Differentiation between Benign and Malignant Adrenal Mass using Contrast-Enhanced Ultrasound
Kontrastmittel-Sonografie zur Differenzierung von benignen und malignen Nebennierenraumforderungen
Authors
M. Friedrich-Rust1, T. Glasemann1, A. Polta1, K. Eichler2, K. Holzer3, S. Kriener4, E. Herrmann5, J. Nierhoff5, D. Bon5, W. O. Bechstein3, T. Vogl2, S. Zeuzem1, J. Bojunga1
Affiliations
Affiliation addresses are listed at the end of the article.
Key words
abdomen
adrenal gland
ultrasound
ultrasound power Doppler
received accepted
2.6.2010
25.4.2011
Bibliography
DOI http://dx.doi.org/10.1055/ s-0031-1273408 Published online June 10, 2011 Ultraschall in Med 2011; 32: 460-471 @ Georg Thieme Verlag KG Stuttgart · New York · ISSN 0172-4614
Correspondence
PD Dr. Mireen Friedrich-Rust Department of Internal Medicine 1, J. W. Goethe-University Hospital Theodor-Stern-Kai 7 60590 Frankfurt Tel .: ++ 49/69/63 01 52 97
Fax: ++ 49/69/63 01 51 22
Zusammenfassung
!
Ziel: Nebennierenraumforderungen können so- nografisch mit guter Sensitivität und Spezifität detektiert werden. Ziel der vorliegenden Studie war es, die Wertigkeit der Kontrastmittel-Sono- grafie (CEUS) unter Verwendung von Zeit-Intensi- täts-Kurven für die Charakterisierung von Neben- nierenraumforderungen zu evaluieren.
Material und Methoden: 108 Patienten mit 116 Raumforderungen der Nebennieren erhielten einen Ultraschall der Nebenniere, inklusive einer Kontrastmittel-Sonografie mit dem Kontrastmittel Sonovue. Die Kontrastmitteldynamik wurde mit- tels Zeit-Intensitätskurven aufgezeichnet. Anhand der Kontrastmittelanflutungszeit (CE) wurden fol- gende 4 Kontrastmittel-Muster verwendet, welche in einer Pilotstudie entwickelt wurden: pattern I=früh arterielle CE, pattern II = arterielle CE, pat- tern III = späte CE, pattern IV = keine CE. Zusätzlich erhielten alle Patienten eine Beurteilung der Nebenniere mit einer 2. Bildgebung (CT oder MRT) sowie eine laborchemische Hormondiagnostik. In Fällen mit malignitätsverdächtigem Befund erfolgte eine Biopsie oder Adrenalektomie.
Ergebnisse: CEUS-pattern-I&II konnte bei allen Patienten mit primärem oder sekundärem Malig- nom der Nebenniere (n=16) nachgewiesen wer- den. Die Sensitivität der CEUS für die Diagnose ei- ner malignen Nebennierenraumforderung betrug 100% (CI [75;100]) und die Spezifität 67% (CI [56;75]). Bei 40 Nebennierenraumforderungen war die Histologie die Referenzmethode. In 68% dieser Nebennierenraumforderungen waren MRT/ CTund CEUS kongruent in der Beurteilung der Dig- nität.
Schlussfolgerung: Die Kontrastmittel-Sonografie ist eine nützliche ergänzende Methode in der Aufarbeitung von Nebennierenraumforderungen mit exzellenter Sensitivität für die Diagnose eines Malignoms.
Abstract
!
Purpose: Adrenal masses can be detected by ul- trasound with high sensitivity and specificity. The aim of the present study was to evaluate CEUS in a large patient population using CEUS patterns identified in a previous pilot study.
Materials and Methods: 116 adrenal masses were evaluated by ultrasound, including CEUS with the contrast agent Sonovue®. The dynamic of contrast enhancement (CE) was analyzed using time-intensity curves. The time of the first CE in the adrenal mass was used to define four CEUS patterns: pattern I=early arterial CE, pattern II = arterial CE, pattern III = late CE, pattern IV = no CE. In addition, all patients received CT/MRI and hormonal testing. In suspicious cases biopsy or adrenalectomy was performed.
Results: CEUS patterns I&II were seen in all pa- tients with primary or secondary malignant le- sions of the adrenal gland (n= 16). The sensitivity and specificity of CEUS for the diagnosis of malig- nant adrenal mass were 100% (CI [75;100]) and 67% (CI [56;75]), respectively. Overall histology was available as a reference method for 40 adre- nal masses. In 68% of histologically diagnosed adrenal masses, MRI/CT and CEUS were congru- ent concerning the characterization of malignant versus benign adrenal mass.
Conclusion: Contrast-enhanced ultrasound may be a useful method in the diagnostic work-up of adrenal mass with excellent sensitivity for the diagnosis of malignancy.
Introduction
!
Adrenal masses are one of the most common endocrine neo- plasms with a reported overall frequency of about 6% (range 1-32) in autopsy studies [1, 2]. Besides classic hormone se- creting tumors, adrenocortical carcinomas and metastatic car- cinomas, the vast majority of adrenal masses are non-hyperse- creting, benign adrenocortical adenomas. The widespread use of imaging modalities - ultrasound, CT and MRI - has resulted in an increasing clinical necessity for a rational diagnostic work-up. In addition to a careful medical history and hormo- nal testing, imaging modalities like CT and MRI are the corner- stone of diagnosing adrenal masses.
However, adrenal masses can be detected by ultrasound with high sensitivity and specificity as well [3-7] and visualization by transabdominal B-mode ultrasound has been reported in 99% (79/80) for the right adrenal gland and 69% (55/80) for the left adrenal gland [3]. Although studies showed that be- nign adrenal masses are on average smaller than malignant adrenal masses, no conclusion concerning the risk for malig- nancy of an adrenal mass could be drawn solely from the size of the adrenal mass [5]. Therefore, the differentiation be- tween benign and malignant adrenal masses by conventional ultrasound was previously not sufficiently possible.
