European Journal of Nuclear Medicine and Molecular Imaging

, Volume 37, Issue 3, pp 484–493

Clinical value of 18F-fluorodihydroxyphenylalanine positron emission tomography/computed tomography (18F-DOPA PET/CT) for detecting pheochromocytoma

Authors

    • Department of Nuclear MedicineUniversity of Ulm
  • Wolfram Karges
    • Division of Endocrinology and DiabetesRWTH Aachen
  • Katrin Zeich
    • Department of Nuclear MedicineUniversity of Ulm
  • Sandra Pauls
    • Department of RadiologyUniversity of Ulm
  • Frederik A. Verburg
    • Department of Nuclear MedicineUniversity of Würzburg
  • Henning Dralle
    • Department of General, Visceral and Vascular SurgeryUniversity Halle-Wittenberg
  • Gerhard Glatting
    • Department of Nuclear MedicineUniversity of Ulm
  • Andreas K. Buck
    • Department of Nuclear MedicineUniversity of Ulm
  • Christoph Solbach
    • Department of Nuclear MedicineUniversity of Ulm
  • Bernd Neumaier
    • Department of Nuclear MedicineUniversity of Ulm
    • Section for RadiochemistryMax-Planck-Institut für neurologische Forschung
  • Sven N. Reske
    • Department of Nuclear MedicineUniversity of Ulm
  • Felix M. Mottaghy
    • Department of Nuclear MedicineUniversity of Ulm
    • Department of Nuclear MedicineRWTH Aachen
Original Article

DOI: 10.1007/s00259-009-1294-7

Cite this article as:
Luster, M., Karges, W., Zeich, K. et al. Eur J Nucl Med Mol Imaging (2010) 37: 484. doi:10.1007/s00259-009-1294-7
  • 272 Views

Abstract

Purpose

In detecting pheochromocytoma (PHEO), positron emission tomography (PET) with the radiolabelled amine precursor 18F-fluorodihydroxyphenylalanine (18F-DOPA) offers excellent specificity, while computed tomography (CT) provides high sensitivity and ability to localize lesions; therefore, the combination of these modalities could be advantageous in this setting. The aim of this study was to investigate whether combined 18F-DOPA PET/CT more accurately detects and localizes PHEO lesions than does each modality alone.

Methods

18F-DOPA PET, CT and 18F-DOPA PET/CT images of 25 consecutive patients undergoing diagnostic scanning of suspected sporadic or multiple endocrine neoplasia type 2 syndrome-associated PHEO were reviewed retrospectively in randomized sequence. Two blinded observers scored the images regarding the likelihood of PHEO being present and localizable. Results were correlated with subsequent clinical history and, when available, histology.

Results

Of the 19 lesions detected by all three modalities, PET identified each as positive for PHEO, but was unable to definitively localize 15 of 19 (79%). CT could definitively localize all 19 lesions, but could not definitively diagnose or exclude PHEO in 18 of 19 (95%) lesions. Furthermore, CT falsely identified as negative for PHEO one lesion which was judged to be positive for this tumor by both PET and PET/CT. Only in PET/CT scans were all 19 lesions accurately characterized and localized. On a per-patient basis, the sensitivity of 18F-DOPA PET/CT for PHEO was 100% and the specificity 88%, with a 100% positive predictive value and an 88% negative predictive value.

Conclusion

18F-DOPA PET/CT more accurately diagnoses and localizes adrenal and extra-adrenal masses suspicious for PHEO than do 18F-DOPA PET or CT alone.

Keywords

PheochromocytomaMultiple endocrine neoplasia type 2 (MEN2)18F-fluorodihydroxyphenylalanine (18F-DOPA)Positron emission tomography (PET)Computed tomography (CT)Staging

Introduction

Pheochromocytoma (PHEO) is a catecholamine-producing neuroendocrine tumor arising from adrenal and, more rarely, extra-adrenal chromaffin cells. Approximately 90% of PHEOs are sporadic, and the remaining 10% occur as part of hereditary tumor syndromes including multiple endocrine neoplasia type 2 (MEN2) [1]. To prevent morbidity from hormonal excess caused by neoplastic norepinephrine or epinephrine secretion, complete surgical resection is the therapy of choice for benign or malignant PHEOs [2].

