KIDNEYS, URETERS, BLADDER, RETROPERITONEUM
Check for updates
Inter-individual comparison of diagnostic accuracy of adrenal washout CT compared to chemical shift MRI plus the T2-weighted (T2W) adrenal MRI calculator in indeterminate adrenal masses: a retrospective non-inferiority study
Rachel McPhedran1 . Rosalind Gerson1 . Hana Alfaleh1 . Paul Sathiadoss1 . Nicola Schieda1(D
Received: 24 January 2022 / Revised: 12 April 2022 / Accepted: 13 April 2022 / Published online: 10 May 2022 @ The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2022
Abstract
Objective To compare diagnostic accuracy of washout (WO)-CT to chemical shift (CS)-MRI +T2W adrenal MRI Calculator (T2W-Calculator) to diagnose adrenal adenoma in indeterminate adrenal masses.
Methods This retrospective, cross-sectional, non-inferiority study evaluated 40 consecutive indeterminate adrenal masses; each with WO-CT and MRI. Two blinded radiologists independently evaluated in mixed order: pre-contrast attenuation (Hounsfield Units, HU) and absolute WO ([Peak.HU-Delay.HU]/[Peak.HU-Pre.HU] × 100%), Chemical Shift Signal Inten- sity (CS-SI) Index, T2W SI ratio, and Entropy (which were imputed into the T2W-Calculator). Diagnostic accuracy for adrenal adenoma was tabulated using 2× 2 tables. True -positive diagnoses of adenoma were CT =Pre-HU < 10 or absolute WO≥60%, MRI=SI index≥16.5% or T2W-Calculator<0.631.
Results There were 73% (29/40) adenomas and 27% (11/40) other masses (5 pheochromocytoma, 3 solitary fibrous tumor, 1 metastasis, 1 cavernous hemangioma, and 1 adrenocortical carcinoma). Sensitivity, specificity, and accuracy for diagnosis of adenoma using CT-WO were 78% (95% confidence intervals [CI] 56-93%), 35% (14-62%), and 57% (42-71%) Reader 1 and 72% (53-87%), 46% (17-77%), and 59% (41-76%) Reader 2. Sensitivity, specificity, and accuracy for diagnosis of adenoma using MRI were 100% (88-100%), 64% (34-90%), and 82% (67-97%) Reader 1 and 86% (68-96%), 73% (39-94%), and 80% (64-95%) Reader 2. MRI had higher overall accuracy (p=0.02 Reader 1, 0.05 Reader 2) compared to CT-WO. Conclusion Chemical shift MRI combined with the T2W adrenal MRI calculator is not inferior to CT Washout for diagnosis of adrenal adenoma among indeterminate adrenal masses.
Keywords Adrenal . Adenoma . Pheochromocytoma . Washout . Computed tomography · MRI
Introduction
Incidental adrenal masses detected at cross-sectional imag- ing are common in clinical practice [1-3]. Incidental adrenal masses most commonly represent benign adrenal adenomas; however, in patients with cancer history, the risk of malig- nancy increases substantially [2, 4]. Guidelines vary [5]; however, imaging diagnosis of adrenal adenoma is typi- cally recommended, particularly in patients with underlying
malignancy, among heterogeneous and in larger nodules [5]. The imaging diagnosis of adrenal adenoma depends, in large part, on the detection of microscopic fat. Microscopic fat is characteristic of adenoma (with few exceptions) and can be diagnosed either using a threshold of < 10 Hounsfield Units (HU) at unenhanced CT or by chemical shift MRI [6]. In approximately 30% of adenomas, there is insufficient microscopic fat to be reliably diagnosed with imaging [6]. MRI may detect microscopic fat in an additional proportion of ‘lipid-poor’ adenomas particularly those which measure between 10 and 30 HU [7, 8]. Nevertheless, even with chem- ical shift MRI, some adenomas may be lipid poor.
