EJE
Agouti-related peptide: a promising prognostic biomarker in adrenocortical carcinoma. Results from a large single-centre cross-sectional analysis
Mario Detomas, 1,*[D Lisa R. Kagan,1 James F.H. Pittaway,2 Marc P. Schauer,1 Simon Kloock,10 Niklas Geiger,1 Otilia Kimpel,10 Martin Fassnacht, 1,3D Barbara Altieri, 1,4[D and Ulrich Dischinger1
1Department of Internal Medicine I, Division of Endocrinology & Diabetes, University Hospital Würzburg, Würzburg 97080, Germany 2Centre for Endocrinology, William Harvey Research Institute, Barts and the London School of Medicine and Dentistry, Queen Mary University of London, London EC1M 6BQ, United Kingdom
3Comprehensive Cancer Center Mainfranken, University of Würzburg, Würzburg 97080, Germany
4Bavarian Cancer Research Center (BZKF), University Hospital of Würzburg, Würzburg 97080, Germany
*Corresponding author: Department of Internal Medicine I, Division of Endocrinology & Diabetes, University Hospital Würzburg, Oberdürrbacher Straße 6, Würzburg 97080, Germany. Email: detomas_m@ukw.de
Abstract
Background: Adrenocortical carcinoma (ACC) is a rare but aggressive malignancy with a heterogeneous prognosis. To date, there are no reliable circulating biomarker to predict clinical outcomes in ACC. Agouti-related peptide (AgRP), an orexigenic hypothalamic neuropeptide, is also produced by the adrenal gland. However, its expression in adrenal tumours remains unexplored.
Objective: To evaluate AgRP levels in tissue and plasma of patients with ACC, benign adrenal tumours (BAT), and other malignancies. Methods: AgRP mRNA expression was assessed in silico, and protein levels were analysed by immunohistochemistry in ACC and BAT. Plasma AgRP (pAgRP) was measured using ELISA in patients with ACC, BAT, and other cancers. Tumour burden of ACC was quantified using RECIST1.1.
Results: AgRP mRNA levels were higher in ACC than in BAT (3.0 vs 2.4, P <. 0001). AgRP tissue expression was higher in ACC compared to BAT (H-score 120 [49] vs 80 [48], P <. 0001). pAgRP was more elevated in ACC than BAT and other malignancies (1.3 vs 0.3 vs 0.2 ng/ml, all P <. 0001) and was higher in ACC with high tumour burden than those with low and those ACC-free after surgery (2.7 vs 0.6 vs 0.3 ng/ml, P <. 0001). pAgRP increased with the number of metastatic sites (24 sites: 2.7 ng/ml vs 2-3: 1.5 ng/mL vs 1: 0.5 ng/ml, P <. 05). In multivariable Cox regression, elevated pAgRP was independently associated with worse progression-free (HR = 2.654) and overall survival (HR = 2.624) (both P <. 05).
Conclusions: AgRP is more highly expressed in both tissue and plasma of ACC compared to BAT and other cancers. Its strong association with tumour burden and prognosis highlights its potential as biomarker in ACC.
Keywords: survival, cancer, adrenal, Cushing’s syndrome, AgRP
Significance
Agouti-related peptide (AgRP) is markedly higher in both tissue and plasma of adrenocortical carcinoma (ACC) than in be- nign adrenal tumours and other cancers. Circulating AgRP tracks tumour burden, falls after curative surgery, and independ- ently predicts progression-free and overall survival beyond established factors. These data position AgRP as a practical, minimally invasive biomarker for differential diagnosis and risk stratification in ACC, with potential utility for monitoring disease course and treatment response. If validated prospectively, AgRP could complement current clinicopathological mod- els to improve prognostication and guide management in ACC.
Introduction
Adrenocortical carcinoma (ACC) is a rare but aggressive ma- lignant adrenal tumour with an annual incidence of 0.7-2.0 cases per million people.1,2 Early and accurate diagnosis of ACC is crucial, as delays can lead to poor outcomes, with a 5-year overall survival rate of <15% in advanced cases.2 The rarity of ACC and its individual manifestation make
diagnosis challenging; while over 50% of cases are associated with hypercortisolism and hyperandrogenism, about 10% progress silently,3,4 complicating early detection.2 Distin- guishing ACC from benign adrenal tumours, especially adre- nocortical adenoma (ACA), remains a critical diagnostic issue. Although imaging techniques represent important tools in differentiating ACC from ACA, they are still not as reliable
as postoperative histological analysis (including Weiss score and Ki67 index).2,5,6
Recent studies have identified several immunohistochemis- try markers, such as FSCN1, Livin/BIRC7, SF-1, VAV2, and TOP2A,7-11 that are characteristic of ACC. However, none of them is sufficiently validated to be introduced in standard pathological work-up mostly because they lacked robust val- idation with larger cohorts and they offered no clear advan- tage over established markers (like, eg, Ki-67). Moreover, while recent studies have shown that DLK-1 is expressed more in tissue and elevated in blood of patients with ACC than in those with benign adrenal lesions,12 high expression has been also reported in other types of cancer,13 limiting its specificity.
Agouti-related peptide (AgRP) is a neuropeptide that acts as an inverse agonist at the melanocortin receptor 4 (MC4R), mediating orexigenic effects.14,15 Although mainly produced in the hypothalamus, AgRP is significantly expressed in the ad- renal medulla (where it is involved in the sympathetic response to fasting)16 and in the zona fasciculata and glomerulosa of the adrenal cortex, where it acts as presumably a regulator of cell growth and steroidogenesis.17,18 The link between AgRP and steroidogenesis is not just observed at the level of the adrenal gland. In fact, glucocorticoids modulate AgRP synthesis and release from the hypothalamus, as AgRP neurons express glucocorticoid receptors and response elements, indicating a feedback mechanism.19,20 For these reasons, AgRP has been studied already in the context of hypercortisolism,20-22 but never in ACC.
