KIDNEYS, URETERS, BLADDER, RETROPERITONEUM
Check for updates
Validation of the modified CT criteria for identifying non-adenomas
Min Hwan Kwak1 . Ji-Sup Yun2 . Ji Na Kim3 . Young Rae Lee3 . Chan Kyo Kim1 . Kyung A. Kang1(D
Received: 14 August 2023 / Revised: 30 December 2023 / Accepted: 12 January 2024 / Published online: 27 February 2024 @ The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2024
Abstract
Purpose Although adrenal computed tomography (CT) percentage washout is a potentially powerful imaging technique for differentiating adrenal adenomas from non-adenomas, its application to non-adenomas can be problematic. Recently, modified criteria for diagnosing pheochromocytomas using adrenal CT were developed based on data from 199 patients with surgically proven pheochromocytomas and adenomas. However, these criteria have not been thoroughly validated. The purpose of this study was to validate the performance of the modified criteria for diagnosing non-adenomas including pheochromocytomas. Methods The conventional and modified criteria were applied to 266 patients from two cohorts who had surgically proven lipid-poor adenomas (155/266, 58.3%) and non-adenomas (111/266, 41.7%) and underwent adrenal CT. Two radiologists calculated the attenuation on each dynamic phase and percentage washout of adrenal masses. The final assessments based on the conventional and modified criteria were categorized into adenomas or non-adenomas. The diagnostic performance of each criterion for diagnosing non-adenomas was evaluated using the area under the receiver operating characteristic curve (AUC). False negatives and positives were also compared.
Results The AUC for the diagnosis of non-adenomas was 0.806 for conventional criteria and 0.858 for modified criteria (p=0.047). The false-negative rate of conventional criteria for the diagnosis of non-adenomas was 29.7%. Use of modified criteria could have reduced the false-negative rate by to 7.2%. The false-positive rate increased from 9% to 21.3% when using the modified criteria.
Conclusion The utilization of modified criteria has the potential to identify additional non-adenomas that would otherwise be misdiagnosed as adenomas using conventional criteria alone.
Min Hwan Kwak and Ji-Sup Yun have equally contributed co-first authors to this work.
☒ Chan Kyo Kim chankyo.kim@samsung.com
☒ Kyung A. Kang Kyunga03.kang@samsung.com
1 Department of Radiology, Samsung Medical Center, Sungkyunkwan University School of Medicine, 81 Irwon-Ro, Gangnam-Gu, Seoul 06351, Korea
2 Department of Surgery, Kangbuk Samsung Hospital, Sungkyunkwan University School of Medicine, Seoul, Korea
3 Department of Radiology, Kangbuk Samsung Hospital, Sungkyunkwan University School of Medicine, Seoul, Korea
Graphical abstract
Validation of the modified CT criteria for identifying non-adenomas
False-negative rate of CT criteria for diagnosing non-adenomas
Modified CT criteria could identify an additional 23% of non-adenomas.
· Conventional criteria: lesion attenuation on UCT >10 HU; APW < 60%; AND RPW <40%
29.7%
23%
· Modified criteria: (a) conventional criteria; OR (b) one of following findings: (i) lesion attenuation on UCT ≥ 40 HU, (ii) 1-min CECT ≥ 160 HU, (iii) 15-min CECT≥70 HU, OR (iv) intralesional cystic degeneration
7.2%
Conventional
Modified
Abdominal Radiology The Official Journal of the Society of Abdominal Radiology www.abdominalradiology.org
Kwak et al; 2024
Keywords Adrenal glands · Adenoma · Pheochromocytoma · Adrenocortical carcinoma · Neoplasm metastasis
Introduction
Adrenal adenomas are the most common type of adrenal gland tumors, representing 50-80% of adrenal lesions [1]. Dedicated adrenal computed tomography (CT) has helped radiologists differentiate adenomas from non-ade- nomas in patients with incidental adrenal masses. Adrenal adenomas are typically diagnosed when a lesion meets one of the following criteria: (a) ≤ 10 Hounsfield unit (HU) in unenhanced CT (UCT); (b) absolute percentage washout (APW) ≥ 60%; or (c) relative percentage wash- out (RPW) ≥ 40% [2]. However, a recent meta-analysis including 10 studies (114 pheochromocytomas) revealed relatively low specificity (0.67), although the overall diag- nostic performance was excellent, with an area under the curve (AUC) of 0.97 [3]. This is probably due to a signifi- cant overlap between adenomas and pheochromocytomas on CT washout. Approximately one-third of pheochro- mocytomas exhibit washout within the adenoma range [3-6]. This indicates that a non-negligible proportion of pheochromocytomas can be falsely diagnosed as adenomas using this criterion alone. Therefore, findings other than the washout percentage should be used when diagnosing pheochromocytoma.
Recently, modified criteria using adrenal CT was devel- oped to detect more pheochromocytomas, based on data from 199 patients who surgically proven pheochromo- cytomas and adenomas [7]. These modified criteria pro- vide optimal threshold values for each dynamic phase in
addition to the washout percentage. This improved the diagnostic performance to distinguish pheochromocyto- mas from adrenal adenomas.
However, its performance has not been externally vali- dated. Furthermore, the modified criteria were limited to pheochromocytomas in original study. In real-world clinical practice, adenomas and pheochromocytomas are not the sole types of adrenal masses. When incidental adrenal masses do not meet the adenoma washout criteria on CT, it is fre- quently problematic to differentiate pheochromocytomas from other non-adenomas (e.g., adrenocortical carcinoma or metastasis). Therefore, this study also aimed to determine whether the previously proposed criteria could be extended to non-adenomas and employed to differentiate pheochro- mocytomas from other non-adenomas.
The purpose of our study was to validate the performance of the modified criteria in the diagnosis of non-adenomas, including pheochromocytomas.
