell. Chemicals and solvents were obtained from Sigma (Zwijndrecht, The Netherlands), Merck (Amsterdam, The Netherlands), Janssen (Geel, Belgium) and Rathburn (Walkerburn, Scotland). Enzymes were purchased from Sigma except tryptophanase (TRP). DAO from porcine kidney was dissolved in 3.6 M (NH 4 ) 2 SO 4 and ph adjusted to 6.5. CAT from Aspergillus niger in 3.2 M (NH 4 ) 2 SO 4 solution ph 6.0 and GPT from porcine heart in 1.8 M (NH 4 ) 2 SO 4 solution ph 6.0 were used without further treatment. TRP dissolved in 20 mm potassium phosphate buffer ph 7.5, 0.1 mm pyridoxal 5 -phosphate, 0.1 mm dithiotheitol and 20% glycerol was purchased from Ikeda (Hiroshima, Japan) and used without further treatment. Tris(hydroxymethyl)-aminomethane (TRIS), ammonium sulfate and α-ketoglutaric acid were dissolved in distilled water. Pyridoxal 5 -phosphate and flavin adenine dinucleotide were dissolved in 0.1 M sterile phosphate buffer ph 7.2 and 5- hydroxyindole (HIn) was dissolved in ethanol. Preparation of the synthesis is very time consuming compared to other PET tracers. Most of the necessary solutions have to be prepared manually the day before synthesis and are stable between 1 day and 4 weeks when cooled or frozen. Depending on the frequency of syntheses in a period of time, preparation can be more or less time-consuming by repeated use of the reagents. 36

52 Development of a reliable remote-controlled synthesis of β-[ 11 C]-5-Hydroxy-L-tryptophan on a Zymark robotic system Preparative HPLC was performed on a Waters (Etten-Leur, The Netherlands) 515 system using an Alltech (Breda, The Netherlands) 250 x 10 mm Econosphere C18 10 µm column (5 ml/min) with 0.1 M NaH 2 PO 4 containing 2% ethanol as mobile phase in series with a Waters 481 UV detector (280 nm) and an Eberline (Erlangen, Germany) RM25 radiation detector. Analytical HPLC was performed on a Waters 515 system equipped with a Waters 150 x 3.9 mm NovaPak C18 4 µm column (flow 1 ml/min) in series with a Waters 486 UV detector (280 nm) and an Eberline RM25 radiation detector and 0.1 M NaH 2 PO 4 as mobile phase. Enantiomeric purity was assessed by HPLC on a Waters 600 system using a Serva (Heidelberg, Germany) pre-conditioned (0.1 M CuNO 3 ) 250 x 4.6 mm ChiralProCu 6 µm column (1 ml/min) in series with a Waters 2487 UV detector (280 nm) and a Bicron (Paris, France) frisk-tech radiation detector with 0.1 M NaH 2 PO 4 as mobile phase. C18 Solid phase extraction columns were pre-conditioned with 10 ml ethanol and 10 ml distilled water. [ 11 C]Methyl iodide [ 11 C]Methane was produced in an aluminum target by the 14 N(p,α) 11 C nuclear reaction in N 2, containing 10% H 2, using 17 MeV protons and a Scanditronix MC-17 cyclotron (1 hour, 30 μa). After trapping [ 11 C]methane in a software-controlled module, [ 11 C]methyl iodide was formed in average yields of 50 % by reaction with iodide vapour at 725 C in 10 min. Obtained [ 11 C]methyl iodide was transferred to a second hot cell via PEEK tubings into a minivial containing the precursor solution. β-[ 11 C]-5-Hydroxy-L-tryptophan 3 mg of N-(diphenylmethylene)glycine tert-butyl ester (10 µmol) was dissolved in 290 µl dimethylformamide. 7 µl potassium hydroxide (KOH) 5 M was added manually just before 11 trapping [ C]methyl iodide by the needle device at 0 C. The needle device contains one long needle to trap [ 11 C]methyl iodide and one short needle equipped with a carbosphere trap. After trapping of [ 11 C]methyl iodide the minivial containing the yellow mixture was moved to the preheated oven (60 C) with the finger-hand to react for 90 sec. The minivial was then placed into the minivial holder and a mixture of water/0.1 M phosphate buffer ph 7.2/ethanol (2.27 ml/0.77 ml/0.3 ml) was added by the needle-hand. The cloudy liquid was transferred to the C18 SPE column, passed by air pressure via a three-way valve to a waste bottle and washed with 2 ml water. After switching the three-way valve the radioactive compounds were eluted with 2 ml dichloromethane to a minivial containing 200 µl HCl 6 M. The minivial was set into a pre-heated oven (70 C) by the finger-hand and then heated up to 160 C in a stream of helium gas. During evaporation the SPE-column was replaced manually by a 6 ml syringe connected with a 22 µm sterile pore filter. After complete evaporation the minivial was placed in a cooled down oven (0 C) for 1 min. 1 ml of 0.1 M TRIS was added by the needle-hand after placing the minivial into the vial rack. The minivial was slightly shaken manually and 1 ml air was added by the needle-hand to avoid low pressure in the minivial after rapid cooling down. The clear solution was transferred with the needle-hand to a minivial containing 100 µl ammonium sulfate 1.5 M, 80 µl TRIS 0.5 M ph 9, 15 µl KOH 5 M, 50 µ l α-ketoglutaric acid 0.2 M, 10 µl pyridoxal 5 -phosphate 10.7 mm and 10 µl flavin adenine dinucleotide 1.8 mm. ph was measured with a hand-held ph-meter and adjusted to by adding 37

53 Chapter 2 KOH 5 M manually with a pipette. A prepared mixture of 11 µl GPT (20 units), 20 µl CAT (3,600 units), 250 µl DAO (12 units) and 50 µl TRP (8 units) was added manually with a pipette. Finally, 10 µl HIn 0.5 M were added with a pipette, slightly shaken manually and the solution was incubated for 8 minutes. Enzymes were denatured by manual addition of 3 drops of HCl 6 M. The suspension was transferred by the needle-hand to the SPE-station and filtered into an empty minivial via a 22 µm sterile pore filter by air pressure and washed with 0.6 ml HPLC eluens. The minivial containing a clear solution was transferred to the HPLC injection station by the finger-hand and injected on the preparative HPLC column. β-[ 11 C]-5-Hydroxy-Ltryptophan was obtained after purification on the described preparative HPLC system with a retention time of 9 minutes. The radioactive fraction of 5-HTP was collected by passage through a 22 μm sterile pore filter in a sterile vial ready for use in human studies. The sterile vial was placed in a lead container near the door of the hot cell by the finger-hand. Conclusion Because of its multi-enzymatic steps [ 11 C]-5-HTP synthesis is difficult, but with slight modifications the end result is suitable for reliable patient studies. With a high specific activity and increasing yields this tracer can be successfully applied for patient studies. Acknowledgements We would like to thank the group of Bengt Långström, Uppsala, Sweden for assistance in implementation of [ 11 C]-5-HTP. This work was supported by grant of the Dutch Cancer Society, Amsterdam, The Netherlands. References 1. Örlefors H, Sundin A, Garske U, Juhlin C, Öberg K, Skogseid B, Långström B, Bergström M, Eriksson B. Whole-Body 11 C-5-Hydroxytryptophan Positron Emission Tomography as a Universal Imaging Technique for Neuroendocrine Tumors: Comparison with Somatostatin Receptor Scintigraphy and Computed Tomography. J Clin Endocrinol Metabol 2005; 90: Uchino H, Kanai Y, Kim D, Wempe M, Chairoungdua A, Morimoto E, Anders MW, Endou H. Transport of amino acid-related compounds mediated by L-type amino acid transporter 1 (LAT1): insights into the mechanisms of substrate recognition. Mol Pharmacol 2002; 61: De Vries E, Luurtsema G, Brussermann M, Elsinga P, Vaalburg W. Fully automated synthesis module for the high yield one-pot preparation of 6- [ 18 F]fluoro-L-DOPA. Appl Radiat Isot 1999; 51: Sundin A, Eriksson B, Bergström M, Bjurling P, Lindner K, Öberg K, Långström B. Demonstration of [ 11 C] 5-hydroxy-L-tryptophan uptake and decarboxylation in carcinoid tumors by specific positioning labeling in positron emission tomography. Nucl Med Biol 2000; 27:

54 Development of a reliable remote-controlled synthesis of β-[ 11 C]-5-Hydroxy-L-tryptophan on a Zymark robotic system 5. Eriksson B, Örlefors, H, Öberg, K, Sundin A, Bergström M, Långström B. Developments in PET for the detection of endocrine tumours. Best Pract Res Clin Endocrinol Metabol 2005; 19: Öberg K, Eriksson B. Nuclear medicine in the detection, staging and treatment of gastrointestinal carcinoid tumours. Best Pract Res Clin Endocrinol Metabo 2005; 19: Harada N, Nishiyama S, Sato K, Tsukada H. Development of an automated synthesis apparatus for L-[3-11 C] labeled aromatic amino acids. Appl Radiat Isot 2000; 52: Ikemoto M, Sasaki M, Haradahira T, Yada T, Omura H, Furuya Y, Watanabe Y, Suzuki K. Synthesis of L-[β- 11 C]amino acids using immobilized enzymes. Appl Radiat Isot 1999; 50: Sasaki M, Ikemoto M, Mutoh M, Haradahira T, Tanaka A, Watanabe Y, Suzuki K. Automatic synthesis of L-[β- 11 C]amino acids using an immobilized enzyme column. Appl Radiat Isot 2000; 52: Bjurling P, Watanabe Y, Tokushige M, Oda T, Långström B. Synthesis of β- 11 C- L-tryptophan and 5-hydroxy-L-tryptophan using a multi-enzymatic reaction route. J Chem Soc Perkin Trans I 1989; Bjurling P, Antoni G, Watanabe Y, Långström B. Enzymatic synthesis of carboxy- 11 C-labelled L-tyrosine, L-DOPA, L-tryptophan and 5-hydroxy-L-tryptophan. Acta Chem Scand 1990; 44 : Watanabe T, Snell EE. Reversibility of the tryptophanase reaction. Synthesis of tryptophan from indole, pyruvate, and ammonia. Proc Nat Acad Sci USA 1972; 69:

55 40

56 Chapter 3 nipulation of [ Ma C]HTP and [ F]FDOPA accumulation in neuroendocrine tumor cells Oliver C. Neels 1, Klaas P. Koopmans 1, Pieter L. Jager 1, Laya Vercauteren 2, Aren van Waarde 1, Janine Doorduin 1, Hetty Timmer-Bosscha 3, Adrienne H. Brouwers 1, Elisabeth G.E. de Vries 3, Rudi A. Dierckx 1, Ido P. Kema 4, Philip H. Elsinga 1. Departments of Nuclear Medicine and Molecular Imaging 1, Medical Oncology 3 and Pathology and Laboratory Medicine 4, University of Groningen and University Medical Center Groningen, The Netherlands Department of Pharmacy 2, University of Ghent, Belgium Submitted for publication. 41

57 Chapter 3 Summary [ 11 C]-5-Hydroxytryptophan ([ 11 C]HTP) and 6-[ 18 F]fluoro-levodopa ([ 18 F]FDOPA) are used to image neuroendocrine tumors (NETs) with positron emission tomography (PET). The precise mechanism by which these tracers accumulate in tumor cells is unknown. We aimed to study tracer uptake via large amino acid transporters (LAT), peripheral decarboxylation (inhibited by carbidopa) and intracellular breakdown by monoamine oxidase (MAO). [ 11 C]HTP and [ 18 F]FDOPA tracer accumulation was assessed in a human neuroendocrine tumor cell line BON. The carbidopa experiments were performed in a tumor bearing mouse model. Intracellular [ 11 C]HTP accumulation was 2-fold higher than [ 18 F]FDOPA. Cellular transport of both tracers was inhibited by amino-2-norbornanecarboxylic acid. The MAO inhibitors clorgyline and pargyline increased tracer accumulation in vitro. Carbidopa did not influence tracer accumulation in vitro but improved tumor imaging in vivo. Despite lower accumulation in vitro, visualization of [ 18 F]FDOPA is superior to [ 11 C]HTP in the neuroendocrine pancreatic tumor xenograft model. This could be a consequence of the serotonin saturation of BON cells in the in vivo model. Introduction Because of high sensitivity positron emission tomography (PET) studies for imaging of neuroendocrine tumors (NETs) have recently raised interest 1,2,3,4,5,6,7. To visualize these tumors the principle of amine precursor uptake and decarboxylation (APUD) 8 and imaging of the somatostatin receptor plays a major role. Neuroendocrine tumors possess the unique property of synthesis, storage and secretion of biogenic amines. Clinically applied tracers to visualize this are 6-[ 18 F]fluoro-levodopa ([ 18 F]FDOPA) and [ 11 C]-5-hydroxytryptophan ([ 11 C]HTP). Despite the clinical utility of these tracers, little is known about the precise mechanisms that govern their accumulation in tumor cells. Several factors are potentially involved in this accumulation. The first factor is the fact, as both tracers are amino acid derivatives, that they may well be a substrate for transmembrane amino acid transporters. The Na + - independent transporters LAT1, LAT2, LAT3 and LAT4 are defined as system L transporters and can be blocked by the model substrate amino-2-norbornanecarboxylic acid (BCH) 9. LAT transporters are responsible for the transport of large neutral amino acids across the cellular membrane. Secondly, the precise effects of the peripheral decarboxylation inhibitor carbidopa, generally given to patients before [ 11 C]HTP or [ 18 F]FDOPA tracer injection in order to improve uptake, are poorly understood. A third factor involved can be monoamine oxidase (MAO). It plays a major role in the metabolism of tryptophan and levodopa and is the enzyme responsible for the degradation of serotonin (5-HT) to 5-hydroxyindole acetic acid (5-HIAA) and of dopamine to homovanillic acid. Based on inhibitor sensitivity and substrate selectivity MAOs are subtyped as MAO A and B 10,11. While MAO A is mainly responsible for the degradation of 5-HT, MAO B breaks down both 5-HT and dopamine 12. With the use of selective and nonselective MAO inhibitors like clorgyline (MAO A) and pargyline (MAO A & B) the effect of the MAOs on tracer trapping can be examined 13,14. An increased accumulation due to reduced intracellular metabolism of [ 11 C]HTP and [ 18 F]FDOPA can be expected by the use of MAO inhibitors. 42

58 Manipulation of [ 11 C]HTP and [ 18 F]FDOPA accumulation in neuroendocrine tumor cells The aim of the present study was the analysis of the uptake of [ 11 C]HTP and [ 18 F]FDOPA by large amino acid transporters (LAT), peripheral decarboxylation by amino acid d ecarboxylase (AADC) and intracellular breakdown by monoamine oxidase (MAO) in vitro. Finally, experiments with carbidopa were extended to a tumor -bearing animal model to get a better understanding of the in vivo metabolism, using micropet. Materials and Methods Materials Tracers Synthesis of [ 18 F]FDOPA was carried out as described earlier 15 with an average specific a ctivity of 9.8 GBq/mmol after end of synthesis. The tracer was diluted with 0.9 % saline to the required concentration of ~ 1 MBq/ml ± 0.01 ml [ 18 F]FDOPA corresponding to a dose of 6.17 ± 0.63 MBq was injected per animal. [ 11 C]HTP was synthesized via an enzymatic method with an average specific activity of 30,000 GBq/mmol after end of sy nthesis as recently described 16. It was used at a concentration of ~ 20 MBq / ml ± ml [ 11 C]HTP, corresponding to a dose of ± 1.18 MBq were injected per animal. Chemicals 5-HTP, 5-HT, 5-HIAA, carbidopa, clorgyline, pargyline hydrochloride and BCH were purchased from Sigma (Zwijndrecht, The Netherlands). GMC (5.6 mm D-glucose, 0.49 mm MgCl 2, 0.68 mm CaCl 2 ) was added to phosphate-buffered saline solution (PBS; 140 mm NaCl, 2.7 mm KCl, 6.4 mm Na 2 HPO 4, 0.2 mm KH 2 PO 4 ). ph of the resulting PBS- GMC solutions was adjusted to 7.4 with sodium hydroxide. PBS-GMC was used to deplete internal amino acid pools 17,18. Matrigel was purchased from BD Biosciences (Erembodegem, Belgium). In vitro experiments Cell culture method Experiments were performed with the human neuroendocrine pancreatic tumor cell line BON 19. Cells were maintained in 25 cm 2 culture flasks in 5 ml D-MEM/F-12 (1:1) medium supplemented with 10 % fetal calf serum (FCS) containing the amino acids L-phenylalanine (215 µmol) and L-tryptophan (44 µmol) amongst others. Cells were grown in a humidified atmosphere containing 5 % CO 2 and were routinely subcultured every 3-4 days. Cultures grown to a cell density of 1.0 x x 10 6 cells per ml were used for experiments. Cells were harvested by trypsin treatment, resuspended and diluted 3:7 in culture medium on 12 wells plates 1 day before the experiments (1 ml per well). Viability and number of cells were determined by the trypan blue exclusion technique. Cell viability 1-2 hours after experiments was over 90 %. At the day of experiment, cells were washed with warm PBS (3 x 2 ml) and 1 ml of culture medium or PBS-GMC buffer per well was added. The cells were then placed in a water bath for 1 hour at 37 C before start of the experiments to allow depletion of internal amino acids. Accumulation experiments were started by addition of 150 µl [ 18 F]FDOPA (~ 0.2 MBq) or 60 µl [ 11 C]HTP (~ 1.2 MBq) of the solution in each well. 43

59 Chapter 3 Determination of intracellular tracer accumulation After completion of the experiment, buffer was removed and cells were washed with icecold PBS (3 x 2 ml) and harvested by addition of 200 µl trypsin per well. 1 ml growth medium per well was added, cells were resuspended, transferred to 10 ml tubes and counted in a gamma counter (Compugamma, LKB Wallac, Finland). Measurements of tracer accumulation were expressed as percentage of the radioactive dose per 1 x 10 5 cells. All results were corrected for non-specific accumulation. For the determination of non-specific tracer accumulation all washing procedures were done with ice-cold PBS. Experiments were performed in ice-cold culture medium or PBS-GMC buffer and wells-plates were placed on ice. Tracer accumulation at 0 C was considered as non-specific binding. Results represent the mean of 3-4 experiments ± standard error of the mean. Individual experiments were performed in duplicate. Inhibition experiments Various concentrations of the blocking agent BCH (0-20 mm) were applied to determine an adequate blocking concentration. To maintain cell viability carbidopa was used at a maximum concentration of 0.08 mm as higher concentrations of carbidopa induced 20,21 apoptosis. Clorgyline and pargyline were added in a concentration of 0.1 mm. Inhibition experiments were carried out by adding 1 ml of culture medium (control and carbidopa only) or PBS-GMC containing the blocking agent in the relevant concentration to each well. Afterwards cells were incubated for 1 hour in order to achieve the required amino acid depletion. Subsequently, tracer incubation was performed and intracellular accumulation determined as described above. A tracer incubation time of 15 minutes was used for the blocking agent BCH. Tracer incubation times ranging of 5-60 minutes with AADC and MAO inhibitors carbidopa, clorgyline and pargyline were used. Due to the short half-life of 11 C, incubation periods longer than 60 minutes were not applied. Non-labeled tracer accumulation Because of the short half-life of PET isotopes the detection and quantification of radioactive 5-HTP metabolites like 5-HT or 5-HIAA is limited. A sensitive automatic detection method of 5-HTP and its metabolites used in carcinoid patients was used for the detection of non-labeled tracer in vitro 22. For these experiments culture medium was 2 removed from the 25 cm culture flasks. Cells were then washed with PBS (3 x 2 ml). 5 ml PBS-GMC buffer containing the carbidopa (0.08 mm), the MAO A inhibitor clorgyline (0.1 mm) or the non-selective MAO inhibitor pargyline (0.1 mm) was added. After the 1 hour depletion period, non radioactive labeled 5-HTP (55 nm) was added in 5 ml PBS- GMC buffer. After 15 and 60 minutes of tracer incubation, PBS-GMC buffer was removed. The buffer supernatant was analyzed for the amounts of 5-HIAA. Pre concentration was performed by liquid liquid extraction in the following way. Per sample 2 drops of glacial acetic acid, 1 g NaCl, 5 ml diethyl ether were added and slightly shaken. After centrifugation at 2,000 g for 5 minutes the organic phase was transferred into a test-tube and diethyl ether evaporated in a slight stream of nitrogen. Samples were dissolved in eluent and analyzed as described earlier

60 Manipulation of [ 11 C]HTP and [ 18 F]FDOPA accumulation in neuroendocrine tumor cells For the analysis of 5-HTP metabolites supernatant was removed, the cells were harvested with 1 ml trypsin and resuspended in 1 ml PBS containing 10 % FCS. Cells were washed three times with ice-cold PBS (1 ml) and centrifuged for 10 minutes at 10,000 g. Cells were lysed using liquid nitrogen. The concentrations of 5-HTP, 5-HT and 5-HIAA present in the lysed BON cells dissolved in 1 ml saline were determined by the method described above. Cellular accumulation of non-labeled tracer was defined as amount of substance detected in lysed cells divided by the total amount of 5-HTP in incubation medium at the start of the experiment. All results were represented using the mean of 3 experiments ± standard error of the mean. In vivo experiments Animals Nude male mice (BALB/c, age 6-8 weeks, body weight 18 to 24 grams) were obtained from Harlan Netherlands BV (Horst, The Netherlands). Experimental groups consisted of 4-5 animals to perform micropet scanning after injection of [ 11 C]HTP and [ 18 F]FDOPA. The total number of studied animals was 36. Animals were housed in temperature and humidity controlled rooms with 12-hours day and 12-hours night cycles and were provided with forage and water ad libitum. Animals were housed in hepa-filtered cages in the animal research facility of the University Medical Center Groningen under controlled water, lab chow, humidity and temperature conditions. At least 2 hours before starting experiments the animals were acclimated to laboratory conditions. All animal experiments were performed by licensed investigators in compliance with the Law on Animal Experiments of The Netherlands. The study protocol was approved by the Committee on Animal Ethics of the University of Groningen. MicroPET scanning 60 minutes dynamic scanning followed by 10 minutes transmission scanning was performed using a Concorde micropet Focus 220 system equipped with micropet manager for data acquisition in list mode and ASIPro for preparing sinograms and image reconstruction. Using ASIPro's clipping tool, areas with very high activity that were not relevant to the current study, such as the liver region, were removed as to yield a more cleaned up version of the scan that is, therefore, easier to evaluate. Ordered subset expectation maximization (OSEM2D) statistics was applied for the quantitative analysis. The PET acquisition data were fully corrected for dead time, random coincidences, attenuation and scatter. PET image size was 128 x 128 x 95 voxels. In all experiments, cells were harvested with trypsin and resuspended in 1 ml growth medium/matrigel 1:1 and injected subcutaneously (SC, 1 x 10 6 cells/injection) into the right shoulder of the animals to establish tumors. Growth of tumors was checked three times a week. After approximately 3 weeks growth, a 10 mm tumor size was reached and considered suitable for experiments. Intravenous (IV) injection and scanning procedure were performed under isoflurane anesthesia. 4-5 animals from each group received carbidopa (1 mg/kg) intraperitoneally (IP) in the abdominal region 1 hour before tracer injection. Thereafter, radioactive tracers ([ 11 C]HTP: 0.50 ± 0.24 MBq/g bodyweight; [ 18 F]FDOPA: 0.29 ± 0.13 MBq/g bodyweight) were injected IP in the abdominal region or IV via the penile vein. After scanning for 70 minutes (60 minutes dynamic scanning, 10 45