The aim of this study were (2) to compare tissue and plasma AgRP levels between ACC and other adrenal lesions, (2) to evaluate the association of AgRP with tumour burden, and (3) to assess its prognostic value in ACC.
Materials and methods
Study design and population
Cross-sectional analysis of patients with ACC treated at the University Hospital of Würzburg from January 1, 2009 until January 31, 2024. Inclusion criteria for ACC patients were as follows: (1) age over 18 years; (2) histologically confirmed diagnosis of ACC; (3) available clinical, radiological, histo- logical data at time of diagnosis and/or during follow-up; and (4) access to blood samples collected in a fasting state of at least 6 hours. Disease status and survival information were gathered until September 15, 2024. Fasting blood samples of patients with benign adrenal tumours and other (non-adrenal) malignancies were used as controls. All patients provided written informed consent for collecting blood, tissue samples, and clinical data, including follow-up and survival data. All patients with ACC and benign adrenal lesions were included in the ENSAT registry (approved by the ethic com- mittee of the University Hospital of Würzburg, #88/11). As a control group, 15 patients with non-adrenal malignancies were enrolled. Ten of them had a thyroid cancer (medullary thyroid cancer n = 5, anaplastic thyroid cancer n = 2, follicu- lar thyroid cancer n = 1, poorly differentiated thyroid cancer n = 1), 5 had a small cell lung cancer, and 1 had parathyroid cancer. Of note, all of them had presence of metastasis at the time of the blood analysis. All of these patients were included in the Thyroid and Parathyroid Tumors Registry (96/13) and the Network of Excellence for Neuroendocrine Tumors, NeoExNET, Registry (85/12).
This study was conducted and reported in accordance with the REMARK (REporting recommendations for tumor MARKer prognostic studies) guidelines to ensure transparency, reproducibility, and completeness in biomarker research.23
This study was conducted in accordance with the Declaration of Helsinki.
In silico analysis and immunohistochemistry
In order to evaluate the presence of AgRP in the adrenal tissue, in silico analysis was performed using data from Affymetrix Human Genome U133 Plus 2.0 oligonucleotide arrays from Giordano and colleagues.24
Chromogenic immunohistochemical (IHC) staining on hu- man formalin-fixed, paraffin-embedded (FFPE) tumour slides was performed as previously reported.25 The primary anti- body that has been used was AgRP (1:2000, anti-rabbit, #HPA041017, Sigma). The N-Universal Negative Control Anti-Rabbit (IS600, respectively, Dako, Glostrup, Denmark) was used for the negative control. Primary antibodies were in- cubated for 1 hour at room temperature. Microscopic assess- ment and H-score evaluation was performed using Aperio Leica scanner and Aperio Leica ImageScope software analysis (Leica).
Measurement of plasma AgRP
Plasma AgRP was quantified using the Phoenix Pharmaceuti- cals AgRP (83-132) Human EIA (catalog EK-003-53) accord- ing to the manufacturer’s instructions. Analytical sensitivity (limit of detection) was 0.09 ng/ml with a linear reportable range of 0.09-0.61 ng/ml, intra-assay CV <10%, and inter- assay CV <15%.
EDTA plasma was drawn after ≥6 hours fasting and stored at -20 ℃ until analysis. After an incubation time of 2 hours at room temperature on a microplate shaker, plates were washed 4 times with EIA Assay buffer. Streptavidin-horseradish perox- idase (SA-HRP), which catalyses the tetramethylbenzidine (TMB) substrate solution, was added, incubated for 1 hour at room temperature on a microplate shaker, and washed 4 times before adding TMB. The reaction was stopped by 2N HCI. Two positive controls were used on each plate. Samples and stand- ards were run in duplicate with plate controls. All used deter- gents were provided with the kit. The intensity of absorbance was measured by Victor microplate reader (PerkinElmer 2030, Waltham, USA) at 450 nm. Levels of AgRP were extrapo- lated from AgRP standard concentrations. When initial back- calculated concentrations exceeded the upper calibration point, samples were diluted and re-assayed.
Definition of tumour burden
Considering that AgRP analysis was not always performed at diagnosis of ACC, tumour burden was assessed using baseline RECIST 1.1 criteria.26 The sum of the longest diameters (SLD) of up to 5 target lesions (maximum 2 per organ) was calcu- lated, including the primary tumour when applicable. Target lesions were eligible if ≥10 mm in 1 dimension on CT and ma- lignant lymph nodes if >15 mm in short axis.26 Patients were classified as having low or high tumour burden based on the median SLD.
In addition, tumour burden was also assessed by counting the number of metastatic organ sites, as described in previous studies. 27-30
Statistical analysis and correlation with clinical/ histopathological data
Categorical variables were expressed as numbers with per- centage and were compared with the chi-square (x2) test. Continuous variables were tested for Gaussian distribution with the Shapiro-Wilk test and were presented as mean and standard deviation (SD). Parametric and non-parametric data were analysed with Student’s t-tests or ANOVA followed by Tukey post hoc test or Mann-Whitney U test or Kruskal- Wallis test followed by Dunn’s post hoc test, respectively, and reported as mean ± SD.