Contrast-enhanced ultrasound as a tool for the characteriza- tion of adrenal masses has only been evaluated in a few pub- lished studies. While a first pilot study using Levovist® only showed improved visualization of the vascularization of adre- nal masses without differentiation between benign and malig- nant adrenal masses [8], two other studies [9, 10] using Sono- Vue® showed that although there are no specific CEUS patterns distinguishing benign from malignant adrenal masses [10] the analysis of the dynamics of contrast enhancement using time-intensity curves showed (early) arterial contrast enhancement and rapid wash-out in all malignant adrenal masses resulting in a sensitivity of 100% and specificity of 82% in a pilot study [9]. Based on the results of the pilot study, a sample size calculation was performed, suggesting the evaluation of 100 patients by assuming a number of 15 malignant lesions. This sample size allows estimation of sensi- tivity and specificity with 95% confidence intervals with a re- liable overall length (a length of 20% is then possible for a sensitivity of around 96% and a specificity of around 67%). The aim of the present study was to prospectively evaluate CEUS of adrenal masses in the calculated sample size and to validate the CEUS patterns identified in the pilot study.
Materials and Methods
!
Patients
Between April 2006 and December 2009, all patients referred to our endocrinology referral center for further diagnostic evaluation of adrenal masses were screened for inclusion in the present study. The inclusion criteria were: presence of adrenal mass, reference method (CT/MRI or histology) per- formed or planned, age ≥18 years of age, written informed consent. The exclusion criteria were: missing reference method (CT/MRI or histology for benign adrenal mass and histology for malignant adrenal mass), patients with adrenal metastases who received chemotherapy within the last 6 months, preg- nancy or lactation, suspected allergy to the contrast agent
Sonovue®, cardiac insufficiency NYHA II-IV, acute coronary artery syndrome or clinically unstable ischemic heart disease within the previous 7 days, severe cardiac arrhythmias, acute heart failure, severe pulmonary hypertension (pulmonary ar- terial pressure >90 mmHg), uncontrolled systemic blood pres- sure, acute respiratory distress syndrome.
108 consecutive patients (47 men, 61 women; age range 27 - 87, mean age 59 years) fulfilled the inclusion criteria and were enrolled in the study. The adrenal masses were detected pri- marily at ultrasound (n=46), magnetic resonance imaging (MRI) (n=21) or computed tomography (CT) (n=41). The adrenal mass was detected incidentally (n=98), during screen- ing of patients with primary malignoma of an organ other than the adrenal gland (n=6), or during work-up of patients with therapy refractory hypertension (n=4). 34 of the 108 pa- tients had already been included in the pilot study [9]. The pi- lot study consisted of 35 patients. However one patient with metastasis of the adrenal gland, who had received successful chemotherapy, was excluded from the present study since this was one of the exclusion criteria of the present study.
In addition to the overall analysis, a subanalysis was per- formed analyzing only the 74 patients with 82 adrenal masses, and excluding the patients of the pilot study. The aim of this subanalysis was to validate the results of the pilot study in an independent patient group.
All patients received an ultrasound of the adrenal mass, in- cluding contrast-enhanced ultrasound. In addition, all patients underwent CT or MRI. Patients without clear signs of a benign adrenal mass on MRI or CT scan underwent fine needle biopsy or surgery with adrenalectomy and histological evaluation of the specimen. To analyze hormonal activity, detailed labora- tory testing was performed (see below).
The present study was performed in accordance with the ethi- cal guidelines of the Helsinki Declaration and was approved by the local ethical committee. Written informed consent was ob- tained from each patient.
Hormonal evaluation
Besides routine laboratory testing, a specific hormonal evalua- tion was performed after overnight fasting to detect hormonal activity of the adrenal mass. Hormonal evaluation included ba- sal parameters for adrenocorticotropic hormone (ACTH), corti- sol, dehydroepiandrosterone (DHEAS), aldosterone, renin activ- ity, plasma-free normetanephrines, plasma-free metanephrines and chromogranin A. In addition, an overnight dexamethasone suppression test using 1 mg dexamethasone was performed to rule out Cushing’s syndrome if suspected. The laboratory re- sults were judged by an experienced endocrinologist.
Ultrasound examination
The ultrasound examination (Hitachi EUB-8500/ EUB-900, Hi- tachi, Tokyo, Japan) was performed by two experienced exam- iners blinded to the results of MRI/CT. The adrenal mass was examined for size, echogenicity, echotexture, margins and per- fusion pattern. Using power Doppler ultrasound, the vasculari- zation was classified as hypovascular, isovascular or hypervas- cular in comparison to the vascularization of the liver or spleen. Specific vascular patterns, such as an afferent blood vessel or irregular tumor vessels, were recorded.
Contrast-enhanced dynamic ultrasound was performed with contrast-specific, continuous-mode software (Hitachi EUB- 8500/ EUB-900, Hitachi, Tokyo, Japan) using a low mechanical
index (MI <0.11). Sulfur hexafluoride (SonoVue®, Bracco Inter- national, Milan, Italy) was used as the ultrasound contrast agent. SonoVue® was injected once as an intravenous bolus of 4.8 ml via a 20-gauge cannula into an antecubital vein within less than 2 sec, followed by a 10 ml saline flush. The time of injection was recorded starting with - 5sec on the display of the ultrasound machine. The adrenal mass was scanned con- tinuously for 5 min using the liver or the spleen as an in vivo reference. The patients were asked to breath superficially only to prevent severe motion of the adrenal gland during the 5 min of acquisition. In case of very small lesions or heavy mo- tion during breathing, the patients were asked to hold their breath intermittently.
The video clip was digitally recorded. After the real-time ex- amination the video clips were reviewed and the contrast en- hancement pattern within the adrenal mass was judged as homogeneous or inhomogeneous, and specific perfusion pat- terns like central, peripheral or chaotic contrast enhancement were assessed. Regions-of-interest (ROI) were then placed in the adrenal mass and in the reference organ (spleen, liver) and analyzed over time by creating time-intensity curves (TIC). In the case of homogenous contrast enhancement within the adrenal mass, the ROI was placed over the entire adrenal mass. In the case of inhomogeneous contrast enhancement, multiple ROIs with a size of at least 5 mm were placed at dif- ferent sites within the adrenal mass. The time-intensity curve with the quickest dynamic was judged as relevant and was in- cluded in the final analysis, since less or no contrast enhance- ment was suggestive of scar tissue or necrotic tissue. Intermit- tent blurring caused by intermittent breathing in small lesions was neglected. The time of first contrast enhancement was ca- tegorized according to the four patterns described earlier using the TIC analysis [9]: pattern I=early arterial contrast enhance- ment (<20 sec), pattern II =arterial contrast enhancement be- ginning in the arterial phase (21-40 sec), pattern III = late and little contrast enhancement in the late phase (>40sec), and pattern IV = no recordable contrast enhancement at all. The ra- tio of maximum contrast enhancement intensity in the adre- nal mass and the reference organ was calculated. The time of contrast washout was documented.