61 Chapter 3 minutes transmission scan), animals were sacrificed to determine tracer accumulation in tumor tissue and in different body parts e.g. liver, kidney, pancreas, intestines, brain. Radioactivity was measured in a gamma counter. Measurements of tracer accumulation were expressed as percentage of the injected dose/g bodyweight. Statistics Differences between various groups were tested for statistical significance using Student s t-test for independent samples. P-values < 0.05 were considered significant. Results In vitro experiments Time-course of tracer accumulation Cellular accumulation of [ 18 F]FDOPA and [ 11 C]HTP of cells incubated in culture medium over a period of 60 minutes was rapid, but low ([ F]FDOPA 0.07 ± 0.01 %, [ C]HTP 0.15 ± 0.01 %/10 5 cells). However, in amino acid free PBS-GMC buffer both tracers showed very rapid accumulation and under these conditions considerably higher levels of accumulation were reached. [ 18 F]FDOPA was accumulated to 1.2 ± 0.2 %/10 5 cells after 15 minutes and remained constant up to the end of the 60 minutes incubation period. [ 11 C]HTP accumulation was much higher than [ 18 F]FDOPA (ratio 5:1) with a maximum tracer accumulation of 5.3 ± 0.8 %/10 5 cells at 60 minutes (figure 1). Inhibition experiments Incubation of cells with carbidopa did not affect accumulation of both [ 11 C]HTP and [ 18 F]FDOPA after 60 minutes, neither in full culture medium nor in PBS-GMC buffer. Clorgyline preincubation led to significant higher accumulation compared to control for [ 11 C]HTP (14.2 ± 3.8 %/10 5 cells after 60 minutes) and [ 18 F]FDOPA (9.2 ± 2. 9 %/10 5 cells after 60 minutes) in PBS-GMC buffer. With pargyline slightly higher accumulation 18 5 compared to control was obtained for [ F]FDOPA (3.7 ± 0.7 %/10 cells after 60 minutes) (figure 1). Accumulation experiments with BCH were performed at the time point of 15 minutes since [ 18 F]FDOPA reached a plateau phase of accumulation at 15 minutes in untreated BON cells. At 1.0 mm BCH, the accumulation of [ 18 F]FDOPA and [ 11 C]HTP was inhibited to levels of 0.43 ± 0.02%/10 5 cells ([ 18 F]FDOPA) and 0.22 ± 0.08 %/10 5 cells ([ 11 C]HTP) and suppressed at 20 mm BCH. The BCH IC 50 value for [ 11 C]HTP is 0.12 mm (figure 2). The lowest used concentration of 0.03 mm BCH resulted in a reduction of tracer accumulation to 45 % compared to control for [ 18 F]FDOPA and a BCH IC 50 value of 0.01 mm. Non-labeled tracer accumulation In BON cells (control) apart from 5-HTP both 5-HT and 5-HIAA were detected, 15 and 60 minutes after starting incubation. Only low cellular 5-HT levels (0.6 ± 0.1 % after 60 minutes) were found compared to 5-HTP (12.0 ± 0.0 % after 60 minutes) and 5-HIAA (11.9 ± 0.8 % after 60 minutes) levels. Treatment with carbidopa did neither increase 5- HTP (3.0 ± 0.9 % after 60 minutes) nor decrease cellular levels of 5-HT (3.0 ± 0.3 % after 60 minutes) or 5-HIAA (11.8 ± 1.4 % after 60 minutes). Clorgyline however increased 5- HT (18.4 ± 2.9 % after 60 minutes) and decreased 5-HIAA levels (0.9±0.0 % after 60 46

62 Manipulation of [ 11 C]HTP and [ 18 F]FDOPA accumulation in neuroendocrine tumor cells minutes). Similar results were obtained after pargyline treatment (5-HT: 29.3 ± 5.0 %, 5- HIAA: 2.4 ± 0.2 % after 60 minutes) (figure 3) Control Carbidopa Pargyline Clorgyline * * % / 10E5 cells 10 * * 5 * incubation time (min) Control Carbidopa Pargyline Clorgyline % / 10E5 cells 10 * * * * 5 * * * * * incubation time (min) Figure 1. Accumulation of [ 11 C]HTP (top) and [ 18 F]FDOPA (bottom) in BON cells. Tracer accumulation was measured in amino acid free medium and was corrected for non-specific binding. Mean ± SEM of 3-4 experiments. *P 0.05 compared to control. 47

63 Chapter [ 11 C]HTP [ 18 F]FDOPA 80 % of control concentration BCH (mm) Figure 2. Accumulation of [ 11 C]HTP and [ 18 F]FDOPA in BON cells after inhibition of LAT. Tracer accumulation was measured in amino acid free medium and was corrected for non-specific binding. A concentration of mm BCH is set as control. Incubation time: 15 minutes. Mean ± SEM of 3 experiments. BCH IC 50 = 0.12 mm for [ 11 C]HTP, 0.01 mm for [ 18 F]FDOPA. Animal experiments 36 mice were divided in 8 groups of 4-5 animals (control and carbidopa) and given [ 18 F]FDOPA or [ 11 C]HTP, IP or IV. In 33 of 36 mice micropet scans, tumors were visualized. Tumor weights on day of experiments ranged from 19 mg to 311 mg (average 118 mg) after decapitation. In 3 mice ([ 18 F]FDOPA IP control, [ 18 F]FDOPA IV carbidopa, [ 11 C]HTP IV carbidopa) scans, tumors were not visualized. In the tumor region following IV and IP administration [ 18 F]FDOPA in combination with carbidopa pretreatment gave the highest specific uptake values (SUVs) (figure 4). Within 5 minutes after IV tracer injection the plateau phase of the maximum tumor SUV for all experiments was reached except for [ 18 F]FDOPA combined with carbidopa given IP in the abdominal region. IP tracer injection showed increasing tumor SUV over time reaching a plateau phase later than 1 hour. Treatment with carbidopa resulted in higher SUVs for tumors compared to controls for both For IV injections both carbidopa treated and control, tracers within 60 minutes time. 18 [ F]FDOPA generated significant higher tumor SUVs than those obtained after [ C]HTP. Biodistribution data after scanning demonstrated high SUVs for liver, kidney and pancreas for both tracers (IV). SUVs were not significantly altered after carbidopa treatment compared to control (figure 5). Although tumor SUVs in micropet images were higher for [ 18 F]FDOPA, [ 11 C]HTP resulted in higher uptake in most organs and significant differences in uptake were noted for spleen, red blood cells and blood plasma irrespective of carbidopa treatment

64 Manipulation of [ 11 C]HTP and [ 18 F]FDOPA accumulation in neuroendocrine tumor cells Control Carbidopa Clorgyline Pargyline % of administered 5-HTP * * * * 0 * 5-HTP 5-HT 5-HIAA * Control Carbidopa Clorgyline Pargyline * % of administered 5-HTP * 0 * * * * * * 5-HTP 5-HT 5-HIAA Figure 3. 5-HTP and metabolites in BON cells after incubation with 5-HTP. Top: 15 minutes incubation time. Bottom: 60 minutes incubation time. Mean ± SEM of 3 experiments. *P 0.05 compared to control. 49

65 Chapter 3 1,0 0,8 [ 11 C]HTP [ 11 C]HTP with Carbidopa [ 18 F]FDOPA [ 18 F]FDOPA with Carbidopa SUV tumour 0,6 0,4 * ** * * * * * ** 0,2 0, time in sec SUV tumour 1,0 0,8 0,6 0,4 *** *** *** [ 11 C]HTP [ 11 C]HTP with Carbidopa [ 18 F]FDOPA [ 18 F]FDOPA with **** Carbidopa *** *** *** *** *** **** **** **** *** **** 0,2 0, time in sec Figure 4. Tracer accumulation in tumor after IP (top) and IV (bottom) injection expressed as SUV. Graphics represent mean ± SEM of 4-5 animals. P 0.05: *[ 18 F]FDOPA: Carbidopa vs. control, **Control: [ 11 C]HTP vs. [ 18 F]FDOPA, ***Control: [ 18 F]FDOPA vs. [ 11 C]HTP, ****Carbidopa: [ 18 F]FDOPA vs. [ 11 C]HTP. 50

66 Manipulation of [ 11 C]HTP and [ 18 F]FDOPA accumulation in neuroendocrine tumor cells 14 [ 11 C]HTP [ 11 C]HTP with Carbidopa [ 18 F]FDOPA 12 [ 18 F]FDOPA with Carbidopa 10 SUV [ 11 C]HTP [ 11 C]HTP with Carbidopa [ 18 F]FDOPA [ 18 F]FDOPA with Carbidopa Tumor Liver Kidney Pancreas Cerebellum Cerebral cortex Rest brain Adipose tissue Bone Heart Large intestine Small intestine Lung Muscle Plasma Red blood cells Spleen Subm. gland 10 SUV Tumor Liver Kidney Pancreas Cerebellum Cerebral cortex Rest brain Adipose tissue Bone Heart Large intestine Small intestine Lung Muscle Plasma Red blood cells Spleen Subm. gland Figure 5. Accumulation of [ C]HTP and [ F]FDOPA in different organs 70 min after IP (top) and IV (bottom) injection, expressed as SUV. In separate experiments animals were pretreated with carbidopa. Values express mean ± SEM from 4-5animals. 51

67 Chapter 3 Discussion Cellular accumulation of [ 11 C]HTP and [ 18 F]FDOPA in neuroendocrine tumor cells is rapid and reaches a plateau after 15 minutes in vitro and within less than 5 minutes in the tumor in animal model after IV injection. We identified LAT, AADC and MAO as factors affecting intracellular tracer accumulation. Inhibition of amino acid transport resulted in a nearly complete shutdown of accumulation, illustrating that this mechanism is a key factor in intracellular accumulation. Carbidopa did not influence cellular accumulation of both tracers in tumor cells in vitro but did increase tumor accumulation of radioactivity in animals. In vitro, selective inhibition of MAO A by clorgyline induced increased accumulation of both tracers, confirming that MAO is a third important factor affecting their biodistribution. Non-selective inhibition of MAO by pargyline only increased [ 18 F]FDOPA accumulation. We performed in vitro cellular accumulation experiments in culture medium because the levels of large amino acids as present in D-MEM/F-12 culture medium are similar to the ones in blood plasma 23. The very low rates of [ 18 F]FDOPA and [ 11 C]HTP accumulation suggest that other amino acids in culture medium are competing for accumulation with these radiolabeled tracers. Thereafter the amino acid free PBS, in which cells remain viable during the 2 hours test period, was used. PBS was supplemented with D-glucose, magnesium chloride and calcium chloride as used by Jager et al. 18. [ 11 C]HTP was accumulated twice as much as [ 18 F]FDOPA over a period of 60 minutes. This could be a consequence of the fact that BON cells produce more 5-HT than dopamine 24,25. Retention of neurotransmitters is considered to be the resultant of uptake (LAT), decarboxylation (AADC) and granular storage by vesicular monoamine transporters (VMAT). The latter process prevents enzymatic breakdown in the cytoplasm (MAO) and subsequent secretion. One of the most important factors of VMAT activity is the amount of 5-HT or catecholamines in secretory vesicles 26,27. In our cell studies the medium was free of amino acids and metabolites tentatively resulting in reduced filling of secretory granules. This theoretically led to higher VMAT-substrate activity. Separate granular storage of 5-HT and catecholamines and may explain differences in retention for [ 11 C]HTP and [ 18 F]FDOPA The cellular accumulation of [ 11 C]HTP and [ 18 F]FDOPA is blocked by low concentrations of BCH, a conventional inhibitor of the amino acid transporter system L 18. An interesting aspect regarding the growth of BON tumor cells is its dependence on large amino acids. If the supply of those amino acids could be disrupted e.g. by BCH, proliferation would possibly be slowed down or stopped. Recently a dose-dependant inhibition of growth of C6 glioma cells was observed following BCH exposure in vitro and in vivo 22. Örlefors et al. reported in patients an improved uptake of [ 11 C]HTP in carcinoid tumors and decreased urinary 5-HIAA levels when [ 11 C]HTP was administered after oral administration of carbidopa 31. This was suggested to be the result of decreased conversion of [ 11 C]HTP and [ 18 F]FDOPA to [ 11 C]5-HT/[ 18 F]dopamine by activity of the AADC enzyme in peripheral tissues such as the liver and kidneys 32,33,34. They hypothesized that the degradation of [ 11 C]HTP to [ 11 C]-5-HT in peripheral organs is blocked by carbidopa and thus increases the availability of [ 11 C]HTP resulting in higher accumulation of radiolabeled tracer in tumor lesions. Our results validate this hypothesis. Carbidopa did not affect the accumulation of [ 11 C]HTP or [ 18 F]FDOPA in BON cells. This suggests that intracellular 52

68 Manipulation of [ 11 C]HTP and [ 18 F]FDOPA accumulation in neuroendocrine tumor cells decarboxylation in BON cells was not inhibited. AADC activity was found to be upregulated in NET cells 36. In nude mice bearing a BON tumor, carbidopa increased tracer tumor accumulation within the tumor resulting in better imaging (figure 6). Our mice PET images show high accumulation in the abdominal region. This is in line with our mice biodistribution data with high SUVs for organs such as liver and kidney located in this region. In the tumor model which we used, which was derived from a human pancreatic tumor cell line [ 18 F]FDOPA accumulation was higher than [ 11 C]HTP. Due to the high 5-HT production by carcinoid tumors 22, the granules in tumor cells of our animals are expected to be saturated with 5-HT. This may reduce granular storage of [ 11 C]-5-HT which results in cytoplasmic breakdown (MAO) and subsequent excretion of metabolites. As noted above, different granular storage of 5-HT and catecholamines may explain differences in tracer retention. Figure 6. Full-color in appendix. A - [ 11 C]HTP PET after IV injection, coronal view. Left: control 22.6 g, 9.9 MBq, tumor weight 23.5 mg. Right: carbidopa treated 21.6 g, 6.2 MBq, tumor weight 61.1 mg. B - [ 18 F]FDOPA PET after IV injection, coronal view. Left: control 20.5 g, 8.4 MBq, tumor weight 109 mg. Right: carbidopa treated 23.6 g, 8.4 MBq, tumor weight 57.9 mg. Summed frames. Hot spots in abdominal region were cleaned up using ASIPro s clipping tool. Arrows point at tumors located in the right shoulder. 53

69 Chapter 3 Increased in vitro accumulation was noticed after exposure to the MAO A inhibitor clorgyline. This result differs from reported PET studies with the MAO inhibitors harmine and deprenyl. Harmine and deprenyl both decreased the accumulation of [ 18 F]FDG and [ 11 C]DOPA in BON cells 37. Degradation of [ 18 F]fluorodopamine and [ 11 C]-5-HT appears to be blocked by clorgyline. Pargyline gives a slightly higher [ 18 F]FDOPA accumulation compared to control in accordance with the fact that pargyline is an inhibitor for MAO B. Higher concentrations of pargyline could probably also lead to an increased [ 11 C]HTP accumulation. In our institution, sensitive analytical methods to profile tryptophan related plasma indoles in patients 22 are available. This allowed us, as the half-life of 11 C is short, to determine the transformation of cold 5-HTP into 5-HT and 5-HIAA at 2 different time points (15 and 60 min) at equal concentrations as used for radioactive labeled [ 11 C]HTP. The use of selective AADC and MAO inhibitors gave insight into intracellular processes. Presence of AADC in BON cells is confirmed by the intracellular synthesis of 5-HT. No increase in 5-HTP is noticed after carbidopa treatment most likely as carbidopa is not a substrate for LAT transporters in BON cells and therefore not being transported into these cells. NET cells express upregulated AADC activity 36. Inhibition of AADC will therefore not be inhibited by carbidopa within the tumor cell and intracellular conversion of 5-HTP into 5-HT will still be possible. Use of the MAO A inhibitor clorgyline and non-selective MAO inhibitor pargyline resulted in higher intracellular tumor 5-HT and lower 5-HIAA level. Once 5-HT is formed inside the cell, it appears not to be transported outside the cell before being deaminated by MAO into 5-HIAA. Conclusion Inhibition of LAT in BON cells leads to decreased tracer accumulation in vitro. Carbidopa does not influence tracer accumulation in tumor cells in vitro but improves tumor imaging in vivo. [ 18 F]FDOPA is superior to [ 11 C]HTP in tumor SUVs in this neuroendocrine pancreatic tumor xenograft model. MAO inhibition improves tracer accumulation in vitro. Acknowledgements BON cells were kindly provided by Dr. Ahlman from the Department of Surgery, Sahlgrenska University Hospital, Gothenburg, Sweden. We thank E.E.A. Venekamp and J. Krijnen for assistance in analyzing 5-HTP, 5-HT and 5-HIAA samples. This work was supported by grant of the Dutch Cancer Society. References 1. Rufini V, Calcagni ML, Baum RP. Imaging of neuroendocrine tumors. Semin Nucl Med 2006; 36: Eriksson B, Örlefors H, Öberg K, Sundin A, Bergström M, Långström B. Developments in PET for the detection of endocrine tumours. Best Pract Res Clin Endocrinol Metabol 2005; 19:

70 Manipulation of [ 11 C]HTP and [ 18 F]FDOPA accumulation in neuroendocrine tumor cells 3. Öberg K, Eriksson B. Nuclear medicine in the detection, staging and treatment of gastrointestinal carcinoid tumours. Best Pract Res Clin Endocrinol Metabol 2005; 19: Örlefors H, Sundin A, Garske U, Juhlin C, Öberg K, Långström B, Bergström M, Eriksson B. Whole-body 11 C-5-hydroxytryptophan positron emission tomography as a universal imaging technique for neuroendocrine tumors comparison with somatostatin receptor scintigraphy and computed tomography. J Clin Endocrinol Metab 2005; 90: Montravers F, Grahek D, Kerrou K, Ruszniewski P, de Beco V, Aide N, Gutman F, Grangé JD, Talbot JN. Can fluorodihydroxyphenylalanine PET replace somatostatin receptor scintigraphy in patients with digestive endocrine tumors? J Nucl Med 2006; 47: Koopmans KP, de Vries EGE, Kema IP, Elsinga PH, Neels OC, Sluiter WJ, van der Horst-Schrivers ANA, Jager PL. Staging of carcinoid tumours with 18 F-DOPA PET: a prospective, diagnostic study. Lancet Oncol 2006; 7: Sundin A, Garske U, Örlefors H. Nuclear imaging of neuroendocrine tumors. Best Pract. Res Clin Endocrinol Metab 2007; 21: Pearse AG. The APUD concept and hormone production. Clin Endocrinol Metab 1980; 9: Fuchs BC, Bode BP. Amino acid transporters ASCT2 and LAT1 in cancer: partners in crime? Sem Cancer Biol 2005; 15: Johnston JP. Some observations upon a new inhibitor of monoamine oxidase in brain tissue. Biochem Pharmacol 1968; 17: Knoll J, Magyar K. Some puzzling effects of monoamine oxidase inhibitors. Adv Biochem Psychopharmcol 1972; 5: Shih JC, Chen K, Ridd MJ. Monoamine oxidase: from genes to behavior. Annu Rev Neurosci 1999; 22: Saura J, Kettler R, Da Prada M, Richards JG. Quantitative enzyme radioautography with 3 H-Ro and 3 H-Ro in vitro: localization and abundance of MAO-A and MAO-B in rat CNS, peripheral organs, and human brain. J Neurosci 1992; 12: Saura J, Nadal E, van den Berg B, Vila M, Bombi JA, Mahy N. Localization of monoamine oxidase in human peripheral tissues. Life Sci. 1996; 59: De Vries EFJ, Luurtsema G, Brussermann M, Elsinga PH, Vaalburg W. Fully automated synthesis module for the high yield one-pot preparation of 6- [ 18 F]fluoro-L-DOPA. Appl Radiat Isot 1999; 51: Neels OC, Jager PL, Koopmans KP, Eriks E, de Vries EGE, Kema IP, Elsinga PH. Development of a reliable remote-controlled synthesis of β-[ 11 C]-5-hydroxy-Ltryptophan on a Zymark robotic system. J Label Compd. Radiopharm 2006; 49: Shotwell MA, Jayme DW, Kilberg MS, Oxender DL. Neutral amino acid transport systems in chinese hamster ovary cells. J Biol Chem 1981; 256: Jager PL, de Vries EGE, Piers DA, Timmer-Bosscha H. Uptake mechanisms of L- 3-[ 125 I]iodo-alpha-methyl-tyrosine in a human small-cell lung cancer cell line: comparison with L-1-[ 14 C]tyrosine. Nucl Med Comm 2001; 22:

71 Chapter Evers BM, Townsend CM, Upp JR, Allen E, Hurlbut SC, Kim SW, Rajamaran S, Singh P, Reubi JC, Thompson JC. Establishment and characterization of a human carcinoid in nude mice and effect of various agents on tumor growth. Gastroenterology 1991; 101: Pino R, Lyles GA. Effects of activity of semicarbazide-sensitive aminooxidases and cellular glutathione on the cytotoxic effect of allylamine, acrolein, and formaldehyde in human cultured endothelial cells. Vopr Med Khim 1997; 43: Basma AN, Morris EJ, Nicklas WJ, Geller HM. L-DOPA cytotoxicity to PC12 cells in is via its autooxidation. J Neurochem 1995; Kema IP, Meijer WG, Meiborg G, Ooms B, Willemse PHB, de Vries EGE. Profiling of tryptophan-related plasma indoles in patients with carcinoid tumors by automated, on-line, solid-phase extraction and HPLC with fluorescence detection. Clin Chem 2001; 47: Hagberg GE, Torstenson R, Marteinsdottir I, Fredrikson M, Långström B, Blomqvist G. Kinetic compartment modeling of [ 11 C]-5-hydroxy-L-tryptophan for positron emission tomography assessment of serotonin synthesis in human brain. J Cereb Blood Flow Metab 2002; Parekh D, Ishizuka J, Townsend CM, Haber B, Beauchamp RD, Karp G, Kim SW, Rajamaran S, Greeley G, Thompson JC. Characterization of a human pancreatic carcinoid in vitro: morphology, amine and peptide storage, and secretion. Pancreas 1994; 9: Lemmer K, Ahnert-Hilger G, Höpfner M, Hoegerle S, Faiss S, Grabowski P, Jockers-Scherübreceptors and transporter in neuroendocrine gastrointestinal tumor cells. Life Sci M, Riecken E, Zeitz M, Scherübl H. Expression of dopamine 2002; 71: Brunk B, Blex C, Rachakonda S, Höltje M, Winter S, Pahner I, Walther DJ, Ahnert-Hilger G. The first luminal domain of vesicular monoamine transporters mediates G-protein-dependent regulation of transmitter uptake. J Biol Chem 2006; 281: Höltje M, Winter S, Walther D, Pahner I, Hörtnagl H, Ottersen OP, Bader M, Ahnert-Hilger G. The vesicular monoamine content regulates VMAT2 activity through Gα q in mouse platelets. J Biol Chem 2003; 278: Feldman JM. Monoamine and dimonoamine oxidase activity in the diagnosis of carcinoid tumors. Cancer 1985, 56: Smith CC. Evidence for separate serotonin and catecholamine compartments in human platelets. Biochem Biophys Acta 1996; 1291: Eissele R, Anlauf M, Schärfer MKH, Eiden LE, Arnold R, Weihe E. Expression of vesicular monoamine transporters in endocrine hyperplasia and endocrine tumors of the oxyntic stomach. Digestion 1999; 60: Örlefors H, Sundin A, Lu L, Öberg K, Långström B, Eriksson B, Bergström M. Carbidopa pretreatment improves image interpretation and visualisation of carcinoid tumours with 11 C-5-hydroxytryptophan positron emission tomography. Eur J Nucl Med Mol Imaging 2006; 33:

