Klaas Pieter Koopmans
1, Rudi A. Dierckx
2, Philip H. Elsinga
2, Thera P. Links
3, Ido P. Kema
4, Helle-Brit Fiebrich
5,
Annemiek M.E. Walekamp
5, Elisabeth G.E. de Vries
5, and Adrienne H. Brouwers
21Department of Radiology and Nuclear Medicine, Martini Hospital Groningen, Groningen, The Netherlands
2Department of Nuclear Medicine and Molecular Imaging, University of Groningen, and University Medical Center of Groningen, Groningen, The Netherlands
3Department of Endocrinology, University of Groningen, and University Medical Center of Groningen, Groningen, The Netherlands
4Department of Laboratory Center, University of Groningen, and University Medical Center of Groningen, Groningen, The Netherlands
5Department of Medical Oncology, University of Groningen, and University Medical Center of Groningen, Groningen, The Netherlands
4.3
GEP gastroenteropancreatic
LAT system l large amino acid transporters L‐DOPA l‐3,4 ‐dihydroxyphenylalanine MIBG metaiodobenzylguanidine MIP maximum intensity projection nET neuroendocrine tumor PET positron emission tomography
sRs somatostatin receptor scintigraphy with 111In‐octreotide VMAT vesicular monoamine transporters
introduction
Gastroenteropancreatic neuroendocrine tumors (GEP‐nETs) are tumors that arise in the pancreas and gastrointestinal system and are derived from neuroendocrine cells.
In contrast to many other malignancies, well‐differentiated GEP‐nETs generally have a low glucose metabolism [1, 2]. 18F‐fluorodexyglucose (18F‐FDG) PET scanning has limited value for the primary staging of patients with well‐differentiated GEP‐nET, showing only moderately increased glucose uptake in primary tumors and also often missing metastases. Besides, these primary tumors can be very small in size, thus failing the detection limit of PET camera systems. Therefore, as a staging tool, 18F‐
FDG does not seem to have a place in staging the general GEP‐nET patient population [1, 3]. A few studies with small heterogeneous GEP‐nET tumor groups have shown that in patients with rapidly progressive disease, dedifferentiation of GEP‐nET tumors can lead to a higher cellular glucose metabolism in tumor cells. In these patients, 18F‐
FDG PET can be of benefit for tumor staging. The 18F‐FDG uptake in tumor lesions in these patients could possibly play a role in predicting outcome and in assessing therapy response [4]. 18F‐FDG PET can be of value when other malignancies are sus-pected in patients with GEP‐nETs, since these patients experience a higher incidence of these malignancies compared to the general population [5]. However, data for this indication are scarce. Due to the limited applicability of 18F‐FDG PET in GEP‐nETs, this technique will not be discussed in depth.
In contrast to most tissues in the human body, GEP‐nETs often show increased synthesis and secretion of hormones and neurotransmitters. nontumorous neuroen-docrine cells regulate a variety of body functions through paracrine stimulation with a large variety of hormones and neurotransmitters. serotonin, catecholamine, and histamine are examples of compounds that share specific steps in their biosynthesis and storage, such as decarboxylation prior to storage in granules [6]. In GEP‐nETs, especially the catecholamine and serotonin biosynthetic pathways are upregulated.
Therefore, increased biosynthesis of these specific amines in GEP‐nETs enables imaging with specific amine precursors. In this chapter, imaging of cells that show increased production of catecholamine and/or serotonin will be discussed for GEP‐
nETs encompassing neuroendocrine tumors (nETs) from foregut origin (bronchus, lung, thymus, stomach, pancreas, and proximal duodenum), midgut nETs (from distal half of second part of the duodenum to the proximal two‐thirds of the transverse colon), and hindgut nETs (descending colon, sigmoid, and rectum).
CATECHOLAMInE PATHWAY 77
cAtecholAmine pAthwAy
Catecholamines act as neurotransmitters especially in the brain, or as hormones, for example, adrenaline when it is released from the adrenals, via α‐ and β‐adrenergic receptors located on vessels and internal organs.