Kaplan-Meier survival analysis was used to estimate event- free survival and to investigate the correlation between AgRP and prognosis, with differences between survival curves as- sessed by the log-rank (Mantel-Cox) test and the Gehan- Breslow-Wilcoxon test. To identify factors independently influencing patient outcomes, Cox proportional hazards re- gression modelling was performed, reporting hazard ratios (HR), and 95% confidence intervals (CI). Initially, univariable analyses were conducted, including all known or potentially relevant prognostic factors such as age, sex, tumour burden, Ki-67 proliferation index, and glucocorticoid excess. AgRP was dichotomized into “high” and “low” according to the me- dian, and additional subdivision according to interquartile ranges was performed. Subsequently, variables with a P-value <0.10 were included in multivariable Cox regression models to identify independent predictors. For subgroup com- parisons and analyses involving multiple statistical tests, Bonferroni correction was applied. After fitting the Cox pro- portional hazards models, the proportional hazards assump- tion was evaluated by plotting scaled Schoenfeld residuals versus time for each covariate and by inspecting deviance re- siduals versus the linear predictor (BX). Multicollinearity among the Cox model covariates was assessed using variance inflation factors (VIF) derived from a linear regression of the covariate matrix; VIF > 5 (tolerance < 0.20) was predefined as indicative of problematic collinearity. If any assumption had been violated, it was planned a priori to address this by stratification, inclusion of time-dependent effects, or re- specification of covariates. Outliers were assessed on log10- transformed values using Tukey’s 1.5x IQR criterion.
Progression-free survival (PFS) and overall survival (OS) were calculated. PFS was defined as the time between blood sampling for the AgRP analysis and the date of progression or last follow-up. OS was defined as the time from analysis of AgRP until death or last follow-up.
A P-value <. 05 was considered statistically significant. Statistical analysis was performed with SPSS version 29 (IBM Corporation, Armonk, NY, USA) and GraphPad Prism version 10 (GraphPad Software, San Diego, CA, USA).
Results
AgRP in adrenocortical tissue
In order to evaluate AgRP expression in adrenocortical tissues, bulk RNA dataset from Giordano et al. including 65 adrenal tissues was analysed.24 Here, AgRP mRNA levels of n = 10 normal adrenal glands (NAG), n =22 adrenocortical aden- omas (ACA), and n=33 ACC were compared. AgRP mRNA levels were found to be significantly upregulated in ACC when compared to NAG (median [IQR] 3.0 [1.0] vs 2.6 [0.3], P =. 029) and ACA; 3.0 [1.0] vs 2.4 [0.3], P <. 0001) (Figure 1A).
Immunohistochemistry (IHC) for the evaluation of AgRP at protein levels was performed in different adrenal tissues, in- cluding NAG (n =2), endocrine inactive adenoma (EIA, n = 6), cortisol producing adenoma (CPA, n = 6), pheochromocy- toma (Pheo, n = 9), and ACC (n = 36). AgRP tissue expression was higher in ACC (median H-score = 120 [49]) than in other adrenal tissues (NAG=63 [21]; EIA = 52 [41], CPA = 96 [30], Pheo = 96 [69]), although the difference was only signifi- cant in the comparison ACC vs EIA (P =. 003) (Figure 1B). Nevertheless, when all 21 benign adrenal tumours were pooled and compared to ACC, AgRP tissue expression was significantly higher in the ACC group (120 [49] vs 80 [48], P <. 0001).
Plasma AgRP in adrenocortical carcinoma vs other adrenal lesions and other malignant entities
Plasma AgRP was analysed in 106 ACC patients (63% wom- en). Of note, 26 (25%) were tumour-free after primary sur- gery. Table 1 summarizes clinical and tumour characteristics at ACC diagnosis and at the time of plasma AgRP sampling. At diagnosis, 60% of patients had Cushing’s syndrome and 36% had androgen excess (Table 1). At the time of AgRP sam- pling, 3 patients (3%) had primary ACC without metastases, 28 (26%) had a primary tumour with metastases, 49 (46%) had recurrent ACC after primary surgery, and 26 (25%) were disease-free after primary surgery.
Fifty-sevem patients with benign adrenal tumours [n= 18 with aldosterone-producing adenoma (APA), n=17 with CPA (including 8 with overt adrenal CS), n= 12 with EIA, and n = 10 with Pheo] were analysed. AgRP plasma levels were significantly higher in patients bearing ACC (median and IQR 1.3 [2.3] ng/ml) than those with APA 0.29 [0.21] ng/ml), CPA (0.35 [0.36] ng/ml), EIA (0.28 [0.23] ng/ml) and Pheo (0.28 [0.22] ng/ml) (all P <. 005) (Figure 2A).
If taken separately, the 8 patients with benign overt CS showed higher levels of plasma AgRP than the other benign le- sions but still significantly lower plasma AgRP than ACC (0.52 (0.31) vs 1.3 (2.3) ng/mL, P =. 0039).
When plasma AgRP of 77 metastatic ACC patients were compared with those of the 16 patients with other metastatic malignant tumours, significantly higher plasma AgRP levels were observed in the ACC group (0.17 [0.28] vs 1.3 [2.3], P <. 0001) (Figure 2B).
Plasma AgRP and tumour burden in adrenocortical carcinoma
Patients who underwent primary surgery and were tumour- free had significant lower plasma levels of AgRP than patients bearing ACC (0.3 [0.2] vs 1.3 [2.3] ng/mL, P <. 0001) (Figure 3A).
In order to define tumour burden, we applied RECIST 1.1, calculating the sum of longest diameters (SLD) across up to 5 target lesions (maximum 2 per organ; as reported in the Methods). Using the cohort median SLD of 153 mm to dichot- omize patients, those with high tumour burden showed higher plasma AgRP than those with lower one (2.7 [4.5] vs 0.6 [0.5], P <. 0001) and with those tumour-free (2.7 [4.5] vs 0.3 [0.4], P <. 0001) (Figure 3B). Among the metastatic ACC, patients with ≥4 metastatic organ sites showed higher plasma AgRP than those with 2-3 metastatic organs and those with 1
A
B
*
normalized AgRP mRNA level
AgRP staining
5
250
4
200
3
o
IHC score
150
2
100
1
50-
0
0
NAG (n=10)
ACA (n=22)
ACC (n=33)
NAG (n=2)
EIA (n=6)
CPA (n=6)
Pheo (n=9)
ACC (n=36)
C
NAG (score 53)
CPA (score 80) with adjacent NAG
ZG
c
CPA
ZF
NAG
NAG
CPA
ZR
M
5x
5x
0.2x
0.2x
Ç
ZG
ZF
10x
10x
ACC (score 135)
ACC (score 177)
5x
5X
10x
20x
10x
20x
(respectively (2.7 [6.3] vs 1.5 [2.0] vs 0.5 [0.4], both P <. 05) (Figure 3C).
