[11C]Metahydroxyephedrine and [18F]Fluorodeoxyglucose Positron Emission Tomography Improve Clinical Decision Making in Suspected Pheochromocytoma
Pheochromocytomas are rare tumors of chromaffin cells for which the optimal management is surgical resection. Precise diagnosis and localization may be elusive. We evaluated whether positron emission tomography (PET) scanning with the combination of [18F]fluorodeoxyglucose (FDG) and the norepinephrine analogue [11C]metahydroxyephedrine (mHED) would allow more exact diagnosis and localization.
Fourteen patients with suspected pheochromocytoma were evaluated by anatomical imaging (computed tomography or magnetic resonance imaging) and [131I]metaiodobenzylguanidine (MIBG) planar imaging. PET imaging was performed by using mHED with dynamic adrenal imaging, followed by a torso survey and FDG with a torso survey. Images were evaluated qualitatively by an experienced observer.
Eight patients had pathology-confirmed pheochromocytoma. Of the other six, two patients had normal adrenal tissue at adrenalectomy, and the other four had subsequent clinical courses inconsistent with a diagnosis of pheochromocytoma. In four of eight patients with pheochromocytoma, MIBG failed to detect one or more sites of pathology-confirmed disease. The mHED-PET detected all sites of confirmed disease, whereas FDG-PET detected all sites of adrenal and abdominal disease, but not bone metastases, in one patient. MIBG and FDG-PET results were all negative in the six patients without pheochromocytoma. One patient with adrenal medullary hyperplasia had a positive mHED-PET scan. PET scanning aided the decision not to operate in three of six patients. The resolution of PET functional imaging was superior to that of MIBG.
PET scanning for pheochromocytoma offers improved quality and resolution over current diagnostic approaches. PET may significantly influence the clinical management of patients with a suspicion of these tumors and warrants further investigation.
KeywordsPheochromocytoma Metahydroxyephedrine Fluorodeoxyglucose Positron emission tomography Adrenal surgery Laparoscopy
Accurate diagnosis and localization of pheochromocytomas is of the utmost importance. Pheochromocytomas are rare tumors of chromaffin cells that secrete excessive catecholamines. They are one of the few surgically treatable causes of hypertension, and, if they are unrecognized, fatal cardiovascular and cerebrovascular sequelae can ensue. However, because of their low prevalence, the frequently episodic nature of tumor catecholamine secretion, and the difficulty encountered in anatomical localization, diagnosis and appropriate treatment can often be delayed, with potentially disastrous consequences for the patient.
Diagnosis of pheochromocytoma relies on the biochemical demonstration of excessive catecholamines or their metabolites in urine, blood, or both.1 Because treatment is surgical resection, once a clear determination of excessive catecholamine production has been made, localization of the tumor is required. Pheochromocytomas have been defined in surgical dogma by the “rule of tens.” Ten percent of tumors have been traditionally believed to be bilateral, extra-adrenal, familial, or malignant. These numbers have recently been disputed.2,3 Nevertheless, they were the motivation for the past surgical approach of an extensive exploratory celiotomy with exploration of both adrenal glands, the para-aortic and para-caval chains, the region of the organ of Zuckerkandl, the base of the bladder, and the pelvis.