72 Manipulation of [ 11 C]HTP and [ 18 F]FDOPA accumulation in neuroendocrine tumor cells 32. Hoffman JM, Melega WP, Hawk TC, Grafton SC, Luxen A, Mahoney DK, Barrio JR, Huang SC, Mazziotta. JC, Phelps ME. The effects of carbidopa administration on 6-[ 18 F]fluoro-L-DOPA kinetics in positron emission tomography. J Nucl Med 1992; 32: Eriksson B, Bergström M, Örlefors H, Sundin A, Öberg K, Långström B. Use of PET in neuroendocrine tumors. In vivo applications and in vitro studies. Q J Nucl Med 2000; 44: Bergström M, Lu L, Eriksson B, Marques M, Bjurling P, Andersson Y, Långström B. Modulation of organ uptake of 11 C-labelled 5-hydroxytryptohan. Biogenic Amines 1996; 12: Nawashiro H, Otani N, Shinomiya N, Fukui S, Ooigawa H, Shima K, Matsuo H, Kanai Y, Endou H. L-type amino acid transporter 1 as a potential molecular target in human astrocytic tumors. Int J Cancer 2006; 119: Vachtenheim J, Novotná H. Expression of the aromatic L-amino acid decarboxylase mrna in human tumour cell lines of neuroendocrine and neuroectodermal origin. Eur J Cancer 1997; 33: Xing T, Wu F, Brodin O, Fasth KJ, Långström B, Bergström M. In vitro PET evaluations in lung cancer cell Lines. Anticancer Res 2000; 20:

73 58

74 Chapter 4 Staging of carcinoid tumours with 18 F-DOPA PET: a prospective, diagnostic accuracy study Klaas P. Koopmans 1, Elisabeth G.E. de Vries 2, Ido P. Kema 3, Philip H. Elsinga 1, Oliver C. Neels 1, Wim J. Sluiter 4, Anouk N.A. van der Horst-Schrivers 3, Pieter L. Jager 1. Departments of Nuclear Medicine and Molecular Imaging 1, Medical Oncology 2, Pathology and Laboratory Medicine 3 and Endocrinology 4, University of Groningen and University Medical Center Groningen, The Netherlands The Lancet Oncol 2006; 7:

75 Chapter 4 Summary Background To assess individual treatment options for patients with carcinoid tumours, accurate knowledge of tumour localisation is essential. We aimed to test the diagnostic sensitivity of 6-[fluoride-18]fluoro-levodopa ( 18 F-DOPA PET), compared with conventional imaging methods, in patients with carcinoid tumours. Methods In a prospective, single-centre, diagnostic accuracy study, 18 F-DOPA PET with carbidopa pretreatment was compared with somatostatin-receptor scintigraphy (SRS), CT, and combined SRS and CT in 53 patients with a metastatic carcinoid tumour. The performance of all imaging methods was analysed for individual patients, for eight body regions, and for the detection of individual lesions. PET and CT images were fused to improve localisation. To produce a composite reference standard, we used cytological and histological findings; all imaging tests, including secondary assessments for newly found lesions; follow-up; and biochemical data. Sensitivities were calculated and compared. Findings In patient-based analysis, we recorded sensitivities of 100% (95% CI ) for 18 F- DOPA-PET, 92% (82 98) for SRS, 87% (75 95) for CT, and 96% (87 100) for combined SRS and CT (p=0.45 for 18 F-DOPA PET vs combined SRS and CT). However, 18 F-DOPA PET detected more lesions, more positive regions, and more lesions per region than combined SRS and CT. In region-based analysis, sensitivity of 18 F-DOPA PET was 95% (90 98) versus 66% (57 74) for SRS, 57% (48 66) for CT, and 79% (70 86) for combined SRS and CT (p=0.0001, PET vs combined SRS and CT). In individual-lesion analysis, corresponding sensitivities were 96% (95 98), 46% (43 50), 54% (51 58), and 65% (62 69; p< for PET vs combined SRS and CT). Interpretation If the improved tumour localisation seen with 18 F-DOPA-PET compared with conventional imaging is confirmed in future studies, this imaging method could replace use of SRS, help improve prediction of prognosis, and be used to assess patients' response to treatment for carcinoid tumours. Introduction Neuroendocrine tumours are a heterogeneous group of slow-growing lesions arising from neuroendocrine cells, of which carcinoid tumours are the most common. These tumours are often located in the abdomen and can produce and secrete a large variety of products because of their intrinsic ability to take up, accumulate, and decarboxylate amine precursors 1. In metastatic disease, these products, such as serotonin and catecholamines, can bypass the first-pass metabolisation and inactivation by the liver and can cause symptoms. Treatment options for carcinoid tumours include curative or debulking surgery, medical treatment with somatostatin analogues, and interferon 2. To assess individual treatment options, accurate knowledge of tumour localisation, biochemical activity, and progression is essential. The initial work-up for patients with 60

76 Staging of carcinoid tumours with 18 F-DOPA PET: a prospective, diagnostic accuracy study carcinoid tumours consists of morphological imaging methods such as CT, combined with functional whole-body imaging with somatostatin-receptor scintigraphy (SRS) 3,4,5. However, CT and MRI of the abdomen have difficulties in correctly separating tumours and mesenterial metastases from intestinal structures 6,7,8. Furthermore, SRS can produce false-negative findings, because of the variable affinity and expression of somatostatin receptors and the restricted resolution of gamma cameras and single-photon-emission tomography (SPECT) methods 9,10. PET using the catecholamine precursor 6-[ fluoride-18]fluoro-levodopa ( 18 F-DOPA) has emerged as a new imaging method for neuroendocrine tumours 6. By contrast with other methods, this procedure is based on the intrinsic property of neuroendocrine tumours to 11,12,13,14,15 take up amine precursors, such as 18 F-DOPA. The combination of this specific tracer with the high resolution provided by PET could lead to a clinically relevant improvement in the detection and staging of neuroendocrine tumours. A few small 6,16,17 18 studies have shown some potential of F-DOPA PET in small and heterogeneous groups of patients with neuroendocrine tumours. T herefore, the aim of this study was to compare the diagnostic sensitivity of 18 F-DOPA PET with that of conventional imaging methods such as SRS and CT, in a large and homogeneous population of patients with carcinoid tumours. Methods Patients Eligible patients for this prospective single-centre diagnostic accuracy study included: those who were newly referred to our centre (which serves the northern region of the Netherlands) with a carcinoid tumour, based on clinical or biochemical findings, and at least one abnormal lesion detected on CT, MRI, sonography, or SRS; and those known to have a histopathologically proven carcinoid tumour, who had a clinical indication for restaging, and who had at least one abnormal lesion on conventional imaging studies. We excluded patients younger than 18 years, those who were pregnant, and those in whom an additional non-carcinoid tumour had been diagnosed. Every consecutive patient underwent 18 F-DOPA PET, SRS, and CT scanning of the abdomen and (if needed) of the chest, and biochemical analysis. Imaging methods were undertaken in a random order. The local medical ethics committee approved the study and all patients gave written informed consent. Procedures 18 F-DOPA was produced in the radiochemical laboratory of our hospital as described previously 18. Patients fasted for 6 h before the examination and were allowed to continue all medication. Whole-body two-dimensional PET images were acquired 60 min after intravenous use of 18 F-DOPA ( MBq, radiation dose msv) 19, on a Siemens ECAT HR+ (high-resolution) positron camera (Siemens, Knoxville, TN, USA) with attenuation correction (7 10 bedpositions of 5 min emission and 3 min transmission scan). For the reduction of tracer decarboxylation and subsequent renal clearance, all patients received 2 mg/kg carbidopa orally as pre-treatment, 1 h before the 18 F-DOPA injection, to increase tracer uptake in tumour cells 19,20,21. 61

77 Chapter 4 Two nuclear medicine physicians (KPK, PLJ), who were masked to the results of other imaging examinations and to the extension of tumour spread in study patients, interpreted the 18 F-DOPA PET images independently. Only lesions in every body region that clearly showed more activity than that seen in patients and regions not known to contain tumours were regarded as abnormal. If discrepancies were found, a consensus reading was done. Since 18 F-DOPA PET is a new test, these physicians built expertise in the first 20 cases, and then reviewed these early cases again in the second half of According to Dutch standards, we obtained planar total-body and SPECT images 24 h after intravenous administration of 200 MBq indium-111-octreotide (Octreoscan; Mallinckrodt, Petten, Netherlands; radiation dose 10 msv) 22, using standard methods (Siemens Multispect 2 gamma camera, medium-energy collimator, 10 min spotviews, 64 projections of 30 s). If interfering bowel activity was seen, images were recorded again at 48 h 23. We withheld laxatives only if patients presented with diarrhoea. All patients were allowed to continue their treatment. SRS scans were interpreted by dedicated specialists as part of routine care and independently reread by a nuclear medicine physician (PLJ), who was masked to the results of other imaging examinations and to the extension of tumour spread in the study patients. CT (4 16 slice, Siemens Somatom Sensation, Siemens Medical Systems, Erlangen, Germany; radiation dose 8 20 msv) 24 was done with oral contrast and intravenous contrast enhancement (Visipaque 270, 120 ml, 2.5 ml/s). The reconstruction interval was 3 8 mm. All patients underwent CT of the entire abdomen and pelvis. The CT imaging area was extended to include the chest in 26 patients, and the neck and chest in three patients because of clinical suspicion of tumours in those regions. CT scans were interpreted by dedicated specialists as part of routine care. At the time of image fusion, results were reviewed again by the investigators, and for discrepancies, consensus was reached after multidisciplinary discussion. As a composite reference standard for the presence of tumour lesions, we used all available cytological, histological, follow-up, and imaging findings, because cytological or histological verification of every lesion is not feasible and not justifiable ethically in all patients because of the tumour load in many of these patients. If possible, new findings on PET were verified by other investigations other than CT and PET CT fusion. These were: MRI (n=8), bone scintigraphy (n=9), planar radiographs (n=13), sonography (n=4), surgery (n=10), or biopsy (n=5). These investigations included verification of lesions in body regions that were outside the CT field. However, in many cases, the number of new and previously unknown lesions on PET imaging was high, which led to the analysis of every individual localisation. After images had been interpreted, CT and 18 F-DOPA PET images were fused automatically by use of three-dimensional fusion software (Siemens Leonardo workstation) with manual fine adjustments. Experienced physicians compared the fusion images with the results of visual matching for the accuracy of lesion localisation. As markers for serotonin metabolism, we measured serotonin concentrations in platelets and urinary 5-hydroxyindole acetic acid (5-HIAA) in a 24-h urine sample (upper reference limits 5.4 nmol/10 9 platelets and 3.8 mmol/mol creatinine, respectively). As markers of catecholamine metabolism, we measured urinary concentrations of metanephrine, normetanephrine, and 3-methoxytyramine in a 24-h urine collection (upper reference limits 62

78 Staging of carcinoid tumours with 18 F-DOPA PET: a prospective, diagnostic accuracy study 99, 260, and 197 μmol/mol creatinine, respectively) 14,25. Sampling procedures and analytical methods were done as previously described 24,25,26,27,28,29,30. We measured serum concentrations of chromogranin A by use of a radioimmunoassay (Cga-React, Cis Bio International, Marcoule, France) as a marker for tumour volume (reference interval mg/l). Statistical analysis Analysis was done at three levels. At the first level, individual patients were analysed. Image studies were regarded as positive if a patient had at least one lesion. The second level of analysis addressed body regions head and neck, mediastinum, lungs, liver, abdomen and pelvis, bone, and soft tissue of the extremities. A region was regarded as positive, if at least one lesion was detected in that region. The third level analysed the individual lesions that were counted for all imaging methods. If the number of lesions in one region (e.g., liver) was more than ten, the number of lesions was truncated at ten lesions for that region to avoid bias. SRS is a whole-body procedure, whereas CT covers only the most relevant parts of the body. To eliminate possible bias towards whole-body imaging methods, we only analysed regions for which all three imaging methods were available. Sensitivities were calculated with the composite reference standard and were compared with paired observations and McNemar's test. Patient-based sensitivity was calculated as the proportion of patients with at least one lesion detected. Regional sensitivity was calculated by dividing the number of patients with a positive region (detected with that particular method) by the total number of patients in whom that region was positive by any imaging method. We calculated lesion-based sensitivity by dividing the number of lesions detected with a particular method by the total number of lesions detected by any method. Pitman's test for paired data was used to compare the number of lesions per region. Wilcoxon's test was used to compare the number of patients with five or fewer positive body regions detected by PET and by combined SRS and CT. For correlations, Spearman's r test was calculated. Significance level was 0.05, two-sided. We did statistical analysis by using the SPSS package version Results Between October, 2003, and February, 2006, we asked 68 consecutive patients to participate in the study (figure 1); however, three declined PET scanning, and we could not obtain all required information for 12, because of various logistical reasons (e.g., no biochemistry or pathology findings, no SRS). Sensitivity was calculated in the remaining 53 patients assessed, of whom 25 were newly diagnosed with carcinoid disease (table 1). The median time between PET and CT was 59 days (range 1 191) and between PET and SRS was 47 days (1 206). Mean values for these intervals were 25 days (SD 57) and 42 days (75), respectively. In retrospect, the interval was short compared with disease progression in all patients. One patient developed a carcinoid crisis after intravenous administration of 18 F-DOPA, which was treated successfully

79 Chapter 4 68 patients included 3 patients declined PET scan 12 patients has incomplete informations 53 patients assessed Figure 1. Study flowchart of patient Table 1. Patients' characteristics (n=53) Value Sex (male/female) 25/28 Age (years) 59 (35-77) Newly diagnosed patients vs. known disease 25/28 Histological vs. biochemical diagnosis 52/1 Primary localisation Lung 5 Duodenum 3 Jejunum 3 Ileum 25 Colon 1 Unknown 16 Carcinoid syndrome 21 Treatment during scan Somatostatin analogues only 15 Somatostatin analogues and interferon 1 Biochemical variables Platelet serotonin >5.4 nmol/10 9 platelets 42/51 Urinary 5-HIAA >3.8 mmol/mol creatinine 35/51 Urinary metanephrine >9.9 µmol/mol creatinine 5/48 Urinary normetanephrine >260 µmol/mol creatinine 7/48 Urinary 3-methoxytyramine >197 µmol/mol creatinine 16/48 Serum chromogranin A >100 mg/l 21/31 Data are number of patients or median (range). 64

80 Staging of carcinoid tumours with 18 F-DOPA PET: a prospective, diagnostic accuracy study 18 F-DOPA PET produced high-quality tomographical images that were easily interpretable (figure 2). More patients had positive lesions detected by 18 F-DOPA PET than by SRS or by combined SRS and CT (table 2; 18 F-DOPA PET vs combined SRS and CT, p=0.45). Four patients were recorded as negative on SRS, seven on CT, and two on combined SRS and CT (both of whom were shown to have tumours when assessed 6 months later with SRS). Figure 2. Full-color in appendix. Imaging of a patient with carcinoid disease and metastases in the bone, mediastinum, liver, and abdomen. (A) 18 F-DOPA PET imaging. Red arrows indicate areas with physiological 18 F-DOPA uptake (striatum, kidneys, ureter, bladder), whereas all other black spots are tumour lesions. (B) Planar SRS imaging. Arrows indicate mediastinal tumour lesions. (C) CT PET fusion imaging. Coloured areas indicate tumour lesions. In this patient, both planar and SPECT SRS missed most lesions found with 18 F-DOPA PET imaging. Abdominal and femoral lesions were not recorded on CT. Table 2. Patient-based analysis Positive lesions (n) Sensitivity (95% CI) Lesions detected Per patient (median [range]) 18 F-DOPA PET % (93-100) 12 (1-36) SRS only 49 92% (82-98) 4 (0-20) CT only 46 87% (75-95) 6 (0-20) SRS and CT 51 96% (87-100) 10 (1-36) 65

81 Chapter 4 Table 3 shows region-based and lesion-based sensitivities. Of 326 regions that were assessable, 122 (37%) were judged as positive for tumour. 18 F-DOPA PET detected 117 of these positive regions (sensitivity 95%), whereas SRS only detected 80 (sensitivity 66%; table 3). When data from SRS and CT were combined, sensitivity reached 79% but was substantially lower than that for 18 F-DOPA PET ( 18 F-DOPA PET vs combined SRS and CT, p=0.0001). Table 3. Sensitivity of imaging methods in patients with carcinoid tumours 18 F-DOPA PET SRS (95% CI) CT (95% CI) SRS and CT (95% CI) (95% CI) Positive regions or lesions (n) Region-based analysis Mediastinal * 92% (74-99) 69% (38-91) 15% (1-46) 69% (38-91) 13 Lung * 60% (13-96) 80% (27-99) 40% (40-87) 80% (27-99) 5 Liver 98% (88-100) 77% (62-89) 75% (60-87) 89% (75-96) 44 Pancreas 100% (26-100) 0% (0-74) 67% (7-100) 67 % (7-100) 3 Abdomen or pelvis 97% (86-100) 72% (55-85) 62% (44-77) 87 % (72-96) 39 B one 92% (63-100) 31% (1-62) 46% (19-75) 54% ( 25-81) 13 Extremities 100% (45-100) 20% (0-73) 20% (0-73) 20 % (0-73) 5 Total 95% (90-98) 66% (57-74) 57% (48-66) 79% (70-86) 122 Lesion-based analysis Mediastinal * 95% (82-100) 41% ( 25-58) 8% (1-21) 44 % (28-60) 39 * L ung 55% (23-84) 53% (16-77) 36% (10-70) 64% (30-90) 11 Liver 97% (95-94) 53% (48-59) 66% (61-71) 73% ( 68-78) 360 Pancreas 100% (26-100) 0% (0-74) 67% (7-100) 67% (7-100) 3 Abdomen or pelvis 96% (93-98) 41% (35-48) 49% (42-56) 64% (57-71) 208 B one 97% (88-100) 31% (20-44) 41% (29-54) 48% (35-61) 61 Extremities 100% (45-100) 20% (0-73) 20% (0-73) 20 % (0-73) 5 Total 96% (95-98) 46% (43-50) 54% (51-58) 65% (62-69) 687 No lesions were found in the head and neck region. * Only regions in field of view of all imaging procedures compared. 18 F-D OPA PET vs combined SRS and CT, p= F-D OPA PET vs combined SRS and CT, p< lesions were regarded as positive for tumour (table 3). 18 F-DOPA PET detected 658 lesions (sensitivity 96%), and SRS detected 315 (sensitivity 46%). Combined SRS and CT detected 450 lesions (sensitivity 65%; 18 F-DOPA PET vs combined SRS and CT, p<0.0001). Most positive lesions were found in the liver and abdomino-pelvic regions. 18 F- DOPA PET showed more lesions in these two regions than did combined SRS and CT (liver, 348 vs 261, p<0.0001; abdomen/pelvis, 203 vs 135, p<0.0001, respectively). 18 F-D OPA PET detected a mean of 2.2 (SD 0.93) positive regions per patient, versus 1.8 ( 0. 81) detected by combined SRS and CT (p=0.0007). Based on our reference standard and follow-up, we could not record any false-positive lesions. The median number of lesions per patient was 12 for 18 F-DOPA PET and ten for combined SRS and CT (table 2). A mean of 13 5 lesions (SD 7.9) per patient were fo und overall ( 18 F-DOPA PET, 12.4 [7. 4]; SRS, 6.2 [5. 6]; combined SRS and CT, 8.8 [6.4]; 18 F-DOPA PET vs combined SRS and CT, 18 p<0.0001). Thus, F-DOPA PET detects an additional tumour-positive region in one of three patients, and detects four additional lesions per patient. 66

82 Staging of carcinoid tumours with 18 F-DOPA PET: a prospective, diagnostic accuracy study Urinary 5-HIAA excretion correlated with the number of tumour lesions detected by 18 F- DOPA PET (r=0.41, p=0.003), by CT (r=0.40, p=0.003), and by combined SRS and CT (r=0.38, p=0.006). Platelet serotonin concentrations correlated only with the number of tumour lesions detected by 18 F-DOPA PET (r=0.45, p=0.001). We recorded no correlation between the total number of tumours detected by any imaging method and concentrations of serum chromogranin A, urinary metanephrine, normetanephrine, or 3-methoxytyramine. Discussion We showed improved diagnostic sensitivity of 18 F-DOPA PET in staging and identification of carcinoid tumours, compared with currently applied, standard whole-body imaging with SRS. Compared with the combination of SRS with CT, 18 F-DOPA PET detected substantially more individual tumour lesions, more affected body regions, and more lesions per region. The improved lesion detection of carcinoid tumours with 18 F-DOPA PET provides a better understanding of the true extent of tumour spread in patients. The precise mechanisms that determine the uptake of 18 F-DOPA in neuroendocrine tissues are not yet fully elucidated. The increased demand for amino acids, precursors in the overactive secretory pathways in neuroendocrine tumours 14, probably induces high uptake of this amino acid tracer in tumours by upregulation of transmembrane amino acid transporters. However, intracellular mechanisms, such as the highly active amino acid decarboxylase enzyme that is specifically active in neuroendocrine tumours, probably contribute to tracer uptake 11,13. Overactivity of the catecholamine pathway could induce uptake of the catecholamine precursor tracer 18 F-DOPA, but 18 F-DOPA uptake was also present in the absence of increased urinary catecholamine metabolite secretion. Only two small studies 6,17 have been reported on 18 F-DOPA PET scanning in patients with carcinoid tumours. Hoegerle and colleagues 6 did a lesion-based analysis (n=17) and found that 18 F-DOPA PET was more sensitive than SRS, CT, and MRI in detecting primary tumours and lymph-node metastases 6. However, the performance of 18 F-DOPA PET for the detection of organ metastases was similar to that of SRS and worse than that of CT and MRI combined. Our improved results might be due to the use of oral carbidopa pretreatment, which increases the concentration and availability of 18 F-DOPA, thereby improving lesion detectability 32. Hoegerle and co-workers 6 also used either CT or MRI, and studied fewer patients than we did. Becherer and colleagues 17 studied 23 patients, in whom 18 carcinoid tumours had been detected. 18 F-DOPA PET yielded high sensitivities in a r egion-based analysis similar to that used in our study. However, th ey detected fewer lesions in the lung (one of five patients who had lung tumours) than that seen in our study (three of five patients with lung tumours), although the numbers for this region were low in both studies. In the Becherer study 17, CT was the gold standard, and only lesions visible on 18 F-DOPA PET were regarded as false-positive. A notable alternative for the imaging of neuroendocrine tumours is the use of a direct precursor for the serotonin pathway, 11 C-5-hydroxytryptophan ( 11 C-5-HTP). This tracer has been inves tigated in 42 patients with various neuroendocrine tumours, after pretreatment with carbidopa C-5-HTP PET detected carcinoid tumours in 13 of these patients, which was similar to our results with 18 F-DOPA. General applicability was limited by the difficult tracer synthesis of 11 C-5-HTP and the short half-life of an 11 C-based tracer of 20 min (halflife of an 18 F-based tracer is 110 min). Other developments include the search for new 67