Conventional morphological imaging, including computed tomography (CT), detects PHEOs with 93–100% sensitivity, as most of these tumors exceed 2 cm in diameter [3]. However, CT has a low specificity for PHEOs, since morphological imaging cannot distinguish these tumors from other types of adrenal masses.

The 123I-labelled adrenaline precursor metaiodobenzylguanidine (mIBG) or 131I-labelled mIBG are traditional tracers of choice for imaging neuroendocrine tumors [48]. Unfortunately, the gamma-emitting nature of these radiotracers restricts their use to single photon emission computed tomography (SPECT), which, because of limited spatial resolution, has low sensitivity, especially with respect to smaller lesions (<1 cm). Moreover, the relatively high uptake of mIBG by healthy adrenomedullary tissue reduces the specificity of these tracers for PHEO.

Of interest, neuroendocrine tumors are able to take up, decarboxylate and store amino acid precursors, allowing the use of tracers derived from these substances, such as 18F-fluorodihydroxyphenylalanine (18F-DOPA) [9]. While initially developed for neurological imaging [10, 11], 18F-DOPA was shown to accumulate in a larger percentage of neuroendocrine tumors than do somatostatin analogues or 18F-fluorodeoxyglucose (FDG) [12]. Their relatively abundant dihydroxyphenylalanine decarboxylase expression theoretically makes PHEO cells excellent targets for imaging using 18F-DOPA. A study by Hoegerle et al. [13] showed that 18F-DOPA positron emission tomography (PET) had a higher spatial resolution and a more selective, clearer radiotracer accumulation in PHEOs than did 123I-mIBG SPECT. The sensitivity and specificity of 18F-DOPA PET were 95–99%. An additional advantage of 18F-DOPA PET is that due to the high tissue selectivity of 18F-DOPA, this imaging procedure takes much less time than does 123I-mIBG SPECT: well within 4 h versus 24 h.

Given the complementary advantages of these imaging modalities in PHEO, i.e., high specificity with 18F-DOPA PET and morphological information and high sensitivity with CT, combined 18F-DOPA PET and CT could prove clinically useful in detecting and localizing PHEO, especially the malignant variety. We therefore conducted a retrospective analysis comparing the value of these two imaging modalities together versus that of either modality alone in diagnosing or excluding and in localizing PHEO in patients suspected to have that condition.

Materials and methods

Patients

Twenty-five consecutive patients [16 women, 9 men, mean ± standard deviation (SD) age: 45 ± 13 years, range: 12–68 years] who were referred to the University of Ulm Nuclear Medicine Department between 1 July 2002 and 30 November 2005 for imaging of suspected PHEO were included in the analysis. The clinical characteristics of these patients are summarized in Table 1. Suspicion of newly presenting or recurrent PHEO, either sporadic (n = 13 patients) or related to MEN2A syndrome (n = 11 patients) or MEN2B syndrome (n = 1 patient), was based on clinical findings, hormonal analyses or both.
Table 1

Summary of patient demographics, history and findings of PHEO

Pat. No.

Sex

Age (years)

Time since PD (years)

ID/R

PHEO localization

Maximum lesion diameter (cm)

18F-DOPA PET/CT diagnosis

Correlation with histopathological/clinical results

1

M

55

0

ID

Right AG

1.1

PHEO

TP

2

F

54

0

ID

Left AG

3.3

PHEO

TP

3

F

42

0

ID

Right AG

5.0

PHEO

TP

4

F

39

0

ID

Left AG

2.2

PHEO

TP

5

F

68

0

ID

Left AG

3.8

PHEO

TP

6

M

54

0

R

Left AG

2.1

PHEO

TP

7

F

41

0

ID

-

 

No PHEO

TN

8

F

56

1

R

-

 

No PHEO

TN

9

M

37

6

R

Left AG

0.8

PHEO

TPa

10

F

25

7

R

Right AG

13.0

PHEO

TP

11

F

48

13

R

Metastases

 

PHEO metastases

TP

12

M

60

10

R

-

  

TN

13

F

54

0

ID

Right AG

4.0

PHEO

TP

14

M

49

0

ID

Right AG

5.6

PHEO

TP

15

M

56

0

ID

-

 

No PHEO

TN

16

M

26

0

ID

Right AG

1.5

Bilateral PHEO

TP

Left AG

2.2

17

F

39

0.1

R

-

 

No PHEO

TN

18

M

46

10

R

-

 

No PHEO

TN

19

F

27

4

R

Left AG

0.4

PHEO

TP

20

F

33

0

R

-

 