With the advent of laparoscopic adrenalectomy, the playing field has been dramatically altered. The laparoscopic approach is a completely directed one, and accurate localization of pheochromocytoma, whether unilateral or bilateral, adrenal or extra-adrenal, has become even more important. Currently, the standard approach is anatomical imaging with computed tomography (CT), magnetic resonance imaging (MRI), or both, coupled with functional imaging by using planar metaiodobenzylguanidine (MIBG) scanning. Single-photon emission computed tomography (SPECT) MIBG scanning has excellent specificity but is still hampered by a lack of resolution, imperfect sensitivity, and subjective poor image quality. Some have argued that in patients with a firm biochemical diagnosis of pheochromocytoma and a unilateral adrenal mass on CT or MRI, a functional study may be unnecessary.4
Positron emission tomography (PET) scanning, usually with [18F]fluorodeoxyglucose (FDG), is becoming a more common imaging modality in oncology and even in the imaging of endocrine tumors. It has shown value in the localization and prognostic determination of thyroid cancer.5 PET scanning allows accurate qualitative functional imaging with excellent image resolution. It also allows quantitation of results. Several groups have investigated its role in the management of pheochromocytoma by using a variety of tracers.6, 7, 8, 9, 10, 11 [11C]Metahydroxyephedrine (mHED) is a norepinephrine analogue and was the first positron-emitting probe of the sympathoadrenal system administered in humans.12,13 As such, it holds promise for imaging catechol-secreting tumors such as pheochromocytoma and neuroblastoma.14 FDG-PET has been of varying utility for pheochromocytoma; however, some studies have suggested that malignant pheochromocytomas have enhanced visualization compared with nonmalignant tumors.9 We undertook this study to determine whether PET, by using a combination of FDG and mHED in confirmed or suspected pheochromocytoma, would allow more definitive diagnosis or improved localization. Our hypothesis was that the combination of a specific agent (mHED) and a nonspecific marker associated with more aggressive tumors (FDG) would be able to localize pheochromocytoma with a broad range of biologic characteristics and that the biological characteristics of the tumors would be reflected by the level of uptake of the two tracers.
MATERIALS AND METHODS
Studies were performed under the auspices of a National Institutes of Health grant with the approval of the Human Subjects Division of our institutional review board and our radiation safety committee. Written informed consent was obtained from all subjects.
At least one 24-hour urine analysis for fractionated catecholamines, metanephrines, or both was performed on each of the subjects before imaging studies. The choice of laboratory for these studies was at the referring physician’s discretion. Ten of the 14 subjects also had plasma metanephrine studies (metanephrine and normetanephrine). When performed, these were performed by the Mayo Medical Laboratories (the normal range for metanephrine is <.50 nmol/L, and for normetanephrine it is <.90 nmol/L). Four patients also had serum catecholamine studies; however, these had a role in the decision to proceed with PET studies in only one subject (patient 12).
All 14 patients underwent abdominal CT scan, with or without contrast, and/or contrast-enhanced MRI within 3 months of PET imaging. If patients were referred from outside institutions with adequate anatomical images of the adrenal glands, the studies were not repeated. CT scans at our institution were performed by using spiral scanners with slice spacing of 5 mm. All subjects received noncontrast studies, and some subjects in whom diagnoses other than pheochromocytoma were suspected also underwent scans after intravenous contrast injection. In addition to CT, eight subjects underwent MRI, which included T1- and T2-weighted images before contrast and a T1-weighted image after intravenous gadolinium-diethylene-triamine-pentaacetic acid contrast, acquired in both the axial and coronal planes. CT and MRI scans underwent prospective clinical interpretation by board-certified radiologists without information from the PET studies. The CT and MRI scans were also used to provide correlative anatomical information in the interpretation of the PET studies.
FDG was made by the Hamacher method.15 The [11C]mHED was made according to the method of Rosenspire et al.13 The [11C]mHED product was purified by semipreparative high-performance liquid chromatography by using an Inertsil ODS-2 C18 column (Metachem, Torrance, CA) 10 × 250 mm, 5-μm particles, eluted with .15 M of phosphate-buffered saline (USP):ethanol (USP) (95:5 v/v). Chemical and radiochemical purity were assessed by using a C18 column (betabasic 2.1 × 150 mm) eluted with .8 mM of formate (pH 7):methanol (95:5 v/v) with radioactivity and mass spectral (ES+) detection. The [11C]mHED was >95% pure with <10 μg of mHED per injection.
MIBG scintigraphy was performed after an injection of .5 mCi of [131I]MIBG, followed by whole-body planar imaging by using a dual-head gamma camera (VG SPECT Tomograph; General Electric Medical Systems, Waukesha, WI) at 24 and 48 hours. Acquisition was performed by using a continuous sweep at 5 cm/min. To prevent thyroid uptake of radioactive iodine and subsequent dysfunction, Lugol’s solution was administered as three drops in water or juice three times daily (or saturated solution of potassium iodide, one drop in water or juice three times a day). This was begun one day before dosing and continued for 15 days afterward. Laxatives were given between the 24- and 48-hour images for bowel clearance.
