ELSEVIER

NOVEMBER 2018 ===== Surgery

journal homepage: www.elsevier.com/locate/surg

SURGERY

-

===

The effect of hormonal secretion on survival in adrenocortical carcinoma: A multi-center study

Check for updates

Alaa Sada, MD, MSª, Trenton R. Foster, MDª, Ruaa Al-Ward, MDb, Sahar Sawani, MDC, HElaine Charchar, MDª, Reza Pishdad, MDe, Anat Ben-Shlomo, MDf, Benzon M. Dy, MDª, Melanie L. Lyden, MDª, Emily Bergsland, MD8, Sina Jasim, MD, MPHh, Nitya Raj, MDi, Jessica B. Shank, MD), Oksana Hamidi, DOK, Amir H. Hamrahian, MDe, José L. Chambô, MD1, Victor Srougi, MD1,m, Maria CBV. Fragoso, MD, PhDd,n, Paul H. Graham, MDº, Mouhammed Amir Habra, MDb, Irina Bancos, MDP,*,*, Travis J. Mckenzie, MDa,*,*

ª Department of Surgery, Mayo Clinic, Rochester, MN

b Department of Endocrine Neoplasia and Hormonal Disorders, The University of Texas MD Anderson Cancer Center, Houston, TX

Department of Medicine, Baylor College of Medicine, Houston, TX

d Unidade de Suprarrenal, Hospital das Clinicas da Faculdade de Medicina da Universidade de São Paulo, São Paulo, Brazil

e Division of Endocrinology, Diabetes and Metabolism, Johns Hopkins University, Baltimore, MD

f Adrenal Program, Division of Endocrinology, Diabetes, and Metabolism, Cedars Sinai Medical Center, Los Angeles, CA

% Department of Medicine, University of California San Francisco, San Francisco, CA

h Division of Endocrinology, Metabolism and Lipid Research, Washington University in St. Louis, Saint Louis, MO

Department of Medicine, Memorial Sloan Kettering Cancer Center, New York, NY

¿ Division of Surgical Oncology, University of Nebraska Medical Center, Omaha, NE

k Division of Endocrinology and Metabolism, University of Texas Southwestern Medical Center, Dallas, TX

1 Division of Urology, Hospital das Clínicas da Faculdade de Medicina da Universidade de São Paulo, São Paulo, Brazil

m Division of Urology, Hospital Moriah, São Paulo, Brazil

n Instituto do Câncer do Estado de São Paulo da Faculdade de Medicina da Universidade de São Paulo, São Paulo, Brazil

° Department of Surgical Oncology, The University of Texas MD Anderson Cancer Center, Houston, TX

P Division of Endocrinology, Mayo Clinic, Rochester, MN

ARTICLE INFO

Article history: Accepted 27 April 2023 Available online 8 November 2023

ABSTRACT

Background: Current evidence suggests that cortisol secreting adrenocortical carcinoma has worse prognosis compared to non-secreting adrenocortical carcinoma. However, the effect of other secretory subtypes is unknown.

Methods: This multicenter study within the American-Australian-Asian Adrenal Alliance included adults with adrenocortical carcinoma (1997-2020). We compared overall survival and disease-free survival among cortisol secreting, mixed cortisol/androgen secreting, androgen secreting, and non-secreting adrenocortical carcinoma.

Results: Of the 807 patients (mean age 50), 719 included in the secretory subtype analysis: 24.5% were cortisol secreting, 13% androgen secreting, 28% mixed cortisol/androgen, 32.5% non-secreting, and 2% were mineralocorticoid secreting.

Median overall survival and disease-free survival for the entire cohort were 60 and 9 months, respec- tively. Median overall survival was 36 months for cortisol, 30 for mixed, 60 for androgen secreting, and 115 for non-secreting adrenocortical carcinoma, P < . 01. Median disease-free survival was 7 months for cortisol, 8 for mixed, 10 for androgen, and 12 for non-secreting adrenocortical carcinoma, P = . 06.

On multivariable analysis of age, sex, Ki67%, secretory subtype, stage, resection, and adjuvant therapy, predictors of worse overall survival were older age, higher Ki67%, stage IV, mixed secreting, R1, and no

The 43rd Annual Meeting of the American Association of Endocrine Surgeons in Birmingham, Alabama from April 29 to May 1, 2023.

* Reprint requests: TJ Mckenzie, MD, Department of Surgery, Mayo Clinic, 200 1st ST SW, Rochester, MN 55905.

* Irina Bancos, MD, Division of Endocrinology, Mayo Clinic, Rochester, MN, 200 1st ST SW, Rochester, MN 55905. E-mail addresses: Mckenzie.Travis@mayo.edu (T.J. Mckenzie), Bancos.Irina@mayo.edu (I. Bancos);

Twitter: @Aabdusada, @rpishdad, @Benzon_dy, @Sina_jasim, @Fragoso_mc, @irinabancos, @theaaes, @A5_adrenal, @mayoclinicendo, @mayoclinicsurg, @endocrinofmusp, @washuendo, @endohopkins

Dr. Irina Bancos and Dr. Travis J. Mckenzie share the last author position.

https://doi.org/10.1016/j.surg.2023.04.070 0039-6060/ 2023 Elsevier Inc. All rights reserved.

adjuvant therapy, P < .05. On subgroup analysis of R0 resection, predictors of worse overall survival included older age and higher Ki67%. Ki67% >40, stage III and cortisol secretion were associated with worse disease-free survival.

Conclusion: Mixed cortisol/androgen secreting adrenocortical carcinoma was associated with worse overall survival, while cortisol or androgen secreting alone were not. Notably, among patients after R0 resection, secretory subtype did not affect overall survival. Cortisol secreting adrenocortical carcinoma demonstrated worse disease-free survival. Ki67% remained a strong predictor of worse overall survival and disease-free survival independent of stage.

@ 2023 Elsevier Inc. All rights reserved.