Materials and methods
Study population
For temporal validation, more recent patient data were col- lected from settings similar to those in which the model was developed. We included 298 patients who had undergone adrenal CT and adrenalectomy between January 2018 and December 2021. Patients were excluded if lesion attenua- tion was ≤ 10 HU in unenhanced CT (n=81). Furthermore,
the following patients were excluded; (a) non-neoplastic diseases, such as cortical hyperplasia, endothelial cyst, lymphangioma, or tuberculosis (n=29); (b) myelolipoma (n=4); (c) extra-adrenal lesion (n=3); (d) lesion containing macroscopic fat, measuring ← 20HU [8] (n=5; all adeno- mas with or without myelolipomatous metaplasia); (e) pre- dominantly cystic mass (n=3); and (f) rare histopathologic subtypes [(n=8); including cortical neoplasm with uncertain malignant potential (n=4), hemangioma (n=2), solitary fibrous tumor (n=1), and adenomatoid tumor (n=1)].
A spatially independent dataset was prepared at Kangbuk Samsung Hospital. Between January 2012 and December 2021, 154 patients with adrenal masses who underwent adrenal CT and adrenalectomy were included. Of the 154 patients, 27 with lipid-rich adenomas were excluded. Subse- quently, 26 patients were excluded for the following reasons: (a) non-neoplastic disease (n=18); (b) myelolipoma (n=4); (c) lesions containing macroscopic fat, measuring ← 20HU [8] (n= 1, adenoma with myelolipomatous metaplasia); and (d) rare histopathologic subtypes [n=3, including cortical neoplasm with uncertain malignant potential (n=2), and ganglioneuroblastoma (n=1)].
Our study excluded adrenal masses with an attenuation of ≤ 10 HU in UCT. Attenuation measurements ≤ 10 HU are highly specific for diagnosing of lipid-rich adenomas [4, 9, 10]. The updated 2023 guideline from the European Society of Endocrinology provides advanced insights into the management of adrenal incidentalomas [11]. In terms of assessing malignant risk, homogeneous lesions with low attenuation on unenhanced CT below 10 are con- sidered benign and do not require any additional imag- ing, regardless of their size. According to the systematic reviews, CT density > 10 HU exhibits remarkably high
sensitivity (100% in all 6 studies [95% CI 100%-100%] for detection of adrenal malignancy [12-17]. This implies that adrenal masses with a density of ≤ 10 HU are vir- tually never malignant. In our study, all adrenal masses with an attenuation of ≤ 10 HU (n= 108) were surgically confirmed as adenomas.
Our study excluded non-neoplastic adrenal masses (n=47), including adrenal hyperplasia (n=22), cystic lesions (n=20), adrenal tuberculosis (n=2), and cases with pathological findings not explicitly outlined (n=3). The majority consisted of cortical hyperplasia (n =22). The enhancement pattern or morphological features of cortical hyperplasia cannot reliably distinguish between cortical hyperplasia and adenomas [18-20]. However, our research primarily aimed to distinguish non-adenomas, such as pheochromocytoma or adrenocortical carcinoma, from adenomas; therefore, we excluded this criterion. Cystic lesions were also excluded [n = 20; endothe- lial cysts (n=12), lymphangiomas (n=6), pseudocysts (n =2)], because the imaging features of uncomplicated cysts are usually straightforward, and they can be diag- nosed by lack of enhancement on CT. Adrenal tuberculosis could be misdiagnosed as primary adrenal tumor on CT; however, the numbers were too small to establish statis- tical significance (n =2). In a few instances, the patho- logical findings were not explicitly outlined (n=3, e.g., no evidence of malignancy).
We recruited 266 patients from both cohorts (Fig. 1). Two participants aged 16 and 17 were included, with the rest of the cohort being 20 years old or older. The study protocol was approved by each hospital’s institutional review board, and the requirement for informed consent was waived.
Patients who underwent adrenal CT and adrenalectomy
Center I (n = 298) from Jan. 2018 to Dec. 2021
Center II (n = 154) from Jan. 2012 to Dec. 2021
81 patients excluded - Lipid-rich adenoma (Unenhanced CT density less than 10 HU)
27 patients excluded - Lipid-rich adenoma (Unenhanced CT density less than 10 HU)
52 patients excluded
- Non-neoplastic disease (n = 29)
26 patients excluded
- Myelolipoma (n = 4)
- Extra-adrenal lesions (n=3)
- Non-neoplastic disease (n = 18)
- Lesion containing macroscopic fat (n = 5)
- Myelolipoma (n = 4)
- Predominantly cystic mass (n = 3)
- Lesion containing macroscopic fat (n = 1) - Rare histopathologic subtypes (n = 3)
- Rare histopathologic subtypes (n = 8)
Study population (n = 165)
Study population (n = 101)
Final inclusion (n = 266)
CT protocol
CT examinations were performed using one of five multi- detector CT scanners in Center I (Discovery CT750 HD, Revolution Frontier, GE Healthcare; Revolution Apex, GE Healthcare; Somatom Definition Flash, Siemens Healthcare; Somatom Force, Siemens Healthcare; Aquilion ONE(TSX- 305A), Canon Medical Systems) and one of two multidetec- tor CT scanners in Center II (Brilliance 64, Phillips Medi- cal Systems; Brilliance iCT SP, Phillips Medical Systems). All adrenal CT scans were conducted using single-energy CT. Imaging parameters were as follows: slice thickness, 2.0-2.5 mm; tube voltage, 120 kVP; variable tube current determined by automated tube current modulation. A total of 120 mL of nonionic iodine contrast agent was intravenously administered at a rate of 3.0 mL/s using a power injector. The adrenal CT protocol comprised an unenhanced scan, followed by a venous phase scan (60 s) and a delayed phase scan (15 min) after intravenous contrast material administra- tion [2, 21].
