Key words
adrenal Cushing’s syndrome - adrenocorticotrophic hormone - Cushing's syndrome - adrenocortical
carcinoma
Introduction
Endogenous Cushing’s syndrome (CS) is characterized by excess and unregulated
cortisol secretion leading to adverse clinical outcomes. Hypersecretion of cortisol
may be driven either by an excessive adrenocorticotrophic hormone (ACTH) from the
pituitary/ectopic source or by a primary adrenal pathology – adrenal
CS. Adrenal CS is a less frequent cause of CS, accounting for 20% of adult
patients, whereas it is more common in children (50% in children<7
years of age) [1]
[2]. Unilateral pathology (tumor) is
predominant, whereas bilateral pathology (hyperplasia) accounts for a smaller
proportion of patients. Adrenocortical carcinoma (ACC) has a bimodal age
distribution (first and 5–6th decade), whereas adrenocortical adenoma (ACA)
is more prevalent in the 4–5th decade [3]. Bilateral causes also have age predilection with primary pigmented
nodular adrenocortical disease (PPNAD) manifesting in the first three decades and
primary bilateral macronodular adrenal hyperplasia (PBMAH) presenting in the
5–6th decade [4]. Understanding the
differences in presentation, biochemistry, and radiological evaluation of adrenal
CS
is vital for their appropriate management.
Measurement of plasma ACTH level by immunometric assay helps in the etiological
classification of endogenous CS. Plasma ACTH level>20 pg/ml
indicates ACTH-dependent CS, while≤10 pg/ml indicates
ACTH-independent (adrenal) etiology. ACTH levels
> 10–20 pg/ml fall in the ‘grey
zone’, and an additional battery of tests is warranted for distinguishing
ACTH-dependent from ACTH-independent etiologies [5]
[6]. Inappropriate sample
collection and storage conditions may lead to falsely low ACTH levels due to
degradation of this labile analyte, which may erroneously suggest an
ACTH-independent CS. On the other hand, adrenal CS may be misdiagnosed as
ACTH-dependent by an unsuppressed plasma ACTH, as reported in a few case
reports/series [7]
[8]
[9]
[10]. Here, we describe our
experience of adrenal CS from a single center, emphasizing caution for
interpretation of ACTH.
Ethics Approval
The study was approved by the Institutional Ethics Committee-II
(EC/OA-101/2019) of Seth GS Medical College and KEM hospital
with a waiver of consent.
Patients and Methods
A retrospective data analysis of patients with adrenal CS, diagnosed and managed at
our institute between January 2006 and March 2020, a tertiary referral health care
center in western India, was done. The final etiological diagnosis of adrenal CS was
based on histopathology in all, except for five metastatic ACC and two PBMAH
patients in whom the diagnosis was based on clinical, hormonal, and imaging
characteristics. In adrenocortical tumor patients with available histopathology
(surgical specimen in 36 cases and biopsy in 4 cases), the presence of metastasis,
local invasion or recurrence, and/or a Weiss score≥4, were used to
diagnose ACC whereas in those without metastasis, local invasion or recurrence and
Weiss score of≤3, a diagnosis of ACA were made. Data on demography, clinical
evaluation, biochemistry, imaging, management, histopathology, and outcomes were
recorded in a standard format. In a patient presenting with adrenal mass, autonomous
cortisol secretion was diagnosed by an overnight dexamethasone suppression test
(ODST) serum cortisol value was>5 μg/dl
and/or 24-hour urinary free cortisol (UFC) was more than the upper limit of
normal (ULN). Patients with possible autonomous cortisol secretion (ODST cortisol
1.9–5.0 μg/dl) were excluded. Moon facies with
plethora, easy bruising, wide livid striae, proximal myopathy, and weight gain with
reduced growth velocity in children were considered as discriminatory signs of CS
(DSCS) [11]. In clinically suspected CS with
biochemically proven endogenous hypercortisolism and suppressed plasma
ACTH≤10 pg/ml were subjected to adrenal imaging. Plasma ACTH
level was repeated in patients with borderline plasma ACTH
(10–20 pg/ml). Contrast-enhanced CT adrenal imaging was
performed with a 64-slice multidetector CT system (Brilliance 64, Philips
Healthcare, Best, and The Netherlands), and absolute and relative washout
characteristics of adrenal masses were noted as per standard protocol [12]. 18F-Fluorodeoxyglucose positron
emission tomography-computed tomography (FDG-PET/CT) was performed as per
standard protocol for patients with suspected malignancy. The highest standardized
uptake value (SUVmax) was determined by software incorporated in the PET
workstation. It was defined as a focal area of abnormal uptake in the region of
interest (ROI) compared to the surrounding.