No PHEO

TN

21

F

42

0

R

-

 

No PHEO

TN

22

F

43

0

R

Right AG

1.5

Bilateral PHEO

TP

Left AG

1.0

23

F

49

17

R

Right AG

1.8

PHEO

TPa

24

M

12

0

R

-

 

No PHEO

TN

25

F

61

0

R

Left AG

 

PHEO

FPa

18F-DOPA18F-fluorodihydroxyphenylalanine, AG adrenal gland, F female, FP false positive, ID/R initial diagnosis or restaging/diagnostic follow-up, M male, Pat. No. patient number, PD primary diagnosis, PHEO pheochromocytoma, TP true positive, TN true negative

aDiagnosed only on PET/CT

All patients received 18F-DOPA in accordance with the German Federal Law on the Compassionate Use of Medical Substances; all patients or their parents/guardians gave prior written informed consent for the 18F-DOPA PET/CT.

18F-DOPA production

[18F]Fluorine was produced in the Ulm University PETtrace cyclotron (GEMS, Uppsala, Sweden). With CDCl3 used as a solvent, the protected trimethylstannyl precursor (N-formyl-3,4-di-tert-butoxycarbonyloxy-6-[trimethylstannyl]-L-phenylalanine ethyl ester; ABX Advanced Biochemical Compounds, Radeberg, Germany) was reacted in a commercial automatic system (FX-FDOPA, GEMS, Münster, Germany) with the 18F2 at 5°C. A deprotection step was then performed using 5 M HCl according to a synthesis procedure adapted from that published by de Vries et al. [14].

The 18F-DOPA-containing fraction was separated from the reaction mixture using a semipreparative ultraviolet/radio-high-performance liquid chromatography system followed by sterile filtration. After quality control, the product solution was ready for intravenous (i.v.) injection. Radiochemical purity always was >95%; total synthesis time was 50 min from the end of irradiation.

PET/CT imaging

One hour before tracer injection, patients were given 100 mg carbidopa to increase tracer accumulation in the lesion [15, 16]. Subsequently, patients received an i.v. injection with a mean ± SD 309 ± 53 MBq (range: 208–399 MBq) 18F-DOPA and underwent CT followed by PET. All imaging was performed using an integrated PET/CT camera (Discovery LS, GE Healthcare Medical Systems, Munich, Germany) equipped with a full-ring dedicated PET scanner and a four-slice CT scanner [17].

CT scans were acquired with a 2.5-mm slice thickness and a rotational speed of 0.5 s/rotation, using 140 keV X-rays with a 290 mA current. Scanning speed was 15 mm/s. During CT scanning, a water-soluble CT contrast agent (Schering Ultravist 300®, Bayer Schering Pharma, Leverkusen, Germany) was administered intravenously with an automated injector (MEDRAD®, EnVision CT™, Warrendale, PA, USA) using a biphasic contrast injection protocol, resulting in a venous phase CT acquisition. CT images were reconstructed with the smoothing “standard” and the edge-enhancing “bone” kernels.

The PET scanner was equipped with a 14.6-cm axial field of view. Imaging time was 4 min per bed position. Patients were scanned from the top of the skull to the inguinal region. Images were acquired in a 128 × 128 matrix with a 4.29-mm pixel size. Subsequently, the acquired data set was reconstructed using a two-dimensional ordered subset expectation maximization algorithm with 2 iterations and 28 subsets; attenuation correction was performed using the previously obtained CT scan. Images were processed with a smoothing filter assuming a full-width at half-maximum (FWHM) of 3.91 mm and with a post-filter assuming a FWHM of 6.5 mm.

After processing of stand-alone CT and PET images, a fusion data set with a transaxial slice thickness of 4.25 mm was assembled. Mean and maximum tissue standardized uptake values (SUVmean and SUVmax, respectively) were based on elliptical regions of interest manually drawn around foci of 18F-DOPA uptake.

Study analysis

For the present study, a nuclear medicine physician and a radiologist, both board certified and with lengthy experience in their respective medical specialties, separately classified each lesion on the PET or CT images as well as the integrated PET/CT images regarding the likelihoods that PHEO was present (lesion characterization, LC) and that the lesion could be localized. LC was performed visually, using a 5-point scale of 0 = negative, 1 = probably negative, 2 = ambiguous, 3 = probably positive and 4 = positive for PHEO. No further criteria were defined for each of these scores. However, we defined as PHEO (1) a tumor in the adrenal gland suspected of causing catecholamine excess or (2) a lesion outside the adrenal gland causing such excess, if one or more tumors in the adrenal gland causing catecholamine excess also were present or had been resected and histologically confirmed.

