Adrenal adenoma and metastasis from clear cell renal cell carcinoma: can they be differentiated using standard MR techniques?

Abstract

Background

Chemical-shift magnetic resonance imaging (MRI) has been known to successfully differentiate adenomas from metastases. However, there has been concern that metastasis from extra-adrenal primary malignancies which contain high lipid content such as clear cell renal cell carcinoma (RCC) could mimic adrenal adenomas.

Purpose

To evaluate the ability of MR to differentiate adrenal adenoma from metastasis using chemical-shift imaging and MR feature analysis in patients with clear cell RCC.

Material and Methods

This study was institutional review board-approved; informed consent was waived. Eleven patients with 13 metastases and 13 patients with 15 adrenal adenomas in patients with clear cell RCC for evaluation of an adrenal mass underwent MR. Signal intensity on in- and opposed-phases, signal intensity index (SII), size, T2 SI, cystic change, necrosis, and hemorrhage were evaluated. Statistical analyses included Student t-test and Fisher exact test. If available, precontrast CT attenuation of the adrenal adenomas was measured. SII was correlated with attenuation using Pearson correlation coefficient.

Results

Mean size of adenomas was smaller than that of metastases (P < 0.002). Mean SII of adenomas (45.0% ± 24.6) was significantly greater than that of metastases (6.6% ± 4.7; P < 0.001). With a threshold of 16.5% for SII, the sensitivity, specificity, and accuracy for adenomas were 80%, 100%, and 89.2%, respectively. All six lipid-rich adenomas were diagnosed as adrenal adenoma. Three of eight (37.5%) lipid-poor adenomas were misdiagnosed as metastases. While up to 53.8% (7/13) of the metastases demonstrated cystic change, necrosis, or hemorrhage, only one (6.7%) adenoma exhibited cystic change or necrosis (P < 0.05 for all). Precontrast attenuation and SII were significantly correlated: r = –0.810 (P < 0.001).

Conclusion

In patients with clear cell RCC who underwent MR for adrenal masses, SII and MR features such as cystic change, necrosis, and hemorrhage were helpful in differentiating adenomas from metastases.

Keywords

Adrenal MRI adenoma metastases chemical shift

Introduction

When an incidental adrenal mass is detected in patients with a known extra-adrenal primary malignancy, it is crucial to differentiate between an adrenal adenoma and metastasis. Depending on whether the mass is an adenoma or metastasis, staging and the treatment strategy of the primary tumor may be significantly different.

Opposed-phase and in-phase gradient-echo (GRE) imaging has been proposed as a method that could successfully differentiate adenomas from metastases (1 –3). This has been explained to be due to the abundant intracellular lipid content in adrenal adenomas (4). Reinig et al. (1) reported that chemical-shift-induced signal change of the adrenal masses enabled differentiation between adenomas, metastasis, and pheochromocytomas. Mayo-Smith et al. (2) suggested that both quantitative and qualitative chemical-shift magnetic resonance imaging (MRI) were able to differentiate adrenal adenomas from metastases. However, there has been concern that metastasis from extra-adrenal primary malignancies, which contain high lipid content such as clear cell renal cell carcinoma (RCC) could mimic adrenal adenomas in chemical-shift MRI (5). Shinozaki et al. (5) reported a case of adrenal metastasis from clear cell RCC showing significant signal drop from in- to opposed-phased images. To the best of our knowledge, the diagnostic value of chemical-shift imaging in differentiating between adrenal adenomas and metastasis from clear cell RCCs has not been previously tested. Furthermore, the imaging features using standard MR techniques that may be helpful in differentiating adrenal adenoma and metastasis from clear cell RCC have not been well described in the literature.

Therefore, the purpose of our study was to evaluate the ability of MR to differentiate adrenal adenoma from metastasis using chemical-shift imaging and MR feature analysis in patients with clear cell RCC.