Patients with ACA underwent laparoscopic adrenalectomy, while adrenal tumors
suspicious of ACC underwent open surgical resection. Patients with unilateral
lesions who had post-surgery 8:00 AM serum cortisol
level<5 μg/dl were diagnosed to have suppressed
hypothalamic-pituitary-adrenal (HPA) axis and were replaced with oral
glucocorticoid. Patients with uncured/inoperable ACC were managed with local
bed radiotherapy and/or chemotherapy regimen with etoposide, doxorubicin,
and cisplatin with or without mitotane.
Cortisol was measured by a solid-phase competitive chemiluminescent enzyme
immunoassay (Siemens Healthcare) with an analytical sensitivity of
0.2 μg/dl. This assay’s intra-assay and inter-assay
coefficients of variability (CV) were 6.9 and 7.3%, respectively. Plasma
ACTH was measured on a solid-phase, two-site sequential chemiluminescent assay.
Immulite (Siemens Healthcare) assay has been used since 2006, and Liaison (Diasorin)
was added since December 2017. The intra-assay, interassay CV, and analytical
sensitivity were 9.6%, 8.8%, 0.5 pg/ml, and
4.9%, 8.9%, 1.6 pg/ml for Siemens Immulite and
Liaison assays, respectively. The lowest plasma ACTH value was considered for
analysis when multiple values were available. Plasma ACTH was measured with adequate
pre-analytic care (collection of plasma sample, maintaining a cold temperature, and
immediate processing). Serum DHEAS was measured by Chemiluminescence Microparticle
Immunoassay (CMIA) on Rosch Cobas platform with intra-assay and inter-assay (CVs)
of
4 and 4.6%, respectively, and analytical sensitivity of
0.2 μg/dl. We have used the age and gender-specific
normative range for serum DHEAS as described previously [13]
[14].
Statistical analysis
Qualitative data were represented as frequency and percentage. Quantitative data
were described using mean±standard deviation for normally distributed
data, otherwise as median with range. Association between qualitative variables
was assessed by the chi-square test or Fisher’s exact test. Analysis of
quantitative data between two groups was carried out using unpaired
t-test or Mann–Whitney test. For all statistical tests,
p<0.05 was considered significant. Receiver operating characteristic
(ROC) curve analysis was performed to differentiate ACC from ACA using tumor
size and relative washout. Statistical analysis was performed using IBM SPSS
Statistics (version 23.0) and MedCalc for Windows (version 19.8).
Results
Of the total endogenous CS patients (n=310), 18.7% (n=58) had
adrenal CS. The characteristics of the study cohort are described in [Table 1]. Unilateral adrenal pathology was
more common [ACC=30 (51.7%), ACA=15 (25.9%)]
compared to bilateral causes [PPNAD=10 (17.2%), PBMAH=3
(5.2%)]. Patients with PPNAD were the youngest [19 (1.2–48) years],
whereas those with PBMAH were the oldest [57 (30–60) years]. There was
female preponderance, which was most marked in the ACA group. DSCS was more frequent
in ACA (93.3%) and PPNAD (100%) than ACC (46.7%) and PBMAH
(1/3 patient).
Table 1 Baseline characteristics of adrenal Cushing’s
syndrome (CS).
