ELSEVIER
EJR EUROPEAN JOURNAL OF RADIOLOGY
How accurate is MR imaging in characterisation of adrenal masses: update of a long-term study
Selma Hönigschnabl a, Sylvia Gallo ª, Bruno Niederle b, Gerhard Prager b, Klaus Kaserer ℃, Gerhard Lechner ª, Gertraud Heinz-Peer a,*
ª Division of Surgical Radiology, Department of Radiology, University of Vienna, Währinger Gürtel 18-20, A-1090 Vienna, Austria
b Department of Surgery, University of Vienna, Vienna, Austria
” Department of Pathology, University of Vienna, Vienna, Austria
Received 8 October 2001; received in revised form 9 October 2001; accepted 10 October 2001
Abstract
Objective: The purpose of this study was to update a long-term study that evaluates the accuracy of MR imaging in the characterisation of adrenal tumours. In all patients, MR imaging findings were correlated with histopathologic results. Patients: In 204/560 patients who underwent MR imaging for characterisation of an adrenal mass, histopathologic results were available. The final study group consisted of 229 adrenal masses in 204 patients. MR imaging was performed using T2-weighted fast spin-echo imaging and unenhanced and gadolinium-enhanced T1-weighted spin-echo imaging in all patients. In addition, chemical shift imaging was performed in 182 patients and dynamic gadolinium-enhanced studies in 198 patients. Chemical shift images and dynamic studies were qualitatively assessed. All images were reviewed by an experienced investigator (Gertraud Heinz-Peer) who was blinded to the clinical history and the results of prior imaging studies. Results: The sensitivity of MR imaging for the differentiation of benign and malignant adrenal masses was 89%, the specificity 99%, and the accuracy was 93.9%. This results in a positive predictive value (PPV) of 90.9% and a negative predictive value (NPV) of 94.2%. These results are comparable to the data published previously by our study group with a lower number of cases. Conclusion: Large study numbers show that MR imaging is a reliable method in characterisation of benign and malignant adrenal masses. Since laparoscopic adrenalectomy has become the new gold standard in the surgical treatment of benign adrenal lesions, the high accuracy of MR imaging in characterisation of those lesions offers even patients with large adrenal masses (> 5 cm) the advantages of the minimally invasive technique. @ 2002 Elsevier Science Ireland Ltd. All rights reserved.
Keywords: MR imaging; Adrenal mass; Laparoscopic adrenalectomy
1. Introduction
With the advent of dynamic gadolinium enhanced- and chemical shift (CSI)- techniques, MR imaging has become a well-accepted diagnostic method in the char- acterisation of adrenal masses. The accuracy of MR imaging in the differentiation between benign and ma- lignant adrenal tumours has been reported in many studies [1-4]. However, not all of the reported cases have been proven histologically.
Korobkin et al. [5] and Outwater et al. [6] showed that the presence of histologic lipid in many of the examined adenomas accounted for the low attenuation on unenhanced CT, causing a loss in signal intensity on chemical shift MR imaging.
Besides MR imaging, unenhanced CT scans and con- trast media wash-out characteristics on delayed en- hanced CT scans have become widely accepted in the differentiation of benign and malignant adrenal masses. Similar to MR studies characterisation of adrenal masses in these studies was not fully based on histolog- ical proof. However, some investigators have claimed superiority of these techniques over MR imaging tech- niques [7-9].
* Corresponding author. Tel .: + 43-1-40400-4818; fax: +43-1- 40400-4898.
E-mail address: gertraud.heinz@univie.ac.at (G. Heinz-Peer).
In a previous study performed by our study group, 134 adrenal masses characterised by conventional MR imaging, chemical-shift and gadolinium-enhanced imag- ing techniques were correlated with the results of histo- pathologic examination. The accuracy in differentiation between benign and malignant adrenal masses was 93% [10].
Thus the aim of this study was to re-evaluate the reliability of MR imaging in characterisation of adrenal masses on a large number of adrenal tumours by correlating every single case with the histopathologic result.
2. Subjects and methods
2.1. Patients
Between October 1995 and April 2001, 560 unse- lected patients with an adrenal mass detected on ultra- sound (US) or computed tomography (CT) or with clinically suspected adrenal mass were referred to us for MR imaging.