Image analyses
All images from Center I were independently reviewed by two radiologists (Reader 1, Faculty of Genitourinary Radiology; Reader 2, a radiology resident in their fourth year of training). Two readers were blinded to clinical and pathologic information. The image analysis of Center II was conducted in person by Reader 1 because the external dis- tribution of the images was prohibited. For each CT phase, a round region of interest (ROI) was manually drawn in the center of the adrenal mass to measure the CT attenuation values. The ROI was drawn at the same section and loca- tion of the lesion on each phase of adrenal CT. ROI was placed to encompass two-thirds of the lesion while avoid- ing calcifications [4, 22, 23]. For heterogeneous lesions, the well-enhancing solid component were indispensably included when drawing the ROI, while incorporating some of the adjacent cystic, necrotic, and hemorrhagic compo- nents to cover two-thirds of the lesion [24, 25]. Because Park et al. suggested that an ROI should cover more than one-half of the mass to maximize the detection of cortical carcinoma [26]. In terms of size, adrenal masses measuring less than 1 cm were infrequent but included (8 out of 266 cases for Reader 1 and 5 out of 165 cases for Reader 2). However, all of them were above 0.7 cm. For small lesions less than 1 cm, the largest ROI that could be accommodated within the enhancing portion of the mass was used, while avoiding calcification or partial volume effects from adjacent structures [27]. If multiple or bilateral lesions had similar characteristics, measurements were obtained from the larg- est lesion.
As described in original study, at least one of the fol- lowing should be satisfied for the diagnosis of pheochro- mocytomas (modified CT criteria): (1) conventional crite- ria for pheochromocytoma (lesion attenuation in UCT> 10 HU, APW <60%, and RPW <40%); (2) lesion attenua- tion on unenhanced CT≥ 40 HU; (3) 1-min enhanced CT≥ 160 HU; (4) 15-min enhanced CT ≥70 HU; or (5) intralesional cystic degeneration seen in both 1-min and 15-min enhanced CT [7].
To more precisely investigate which of the abovemen- tioned components actually impacted diagnostic perfor- mance, we subdivided them into three categories to clas- sify whether adrenal tumors were non-adenomas.
Criterion I (conventional criteria): lesion attenuation in UCT>10 HU, APW <60%, and RPW <40%.
Criterion II (cystic degeneration criteria): lesions show- ing focal areas of low attenuation at the same site on 1-min and 15-min CT scans.
Criterion III (mean attenuation criteria): lesions show- ing attenuation on UCT ≥ 40 HU; 1-min enhanced CT ≥ 160 HU; or 15-min enhanced CT ≥ 70 HU.
The modified CT criteria were defined as a combination of criteria I, II or III.
APW and RPW were calculated using the following formula:
APW = (1 - min enhanced HU - 15 - min delayed HU) /(1 - min enhanced HU - unenhanced HU) × 100%
RPW = (1 - min enhanced HU - 15 - min delayed HU) /(1 - min enhanced HU) × 100%
Statistical analysis
Continuous variables were expressed as means ± stand- ard deviations or as medians (interquartile ranges [IQRs]) according to the data distribution. Categorical variables are expressed as frequencies (percentages). For compari- son between the two groups (adenomas vs. non-adeno- mas), an independent t-test or Mann-Whitney U test was used for continuous variables, and the chi-square test or Fisher’s exact test was used for categorical variables.
For the subgroup analysis, we subdivided non-ade- nomas into four groups based on histopathology (pheo- chromocytomas, metastasis, adrenocortical carcinomas, and neurogenic tumors). For comparison between the five groups, analysis of variance or the Kruskal-Wallis test was used for continuous variables, and the chi-square or
Fisher’s exact test was used for categorical variables. For pairwise comparisons, post-hoc analysis with Bonferroni correction was used.
Receiver operating characteristic (ROC) curves for the diagnosis of non-adenomas were analyzed. We compared the sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and accuracy of the three criteria.
Regarding the inter-reader agreement, the intraclass cor- relation coefficient was used for continuous variables, and weighted kappa was used for categorical variables. The cor- relation coefficient <0.2 was considered poor, 0.21-0.40, fair; 0.41-0.60, moderate; 0.61-0.80, good; and 0.81-1.0 excellent agreement [28]. All statistical analyses were performed with SAS version 9.4 (SAS Institute, Cary, NC, USA). A p-value <0.05 was considered statistically significant.
Results
Baseline characteristics
A comparison of the baseline characteristics of patients with lipid-poor adenomas and non-adenomas, evaluated by reader 1, is presented in Table 1. Of the 266 patients, the propor- tion of lipid-poor adenomas and non-adenomas was 58.3% (155/266) and 41.7% (111/266), respectively. Non-adeno- mas (3.8 cm [IQR, 2.6-5.2]) were significantly larger than adenomas (2.1 cm [IQR, 1.5-2.6]) (p<0.001). In terms of imaging characteristics, heterogeneous enhancement (67.6% vs. 33.6%, p<0.001), cystic degeneration (27.0% vs. 0.7%, p<0.001), and hemorrhage (8.1% vs. 0.7%, p=0.002) were observed more frequently in non-adenomas than they were in adenomas. Calcifications were more prevalent in non-adenomas than in adenomas (10.8% vs. 3.9%), and the P-value approximated 0.05 (p=0.048). Data from Center
I, evaluated by reader 2, is presented in Supplementary Table 1.
The characteristics of each non-adenoma subtype are listed in Supplementary Table S2. Non-adenomas include pheochromocytomas (n=66), metastases (n=15), adreno- cortical carcinomas (n=9), and neurogenic tumors (n=21). Neurogenic tumors included schwannomas (n=6), gangli- oneuromas (n= 14), and neurogenic tumors with uncertain malignant potential (n=1). Adrenal metastases originated from lung cancer (n=8), renal cell carcinoma (n=3), hepa- tocellular carcinoma (n=2), colon cancer (n= 1), and ovar- ian cancer (n=1). The median diameter of adrenocortical carcinomas (6.7 cm [IQR, 4.2-8.9]) and pheochromocyto- mas [3.3 cm (IQR, 2.2-5.0)] were significantly greater than that of the adenomas [2.1 cm (IQR, 1.5-2.7)] (p<0.001, respectively). Adrenocortical carcinomas and pheochro- mocytomas showed heterogeneous enhancement and cystic changes more frequently than adenomas (88.9% [8/9], 77.3% [51/66] vs. 33.6% [52/155] for heterogeneous enhancement, 33.3% [3/9], 28.8% [19/66] vs. 0.7% [1/155] for a cystic change). A relatively large portion of the adenomas appeared heterogeneous (33.6% [52/155]) but rarely contained cystic degeneration (0.7% [1/155]) or hemorrhage (0.7% [1/155]).
