| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Turku PET Centre (S.K., M.S., A.S., R.P., P.N.), and Departments of Surgery (S.K., R.G.), Oncology and Radiotherapy (H.M.), and Pathology (K.A.), Turku University Hospital, FIN-20520 Turku, Finland; Turku PET Centre (O.S., J.B.), Radiopharmaceutical Chemistry Laboratory, University of Turku, FIN-20520 Turku, Finland; Department of Medicine (T.S., M.V.), Helsinki University Hospital, FIN-00029 Helsinki, Finland; and Department of Medicine (J.S.), Tampere University Hospital, FIN-33014 Tampere, Finland
Address all correspondence and requests for reprints to: Pirjo Nuutila, M.D., Turku PET Centre, Turku University Hospital, P.O. Box 52, FIN-20521 Turku, Finland. E-mail: pirjo.nuutila{at}tyks.fi.
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Patients and Methods: We prospectively studied 10 patients with confirmed hyperinsulinemic hypoglycemia and presumed insulin-secreting tumor using 18F-DOPA PET. Anatomical imaging was performed with computed tomography (CT) and magnetic resonance imaging (MRI). All patients were operated on, and histological verification was available in each case. Semiquantitative PET findings in the pancreas using standardized uptake values were compared to standardized uptake values of seven consecutive patients with nonpancreatic neuroendocrine tumors.
Results: By visual inspection of 18F-DOPA PET images, it was possible in nine of 10 patients to localize the pancreatic lesion, subsequently confirmed by histological analysis. 18F-DOPA uptake was enhanced in six of seven solid insulinomas and in the malignant insulinoma and its hepatic metastasis. Two patients with ß-cell hyperplasia showed increased focal uptake of 18F-DOPA in the affected areas. As compared to CT or MRI, 18F-DOPA PET was more sensitive in localizing diseased pancreatic tissue.
Conclusion: 18F-DOPA PET was useful in most patients with insulinoma and negative CT, MRI, and ultrasound results. In agreement with previous findings in infants, preoperative 18F-DOPA imaging seems to be a method of choice for the detection of ß-cell hyperplasia in adults. It should be considered for the detection of insulinoma or ß-cell hyperplasia in patients with confirmed hyperinsulinemic hypoglycemias when other diagnostic work-up is negative.
|
Introduction |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
After diagnosis of organic hyperinsulinism, surgical enucleation of the tumor or partial pancreatic resection is the primary mode of treatment. The conventional imaging studies [ultrasound, computed tomography (CT), magnetic resonance imaging (MRI), angiography, intraarterial calcium stimulation with venous sampling, endoscopic ultrasound (EUS), and somatostatin receptor scintigraphy (SRS)] have been the cornerstone for tumor localization in patients with insulinoma or ß-cell hyperplasia. However, so far, all these methods either have limited sensitivity or are invasive procedures (11).
Positron emission tomography (PET) combined with fluorine-18-L-dihydroxyphenylalanine (18F-DOPA) has been used for more than two decades to investigate the activity of the dopaminergic system in vivo in neurological disorders. Neuroendocrine tumors (NETs) have the capacity to take up and decarboxylate amine precursors like L-DOPA and 5- hydroxy-L-tryptophan (12). Furthermore, the endocrine pancreas has been shown to have an efficient mechanism for the uptake and decarboxylation of L-DOPA, converting it to dopamine (13, 14). There has also been some promising experience of diagnosing pancreatic NETs using tracers like 11C-5-hydroxy-L-tryptophan and 64Cu-TETA-octreotide, in addition to 11C-DOPA (15, 16, 17). 18F-DOPA PET has been successfully used in the imaging of carcinoid tumors, glomus tumors, pheochromocytomas, and medullary thyroid cancers (18, 19, 20, 21). In 2004 Becherer et al. (22) studied 23 patients with advanced NETs using 18F-DOPA, including three successful diagnoses of pancreatic NET, and found 18F-DOPA to be useful in the staging of diseases.