The patient with the highest plasma AgRP levels (32.5 ng/ml) was a young female with a metastatic and aggressive cortisol-secreting ACC and a Ki67 of 30% who unfortunately died 10 months from diagnosis and 6 months after the ana- lysis of plasma AgRP. On the other hand, the lowest plasma AgRP (0.06 ng/ml) was observed in a patient tumour-free at the time of the analysis, who underwent a surgical resection of cortisol-secreting, ENSAT stage II ACC. Of note, only 1 outlier (32.5 ng/ml) was detected; sensitivity analyses exclud- ing it showed no differences, so it was not excluded from the analysis.
Association of plasma AgRP with pathological and clinical parameters
Ki-67 of primary tumour at the time of diagnosis of ACC was available in 77 cases (73%). As shown in Figure S1A, ACC with lower Ki67 did not show significantly lower AgRP than those with high Ki-67. AgRP levels were significantly higher in patients with cortisol-producing ACC compared to those not secreting cortisol (1.3 (2.4) vs 0.64 (0.7) ng/ml, P =. 0083) (Figure S1B). Plasma AgRP levels were significantly higher in women with ACC than in men (1.1 (2.3) vs 0.63 (1.0) ng/ml, P = . 0373) (Figure S2A). Body mass index (BMI) of 103 ACC patients was available. Of those, 29 (25%) were classified as obese (with a BMI over 30 kg/m2), while the
remaining 74 patients had a BMI below 30 kg/m2 (including 46 individuals with a BMI under 25 kg/m2). No significant differ- ence was observed in AgRP plasma levels between overweight/ obese and normal-weight patients (Figure S2B).
Patients who received mitotane alone or in combination with chemotherapy showed significant lower levels of plasma AgRP than untreated patients (Figure S3).
Correlation of plasma AgRP with prognosis
The association between plasma AgRP levels and both progression-free survival (PFS) and overall survival (OS) was
| Clinical characteristics at the time AgRP analysis | |
| Patients with ACC included, n (%) | 106 (100) |
| Women, n (%) | 67 (63) |
| Age, years (mean [SD]) | 51 (16) |
| BMI, kg/m2 (median [IQR]) | 27 (8) |
| ACC-tumour stage at the time of AgRP analysis | |
| Tumour-free after primary surgery, n (%) | 26 (25) |
| Primary tumour without metastases, n (%) | 3 (3) |
| Primary tumour with metastases, n (%) | 28 (26) |
| Recurrence after primary surgery, n (%) | 49 (46) |
| Tumour burden at the time of AgRP analysis of patients bearing ACC | |
| Median sum of long diameters, mm (median [IQR]) | 153 (160) |
| Median number of tumoural lesions (median [IQR]) | 3 (2) |
| Treatment performed before/during AgRP analysis | |
| First diagnosis (no treatment), n (%) | 31 (29) |
| Surgery, n (%) | 75 (71) |
| Mitotane, n (%) | 39 (37) |
| Chemotherapy, n (%) | 26 (25) |
| Radiotherapy, n (%) | 9 (8) |
| ENSAT stage at the time of diagnosis of ACC | |
| I, n (%) | 4 (1) |
| II, n (%) | 15 (12) |
| III, n (%) | 19 (11) |
| IV, n (%) | 68 (76) |
| Other tumour characteristics at diagnosis of ACC | |
| Ki 67%, (mean [SD]) | 28 (20) |
| Cortisol secretion, n (%) | 64 (60) |
| Androgen secretion, n (%) | 38 (36) |
assessed from the time of analysis. AgRP expression was di- chotomized into low and high levels, based on the median val- ue within the cohort (0.782 ng/ml). High plasma AgRP levels were significantly associated with shorter PFS and OS. Patients with high AgRP experienced a markedly shorter PFS com- pared to those with low levels (3 vs 13 months, log-rank: P <. 0001) (Figure 4A). In univariable analysis, high AgRP lev- els were linked to an increased risk of progression (HR 2.940, 95% CI 1.886-4.629, P <. 0001). Multivariable Cox models included AgRP and established prognostic covariates (age, sex, tumour burden, Ki-67%, and cortisol excess) (see Methods). After adjustment for these factors, AgRP remained independently associated with an increased risk of progression (HR 2.654, 95% CI 1.371-5.083, P =. 0040) (Table 2). Notably, when AgRP levels were divided into quartiles, pa- tients in the high and intermediate-high groups had a signifi- cantly higher risk of progression compared to those in the low group (both adj. P <. 0005) (Figure 4A).
A similar pattern was observed for OS. Patients with high AgRP had a median survival of 11 months, compared to 23 months in the low AgRP group (log-rank: P =. 0001) (Figure 4B). In univariable Cox analysis, high AgRP levels were associated with a worse OS (HR 2.675, 95% CI 1.733-5.643, P <. 0001), and this remained significant in multivariate analysis (HR 2.624, 95% CI 1.021-6.917, P =. 0122). When analysing AgRP quartiles, patients with high and intermediate-high levels had significantly shorter OS than those with low or intermediate-low levels (both P <. 005) (Figure 4B). Of note, proportional hazards assump- tion was not violated for any model.
