Key words
ACC - melan A - p53 - Weiss score - Helsinki score
ACCAdrenocortical carcinoma
ENSATEuropean Network for the Study of Adrenal Tumors Classification
IQRInterquartile range
OSOverall survival
HUHounsfield units
INHAInhibin Alpha-Subunit
Introduction
Adrenocortical carcinoma (ACC) is a malignant neoplasm of adrenocortical origin. It
affects 1–2 patients per million individuals per year and accounts for 0.02–0.2% of
all cancer deaths. It tends to be slightly more common in women [1]
[2]
[3]. ACC can occur at any age, but there is a relatively higher incidence in children<5
years of age and adults in their fourth to fifth decade of life. The long-term outcome
is poor with a 5-year survival rate of 60–80% for ACC localized to the adrenal gland,
35–50% for ACC with locally advanced disease and 0–28% for metastatic ACC, indicating
the prognostic relevance of tumor staging [4]. ACC incidence is 10-fold higher in Southern Brazil children, where a specific TP53
germline mutation (R337H) has been detected. Although most ACCs are sporadic,<10%
of the cases can occur as part of hereditary syndromes including Li–Fraumeni syndrome
(TP53 mutations), Beckwith–Wiedemann syndrome (IGF-II overexpression), and familial
adenomatous polyposis coli (APC and CTNNB1 genes) [5].
Clinical presentation of ACC is variable. Many ACCs are biochemically functional and
are associated with glucocorticoid excess (Cushing’s syndrome), sex steroid hypersecretion
(virilizing or rarely feminizing signs), combined glucocorticoid and sex steroid excess,
or mineralocorticoid overproduction [6]. However, a significant proportion of these tumors may be unassociated with syndromes
of steroid excess and typically present with abdominal or flank pain due to a large
retroperitoneal mass [7]
[8]. There are also cases where clinical manifestations are totally absent, and the
tumor is discovered incidentally or even metastatic at time of presentation [9].
The diagnosis of ACC is not always easy since there are cases that are difficult to
be distinguished from benign adenomas and from tumors of other origin such as pheochromocytoma,
liposarcoma or malignant melanoma [10]. SF1 remains a valid marker to document the adrenal origin with a high sensitivity
and specificity, while if it is not available, others markers including inhibin-A,
melan A, and calretinin can be useful when combined [11]. Weiss score, based on a combination of 9 histological criteria remains the best
validated score to distinguish adenomas from ACCs. A Weiss score>3 and a Helsinki
score>8.5 are strongly suggestive of malignancy [12]
[13]. A Weiss score>3 and recently a Ki67 labeling index>10% are strongly suggestive
of malignancy. Of interest, other biomarkers have recently emerged, amongst them proteins
involved in cell proliferation and mitotic spindle regulation, beta-catenin pathway,
DNA damage repair, telomere regulation and PI3K signaling pathway, with IGF-2 being
the most promising. Indeed, juxtanuclear Golgi pattern of IGF-2 expression along with
altered reticulin framework and angioinvasion were found to be the most useful diagnostic
markers of ACC [14].
The most important clinical factors that determine long-term prognosis of ACC are
the disease stage at presentation (according to ENSAT system), radical surgery and
age. However, Ki67 remains the most powerful prognostic marker in both localized and
advanced ACC. Of interest, proliferative activity/tumor grade and cortisol excess
have also been recognized as independent prognostic parameters [15].
Surgery includes complete margin-negative resection of the tumor with the adrenal
gland and when necessary, en bloc resection of invaded organs and peri-aortic retroperitoneal
lymphadenectomy [16]
[17]. Open adrenalectomy with locoregional lymph node removal is recommended when preoperative
diagnosis or high level of suspicion of ACC exists.
In the current study, we aimed to present the 20-year experience at a tertiary Hospital
in patients with ACCs and correlate the immunohistochemical (IHC) features of ACCs
with the clinical – morphological data and survival of patients.
Ethics Approval
This is a retrospective cohort study conducted at the Third Department of Surgery
at General Hospital of Athens “G. Gennimatas”. Ethical approval for this study was
granted upon evaluation by the Scientific Committee of Athens General Hospital. The
study was conducted according to the guidelines of the Declaration of Helsinki and
its later amendments or comparable ethical standards, and it was also approved by
the Institutional Ethics Committee of G. Gennimatas General Hospital of Athens, Greece
(9253/17.03.2017). Written consent was obtained from all patients. A total of 45 cases
of ACC were operated between June 1997 and September 2019, by the same surgeon.