|
Adrenocortical carcinoma n=30
|
Adrenocortical adenoma n=15
|
Primary pigmented nodular adrenal hyperplasia n=10
|
Primary bilateral macronodular adrenal hyperplasia
n=3
|
|
Age (years), median (range)
|
38 (1–50)*
|
25 (4.9–40)*
|
19 (1.2–48)
|
57 (30–60)
|
|
Gender, (Female/Male)
|
17/13*
|
14/1*
|
6/4
|
2/1
|
|
Presentation
|
Abdominal mass
|
26/30 (86.7%)*
|
1/15 (6.7%)*
|
–
|
–
|
|
Cushing’s syndrome
|
4/30 (13.3%)*
|
12/15 (80%)*
|
7/10 (70%)
|
1/3
|
|
Incidental
|
–
|
2/15 (13.3%)
|
–
|
2/3
|
|
Family screening
|
–
|
–
|
3/10 (30%)
|
–
|
|
Discriminatory signs of CS
|
14/30 (46.7%)*
|
14/15 (93.3%)*
|
10/10 (100%)
|
1/3
|
|
|
14/30 (46.7%)*
|
14/15 (93.3%)*
|
10/10 (100%)
|
1/3
|
|
|
3/30 (10%)*
|
10/15 (66.7%)*
|
4/10 (40%)
|
1/3
|
|
|
6/30 (20%)
|
7/15 (46.7%)
|
3/10 (30%)
|
2/3
|
|
|
1/6 (16.7%)
|
2/3 (66.7%)
|
6/9 (66.7%)
|
–
|
|
Irregular menses
|
5/10 (50%)
|
10/13 (79.6%)
|
1/2 (33.3%)
|
0/1
|
|
Hirsuitism
|
6/15 (40%)
|
8/12 (66.7%)
|
2/6 (33.3%)
|
0/2
|
|
Gynecomastia
|
4/13 (30.8%)
|
1/1
|
0/4
|
0/1
|
|
Osteoporosis
|
4/12 (33.3%)
|
2/8 (25%)
|
4/5 (80%)
|
1/2
|
|
Diabetes mellitus
|
4/30 (13.3%)
|
4/15 (26.7%)
|
1/10 (10%)
|
1/3
|
|
Hypertension
|
15/30 (50%)
|
6/15 (40%)
|
2/10 (20%)
|
2/3
|
|
Serum potassium (mEq/l), mean (±SD)
|
3.4 (±0.67)
|
3.7 (±0.78)
|
4.3 (±0.56)
|
4.2 (n=1)
|
|
8 AM serum cortisol (μg/dl), mean
(±SD)
|
24.3 (±11.9)
|
21.7 (±12.5)
|
19.6 (±7)
|
18 (±6.3)
|
|
ODST serum cortisol (μg/dl), Mean
(±SD)
|
19.6 (±10.5)
|
19 (±8.5)
|
21.3 (±8.3)
|
13.7 (±10.2)
|
|
Plasma
Adrenocorticotrophic
hormone (pg/ml)
|
Mean ( ± SD)
|
15.9 (±13.3)*
|
8.8 (±12)*
|
7.2 (±3.7)
|
4.5 (±5.1)
|
|
Median (range)
|
11.5 (1.6–52.9)*
|
5 (1.6–49.6)*
|
7.1 (1.6–12.1)
|
1.6 (1.6–10.4)
|
|
≤ 10 pg/ml
|
12 (40%)
|
12 (80%)
|
8 (80%)
|
2 (66.7%)
|
|
> 10–20 pg/ml
|
10 (33.3%)
|
2 (13.3%)
|
2 (20%)
|
1 (33.3%)
|
|
> 20 pg/ml
|
8 (26.7%)
|
1 (6.7%)
|
0
|
0
|
|
Serum DHEAS (μg/dl), median (range)
|
307.1 (29.6–1500)*
|
28.5 (16.2–367)*
|
38.8 (2.5–203)
|
22 (15–19)
|
|
Elevated DHEAS>Upper limit of normal
|
9/18 (50%)
|
2/11 (18.2%)
|
1/6 (16.7%)
|
0/2
|
|
Low DHEAS<Lower limit of normal
|
3/18 (16.7%)*
|
6/11 (54.5%)*
|
2/6 (33.3%)
|
2/2
|
|
Maximum lesion size (cm), mean (±SD)
|
13.3 (±4.2)*
|
3.5 (±0.9)*
|
–
|
3.8 (±1.2)
|
|
Tumor size>6 cm
|
28/30 (93.3%)
|
0/15
|
–
|
–
|
|
Tumor size>4 cm
|
29/30 (96.7)
|
4/15 (26.7%)
|
–
|
–
|
|
Metastasis/Local invasion
|
16/5
|
0/0
|
–
|
–
|
|
Tumor side (Right/Left)
|
11/19
|
10/5
|
–
|
–
|
|
Unenhanced CT density (HU), mean (±SD)
|
39.8 (±6)*
|
23 (±10.8)*
|
–
|
–
|
|
Unenhanced CT density<10 HU
|
0/18
|
1/11
|
–
|
–
|
|
Absolute washout (%)
|
42±23.8* (n=18)
|
66.5±17.9* (n=11)
|
–
|
–
|
|
Relative washout (%)
|
16.7±11.3*(n=18)
|
49.5±17.8* (n=11)
|
–
|
–
|
|
Absolute washout>60%
|
5/18
|
7/11
|
–
|
–
|
|
Relative washout>40%
|
0/18
|
8/11
|
–
|
–
|
|
Lesion SUVmax (FDG PET)
|
12.9±4.8* (n=18)
|
6.2±1.8* (n=3)
|
–
|
–
|
|
Lesion SUVmax to liver SUVmean ratio (FDG PET)
|
6.5±2.6 (n=10)
|
1.9 (n=1)
|
–
|
–
|
*Statistically significant difference between ACC and ACA
cohorts. Variables, where units are not specified are described as
n/N (%); ODST: Overnight dexamethasone suppression test;
DHEAS: Dehydroepiandrosterone sulfate; SUVmax: Highest standardized uptake
value; FDG PET: Fluorodeoxyglucose (FDG)-positron emission tomography.