All PET studies were performed by using a two-dimensional acquisition mode (Advance PET tomograph; General Electric Medical Systems). In most instances, [11C]mHED and [18F]FDG PET studies were performed in a single imaging session to provide precise coregistration. In three patients, FDG and mHED PET studies were performed on consecutive days because of scheduling constraints. To clear the urinary background, a Foley catheter was placed before imaging. Intravenous fluids (500 mL of normal saline) were given to ensure proper hydration and urine flow. For the FDG study, in which renal uptake and urinary excretion of tracer cause a more significant background, intravenous furosemide (20 mg) and dilution of the bladder contents with sterile water were used to clear renal and urinary FDG. Patients were encouraged to undergo conscious sedation with oral lorazepam to improve comfort.
The mHED scan consisted of early dynamic imaging over the adrenal glands, followed by a torso survey from the pelvis to the upper neck, covering the potential sites of extra-adrenal pheochromocytoma. On the General Electric Advance tomograph, this typically required five bed positions. The adrenal bed position was identified by reference to the diaphragm on the transmission study. After the positioning scans, short transmission scans of each bed position were acquired by using rotating germanium 68 rod sources. [11C]mHED .2 mCi/kg, up to a maximum of 20 mCi, was infused over 1 minute. The initial 25 minutes consisted of dynamic imaging at the level of the adrenals. After the dynamic imaging, the torso survey was performed, with increasing acquisition times from 7 to 11 minutes per bed position to account for 11C decay. The total patient scan time did not exceed 65 minutes.
After the completion of the mHED study, a 1-minute infusion of 10 mCi of [18F]FDG was administered. Dynamic imaging was again performed over the adrenals. The patient remained on the imaging table to maintain alignment with the mHED scan. Starting at 45 minutes after injection, a torso survey was performed, with 7-minute emission scans per bed position. This was followed by 3-minute postinjection transmission studies per field of view. The total imaging time was 175 minutes.
After correction for scattered and random coincidences, PET projection data were reconstructed by using filtered back-projection and a Hanning filter onto a 128 × 128 × 35 matrices per 15 cm axial field of view. The approximate reconstructed image spatial resolution was 10 mm. For viewing, dynamic data were summed from 15 to 25 minutes for mHED and 30 to 60 minutes for FDG.
Data and Statistical Analysis
The fraction of pathology-proven pheochromocytoma detected by MIBG scintigraphy and mHED and FDG PET was calculated. The SUVs for adrenal foci proven to be pheochromocytoma versus those without pathology confirmation were compared and assessed for significant differences between groups by using Student’s t-test. P < .05 was considered significant.
Clinical and imaging characteristics of subjects enrolled in the PET imaging protocol
Urinary/plasma catechols or metanephrines
Anatomical imaging (cm)
R = 14; L = 0.8
B lap adr
HTN, HA, subarach. hem.
R = 14; AoC = 6
L = 7.1
Lap to open adr
± ↑/↑; lateralizing adr vein sampling
R = 2
L = 6
R = 2.9
HTN, Hx of pheo
BM biopsy, chemo
R = 1.5
R = 2.5
R = 4
Equivocal (postop L adrenal)
Lap to open R adr
Initial pheo only, no recurrence
HTN, chest pain
N/-(↑ serum catechols)
R = 9
HTN, Hx of pheo
Pheo with lung mets
HTN, MEN 2A
R = 6; L = 5
B open adr
Urine biochemical evaluation was positive in all eight patients with pheochromocytoma. Abnormally increased urine metanephrines were present in seven (1.1- to 33.0-fold the upper limit of normal [ULN]), and abnormally increased normetanephrine was present in five (1.2- to 16.0-fold the ULN). One patient had an elevation of 3-methoxytyramine as the only urine abnormality but had abnormal plasma normetanephrines. Seven of the eight patients with pheochromocytoma also had plasma metanephrine studies. Normetanephrine was increased in all patients (range, 2.27–106 nmol/L); metanephrine was increased in three of the seven.
Four of the six patients without pheochromocytoma had definite or borderline increases of their urinary catecholamines, metabolites, or both, with 24-hour metanephrines 1.1- to 1.8-fold the ULN and normetanephrines 1.9- to 5.0-fold the ULN. The remaining two had normal levels. Plasma metanephrines were obtained in three of six patients. Metanephrines were normal in all, whereas normetanephrine ranged from 1.28 to 3.71 nmol/L.