Patients and Methods
The preoperative diagnosis, operative details, complications, length of hospital stay,
morbidity, pathology report, and follow up were retrieved from the hospital records
of 742 patients who underwent 758 adrenalectomies for benign and malignant tumors
in the same period. Preoperative localization was established in all patients by computerized
tomography (CT) or magnetic resonance imaging (MRI). Endocrinological evaluation and
complete adrenal dynamic testing were performed to determine whether the tumor was
functional or not.
The cases were reviewed and a diagnosis of ACC was confirmed using histopathological
examination after surgical resection. Immunohistochemical analysis for Inhibin-A,
Melan A, Calretinin, Ki67, Synaptophysin, CKAE1/AE3, p53, Vimentin has also been performed.
Demographic and clinical characteristics of the patients including also the stage
of the disease, the localization, the weight and size of tumor, type of surgical procedure
and the type of hypersecreted steroids, in cases of functional tumor, are provided
in [Table 1].
Table 1 Demographic, clinical, and histologic characteristics of patients with ACC.
|
Number
|
%
|
|
Gender
|
|
|
|
women
|
26
|
57.7
|
|
men
|
19
|
42.3
|
|
Age
|
|
|
|
<20
|
1
|
2.2
|
|
20–30
|
2
|
4.45
|
|
31–40
|
2
|
4.45
|
|
41–50
|
14
|
31.1
|
|
51–60
|
10
|
22.2
|
|
61-70
|
8
|
17.8
|
|
>70
|
8
|
17.8
|
|
Tumor size (greatest dimension)
|
|
|
|
<10 cm
|
30
|
66.7
|
|
≥10 cm
|
15
|
33.3
|
|
Tumor weight
|
|
|
|
<300 g
|
31
|
68.9
|
|
≥300 g
|
14
|
31.1
|
|
Lateralization
|
|
|
|
Left adrenal
|
26
|
57.7
|
|
Right adrenal
|
19
|
42.3
|
|
Stage (ENSAT)
|
|
|
|
I
|
3
|
6.6
|
|
II
|
28
|
62.3
|
|
III
|
12
|
26.7
|
|
IV
|
2
|
4.4
|
|
Surgical approach
|
|
|
|
Open adrenalectomy
|
30
|
66.7
|
|
Laparoscopic adrenalectomy
|
12
|
26.7
|
|
Conversion to open
|
3
|
6.6
|
|
Resection of additional organs (kidney, spleen)
|
10
|
22.2
|
|
Hormonal activity
|
|
|
|
Functional
|
17
|
37.7
|
|
Nonfunctional
|
28
|
62.3
|
All the operations, open or laparoscopic, were performed with the patient in the lateral
decubitus position. Therefore, there was no need to change the position of the patient
in the event of a conversion to an open operation. A transperitoneal lateral approach
was used to perform laparoscopic adrenalectomy. The adrenal gland was not grasped
during dissection in order to avoid troublesome bleeding or tumor disruption. The
specimen in laparoscopic procedures was placed in a special bag and was extracted
through minimal extension of the initial incision.
Thirty surgeries were open adrenalectomies, 12 laparoscopic and 3 other laparoscopic
surgeries were converted to open surgery. Additional organs (kidney, spleen) were
resected in 10 operations. There was no surgical mortality. Median hospital stay was
significantly longer in the open compared to laparoscopic surgery group (median 7
compared to 3 days respectively).
Routine lymphadenectomy during surgical resection of ACC has not been performed in
any of our cases. A few lymph nodes (1–5) were included in specimens where resection
of additional organs was performed (peripheral pancreatectomy and splenectomy or nephrectomy
). Para-aortic lymphadenectomy is performed when indicated from imaging or intraoperative
inspection. The role of lymph node dissection in the management of ACC remains controversial
[18]
[19].