Lesion enhancement
The results of the quantitative attenuation analysis of the adrenal masses, evaluated by reader 1, are shown in Table 2 and Fig. 2. The mean attenuation of adenomas (23.9 HU [IQR, 16.8-31.5]) was significantly lower than that of non- adenomas (34.2 HU [IQR, 27.5-41.1]) in UCT (p<0.001). The mean attenuation was significantly higher in adenomas (101 HU [IQR, 85.9-119.7]) than that of non-adenomas (81.4 HU [IQR, 54.3-108.0]) on the 1-min scan (p<0.001). On the 15-min scan, the mean attenuation of adenomas (46.7 HU [IQR, 36.5-54.6]) was significantly lower than that of non-adenomas (65.6 HU [IQR, 59.7-75.2]) (p<0.001).
| Adenoma (N=155) | Non-adenoma (N=111) | p-value | |
|---|---|---|---|
| Age (year) | 52.3±10.8 | 53.0±15.5 | 0.658 |
| Size (cm)ª | 2.1 (1.5, 2.6) | 3.8 (2.6, 5.2) | <0.001 |
| Sex (Male) | 54 (34.8%) | 51 (45.9%) | 0.089 |
| Bilaterality | 8 (5.2%) | 6 (5.4%) | >0.99 |
| Multiplicity | 11 (7.1%) | 8 (7.2%) | >0.99 |
| Heterogeneity | 52 (33.6%) | 75 (67.6%) | <0.001 |
| Cystic degeneration | 1 (0.7%) | 30 (27.0%) | <0.001 |
| Calcification | 6 (3.9%) | 12 (10.8%) | 0.048 |
| Hemorrhage | 1 (0.7%) | 9 (8.1%) | 0.002 |
Unless indicated otherwise, data are presented as the number of patients, with percentages in parentheses “Data are presented as means ± standard deviations
aData are presented as medians, with ranges in parentheses
| Adenoma (N=155) | Non-adenoma (N=111) | p-value | |
|---|---|---|---|
| Mean attenuation | |||
| Unenhanced scan | 23.9 (16.8, 31.5) | 34.2 (27.5, 41.1) | <0.001 |
| 1-min scan | 101 (85.9, 119.7) | 81.4 (54.3, 108.0) | <0.001 |
| 15-min scan | 46.7 (36.5, 54.6) | 65.6 (59.7, 75.2) | <0.001 |
| APW | 75.3 (67.8, 79.2) | 40.7 (-19.0, 60.9) | <0.001 |
| RPW | 56.1 (49.9, 63.2) | 21.5 (- 12.3, 38.4) | <0.001 |
APW absolute percentage washout, RPW relative percentage washout Data are presented as medians, with ranges in parentheses (in Hounsfield units)
Data from Center I, evaluated by reader 2, is presented in Supplementary Table 1.
The mean attenuation of each non-adenoma subtype is listed in Supplementary Table S3. There was no significant difference between the subtypes of non-adenomas in UCT and 15-min scans. On a 1-min scan, pheochromocytomas (100.3 HU [IQR, 70.5-113.4]) showed significantly high attenuation than adrenocortical carcinomas (68.8 HU [IQR, 47.4-75.8]) (p<0.008) and neurogenic tumors (43.8 HU [IQR, 37.0-60.1]) (p<0.001). Metastasis (81.4 HU [IQR, 73.0-99.8]) showed significantly high attenuation than neu- rogenic tumor 43.8 HU [IQR, 37.0-60.1]) (p<0.013).
The percentage washout values of adenomas were sig- nificantly higher than those of non-adenomas on APW (75.3% [IQR, 67.8-79.2] vs. 40.7 [IQR, -19.0-60.9]) and RPW (56.1 [IQR, 49.9-63.2] vs. 21.5 [IQR, - 12.3-38.4]) (p<0.001). In the non-adenoma group, no significant dif- ferences were identified between pheochromocytomas, metastases, and adrenocortical carcinomas in the pairwise comparison. All neurogenic tumors showed progressive con- trast enhancement patterns; therefore, the median APW and RPW of the neurogenic tumors were negative (Supplemen- tary Figs. S1 and 2). Neurogenic tumors and adrenocorti- cal carcinomas did not show significant difference in terms of RPW (p=0.421). However, apart from this, neurogenic tumors displayed lower values in both APW and RPW when compared to the other groups.
Diagnostic performance of adrenal CT criteria
A summary of the diagnostic performance of the adrenal CT criteria for the diagnosis of non-adenomas is presented in Table 3. The AUCs of the ROC curves were 0.806 (95% confidence interval [CI], 0.758-0.855), 0.829 (95% CI, 0.782-0.875), and 0.858 (95% CI 0.817-0.898) for Cri- terion I alone, Criterion I or II, and Criterion I, II, or III, respectively (Fig. 3). Sensitivities for the diagnosis of non- adenoma were 70.3%, 74.8%, and 92.8%, and NPVs were 81.0%, 83.4%, and 93.9% for Criterion I alone, Criterion I or II, and Criterion I, II, or III, respectively. The specificities
for the diagnosis of non-adenoma were 91.0%, 91.0%, and 78.7%, and PPVs were 84.8%, 85.6%, and 75.7% for Cri- terion I alone, Criterion I or II, and Criterion I, II, or III, respectively.
False-negative rates of adrenal CT criteria
Based on the complement of sensitivity (1-sensitivity), the false-negative rate was 29.7% when using Criterion I alone (conventional criteria) for the diagnosis of non-adenomas. This meant that 30% of non-adenomas were falsely diag- nosed as adenomas by conventional criteria. Adding crite- ria for cystic degeneration (Criterion I or II) reduced the false-negative rate (25.2%) slightly. The false-negative rate was reduced to 7.2% using modified criteria (Criterion I, II, or III) (Fig. 4A). Based on the complement of specificity (1-specificity), the false-positive rate was 9% according to conventional criteria. This meant that 9.0% of the adeno- mas were falsely diagnosed as non-adenomas according to conventional criteria. Adding the criteria for cystic degen- eration (Criterion I or II) did not change the false-positive rate (9.0%). However, the false-positive rate increased to 21.3% when using the modified criteria (Criterion I, II, or III) (Fig. 4B).