Lesion localization was rated on a 3-point scale of 0 = unknown, 1 = probable and 2 = definite localization, without additional criteria for these scores.

The stand-alone PET images, the stand-alone CT images and the fused PET/CT images each were reviewed using a dedicated computer workstation (eNTEGRA, GE Healthcare Medical Systems, Munich, Germany) in separate sessions at least 7 days apart. The readers were blinded to the identity and history of the patients as well as to the original clinical reports of the scans; for each session, the 25 patients’ images were reviewed in a different, randomized sequence. We chose to re-read the scans for the present study for two reasons. First, we wished to provide a uniform scoring system across imaging modalities in order to compare the modalities with respect to observer-dependent diagnostic certainty. Second, we sought to remove two potential sources of bias that might have been introduced by using the original readings: variability associated with the larger number of original readers and influence of the original readers’ knowledge of the patient(s) on scan interpretation.

For further analysis, the readers’ interpretations in the present study were compared to the post-18F-DOPA PET/CT pathology reports on the histology of surgical/biopsy specimens where such reports were available. If no material was obtained for pathological analysis, the study scan interpretations were compared to the post-18F-DOPA PET/CT clinical endocrinological follow-up and laboratory results. Patients without histological data were considered PHEO negative if throughout the post-imaging endocrinological follow-up no further plasma catecholamine or free metanephrine elevations were encountered; if catecholamine or metanephrine levels were elevated at any time during follow-up, the patient was considered PHEO positive.

Statistical analysis

To identify a possible added value of PET/CT versus PET or CT alone, we applied the McNemar’s test. A p value <0.05 was considered to indicate statistical significance. Statistical analysis was performed using the SPSS 17.0 software package (SPSS Inc., Chicago, IL, USA).

Results

Lesion counts and LC and localization scores

Table 1 summarizes 18F-DOPA PET/CT findings for each patient. In five patients, no pathological accumulation of 18F-DOPA was encountered. In the remaining 20 patients, a total of 27 lesions (of which all were rated as positive or probably positive for PHEO) was found by 18F-DOPA PET. CT revealed 28 lesions (of which 20 were rated as probably positive for PHEO), whereas a total of 40 lesions (of which 33 were rated as positive or probably positive for PHEO) was identified by combined 18F-DOPA PET/CT. Four lesions were only identified by combined imaging. Three of these lesions were solitary and formed the basis of a diagnosis of PHEO; of the three patients involved (identified by a in Table 1), two turned out to be true positives and one a false positive. The fourth lesion identified only by combined imaging was an additional metastasis in an individual (patient 11). already known to have malignant PHEO. Nineteen lesions were found concurrently on all three imaging modalities.

Table 2 categorizes the LC and localization scores of the suspicious lesions identified by each of the three imaging methods. 18F-DOPA PET identified 26 lesions as positive for PHEO and 1 lesion as probably positive. The LC scores for CT showed less diagnostic certainty. Notably, even combined PET/CT scanning did not allow definitive classification of all lesions, as 10 of 40 (25%) were described as probably positive, probably negative or ambiguous. Regarding localization, 18F-DOPA PET alone was only able to clearly situate 5 of 27 (19%) lesions, whereas both the CT scan and the PET/CT scan could definitively localize all lesions (Table 2).
Table 2

Number of lesions and lesion characterization and localization scores for lesions discovered on one or more of 18F-DOPA PET, CT or 18F-DOPA PET/CT

Modality

PET

CT

PET/CT

Characterization score

 4 = positive for PHEO

26

0

29

 3 = probably positive

1

20

4

 2 = ambiguous

0

3

2

 1 = probably negative

0

3

4

 0 = negative

0

2

1

Total no. of lesions

27

28

40

Localization score

 2 = definite localization

5

28

40

 1 = probable localization

22

0

0

 0 = unknown localization

0

0

0

Total no. of lesions

27

28

40

18F-DOPA18F-fluorodihydroxyphenylalanine, CT computed tomography, PET positron emission tomography, PET/CT combined positron emission tomography and computed tomography