Material and Methods

Patients

Institutional Review Board approval was obtained for this retrospective study; requirement for informed consent was waived. Through a computerized search of our medical database from January 2002 to December 2012, we identified 29 patients with pathologically proven clear cell RCC who underwent MR for the evaluation of adrenal masses at our institution. Inclusion criteria were availability of MR images that included in- and opposed-phase sequences and proof of diagnosis. Proof of diagnosis required either histopathological evaluation or imaging and clinical follow-up. The criteria for imaging and clinical follow-up were as follows: (i) if the mass was stable in size for at least 6 months, it was considered an adenoma (6); (ii) with evidence of underlying widespread metastatic disease, if the mass displayed an increase in size (>2 mm/year), was newly appeared during follow-up, or showed interval regression in size following targeted chemotherapy, the mass was accepted as metastasis (7 –9); (iii) for these cases in which pathologic confirmation was absent, a consensus was reached among the departments of radiology, surgery, radiation oncology, and medical oncology. However, masses <8 mm were excluded, because obtaining accurate measurements of chemical-shift MR may be difficult in smaller masses in that they are prone to volume averaging with the signal intensity of the adrenal margin, rendering spurious reduction of the SI of the adrenal mass in the opposed phase (10). Therefore, among the 29 patients, we excluded four patients with pathology other than adenoma or metastasis. In addition, one patient was excluded because of insufficient proof of diagnosis. There were no adrenal masses excluded due to small size. Ultimately, 24 patients (18 men and 6 women) with a mean age of 57.7 years (range, 31–77 years) comprised our study population. Of the 24 patients, 11 had 13 metastases (surgically confirmed in 7 and based on follow-up imaging in 6 tumors), and the remaining 13 patients had 15 adenomas (surgically confirmed in 7 and based on follow-up imaging in 8 tumors). A total of 28 adrenal masses in 24 patients (four patients with two tumors) were evaluated.

MR protocol

The patients underwent MR examinations using 1.5-T systems (Magnetom Vision [in 19 patients] or Magnetom Sonata [in 5 patients], Siemens Medical Solutions, Erlangen, Germany). The following sequences were acquired in the axial or coronal plane (chemical-shift MRI was acquired in the coronal plane for the following reasons: (i) lower number of images for coverage of both the kidney and adrenal glands; (ii) easier comparison of the signal intensity of the adrenal and renal lesion with that of the liver and spleen) as follows: (a) coronal T1-weighted (T1W) opposed- and in-phase GRE imaging with a TR of 100 ms, and a TE of 2.4 and 5.0 ms for in- and opposed-phase imaging, respectively; flip angle, 70°; field of view [FOV], 30–35 cm; slice thickness, 4 mm; intersection gap, 0.8 mm; matrix, 256 × 179; and one signal acquisition; (b) axial T2-weighted (T2W) turbo spin echo without fat saturation with a TR of 3500 ms, and a TE of 98 ms; flip angle, 150°; FOV, 30–35 cm; slice thickness, 4 mm; intersection gap, 0.8 mm; matrix, 320 × 205; and one signal acquisition; (c) coronal T1W 3D FLASH breath hold technique with fat selective prepulse (VIBE) using a TR of 4.8 ms, and a TE of 2.3 ms; flip angle, 10°; FOV, 30–35 cm; slice thickness, 2.5 mm; matrix, 256 × 145; and one signal acquisition obtained before and after intravenous bolus administration of 0.1 mmol/kg of gadopentetate dimeglumine (Magnevist; Bayer Schering Pharma AG, Berlin, Germany) at a rate of 2 mL/s, and followed by a 20-mL saline flush by a power injector. Scanning delay times for dynamic contrast-enhanced images determined by real-time MRI fluoroscopic monitoring were 7 s after contrast media arrival at a distal thoracic aorta (immediate), and then 1, 3, and 5 min after contrast medium injection.