Histopathological evaluation and Immunohistochemical staining
Histological diagnosis of ACC was recorded by a reference pathologist on tumor tissue
removed at surgery. All cases were evaluated on a routine hematoxylin-eosin-stained
slide according to the Weiss criteria in which the presence of three or more criteria
highly correlates with malignant behavior [13]. The Helsinki score was determined as follows: (5×presence of necrosis)+(3×mitotic
rate>5 per 50 high-power fields)+proliferation index value in the most proliferative
area of tumor. The thresholds used for malignancy were a total score≥3 for the Weiss
score and a total score>8.5 for the Helsinki score.
Immunohistochemical studies were performed on a representative deparaffinized tissue
section after antigen retrieval using appropriate positive and negative controls in
all cases. The Ki67 index was evaluated as a proliferation marker to assess ACC prognosis
using the anti-human Ki67 antibody (clone MIB-1, DAKO). Ki-67 index was estimated
as the percentage of the number of immunostained nuclei among the total number of
nuclei of tumor cells (count 500–2000 tumor cells) in the high-power field measured
in chosen areas of the highest density of positive nuclei (hotspots). Tumor stage
was evaluated according to the classification of ACC proposed by the European Network
for the Study of Adrenal Tumours (ENSAT). Immunohistochemical analysis was performed
using anti p53 (clone DO-7, DAKO), CK-PAN (clone AE1/AE3, DAKO), Inhibin-A (clone
EP378, BioSB), Melan A (clone A103, DAKO), Calretinin (clone DAK-Carelt1, DAKO), Synaptophysin
(clone DAK-SYNAP, DAKO) and Vimentin (clone V9, DAKO) antibodies. Immunohistochemical
analysis was conducted using DAKO EnVision FLEX (Dako, Carpenteria, CA, USA). The
staining intensity in each slide was evaluated independently by two different pathologists,
blinded to any clinical information prior to pathological review and interpretation
for the purpose of this study. Immunohistochemical staining was considered positive
when more than 10% of them were stained with the respective antibodies. For the evaluation
of p53, the nuclear expression of p53 protein in few or in almost all tumor cells
but with variable intensity was considered as normal/wild type pattern indicative
of no underlying TP53 mutation. The strong nuclear expression in at least 50% of the
tumor cell was defined as overexpression while the overexpression in at least>80%
was considered strongly associated with TP53 mutation [20]
[21]
[22].
Statistical analysis
The Shapiro-Wilk test was used for the assessment of normality of data distribution.
For qualitative variables, comparisons among two or more groups were made using Chi-squared
test or Fisher’s exact test. For quantitative parameters, comparisons among two groups
were made with Student’s t-test, Welch’s t-test or Mann-Whitney U-test, as appropriate,
whereas comparisons among three or more groups were made with ANOVA or Kruskal-Wallis
test, as appropriate. Correlations between two qualitative variables were tested with
Pearson’s correlation coefficient or Spearman's rank correlation coefficient, as indicated.
Kaplan-Meier curves and the log-rank test were used for the assessment and comparison
of overall survival among groups. Cox regression was performed for multivariate survival
analysis. All the tests were two-tailed and the results were considered statistically
significant if the p-value was less than 0.05. The 25th edition of Statistical Package
for Social Sciences (SPSS) (IBM Corporation, Armonk, NY, USA) was used for statistical
analysis.
Results
Demographic and clinical data
Of a total of 45 patients who diagnosed with ACC the majority were female (1.37:1)
with a mean age of 54.64±13.3 years and median age of 55.5 years (IQR 19–77) at diagnosis.
All the resections were R0. There were 2 cases of inoperable ACC, where only biopsy
of the tumor was performed, but these 2 cases were not included in our study. The
surgeon systematically removed the adrenal tumor with the surrounding fat.
In all patients, routine biopsies were taken from the remnant adipose tissue in the
four dimensions. To note, there was a case of laparoscopic adrenalectomy for benign
disease in a patient who presented with ACC at the site of the resection, three years
later. The patient underwent excision of the tumor, upper pole nephrectomy and distal
pancreatectomy. The initial histology slides were reviewed and showed adenoma Weiss
score II.
Tumors were located in the left adrenal gland in 26 (57.7%) patients and in the right
adrenal gland in 19 (42.3%). In our cohort no cases of bilateral ACC, or hereditary
syndrome were identified.
