CrossMark
The computed tomography adrenal wash-out analysis properly classifies cortisol secreting adrenocortical adenomas
Anne-Laure Humbert1 . Guillaume Lecoanet2 . Sophie Moog1 . Fehd Bouderraoui3 . Laurent Bresler4 . Jean-Michel Vignaud5 . Elodie Chevalier3 . Laurent Brunaud4 . Marc Klein1 . Thomas Cuny1
Received: 4 October 2017 / Accepted: 3 January 2018 @ Springer Science+Business Media, LLC, part of Springer Nature 2018
Abstract
Purpose Adrenocortical lesions are characterized through imaging, hormonal and histopathological analysis. Our aim was to compare the radiological features of adrenocortical lesions with their cortisol-secreting status and histopathological Weiss score.
Methods Seventy five patients operated between 2004 and 2016 in the University Hospital of Nancy for either adreno- cortical carcinomas (ACC) or adrenocortical adenomas (ACA) were enrolled in this study. We collected cortisol parameters, Computed Tomography (CT) scans (unenhanced density, wash-out (WO) analysis) and 18F-Fluorodeoxyglucose positron emission tomography/computed tomography (18F-FDG PET/CT) datas. The histopathological Weiss score ultimately dif- ferentiates ACA (score ≤2) from ACC (score ≥3). One-way ANOVA, Fisher’s exact and unpaired t tests were used for statistical analysis with significancy reached at p < 0.05.
Results There were 23 ACC and 52 ACA with 40 patients (53%) who had an autonomous secretion of cortisol. On CT scan, ACC were larger compared to ACA (108 vs. 37 mm, p <0.0001). A roughly similar proportion of cortisol-secreting (22/25) and non-secreting (15/19) ACA were atypical (i.e., unenhanced density value ≥ 10 Hounsfield Units [HU]), however 85% of cortisol-secreting vs. 40% of non-secreting ACA were classified as benigns by the relative WO analysis (p=0.08). Likewise, there was a trend for a higher 18F-FDG uptake in cortisol-secreting ACA compared to non-secreting ACA (p= 0.053).
Conclusions The relative adrenal WO analysis consolidates the benign nature of an ACA, especially in case of cortisol oversecretion, a condition known to compromise the diagnostic accuracy of the 10 HU unenhanced CT attenuation threshold.
Keywords Incidentaloma . Adrenal adenoma . Cushing’s syndrome . 18F-FDG/PET . Weiss score
Electronic supplementary material The online version of this article (https://doi.org/10.1007/s12020-018-1522-7) contains supplementary material, which is available to authorized users.
☒ Thomas Cuny
1 Department of Endocrinology, University Hospital of Nancy, Nancy, France
2 Department of Radiology, University Hospital of Nancy, Nancy, France
3 Department of Nuclear Medicine, University Hospital of Nancy, Nancy, France
4 Department of Endocrine and General Surgery, University Hospital of Nancy, Nancy, France
5 Department of Anatomopathology, University Hospital of Nancy, Nancy, France
Introduction
Diseases consecutive to adrenal disturbances are frequent in clinical practice and abnormalities can affect the adrenal gland(s) in nearly 10% of autopsy series [1, 2]. Incidenta- lomas, defined as adrenal masses greater than 1 cm in dia- meter detected on imaging performed for other indications, are currently found in up to 4-10% of computed tomo- graphy (CT) examinations [2, 3]. In most cases, adrenal incidentalomas are nonfunctioning, benign adrenocortical adenomas (ACA), but may also represent conditions requiring therapeutic intervention (e.g., adrenocortical car- cinoma [ACC], phaeochromocytoma, hormone-producing adenoma or metastasis). In particular, the incidence of metastatic adrenal lesions increases to 71% in patients with adrenal incidentalomas and history of malignancy [4].
According to the recent European Society of Endocrinology (ESE) guidelines, an initial diagnostic evaluation is man- datory to determine whether the adrenal lesion is benign or malignant and/or hormonally active [5].
