FDG-PET and CT characterization of adrenal lesions in cancer patients

Original article

FDG-PET and CT characterization of adrenal lesionsin cancer patientsSuman Jana1, Tong Zhang1, David M. Milstein1, Carmen R. Isasi1, M. Donald Blaufox1

1 Department of Nuclear Medicine, Montefiore Medical Center and Albert Einstein College of Medicine, Bronx, NY 10461, USA

Received: 15 May 2005 / Accepted: 14 July 2005 / Published online: 29 September 2005© Springer-Verlag 2005

Abstract. Purpose: Fluorine-18 fluorodeoxyglucose pos-itron emission tomography (FDG-PET) may differentiatebenign from malignant adrenal lesions. In this study,standardized uptake values (SUVs), visual interpretation,and computed tomography (CT) data were correlated withthe final diagnosis to determine the contribution of adrenalFDG-PET in patients with known non-adrenal cancer.Methods: Ninety-two patients with adrenal lesions on CTunderwent FDG-PET. Eighty adrenals in 74 patients metthe inclusion criteria (PET scan within 4 weeks of CT plus>1 year of follow-up after PET scan with repeat CT orbiopsy for final diagnosis). CT was considered positive formetastases (CT+) based on two of the following threecriteria: >4 cm, Hounsfield units (HU) >30, and delayedcontrast enhancement. Lesions with <2 cm, with HU <20,and showing no enhancement were considered benign(CT−). Remaining lesions were considered indeterminate(CT-Ind). Visually, adrenal uptake exceeding liver uptakewas considered PET positive (PET+). Diagnosis of metas-tases was based on biopsy or interval CT growth (unchanged>1 year=benign). SUVmax and SUVavg were calculated froma 4×4 pixel region of interest drawn from CT, PET, and fusedimages. A receiver operator curve (ROC) determined theSUV with the best sensitivity and specificity.Results: Overall, PET was 93% sensitive and 96% specificfor metastases. A SUVmax of 3.4 was 95% sensitive and 86%specific. A SUVavg of 3.1 was 95% sensitive and 90%specific. There was no significant difference between visualinterpretation and SUV (SUVmax or SUVavg). Among CT+and CT− lesions, PETwas 100% sensitive and 96% specific;CT was 86% sensitive and 100% specific. In the CT-Indgroup, PET was 88% sensitive and 96% specific.Conclusion: PET accurately characterized adrenal lesions.Visual interpretation was as accurate as SUV. FDG-PET was

most useful in the 52.5% of cancer patients with inconclusiveadrenal lesions on CT.

Keywords: Adrenal lesions – FDG-PET – PET/CT –Adrenal metastases – Image fusion

Eur J Nucl Med Mol Imaging (2006) 33:29–35DOI 10.1007/s00259-005-1915-8

Introduction

The adrenal glands are a common site of metastases,adrenal metastases occurring in approximately 27% ofpostmortem examinations of patients with malignantneoplasms of epithelial origin [1, 2]. The most commonneoplasms associated with adrenal metastases are lung,breast, and melanoma [1–3]. Adrenal masses are frequentlyfound incidentally on computed tomography (CT), mag-netic resonance imaging (MRI), or ultrasound. The wide-spread use of high-resolution anatomic imaging modalitiesin patients with suspected abdominal and lower chestdisease commonly leads to the detection of an unexpectedadrenal mass [4–6]. These adrenal masses have been called“incidentalomas” by virtue of their serendipitous detection[7]. The prevalence of incidentalomas on CT scan has beenreported to be 0.35%–5.0% [7–10]. Herrera et al. reportedan adrenal mass in 3.4% of 61,054 abdominal CT scans[11].