CT/MRI
Triple-phase spiral CT examination together with a delayed contrast-enhanced CT scan (Imeron 400 [Bracco Altana Phar- ma, Konstanz, Germany] [400 mgI/mL] or Ultravist 370 [Bayer Schering Pharma, Berlin, Germany] [370 mg/mL]) were exclu- sively accepted for analysis. 90-100 ml of contrast agent was injected at a rate of 2-4ml/sec. The contrast phase was run 20 sec after injection for the arterial phase, and 85 sec after injection for the venous phase. The percentage of washout was calculated by the following formula: (1-delayed enhanced HU value/initial enhanced HU value)× 100 [11]. If no clear cri- teria for benign adrenal mass were present (unenhanced CT attenuation ≤10 Hounsfield units and CT contrast medium washout ≥50% at 6 min after contrast medium administration) [1, 11], an additional MRI examination was performed. MR imaging was performed on a 1.5-Tesla scanner (Avanto, Sonata, Symphony, Siemens, Erlangen, Germany). The MRI examina- tions included unenhanced T1-weighted and T2-weighted scans, chemical shift imaging (in-phase and out-of-phase im- aging) and dynamic T1-weighted contrast-enhanced imaging
(Gadolinium (Gadopentetat) Bayer Schering Pharma, Leverku- sen, Germany).
MRI was performed by using T2-HASTE-sequence coronal and transversal, T1-weighted FLASH-2D-sequence, T1-weighted turbo spin echo-sequences and T1-weighted FLASH-2D in- phase images. For dynamic contrast-enhanced T1w vibe se- quences during the arterial, portal venous, and equilibrium phases of contrast enhancement, an additional T1-weighted FLASH-2D sequence was applied. All sequences were per- formed during breath-hold. Administration of the contrast agent was carried out using a power injector at a constant injection rate of 2.0 ml/sec. The dose administered was 0.1 mmol/kg bodyweight corresponding to 0.2 mL/kg body- weight of a 0.5 mol/L formulation injected in a volume based on bodyweight and followed by a 20-mL saline flush.
The criteria for the diagnosis of adenoma on MRI were: (1) homogeneous SI on unenhanced T2-weighted and T1-weigh- ted images; (2) homogeneous enhancement on dynamic ima- ging post-contrast medium injection (0.1 mmol/kg bodyweight Gd-DTPA); and (3) loss of SI of the lesion on out-of-phase ima- ges [12]. As an internal standard, the SI of the adrenal mass was compared to the SI of the spleen.
Final reference diagnosis
The reference standard was determined by an experienced radiologist together with an experienced endocrinologist. The final reference diagnosis was defined by combining all avail- able information from imaging (CT/MRI) plus additional infor- mation from the hormonal evaluation and from histology. Histological verification of adrenal masses after adrenalectomy was available in 30 adrenal masses of 28 patients following surgery and in 10 adrenal masses of 10 patients following fine needle biopsy (6 metastases, 1 benign follicular spreading tumor, 1 adenoma, 1 B-cell lymphoma and 1 metastasized adrenocortical carcinoma). In the remaining 76 adrenal masses of 70 patients, the diagnosis was confirmed by means of CT (n=26) or MRI (n=50). In the 6 patients with metastases, the primary malignoma was diagnosed prior to the diagnosis of adrenal mass. The primary malignoma was bronchial carcino- ma in 4 cases, renal cell carcinoma in 1 case and a neuroendo- crine tumor in 1 case. The patients with T-cell and B-cell lym- phoma did not have extra adrenal malignant tissue diagnosed prior to the evaluation of the adrenal mass. The diagnosis was made after histology of the adrenal mass.
Statistical analysis
Statistical analysis was performed using SigmaPlot and Sigma- Stat for Windows (version 11.0, Systat Software, Inc. Germa- ny), BiAS for Windows (version 9.08, epsilon 2010, Frankfurt, Germany) and R (version 2.12.1, R Foundation for Statistical Computing, Vienna, Austria). Clinical and laboratory character- istics of patients were expressed as mean +SD, median and range. Fisher’s exact test was used to evaluate the criteria for the diagnosis of malignant adrenal masses and non-adenoma. The baseline characteristics of patients with benign and malig- nant adrenal masses were compared using the Wilcoxon- Mann-Whitney two-sample rank-sum test. A p-value less than 0.05 was judged to be statistically significant. Areas under the ROC curves (AUROC) as well as 95% CI of AUROC were calcul- ated. Here, values of AUROC close to 1.0 indicate the highest diagnostic accuracy.
Downloaded by: National University of Singapore. Copyrighted material.
Results
!
108 patients with 116 adrenal masses were included in the overall analysis. The right and left adrenal glands were in- volved equally in 58 cases each. All lesions which were pri- marily detected on CT or MRI could be visualized with ultra- sound. In three additional patients with a right adrenal mass with BMI >35 kg/m2, CEUS was not possible due to steatosis of the liver without contrast enhancement of the dorsal liver area. The contrast enhancement in the adrenal gland could not be assessed. Therefore these patients were not included in the present study.
The adrenal masses consisted of 16 malignant and 100 benign adrenal masses (79 adenomas and 21 non-adenomas). The overall 37 non-adenomas (16 malignant and 21 benign) were in detail: 11 benign pheochromocytomas, 1 malignant pheo- chromocytoma, 6 adreno-cortical carcinomas, 6 metastases, 5 (angio)myelolipomas, 3 hemangiomas, 1 B-cell lymphoma, 1 T-cell lymphoma, 1 neuroendocrine carcinoma, 1 benign fol- licular spreading tumor, and 1 ganglioneuroma.
The 16 patients (50% male) with 16 malignant adrenal masses had a median age of 62 years (mean + SD: 63 +10, range: 47- 87), a median bodyweight of 72 kg (mean + SD: 72±15, range: 48-106), a median height of 169cm (mean+ SD: 170±8, range: 160-191 cm), and a median BMI of 25 (mean ± SD: 25 +5, range: 17-33). The 92 patients (48% male) with 100 be- nign adrenal masses had a median age of 62 years (mean ± SD: 59+14, range: 27-87), a median bodyweight of 78 kg (mean+ SD: 77±14, range: 48-110), a median height of 168 cm (mean ± SD: 169+9, range: 138-190cm), and a me- dian BMI of 27 (mean ± SD: 27+4, range: 18-38). No signifi- cant difference was found for these characteristics between the group of patients with benign and malignant adrenal mas- ses.