Six patients without pheochromocytoma, and with no history of pheochromocytoma, were imaged. Two of these patients underwent adrenalectomy on the basis of clinical suspicion and underwent further work-up. One had mild abnormalities of urine metanephrines, increased plasma normetanephrine of 3.71 nmol/L, negative MIBG and PET imaging, and lateralizing adrenal vein sampling (patient 4). The second had only mild abnormalities of urine and plasma metanephrines and equivocal functional imaging by MIBG and FDG-PET but was believed to have a positive mHED-PET (patient 9). She underwent a laparoscopic adrenalectomy on the basis of clinical symptoms, biochemical data, and imaging results and did not have a pheochromocytoma. It is interesting to note that the pathologic tissue examination demonstrated adrenal medullary hyperplasia. One subject (patient 11) had previously undergone adrenalectomy for pheochromocytoma before our PET imaging protocol becoming available. During follow-up, fluctuating increases of urinary catecholamines were suggestive of recurrence. MIBG scanning showed equivocal uptake in the lungs, and FDG-PET showed mildly increased uptake in the remaining (left) adrenal, but mHED-PET was completely negative. The remaining three patients (patients 7, 10, and 12) had equivocal or increased biochemistry results and negative functional imaging, including PET. Subsequent clinical follow-up for all four subjects has suggested that these patients do not have pheochromocytoma. In these patients, the functional imaging, particularly mHED-PET, was key in the decision not to operate.
Quantification of Functional Imaging
In this prospective study of a novel functional imaging approach to patients with suspected pheochromocytoma, we found PET scanning to be very useful clinically. The identification of pheochromocytoma can often be elusive, given its low prevalence together with vague and nonspecific symptoms and signs. Definitive diagnosis relies on increases in urine and/or plasma catecholamines and their metabolites.1 However, increased catecholamines are not specific to pheochromocytoma, and, hence, false-positive results are a significant problem. In patients with unequivocal pheochromocytoma, localization is the next step, so that surgical resection can be undertaken after careful preoperative hydration and adrenergic blockade. Despite initial concerns about the safety of the laparoscopic approach to resection of pheochromocytoma, several reports have shown this technique to be safe and effective, with good outcomes, although for most series follow-up is relatively limited.16, 17, 18
We found all patients with biopsy-proven pheochromocytoma to have uptake on mHED-PET scanning. Pheochromocytoma was more accurately detected with our PET imaging protocol compared with standard anatomical imaging and [131I]MIBG. Bilateral and extra-adrenal disease were detected in several patients by PET but not by anatomical imaging or MIBG scintigraphy. This had important beneficial clinical consequences for these patients. Even in the one case that we scored as a false positive on mHED-PET, final pathologic analysis revealed adrenal medullary hyperplasia that potentially contributed to the increased mHED uptake. This is an as-yet unproven interpretation.
There are limitations to this study. Patient numbers were small and, as such, are insufficient to calculate diagnostic sensitivity and specificity. Our numbers are, however, comparable to those in other studies that have evaluated PET scanning of pheochromocytoma by using several radiopharmaceuticals. This study looked at all patients undergoing PET scanning for suspected pheochromocytoma, including some who did not, in fact, have a pheochromocytoma. As noted in our study, even some patients without a pheochromocytoma had increases in their catecholamines or metabolites, thus making its exclusion difficult at times. Forty-three percent (6 of 14) of patients in this study did not have pheochromocytoma at the time of imaging. An important finding was that PET imaging was very useful in helping to exclude pheochromocytoma in most of these patients, thus avoiding surgery or potentially morbid therapy. Patient follow-up is limited (<3 years); therefore, rates of recurrence or subsequent pheochromocytoma development may be an underestimation. All patients tolerated PET scanning well.