83 Chapter 4 radiolabeled somatostatin analogues for improved SRS SPECT imaging and SRS PET imaging 34,35,36. A perfect gold standard is difficult to establish in any diagnostic accuracy study. In our study, new diagnostic methods might have been much better than current standard methods and might detect many unknown lesions that can never all be verified by cytological or histological analysis. Where possible, new findings were verified but we also assumed that when several lesions were verified by one technique, other lesions with identical and unequivocal uptake of this same tracer in the same patient could also be regarded as true tumours. The compo site reference standard depended to some extent on the 18 F -DOPA PET results, and also on the CT and SRS results. Thus, our sensitivity valu es should be interpreted with caution. In view of the many new lesions detected by 18 F-DOPA PET and the fact that cytological and histological verification has a risk of bleeding complications in these highly vascularised lesions, we did not consider the undertaking of ten biopsies in one patient as feasible. Enhanced detection of lesions with 18 F-DOPA PET can lead to improvemen ts in patients' c are. In particular, the imaging method's excellent detection properties for liver, bone, and a bdominal lesions could lead t o alterations in surge ry, medical treatment, and radiotherapy p lans. Neuroendocrine t umours are currently classified by the WHO framewor k, which is based on morphological, clinical, and functional aspects of the tumour and i ts metastases. Treatment options depend on the tumour mass, functional ac tivity, and grow th behaviour of these tumours 37, F-DOPA PET could add important information on tumour localisations and progno sis and be used to aid research in the response to n ew molecular-targeted drugs, possibly even replacing SRS. Howev er, we did not aim to rec ord how use of this technique c ould change management, since we included patients with at least one histologically c onfirmed lesion or with known extensive disease. Furthermore, image interpreters need to devel op experience with the technique before being completely accurate. With the excellent sensitivity of 18 F-DOPA PET recorded in patients with proven tumours, future studies are under way to measure the technique's detection capability in patients suspected of having a neuroendocrine tumour. 18 In summary, F-DOPA PET significantly improves the detection of carcinoid tumours and their metastases compared with conventional techniques, and detects an additional tumourpositive region in one of three patients, and a mean of four additional lesions per patient. This technique might also contribute greatly to the staging of these patients. With the rapidly expanding availability of PET and increasing commercial production and distribution of radiotracers, the availability of tracers such as 18 F-DOPA will probably also increase. Furthermore, treatment of these patients is often centralised in well-equipped hospitals. Although 18 F-DOPA PET is almost sufficient for staging, addition of CT improves the localisation of lesions, which is relevant to guide surgical and radiotherapeutical procedures. The new trend of combined PET CT scanning could therefore become a one-stop procedure in the staging of carcinoid tumours. Acknowledgments Our work was supported by grant from the Dutch Cancer Society. We thank WW de Herder for his valuable advice. 68

84 Staging of carcinoid tumours with 18 F-DOPA PET: a prospective, diagnostic accuracy study References 1. Pearse AG. The APUD cell concept and its implications in pathology. Pathol Annu 1974; 9: Moertel CG, Lefkopoulo M, Lipsitz S et al. Streptozocin-doxorubicin, streptozocin-fluorouracil or chlorozotocin in the treatment of advanced islet-cell carcinoma. N Engl J Med 1992, 326: Öberg K, Kvols L, Caplin M et al. Consensus report on the use of somatostatin analogs for the management of neuroendocrine tumors of the gastroenteropancreatic system. Ann Oncol 2004; 15: Plockinger U, Rindi G, Arnold R et al. Guidelines for the diagnosis and treatment of neuroendocrine gastrointestinal tumours. A consensus statement on behalf of the European Neuroendocrine Tumour Society (ENETS). Neuroendocrinology 2004; 80: Modlin IM, Kidd M, Latich I et al. Current status of gastrointestinal carcinoids. Gastroenterology 2005; 128: Hoegerle S, Altehoefer C, Ghanem N et al. Whole-body 18 F dopa PET for detection of gastrointestinal carcinoid tumors. Radiology 2001; 220: Kaltsas G, Rockall A, Papadogias D et al. Recent advances in radiological and radionuclide imaging and therapy of neuroendocrine tumours. Eur J Endocrinol 2004; 151: Kumbasar B, Kamel IR, Tekes A et al. Imaging of neuroendocrine tumors: accuracy of helical CT versus SRS. Abdo m Imaging 2004; 29: Fahey FH, Harkness BA, Keyes Jr JW et al. Sensitivity, resolution and image quality with a multi-head SPECT camera. J Nucl Med 1992; 33: De Herder WW, Hofland LJ, van der Lely AJ, Lamberts SW. Somatostatin receptors in gastroentero-pancreatic neuroendocrine tumours. Endocr Relat Cancer 2003; 10: Meijer WG, Copray, SC Hollema H et al. Catecholamine-synthesizing enzymes in carcinoid tumors and pheochromocytomas. Clin Chem 2003; 49: Pearse AG. The APUD cell concept and its implications in pathology. Pathol Annu 1974; 9: Gilbert JA, Bates LA, Ames MM. Elevated aromatic-l-amino acid decarboxylase in human carcinoid tumors. Biochem Pharmacol 1995; 50: Kema IP, de Vries EGE, Slooff MJ et al. Serotonin, catecholamines, histamine, and their metabolites in urine, platelets, and tumor tissue of patients with carcinoid tumors. Clin Chem 1994; 40: Feldman JM, Moore JO. Biogenic amines in carcinoid tumors. Biog Amines 1989; 6: Ahlstrom H, Eriksson B, Bergström M et al. Pancreatic neuroendocrine tumors: diagnosis with PET. Radiology 1995; 195: Becherer A, Szabo M, Karanikas G et al. Imaging of advanced neuroendocrine tumors with 18 F-FDOPA PET. J Nucl Med 2004; 45:

85 Chapter De Vries EFJ, Luurtsema G, Brussermann M et al. Fully automated synthesis 18 module for the high yield one-pot preparation of 6-[ F]-fluoro-L-DOPA. Appl Radiat Isot 1999; 51: Brown WD, Oakes TR, DeJesus OT et al. Fluorine-18-fluoro-L-DOPA dosimetry with carbidopa pretreatment. J Nucl Med 1998; 39: Bergström M, Lu L, Eriksson B et al. Modulation of organ uptake of 11 C-labelled 5-hydroxytryptophan. Biog Amines 1996; 12: Ishikawa T, Dhawan V, Chaly T et al. Fluorodopa positron emission tomography with an inhibitor of catechol-o-methyltransferase: effect of the plasma 3-Omethyldopa fraction on data analysis. J Cereb Blood Flow Metab 1996; 16: International Commission on Radiological Protection, ICRP publication 80: radiation dose to patients from radiopharmaceuticals. Annals of the ICRP Volume 28/3, Elsevier, Pergamom, Oxford, Balon HR, Goldsmith SJ, Siegel BA et al. Procedure guideline for somatostatin receptor scintigraphy with 111 In-pentetreotide. J Nucl Med 2001; 42: International Commission on Radiological Protection, ICRP Publication 87: managing patient dose in computed tomography. Annals of the ICRP Volume 30/4, ICRP Online, Elsevier, Pergamom, Oxford, Eisenhofer G, Kopin IJ, Goldstein DS. Catecholamine metabolism: a contemporary view with implications for physiology and medicine. Pharmacol Rev 2004; 56: Kema IP, Schellings AM, Hoppenbrouwers CJ et al. High performance liquid chromatographic profiling of tryptophan and related indoles in body fluids and tissues of carcinoid patients. Clin Chim Acta 1993; 221: Willemsen JJ, Ross HA, Wolthers BG et al. Evaluation of specific highperformance liquid-chromatographic determinations of urinary metanephrine and normetanephrine by comparison with isotope dilution mass spectrometry. Ann Clin Biochem 2001; 38: Kema IP, Meiborg G, Nagel GT et al. Isotope dilution ammonia chemical ionization mass fragmentographic analusis of urinary 3O-methylated catecholamine metabolites. Rapid sample clean-up by derivatization and extraction of lyophilic samples. J Chromatogr Biomed Appl 1993; 671: Kema IP, Meijer WG, Meiborg G et al. Profiling of tryptophan-related plasma indoles in patients with carcinoid tumors by automated, on-line, solid-phase extraction and HPLC with fluorescence detection. Clin Chem 2001; 47: Mulder EJ, Oosterloo-Duinkerken A, Anderson GM et al. Automated on-line solid-phase extraction coupled with HPLC for measurement of 5-hydroxyindole-3- acetic acid in urine. Clin Chem 2005; 51: Koopmans KP, Brouwers AH, De Hooge MN et al. Carcinoid crisis after injection 18 of 6 F-fluorodihydroxyphenylalanine in a patient with metastatic carcinoid. J Nucl Med 2005; 46:

86 Staging of carcinoid tumours with 18 F-DOPA PET: a prospective, diagnostic accuracy study 32. Örlefors H, Sundin A, Lu L et al. Carbidopa pretreatment improves image interpretation and visualisation of carcinoid tumours with 11 C-5- hydroxytryptophan positron emission tomography. Eur J Nucl Med Mol Imaging 2006, 33: Örlefors H, Sundin A, Garske U et al. Whole-body C-5-hydroxytryptophan positron emission tomography as a universal imaging technique for neuroendocrine tumors: comparison with somatostatin receptor scintigraphy and computed tomography. J Clin Endocrinol Metab 2005; 90: Hubalewska-Dydejczyk A, Fross-Baron K, Mikolajczak R et al. 99m Tc- EDDA/HYNIC-octreotate scintigraphy, an efficient method for the detection and staging of carcinoid tumours: results of 3 years' experience. Eur J Nucl Med Mol Imaging 2006; 33: Hofmann M, Maecke H, Borner R et al. Biokinetics and imaging with the somatostatin receptor PET radioligand 68 Ga-DOTATOC: preliminary data. Eur J Nucl Med 2001; 28: Seemann MD, Meisetschlaeger G, Gaa J, Rummeny EJ. Assessment of the extent of metastases of gastrointestinal carcinoid tumors using whole-body PET, CT, MRI, PET/CT and PET/MRI. Eur J Med Res 2006; 11: In: Solcia E, Kloppel G, Sobin LH. Editors. Histological typing of endocrine tumours (2nd edn.), WHO, Heidelberg, Modlin IM, Kidd M, Latich I et al. Current status of gastrointestinal carcinoids, Gastroenterology 2005; 128:

87 72

88 Chapter 5 Improved staging and characterization of lesions in patients with carcinoid and islet cell tumors with 18 F-DOPA and 11 C-5-HTP positron emission tomography Klaas P. Koopmans 1, Oliver C. Neels 1, Ido P. Kema 3, Philip H. Elsinga 1, Wim J. Sluiter 4, Koen Vanghillewe 5, Adrienne H. Brouwers 1, Elisabeth G.E. de Vries 2, Pieter L. Jager 1. Departments of Nuclear Medicine and Molecular Imaging 1, Medical Oncology, Pathology and Laboratory Medicine 3, Endocrinology 4, University of Groningen and University Medical Center Groningen, The Netherlands. Department of Radiology 5, Martini Hospital Groningen, The Netherlands. In press for J Clin Oncol 2 73

89 Chapter 5 Summary Background To evaluate and compare diagnostic sensitivity of PET scanning in carcinoid and islet cell tumor patients with a serotonin and a catecholamine precursor as tracers. Methods Carcinoid (n=24) or pancreatic islet cell tumor (n=23) patients with at least one lesion on conventional imaging including somatostatin receptor scintigraphy (SRS) and CT scan underwent 11 C-5-hydroxytryptophan ( 11 C-5-HTP) PET and 6-[F-18]fluoro-Ldihydroxyphenylalanin ( 18 F-DOPA) PET. PET findings were compared with a composite reference standard derived from all available imaging, clinical and cytological/histological information. Findings In carcinoid tumor patients per patient analysis showed sensitivities for 11 C-5-HTP PET, 18 F-DOPA PET, SRS and CT of 100, 96, 86, 96 % respectively and in islet cell tumors of 100, 89, 78, 87%. In carcinoid patients per-lesion analysis revealed sensitivities for 11 C-5- HTP PET, 11 C-5-HTP PET/CT, 18 F-DOPA PET, 18 F-DOPA PET/CT, SRS, SRS/CT and CT alone of respectively 78, 89, 87, 98, 49, 73 and 63% and in islet cell tumors of 67, 96, 41, 80, 46, 77 and 68%. In all carcinoid patients 18 F-DOPA PET and 11 C-5-HTP PET detected more lesions than SRS (p <0.001). 11 C-5-HTP PET was superior to 18 F-DOPA PET in islet cell tumors (p <0.0001). In all cases CT improved the sensitivity of the nuclear scans. Interpretation 18 F-DOPA PET/CT is the optimal imaging modality for staging in carcinoid patients and 11 C-5-HTP PET/CT in islet cell tumor patients. Introduction Carcinoid tumors and pancreatic islet cell tumors are relatively indolent tumors. They belong to the group of neuroendocrine tumors that arise from neuroendocrine cells. These tumors can produce and secrete a large variety of products because of their intrinsic ability to take up, accumulate and decarboxylate amine precursors 1. Treatment options for these tumors include curative or debulking surgery, systemic treatment with somatostatin analogues, interferon and chemotherapy 2. To assess individual treatment options, accurate knowledge of tumor localization, biochemical activity and rate of progression is essential. The initial work-up for patients with carcinoid and islet cell tumors consists of morphological imaging methods such as CT, combined with functional whole body imaging using somatostatin receptor scintigraphy (SRS) 3,4. However, on CT and MR imaging of the abdomen it can be difficult to correctly distinguish tumors and mesenterial metastases from intestinal structures. In addition, CT and MR lesions cannot always be perfectly characterized as being malignant, especially in the pancreas, as frequently benign lesions or cysts may have rather similar or a mixed appearance 5,6,7. 74

90 Improved staging and characterization of lesions in patients with carcinoid and islet cell tumors with 18 F-DOPA and 11 C-5-HTP positron emission tomography Apart from the advantage of coveri ng the whole body in a single investigation, functional imaging methods also allow charact erization of lesions on CT or MR. SRS is often used for this purpose. However, it may produce false negative findings, due to variable affinity and expression levels of somatostatin receptors or small size of lesions because of the limited reso lution of the gammacamera and single photon emission tomography (SPECT) methods 8,9. Recently, two positron emission tomography (PET) tracers have emerged as potential fu nctional imaging modalities in neuroendocrine tumors. In combination with the high resolution of PET this may lead to a clinically relevant impro vement in detection, characterization and staging of these tumors. The first new tracer method is 18 F-DOPA PET, employing the catecholamine precursor 6-[F-18]fluoro-L-dihydroxyphenylalanin ( 18 F- DOPA) 5,6,10 whose uptake is based on the property of neuroendocrine tumors to take up amine precursors 1,11. For the detection of carcinoid disease, its superiority over presently 5,6 used modalities has been shown, but this advantage is less clear for islet cell tumours. The second metabolic PET tracer 11 C-5-hydroxytryptophan ( 11 C-5-HTP) is a direct precursor for the serotonin pathway and therefore a potentially sensitive universal method for neuroendocrine tumor detection. However, availability and experience with 11 C-5-HTP is limited due to its complex production 12,14. Currently, there are no head-to-head studies av ailable in which 18 F-DOPA is compared to 11 C-5-HTP PET in their ability to detect neuroendocrine tumors. Therefore, the aim of this study was to evaluate the diagnostic sensitivity of 11 C-5-HTP in comparis on with 18 F-DOPA PET in a large population of patients with a carcinoid or islet cell tumor. Methods Patients Patients eligible for this prospective single-centre diagnostic accuracy study were: new patients referred to our centre with a carcinoid or pancreatic islet cell tumor, based on clinical, histological and/ or biochemical findings and at least one abnormal lesion detected on CT, MRI, sonography or SRS, and patients known to have a histopathologically proven neuroendocrine tumor, who had a clinical indication for (re)staging and who had at least one abnormal lesion on conventional imaging studies. We excluded patients under 18 years of age, pregnant patients and those with an additional non-neuroendocrine tumor. Each consecutive patient underwent 11 C-5-HTP PET, 18 F-DOPA PET, SRS, CT of the abdomen and also the chest, when indicated, within a short interval, and in random order. Biochemical analysis for relevant tumor markers in blood and urine was performed. All patients were allowed to continue their medication. The local medical ethics committee approved the study and all patients gave written informed consent. Procedures 11 C-5-HTP PET and 18 F-DOPA PET For the reduction of tracer decarboxylation and subsequent renal clearance all patients received 2 mg/kg carbidopa orally as pre-treatment 1 h prior to the 11 C-5-HTP and 18 F- DOPA injection to increase tracer uptake in tumor cells 16, C-5-HTP was produced using 75

91 Chapter 5 a multi-enzymatic synthesis of enantiomerically pure 11 C-5-HTP on a Zymark robotic system 14. Patients fasted for 2 h before the examination. Whole body 2D-PET images were acquired 10 min after the intravenous (IV) administration of 11 C-5-HTP (200 ± 50 MBq, with an estimated mean radiation dose of 0.67 msv) on a Siemens ECAT HR+ positron camera (Siemens, Knoxville, TN, USA) with attenuation correction (7-10 bed positions of 5 min emission and 3 min transmission scan). 18 F-DOPA was produced as described earlier 17. Patients fasted for 6 h before the examination. Whole body 2D-PET images were acquired as described for 11 C-5-HTP PET 60 min after the IV administration of 18 F-DOPA (180 ± 50 MBq, with mean radiation dose of 4 msv) 18. Two nuclear medicine physicians (KPK, PLJ) blinded for the results of other imaging examinations and clinical information interpreted the sets of 11 C-HTP and 18 F-DOPA PET images independently. Discrepant cases were reviewed in a multidisciplinary team and a consensus was reached. Only lesions with an unequivocal visibility clearly above normal activity in that body region were considered abnormal. Somatostatin receptor scintigraphy According to Dutch standards, 24 h after IV administration of 200 MBq 111 In-octreotide (Octreoscan; Mallinckrodt, Petten, The Netherlands with an estimated mean radiation dose of 10 msv 19 ), planar total-body and SPECT images were obtained using standard methods. (Siemens Multispect 2 gammacamera, medium energy collimator, 10 min spotviews, 64 projections of 30 s). If interfering bowel activity was observed, 48 h images were recorded 20. SRS scans were interpreted by dedicated specialists as part of routine patient care and subsequently independently reread by a nuclear medicine physician (PLJ) blinded for the results of other imaging examinations and clinical information. CT CT (4-16 slice, Siemens Somatom Sensation, Siemens Medical Systems, Erlangen, Germany with an estimated mean radiation dose of 8-20 msv) 21 was performed using oral contrast and IV contrast (Visipaque 270, 120 ml, 2.5 ml/s) enhancement. The reconstruction interval was mm. All patients underwent an abdominal CT, 42 patients also a chest CT. CT scans were interpreted as part of routine patient care and were reread by an experienced radiologist (KV) blinded for the clinical information. In discrepant cases consensus was reached after multidisciplinary discussion. Composite Reference standard As a composite reference standard for presence of tumor lesions, all available cytological, histological, follow-up findings and all imaging findings were used. This is considered the optimal gold standard, as cytological or histological verification of every lesion is not feasible and not justifiable in these patients 5. Whenever possible, new findings on PET were verified with additional investigations. 76

92 Improved staging and characterization of lesions in patients with carcinoid and islet cell tumors with 18 F-DOPA and 11 C-5-HTP positron emission tomography Biochemical analysis As markers for serotonin metabolism we measured serotonin levels in platelets and urinary 5-hydroxy indole acetic acid (5-HIAA) in a 24 h urine collection (upper reference limits 5.4 nmol/10 9 platelets and 3.8 mmol/mol creatinine, respectively) 22,23,24,25. Serum chromogranin A was determined using a radioimmunoassay (Cga-React, Cis Bio International, Gif-sur- Yvette, France) as a marker for general neuroendocrine tumor activity (reference interval mg/l). Data and statistical analysis The STARD checklist was used during design and writing of this report 26. Based on earlier studies, 18 F-DOPA PET and 11 C-5-HTP PET are both accurate techniques for staging of neuroendocrine tumors, but results may differ in subgroups 5,6,14. Therefore we aimed to study approximately 25 carcinoid and 25 islet cell tumor patients to make a statistically meaningful comparison between both diagnostic methods. We wanted to be able to document an increase in sensitivity from 65% (average value, for islet cell tumor patients for conventional imaging) to 90% with 11 C-5-HTP PET, using McNemar s test for comparison with 80% power and 5% two-sided significance levels. Analyses were performed at the level of individual patients and individual lesions. When the number of lesions in one organ (e.g. liver) was higher than 10, the number of lesions was truncated at 10 for that region to avoid bias. PET and SRS are whole body modalities, while CT only covers the most relevant parts of the body. In order to eliminate bias towards total body imaging methods, only body areas for which all four imaging modalities were available have been evaluated. Sensitivities were calculated using the composite reference standard, and were compared using paired observations and McNemar s test. Patient based sensitivity was calculated as number of patients with a positive test (at least one lesion detected) by total number of patients. Per lesion sensitivity of a modality was calculated by dividing the number of positive lesions detected with that modality by the total number of positive lesions. Significance level was 0.05, two-sided. The statistical tests were carried out using the SPSS package version Results Patients Between February 2005 and February 2007, 50 consecutive patients were recruited, of which 3 patients declined one or more of the imaging procedures leaving full data of 24 patients with a carcinoid tumor and 23 patients with an islet cell tumor for analysis, as presented in the flow diagram (figure 1). Patient characteristics are presented in table 1. 77

93 Chapter 5 50 patients 47 patients 3 patients refused complete imaging 24 patients with carcinoid tumor 23 patients with islet cell tumor Figure 1. In this flow diagram a schematically view of the recruited patients is given. All carcinoid patients and 39% of islet cell tumor patients had biochemical proof of increased serotonin metabolism. CT, SRS, 11 C-5-HTP PET, 18 F-DOPA PET were carried out within a median of 55 days. In 29 patients both PET scans were performed on the same day. The mean interval between both PET scans was 18 days. Newly detected lesions were confirmed with MRI (n=2), bone scintigraphy (n=2), planar X-ray (n=3) and sonography (n=3), surgery (n=3) or biopsy (n=1). Results in a representative patient are shown in figure 2. Table 1. Patient characteristics. Characteristics Value Sex (n of patients) Male/Female 29/18 Median age in years (range) 56 (18-79) New patients vs. patients with known disease (no patients) 11/36 Patients with abdominal carcinoid (n patients) 24 Patients with carcinoid syndrome 12 Treatment during scan Somatostatin analogues (n) 17 Somatostatin analogues + interferon (n) 2 Patients with islet cell tumors (n patients) 23 Treatment during scan Somatostatin analogues (n) 4 Somatostatin analogues + interferon (n) 1 Chemotherapy 1 Radiotherapy on bone metastases 1 78