In lipid-poor adenomas, diagnosis can be established by adrenal washout CT (WO-CT) which evaluates the attenu- ation of an adrenal mass over time during peak and 15-min
☒ Nicola Schieda
1 Department of Medical Imaging, The Ottawa Hospital, University of Ottawa, 1053 Carling Avenue, C1 Radiology, Ottawa, ON K1Y 4E9, Canada
delayed enhancement. CT-WO is purported to be highly accurate to diagnose adenoma [9] and is recommended as the first-line imaging test for characterization of indetermi- nate 1-4-cm adrenal masses by the American College of Radiology (ACR) [10]. The accuracy of CT-WO has been re-evaluated in recent years, given more liberal use of the exam in clinical practice [11]. In the incidental adrenal mass population, CT-WO may fail to diagnose adrenal adenoma due to the fact that up to one-third of lipid-poor adenomas do not washout in the adenoma range [12]. Moreover, it is known that up to one-third of pheochromocytoma may washout in the adenoma range and that there is substantial overlap in peak attenuation comparing pheochromocytoma and adenoma [13, 14]. Adrenocortical carcinoma (ACC) are very rare and their imaging parameters at CT-WO unreli- ably described in the literature; however, in a study evaluat- ing 41 ACC, nearly one-third of ACC also showed abso- lute washout in the adenoma range [15]. In cancer patients, CT-WO suffers from the same limitations, namely, that up to one-third of adenomas do not washout, thereby simulating metastases; and metastases from hypervascular tumors may washout in the adenoma range [16].
MRI may represent an alternative to CT-WO for char- acterization of adrenal masses. In addition to increased sensitivity for detection of microscopic fat compared to unenhanced CT, Tu et al. have recently described the use of a T2-weighted (T2W) adrenal MRI calculator which can be used to differentiate lipid-poor adenomas from other masses. The T2W adrenal MRI calculator is a logistic regression model which encompasses T2W signal intensity and entropy (a measure of heterogeneity). In a prelimi- nary study evaluating the T2W adrenal MRI calculator, Tu et al. report high accuracy to differentiate lipid-poor adenomas from metastases [17]. The T2W calculator has also been piloted to differentiate lipid-poor adenomas from pheochromocytoma with high accuracy [18]. The pur- pose of the present study was to compare the diagnostic accuracy of CT-WO to chemical shift MRI plus the T2W adrenal MRI calculator to diagnose adrenal adenoma in the same group of patients who underwent both exams, including both incidental adrenal masses (incidentalomas) and masses detected in patients with underlying primary malignancy.
Materials and methods
Patients
This retrospective cross-sectional study was approved by our local research ethics board who waived the need for informed consent. We searched through our institutional
570 studies
Excluded: 506 studies
409 - Washout CT or MRI not performed 44 - Duplicate patient
2 - Study images unretrievable
32 - No adrenal nodule
18 - Macroscopic fat content
1 - Excessive image artifact
65 patients (71 nodules) with both washout CT & MRI
Excluded: 31 nodules
20 - Incomplete MRI protocol
1 - Study images unretrievable
2 - Excessive image artifact
2 - No reference standard
1-MRI & CT interval >15 years
4-Cysticlesions
1 -Macroscopic fat content
37 patients (40 nodules) Known malignancy: 18 patients
Picture Archive and Communication System (PACS; Change Healthcare Radiology Solutions) for patients who underwent a dedicated adrenal washout CT protocol at our institution between December 2006 and October 2021. We identified 570 examinations. We then searched this provi- sional database for patients who also underwent abdominal MRI, identifying 70 total patients (77 nodules). A sum- mary of patient inclusion and exclusion criteria is pro- vided in Fig. 1. 32 studies had an interval between CT-WO and MRI of less than 2 years (mean 100.8 ± 105 days, range 0-469 days, calculated as absolute number of days between studies). There were 8 studies with a mean time interval between studies of greater than 2 years that were included as the nodules demonstrated size stability or a growth rate of <3 mm/year (mean 1526.1 ±754.2 days, range 878-2693).
Reference standard
The reference standard for adrenal adenoma was established by histopathology (N=1), by minimum of 1-year imaging interval stability in size or growth rate <3 mm/year [10] (N=13) or by presence of microscopic fat (unenhanced
CT attenuation < 10 HU or chemical shift SI index ≥ 16.5% [19]) in patients with a cancer history but without a primary malignancy known to demonstrate microscopic fat (N=15; N=10 for UECT < 10HU, N=5 for chemical shift MRI signal drop) determined by a radiologist with 15 years of experience in adrenal CT and MRI [NS]. In cases where unenhanced CT attenuation < 10 HU was used to establish a diagnosis of adenoma, the indeterminate adrenal nodule being investigated was depicted incidentally on a preced- ing single-phase-enhanced CT where the attenuation value was> 10 HU. The reference standard for pheochromocytoma was histopathology (N=4) or positive biochemical testing and MIBG (N=1). The reference standard for metastases was histopathology (N=1). The diagnosis for all other diag- noses were established by histopathology (N=5).