On comparing ACC with ACA, ACC primarily presented as abdominal mass (86.7%)
while most ACA presented with CS (80%). ACA presented at a younger age and
had female predilection. Basal and ODST serum cortisol were comparable between the
two groups. Plasma ACTH levels were significantly higher (15.9±13.3 vs.
8.8±12.0 pg/ml) in the ACC group than in the ACA group.
Twelve patients (ACC=10, ACA=2) had plasma ACTH levels between 10
and 20 pg/ml and nine patients (ACC=8, ACA=1) had
levels>20 pg/ml.
Low serum DHEAS had a sensitivity of 54.5% (6/11) and specificity of
83.3% (15/18) for the diagnosis of ACA. Elevated serum DHEAS had a
sensitivity of 50% (9/18) and specificity of 81.8%
(9/11) for the diagnosis of ACC. A patient of ACA (5-year-old girl) had
presented with abdominal mass (size: 5.6 cm) and also had pubarche and
clitoromegaly with a serum testosterone level of 1.1 ng/ml and DHEAS
level of 185.3 μg/dl (normal range: 7.4–46.8).
Another 13-year-old boy with the final diagnosis of ACA (size: 5.2 cm) had
presented with CS; also had gynecomastia with a serum estrogen level of
246 pg/ml and DHEAS level of 367 μg/dl
(normal range: 13.6–288.9). Both these patients had a Weiss score of 3 on
histopathological examination. Adrenal tumor size>4 cm had a
sensitivity of 96.7% (29/30) and specificity of 73.3%
(11/15) for the diagnosis of ACC. This specificity increased to
100%, with a sensitivity of 93.3% (28/30) when tumor size
cut off was increased to>6 cm. Basal HU was>10, for all
adrenal tumors, except one ACA (1/11, 9.1%). Absolute washout
of>60 and≥55% had sensitivities of 63.6%
(7/11) and 90.9% (10/11) and specificities of 72.2%
(13/18) and 66.7% (12/18), respectively, for the diagnosis
of ACA. Relative washout of>40% had a sensitivity of 72.7%
(8/11) and specificity of 100% (18/18) for the diagnosis of
ACA. On FDG-PET/CT, the lesion SUVmax of ACC was significantly higher than
that of ACA (12.9±4.8 vs. 6.2±1.8, p=0.019), and the ratio
of lesion SUVmax to liver SUVmean was 6.5±2.6 (n=10). In the ACA
cohort, lesion SUVmax was 6.2±1.8 (n=3), and lesion SUVmax to liver
SUVmean ratio was 1.9 (n=1). Using ROC curve analysis, a lesion size of more
than of 5.4 cm had sensitivity and specificity of 93.3 and 99.93%
respectively, for the diagnosis of ACC whereas a relative washout of more than
31.8% had sensitivity and specificity of 90.9 and 95%, respectively,
for the diagnosis of ACA.
The plasma ACTH values for adrenal CS patients are depicted in [Fig. 1]. One ACA and eight ACC patients had
plasma ACTH of>20 pg/ml despite hypercortisolemic state as
defined by ODST serum cortisol value>5 μg/dl
and/or UFC>ULN ([Table 2]).
Repeat ACTH values were not available for these ACC patients as they had presented
with abdominal pain leading to detection of large adrenal mass
(8.9–19 cm) on ultrasonogram. Of these eight ACC patients, only
three had DSCS. While the ACA patient with ACTH above 20 pg/ml was a
21-year-old female, presented with CS and an unsuppressed plasma ACTH
(49.6 pg/ml). Her pituitary imaging was normal, and inferior
petrosal sinus sampling was planned. Meanwhile, contrast-enhanced CT chest and
abdomen was done to localize an obvious ectopic source, which revealed a
3.2 cm lipid-poor (unenhanced CT density: 27 HU) left adrenal mass with
relative washout of 47.1%, with a thinned out right adrenal. She was cured
after the left adrenalectomy. ACTH values in these patients were obtained from the
Seimens Immulite platform. In another five patients with initial ACTH
of>13 pg/ml (13.1–149 pg/ml)
(13.1–149 pg/ml) by Siemens Immulite assay, repeat ACTH
measurements by Liaison assay were<13 pg/ml
(1.6–12.7 pg/ml).
Fig. 1 Scatter diagram of adrenocorticotrophic hormone (ACTH) levels
in the total study cohort (n=58).