Quantitative image analysis for chemical-shift imaging

All images were reviewed by two radiologists (JYC and SW) blinded to the final diagnosis. First, the maximal diameter of the mass was measured as the greater of the longest diameters in the axial and coronal planes. Then, a free-drawn region of interest (ROI) was placed on solid portion of the tumor referencing T2W and contrast-enhanced VIBE images, and avoiding the peripheral areas to avoid causing partial volume effect from phase cancellation artifacts at the interface between fatty tissue (11). The area, location, and size of the ROI were constant between in- and opposed-phase images. The signal intensity index (SII), a quantitative measure of signal intensity change between in- and opposed-phases was calculated using the following equation (12,13): SII = [(SI_in adrenal - SI_opp adrenal)/SI_in adrenal] × 100. A SII ≥ 16.5% was considered a diagnostic of adenoma (12). Furthermore, in the patients in the adenoma group who also underwent CT scans including a precontrast phase, a separate ROI was placed on the tumor in the same manner as on MR in order to determine if the adenoma was lipid-rich (<10 HU) or not (14). The CT measurement was performed by another radiologist (SYK) unaware of the chemical-shift MR findings.

Qualitative image analysis

The opposed- and in-phase, T2W, T2*-weighted, and pre- and postcontrast VIBE images were analyzed by the same radiologists (JYC and SW) to assess the following characteristics of the adrenal mass (15 –18): (i) SI of the tumor on T2W images as compared with that of the liver as definitely or slightly hyperintense, isointense, or hypointense; (ii) the presence of cystic change or necrosis, defined as a region within the tumor with SI identical to that of cerebrospinal fluid on T2W images, low SI on T1W images, without enhancement, and lobulating shape for cystic change, and as a region with high SI but with a lesser degree than that of the cerebrospinal fluid on T2W images, low SI on T1W images, absence of enhancement, and with a central location within the tumor for necrosis; (iii) hemorrhage, defined as a non-enhancing region of high SI on T1W images, variable SI on T2W images, and without suppression on fat-saturated sequences; and (iv) lateralization of the adrenal mass with regard to the primary clear cell RCC.

Statistical analysis

All statistical analyses were done with PASW statistical software (version 18.0; SPSS Inc., Chicago, IL, USA). A two-tailed P value of <0.05 was considered to indicate a statistically significant difference. The Student t test was used to compare the quantitative variables between the two groups. For the qualitative variables, the Fisher exact test was used to compare the proportions between the adenomas and metastases. The Pearson correlation coefficient was used to assess the relationship between precontrast CT attenuation and SII. In addition, a subgroup analysis for SII was done in small (<3 cm) and homogeneous (without qualitative features of cystic change, necrosis, or hemorrhage) tumors.

Results

Patient characteristics

The mean patient age was 57.4 years ± 10.8 in the adenoma group and 58 years ± 12.2 in the metastasis group (P = 0.90). The ratio of men to women was 1.6:1 in the adenoma group and 10:1 in the metastasis group (P = 0.166).

Quantitative image analysis

The results of the quantitative image analysis are summarized in Table 1. The mean tumor size of metastases was significantly greater than that of adenomas. The mean size of adenomas and metastases were 19.9 mm ± 8.7 and 41.7 mm ± 18.2, respectively (P = 0.001).

Table 1.

Analyses of quantitative characteristics of adenomas and metastases.

Parameter	Adenomas (n = 15)	Metastases (n = 13)	P
Size (mm)	19.9 ± 8.7 (10.3–34.9)	41.7 ± 18.2 (16.3–72.0)	0.001
In-phase SI	233.5 ± 87.1 (137–382)	218.6 ± 57.2 (124–323)	0.593
Opposed-phase SI	124.7 ± 74.1 (55–355)	204.4 ± 52.2 (116–289)	0.004
Signal intensity index	45.0 ± 24.6 (4.7–71.5)	6.6 ± 4.7 (0.6–13.7)	< 0.001

Data are means ± standard deviations, and ranges are in parentheses.

