Immunohistochemical evaluation of apelin and VEGF expression in cortisol-producing adrenal adenomas Authors: Hatice Çalişkan Burgucu, Tuğba Günler, Ethem Ömeroğlu, Yaşar Ünlü, Oğuzhan Aksu Published: 2026 ABSTRACT AbstractPurposeIn this study, we aimed to investigate the immunohistochemical expression of apelin in tissue samples obtained from patients with Cushing’s syndrome (CS) due to cortisol-producing adrenal adenoma (CPA) and to compare these findings with normal adrenal cortex tissue. Additionally, we sought to evaluate the association of apelin expression with clinical, hormonal, and histopathological parameters. The expression profile of vascular endothelial growth factor (VEGF) was also assessed to explore potential angiogenic interactions.MethodsThis retrospective, single-center study included 11 patients diagnosed with CPA and 15 control subjects with histologically normal adrenal cortex tissue. Formalin-fixed, paraffin-embedded tissue sections were subjected to immunohistochemical staining for apelin and VEGF. Immunohistochemical evaluation was performed by assessing staining intensity and the percentage of positively stained cells. An immunoreactivity score (IRS) was calculated by combining these parameters to provide a semi-quantitative measure of protein expression.ResultsApelin staining intensity was significantly higher in CPA tissues compared with controls (p = 0.026), whereas the percentage of apelin-positive cells and IRS values did not differ significantly between groups. In an exploratory ROC analysis, apelin staining intensity showed moderate discrimination between CPA and control adrenal cortex tissues (AUC = 0.776, 95% CI: 0.581–0.971, p = 0.018). In bias-reduced univariable logistic regression, apelin staining intensity was associated with CPA (OR = 2.96, 95% CI: 1.02–8.60, p = 0.046), although the wide confidence interval indicated considerable uncertainty. VEGF staining parameters did not differ significantly between CPA and control tissues. No statistically significant correlations could be demonstrated between apelin or VEGF IRS and the evaluated clinicopathological or hormonal parameters.ConclusionApelin staining intensity was higher in CPA than in normal adrenal cortex, whereas the proportion of apelin-positive cells and IRS did not differ significantly between groups. VEGF staining parameters also did not differ significantly between CPA and control tissues. These findings suggest altered apelin immunoreactivity in CPA; however, given the exploratory nature and small sample size of the study, larger studies are required to confirm these observations and clarify their biological significance.Clinical trial numberNot applicable.Supplementary InformationThe online version contains supplementary material available at https://doi.org/10.1186/s12902-026-02563-w.Keywords: Cortisol-producing adrenal adenoma, Cushing’s syndrome, Angiogenesis, Apelin, Vascular endothelial growth factor, Immunohistochemistry Introduction Cushing’s syndrome (CS) is an endocrine disorder caused by chronic exposure to excessive glucocorticoids and associated with significant multisystem morbidity. Its clinical presentation is heterogeneous and depends on the duration and severity of cortisol excess. No single clinical feature is pathognomonic, and symptoms range from mild biochemical hypercortisolism to overt clinical disease [1]. Endogenous CS is etiologically classified into adrenocorticotropic hormone (ACTH)–dependent forms, predominantly Cushing’s disease (CD), and ACTH-independent forms resulting from primary adrenal pathology [2]. ACTH-independent CS most commonly arises from cortisol-producing adrenal adenomas (CPA) and, less frequently, from nodular adrenal hyperplasia or adrenocortical carcinoma (ACC), accounting for approximately 15–20% of endogenous cases [3]. Although endogenous CS is rare, with an estimated annual incidence of 0.7–2.4 cases per million population, delayed diagnosis and persistent hypercortisolism contribute to increased cardiovascular, metabolic, and overall mortality risk. Surgical resection of the adrenal lesion remains the first-line treatment for ACTH-independent CS [4]. Angiogenesis is a fundamental biological process involved in both normal adrenal cortical development and adrenocortical tumorigenesis. The adrenal cortex represents one of the most highly vascularized endocrine tissues, where dynamic vascular remodeling is tightly coordinated with hormonal activity to ensure adequate hormone synthesis and systemic delivery [5]. Among the mediators supporting angiogenesis, vascular endothelial growth factor (VEGF) is considered a key regulator of adrenal vascular architecture and endothelial cell proliferation. Experimental and histopathological evidence demonstrates that VEGF expression is closely related to adrenocortical cellular function and is frequently upregulated in neoplastic adrenal tissues, particularly in adrenocortical tumors. Increased VEGF signaling has been associated with higher microvascular density, supporting a potential role in facilitating tumor growth by enhancing oxygen and nutrient delivery to steroidogenic cells [6]. In ACTH-independent Cushing syndrome (AICS), especially in CPA, tumor expansion and persistent steroid hypersecretion occur within a highly vascularized microenvironment. Collectively, these findings support the concept that VEGF-driven angiogenesis may contribute not only to tumor progression but also to the maintenance of cortisol overproduction in adrenal neoplasms [5, 6]. Apelin is an endogenous peptide