Introduction

Adrenocortical carcinoma (ACC) is a malignant neoplasm arising from the adrenal cortex with an estimated incidence of 0.5 to 2 cases per million person years.1,2 In addition to being rare, ACC is associated with poor outcomes.3,4 For decades, limited progress has been made to improve outcomes of patients with ACC. A population-based study using the Surveillance, Epidemiology, and End Results Registry (SEER) evaluated patients with ACC between 1973 and 2014 and found that survival of patients with ACC not only continue to be poor but also have been declining.1

Several factors are linked to poor survival of patients with ACC including advanced stage at presentation, incomplete surgical resection, and higher tumor grades or high Ki67%.5-7 The effect of various hormone secreting ACC on its presentation, characteristics and survival is not well studied. Some studies suggest that cortisol- secreting ACC is associated with worse survival compared to non- cortisol secreting ACC.2,6,8 However, the effect of other secretory subtypes of ACC on outcomes is poorly understood given the rarity of disease and as large cancer databases lack the data about ACC secretory subtypes. In an effort to better understand the clinical behavior, tumor characteristics and survival outcomes between different secretory subtypes of ACC, we conducted a multicenter study within the American-Australian-Asian Adrenal Alliance (A5) to evaluate the clinical behavior of different secretory subtypes of adrenocortical carcinomas.

Methods

Study design

We conducted a retrospective multicenter study including par- ticipants from 10 academic tertiary centers within the American- Australian-Asian Adrenal Alliance (A5) research collaboration. Local Institutional Review Board (IRB) approval was obtained by each center. Informed consents were either obtained or waived based on each institution’s IRB protocols and ethical committees.

Adrenocortical carcinoma patients were included in this study if they meet the following criteria:

- Age ≥ 18 years

- Initial ACC diagnosis between January 1,1997 and December 31, 2020

- Confirmed ACC diagnosis either with (1) biopsy of primary tu- mor, (2) biopsy of metastasis, (3) surgical pathology, or (4) locally invasive or metastatic adrenal mass with clear hormone excess without biopsy

Baseline characteristics were reported at the time of initial diagnosis. Tumor stage was reported based on the eighth edition of the TNM staging at the time of initial diagnosis. Tumor size was reported based on the greatest tumor dimension on final pathology for patients who underwent surgery or based on the greatest

dimension on diagnostic cross-sectional images for patients who did not undergo surgery. Surgical resection was categorized as: R0, R1, R2, Rx. R0: complete resection with negative margins, R1: positive microscopic margins, R2: positive macroscopic margins or unresected distant metastatic disease, Rx: margins status is un- known. Regional lymph node status (N) at the time of initial diag- nosis was categorized as Nx if regional lymph nodes could not be assessed, N0 if regional lymph nodes were examined and negative, and N1 if regional lymph nodes were involved with metastatic tumor. Distant metastasis (M) at the time of initial diagnosis was classified as no distant metastasis (M0), distant metastasis (M1), or unknown distant metastatic status (Mx).

Neoadjuvant therapy was categorized as receiving either chemotherapy, radiation therapy, or a combination of both (chemotherapy and radiation therapy) in the neoadjuvant setting for those who underwent surgery. Adjuvant treatment was cate- gorized as receiving any of the following treatments in the adjuvant setting: mitotane only, systematic chemotherapy without mito- tane, both systematic chemotherapy and mitotane, mitotane and tumor bed external beam radiation, systematic chemotherapy and tumor bed external beam radiation, systematic chemotherapy and mitotane and tumor bed external beam radiation, tumor bed external beam radiation only, or other.

Recurrence was defined as recurrence after the initial tumor resection and was categorized as local recurrence only, distant recurrence only or simultaneous local and distant recurrence. Disease-free survival was defined as interval between date of initial surgical resection and date of first recurrence for TNM stages I/II and III who underwent R0 resection. Overall survival was defined as the interval between date of diagnosis and date of last contact or death. Mineralocorticoid excess group and male patients with elevated estradiol levels were excluded from Kaplan survival analysis and multivariable given the limited number of cases but they were reported in the overall cohort description.

Definitions

Given that the hormonal evaluation protocols differed by cen- ters, we reported laboratory results obtained at the time of initial ACC diagnosis as: increased, normal, or not assessed based on each institution testing protocols and reference ranges. We did not specify normal values for any hormone so elevated values were determined by each center according to their protocols. Hyper- cortisolism was confirmed through the use of late-night salivary cortisol, 24-hour urinary free cortisol (UFC) excretion, or the overnight 1 mg dexamethasone suppression test (DST), in accor- dance with the Endocrine Society Clinical Practice Guidelines. These results were reported as a categorical variable (increased, normal, or not assessed).9

The following variables were collected: cortisol, aldosterone, de- hydroepiandrosterone sulfate (DHEAS), androstenedione, testos- terone, 17-hydroxy-progesterone (17OH-prog), 17-hydroxy- pregnenolone (17OH-preg), 11-deoxycortisol, 11 deoxycorticosterone,

and estradiol. Elevated precursors referred to an elevation of any of the following (17OH-preg, 17OH-prog, or 11-deoxycortisol). Miner- alocorticoid excess was defined by an elevation in aldosterone or 11- deoxycorticosterone with suppressed plasma renin activity. Plasma renin activity results were reported from one center.

Hyperandrogenism referred to an elevation of any of the following (DHEAS, testosterone, or androstenedione). To be included in the secretory subtypes, all patients had at least available cortisol levels as part of the hormonal evaluation.

In addition to hormonal evaluation, the presence of the following hormonal excess symptoms was also reported at the time of initial diagnosis as (yes, no, or unknown) based on each center clinical evaluation: Cushing syndrome, androgen excess symptoms in women, mineralocorticoid excess symptoms, and estrogen excess symptoms in men.

We categorized hormonal secretory subtypes into 5 groups: cortisol secreting (group C), mixed cortisol and androgen secreting (group A/C), androgen secreting (group A), mineralocorticoid secreting (group M), and non-secreting (group NS). Patients with mineralocorticoid excess were excluded if they did not have available plasma renin activity levels.