Hormonal evaluation
11 patients with 12 adrenal masses presented with clinical or laboratory signs of pheochromocytoma with plasma meta- nephrine levels of 883-3788 pg/ml (normal range <90 pg/ml) and plasma normetanephrine levels of 408 -3000 pg/ml (nor- mal range <200 pg/ml). While 4 patients were evaluated due to therapy refractory hypertension, no clinical signs of pheo- chromocytoma were present in 7 patients. The diagnosis was made during hormonal evaluation after the diagnosis of an in- cidentally detected adrenal mass. All these patients received adrenalectomy and the diagnosis of pheochromocytoma was confirmed. 2 patients with adenoma presented with clinical and laboratory signs of Cushing’s syndrome. The dexametha- sone overnight suppression test was pathological with a corti- sol level of 24 and 33 ug/dl (normal range <1.8ug/dl) at 8:00 a.m. following a dose of 1 mg dexamethasone at midnight. Both patients received adrenalectomy and histology revealed adenoma without signs of malignancy. One patient without clinical symptoms had laboratory signs of subclinical Cushing’s syndrome with a cortisol level of 5.1 ug/dl after a dexametha- sone test.
8 patients with adrenal masses presented with laboratory signs of Conn’s syndrome. Aldosterone levels were significantly elevated with 483 - 1162 pg/ml (normal range <285 pg/ml) and a plasma-aldosterone concentration to plasma-renin-activity ratio of 723-3195 (normal range <20). 3 patients received adrenalectomy and histology revealed adenoma without signs
of malignancy and 5 patients were treated medically. Hormo- nal analysis was not available in a few patients (n=21) with- out clinical signs of endocrine activity and with more than six months of follow-up. Chromogranin A was significantly eleva- ted in the patient with the neuroendocrine carcinoma of the adrenal gland. 3 of the overall 6 adrenocortical carcinomas were functioning. 1 presented with increased cortisol produc- tion and 2 with increased metanephrine levels.
B-mode and power Doppler ultrasound
108 patients with 116 adrenal masses were included in the study. No specific shape (round, oval, polycyclic) of adrenal mass in B-mode was found for malignant and benign lesions (p=0.59). However, malignant lesions showed significantly more often an inhomogeneous echotexture (56 vs. 23%; p=0.013) on B-mode and a hypoechoic or echocomplex echo- genicity (94 vs. 67%; p=0.037). The mean size of benign le- sions was 33 mm length×24 mm width (median 19×27 mm, range 5-126mm); the mean size of malignant lesions was 63 mm length × 80 mm width (median 55×72 mm, range 18 - 190 mm). A size greater than 4cm was found significantly more often in malignant than in benign lesions (81 vs. 21%; p=0.0000050). The area under the ROC curve (AUROC) for the diagnosis of malignant adrenal masses was 86% (95% CI: 0.75-0.97) for the length of the adrenal mass with an optimal cut-off of 57.5 mm (sensitivity of 82% and specificity of 91%). The area under the ROC curve (AUROC) for the diagnosis of malignant adrenal masses was 89% (95% CI: 0.81 -0.96) for the width of the adrenal mass with an optimal cut-off of 32 mm (sensitivity of 88% and specificity of 82%). Power Dop- pler examination revealed hypervascularization or an afferent blood vessel in 12 cases. Hypervascularization of the adrenal mass and/or an afferent blood vessel was found significantly more often in malignant lesions than in benign lesions (31% vs. 7%; p=0.012). The 5 (angio)myelolipomas showed a charac- teristic homogenous hyperechoic echogenicity with regular margins. The diagnosis could be made by B-mode ultrasound in all 5 cases.
Details are shown in . Table 1 for the overall analysis and in Table 2 excluding the patients of the pilot study.
Contrast-enhanced ultrasound
No specific perfusion patterns like central, peripheral or cha- otic contrast enhancement could be found to differentiate benign and malignant adrenal masses. 32 adrenal masses showed early arterial contrast enhancement (time of first con- trast enhancement <20 sec=pattern I) with an intensity peak between 11 and 50 sec, followed by a rapid washout, 17 adre- nal masses showed arterial contrast enhancement (21 - 40 sec = pattern II) with an intensity peak after 29-110 sec, 32 adrenal masses demonstrated late and little contrast enhancement in the late phase (>40 sec =pattern III) with an intensity peak after 50-193 sec, and 35 adrenal masses showed no recordable contrast enhancement at all (pattern IV) ( Fig. 1). CEUS patterns according to the tumor entity are shown in @ Table 3.
Using pattern III&IV as criteria for benign adrenal masses and pattern I&II as criteria for malignant lesions according to the pilot study [9], the diagnosis of a malignant lesion could be made by CEUS with a sensitivity of 100% (95% CI: 0.79 -1.0), a specificity of 67% (95% CI: 0.57 -0.76), a PPV of 33% (95% CI: 0.20-0.48), and an NPV of 100% (95% CI: 0.95-1.0).
| US pattern | benign (n= 100) malignant (n=16) sum (n=116) | p-value | sens (%) (95% CI) | spec (%) (95% | CI) P PPV (%) (95% CI) | NPV (%) (95% CI) | Odds ratio |
|---|---|---|---|---|---|---|---|
| increased vascularization using power | |||||||
| Doppler | 42 (15- 72) | 89 (82- 95) | |||||
| yes | 7 5 12 | 0.012 | 31 (11 -59) | 93 (86- 97) | 6.04 (1.82 - 20.0) | ||
| no | 93 11 104 | ||||||
| hypoechoic/echocomplex echogenicity | |||||||
| yes | 67 15 82 | 0.037 | 94 (70 - 100) | 33 (24- 43) | 18 (11-28) | 97 (85- 100) | 7.39 (1.22 - 44.9) |
| no | 33 1 34 | ||||||
| inhomogeneity | |||||||
| yes | 23 9 32 | 0.013 | 56 (30-80) | 77 (68 - 85) | 28 (14-47) | 92 (84-97) | 4.3 (1.52-12.2) |
| no | 77 7 84 | ||||||
| irregular margin | |||||||
| yes | 24 6 30 | 0.35 | 38 (15-65) | 76 (66-84) | 20 (8-39) | 88 (80-94) | 1.9 (0.63 -5.73) |
| no | 76 10 86 | ||||||
| round or polycyclic shape | |||||||
| yes | 61 11 72 | 0.78 | 69 (41 - 89) | 39 (29 - 49) | 15 (8 -26) | 89 (75- 96) | 1.41 (0.45 - 4.36) |
| no | 39 5 44 | ||||||
| size ≥4 cm | |||||||
| yes | 21 13 34 | 0.0000050 | 81 (54-96) | 79 (70-87) | 38 (22 - 56) | 96 (90-99) | 16.3 (5.33 - 49.9) |
| no | 79 3 82 | ||||||
| CEUS pattern I/ II | |||||||
| yes | 33 16 49 | 1.9×10-7 | 100 (79- 100) | 67 (57-76) | 33 (20 - 48) | 100 (95-100) | 66.5 (12.9- 342) |
| no | 67 0 67 | ||||||
| CEUS pattern I/II without 5 (angio)myeloli- poma and without 12 pheochromocytoma (n= 99) | |||||||
| yes | 17 15 32 | 2.6×10-9 | 100 (78 -100) | 80 (70-88) | 47 (29-65) | 100 (95-100) | 119 (25.4-563) |
| no | 67 0 67 |
1 US = ultrasound; benign = number of benign adrenal masses; malignant = number of malignant adrenal masses; sens = sensitivity; spec = specificity; PPV = positive predictive value; NPV = negative predictive value; CEUS = con- trast-enhanced ultrasound; CI = confidence interval.