Discussion
This large monocentric cross-sectional study analyses for the first time the role of the orexigenic neuropeptide AgRP in the context of ACC. In patients with ACC, plasma AgRP was not only observed to be higher than in patients with other adrenal lesions but was also significantly higher than in pa- tients with other types of cancer. ACC patients who were tumour-free following surgical resection showed significantly lower levels than patients bearing the disease, and the levels
A
B
**
35
o
35-
O
20-
20-
AgRP [ng/ml]
55
AgRP [ng/ml]
55
4-
4
3.
3-
2.
2-
o
1.
1.
0
0
V
APA
CPA
ΕΙΑ
Pheo
ACC
Other metastatic malignancies
Metastatic ACC
A
B
C
ns
*
35
35
35
20
20
20
UIUI
con
Ch CH
AgRP [ng/ml]
AgRP [ng/ml]
AgRP [ng/ml]
4
4
4
3
3
3
2
2
2
1
1
1
0
ACC-free
Bearing ACC
0
ACC-free
Low tumour burden
High tumour burden
0
1 metastatic organ sites
2-3 metastatic organ sites
≥4 metastatic organ sites
A
Progression free survival from AgRP analysis (median)
Progression free survival from AgRP analysis (interquartiles)
100
Low
100
Low
Probability of Survival
80-
High
Probability of Survival
80-
Trend p<0.0001
Intermediate Low
Intermediate High
60
logrank HR 2.600 (CI 95% 1.659-4.074), p<0.0001
60
- High
40
40-
Low vs High adj. p<0.0001 Int. High vs Low adj. p=0.0003
20
20
0
0
0
10
20
30
40
50
60
70
80
0
10
20
30
40
50
60
70
80
No at risk
Time(months)
53
15
2
No at risk
Time(months)
53
0
6
2
27
0
26
9
1
0
26
6
1
0
27
3
0
0
3
2
0
B
Overall survival from AgRP analysis (median)
Overall survival from AgRP analysis (interquartiles)
100
Low
100
Low
Probability of Survival
80
High
Probability of Survival
80
Trend p<0.0001
Intermediate Low
logrank HR 3.614 (CI 95% 2.184- 5.982), p<0.0001
Intermediate High
- High
60
60
40
40
Low vs High adj. p=0.0006
Int. High vs Low adj. p=0.0066
Int.Low vs Int.High adj. p=0.003
20
20
Int.Low vs High adj. p=0.00042
0
0
0
50
100
150
0
50
100
150
No at risk
Time(months)
Time(months)
53
53
4
No at risk
2
0
0
1
27
0
0
0
0
26
1
0
0
26
0
0
0
27
2
1
0
of AgRP were associated to the tumour burden. Moreover, the overall survival and the progression-free survival were signifi- cantly worse in patients with high AgRP levels.
While AgRP is in healthy subjects is predominantly produced in the hypothalamus, previous studies have demonstrated its ex- pression and secretion in the adrenal glands.16-18 Given that no
| Variable | n | Median time to progression (months) | Univariable | Multivariable | ||||
|---|---|---|---|---|---|---|---|---|
| HR | 95% CI | P-value | HR | 95% CI | P-value | |||
| Factors influencing progression-free survival | ||||||||
| Age | ||||||||
| ≤49 | 42 | 8 | 1.07 | 0.692-1.69 | .754 | — | — | — |
| >50 | 64 | 6 | ||||||
| Sex | ||||||||
| Female | 67 | 6 | 0.99 | 0.621-1.55 | .968 | — | — | — |
| Male | 39 | 9 | ||||||
| Tumour burden | ||||||||
| Tumour-free | 26 | 19 | 3.83 | 1.95-8.20 | <. 0001 | 4.63 | 2.09-11.32 | <. 0001 |
| Low tumour burden | 42 | 7 | ||||||
| High tumour burden | 38 | 2 | 7.51 | 3.83-16.13 | <. 0001 | 10.88 | 4.32-29.13 | <. 0001 |
| Ki67% | ||||||||
| 1-19 | 24 | 14 | 1.93 | 1.08-3.67 | .0264 | 1.51 | 0.82-2.96 | .193 |
| ≥20 | 52 | 7 | ||||||
| Cortisol secretion | ||||||||
| No | 33 | 9 | 1.58 | 0.97-2.67 | .0640 | 0.93 | 0.50-1.79 | .831 |
| Yes | 73 | 7 | ||||||
| Plasma AgRP | ||||||||
| Low | 53 | 13 | 2.94 | 1.88-4.63 | <. 0001 | 2.65 | 1.37-5.08 | .0040 |
| High | 53 | 3 | ||||||
| Variable | n Median | overall survival (months) | Univariable | Multivariable | ||||
|---|---|---|---|---|---|---|---|---|
| HR | 95% CI | P-value | HR | 95% CI | P-value | |||
| Factors influencing overall survival | ||||||||
| Age | ||||||||
| ≤49 | 42 | 19 | 1.09 | 0.62-1.94 | 0.769 | — | — | — |
| >50 | 64 | 14 | ||||||
| Sex | ||||||||
| Female | 67 | 15 | 0.73 | 0.39-1.32 | 0.309 | — | — | — |
| Male | 39 | 21 | ||||||
| Tumour burden | ||||||||
| Tumour-free | 26 | 33 | 4.06 | 1.53-13.98 | .0036 | 2.93 | 0.99-10.78 | .0549 |
| Low tumour burden | 42 | 16 | ||||||
| High tumour burden | 38 | 7 | 10.11 | 3.91-34.45 | <. 0001 | 7.22 | 2.09-29.63 | .0015 |
| Ki67% | ||||||||
| 1-19 | 24 | 24 | 2.01 | 0.92-5.05 | .0829 | 1.22 | 0.54-3.16 | .6432 |
| ≥20 | 52 | 16 | ||||||
| Cortisol secretion | ||||||||
| No | 33 | 14 | 2.67 | 1.39-5.68 | .0026 | 2.98 | 1.26-7.88 | .0122 |
| Yes | 73 | 32 | ||||||
| Plasma AgRP | ||||||||
| Low | 53 | 23 | 3.07 | 1.73-5.64 | <. 0001 | 2.62 | 1.02-6.91 | .0452 |
| High | 53 | 11 | ||||||
prior analysis has focused on AgRP in adrenal lesions, we aimed to investigate its mRNA levels and protein expression in adrenal tumours. Consistent with in silico RNA expression data, our im- munohistochemical analysis revealed higher AgRP expression in ACC compared to other adrenal lesions, although this difference was not statistically significant. Of note, this was also probably related to the elevated AgRP levels observed in pheochromocyto- mas, due to its expression in neuroendocrine chromaffin cells of the adrenal medulla.16 In support of our findings, other studies have shown AgRP expression in the adrenal cortex, reporting also a role in the regulation of steroidogenesis.17
Given that previous studies have reported AgRP upregula- tion in CS2º and demonstrated that glucocorticoids signifi- cantly influence AgRP synthesis and release,20,31,32 one might assume that the elevated AgRP levels observed in ACC are merely a result of cortisol excess (especially since 60% of the ACC cases analysed were cortisol-secreting). However, some important aspects have to be considered.