The nature of the adrenal lesion is already suggested by its morphological aspects: a regular border, a maximum size diameter less than 4 cm, and/or an homogeneous tissue content (without history of known malignancy) are indicative of benignity [4, 6, 7]. However, none of them are specific enough to accurately differentiate malignant from benign tumors and the identification of small malignant ACC remains really challenging. In general, ACA which are of lipid-rich compo- sition have an attenuation value density ≤10 hounsfield units (HU) on noncontrast (or “unenhanced”) CT scan. Although this cut-off is routinely used to predict the benign nature of an adrenal mass, a third of benign ACA present with a poor content of lipids and may have an attenuation value greater than 10 HU, thus overlapping in density with malignant lesions and phaeochromocytomas. The same applies to a majority (80%) of cortisol-secreting ACA which had an unenhanced density > 10 HU [8]. In this situation, the quan- tification (in %) of the adrenal contrast enhancement washout (WO) after contrast medium injection helps to differentiate benign from malignant tumors as ACA generally take up intravenous CT contrast rapidly, but also have a rapid loss of contrast [9]. It is therefore considered that a relative WO≥ 40% and/or an absolute WO ≥60% are in favor of benignity [10]. In cases of atypical lesions on CT scan, a positron emission tomography/computed tomography using 18F- Fluorodeoxyglucose (18F-FDG PET/CT) may provide addi- tional information as an adrenal to liver maximum standar- dized uptake value (maxSUV) ratio below 1.52 (directly calculated from the 18F-FDG uptake) suggests benignity of the adrenal lesion [11, 12]. Nonetheless, an increase of 18F-FDG uptake in the adrenal gland(s) can also be observed in various conditions other than ACC, such as its diagnosis performance can be undermined in non cancer situations [13]. Given these areas of uncertainty that may occur during CT scan and/or 18F- FDG PET/CT procedures, it seemed interesting i/ to analyze to what extent the existence of a secretion of cortisol could influence the imaging features of adrenocortical lesions and ii/ to compare the results of the CT scan and/or 18F-FDG PET/CT exams performed before surgery, to the final 9-criteria Weiss score system, used to classify an adrenocortical lesion tumor as benign (Weiss score ≤2) or malignant (Weiss score ≥3)[14].
Therefore, our aim was to compare the radiological features of 75 operated adrenocortical lesions (CT scan and 18F-FDG PET/CT) with their cortisol-secreting status and histopathological Weiss score.
Subjects and methods
Patients
The present study was approved by the Ethics Committee of the University Hospital of Nancy (Nancy, France) and informed consent was obtained from each patient. A total of 75 patients enrolled in this study underwent resection of either an ACA or an ACC between January 2004 and June 2016. The endocrine and radiological characteristics of each tumor were documented before surgery. A complete screening for autonomous cortisol oversecretion was per- formed in every patient and included i) 24h-urinary free cortisol (24h-UFC) ii) measurement of plasma cortisol cir- cadian rhythm (8am, 12 pm, 4 pm, 8 pm, 12 am) and iii) 1- mg overnight dexamethasone suppression test (1mg-DST). We considered patients as having non-secreting adrenal lesions if simultaneously, the 24h-UFC was normal, morning plasma cortisol was below 1.8 µg/dL (50 nmol/l) after 1mg-DST and if the cortisol circadian rhythm was normal. On the other hand, patients had an overt autono- mous secretion of cortisol (ASC) if 24 h-UFC was increased, morning cortisol after 1mg-DST ≥5.0µg/dL (138 nmol/L) and if they presented with an abolished cir- cadian rhythm of cortisol secretion. According to our cri- teria, seven patients had an intermediate functional status with a morning cortisol after 1mg-DST between 1.8-5.0 µg/ dL (50-138 nmol/L) with normal or slightly increase of 24h-UFC. However, their respective circadian rhythm of cortisol was normal. Theses cases were considered as hav- ing a possible ASC and were excluded from the statistical analysis. Primary hyperaldosteronism and phaeochromo- cytoma were ruled out by plasma measurement of renine/ aldosterone and urinary metanephrines levels in all patients of the study. All patients underwent adrenal surgery because of relevant hormone excess and/or suspicion of malignant tumor (i.e., indeterminate mass with a size ≥4 cm and/or rapidly growing tumor).
Computed Tomography images analysis
Features analyzed on routine Computed Tomography (CT) scanning included the maximum diameter size, shape and unenhanced attenuation value of the adrenal mass. The patients were examined with a 64-detector device (Light- speed VCT 64, General Electric Healthcare, Milwaukee, USA) and an unenhanced CT scan was performed on the kidney and adrenal glands with the following parameters:
| Collimation | Thickness | Gap | kV | mA min/max | Noise index | Rotation time | Pitch | Filter |
|---|---|---|---|---|---|---|---|---|
| 64 × 0625 | 1.25 | 0.9 | 120 | 200/700 | 24 | 0.6 | 1.375 | Std |
To evaluate the unenhanced density value of the lesion as well as its density value during the enhanced phases of the CT scan (i.e., portal enhanced and delayed enhanced pha- ses, respectively), the radiological examiner set up a region of interest (ROI) which covered two third of the adrenal mass. The unenhanced density of the adrenal lesion was firstly calculated without further analysis required if the density value was ≤ 10HU. Otherwise, if the unenhanced density exceeded 10HU, the portal enhanced phase density (or early density) and the delayed density (or late density) of the adrenal mass were respectively measured 60 s and 10 min after administration of iodinated contrast media. The latter consisted of 1.5 mL/kg iodinated contrast media (Ioméron 400, Iomeprol, Bracco Imaging France or Xenetix 350, Iobitridol, Guerbet, France Or Omnipaque, Iohexol, General Healthcare, Milwaukee, USA) at a rate of 4-5 mL/ min. Note that for all phases of the CT scan (unenhanced and enhanced), we used the same calibration parameters as presented in the table above.