Patients in whom neither cytologic nor histologic analysis was possible were excluded from this study (n = 356). Thus, our study group consisted of 229 adrenal masses in 204 consecutive patients (72 men, 132 women; age range, 2.0-83.0 years; median, 54.1 years). One hundred and ten out of 229 (48.0%) adrenal masses were located on the right side. In 23 patients a bilateral adrenal mass was diagnosed.
Approval by the investigative review board was not considered necessary since MR imaging is a non-inva- sive technique and an accepted method in the preopera- tive evaluation of adrenal masses. In addition, no study drug was used.
The indications for MR imaging included adrenal masses seen in patients with abnormal results at chemi- cal analysis of blood and urine, incidentally (i.e. found during CT or sonography performed for other reasons) detected adrenal masses, and known malignant disease.
2.2. MR imaging technique
MR imaging was performed with either a 1.0-T unit (Gyroscan T10-NT; Philips, Eindhoven, the Nether- lands) or a 1.5-T unit (Siemens, Magnetom-Vision, Erlangen, Germany). T1-weighted spin-echo axial im- ages were obtained with a TR of 425 ms and a TE of 15 ms. T2-weighted axial and coronal images were ob- tained with the fast spin-echo technique by using a TR of 1800 ms and a TE of 100 ms. The section thickness was 4 mm with an intersection gap of 0.4 mm for both T1- and T2-weighted imaging. The matrix size was 204 × 256, and four signals were acquired. Respiratory- ordered phase-encoding and superior and inferior spa-
tial saturation pulses were used for spin-echo imaging, with flow compensation being used for most of the spin-echo images.
Chemical-shift imaging was performed using a T1- weighted gradient-recalled echo sequence with breath- holding with a TR of 25.0 ms and a TE of 6.9 ms for in-phase images and a TE of 3.5 ms for opposed-phase images on the 1.0-T scanner, and a TR/TE = 150/4.6 for in-phase and a TR/TE = 133/2.2 for opposed-phase images on the 1.5-T scanner, respectively. The flip angle was 30°, the matrix size 128 x 256, and one signal was acquired. The field of view varied from 320 to 375 cm, depending from the size of the patient.
In all 0.1 mmol/kg of gadopentetate dimeglumine were administrated for the dynamic studies, which were performed by obtaining more than one image of single sections in the same position in 13 s intervals. A gradient-recalled echo sequence was used with a TR of 9.6 ms, a TE of 4.6 ms, a flip angle of 25°, a matrix size of 128 × 256, a field view of 375 cm, a section thickness of 6 mm, and an intersection gap of 0.6 mm. The T1-weighted spin-echo images were obtained 15 min after the injection of gadolinium. The section thickness was 4 mm with an intersection gap of 0.4 mm.
2.3. MR imaging criterias
As adenomas were diagnosed masses that appeared hypo- or isointense relative to the liver on T1-weighted images and hyper- or isointense to the liver on T2- weighted images, and that lost signal intensity on op- posed-phase images compared with in-phase images (Fig. 1a-c) [11]. In addition, a quick washout on gadolinium-enhanced studies was considered more typi- cal of benign than malignant lesions [12].
Lesions with marked enhancement on delayed gadolinium series were considered more likely to be malignant. The diagnosis of carcinomas and metastases was based more on findings from chemical shift and gadolinium enhanced studies than on the signal intensi- ties of conventional techniques (Fig. 2a-c, Fig. 3a-c) [11].
High signal intensity of homogenous adrenal masses on T2-weighted images and no signal loss on opposed- phase images compared with in-phase images was con- sidered to be a pheochromocytoma [13,14].
In addition, an atypical pheochromocytoma may be of medium signal intensity on T2-weighted images or may appear inhomogeneous, especially when they are cystic pheochromocytomas (Fig. 4a-b) [14].
Lesions appearing heterogeneous on T1 and hetero- geneous on T2 and showed a peripheral nodular en- hancement and central hypoperfusion on contrasted MRI were characterised as adrenal carcinoma. The invasion of adjacent organs or the inferior vena cava was also considered typical of adrenal carcinoma [14].
The presence of fat-containing areas that showed signal intensity equal to those of subcutaneous and
(a)Sc3
TSE/M
H
SL 12
L
5 cm
AP 25 pos
(b)
(c)
(a)
(b)
(c)
retroperitoneal fat at all pulse sequences was the crite- rion for myelolipoma. The suppression of fat-contain-
ing areas on fat-saturation MR images, in addition, was considered typical of myelolipoma [14].
(a)
(b)
(c)
The signal intensity of cysts may depend on their content [14], however, round adrenal masses with sharp margins in which there was no gadolinium enhance- ment were diagnosed as adrenal cysts.