Recently, we and others have demonstrated the usefulness of 18F-DOPA PET in the detection of focal ß-cell hyperplasia (23, 24) and ectopic lesion (25) in congenital hyperinsulinism of infancy. Concurrently, with infant studies, we have conducted another prospective research protocol to determine the potential of 18F-DOPA PET to identify and precisely localize the insulin-secreting tumors of the pancreas in adults. Our aim has been to compare PET findings with the established imaging methods. This report summarizes our current results from eight consecutive adult patients with insulinoma and two patients with ß-cell hyperplasia. We compare the diagnostic sensitivity of 18F-DOPA PET with conventional methods in hyperinsulinemic and hypoglycemic adults. Furthermore, the outcome of these 10 patients is described.
|
Patients and Methods |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Ten consecutive patients (five females, five males; mean age, 49 ± 13 yr) with confirmed hyperinsulinemic hypoglycemic episodes were referred to the PET Centre in Turku, Finland, between April 2001 and January 2006 (Table 1
). All patients had symptoms of neuroglycopenia associated with inappropriately elevated levels of insulin in relation to decreased glucose levels. The study patients fulfilled the diagnostic biochemical criteria of hyperinsulinemic hypoglycemia, with a mean glucose value of 2.2 mmol/liter, a concomitant mean insulin level of 41 mU/liter, and a mean C-peptide level 1.69 nmol/liter. The mean duration of symptoms before diagnosis was more than 3 yr, and mean body mass index was 29.1 ± 5.4 kg/m2 (Table 1). More details about patient histories are given in Results. All patients were studied with 18F-DOPA PET. PET results were compared with the pancreatic 18F-DOPA uptake of seven consecutive patients (three females, four males; mean age, 47 ± 16 yr; body mass index, 24.6 ± 1.9 kg/m2) with suspected extraadrenal or adrenal neuroendocrinological disease. None of these had any suspicion of pancreatic disease. Two of seven control patients had cortical adrenal adenoma, one patient had a malignant pheochromocytoma, one patient had a suspicion of recurrent adrenal pheochromocytoma, and three patients had no NETs.
|
18F-DOPA PET
Patients fasted for at least 6 h before the PET scan. If the patient was using medication that prevents pancreatic insulin release (diazoxide, somatostatin analogs, or cortisone), the medication was withdrawn 24 h before the study. Plasma glucose level was monitored, and a glucose (G5%–G10%, 40–100 ml/h) infusion was given, if needed, to keep plasma glucose between 4.0 and 5.0 mmol/liter.
18F-DOPA was synthesized according to previously described methods (26). The average iv injected dose of 18F-DOPA was 234 ± 71 MBq (range, 160–362). Scanning began 60 min after injection. Patients underwent a whole-body PET scan from the level of the eyes to the thigh with a GE Advanced PET scanner (General Electric Medical Systems, Milwaukee, WI) operated in two-dimensional mode. The scanner consists of 18 rings of bismuth germinate detectors yielding 35 transverse slices spaced at 4.25-mm intervals. The diameter of the imaging field of view is 55 cm, and the axial length is 15.2 cm.
To obtain images for visual and semiquantitative analysis, the data were corrected for dead time, decay, and photon attenuation, and reconstructed in a 128 x 128 matrix. The final in-plane resolution in segmented attenuation correction and iterative-reconstructed and Hann-filtered (4.6 mm) image was 5 mm (full-width half-maximum).
PET images were analyzed visually and semiquantitatively by calculating mean and maximum standardized uptake values (SUVs) in the region of interest drawn separately on the pancreatic head, body, and tail. Axial, coronal, and sagittal views were evaluated, and the pancreas invariably had a sufficiently high uptake of 18F-DOPA to distinguish it from the liver, duodenum, and kidneys. Variable uptake was seen in the gallbladder and biliary duct. Pancreatic tissue uptake of PET images was always correlated side by side with anatomical reference images of CT or MRI. Interpretation was based on consensus of two specialists with significant experience in PET imaging (M.S. and H.M.), and there was no disagreement.
