Diagnostic Value of Serum Neutrophil Gelatinase-Associated Lipocalin for Adrenal Insufficiency

Abstract

The adrenocorticotropic hormone (ACTH) stimulation test is the gold standard for confirming adrenal insufficiency (AI); however, its requirement for frequent blood draws and pharmacologic stimulation can be a barrier to early diagnosis. This study aims to evaluate the diagnostic value of serum neutrophil gelatinase-associated lipocalin (NGAL) as a novel biomarker for diagnosing AI. This prospective study was conducted at Chonnam National University Hwasun Hospital between June 2020 and November 2021. Patients with either pituitary tumors or a history of exogenous steroid use who underwent the rapid ACTH stimulation test were included in the analysis. A total of 75 patients suspected of having AI were analyzed, with 31 patients (41.3%) being diagnosed with AI. Serum NGAL levels were significantly higher in the secondary AI group with pituitary tumors compared to the non-AI group (92.2±33.6 ng/mL vs. 74.7±24.1 ng/mL, p=0.037). There was a negative correlation between serum NGAL levels and cortisol levels 60 minutes after ACTH administration in patients with pituitary tumors (r=-0.246, p=0.043). A positive correlation was found between serum NGAL and baseline serum creatinine levels in the secondary AI group with pituitary tumors (r=0.400, p=0.002). The Receiver operating characteristic curve analysis revealed that the area under the curve for secondary AI was 0.673, with an optimal cut-off value of 78.4 ng/mL, yielding a sensitivity of 60.0% and specificity of 59.4% (p=0.038). Serum NGAL may serve as a supportive diagnostic tool for secondary AI when confirmatory tests are not feasible.

INTRODUCTION

Adrenal insufficiency (AI) is a complex condition characterized by an absolute or relative deficiency of cortisol production from the adrenal glands. This can result from primary adrenal disorders, secondary AI due to adrenocorticotropic hormone (ACTH) deficiency, or suppression of ACTH by exogenous corticosteroids.1 The most common cause of AI is tertiary AI, which occurs due to exogenous corticosteroid use (also referred to as drug-induced or iatrogenic AI), followed by secondary AI, while primary AI is relatively rare.2 Several long-term follow-up studies have demonstrated increased morbidity and mortality, along with a decreased quality of life, in patients with AI.3,4,5 Therefore, early recognition and diagnosis of AI, coupled with appropriate corticosteroid replacement, are essential to reduce the risk of mortality. AI can be confirmed through specific ACTH stimulation tests, such as the cosyntropin stimulation test. However, diagnosing AI remains challenging due to the variable and often vague symptoms and signs associated with the condition. Additionally, the need for cosyntropin injections and repeated blood draws can hinder timely diagnosis.

Recent studies have evaluated the diagnostic and prognostic value of neutrophil gelatinase-associated lipocalin (NGAL) in relation to inflammation, metabolic complications, and various conditions, including renal, cardiovascular, and malignant diseases.6,7 NGAL, also known as lipocalin-2, is a protein that transports small, hydrophobic molecules, including steroids, free fatty acids, and hormones, to target organs.8 It is primarily secreted by immune cells such as activated neutrophils and macrophages, but is also produced by various human tissues, including tubular cells in the kidneys, heart, lungs, and adipocytes.7,9,10

While corticosteroids are associated with inflammation and various metabolic complications, research on the relationship between NGAL and corticosteroid deficiency is limited. A previous study found that dexamethasone, a synthetic glucocorticoid, increased lipocalin-2 gene expression in a dose-dependent manner in the adipose tissue of premenopausal women.11 To our knowledge, no studies have specifically investigated the relationship between NGAL levels and corticosteroid deficiency status. Thus, this study aimed to evaluate the association of serum NGAL levels in patients with AI, including secondary and drug-induced AI.