Overt hypercortisolism was defined by the presence of elevated cortisol levels and Cushing syndrome at the time of initial diagnosis while mild hypercortisolism was defined by elevated cortisol levels in the absence of Cushing syndrome.

Cortisol secreting (group C)

Patients were categorized into this group if they had any of the following:

- Elevated cortisol

- Elevated cortisol and any precursor (17OH preg, 17OH prog, or 11-deoxycortisol)

- Elevated cortisol and mineralocorticoid

- Elevated cortisol, mineralocorticoid, and any precursor

- Elevated cortisol and mineralocorticoid but were not tested for androgen

- Elevated cortisol but were not tested for androgen or mineral- ocorticoid elevation

Mixed cortisol and androgen secreting (group A/C)

Patients were categorized into this group if they had any of the following:

- Elevated cortisol and androgen levels (DHEAS, testosterone, or androstenedione) with or without elevated any precursor

- Male patients with elevated cortisol, androgen, and estradiol levels

- Elevated cortisol, mineralocorticoid, any precursor, and androgen

- Elevated cortisol and androgen but were not tested for mineralocorticoid

Androgen secreting (group A)

Patients were categorized into this group if they had any of the following:

- Hyperandrogenism (elevation in DHEAS, testosterone, or an- drostenedione) with normal cortisol and mineralocorticoid levels

- Hyperandrogenism with normal cortisol levels but were not tested for mineralocorticoids

Mineralocorticoid secreting (group M)

Patients were categorized into this group if they had any of the following:

- Elevated aldosterone or 11-deoxycorticosterone, suppressed plasma renin activity, and normal cortisol levels

- Elevated aldosterone or 11-deoxycorticosterone, elevated pre- cursors, suppressed plasma renin activity and normal cortisol levels

Non-Secretory ACC (group NS)

Patients were categorized into this group if they did not have an elevation of any of the previously listed variables except precursors as they can have normal or elevated precursors.

Statistical analysis

Variables were analyzed using x2, Fisher exact, and Wilcoxon rank sum tests. Kaplan-Meier analysis along with log-rank test and multivariable Cox proportional hazards regression models were used to assess disease-free survival and overall survival. In the multi- variable analysis (Cox proportional hazards regression models), age was analyzed as per 10-year increase while Ki67% was catego- rized into four groups as <10, ≥10 to <20, >20 to <40, and ≥40.

Analysis was completed using JMP, version 16.2.0. SAS Institute, Inc, Cary, NC, 1989-2021.

Results

Overall cohort

A total of 807 patients were identified: the mean age was 50 ± 15 years, and 505 (63%) were female. Androgen levels at the time of diagnosis were available for 612 patients (76%). On the other hand, cortisol levels at the time of initial diagnosis were available for 737 (91%) of patients. In patients with hypercortisolism and available data about their hypercortisolism symptoms, 89 patients had mild hypercortisolism and 277 had overt hypercortisolism.

Other laboratory results for the entire cohort are summarized in Table I.

At the time of initial diagnosis, 53 (6.5%) patients presented with TNM stage I, 232 (29%) stage II, 237 (29%) stage III, 248 (31%) stage IV, while the stage was not available for 37 (4.5%). Tumor size was ≤5 cm in 14 patients with TNM stage III, 11 patients with TNM stage IV, 1 patient with unknown stage vs 0 patients with TNM stage I or II.

Ki67% results were available for 465 patients. Median (IQR) Ki67% increased by stages as follows: Median (IQR) Ki67% was 6 (4.5,15) for stage I, 15 (7.25, 29) for stage II, 20 (13, 37) for stage III and 30 (15, 50) for stage IV, P < . 01.

Surgical resection was undertaken in 714 (88%) patients and of those 436 (61%) underwent complete R0 resection at their initial resection as shown in Table I. Neoadjuvant treatment was admin- istered to 44 patients (5%): 32 patients received neoadjuvant sys- tematic chemotherapy; 8 patients received preoperative radiation therapy, while 4 patients received both. Of patients who received neoadjuvant treatment, 27 (61%) presented with stage IV, 7 (16%) stage III, 6 (14%) their stage was not available while the rest 4 (9%) had stage I/II. After neoadjuvant treatment, 18 patients underwent