In the catecholamine pathway, phenylalanine and intermediate products such as l‐3,4 ‐dihydroxyphenylalanine (L‐DOPA) are taken up via system l large amino acid transporters (LAT) (Fig. 4.3.1). After entering the cell, decarboxylation to dopamine takes place intracytoplasmatically via the enzyme aromatic amino acid decarbox-ylase (AADC). Vesicular monoamine transporters (VMAT) then transport dopamine into intracellular storage vesicles. In these vesicles, dopamine can, dependent on the activity of specific enzymes, be further metabolized to noradrenaline and adrenaline. The resulting end products dopamine, noradrenaline, and adrenaline can be released in the extracellular environment from these vesicles. selective
Secretory
Serotonin metabolism
Catecholamine metabolism
Active transport VMAT transporter, located on
secretory vesicle
figure 4.3.1 Overview. In this figure, a schematic overview is presented on the uptake mechanisms of the main tracers used in GEP‐nET imaging. GEP‐nET, gastroenteropancreatic neuroendocrine tumor.
reuptake transporter systems, for example, dopamine and noradrenaline trans-porters, can thereafter transport these hormones back into the cell.
Catecholamines can be metabolized mainly intracellularly but also extracellularly into various metabolites, which can be measured in plasma and urine as acid and basic cate-cholamine metabolites, such as homovanillic acid (HVA), vanillylmandelic acid (VMA), and metanephrines. Increased LAT activity seems to play a role in nETs, where due to a high precursor turnover and high AADC activity, a high precursor need has to be satis-fied. The exact mechanism for precursor uptake regulation is not yet clear though [6, 7].
For the catecholamine pathway, 6‐18F ‐l‐3,4‐dihydroxyphenylalanine (18F‐DOPA) and 6‐18F‐dopamine are available as tracers. since 6‐18F‐dopamine will be described extensively in Chapter 4.4.2, we will limit this chapter to 18F‐DOPA.
18F‐DOPA is most commonly produced via regioselective fluorodestannylation (electrophilic fluorination). Another possibility is nucleophilic fluorination, a multi-step procedure that, although more time consuming, has the advantage of having readily available large quantities of no‐carrier‐added 18F‐fluoride [6, 8].
18F‐DOPA tracer has been developed and first studied in men at McMaster university, Hamilton, Canada, during the 1970s and 1980s [9]. It has since then first been used for the imaging of the dopaminergic system in the striatum in patients with Parkinson’s disease and related disorders. several years later, this tracer has been used as a whole‐
body imaging technique in patients with nETs (Fig. 4.3.2) [10, 11].
figure 4.3.2 18F‐DOPA PET/CT. Images of a 73‐year‐old female with a serotonin‐pro-ducing metastasized well‐differentiated nET of the small bowel (carcinoid). On the left, a maximum intensity projection (MIP) image is depicted; in the middle, a coronal slice; and on the right, a coronal fusion slice of PET and low‐dose CT. Physiological uptake of 18F‐DOPA is visible in the striata and excretion via kidneys with physiological activity in the ureters and urinary bladder. Metastases are visible in the liver and mesenteric tissue/lymph nodes.
18F‐DOPA, 6‐18F‐l‐3,4‐dihydroxyphenylalanine; PET/CT, positron emission tomography/
computed tomography. (See insert for color representation of the figure.)
CATECHOLAMInE PATHWAY 79
In most centers, patients are required to fast 2–6 h with amino acid‐free fluid intake before the 18F‐DOPA PET scan is performed [12, 13]. 18F‐DOPA doses used are either a fixed dose or a body weight‐dependent dose, up to 5 MBq/kg. scanning after injection of 18F‐DOPA PET scan is usually performed 60 min after tracer injec-tion with or without suV quantificainjec-tion. scanning at other time points has not been proven to be advantageous [14].
The use of carbidopa as pretreatment is still in debate. When used, it is given as 2 mg/kg body weight or a fixed dose (i.e., 100–200 mg) 1 h before tracer injection [15–17]. Carbidopa is a peripheral inhibitor of the enzyme AADC. It decreases the peripheral decarboxylation of 18F‐DOPA and also β‐[11C]‐5‐hydroxy‐l‐tryptophan (11C‐5‐HTP), a tracer of the serotonin pathway, thus reducing renal excretion and subsequently improving tracer uptake in metastatic carcinoid tumors most likely due to increased tracer availability in the circulation. This resulted in better image quality due to decreased streaky image reconstruction artifacts caused by high physiological excretion of the radiotracer via kidneys and urinary bladder and increasing suV values of tumor lesions [17]. Why carbidopa pretreatment results in a generally decreased tracer uptake in pancreatic tissue is not quite understood, but may be related to differences in AADC activity in nETs and differences in metabolic handling of these PET tracers by exocrine and endocrine pancreatic tissue [16].