MATERIALS AND METHODS

1. Study population

We conducted a prospective study between June 2020 and November 2021 at Chonnam National University Hwasun Hospital, a tertiary referral center in South Korea. The study included patients with suspected AI based on clinical symptoms and signs, or a history of chronic exogenous corticosteroid use, who visited the endocrinology department prior to pituitary or orthopedic surgery. The eligibility criteria were as follows: (1) patients over 18 years of age, (2) patients who underwent a rapid ACTH stimulation test to determine the presence of AI, and (3) patients who provided informed consent. Patients with medical conditions that could affect serum NGAL levels were excluded from the study. The exclusion criteria included individuals with active malignancies (solid or hematologic), those presenting with acute infections or inflammation, patients currently receiving corticosteroid or immunosuppressive therapy, and those with chronic autoimmune diseases such as systemic lupus erythematosus. Additionally, we excluded patients who were suspected of having or had already been diagnosed with Cushing’s syndrome.

A total of 75 patients were included in the analysis. We evaluated the clinical characteristics of these patients, including age, sex, body mass index (BMI), and the prevalence of comorbidities such as hypertension, type 2 diabetes mellitus (T2D), dyslipidemia, cardiovascular disease (CVD), and chronic kidney disease (CKD) based on medical records. The estimated glomerular filtration rate (eGFR) was calculated using the Modification of Diet in Renal Disease equation (mL/min/1.73 m²), with CKD defined as an eGFR of less than 60 mL/min/1.73 m².12 We compared serum NGAL concentrations between the AI group and the non-AI group, and conducted a subgroup analysis between the secondary AI group and the drug-induced AI group.

This study was approved by the Institutional Review Board of Chonnam National University Hwasun Hospital (IRB No. CNUHH-2020-117), and it was conducted in accordance with the guidelines of the Declaration of Helsinki. All patients provided written informed consent.

2. Rapid ACTH stimulation test and biochemical measurements

To confirm AI, a rapid ACTH stimulation test was conducted using 250 µg of synthetic ACTH (Synacthen; Novartis, Basel, Switzerland). Considering the biological half-life of exogenously administered corticosteroids, the test was performed at least 48 hours after the last corticosteroid dose. The test was conducted in the morning between 8 and 9 a.m. with freshly prepared 250 µg of Synacthen. Intravenous synthetic ACTH was administered after baseline blood samples were drawn for serum cortisol and plasma ACTH. Additional blood samples for serum cortisol were collected at 30 and 60 minutes after ACTH administration. Plasma ACTH levels were measured using an immunoradiometric assay (Cisbio Bioassays International, Codolet, France), and serum cortisol was measured by radioimmunoassay (RIAZENco CORTISOL, R-JG-100, ZenTech, Liege, Belgium). AI was defined as a peak cortisol level of less than 18 µg/dL (500 nmol/L) at either 30 or 60 minutes after ACTH injection.13

3. Measurement of serum NGAL concentration

Blood samples were collected in serum-separating tubes and centrifuged. The supernatant was aliquoted into cryotubes and stored in a -80℃ deep freezer. Samples for the quantitative analysis of NGAL were collected at once and measured using the ELISA method (R&D Systems, human lipocalin-2/NGAL Immunoassay, DLCN20). The detection limit was 0.014 ng/mL, with an assay range of 0.2-10 ng/mL. The coefficient of variation (CV) for intra-assay precision ranged from 3.6% to 4.4%, and for inter-assay precision, from 5.6% to 7.9%. For enhanced reliability, NGAL tests were performed in duplicate, and the mean value was used for analysis.

4. Statistical analysis

Data are expressed as mean±standard deviation or as n (%). Continuous variables were analyzed using Student’s t-test, while categorical variables were analyzed with the Chi-square test. Pearson correlation was used for correlation analysis. The diagnostic accuracy of NGAL for AI was assessed using receiver operating characteristic (ROC) curve analysis, with the area under the curve (AUC) employed to evaluate diagnostic performance. The optimal cut-off value was determined as the concentration at which the Youden index (sensitivity+specificity-1) was maximized. The observed difference in NGAL between the AI and non-AI groups corresponded to a moderate effect size (Cohen’s d=0.60), with an estimated statistical power of approximately 72%-78% for the available sample size of 75 participants. Post-hoc subgroup analyses demonstrated a moderate-to-large effect size in the pituitary tumor subgroup (Cohen’s d=0.70) and a small effect size in the exogenous steroid subgroup (Cohen’s d=0.27), indicating that the latter subgroup was likely insufficiently powerful to detect small between-group differences. All statistical analyses were performed using SPSS Statistics, version 26 (IBM, Armonk, NY) and MedCalc. A p-value <0.05 was considered statistically significant.