Table I Summary of the entire cohort (N = 807)
All cohort N = 807Men N = 302Women N = 505P value
Age at diagnosis.02
Mean + SD50 ± 1552 ± 1549 ± 16
Median (IQR)52 (38, 62)53 (40, 63)51 (36, 61)
Range18, 8518, 8518, 85
Sex-
Male302 (37%)302 (100%)-
Female505 (63%)-505 (100%)
Laterality.93
Right ACC370 (46%)139 (46%)231 (46%)
Left ACC437 (54%)163 (54%)274 (54%)
Cortisol
Elevated382 (47%)127 (42%)255 (50.5%).01+
Normal355 (44%)139 (46%)216 (42.8%)
Not assessed70 (9%)36 (12%)34 (6.7%)
Aldosterone.34+
Elevated47 (6%)21 (7%)26 (5%)
Normal528 (65%)189 (62.5%)339 (67%)
Not assessed232 (29%)92 (30.5%)140 (28%)
11-deoxycorticosterone.82
Elevated9 (1%)4 (1.3%)5 (1%)
Normal9 (1%)4 (1.3%)5 (1%)
Not assessed789 (98%)294 (97.4%)495 (98%)
DHEAS<. 01+
Elevated254 (31%)74 (24.5%)180 (36%)
Normal322 (40%)119 (39.4%)203 (40%)
Not assessed231 (29%)109 (36.1%)122 (24%)
Androstenedione.11+
Elevated106 (13%)30 (10%)76 (15%)
Normal122 (15%)46 (15%)76 (15%)
Not assessed579 (72%)226 (75%)353 (70%)
Testosterone<. 011
Elevated176 (22%)15 (5%)161 (32%)
Normal316 (39%)144 (48%)172 (34%)
Not assessed315 (39%)143 (47%)172 (34%)
17OH-progesterone.331
Elevated66 (8%)21 (7%)45 (9%)
Normal128 (16%)43 (14%)85 (17%)
Not assessed613 (76%)238 (79%)375 (74%)
17OH-pregnenolone.78
Elevated17 (2%)5 (2%)12 (2.4%)
Normal26 (3%)10 (3%)16 (3.1%)
Not assessed764 (95%)287 (95%)477 (94.5%)
11-deoxycortisol.921
Elevated79 (10%)29 (10%)50 (10%)
Normal54 (7%)19 (6%)35 (7%)
Not assessed674 (83%)254 (84%)420 (83%)
Estradiol<. 01+
Elevated41 (5%)25 (8%)16 (3%)
Normal208 (26%)74 (25%)134 (27%)
Not assessed558 (69%)203 (67%)355 (70%)
Tumor size (cm)<. 01*
N795296499
Mean ± SD11.9 ± 5.613 ± 6.211.3 ±5.2
Median (IQR)11 (7.9, 15)11.5 (8, 16)11 (7.5, 14.5)
Range1.5, 38.51.5, 38.51.5, 28
Regional lymph nodes.21+
N0356 (44%)143 (47%)213 (42%)
N166 (8%)27 (9%)39 (8%)
Nx385 (48%)132 (44%)253 (50%)
Distant metastasis.69
M0554 (68.4%)206 (68.2%)348 (69%)
M1248 (31%)95 (31.5%)153 (30%)
Mx5 (0.6%)1 (0.3%)4 (1%)
Single versus multi-metastatic sites (in M1 group).50+
N24093147
Single site114 (47.5%)47 (50.5%)67 (45.6%)
Multi-sites126 (52.5%)46 (49.5%)80 (54.4%)
TNM stage.69
I53 (6.5%)15 (5%)38 (8%)
II232 (29%)86 (28.5%)146 (29%)
III237 (29%)91 (30%)146 (29%)
(continuedon next page)
Table I (continued )
All cohort N = 807Men N = 302Women N = 505P value
IV248 (31%)95 (31.5%)153 (30%)
Unknown37 (4.5%)15 (5%)22 (4%)
Lymphovascular invasion.42
Present336 (42%)134 (44%)202 (40%)
Absent251 (31%)87 (29%)164 (32.5%)
Unknown220 (27%)81 (27%)139 (27.5%)
Ki67%
N465173292.85"
Mean ± SD24.7 ± 19.124.5 ± 1925 ± 19
Median (IQR)20 (10, 39.5)20 (10, 35.5)20 (10, 40)
Range<1, 80<1,80<1, 80
Surgical resection.17
Yes714 (88%)261 (86%)453 (90%)
No93 (12%)41 (14%)52 (10%)
Type of resection.70
N714261453
R0436 (61%)162 (62%)274 (60%)
R1103 (14.4%)41 (16%)62 (14%)
R266 (9.2%)22 (8%)44 (10%)
Rx109 (15.4%)36 (14%)73 (16%)
Neoadjuvant therapy.871
Yes44 (5%)17 (6%)27 (5%)
No763 (95%)285 (94%)478 (95%)
Adjuvant therapy.42
Yes482 (60%)172 (57%)310 (61.4%)
No316 (39%)127 (42%)189 (37.4%)
Unknown9 (1%)3 (1%)6 (1.2%)

* Wilcoxon rank sum test.

* Pearson x2 test.

R0 resection, 6 patients R1 10 had R2, while 10 patients had Rx resection.

A total of 482 (60%) patients received adjuvant treatments: 192 (40%) patients received adjuvant mitotane only, 158 (33%) received adjuvant mitotane and systematic chemotherapy, 52 (11%) patients received mitotane and tumor bed external beam radiation, 30 (6%) patients received systematic chemotherapy without mitotane, 21 (4%) patients had tumor bed external beam radiation only, 21 (4%) patients received a combination of adjuvant mitotane, systematic chemotherapy and tumor bed external beam radiation, while 8 (2%) received adjuvant tumor bed external beam radiation and sys- tematic chemotherapy without mitotane.

Secretory subtypes (groups)

A total of 719 patients were included in the secretory subtype groups: 176 (24.5%) were in group C, 202 (28%) group A/C, 96 (13%) group A, 11 (2%) group M, while 234 (32.5%) were group NS. Baseline characteristics of the secretory subtypes are summarized in Table II. The median follow-up (IQR) duration was 25 (11, 60) months. Various secretory subtypes differed by their presented stage. Patients with non-secretory ACC presented with stages I/II in 47% of cases and with stage IV in 19% while group M presented with stages I/II in 64% of cases and with stage IV in 18%. Other subtypes of ACC were more likely to present with stage IV and less likely to present with stages I/II compared to NS and M groups as shown in Table II. Median (IQR) Ki67% did not differ between groups. The rate of complete surgical resection (R0) was the highest in Group M (89%) followed by group A (70%).

Survival

Overall Survival

The median overall survival (OS) was 60 months as shown in Figure 1. One-year, 2-year, and 5-years survival rates were 83%, 69%,

and 50%, respectively. Median OS was 36 months for group C, 30 months for A/C, 60 months for group A, and 115 for group NS, P < .01. Median OS was worse for patients with overt hypercortisolism (32 months) compared to those with mild (37 months) or no hypercortisolism (97 months), P < . 01 as shown in Figure 1.

On multivariable analysis of age, sex, Ki67%, secretory subtype, TNM stage, resection status, and adjuvant therapy, independent predictors of worse OS were older age, higher Ki67%, stage IV (versus I/II), Group A/C (versus Group NS), R1 (versus R0) and the absence of adjuvant therapy, P < . 05 as shown in Table III. The hazard ratio (HR) of worse overall survival was proportional to the elevation in Ki67% as HR was 5.73 for Ki67% (≥10 to <20), 6.02 for Ki67% (≥20 to <40) and 10.82 for Ki67% (≥40) after adjusting for stage and other important variables.