The mean SII of adenomas was significantly greater than that of metastases (Figs. 1 and 2). The mean SII of the adenomas and metastases were 45.0% ± 24.6 and 6.6% ± 4.7, respectively (P < 0.001). Using the diagnostic criteria of SII >16.5%, the sensitivity, specificity, and accuracy in determining adenomas were 80% (12/15), 100% (13/13), and 89.3% (25/28), respectively.

Fig. 1.

Scatterplots of signal intensity index (SII) of adrenal adenomas and metastases.

Fig. 2.

Chemical-shift MR and precontrast CT scans of right lipid-rich adrenal adenoma in a 56-year-old woman with clear cell RCC. (a, b) Adrenal mass (arrow) shows overt signal intensity (SI) loss between in-phase (a) and opposed-phase (b) gradient-echo coronal MR images with a calculated SII of 71.48 (repetition time ms/echo time ms, 100/5.0 and 100/2.4 for in- and opposed -phases, respectively). (c) Coronal T2W image shows adrenal mass (arrow) demonstrating slightly higher signal intensity than the liver without hemorrhage, cystic change, or necrosis (repetition time ms/echo time ms, 3500/98). (d) Transverse precontrast CT image shows hypoattenuating (7 HU) adrenal mass (arrow).

Fourteen of the 15 adrenal adenomas had available precontrast CT scans. The mean precontrast attenuation was 14.9 HU ± 14.2 (range, –9 to 37 HU). Among them six adenomas were lipid-rich (mean, 1.5 HU ± 6.9, range, –9 to 9 HU) and eight were lipid-poor (mean, 24.9 HU ± 8.5, range, 12–37 HU). All six lipid-rich adenomas demonstrated a SII >16.5%, therefore resulting in 100% sensitivity for the diagnostic criterion. Regarding the lipid-poor adenomas, it was observed that 62.5% (5/8) had a SII >16.5%. Therefore these three adenomas were misclassified as metastases (Fig. 3). There was a significant negative correlation between precontrast CT attenuation and SII (Fig. 4): rho (r) = –0.8099 (P = 0.0004).

Fig. 3.

Chemical-shift MR and precontrast CT scans of right lipid-poor adrenal adenoma in a 31-year-old woman with clear cell RCC. (a, b) Adrenal mass (arrow) shows no overt signal intensity loss between in-phase (a) and opposed-phase (b) gradient-echo coronal MR images with a calculated SII of 5.84 (repetition time ms/echo time ms, 100/5.0 and 100/2.4 for in- and opposed-phases, respectively). (c) Transverse precontrast CT image shows hyperattenuating (37 HU) adrenal mass (arrow).

Fig. 4.

Scatter plot of signal intensity index versus precontrast CT attenuation values of adrenal adenomas. Vertical reference line indicates the 10-HU cut-off value used for determining lipid-rich (<10 HU) and lipid-poor (≥10 HU). Horizontal reference line indicates the 16.5% cut-off value used in the diagnostic criteria for determining adenomas.

Qualitative image analysis

The results of the qualitative image analysis are summarized in Table 2. All adenomas were slightly higher (40%, 6/15) or similar (60%, 9/15) in SI to the liver, whereas the metastases showed definitely higher SI (46.2%, 6/13), slightly higher SI (46.2%, 6/13), and lower SI (7.7%, 1/13) compared with the liver (P = 0.001). The metastases (46.2%, 6/13) had a significantly greater proportion of masses with a SI definitely higher than the liver compared with adenomas (0%, 0/0; P = 0.005). Up to 53.8% (7/13) of the metastases demonstrated cystic change, necrosis, or hemorrhage (Fig. 5). On the other hand, none of the adenomas demonstrated such MR features (P <0.05 for all features) except for one (6.7%) of the 15 adenomas, which demonstrated cystic change or necrosis. There was no significant difference in the lateralization of the adrenal masses with regard to the primary clear cell RCC (P = 1.000).

Fig. 5.