Table 3 compares among the imaging modalities the distribution of LC and localization scores for the 19 suspicious lesions detected concurrently by all three scan types. PET identified all 19 as positive for PHEO, but was unable to definitively localize 15 of 19 (79%). The CT scan definitively localized all 19 lesions, but could not definitively diagnose or exclude PHEO in 18 of 19 (95%) lesions. Furthermore, the CT scan falsely identified as negative for PHEO one lesion which was judged to be positive for this tumor by both PET and PET/CT. Only in PET/CT scans were all 19 lesions clearly characterized and localized.
Table 3

Lesion characterization and localization scores for 19 lesions discovered on all of 18F-DOPA PET, CT and 18F-DOPA PET/CT

Modality

PET

CT

PET/CT

Characterization score

 4 = positive for PHEO

19

0

19

 3 = probably positive

0

16

0

 2 = ambiguous

0

2

0

 1 = probably negative

0

0

0

 0 = negative

0

1

0

Total no. of lesions

19

19

19

Localization score

 2 = definite localization

15

19

19

 1 = probable localization

4

0

0

 0 = unknown localization

0

0

0

Total no. of lesions

19

19

19

18F-DOPA18F-fluorodihydroxyphenylalanine, CT computed tomography, PET positron emission tomography, PET/CT combined positron emission tomography and computed tomography

Tracer uptake

For lesions diagnosed in the study readings as definitely PHEOs, the mean ± SD SUVmean was 8.9 ± 4.3 and the mean ± SD SUVmax was 17.0 ± 9.3. 18F-DOPA uptake in normal adrenal glands did not exceed background uptake; we were therefore unable to draw regions of interest to determine the SUV in these glands.

Patients undergoing 18F-DOPA PET/CT for primary diagnosis/exclusion of PHEO

All ten patients who were referred for initial diagnosis/exclusion of PHEO subsequently underwent surgery. In eight of them, PET/CT found an 18F-DOPA-accumulating mass in the adrenal region, which was subsequently histologically confirmed (Table 1). In one patient, an 18F-DOPA-accumulating mass was seen in the cervical region. Histological analysis revealed a paraganglioma; based on the clinical history, this patient was classified as truly negative for PHEO. In the remaining patient, PET/CT scanning identified a mass that did not accumulate 18F-DOPA and histological analysis showed an adrenocortical adenoma (Fig. 1). Thus on a per-patient basis, the sensitivity and specificity of PET/CT for the primary identification of PHEO were each 100%.
https://static-content.springer.com/image/art%3A10.1007%2Fs00259-009-1294-7/MediaObjects/259_2009_1294_Fig1_HTML.gif
Fig. 1

18F-DOPA PET, CT and 18F-DOPA PET/CT images of an adrenocortical adenoma (from patient 12); there is no 18F-DOPA uptake in the enlarged right adrenal gland. The red arrow indicates the lesion of interest

Patients scanned for diagnostic follow-up or restaging

As also seen in Table 1, of 15 patients scanned for diagnostic follow-up or restaging, 7 did not show any 18F-DOPA uptake that was rated as suspicious for PHEO. In one of these seven patients, PET/CT showed an osteolytic lesion that subsequent biopsy revealed to be an adrenocortical carcinoma metastasis. In another of the seven patients, an individual with MEN2A, 18F-DOPA revealed a lesion in the cervical region that was later histologically confirmed to be a medullary thyroid carcinoma recurrence. In a median follow-up of 14 months (range: 1–40 months), no further catecholamine elevations have been observed in these seven patients; therefore, it is likely that 18F-DOPA PET/CT was truly negative for PHEO in all these cases.

In the other 8 of the 15 patients scanned for diagnostic follow-up or restaging, pathological 18F-DOPA accumulations were encountered. In two patients with MEN2 syndrome, a bilateral malignant PHEO was identified (example in Fig. 2) and in two others a unilateral PHEO. All these PHEOs were surgically removed and histologically confirmed. Two additional patients showed histologically confirmed local PHEO recurrence. In yet another patient, foci of pathological 18F-DOPA uptake represented additional metastatic dissemination of a previously resected malignant PHEO (Fig. 3). In the last of the eight patients, an adrenal mass with a mildly intense 18F-DOPA accumulation (size: 16 mm, SUVmean: 1.8, SUVmax: 2.9) was judged to be positive, but clinical follow-up thus far revealed no further evidence of PHEO. This case thus appears to be a false positive (Fig. 4).
https://static-content.springer.com/image/art%3A10.1007%2Fs00259-009-1294-7/MediaObjects/259_2009_1294_Fig2_HTML.gif
Fig. 2