ligand of the G protein–coupled receptor APJ, also known as the angiotensin-like receptor [7]. The apelin–APJ signaling pathway is widely expressed across various tissues and has been implicated in the regulation of multiple physiological processes, including cardiovascular function, angiogenesis, and energy metabolism [8]. Beyond its vascular actions, emerging evidence suggests that apelin also plays a significant role in metabolic homeostasis by modulating glucose metabolism, insulin sensitivity, and lipid metabolism. Notably, circulating apelin levels have been reported to be elevated in conditions such as obesity, insulin resistance, and type 2 diabetes mellitus, where they show positive associations with markers of metabolic dysfunction [9]. Given that CS is characterized by central obesity, impaired glucose tolerance, and insulin resistance, dysregulation of the apelin axis may contribute to its complex cardiometabolic phenotype. Beyond its metabolic role, the apelin–APJ system has also been recognized as a modulator of angiogenesis. Apelin expression is inducible under hypoxic conditions through hypoxia-inducible factor (HIF)-dependent pathways, placing it within the transcriptional network that also regulates VEGF [10]. Experimental and translational evidence supports functional cross-talk between apelin signaling and VEGF-mediated angiogenic pathways, with apelin being associated with enhanced VEGF expression and augmented neovascular responses in various tissues [11]. Apelin has been reported to promote angiogenesis through activation of multiple signaling pathways, including AMP-activated protein kinase (AMPK)/endothelial nitric oxide synthase (eNOS), phosphatidylinositol 3-kinase (PI3K)/Akt/eNOS, and HIF-1/VEGF/VEGF receptor (VEGFR) cascades. This supports apelin’s role in vascular remodeling and endothelial proliferation [12, 13]. Taken together, these findings suggest that apelin may cooperate with VEGF-driven vascular remodeling and contribute to the proangiogenic microenvironment observed in endocrine tumors. However, the expression profile of apelin in CPA and its potential interaction with VEGF-driven angiogenic signaling have not been systematically investigated. Clarifying this relationship may provide novel insights into the angiogenic and metabolic microenvironment of CPA. To the best of our knowledge, this is the first study investigating apelin expression in cortisol-producing adrenal adenomas. In this study, we aimed to investigate apelin and VEGF expression in tissue samples obtained from patients diagnosed with ACTH-independent Cushing syndrome due to CPA. Furthermore, we aimed to investigate the relationships between apelin and VEGF expression and clinicopathological characteristics in this patient cohort. We hypothesized that apelin immunoreactivity may be altered in CPA tissues and explored its potential relationship with VEGF immunoreactivity and clinicopathological characteristics. Materials and methods This retrospective, single-center study included patients who underwent adrenalectomy for CPA at Konya City Hospital between August 2020 and May 2025. The study protocol was approved by the Non-Interventional Clinical Research Ethics Committee of Konya City Hospital (Approval No. 2025/116). Owing to the retrospective design, all data were fully anonymized before analysis to ensure patient confidentiality. Demographic, biochemical, and clinicopathological data were obtained from electronic medical records. The study was conducted in accordance with the ethical principles of the Declaration of Helsinki. The study was reported in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines. Patients were included only if they had clinically overt Cushing syndrome (CS) due to a cortisol-producing adrenal adenoma. The diagnosis of CS was based on compatible clinical features of hypercortisolism together with biochemical evidence of ACTH-independent cortisol excess. Biochemical evaluation included basal serum ACTH, 24-hour urinary free cortisol, midnight serum cortisol for assessment of the diurnal cortisol rhythm, and the 1-mg overnight dexamethasone suppression test. All included patients demonstrated inadequate cortisol suppression following the 1-mg overnight dexamethasone suppression test (serum cortisol ≥ 1.8 µg/dL) and disruption of the normal diurnal cortisol rhythm. Basal ACTH concentrations were suppressed below the lower limit of the assay-specific reference range (10–60 pg/mL) in all patients, supporting ACTH-independent cortisol excess. Twenty-four-hour urinary free cortisol was considered as part of the overall biochemical assessment [2]. Histopathological confirmation of CPA was established following adrenalectomy. To exclude malignant adrenal cortical tumors, all specimens were evaluated according to the Weiss scoring system, a widely used histopathological algorithm for differentiating adrenocortical adenoma from carcinoma. The Weiss system evaluates nine histological parameters, including nuclear grade, mitotic rate, atypical mitoses, clear cell component, diffuse architecture, necrosis, venous invasion, sinusoidal invasion, and capsular invasion. Tumors with a Weiss score ≥ 3 are considered suggestive of adrenocortical carcinoma. In the present study, only tumors with a Weiss score ≤ 2 were included in the final analysis to ensure inclusion of benign adrenal cortical adenomas [14]. For immunohistochemical analysis, formalin-fixed, paraffin-embedded tissue sections obtained from patients diagnosed with CPA were used, along with control