| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Editorial |
1 From the Department of Radiology, University of Michigan Medical Center, UH-B1D520, Box 0030, 1500 E Medical Center Dr, Ann Arbor, MI 48109. Received August 29, 2000; accepted August 31. Address correspondence to the author (e-mail: korobkin@umich.edu).
Index terms: Adrenal gland, CT, 86.12111, 86.12113, 86.12114 • Adrenal gland, neoplasms, 86.317 • Computed tomography (CT), contrast enhancement, 86.12113, 86.12114 • Editorials • Fat
Whenever an adrenal mass is detected at abdominal computed tomography (CT), three questions should be asked. Does the mass have CT features that permit a specific CT diagnosis, such as hemorrhage, myelolipoma, and some cysts? Does the patient have clinical or biochemical evidence of a hypersecretory adrenal disorder, such as pheochromocytoma, Cushing syndrome, or primary hyperaldosteronism? Does the patient have a known extraadrenal primary neoplasm; in other words, should the possibility of adrenal metastasis be strongly considered? In the absence of evidence of a hypersecretory adrenal endocrine syndrome, an unsuspected adrenal mass is often referred to as an adrenal "incidentaloma."
Since the early 1980s, incidentally discovered adrenal masses have become a widespread clinical problem largely due to the widespread use of abdominal CT. Up to 5% of abdominal CT scans obtained for reasons other than suspected hypersecretory adrenal disease will demonstrate an adrenal mass (1). Detection of such incidental masses raises problems because it is usually impossible to differentiate a benign from a malignant adrenal mass on the basis of standard morphologic features.
In patients without a known extraadrenal primary neoplasm, an adrenal mass depicted at CT is usually a benign adrenal cortical adenoma. Because of the low prevalence of adrenal cortical carcinoma, serial imaging rather than surgical removal has often been recommended to ensure stability in size, especially if the mass is smaller than 5–6 cm. Even with 6-cm adrenal masses, the diagnosis is much more likely to be adrenal adenoma than adrenal cortical carcinoma (1). Most serial imaging protocols, however, typically involve the acquisition of two to five CT scans, which subjects many patients with benign adenomas to substantial financial expense, as well as cumulative radiation exposure (1).
In patients with a known extraadrenal primary malignancy, detection of an adrenal mass often requires diagnosis or exclusion of an adrenal metastasis to determine appropriate therapy for the primary tumor. In the past, this requirement has led to a large number of percutaneous adrenal biopsies despite the small but not negligible rate of both minor and major complications (2).
To decrease the number of percutaneous adrenal biopsies performed in patients with cancer at the time of initial tumor staging and to decrease the time and expense of serial imaging studies of an adrenal mass in patients without an extraadrenal malignancy, extensive research has been conducted to determine if noninvasive imaging could be used to characterize adrenal masses. Most of these studies have centered on CT densitometry and chemical shift magnetic resonance (MR) imaging.
During the past 10 years, evidence has shown that CT densitometry can be used to accurately differentiate benign from malignant adrenal masses. To our knowledge, Lee et al (3) reported the first large series in which the clinical use of nonenhanced CT attenuation values for the characterization of adrenal masses was assessed; they showed that most adenomas had attenuation values lower than those of malignant masses. They also showed how the scatterplot data from such a study could be used to assign threshold values for the calculation of sensitivity and specificity for the diagnosis of adenoma. Korobkin et al (4) confirmed these findings and also showed that, unlike nonenhanced attenuation values, contrast material–enhanced CT values showed too much overlap between the two groups to allow an accurate differentiation between adrenal adenomas and nonadenomas. Boland et al (5) pooled the data from 10 previous articles and showed that a sensitivity of 71% and a specificity of 98% result from choosing a threshold value of 10 HU for the diagnosis of adrenal adenoma. Ninety-eight percent of homogeneous adrenal masses with a nonenhanced CT attenuation value of 10 HU or less will be benign (most will be adenomas), whereas 29% of adenomas will have an attenuation value of more than 10 HU and will be indistinguishable from most nonadenomas, including metastases.