2.4. Data analysis
All MR images were reviewed by an experienced investigator who did not know the clinical history and the results of prior imaging studies and the histopatho- logic results. Gadolinium enhanced studies were per- formed in 198 patients. The signal intensity of these gadolinium enhanced images was compared with unen- hanced images, and maximal gadolinium washout was assessed by comparing the signal intensity at maximum enhancement with the one on the final image.
In addition, 92 adrenal masses were evaluated quan- titatively for the same signal intensity ratios for which qualitative assessment was performed.
The percentage change in the signal intensity ratio of the mass to spleen, when comparing opposed-phase to in-phase images, was calculated on the basis of a for- mula used by Mitchell et al. [11], ([SIm/SIs]opposed/[SIM/ SIslin - 1) × 100. SI represents signal intensity; M the mass; and S, the spleen. The signal intensity was calcu- lated in this way: SI index = ([SIin adrenal - SI adrenal]/[SI __ adrenal]) × 100. opposed
2.5. Standard of reference
For the proof of diagnosis, a histopathologic exami- nation was established after surgical resection (n = 227) and CT-guided percutaneous biopsy (n = 2).
We operated on all functional lesions independent of their size and on all non-functional lesions that were larger than 3 cm, as part of our clinical and surgical protocol. For invasive procedures, informed consent was obtained from all patients.
2.6. Statistical analysis
Data analysis was performed using statistical soft- ware SAS 6.12® (Statistical Analysis System version 6.12; SAS Institute, Cary, NC, USA) for windows (Microsoft, Redmond, CA, USA).
Sensitivity, specificity, positive and negative predic- tive values, and accuracy for MR findings were calcu- lated using 2 × 2 contingency tables. For the computation of the positive and negative predictive values, the prevalence was taken from the sample. Results of histopathologic examination were considered as gold standard. Categorical variables such as gender were analysed using frequency tables. A P < 0.05 was considered statistically significant.
(a)
(b)
SP
3. Results
On the basis of conventional MR imaging, qualita- tive and quantitative analysis of chemical-shift tech- niques, and on qualitative dynamic studies after intravenous administration of gadolinium, the diagnosis of adenoma was made in 131 cases (57.2%). Forty-five adrenal masses were diagnosed as pheochromocytomas (19.7%), 15 as carcinomas (6.6%), 14 as metastases (6.1%), adrenal cysts in eight cases (3.5%), myelolipo-
mas in four cases (1.7%), collision tumour of the adrenal gland in one patient (0.4%), one neurogenic tumour (0.4%), one lymphoma (0.4%), two lesions were classified as epithelial tumours with high malignant potential (borderline tumours) (0.9%), and finally seven lesions were classified as haemorrhages (3.1%).
Histopathologic results showed adenomas in 136 cases (59.4%), pheochromocytomas in 45 adrenal masses (19.7%), adrenal carcinomas in ten cases (4.4%), metastases in eight cases (3.5%), adrenal cysts in seven
cases (3.1%), adrenal haemorrhages in seven cases (3.1%), myelolipoma in five cases (2.2%), neurogenic tumours in five cases (2.2%), epithelial tumours with high malignant potential (borderline tumours) in two cases (0.9%), two adrenal masses (bilateral in the same patient) proved to be granulomatous tissue suspicious for tuberculosis (0.9%), lymphoma in one case (0.4%), and finally one collision tumour (0.4%).
The sensitivity of MR imaging for the differentiation of benign and malignant adrenal masses was 89%, the specificity was 99%, and the accuracy was 93.9%. This results in a positive predictive value (PPV) of 90.9% and a negative predictive value (NPV) of 94.2%.
The sensitivity of MR imaging for the diagnosis of adenoma was 96.2%, the specificity was 89.8%, and the accuracy was 93.9% (PPV = 91.0%, NPV = 94.6%). Pheochromocytomas were diagnosed with a sensitivity of 93.3%, a specificity of 98.4%, and an accuracy of 97.4% (PPV = 93.3%, NPV = 96.5%). The sensitivity for adrenal carcinomas was 93.2%, the specificity was 100%, and the accuracy was 97.8% (PPV = 100%, NPV = 96.7%). Adrenal metastases were diagnosed with a sensitivity of 84%, a specificity of 100%, and an accuracy of 97.4% (PPV = 100%, NPV = 95.3%). Myelolipomas had a sensitivity of 100%, a specificity of 99.6% and an accuracy of 99.6%. The sensitivity, spe- cificity, accuracy, PPV, and NPV for all entities are summarised in Table 1.