Two of the major issues with novel imaging agents are the availability of the agent and the generalizability of the results. [11C]mHED has a short half-life (20 minutes) and requires an on-site cyclotron, which limits its widespread use. There are several centers in the United States with such a capacity, and regionalized imaging of these tumors is possible. More important is the proof of concept that PET imaging with a combination of radiopharmaceuticals may improve detection of pheochromocytoma and possibly help to predict its behavior. Although the half-life of 11C limits its more generalized application, its use does allow single-session imaging with other 18F-based radiotracers, such as FDG, because of their longer half-lives. Conceivably, this dual study may allow improved detection and localization in difficult cases, although this would admittedly be the exception. What is compelling is the potential use of combined agents in helping to predict behavior and outcome in a fashion analogous to what is seen in thyroid cancer.5 Because the malignant phenotype of pheochromocytoma is predicated on its biology, i.e., recurrence or metastasis, and not its histological characteristics, a secondary goal is to ultimately see whether the combination of tracers will allow us to better predict the behavior of pheochromocytoma. The use of two 18F-based tracers would necessitate multiple imaging days, thus obviating one of the benefits of dual tracer imaging with 11C and 18F radioisotopes. An adequate test of this hypothesis requires larger patient numbers, with more long-term follow-up, and active further study of this is ongoing. Only one study has evaluated the use of FDG-PET in imaging benign and malignant pheochromocytoma.9 Even though more malignant pheochromocytomas than benign pheochromocytomas were detected, quantitative SUVs between the two biologically different tumors were similar.
Two other small series have reported the use of [11C]mHED in imaging of pheochromocytoma. In an initial series, using an earlier PET tomograph with relatively thick image slices, Shulkin et al.10 were able to localize pheochromocytoma in 9 of 10 patients who were imaged. Compared with [131I]MIBG, PET scanning detected more lesions with greater contrast and in a much more timely fashion. The single false-negative result was in a patient whose tumor secreted dopamine and little norepinephrine or epinephrine, in keeping with mHED’s mechanism of action, which reflects catecholamine uptake and storage, rather than synthesis. In a more contemporary series with 19 patients, Trampal et al.11 found equally impressive results with detecting pheochromocytoma by using [11C]mHED. Sensitivity was found to be 92%, specificity was 100%, and accuracy was 95%. The single patient with a false-negative study had a tumor that secreted both norepinephrine and epinephrine, unlike the patient in the Shulkin study. Our study had no false negatives and only a single false positive (in the patient with adrenal medullary hyperplasia).
Other PET radiotracers have been used to image and localize pheochromocytoma, most of them labeled with 18F. Pacak et al.7 demonstrated a high positive and negative predictive value of PET scanning of pheochromocytoma with [18F]fluorodopamine, noting positive scans in all 17 patients with pheochromocytoma and negative scans in 9 of 11 patients without pheochromocytoma. A follow-up study by the same authors of 16 patients with metastatic pheochromocytoma confirmed the superiority of [18F]fluorodopamine to [131I]MIBG; that study found disease in more patients and in more sites.19 In a series of 14 patients, Hoegerle et al.6 found that the novel radiotracer [18F]dihydroxyphenylalanine had a higher sensitivity than [123I]MIBG for detecting pheochromocytoma (100% vs. 71%). These results await confirmation in a larger series and more widespread availability of these radiotracers.
There are many advantages of PET scanning over [131I]MIBG. PET exposes patients to lower radiation doses than [131I]MIBG. Because PET uses no iodine, there are no adverse effects on the thyroid, and thyroid blocking is not required. Patient throughput is significantly faster with PET, because scans can be obtained immediately or within a few hours of administration of the analogue. [131I]MIBG requires delayed scans at 24, 48, and sometimes 72 hours. Finally, and importantly, PET scanning is superior in its spatial resolution, as opposed to the planar imaging of [131I]MIBG. Improved-quality images with [123I]MIBG are possible,20 but image resolution for SPECT is inferior to that for PET, and image quantification, which was important in our study, is not accurate with SPECT.
In conclusion, this study demonstrates the feasibility and usefulness of FDG-PET and especially mHED-PET scanning in the high-quality efficient detection of pheochromocytoma. Results suggest that PET imaging may play a crucial role in surgical approach and outcome. In certain patients, these techniques may also prove invaluable in excluding the diagnosis of pheochromocytoma and, hence, avoiding unnecessary surgery. PET scanning with specific tracers for the localization of pheochromocytoma may well become the future functional imaging modality of choice.
The authors thank Joanne Wells for help with the time course data and Dr. James Caldwell for the initial studies that allowed comparison with normal adrenals. Supported by National Institutes of Health grants P01 CA042045 and RO1 HL50239.