Specificity of attenuation threshold in each phase
The specificities for the diagnosis of non-adenomas were 92.9% (95% CI, 0.877-0.964), 94.8% (95% CI, 0.901-0.978), and 94.8% (95% CI, 0.901-0.978) when using thresholds of 40 HU in UCT, 160 HU in 1-min enhanced CT, and 70 HU in 15-min enhanced CT.
Subgroup analyses
The diagnostic performance was also evaluated in the sub- group that included only pheochromocytomas and adeno- mas. The AUCs of the ROC curve were 0.742 (95% CI, 0.674-0.797), 0.773 (95% CI, 0.727-0.844), and 0.847 (95% CI, 0.802-0.897) for Criterion I alone, Criterion I or
A
100.0
209
*
80.0
Attenuation on UCT (HU)
60.0
205
o
40.0
20.0
.0
adenoma
non-adenoma
B
250.0
Attenuation on 1-min scan (HU)
99
200.0
o
112 72
3
150.0
100.0
50.0
.0
adenoma
non-adenoma
C
100.0
209
o
Attenuation on 15-min scan (HU)
80.0
60.0
40.0
183
105
220
68
20.0
.0
adenoma
non-adenoma
| Criterion I alone | Criterion I or II | Criterion I or II or III | |
|---|---|---|---|
| Sensitivity | 0.703 (0.612, 0.780) | 0.748 (0.660, 0.819) | 0.928 (0.864, 0.963) |
| Specificity | 0.910 (0.854, 0.945) | 0.910 (0.854, 0.945) | 0.787 (0.716, 0.844) |
| PPV | 0.848 (0.761, 0.907) | 0.856 (0.772, 0.912) | 0.757 (0.679, 0.822) |
| NPV | 0.810 (0.746, 0.862) | 0.834 (0.771, 0.883) | 0.939 (0.883, 0.969) |
| Accuracy | 0.823 (0.773, 0.864) | 0.842 (0.794, 0.881) | 0.846 (0.798, 0.884) |
| AUC | 0.806 (0.758, 0.855) | 0.829 (0.782, 0.875) | 0.858 (0.817, 0.898) |
PPV positive predictive value, NPV negative predictive value, AUC area under curve of receiver-operating characteristic curve
Data are presented with 95% confidence interval in parentheses
1.0
0.8
Sensitivity
0.6
0.4
0.2
Criterion I
-Criterion I or II
- Criterion I or II or III
0.0
0.0
0.2
0.4
0.6
0.8
1.0
1-specificity
II, and Criterion I, II, or III, respectively, for the diagnosis of pheochromocytomas (Supplementary Table S4).
Data from 66 patients with pheochromocytoma, 15 with metastasis, 9 with adrenocortical carcinoma, and 21 with neurogenic tumors were merged for second subgroup analysis. Diagnostic performance was 0.343 (95% CI, 0.268-0.419), 0.363 (95% CI, 0.291-0.435), and 0.477 (95% CI, 0.430-0.523) for Criterion I alone, Criterion I or II, and Criterion I, II, or III, respectively, for the diagnosis of pheochromocytomas (Supplementary Table S4).
Biochemical test for pheochromocytomas
It is well established that plasma free metanephrines exhibit high sensitivity in confirming pheochromocytomas [99% (95% CI, 96%-100%)] [29]. In our study, 130 out of 165 patients from center I underwent plasma free metanephrines
testing. The sensitivity, specificity, PPV, and NPV for diag- nosing pheochromocytoma were 0.979, 0.868, 0.807, and 0.986, respectively (applying a cutoff value of plasma nor- metanephrine > 0.6 nmol/L and metanephrine > 0.3 nmol/L). The AUC for pheochromocytoma diagnosis increased from 0.734 to 0.846 when combining imaging (criteria I or II or III) and biochemical tests (Supplementary Table S5). Among the 48 patients with pheochromocytoma, only one had normal levels of plasma free metanephrines (norme- tanephrine = 0.46, metanephrine =0.3). In this particular case, the mass was initially diagnosed as an adenoma based on conventional criteria, with an APW of 65.2% and RPW of 39.5%, as assessed by reader 1. The mass showed het- erogeneous enhancement without cystic degeneration. How- ever, it demonstrated > 40 HU on unenhanced CT (42.9 HU), ultimately leading to a revised diagnosis of non-adenoma (Fig. 5).
Inter-reader agreement
Inter-reader agreement regarding the lesion attenuation was 0.925 (95% CI, 0.900-0.944), 0.943 (95% CI 0.924-0.957), and 0.949 (95% CI, 0.931-0.962) for UCT, 1-min enhanced CT, and 15-min enhanced CT, respectively. Inter-reader agreement regarding the percentage washout value was 0.861 (95% CI, 0.817-0.896) and 0.929 (95 CI, 0.905-0.947) for APW and RPW, respectively. The weighted kappa for cystic degeneration was 0.836 (95% CI, 0.685-0.988). The weighted kappa was 0.825 (95% CI, 0.672-0.978), 0.864 (95% CI, 0.711-1.016), and 0.866 (95% CI, 0.713-1.018) for Criterion I, Criterion I or II, and Criterion I, II, or III, respectively.