CT
All patients were examined with CT. Seven of 10 patients were imaged on a four-row CT scanner (Siemens Volume Zoom; Siemens, Erlangen, Germany), and three patients were imaged on a one-row CT scanner (Siemens Somatom Plus, Siemens). In all examinations, the reconstructed final transverse slice thickness was either 5 or 6 mm. Four patients (no. 1, 2, 3, and 5) were imaged without contrast injection and after contrast administration both in arterial and venous phase. Six patients were imaged without contrast and postcontrast venous phase.
MRI
MRI imaging was performed at 1.5 T (Vision or Symphony; Siemens Medical System, Erlangen, Germany) on six patients and at 1.5 T (Genesis Sigma; General Electric Medical Systems) on one patient. Five patients (no. 1, 2, 5, 9, and 10) were imaged on Symphony with the following pulse sequences: transverse T2-weighted images with and without fat saturation, and transverse T1-weighted images with slice thickness of 5 mm, in addition to axial T1-weighted images with and without administration of contrast agent, gadopentetate dimeglumine (Magnevist; Schering AG, Berlin, Germany). The T1-weighted sequence was performed after iv administration of Magnevist followed by a saline flush, i.e. in precontrast, arterial, venous, and steady-state phases. One patient (no. 7) was imaged with Vision with the following pulse sequences: transverse and coronal T2-weighted images with slice thickness of 5 mm, transverse T2-weighted images with slice thickness of 6 mm, and T1-weighted images with fat saturation with and without Magnevist. One patient (no. 8) was imaged with Genesis Sigma with the following pulse sequences: transverse T2-weigted images with and without fat saturation, transverse T1-weigted phase images, coronal T2-weighted images with fat saturation, and transverse T1-weighted gradient-refocused images with fat saturation. This last sequence was performed with and without administration of Magnevist in precontrast, arterial, venous, and steady-state phases.
Histological analysis
The primary diagnosis was performed using 4-µm-thick, van Gieson-stained sections cut from paraffin blocks. The immunohistochemical staining was performed using automated immunostainer (TechMate 500, DakoCytomation; Dako, Glostrup, Denmark). Sections were dried overnight at room temperature and kept at 60 C for 1 h. This was followed by antigen retrieval in ChemMate (DakoCytomation) solution in a microwave oven twice before immunohistochemical staining. Primary antibody incubation was followed by incubation with a ChemMate detection kit, based on the indirect streptavidin biotin method, with a biotinylated link antibody. Endogenous peroxidase treatment was followed by incubation with streptavidin conjugate labeled with horseradish peroxidase. Diaminobenzedine was used as a chromogen, followed by light Mayer’s hematoxylin nuclear counterstaining.
Statistical analysis
Coefficient of variation (CV) was used to assess intrapatient variations. The Mann-Whitney U test was used to compare the variables (CV%) between groups. Sensitivity was calculated in each imaging method (PET, CT, and MRI). A McNemar test was performed to compare PET, CT, and MRI, and 
coefficient was determined to quantify agreement of these imaging methods. P < 0.05 was considered statistically significant. All statistical analyses were conducted using SPSS statistical software (version 13.0; SPSS Inc., Chicago, IL).
|
Results |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Histological confirmation of diagnosis was obtained in all of 10 study patients. Two patients had focal ß-cell hyperplasia (patients 1 and 2), seven had solitary benign insulinoma, and one patient (no. 5) had malignant insulinoma.
Patients with ß-cell hyperplasia.