| US pattern | benign (n=72) | malignant (n= 10) | sum (n=82) | p-value | sens (%) (95% CI) | spec (%) (95% CI) | PPV (%) (95% CI) | NPV (%) (95% CI) | Odds ratio |
|---|---|---|---|---|---|---|---|---|---|
| increased vascularization using power | |||||||||
| Doppler | |||||||||
| yes | 5 | 2 | 7 | 0.20 | 20 (2.5- 56) | 93 (85-98) | 29 (3.7- 71) | 89 (80-95) | 3,35 (0.60 - 18.8) |
| no | 67 | 8 | 75 | ||||||
| hypoechoic/echocomplex echogenicity | |||||||||
| yes | 53 | 10 | 63 | 0.11 | 100 (69- 100) | 26 (17- 38) | 16 (8-27) | 100 (82-100) | 7,65 (0.88 - 66.8) |
| no | 19 | 0 | 19 | ||||||
| inhomogeneity | |||||||||
| yes | 17 | 5 | 22 | 0.12 | 50 (19-81) | 76 (65-86) | 23 (8 - 45) | 92 (82-97) | 3,24 (0.87 - 12.0) |
| no | 55 | 5 | 60 | ||||||
| irregular margin | |||||||||
| yes | 14 | 3 | 17 | 0.43 | 30 (7- 65) | 81 (70- 89) | 18 (4-43) | 89 (79-96) | 1,78 (0.41 - 7.70) |
| no | 58 | 7 | 65 | ||||||
| round or polycyclic shape | |||||||||
| yes | 43 | 6 | 49 | 1.00 | 60 (26 -88) | 40 (29- 53) | 12 (5 - 25) | 88 (72-97) | 1,01 (0.26 - 3.93) |
| no | 29 | 4 | 33 | ||||||
| size ≥ 4 cm | |||||||||
| yes | 39 | 10 | 49 | 0.0048 | 100 (69- 100) | 46.0 (34-58) | 20 (10-34) | 100 (89-100) | 17,8 (2.29 - 138) |
| no | 33 | 0 | 33 | ||||||
| CEUS pattern I/II | |||||||||
| yes | 27 | 10 | 37 | 0.00016 | 100 (69 - 100) | 63 (50 - 74) | 27 (14-44) | 100 (92 - 100) | 34,8 (5.30 - 228) |
| no | 45 | 0 | 45 | ||||||
| CEUS pattern I/II without 4 (angio)mye- lolipoma and without 11 pheochromo- cytoma (n = 67) | |||||||||
| yes | 16 | 9 | 25 | 0.000048 | 100 (66- 100) | 72 (59- 83) | 36 (18-57) | 100 (92- 100) | 48,9 (7.79 - 308) |
| no | 42 | 0 | 42 |
1 US = ultrasound; benign = number of benign adrenal masses; malignant = number of malignant adrenal masses; sens = sensitivity; spec = specificity; PPV = positive predictive value; NPV = negative predictive value; CEUS = contrast-enhanced ultrasound; CI = confidence interval.
Intensity of contrast-enhancement
seconds
0
20
40
60
80
100
120
Downloaded by: National University of Singapore. Copyrighted material.
Calip
000
D1:
31.3 mm
06
D2:
24.3 mm
D1
D2
No.122/122
042
T1
T2
T3
T4
T5
T6
17
T8
T9
T10
G
>
When excluding the 5 (angio)myelolipomas, which could be cor- rectly diagnosed by B-mode ultrasound and the 12 pheochromo- cytomas, which are diagnosed by hormonal and clinical evalua- tion, the specificity for the diagnosis of a malignant lesion made
by CEUS increased to 80% (95% CI: 0.70-0.88) without a reduc- tion of sensitivity. Details are shown in . Table 1 for the overall analysis and in @ Table 2 excluding patients of the pilot study.
Calip
D1:
19.7 mm
000
D2:
15.1 mm
+
D1
+
a
D2
b
C
d
b
T
(s)
a
C
TT
1217
T4
T1
19.9
T2
39.8
T3
44.1
T4
79.7
Add
Dilere
TO
Set TO
When using pattern III and IV as criteria for adenoma and pat- tern I and II as criteria for non-adenoma, the diagnosis of a non- adenoma could be made by CEUS with a sensitivity of 87% (95% CI: 0.71 - 0.95), a specificity of 78% (95% CI: 0.68 - 0.87), an PPV of 65% (95% CI: 0.50-0.78), and a NPV of 93% (95% CI: 0.83 - 0.98).
The area under the ROC curve (AUROC) for the diagnosis of malignant adrenal masses was 69% (95% CI: 0.57 -0.81) for the ratio of maximum contrast enhancement in the adrenal mass and the adjacent organ (spleen/liver) with an optimal cut-off of 0.1 (sensitivity of 94% and specificity of 42%). Exam- ples are shown in . Fig. 2-6.
Contrast-enhanced ultrasound in comparison to CT, MRI using histology as a reference method
In 40/116 of adrenal masses, histological diagnosis (30 from surgery and 10 from biopsy) was available. CT was available in 21/40 cases and MRI in 19/40 cases. CT/MRI was indeter- minate with respect to the diagnosis of benign or malignant adrenal masses in 12 cases (9 pheochromocytomas, 2 adeno- mas, 1 T-cell lymphoma). CEUS could correctly diagnose 4 adrenal masses among these indeterminate lesions.