The adrenal wash-out (WO) analysis was performed exclusively in homogeneous lesions and both relative and absolute WO were calculated by these mathematical for- mulas: relative WO=(early density-late density)/(early density) x 100%; absolute WO = (early density-late den- sity)/(early density-unenhanced density) x 100%. When no adrenal WO was observed by the radiologist, the results were expressed as 0%. Adrenal lesions were preoperatively considered as benigns if the relative WO and absolute WO were ≥ 40% and ≥60%, respectively. In our series, all the ACC were heterogeneous and therefore did not undergo WO analysis.
Finally, in a subgroup of adrenal lesions, the adrenal wash-out analysis was performed in spite of an unenhanced density < 10HU.
18F-FDG PET/CT images analysis
18F-FDG PET/CT was performed on a hybrid PET camera which included a six-detector CT (Biograph 6 True Point, SIEMENS, Knoxville, Tennessee, USA). An activity of 4 MBq/kg of 18F-FDG was injected intravenously and a CT recording, used for attenuation correction, was initiated 60 min later. The latter was followed by a 3D FDG-PET recording, which involved four to six bed positions lasting 210 s each. FDG-PET images were reconstructed by the OSEM method. A 6-hour fast and a glycemia less than 200 mg/dl were necessary before the 18F-FDG PET/CT was carried out. The imaging parameters were constant during the period of the study. The calculation of the adrenal maxSUV to liver maxSUV ratio was performed as pre- viously described [11].
Histological analysis
The nine histological parameters were evaluated for each tumor sample, according to the Weiss score system: high nuclear grades, mitotic rate > 5/50 high power fields (HPF), atypical mitotic figures, clear tumor cell cytoplasm (less than 25% tumor cells), diffuse architecture (greater than 33% of tumor), necrosis, venous invasion, sinusoidal invasion, and capsular invasion. Benigns ACA were defined by a Weiss score ≤2 and ACC by a Weiss score ≥3.
Statistical analysis
For statistical analysis, GraphPad Prism® version 6.04 (Graph-Pad Software, San Diego, Calif., USA) was used. Comparative statistical evaluations between ACA vs. ACC or secreting vs. non-secreting lesions were accomplished with unpaired t tests or one-way ANOVA followed by Tukey’s tests for multiple post hoc comparisons. Data are reported as means + SD. A P value <0.05 was considered to be significant for all tests. To measure the strength of association between the maximal diameter or the maxSUV (lesion/liver) of the lesions and their respective Weiss score, without specifying dependency, Spearman’s rank-order correlations were run. The differences were considered to be statistically significant at P < 0.05. For qualitative data (Tables 1 and 2), comparisons between groups were based on the Fisher’s exact test.
Results
Patients characteristics
There were 44 women and 31 men and the mean age at diagnosis was 52.8 + 12.6 years, roughly the same between the ACC and the ACA group (50.6 + 13.7 years vs. 53.8 ± 12 years, respectively, NS, Table 1). A total of 52 patients had an ACA (68%) and 23 an ACC (32%). A history of extra adrenal cancer was found in 5 patients with ACA.
Incidentaloma was the mode of revelation in 34 ACA and 9 ACC (p=0.04). Cushing symptoms led to the diagnosis of 25 secreting adrenal lesions (13 ACA, 12 ACC). Overall, 28 ACA and 12 ACC were biologically confirmed with an autonomous secretion of cortisol (ASC). A total of 7 patients were considered as having a possible ASC and two of them had an ACC (Table 2). In three patients (pt. 63, 66, and 75), the secretory status was not available because not assessed pre-operatively in 2 cases and because of the existence of a 21-hydroxylase deficiency in pt 63 (Supplemental Table).