MRscans in a 68-year-old man with 4.7-cm sized right adrenal metastasis from clear cell RCC. (a) Coronal T2W image shows adrenal mass (arrow) with multiple areas of cystic change and necrosis (repetition time ms/echo time ms, 3500/98). (b, c) Adrenal mass (arrow) shows no overt signal intensity loss between in-phase (b) and opposed-phase (c) gradient-echo coronal MR images with a calculated SII of 1.18 (repetition time ms/echo time ms, 100/5.0 and 100/2.4 for in- and opposed-phases, respectively).

Table 2.

Analyses of qualitative characteristics of adenomas and metastases.

Parameter	Adenomas (n = 15)	Metastases (n = 13)	P
T2-weighted SI
>> Liver	0 (0.0%)	6 (46.2%)	0.001
> Liver	6 (40.0%)	6 (46.2%)	(0.005)*
= Liver	9 (60.0%)	0 (0.0%)
< Liver	0 (0.0%)	1 (7.7%)
Cysts or necrosis
Yes	1 (6.7%)	7 (53.8%)	0.011
No	14 (93.3%)	6 (46.2%)
Hemorrhage
Yes	0 (0%)	5 (38.5%)	0.013
No	15 (100%)	8 (61.5%)
Lateralization
Ipsilateral	7 (46.7%)	7 (53.8%)	1.000
Contralateral	8 (53.3%)	6 (46.2%)

Data are number of lesions, and percentages are in parentheses.

*P values calculated by comparing the ratio of definitely higher than liver with the rest.

Subgroup analysis of small (<3 cm) and homogeneous adrenal tumors

There were a total of 11 adenomas and four metastases that met the criteria of small and homogeneous adrenal tumors. Within this subgroup, the mean SII of adenomas were significantly larger than that of metastases. The mean SII of the adenomas and metastases were 41.7% ± 27.0 and 5.1% ± 6.0, respectively (P = 0.001). Using the criteria of SII <16.5%, the sensitivity, specificity, and accuracy in determining adenomas were 72.7% (8/11), 100% (4/4), and 80% (12/15), respectively.

Discussion

In this retrospective study, the mean SII values of adrenal metastases versus adenomas in the patients with clear cell RCC were 45% versus 6.6%; no metastasis had an SII >16.5%. Therefore, by using a diagnostic criterion of SII >16.5% adenomas could be differentiated from metastases with a sensitivity, specificity, and accuracy of 80%, 100%, and 89.2%, respectively. In fact, there were no cases of metastasis from clear cell RCC that exhibited abundant lipid. In addition, the sensitivity of detecting lipid-rich adenomas was 100%. Therefore, unlike the concern that metastases from clear cell carcinoma may occasionally mimic adenomas because of their abundant microscopic fat content, the SII of adrenal masses may be helpful in differentiating adenomas (especially those that are lipid-rich) from metastases. Accurate characterization of the adrenal mass in this clinical setting is critical for staging and therapeutic planning and with our study results, unnecessary invasive procedures such as percutaneous biopsy or surgery may be avoided in selected patients.

However, our study results demonstrated that chemical-shift imaging had limited value in the subgroup of lipid-poor adenomas with a sensitivity of 62.5%. Furthermore, there was a significant negative correlation between the precontrast CT attenuation and SII. This is in agreement with previous studies that have reported that the role of MR for characterizing hyperattenuating adrenal masses is limited owing to a strong correlation between decrease in SI on the opposed phase and CT attenuation (10,13,19). Taking this into consideration, it has been proposed that dedicated adrenal CT including a delayed wash-out phase may be the modality of choice in diagnosing adrenal adenomas, since the excellent diagnostic performance of the relative percentage wash-out in adrenal adenomas are maintained at all attenuation values (6,20). However, recently Choi et al. (8) observed that metastases from hypervascular tumors such as RCC may exhibit similar percentage of enhancement wash-out to that of lipid-poor adrenal adenomas. In their study, 95% of metastases from RCC were falsely diagnosed as lipid-poor adenomas when the threshold of 60% for absolute percentage wash-out or 40% for relative percentage wash-out was used. Therefore, our study results along with that of Choi et al. (8) suggest that if a lipid-poor adrenal mass is detected using chemical-shift MR in patients with underlying clear cell RCC, dedicated adrenal CT may not be helpful as the next diagnostic test. In this clinical setting either biopsy or close follow-up may be recommended.