A noninvasive evaluation of an adrenal mass isdesirable to avoid complications associated with adrenalbiopsy. Recently, 18F-fluorodeoxyglucose positron emis-sion tomography (FDG-PET) has shown great potential indifferentiating malignant from benign adrenal lesions inpatients with proven malignancy and in patients withincidentally detected adrenal masses on CT or MRI [12–15]. However, recent publications suggest that even non-contrast CT can differentiate a benign lesion from metas-tases with good accuracy [3, 16–18]. Most of the reports inthe literature have evaluated the ability of FDG-PET todifferentiate benign lesion from metastases; CT findingswere used only to detect the lesion. CT is the most

Suman Jana (*).Department of Nuclear Medicine,Montefiore Medical Center and Albert EinsteinCollege of Medicine,Bronx, NY 10461, USAe-mail: [email protected].: +1-718-4058461, Fax: +1-718-8240830

European Journal of Nuclear Medicine and Molecular Imaging Vol. 33, No. 1, January 2006

commonly used imaging modality in oncology and most ofthe patients studied in the nuclear medicine department byPET have already undergone at least one non-contrast CTscan. In this study we utilized the available CT data andpublished CT criteria [16–18] to achieve a CT diagnosis foreach lesion. CT diagnosis was used to identify differentgroups of patients, and the FDG-PET findings were thenevaluated to determine the value of FDG-PET in char-acterizing adrenal lesions. Most of the reported studiesdescribe the difficulty in drawing a region of interest (ROI)for standardized uptake value (SUV) measurement, par-ticularly in FDG-negative lesions. In one recent publicationusing image fusion it was possible to calculate the SUV innormal adrenals [19]. This study also used the principles ofimage fusion to draw an ROI around a negative lesion. Thepresent study also evaluated whether SUV measurementprovides any extra advantage over visual interpretation.

Materials and methods

A total of 1,684 charts of known cancer patients referred to our centerfor FDG-PET from October 2000 to July 2004 were reviewedretrospectively. Patients with adrenal lesions on CT who also had anFDG-PETscan within 4 weeks of the CTwere included in this study ifthey had either >1 year of follow-up after the first PETscanwith repeatCT or biopsy to provide a final diagnosis. Final clinical diagnosis ofadrenal metastases or a benign lesion was based on adrenal biopsy (13lesions), or a follow-upCTshowing significant change (metastases) orno change after >1 year (benign), or conclusive MRI chemical shift(four lesions). The minimum follow-up period ranged from 14 to 26months with a mean of 17.3.

PET imaging was performed using a C-PET scanner (ADAC/Philips) or Allegro (ADAC/Philips) or Gemini (ADAC/Philips). Allimages were acquired in three-dimensional mode. The patients fastedfor at least for 4 h and were advised to drink plenty of water. PETscanning was started 60 min after an intravenous injection of 185–370 MBq (5–10 mCi) of 18F FDG. Sequential overlapping scanswere acquired to include the neck, chest, abdomen, and pelvis.Transmission scans using a 137Cs source were interleaved betweenthe multiple emission scans to correct for nonuniform attenuationusing a segmented attenuation correction algorithm. The images werereconstructed using an iterative reconstruction method. The PETimages were displayed on a high-resolution computer screen as 3-Dvolume, coronal, sagittal, and transaxial views. The images werequalitatively evaluated by two nuclear medicine physicians. TheFDG PET scan was considered positive if the relative uptake in theadrenal region was more than the uptake in the liver in the samecoronal or transaxial section, as suggested by Yun et al. [12].

SUV max and SUV avg were calculated for all positive lesions bydrawing an ROI of 4x4 pixels on the lesion, and those of a negativelesion were calculated by drawing a similar ROI on the PET imagearound the region of the abnormal adrenal seen on CT by eithervisual or software image fusion. All SUV calculations wereperformed on the transaxial slices.

CT results were collected from the printed report and from reviewof the films (56 scans) by a board-certified radiologist whennecessary. Among 74 patients, 36 had contrast studies to determinethe delayed enhancement of the adrenals (among these 36, 32 filmswere available for review). CT scans was defined as positive formetastases (CT+) if the lesion met two of the following three criteria:>4 cm, Hounsfield units (HU) >30, and delayed enhancement 15–30 min after contrast. Any lesion that met all three of the following

criteria on CT was considered benign: <2 cm, HU <20, and noenhancement. Lesions that did not fulfill the criteria of CT+ or CT−were classed as indeterminate (CT-Ind).