Table 2 Adrenal Cushing syndrome (CS) patients with plasma
adrenocorticotrophic hormone (ACTH)>20 pg/ml
from this study.
|
Patient No
|
Age/sex
|
Presentation
|
Discriminatory signs of CS
|
8 AM serum cortisol (μg/dl)
|
ODST Serum Cortisol (μg/dl)
|
Plasma ACTH pg/ml (Assay platform)
|
Tumor size (cm)
|
Final diagnosis
|
|
P 1
|
45/F
|
Abdominal mass
|
No
|
20.3
|
7
|
35.2 (SI)
|
12
|
ACC
|
|
P 2
|
40/F
|
Abdominal mass
|
No
|
17.9
|
17.8
|
26.1 (SI)
|
11.6
|
ACC
|
|
P 3
|
46/M
|
Abdominal mass
|
No
|
11.3
|
11.8
|
25.4 (SI)
|
10.6
|
ACC
|
|
P 4
|
30/M
|
Abdominal mass
|
No
|
10.8
|
8.4
|
46 (SI)
|
19
|
ACC
|
|
P 5
|
45/M
|
Abdominal mass
|
No
|
–
|
9
|
27.7 (SI)
|
18.9
|
ACC
|
|
P 6
|
45/F
|
Abdominal mass
|
Yes
|
19.3
|
25.1
|
27.9 ,(SI)
|
13
|
ACC
|
|
P 7
|
29/F
|
Abdominal mass
|
Yes
|
–
|
–a
|
52.9 (SI)
|
10
|
ACC
|
|
P 8
|
42/M
|
Abdominal mass
|
Yes
|
32.8
|
–a
|
37.5 (SI)
|
8.9
|
ACC
|
|
P 9
|
21/F
|
Cushing’s syndrome
|
Yes
|
28.5
|
28
|
49.6 (SI)
|
3.2
|
ACA
|
a Urinary free cortisol was 1.7 and 7.3 times the upper limit of
normal for P7 and P8, respectively; F: Female, M: Male; ACC: Adrenocortical
carcinoma; ACA: Adrenocortical adenoma; ODST: Overnight dexamethasone
suppression test; SI: Siemens Immulite.
Most ACC patients (53.3%) were of ENSAT (European Network for the Study of
Adrenal Tumors) stage IV, while stage II and III comprised 20% patients
each, and only two patients presented with stage I disease. Surgery was not
considered feasible in nine out of the 16 metastatic patients, of which 5 received
palliative chemotherapy. Twenty-one patients underwent open surgery followed by
local bed radiotherapy (n=9) and chemotherapy (n=13). The median
survival of the ACC group was 23 (95% CI: 3 to 43) months ([Fig. 2]). All ACA patients were cured
following laparoscopic excision of the tumor. The mean time to HPA axis recovery
post-surgery was 10.6±9.1 months.
Fig. 2 Kaplan–Meier survival curve of the total cohort of
adrenocortical carcinoma (n=30).
Most (9/10) of the PPNAD patients have been described previously [15]. Two patients of AIMAH presented
incidentally on abdominal imaging, while one presented with CS. The latter, who
underwent bilateral adrenalectomy, were cured, while the other two were managed with
watchful observation (annual monitoring with 24-hour UFC) and medical therapy for
comorbidities (diabetes mellitus and hypertension).
Discussion
In our cohort, adrenal CS comprised 18.7% of all endogenous CS, with
unilateral etiology (77.6%) being more common than bilateral
(22.4%). We report that plasma ACTH level
of>20 pg/ml is not uncommon in adrenal CS when measured by
Siemens Immulite assay, especially in patients with ACC. All ACC with an ACTH
of>20 pg/ml were detected by ultrasonogram. Washout
characteristics (relative washout<40%, absolute washout
of<60%) and size (>5.4 cm), but not baseline HU, had
good diagnostic accuracy to differentiate cortisol- secreting ACC from ACA.
The proportion of adrenal CS amongst endogenous CS is similar to that reported in
the
literature (≈20%) [1]
[2]
[6]. In
contrast, amongst the adrenal CS, the proportion of ACC (despite being a predominant
adult cohort) was more than ACA, which may be due to referral bias.
The younger age of our ACC cohort may represent the earlier age at presentation of
cortisol-secreting ones, as also reported in a previous Indian study [16], than unselected ACC cohorts. Despite
cortisol hypersecretion in all, clinical features of CS were present in only
46.7% of patients. This observation suggests considering a biochemical
evaluation for CS in all ACC patients, irrespective of the clinical features. The
majority (86.7%) of patients in our cohort had presented with abdominal
symptoms, possibly due to delay in seeking healthcare or inefficient cortisol
secretion or action (masked by the anabolic action of androgens). In addition,
53.3% of patients had distant metastases at presentation, compared to lower
metastasis rates (25 to 47.4%) in larger studies [3]
[17].