The main clinical symptoms were hormone related (hypercortisolism) while no other
symptoms such as abdominal or flank pain and fatigue were observed. ACC was associated
with hormonal overproduction in 17 patients (37.7%) as follows: cortisol overproduction
in 15 patients (33.3%), and overproduction of more than one hormone (cortisol and
androgens or estrogens) in 2 patients (4.4%). There was no case of aldosterone overproduction.
Clinical follow up was available for 38 out of 45 patients with a median follow up
period of 18 months (IQR 1–96 months). Patients were staged according to the ENSAT
staging system; 3 (6.6%) cases were classified as stage I, 28 (62.3%) as stage II,
12 (26.7%) as stage III and 2 (4.4%) as stage IV ([Table 1]). Seven patients (15.5%) lost to follow up.
From the 12 cases of stage III by the ENSAT classification, only 2 were considered
stage II preoperatively and proved to be infiltrative to surrounding tissues (stage
III) during the operation. There was one recurrence in the tumor bed, treated with
reoperation and one recurrence with multiple bilateral lung metastases treated conservatively.
All the patients received mitotane postoperatively with the exception of two patients
with ACCs smaller than 5 cm diameter. Furthermore, two patients with stage IV and
9 out of 12 patients with stage III received chemotherapy in addition to mitotane.
Sixteen patients have died of their disease 1 to 36 months after diagnosis while the
1-year survival was 60.5% and 5-year survival was 18.4%.
Macroscopic and histopathological characteristics of the ACCs
The characteristic gross features of malignancy (such as hemorrhage and necrosis)
were detected in all cases of ACCs. Normal adrenal tissue was not distinguished in
all the resected samples. The median size of ACCs was 9 cm (IQR 3.5–22 cm) in greatest
dimension and their weight ranged from 18 to 1400 g (median weight of 127 g). The
median size of right-sided lesions was 8.5 cm (range: 5–22 cm) and the median diameter
of left-sided lesions was 8 cm (range: 1.5–18 cm).
Correlation between morphological characteristics of ACCs and Weiss score
The ACC tumor volume and diameter was higher in right-sided lesions reaching statistical
significance (p=0.08 and p=0.07, respectively). Moreover, a significant correlation
between tumor volume (p=0.011, r=0.418), diameter (p=0.005, r=0.449), weight (p=0.04,
r=0.339) and Weiss score was observed.
Immunohistochemical markers of ACCs
The IHC characteristics of the 45 adrenocortical tumors are presented in Supplementary
Table 1 S. The diagnosis of ACC was confirmed histologically by Weiss score>3, with a range
4–8. The mitotic count ranged from 2 to 28 mitoses per 50 high power fields. Twenty-eight
of ACC tumors were surrounded by fibrous capsule while 3 of our cases showed capsular
invasion (no data regarding the presence of fibrous capsule was available in 14 patients’
histological reports). In our series, Ki67 varied from<1% to 75% (median:16.4%).
All except 5 tumors were stained positive for synaptophysin, while nine (20%) tumors
for CKAE1/AE3, 34 (75.5%) tumors for Melan-A, 41 (91.11%) tumors for inhibin A, 37
(82.22%) tumors for vimentin, and 40 (88.88%) tumors for calretinin were positively
stained.
Statistical analysis showed that the expression of Ki-67 was positively correlated
with Weiss score (p=0.25, r=0.383) as it was expected. Our data also provide evidence
for an association between Weiss score and the expression of vimentin and synaptophysin
(p=0.02 and p=0.07, respectively). The correlation between the expression of CKAE1/AE3
as well as Inhibin A and Weiss score shows a trend towards significance (p=0.14 and
p=0.14 respectively). Moreover, the relation of CKAE1/AE3 and Inhibin A expression
with Helsinki score missed statistical significance threshold (p=0.11 and p=0.13,
respectively). P53 overexpression was related with higher Ki-67 (p<0.001) as well
as with higher Helsinki score (p<0.001). No statistical significance was observed
regarding the correlation between IHC markers and ENSAT staging (I/II vs. III/IV).
Correlation of clinical and IHC features with OS in ACC Patients
The median follow-up was 18 months (IQR 1–96). Seven patients (15.5%) lost to follow
up, the 1-year survival was 60.5% and 5-year survival was 18.4%. Survival was significantly
higher in patients with stage I/II as compared to patients with stage III/IV ACC (62±8.7
months vs. 31±12.1 months, p=0.02), as expected.