Adrenal lesions ranged in size between 9 and 300 mm, mean diameter was 41.8 mm. Thirteen adenomas were larger than 50 mm. Pheochromocytomas showed a mean diameter of 43 mm; two pheochromocytomas were smaller than 20 mm in diameter. The diameters of adrenal carcinomas ranged between 20 and 300 mm, mean 107 mm. Mean diameter of metastases was 46 mm, range 20-70 mm.
Quantitative analysis for dynamic Gadolinium-en- hanced studies was reported in our previous study. According to the study of Korobkin et al. [4] that showed no significant difference between quantitative and qualitative evaluation, quantitative analysis is no longer part of our routine protocol. With histology as the gold standard, the correct diagnosis was missed in 21/229 (9.2%) cases by MR imaging.
4. Discussion
Adrenal masses are a relatively common finding on cross sectional imaging methods. Adrenal biopsy, surgi- cal resection or clinical follow-up has been widely used to distinguish between benign and malignant tumours. Since the introduction of chemical shift imaging [11], the reliance on MRI in characterisation of adrenal masses has steadily increased. Mayo-Smith et al. [2] and Outwater et al. [3] described a refinement of the CSI technique described by Mitchell that eliminated the need for complex data analysis and at the same time greatly reduced the time required to obtain the image. Sensitivity of 87% and specificity of 92% in differentia- tion between adenomas and non-adenomas were achieved with this technique [15,16]. Korobkin et al. [4] have demonstrated that visual assessment of signal loss within the adrenal is considered as accurate as the region of interest measurements. In that study, 81% sensitivity and 100% specificity were reported [4]. How- ever, histological proof was not achieved for all of the cases reported in these studies. A previous study from our study group described a sensitivity of 91%, a spe- cificity of 94%, and an accuracy of 93% in differentiat- ing between benign and malignant adrenal masses [10]. That study included 134 adrenal tumours with histolog- ical correlation in all cases. The present study includes 229 adrenal masses and proof of diagnosis has been established by surgical resection in 227 cases, and by fine needle aspiration in two cases. The long-term up- date showed a slight, but not significant decrease in sensitivity of differentiating between benign and malig- nant adrenal masses, however, the specificity and the accuracy were nearly unchanged (Table 1). In addition, the high sensitivity, specificity, and accuracy in the characterisation of adrenal adenomas, pheochromocy- tomas, adrenal carcinomas, and metastases remained comparable to the previous study (Figs. 5-9). The low number of metastases in our study is still a crucial point. This may be explained by the rareness of solitary adrenal metastasis which are operated on, and the enormous increased reliance on both, MRI and CT that has led to a significant less number of adrenal biopsies.
| Finding | Number of adrenal masses | Sensitivity (%) | Specificity (%) | Accuracy (%) | Predictive value (%) | |
|---|---|---|---|---|---|---|
| Positive | Negative | |||||
| Malignant neoplasms | 25 | 89 | 99 | 93.9 | 90.9 | 94.2 |
| Adenoma | 136 | 96.2 | 89.8 | 93.9 | 91.0 | 94.6 |
| Pheochromocytoma | 45 | 93.3 | 98.4 | 97.4 | 93.3 | 96.5 |
| Carcinoma | 10 | 93.2 | 100 | 97.8 | 100 | 96.7 |
| Metastasis | 8 | 84 | 100 | 97.4 | 100 | 95.3 |
0 1999 2001
%
100
90
80
70
60
50
40
30
20
10
0
Sensitivity
Specificity
Accuracy
o Adenoma 1999
Adenoma 2001
100
90
80
70
60
%
50
40
30
20
10
0
Sensitivity
Specificity
Accuracy
Characterisation of adrenal masses might even have been improved in our study, if the reviewer had not been blinded to clinical data and prior imaging studies. Usually a careful approach to adrenal imaging begins with knowledge of the clinical history surrounding the
diagnosis. A large irregular adrenal mass discovered as an incidental finding, has different implications than a small adrenal lesion in a patient with cancer in the clinical history or in a patient with paroxysmal hyper- tension. In addition, characteristic laboratory findings
are diagnostic for functional adenoma and pheochro- mocytoma. Our study includes 87 functional adenomas and 45 pheochromocytomas of which most have been correctly diagnosed without knowledge of the hor-
monal status and the clinical history. Adenomas usually contain large amounts of intracellular lipids, which can be easily detected by chemical shift imaging and thus, allows us to make this specific diagnosis. However, a
o 1999 2001
100
90
80
70
60
50
40
30
20
10
0
Sensitivity
Specificity %
Accuracy
1999 =2001
100
90
80
70
60
%
50
40
30
20
10
0
Sensitivity
Specificity
Accuracy
o 1999 2001
%
100
90
80
70
60
50
40
30
20
10
0
Sensitivity
Specificity
Accuracy
small number of adenomas is ‘lipid-poor’, and may be very difficult to diagnose. These lesions are variable sized, however, most of them are large (>5 cm) and show necrotic areas. In our study 13/136 adenomas were larger than 5 cm and did not show the typical appearance on MR imaging. In 9/13 of these histologi- cally proven adenomas, the right MR diagnosis was missed. Most of these lesions were interpreted as bor- derline or malignant tumours on MR imaging. This may be explained by the absence of drop in signal intensity on opposed phase images. However, our study showed no malignant lesion with decreased signal in- tensity on opposed phase images.