This could reflect either delayed diagnosis or more aggressive nature of the disease
in cortisol-secreting ACC. Corresponding with the presentation, the median survival
in our cohort was also shorter than reported cohorts [17]
[18].
The ACA subgroup in our study had a mean age of 25.7±10.5 years, and
predominant presentation due to overt CS in 80% of patients, which could be
due to selection bias as patients with possible autonomous cortisol production (ODS
cortisol 1.8 to 5 μg/dl) were not included in the cohort. In
addition, we had two pediatric patients who presented with co-secretion of sex
steroids. Co-secretion of cortisol with sex steroids by ACA is rarely described and
should raise the suspicion for malignancy [3]
[19]
[20]. Both of these patients had a Weiss score
of 3 on histopathological examination and were disease-free at the last follow-up
(5
and 5.5 years). However, low-grade ACC cannot be ruled out in both these patients
and are under close surveillance for recurrence of the disease. The mean duration
for recovery of the HPA axis after surgery in our cohort was similar to the reported
11.2 months in literature [21].
In general, 71% of benign adrenal masses have a basal HU of<10
(lipid-rich) [22]. In contrast, only
9.1% of cortisol secreting adenoma were lipid-rich in our study. Similar
observations were made in two studies from France, one reporting 12%
(3/25) and another reporting 20.6% (7/34) of
cortisol-secreting ACA having a basal HU of<10 [23]
[24].
In a clinicopathological correlation study, unlike aldosterone-secreting ACA that
were characterized by lipid-rich clear cells, most of cortisol-secreting ACA were
characterized by the presence of granule cells [25]. Similarly, in another study from France, lipid-poor (baseline
attenuation:>10 HU) cortisol-secreting ACA had<25%
lipid-rich clear cells in 50% [24]. In
this study, the magnitude of cortisol hypersecretion in cortisol-secreting ACA was
associated with lipid-poor nature, both on histopathology (<25%
clear cells) and radiology (baseline HU). Hyperfunctioning nature with depletion of
intracytoplasmic lipid droplets containing cholesterol esters necessary for cortisol
synthesis or upregulation of cAMP pathway with a change from lipid-rich clear cell
to lipid-poor compact cell have been hypothesized to account for the lipid poor
nature of cortisol-secreting ACA. Hence, basal attenuation of<10 HU has less
diagnostic sensitivity for cortisol secreting ACA. However, a baseline attenuation
of<10 HU excludes ACC, irrespective of secretory status, and can be useful
to differentiate ACA, including a small proportion of cortisol-secreting ones, with
this radiological characteristic (10–20%) from ACC [23]
[24]
[26].
Size of the tumor>6 cm was 100% specific for the diagnosis of
ACC, but sensitivity was 93.3% (28/30). Poor relative contrast
washout (<40%) was 100% sensitive for the diagnosis of ACC,
but specificity was 72.7%, as few [3/11 (27.3%)] ACA
exhibited poor washout. This is in concordance with data in general, where
lipid-poor adenoma has poor washout in 41% of the patients [27]. Hence, a role FDG-PET/CT has been
suggested by He et al. to differentiate the malignant lesions (higher SUVmax and
higher lesion SUVmax to liver SUVmean ratio) from benign ones [28]. Similar values of these two parameters
were noted in our study too. Notably, a patient with ACC had a lesion size
of<6 cm, but the lesion SUVmax of 11.6 was suggestive of malignancy.
Similarly, a patient of ACA with poor washout characteristics had a lesion SUVmax
of
6.7, which was indicative of its benign nature. Also, higher FDG uptake has been
observed in cortisol-secreting ACA in comparison to non-secreting ones [23].
PBMAH patients of our study presented later in life either with subclinical
(n=2) or clinical CS (n=1), which is similar to the available
literature. Unlike ACC, ACTH was suppressed in all three patients despite
subclinical CS in two patients. The decision for bilateral adrenalectomy or medical
management was individualized based on the clinical severity of hypercortisolemia
and comorbidities, as suggested previously [4].
We report that plasma ACTH of>20 pg/ml is not uncommon in
patients with adrenal CS, especially in ACC. In contrast, unsuppressed ACTH in
adrenal CS is limited to a few cases in the literature. Adrenal CS cases reported
in
the literature (ODST serum cortisol value>5 μg/dl
and/or 24-hour UFC more than the upper limit of normal) with unsuppressed
ACTH are summarized in [Table 3].