tissues containing morphologically normal adrenal cortex. The control group consisted of 15 histologically normal adrenal cortex samples, of which 12 were obtained from nephrectomy specimens and 3 from adrenal cortex adjacent to benign adrenal lesions. Serial sections of 4–5 μm thickness were cut from each paraffin block and mounted on positively charged slides. Two slides were prepared from each specimen. Immunohistochemical staining was performed using a closed-system automated staining platform (DAKO, Glostrup, Denmark) in accordance with the manufacturer’s instructions. One slide was incubated with a primary antibody against VEGF (RB-9031-PO, Thermo Fisher Scientific, Waltham, MA, USA), and the other with a primary antibody against apelin (Apelin Polyclonal Antibody, orb1411607, Biorbyt Ltd., Cambridge, UK). For VEGF immunohistochemistry, heat-induced epitope retrieval (HIER) was performed in 1 mM EDTA buffer (pH 8.0) at boiling temperature for 10–20 min, followed by cooling at room temperature for 20 min. The VEGF primary antibody was applied at a dilution of 1:100 and incubated for 30 min at room temperature. For apelin immunohistochemistry, HIER was performed in citrate buffer (pH 6.0) at boiling temperature for 15 min, followed by cooling at room temperature. The apelin primary antibody was applied at a dilution of 1:200 and incubated overnight at 4 °C. Immunohistochemical staining was performed using the DAKO Autostainer Link 48 automated staining platform (Dako/Agilent Technologies, Glostrup, Denmark), with the EnVision FLEX detection/visualization system. 3,3′-Diaminobenzidine (DAB) was used as the chromogen, and hematoxylin was used for counterstaining. Placental tissue was used as a positive control for both apelin and VEGF, with strong cytoplasmic immunoreactivity in villous trophoblasts and intravillous vascular endothelial cells considered indicative of appropriate staining. Negative controls were processed using the same immunohistochemical procedure with omission of the primary antibody, and no specific immunoreactivity was observed. Brown cytoplasmic immunoreactivity was considered indicative of positive expression for both antibodies. Each slide was initially scanned in its entirety at low magnification. A fixed number of high-power fields (HPFs) was not used. Representative areas of viable tumor tissue without necrosis, hemorrhage, or technical artifacts were evaluated at ×200 and ×400 magnification to determine the percentage of positively stained cells and staining intensity. For each specimen, both the proportion of positively stained tumor cells and the staining intensity were evaluated. The percentage of positive tumor cells was graded using a semi-quantitative scale as follows: 0 (no positive cells), 1 (< 25% positive tumor cells), 2 (25–50% positive tumor cells), 3 (51–75% positive tumor cells), and 4 (> 75% positive tumor cells). The proportion score cut-offs were defined according to the routine semi-quantitative assessment approach of the evaluating pathologist. In addition, the percentage of positive cells was also recorded as a continuous variable for statistical comparisons. Staining intensity was independently classified as 0 (no staining), 1 (weak), 2 (moderate), or 3 (strong) [15]. An immunoreactivity score (IRS) was calculated by multiplying the proportion score by the intensity score, yielding a total score ranging from 0 to 12. Based on previously published IRS categorization, immunostaining was classified as negative (IRS 0–1), low (IRS 2–3), moderate (IRS 4–8), and high (IRS 9–12) [16]. All sections were independently evaluated by three experienced pathologists who were blinded to the patients’ clinical data and histopathological parameters, including Weiss scores. Each slide was assessed independently, and discordant scores were subsequently resolved by consensus evaluation using a multi-headed microscope. Interobserver agreement for the four-level ordinal apelin and VEGF staining intensity scores was evaluated using pairwise linearly weighted Cohen’s kappa coefficients based on the three pathologists’ independent pre-consensus assessments. Statistical analysis All primary analyses and visualizations were performed in R (version 4.5.2), whereas Spearman correlation analyses were conducted in IBM SPSS Statistics (version 22.0; IBM Corp., Armonk, NY, USA). The use of two statistical software packages reflected the analytical workflow of the study, with correlation analyses performed separately from the subsequent R-based analyses. Categorical variables were compared using Fisher’s exact test. Continuous variables were assessed for normality, and when parametric assumptions were not satisfied, the Wilcoxon rank-sum test was applied. Effect sizes for non-parametric comparisons were reported as rank-biserial correlation coefficients. To address potential statistical instability related to the small sample size, resampling-based approaches were implemented. Permutation testing with 10,000 iterations was used to validate categorical comparisons. In logistic regression analyses, bias-reduced estimation methods (brglm2) were applied to minimize small-sample bias and separation issues. Confidence intervals for regression and receiver operating characteristic (ROC) analyses were calculated using the bootstrap percentile method with 5,000 resamples. A two-sided p value < 0.05 was considered statistically significant. Results A total of 11 patients with CS due to CPA and 15 control subjects with normal adrenal cortex tissue were included in the study. In