First, plasma AgRP was significantly higher in ACC than in those with benign hypercortisolism, including overt CS. Second, levels of AgRP in patients with benign overt CS were not significantly different from ACC without hypercor- tisolism (median 0.52 vs 0.64 ng/ml). Third, AgRP levels were negatively affecting the prognosis of ACC patients, inde- pendently from hypercortisolism. Finally, preclinical work showed AgRP expression in rat adrenal cortex, presence of MC3/MC4 receptor, and functional antagonism of a-MSH-stimulated corticosterone secretion by AgRP in dis- persed adrenal cells (consistent with a local inhibitory para- crine mechanism).33 Moreover, the authors reported that glucocorticoids upregulate adrenal AgRP mRNA (indicating feedback regulation).33 Together, these data support a model in which AgRP participates in adrenal autocrine/paracrine sig- nalling and glucocorticoid-responsive feedback.33
Given these results and by looking at the mRNA and immu- nohistochemical analysis, it seems that AgRP may be
expressed and secreted from ACC cells, and not be completely cortisol-related. Considering that ACC are larger than benign cortisol secreting adrenal adenomas34 they might have greater capacity to produce AgRP. This can be also explained by the fact that higher tumour burden ACC show higher levels of cir- culating AgRP, than those with lower burden and those who were tumour-free at the time of AgRP analysis.
Interestingly, AgRP has been reported to be elevated in cer- tain forms of cancer, probably as part of the physiological mechanism to prevent cachexia.35 Nevertheless, in our study patients with ACC exhibited higher levels of plasma AgRP compared to those with other malignancies, which can be ex- plained by the secretion of AgRP directly from tumour tissue and not as a compensatory mechanism in the case of other can- cers in which probably most of AgRP is of hypothalamic origin.
Another intriguing finding concerns the prognostic role of AgRP following the diagnosis and during the treatment of ACC. Elevated levels of AgRP were found to correlate nega- tively with patient survival and disease progression during follow-up, even by controlling for tumour burden and other factors known to influence prognosis in ACC. These observa- tions could be explained by different mechanisms directly link- ing AgRP to tumour progression. For example, previous studies have demonstrated that mutated forms of AgRP bind to the cell adhesion receptor integrin avB3, which could play a role in tumour angiogenesis.36 Furthermore, considering the orexigenic role of AgRP in energy expenditure,37 we might speculate that AgRP could influence energy metabolism dur- ing ACC progression and prevents body weight loss, which is much rarer in ACC than in other advanced cancers. Although these hypotheses are interesting, higher levels of AgRP may be simply related to the burden of ACC, since it is expressed in the adrenal tissue. Either ways, if it is produced in the context of cancer progression or if it is a sort of marker of adrenocortical tissue in ACC, it can be interesting in pre- dicting the prognosis and the tumour stage of ACC.
Although these results are of interest, some limitations need to be pointed out. First, our cohort predominantly includes pa- tients with advanced stages of ACC, with a limited number of patients with localized tumour. This could have affected the results, particularly the survival curves. Second, AgRP was measured in different patients, and not longitudinally in the same ones. Repetitive measurements of AgRP in the same pa- tient throughout the different phases of cancer could confirm the validity of our findings. Nevertheless, our single-centre study encompasses a large cohort of ACC patients with a uni- form approach to experimental analysis and data collection. Third, some groups of the immunohistochemistry analysis (eg, NAG, EIA, CPA) were very small, which limits statistical power and may affect the robustness of subgroup analyses
In conclusion, this study represents the first investigation of AgRP in the context of ACC. AgRP appears to be a promising biomarker for differentiating ACC from other adrenal lesions and malignant entities, and it may also serve as a prognostic marker for survival. However, these findings warrant pro- spective, longitudinal, multicentre validation to test AgRP as a translational biomarker for disease monitoring and risk stratification in ACC. Priority areas include serial sampling, assay standardization, and integration with clinicopathologi- cal models (eg, ENSAT stage, Ki-67, imaging-derived tumour burden and treatment response) to define actionable thresh- olds and clinical utility.
Supplementary material
Supplementary material is available at European Journal of Endocrinology online.
Funding
This work was supported by the DFG German Research Foundation Project 314061271 (TRR 205) (M.F.), the Clinician Scientist program RISE funded by the Else- Kroner-Fresenius-Stiftung and the Eva-Luise-und-Horst- Köhler Stiftung (M.D.) and by the Bavarian Cancer Research Center (BZKF) (B.A.).