Statistical analysis

Data were summarized calculating proportions for categoricalvariables and means and their standard deviation for continuousvariables. To evaluate the diagnostic performance of PET wecalculated the sensitivity (proportion of true positives among alllesions confirmed as malignant), specificity (proportion of truenegatives among all non-malignant lesions), and accuracy (pro-portion of correctly classified lesions among all lesions). In addition,receiver operating characteristic (ROC) curves were calculated andthe areas under the ROC curve were compared using methods forcorrelated data. All statistical analyses were performed using StataSoftware (version 7.0).

Results

Ninety-two patients who were found to have adrenallesions by CT underwent FDG-PET scan. Eighty adrenallesions (68 unilateral and six bilateral; 47 in females and 33in males) in 74 patients (age range 49–88 years; mean±SD68.57±9.2 years) met our inclusion criteria. All 74 patientshad proven malignancy (42 lung cancer, 12 lymphoma,seven colon cancer, six breast cancer, four melanoma, andthree esophageal cancer). The 80 lesions fell into three CTgroups. Twelve lesions (15%; largest dimension range 1.8–6.5 cm, mean 4.6) were determined as CT+, 26 (32.5%;largest dimension range 1.2–3.4 cm, mean 2.1) were CT−,and 42 (52.5%; largest dimension range 1.2–4 cm, mean2.8) failed to meet the criteria for either CT+ or CT− andwere classified as CT-Ind. Table 1 shows the distribution ofPET findings and the final diagnoses in these three groupsof lesions.

Table 1. The distribution of adrenal lesions in three groupsaccording to the CT findings, with their PET findings and finaldiagnosis

CT findings (groups) PET findings Final diagnosis

Group 1: CT+ (n=12) PET+ (n=12) Dx+ (n=12) (Bx=8)Dx− (n=0)

PET− (n=0) Dx+ (n=0)Dx− (n=0)

Group 2: CT− (n=26) PET+ (n=3) Dx+ (n=2)Dx− (n=1)

PET− (n=23) Dx+ (n=0)Dx− (n=23)

Group 3: CT-Ind (n=42) PET+ (n=15) Dx+ (n=14) (Bx=2)Dx− (n=1)

PET− (n=27) Dx+ (n=2)Dx− (n=25) (Bx=3)

Dx+ malignant, Dx− benign, Bx biopsy proven, n number oflesions

30


Thirty lesions (37.5%) were found to be metastaticcancer at the final diagnosis. Two of these metastaticlesions fell into the category of CT-Ind; both were PETpositive. Two of the 50 benign lesions showed increasedFDG uptake; one was a CT− and the other, a CT-Ind lesion.Among all 80 lesions, PETwas 93% (28/30) sensitive, 96%(48/50) specific, and 95% (76/80) accurate.

Among the CT+ and CT− lesions, PET was 100%(14/14) sensitive, 96% (22/23) specific, and 97% (37/38)accurate. In this group, CT was 86% (12/14) sensitive and100% (24/24) specific with an accuracy of 95% (36/38).Figure 1 shows the CT and PET scan findings of a patientwith bilateral adrenal lesions who was CT+ and PET+, andwas subsequently found to be positive for metastases bybiopsy.

CT was indeterminate in 52.5% (42/80) of the lesions,including five biopsied lesions of which two were positivefor metastases. In this CT-Ind group PET was 88% (14/16)sensitive, 96% (25/26) specific and 93% (39/42) accurate.Figure 2 shows CTand PET findings of a patient whose leftadrenal was in the CT-Ind group but the PET+ group andsubsequently found to be positive for metastasis at biopsy.Among the biopsy-proven group (13 lesions), PET was100% specific (3/3) and sensitive (10/10).