CT and MRI technique
Adrenal washout CT and MRI were performed at a single referral center and protocols are summarized in Supplemen- tary Tables 1 and 2. Although techniques varied slightly by CT and MRI system over the time period, all CT-WO were required to have a dedicated unenhanced peak (70 s) and 15-min delayed phase with matched CT parameters between phases to enable accurate attenuation comparisons. Simi- larly, MRI were typically performed with either 2-D or 3-D chemical shift MRI and with axial T2-weighted single-shot turbo or fast spin-echo imaging.
Radiologist interpretation
Two abdominal radiologists (HA and PS) with 1 and 4 years of post-fellowship experience, blinded to all clinical infor- mation and the reference standard diagnoses, were provided only with location of the adrenal mass. Each radiologist independently evaluated each mass in a mixed order to mini- mize bias. Reader 1 evaluated CT-WO for the first half of the study cohort and MRI for the second half of the study cohort at interpretation session 1 and reversed the order for interpretation 2. The examinations were evaluated in 2 ses- sions after a 1-2-week washout period. The opposite read- out strategy was employed for Reviewer 2.
Adrenal washout CT
Radiologists recorded the mean attenuation on unenhanced CT, 70-s peak contrast-enhanced (CE)-CT and 15-min delayed CE-CT images by placing a region of interest (ROI) within the lesion on matched CT images from each phase in the axial plane on the axial image where the nodule appeared the largest. For homogeneous masses, the radiolo- gists were instructed to place an ROI measurement, which encompassed at least 2/3 of the largest portion of the tumor,
Fig. 2. For heterogeneous masses, the radiologists were instructed to place an ROI in the most enhancing area on 70-s peak CE-CT images with a minimum ROI diameter of 5 mm, Fig. 3. Absolute % washout is calculated using the following equation [19]:
(70 sec. density - 15 min. density)
Absolute % washout at 15 min = (70 sec. density - NECT density) × 100.
Chemical shift signal intensity index
Radiologists placed region of interest (ROI) measurements in the nodule on chemical shift MRI, avoiding the edges of the lesion to not include areas of chemical shift artifact of the second kind in the measurement [20]. For nodules that were homogeneous on opposed-phase (OP) images, the ROI was placed on the axial image where the nodule appeared the largest using an ROI that encompassed 2/3 of the area of the nodule and then copy pasted onto the IP image at the same level, Fig. 2. For nodules that were heterogeneous on OP images, an ROI was placed within the area subjectively showing signal intensity loss on OP images (measuring at least 5 mm in size) and copy pasted onto the in-phase (IP) image at the same level, Fig. 4. The chemical shift SI index is calculated, as described previously [21-23] as follows:
Chemical Shift SI index = (SI.mass.IP - SI.mass.OP) SI.mass.IP × 100.
T2W adrenal MRI calculator
Measurements were performed on axial T2W, selecting the center slice of the nodule where it appeared the larg- est, similar to CT-WO. For homogeneous nodules, a cir- cular ROI was placed within the nodule encompassing at least 2/3 of the lesion, Fig. 2. For heterogeneous nodules, a circular ROI was placed within the nodule encompassing the most T2W hyperintense aspect of the nodule, judged subjectively, and measuring at least 5 mm in diameter, as described previously [17], Fig. 3. A fixed diameter (5 mm) circular ROI was placed in the ipsilateral skeletal muscle so that the T2W nodule-to-muscle SI ratio could be calculated: (SInodule/SImuscle) [21].
The axial image used for T2W SI measurements was then segmented manually using ImageJ (version 1.52, National Institutes of Health). Manual segmentation was performed by each reader contouring the entire mass, avoiding the edges of the mass and adjacent structures, Fig. 2. 2-Dimen- sional Histogram analysis was performed and Entropy (a measure of histogram irregularity) was obtained using a free
a
b
C
d
e
f
a
b
C
d
e
f
a
b
C
d
e
f
downloadable plugin software plugin created for ImageJ [24]. Settings were changed from the original description by Tu et al. to use the natural logarithm [25] instead of the original log base 2 and to remove the constraint of con- stant bin widths. The process of exporting an anonymized DICOM image, segmentation, and histogram analysis required approximately 5 min per patient and the use of a separate laptop operating ImageJ. The first-order texture feature Entropy was used in accordance with the study by Tu et al. since the prior study showed Entropy to be the best performing first-order texture analysis feature to differentiate lipid-poor adenomas from metastases.