Authors’ contributions
Mario Detomas (Conceptualization [lead], Data curation [lead], Formal analysis [lead], Investigation [lead], Methodology [lead], Validation [lead], Visualization [lead], Writing-original draft [lead], Writing-review & editing [lead]), Lisa Kagan (Formal analysis [supporting], Investigation [supporting], Methodology [supporting], Writing-original draft [supporting]), James Pittaway (Formal analysis [supporting], Validation [supporting], Writing-review & editing [supporting]), Marc Schauer (Visualization [supporting], Writing-review & editing [sup- porting]), Simon Kloock (Writing-review & editing [sup- porting]), Niklas Geiger (Investigation [supporting]), Otilia Kimpel (Formal analysis [supporting]), Martin Fassnacht (Funding acquisition [lead], Resources [lead], Writing- review & editing [supporting]), Barbara Altieri (Formal analysis [supporting], Investigation [lead], Methodology [lead], Writing-review & editing [supporting]), and Ulrich Dischinger (Conceptualization [lead], Funding acquisition [equal], Supervision [lead], Writing-original draft [support- ing], Writing-review & editing [supporting])
M.D. and U.D. designed the research. M.D., J.F.H.P., and L.R.K. performed the statistical analyses. M.D., L.R.K., and U.D. drafted the first version of the manuscript. M.D., L.R.K., B.A., M.P.S., S.K., N.G., O.K., M.F., and U.D. col- lected clinical data from patients and performed the experi- ments. All authors contributed to writing the manuscript and approved the final version to be published.
Conflict of interest: The authors declare no conflict of interest.
References
1. Kerkhofs TM, Verhoeven RH, Van der Zwan JM, et al. Adrenocortical carcinoma: a population-based study on incidence and survival in The Netherlands since 1993. Eur J Cancer. 2013;49(11):2579-2586. https://doi.org/10.1016/j.ejca.2013.02. 034
2. Fassnacht M, Dekkers OM, Else T, et al. European Society of Endocrinology Clinical Practice Guidelines on the management of adrenocortical carcinoma in adults, in collaboration with the European Network for the Study of Adrenal Tumors. Eur J Endocrinol. 2018;179(4):G1-G46. https://doi.org/10.1530/EJE- 18-0608
3. Altieri B, Ronchi CL, Kroiss M, Fassnacht M. Next-generation ther- apies for adrenocortical carcinoma. Best Pract Res Clin Endocrinol Metab. 2020;34(3):101434. https://doi.org/10.1016/j.beem.2020. 101434
4. Detomas M, Pivonello C, Pellegrini B, et al. MicroRNAs and long non-coding RNAs in adrenocortical carcinoma. Cells. 2022;11(14):2234. https://doi.org/10.3390/cells11142234
5. Mete O, Erickson LA, Juhlin CC, et al. Overview of the 2022 WHO classification of adrenal cortical tumors. Endocr Pathol. 2022;33(1):155-196. https://doi.org/10.1007/s12022-022-09710-8
6. Beuschlein F, Weigel J, Saeger W, et al. Major prognostic role of Ki67 in localized adrenocortical carcinoma after complete resec- tion. J Clin Endocrinol Metab. 2015;100(3):841-849. https://doi. org/10.1210/jc.2014-3182
7. Cantini G, Fei L, Canu L, et al. Circulating Fascin 1 as a promising prognostic marker in adrenocortical cancer. Front Endocrinol (Lausanne). 2021;12:698862. https://doi.org/10.3389/fendo.2021. 698862
8. Altieri B, Sbiera S, Della Casa S, et al. Livin/BIRC7 expression as malignancy marker in adrenocortical tumors. Oncotarget. 2017;8(6):9323-9338. https://doi.org/10.18632/oncotarget.14067
9. Roca E, Berruti A, Sbiera S, et al. Topoisomerase 2alpha and thymi- dylate synthase expression in adrenocortical cancer. Endocr Relat Cancer. 2017;24(7):319-327. https://doi.org/10.1530/ERC-17- 0095
10. Sbiera S, Schmull S, Assie G, et al. High diagnostic and prognostic value of steroidogenic factor-1 expression in adrenal tumors. J Clin Endocrinol Metab. 2010;95(10):E161-E171. https://doi.org/ 10.1210/jc.2010-0653
11. Sbiera S, Sbiera I, Ruggiero C, et al. Assessment of VAV2 expression refines prognostic prediction in adrenocortical carcinoma. J Clin Endocrinol Metab. 2017;102(9):3491-3498. https://doi.org/10. 1210/jc.2017-00984
12. Mariniello K, Pittaway JFH, Altieri B, et al. Dlk1 is a novel adreno- cortical stem/progenitor cell marker that predicts malignancy in adrenocortical carcinoma. Cancer Commun (Lond). 2025;45(6): 663-668. https://doi.org/10.1002/cac2.70012
13. Pittaway JFH, Lipsos C, Mariniello K, Guasti L. The role of delta- like non-canonical Notch ligand 1 (DLK1) in cancer. Endocr Relat Cancer. 2021;28(12):R271-RR87. https://doi.org/10.1530/ERC- 21-0208
14. Mercer AJ, Hentges ST, Meshul CK, Low MJ. Unraveling the cen- tral proopiomelanocortin neural circuits. Front Neurosci. 2013;7: 19. https://doi.org/10.3389/fnins.2013.00019
15. Lee M, Wardlaw SL. The central melanocortin system and the regu- lation of energy balance. Front Biosci. 2007;12(8-12):3994-4010. https://doi.org/10.2741/2366
16. Gupta R, Ma Y, Wang M, Whim MD. AgRP-expressing adrenal chromaffin cells are involved in the sympathetic response to fasting. Endocrinology. 2017;158(8):2572-2584. https://doi.org/10.1210/ en.2016-1268