At the same time, Mitchell et al (6) and others used chemical shift MR imaging to show that a relative loss in signal intensity in an adrenal mass, when opposed-phase and in-phase images are compared, could also be used to characterize many adrenal masses as benign adenomas. The sensitivity and specificity for the diagnosis of adrenal adenoma were remarkably similar for chemical shift MR imaging and nonenhanced CT densitometry. Because adrenal adenomas, unlike most metastases and other nonadenomas, often contain large amounts of histologic lipid, many authors speculated that both nonenhanced CT and chemical shift MR imaging can be used to characterize adrenal masses as adenomas by detecting lipid in them.
Evidence suggesting that both techniques are used to detect the same property of adrenal adenomas was reported in a article by Outwater et al (7), who studied 47 adrenal masses at both nonenhanced CT and chemical shift MR imaging. The results of both techniques were highly correlated, and most of the same adenomas that were indeterminate with one technique were also indeterminate with the other. Further evidence that both techniques can be used to detect both the presence and amount of lipid in adrenal adenomas came from a histologic study (8) of resected adenomas that had undergone presurgical CT and/or MR imaging; the findings showed a linear correlation between the percentage of lipid-rich cortical cells and both the nonenhanced CT attenuation value and the relative change in signal intensity on opposed-phase chemical shift MR images.
There are two limitations to the use of nonenhanced CT attenuation values to characterize adrenal masses. First, most incidental adrenal masses are initially detected at enhanced CT, where attenuation values show too much overlap to allow differentiation of adenomas from nonadenomas. Characterization with nonenhanced CT densitometry requires an additional examination on another day. Second, a substantial minority of adrenal adenomas are lipid-poor and cannot be characterized by means of their nonenhanced CT attenuation. The possibility of characterizing adrenal adenomas without acquisition of a nonenhanced scan was first proposed in a preliminary study (9) with findings that showed that the CT attenuation values of adenomas were substantially lower than those of metastases on scans obtained 1 hour after contrast enhancement. Subsequent study findings suggested that delayed enhanced CT densitometry performed as early as 30 minutes (10) or 15 minutes (11) after enhancement could also be used to differentiate adenomas from nonadenomas, with optimal threshold values of 40 HU at 60 and 30 minutes and of 30 HU at 15 minutes.
Two groups of investigators (12,13) then reported the contrast enhancement washout curves of adrenal adenomas and nonadenomas. As early as 3 minutes after contrast enhancement, the mean CT attenuation value of adenomas is already substantially lower than that of nonadenomas (3). In fact, the mean attenuation value of adenomas was significantly lower than that of nonadenomas at all delay times of up to 45 minutes. Only at the time of a standard enhanced scan, obtained at 60 seconds after the injection of contrast material, is the attenuation value not significantly different in the two groups of adrenal masses. Because the optimal threshold attenuation values may depend on the type (ionic vs nonionic), total dose, and injection rate of intravenously administered contrast material, both groups of investigators also reported the percentage washout and relative percentage washout of enhancement of the adrenal masses at the different times studied. They described threshold enhancement washout values for both calculations with the corresponding sensitivities and specificities for the diagnosis of adrenal adenoma.
The concept of enhancement washout is easy to understand, and washout is simple to calculate. Enhancement can be defined as the difference between the standard enhanced attenuation value (about 60 seconds after the start of bolus intravenous injection of contrast material) and the nonenhanced value of the mass, as follows: enhancement equals enhanced attenuation minus nonenhanced attenuation.
In other words, enhancement is a quantitative measurement of the wash-in of contrast material. Enhancement washout can be defined as the difference between the enhanced attenuation value and the delayed enhanced attenuation value of the mass, as follows: enhancement washout equals enhanced attenuation minus delayed enhanced attenuation.