RESULTS

1. Baseline characteristics of study population

Of the 75 patients who underwent a rapid ACTH stimulation test before surgery, 31 patients (41.3%) were diagnosed with AI. The mean age was 59.7±18.0 years, and 46 patients (61.3%) were women. Among patients with AI, serum cortisol levels were 6.5±4.4 µg/dL at baseline, 11.3±4.6 µg/dL at 30 minutes, and 12.7±5.2 µg/dL at 60 minutes following ACTH stimulation. There were no significant differences in age, sex, BMI, or the prevalence of comorbidities–including hypertension, T2DM, CVD, dyslipidemia, and CKD–between the AI and non-AI groups (all p>.05). Additionally, biochemical parameters such as neutrophil count, blood glucose, sodium, and albumin levels did not differ between the two groups (all p>0.05). However, baseline serum creatinine levels were higher in the AI group compared to the non-AI group (0.88±0.27 mg/dL vs. 0.73±0.24 mg/dL, p=0.011), and the eGFR was lower in the AI group (83.5±23.0 mL/min/1.73 m² vs. 97.9±25.3 mL/min/1.73 m², p=0.015) (Table 1). Patients were categorized into two groups based on the suspected cause of AI: those with pituitary tumors (n=50) and those with a history of exogenous steroid use (n=25). In patients with pituitary tumors, those with secondary AI had higher baseline serum creatinine levels (0.93±0.30 mg/dL vs. 0.70±0.24 mg/dL, p=0.003) and lower eGFR (82.7±24.7 mL/min/1.73 m² vs. 104.4±25.6 mL/min/1.73 m², p=0.005) compared to the non-AI group. Among patients with a history of exogenous steroid use, there were no significant differences in baseline serum creatinine levels or eGFR between those with drug-induced AI and the non-AI group (Supplementary Table 1). Acute kidney injury (AKI) was defined by the following criteria: (1) an increase in serum creatinine by ≥0.3 mg/dL within 48 hours, (2) an increase in serum creatinine ≥1.5 times the baseline value within the prior 7 days, or (3) urine output <0.5 mL/kg/h for 6 hours.14 Based on these criteria, no patients had AKI at the time of their serum NGAL level measurements.

TABLE 1. Baseline characteristics of patients according to the presence of AI.

Values are presented as mean±standard deviation or n (%). AI: adrenal insufficiency, BMI: body mass index, CVD: cardiovascular disease, CKD: chronic kidney disease, HDL-C: high-density lipoprotein cholesterol, LDL-C: low-density lipoprotein cholesterol, eGFR: estimated glomerular filtration rate, WBC: white blood cell count, ACTH: adrenocorticotropic hormone, NGAL: neutrophil gelatinase-associated lipocalin. ^aHbA1c was measured in 59 patients.

2. Comparison of serum NGAL levels between the AI and non-AI groups

Serum NGAL levels were higher in the AI group compared to the non-AI group (96.7±32.0 ng/mL vs. 79.6±29.0 ng/mL, p=0.018) (Table 1 and Fig. 1A). Among patients with pituitary tumors, serum NGAL levels in the AI group were significantly higher than in the non-AI group (92.2±33.6 ng/mL vs. 74.7±24.1 ng/mL, p=0.037; Fig. 1B). However, there were no significant differences in serum NGAL levels between the drug-induced AI group and the non-AI group (103.9±29.3 ng/mL vs. 91.4±36.7 ng/mL, p=0.359; Fig. 1C).

FIG. 1. Comparison of serum NGAL levels according to the presence of AI. (A) All patients. (B) Patients with pituitary tumors. (C) Patients with a history of exogenous steroid use. AI: adrenal insufficiency. ^*p<0.05.