A larger tumor diameter>10 cm (p=0.007), tumor volume > 500 cm3 (p=0.0003), tumor weight>300 g (p=0.003), Ki-67 index>4% (p=0.04), Weiss score>5
(p=0.01), Helsinki score>8 (p=0.06) were significantly associated with shorter overall
survival (OS) in the univariable analysis. On the contrary, the expression of Melan
A was independently associated with longer OS time (p=0.01). The relation between
average and median survival period and the expression of immunohistochemical markers
in patients with ACC are summarized in [Table 2].
Table 2 Average and median of survival period in patients with ACC in relation to immunohistochemical
markers.
|
Immunohistochemical staining
|
Average period of survival (months)
|
Median Period of survival (months)
|
Standard error
|
95% Confidence interval
|
Log Rank test
|
|
CKAE1/AE3
|
Yes (n=9)
|
61.5
|
NR*
|
15.3
|
31.5–91.5
|
p=0.88
|
|
No (n=28)
|
53.9
|
30
|
8.6
|
37.1–70.7
|
|
Inhibin A
|
Yes (n=35)
|
56.3
|
NR
|
8.1
|
40.4–72.2
|
p=0.17
|
|
No (n=3)
|
13
|
13
|
0
|
13–13
|
|
P53
|
Yes (normal) (n=31)
|
57.5
|
NR
|
8.7
|
40.5–74.5
|
p=0.29
|
|
Overexpression (n=7)
|
24
|
11
|
8.5
|
7.4–40.6
|
|
Melan A
|
Yes (n=32)
|
61.8
|
NR
|
8.7
|
45.2–78.4
|
p=0.01
|
|
No (n=6)
|
15.1
|
15
|
8.5
|
4–26.1
|
|
Synaptophysin
|
Yes (n=33)
|
53.9
|
30
|
8.3
|
37.6–70.3
|
p=0.76
|
|
No (n=5)
|
60
|
NR
|
19.6
|
21.6–98.4
|
|
Vimentin
|
Yes (n=33)
|
57.2
|
NR
|
8.2
|
41.1–73.4
|
p=0.1
|
|
No (n=5)
|
10.7
|
12
|
3.8
|
3.2–18.1
|
|
Ki67
|
≤4% (n=9)
|
81
|
NR
|
13
|
55.5–106.5
|
p=0.04
|
|
>4% (n=29)
|
40.4
|
24
|
7.9
|
24.8–55.9
|
The patients lost to follow-up have been excluded from the analysis. * NR: Not reached yet.
In multivariable analysis, tumor volume>400 cm3 (p=0.046), Weiss score>5 (p=0.007) and overexpression of p53 (p=0.036) were independent
risk factors for shorter survival. In more details, Cox regression analysis showed
that patients with tumor volume (>400 cm3), p53 overexpression or higher Weiss score (>5) seem to imply a higher risk of death
by 11.47, 21.29 and 212.5 times, respectively ([Table 3]).
Table 3 Multivariable Cox-regression analysis in 45 ACC patients.
|
p-Value
|
HR
|
95.0% CI for Exp (B)
|
|
Lower
|
Upper
|
|
Gender (male)
|
0.599
|
0.557
|
0.063
|
4.940
|
|
Age>50
|
0.162
|
0.171
|
0.014
|
2.032
|
|
Stage III/IV
|
0.772
|
0.703
|
0.065
|
7.650
|
|
Volume>400 cm3
|
0.046
|
11.472
|
1.046
|
125.799
|
|
CKAE1/AE3
|
0.098
|
0.073
|
0.003
|
1.614
|
|
p53 overexpression
|
0.036
|
21.287
|
1.214
|
373.182
|
|
Melan A
|
0.054
|
0.103
|
0.010
|
1.040
|
|
Synaptophysin
|
0.984
|
0.954
|
0.011
|
84.638
|
|
Vimentin
|
0.050
|
0.019
|
0.000
|
1.003
|
|
Weiss score>5
|
0.007
|
212.502
|
4.410
|
10 240.462
|
Discussion
This study describes a single center experience in adrenocortical tumors overlooking
a 20-year period. Forty-five ACCs (6%) were diagnosed among 742 patients who underwent
adrenalectomy due to adrenal masses.