The SUVmax ranged from 2.3 to 16 in the metastasisgroup, with a mean of 5.5. The SUVmax of the benign

lesions ranged from 1.3 to 3.6 with a mean of 2.6. Inmalignant lesions the SUVavg ranged from 2.1 to 13 with amean of 4.81 and in benign lesions it ranged from 1.2 to 3.3with a mean of 2.2. Figures 3 and 4 show the ROCsobtained from SUV max and SUVavg respectively. From thecurve in Fig. 3 it is evident that if 3.1 is taken as the cutofffor SUVmax, there is 100% sensitivity but only 73%specificity. On the other hand, if 5.5 is used as the cutoff ityields 100% specificity but the sensitivity falls to 55%. Acutoff value of 3.4 provides 95% sensitivity and 86%specificity. For a SUVavg of 3.1 the sensitivity is 95% andthe specificity, 90% (Fig. 4). There was no significantdifference between SUV max and SUVavg statistically.Comparison with the sensitivity and specificity of visualinterpretation revealed no significant difference with theSUV measurements.

Discussion

The prevalence of adrenal adenoma is high; small homo-geneous adrenal masses discovered incidentally are likelyadenomas. If a patient has other metastases and the pres-ence of adrenal metastases will not alter therapy, furtherevaluation is not necessary [1]. However, if there is noother evidence of metastases, an incidentally found adrenal

Fig. 1. A 75-year-old woman with lung cancer. a PET images werepositive for bilateral adrenal uptake (arrow and arrowhead) withSUVmax values of 9 (right adrenal) and 11 (left adrenal). b Ontransaxial CT, the right adrenal (arrow) measured 2.5×2.8 cm with an

attenuation value of 43 HU, the left adrenal (arrow) measured5×3.5 cm with an attenuation value of 48 HU, and both showeddelayed enhancement; therefore, both lesions fell into the CT+ group.Biopsy was positive for bilateral metastases

31


mass requires further evaluation as it significantly in-fluences the management of that patient.

In patients without a known extra-adrenal primarymalignancy, more than 80% of incidentally discoveredadrenal masses are benign nonfunctioning adenomas [10,

Fig. 2. A 61-year-old man with lung cancer. a PET images werepositive for left adrenal uptake (arrowhead) with a SUVmax of5.6. b Transaxial CT showed a normal right adrenal but the leftadrenal (arrow) measured 3.5 cm with an attenuation value of

37 HU and equivocal delayed enhancement. Therefore, the leftadrenal fell into the CT-Ind group. The final diagnosis was provenby biopsy to be metastatic left adrenal

Area under ROC curve = 0.9646

Sen

sitiv

ity

1 - Specificity0.00 0.25 0.50 0.75 1.00

0.00

0.25

0.50

0.75

1.00

Fig. 3. ROC generated fromSUVmax data. A SUVmax of 3.1(arrowhead) provides a sensi-tivity of 100% and a specificityof only 73.33%. On the otherhand, a SUVmax of 5.5 (thinarrow) provides a specificity of100% but a sensitivity of only55%. A SUVmax of 3.4 (openarrow) provides a sensitivity of95% with a specificity of86.25%

32


20, 21]. Even in patients with a primary neoplasm in whoman adrenal metastasis is an important consideration, themajority of adrenal masses (about 40–57%) are benign[22–24]. However, despite the high frequency with which abenign lesion is the cause of an incidentally detected mass,it is important to distinguish the benign lesions from theless common malignant masses, which might requireintervention [1].

Adrenal biopsy is a relatively invasive procedure and isassociated with a failure rate of 14–50% after the firstattempt due to either non-diagnostic pathology or the failureto obtain adequate tissue [22, 23, 25]. This procedure is alsoassociated with a reported complication rate of 8.4–11.3%[22, 26]. The commonly reported complications are pneu-mothorax, pain, perinephric hemorrhage, subcapsular andintrahepatic hematoma, and hepatic needle tract metastasesdepending on the approach to the adrenal gland [22, 23, 25,26]. There are also reported cases of hypotension and a fallin hematocrit requiring blood transfusion as complicationsof adrenal biopsy [22].