A 34-yr-old woman (patient 1) had been otherwise healthy except for a history of preeclampsias during pregnancies and low plasma glucose concentrations after delivery. In March 2001, during a varicose vein operation, the patient had a hypoglycemic episode with seizure and she was admitted to our hospital. She needed continuous glucose infusion to maintain euglycemia. During hypoglycemia, serum insulin level was 220 mU/liter and that of plasma C-peptide 4.4 nmol/liter (Table 1
). CT, MRI, angiography, and SRS were all normal. At explorative laparotomy, the pancreas was regarded as normal including a negative intraoperative ultrasound finding. Before pancreatectomy, it was decided to try to localize the hyperfunctional focus using 18F-DOPA PET. PET imaging showed enhanced tracer uptake in the head of the pancreas (Fig. 1 and Table 2). The mean SUV in head, body, and tail was 4.5, 2.4, and 4.1 g/ml, respectively. In 2001 relaparotomy was performed and a near-total pancreatectomy, guided by PET and preserving only the unaffected part of the head of the pancreas, was performed. Histological analysis showed typical findings of focal ß-cell hyperplasia (Fig. 2). Thereafter, the patient was entirely asymptomatic and normoglycemic, and follow-up was discontinued in June 2005.
|
|
|
). The mean SUV in head, body, and tail was 3.0, 2.4, and 2.7 g/ml, respectively. The patient underwent laparotomy for a presumed insulinoma, but no palpable abnormality in the pancreas was found. Under a suspected diagnosis of insulinoma, a subtotal standard resection of the pancreas with splenectomy was performed, leaving the head of the pancreas intact. Histological examination of the resected specimen revealed focal hyperplasia of islet cells, and also the proximal resection margin contained a microscopic focus of nesidioblastosis (Fig. 2), whereas the major part of resected pancreas was considered to be normal. Postoperatively, the patient has remained symptomatic. The most effective medication during these years has been a long-acting calcium channel blocker. Repeated 18F-DOPA PET showed intense uptake in the pancreas remnant consistent with residual nesidioblastosis. Both patients with ß-cell hyperplasia (no. 1 and 2) had the genomic DNA analyzed for mutations in the SUR1 and the Kir6.2 genes, which encode the subunits of the pancreatic ATP-sensitive potassium channel responsible for glucose-induced insulin secretion (1). No mutations were found in either of these genes.
Patients with insulinomas had a varying degree of symptoms and signs (Table 1). A typical case is a previously healthy 53-yr-old man (patient 9) who presented in July 2004 with a 2-yr history of episodes of seizures. Laboratory studies during evaluation revealed a plasma glucose level of 2.2 mmol/liter with a concomitant serum insulin level of 8 mU/liter. Neither contrast CT nor MRI was able to detect a pancreatic tumor. Four months later, 18F-DOPA PET showed enhanced focal uptake in the head of pancreas (Fig. 3). When laparotomy was performed in January 2005, a 1-cm tumor in the head of the pancreas was found, and pancreaticoduodenectomy was performed. The pathological diagnosis was insulinoma. Postoperatively, the patient has remained entirely asymptomatic.
|
). Thus, PET strongly suggested that hypoglycemic episodes were due to an insulinoma in the pancreatic body. In September 2002, a firm tumor smaller than 3 cm was found at laparotomy in agreement with the findings in PET. The body and tail of the pancreas were resected, and final diagnosis was malignant insulinoma with metastases in the liver. After 4 months of remission, the disease progressed. Thereafter she was treated with chemotherapy until late 2003, and she died 2 months after finishing the oncological treatment.
|
The pancreas was visible in all patients on 18F-DOPA PET scans. Both patients with focal ß-cell hyperplasia (patients 1 and 2) had an area of increased uptake of 18F-DOPA (Fig 1). By visual inspection, it was possible to localize the lesion in six of seven patients with insulinoma using 18F-DOPA PET (Table 2 and Figs. 3 and 4).