94 Improved staging and characterization of lesions in patients with carcinoid and islet cell tumors with 18 F-DOPA and 11 C-5-HTP positron emission tomography A B C D Figure 2. Full-color in appendix. Fused 18 F-DOPA PET CT scan (A), SRS (B), 18 F-DOPA PET (C) and 11 C-5-HTP PET (D) of a 80 year old male patient with metastatic carcinoid tumor. The CT scan shows a mesenterial mass and two smaller lesions in the upper mediastinum. On SRS (both planar and SPECT, not shown here) only the larger mediastinal mass, the large mesenterial mass and a small lesion on the left cranial side of the urinary bladder could be found. Both 18 F-DOPA PET and 11 C-5-HTP PET showed a number of smaller lesions in the upper mediastinum and upper lobes of both right and left lung, with 18 F-DOPA yielding the best contrast. Note that the small lung lesions show less 11 C-5-HTP uptake than 18 F -DOPA uptake. Patient based analysis Based on our selection criteria, all patients were considered positive for tumor. In a per patient analysis in carcinoid tumor patients, 11 C-5-HTP PET detected one or more tumor lesions in all 24 patients (sensitivity 100%, table 2), whereas 18 F-DOPA PET and CT detected one or more tumor lesions in 23 of 24 patients (sensitivity 96%) and SRS detected one or more tumor lesions in 18 of 21 patients (sensitivity 86%). In a per patient analysis in patients with islet cell tumors, 11 C-5-HTP detected one or more tumor lesions in 23 of a total of 23 patients (sensitivity 100%) CT detected one or more tumor lesions in 20 of 23 patients (sensitivity 87%), SRS in 14 of 19 patients (sensitivity 78%) and 18 F-DOPA PET 16 of 23 patients (sensitivity 89). However, there were no statistically significant differences. Lesion based analysis In patients with a carcinoid tumor, 371 tumor lesions were detected based on the composite reference standard (table 3). The largest number of lesions was present in liver and abdomen (75% of all). 18 F-DOPA PET and 11 C-5-HTP PET had the highest sensitivity for the detection of these lesions compared to the other imaging modalities. The smallest lesion size that could be detected with 18 F-DOPA PET and 11 C-5-HTP PET was approximately 5 mm, as measured and confirmed on the PET-CT fused images. Overall 18 F-DOPA PET found most lesions, followed by 11 C-5-HTP PET. However, this difference was not statistically significant. 18 F-DOPA PET and 11 C-5-HTP PET were both significantly better in detecting tumor lesions than SRS ( 18 F-DOPA PET: p=0.001 for 11 C-5-HTP PET: p=0.008). The combination 18 F-DOPA PET with CT had the highest sensitivity for 79

95 Chapter 5 detection of carcinoid lesions ( 98%), as CT detected lesions missed by nuclear medicine techniques, and vice versa. Therefore, combining nuclear medicine techniques with CT yielded more lesions. When SRS would have been left out, not a single lesion would have been missed. In patients with islet cell tumors, a total of 294 tumor lesions were detected. Most lesions (71%) were found in the liver and abdomen. In these patients 11 C-5-HTP PET and CT performed equally well, and were both better than the other imaging modalities, although not statistically significant. Both SRS and 18 F-DOPA PET had a relatively poor performance for islet cell tumor detection. Again, combining SRS and PET with CT led to an increased number of detected islet cell tumor lesions and therefore increased sensitivity (table 3) The combination of 11 C-5-HTP PET with CT had the highest sensitivity. When SRS would have been left out, only 8% of all lesions would have been missed. These PET negative lesions were found in two patients. There was no statistical relationship between elevation of biochemical parameters and imaging results of 18 F-DOPA PET, 11 C-5-HTP PET, SRS, or CT. Table 2. Patient based analysis. Imaging modality Number of Sensitivity patients with (95%CI) positive lesions Mean and median number of lesions per patient (range) Carcinoid tumor (n=24) CT 23 96% (78-100) 7.5; 9.7 (0 30) SRS 18 86% (62 95) 5.5; 6.9 (0 30) 18 F -DOPA PET 23 96% (73 100) 11.0; 13.4 (0 33) 11 C-5-HTP PET % (85 100) 10.0; 12.1 (0 33) Islet cell tumor (n=23) CT 20 87% (66 97) 7. 0; 8.7 ( 0 41) SRS 14 78% (56 97) 1. 0; 5.1 ( 0 40) 18 F-DOPA PET 16 89% (66 97) 1. 0; 5.2 (0 40) 11 C-5-HTP PET % (84 100) 3. 0; 8.7 ( 1 40) T he results for the patient based analysis are presented with the number of tumor positive patients, patient based sensitivity with 95%CI and the mean and median number of lesions per patient. 80

96 Improved staging and characterization of lesions in patients with carcinoid and islet cell tumors with 18 F-DOPA and 11 C-5-HTP positron emission tomography Table 3. Lesion based analysis. CT Sensitivity (95%CI) SRS Sen (95% CI) sitivity SRS + CT Sensitivity (95%CI) 18 F -DOPA Sensitivity (95%CI) PET 18 F -DOPA CT Sensitiv (95%CI) PET + ity 11 C-5-HTP PET Sensitivity (95%CI) 11 C-5-HTP PET + CT Sensitivity (95%CI) Number of positive regions Carcinoid tumors Head and neck 46% (19-75) 31% (8 7-62) 69% (3 8-91) 85% (5 4-99) 92% ( ) Mediastinal* Lung* Liver 75% (53-90) 60% (34-84) 67% (59-74) 25% 50% 62% 0-47) 1-74) 4-70)* 100% ( ) 73% (44-93) 100% ( % (12-1 Abdomen / pelvis 68% (59-76) 5 0% (4 1-59) 83% (7 5-89) 87% (79-92 ) 99% ( ) 91% (6 90% (5 78% (7 58% (3 40% (1 93% (8 Pancreas 50% (0-100) 0% (0-88) 50% (0-1 00) 50% (0-1 00) Bone 28% (14-45) 6 9% (5 Extremities Total 63% (58-68) 4 9% (44-54 (1 (2 (5 2-84) 69% (5 100 ) 7-98) 9-99) 1-85) 6-78) 6-68) 8-96) 2-84) 100% (89-100) 0 73% (6 8-78) 87% (84-91 ) 00) 00) 100% ( ) 0 98% (96-99) 85% (54-99) 54% (32-75) 13% (1-41) 84% (78-90) 50% (0-100) 81% (73-87) 92% (77-98) 0 78% (74-83) 92% (63-100) 54% (32-75) 73% (44-93) 91% (86-95) 100% (12-100) 95% (89-98) 92% (77-98) 0 89% (86-92) Islet cell tumors Head and neck 0% (0-63) 10 0% ( ) 100% ( ) Mediastinal * Lung* Liver Pancreas Abdomen / pelvis Bone 61% (40-79) 0 77% (69-84) 50% (26-74) 80% (6-89) 32% (17-51) 7% (18-56 ) 7% (48-65 ) 0% (26-74 ) 1% (20-42 ) 6% (27-62 ) 25% (0-82) 43% ( % ( % ( % ( % ( % (0-82) 68% (47-84) 0 86% ( ) 83% (58-97) 83% (72-91) 68% (49-83) Extremities 50% (0-100) 0% (0-89) 0% (0-89 ) 50% (0-1 00) 100% ( ) 63% (4 0 85% (7 86% (5 81% (6 46% (2 0-79) 8-91) 8-97) 9-89) 7-62) ) ) ) ) ) % (5-95) 79% (59-92) 0 67% (59-75) 79% (52-94) 50% (38-62) 97% (84-100) 50% (0-100) 50% (5-94) 93% (76-99) 0 96% (91-99) 94% (72-100) 100% (95-100) 97% (84-100) 50% (0-100) Total 68% (63-74) 4 6% (40-52 ) 77% (72-82 ) 41% (36-47 ) 80% (75-85) 67% (62-73) 96% (93-98) 294 The sensitivities for the combination SRS w other, the results for the combination of nucl with 18 F-DOPA PET and p=0.008 for the co ith CT, 18 F-DOPA PET with CT and 11 C-5-HTP PE T wi th CT are shown. To illustrate the additional value of combining these scans with each ear imagi ng methods with CT are shown. * p = for the comparison o f SRS with 18 F -DOPA PET; p=0.001 for the comparison of SRS mparison of SRS with 11 C-5- HTP PET. 81

97 Chapter 5 Discussion This study demonstrates that both 11 C-5-HTP PET and 18 F-DOPA PET have excellent sensitivity to detect carcinoid and islet cell tumors lesions. 11 C-5-HTP PET was the only imaging method, which was able to detect tumor lesions in all carcinoid and islet cell tumor patients. In carcinoid patients 18 F-DOPA PET was the best modality as it detected more lesions compared to all other modalities including 11 C-5-HTP PET, CT and SRS. In islet cells tumors 11 C-5-HTP PET detected more tumor positive patients and lesions than 18 F- DOPA PET and SRS (figure 3). Adding CT to both PET techniques resulted in a slight further improvement in sensitivity (table 3). Therefore 18 F-DOPA PET-CT is considered the optimal technique for staging of patients with carcinoid tumors, and 11 C-5-HTP PET- CT for islet cell tumor patients. In patients with carcinoid tumors, SRS scanning can be omitted without missing any lesions. For islet cell tumors, this is less clear-cut. 11 C-5-HTP PET combined with CT gives the best tumor detection for most patients. However, in a minority, namely 8% of patients, SRS performs equal or better then metabolic PET imaging methods. Therefore, in patients with islet cell tumors SRS remains of additional value. Overactivity of the serotonin and most likely also the catecholamine pathway appears to be the key factor that determines the intracellular tracer c oncentration. Increased activity of transmembrane amino acid tr ansporters results in high entry of both tr acers in cells. In the tumoral cytoplasm 11 C-5-HTP and 18 F-DOPA PET are metabolized via th e abundantly present enzyme aromatic amino acid decarboxylase (AADC) to hormonal products which can be stored in pathway specific secretory vesicles. In contrast to islet cell tumors, most patients with carcinoid tumors the serot onin pathway is highly active. In these cells, the storage capability for the 11 C-5-HTP metabolites is relatively saturate d by endogenous serotonin. The 11 C-5-HTP metabolites are therefore rapidly degrade d via mono amino oxidase activity and subsequently excreted from the cell. This may explain the superior d ance for iagnostic perform F-DOPA PET in carcinoid and C-5-HTP in islet cell tumors 27. The intracellular tracer concentration is directly related to the probabilit y of visualization using the PET scanner. These high tracer concentr ations allow the det ection of smaller l esions, up to 5 mm in diameter. In both patient groups CT detected additional lesions and was therefore complementary to the PET techniques. The combination of 11 C-5-HTP P ET with CT prove d to be the best method to detect islet cell tumor lesions, whereas the combination 18 F-DOPA PET with CT detected most tumor lesions in patients with carcinoid disease. Both PET combinations performed better then the combination of SRS with CT in both tumor types. Although difficult to quantify, another advantage of combining CT with 11 C-5-HTP or 18 F-DOPA PET is the ability to characterize neuroendocrine origin of lesions of lesions found on CT. Both PET methods allow better staging and estimation of the total body tumor load. The addition of 11 C-5-HTP PET to CT in islet cell tumor patients clearly helps to provide a better understanding of the number of lesions and their distribution. This will support treatment decisions. In addition, better estimation of the total body tumor load and the detection of metastases in unknown regions may refine clinical management. Finally, the recent development of combined PET-CT scanning gives superior diagnostic information in 82

98 Improved staging and characterization of lesions in patients with carcinoid and islet cell tumors with 18 F-DOPA and 11 C-5-HTP positron emission tomography a single session, largely obviates the need for SRS, and reduces the burden of multiple diagnostic tests. 11 All other published data regarding C-5-HTP are from the group of Uppsala, Sweden. The y studied 42 patie nts with a m ixture of neuroendocrine and non-endocrine tumor patients. They concluded that 11 C-5-HTP PET was superior to SRS and CT for neuroendocrine tumor lesions and could be regarded as a universal imaging agent for these tumours 13. No head to head comparison of 18 F-DOPA and 11 C-5-HTP versus CT and SRS was performe d. Recent data also point t o the utility of 18 F-DOPA PET in assessing pancreatic lesions in infants and adults with hyperinsulinism 28. A B C D Figure 3. Full-color in appendix. CT scan (A), SRS (B), 18 F-DOPA PET ( C) and 11 C-5-HTP PET (D) of a 54 year old male patient with metastatic islet cell tumor. The CT scan shows a large mass in th e pancreatic head region (a rrow), SRS shows equivocal (arrow) and 18 F-DOPA PET shows low uptake in the pancreatic region and minor uptake in the upper chest and in two thoracic vertebrae. 11 C-5-HTP PET, however, shows numerous bone, liver and abdominal lesions, including the pancreatic region with much higher contrast. In our study in both patient g roups 11 C-5-HTP PET was far superio r to SRS and in islet cell tumor nts patie C-5-HTP PET performed even bett er than F-DOPA PET. Therefore, 11 C- 5-HT P PET could indeed be seen as a universal imaging agent for carcinoid and islet cell 11 tumors. However, the synthesis of C-5-HTP PE T is complex. For efficient use of available resources and time it seems logical to use 18 F-DOPA PET for all patients with non-islet cell tumors and 11 C-5-HTP PET only for those with a proven or a suspected islet cell tumor. As PET is now in general perfo rmed in combination with C T, this even further improves the lesion detection and characterization properties of both PET scans and provides anatomical information all in a single and rapid session. Acknowledgements Supported by grant from the Dutch Cancer Societ y. 83

99 Chapter 5 References 1. Pearse AG. The APUD cell concept and its implications in pathology. Pathol Annu 1974; 9: Moertel CG, Lefkopoulo M, Lipsitz S et al. Streptozocin-doxorubicin, streptozocin-fluorouracil or chlorozotocin in the treatment of advanced islet-cell carcinoma. N Engl J Med 1992; 326: Modlin IM, Kidd M, Latich I et al. Current status of gastrointestinal carcinoids. Gastroenterology 2005; 128: Plockinger U, Rindi G, Arnold R et al. Guidelines for the diagnosis and treatment of neuroendocrine gastrointestinal tumors. A consensus statement on behalf of the European Neuroendocrine Tumour Society (ENETS). Neuroendocrinology 2004; 80: Koopmans KP, de Vries EG, Kema IP et al. Staging of carcinoid tumors with 18 F- DOPA PET: a prospective, diagnostic accuracy study. Lancet Oncol 2006; 7: Hoegerle S, Altehoefer C, Ghanem N et al. Whole-body 18 F dopa PET for detection of gastrointestinal carcinoid tumors. Radiology 2001; 220: Kumbasar B, Kamel IR, Tekes A et al. Imaging of neuroendocrine tumors: accuracy of helical CT versus SRS. Abdom Imaging 2004; 29: De Herder WW, Hofland LJ, van der Lely AJ et al: Somatostatin receptors in gastroentero-pancreatic neuroendocrine tumors. Endocr Relat Cancer 2003; 10: Fahey FH, Harkness BA, Keyes JW, Jr. et al. Sensitivity, resolution and image quality with a multi-head SPECT camera. J Nucl Med 1992; 33: Becherer A, Szabo M, Karanikas G et al. Imaging of advanced neuroendocrine tumors with (18)F-FDOPA PET. J Nucl Med 2004; 45: Meijer WG, Copray SC, Hollema H et al. Catecholamine-synthesizing enzymes in carcinoid tumors and pheochromocytomas. Clin Chem 2003; 49: Bjurling P, Watanabe Y, Tokushige M et al. Synthesis of β- 11 C-labelled L- tryptophan and 5-hydroxy-L-tryptophan using a multi-enzymatic reaction route. J Chem Soc 1981; Örlefors H, Sundin A, Garske U et al. Whole-body 11 C-5-hydroxytryptophan positron emission tomography as a universal imaging technique for neuroendocrine tumors: comparison with somatostatin receptor scintigraphy and computed tomography. J Clin Endocrinol Metab 2005; 90: Neels OC, Jager PL, Koopmans KP et al. Development of a reliable remotecontrolled synthesis of β-[ 11 C] - 5-hydroxy-L-tryptophan on a Zymark robotic system. J Labelled Comp Radiopharm 2006; 49: Bergström M, Lu L, Eriksson B et al. Modulation of organ uptake of 11 C-labelled 5-hydroxytryptophan. Biog Amines 1996; 12: Örlefors H, Sundin A, Lu L et al. Carbidopa pretreatment improves image interpretation and visualisation of carcinoid tumors with 11 C-5-hydroxytryptophan positron emission tomography. Eur J Nucl Med Mol Imaging 2005; 33:

100 Improved staging and characterization of lesions in patients with carcinoid and islet cell tumors with 18 F-DOPA and 11 C-5-HTP positron emission tomography 17. Vries EFJ, Luurtsema G, Brussermann M et al. Fully automated synthesis module 18 for the high yield one-pot preparation of 6-[ F]-fluoro-L-DOPA. Appl Radiat Isot 1999; 51: Brown WD, Oakes TR, DeJesus OT et al. Fluorine-18-fluoro-L-DOPA dosimetry with carbidopa pretreatment. J Nucl Med 1998; 39: International Commission on Radiological Protection: ICRP publication 80: radiation dose to patients from radiopharmaceuticals. Pergamom, Oxford: Elsevier, Balon HR, Goldsmith SJ, Siegel BA et al. Procedure guideline for somatostatin 111 receptor scintigraphy with In-pentetreotide. J Nucl Med 2001; 42: International Commission on Radiological Protection: ICRP Publication 87: Managing patient dose in computed tomography. ICRP Online. Elsevier, Kema IP, Schellings AM, Hoppenbrouwers CJ et al. High performance liquid chromatographic profiling of tryptophan and related indoles in body fluids and tissues of carcinoid patients. Clin Chim Acta 1993; 221: Kema IP, Meiborg G, Nagel GT et al. Isotope dilution ammonia chemical ionization mass fragmentographic analysis of urinary 3O-methylated catecholamine metabolites. Rapid sample clean-up by derivatization and extraction of lyophilic samples. J Chromatogr Biomed Appl 1993; 671: Kema IP, Meijer WG, Meiborg G et al. Profiling of tryptophan-related plasma indoles in patients with carcinoid tumors by automated, on-line, solid-phase extraction and HPLC with fluorescence detection. Clin Chem 2001; 47: Willemsen JJ, Ross HA, Wolthers BG et al. Evaluation of specific high- liquid-chromatographic determinations of urinary metanephrine and performance normetanephrine by comparison w ith isotope dilution mass spectrometry. Ann Clin Biochem 2001;38: Bossuyt PM, Reitsma JB, Bruns DE et al. Towards complete and accurate reporting of studies of diagnostic accuracy: the STARD initiative. Standards for reporting of diagnostic accuracy. Clin Chem 2003; 49: Smith CC. Evidence for separate serotonin and catecholamine compartments in human platelets. Biochim Biophys Acta 1996; 1291: Mohnike K, Blankenstein O, Christesen HT et al. Proposal for a standardized protocol for 18 F-DOPA-PET (PET/CT) in congenital hyperinsulinism. Horm Res 2006; 66:

101 86

102 Chapter 6 5-Fluorotryptophan as potential PET tracer for neuroendocrine tumors Oliver C. Neels 1, Klaas P. Koopmans 1, Pieter L. Jager 1, Laya Vercauteren 2, Hetty Timmer- Bosscha 3, Adrienne H. Brouwers 1, Elisabeth G.E. de Vries 3, Rudi A. Dierckx 1, I.P. Kema 4, 1 Philip H. Elsinga. Departments of Nuclear Medicine and Molecular Imaging 1, Medical Oncology 3, Pathology and Laboratory Medicine 4, University of Groningen and University Medical Center Groningen, The Netherlands Department of Pharmacy 2, University of Ghent, Belgium Submitted for publication. 87

103 Chapter 6 Summary [ 11 C]-5-HTP scanning in patients showed powerful results in the detection of neuroendocrine tumors. An 18 F labeled tryptophan analog with a longer radioactive half-life than [ 11 C]-5-HTP would increase clinical applicability because availability of [ 18 F]fluorotryptophan enables widespread use. Our results show that 5-fluorotryptophan (5- FTP) has, similar to 5-HTP, favorable characteristics to accumulate in a human neuroendocrine tumor cell line. An enzymatic synthesis towards 5-FTP gave high yields (73 ± 6 %) showing that an analogous synthesis of [ 18 F]-5-FTP is feasible. 5-FTP is a potential tracer for the detection of neuroendocrine tumors. Introduction Positron emission tomography (PET) of neuroendocrine tumors (NETs) demonstrated high contrast images by using the labeled large amino acids 6-[ 18 F]fluoro-levodopa ([ 18 F]FDOPA) and [ 11 C]-5-hydroxytryptophan ([ 11 C]-5-HTP). While [ 18 F]FDOPA is widely used in imaging NETs, the use of [ 11 C]-5-HTP is limited to centers with a cyclotron on site due to the short half-life of 11 C (20.4 min). To date, only a few PET-centers worldwide are performing successful studies with [ 11 C]-5-HTP PET for staging of neuroendocrine tumors 1-6. Given the interesting results with [ 11 C]-5-HTP scanning in patients and the short half-life of [ 11 C]-5-HTP, the interest in an 18 F labeled tryptophan analog (half-life 18 F is 110 min) has increased in the past years 7,8. Amine precursor uptake and decarboxylation (APUD) 9 in NET plays a major role in developing new amino acids for precise staging of NETs. Labeling of amino acids with 18 F has been described by using [ 18 F]fluoride as well as [ 18 F]F 2 gas 8. The nucleophilic substitution reaction with [ 18 F]fluoride is usually the preferred labeling method because of its high specific radioactivity. However, direct ring labeling is not possible because these aromatic rings are not activated for nucleophilic substitution. The electrophilic substitution reaction with [ 18 F]fluorine gas is the most widely used method to label large amino acids containing an aromatic ring structure like [ 18 F]FDOPA directly at the ring position Fluorotryptophan (5-FTP) is an analogue of tryptophan with the characteristic properties of an amino acid. 5-FTP contains a fluorine atom that is not present in natural amino acids and therefore does not belong to the group of proteinogenic amino acids. To verify if 5-FTP is a potential PET tracer following the APUD principle we studied non-labeled tracer accumulation in a human NET cell line (BON) with 5-hydroxytryptophan (5-HTP) and 5- FTP using amino acid decarboxylase (AADC) and monoamine oxidase (MAO) inhibitors. AADC is responsible for the decarboxylation of amino acids such as 5-HTP and levodopa into their corresponding amines. MAO metabolizes these amines to 5-hydroxyindole acetic acid, respectively homovanillic acid. A tentative view on the metabolism of 5-FTP in neuroendocrine tumors is shown in figure 1. For the synthesis of [ 18 F]-5-FTP our strategy was to combine methods used in the synthesis of [ 11 C]HTP 1 and [ 18 F]FDOPA 14. The synthesis can be divided into two steps (figure 2), namely 1) the formation of [ 18 F]fluoroindole and 2) the reversed tryptophanase reaction 15 giving [ 18 F]-5-FTP. For the first step, 5-trimethylstannylindole will be investigated as a starting compound for fluorodestannylation to from 5-fluoroindole 16. Fluorodestannylation is a method to label aromatic ring structures with F 2 gas from stannylated compounds. To investigate if 5-fluoroindole is a good substrate for the tryptophanase reaction, different 88