Discussion
The adrenal percentage washout is crucial to narrowing the differential diagnosis of an adrenal mass. However, it should be noted that 30% of non-adenomas can be falsely diag- nosed as adenomas by washout measurements alone [4-6,
A
B
30
29.7
25
25
False-negative rate(%)
25.2
20
21.3
False-positive rate(%)
20
15
15
10
10
9
9
5
7.2
5
0
0
☒ Criterion I
☐ Criterion I or II
☐ Criterion I or II or III
☒ Criterion I
☐ Criterion I or II
☐ Criterion I or II or III
Criteria
Criteria
A
B
C
30]. Our finding of a high false-negative rate when using criterion I to diagnose non-adenomas is consistent with a previous meta-analysis that documented a false-positive rate of 33% when employing adrenal washout criteria to diagnose adenomas [3]. This indicates that even if adrenal masses exhibit an adenoma washout pattern, this group likely has a non-negligible proportion of non-adenomas [30]. This limi- tation implies the need for additional diagnostic criteria to accurately identify non-adenomas.
focal area of low attenuation on 1-min enhanced CT (arrowheads) and demonstrated gradual enhancement on 15-min enhanced CT. Although not meeting criteria I or II for pheochromocytoma, attrib- uted to an APW (>60%) and the absence of cystic degeneration, the lesion satisfied criterion III based on its attenuation on unenhanced CT (>40 HU)
In this study, the false-negative rate significantly reduced from 30 to 7% with an increase in AUC from 0.806 to 0.858 (Table 3, Fig. 4A) when the mean attenuation criteria were added. This suggests that the modified criteria could identify an additional 23% of non-adenomas that would otherwise be misdiagnosed as adenomas using conventional criteria alone. Regarding the mean attenuation, previous studies have noted that pheochromocytomas demonstrate higher attenuation in UCT and 1-min enhanced CT compared to
adenomas [31, 32]. However, no optimal cut-off values have been proposed so far because of the unacceptable overlap between the two groups. The threshold values of 40 HU, 160 HU, and 70 HU for UCT, 1-min enhanced CT, and 15-min enhanced CT proposed by Kang et al. [7] were set at 95% specificity, respectively, to diagnose more pheochromocy- tomas. In our study, using this threshold, the specificities were similar to the previous study (92.9%, 94.8%, and 94.8% for UCT, 1-min enhanced CT, and 15-min enhanced CT, respectively). A reduction in false negatives is a predicted result that satisfies the objective of the original study. Of course, the number of patients who were misdiagnosed with non-adenomas and had to undergo unnecessary biochemical tests increased from 9.0% to 21.3% (Fig. 4B). However, the clinical implications of missing non-adenomas (e.g., pheo- chromocytomas, adrenal cortical carcinomas, or metastasis) can be significant, and the increase in false positives (12%) was less than the decrease in false negatives (23%). For the general use of modified criteria in screening adrenal masses, cost-effectiveness should be addressed in future studies.
Cystic changes are well-known findings in pheochromo- cytomas and adrenal cortical carcinomas [10, 33]. However, in the present study, adding the criterion of cystic degenera- tion to the conventional criteria did not result in a significant difference in diagnostic performance, with a slightly lower false-negative rate (Table 3 and Fig. 4). This was probably because the ROI was drawn as large as possible to cover the entire lesion, including cystic changes. APW and RPW are expected to be low and not significantly different from conventional criteria. The previous study demonstrated that placing a small ROI to fit the highest attenuated area on the early enhancement scan resulted in higher values for APW and RPW compared to those obtained by covering more than half of the lesion [26].
It is unlikely that the modified criteria proposed by Kang et al. [7] can distinguish pheochromocytomas from other non-adenomas. The mean attenuation and washout val- ues overlapped considerably among pheochromocytomas, metastases, and adrenocortical carcinomas. These results compare favorably with those published previously [32]. Furthermore, in a subgroup analysis of the non-adenoma group, the AUC of the modified criteria was low (0.477) to distinguish pheochromocytomas from other non-adenomas. Non-adenomas misdiagnosed as adenomas using conven- tional criteria alone included metastasis (n=3), adreno- cortical carcinoma (n=2), as well as pheochromocytomas (n=28). Conversely, these findings also suggest that the modified criteria can be applied more comprehensively to non-adenomas and are not limited to pheochromocytomas. Of note, neurogenic tumors were characterized by progres- sive enhancement, distinct from other non-adenomas (Sup- plementary Fig. S2). Although the available evidence is
limited to a few observational studies, adrenal schwanno- mas, and ganglioneuromas have been documented to show progressive or persistent enhancement on delayed CT [34, 35]. In our study, all neurogenic tumors (n=21) displayed progressive or persistent enhancement during the portal venous and delayed phases.
The mean attenuation value on a 1-min enhanced scan for each subtype differed from those reported in previous studies [30-32]. Szolar et al. [32] observed that the mean attenua- tion values for all non-adenomas (i.e., adrenocortical carci- nomas, pheochromocytomas, and metastases) were signifi- cantly higher than those of adenomas in a 1-min enhanced scan. In contrast, in our study, the mean attenuation value of adenomas was higher than that of adrenocortical carci- nomas (p=0.008) and metastases (p=0.275). There was no significant difference in attenuation between pheochro- mocytomas and adenomas (p>0.999). This was probably because lipid-rich adenomas were excluded from the study population. When lipid-rich adenomas were included, the median attenuation of adenomas on 1-min images was 87.3 HU (IQR, 70.3-107.4). This result is comparable to that reported by Schieda et al. [30]. Kodama et al. reported a significant association between pathologic fat content and attenuation measurements on both unenhanced and contrast- enhanced CT in the liver [36].