the CPA group, 10 of 11 patients (90.9%) were female and 1 (9.1%) was male, whereas in the control group 10 of 15 subjects (66.7%) were female and 5 (33.3%) were male. There was no statistically significant difference in sex distribution between groups (p = 0.197), with a small-to-moderate effect size (Cramer’s V = 0.284). The mean age was 46.09 ± 12.36 years in the CPA group and 45.27 ± 10.48 years in the control group, with no significant difference between groups (p = 0.856). The effect size for age comparison was negligible (Cohen’s d = − 0.073). Among patients with CPA, tumor localization was right-sided in 8 cases (72.7%) and left-sided in 3 cases (27.3%). Tumor size ranged from 30 to 60 mm, with a median diameter of 35 mm. The Ki-67 proliferation index ranged from < 1% to 3%, with a median value of 1%, and was reported as < 1% in two cases. Necrosis was identified in one patient, and mitotic activity ranged from 0 to 4 per high-power field, with most cases demonstrating 0–1 mitoses (Table 1). Table 1Demographic characteristics of study groups and clinicopathological features of CPA patientsVariablesCPA n = 11Control n = 15 p Age46.09 ± 12.3645.27 ± 10.480.856*(d=-0.073)Sex (female/male)10/110/50.197†(V = 0.284)Tumor location (right/left)8/3NATumor size (mm)35 (30–60)NAKi-67%1 (< 1–3)NAWeiss score0 (0–2)NANecrosis, (present/absent)1/10NAMitotic activity (per HPF)1 (0–4)NA* Student’s t-test; effect size reported as Cohen’s d† Fisher’s exact test; effect size reported as Cramer’s VKi-67 was < 1% in two casesNA: not applicable; variables not assessed in the control group Demographic characteristics of study groups and clinicopathological features of CPA patients * Student’s t-test; effect size reported as Cohen’s d † Fisher’s exact test; effect size reported as Cramer’s V Ki-67 was < 1% in two cases NA: not applicable; variables not assessed in the control group Hormonal evaluation showed a median basal morning ACTH level of 2.7 (1.5-9) pg/mL (normal range 10–60 pg/ml), consistent with ACTH-independent hypercortisolism. The mean serum cortisol concentration was 15.84 ± 3.8 µg/dL. Following the 1-mg dexamethasone suppression test, cortisol levels remained elevated, with a median value of 5.77 µg/dL. Twenty-four-hour urinary free cortisol levels demonstrated substantial variability, with a median value of 141.5 (13.68–570) µg/24 h (normal range: 55.5–286 µg/24 h). Midnight serum cortisol levels were also elevated, with a median value of 8.94 (4.4–9.5) µg/dL. The percentage of cells stained with apelin was comparable between the CPA and control groups (28.86 ± 24.88 vs. 31.33 ± 23.26; p = 0.685; r = 0.085), indicating no significant difference in the overall proportion of stained cells. In contrast, the distribution of apelin staining intensity differed significantly between groups (p = 0.026; V = 0.591). Strong apelin staining was observed in 5 of 11 CPA tissues (45.5%), whereas no strong staining was detected in the control group. Moderate staining was present in 4 of 11 CPA samples (36.4%) and in 7 of 15 control tissues (46.7%), while weak staining occurred in 1 of 11 CPA samples (9.1%) and in 5 of 15 control samples (33.3%). No staining was observed in 1 of 11 CPA tissues (9.1%) and in 3 of 15 control tissues (20.0%) (Fig. 1A-F). The mean apelin IRS was higher in the CPA group compared with the control group (5.09 ± 3.36 vs. 2.93 ± 2.12); however, this difference did not reach statistical significance (p = 0.088; r = 0.340). Similarly, apelin expression categories derived from IRS classification did not differ significantly between groups (p = 0.189; V = 0.446). Moderate apelin expression was detected in 7 of 11 CPA tissues (63.6%) and in 7 of 15 control tissues (46.7%), whereas high expression was observed in 2 of 11 CPA samples (18.2%) and was not detected in the control group (Table 2; Fig. 2). In a sensitivity analysis restricted to nephrectomy-derived controls (n = 12), the higher apelin staining intensity in CPA tissues remained statistically significant (p = 0.013). A Normal adrenal cortex (hematoxylineosin, H&E, ×10). B Cortisol-producing adrenal adenoma (H&E, ×20). C No apelin staining in normal adrenal cortex. D Weak apelin staining intensity in CPA. E Moderate apelin staining intensity in CPA. F Strong apelin staining intensity in CPA Table 2Comparison of apelin and VEGF immunohistochemical expression between CPA and control groupsVariablesCPA n = 11Control n = 15 p Apelin %28.86 ± 24.8831.33 ± 23.260.685¥ (r = 0.085)Apelin staining intensity 0.026 † (V = 0.591) – No staining n (%)1 (9.09)3 (20)– Weak n (%)1 (9.09)5 (33.33)– Moderate n (%)4 (36.36)7 (46.67)– Strong n (%)5 (45.45)0 (0)Apelin IRS5.09 ± 3.362.93 ± 2.120.088¥ (r = 0.340)Apelin expression group n (%)0.189 † (V = 0.446)– Negative apelin expression1 (9.09)3 (20)– Low apelin expression1 (9.09)5 (33.33)– Moderate apelin expression7 (63.64)7 (46.67)– High apelin expression2 (18.18)0 (0)VEGF %54.77 ± 28.2341.00 ± 22.520.140¥ (r = 0.294)VEGF staining intensity0.781 † (V = 0.255)– No staining n (%)1 (9.09)0 (0)– Weak n (%)2 (18.18)4 (26.67)– Moderate n (%)6 (54.55)9 (60)– Strong n (%)2 (18.18)2 (13.33)VEGF IRS5.45 ± 3.364.67 ± 3.180.309¥ (r = 0.205)VEGF expression group n (%)0.500 † (V = 0.308)– Negative VEGF expression1 (9.09)0 (0)– Low VEGF expression3 (27.27)6 (40)– Moderate VEGF expression4 (36.36)7 (46.67)– High VEGF expression3 (27.27)2 (13.33)†Fisher exact test; effect size reported as Cramer’s V. ¥ Wilcoxon rank-sum test; effect size reported as rank-biserial correlation Comparison of apelin and VEGF