ACCs are typically large tumors with diameter>8 cm usually weighing more than 100 g
[23]
[24]. In our series, ACCs weight ranged from 18 to 1400 g (median weight=127 g) and the
median tumor size was 9 cm (IQR 3.5–22 cm) in greatest dimension. Tumor size is a
good predictor of malignancy since tumors with diameter>6 cm have shown a 25% chance
of being malignant compared to 2% of those with a size<4 cm [25]. However, recent studies demonstrated that ACC risk per size in adrenal incidentaloma
(AI) is less than previously reported. In a retrospective review of 2219 AIs, the
optimal cut-off diameter for ACC was 4.6 cm and a combination of thorough hormonal
evaluation and imaging characteristics along with the size of the tumor should better
guide the management of AI [26].
According to our data, tumor volume, diameter, weight were significantly correlated
with Weiss score as well as with ENSAT stage. Moreover, tumor volume>500 cm3 was related with shorter survival rate in the multivariable analysis. In line with
previous reports [27]
[28], we found that most of the ACC cases were women at forth-fifth decade of life, with
stage II (ENSAT criteria) and located on the left, albeit not all studies agree with
the left sided preponderance [29].
We also observed a trend for positive correlation between tumor size (volume and diameter)
and left localization. Moreover, left ACCs had higher Helsinki score, while no correlation
between tumor laterality and Weiss score, ki-67, and stage was indicated.
In line with the literature [30], we found that the mean survival of ACCs with cortisol hypersecretion was shorter
(31.4 months) compared to non-secreted ACCs ( 61.5 months).
The histopathologic diagnosis of ACC is based on the recognition of several morphologic
parameters none of which is in itself pathognomonic for malignancy. Moreover, there
is no single distinctive immunohistochemical marker indicative of malignancy, often
making it necessary to use a wider panel of parameters for confirmation of the diagnosis
[20]. For this reason the Weiss score that combined morphologic and immunohistochemical
criteria was introduced. A Weiss score>3 criteria has a specificity>95% and a sensitivity
of 100% for malignancy in many series [31].
ENSAT has shown that Ki-67 proliferation index is the most powerful prognostic marker
in both localized and advanced ACC indicative of aggressive behavior and that higher
Ki-67 levels are consistently associated with a worse prognosis [9]. Of note, different cut off points for Ki-67 have been suggested as predictive of
overall survival and disease-free survival, by various studies [28]
[32]
[33]
[34], which could be attributed – among others – to differences in the number of patients
and in the follow-up period among studies, as well as possible inter-laboratory variations
in the Ki-67 assessment in ACCs.
In our series, Ki-67 was varied from 1% to 75% and its expression was positively correlated
with Weiss score (p=0.025, r=0.383) and strongly correlated with Helsinki score (p<0.001,
r=0.979) as expected. Moreover, we found that a Ki-67>4% was significantly associated
with shorter survival in patients with ACC.
The Helsinki score that associates two Weiss criteria and the Ki-67 with a cut-off
value of 8.5, can diagnose metastatic adrenocortical carcinoma with 100% sensitivity
and 99.4% specificity [35]; it also is correlated to survival in patients with ACC [35]. In our series Weiss score>5 (p=0.01) and Helsinki score>8 (p=0.06) were significantly
associated with lower overall survival, in the univariable analysis.
In an attempt to find predictive and prognostic immunohistochemical markers of adrenocortical
carcinoma, we proceed to investigate the expression of several markers including melan-A,
synaptophysin, calretinin, inhibin-A and to correlate it with Ki-67, Weiss score,
Helsinki score and overall survival [36].
Inhibin-A was positive in 95.5% of our cases and its correlation with Weiss score
showed a trend towards significance. In consistent with our findings, Weissferdt et
al. showed largely diffuse cytoplasmic reactivity of inhibin-A in 92.5% of cases of
ACC [37].
Melan A (MART-1) mainly expressed in malignant melanomas, is also a widely used biomarker
for adrenal cortical and other streoidogenic tissues with high specificity and sensitivity.
We found positive melan-A expression in 75.5% of ACCs. In line, small series with
10 and 30 ACCs have shown positive reactivity to melan-A in 60 and 97%, respectively
[28]
[38]. Melan-A expression was not related with stage, Ki-67, Weiss score, Helsinki score,
in our cases.