Statistical analysis
This study was designed as a non-inferiority assessment to determine if the T2W adrenal MRI calculator is not inferior to CT-WO for diagnosis of adrenal adenoma. We used the study by Akbulut et al. which reported an overall accuracy of 69.7% for diagnosis of adrenal adenoma using CT-WO [12] and the accuracy of 95% in the study by Tu et al. who evalu- ated the T2W calculator [17]. The non-inferiority limit was set at 5% with a significance level (alpha) of 5% and a Power (beta) of 90%. A minimum sample of 25 masses evaluated by CT-WO and MRI was required to be 90% certain that the upper limit of a one-sided 95% confidence interval (CI) or
opposed-phase (e) T1W GRE images depict the mass has heteroge- neous signal intensity on both in-phase (d) and opposed-phase (e) images (arrows) with a small focus of microscopic fat (arrowhead) at the medial margin of the mass in (e). Axial T2W image (f) shows the mass (arrow) as heterogeneously hyperintense and T2W calculator- predicted non-adenoma
equivalently a 90% two-sided 95% CI will exclude a differ- ence between CT-WO and T2W adrenal MRI calculator of more than 5%.
For CT-WO, we constructed 2×2 tables for each radi- ologist including the number of true -positive, false-posi- tive, true-negative, and false-negative diagnoses of adrenal adenoma compared to reference standards. A diagnosis of adrenal adenoma at CT-WO was considered when the nodule measured < 10 HU at unenhanced CT or in cases where the nodule measured ≥ 10 HU at unenhanced CT had absolute washout ≥ 60%, Fig. 5. For MRI, a similar strategy of con- structing 2×2 tables for each radiologist was performed. A diagnosis of adrenal adenoma was considered when the nodule showed chemical shift SI index ≥ 16.5%. In nodules with chemical shift SI index < 16.5%, the T2W adrenal MRI calculator was used. T2W signal intensity ratio and entropy were imputed into an equation derived from the original logistic regression model described by Tu et al. [17]. The methodology for derivation of the T2W adrenal MRI calcu- lator equation from the original logistic regression model is described in Supplementary Text 2. The calculator yields a predicted probability that a given nodule represents a metas- tasis or pheochromocytoma using a threshold of 0.652 with sensitivity of 88% and specificity of 100%. The diagnostic accuracy was then compared between CT-WO and MRI for both readers and overall accuracy compared using the
Indeterminate adrenal nodule
CT WO
MRI
Attenuation on unenhanced phase
Chemical shift SI index
<10 HU
>10 HU
>16.5%
<16.5%
Adenoma
Proceed to washout
Adenoma
Proceed to adrenal T2W calculator (T2WC)
>60% absolute washout
<60% absolute washout
T2WC <0.652
T2WC >0.652
Adenoma
Indeterminate; consider alternate Dx
Adenoma
Indeterminate; consider alternate Dx
McNemar test. A p value ≤0.05 was considered statistically significant. Data analyses were conducted using STATA v15.1 (Statacorp, College Station, TX, USA).
Results
A summary of patient demographic factors, adrenal nod- ule size, and diagnosis is provided in Tables 1 and 2. 45% (18/40) patients had underlying malignancy and 55% (22/40) nodules were incidentalomas.
