The Biochemical Diagnosis of Adrenal Insufficiency With Modern Cortisol Assays

Reappraisal in the Setting of Opioid Exposure and Hospitalization

Caitlin Colling; Lisa Nachtigall; Beverly M. K. Biller; Karen K. Miller

Disclosures

Clin Endocrinol. 2022;96(1):21-29. 

In This Article

Results

Cohort 1 Analysis: Diagnosis of Adrenal Insufficiency in Patients With Chronic Opioid Exposure

Clinical Characteristics. Clinical characteristics of Cohort 1 (N = 75) are shown in Table 1. The majority of patients (57, 76%) were prescribed a single opioid; 18 (24%) were prescribed two opioids. Oxycodone was the most commonly prescribed opioid, in 29 (39%) patients. Fifty-seven percent of patients had a diagnosis of musculoskeletal pain and 48% had abdominal pain. Cancer or opioid dependence/use disorder was diagnosed in 35% and 19% of the cohort, respectively. No patient was diagnosed with adrenal crisis. Overall, 87% of patients had a visit diagnosis of at least one major sign or symptom of adrenal insufficiency.[26] Diagnoses included hypotension in 52%, abdominal pain in 29%, hyponatremia in 24%, and fatigue in 24% of patients (Table S1).

Opioid-induced Adrenal Insufficiency. The prevalence of adrenal insufficiency as indicated by a peak cortisol <405 nmol/L (14.7 μg/dl) in the 75 patients with chronic opioid exposure who underwent cosyntropin stimulation testing was 4% (N = 3). Applying a cutoff of <500 nmol/L (18.1 μg/dl), 18.7% of patients had abnormal cortisol response to exogenous ACTH stimulation (Table 1). Risk factors for adrenal insufficiency were not compared due to low number of patients with the diagnosis.

Cohort 2 Analysis: Diagnosis of Adrenal Insufficiency in Opioid-exposed and Unexposed Patients

Clinical Characteristics. Of the 854 patients who met inclusion criteria, 472 (55%) were female with a median age of 58 (42–72) years (Table 2). The majority of opioid-exposed inpatients (121, 78%) were prescribed a single opioid; 35 (22%) were prescribed two opioids. Oxycodone was the most commonly prescribed opioid with 70 (45%) patients with a prescription.

More than 89% of inpatients undergoing cosyntropin stimulation testing had a visit diagnosis of at least one major sign or symptom of adrenal insufficiency, with opioid-exposed and unexposed patients as likely to have diagnoses of a sign or symptom of adrenal insufficiency (89% vs. 89%).[26] Hypotension, abdominal pain and hyponatremia were the most common signs experienced, diagnosed in 60%, 25%, and 24% of patients, respectively. Only 41% of outpatients undergoing HPA axis testing had a diagnosis of a sign or symptom of adrenal insufficiency associated with the cosyntropin stimulation testing visit (Table S1).

Overall, 4.6% of patients undergoing cosyntropin stimulation testing who also had morning cortisol levels collected were diagnosed with adrenal insufficiency; no difference was observed in the proportion of inpatients and outpatients with adrenal insufficiency (5.1% vs. 4.0%, p = .54). Low morning cortisol was more commonly found in inpatients compared to outpatients (5.4% vs. 1.6%, p = .004).

Performance of Morning Cortisol for the Diagnosis of Adrenal Insufficiency. One out of 7 (14%) outpatients with morning cortisol <83 nmol/L (3 μg/dl) achieved a stimulated value >405 nmol/L (14.7 μg/dl) compared with 14 out of 22 (64%) inpatients. The sensitivity of a low morning serum cortisol (<83 nmol/L [3 μg/dl]) for abnormal cosyntropin stimulation test (peak cortisol <405 nmol/L [14.7 μg/dl]) was 41.0% (95% confidence interval [CI]: 27.1%–56.6%) with a specificity of 98.2% (97.0%–98.9%) in the entire sample. For inpatients, the specificity was 96.4% (94.0%–97.8%) with a specificity of 95.1% (90.3%–97.6%) in opioid-exposed and 97.1% (94.2%–98.6%) in opioid-unexposed inpatients. For opioid-unexposed outpatients, the specificity was 99.8% (98.7%–100.0%). Only one opioid-exposed outpatient had a morning cortisol <83 nmol/L (3 μg/dl), thus specificity was not calculated.

Use of morning cortisol to rule out adrenal insufficiency was examined in hospitalized patients. A morning basal cortisol of <348 nmol/L (12.6 μg/dl) had 100% sensitivity (95% CI: 84.5%–100%) for adrenal insufficiency. A cutoff of <276 nmol/L (10 μg/dl) had a sensitivity of 90.5% (71.1%–97.3%) for the diagnosis of adrenal insufficiency (Table 3, Figure 2).

Figure 2.

ROC curve of morning cortisol level for the diagnosis of adrenal insufficiency with the cosyntropin stimulation test as the reference test. Black line represents ROC curve for hospitalized patients (AUC = 0.86). Grey line represents ROC curve for outpatients (AUC = 0.85). AUC, area under the curve; ROC, receiver operating characteristic

processing....