In our study, certain MRI features were helpful in the differentiation between adenomas and metastases. While metastases mostly exhibited T2 SI that was definitely or slightly hyperintense to the liver, adenomas were mostly isointense or slightly hyperintense to the liver. This is concordant with previous reports, which observed that the SI of metastases tended to be higher than adenomas (21,22). However, they also acknowledged that the overlap between the two entities were substantial. Furthermore, the liver as a reference value for T2 SI may not be accurate, because of the large SI changes that can be seen with fatty infiltration or iron overload (2). The presence of necrosis, cystic change, and hemorrhage was also significantly predictive for metastases. Of note, in our study we did not separately evaluate cystic change and necrosis, because it sometimes may be difficult to exactly distinguish the two imaging features. A comprehensive analysis of these features and their prevalence in the adenomas or metastases has not been well defined in the literature. Yet, several investigators have reported that compared to adenomas, metastases tend to be more heterogeneous on CT and MR (7,8,23). We speculate that due to the higher prevalence of cystic change, necrosis, and hemorrhage, the metastases may demonstrate a heterogeneous appearance. However, we must note that the higher prevalence of these MR findings in metastases may have been partly attributed to their larger size compared with adenomas. It is well known that as a mass increases in size, large necrotic areas within the mass may indicate malignancy (8). Subsequently, small adrenal metastases may often exhibit homogeneous appearance (24,25). A larger study including a subgroup analysis of small adrenal masses would provide a more definitive conclusion regarding the diagnostic value of cystic change, necrosis, and hemorrhage.

Regarding the size discrepancy between the adenomas and metastases, one could raise the issue that the larger size of the metastases (and possibly the associated higher frequency of aggressive features) would skew our cases toward those that are not diagnostic dilemmas. Therefore, we performed a subgroup analysis for SII in small and homogeneous tumors. As a result it was demonstrated that the mean SII was significantly different between the two entities and the criteria of SII >16.5% for adenoma yielded a specificity of 100%.

There were some limitations in our study. First, the retrospective design may have introduced selection bias. Eight (57.1%) of the 14 adenomas with precontrast CT scans demonstrated a precontrast attenuation of >10 HU. However, in the literature, such hyperattenuating or lipid-poor adenomas are known to consist of about 29% of the adenomas (14). Additional MRI may not have been performed on the patients with adrenal adenomas with a precontrast attenuation <10 HU, since they were not considered to be indeterminate adrenal lesions. However, if such patients had undergone MRI and were included in our study population, it would have resulted in greater differences in the SII between adenomas and metastases. Second, not all masses were histologically confirmed. However, several studies have reported that adrenal adenomas can be diagnosed on the basis of stability of size over a follow-up of 6 months or more (6,24,26,27). In our study all adenomas were stable in size (n = 8, mean size change, 0.2 mm; mean follow-up, 19.9 months) whereas all metastases (except one which demonstrated a significant decrease in size [17 mm during 6 months] after chemotherapy ) revealed significant growth (n = 5; mean size increase, 14.5 mm; mean follow-up, 3.5 months). In addition, we believe that it would be acceptable to clinically diagnosis metastasis on the basis of increasing size, interval new appearance, or interval regression after chemotherapy given that that the patient has evidence of underlying widespread metastatic disease (7 –9). Third, we did not perform radiologic-pathologic correlation between chemical-shift imaging and fat content on microscopy. Further studies with evaluation of the actual fat content may be able to provide additional information. Fourth, the number of adrenal metastases (n = 13) and adenomas (n = 15) in our study was relatively small. Furthermore, only 11 adenomas and four metastases were included in the subgroup analysis for homogeneous and small tumors. Analytic results from such a small group may not be able to represent the true significance in a larger population. It is warranted that future studies with a larger study population, possibly a multicenter study, may be needed in order to have more convincing power. However, we must also note that the maximum number of metastatic adrenal masses from RCC in the studies using MR, which had specified the origin of metastases, was one (2,3,28,29), rendering our study population the largest to date. Fifth, we could not compare the diagnostic capacities of CT and MR in differentiating adenomas and metastases from clear cell RCC, owing to a lack of dedicated adrenal CT scans. However, this issue was not the focus of our study. Sixth, regarding the analysis of chemical-shift MRI, we used only the SII. Although other methods (adrenal-spleen ratio, adrenal-liver ratio, and adrenal-muscle ratio) have been suggested, the aim of our study was not to compare the diagnostic capacity of these methods. Furthermore, among them, SII has been reported to be superior to others in both clinical and phantom-based studies (12,13,30). In addition, we did not include a qualitative approach to identify signal loss in the adrenal masses. While some have reported that qualitative analysis was less sensitive than quantitative analysis for the characterization of adenoma, others have suggested that both demonstrated similar accuracy (2,31). This may be partly attributed to the level experience of the observers, and the fact that qualitative analysis may be prone to subjectivity. Therefore we concentrated on detection of quantitative SI loss using the SII as an objective method. Seventh, since we included only patients with clear cell RCC, our study results may not be extended to patients with other lipid-rich primary malignancies such as liposarcoma and hepatocellular carcinoma.