104 5-Fluorotryptophan as potential PET tracer for neuroendocrine tumors parameters like ph and concentrations of reactants will be used to optimize this latter reaction to obtain 5-FTP. NH 2 NH2 F COOH AADC F MAO F COOH N H 5-Fluorotryptophan N H 5-Fluorotryptamine N H 5-Fluoroindole acetic acid Figure 1. Tentative view on the metabolism of 5-fluorotryptophan. The aim of these experiments is to investigate if an enzymatic synthesis of an 18 F tryptophan analogue is feasible, and if the resulting 5-fluorotryptophan is accumulated in a BON cell line. Br Pd[P(C 6 H 5 ) 3 ] 4 (H 3 C) 3 Sn F A (Sn(CH 3 ) 3 ) 2 F 2 N H 5-Bromoindole N H 5-Trimethylstannylindole N H 5-Fluoroindole COOH B F N H 5-Fluoroindole + O OH O Pyruvic acid Tryptophanase Pyridoxal 5'-phosphate (NH 4 ) 2 SO 4 ph temperature TRIS 0.1 M/EtOH (75/25) F N H 5-Fluorotryptophan NH 2 Figure 2. 5-Fluorotryptophan synthesis: A Stannylation of 5-bromoindole and subsequent fluorination with F 2 gas; B the reversed tryptophanase reaction applied on 5-fluoroindole. Materials & Methods Chemicals Chemicals and solvents were obtained from Sigma (Zwijndrecht, The Netherlands), Merck (Amsterdam, The Netherlands), Janssen (Geel, Belgium) and Rathburn (Walkerburn, Scotland). 0.3 % F 2 in neon (6.2 bar) was used. Tris(hydroxymethyl)-aminomethane (TRIS) and ammonium sulfate (AMS) were dissolved in distilled water. Pyridoxal 5 -phosphate (PLP) was dissolved in 0.1 M sterile phosphate buffer ph 7.2. Tryptophanase (TRP) dissolved in 20 mm potassium phosphate buffer ph 7.5, 0.1 mm pyridoxal 5 -phosphate, 0.1 mm dithiotheitol and 20% glycerol was purchased from Ikeda (Hiroshima, Japan) and used without further treatment. GMC (5.6 mm D-glucose, 0.49 mm MgCl 2, 0.68 mm 89

105 Chapter 6 CaCl 2 ) was added to phosphate-buffered saline solution (PBS; 140 mm NaCl, 2.7 mm KCl, 6.4 mm Na 2 HPO 4 x2h 2 O, 0.2 mm KH 2 PO 4 ). ph of the resulting PBS-GMC solutions was adjusted to 7.4 with sodium hydroxide. PBS-GMC was used to deplete internal amino acid pools 17,18. Cell culture method Cellular accumulation experiments were performed with cells derived from a human neuroendocrine pancreatic tumor (BON) 19. Cells were maintained in 25 cm 2 culture flasks in 5 ml D-MEM/F-12 (1:1) medium supplemented with 10 % fetal calf serum (FCS). They were grown in a humidified atmosphere containing 5 % CO 2 and were routinely subcultured every 3-4 days. For experiments, the cells were grown to a concentration of 1.4 x 10 6 cells/ml. Viable cell number was determined by the trypan blue exclusion technique. Cell viability during and 1 hour after experiments was over 90%. Analytical methods Semipreparative HPLC purification: System A: Column: Chromspher Si (250 x 10 mm), 5 µm; Eluent: Hexane/EtOAc : 9/1; Flow rate: 5 ml/min; Temperature: RT; UV detection at λ: 280 nm; Retention times: 5- trimethylstannylindole 13 min, 5-fluoroindole 14 min. System B: Column: Econosphere C18 (250 x 10 mm), 10 µm; Eluent: NaH 2 PO 4 /EtOH: 9/1; Flow rate: 5 ml/min; Temperature: RT; UV detection at λ: 280 nm; Retention time 5- fluorotryptophan 19 min. An electrospray mass spectrum was recorded on an API 3000 triple quadrupole mass spectrometer (MDS-Science, Concord, Ontario, Canada). Nuclear magnetic response spectroscopy was performed on a Bruker 200 MHz. Experimental Cell experiments Experiments were performed in 25 cm 2 culture flasks. Culture medium was removed and cells were washed with warm PBS (3 x 2 ml). PBS-GMC buffer (5 ml) containing the AADC inhibitor carbidopa (0.08 mm), the MAO A inhibitor clorgyline (0.1 mm) or the MAO B inhibitor pargyline (0.1 mm) was added. The cells were then placed in a water bath at 37 C. After 1 hour depletion period PBS-GMC buffer was removed and PBS-GMC buffer containing unlabeled 5-HTP or 5-FTP (55 µm) was added. After 15 and 60 min of tracer incubation PBS-GMC buffer was removed. BON cells were harvested with 1 ml trypsin and taken up in 1 ml PBS containing 10% FCS. BON cells were washed three times with ice-cold PBS (1 ml) and centrifuged 10 min at 10,000 g. Lysis was performed with liquid nitrogen. Concentrations of 5-HTP and 5-FTP from lysed BON cells dissolved in 1 ml saline were determined as described earlier 20. Results represent the mean of 3 experiments ± standard error of the mean (SEM). Statistics Differences between various groups (control and treated) were tested for statistical significance using Student s paired t-test. P-values < 0.05 were considered as significant. 90

106 5-Fluorotryptophan as potential PET tracer for neuroendocrine tumors Chemistry 5-Trimethylstannyl-1H-indole (1) 1.53 g 5-Bromoindole (7.8 mmol) was dissolved in 20 ml anhydrous 1,4-dioxane and 909 mg tetrakis triphenylphosphine palladium (0) (0.4 mmol) in 5 ml anhydrous 1,4-dioxane was added. After addition of 2.1 g hexamethylditin (15.3 mmol) the solution was refluxed for 2 days under argon. After cooling down, the mixture was taken up in 80 ml ethyl acetate, filtered and washed with water and brine. The organic residues were collected, dried over Na 2 SO 4 and evaporated to dryness. The crude product was purified by silica gel chromatography (eluent: Hexane/EtOAc : 3/1; Rf: 0.79) to give 330 mg 5-trimethylstannyl- 1H-indole as a solid in a yield of 15 %. Mp.: 48.3 C. 1 H-NMR (CD 3 OD, 298 K): 7.39 (s, 1H, H-4), 7.13 (m, 1H, H-7), 7.09 (m, 1H, H-2), 6.88 (m, 1H, H-6) 6.14 (m, 1H, H-3), 0.13 ( s, 9H, H-8); 13 C NMR (CD 3 OD, 298 K): (C-8), (C-5), (C-4), (C- 3), (C-6), (C-1), (C-7), (C-2), (C-9). 5-Fluoroindole (2) mg ( µmol) 1 was dissolved in 2 ml dry acetonitrile (from 4Å molecular sieves) and cooled down to 0 C. 0.3 % F 2 in neon gas was bubbled through the solution at a flow rate of 200 ml/min within 10 min including one purge of pure neon gas. After heating at 50 C for 10 min in a closed reaction vessel the mixture was filtered via a 22 µm LG pore filter. The mixture was purified by HPLC (system A) and evaporated to dryness with a rotary evaporator to obtain 5-fluoroindole. Enzymatic synthesis of 5-fluorotryptophan The reversed tryptophanase reaction is controlled by different parameters like the concentrations of used chemicals and enzyme, ph and reaction temperature (figure 2). Fixed parameters were the amount of 1.35 mg 5-fluoroindole (10 µmol), 50 µl TRP, 10 µl PLP 10.7 mm and 1.5 ml TRIS 0.1 M solvent containing 25 % ethanol. First, AMS 1.5 M was added to the solution of 5-fluoroindole followed by pyruvic acid. ph was then adjusted, tryptophanase and PLP were added and the mixture was incubated for 10 min. Reaction was stopped by addition of 2 drops HCl 6 M followed by filtration through a 22 µm GP pore filter and subsequent adjustment to ph ~ 9 and injection on HPLC system B. ph, temperature and the amounts of AMS 1.5 M and pyruvic acid were varied in several experiments to optimize yields. Yields were determined by HPLC. The experiment was repeated five times under optimal conditions times to determine yield as mean ± SEM. 5-Fluorotryptophan (3) Purified 2 was dissolved in 1.5 ml TRIS 0.1 M/ethanol (75/25), 400 µl AMS 1.5 M and 20 µl pyruvic acid. The mixture of the ph was adjusted to ph and 50 µl TRP and 10 µl PLP were added and warmed at 40 C for 10 min in a metal heating block. The reaction was stopped by addition of 3 drops HCl 6 M and filtration on a 22 µm GP pore filter. The colorless clear mixture was purified on HPLC (system B) and the corresponding peak for 5- fluorotryptophan was analyzed by mass spectroscopy. 91

107 Chapter 6 Results Cellular accumulation of 5-HTP and 5-FTP After 15 min of incubation, significant levels of 5-FTP (0.6 % of all 5-FTP in medium) were detected in cell lysates compared to 5-HTP (3.4 % of all 5-HTP in medium). After 60 min incubation period 5-FTP levels remained constant (0.5 % of all 5-FTP in medium) while 5-HTP levels increased over time (12.0 % of all 5-HTP in medium). Treatment with AADC and MAO inhibitors lowered 5-HTP levels after 60 min (carbidopa: 3.0 %, clorgyline: 1.5 %, pargyline: 3.1 % of all 5-HTP in medium) but led to an increase of 5- FTP levels (carbidopa: 0.9 %, clorgyline: 1.1 %, pargyline: 2.1 % of all 5-FTP in medium) (figure 3) Control Carbidopa Clorgyline Pargyline % % HTP 5-FTP 0 5-HTP 5-FTP Figure 3. Accumulation of 5-HTP and 5-FTP in BON cells. Left: 15 min incubation time. Right: 60 min incubation time. Values express percentage of amounts 5-HTP and 5-FTP used for incubation. Mean ± SEM of 3 experiments. Chemistry 5-trimethylstannylindole was obtained from its corresponding bromoindole in a yield of 15 1 %. H-NMR analysis showed shifted signals for protons H2-H7 and a characteristic signal for the protons of the trimethylstannylgroup appearing in the product only. 13 C-NMR analysis showed a characteristic signal for the primary carbon atoms of the trimethylstannylgroup. Reaction of F 2 gas with 5-trimethylstannylindole gave 15 % yield of 5-fluoroindole. Optimal conditions for the synthesis of 5-FTP using the reversed tryptophanase reaction were determined by using 1.35 mg 5-fluoroindole dissolved in TRIS 0.1 M/ethanol (75/25) and addition of 400 µl AMS 1.5 M and 20 µl pyruvic acid. The effect on the ph on the 5-FTP yield was significant (figure 4). Maximal yields were found at To estimate the yield, ph was adjusted to and 50 µl TRP and 10 µl PLP were added and the whole mixture was incubated for 10 min at 40 C. Under these conditions, yields of 73 ± 6 % were obtained after 5 experiments. The reversed 92

108 5-Fluorotryptophan as potential PET tracer for neuroendocrine tumors tryptophanase reaction was performed on 5-fluoroindole giving 5-FTP confirmed by mass spectroscopy of a HPLC sample resulting in a mass of (mass 5-FTP +1: 223.1) rcentage Yield in pe ph Figure 4. Enzymatic synthesis of 5-fluorotryptophan depending on ph. Optimal conditions were determined: 1.35 mg 5- fluoroindole in 1.5 ml TRIS 0.1 M/ethanol (75/25), 400 µl AMS 1.5 M, 20 µl pyruvic acid, 50 µl TRP, 10 µl PLP 10.7 mm, ph , 40 C and 10 min reaction time. Yield after 5 experiments: 73 ± 6 %. Discussion The aim of this study was to develop an approach for the synthesis of the tracer [ 18 F]-5-FTP for the detection of neuroendocrine tumors with PET and to determine if 5-FTP accumulates in a BON cell line according to the APUD principle. We showed that 5-FTP is taken up into the neuroendocrine pancreatic tumor cell line BON while cellular levels of 5-FTP were lower than 5-HTP levels. It would have been even more informative if in addition 5-fluorotryptamine levels, the cellular metabolic intermediateproduct, were measured (figure 1). However as currently no suitable method is available, we were not able to identify 5-fluorotryptamine and 5-fluoroindole acetic acid in cell lysates. A reason for this could be the fact that fluorination of indoles probably leads to different retention times with our HPLC analysis. Detection of 5-FTP proved to be possible in this setting, as shown. Quantification of the compounds 5-fluorotryptamine and 5-fluoroindole acetic acid is necessary to obtain more insight in the metabolic fate and cellular handling of 5-FTP. It can be envisaged that 5-FTP 93

109 Chapter 6 is a substrate for AADC and that 5-fluorotryptamine is analogous to serotonin, stored in secretory granules 21,22. The exposure of BON cells to specific AADC and MAO inhibitors led to an increase of cellular 5-FTP levels. From these data it can be concluded that uptake can be manipulated, but reasons for this increase could be further analyzed when also other metabolically related compounds are quantified in future experiments. A profiling method for the detection of 5- fluorotryptamine and other metabolites of 5-FTP can give more insights in intracellular accumulation. A potential synthesis route to obtain 5-FTP is an enzymatic synthesis analogous to [ 11 C]-5- HTP. Tryptophanase activity is dependent on ph and temperature. Because of its fluorine atom, 5-fluoroindole has different physiological and charge properties than indole or 5- hydroxyindole. Therefore we determined if 5-fluoroindole was a good substrate for tryptophanase. High yields in the reversed tryptophanase reaction on 5-fluoroindole indicate that the fluorine atom was not detrimental to the enzymatic synthesis of 5-FTP from 5-fluoroindole. ph is the key factor in the enzymatic synthesis and if adjusted precisely, 5-FTP can be obtained in high and reliable yields. As a prerequisite for this approach we investigated the synthesis of 5-fluoroindole by electrophilic fluorodestannylation. Major challenge is the separation of fluoroindole from its stannylated precursor, which is used in amounts of mg. Simple solid phase extraction of precursor and product was not satisfying as both precursor and 5-fluoroindole were not sufficiently trapped on the extraction column. HPLC proved to be satisfactory. However, as described above, high amounts of precursor are used and therefore only small volumes of the reaction mixture have been injected on HPLC to avoid column capacity overload. We performed the enzymatic tryptophanase reaction with HPLC purified 5- fluoroindole giving 5-FTP, confirmed by mass spectroscopy. 18 F-labeling work towards [ 18 F]-5-FTP applying a similar approach is in progress. Conclusion 5-FTP is accumulated in a NET cell line. 5-Fluoroindole is a good substrate for the enzymatic synthesis towards 5-FTP. 5-Trimethylstannylindole has been developed as useful precursor for the enzymatic synthesis towards 18 F labeled 5-FTP via 5-fluoroindole. Acknowledgement BON cells were kindly provided by Dr. Ahlman from the Department of Surgery, Sahlgrenska University Hospital, Gothenburg, Sweden. We thank E.E.A. Venekamp and J. Krijnen for assistance in analyzing 5-HTP and 5-FTP samples. This work was supported by grant of the Dutch Cancer Society. References 1. Neels OC, Jager PL, Koopmans KP, Eriks E, de Vries EGE, Kema IP, Elsinga PH. Development of a reliable remote-controlled synthesis of β-[ 11 C]-5-hydroxy-Ltryptophan on a Zymark robotic system. J Label Compd Radiopharm 2006; 49:

110 5-Fluorotryptophan as potential PET tracer for neuroendocrine tumors 2. Vallabhajosula S, Kothari PJ, Suehiro M, Lampiri E, Goldsmith SJ. Enzymatic synthesis of 5-Hydroxy-L-( 11 C)tryptophan (5-HTP). J Label Compd Radiopharm 2007; 50: 177 (abstract). 3. Örlefors H, Sundin A, Lu L, Öberg K, Långström B, Eriksson B, Bergström M. Carbidopa pretreatment improves image interpretation and visualisation of carcinoid tumours with 11 C-5-hydroxytryptophan positron emission tomography. Eur J Nucl Med Mol Imaging 2006; 33: Örlefors H, Sundin A, Garske U, Juhlin C, Öberg K, Långström B, Bergström M, Eriksson B. Whole-body 11 C-5-hydroxytryptophan positron emission tomography as a universal imaging technique for neuroendocrine tumors comparison with somatostatin receptor scintigraphy and computed tomography. J Clin Endocrinol Metab 2005; 90: Sundin A, Sörensen J, Örlefors H, Eriksson B, Bergström M, Fasth KJ, Långström B. Whole-body PET with [ 11 C]-5-hydroxytryptophan for localization of neuroendocrine tumors. Clin Positron Imaging 1999; 2: 338 (abstract). 6. Örlefors H, Sundin A, Ahlström H, Bjurling P, Bergström M, Lilja A, Långström B, Öberg K, Eriksson B. Positron emission tomography with 5-hydroxytryptophan in neuroendocrine tumors. J Clin Oncol 1998; 16: Sundin A, Garske U, Örlefors H. Nuclear imaging of neuroendocrine tumors. Best Pract Res Clin Endocrinol Metab 2007; 21: Laverman P, Boerman OC, Corstens FHM, Oyen WJG. Fluorinated amino acids for tumour imaging with positron emission tomography. Eur J Nucl Med 2002; 29: Pearse AG. The APUD cell concept and its implications in pathology. Pathol Annu 1974; 9: VanBrocklin HF, Blagoev M, Hoepping A, O Neil JP, Klose M, Schubiger PA, Ametamey S. A new precursor for the preparation of 6-[ 18 F]fluoro-L-m-tyrosine ([ 18 F]FMT): efficient synthesis and comparison of radiolabelling. Appl Radiat Isot 2004; 61: Chirakal R, Schrobilgen GJ, Firnau G, Garnett S. Synthesis of 18 F labelled fluoro- fluoro-m-tyramine and fluoro-3-hydroxyphenylacetic acid. Appl m-tyrosine, Radiat Isot 1991; 42: De Vries EFJ, Luurtsema G, Brussermann M, Elsinga PH, Vaalburg W. Fully automated synthesis module for the high yield one-pot preparation of 6- [ 18 F]fluoro-L-DOPA. Appl Radiat Isot 1999; 51: Dollé F, Demphel S, Hinnen F, Fournier D, Vaufrey F, Crouzel C. 6-[ 18 F]Fluoro- L-DOPA by radiofluorodestannylation: A short and simple synthesis of a new labelling precursor. J Label Compd Radiopharm 1998; 41: Füchtner F, Steinbach J. Efficient synthesis of the 18 F-labelled 3-O-methyl-6- [ 18 F]fluoro-L-DOPA. Appl Radiat Isot 2003; 58: Watanabe T, Snell EE. Reversibility of the tryptophanase reaction: synthesis of tryptophan from indole, pyruvate and ammonia. Proc Nat Acad Sci 1972; 69:

111 Chapter Azizian H, Eaborn C, Pidcock A. Synthesis of organotrialkylstannylalkanes. The reaction between organic halides and hexaalkyldistannanes in the presence of palladium complexes. J Organometall Chem 1981; 215: Shotwell MA, Jayme DW, Kilberg MS, Oxender DL. Neutral amino acid transport systems in chinese hamster ovary cells. J Biol Chem 1981; 256: Jager PL, de Vries EGE, Piers DA, Timmer-Bosscha H. Uptake mechanisms of L- 3-[ 125 I]iodo-alpha-methyl-tyrosine in a human small-cell lung cancer cell line: 14 comparison with L-1-[ C]tyrosine. Nucl Med Comm 2001; 22: Evers BM, Townsend CM, Upp JR, Allen E, Hurlbut SC, Kim SW, Rajamaran S, Singh P, Reubi JC, Thompson JC. Establishment and characterization of a human carcinoid in nude mice and effect of various agents on tumor growth. Gastroenterology 1991; 101: Kema IP, Meijer WG, Meiborg G, Ooms B, Willemse PHB, de Vries EGE. Profiling of tryptophan-related plasma indoles in patients with carcinoid tumors by automated, on-line, solid-phase extraction and HPLC with fluorescence detection. Clin Chem 2001; 47: Brunk B, Blex C, Rachakonda S, Höltje M, Winter S, Pahner I, Walther DJ, Ahnert-Hilger G. The first luminal domain of vesicular monoamine transporters mediates G-protein-dependent regulation of transmitter uptake. J Biol Chem 2006; 281: Höltje M, Winter S, Walther D, Pahner I, Hörtnagl H, Ottersen OP, Bader M, Ahnert-Hilger G. The vesicular monoamine content regulates VMAT2 activity through Gα q in mouse platelets. J Biol Chem 2003; 278:

112 97

113 98

114 Summary and future perspectives 99

115 Summary and future perspectives Neuroendocrine tumors are slowly growing tumors which originate from neuroendocrine cells. These tumors can secrete several products. In case of overproduction of serotonin, symptoms such as flushing, diarrhea and right-sided heart disease can occur. Next to serotonin, other well known products are e.g. catecholamines. Amine precursors such as levodopa and 5-hydroxytryptophan are taken up into neuroendocrine tumor cells by large amino acid transporters (LAT) and are decarboxylated by amino acid decarboxylase (AADC) to serotonin and dopamine. These amines are stored in cellular vesicles via the vesicular monoamine transporter VMAT. After release the amines are metabolized by the enzyme monoamine oxidase (MAO) to respectively 5- hydroxyindole acetic acid and homovanillic acid and subsequently excreted. Staging of neuroendocrine tumors is required for optimal treatment decisions. Because of their slow growth and frequently the induction of non-specific symptoms, these tumors are often metastasized at diagnosis. Different imaging methods for detection of neuroendocrine tumors are currently available. Routine imaging techniques consist of morphological techniques such as computed tomography (CT) or magnetic resonance imaging (MRI) and functional imaging such as somatostatin receptor scintigraphy (SRS). Besides SRS, in neuroendocrine tumors positron emission tomography (PET) is another functional imaging technique which can be useful to improve tumor image quality and sensitivity of tumor detection. Two new interesting PET tracers are available. The PET tracers L-6-[ 18 F]fluorolevodopa ([ 18 F]FDOPA) and L-[ 11 C]-5-hydroxytrytophan ([ 11 C]HTP) are of special interest as they take actively part in the biochemical pathways of neuroendocrine tumors. Therefore, the aim of this thesis was to study the development, biochemical behavior and diagnostic value of new PET tracers for imaging neuroendocrine tumors. In chapter 1 a literature review is given of the techniques currently used for staging of neuroendocrine tumors in nuclear medicine. First, an overview of the mechanisms involved in tracer uptake is presented. The different morphological and functional detection methods are described and results obtained in the last decade are shown. The sensitivities of the various tracers in several subtypes of tumors have been compared in Forest plots. This comparison showed, that the metabolic PET tracers [ 18 F]FDOPA and [ 11 C]HTP perform with higher sensitivity than the currently used standard SRS. These studies with [ 11 C]HTP have been described by the Uppsala group exclusively, as they have been the only PET center worldwide with the capability to produce this tracer. Given their interesting results we decided to investigate the production of [ 11 C]HTP on a Zymark robotic system. The synthesis described in chapter 2 was started by the production of [ 11 C]methyl iodide and labeling of the precursor N-(diphenylmethylene)glycine tert-butyl ester. Subsequent hydrolysis gave racemic [ 11 C]alanine. [ 11 C]HTP was obtained from [ 11 C]alanine in a one-pot synthesis using 4 different enzymes. Average radiochemical yield after HPLC purification was 15 ± 12 % from the time of release of [ 11 C]methyl iodide. Radiation exposure for the radiochemist was reduced to a minimum of 260 µsv for the skin and 40 µsv for the whole body. Currently, [ 11 C]HTP is produced reliably in doses of 400 MBq which is sufficient for patient studies. Although [ 18 F]FDOPA and [ 11 C]HTP showed interesting results in humans, still little is known about the processes involved in accumulation of these tracers by neuroendocrine tumor cells. In chapter 3 we therefore evaluated the tracer uptake via LAT transporters, the 100