On subgroup analysis of patients with stages I, II, and III who underwent R0 resection, predictors of worse OS included older age, and higher Ki67% (Table IV). Similar to overall survival in all pa- tients, the hazard ratio of worse overall survival in the R0 group was proportional to the elevation in Ki67% as shown in Table IV.

As group C included some patients who were not tested for hyperandrogenism, we ran the multi-variable analysis models after excluding patients who were not tested for hyperandrogenism, and we found similar results as shown in Tables VI, VII, and VIII in the Appendix.

Disease-free survival

After excluding the mineralocorticoid secreting ACC (group M), the median disease-free survival (DFS) was 9 months as shown in Figure 2. Median DFS was 7 months for group C, 8 months for group A/C, 10 months for group A, and 12 months for group NS, P = . 06. Median DFS was worse for patients with overt hypercortisolism (7 months) compared to those with mild (9 months) or no hyper- cortisolism (12 months), P < . 01 as shown in Figure 2. Appendix Figure 3 is a consort diagram of the OS and DFS included groups.

Table II Characteristics of secretory subtypes
All N = 719Group C N = 176Group A/C N = 202Group A N = 96Group M N = 11Group NS N = 234P value
Age.03*
Mean ± SD50±1549±1648±1647+1544±1452+15
Median (IQR)51 (37, 61)50 (36, 61)50 (36, 61)47 (34, 59)49 (35, 58)52 (42, 63)
Range18, 8518,8518, 8320, 7520, 5918, 85
Sex<. 011
Male255 (35%)69 (39%)54 (27%)26 (27%)8 (73%)98 (42%)
Female464 (65%)107 (61%)148 (73%)70 (73%)3 (27%)136 (58%)
Laterality.951
Right ACC325 (45%)77 (44%)95 (47%)45 (47%)5 (45 %)103 (44%)
Left ACC394 (55%)99 (56%)107 (53%)51 (53%)6 (55%)131 (56%)
Tumor size (cm)<. 01
N7111722029611230
Mean ± SD11.7+5.510.5+5.212.3+5.212.9±5.89.1±4.111.8±5.7
Median (IQR)11 (7.7, 15)9.8 (6.3, 14)11.5 (9, 14.8)12.3 (8.6, 16)9.8 (5.7, 12)11 (8, 15)
Range1.5, 38.51.8, 27.22.2, 343, 294,181.5, 38.5
Regional lymph nodes<. 011
N0325 (45%)62 (35%)87 (43%)49 (51%)3 (27%)124 (53%)
N160 (8%)17 (10%)26 (13%)5 (5%)1 (9%)11 (5%)
Nx334 (47%)97 (55%)89 (44%)42 (44%)7 (64%)99 (42%)
Distant metastasis<. 01
M0490 (68%)115 (65%)113 (56%)65 (68%)9 (82%)188 (80.3%)
M1225 (31%)59 (34%)88 (43.5%)31 (32%)2 (18%)45 (19.2%)
Mx4 (1%)2 (1%)1 (0.5%)0 (0%)0 (0%)1 (0.5%)
Single versus multi-metastatic sites.091
N218598530242
Single site105 (48%)34 (58%)34 (40%)13 (43%)0 (0%)24 (57%)
Multi-sites113 (52%)25 (42%)51 (60%)17 (57%)2 (100%)18 (43%)
TNM stage<. 011
I48 (7%)21 (12%)7 (3%)3 (3.1%)1 (9%)16 (7%)
II214 (30%)38 (21.5%)47 (23%)29 (30.2%)6 (55%)94 (40%)
III202 (28%)51 (29%)52 (26%)29 (30.2%)2 (18%)68 (29%)
IV225 (31%)59 (33.5%)88 (44%)31 (32.3%)2 (18%)45 (19%)
Unknown30 (4%)7 (4%)8 (4%)4 (4.2%)0 (0%)11 (5%)
Lymphovascular invasion.14+
Present289 (40%)70 (40%)89 (44%)43 (45%)3 (27%)84 (36%)
Absent237 (33%)52 (29.5%)56 (28%)29 (30%)5 (46%)95 (41%)
Unknown193 (27%)54 (30.5%)57 (28%)24 (25%)3 (27%)55 (23%)
Ki67%.53*
N424100119642139
Mean ± SD25±1924±1927±2024.5+2025±2123±18
Median (IQR)20 (10, 39)20 (10, 34.7)20 (10, 40)20 (7, 40)25 (10, 40)20 (10, 30)
Range<1,801, 751,80<1,8010,40<1, 79
Surgical resection.151
Yes637 (88.6%)160 (91%)172 (85%)82 (85%)9 (82%)214 (91.5%)
No82 (11.4%)16 (9%)30 (15%)14 (15%)2 (18%)20 (8.5%)
Type of resection.021
637160172829214
R0393 (62%)89 (56%)94 (55%)57 (70%)8 (89%)145 (68%)
R183 (13%)24 (15%)30 (17%)3 (4%)1 (11%)25 (12%)
R264 (10%)16 (10%)22 (13%)11 (13%)0 (0%)15 (7%)
Rx97 (15%)31 (19%)26 (15%)11 (13%)0 (0%)29 (13%)
Neoadjuvant therapy.69
Yes40 (5.6%)10 (6%)14 (7%)6 (6%)0 (0%)10 (4%)
No679 (94.4%)166 (94%)188 (93%)90 (94%)11 (100%)224 (96%)
Adjuvant therapy.051
Yes431 (60%)103 (59%)141 (70%)58 (60%)5 (45.4%)124 (53%)
No280 (39%)71 (40%)58 (29%)37 (39%)6 (54.6%)108 (46%)
Unknown8 (1%)2 (1%)3 (1%)1 (1%)0 (0%)2 (1%)

* Wilcoxon rank sum test.

+ Pearson x2 test.