CT, 18F-DOPA PET (coronal view) and 18F-DOPA PET/CT images of a bilateral pheochromocytoma (red arrows) (from patient 16)

https://static-content.springer.com/image/art%3A10.1007%2Fs00259-009-1294-7/MediaObjects/259_2009_1294_Fig3_HTML.gif
Fig. 3

Whole-body 18F-DOPA PET scan (of patient 12) with widely metastatic pheochromocytoma (left, red arrows indicate lesions) with CT (upper right) and 18F-DOPA PET/CT (lower right) images of a bone metastasis (indicated by red arrows)

https://static-content.springer.com/image/art%3A10.1007%2Fs00259-009-1294-7/MediaObjects/259_2009_1294_Fig4_HTML.gif
Fig. 4

18F-DOPA PET/CT(coronal view) image of a false-positive lesion (in patient 25)

Therefore, on a per-patient basis, sensitivity of 8F-DOPA PET/CT was 100%, specificity was 88%, negative predictive value (NPV) was 88% and positive predictive value (PPV) was 100% in individuals scanned for follow-up or restaging.

Overall analysis

For our overall series, based on correlation with pathology reports and clinical outcome data per discovered lesion, sensitivity of 18F-DOPA PET for PHEO was 90% and the specificity was 93%, with a PPV of 81% and an NPV of 89%. Sensitivity of CT for PHEO was 66% and the specificity was 81%, with a PPV of 91% and an NPV of 45%. The overall sensitivity of 18F-DOPA PET/CT for PHEO was 100% (95% confidence interval: 78–100%) and the specificity was 90% (95% confidence interval: 55–100%), with a PPV of 91% and an NPV of 100%.

Additional value of PET/CT

McNemar’s test showed that combined 18F-DOPA PET/CT imaging had a clear additional relevance over stand-alone CT imaging (p significant at the <0.05 level); the same held true for 18F-DOPA PET imaging versus stand-alone CT imaging. However, 18F-DOPA PET/CT imaging did not show a statistically significant additional clinical relevance over stand-alone 18F-DOPA PET imaging.

Discussion

The present study shows that integrated 18F-DOPA PET/CT scanning is superior in the clinical detection of PHEO compared to each of these modalities by themselves. Obvious advantages of PET/CT are a better sensitivity for PHEO as well as a more precise anatomical localization of lesions than is seen with CT or PET alone. Especially important is the ability of 18F-DOPA PET/CT to immediately provide information about the nature of a lesion seen outside the adrenal glands in patients with a primary PHEO.

To our knowledge, this is the first published study to compare the integrated versus the stand-alone use of 18F-DOPA PET and CT for detecting PHEO. Our finding of added value for the combination of these imaging methods versus either method alone resembles observations by Beheshti et al. [18]. These investigators studied the same three modalities as well as PET with another tracer, 18F-FDG, in the detection of another neuroendocrine tumor, medullary thyroid cancer. Additionally, the 100% sensitivity and 89% specificity of 18F-DOPA PET/CT in detecting PHEO in our study are in line with the 95–99% sensitivity and specificity of 18F-DOPA PET/CT in the same setting that were observed by Hoegerle et al. in their comparison of 18F-DOPA PET versus 123I-mIBG SPECT [13]. Several other studies showed a highly selective 18F-DOPA accumulation in PHEO lesions, with little physiological uptake in healthy adrenal tissue [12, 19, 20], suggesting that even mildly elevated 18F-DOPA accumulation should be considered pathological. This assumption was, however, disproven to a degree in the present study, as a slightly enlarged adrenal gland showing an 18F-DOPA uptake exceeding the background turned out to be false positive. Further investigation in larger patient groups seems warranted to clearly delineate the upper limit of physiological 18F-DOPA uptake. However, the accumulation of 18F-DOPA in PHEOs is certainly far superior to and more selective than that of 123I-mIBG [21]. With the use of 123I-mIBG SPECT, a difference in radiotracer uptake intensity between the left and right adrenal glands is the most important diagnostic criterion, which leads to problems, especially in the case of bilateral PHEO [13] or after unilateral adrenalectomy.