immunohistochemical expression between CPA and control groups †Fisher exact test; effect size reported as Cramer’s V. ¥ Wilcoxon rank-sum test; effect size reported as rank-biserial correlation Distribution of apelin immunoreactivity scores in CPA and control tissues Interobserver agreement was high for both markers. For apelin staining intensity, pairwise linearly weighted Cohen’s kappa coefficients were 0.890 for Pathologists 1 and 2, 0.889 for Pathologists 1 and 3, and 0.927 for Pathologists 2 and 3 (all p < 0.001). For VEGF staining intensity, the corresponding kappa coefficients were 0.805, 0.852, and 0.948, respectively (all p < 0.001). In an exploratory ROC analysis, apelin staining intensity showed moderate discrimination between CPA and control adrenal cortex tissues, with an area under the curve (AUC) of 0.776 (SE = 0.099, 95% CI: 0.581–0.971, p = 0.018) (Fig. 3). Bootstrap resampling (5,000 iterations) yielded a 95% confidence interval for the AUC of 0.573–0.936, consistent with the asymptotic confidence interval. ROC curve illustrating the discriminatory ability of apelin staining intensity to distinguish CPA from normal adrenal cortex tissues. The AUC was 0.776 (95% CI: 0.581–0.971, p = 0.018), indicating moderate discrimination For VEGF, the proportion of VEGF-stained cells did not differ significantly between CPA and control tissues (54.77 ± 28.23 vs. 41.00 ± 22.52; p = 0.140; r = 0.294). Although the mean value appeared higher in the CPA group, this difference was not statistically confirmed (Table 2). Consistent with this finding, VEGF staining intensity did not differ significantly between the groups (p = 0.781; Cramer’s V = 0.255). In the CPA group, moderate staining was the most frequent pattern and was observed in 6 of 11 tissues (54.5%), followed by weak staining in 2 of 11 samples (18.2%) and strong staining in 2 of 11 samples (18.2%), while no staining was detected in 1 of 11 samples (9.1%). In the control group, moderate staining was also the predominant pattern and was present in 9 of 15 tissues (60.0%), followed by weak staining in 4 of 15 preparations (26.7%) and strong staining in 2 of 15 preparations (13.3%); no negative staining was observed in the control tissues (Fig. 4A-D). The mean VEGF IRS was 5.45 ± 3.36 in the CPA group and 4.67 ± 3.18 in the control group, with no statistically significant difference between groups (p = 0.309; r = 0.205), although the mean score appeared higher in the CPA group. Likewise, VEGF expression categories derived from IRS classification did not differ significantly between CPA and control tissues (p = 0.500; V = 0.308). Based on IRS classification, negative VEGF expression was observed in 1 of 11 CPA tissues (9.1%) and in none of the control tissues. Low expression was detected in 3 of 11 CPA samples (27.3%) and in 6 of 15 control tissues (40.0%). Moderate expression was present in 4 of 11 CPA samples (36.4%) and in 7 of 15 control tissues (46.7%), whereas high expression was observed in 3 of 11 CPA tissues (27.3%) and in 2 of 15 control tissues (13.3%) (Fig. 5). A No VEGF staining in CPA. B Weak VEGF staining intensity in CPA. C Moderate VEGF staining intensity in CPA. D Strong VEGF staining intensity in CPA Distribution of VEGF immunoreactivity scores in CPA and control tissues Bias-reduced univariable logistic regression analysis was performed to evaluate the association between immunohistochemical parameters and the presence of CPA. Apelin staining intensity was associated with CPA (OR = 2.96, 95% CI: 1.02–8.60, p = 0.046); however, the wide confidence interval indicates considerable uncertainty in the effect estimate. In contrast, apelin percentage (OR = 1.00, p = 0.826), apelin IRS (OR = 1.31, p = 0.114), VEGF percentage (OR = 1.02, p = 0.220), and VEGF IRS (OR = 1.07, p = 0.575) were not significantly associated with CPA (Table 3). Table 3Univariable bias-reduced logistic regression analysis of apelin and VEGF immunohistochemical parameters associated with CPAVariableOR95% CI p Apelin %1.000.96 – 1.030.826Apelin IRS1.310.94 – 1.830.114Apelin staining intensity2.961.02–8.60 0.046 VEGF %1.020.99 – 1.050.220VEGF IRS1.070.84 – 1.380.575For Apelin IRS and VEGF IRS, ORs represent the change in the odds of CPA per 1-unit increase in the respective IRS (0–12) scaleAbbreviations: OR, odds ratio; CI, confidence interval; IRS, immunoreactivity score Univariable bias-reduced logistic regression analysis of apelin and VEGF immunohistochemical parameters associated with CPA For Apelin IRS and VEGF IRS, ORs represent the change in the odds of CPA per 1-unit increase in the respective IRS (0–12) scale Abbreviations: OR, odds ratio; CI, confidence interval; IRS, immunoreactivity score Overall, among the investigated markers, only apelin staining intensity demonstrated a statistically significant difference between CPA and control adrenal cortex tissues, with a large effect size. Other apelin-related parameters and all VEGF-related variables did not show statistically significant differences between groups. Apelin and VEGF immunohistochemical expression profiles in CPA and control adrenal cortex tissues are summarized in Table 2. Spearman correlation analysis was performed to evaluate the associations between apelin IRS, VEGF IRS, and clinical, pathological, and hormonal parameters in the CPA group. No significant correlation was observed between apelin IRS and VEGF IRS (r = 0.210, p = 0.535). As shown in Table 4, no statistically significant correlations were identified between apelin IRS and tumor size (r = 0.017; p = 