It follows that the percentage enhancement washout represents the percentage of the initial wash-in of enhancement that is washed out at the time of delayed scanning, as follows: percentage enhancement washout equals (enhancement washout divided by enhancement) multiplied by 100.
The relative enhancement washout is an approximation of the true enhancement washout; it relates the enhancement washout to the enhanced attenuation value instead of the enhancement wash-in, as follows: relative percentage enhancement washout equals (enhancement washout divided by enhanced attenuation) multiplied by 100.
In the current issue of Radiology, Peña et al (14) confirm the accuracy of the relative percentage enhancement washout in the characterization of adrenal masses. The equation they used to calculate the washout is easily derived from that shown previously and yields identical results. They report that 99 of 101 adrenal masses were correctly characterized as benign or malignant by using a threshold of 50% washout on a 10-minute–delayed enhanced scan. Benign lesions showed more than 50% washout, and metastases (plus one pheochromocytoma) demonstrated less than 50% washout. If the two benign masses with relative washout values of less than 50% are excluded because their dynamic enhanced attenuation values of 12 and 11 HU alone suggest that they are benign, the sensitivity and specificity are both 100%.
In our previous article (12), we also described a sensitivity and specificity of 100% at 10 minutes, but at a threshold of 40% washout rather than 50% washout. Any report of perfect accuracy must be viewed with great caution, of course, and, in the article by Peña et al (14), careful inspection of Figure 4 shows that many of the benign and malignant lesions appear to be clustered near the 50% mark. Because there is nothing magical about the 50% level, it is likely that the sensitivity and/or specificity of this calculation will decrease somewhat as more cases are accumulated.
Although there are small differences between the techniques and results in previous studies and those of Peña et al (14), the findings are similar enough to indicate that the more rapid washout of CT enhancement in adrenal adenomas compared with nonadenomas is a powerful and accurate method for use in the characterization of adrenal masses. The anatomic or physiologic mechanisms underlying this difference in response to contrast enhancement have not yet been elucidated, but some authorities (10) have speculated that nonadenomas have a disturbed capillary permeability, with prolonged retention of contrast material in the effective extracellular space.
It is useful to refer to adenomas with nonenhanced CT attenuation values of 10 HU or less as lipid-rich and to refer to those with values of more than 10 HU as lipid-poor. The lipid-poor adenomas are an important subgroup because it is precisely these adenomas that cannot be characterized at nonenhanced CT densitometry. We recently studied 18 such lipid-poor adrenal adenomas on nonenhanced, enhanced, and 15-minute–delayed enhanced CT scans and compared them with 56 lipid-rich adenomas and 40 nonadenomas (15). We found that the lipid-poor adenomas showed enhancement washout features that were statistically identical to those of lipid-rich adenomas. Although lipid-poor adenomas cannot be differentiated from nonadenomas by means of their CT attenuation values, differentiation is possible by using their larger percentage enhancement washout and relative enhancement washout.
We also found that, despite a significantly larger relative percentage enhancement washout than nonadenomas, the lipid-poor adenomas did have a relative percentage washout that was significantly smaller than that of lipid-rich adenomas (15). Peña et al (14) also report that six benign adrenal lesions with nonenhanced attenuation values of more than 10 HU had a relative enhancement washout value that was somewhat smaller than that of lesions with nonenhanced values of less than 10 HU, although statistical analysis of their data was not described. This finding is not surprising because the relative washout is only an approximation to the true washout; it is calculated by deleting the nonenhanced attenuation value from the formula for percentage enhancement washout. The two calculations of enhancement washout are equal only when the nonenhanced attenuation is 0 HU, which is close to the mean attenuation value of lipid-rich adenomas. By not including the higher mean nonenhanced attenuation value of lipid-poor adenomas in the denominator of the equation, the relative enhancement washout becomes lower than the true enhancement washout. The higher the nonenhanced attenuation value of lipid-poor adenoma, the lower the relative percentage enhancement washout, and the closer it will be to the mean value for nonadenomas. In theory at least, use of the true enhancement washout should be more accurate than the relative enhancement washout for the differentiation of lipid-poor adenomas from nonadenomas.