In the 28 remainders CT/MRI was correct with respect to the diagnosis of malignant adrenal masses in 21 cases and incor- rect in 7 cases (5 adenomas, 1 benign follicular spreading tu- mor, 1 hemangioma). CEUS could correctly classify 4 lesions among these 7 incorrectly diagnosed lesions.
Downloaded by: National University of Singapore. Copyrighted material.
a
b
C
1000
H
000
T
TI
12
T.TE
T1
T2
14.2
25.6
37.0
A
T3
T4
39.8
T5
79.7
FR Interpolate Base Frame
Add
Set Start Iyect
Mag X1
MOVE
Current Frame
Delete
1500
To
d
৳
D
D
WIDE
TIC Analysis
Close
e
Set TO
Calip
D1:
78.1 mm
063
000
D2:
35.2 mm
¢
D1
৳
5.
Y
D2
a
b
C
A
Downloaded by: National University of Singapore. Copyrighted material.
B-mode ultrasound and CEUS were incorrect with respect to the diagnosis of malignant adrenal masses in 16 cases (8 pheochro- mocytomas, 4 adenomas, 3 hemangiomas, 1 ganglioneuroma). CT/MRI was correct in 2 cases among these 16 incorrectly diag- nosed lesions.
CT/MRI and ultrasound (B-mode +CEUS) were congruent in 68% of adrenal masses with respect to the diagnosis of benign and malignant adrenal masses and in 70% with respect to the diagnosis of adenomas and non-adenomas.
Discussion
!
The incidence of adrenal masses reported from autopsy and CT studies is 4-6% [2, 13, 14]. Non-invasive differentiation be- tween benign and malignant adrenal masses with a high sen- sitivity for the diagnosis of malignant adrenal masses is essen-
tial to reduce the number of expensive and/or invasive procedures in patients with benign adrenal masses. The pre- sent study shows that contrast-enhanced ultrasound using TIC analysis can be performed with excellent sensitivity (100%) for the diagnosis of malignant adrenal masses and may therefore be a useful method in the diagnostic work-up of adrenal mas- ses. This study was based on a proof of principle study, which revealed 4 CEUS patterns enabling the differentiation of be- nign and malignant adrenal masses with a sensitivity of 100% and a specificity of 82% [9]. The validation study in 74 patients with 82 adrenal masses, as well as the overall analysis in 108 patients with 116 adrenal masses confirmed the sensitivity of 100%. Two previous studies [10, 15] looked at contrast en- hancement patterns observed by the ultrasound examiners. They reported an improvement in the visualization of vascu- larization of adrenal masses without adequate differentiation between benign and malignant adrenal masses. Moreover, in
| diagnosis | no. of adrenal masses with CEUS pattern I | max. size (mm) | no. of adrenal masses with CEUS pattern II | max. size (mm) | no. of adrenal masses with CEUS pattern III | max. size (mm) | no. of adrenal masses with CEUS pattern IV | max. size (mm) |
|---|---|---|---|---|---|---|---|---|
| adenoma | 8 | 13-61 | 9 | 12-35 | 29 | 10- 104 | 33 | 12- 43 |
| (angio)myelolipoma | 4 | 19- 78 | 2 | 32- 80 | 1 | 55 | 0 | |
| hemangioma | 2 | 49 - 114 | 1 | 32 | 0 | 0 | ||
| benign follicular spreading tumor | 0 | 0 | 0 | 1 | 23 | |||
| ganglioneuroma | 1 | 90 | 0 | 0 | 0 | |||
| pheochromocytoma, benign | 5 | 27 - 100 | 3 | 14-55 | 2 | 17-41 | 1 | 34 |
| pheochromocytoma, malignant | 1 | 62 | 0 | 0 | 0 | |||
| adreno-cortical carcinoma | 6 | 62 - 190 | 0 | 0 | 0 | |||
| metastasis | 4 | 19- 78 | 2 | 37-66 | 0 | 0 | ||
| B-cell lymphoma | 1 | 60 | 0 | 0 | 0 | |||
| T-cell lymphoma | 1 | 93 | 0 | 0 | 0 | |||
| neuroendocrine carcinoma | 1 | 62 | 0 | 0 | 0 |
1 No. = number; CEUS = contrast-enhanced ultrasound.
the present study no specific perfusion patterns like central, peripheral or chaotic contrast enhancement could be found to differentiate between benign and malignant adrenal masses. However, only one published abstract [16] besides our pilot study and the present study evaluated time-intensity curve analysis in the evaluation of adrenal masses. TIC analysis al- lows the quantification of contrast enhancement over time and detects changes in perfusion in a more detailed manner than the qualitative analysis of an examiner. In the published abstract [16] benign and malignant adrenal masses could be differentiated using TIC analysis, therefore confirming the re- sults of the present study.
Hormonal evaluation was performed according to current re- commendations [1] following history and physical examina- tion. In the present study 23 adrenal masses (20%) were func- tional. According to the literature, approx. 85% of adrenal incidentalomas are non-functional, and 15% are functional [17].
Following the current recommendations [1] all patients with negative hormonal testing either presented with a high-quality CT or received an MRI scan at our clinic to differentiate be- tween benign and suspicious lesions. Contrast-enhanced CT and gadolinium-enhanced MRI have shown excellent sensitiv- ity, specificity and PPV for the diagnosis of benign adrenal masses and are currently accepted as a reference method for non-functional adrenal masses less than 4cm in size [1, 11, 18-23]. Therefore, CT or MRI was used as the reference stan- dard for the diagnosis of adenomas. Imaging follow-up (after 6, 12, and 24 months) and laboratory testing (annually for 4 years) are suggested in current recommendations [1, 18].