Even though sensitivity of anatomical imaging modalities such as CT and magnetic resonance imaging (MRI) in detecting PHEO also lies at 98–100%, the frequent findings of incidental adrenal masses (“incidentalomas”) of non-adrenomedullary origin with these methods lead to a specificity which at best lies around 70%. The added value of combining anatomical and functional imaging is especially apparent in the case of such incidentalomas. The present study included two patients in whom 18F-DOPA PET suggested that adrenal masses were of non-adrenomedullary origin, a finding that was later histologically confirmed. Correct localization of positive 18F-DOPA PET findings by means of concurrent CT imaging also facilitates a correct diagnosis, especially in the case of metastatic disease or in patients with MEN2 syndrome, who often simultaneously present with an 18F-DOPA-accumulating medullary thyroid carcinoma.

The present study has certain limitations. The rarity of PHEO resulted in inclusion of a relatively low number of patients. The disadvantage of small patient cohorts is exemplified by a single false-positive case causing the specificity to drop from 100% to 90%. Based on the experience of others with 18F-DOPA PET/CT [13], this value is not unlikely an underestimation of specificity. The small study sample size may also have resulted in insufficient power of the McNemar’s test to find a statistically significant advantage in clinical relevance for 18F-DOPA PET/CT versus 18F-DOPA PET alone.

Another limitation of our study was its retrospective nature, which forced us to rely on clinical follow-up data from several patients who did not undergo surgery. Even though catecholamine levels are the primary and most sensitive markers for PHEO and remained normal throughout follow-up for all our 18F-DOPA PET/CT-negative patients, it is conceivable that a clinically detectable PHEO may still develop later on.

Noteworthy in the present study is the relatively high number of lesions (10/40 or 25%) which were not classified with certainty even with PET/CT. The exact cause of this high degree of uncertainty is not entirely clear; it might be due to an observer bias towards reading the scans specifically rather than sensitively. Such a bias may have arisen because the readers, although blinded to individual patient data, were aware that the series had a high prevalence of PHEO.

Alternatives to 18F-DOPA imaging may become more widely available. Such approaches include 6-18F-fluorodopamine PET [22, 23], which also showed a high potential for differentiating between healthy adrenal tissue and PHEO. In a recent study by Timmers et al. [24], 6-18F-fluorodopamine PET showed a similar sensitivity in detecting non-metastatic PHEO, but a clearly higher sensitivity in identifying metastases, compared to 123I-mIBG scintigraphy. In the same study, even though 6-18F-fluorodopamine sensitivity for both the detection and the localization of PHEO lesions was slightly lower than that of either whole-body CT or whole-body MRI, specificity values were much higher.

The role of the much-used 18F-FDG in the diagnosis of PHEO seems rather limited. Although a study by Taïeb et al. showed that all PHEO lesions had an elevated 18F-FDG accumulation, that study as well as another study by the same investigators suggested that, compared to 18F-DOPA PET, FDG-PET may underestimate the extent of disease [25, 26].

11C-hydroxyephedrine PET likewise demonstrated sensitivity and specificity in detecting PHEO that were high [2729] and superior to those of 123I-mIBG scintigraphy [30]. However, the requirement of an on-site cyclotron to produce the short-lived 11C (half-life: 20.4 min.) will hinder the widespread adoption of 11C-hydroxyephedrine, especially with the ready availability of 18F-containing tracers such as 18F-DOPA, which seem to have similarly good results. Alternatively, compounds labelled with the generator-derived radionuclide 68Ga, such as the somatostatin receptor agonists 68Ga-DOTA-DPhe1,Tyr3-octreotate (DOTATATE) or 68Ga-DOTA0-D-Phe1-Tyr3-octreotide, may be suitable for the diagnosis and localization of PHEO lesions; a small initial study by Win et al. showed that 68Ga-DOTATATE PET was superior to 123I-mIBG scintigraphy for PHEO imaging [31].

Conclusion

18F-DOPA PET/CT more accurately characterizes and localizes adrenal and extra-adrenal masses suspicious for PHEO than do 18F-DOPA PET or CT alone. The combined procedure appears to be extremely useful in the initial diagnosis as well as in the follow-up of this neuroendocrine tumor.

Acknowledgement

The authors wish to thank Robert J. Marlowe for his invaluable and extensive editorial assistance.

Copyright information

© Springer-Verlag 2009