The relative enhancement washout of an adrenal mass was a concept introduced as an approximation to the true enhancement washout that could to be used when a delayed enhanced scan is obtained after an adrenal mass is depicted on a standard enhanced CT scan, without knowledge of the nonenhanced attenuation. It is not a physiologic construct, however, because it relates the amount of attenuation loss to only the enhanced attenuation value, not to the gain in attenuation value after the administration of contrast material.
With the pressure to quickly examine patients at CT, few of our nonemergency CT scans are now interpreted before the patient leaves the scanning table, so most of the adrenal masses are detected only when enhanced scans are reviewed and interpreted at a later time. As a result, characterization of adrenal masses is now typically performed on another day, with a nonenhanced scan obtained first. Whenever enhancement washout is necessary, both true and relative enhancement washouts can be calculated. For the occasional adrenal mass that is detected at enhanced CT before the patient leaves the scanning table, we obtain a 15-minute–delayed scan and use the relative enhancement washout calculation to characterize the lesion.
The ability to characterize lipid-poor adrenal masses by using early delayed enhanced CT attenuation washout makes CT potentially more useful than MR imaging. Although chemical shift MR imaging can be used to characterize lipid-rich adenomas with accuracy similar to that of nonenhanced CT, adenomas with only small amounts of lipid will not be depicted. Although 10 years ago Krestin et al (16) reported that gadolinium enhancement washout curves at MR imaging could be used to differentiate adenomas from nonadenomas, findings of subsequent studies (17) have failed to confirm those observations, and to my knowledge, that technique is not currently used to characterize adrenal masses.
Rapid enhancement washout of adrenal masses may be unique to cortical adenomas. Although most research on CT characterization of adrenal masses has focused on the adenoma versus metastasis problem, results of one recent study (13) showed that five of six benign adrenal pheochromocytomas had enhancement washout curves similar to those of metastases rather than adenomas. The ability to use CT enhancement washout calculations to differentiate most adrenal adenomas from metastases (and probably pheochromocytomas) now seems established, but little is known about the enhancement washout of adrenal cortical carcinoma. Among the articles about delayed enhanced attenuation values and enhancement washout curves, only one (10) included three cases of adrenal carcinoma. In that article, the carcinomas were included in the data set for nonadenomas, but their location on the scatterplots of the individual cases was not provided.
Most adrenal cortical carcinomas are larger than 6 cm at presentation and often have demonstrable metastases. Typically, these tumors also have large amounts of necrosis, which would invalidate attempts to assess enhancement washout. Despite this more common presentation of adrenal cortical carcinoma, the possibility of differentiating the rare small adrenal cortical carcinoma from the more common adrenal adenoma between 3 and 6 cm in diameter by using enhancement washout calculations to our knowledge has not yet been systematically assessed.
Peña et al (14) correctly emphasized the importance of accurate placement of the region-of-interest electronic cursor in the central one-half to two-thirds of the mass, as well as the necessity of excluding any small regions of calcification or likely necrosis. Adrenal masses that contain substantial portions of inhomogeneously low attenuation, which indicate substantial components of necrosis or cystic change, cannot be characterized by means of delayed enhancement washout calculations. The abnormal or absent capillary beds in these excavated regions will probably show slow enhancement washout regardless of the original underlying histologic cause.