116 Summary and future perspectives influence of the decarboxylase inhibitor carbidopa and of the MAO inhibitors clorgyline and pargyline on tracer accumulation. The effect of carbidopa in vivo on metabolism of [ 18 F]FDOPA and [ 11 C]HTP in small animals was studied with a micropet camera. The cellular tra nsport of both PET tracers in the neuroendocrine BON tumor cell line was inhibited by amino-2-norbornanecarboxylic acid and resulted in low IC 50 values ([ 18 F]FDOPA: 0.01 mm; [ 11 C]HTP: 0.12 mm) after 15 minutes tracer incubation. Inhibition of MAO by clorgyline led to a significant increase in tracer accumulation compared to control ([ 18 F]FDOPA: P=0.02; [ 11 C]HTP: P=0.02) in vitro after 60 minutes tracer incubation. We showed that carbidopa did not influence accumulation in tumor cells in vitro, but did increase uptake in tumor bearing mice. This increased uptake most likely is the result of inhibition of decarboxylation in peripheral organs which results in better availability of tracer for accumulation in tumor cells. Standardized uptake values (SUVs) of [ 18 F]FDOPA were superior to [ 11 C]HTP 60 minutes after intravenous injection in a neuroendocrine pancreatic tumor animal model (P=0.03). In small animals, neuroendocrine tumors from human origin with weights of less than 20 mg were still clearly visualized with both PET tracers. In chapter 4 the diagnostic value of [ 18 F]FDOPA for the detection of patients with carcinoid disease was studied. 53 patients underwent a [ 18 F]FDOPA PET scan which was compared with the standard imaging methods such as SRS and CT. In a patient based analysis [ 18 F]FDOPA had a sensitivity of 100% (95%CI ), SRS of 93% (95%CI 82-98), CT of 87% (95%CI 75-95) and the combination of SRS and CT 96% (95%CI ) (P=0.45 for PET versus SRS with CT). However, [ 18 F]FDOPA alone detected many more lesions, more positive regions and more lesions per region than SRS combined with CT. In regional analysis sensitivity of [ 18 F]FDOPA was 95% (95%CI 90-98) versus 66% (95%CI 57-74) for SRS alone, 57% (95%CI 48-66) for CT and 79% (95%CI 70-86) for SRS with CT (P=0.0001, PET versus SRS with CT). In individual lesion analysis, sensitivities were 96% (95%CI 95-98), 46% (95%CI 43-50), 54% (95%CI 51-58) and 65% (95%CI 62-69) for PET, SRS, CT and SRS with CT respectively (P< for PET versus SRS with CT). The results of this study showed that PET imaging with [ 18 F]FDOPA is superior for staging carcinoid tumor lesions compared to SRS and CT. In chapter 5, a study is described analyzing the PET tracers [ 18 F]FDOPA and [ 11 C]HTP. As there were no head to head comparisons, it was until now not clear from the literature which PET tracer was superior in which setting. 24 patients with a carcinoid tumor and 23 patients with an islet cell tumor were included in the study. Patients had to have, before performing the PET scans, at least one lesion based on clinical, histological and/ or biochemical findings and detected one abnormal lesion on SRS, CT or MRI. All patients underwent [ 11 C]HTP and [ 18 F]FDOPA PET scans and additional SRS and CT scan. [ 18 F]FDOPA and [ 11 C]HTP showed more carcinoid tumor positive regions than SRS as [ 18 F]FDOPA sensitivity was 81%, and [ 11 C]HTP sensitivity was 77% (P=0.001 and for comparison with SRS respectively). Sensitivity of SRS was 58% and CT 70% (P=ns for the comparison of both PET scans with CT). In islet cell tumors [ 11 C]HTP detected more tumor positive regions (sensitivity 70%) than [ 18 F]FDOPA (sensitivity 44%, P= for the comparison with [ 18 F]FDOPA). In carcinoid patients, both [ 11 C]-5-HTP and [ 18 F]FDOPA revealed more tumor positive lesions (sensitivity 78% and 87%) than SRS (sensitivity 49%, P < for the comparison of both [ 11 C]HTP and [ 18 F]FDOPA with 101

117 SRS) in carcinoid patients. These results show that [ 18 F]FDOPA performs best in carcinoid tumors and is superior to SRS and CT. Imaging of islet cell tumors gave best results with [ 11 C]HTP. This tracer was superior to SRS and [ 18 F]FDOPA while CT gave similar results as [ 11 C]HTP. For staging both carcinoid and islet cell tumors, these PET scans are, in combination with CT, a promising tool for anatomical localization of the tumor. [ 11 C]HTP shows interesting results for the detection of neuroendocrine tumors and in particular islet cell tumors. However, only a few centers are equipped with a cyclotron and are therefore capable to produce tracers with short half-lives. If a fluorine-18 labeled tryptophan analogon could be produced with a half-life of 110 minutes this could be transported to other hospitals. It is therefore an attractive alternative for [ 11 C]HTP. We investigated a synthesis route of 5-fluorotryptophan by combining reliable methods used in the synthesis of [ 11 C]HTP and [ 18 F]FDOPA (chapter 6). We demonstrated that non-labeled 5-fluorotryptophan accumulates in neuroendocrine tumor cells. Electrophilic fluorodestannylation, a method used in the synthesis of [ 18 F]FDOPA, was chosen to obtain 5-fluoroindole with yields of 15 % from the newly developed precursor 5- trimethylstannylindole. The reversed tryptophanase reaction was used for the synthesis of [ 11 C]HTP. We optimized the synthesis of 5-fluorotryptophan from 5-fluoroindole so far, that 5-fluorotryptophan can now be obtained in reliable yields over 70 % using the reversed tryptophanase reaction. Labeling work is in progress to synthesize [ 18 F]-5-fluorotryptophan from 5-trimethylstannylindole. If successfully developed, [ 18 F]-5-fluorotryptophan might well be a feasible tool for staging neuroendocrine tumors. Future perspectives In this thesis it is shown that [ 18 F]FDOPA and [ 11 C]HTP are relevant PET tracers for the visualization of neuroendocrine tumors (chapter 3, 4 and 5). Tracer accumulation in neuroendocrine tumor cells results from the sum of tracer uptake by LAT, decarboxylation AADC and subsequent storage of amines in granules (VMAT). We showed that carbidopa, administered to patients before the imaging to increase tracer uptake, does not influence decarboxylation of tracer in tumor cells. Carbidopa exerts its effect by inhibition of tracer decarboxylation in peripheral organs and therefore more tracer can be taken up by the neuroendocrine tumor. Inhibition of MAO increased tracer accumulation in vitro (chapter 3). MAO was overexpressed in the neuroendocrine BON tumor cell line used in our studies. If inhibited, it resulted in high intracellular levels of serotonin or catecholamines. Further preclinical research is required to evaluate whether MAO inhibition should be considered in the clinic to further improve image quality and sensitivity of neuroendocrine tumor detection. In our studies metabolic tracers [ 18 F]FDOPA and [ 11 C]HTP PET are superior to SRS or morphological techniques such as CT and MRI. However, SRS is used with a gamma camera with lower resolution than PET cameras. 68 Ga-octreotide analogues are in development and can be used to image receptor expression in neuroendocrine tumors with PET. To know which technique is superior, a head to head comparison with metabolic tracers such as [ 18 F]FDOPA and [ C]HTP is required. [ C]HTP showed good results for visulalization of neuroendocrine tumors but because of the short tracer half-life the scans have to be started directly after injection and the tracer can be administered to one patient per synthesis only. Tryptophan labeled with 18 F would not only allow longer distribution of 102

118 Summary and future perspectives tracer but it would also give the possibility to scan for a longer period and would also make the tracer available for two or more patients per synthesis. We showed that 5- fluorotryptophan has favorable characteristics to enter human neuroendocrine tumor cells. A profiling method for metabolites such as 5-fluorotryptamine and 5-fluoroindole acetic acid could show the total accumulation of the tracer in tumor cells and give more insights in the metabolism and biochemical behavior of 5-fluorotryptophan in neuroendocrine tumors. To complete the synthesis towards [ 18 F]-5-fluorotryptophan, labeling of 5- trimethylstannylindole using [ 18 F]F 2 gas should be investigated. Once established, the value of [ 18 F]-5-fluorotryptophan for imaging neuroendocrine tumors can be determined in vitro and in a small animal model to get a better understanding of tracer metabolism and accumulation. As [ 18 F]-5-fluorotryptophan contains a fluorine atom further studies with regard to biochemical behavior in organisms are necessary. Other isomers of fluorotryptophan such as 4-,6- or 7-fluorotryptophan may show also affinity for neuroendocrine tumors and therefore, labeled with 18 F, they can be attractive tracers in this setting yielding comparable results as [ 18 F]FDOPA and [ 11 C]HTP. 103

119 104

120 Samenvatting / Zusammenfassung 105

121 Samenvatting Neuroendocriene tumoren zijn meestal langzaam groeiende tumoren die afkomstig zijn van neuroendocriene cellen. Deze tumoren kunnen verschillende producten uitscheiden waaronder serotonine. De overproductie van serotonine kan o.a. flushes, diarree en hartziekte veroorzaken. Andere uitscheidingsproducten zijn catecholamines. Omdat de tumoren langzaam groeien en de symptomen wisselend verschijnen zijn deze tumoren vaak al uitgezaaid als de diagnose gesteld wordt. Voor een optimale behandeling is het nodig uitzaaiingen in kaart te brengen, het zogenaamde stagerende onderzoek. Hiervoor zijn verschillende methoden beschikbaar. Naast standaardtechnieken, die de anatomie laten zien zoals computertomografie (CT) en magnetic resonance imaging (MRI) bestaan ook afbeeldingsmethoden, die een bepaalde eigenschap in het weefsel laten zien (functioneel) zoals de somatostatine receptor met somatostatine receptor scintigrafie (SRS). Positron emissie tomografie (PET) is een tweede nucleaire methode naast SRS om neuroendocriene tumoren te visualiseren. PET zou de beeldkwaliteit en de sensitiviteit van de tumordetectie kunnen verbeteren ten opzichte van SRS. Twee nieuwe PET tracers om toe te passen bij dit tumor type zijn sinds kort beschikbaar. Deze PET tracers zijn L-6-[ 18 F]fluoro-levodopa ([ 18 F]FDOPA) en L-[ 11 C]-5-hydroxytryptophan ([ 11 C]HTP). Zij maken beide een actief onderdeel uit van het metabolisme in neuroendocriene tumoren. Amine precursors zoals levodopa en 5-hydroxytryptofaan worden in neuroendocriene tumorcellen door zogenaamde large amino acid transporters (LAT) opgenomen en door amino acid decarboxylase (AADC) gedecarboxyleerd naar serotonine en dopamine. Deze amines worden in cellulaire vesicles opgeslagen. Het enzym monoamine oxidase (MAO) zorgt ervoor dat na vrijkomen van deze amines omzetting plaatsvindt na 5- hydroxyindoolazijnzuur en homovanillinezuur met vervolgens uitscheiding via de nieren. Het doel van dit proefschrift is het gedrag en de diagnostische waarde van nieuwe ontwikkelde PET tracers voor het afbeelden van neuroendocriene tumoren te bestuderen. In hoofdstuk 1 wordt een literatuuroverzicht gegeven van de huidige technieken die in de nucleaire geneeskunde gebruikt worden om neuroendocriene tumoren op te sporen. De cellulaire opnamemechanismen van traceropname worden gepresenteerd. Verschillende morfologische en functionele detectie methodes worden beschreven en de resultaten met deze technieken in patiënten met neuroendocriene tumoren gedurende de laatste tien jaar worden gepresenteerd. De metabole PET tracers [ 18 F]FDOPA en [ 11 C]HTP laten een hogere sensitiviteit (gevoeligheid) zien dan de gebruikelijke standaard SRS. Studies met [ 11 C]HTP zijn alleen beschreven door een instituut uit Uppsala in Zweden. Om klinische studies te kunnen uitvoeren hebben wij de synthese van [ 11 C]HTP met behulp van een Zymark robot systeem opgezet. Bereiding op deze wijze geeft een lage stralingsdosis voor de radiochemicus. De opzet van deze synthese wordt beschreven in hoofdstuk 2 en begint met de productie van [ 11 C]methyliodide en de koppeling aan de precursor N- (diphenylmethyleen)glycine tert-butyl ester. Hydrolyse levert [ 11 C]alanine op dat in een zogenaamde eenpotsynthese middels 4 verschillende enzymen omgezet wordt naar [ 11 C]HTP. De radiochemische opbrengsten van [ 11 C]HTP zijn 15 ± 12 % berekend vanaf het inleiden van [ 11 C]methyliodide. De stralenbelasting voor de radiochemicus werd gereduceerd naar een minimum van 260 µsv voor de huid en tot 40 µsv voor het gehele lichaam. Tegenwoordig kan [ 11 C]HTP op betrouwbare wijze met doses van 400 MBq geproduceerd worden. Dit is voldoende voor een patiëntenscan. 106

122 Samenvatting / Zusammenfassung Hoewel [ 18 F]FDOPA en [ 11 C]HTP neuroendocriene tumoren in patiënten goed kunnen afbeelden is nog verrassend weinig bekend over de processen die de opname van deze tracers in deze tumoren bepalen. In hoofdstuk 3 worden de tracer opname via LAT transporters en de invloed van de decarboxylaseremmer carbidopa en de MAO remmers clorgyline en pargyline geëvalueerd. De effecten van carbidopa op het metabolisme van [ 18 F]FDOPA en [ 11 C]HTP werden in vivo in proefdieren bestudeerd met behulp van een micropet camera. Het transport van beide PET tracers in de neuroendocriene tumorcellijn BON werd zeer goed geblokkeerd door amino-2-norbornaancarbonzuur (helft maximale remmende concentratie (IC 50 ) [ 18 F]FDOPA: 0,01 mm; [ 11 C]HTP: 0,12 mm). Remming van MAO door clorgyline leidt tot een toegenomen traceropname vergeleken met controlewaarden ([ 18 F]FDOPA: P=0,02; [ 11 C]HTP: P=0,02) in vitro na 60 minuten tracerincubatie. Het toedienen van carbidopa had geen invloed op de opname van tracer in tumorcellen in vitro, maar zorgde voor een verhoogde opname in tumoren van proefdieren. Deze verhoogde opname is mogelijk het resultaat van remming van het enzym AADC in perifere organen wat ervoor zorgt dat meer tracer beschikbaar gemaakt kan worden voor opname in tumorcellen. Traceropnames zijn 60 minuten na toediening hoger voor [ 18 F]FDOPA dan voor [ 11 C]HTP in een neuroendocriene pancreastumor diermodel (P=0,03). Humane neuroendocriene tumoren lichter dan 20 mg konden in proefdieren nog met beide PET tracers duidelijk zichtbaar gemaakt worden. Hoofdstuk 4 beschrijft de diagnostische waarde van [ 18 F]FDOPA bij patiënten met een gemetastaseerd carcinoid. 53 patiënten hebben een [ 18 F]FDOPA PET scan ondergaan. Deze werd vergeleken met standaarddiagnostiek zoals SRS en CT. In analyse op patiënt niveau had [ 18 F]FDOPA een sensitiviteit van 100% (95% confidence interval (CI) ), SRS van 93% (95%CI 82-98), CT van 87% (95%CI 75-95) en de combinatie van SRS en CT 96% (95%CI ) (P=0,45 voor PET vergeleken met SRS en CT). [ 18 F]FDOPA detecteerde meer laesies, meer positieve regio s en meer laesies per regio dan SRS gecombineerd met CT. Per regio was de sensitiviteit van [ 18 F]FDOPA 95% (95%CI 90-98) versus 66% (95%CI 57-74) voor SRS, 57% (95%CI 48-66) voor CT en 79% (95%CI 70-86) voor SRS en CT (P=0,0001, PET versus SRS en CT). In analyse op laesie niveau was de sensitiviteit 96% (95%CI 95-98), 46% (95%CI 43-50), 54% (95%CI 51-58) en 65% (95%CI 62-69) voor respectievelijk PET, SRS, CT en SRS met CT (P<0,0001 voor PET versus SRS en CT). De resultaten van deze studies lieten zien dat PET imaging met [ 18 F]FDOPA beter is voor het zichtbaar maken van carcinoidhaarden dan SRS en CT. Hoofdstuk 5 beschrijft een vergelijkende studie tussen de PET tracers [ 18 F]FDOPA en [ 11 C]HTP. 24 patiënten met een carcinoidtumor en 23 patiënten met een eilandjesceltumor werden bestudeerd. Alle patiënten ondergingen ook een SRS en CT scan. Er werden meer haarden gevonden met [ 18 F]FDOPA en [ 11 C]HTP dan met SRS. De sensitiviteit van [ 18 F]FDOPA was 81%, en de sensitiviteit van [ 11 C]HTP was 77% (respectievelijk P=0,001 en 0,004 vergeleken met SRS). De sensitiviteit van SRS was 58% en van CT 70% (P=ns voor de vergelijking van beide PET scans met CT). In eilandjesceltumoren kon [ 11 C]HTP meer tumor positieve regio s (sensitiviteit 70%) detecteren dan [ 18 F]FDOPA (sensitiviteit 44%, P=0,0001 in vergelijking met [ 18 F]FDOPA). In carcinoidpatiënten lieten zowel [ 11 C]-5-HTP en [ 18 F]FDOPA meer tumor positieve laesies zien (sensitiviteit 78% en 87%) dan SRS (sensitiviteit 49%, P < 0,001 voor de vergelijking van zowel [ 11 C]HTP en [ 18 F]FDOPA met SRS) in carcinoidpatiënten. Deze 107

123 resultaten laten zien dat [ 18 F]FDOPA de beste tracer is voor het afbeelden van carcinoide laesies en dat het superieur is aan SRS en CT. Afbeelden van eilandjesceltumoren leverde de beste resultaten met [ 11 C]HTP. Deze tracer was beter dan SRS en [ 18 F]FDOPA maar CT leverde vergelijkbare resultaten op als [ 11 C]HTP. Voor het afbeelden van carcinoiden en eilandjesceltumoren zijn deze PET scans in combinatie met CT voor de anatomische lokalisatie van de tumor veelbelovend. Voor de detectie van neuroendocriene tumoren en in het bijzonder voor eilandjesceltumoren geeft [ 11 C]HTP interessante resultaten. Weinig centra bezitten een cyclotron en kunnen derhalve geen tracers met korte halfwaardetijden produceren. Als een met fluor-18 gelabelde tryptofaanvariant met een halfwaardetijd van 110 minuten gesynthetiseerd zou kunnen worden, zou deze tracer ook naar andere ziekenhuizen vervoerd kunnen worden. Dit zou daarom een attractief alternatief kunnen zijn voor [ 11 C]HTP. Wij hebben een nieuwe route voor de synthese van 5-fluorotryptofaan ontworpen door betrouwbare methodes uit de syntheses van [ 11 C]HTP en [ 18 F]FDOPA te combineren (hoofdstuk 6). Wij lieten zien dat niet gelabelde 5-fluorotryptofaan in neuroendocriene tumorcellen wordt opgenomen. De electrofiele fluorodestannylering, een methode die gebruikt wordt voor de synthese van [ 18 F]FDOPA, werd gekozen om 5- fluoroindool met opbrengsten van 15% vanuit de nieuw ontwikkelde precursor 5- trimethylstannylindool te maken. De reversed tryptophanase reactie is een onderdeel van de synthese tot [ 11 C]HTP. We hebben de synthese van 5-fluorotryptofaan uitgaande van 5- fluoroindole zover geoptimaliseerd dat 5-fluorotryptofaan nu in betrouwbare opbrengsten van 70 % middels deze reversed tryptophanase reactie gemaakt kan worden. Bij een succesvolle ontwikkeling zou [ 18 F]-5-fluorotryptofaan een goede tracer kunnen zijn voor het opsporen van neuroendocriene tumoren. Zusammenfassung Neuroendokrine Tumore sind meistens langsam wachsende Tumore. Sie stammen ab von neuroendokrinen Zellen. Diese Tumore können u.a. Serotonin und Katecholamine ausscheiden. Eine Überproduktion von Serotonin kann u.a. zu Symptomen wie Hitzewallungen, Durchfall und Herzbeschwerden führen. Für eine optimale Behandlungsmethode von neuroendokrinen Tumoren ist eine Stadienbestimmung (Staging) erforderlich. Wegen des langsamen Wachstums und der unregelmässig auftretenden Symptomen sind diese Tumore zum Zeitpunkt der Diagnose bereits metastasiert. Zur Detektion von neuroendokrinen Tumoren sind derzeit verschiedene bildgebende Verfahren (Imaging) verfügbar. Neben den morphologischen Methoden der Computertomographie (CT) und der Magnetresonanztomographie (MRT) ist auch die Somatostatin Rezeptor Szintigraphie (SRS) zum Abbilden von Somatostatin Rezeptoren gebräuchlich. Neben der SRS ist die Positronen-Emissions-Tomographie (PET) eine geeignete funktionale Methode um die Bildqualität und Sensitivität der Tumordetektion zu verbessern. Gegenwärtig sind zwei neue PET-Tracer verfügbar. Von speziellem Interesse sind die PET-Tracer L-6-[ 18 F]fluoro-levodopa ([ 18 F]FDOPA) und L- [ 11 C]-5-hydroxytrytophan ([ 11 C]HTP), da beide aktiv an den biochemischen Stoffwechseln der neuroendokrinen Tumore beteiligt sind. Prekursoren dieser Amine, wie Levodopa oder 5-Hydroxytryptophan, werden über so genannte Large Amino acid Transporter (LAT) in die neuroendokrinen Tumorzellen aufgenommen und anschließend durch das Enzym 108