0.961), Ki-67 index (r = 0.493; p = 0.124), mitotic count (r = 0.389; p = 0.236), serum ACTH level (r = − 0.019; p = 0.955), basal morning cortisol level (r = − 0.267; p = 0.428), 24-h urinary cortisol level (r = − 0.304; p = 0.364), or midnight cortisol level (r = − 0.459; p = 0.214). Similarly, VEGF IRS was not significantly associated with tumor size (r = 0.286; p = 0.394), Ki-67 index (r = 0.061; p = 0.859), mitotic count (r = − 0.202; p = 0.552), ACTH level (r = − 0.024; p = 0.944), basal morning cortisol level (r = 0.263; p = 0.435), 24-h urinary cortisol level (r = 0.368; p = 0.266), or midnight cortisol level (r = − 0.031; p = 0.937). Overall, no statistically significant association was detected between apelin IRS or VEGF IRS and tumor characteristics, proliferative markers, or hormonal parameters in the CPA group. These findings should be interpreted cautiously given the limited sample size. Table 4Spearman correlation analysis between apelin IRS, VEGF IRS, and clinicopathological / hormonal parameters in the CPA groupVariableApelin IRSVEGF IRS r p r p VEGF IRS0.2100.535Tumor size0.0170.9610.2860.394Ki-67%0.4930.1240.0610.859Mitotic count0.3890.236-0.2020.552ACTH-0.0190.955-0.0240.944Serum cortisol-0.2670.4280.2630.43524-h urinary cortisol-0.3040.3640.3680.266Midnight cortisol-0.4590.214-0.0310.937Correlations were calculated using Spearman’s rho Spearman correlation analysis between apelin IRS, VEGF IRS, and clinicopathological / hormonal parameters in the CPA group Correlations were calculated using Spearman’s rho Discussion In the present study, we evaluated the immunohistochemical expression patterns of apelin and VEGF in CPA and normal adrenal cortex tissues. Apelin staining intensity was significantly higher in CPA tissues compared with normal adrenal cortex. Although the overall proportion of apelin-positive cells did not differ substantially between groups, strong apelin staining was observed predominantly in CPA tissues, indicating a difference in staining intensity rather than an increase in overall apelin expression. The discrepancy between staining intensity and IRS findings may be explained by the composite nature of the IRS, which incorporates both staining intensity and the proportion of positive cells. In our cohort, while staining intensity differed significantly between groups, the proportion of stained cells remained comparable, potentially diluting the overall IRS difference. In contrast, VEGF expression was detected in both CPA and control adrenal cortex samples, and no statistically significant differences were identified between the groups in terms of VEGF staining parameters. These findings suggest that VEGF expression may reflect the intrinsically rich vascular microenvironment of the adrenal cortex rather than a tumor-specific alteration in CPA. Overall, these findings indicate higher apelin staining intensity in CPA tissues without evidence of increased overall apelin expression or activation of apelin signaling. However, the wide confidence interval of the logistic regression estimate indicates considerable uncertainty, and this analysis should be regarded as complementary to, rather than an independent confirmation of, the between-group comparison of apelin staining intensity. Similarly, the ROC analysis should be considered exploratory, given the small sample size and wide confidence interval around the AUC estimate. Moreover, because apelin staining was assessed in postoperative histological specimens, its discriminatory ability should not be interpreted as evidence of clinical diagnostic utility. Taken together, the higher apelin staining intensity observed in CPA tissues should be regarded as a nominal exploratory finding in the context of multiple statistical comparisons and should be independently replicated. Its biological significance and potential relationship with angiogenic or metabolic pathways also require further investigation. Endocrine glands are among the most highly vascularized organs in the human body, and an adequate vascular network is essential for maintaining hormonal secretion and precise endocrine feedback regulation. The adrenal cortex, similar to other endocrine tissues, possesses a dense and highly organized microvascular structure that supports efficient steroid hormone synthesis and systemic hormone delivery. Consequently, both normal adrenal tissue and adrenal tumors represent highly vascularized structures. There are few studies investigating angiogenesis and angiogenesis-related markers in CS due to CPA and adrenocortical tumors. In a previous immunohistochemical study evaluating endocan, an endothelium-derived proteoglycan involved in angiogenesis, endocan expression and staining intensity were significantly higher in CPA tissues than in normal adrenal cortex [17]. Another study assessing angiogenic activity with the endothelial marker CD31 reported significantly higher CD31 expression in adrenocortical carcinoma (ACC) compared with CS, supporting the notion that angiogenesis may influence the biological behavior of adrenocortical tumors. In that study, lymphangiogenesis was additionally evaluated using podoplanin (D2-40), and D2-40 expression was found to be significantly lower in ACC than in CS, indicating that lymphatic vessel density might be more closely linked to the functional characteristics of the tumor rather than its malignant nature [18]. Previous studies have demonstrated that apelin expression is frequently upregulated in various tumor types and may contribute to tumor progression through its