The semiquantitative analysis was performed in nine patients, whereas in one patient (no. 7) high tracer uptake in bile made the analysis unreliable. In the remaining nine study patients, the mean and maximum SUV was 4.4 ± 1.0 and 6.7 ± 2.0 g/ml in the affected pancreas and 3.2 ± 1.0 and 5.1 ± 1.6 g/ml, respectively, in other parts of the pancreas. In control patients with nonpancreatic NET, SUVmean of head, body, and tail was 4.4 ± 1.0, 3.9 ± 1.6, and 4.6 ± 1.7 g/ml, and SUVmax was 6.8 ± 1.4, 5.6 ± 2.3, and 7.2 ± 2.6 g/ml, respectively. Accumulation of 18F-DOPA seemed to be more homogenous in control patients, but no significant difference was found in maximum or mean SUV values between patients with and without pancreatic tumor lesions. In general, the body of the pancreas showed 18 ± 8% lower tracer uptake compared with the rest of the pancreas in both study and control patients.
18F-DOPA PET vs. CT and MRI
The sensitivity of 18F-DOPA PET was 90% [95% confidence interval (CI), 56–99%] to identify the disease focus in patients with PHH, whereas the sensitivity of CT was 30% (95% CI, 7–65%) and MRI was 40% (95% CI, 12–74%). CT and MRI were negative in both patients with ß-cell hyperplasia. By contrast, in four patients with insulinoma, CT (three of eight) or MRI (one of five) also showed the lesion (Table 2). In five of 10 patients, 18F-DOPA PET provided additional information for surgery that was obtained with none of the other imaging procedures. Difference in classification between PET and CT was significant (P = 0.031), and between PET and MRI the difference remained nonsignificant (P = 0.125). The coefficient for agreement was negligible ( = 0.09) for PET and CT, and that for PET and MRI ( = –0.21) was negative, respectively. The latter was due to being as conservative as possible in statistical analysis and required MRI to be considered as positive whenever it was not actually performed (three cases). Similarly, CT was regarded as positive in patient 4, in whom the lesion was only detected in retrospect with knowledge of PET results.
Management, histology, and outcome
All patients underwent operation. The tumors were bimanually palpable in eight patients, whereas in the case of ß-cell hyperplasia (patients 1 and 2), no palpable tumors were found. Insulinomas were enucleated in three patients. A subtotal pancreatic resection was necessary in seven patients, and three of those patients also underwent splenectomy. Only one patient required a pancreaticoduodenectomy (Table 1). The primary tumors were distributed throughout the pancreas and ranged from 1.0–3.0 cm in diameter. In seven of 10 patients, the tumors had typical histological characteristics of insulinoma, such as insulin-producing ß-cells with amyloid stroma. One patient (no. 5) had a malignant insulinoma in addition to liver metastases. Two patients had a focal form of ß-cell hyperplasia. In those patients, morphological findings included hypertrophic islets showing ß-cells with enlarged nuclei, increased numbers of islet cells budding off ducts, and neoformation of islets from ducts (Fig. 2).
Hypoglycemia symptoms disappeared after operation in nine patients. Only one patient (no. 2), who had ß-cell hyperplasia of the head of the pancreas, remained symptomatic after operation. One patient (no. 6) had a complication related to the pancreatic enucleation, resulting in severe pancreatitis, leading to death. Patient 5, who had malignant insulinoma with liver metastases, died 15 months postoperatively.
|
Discussion |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
PET provides the opportunity to design tracers chemically close or identical to physiological substrates, which is of interest in NET research (12). Increased activity of L-DOPA decarboxylase was found to be characteristic of NET (28). According to previous immunohistochemical studies, pancreatic ß-cells can also take up amine precursors, like L-DOPA, and convert them into dopamine by L-DOPA decarboxylase (13, 29, 30). The radiolabeled L-DOPA is transported across the ß-cell membrane by an amino acid transporter, after which it is decarboxylated into 18F-fluoro-dopamine and stored in vesicles.