However, CT and MRI are expensive examinations compared with ultrasound. CT is associated with significant radiation ex- posure and both CT and MRI are commonly performed in ad- dition to screening ultrasound. Furthermore, adenomas are much more common than non-adenomas with a prevalence of up to 6% [1] and many CT and MRI examinations are thus performed simply to rule out malignancy. Only approx. 7% of adrenal masses are malignant [1]. Therefore, a pre-selection of patients with suspicious adrenal masses during screening ul- trasound would be favorable and could be performed with
CEUS at this point. Since CEUS demonstrates excellent sensitiv- ity of 100%, the number of patients requiring further diagnos- tic imaging by CT/MRI could be significantly reduced to ap- prox. 42% (49/116) without missing patients with malignant lesions. A cost reduction is expected by using this approach and cost analyses should be performed in future studies. When excluding patients with pheochromocytomas (which are diagnosed on the basis of clinical signs and/or laboratory testing and treated surgically and for which benignity or ma- lignancy can often not be determined even histologically) and when taking into account the classical B-mode ultrasound diagnostic criteria of (angio)myelolipoma, the specificity of ultrasound with CEUS increased from 67% to 79% without a reduction of the sensitivity. Hormonal evaluation cannot dif- ferentiate between benign and malignant adrenal masses.
In accordance with present recommendations [1], all patients with clinical or laboratory signs of pheochromocytoma or sus- picious appearance on CT/MRI imaging received histological verification of diagnosis.
As reported in previous studies [5] as well as in the present study, the average size of benign adrenal masses was smaller than that of malignant adrenal masses. In previous studies a size greater than 4cm was shown to have a 71% sensitivity and 79% specificity for the detection of malignancy in patients with adrenal masses [14]. In the present study 33 adrenal masses were larger than 4cm in size and had a sensitivity of 81% and specificity of 79%. Therefore, in concordance with previously published data [14, 17, 24], no sufficient conclusion concerning benignity or malignancy could be drawn solely from the size of the adrenal mass. In addition, with improve- ment in technology and experience, the aim should be to de- tect malignant adrenal masses at an earlier stage during ultra- sound examination before they reach a size of 4cm.
A recently published algorithm for the evaluation of patients with an adrenal incidentaloma starts with history and physical examination and hormonal testing [1]. The next step, if these findings are negative, is the imaging phenotype. At this point - taking into account the results of the present study - B- mode and contrast-enhanced ultrasound could be used as the first and most cost-effective imaging method to differentiate
Downloaded by: National University of Singapore. Copyrighted material.
between benign and suspicious appearance. Since the sensitiv- ity for the diagnosis of a malignant lesion in the present study was 100%, no malignant lesions were missed. Since adenomas are common and malignancies of adrenal masses are rare, con- trast-enhanced ultrasound could be a useful tool to prevent most patients with adenomas from extensive diagnostic work-up and to select the patients that need further diagnosis. Since the specificity of MRI and CT is higher than that of con- trast-enhanced ultrasound, the next step for all adrenal masses with CEUS pattern I or II vascularization should be a second imaging study (CT/MRI). Hereby, the number of CT/MRI exam- inations required will be significantly reduced due to the pre- selection using contrast-enhanced ultrasound. If CT/MRI shows a suspicious lesion, biopsy or operation is advised.
In patients with primary malignancy and suspected adrenal metastasis, further differentiation of the adrenal tumor is ad- vised if the diagnosis would change the therapeutic concept of these patients. The present study is too small to draw con- clusions concerning the value of CEUS in the work-up of pa- tients with primary malignancy and suspected adrenal metas- tasis. Larger studies are awaited.
In clinical practice CEUS of the adrenal mass should be per- formed by injecting 1.2-4.3 ml Sonovue with continuous ul- trasound scanning for approx. 3 min without relevant motion to allow TIC analysis.
The present study has several limitations.
In the present study, ultrasound at our center was focused on the visualization of the adrenal gland and detection of adrenal masses. This of course does not represent the use of ultrasound in clinical practice which is often limited with respect to time. While the right adrenal gland can be quickly visualized, the visualization of the left adrenal gland can take quite some time. In addition, many patients were referred to our center in which the primary diagnosis of an adrenal mass was detected by CT or MRI. Therefore, in the present study no conclusion can be drawn regarding the detection rate of adrenal mass using ultrasound. Nevertheless, the focus of the study was the characterization and differentiation of adrenal masses.
For the same reason (primary diagnosis of adrenal masses on CT/MRI), the CEUS examiners were only blinded to the final reference method (CT/MRI or histology) in 62% of cases: in 100% of patients with histology as the reference method and in 41% of patients with CT/MRI as the reference method. Nevertheless, TIC analysis allows objectification of the subjec- tive observation of perfusion by the CEUS examiner and the evaluation of interobserver variability was not the aim of the present study.
Due to the continuous ultrasound examination which was ne- cessary to enable TIC analysis, contrast dynamics in the late phase (after 3-5min) could not be judged sufficiently in many cases due to microbubble disruption despite the low MI. Nevertheless, the contrast enhancement patterns used for the interpretation of adrenal masses mainly focused on the time of first contrast enhancement which was not influenced negatively by the continuous examination.
While the contrast enhancement pattern of hypervascular le- sions using CEUS was comparable with vascularization pat- terns using CT or MRI, adrenal adenoma showed no or mini- mal contrast enhancement using CEUS but specific contrast enhancement using CT or MRI. The different contrast enhance- ment patterns of CEUS and CT/MRI in adrenal adenomas might be due to the amount of detectable contrast agent due
to the anatomical site of the adrenal gland in CEUS. This might also explain why TIC analysis shows better results in differen- tiating between benign and malignant adrenal masses as com- pared to the visualization of the examiner only. Nevertheless, as demonstrated in . Fig. 2, 3, the perfusion in the adjacent liver or spleen at the same anatomical depth as the adrenal mass can be used as a quality criterion to exclude a microbub- ble detection problem. Future improvement of ultrasound technology might improve the visualization of perfusion in small adrenal adenomas as well as in dorsal areas of the liver and spleen. Therefore, the quantitative difference of perfusion between the adrenal mass and the adjacent liver or spleen will not change. In addition, the aim of the present study was to evaluate patterns of contrast enhancement in different adrenal masses and compare them with each other and these limita- tions counted for all adrenal masses equally.
Severe adipositas with steatosis of the liver is a limitation con- cerning not only CEUS of the adrenal gland which is located dorsal to the liver, but also a well known limitation of CEUS of dorsal liver lesions. In those patients CT/MRI is essential. 6 patients with primary malignancy and adrenal mass metas- tases were included in the present study. In patients with a primary malignoma and an adrenal mass, the diagnosis of adrenal mass malignoma (metastasis) is higher than in pa- tients without a primary malignancy and with an adrenal inci- dentaloma. Nevertheless, patients with a primary malignancy and patients with an incidentaloma were both included in the present study representing the population of patients with an adrenal mass.