124 Samenvatting / Zusammenfassung Amino Acid Decarboxylase (AADC) zu Serotonin bzw. Dopamin umgesetzt. Die Amine werden mittels vesikulärer Monoamine Transportern (VMAT) in den zellulären Vesikeln gespeichert. Nachdem die Monoamine die Zellen verlassen haben, erfolgt eine Metabolisierung zu 5-Hydroxyindolessigsäure bzw. Homovanillinsäure durch das Enzym Monoamine Oxidase und abschliessender Exkretion. In dieser Doktorarbeit werden die Entwicklung, das biochemische Verhalten und der diagnostische Wert neuer PET Tracer zum Abbilden neuroendokriner Tumore beschrieben. Kapitel 1 beschreibt die Techniken zum Aufspüren von neuroendokrinen Tumoren wie sie heutzutage in der Nuklearmedizin angewandt werden. Die vorherrschenden Mechanismen in der Traceraufnahme werden in einer Übersicht präsentiert. Verschiedene morphologische und funktionale Detektionsmethoden und ihre Resultate der letzten Dekade werden beschrieben. Die Sensitivität der verschiedenen Tracer in unterschiedlichen Subtypen von Tumoren wird in Forest Plots verglichen. Dieser Vergleich zeigt, das die metabolischen PET Tracer [ 18 F]FDOPA und [ 11 C]HTP eine höhere Sensitivität aufweisen als der momentan gebräuchliche Standard SRS. Studien mit [ 11 C]HTP werden einzig von einer Arbeitsgruppe aus Uppsala beschrieben, welche weltweit das einzige PET Zentrum ist, das diesen Tracer synthetisieren kann. Aufgrund der interessanten Resultate haben wir uns entschieden die Produktion von [ 11 C]HTP auf einem Zymark Roboter System auszuführen. Kapitel 2 beschreibt die Synthese von [ 11 C]HTP und beginnt mit der Produktion von [ 11 C]Iodmethan und der Markierung des Precursors N-(diphenylmethylen)glycin tert-butyl ester. Die darauffolgende Hydrolyse resultiert im racemischen [ 11 C]Alanin. [ 11 C]HTP wird aus [ 11 C]Alanin in einer Eintopfsynthese mit vier verschiedenen Enzymen erhalten. Die durchschnittliche radiochemische Ausbeute nach HPLC Säuberung beträgt 15 ± 12 % nach Einleiten von [ 11 C]Iodmethan. Die Strahlenbelastung für den Radiochemiker konnte dabei auf ein Minimum von 260 µsv für die Haut und 40 µsv für den ganzen Körper reduziert werden. [ 11 C]HTP wird heute in zuverlässigen Dosen von 400 MBq produziert, was für Patientenstudien ausreichend ist. Obwohl [ 18 F]FDOPA und [ 11 C]HTP in humanen Studien interessante Resultate aufweisen, ist überraschend wenig über die Prozesse bekannt, die für die Akkumulation dieser Tracer in neuroendokrinen Tumorzellen verantwortlich sind. Kapitel 3 beschreibt die Traceraufnahme durch LAT Transporter und den Einfluss des Decarboxylase Inhibitors Carbidopa sowie der MAO Inhibitoren Clorgyline und Pargyline auf die Tracerakkumulation. Der Effekt von Carbidopa auf den Metabolismus von [ 18 F]FDOPA und [ 11 C]HTP in vivo in Kleintieren wurde mit einer micropet Kamera untersucht. Der Transport beider Tracer in die neuroendokrine Zelllinie BON wird durch Amino-2- norbornancarbonsäure unterdrückt und resultiert in niedrigen IC 50 (Konzentration eines Inhibitors um 50 % Blockade zu erzielen) Werten ([ 18 F]FDOPA: 0,01 mm; [ 11 C]HTP: 0,12 mm) nach 15 Minuten Tracerinkubationszeit. Die Inhibition von MAO durch Clorgyline führte zu einem signifikanten Anstieg in der Tracerakkumulation im direkten Vergleich mit Kontrollwerten ([ 18 F]FDOPA: P=0,02; [ 11 C]HTP: P=0,02) nach 60 Minuten Tracerinkubationszeit in vitro. Wir konnten zeigen, dass Carbidopa in vitro keinen Einfluss auf die Tracerakkumulation hat, jedoch eine erhöhte Traceraufnahme in Tumoren bei Mäusen festzustellen ist. Die Decarboxylierung in peripheren Organen ist blockiert und resultiert in einer besseren Verfügbarkeit der Tracer für die Akkumulation in Tumorzellen. 109

125 Standardisierte Aufnahmewerte (SUV) waren im Kleintiermodell einer neuroendokrinen Zelllinie nach 60 Minuten intravenöser Injektion für [ 18 F]FDOPA höher als für [ 11 C]HTP (P=0,03). In Kleintieren konnten neuroendokrine Tumore von humanem Ursprung mit einem Tumorgewicht von weniger als 20 mg mit beiden PET Tracer deutlich visualisiert werden. Kapitel 4 beschreibt den diagnostischen Nutzen von [ F]FDOPA für die Detektion von Patienten mit karzinoiden Tumoren. Bei 53 Patienten wurden [ 18 F]FDOPA PET Scans durchgeführt und mit Standard Imaging Methoden wie der SRS und der CT verglichen. In einer auf Patienten basierten Analyse erreichte [ 18 F]FDOPA eine Sensitivität von 100% (95%CI ), die SRS 93% (95%CI 82-98), die CT 87% (95%CI 75-95) und die Kombination aus SRS und CT 96% (95%CI ) (P=0,45 für PET versus SRS und CT). [ 18 F]FDOPA allein detektierte viel mehr Läsionen, mehr positive Regionen und mehr positive Läsionen pro Region als SRS kombiniert mit CT. In der regionalen Analyse erreichte [ 18 F]FDOPA eine Sensitivität von 95% (95%CI 90-98) gegenüber 66% (95%CI 57-74) für SRS allein, 57% (95%CI 48-66) für CT und 79% (95%CI 70-86) für SRS und CT kombiniert (P=0,0001, PET versus SRS und CT). In der individuellen Läsionenanalyse wurden Sensitivitäten von 96% (95%CI 95-98), 46% (95%CI 43-50), 54% (95%CI 51-58) und 65% (95%CI 62-69) für PET, SRS, CT und SRS kombiniert mit CT (P<0,0001 für PET gegenüber SRS und CT) ermittelt. Die Ergebnisse dieser Studie zeigen, dass die PET zum Aufzeigen karzinoider Tumorläsionen mit [ 18 F]FDOPA im Vergleich besser ist als SRS und CT. In Kapitel 5 wird eine vergleichende Studie zwischen den PET Tracern [ 18 F]FDOPA und [ 11 C]HTP beschrieben. Da bisher keine vergleichenden Studien durchgeführt wurden, ist bis heute unklar welcher PET Tracer in bestimmten Fällen besser einzusetzen ist. In der in dieser Arbeit beschriebenen Studie wurden 24 Patienten mit einem karzinoiden Tumor und 23 Patienten mit einem Inselzelltumor untersucht. Bevor die Patienten einem PET Scan unterzogen wurden, musste mindestens eine Läsion basierend auf klinischen, histologischen und/oder biochemischen Befunden erwiesen sein sowie eine abnormale Läsion mit Hilfe von SRS, CT oder MRT detektiert worden sein. Alle Patienten erhielten [ 11 C]HTP und [ 18 F]FDOPA PET Scans und zusätzliche SRS und CT Scans. [ 18 F]FDOPA und [ 11 C]HTP zeigen mehr positive Regionen für karzinoide Tumore als SRS mit einer Sensitivität von 81% für [ 18 F]FDOPA und einer Sensitivität von 77% für [ 11 C]HTP (P=0,001 und 0,004 im Vergleich mit SRS). Die Sensitivität für SRS betrug 58% und für CT 70% (P=ns für den Vergleich beider PET scans mit CT). In Inselzelltumoren wurden mehr positive Regionen durch [ 11 C]HTP detektiert (Sensitivität 70%) als für [ 18 F]FDOPA (Sensitivität 44%, P=0,0001 im Vergleich mit [ 18 F]FDOPA). Sowohl [ 11 C]-5- HTP als auch [ 18 F]FDOPA zeigten mehr positive Läsionen (Sensitivität 78% und 87%) als SRS (Sensitivität 49%, P < 0,001 im Vergleich von [ 11 C]HTP und [ 18 F]FDOPA und SRS) in Patienten mit karzinoidem Tumor. Diese Ergebnisse zeigen, dass [ 18 F]FDOPA am besten für karzinoide Tumore einsetzbar ist verglichen mit der SRS und der CT. Das Imaging von Inselzelltumoren lieferte die besten Resultate mit [ 11 C]HTP. Dieser Tracer war besser geeignet als SRS und [ 18 F]FDOPA, während CT ähnliche Resultate wie [ 11 C]HTP erzielen konnte. Für die Stadienbestimmung von sowohl karzinoiden Tumoren und Inselzelltumoren, sind diese PET Scans in Kombination mit der CT eine vielversprechende Methode für die anatomische Lokalsierung des Tumors

126 Samenvatting / Zusammenfassung [ 11 C]HTP liefert interessante Ergebnisse bei der Detektion von neuroendokrinen Tumoren, im Speziellen für Inselzelltumore. Jedoch sind nur wenige PET-Zentren mit einem Zyklotron ausgestattet und somit nicht in der Lage Tracer mit einer kurzen Halbwertszeit zu produzieren. Fluor-18 markiertes Tryptophan mit einer Halbwertszeit von 110 Minuten könnte in andere Krankenhäuser transportiert werden und wäre damit eine attraktive Alternative zu [ 11 C]HTP. Wir konnten aus bestehenden zuverlässigen Methoden der [ 11 C]HTP und [ 18 F]FDOPA Synthesen eine neuartige Syntheseroute für 5- Fluorotryptophan entwickeln (Kapitel 6). Wir haben gezeigt, dass 5-Fluorotryptophan durch neuroendokrine Tumorzellen aufgenommen wird. Die elektrophile Fluorodestannylierung wurde gewählt um 5-Fluoroindol in Ausbeuten von 15% vom neu entwickelten Prekursor 5-Trimethylstannylindol zu erhalten. Die umgekehrte Tryptophanasereaktion findet ihre Anwendung in der Synthese von [ 11 C]HTP. Wir haben die Synthese von 5-Fluorotryptophan ausgehend von 5-Fluoroindol soweit optimiert, dass 5-Fluorotryptophan durch die umgekehrte Tryptophanasereaktion in Ausbeuten von mehr als 70% erhalten wird. Die Markierung von 5-Trimethylstannylindole mit Fluor-18 ist in Arbeit um [ 18 F]-5-Fluorotryptophan als möglichen Tracer für das Aufspüren von neuroendokrinen Tumoren zu synthesieren. 111

127 112

128 Dankwoord 113

129 Dankwoord Dit boekje is tot stand gekomen door het interdisciplinaire samenwerken van diverse mensen met liefde voor hun vak. Mijn eerste copromotor, Dr. Philip Elsinga, wil ik bedanken voor het intensieve samenwerken. Beste Philip, het is mij een genoegen te zeggen dat ik onder jouw begeleiding vier leerrijke en plezante jaren heb gehad. Door jou ben ik een radiochemicus geworden en met enkele trots kan ik zeggen dat ik geen betere leermeester voor dit vak had kunnen hebben. Je deur stond altijd voor me open, letterlijk en figuurlijk. Jij was diegene die me erop wees dat ik weer eens een Germanisme had gebruikt. Dank jou weet ik ook het woord enzym goed uit te spreken. Uiteindelijk zou ik zonder jou nu niet aan de andere kant van de aardbal zijn. Ik wil mijn eerste promotor bedanken, Prof. dr. Elisabeth de Vries. Beste Liesbeth, je energie en je doorzettingsvermogen hebben altijd indruk op mij gemaakt. Je directheid liet me soms even schrikken. Maar als ik er goed ging over nadenken had je toch wel altijd gelijk. Je kon me altijd dingen uitleggen die voor mij onduidelijk waren, simpelweg omdat ik niet de medische achtergrond had. Je hebt in de afgelopen jaren ervoor gezorgd dat mijn medische horizon enorm verbreedt is. Mijn tweede promotor, Prof. dr. Pieter Jager, wil ik bedanken voor zijn enthousiasme met betrekking tot mijn onderzoek. Beste Piet, de eerste Megabecquerel van in Groningen gemaakte HTP maakten je zichtbaar vrolijk. Het duurde even voordat de tracer klaar was voor de eerste patiënt maar uiteindelijk hebben we toch mooie resultaten gekregen. We hebben ook leuke NET-etentjes gehad waar duidelijk werd dat je een echt familiemens bent. Ik ben blij dat ik met jou een aantal jaren mocht samenwerken en wens jou en jouw familie heel veel succes in Canada. In het buitenland werken is even wennen maar uiteindelijk denk je het hele leven graag aan terug. Maar dat hoef ik jou niet te vertellen. KPK, beste Klaas Pieter, we hebben een hele tijd samengewerkt op dit project. Van de BONs tot de BALBs en de eerste HTP-patiëntenscans hebben wij een heel mooie periode gehad. Ik ben heel erg blij dat je ondanks jouw eigen promotie mijn paranimf kunt zijn. Een betere afsluiting kan ik me niet voorstellen. Mijn tweede copromotor, Dr. Ido Kema, wil ik graag bedanken voor de goede ondersteuning en uitgebreide gesprekken over metabolismen in de cellulaire wereld. Beste Ido, je hebt mij de brug laten zien die de chemie en de biochemie verbindt. Ik sta er nu midden op met een uitzicht om te genieten. Prof. dr. Rudi Dierckx, mijn derde promotor, wil ik bedanken voor de leuke gesprekken tijdens en buiten werktijd. Met je komst is de afdeling NGMB een stuk internationaler geworden. Dit niet alleen omdat je Belg bent maar vooral om dat je met succes het percentage buitenlands personeel verhoogd hebt. Adrienne, je hebt de staf van Piet overgenomen. Je zal het project verdere goede impulsen kunnen geven. Mijn dank gaat aan de leden van de beoordelingscommissie, Prof. dr. ir. Marion de Jong van het Erasmus MC te Rotterdam, Prof. dr. Pax Willemse van het UMC Groningen en Prof. dr. Jürgen Martens van de Carl von Ossietzky Universität Oldenburg in Duitsland, voor het lezen en beoordelen van mijn proefschrift. Voordat ik aan dit onderzoek mocht beginnen was er nog een sollicitatiegesprek te voeren onder de leiding van Wim Vaalburg. Beste Wim, tijdens dit drietalige gesprek liet je mij weten dat mijn Nederlands binnen een jaar veel beter zou zijn. Toen geloofde ik er niks van 114

130 Dankwoord maar jij had gelijk. Je gaaf mij en Henriette vanaf begin het gevoel welkom te zijn. Het is jammer dat ik maar twee van mijn 4 jaren ond er jou leiding mocht werken. Bram Maas wil ik bedanken voor vier heerlijke jaren. Beste Bram, wat hebben wij plezier gehad op het lab, en ook daarbuiten. Vroeg zorgde jij ervoor dat ik het shirtje met het nummer 14 ging dragen. Het einde was dat we samen het Wilhelmus in de Euroborg zangen. Als ik ooit een goede analist nodig zou hebben dan weet ik jou te vinden. Alle mensen die met mij op de kamer van 6 hebben gezeten wil ik bedanken. Anne Rixt, voor de eerste 2 jaren en Janine voor de laatste 2 jaren. Jullie twee zijn echte snoepliefhebbers en ik ben blij dat ik af en toe mocht meegenieten. Jullie manier om onderzoek aan te gaan was ook voor mij leerzaam. DE fles heeft mij succesvol gepasseerd. Met de parttime dames Hilde, Bertha en Lizette was het altijd leuk om even bij te kletsen. De letters G, M en P zijn geen onbelangrijke in de radiofarmaceutische wereld. Harry Hendrikse, Marjolein de Hooge en Marieke Sturkenboom, jullie hebben mij dat op indrukwekkende wijze geleerd. Beste Jitze, dank jou technische en nauwkeurige vaardigheden kon ik altijd mijn proeven uitvoeren. Erik, met je nuchterheid en jouw uitspraak je moet nog eerst promoveren heb je bijgedragen dat het boekje toch nog rond is gekomen. Aren, je deur stond altijd voor mij open. Ik wil jou bedanken voor jouw ondersteuning op alle gebieden. Je liet me zien dat men ook op meer dan een vakgebied succesvol bezig kan zijn. De MNWers wil ik bedanken voor hun geduld, vooral als ik er weer met een pot C11-tracer stond te trippelen. Arja en Erna, jullie hebben ervoor gezorgd dat een publicatie Burn-out door HTP nooit zal verschijnen. Gerda, jouw vrolijke lach was altijd goed te horen en liet zelfs een trieste dag vrolijk worden. Mijn dank gaat aan de fysici Anne Paans en Johan de Jong. Beste Anne, ik wil je bedanken voor je steun tijdens de cursus stralingsveiligheid. Ik kon altijd met een vraagje bij jou terecht. Johan, bedankt voor je hulp bij het uitvoeren van de micropet scans. Annie van Zanten, je hebt ervoor gezorgd dat ik me nooit gedachtes hoefde te maken over administratieve dingen. Beste Annie, hiervoor van harte bedankt. Hetty Timmer-Bosscha en hun medewerkers van het lab de vitrine, Enge Venenkamp en Jasper Krijnen van het laboratorium medisch centrum en de medewerkers van het CDL wil ik bedanken voor hun bijdrage aan dit proefschrift. Iedereen met wie ik mocht samenwerken de afgelopen jaren op welke manier dan ook, bedankt: Antoon Willemsen (je initialen kom ik nu dagelijks tegen), Jan Pruim, Riemer Slart, Klaas Willem Sietsema, Liesbeth Ruytjens, David Cobben, Lukas Been, Laya Vercauteren, de jonge honden Nanne Brattinga en Xi-En Master Yang, Vera Rutgers en alle medewerkers van de afdeling NGMB die niet werden genoemd. Kurz vor Ende dieses Dankwortes möchte ich mich bei allen bedanken, die uns schöne Tage in der Heimat oder op Schier bescherten und Unterschlupf gewährt haben: Herms und Sabine, Kirsten und Joosten, Meike, Thomas und Ruben, Torsten und Svenja, Hillu und Barbro, Stefan, Chrischi, Antje und Paula. Unseren Familien möchte ich danken, einfach dafür das sie für uns da sind. Meinen Großeltern, Holger und Annette, den Heberts, Elisabeth, Malte und Sara und besonders meinen Eltern Reinhard und Brigitte. Mein letzter Dank geht an Dich, liebe Henriette. Ohne Deine Unterstützung, Wärme und Liebe wären sowohl die letzten 7 Jahre als auch dieses Buch nicht so geworden wie sie sind. Ohne Dich wären wir nie in Groningen gelandet und auch jetzt nicht hier in Melbourne. Ich liebe Dich! 115

131 116

132 117

133 118

134 Appendix 119

135 Chapter 1. Figure 1. Metastasized carcinoid. 18 F-DOPA PET scan (A), Octreotide scan (B) and 18 F-DOPA PET-CT fusion image (C) of a female patient presenting with a metastasized carcinoid. This patient illustrates the intra-individual heterogeneity in the uptake of different tracers by tumor metastases. 120

136 Appendix 99m Tc-V-DMSA Passive diffusion Active transport 11 C-5- HTP 18 F-DOPA Phosphate metabolism Serotonin Catecholamine pathway IMT Lysosome Nucleus Secretory vesicle Internalisation of receptorligand complex LAT 1 4F2HC complex transporter (family of LAT) Noradrenalin transporter NaPi co-transporter Glucose transporter VMAT transporter, located on secretory vesicle L-3-[ 123 I]-iodo-alphamethyltyrosine (IMT) Glucose metabolism Somatostatin receptor with (labelled) ligand Bombesin receptor with (labelled ligand) 18 F-Dopamine MIBG 18 FDG CCK receptor with (labelled) ligand VIP receptor with (labelled) ligand Antigen with (labelled) antibody Chapter 1. Figure 2. Metabolic pathways. In this figure the different metabolic pathways by which neuroendocrine tumors can be visualized using nuclear medicine imaging techniques are schematically depicted. Three major routes can be identified: receptor based techniques, techniques which use the metabolic properties of these tumors and labeled antibody based techniques 121

137 122

138 Appendix 123

139 Chapter 3. Figure 6. A - [ 11 C]HTP PET after IV injection, coronal view. Left: control 22.6 g, 9.9 MBq, tumor weight 23.5 mg. Right: carbidopa treated 21.6 g, 6.2 MBq, tumor weight 61.1 mg. B - [ 18 F]FDOPA PET after IV injection, coronal view. Left: control 20.5 g, 8.4 MBq, tumor weight 109 mg. Right: carbidopa treated 23.6 g, 8.4 MBq, tumor weight 57.9 mg. Summed frames. Hot spots in abdominal region were cleaned up using ASIPro s clipping tool. Arrows point at tumors located in the right shoulder. 124

140 Appendix Chapter 4. Figure 2. Imaging of a patient with carcinoid disease and metastases in the bone, mediastinum, liver, and abdomen. (A) 18 F-DOPA PET imaging. Red arrows indicate areas with physiological 18 F-DOPA uptake (striatum, kidneys, ureter, bladder), whereas all other black spots are tumour lesions. (B) Planar SRS imaging. Arrows indicate mediastinal tumour lesions. (C) CT PET fusion imaging. Coloured areas indicate tumour lesions. In this patient, both planar and SPECT SRS missed most lesions found with 18 F-DOPA PET imaging. Abdominal and femoral lesions were not recorded on CT. 125

141 126

142 Appendix 127

143 A B C D Chapter 5. Figure 2. Fused 18 F-DOPA PET CT scan (A), SRS (B), 18 F-DOPA PET (C) and 11 C-5-HTP PET (D) of a 80 year old male patient with metastatic carcinoid tumor. The CT scan shows a mesenterial mass and two smaller lesions in the upper mediastinum. On SRS (both planar and SPECT, not shown here) only the larger mediastinal mass, the large mesenterial mass and a small lesion on the left cranial side of the urinary bladder could be found. Both 18 F-DOPA PET and 11 C-5-HTP PET showed a number of smaller lesions in the upper mediastinum and upper lobes of both right and left lung, with 18 F-DOPA yielding the best contrast. Note that the small lung lesions show less 11 C-5-HTP uptake than 18 F -DOPA uptake. 128

144 Appendix A B C D Chapter 5. Figure 3. CT scan (A), SRS (B), 18 F-DOPA PET (C) and 11 C-5-HTP PET (D) of a 54 year old male patient with metastatic islet cell tumor. The CT scan shows a large mass in the pancreatic head region (arrow), SRS shows equivocal (arrow) and 18 F-DOPA PET shows low uptake in the pancreatic region and minor uptake in the upper chest and in two thoracic vertebrae. 11 C-5-HTP PET, however, shows numerous bone, liver and abdominal lesions, including the pancreatic region with much higher contrast. 129