Solitary adrenal metastases from lung cancer and mel-anoma without any evidence of other distance metastasesare well documented [25]. A tumor with an adrenal metas-tasis is considered unresectable and is usually treated withchemotherapy [25]. There is a significant difference insurvival between patients with non-small cell lung cancerassociated with a benign adrenal mass who undergo cura-tive lung resection and patients with the same type ofcancer and adrenal metastases treated with chemotherapy[25].

The diagnostic approach to adrenal lesions may differfrom institution to institution according to the availabilityof imaging methods. Nevertheless, CT scanning is con-sidered the most important and the most commonly usedmodality in oncology. Most patients visiting the nuclearmedicine department for an FDG-PET scan will alreadyhave undergone CT. In the recent literature there isevidence that CT can be helpful in accurate diagnosis of

benign lesions [1, 17, 18]. Lipid within adenomas causes alow attenuation value on CT. Evidence has accumulatedthat unenhanced CT densitometry can be used to accuratelydifferentiate adrenal adenomas from metastases [2, 8, 17,18]. Using pooled data from multiple published studies ofcalculated accuracies and corresponding threshold valuesof unenhanced attenuation values, a group of researchersfound that the optimal sensitivity (71%) and specificity(98%) for the diagnosis of adrenal adenoma resulted fromchoosing a threshold attenuation value of 10 HU onunenhanced CT [18]. We used this cutoff as one of ourcriteria to reach a diagnosis by CT.

Standard contrast-enhanced CT images of the adrenalglands are obtained approximately 60 s after the beginningof intravenous injection of a contrast bolus. Studies suggestthat this time point is the only one when the attenuationvalues of adenomas and metastases are nearly identical.Adenomas lose enhancement more rapidly, as early as5 min after the injection of contrast material, so attenuationvalues at 10–15 min after contrast administration can beused to differentiate adenomas from other masses [1, 27,28]. Although the threshold attenuation value for thediagnosis of an adenoma varies among series, masses withan attenuation value of less than 30–40 HU on a contrast-enhanced CT scan obtained with a 15-min delay are almostalways adenomas [1].

Evidence has accumulated that chemical shift MRimaging can also be used to differentiate adrenal adenomasfrom metastases. Taking advantage of the different reso-nant frequency peaks for the hydrogen atom in water andtriglyceride (lipid) molecules, chemical shift MR imagingresults in a decrease in the signal intensity of tissuecontaining both lipid and water in comparison with tissuecontaining no lipid [29]. The chemical shift change can bedetected by simple visual analysis or by quantitativemethods. Several different formulas have been proposed toquantify the amount of chemical shift change and optimalthreshold values determined by analysis of scatter plot data

Area under ROC curve = 0.9708

Sen

sitiv

ity

1 - Specificity0.00 0.25 0.50 0.75 1.00

0.00

0.25

0.50

0.75

1.00

Fig. 4. ROC generated fromSUVavg data. A SUVavg of 2.7(arrowhead) provides a sensi-tivity of 100% and a specificityof only 76.67%. On the otherhand, a SUVavg of 4.2 (thinarrow) provides a specificity of100% but a sensitivity of only60%. A SUVavg of 3.1 (openarrow) provides a sensitivity of95% with a specificity of 90%

33


[1, 30, 31]. The reported sensitivity and specificity valuesfor a benign lesion were 87–100% and 92–100% re-spectively [32, 33]. At the highest (definite) confidence of abenign lesion, the positive predictive value could be as highas 99–100% [32, 33]. Failure rate for diagnosis could behigh (13–17%) in lipid-poor adenomas [32, 33].