pro-angiogenic effects. Experimental and pathological studies have shown that apelin and its receptor APJ are expressed in tumor cells as well as in endothelial cells within the tumor microenvironment, where they promote endothelial proliferation, vascular remodeling, and neovascularization. Increased apelin signaling has therefore been associated with enhanced angiogenic activity and tumor growth in several malignancies. These findings suggest that apelin may function as an important mediator linking tumor cell activity with angiogenic pathways [19]. However, in our cohort, no significant correlation was observed between Apelin IRS and VEGF IRS. Therefore, although interactions between apelin and VEGF pathways have been proposed in previous studies, our immunohistochemical findings do not demonstrate a direct association between the expression of these two markers in CPA tissues. Interestingly, only a single previous study has evaluated apelin expression in adrenal-related neuroendocrine tumors. In that immunohistochemical study investigating several biomarkers in pheochromocytoma and paraganglioma, apelin expression was reported to be reduced in metastatic tumors compared with non-metastatic lesions, suggesting a potential association between lower apelin expression and tumor aggressiveness [20]. However, in the present study, apelin staining intensity was significantly higher in CPA tissues compared with normal adrenal cortex. This apparent discrepancy may be explained by fundamental differences in tumor biology, as cortisol-producing adrenal adenomas originate from the steroidogenic adrenal cortex, whereas pheochromocytomas and paragangliomas arise from chromaffin cells of the adrenal medulla or extra-adrenal paraganglia. These differences in cellular origin and endocrine function may partly account for the divergent apelin staining patterns observed across tumor types. The presence of apelin immunoreactivity in both CPA and normal adrenal cortex tissues indicates that apelin immunoreactivity is not specific to CPA. In addition to its angiogenic functions, apelin is also involved in metabolic regulation, including glucose homeostasis, insulin sensitivity, and lipid metabolism [21]. Previous studies have demonstrated that circulating apelin levels are elevated in metabolic disorders such as obesity, insulin resistance, and type 2 diabetes mellitus. Experimental and clinical evidence suggests that apelin may act as an adaptive metabolic peptide that is upregulated in response to metabolic stress and altered energy balance [22]. Chronic hypercortisolism in Cushing syndrome is frequently associated with metabolic disturbances, including central obesity and impaired glucose metabolism [4]. Given the established metabolic effects of cortisol excess and the role of apelin in metabolic regulation, a potential relationship between apelin staining intensity and metabolic alterations in CPA may be biologically plausible. However, detailed anthropometric and metabolic parameters were not systematically evaluated in the present study. Therefore, the potential relationship between higher apelin staining intensity and the metabolic consequences of cortisol excess could not be assessed and warrants investigation in future studies. The absence of demonstrable relationships between apelin IRS and clinicopathological or hormonal characteristics in our cohort limits conclusions regarding the potential role of apelin in tumor behavior or hormonal activity. Given the small sample size and limited variability of some pathological features, these relationships remain uncertain and should be further investigated in larger cohorts. Given the reported interactions between apelin and VEGF signaling in angiogenic processes, the relationship between these pathways may also be relevant to adrenal tumor biology. Experimental studies have demonstrated that apelin signaling interacts with VEGF pathways and contributes to endothelial proliferation, vascular maturation, and neovascularization within the tumor microenvironment [10]. Previous studies evaluating VEGF expression in adrenocortical tumors have reported heterogeneous and sometimes conflicting findings. In a recent immunohistochemical study, Karaköse et al. observed significantly higher VEGF expression in adrenocortical carcinoma compared with adrenocortical adenomas [23]. In contrast, Oliveira et al., who evaluated VEGF expression patterns in adrenocortical tumors, found no significant differences among the tumor groups, indicating that the observed VEGF expression patterns were not specifically associated with malignancy [24]. These heterogeneous findings may reflect differences in tumor functionality, biological behavior, and methodological approaches across studies, indicating that the role of VEGF-associated angiogenesis in adrenocortical tumors remains incompletely understood. Consistent with the heterogeneous findings in the literature, our study did not demonstrate a statistically significant difference in VEGF expression between CPA and normal adrenal cortex tissues. This observation may reflect the intrinsically rich vascular microenvironment of the adrenal cortex. Similarly, no significant correlations were observed between VEGF IRS and clinicopathological or hormonal parameters. However, given the small sample size and limited statistical power of the present study, the absence of significant associations should not be interpreted as evidence that VEGF expression is unrelated to clinicopathological or hormonal characteristics in CPA. Although the