On multivariable analysis including patients with stages I, II, and III who underwent RO resection, Ki67% >40, stage III and group C were predictors of worse DFS after adjusting for age, sex, secretory subtype, stage, and adjuvant therapy (Table V).

Discussion

Although it is well known that ACC is associated with poor survival, the effect of various hormonal excess on ACC survival is

not well studied. In this multi-institutional study, we compared the different tumor characteristics and outcomes between various secretory subtypes of ACC and found that mixed cortisol/androgen secretion was associated with worse overall survival compared to non-secreting ACC, while cortisol or androgen secretion alone were not after adjusting for clinically relevant variables (age, sex, Ki67%, TNM stage, resection, and adjuvant therapy). Interestingly, among patients with nonmetastatic ACC who underwent R0 resection, the secretory subtype did not affect overall survival while cortisol-

Figure 1. Overall survival.

A

100.0%

B

100.0%

- Group A

… Group C

-Group NS

80.0%

80.0%

-Group A/C

Survival %

60.0%

Survival %

60.0%

4

1

40.0%

7

40.0%

20.0%

20.0%

0.0%

0

12

24

36

48

60

72

84

96

108

120

0.0%

Survival Time (Months)

0

12

24

36

48

60

72

84

96

108

120

Overall Survival Time (Months)

Overall Survival for All

Overall Survival by Secretory Subtypes

C

100.0%

-Mild Hypercortisolism:

-No Hypercortisolism:

··· Overal Hypercortisolism:

80.0%

Survival %

60.0%

40.0%

20.0%

0.0%

0

12

24

36

48

60

72

84

96

108

120

Overall Survival Time (Months)

Overall Survival by Hypercortisolism

Table III Multi-variable analysis including all patients
VariableHR (95% CI)P value
Age (per 10-year increase)1.19 (1.06, 1.33)<. 01
Sex
FemaleRef
Male1.07 (0.75, 1.53).69
Ki67%
<10Ref
≥10 to <205.73 (2.76, 11.85)<. 01
≥20 to <406.02 (2.94, 12.31)<. 01
≥4010.82 (5.22, 22.43)<. 01
Secretory subtype
Group NSRef
Group C1.47 (0.91, 2.38).11
Group A1.01 (0.53, 1.94).97
Group A/C1.76 (1.12, 2.79).01
TNM stage
I/IIRef
III1.35 (0.87, 2.11).18
IV3.66 (2.33, 5.76)<. 01
Resection status
R0Ref
R11.94 (1.20, 3.14)<. 01
R21.53 (0.77, 3.04).22
Rx1.62 (1.04, 2.53).03
Adjuvant therapy
YesRef
No1.56 (1.09, 2.25).01
Unknown4.30 (1.42, 12.96)<. 01

secreting ACC was associated with worse disease-free survival. High Ki67% was a strong predictor of worse outcomes in all patients and in the R0 group.

Several studies indicated that secretory ACC in general and cortisol-secreting ACC in specific are associated with worse out- comes compared to non-secretory ACC.2,8,10-13 However, as secre- tory ACC is a broad term that represents various subtypes of ACC, we elected to subcategorize secretory ACC into different groups to better understand the individual clinical behavior of each group. Contrary to previous studies, our results suggest that cortisol secretion alone was not associated with worse overall survival. However, mixed cortisol and androgen secreting ACC was associ- ated with worse overall survival in comparison to non-secretory neoplasms after adjusting for other clinically relevant variables. This controversy can be explained by the fact that other studies categorized ACC secretion as cortisol secreting or non-cortisol secreting without analyzing cortisol and androgen co-secreting ACC as a separate entity as we did in the current study. A series of 202 patients with ACC demonstrated that cortisol-secretion was associated with worse survival, compared to non-cortisol secreting ACC. However, in that study about half of patients in the cortisol- secretion group had mixed cortisol and androgen secreting ACC.10 Moreover, a systematic review and meta-analysis included 19 studies with a total of 3,814 patients with ACC and found that cortisol-secreting ACC is associated with worse survival. However, all included studies in the meta-analysis did not differentiate be- tween cortisol-secreting versus cortisol and androgen co-secreting tumors.8 To our knowledge, the current study is the first one to find that worse outcomes of ACC is related to cortisol and androgen co- secretion rather than cortisol or androgen secretion alone. While our results did not indicate worse overall survival in the group C, it is important not to overlook the role of hypercortisolism in deter- mining ACC prognosis especially in cases with overt hyper- cortisolism. Several studies have shown that overt hypercortisolism

Table IV Multi-variable analysis of overall survival including TNM stages I, II, and III who had R0 resection
VariableHR (95% CI)P value
Age (per 10-year increase)1.24 (1.01, 1.53).03
Sex
FemaleRef
Male0.96 (0.53, 1.73).90
Ki67%
<10Ref
≥10 to <204.79 (1.63, 14.09)<. 01
≥20 to <406.80 (2.40, 19.30)<. 01
≥4011.06 (3.78, 32.36)<. 01
Secretory subtype
Group NSRef
Group C1.56 (0.74, 3.31).23
Group A0.56 (0.16, 1.88).34
Group A/C1.87 (0.91, 3.81).08
TNM stage
I/IIRef
III1.80 (0.98, 3.31).05
Adjuvant therapy
YesRef
No1.75 (0.94, 3.24).07
Unknown12.67 (2.32, 69.03)<. 01

affects survival and the survival curves in our study support the findings that OS was worse for patients with overt versus mild hypercortisolism. However, in the current study we were not able to adjust for mild versus overt hypercortisolism in the multi- variable analysis models given the limited number of patients.