In contrast to MRI and CT, FDG PET is based onincreased glucose uptake in malignant lesions. FDG-PET iscurrently used for primary evaluation as well as follow-upof patients with cancer. It has been documented to be veryuseful in patients with melanoma, lung, and breast cancer,all of whom are more likely to have adrenal metastases [1–3]. One of the most important advantages of the PETscan isthat it is a noninvasive whole-body imaging modality andone scan evaluates the overall extent of the cancer. TheFDG-PET scan is used with increasing frequency as anoninvasive imaging modality for evaluating patients withcancer [34]. It is not surprising that FDG PET has also beenused to characterize adrenal lesions. Most of the studies[12, 13, 16] have reported sensitivities and specificities,without comparisons with CT scan findings. In this studywe compared the CT scan findings with the PET scanresults. Most others have not performed SUV measure-ments owing to the difficulty in drawing an ROI around aPET-negative lesion. However, with the increasing popu-larity of PET/CT scanners it is even possible to draw ROIsin FDG-negative adrenal glands using image fusion, asreported by Bagheri et al. [19].

Authors using different visual interpretation criteriahave reported different sensitivities and specificities. In 24lesions of different cancer types in 20 patients, Boland et al.[13] reported 100% accuracy using the tumor-to-back-ground ratio. In 33 lesions of 27 lung cancer patients,Erasmus et al. [14] reported a 100% sensitivity and spec-ificity when using activity higher than background as in-dicative of a positive scan. In another study in 41 patientswith 50 lesions, Yun et al. [12] found a sensitivity of 100%and a specificity of 94% when using lesion uptake morethan or equal to liver uptake as the criterion of a positiveresult. In comparison, we found a sensitivity of 93% and aspecificity of 96% using criteria similar to those of Yun et al.[12] in 80 lesions of 74 patients. We found two adenomaswith higher uptake than the liver, and corresponding SUV max

values were 3.6 and 3.1. Yun et al. [12] also reportedvariable FDG uptake in adenomas, leading to false pos-itive results. The most common type of tumor to causefalse positive results on FDG-PET scan is reported to bepheochromocytoma of the adrenal [12]. In our study therewere also two false negative lesions on FDG PET (SUVmax

2.1 and 2.3) in patients with pulmonary carcinoid. LowerFDG uptake has been reported in the metastases from neu-roendocrine tumors [12, 35]. Hemorrhage and necrosisare reported to be the other common causes of a falsenegative FDG-PET scan [12]. In this study we found nosignificant difference between visual interpretation andSUV measurement.

The prevalence of adrenal masses in autopsy series hasbeen reported to be 1.4–12.4% [7–10, 36]. This suggeststhat the postmortem detection rates are higher than current

imaging detection rates. Detection will likely increase asimaging techniques improve and are used more frequently[7, 32]. This may explain the higher frequency of benignlesions (63%) in our study compared with the 40–57% ratereported in the literature [22–24]. The other reason for thehigh incidence of benign lesions could be the relativelyhigher number of patients with lymphoma in this study.Prior studies have shown that involvement of the adrenalby lymphoma is uncommon, occurring in only 1–4% ofpatients [19, 37]. In this study the CT scan, when con-clusive, provided a sensitivity of 86% with a specificity of100%. This is consistent with the results reported in theradiology literature using similar criteria [1, 18, 27, 28].Therefore, if CT is conclusive, other imaging studies arenot required to characterize the adrenal lesion; however,whole-body FDG-PET may still be useful to evaluate otherorgans for metastases. On the other hand, we found that asignificant number of lesions (52.5%) were CT indetermi-nate, requiring further evaluation, and in this group ofpatients FDG-PET was very useful, providing an accuracyof 93%.

Possible limitations of this study are that it wasretrospective in nature and that adequate comparison ofCT and PET diagnoses was hindered by the fact that not allthe patients had CTwith contrast enhancement. In practice,most patients undergo FDG-PETscan for evaluation of theirdisease status after CT, which may be performed withcontrast but is not specifically performed for adrenalevaluation based on delayed enhancement. Another possi-ble limitation was SUV calculation from three differentscanners (C-PET, Allegro, and Gemini): though they werefrom the same vendor, i.e., Philips/ADAC, and possessedsimilar software, there could be potential differences. Themean±SD of the percent difference in SUVs between theAllegro and the Gemini was found to be 15±1.2% in aphantom study conducted recently in our center (unpub-lished data).