In our study, the sensitivity of plasma free metanephrines for diagnosing pheochromocytomas was 98%, similar to the results of a previous study [29]. Pheochromocytomas can be easily diagnosed through biochemical testing; however, up to 60% of pheochromocytomas can be incidentally detected during CT examinations performed for various reasons [37]. Therefore, suspicion of pheochromocytoma is impor- tant based on adrenal CT findings regardless of biochemical test results. This study had several limitations. Our study has limitations inherent to its retrospective nature. The number of non-adenomas, except pheochromocytomas, might have been too small for statistical analysis. One potential limita- tion of our study is the presence of selection bias, as only patients who underwent surgery were included. In clinical investigations, the proportion of adenoma to non-adenoma is observed at a ratio of 80:20. Conversely, in surgical studies, this ratio converges to a near equivalence, mirroring our own research [38]. Consequently, there exists a propensity for an overestimation in detecting non-adenomas. Another limita- tion of this study is that Reader 2 was unable to analyze the images of Center II. However, the agreement between the two readers regarding the imaging analysis from Center I was excellent. Fifth, data were obtained from different manufacturers’ CT scanners. This can lead to difficulties in maintaining data consistency. For heterogeneous masses, when selecting an ROI, it included a portion of the cystic or necrotic area to cover two thirds of the lesion. This could
decrease the mean attenuation values of the lesions. The optimal location and size of an ROI for CT attenuation meas- urement, particularly for heterogeneously enhancing masses, remain unclear. Therefore, further investigation into the opti- mal ROI location and size for detecting more non-adenomas may be necessary. In summary, the use of modified criteria has the potential to detect more non-adenomas mimicking adenomas compared to conventional criteria.
Supplementary Information The online version contains supplemen- tary material available at https://doi.org/10.1007/s00261-024-04206-x.
Acknowledgements We would like to express our sincere gratitude to Dr. Sung Yoon Park for his valuable insights and guidance throughout this research.
Funding No funds, grants, or other support was received.
Declarations
Conflict of interest The authors have no relevant financial or non-fi- nancial interests to disclose.
Ethical approval Ethical approval was waived by the local Ethics Com- mittee of Samsung Medical center and Kangbuk Samsung Hospital in view of the retrospective nature of the study and all the procedures being performed were part of the routine care.
References
1. Ctvrtlik F, Koranda P, Tichy T (2014) Adrenal disease: a clinical update and overview of imaging. A review. Biomed Pap Med Fac Univ Palacky Olomouc Czech Repub 158:23-34. https://doi. org/https://doi.org/10.5507/bp.2014.010
2. Johnson PT, Horton KM, Fishman EK (2009) Adrenal mass imag- ing with multidetector CT: pathologic conditions, pearls, and pit- falls. Radiographics 29:1333-1351. https://doi.org/https://doi.org/ 10.1148/rg.295095027
3. Woo S, Suh CH, Kim SY, Cho JY, Kim SH (2018) Pheochro- mocytoma as a frequent false-positive in adrenal washout CT: A systematic review and meta-analysis. Eur Radiol 28:1027-1036. https://doi.org/https://doi.org/10.1007/s00330-017-5076-5
4. 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. https://doi.org/https://doi.org/10.2214/AJR.12.9620
5. Park BK, Kim B, Ko K, Jeong SY, Kwon GY (2006) Adrenal masses falsely diagnosed as adenomas on unenhanced and delayed contrast-enhanced computed tomography: pathological correla- tion. Eur Radiol 16:642-647. https://doi.org/https://doi.org/10. 1007/s00330-005-0017-0
6. Park BK, Kim CK, Kwon GY, Kim JH (2007) Re-evaluation of pheochromocytomas on delayed contrast-enhanced CT: washout enhancement and other imaging features. Eur Radiol 17:2804- 2809. https://doi.org/https://doi.org/10.1007/s00330-007-0695-x
7. Kang S, Oh YL, Park SY (2021) Distinguishing pheochromocy- toma from adrenal adenoma by using modified computed tomog- raphy criteria. Abdom Radiol (NY) 46:1082-1090. https://doi. org/https://doi.org/10.1007/s00261-020-02764-4
8. Lanoix J, Djelouah M, Chocardelle L, Deguelte S, Delemer B, Dohan A, Soyer P, Barat M, Hoeffel C (2022) Differentiation
between heterogeneous adrenal adenoma and non-adenoma adre- nal lesion with CT and MRI. Abdom Radiol (NY) 47:1098-1111. https://doi.org/https://doi.org/10.1007/s00261-022-03409-4
9. 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. https://doi.org/https://doi.org/10.2214/ ajr. 171.1.9648789
10. Motta-Ramirez GA, Remer EM, Herts BR, Gill IS, Hamrahian AH (2005) Comparison of CT findings in symptomatic and inci- dentally discovered pheochromocytomas. AJR Am J Roentgenol 185:684-688. https://doi.org/https://doi.org/10.2214/ajr.185.3. 01850684
11. Fassnacht M, Tsagarakis S, Terzolo M, Tabarin A, Sahdev A, Newell-Price J, Pelsma I, Marina L, Lorenz K, Bancos I (2023) European Society of Endocrinology clinical practice guidelines on the management of adrenal incidentalomas, in collaboration with the European Network for the Study of Adrenal Tumors. European Journal of Endocrinology 189:G1-G42
12. Vilar L, da Conceição Freitas M, Canadas V, Albuquerque JL, Botelho CA, Egito CS, Arruda MJ, Moura L, Coelho CE, Casu- lari LA (2008) Adrenal incidentalomas: diagnostic evaluation and long-term follow-up. Endocrine practice 14:269-278