In conclusion, quantitative analysis of chemical-shift MRI using the signal intensity index was useful for differentiating between adenomas and metastases from clear cell RCCs. In addition, MRI features such as cystic change, necrosis, and hemorrhage favored the diagnosis of metastasis from clear cell RCC over that of adenoma.

Footnotes

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

References

Reinig

Stutley

Leonhardt

. Differentiation of adrenal masses with MR imaging: Comparison of techniques. Radiology 1994; 192: 41–46.

Mayo-Smith

Lee

McNicholas

. Characterization of adrenal masses (< 5 cm) by use of chemical shift MR imaging: Observer performance versus quantitative measures. Am J Roentgenol 1995; 165: 91–95.

Savci

Yazici

Sahin

. Value of chemical shift subtraction MRI in characterization of adrenal masses. Am J Roentgenol 2006; 186: 130–135.

Leroy-Willig

Bittoun

Luton

. In vivo MR spectroscopic imaging of the adrenal glands: Distinction between adenomas and carcinomas larger than 15 mm based on lipid content. Am J Roentgenol 1989; 153: 771–773.

Shinozaki

Yoshimitsu

Honda

. Metastatic adrenal tumor from clear-cell renal cell carcinoma: A pitfall of chemical shift MR imaging. Abdom Imaging 2001; 26: 439–442.

Caoili

Korobkin

Francis

. Adrenal masses: characterization with combined unenhanced and delayed enhanced CT. Radiology 2002; 222: 629–633.

Inan

Arslan

Akansel

. Dynamic contrast enhanced MRI in the differential diagnosis of adrenal adenomas and malignant adrenal masses. Eur Radiol 2008; 65: 154–162.

Choi

Kim

Park

. Evaluation of adrenal metastases from renal cell carcinoma and hepatocellular carcinoma: Use of delayed contrast-enhanceded CT. Radiology 2013; 266: 514–520.

McNicholas

Lee

Mayo-Smith

. An imaging algorithm for the differential diagnosis of adrenal adenomas and metastases. Am J Roentgenol 1995; 165: 1453–1459.

10.

Haider

Ghai

Jhaveri

. Chemical shift MR imaging of hyperattenuating (>10 HU) adrenal masses: Does it still have a role? Radiology 2004; 231: 711–716.

11.

Delfaut

Beltran

Johnson

. Fat suppression in MR imaging: Techniques and pitfalls. Radiographics 1999; 19: 373–382.