Interestingly, our results did not show worse overall survival in mixed cortisol and androgen co-secreting ACC when analyzing patients with nonmetastatic disease who underwent complete

resection. This can be explained by the fact that mixed cortisol and androgen secreting ACC were more likely to present at advanced stages; our results suggest that the rate of stage IV in this group was 44% compared to 19% in non-secretory group, and when stage IV cases were excluded from analysis, there was no difference in survival. Furthermore, while secretory ACC can present with different hormonal manifestations that theoretically may lead to diagnosis at earlier stages, this was not the case in our study. Our results indicated that non-secreting ACC were more likely to be diagnosed at stages I/II compared to other secretory subtypes. These findings are in line with other studies demonstrating that secretory ACC were not diagnosed at earlier stages compared to non-secretory tumors. A previous study including 266 patients with ACC found that secretory ACC were diagnosed at TNM stage IV in 41.4% of cases compared to 21.4% of non-secretory tu- mors.2 Another multi-institutional study including 524 patients who underwent R0 resection for ACC found that ACC neoplasms with no overt hypercortisolism presented with stages I/II in 62% of cases compared to 37.5% of ACC neoplasms with overt hypercortisolism.6

In light of the fact that 60% to 70% of ACC cases developed recurrence after complete resection, identifying risk factors asso- ciated with recurrence is critically important as these factors can be taken into consideration when deciding the need for adjuvant treatment.14,15 Our study along with other previous studies showed that Ki67% is a very important prognostic factor of survival and disease free survival.16 Interestingly, while cortisol secretion alone was not associated with worse overall survival, we found that it was associated with worse DFS after adjusting for other important clinical variables. This finding is in contrary to another study including 319 patients from the German ACC registry who under- went complete resection and found that glucocorticoid secretion

A

100.0%

80.0%

Survival %

60.0%

40.0%

20.0%

0.0%

0

12

24

36

48

60

Disease-Free Survival (Months)

Disease-Free Survival (DFS) for All

B

100.0%

- Group A

… Group C

-Group NS

80.0%

-Group A/C

Survival %

60.0%

40.0%

20.0%

0.0%

0

12

24

36

Disease-Free Survival (Months)

DFS by Secretory Subtypes

Figure 2. Disease-free survival.

C

100.0%

-Mild Hypercortisolism

-No Hypercortisolism

--- Overt Hypercortisolism

80.0%

Survival %

60.0%

40.0%

20.0%

0.0%

0

12

24

36

Disease-Free Survival (Months)

DFS by Hypercortisolism

Table V Multi-variable analysis of disease-free survival including TNM stages I, II, and III who had R0 resection
VariableHR (95% CI)P value
Age (per 10-year increase)1.07 (0.91, 1.27).42
Sex
FemaleRef
Male0.97 (0.62, 1.53).92
Ki67%
<10Ref
≥10 to <201.17 (0.53, 2.54).69
≥20 to <401.11 (0.53, 2.31).78
≥403.77 (1.63, 8.70)<. 01
Secretory subtype
Group NSRef
Group C2.09 (1.07, 4.06).02
Group A0.90 (0.37, 2.15).81
Group A/C1.01 (0.56, 1.82).96
TNM stage
I/IIRef
III1.88 (1.10, 3.20).01
Adjuvant therapy
YesRef
No1.33 (0.79, 2.24).27
Unknown4.05 (0.83, 19.7).08

was not associated with higher recurrent risk.16 On the other hand, other studies suggest similar results to ours that cortisol secreting ACC has higher recurrence risk compared to non-cortisol secreting ACC.6,8 A Study including 524 patients with ACC who underwent R0 resection, found that cortisol-secreting ACC was associated with worse DFS. Of note, a limitation of that study is that only symptomatic patients were considered as having hyper- cortisolism.8 We believe that these controversial results about the effect of cortisol-secretion on disease-free survival are mainly related to different utilized definitions of secretory or cortisol- secreting ACC in the literature. Nonetheless, for all types of ACC, Ki67% remains an important prognostic factor for overall and disease-free survival regardless of stage, which highlights the importance of considering Ki67% when choosing subsequent treatments strategy after resection. Notably, we did not find a significant difference in the median Ki67% among different secretory subtypes of ACC but we found a correlation of higher Ki67% with more advanced stages.

Although this is one of the largest studies to evaluate the effect of secretory subtypes of ACC on survival, it has a retro- spective design with several limitations, including selection, and referral bias. Given the multi-institutional nature of the study, the hormonal evaluation protocols, lab assays and reference ranges were not standardized among all patients. Results of mineralocorticoid excess, hyperandrogenism and precursors elevation were not available for all patients and that is mainly because testing for mineralocorticoid excess or hyper- androgenism is not routinely performed in the absence of symptoms or clinical suspicion.17 As patients in group C who were not tested for hyperandrogenism may have had hyper- androgenism that was not diagnosed, we ran the multi-variable analysis models after excluding patients who were not tested for hyperandrogenism and we found similar results as shown in the Appendix (Tables VI, VII, and VIII). The limited comprehensive hormonal evaluation for all patients in this cohort can be associated with selection bias and may have skewed our results. However, contributing centers in the A5 alliance are all tertiary referral centers experienced in diagnosis and managing patients with ACC which suggests that most patients in this cohort received the appropriate hormonal evaluation. Additionally, this

study spans a relatively long period of time which explains the limited data about Ki67% and precursors as those are relatively newer tests. Lastly, while mineralocorticoid secreting and estradiol secreting ACC (in men) are interesting groups to report their survival, we could not include them in the survival analysis given the very limited number of cases.

In conclusion, for all patients with ACC, independent of stage and resection status, mixed cortisol/androgen secretion was associated with worse overall survival, while cortisol or androgen secreting alone were not. Notably, among patients after R0 resection, secretory subtype did not affect overall survival. How- ever, cortisol secreting ACC demonstrated worse DFS after adjusting for clinically relevant variables suggesting that cortisol secretion can affect outcomes of ACC patients. Ki67% remained a strong predictor of worse overall and disease-free survival inde- pendent of stage.

Funding/Support

None declared.

Conflict of interest/Disclosure

Oksana Hamidi reports advisory board participation with Corcept Therapeutics, Strongbridge Pharma, and Recordati Rare Diseases, Amryt Pharma, and Neurocrine Biosciences outside the submitted work. Irina Bancos reports consulting (fee to insti- tution) with Corcept Therapeutics, Recordati Rare Diseases, HRA Pharma, Neurocrine Biosciences, Diurnal, Spruce, Sparrow and Adrenas. Irina Bancos reports funding from Recordati Rare Dis- eases for investigator-initiated project unrelated to this work.