| ACA (n=52) | ACC (n=23) | P value | |
|---|---|---|---|
| Age at diagnosis (years) | 53.8 ±12 | 50.6±13.7 | NS |
| Female | 29 (55%) | 15 (65%) | NS |
| Mode of revelation: | |||
| Incidentaloma | 34 (65%) | 9 (39%) | p =0.04 |
| Cushing symptoms | 13 (25%) | 12 (46%) | p =0.03 |
| Other tumor | 5 (10%) | 0 | NS |
| Tumoral symptoms | 0 | 2 (8%) | NS |
| Other | 0 | 1 (4%) | NS |
| ASC | 28 (53%) | 12 (50%) | NS |
| Possible ASC | 5 (9%) | 2 (8%) | NS |
| Mean maximal diameter on CT scan (mm) | 37 ±10.5 | 108.6±39.6* | p <0.0001 |
ASC is considered if 24h-UFC was increased, morning cortisol after 1mg-DST was ≥ 5.0 µg/dL (138 nmol/L) and circadian rhythm of cortisol secretion was abolished. Subclinical ASC is considered if morning cortisol after 1mg-DST between was between 1.8-5.0 ng/dL (50-138 nmol/L) with normal or slightly increase of 24h-UFC and normal circadian rhythm of cortisol secretion
ACA adrenocortical adenoma, ACC adrenocortical carcinoma, ASC autonomous secretion of cortisol *p < 0.05
| ASC | Non-ASC | P value | |
|---|---|---|---|
| ACA | 28/52 (54%) | 19/52 (36%) | |
| Unenhanced CT attenuation | 19.5 ±10 (n | 22.4 ± 17 (n | p=0.2 |
| (HU) | =25) | =19) | |
| Unenhanced CT > 10HU | 22/25 (88%) | 15/19 (78%) | NS |
| Unenhanced CT > 10HU + | 10/15 (66%) | 1/6 (16%) | NS |
| Absolute washout ≥ 60% | |||
| Unenhanced CT > 10HU + | 13/15 (86%) | 2/7 (28%) | NS |
| Relative washout ≥ 40% | |||
| Adrenal to liver SUVmax | 1.08 ±0.34 | 0.86 ±0.16 | p= |
| ratio | (n=18) | (n=11) | 0.053 |
| ACC | 12/24 (50%) | 10/24 (42%) | |
| Unenhanced CT attenuation | 35.2±2.7 | 35.6±5.1 | NS |
| (HU) | (n=12) | (n=8) | |
| Unenhanced CT > 10HU | 12/12(100%) | 8/8 (100%) | NS |
| Adrenal to liver SUVmax | 2.83±0.9 | 5.9 ±3.8 (n | NS |
| ratio | (n=10) | =6) |
ACC adrenocortical carcinoma, ASC autonomous secretion of cortisol
CT scans data
In ACC (n =23) the mean maximal diameter was sig- nificantly larger than in ACA (n = 52) (108.6 ± 39.6 mm vs. 37 ± 10.5 mm, respectively, p <0.0001). None of the ACA had a maximal diameter above 60 mm. On the opposite, two ACC had a maximal diameter less than 60 mm with respective Weiss scores of 3 (pt. 53) and 4 (pt. 54). In ACC, the maximal size of the tumor was not correlated with the severity of the Weiss score, thus suggesting that the
maximal diameter of the tumor cannot be considered as an independent predictive factor of the tumor aggressiveness (72 = 0.07, P=0.72, Fig. 1a). Unenhanced attenuation values were available in 48 ACA and 37 of them (77%) were greater than 10 HU (Table 2). Among the latter, 22 (60%) were associated with an overt ASC for which both the relative WO and the absolute WO were available in 15 cases. In cortisol-secreting ACA, 13/15 (86%) lesions had a relative WO ≥40%, while the absolute WO was ≥60% in 10/15 (66%) lesions (Table 2). Overall, a relative WO ≥
40% was reached in 15/22 ACA (68%, Fig. 1b) and relative WO and absolute WO results were concordant for benign or atypical lesions in 22/25 cases. All ACCs (n =22) had an unenhanced attenuation value > 10HU. For two of them, the unenhanced density was not available and/or measurable. In ACA as well as in ACC subgroups, the mean unenhanced attenuation value as well as the mean maximal diameter of the lesion did not differ significantly between secreting and non-secreting lesions (Table 2).
Overall, the sensitivity and specificity of the 10 HU unenhanced computed tomography attenuation threshold for the screening of malignant lesions were respectively 100% and 20% with a negative predictive value of 100%. The sensitivity of relative and absolute WO in cortisol- secreting ACA were 87 and 71%, respectively.