17. Doghman M, Soltani Y, Rebuffet V, Naville D, Begeot M. Role of Agouti-related protein in adrenal steroidogenesis. Mol Cell Endocrinol. 2007;265-266:108-112. https://doi.org/10.1016/j. mce.2006.12.025
18. Bicknell AB, Lomthaisong K, Gladwell RT, Lowry PJ. Agouti re- lated protein in the rat adrenal Cortex: implications for novel auto- crine mechanisms modulating the actions of pro-opiomelanocortin peptides. J Neuroendocrinol. 2000;12(10):977-982. https://doi.org/ 10.1046/j.1365-2826.2000.00543.x
19. Shibata M, Banno R, Sugiyama M, et al. AgRP neuron-specific de- letion of glucocorticoid receptor leads to increased energy expend- iture and decreased body weight in female mice on a high-fat diet. Endocrinology. 2016;157(4):1457-1466. https://doi.org/10.1210/ en.2015-1430
20. Page-Wilson G, Peters JB, Panigrahi SK, et al. Plasma agouti-related protein and cortisol levels in cushing disease: evidence for the regu- lation of agouti-related protein by glucocorticoids in humans. J Clin Endocrinol Metab. 2019;104(3):961-969. https://doi.org/10.1210/ jc.2018-01909
21. Page-Wilson G, Freda PU, Jacobs TP, et al. Clinical utility of plasma POMC and AgRP measurements in the differential diagnosis of
ACTH-dependent Cushing’s syndrome. J Clin Endocrinol Metab. 2014;99(10):E1838-E1845. https://doi.org/10.1210/jc.2014-1448
22. Dhillo WS, Gardiner JV, Castle L, et al. Agouti related protein (AgRP) is upregulated in Cushing’s syndrome. Exp Clin Endocrinol Diabetes. 2005;113(10):602-606. https://doi.org/10. 1055/s-2005-872895
23. McShane LM, Altman DG, Sauerbrei W, et al. REporting recommen- dations for tumour MARKer prognostic studies (REMARK). Br J Cancer. 2005;93(4):387-391. https://doi.org/10.1038/sj.bjc.6602678
24. Giordano TJ, Kuick R, Else T, et al. Molecular classification and prognostication of adrenocortical tumors by transcriptome profil- ing. Clin Cancer Res. 2009;15(2):668-676. https://doi.org/10. 1158/1078-0432.CCR-08-1067
25. Altieri B, Secener AK, Sai S, et al. Single-nucleus and spatial tran- scriptome reveal adrenal homeostasis in normal and tumoural ad- renal glands. Clin Transl Med. 2024;14(8):e1798. https://doi.org/ 10.1002/ctm2.1798
26. Eisenhauer EA, Therasse P, Bogaerts J, et al. New response evalu- ation criteria in solid tumours: revised RECIST guideline (version 1.1). Eur J Cancer. 2009;45(2):228-247. https://doi.org/10.1016/j. ejca.2008.10.026
27. Han KR, Pantuck AJ, Bui MH, et al. Number of metastatic sites ra- ther than location dictates overall survival of patients with node- negative metastatic renal cell carcinoma. Urology. 2003;61(2): 314-319. https://doi.org/10.1016/S0090-4295(02)02163-5
28. Oh Y, Taylor S, Bekele BN, et al. Number of metastatic sites is a strong predictor of survival in patients with nonsmall cell lung can- cer with or without brain metastases. Cancer. 2009;115(13): 2930-2938. https://doi.org/10.1002/cncr.24333
29. Yin Z, Tang H, Li L, et al. Impact of sites versus number of metas- tases on survival of patients with organ metastasis from newly diag- nosed cervical cancer. Cancer Manag Res. 2019;11:7759-7766. https://doi.org/10.2147/CMAR.S203037
30. Christ SM, Pohl K, Willmann J, et al. Patterns of metastatic spread and tumor burden in unselected cancer patients using PET imaging: implications for the oligometastatic spectrum theory. Clin Transl Radiat Oncol. 2024;45:100724. https://doi.org/10.1016/j.ctro. 2024.100724
31. Savontaus E, Conwell IM, Wardlaw SL. Effects of adrenalectomy on AGRP, POMC, NPY and CART gene expression in the basal hypothalamus of fed and fasted rats. Brain Res. 2002;958(1): 130-138. https://doi.org/10.1016/S0006-8993(02)03674-0
32. Shimizu H, Arima H, Ozawa Y, et al. Glucocorticoids increase NPY gene expression in the arcuate nucleus by inhibiting mTOR signal- ing in rat hypothalamic organotypic cultures. Peptides. 2010;31(1): 145-149. https://doi.org/10.1016/j.peptides.2009.09.036
33. Dhillo WS, Small CJ, Gardiner JV, et al. Agouti-related protein has an inhibitory paracrine role in the rat adrenal gland. Biochem Biophys Res Commun. 2003;301(1):102-107. https://doi.org/10. 1016/S0006-291X(02)02991-1
34. Fassnacht M, Tsagarakis S, Terzolo M, et al. European Society of Endocrinology clinical practice guidelines on the management of adrenal incidentalomas, in collaboration with the European Network for the Study of Adrenal Tumors. Eur J Endocrinol. 2023;189(1):G1-G42. https://doi.org/10.1093/ejendo/lvad066
35. Bienertova-Vasku J, Tomandl J, Bienert P, et al. Circulating levels of agouti-related peptide in endometrial cancer survivors. World J Oncol. 2011;2(5):232-237. https://doi.org/10.4021/wjon385w
36. Jiang L, Miao Z, Kimura RH, et al. 111In-labeled cystine-knot pep- tides based on the Agouti-related protein for targeting tumor angio- genesis. J Biomed Biotechnol. 2012;2012:368075. https://doi.org/ 10.1155/2012/368075
37. Deem JD, Faber CL, Morton GJ. AgRP neurons: regulators of feed- ing, energy expenditure, and behavior. FEBS J. 2022;289(8): 2362-2381. https://doi.org/10.1111/febs.16176