Supplementary materials

Supplementary material associated with this article can be found, in the online version, at https://doi.org/10.1016/j.surg.2023. 04.070.

References

1. Sharma E, Dahal S, Sharma P, et al. The characteristics and trends in adreno- cortical carcinoma: a United States population based study. J Clin Med Res. 2018;10:636.

2. Sada A, Asaad M, Bews KA, et al. Comparison between functional and non- functional adrenocortical carcinoma. Surgery. 2020;167:216-223.

3. Kebebew E, Reiff E, Duh Q-Y, Clark OH, McMillan A. Extent of disease at pre- sentation and outcome for adrenocortical carcinoma: have we made progress? World J Surg. 2006;30:872-878.

4. Sada A, Glasgow AE, Lyden ML, et al. Positive lymph nodes in adrenocortical carcinoma: what does it mean? World J Surg. 2021;45:188-194.

5. Else T, Kim AC, Sabolch A, et al. Adrenocortical carcinoma. Endocr Rev. 2014;35: 282-326.

6. Berruti A, Fassnacht M, Haak H, et al. Prognostic role of overt hypercortisolism in completely operated patients with adrenocortical cancer. Eur Urol. 2014;65: 832-838.

7. Daher M, Varghese J, Gruschkus SK, et al. Temporal trends in outcomes in patients with adrenocortical carcinoma: a multidisciplinary referral-center experience. J Clin Endocrinol Metab. 2022; 107:1239-12346.

8. Vanbrabant T, Fassnacht M, Assie G, Dekkers O. Influence of hormonal func- tional status on survival in adrenocortical carcinoma: systematic review and meta-analysis. Eur J Endocrinol. 2018; 179:429-436.

9. Nieman LK, Biller BM, Findling JW, et al. The diagnosis of Cushing’s syndrome: an endocrine society clinical practice guideline. J Clin Endocrinol Metab. 2008;93:1526-1540.

10. Abiven G, Coste J, Groussin L, et al. Clinical and biological features in the prognosis of adrenocortical cancer: poor outcome of cortisol-secreting tumors in a series of 202 consecutive patients. J Clin Endocrinol Metab. 2006;91: 2650-2655.

11. Ayala-Ramirez M, Jasim S, Feng L, et al. Adrenocortical carcinoma: clinical outcomes and prognosis of 330 patients at a tertiary care center. Eur J Endo- crinol. 2013;169:891.

12. Puglisi S, Perotti P, Pia A, Reimondo G, Terzolo M. Adrenocortical carcinoma with hypercortisolism. Endocrinol Metab Clin. 2018;47: 395-407.

13. Srougi V, Bancos I, Daher M, et al. Cytoreductive surgery of the primary tumor in metastatic adrenocortical carcinoma: impact on patients’ survival. J Clin Endocrinol Metab. 2021;107:964-971.

14. Glenn JA, Else T, Hughes DT, et al. Longitudinal patterns of recurre- nce in patients with adrenocortical carcinoma. Surgery. 2019;165: 186-195.

15. Erdogan I, Deutschbein T, Jurowich C, et al. The role of surgery in the man- agement of recurrent adrenocortical carcinoma. J Clin Endocrinol Metab. 2013;98:181-191.

16. Beuschlein F, Weigel J, Saeger W, et al. Major prognostic role of Ki67 in localized adrenocortical carcinoma after complete resection. J Clin Endocrinol Metab. 2015;100:841-849.

17. Zeiger MA, Thompson GB, Duh Q-Y, et al. American Association of Clinical Endo- crinologists and American Association of Endocrine Surgeons medical guidelines for the management of adrenal incidentalomas. Endocr Pract. 2009;15:1-20.

Discussion

Check for updates

Dr Matthew Nehs (Boston, MA): These tumors are highly glycolytic. The most aggressive subtypes may be producing cortisol to augment their growth. Did you look at glucose levels? Did you look at Hg A1C, or the presence of diabetes amongst the different hormone secreting subtypes? If so, was there a difference in the groups?

Dr Alaa Sada: Unfortunately, we did not look at glucose levels, HgA1c or diabetes but I agree with you. That is an interesting point and something we should consider for future studies.

Dr Heather Wachtel (Philadelphia, PA): Beautifully presented and this study attests to the power of a collaborative multi-insti- tutional study. You were able to obtain more granular data that

you would not be able to retrieve from a national registry. My question, are these tumors creating a cortisol rich microenviron- ment that is immunosuppressive, which allows the tumors to escape the body’s immune response? Did you look at therapies targeting the cortisol excess and if that impacted your disease free or overall survival?

Dr Alaa Sada: We did not specifically investigate targeted therapies. Instead, we categorized treatment effects into different groups: mitotane only, mitotane + chemotherapy, with or without radiation. We found no significant differences between these groups, but we refrained from delving into further detail due to the limited sample size.

O N THE MOVE?

Send us your new address at least six weeks ahead

Don’t miss a single issue of the journal! To ensure prompt service when you change your address, please photocopy and complete the form below.

Please send your change of address notification at least six weeks before your move to ensure continued service. We regret we cannot guarantee replacement of issues missed due to late notification.

JOURNAL TITLE: Fill in the title of the journal here.

OLD ADDRESS:

Affix the address label from a recent issue of the journal here.

NEW ADDRESS:

Clearly print your new address here.

Name

Address

City/State/ZIP

COPY AND MAIL THIS FORM TO:OR FAX TO:OR PHONE:
Elsevier Health Sciences Division314-447-8029800-654-2452
Subscription Customer ServiceOR E-MAIL: JournalsCustomerService-usa@elsevier.comOutside the U.S., call
3521 Riverport Lane Maryland Heights, MO 63043314-447-8871