Many lesions, which do not meet criteria for ade- nomas or myelolipomas, still represent benign lesions. One has to be aware of hematoma, granulomatous disease, benign neurogenic tumours or just simple adrenal cysts. Pheochromocytoma is another mostly benign tumour of the adrenal gland arising in the adrenal medulla. However, release of catecholamines, produced by this tumour leads to typical clinical symp- toms, and thus makes surgical removal necessary. In addition, approximately 10% of pheochromocytomas are malignant. In our study pheochromocytoma repre- sented the second most common benign adrenal tumour.
Pheochromocytomas have also been reported to show a specific appearance on MR imaging with a characteristic bright T2 signal and bright enhancement of the lesion. Our study includes 45 pheochromocy- tomas ranging in size between 18 and 120 mm, mean
41.8 mm. MR imaging characteristics were atypical in 15 cases. Most of them showed large central necrosis (Fig. 4a-b). In addition, four pheochromocytomas oc- curring in multiple endocrine neoplasia (MEN II) pa- tients appeared inhomogeneous on T1 and T2 images and ranged in size between 18 and 25 mm. Three histologically proven pheochromocytomas were misdi- agnosed on MR imaging as two adenomas and one adrenal carcinoma. In addition, three adrenal masses were falsely classified as pheochromocytomas on MR imaging. Histology revealed two neurogenic tumours and one large adenoma with central necrosis. These incorrect diagnoses on MR imaging could have been avoided, if our study protocol had included knowledge of clinical history and blood and urine analysis.
Non-contrast CT has been reported to be a highly specific test for differentiation of adenomas from malig- nant lesions by several authors [7,9,17-19]. There is a consensus from multiple institutions that lesions of less than 10 HU are most likely benign. However, a Houn- sfield number above the threshold value does not neces- sarily imply a malignancy. In addition, it is unclear what accounts for the occasional case of malignant lesion with very low density. A further criticism is that it requires many patients to return to the CT scanner the next day, since most CT scans are obtained with intravenous contrast media. Boland et al. [8] have shown that delayed enhanced CT scans can lead to significant differences in density between adenomas and metastases. Similar results were reported by several other authors [9,20,21]. One problem with this tech-
nique is that it clearly depends on a well-calibrated CT unit and a uniform injection of contrast media. More- over, it is unclear what effect renal function has on the threshold values evaluated by the above mentioned studies.
In conclusion, the results of our study showed that large study numbers confirm the reliance on MR imag- ing in characterisation of adrenal masses. In addition, the results of this study could have been improved by knowledge of clinical history and laboratory parame- ters. CT is readily available and has good sensitivity in detection and characterisation of adrenal lesions. How- ever, CT also has limitations. MR imaging is a compet- itive method and does not apply radiation dose. The increased reliance on MR imaging, however, is already translating into improved patient care and has clearly reduced the number of invasive procedures such as adrenal biopsy. In case of surgical treatment, the high accuracy of MR imaging offers advantages of the mini- mally invasive technique.
References
[1] Reinig JW, Stutley JE, Leonhardt CM, Spicer KM, Margolis M, Caldwell CB. Differentiation of adrenal masses with MR imag- ing: comparison of techniques. Radiology 1994;192:41-6.