3. Correlation analysis of serum NGAL levels and peak cortisol levels

Pearson correlation analysis showed a negative relationship between serum NGAL levels and cortisol levels at 60 minutes after ACTH administration in patients with pituitary tumors (r=-0.246, p=0.043; Fig. 2A). However, there was no correlation between serum NGAL levels and cortisol levels at 60 minutes in patients with a history of recurrent exogenous steroid use (r=-0.008, p=0.485; Fig. 2B).

FIG. 2. Correlation analysis of serum NGAL level and peak cortisol level. (A) Scatter plot (with 95% confidence interval) in patients with pituitary tumors. (B) Scatter plot (with 95% confidence interval) in patients with a history of exogenous steroid use.

4. Correlation analysis of serum NGAL levels and baseline creatinine levels

Pearson correlation analysis revealed a positive relationship between serum NGAL levels and baseline creatinine levels in patients with pituitary tumors (r=0.400, p=0.002; Fig. 3A). Meanwhile, no correlation was observed between serum NGAL levels and baseline creatinine levels in patients with a history of recurrent exogenous steroid use (r=-0.172, p=0.216; Fig. 3B).

FIG. 3. Correlation analysis of serum NGAL level and baseline creatinine level. (A) Scatter plot (with 95% confidence interval) in patients with pituitary tumors. (B) Scatter plot (with 95% confidence interval) in patients with a history of exogenous steroid use.

5. The diagnostic value of serum NGAL in patients with AI

ROC curve analysis was performed to assess the diagnostic utility of serum NGAL levels in patients with AI. The AUC was 0.657 for AI in the overall patient population (p=0.018, Fig. 4A), 0.673 for secondary AI (p=0.038, Fig. 4B), and 0.609 for drug-induced AI (p=0.343, Fig. 4C). The optimal cut-off value for diagnosing AI across all patients was determined to be 83.7 ng/mL, yielding a sensitivity of 60.6% and specificity of 62.2%. For secondary AI, the cut-off value was established at 78.4 ng/mL, with a sensitivity of 60.0% and specificity of 59.4%. The optimal cut-off value for drug-induced AI was 91.2 ng/mL, demonstrating a sensitivity and specificity of 61.5% (Table 2).

FIG. 4. Receiver operating characteristic curve analysis of serum NGAL level for diagnosing AI. (A) All patients. (B) Patients with pituitary tumors. (C) Patients with a history of exogenous steroid use.

TABLE 2. Diagnostic value of serum NGAL levels for AI.

AUC: area under the curve.

DISCUSSION

This study demonstrated that serum NGAL levels were elevated in the secondary AI group compared to the non-AI group. Additionally, there was a correlation between peak cortisol levels and serum NGAL levels among patients with secondary AI. These findings were associated with higher serum creatinine levels and lower eGFR in patients with secondary AI. However, these correlations were not observed in patients with drug-induced AI.

AI is a potentially life-threatening condition, making early diagnosis and appropriate corticosteroid replacement essential to prevent adrenal crisis and reduce mortality.3,15 Although the rapid ACTH stimulation test is a well-established diagnostic tool for AI,16 it has certain limitations, including the inconvenience of repeated blood sampling and the complexity of the stimulation procedure. Therefore, the development of new, more accessible diagnostic tests for suspected AI could be beneficial. From among the several potential candidates, we selected and evaluated the diagnostic value of serum NGAL in patients with AI.

NGAL, a member of the lipocalin superfamily known as lipocalin-2, is thought to play a role in steroid binding. Serum and urinary NGAL levels have been shown to increase in patients with metabolic complications, such as obesity, diabetes, and nonalcoholic fatty liver disease.17,18 Recent studies have reported serum and urine NGAL as diagnostic biomarkers and predictors for various conditions,19,20 however, the role of NGAL in cortisol metabolism has been rarely evaluated. Only a limited number of studies have investigated the relationship between NGAL and corticosteroids, yielding inconsistent results.18,21 Therefore, this study aimed to investigate the association between serum NGAL levels and the hypothalamic-pituitary-adrenal (HPA) axis in the diagnosis of secondary AI among patients with pituitary tumors. Furthermore, we assessed serum NGAL levels in individuals with a history of exogenous steroid use who were suspected of having tertiary AI.