18F-FDG PET/CT datas
Results of 18F-FDG PET/CT were available for 30 ACA and 18 ACC. The mean adrenal to liver maxSUV ratio was higher in ACC compared to ACA (3.93 ±2.82 vs. 0.98 ± 0.3, respectively, p <0.0001, Fig. 2a) with a positive cor- relation between the adrenal to liver maxSUV ratio of the lesion and its Weiss score (72 = 0.27 p <0.0001, Fig. 2b). However, this positive correlation no longer exist in the specific subgroup of ACC (Weiss ≥3) (Fig. 2c) or ACA (12 =0.07, p=0.16, data not shown). All the ACC in our series had an adrenal to liver maxSUV ratio ≥ 2 (Fig. 2b).
A trend for a highest 18F-FDG uptake was observed in cortisol-secreting ACA in comparison to the non-secreting ones (1.08 ±0.34 vs. 0.86 ±0.13, respectively, p=0.053) (Fig. 2d). This situation was observed in one non-diabetic patient (pt 21) diagnosed with an overt ASC secondary to a cortisol-secreting ACA (Weiss score = 0). The unenhanced attenuation value of the lesion was 20 HU (Fig. 3a) with a relative WO of 79%, in favor of benignity (Fig. 3b, c). The adrenal to liver maxSUV ratio of the lesion was 1.71 (Fig. 3d).
Discussion
Over the past years, the unenhanced density calculated on CT scan progressively emerged as an inescapable diagnostic step of an adrenal lesion. While the 10 HU threshold is routinely used to suggest benignity (at least to rule out malignancy), our study shows that 78 to 88% of ACA present with an atypical aspect (i.e., unenhanced density > 10 HU). Previously, it has been reported an inverse linear relationship between the percentage of lipid-rich cortical cells in the adrenal adenomas and the unenhanced CT attenuation number [15]. Accordingly, nearly 30% of lipid- poor ACA are encountered in clinical practice with an
r2 = 0.07
A
200-
p =0.72
☐
Maximal Diameter (mm)
☐
150
☐
☐
☐
100
☒
☐
☐
0
50
☐
☐
0-
3
4
5
6
7
8
9
Weiss Score
Unenhanced Density Value (HU)
B
Relative Wash Out (%)
80
80
70-
3
60
8
Benign
60
50
0
8
40
8
40
30
9
Undetermined
€
20
Atypical
20
·
0
10
0
0
5
0
unenhanced density > 10 HU [10, 16, 17]. In our study, we did not detail the lipid content of the operated adrenal lesions however the high proportion of atypical ACA we found on CT scan is probably explained by our cohort of patients, all surgically-treated, such as an overestimation of undetermined adrenal lesions is likely to occur. It therefore does not reflect necessarily the real proportion of lipid-poor adenomas encountered in the real life. In the subgroup of cortisol-secreting ACA, for which the surgical resection was indicated for the treatment of a Cushing’s syndrome rather than an atypical aspect, a majority of them also had an unenhanced density > 10 HU. This result is in line with the data from Chambre and coll. where they found that nearly 80% of cortisol-secreting ACA were, due to a low content of lipid-rich cells, greater than 10 HU [8]. The secretion of cortisol thus appears as an interfering metabolic condition which could lead to an overestimation of the atypical nature of an adrenal lesion. In that respect, we here show that
B
12-
4
+
A
11
14-
10
+
Max SUV (lesion/liver)
9
12
Max SUV (lesion/liver)
A
8
10
A
7
ACC
+
6
p = 0.0001
8
r2 = 0.27
5
+
6
44
4
+
3
F
+
4-
+
#
+
+
±
A
5
2
+
2
A44
DADADADA
1
Y
#
0 -
ACA
0
0
1
2
3
4
5
6
7
8
9
Weiss ≤ 2
Weiss ≥ 3
Weiss Score
Max SUV (lesion/liver)
12
D
+
10
2.5
+
p =0.4
p =0.0534
r2 = 0.04
Max SUV (lesion/liver)
8
2.0
6
+
+
3
1.5
4
+
+
2
+ + $ + ¥ ¥
1.0
0
0.5
3
4
5
6
7
8
9
0.0
Weiss Score
Non Secreting
Secreting
analysis of the relative wash-out is of significance in these cases since it successfully puts the benignity aspect of these cortisol-secreting ACA straight in more than 85% of cases. However, in our study both relative and absolute wash-out failed to classify 50% of non-cortisol-secreting ACA as benign tumors but the small number of patients (n =4 for relative WO and n = 5 for absolute WO) could explain this result. In all these patients, though, the maxSUV ratio was, when available, in favor of benign lesions.