Considering washout CT, overall sensitivity, specific- ity, and accuracy for diagnosis of adenoma were 78% (95% confidence intervals 56-93%), 35% (14-62%), and 57% (42-71%) Reader 1 and 72% (53-87%), 46% (17-77%), and 59% (41-76%) for Reader 2. These results are sum- marized in Table 3. Lipid-rich adenomas (38% [11/29] for both Readers) constituted all masses that measured < 10HU. Therefore, non-contrast-enhanced CT attenuation was 100% specific (95% CI 68-100%) for diagnosis of adrenal ade- noma. Conversely, 62% (18/29) adenomas were lipid poor
| Adenoma (N=29) | Non-adenoma (N=11) | |
|---|---|---|
| Age (years) | Mean 61.4±13.3 years (range 31-81) | Mean 48.5±19.6 years (range 21-80) |
| Gender | 69% Female (N=20) 31% Male (N=9) | 45% Female (N=5) 55% Male (N=6) |
| Size (long-axis PV phase, mm) | 24.1±9.6 | 33.1±18.5 |
for both Readers, resulting in a sensitivity of 38% (95% CI 21-58%). There were 45% (5/11; 3 pheochromocytoma, 2 solitary fibrous tumors) masses for Reader 1 and 55% (6/11; 3 pheochromocytoma, 2 solitary fibrous tumors, 1 ACC) masses for Reader 2 which had absolute washout >60%. For reader 1, 41% (12/29) and for reader 2 31% (9/29) ade- nomas had absolute washout <60% and would thus have been classified as non-adenomas. Whereas, 45% (5/11;
| Incidentaloma (N=22) Adenoma (N=16) | Adrenal nodule in patient with a history of malignancy (N=18) Adenoma (N=13) | |
|---|---|---|
| Age (years) | Mean 56.8 ± 14.5 years (range 31-74) | Mean 67.2±9.0 years (range 42-81) |
| Gender | 69% Female (N=11) 31% Male (N=5) | 69% Female (N=9) 31% Male (N=4) |
| Size (long axis, mm) | 25.1±9.9 | 22.9±9.6 |
| Non-adenoma (N=6) | Non-adenoma (N=5) | |
| - Pheochromocytoma (N=2) - Solitary fibrous tumor (N=3) - Thrombosed cavernous hemangioma (N=1) | - Pheochromocytoma (N=3) - Metastasis (N=1) | |
| - Poorly differentiated adreno- cortical carcinoma (N=1) | ||
| Age (years) | Mean 56.3±20 years (range 21-80) | Mean 39.0± 16 years (range 21-56) |
| Gender | 33% Female (N=2) 66% Male (N=4) | 60% Female (N=3) 40% Male (N=2) |
| Size (long axis, mm) | 35.5±23.5 | 30.2±12.1 |
3 pheochromocytoma, 2 solitary fibrous tumor) and 55% (6/11; 3 pheochromocytoma, 2 solitary fibrous tumor, 1 ACC) non-adenomas for reader 1 and reader 2, respectively, had absolute washout> 60% and would have been classified as adenomas.
Considering MRI, overall sensitivity, specificity, and accuracy for diagnosis of adenoma were 100% (88-100%), 64% (31-89%), and 82% (77-97%) Reader 1 and 86% (68-96%), 73% (39-94%), and 80% (64-95%) Reader 2. These results are summarized in Table 3. MRI had higher overall accuracy (p=0.02 Reader 1, 0.05 Reader 2) com- pared to CT-WO. Among adrenal adenomas, 93% (27/29) for Reader 1 and 69% (21/29) for Reader 2 adenomas were lipid rich (i.e., had CSI> 16.5%). For both readers, 27% (3/11) masses had microscopic fat with CSI> 16.5%, 2 solitary fibrous tumors, and 1 ACC. The sensitivity and specific- ity of chemical shift MRI for diagnosis of adrenal adenoma were 93% (95% CI 76-99%) and 73% (95% CI 39-93%) for Reader 1 and 72% (95% CI 53-87%) and 73% (95% CI 39-93%) for Reader 2. The ACC showed a small focus of microscopic fat and was otherwise classified as a non- adenoma using the T2W calculator by both readers, Fig. 4. Among non-adenomas, the T2W calculator classified 73% (8/11) for Reader 1 and 64% (7/11) for Reader 2 masses cor- rectly with false-positive diagnoses in two solitary fibrous tumors and 1 pheochromocytoma for Reader 1 and one addi- tional pheochromocytoma for Reader 2.