NGAL, a glycoprotein with a molecular weight of 25 kDa, is elevated in response to tubular damage and is released during renal regeneration, thereby serving as a recognized biomarker for AKI and a predictor of progressive renal dysfunction and mortality.7,22 Hormonal alterations, such as those involving erythropoietin and thyroid hormones, can affect kidney function, though the causality and underlying mechanisms remain unclear.23,24 While cortisol metabolism is partially influenced by the kidneys,25 the relationship between the HPA axis and kidney dysfunction has been sparsely studied, and the association between cortisol and AKI has not been evaluated.26 Small studies and case reports have shown that primary AI, particularly Addison’s disease, can result in renal injury, but both mineralocorticoid and glucocorticoid pathways are disrupted in primary AI.27 A case report in Korea described a patient with AI presenting hypercalcemia and AKI as initial symptom.28 Our study found higher serum creatinine levels and lower eGFR in the AI group compared to the non-AI group. Subgroup analyses based on the cause of AI revealed reduced renal function in the secondary AI group, even when eGFR remained within the normal range, with no significant difference in the prevalence of CKD. In contrast, the drug-induced AI group did not exhibit any differences in renal function compared to the non-AI group. This discrepancy may indicate that, in pituitary-related secondary AI, more chronic and hemodynamically significant cortisol deficiencies contribute to both tubular injury and functional reductions in renal filtration. In contrast, NGAL elevation in drug-induced AI may be influenced by factors other than current endogenous cortisol levels, including residual effects of prior glucocorticoid exposure and coexisting metabolic conditions such as diabetes. Because NGAL is considered a marker of subclinical tubular injury that rises before overt deterioration in renal function, chronic metabolic dysfunction associated with diabetes may promote oxidative stress, low-grade inflammation, and tubular epithelial injury, leading to elevated urinary or serum NGAL levels even in the absence of measurable reductions in glomerular filtration.29,30 Additionally, serum creatinine may be less sensitive in this population because of altered muscle mass related to prior glucocorticoid exposure.

Previous experimental and clinical studies have established the diagnostic utility of serum and urine NGAL as early biochemical markers for AKI.31 Based on the findings of the present study and the established role of serum NGAL as a biomarker for AKI, serum NGAL may offer diagnostic utility for secondary AI in the context of early renal impairment. The ROC analysis indicated that the AUC for diagnosing secondary AI was 0.673, with sensitivity and specificity both approximating 60%. The optimal cut-off value for serum NGAL was determined to be 78.4 ng/mL. Consequently, these findings suggest that serum NGAL may serve as an adjunctive marker in the diagnosis of secondary AI.

This study has several limitations. First, it was conducted at a single center with a relatively small sample size. Although the overall cohort included 75 patients, the subgroup analysis of exogenous steroid–related AI was comprised of only 25 participants. Accordingly, the findings from the exogenous steroid subgroup should be interpreted as exploratory and hypothesis-generating. Second, it was conducted in patients hospitalized for pituitary or orthopedic surgeries, which may have introduced variations in stress levels. Lastly, follow-up testing for NGAL was not conducted, which limits the ability to evaluate serum NGAL as a predictor of AI and chronic renal dysfunction. Despite these limitations, our study is noteworthy as it is, to our knowledge, the first to investigate the diagnostic role of serum NGAL in AI, emphasizing the reduced renal function in patients with secondary AI. This finding highlights the necessity for careful monitoring of renal function in patients with secondary AI.

In conclusion, serum NGAL levels were found to be elevated and correlated with peak cortisol levels in patients with secondary AI. Moreover, serum NGAL levels were associated with serum creatinine levels, suggesting that NGAL may serve as an early biomarker for reduced renal function in secondary AI patients. Future large-scale, follow-up studies are necessary to further elucidate the diagnostic and predictive value of serum NGAL in patients with AI.

SUPPLEMENTARY MATERIAL

Supplementary Table 1

Comparison of baseline renal function between AI and non-AI group