Unsuppressed ACTH in adrenal CS may prompt unnecessary investigations like MRI
pituitary, IPSS, CT thorax and neck, and/or nuclear scans, and in
unfortunate instances, even an unnecessary pituitary surgery [7]
[8]
[9]. Unsuppressed ACTH has been
reported mostly with ACA, whereas in our series, we observed eight patients of ACC
had plasma ACTH>20 pg/ml. The varying postulates for such
high ACTH values can be assay-related concerns, aberrant precursors secreted by the
tumor, or intra-adrenal ACTH production.
Table 3 Adrenal Cushing syndrome (CS) patients from this study
and those reported in the literature with unsuppressed plasma ACTH as
measured by different immunometric assays.
|
No. [Ref]
|
Age (years)/Sex (M/F)
|
Discriminatory signs of CS
|
Presentation
|
Tumor size (cm)
|
ODST (μg/dl)/UFC (μg/day
or times x ULN)
|
Plasma ACTH pg/ml (Assay platform)
|
Revised Plasma ACTH pg/ml (Assay platform)
|
Investigation done for localizing the site of excess ACTH
|
Final diagnosis
|
|
1 [8]
|
43/F
|
Yes
|
CS
|
3
|
–/–
|
56 (SI)
|
<1 (RE)
|
MRI pituitary Indium111 Pentreotide scan, FDG PET CT,
Chest CT
|
ACA
|
|
2 [8]
|
48/F
|
No
|
Incidentaloma
|
3.8
|
–/>ULN
|
19 (SI)
|
1.76 (RE)
|
–
|
ACA
|
|
3 [8]
|
26/F
|
Yes
|
NA
|
4.2
|
–/>ULN
|
98.5 (SI)
|
<5 (RE)
|
–
|
ACA
|
|
4 [8]
|
46/F
|
Yes
|
NA
|
3.3
|
–/>ULN
|
14 (SI)
|
<1 (RE)
|
MRI pituitary, IPSS
|
ACA
|
|
5 [9]
|
30/F
|
Yes
|
CS
|
2.5
|
20.6/537
|
62 (SI)
|
2.4 (T)
|
MRI pituitary, Gallium-68 DOTATATE scan×2, IJVS, IPSS
|
ACA
|
|
6 [9]
|
55/F
|
Yes
|
CS
|
2.1
|
26.7/132
|
143 (SI)
|
3.9 (T)
|
MRI pituitary, MIBG scan, Gallium-68 DOTATATE scan, IJVS,
IPSS
|
ACA
|
|
7 [9]
|
23/F
|
Yes
|
CS
|
2.5
|
14.4/127
|
21 (SI)
|
1 (R)
|
MRI pituitary
|
ACA
|
|
8 [7]
|
25/M
|
Yes
|
CS
|
11.7
|
24.28/–
|
15.8a
|
<1a
|
Pituitary surgery
|
ACC
|
|
9 [30]
|
44/F
|
No
|
NA
|
2.5
|
13.4/–
|
23.16 (SI)
|
<10 (PEG, SI)
|
–
|
PBMAH
|
|
10 [30]
|
69/F
|
No
|
NA
|
2.7
|
6.6/–
|
22.25 (SI)
|
<5 (HBT, SI)
|
IPSS
|
ACA
|
|
11 [30]
|
28/F
|
Yes
|
CS
|
3.3
|
18.7/–
|
54.95 (SI)
|
<10 (PEG, SI)
|
Gallium-68 DOTATATE scan, FDG PET/CT IPSS
|
ACA
|
|
12 [31]
|
60/F
|
No
|
Incidentaloma
|
2.3
|
20.3/2.3×ULN
|
16 (SI)
|
<0.05 (RE)
|
IPSS
|
PBMAH
|
|
13 [32]
|
54/M
|
Yes
|
NA
|
3.2
|
–/4×ULN
|
65.39 (SI)
|
0.9 (RE)
|
IPSS
|
ACA
|
|
Patients from this study
|
|
1
|
1/F
|
Yes
|
CS
|
3.9
|
9.01/–
|
149 (SI)
|
5.8 (LD)
|
–
|
ACC
|
|
2
|
37/F
|
Yes
|
CS
|
3
|
21.4/–
|
44.2 (SI)
|
1.6 (LD)
|
–
|
ACA
|
|
3
|
30/F
|
Yes
|
CS
|
3.6
|
37.2/–
|
36 (SI)
|
12.7 (LD)
|
–
|
ACA
|
|
4
|
26/M
|
Yes
|
CS
|
–
|
22.8/1500
|
13.7 (SI)
|
2.55 (LD)
|
–
|
PPNAD
|
|
5
|
30/F
|
Yes
|
CS
|
3.4
|
25.4/–
|
13.1 (SI)
|
1.6 (LD)
|
–
|
PBMAH
|
NA: Details not available; ULN: Upper limit of normal; SI: Siemens Immulite
assay; RE: Roche Elecsys assay; T: Tosoh AIA assay; R: Roche Cobas assay;
LD: Liaison DIasorin assay; FDG PET: Fluorodeoxyglucose (FDG)-positron
emission tomography; IPSS: Inferior petrosal sinus sampling; IJVS: Internal
jugular vein sampling; MIBG: Meta-iodobenzylguanidine; ACA: Adrenocortical
adenoma; ACC: Adrenocortical carcinoma; PBMAH: Primary bilateral
macronodular adrenal hyperplasia; a Assay details not
available.