| Sensitivity [95% confidence intervals] | Specificity [95% confidence intervals] | Overall Accuracy [95% confidence intervals] | |
|---|---|---|---|
| Reader 1 | |||
| CT Washout | 78 [56,93] | 35 [14, 62] | 57 [42, 71] |
| MRI | 100 [88, 100] | 64 [34, 90] | 82 [67, 97] |
| Reader 2 | |||
| CT Washout | 72 [53, 87] | 46 [17, 77] | 59 [41, 76] |
| MRI | 86 [68, 96] | 73 [39, 94] | 80 [64, 95] |
Discussion
Adrenal washout CT is recommended as the first-line imag- ing test for indeterminate incidental adrenal masses meas- uring 1-4 cm by the ACR White Paper [10]. The adrenal washout CT test is considered an imaging reference standard for diagnosis of adrenal adenomas based on its description as a highly accurate test to differentiate adenomas from non- adenomas [26]. The washout CT test uses an unenhanced CT threshold of < 10 HU to diagnose microscopic fat, a thresh- old which has been shown to be characteristic of adrenal adenoma [4]. In our study, all masses which measured < 10 HU in attenuation at unenhanced CT were adenomas. Only 30% of adenomas in our study were lipid rich (i.e.,
measured < 10 HU) which is far less than the 70% incidence of lipid-rich adenomas described in the literature [6]. This is almost certainly related to the study population, which is biased since all patients underwent a dedicated adrenal washout CT protocol usually performed for characterization of lipid-poor adrenal masses.
It is known that 30% of adenomas will have insufficient internal fat to be detected at unenhanced CT and will meas- ure ≥ 10 HU [6]. In these cases, chemical shift MRI has been shown to be useful. Two previous studies demonstrated that chemical shift MRI could detect lower levels of micro- scopic fat compared to unenhanced CT using an attenuation threshold of 10 HU and that MRI was particularly in adrenal adenomas measuring between 10 and 30 HU [7, 8]. Our results confirm these findings, as the number of lipid-poor adenomas was only 10% for Reader 1 and 30% for Reader 2. The observed difference between readers may be related to the inherent subjectivity of detection of smaller amounts of microscopic fat in adrenal masses and placement of ROIs within masses with heterogeneous signal intensity drop on opposed phase compared to in-phase images. Our results therefore confirm that there remains definite value in per- forming chemical shift MRI for lipid-poor adrenal masses detected at CT. False-positive diagnoses of adenoma using chemical shift MRI are rare but include primarily fat con- taining metastases and ACC [6]. There were three false-pos- itive diagnoses in our study for both readers which consisted of two solitary fibrous tumors and 1 ACC. The imaging find- ings of solitary fibrous tumors are not well described. In our study, all three masses measured > 10 HU at unenhanced CT, two of three masses had washout in the adenoma range, two of three masses showed microscopic fat on MRI, and two of three masses were classified as adenoma using the T2W calculator. The ACC in our study showed a small focus of microscopic fat. Heterogeneous signal intensity drop on chemical shift MRI remains an indeterminate imaging find- ing. In one study, this was associated with benign disease [27].
For characterization of lipid-poor adrenal masses, adrenal washout CT is usually performed. Most recently, the diag- nostic accuracy of adrenal washout CT has been reconsid- ered. In a recent study, up to one-third of lipid-poor adeno- mas did not washout in the adenoma range [12]. It is also well known that up to one-third of pheochromocytoma may washout in the adenoma range [13, 14]. In one study, nearly one-third of ACC also showed absolute washout in the ade- noma range [15] and metastases from hypervascular tumors may also washout in the adenoma range [16]. Our study re- enforces these concepts. Roughly one-third of adenomas did not have absolute washout in the adenoma range (i.e., <60%) and would have been incorrectly classified as non-adenomas. In addition, nearly half of non-adenomas had washout in the adenoma range (i.e., ≥ 60%) and would have been incorrectly
classified as adenomas using only washout parameters for diagnosis (Table 3).
We sought to determine if a recently described T2W cal- culator which is a two-variable logistic regression model consisting of T2W signal intensity ratio and entropy [17], when combined with chemical shift MRI, is not inferior to adrenal washout CT. Our results confirm that the T2W cal- culator is not inferior to adrenal washout CT for characteri- zation of indeterminate adrenal masses and overall accuracy was actually significantly higher for both readers. To our knowledge, this is the first study to combine chemical shift MRI and the T2W calculator and it remains unclear how to proceed when there is a discrepancy between the pres- ence of microscopic fat (which is diagnostic of adenoma) and T2W calculator results. For example, in the encoun- tered ACC which showed a small focus microscopic fat but with T2W calculator results indicating non-adenoma, which value should be used for clinical diagnosis? This will require further study. Not all lipid-poor adenomas for Reader 2 and roughly one-third of non-adenomas were incorrectly diag- nosed using the T2W adrenal calculator; however, results were significantly better than with washout CT.