13. Hong AR, Kim JH, Park KS, Kim KY, Lee JH, Kong SH, Lee SY, Shin CS, Kim SW, Kim SY (2017) Optimal follow-up strat- egies for adrenal incidentalomas: reappraisal of the 2016 ESE- ENSAT guidelines in real clinical practice. European journal of endocrinology 177:475-483
14. Marty M, Gaye D, Perez P, Auder C, Nunes M, Ferriere A, Hais- saguerre M, Tabarin A (2018) Diagnostic accuracy of computed tomography to identify adenomas among adrenal incidentalo- mas in an endocrinological population. European Journal of Endocrinology 178:439-446
15. Bancos I, Taylor AE, Chortis V, Sitch AJ, Lang K, Prete A, Gilligan LC, Biehl M, Deeks JJ, Arlt W (2020) Urine metabo- lomic phenotyping for detection of adrenocortical carcinoma: still a long way to go-Authors’ reply. The Lancet Diabetes & Endocrinology 8:877-878
16. Duh Q-Y (2020) Understanding the epidemiology of adrenal tumours. The Lancet Diabetes & Endocrinology 8:871-873
17. Schloetelburg W, Ebert I, Petritsch B, Weng AM, Dischinger U, Kircher S, Buck AK, Bley TA, Deutschbein T, Fassnacht M (2022) Adrenal wash-out CT: moderate diagnostic value in distinguishing benign from malignant adrenal masses. European Journal of Endocrinology 186:183-193
18. Park SY, Park BK, Park JJ, Kim CK (2016) Differentiation of adrenal hyperplasia from adenoma by use of CT densitometry and percentage washout. American Journal of Roentgenology 206:106-112
19. Doppman J, Gill Jr J, Miller D, Chang R, Gupta R, Friedman T, Choyke P, Feuerstein I, Dwyer A, Jicha D (1992) Distinc- tion between hyperaldosteronism due to bilateral hyperplasia and unilateral aldosteronoma: reliability of CT. Radiology 184:677-682
20. Lattin Jr GE, Sturgill ED, Tujo CA, Marko J, Sanchez-Mal- donado KW, Craig WD, Lack EE (2014) From the radiologic pathology archives: Adrenal tumors and tumor-like conditions in the adult: radiologic-pathologic correlation. Radiographics 34:805-829
21. Schieda N, Siegelman ES (2017) Update on CT and MRI of Adre- nal Nodules. AJR Am J Roentgenol 208:1206-1217. https://doi. org/https://doi.org/10.2214/AJR.16.17758
22. Korobkin M, Brodeur FJ, Yutzy GG, Francis IR, Quint LE, Dun- nick NR, Kazerooni EA (1996) Differentiation of adrenal adeno- mas from nonadenomas using CT attenuation values. AJR. Ameri- can journal of roentgenology 166:531-536
23. Lee MJ, Hahn PF, Papanicolaou N, Egglin TK, Saini S, Mueller PR, Simeone JF (1991) Benign and malignant adrenal masses: CT distinction with attenuation coefficients, size, and observer analysis. Radiology 179:415-418
24. Kim MJ, Kang KA, Kim CK, Park SY (2023) Inter-method agree- ment between wash-in and wash-out computed tomography for characterizing hyperattenuating adrenal lesions as adenomas or non-adenomas. European Radiology 33:2218-2226
25. Kang S, Oh YL, Park SY (2021) Distinguishing pheochromocy- toma from adrenal adenoma by using modified computed tomog- raphy criteria. Abdominal Radiology 46:1082-1090
26. Park SY, Park BK, Park JJ, Kim CK (2015) CT sensitivities for large (≥ 3 cm) adrenal adenoma and cortical carcinoma. Abdomi- nal imaging 40:310-317
27. Sun MR, Ngo L, Genega EM, Atkins MB, Finn ME, Rofsky NM, Pedrosa I (2009) Renal cell carcinoma: dynamic contrast- enhanced MR imaging for differentiation of tumor subtypes-cor- relation with pathologic findings. Radiology 250:793-802
28. Jakobsson U, Westergren A (2005) Statistical methods for assess- ing agreement for ordinal data. Scand J Caring Sci 19:427-431. https://doi.org/https://doi.org/10.1111/j.1471-6712.2005.00368.x
29. Lenders JW, Pacak K, Walther MM, Linehan WM, Mannelli M, Friberg P, Keiser HR, Goldstein DS, Eisenhofer G (2002) Bio- chemical diagnosis of pheochromocytoma: which test is best? Jama 287:1427-1434
30. Schieda N, Alrashed A, Flood TA, Samji K, Shabana W, McInnes MD (2016) Comparison of Quantitative MRI and CT Washout Analysis for Differentiation of Adrenal Pheochromocytoma From Adrenal Adenoma. AJR Am J Roentgenol 206:1141-1148. https:// doi.org/https://doi.org/10.2214/AJR.15.15318
31. Mohammed MF, ElBanna KY, Ferguson D, Harris A, Khosa F (2018) Pheochromocytomas Versus Adenoma: Role of Venous Phase CT Enhancement. AJR Am J Roentgenol 210:1073-1078. https://doi.org/https://doi.org/10.2214/AJR.17.18472
32. Szolar DH, Korobkin M, Reittner P, Berghold A, Bauernhofer T, Trummer H, Schoellnast H, Preidler KW, Samonigg H (2005) Adrenocortical carcinomas and adrenal pheochromocytomas:
mass and enhancement loss evaluation at delayed contrast- enhanced CT. Radiology 234:479-485. https://doi.org/https://doi. org/10.1148/radiol.2342031876
33. Blake MA, Kalra MK, Maher MM, Sahani DV, Sweeney AT, Mueller PR, Hahn PF, Boland GW (2004) Pheochromocytoma: an imaging chameleon. Radiographics 24 Suppl 1:S87-99. https:// doi.org/https://doi.org/10.1148/rg.24si045506
34. Park BK, Kim CK, Kim B, Kwon GY (2007) Adrenal tumors with late enhancement on CT and MRI. Abdom Imaging 32:515-518. https://doi.org/https://doi.org/10.1007/s00261-006-9156-2
35. Zhang YM, Lei PF, Chen MN, Lv XF, Ling YH, Cai PQ, Gao JM (2016) CT findings of adrenal schwannoma. Clin Radiol 71:464- 470. https://doi.org/https://doi.org/10.1016/j.crad.2016.01.010
36. Kodama Y, Ng CS, Wu TT, Ayers GD, Curley SA, Abdalla EK, Vauthey JN, Charnsangavej C (2007) Comparison of CT methods for determining the fat content of the liver. American Journal of Roentgenology 188:1307-1312
37. Motta-Ramirez GA, Remer EM, Herts BR, Gill IS, Hamrahian AH (2005) Comparison of CT findings in symptomatic and inciden- tally discovered pheochromocytomas. AMERICAN JOURNAL OF ROENTGENOLOGY-NEW SERIES- 185:684
38. Terzolo M, Stigliano A, Chiodini I, Loli P, Furlani L, Arnaldi G, Reimondo G, Pia A, Toscano V, Zini M (2011) AME position statement on adrenal incidentaloma. European journal of endo- crinology 164:851-870
Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.