Assay-related concerns leading to high ACTH can be attributed to the type of ACTH
assay, different ACTH platforms, or heterophile antibody interference [8]
[9]
[29]
[30]
[31]
[32]. Traditionally,
radioimmunoassays (RIA) were used, which also measure clinically insignificant
fragments of POMC-ACTH and caused spurious elevations of ACTH. Immunoradiometric
(IRMA) assays have improved this drawback as the "sandwich"
complexes ensure specificity. As observed in an Italian multicentric study, plasma
ACTH of>20 was observed in four patients of adrenal CS by RIA, as against a
single case by IRMA [33]. Contemporary
chemiluminescence sandwich immunoassay platforms use different capture and detection
antibodies, and this can lead to a difference in ACTH measured, especially at lower
ACTH levels, as shown by a study from Italy [34]. In a recent study, plasma intact ACTH level measured by
LC-MS/MS highly positively correlated with plasma ACTH measured by Roche but
not with that by Siemens; also, among the three discordant samples in which plasma
ACTH by LC-MS/MS was undetectable, that by Roche
was<20 pg/ml in two but was more than
100 pg/ml in all when measured by Siemens [35]. As depicted in [Table 3], in many cases, the ACTH detected by
Siemens Immulite was much greater than Roche Elecsys performed subsequently.
Similarly, all patients with ACTH of>20 pg/ml in our series
were measured on the Siemens Immulite platform. Although a repeat ACTH estimation
was not available for these patients, that by Liaison assay in another five patients
with an initial ACTH of 13 pg/ml by Siemens Immulite assay was
suppressed (<13 pg/ml) in all ([Table 3]) suggesting a major role for
assay-related issues in an apparently unsuppressed ACTH. Notably, most (8/9)
of our patients with this phenomenon had ACC that makes us think of an additional
role for ACC-related, yet-unidentified, factors in causing a plasma ACTH
of>20 pg/ml. As conceptualized in an old study,
steroidogenic precursors (21-deoxycortisol) secreted by ACC may antagonize the
glucocorticoid feedback at the hypothalmic-pituitary level [36]
[37].
Although the intra-adrenal source of ACTH has been reported in PBMAH and mixed
cortico-medullary adrenal tumors (MCMT), the evidence to support the production of
ACTH or ACTH-like substances from ACC is negligible [38]
[39]. The evaluation for the
possible role of heterophile antibodies was not performed in our study.
Hence, when ACTH is measured by the Immulite assay and
is>20 pg/ml in an endogenous CS patient, one must be
cautious and rule out ACC especially in the presence of clinical or biochemical
evidence of androgen excess and/or local abdominal mass effects, before
ordering MRI pituitary. Repeating ACTH on a different platform using a more accurate
ACTH assay and even considering adrenal imaging if clinical suspicion is high may
help in correct diagnosis.
Our study is a retrospective analysis with inherent limitations. Serum DHEAS was not
available in some patients. The lack of LC-MS/MS steroid profile and ACTH
measurements from different platforms are other limitations. Nonetheless, all the
included patients had confirmed adrenal CS with a significant sample size, and
detailed clinical, biochemical and radiological features along with clinical
outcomes are described.
To conclude, adrenal CS has a varied spectrum of clinical, biochemical, and imaging
features that may help to differentiate malignant causes from benign ones.
Cortisol-secreting ACA is usually lipid-poor, and baseline HU is not a sensitive
radiological parameter to distinguish cortisol-secreting ACA from ACC. Plasma ACTH
levels may be>20 pg/ml in adrenal CS when measured by
Siemens Immulite assay, especially in patients with ACC, which can be addressed by
using more accurate ACTH assays.
Data Availability Statement
Data Availability Statement
The data that support the findings of this study are available on request from the
corresponding author. The data is not publicly available due to privacy or ethical
restrictions.