At present, the preoperative planning in the treatment of insulinoma or ß-cell hyperplasia is significantly restricted by the limited preoperative knowledge of the location of the lesion. On the basis of our results, 18F-DOPA PET was a significantly more sensitive imaging method to identify the disease focus in patients with PHH compared with CT or MRI. We found focal accumulation of 18F-DOPA in the pancreas of nine of 10 patients, and the focal lesion was confirmed histologically as an insulinoma or ß-cell hyperplasia in all 10 patients. All patients underwent limited pancreatic resection (seven of 10) or enucleation (three of 10). In line with infants, our experience with adult patients indicates that the functional imaging with 18F-DOPA PET might help to optimize surgical treatment of focal ß-cell hyperplasia. So far, ß-cell hyperplasia is a rare and poorly characterized disease. Recently, there has been rising interest in this disease as a cause of PHH in adults (2, 8, 9, 10). Anlauf et al. (2) reported a series of 232 adult patients with PHH. Ten of those patients (4.3%) histologically showed the features of ß-cell hyperplasia. In a recent study, Service et al. (8) reported six cases of ß-cell hyperplasia occurring in patients who had undergone gastric bypass surgery as treatment for severe obesity.
Surgical excision of the lesion is the primary mode of treatment, and patients are almost invariably cured for life with complete excision of a solitary benign insulinoma. Accurate localization of these small tumors provides important support for the surgeon, limiting the surgical exploration and therewith the morbidity. However, the precise preoperative localization of these lesions has remained a clinical problem. Earlier, routine blind subtotal distal pancreatectomy was the treatment of choice (31, 32, 33). Nowadays, treatment is accomplished by enucleation or partial pancreatic resection. In recent years, however, advanced laparoscopic techniques have been used to resect islet cell tumors in a few patients (34, 35). This novel surgical approach makes preoperative imaging even more important due to lack of bimanual palpation.
The major limitation of the present study was the small number of patients, due to the rarity of the disease. Although the small sample size limits the analysis, our study is unique because it includes a population imaged at one institute. Another limitation is that conventional imaging studies (CT, MRI, SRS, and EUS) were not performed systematically in all patients.
In conclusion, 18F-DOPA PET seems to be superior to CT or MRI in localizing insulinoma or ß-cell hyperplasia. Because L-DOPA also accumulates physiologically in the normal pancreas, 18F-DOPA PET is a useful diagnostic tool only for patients with confirmed inappropriate insulin secretion. Future studies are likely to be performed with hybrid PET/CT scanners, which enable correlation of anatomical and functional information within one imaging session. In addition, externally triggered respiratory gating may further improve the diagnostic accuracy of PET images in the pancreas. It is expected that the results obtained in the present study will provide extensive and useful information for future investigation. Therefore, 18F-DOPA PET has the potential to become the functional imaging method of the future, once the results reported here are confirmed in a larger patient population.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Disclosure Summary: All authors have nothing to declare.
First Published Online January 16, 2007
Abbreviations: CI, Confidence interval; CT, computed tomography; CV, coefficient of variation; EUS, endoscopic ultrasound; 18F-DOPA, fluorine-18-L-dihydroxyphenylalanine; MRI, magnetic resonance imaging; NET, neuroendocrine tumor; PHH, persistent hyperinsulinemic hypoglycemia; PET, positron emission tomography; SRS, somatostatin receptor scintigraphy; SUV, standardized uptake values.
Received July 11, 2006.
Accepted January 5, 2007.
|
References |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  P. L. Jager, R. Chirakal, C. J. Marriott, A. H. Brouwers, K. P. Koopmans, and K. Y. Gulenchyn 6-L-18F-Fluorodihydroxyphenylalanine PET in Neuroendocrine Tumors: Basic Aspects and Emerging Clinical Applications J. Nucl. Med., April 1, 2008; 49(4): 573 - 586. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 O. T. Hardy, M. Hernandez-Pampaloni, J. R. Saffer, J. S. Scheuermann, L. M. Ernst, R. Freifelder, H. Zhuang, C. MacMullen, S. Becker, N. S. Adzick, et al. Accuracy of [18F]Fluorodopa Positron Emission Tomography for Diagnosing and Localizing Focal Congenital Hyperinsulinism J. Clin. Endocrinol. Metab., December 1, 2007; 92(12): 4706 - 4711. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