The number of malignant lesions was low in the present study, especially malignant lesions <3 cm were rare. Further larger studies including more malignant lesions are necessary in the future to answer open questions.
In the present study, CEUS showed excellent sensitivity (100%) but only low specificity (67%) for the diagnosis of malignant adrenal masses. However, a high sensitivity is more important to prevent oversight of a malignant lesion and the aim of the method is to reduce invasive and/or expensive examination of patients with clear signs of adenoma. A possible algorithm would include a second imaging method with high specificity following CEUS as the first method with high sensitivity. A specificity of approx. 96% is reported for CT and MRI examina- tion [1, 22]. Future studies should evaluate the combination of CEUS with CT/MRI as well as an algorithm with CEUS as the first method of choice and CT/MRI in all cases in which CEUS did not show clear signs of a benign adrenal mass. In addition, CEUS as a method with a high sensitivity could be combined with other ultrasound signs with high specificity. Future stu- dies should focus on such a combination of ultrasound signs to further optimize this diagnostic tool in the evaluation of adrenal masses.
In conclusion, contrast-enhanced ultrasound can be a useful method in the diagnostic work-up of adrenal masses with ex- cellent sensitivity for the diagnosis of malignancy.
Affiliations
1 Department of Internal Medicine 1, J. W. Goethe-University Hospital
2 Department of Radiology, J. W. Goethe-University Hospital
3 Department of General and Visceral Surgery, J. W. Goethe-University Hospi- tal
4 Institute of Pathology, J. W. Goethe-University Hospital
5 Institute of Biostatistics and Mathematical Modelling, Faculty of Medicine, J. W. Goethe-University
References
1 YoungJr WF. Clinical practice. The incidentally discovered adrenal mass. N Engl J Med 2007; 356: 601-610
2 Kloos RT, Gross MD, Francis IR et al. Incidentally discovered adrenal masses. Endocr Rev 1995; 16: 460-484
3 Dietrich CF, Wehrmann T, Hoffmann C et al. Detection of the adrenal glands by endoscopic or transabdominal ultrasound. Endoscopy 1997; 29: 859-864
4 Suzuki Y, Sasagawa, Suzuki H et al. The role of ultrasonography in the detection of adrenal masses: comparison with computed tomography and magnetic resonance imaging. Int Urol Nephrol 2001; 32: 303-306
5 Lumachi F, Borsato S, Brandes AA et al. Fine-needle aspiration cytology of adrenal masses in noncancer patients: clinicoradiologic and histolo- gic correlations in functioning and nonfunctioning tumors. Cancer 2001; 93: 323-329
6 Liao JT, Huang TH, Wu BY. Ultrasonic evaluation of adrenal masses. Hu- nan Yi Ke Da Xue Xue Bao 2001; 26: 453-454
7 Trojan J, Schwarz W, Sarrazin C et al. Role of ultrasonography in the de- tection of small adrenal masses. Ultraschall in Med 2002; 23: 96-100
8 Slonina J, Nienartowicz E, Agrawal AK et al. The usefulness of contrast- enhanced sonography in the differential diagnostic of adrenal tumors. Endokrynol Pol 2006; 57: 230-236
9 Friedrich-Rust M, Schneider G, Bohle RM et al. Contrast-enhanced sono- graphy of adrenal masses: differentiation of adenomas and nonadeno- matous lesions. Am J Roentgenol 2008; 191: 1852-1860
10 Dietrich CF, Ignee A, Barreiros AP et al. Contrast-enhanced ultrasound for imaging of adrenal masses. Ultraschall in Med 2010; 31: 163-168
11 Pena CS, Boland GW, Hahn PF et al. Characterization of indeterminate (lipid-poor) adrenal masses: use of washout characteristics at con- trast-enhanced CT. Radiology 2000; 217: 798-802
12 Korobkin M, Lombardi TJ, Aisen AM et al. Characterization of adrenal masses with chemical shift and gadolinium-enhanced MR imaging. Radiology 1995; 197: 411-408
13 Bovio S, Cataldi A, Reimondo G et al. Prevalence of adrenal incidentalo- ma in a contemporary computerized tomography series. J Endocrinol Invest 2006; 29: 298-302
14 Young Jr WF. Management approaches to adrenal incidentalomas. A view from Rochester, Minnesota. Endocrinol Metab Clin North Am 2000; 29: 159-185
15 Bauditz J, Quinkler M, Dammenhayn J et al. Kontrastsonografie bei Ne- bennierenraumforderungen-CEUS in adrenal masses. Ultraschall in Med 2008; 29: S148A
16 Schacherer D, Bollheimer C, Schölmerich J et al. Kontrastmittelsonogra- fie in der Diagnostik von Nebennierenraumforderungen. Ultraschall in Med 2008; 29: V27A
17 Mantero F, Terzolo M, Arnaldi G et al. A survey on adrenal incidentalo- ma in Italy. Study Group on Adrenal Tumors of the Italian Society of En- docrinology. J Clin Endocrinol Metab 2000; 85: 637-644
18 Herrera MF, Grant CS, van Heerden JA et al. Incidentally discovered adrenal tumors: an institutional perspective. Surgery 1991; 110: 1014-1021
19 Fassnacht M, Kenn W, Allolio B. Adrenal tumors: how to establish ma- lignancy ? J Endocrinol Invest 2004; 27: 387-399
20 Grumbach MM, Biller BM, Braunstein GD et al. Management of the clini- cally inapparent adrenal mass (“incidentaloma”). Ann Intern Med 2003; 138: 424-429
21 Hamrahian AH, Ioachimescu AG, Remer EM et al. Clinical utility of non- contrast computed tomography attenuation value (hounsfield units) to differentiate adrenal adenomas/hyperplasias from nonadenomas: Cleveland Clinic experience. J Clin Endocrinol Metab 2005; 90: 871- 877
22 Korobkin M, Brodeur FJ, Francis IR et al. CT time-attenuation washout curves of adrenal adenomas and nonadenomas. Am J Roentgenol 1998; 170: 747-752
23 Szolar DH, Korobkin M, Reittner P et al. Adrenocortical carcinomas and adrenal pheochromocytomas: mass and enhancement loss evaluation at delayed contrast-enhanced CT. Radiology 2005; 234: 479-485
24 Angeli A, Osella G, Ali A et al. Adrenal incidentaloma: an overview of clinical and epidemiological data from the National Italian Study Group. Horm Res 1997; 47: 279-283
Downloaded by: National University of Singapore. Copyrighted material.