Calretinin appears to be a stable marker for ACC and reported series show reactivity
in 70–98% of cases [37]
[39]. In our series, nuclear and cytoplasmic staining was reported in 88.8% of cases.
No correlation between calretinin and stage, Weiss score, Helsinki score and Ki-67
was observed.
We also found a positive synaptophysin staining in 88.6% of ACC cases, while a marginal
positive association between Weiss score and the expression of synaptophysin was observed.
A positive correlation between vimentin expression and Weiss score was also established,
according to our data. Previous study by Satelli et al., reported that vimentin-positive
tumor cells were detected in 73.2% of ACC, while only 14.0% of adrenal cortical adenoma
showed scattered vimentin-positive cells [40].
Vimentin was previously considered as a marker for epithelial-mesenchymal transition
(EMT), however its active role in cancer cell dissemination has recently been recognized
[41]. In a case report by Bulzico et al. on an aggressive metastatic ACC (Weiss-score
of 9), the staining for vimentin showed an intense diffuse cytoplasmic immunohistochemical
pattern [42]. The strong expression of mesenchymal proteins such as vimentin by tumoral cells
could possibly lead to the acquisition of properties such as migration, invasive,
and apoptosis resistance [42]
[43]
[44]. Of note, Sbiera et al. showed that in ACC, there is no indication of EMT as all
adrenocortical tissues lacked expression of epithelial markers and exhibited closer
similarity to mesenchymal tissues. However, higher SLUG expression (mesenchymal marker)
was associated with clinical markers indicating aggressiveness [45].
Of interest, among the studied immunohistochemical markers, only melan-A expression
was independently related with longer overall survival in the univariate analysis
of our series. To note, lack of melan-A immunostaining has been associated with reduced
disease-free and overall survival in melanomas [46]. Zlatibor et al. in a small series of ACC, reported that negative staining for inhibin
A and synaptophysin was significantly associated with shorter survival, which was
not confirmed in our study [28]. Among the main genes have been implicated as tumor drivers in sporadic ACC, is
TP53 [47]. In the context of adult ACC the TP53 somatic mutations are presented in 16 and
21% according to two recent genome-wide studies [48]
[49].
The majority of the studies [50]
[51], albeit not all [52], have shown that p53 somatic mutations are prognostic factor and correlated with
more advanced stages, with shorter disease free survival and poor outcome in ACCs.
In our IHC analysis, we found no relation of p53 overexpression with Weiss score,
as well as disease stage. However, there was a strong relation between p53 overexpression
and higher ki-67, as well as higher Helsinki score. More interestingly, in multivariable
analysis overexpression of p53 was independent risk factor for shorter survival (p=0.036),
implying a higher risk of death by 21.29 times.
In summary, our data reveal that expression of melan-A and lower expression of Ki-67
(≤4) were independently associated with longer OS time. In multivariable analysis,
tumor volume, Weiss score>5 and overexpression of p53 were independent risk factors
for shorter survival.
A limitation of our study is the small number of participants however adrenocortical
carcinoma is a rare cancer. Moreover, the follow up period is not long enough, therefore
in some cases the median survival time have not been reached yet. Finally, we did
not include the inoperable adrenocortical carcinomas.
Strength of our study is that the surgical team as well as the diagnostic pathologists
did not change during the period of 20 years that the cases have been collected.
It is certain that a combination of markers for a more accurate differential diagnosis
is still needed for the appropriate assessment of the adrenal tumors. The appropriate
combination of clinical, histopathological and molecular parameters will improve in
the future the prognostication of patients with ACC and will open the road for personalized
management of these patients. The accurate diagnosis would probably protect the patients
with a better outcome from an aggressive management after complete resection of the
tumor with avoidance of heavy adjuvant treatment, such as mitotane and cytotoxic chemotherapy.
It would also reveal the patients with an intermediate or poor outcome that need a
more intensive follow up strategy and the use of more aggressive systemic adjuvant
therapies.
Until then, the most important factors that determine long-term prognosis of ACC remain
the disease stage at diagnosis, the Weiss score, and the Ki67 index. Immunohistochemical
markers such as Melan A could serve as prognostic factor, however further studies
are needed to prove its usefulness.