Conclusion

The FDG-PET scan accurately characterized the adrenallesions as judged by CT and by biopsy criteria. Visualinterpretation can be as accurate as SUV calculation andcan be used for all clinical purposes. Although this studysuffers from the limitations of being retrospective, itappears that FDG-PET is most useful in patients withknown cancer and inconclusive adrenal lesions on CT. Aconclusive CT using appropriate criteria can characterizeadrenal lesions accurately.

References

1. Dunnick NR, Korobkin M. Imaging of Adrenal Incidentalomas:current status. AJR Am J Roentgenol 2002;179:559–68

2. Abrams HL, Spiro R, Goldstein N. Metastases in carcinoma:analysis of 1000 autopsied cases. Cancer 1950;3:74–85

34


3. Zornoza J, Bracken R, Wallace S. Radiologic features ofadrenal metastases. Urology 1976;8:295–9

4. Falk THM, Sandler MP. Classification of silent adrenal masses:time to get practical. J Nucl Med 1994;35:1152–4

5. Gross MD, Shapiro B. Clinical review 50: clinically silentadenal masses. J Clin Endocrinol Metab 1993;77:885–8

6. Kloos RT, Gross MD, Francis IR, Korobkin M, Shapiro B.Incidentally discovered adrenal masses. Endocr Rev 1995;16:460–84

7. Higgins JC. Diagnosis of renal and adrenal incidentalomas.Clinics in Family Practice 2002;4(3):505

8. Barzon L, Boscara M. Diagnosis and management of adrenalincidentalomas. J Urol 2000;163(2):398–422

9. Cook DM, Loriaux LD. The incidental adrenal mass. Am JMed 1996;101(1) 88–94

10. Ooi TC. Adrenal incidentalomas: in detection, not significance.CMAJ 1997;157(7):903–4

11. Herrera MF, Grant CS, van Heerden JA, Sheedy PF, IlstrupDM. Incidentally discovered adrenal tumors: an institutionalperspective. Surgery 1991;110(6):1014–21

12. Yun M, Kim W, Alnafisi N, Lacorte L, Jang S, Alavi A. 18F-FDG PET in characterizing adrenal lesions detected on CT orMRI. J Nucl Med 2001;42:1795–99

13. Boland GW, Goldberg MA, Lee MJ, Mayo-Smith WW, DixonJ, McNicholas MM, et al. Indeterminate adrenal mass inpatients with cancer: evaluation at PET with 2-[F-18]-fluoro-2-deoxyglucose. Radiology 1995;194:131–4

14. Erasmus JJ, Patz EF Jr, McAdams HP, Murray JG, Herndon J,Coleman RE, et al. Evaluation of adrenal masses in patientswith bronchogenic carcinoma using 18F- fluorodeoxyglucosepositron emission tomography. AJR Am J Roentgenol 1997;168:1357–60

15. Maurea S, Mainolfi C, Bazzicalupo L, Panico MR, Imparato C,Alfano B, et al. Imaging of adrenal tumors using FDG PET:comparison of benign and malignant lesions. AJR Am JRoentgenol 1999;173:25–9

16. Lee MJ, Hahn PF, Papanicolaou N, Egglin TK, Saini S, MuellerPR, et al. Benign and malignant adrenal masses: CT distinctionwith attenuation coefficients, size, and observer analysis. Radiol-ogy 1991;179:415–8

17. Korobkin M, Brodeur FJ, Yutzy GG, Francis IR, Quint LE,Dunnick NR, et al. Differentiation of adrenal adenomas fromnonadenomas using CT attenuation values. AJR Am J Roentgenol1996;166:531–6

18. Boland GW, Lee MJ, Gazelle GS, Halpern EF, McNicholasMM, Mueller PR. Characterization of adrenal masses usingunenhanced CT: an analysis of the CT literature. AJR Am JRoentgenol 1998;171:201–4

19. Bagheri B, Maurer AH, Cone L, Doss M, Adler L. Char-acterization of the normal adrenal gland with 18F-FDG PET/CT.J Nucl Med 2004;45:1340–3