Similarly to CT scan, the results of MRI with chemical shift imaging are based on the lipid content of the adrenal mass. However, the quantitative assessment of loss in signal intensity is still not well standardized between studies and according to the ESE guidelines, evidence bases for per- formance of MRI in the diagnosis of malignancy remains insufficient to establish strong recommendations (4, 12, 13). For these reasons and because of the small number of MRI exams in our series, we only include CT scan datas.
Besides unenhanced density, our study finds that the maximal diameter size remains of importance to distinguish
benign versus malignant lesions. All ACA had a maximum size diameter below 60 mm while only 2/24 (8.3%) ACC presented with a size <60 mm, a proportion which fits with the 5-10% rate of “small” ACC described in european national registries [18, 19]. The 60mm-threshold in dia- meter seems therefore acceptable to indicate adrenal surgery in our institution, even though there is continuing uncer- tainty for small ACC. With the increasing number of CT scan performed, the one could have expected in parallel an increasing number of small ACC diagnosed. However radiological reports showed that the mean maximal dia- meter of ACC at diagnosis remained stable over the years (around 100 mm) including in the children population where the mean maximal diameter of ACC at diagnosis frequently exceeds 60 mm [20, 21]. Moreover, the occur- rence of intratumoral hemorrhage and tissue necrosis in ACC likely suggest rapid, uncontrolled and clinically silent growth phases during the early stages of tumor development until the diagnosis is made years later. Although spec- ulative, such an hypothesis could explain why a the
A
B
C
d
Spin: 0 Tilt: - 90
majority of ACC still presents with an important maximal diameter size at initial diagnosis.
In undetermined cases, 18F-FDG/PET is a nuclear med- icine modality of interest for the characterization of adrenal masses [5]. In a previous study, an adrenal to liver maxSUV ratio above 1.45 has been proposed as a cut off value to distinguish benign from malignant adrenal tumors [11]. Recently, in a prospective study, a threshold value of 1.6 (adrenal to liver SUVmax ratio) for distinguishing ACA from ACC was proposed with a sensibility of 90% and a specificity of 82% [12]. In our study, we observed a sig- nificant higher uptake of 18F-FDG in ACC compared to ACA with a corresponding adrenal to liver max SUV ratio >2 in all malignant tumors. In ACA, we found a trend for a higher uptake of 18F-FDG in cortisol-secreting lesions compared to the non-secreting ones, especially in three patients with respective adrenal to liver max SUV ratio of 1.71, 1.91, and 1.74. The influence of cortisol secretion status over the uptake of 18F-FDG in ACA has been recently discussed [22] and in a previous study conducted by Groussin and coll., there were 5 patients with ACA and respective max SUV ratio of 1.72, 1.83, 1.92, 1.67, and 2.36 [11]. Interestingly, all of them had either a subclinical (n = 3) or an overt autonomous secretion of cortisol (n = 2). One potential explanation for this enhanced uptake of 18F-FDG concerns the overexpression of the glucose transporter GLUT3 both at the DNA and the protein levels, recently observed in a subset of cortisol-secreting ACA [22]. In this
setting, it must be specified that the promoter of the human GLUT3 gene contains three CRE (cAMP response ele- ment)-like elements, such as an overexpression of CRE- binding protein (CREB) or activation of the cAMP- dependant protein kinase (also known as PKA) lead to a significant increase of the GLUT3 expression [23]. It is known that a third of cortisol-secreting ACA carries somatic mutations in the PRKACA gene, responsible for a con- stitutive activation of the PKA signaling pathway within the tumoral cells [24, 25]. Whether this molecular abnormality leads to GLUT3 overexpression in ACA cells remains to be determined, but may be a putative explanation of the high 18F-FDG uptake observed in a subset of cortisol-secreting ACA.
Finally, our study has obvious inherent limitations due to its retrospective nature. One of them is the absence of comparative analysis between CT scan and 18F-FDG PET/ CT datas. In that respect, we included patients with a well characterization of both biological parameters and Weiss score criteria’s. We arbitrarily defined patients with secret- ing adrenal lesions those who had simultaneously three abnormalities of the cortisol secretion namely increase in 24h-UFC, morning cortisol ≥5 µg/dL (138 nmol/L) after 1mg-DST and disruption of the cortisol secretion circadian rhythm. By this mean we wanted to avoid confusion with a potential subclinical cushing syndrome, whose definition is a source of uncertainty [26]. Likewise, non-secreting patients had normal 24h-UFC together with morning
cortisol < 1.8 µg/dL (50 nmol/L) after 1mg-DST and normal cortisol circadian rhythm. Prospective studies analyzing the imaging features of patients with possible ASC compared to non-secreting patients would be of interest to further investigate the potential predictive value of imaging mod- alities in these problematic cases.