sample size was limited, the present cohort consisted exclusively of surgically treated, histopathologically confirmed CPA cases with well-preserved archival tissue suitable for immunohistochemical analysis. Relatively small cohorts are common in immunohistochemical studies of CPA, reflecting the limited availability of well-characterized surgical tissue specimens. Despite this limitation, rigorous pathological assessment, blinded evaluation by three pathologists, and robust statistical approaches including bootstrap resampling and bias-reduced regression were used to maximize the reliability of the present findings. The present study has several limitations. The relatively small sample size limited the statistical power of the analyses and the generalizability of the findings. Given the retrospective and exploratory nature of the study, no formal prespecified primary endpoint was defined. Furthermore, multiple statistical comparisons were performed without formal adjustment for multiplicity, which may have increased the risk of type I error. Accordingly, the significant difference observed in apelin staining intensity should be regarded as a nominal exploratory finding that requires independent replication. The corresponding logistic regression estimate was also imprecise, as reflected by the wide 95% confidence interval (OR = 2.96, 95% CI: 1.02–8.60), with the lower bound only marginally exceeding the null value. Although the sex distribution did not differ significantly between the groups, the predominance of women in the CPA group and the limited sample size precluded adequate adjustment for sex. Therefore, residual confounding by sex, a potential biological modulator of apelin and VEGF expression, cannot be excluded. Concurrent 24-hour urinary creatinine measurements were unavailable because of the retrospective design; therefore, the completeness of 24-hour urine collection could not be objectively verified, which should be considered when interpreting urinary free cortisol measurements. Detailed anthropometric and metabolic parameters were also not systematically evaluated, limiting our ability to determine whether the higher apelin staining intensity observed in CPA tissues was associated with metabolic alterations related to cortisol excess. In addition, apelin and VEGF expression were evaluated using semi-quantitative immunohistochemistry, which, despite blinded assessment by three experienced pathologists and consensus scoring, remains an observer-dependent technique. Finally, direct measures of angiogenesis, including microvessel density and endothelial markers such as CD31 or CD34, as well as circulating apelin levels, APJ/APLNR expression, VEGF receptor expression, and complementary molecular analyses, were not assessed. Therefore, the present immunohistochemical findings should be interpreted as differences in immunohistochemical staining patterns rather than direct evidence of increased angiogenesis. Consequently, the findings of this study should be considered exploratory and hypothesis-generating and require validation in larger multicenter studies incorporating complementary quantitative molecular approaches. Conclusion Our findings suggest that apelin staining intensity is higher in CPA than in normal adrenal cortex, indicating altered apelin immunoreactivity in CPA tissues. In contrast, VEGF expression did not differ significantly between CPA and normal adrenal cortex, and no significant associations with clinicopathological or hormonal parameters could be demonstrated in this cohort. The potential relationship between altered apelin immunoreactivity and angiogenesis-associated pathways remains to be clarified, as angiogenesis was not directly assessed in the present study. These observations should be regarded as preliminary and require validation in larger multicenter studies incorporating complementary molecular and vascular analyses. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Acknowledgements None to declare. Author contributions HÇB conceived and designed the study, performed data collection, statistical analysis, and drafted the manuscript. YÜ, EÖ and TG performed histopathological evaluation and immunohistochemical analyses, and contributed to data interpretation. OA supervised the study and critically revised the manuscript for important intellectual content. All authors read and approved the final manuscript. Funding This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors. Data availability The datasets generated and/or analyzed during the current study are not publicly available due to patient confidentiality and institutional policies but are available from the corresponding author on reasonable request. Declarations Ethics approval and consent to participate This study was approved by the Non-Interventional Clinical Research Ethics Committee of Konya City Hospital (Approval No. 2025/116). The study was conducted in accordance with the ethical principles of the Declaration of Helsinki (as revised in 2013). Consent to participate The requirement for individual informed consent was waived by the ethics committee due to the retrospective nature of the study and the use of fully anonymized data. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Supplementary Materials Supplementary Material 1 Data Availability Statement The datasets generated and/or analyzed during the current study are not publicly available due to patient confidentiality and institutional policies but are available from the corresponding author on reasonable request.