Because of the likely strong similarities between 11C‐5‐HTP and 18F‐DOPA tracer handling in patients with gastrointestinal nETs, carbidopa pretreatment is also advo-cated for 18F‐DOPA PET imaging in this patient group to improve image quality especially in the region of the kidneys and bladder and to improve lesion detectability via further increased suV values of lesions. Indeed, high accuracy rates for the detection of (metastatic) nETs have been reported by the institutes that do use carbi-dopa pretreatment [18]. However, pancreatic islet cell tumors may be an exception to the rule [15]. This should be further investigated [16]. With carbidopa pretreatment, the estimated mean radiation dose is 1.9 msv per 100 MBq 18F‐DOPA [19].
A possible complication of administering catecholamines to patients with a large tumor load is the development of a carcinoid crisis. Thus far, the development of a carcinoid crisis after injection of 18F‐DOPA has only been described in one case and is possibly related to a lower specific activity of 18F‐DOPA tracer [20].
Results of 18F‐DOPA PET(/CT) imaging in patients with midgut well‐differenti-ated nETs (carcinoids) and other GEP‐nETs, such as (non)functioning pancreatic nETs (islet cell tumors), have been described in the literature [6, 18, 21–23]. In these studies, most patients had proven (recurrent) disease, and 18F‐DOPA PET(/CT) has been compared with current anatomical imaging techniques, mostly CT and/or MRI, and other functional imaging techniques, such as somatostatin receptor scintigraphy (sRs) with 111In‐octreotide, 123/131I‐labeled metaiodobenzylguanidine (MIBG), and
18F‐FDG PET, and more recently also with 68Ga‐labeled PET‐based variants of octreotide. Morphological imaging techniques seem to be complementary to the functional imaging techniques under study in these patient series. In carcinoids, patient‐based reported sensitivities for 18F‐DOPA PET(/CT) are (very) high, ranging from 65 to 100% [6, 18, 22, 23], and therefore, it seems to be an excellent staging method for this patient group [6, 23]. 11C‐5‐HTP PET, however, was better than
18F‐DOPA PET on a patient‐based analysis (sensitivity of 100% vs. 96%, respectively), whereas per‐lesion analysis showed the opposite (sensitivity of 78% for 11C‐5‐
HTP and 87% for 18F‐DOPA, respectively) in patients with carcinoids and islet cell tumors [22]. Another study of Fiebrich et al. showed that in patients with carcinoid tumors, 18F‐DOPA uptake, defined as total body tumor load, corre-sponds well to the tumor markers of the serotonin and catecholamine pathway in the urine and plasma in carcinoid patients, thereby reflecting metabolic tumor activity [24].
The 68Ga‐labeled analogues of somatostatin performed better than 18F‐DOPA PET in two studies with mixed nET types. A partial explanation for this result can be that in patients with (non)functioning islet cell tumors, 18F‐DOPA PET does not seem to have a good detection rate for these tumors [13, 21, 22].
since the published results for 18F‐DOPA PET were mainly achieved in patients with proven nET, it can be expected that performance of 18F‐DOPA PET will be lower in clinically more difficult cases where diagnosis of a nET is suspected but has yet to be confirmed [6, 25]. This was addressed in a recent published study in which patients (n = 119) with only biochemical proof of the disease and various types of nETs were investigated [25]. In this study, the diagnostic accuracy of 18F‐DOPA PET/CT in patients who were entered for primary staging of abdominal nETs was 88%, whereas the diagnostic accuracy for patients entered for restaging of abdominal nETs (n = 61) was 92%. specificity of 18F‐DOPA PET in this tumor type has not yet been studied, since most research has been performed in patients with proven nETs where it is unlikely that histological proof will be obtained [6]. Thus far, no studies have been published on the influence of 18F‐DOPA PET on therapeutic management or the prognostic value for early response monitoring with 18F‐DOPA PET. These are all areas in which more research is warranted.
There is a clear need for in‐depth comparison of the other recently emerging functional PET imaging techniques, especially the 68Ga‐labeled somatostatin ana-logues in (subsets of) nETs with 18F‐DOPA PET. Advantages of 68Ga‐labeled somatostatin analogues are the relatively easy generator‐based synthesis and the pos-sibility to evaluate whether peptide receptor radionuclide therapy (PRRT) for nETs can be considered. However, 18F‐DOPA PET may have a broader clinical applica-bility, for example, for studying the dopaminergic system of the human brain, malig-nant tumors that do not (usually) express somatostatin receptors [26], and nononcologic settings, such as infants with hyperinsulinism. since 18F‐DOPA is already commercially available in many countries, the increased availability may lead to a better understanding of the place of 18F‐DOPA PET in clinical decision making for nETs.
serotonin pAthwAy
serotonin is a monoamine neurotransmitter that is present in blood platelets and in the intestinal wall where it regulates contractions. In the brain, it acts as a neurotrans-mitter and plays a role in feelings of well‐being.
sEROTOnIn PATHWAY 81
Both the serotonin and the catecholamine pathways share common transporter systems, the LAT and the VMAT systems. Also, in both pathways, the enzyme AADC plays a key role in the final step for the synthesis of a functional hormone, serotonin and dopamine, respectively.