Rapid early CT enhancement washout is a highly sensitive and specific feature of adrenal adenomas; with nonenhanced CT densitometry in the depiction of lipid-rich adenomas, it makes CT the most useful and accurate imaging method in the characterization of adrenal masses. In our department, we first evaluate known adrenal masses by using nonenhanced CT. If the attenuation of the mass is 10 HU or less, we make a diagnosis of lipid-rich adrenal adenoma (a small fraction of these will be cysts rather than adenomas), and no further evaluation is advised. If the attenuation is more than 10 HU, we consider the mass to be indeterminate and obtain an enhanced and 15-minute–delayed enhanced CT scan. If the enhancement washout is more than about 50%, especially if the delayed attenuation value is less than 35 HU, we make a diagnosis of lipid-poor adenoma, and again, we do not advise further evaluation. If the enhancement washout is less than about 50%, especially if the delayed attenuation value is more than 35 HU, we consider the mass to remain indeterminate. If the patient has a new extraadrenal primary neoplasm with no other evidence of metastases, percutaneous adrenal biopsy is recommended to confirm adrenal metastasis. In a patient without cancer, surgery, follow-up CT, or adrenal scintigraphy with the use of radioiodinated norcholesterol (NP-59) is recommended, depending on the size of the mass and the other specific clinical features.
FOOTNOTES
See also the article by Peña et al (pp 798–802 ) in this issue.
REFERENCES
Related Article
This article has been cited by other articles:
|
|
 |
|
  G. W. L. Boland, M. A. Blake, P. F. Hahn, and W. W. Mayo-Smith Incidental Adrenal Lesions: Principles, Techniques, and Algorithms for Imaging Characterization Radiology, December 1, 2008; 249(3): 756 - 775. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. A. Birnbaum, N. Hindman, J. Lee, and J. S. Babb Multi-Detector Row CT Attenuation Measurements: Assessment of Intra- and Interscanner Variability with an Anthropomorphic Body CT Phantom Radiology, January 1, 2007; 242(1): 109 - 119. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. S. Jhaveri, F. Wong, S. Ghai, and M. A. Haider Comparison of CT histogram analysis and chemical shift MRI in the characterization of indeterminate adrenal nodules. Am. J. Roentgenol., November 1, 2006; 187(5): 1303 - 1308. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. W. Colby, K. P. Banks, and E. Torres AJR teaching file: Incidental adrenal mass and hypertension. Am. J. Roentgenol., September 1, 2006; 187(3 Suppl): S470 - S472. [Full Text] [PDF]
|
|
|
|
 |
|
 F. E. Nwariaku, B. S. Miller, R. Auchus, S. Holt, L. Watumull, B. Dolmatch, S. Nesbitt, W. Vongpatanasin, R. Victor, F. Wians, et al. Primary Hyperaldosteronism: Effect of Adrenal Vein Sampling on Surgical Outcome Arch Surg, May 1, 2006; 141(5): 497 - 503. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. H. Szolar, M. Korobkin, P. Reittner, A. Berghold, T. Bauernhofer, H. Trummer, H. Schoellnast, K. W. Preidler, and H. Samonigg Adrenocortical Carcinomas and Adrenal Pheochromocytomas: Mass and Enhancement Loss Evaluation at Delayed Contrast-enhanced CT Radiology, February 1, 2005; 234(2): 479 - 485. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Minn, A. Salonen, J. Friberg, A. Roivainen, T. Viljanen, J. Langsjo, J. Salmi, M. Valimaki, K. Nagren, and P. Nuutila Imaging of Adrenal Incidentalomas with PET Using 11C-Metomidate and 18F-FDG J. Nucl. Med., June 1, 2004; 45(6): 972 - 979. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. A. Haider, S. Ghai, K. Jhaveri, and G. Lockwood Chemical Shift MR Imaging of Hyperattenuating (>10 HU) Adrenal Masses: Does It Still Have a Role? Radiology, June 1, 2004; 231(3): 711 - 716. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Mansmann, J. Lau, E. Balk, M. Rothberg, Y. Miyachi, and S. R. Bornstein The Clinically Inapparent Adrenal Mass: Update in Diagnosis and Management Endocr. Rev., April 1, 2004; 25(2): 309 - 340. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