12.

Fujiyoshi

Nakajo

Fukukura

. Characterization of adrenal tumors by chemical shift fast low-angle shot MR imaging: comparison of four methods of quantitative evaluation. Am J Roentgenol 2003; 180: 1649–1657.

13.

Park

Kim

. Comparison of delayed enhanced CT and chemical shift MR for evaluating hyperattenuating incidental adrenal masses. Radiology 2007; 243: 760–765.

14.

Boland

Lee

Gazelle

. Characterization of adrenal masses using unenhanced ct: An analysis of the ct literature. Am J Roentgenol 1998; 171: 201–204.

15.

Hindman

Ngo

Genega

. Angiomyolipoma with minimal fat: can it be differentiated from clear cell renal cell carcinoma by using standard MR techniques? Radiology 2012; 265: 468–477.

16.

Heinz-Peer

Honigschnabl

Schneider

. Characterization of adrenal masses using MR imaging with histopathologic correlation. Am J Roentgenol 1999; 173: 15–22.

17.

Glazer

Woolsey

Borrello

. Adrenal tissue characterization using MR imaging. Radiology 1986; 158: 73–79.

18.

Maurea

Imbriaco

D'Angelillo

. Diagnostic accuracy of chemical-shift MR imaging to differentiate between adrenal adenomas and non adenoma adrenal lesions. Radiol Med 2006; 111: 674–686.

19.

Outwater

Siegelman

Huang

. Adrenal masses: correlation between CT attenuation value and chemical shift ratio at MR imaging with in-phase and opposed-phase sequences. Radiology 1996; 200: 749–752.

20.

Pena

Boland

Hahn

. Characterization of indeterminate (lipid-poor) adrenal masses: use of washout characteristics at contrast-enhanced CT. Radiology 2000; 217: 798–802.

21.

Mitchell

Crovello

Matteucci

. Benign adrenocortical masses: diagnosis with chemical shift MR imaging. Radiology 1992; 185: 345–351.

22.

Remer

Weinfeld

Glazer

. Hyperfunctioning and nonhyperfunctioning benign adrenal cortical lesions: characterization and comparison with MR imaging. Radiology 1989; 171: 681–685.

23.

Gufler

Eichner

Grossmann

. Differentiation of adrenal adenomas from metastases with unenhanced computed tomography. J Comput Assist Tomogr 2004; 28: 818–822.

24.

Boland

Blake

Hahn

. Incidental adrenal lesions: principles, techniques, and algorithms for imaging characterization. Radiology 2008; 249: 756–775.

25.

Boland

. Adrenal imaging: why, when, what, and how? Part 1.. Why and when to image? Am J Roentgenol 2010; 195: W377–W381.

26.

Korobkin

Brodeur

Francis

. CT time-attenuation washout curves of adrenal adenomas and nonadenomas. Am J Roentgenol 1998; 170: 747–752.

27.

Caoili

Korobkin

Francis

. Delayed enhanced CT of lipid-poor adrenal adenomas. Am J Roentgenol 2000; 175: 1411–1415.

28.

Bilbey

McLoughlin

Kurkjian

. MR imaging of adrenal masses: value of chemical-shift imaging for distinguishing adenomas from other tumors. Am J Roentgenol 1995; 164: 637–642.

29.

Halefoglu

Altun

Disli

. A prospective study on the utility of diffusion-weighted and quantitative chemical-shift magnetic resonance imaging in the distinction of adrenal adenomas and metastases. J Comput Assist Tomogr 2012; 36: 367–374.

30.

Schindera

Soher

Delong

. Effect of echo time pair selection on quantitative analysis for adrenal tumor characterization with in-phase and opposed-phase MR imaging: initial experience. Radiology 2008; 248: 140–147.

31.

Israel

Korobkin

Wang

. Comparison of unenhanced CT and chemical shift MRI in evaluating lipid-rich adrenal adenomas. Am J Roentgenol 2004; 183: 215–219.