In the serotonin pathway, the amino acid tryptophan and the intermediate product 5‐HTP are precursors for serotonin after uptake via the LAT system. How uptake is regulated is still unclear. After decarboxylation via the enzyme AADC, serotonin is taken up via VMAT into storage vesicles from which it can be released extracellu-larly (Fig. 4.3.1). serotonin is eventually metabolized to 5‐hydroxyindoleacetic acid (5‐HIAA), which is secreted in the urine. The serotonin pathway is overactive in many nETs, which makes this pathway a good candidate for imaging with PET tracers [6].
Thus far, only a carbon‐11‐labeled tracer has been developed for the serotonin pathway, β‐[11C]‐5‐HTP. This tracer was developed in uppsala, sweden, during the 1980s. It is a complex tracer to produce, demanding an on‐site cyclotron for the production of the 11C isotope, which has a half‐life of only 20 min. The synthesis of 11C‐5‐HTP itself is rather complex since it requires two multienzyme steps [27, 28]. Only a few centers worldwide produce 11C‐5‐HTP. However, in experienced hands, quantities up to 1000 MBq can be reliably synthesized for (clinical) use. Image interpretation is mostly easy due to high tumor tracer uptake and low background uptake in normal tissues. Clinical results justify the use of this tracer [7].
In 1993, the first results with this tracer were published [29]. nowadays, 11C‐5‐
HTP PET scanning is typically performed 10–20 min after injection of 11C‐5‐HTP with carbidopa pretreatment, 2 mg/kg body weight or with a fixed dose of 200 mg, 1 h prior to injection (Fig. 4.3.3). In contrary to 18F‐DOPA, no adverse reactions have thus far been reported after intravenous administration of 11C‐5‐HTP.
For 11C‐5‐HTP PET scans, carbidopa proved to be essential in order to improve image quality. Without carbidopa pretreatment, scans showed streaky image recon-struction artifacts due to high urinary tract 11C‐5‐HTP uptake and excretion.
Theoretically, carbidopa decreases the peripheral decarboxylation of 11C‐5‐HTP to
11C‐5‐HT (serotonin) and subsequent urinary excretion of serotonin metabolites. The effect of carbidopa was tested in patients with midgut carcinoids, who were scanned with and without carbidopa pretreatment 1 h prior to injection of 11C‐5‐HTP. It was found that oral pretreatment significantly reduced urinary radioactivity concentration from a mean suV of the pelvis of 155 ± 195 to 39 ± 14 sD. This led to an improved image quality in that area. Tumor uptake of 11C‐5‐HTP significantly increased, from 11 ± 3 to 14 ± 3 sD. Interestingly, pancreatic uptake decreased slightly to an suV of 4.4 ± 0.8 sD. In liver tissue, a small increase of 11C‐5‐HTP uptake was found, now reaching an suV of 3.6 ± 0.8 [17].
Due to the low number of centers using 11C‐5‐HTP PET, only few studies with a reasonable number of patients with GEP‐nETs have thus far been published. In the first published study with 11C‐5‐HTP PET, 18 patients with histopathologically veri-fied nETs were included: midgut (n = 14), foregut (n = 1), and hindgut carcinoid (n = 1) and endocrine pancreatic tumors (n = 2) [30]. 11C‐5‐HTP PET without oral
carbidopa pretreatment was compared to CT. All 18 patients showed increased 11C‐5‐
HTP uptake in tumor tissue; interestingly, a patient with a hindgut carcinoid and a patient with nonfunctioning endocrine pancreatic tumor with normal urinary 5‐HIAA levels also showed increased tumor uptake. 11C‐5‐HTP PET detected more tumor lesions than CT in 10 patients and was equal in five patients (four midgut, one foregut), with missing data in three patients with midgut nET. In the 10 patients that were on treatment (interferon‐α ± octreotide, or somatostatin analogue only), a close correlation between the changes in 11C‐5‐HTP transport rate constant in the tumors and urinary 5‐HIAA was noted. It was therefore suggested that 11C‐5‐HTP PET may serve as a means to monitor therapy, although it is still unknown whether these PET findings under medication relate to changes in tumor metabolism or in changes in amine processing [30].