In conclusion, our study provides a singular analysis of imaging features in well-characterized adrenal lesions according to their respective functional status and histo- pathological Weiss score. Cortisol-secreting adenomas fre- quently present with an unenhanced density value > 10 UH that requires an accurate wash-out analysis to consolidate the benign nature of the lesion.
Compliance with ethical standards
Conflict of interest The authors declare that they have no conflict of interest.
References
1. J.H. Song, F.S. Chaudhry, W.W. Mayo-Smith, The incidental adrenal mass on CT: Prevalence of adrenal disease in 1,049 consecutive adrenal masses in patients with no known malig- nancy. Am. J. Roentgenol. 190, 1163-1168 (2008)
2. W.F. Young, Clinical practice. The incidentally discovered adre- nal mass. N. Engl. J. Med. 356, 601-610 (2007).
3. M. Terzolo, A. Stigliano, I. Chiodini, P. Loli, L. Furlani, G. Arnaldi, G. Reimondo, A. Pia, V. Toscano, M. Zini, G. Borretta, E. Papini, P. Garofalo, B. Allolio, B. Dupas, F. Mantero, A. Tabarin, AME position statement on adrenal incidentaloma. Eur. J. Endocrinol. 164, 851-870 (2011)
4. A. Frilling, K. Tecklenborg, F. Weber, H. Kühl, S. Müller, G. Stamatis, C. Broelsch, Importance of adrenal incidentaloma in patients with a history of malignancy. Surgery 136, 1289-1296 (2004)
5. M. Fassnacht, W. Arlt, I. Bancos, H. Dralle, J. Newell-Price, A. Sahdev, A. Tabarin, M. Terzolo, S. Tsagarakis, O.M. Dekkers, Management of adrenal incidentalomas: European Society of Endocrinology Clinical Practice Guideline in collaboration with the European Network for the Study of Adrenal Tumors. Eur. J. Endocrinol. 175, G1-G34 (2016)
6. J.H. Song, D.J. Grand, M.D. Beland, K.J. Chang, J.T. Machan, W.W. Mayo-Smith, Morphologic features of 211 adrenal masses at initial contrast-enhanced CT: can we differentiate benign from malignant lesions using imaging features alone? Ajr. Am. J. Roentgenol. 201, 1248-1253 (2013)
7. A. Sahdev, J. Willatt, I.R. Francis, R.H. Reznek, The indetermi- nate adrenal lesion. Cancer Imaging 10, 102-113 (2010)
8. C. Chambre, E. McMurray, C. Baudry, M. Lataud, L. Guignat, S. Gaujoux, N. Lahlou, J. Guibourdenche, F. Tissier, M. Sibony, B. Dousset, X. Bertagna, J. Bertherat, P. Legmann, L. Groussin, The 10 Hounsfield units unenhanced computed tomography attenua- tion threshold does not apply to cortisol secreting adrenocortical adenomas. Eur. J. Endocrinol. 173, 325-332 (2015)
9. M.A. Blake, M.K. Kalra, A.T. Sweeney, B.C. Lucey, M.M. Maher, D.V. Sahani, E.F. Halpern, P.R. Mueller, P.F. Hahn, G.W. Boland, Distinguishing Benign from Malignant Adrenal Masses: Multi-Detector Row CT Protocol with 10-Minute Delay 1. Radiology 238, 578-585 (2006)
10. C.S. Peña, G.W. Boland, P.F. Hahn, M.J. Lee, P.R. Mueller, Characterization of indeterminate (lipid-poor) adrenal masses: use of washout characteristics at contrast-enhanced CT. Radiology 217, 798-802 (2000)
11. L. Groussin, G. Bonardel, S. Silvéra, F. Tissier, J. Coste, G. Abiven, R. Libé, M. Bienvenu, J .- L. Alberini, S. Salenave, P. Bouchard, J. Bertherat, B. Dousset, P. Legmann, B. Richard, H. Foehrenbach, X. Bertagna, F. Tenenbaum, 18F- Fluorodeoxyglucose positron emission tomography for the diag- nosis of adrenocortical tumors: a prospective study in 77 operated patients. J. Clin. Endocrinol. Metab. 94, 1713-1722 (2009)