[2] Mayo-Smith WW, Lee MJ, McNicholas MM, Hahn PF, Boland GW, Saini S. Characterization of adrenal masses by use of chemical shift MR imaging: observer performance versus quanti- tative measures. Am J Roentgenol 1995;165:91-5.
[3] Outwater EK, Siegelman ES, Radecki PD, Piccoli CW, Mitchell DG. Distinction between benign and malignant adrenal masses: value of T1-weighted chemical-shift MR imaging. Am J Roentgenol 1995;165:579-83.
[4] Korobkin M, Lombardi TJ, Aisen AM, et al. Characterization of adrenal masses with chemical-shift and gadolinium-enhanced MR imaging. Radiology 1995;197:411-8.
[5] Korobkin M, Giordano TJ, Brodeur F, et al. Adrenal adenomas: relationship between histologic lipid and CT and MR findings. Radiology 1996;200:743-7.
[6] Outwater EK, Siegelman ES, Huang AB, Birnbaum BA. Adrenal masses: correlation between CT attenuation value and chemical shift ratio at MR imaging with in-phase and opposed- phase sequences. Radiology 1996;200:749-52.
[7] Szolar DH, Kammerhuber F. Quantitative CT evaluation of adrenal gland masses: a step forward in the differentiation between adenomas and non-adenomas? Radiology 1997;202:517-21.
[8] Boland GW, Hahn PF, Pena C, Mueller PR. Adrenal masses: characterisation with delayed contrast-enhanced CT. Radiology 1997;202:693-6.
[9] Korobkin M, Brodeur FJ, Francis IR, Quint LE, Dunnick NR, Goodsitt M. Delayed enhanced CT for differentiation of benign from malignant adrenal masses. Radiology 1996;200:737-42.
[10] Heinz-Peer G, Hönigschnabl S, Schneider B, Niederle B, Kaserer K, Lechner G. Characterization of adrenal masses using MR imaging with histopathologic correlation. Am J Roentgenol 1999;173:15-22.
[11] Mitchell DG, Crovello M, Matteucci T, Petersen RO, Miettinen MM. Benign adrenocortical masses: diagnosis with chemical MR imaging. Radiology 1992;185:345-51.
[12] Krestin GP, Steinbrich W, Friedmann G. Adrenal masses: evalu- ation with fast gradient-echo MR imaging and Gd-DTPA-en- hanced dynamic studies. Radiology 1989;171:675-80.
[13] Fink IJ, Reinig JW, Dwyer AJ, Doppman JL, Linehan WM, Keiser HR. MR imaging of pheochromocytomas. J Comput Assist Tomogr 1985;9:454-8.
[14] Lee MJ, Mayo-Smith WW, Hahn PF, et al. State-of-the-art MR imaging of the adrenal gland. RadioGraphics 1994;14:1015-29.
[15] Bilbey JH, McLoughlin RF, Kurkjian PS, et al. MR imaging of adrenal masses: value of chemical-shift imaging for distinguish- ing adenomas from other tumours. Am J Roentgenol 1995;164:637-42.
[16] McNicholas MMJ, Lee MJ, Mayo-Smith WW, Hahn PF, Boland GW, Mueller PR. An imaging algorithm for the differen- tial diagnosis of adrenal adenomas and metastases. Am J Roentgenol 1995;165:1453-9.
[17] Lee MJ, Hahn PF, Papanicolaou N, Egglin TK, Saini S, Mueller PR, Simeone JF. Benign and malignant adrenal masses: CT distinction with attenuation coefficients, size, and observer anal- ysis. Radiology 1991;179:415-8.
[18] Singer AA, Obuchowski NA, Einstein DM, Paushter DM. Metastase or adenoma? Computed tomographic evaluation of the adrenal mass. Cleve Clin J Med 1994;61:200-5.
[19] Boland GWL, Lee MJ, Gazelle GS, Halpern EF, McNicholas MMJ, Mueller PR. Characterization of adrenal masses using unenhanced CT: an analysis of the CT literature. Am J Roentgenol 1998;171:201-4.
[20] Brodeur FJ, Korobkin MT, Francis IR, Quint LE, Dunnick NR. Delayed enhanced CT: a method of differentiating adrenal ade- nomas from non-adenomas (abstract). Radiology 1995;197:185.
[21] Szolar DH, Kammerhuber FH. Adrenal adenomas and non-ade- nomas: assessment of washout at delayed contrast enhanced CT. Radiology 1998;207:369-75.