In another patient series, 38 patients were evaluated, again with a variety of nETs, consisting of midgut carcinoids (n = 13), lung carcinoids (n = 7), nonfunctioning endocrine pancreatic tumors (n = 5), and other nETs [30]. Whole‐body 11C‐5‐HTP PET imaging with oral carbidopa pretreatment was compared to both CT and the functional imaging technique sRs, which is often used in standard clinical care in nETs and more widely available than 11C‐5‐HTP PET. Whole‐body 11C‐5‐HTP PET scanning detected tumor lesions in 36 of 38 (95%) patients. In 84% of patients, sRs was positive, whereas CT was positive in 79%. More lesions were detected with figure 4.3.3 11C‐5‐HTP PET/CT. Images of a 62‐year‐old female with a gastrinoma. In contrary to 18F‐DOPA PET, the striata are not visualized. On the left, an MIP is depicted; in the middle, a coronal slice; and on the right, a coronal fusion slice of PET and low‐dose CT.
Physiological excretion of 11C‐5‐HTP metabolites is via kidneys with physiological activity in the ureters and urinary bladder. Tumor is visible in the head of the pancreas and metastases in the liver.
11C‐5‐HTP, β‐[11C]‐5‐hydroxy‐l‐tryptophan; 18F‐DOPA, 6‐18F‐l‐3,4‐dihydroxyphenylalanine;
PET/CT, positron emission tomography/computed tomography. (See insert for color repre-sentation of the figure.)
sEROTOnIn PATHWAY 83
11C‐5‐HTP PET in 22 of 38 (58%) patients compared to sRs and CT, whereas the imaging modalities showed equal numbers of lesions in 13 of 38 patients (34%). In three patients, sRs or CT showed more lesions than 11C‐5‐HTP PET. Patients with a nonfunctioning endocrine pancreatic tumor and a pancreatic carcinoma with some endocrine differentiation on immunohistochemistry were PET negative. A patient with metastasized thymus carcinoma only showed the primary tumor on 11C‐5‐HTP PET and sRs, while CT scan also detected metastases. From these patients, it was speculated that in the case of high proliferation rate and dedifferentiation of nETs,
18F‐FDG PET is probably the imaging modality of choice. In the 17 patients who had their primary tumor still in situ, 11C‐5‐HTP PET was positive in 16, compared to sRs in 9 and CT in 8 patients. PET could detect surgically removed lesions as small as six mm. The main conclusion of this study was that PET imaging with 11C‐5‐HTP can be universally applied in nETs, also in patients without elevated 5‐HIAA excretion in urine, as long as the tumor is not highly proliferating and/or dedifferentiating [31].
In another study with 24 patients with carcinoid tumors and 23 patients with pancreatic islet cell tumors, 11C‐5‐HTP PET was compared to conventional imaging with CT and sRs and a 18F‐DOPA PET scan [22]. Whole‐body PET images with carbidopa pretreatment were recorded 10 min after intravenous injection of 11C‐5‐
HTP or 60 min after intravenous administration of 18F‐DOPA. The PET findings were compared, per patient and per lesion, with a composite reference standard derived from all available imaging data along with clinical and cytological/histological information.
Results indicated that indeed 11C‐5‐HTP PET can be seen as a universal imaging agent for carcinoid and pancreatic islet cell tumor patients. It was the only imaging modality that was positive in all patients (100% sensitivity). Especially in islet cell tumor patients, more tumor‐positive patients and lesions were found with 11C‐5‐HTP (100 and 67%) compared to 18F‐DOPA (89 and 41%), sRs (78 and 46%), and CT (87 and 68%). In carcinoid patients, the per‐lesion analysis showed that 18F‐DOPA PET outperformed all other imaging techniques. Adding CT to both imaging tech-niques resulted in a further improvement in sensitivity in a per‐lesion analysis, since
Results indicated that indeed 11C‐5‐HTP PET can be seen as a universal imaging agent for carcinoid and pancreatic islet cell tumor patients. It was the only imaging modality that was positive in all patients (100% sensitivity). Especially in islet cell tumor patients, more tumor‐positive patients and lesions were found with 11C‐5‐HTP (100 and 67%) compared to 18F‐DOPA (89 and 41%), sRs (78 and 46%), and CT (87 and 68%). In carcinoid patients, the per‐lesion analysis showed that 18F‐DOPA PET outperformed all other imaging techniques. Adding CT to both imaging tech-niques resulted in a further improvement in sensitivity in a per‐lesion analysis, since