Our study has limitations. The population is biased and not reflective of a consecutive sample of adrenal masses. Patients undergoing adrenal washout CT and MRI likely represent a selected group of patients with indeterminate adrenal masses and this is reflected in the encountered pathologies, including rare solitary fibrous tumors and ACC. Adrenal washout CT may perform better in a consecutive sample of adrenal masses; however, the same could be true for the use of chemical shift MRI and the T2W adrenal calculator. Further verification of our results in less-biased populations would be useful. We included CT and MRI performed over a wide period of time with differences in CT and MRI hardware and software during the study dates. Although this technically should not impact performance of washout CT, the changing parameters encountered between 1.5 and 3 Tesla, use of conventional T2W TSE/ FSE versus single-shot techniques and other differences in software (e.g., signal intensity correction algorithms, par- allel imaging) introduced over time could have impacted performance of the MRI calculator. Reference standards for adrenal masses are challenging, particularly for benign adrenal adenomas which frequently do not undergo histo- logical verification or further follow-up imaging. The use of microscopic fat as an imaging reference standard for adrenal adenoma in patients without a history of primary malig- nancy is not optimal; however, is in our opinion acceptable given the accuracy of diagnosis using both unenhanced CT and MRI in this population [6].
In conclusion, our study reports that the use of chemi- cal shift MRI combined with the T2W adrenal calculator (a logistic regression model encompassing T2W signal
intensity ratio and T2W entropy) is not inferior to adrenal washout CT for characterization of indeterminate adrenal masses. In this study, MRI was significantly more accurate than washout CT for diagnosis of adenoma. Reduced perfor- mance of washout CT stems from a relatively high number of adrenal adenomas which did not washout in the adenoma range and roughly half of non-adenomas washing out in the adenoma range. MRI shows improved ability to detect microscopic fat increasing diagnosis of lipid-rich adenomas and the T2W calculator more accurately differentiated non- adenomas from adenomas in lipid-poor masses. Integrating the T2W MRI calculator with chemical shift imaging shows promise as a potential alternative to CT Washout in evaluat- ing incidental non-metastatic adrenal masses but requires further study.
Supplementary Information The online version contains supplemen- tary material available at https://doi.org/10.1007/s00261-022-03533-1.
References
1. Mansmann G, Lau J, Balk E, Rothberg M, Miyachi Y, Bornstein SR (2004) The clinically inapparent adrenal mass: update in diagnosis and management. Endocr Rev 25:309-340
2. Song JH, Chaudhry FS, Mayo-Smith WW (2008) The incidental adrenal mass on CT: prevalence of adrenal disease in 1,049 con- secutive adrenal masses in patients with no known malignancy. AJR Am J Roentgenol 190:1163-1168
3. Glazer HS, Weyman PJ, Sagel SS, Levitt RG, McClennan BL (1982) Nonfunctioning adrenal masses: incidental discovery on computed tomography. AJR Am J Roentgenol 139:81-85
4. Boland GW, Lee MJ, Gazelle GS, Halpern EF, McNicholas MM, Mueller PR (1998) Characterization of adrenal masses using unenhanced CT: an analysis of the CT literature. AJR Am J Roentgenol 171:201-204
5. Maas M, Nassiri N, Bhanvadia S, Carmichael JD, Duddalwar V, Daneshmand S (2021) Discrepancies in the Recommended Management of Adrenal Incidentalomas by Various Guidelines. J Urol 205:52-59
6. Schieda N, Siegelman ES (2017) Update on CT and MRI of Adrenal Nodules. AJR Am J Roentgenol 208:1206-1217
7. Haider MA, Ghai S, Jhaveri K, Lockwood G (2004) Chemi- cal shift MR imaging of hyperattenuating (>10 HU) adrenal masses: does it still have a role? Radiology 231:711-716