12. C. Guerin, F. Pattou, L. Brunaud, J .- C. Lifante, E. Mirallié, M. Haissaguerre, D. Huglo, P. Olivier, C. Houzard, C. Ansquer, E. Hindié, A. Loundou, C. Archange, A. Tabarin, F. Sebag, K. Baumstarck, D. Taïeb, Performance of 18F-FDG PET/CT in the characterization of adrenal masses in noncancer patients: a pro- spective study. J. Clin. Endocrinol. Metab. 102, 2465-2472 (2017)
13. G.W.L. Boland, B.A. Dwamena, M. Jagtiani Sangwaiya, A.G. Goehler, M.A. Blake, P.F. Hahn, J.A. Scott, M.K. Kalra, Char- acterization of adrenal masses by using FDG PET: a systematic review and meta-analysis of diagnostic test performance. Radi- ology 259, 117-126 (2011)
14. L.M. Weiss, Comparative histologic study of 43 metastasizing and nonmetastasizing adrenocortical tumors. Am. J. Surg. Pathol. 8, 163-169 (1984)
15. M. Korobkin, T.J. Giordano, F.J. Brodeur, I.R. Francis, E.S. Siegelman, L.E. Quint, N.R. Dunnick, J.P. Heiken, H.H. Wang, Adrenal adenomas: relationship between histologic lipid and CT and MR findings. Radiology 200, 743-747 (1996)
16. E.M. Caoili, M. Korobkin, I.R. Francis, R.H. Cohan, N.R. Dun- nick, Delayed enhanced CT of lipid-poor adrenal adenomas. Am. J. Roentgenol. 175, 1411-1415 (2000)
17. B.K. Park, B. Kim, K. Ko, S.Y. Jeong, G.Y. Kwon, Adrenal masses falsely diagnosed as adenomas on unenhanced and delayed contrast-enhanced computed tomography: Pathological correlation. Eur. Radiol. 16, 642-647 (2006)
18. A. Kutikov, K. Mallin, D. Canter, Y .- N. Wong, R.G. Uzzo, Effects of increased cross sectional imaging on the diagnosis and prognosis of adrenocortical carcinoma: analysis of the national cancer data base. J. Urol. 186, 805-810 (2011)
19. T.M.A. Kerkhofs, R.H.A. Verhoeven, J.M. Van der Zwan, J. Dieleman, M.N. Kerstens, T.P. Links, L.V. Van, de Poll-Franse, H.R. Haak, Adrenocortical carcinoma: a population-based study on incidence and survival in the Netherlands since 1993. Eur. J. Cancer 49, 2579-2586 (2013)
20. A.O. Ciftci, M.E. Şenocak, F.C. Tanyel, N. Büyükpamukçu, Adrenocortical tumors in children. J. Pediatr. Surg. 36, 549-554 (2001)
21. T. Kerkhofs, M. Ettaieb, R. Verhoeven, G. Kaspers, W. Tissing, J. Loeffen, M. Van den Heuvel-Eibrink, R. De Krijger, H. Haak, Adrenocortical carcinoma in children: First population-based clinicopathological study with long-term follow-up. Oncol. Rep. 32, 2836-2844 (2014)
22. D. Patel, S.K. Gara, R.J. Ellis, M. Boufragech, N. Nilubol, C. Millo, C.A. Stratakis, E. Kebebew, FDG PET/CT scan and Ffnctional adrenal tumors: a pilot study for lateralization. World J. Surg. 40, 683-689 (2016)
23. N. Jin, W. Qian, X. Yin, L. Zhang, K. Iqbal, I. Grundke-Iqbal, C .- X. Gong, F. Liu, CREB regulates the expression of neuronal glucose transporter 3: a possible mechanism related to impaired brain glucose uptake in Alzheimer’s disease. Nucleic Acids Res. 41, 3240-3256 (2013)
24. F. Beuschlein, M. Fassnacht, G. Assié, D. Calebiro, C.A. Strata- kis, A. Osswald, C.L. Ronchi, T. Wieland, S. Sbiera, F.R. Faucz, K. Schaak, A. Schmittfull, T. Schwarzmayr, O. Barreau, D. Vezzosi, M. Rizk-Rabin, U. Zabel, E. Szarek, P. Salpea, A. For- lino, A. Vetro, O. Zuffardi, C. Kisker, S. Diener, T. Meitinger, M. J. Lohse, M. Reincke, J. Bertherat, T.M. Strom, B. Allolio, Constitutive activation of PKA catalytic subunit in adrenal Cushing’s syndrome. N. Engl. J. Med. 370, 1019-1028 (2014)
25. D. Calebiro, G. Di Dalmazi, K. Bathon, C.L. Ronchi, F. Beus- chlein, cAMP signaling in cortisol-producing adrenal adenoma. Eur. J. Endocrinol. 173, M99-M106 (2015)
26. G. Di Dalmazi, R. Pasquali, F. Beuschlein, M. Reincke, Sub- clinical hypercortisolism: a state, a syndrome, or a disease? Eur. J. Endocrinol. 173, M61-M71 (2015)