20. Prinz R, Brooks M, Churchill R, Graner JL, Lawrence AM,Paloyan E, et al. Incidental asymptomatic adrenal masses detectedby computed tomographic scanning. JAMA 1982;248:701

21. Paulsen SD, Nghiem HV, Korobkin M, Caoili EM, Higgins EJ.Changing role of imaging-guided percutaneous biopsy ofadrenal masses: evaluation of 50 adrenal biopsies. AJR Am JRoentgenol 2004;182:1033–7

22. Bernardino ME, Walther MM, Phillips VM, Graham SD Jr,Sewell CW, Gedgaudas-McClees K, et al. CT-guided adrenalbiopsy: accuracy, safety and indications. AJR Am J Roentgenol1985;144(1):67–9

23. Silverman SG, Mueller PR, Pinkney LP, Koenker RM, SeltzerSE. Predictive value of image-guided adrenal biopsy: analysisof results of 101 biopsies. Radiology 1993;187(3):715–8

24. Welch TJ, Sheedy PF, Stephens DH, Johnson CM, Swensen SJ.Percutaneous adrenal biopsy: review of a 10-year experience.Radiology 1994;193:341–4

25. Ettinghausen SE, Burt ME. Prospective evaluation of unilateraladrenal masses in patients with operable non-small-cell lungcancer. J Clin Oncol 1991;9:1462–6

26. Mody MK, Kazerooni EA, Korobkin M. Percutaneous CT-guided biopsy of adrenal masses: immediate and delayedcomplications. J Comput Assist Tomogr 1995;19(3):434–9

27. Xarli VP, Steele AA, Davis PJ, Buescher ES, Rios CN, Garcia-Bunnel R. Adrenal hemorrhage in the adult. Medicine (Balti-more) 1978;47:211–21

28. Szolar DH, Kammerhuber F. Quantitative CT evaluation ofadrenal gland masses: a step forward in the differentiationbetween adenomas and nonadenomas? Radiology 1997;202:517–21

29. Mitchell DG, Crovello M, Matteucci T, Petersen RO, MiettinenMM. Benign adrenocortical masses: diagnosis with chemicalshift MR imaging. Radiology 1992;185:345–51

30. Mayo-Smith WW, Lee MJ, McNicholas MMJ, Hahn PF,Boland GW, Saini S. Characterization of adrenal masses(<5 cm) by use of chemical shift MR imaging: observerperformance versus quantitative measures. AJR Am J Roent-genol 1995;165:91–5

31. Kreft BP, Muller-Miny H, Sommer T, Steudel A, VahlensieckM, Novak D, et al. Diagnostic value of MR imaging incomparison to CT in the detection and differential diagnosis ofrenal masses. ROC analysis. Eur Radiol 1997;7:542–7

32. Outwater EK, Siegelman ES, Huang AB, Birnbaum R. Adrenalmasses: correlation between CT attenuation value and chemicalshift ratio at MR imaging with in-phase and opposed-phasesequences. Radiology 1996;200:749–52

33. Korobkin M, Lombardi TJ, Aisen AM, Francis IR, Quint LE,Dunnick NR, et al. Characterization of adrenal masses withchemical shift and gadolinium-enhanced MR imaging. Radiol-ogy 1995;197:411–8

34. Jana S, Abdel-Dayem HM. Role of nuclear medicine inevaluating treatment response in oncology. Nuclear MedicineAnnual 2004;1–59

35. Erasmus JJ, McAdams HP, Patz EF Jr, Coleman RE, Ahuja V,Goodman PC. Evaluation of primary pulmonary carcinoidtumors using FDG PET. AJR Am J Roentgenol 1998;170:1369–73

36. Hedeland H, Ostberg G, Hokfelt B. On the prevalence ofadrenocortical adenomas in an autopsy material in relation tohypertension and diabetes. Acta Med Scand 1968;184(3):211–4

37. Pauling M, Williamson B. Adrenal involvement in non-Hodgkin lymphoma. AJR Am J Roentgenol 1983;141: 303–5

35


Documents

FDG-PET and CT characterization of adrenal lesions in cancer patients