8. Israel GM, Korobkin M, Wang C, Hecht EN, Krinsky GA (2004) Comparison of unenhanced CT and chemical shift MRI in evaluating lipid-rich adrenal adenomas. AJR Am J Roent- genol 183:215-219
9. Grajewski KG, Caoili EM (2021) Adrenal Washout CT: Counterpoint-Remains a Valuable Tool for Radiologists Char- acterizing Indeterminate Nodules. AJR Am J Roentgenol 216:1168-1169
10. Mayo-Smith WW, Song JH, Boland GL et al (2017) Manage- ment of Incidental Adrenal Masses: A White Paper of the ACR Incidental Findings Committee. J Am Coll Radiol 14:1038-1044
11. Corwin MT, Remer EM (2021) Adrenal Washout CT: Point- Not Useful for Characterizing Incidentally Discovered Adrenal Nodules. AJR Am J Roentgenol 216:1166-1167
12. Akbulut S, Erten O, Kahramangil B et al (2021) A Critical Analysis of Computed Tomography Washout in Lipid-Poor Adrenal Incidentalomas. Ann Surg Oncol 28:2756-2762
13. Schieda N, Alrashed A, Flood TA, Samji K, Shabana W, McI- nnes MD (2016) Comparison of Quantitative MRI and CT Washout Analysis for Differentiation of Adrenal Pheochro- mocytoma From Adrenal Adenoma. AJR Am J Roentgenol 206:1141-1148
14. Patel J, Davenport MS, Cohan RH, Caoili EM (2013) Can estab- lished CT attenuation and washout criteria for adrenal adenoma accurately exclude pheochromocytoma? AJR Am J Roentgenol 201:122-127
15. Zhang HM, Perrier ND, Grubbs EG et al (2012) CT features and quantification of the characteristics of adrenocortical carcino- mas on unenhanced and contrast-enhanced studies. Clin Radiol 67:38-46
16. Choi YA, Kim CK, Park BK, Kim B (2013) Evaluation of adre- nal metastases from renal cell carcinoma and hepatocellular carcinoma: use of delayed contrast-enhanced CT. Radiology 266:514-520
17. Tu W, Abreu-Gomez J, Udare A, Alrashed A, Schieda N (2020) Utility of T2-weighted MRI to Differentiate Adrenal Metastases from Lipid-Poor Adrenal Adenomas. Radiol Imaging Cancer 2:e200011
18. Gerson R MR, Tu W, Abreu-Gomez J, Udare A, Schieda N (2022) Evaluation of the T2-weighted (T2W) adrenal MRI cal- culator to differentiate adrenal pheochromocytoma from lipid- poor adrenal adenoma. European Radiology
19. Schieda N, Al Dandan O, Kielar AZ, Flood TA, McInnes MD, Siegelman ES (2014) Pitfalls of adrenal imaging with chemical shift MRI. Clin Radiol 69:1186-1197
20. Ramamurthy NK, Moosavi B, McInnes MD, Flood TA, Schieda N (2014) Multiparametric MRI of solid renal masses: pearls and pitfalls. Clin Radiol. https://doi.org/10.1016/j.crad.2014.10.006
21. Sasiwimonphan K, Takahashi N, Leibovich BC, Carter RE, Atwell TD, Kawashima A (2012) Small (<4 cm) renal mass: differentiation of angiomyolipoma without visible fat from renal cell carcinoma utilizing MR imaging. Radiology 263:160-168
22. Hindman N, Ngo L, Genega EM et al (2012) Angiomyolipoma with minimal fat: can it be differentiated from clear cell renal cell carcinoma by using standard MR techniques? Radiology 265:468-477
23. Karlo CA, Donati OF, Burger IA et al (2013) MR imaging of renal cortical tumours: qualitative and quantitative chemical shift imaging parameters. Eur Radiol 23:1738-1744
24. Backer V (2020) Discrete_Histogram_Entropy_Tool. Github, Github. Available via https://github.com/MontpellierRessource sImagerie/imagej_macros_and_scripts/wiki/Discrete_Histo gram_Entropy_Tool. Accessed September 2021
25. Wallis K (2006) A note on the calculation of entropy from histograms.
26. Caoili EM, Korobkin M, Francis IR et al (2002) Adrenal masses: characterization with combined unenhanced and delayed enhanced CT. Radiology 222:629-633
27. Gabriel H, Pizzitola V, McComb EN, Wiley E, Miller FH (2004) Adrenal lesions with heterogeneous suppression on chemical shift imaging: clinical implications. J Magn Reson Imaging 19:308-316
Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.