Workup for Cushing's Syndrome
Initial Step: Rule Out Exogenous Glucocorticoids
Before any biochemical testing, exclude iatrogenic Cushing's syndrome by thoroughly reviewing all glucocorticoid use including oral, inhaled, topical, and injected formulations. 1, 2, 3
- This is a strong recommendation based on high-quality evidence that failure to exclude exogenous glucocorticoid use leads to unnecessary testing without patient benefit. 1
First-Line Screening Tests
For patients with intermediate to high clinical suspicion, perform 2-3 of the following first-line screening tests: 1, 2, 3
Late-Night Salivary Cortisol (LNSC)
- Collect at least 2-3 samples on consecutive days for optimal sensitivity (95%) and specificity (100%). 3, 4
- This test is preferred for low clinical suspicion cases due to easier patient compliance compared to 24-hour urine collection. 2, 3
- Multiple LNSC collections are more convenient than urine collections. 3
24-Hour Urinary Free Cortisol (UFC)
- Collect 2-3 samples to account for day-to-day variability, with diagnostic cutoff >193 nmol/24h (>70 μg/m²). 3, 4
- Sensitivity is 89% and specificity is 100%. 4
Overnight 1 mg Dexamethasone Suppression Test (DST)
- Administer 1 mg dexamethasone at midnight and measure serum cortisol at 8 AM; normal response is cortisol <1.8 μg/dL (50 nmol/L). 3, 4
- Sensitivity is 95% and specificity is 80%. 4
- Measuring dexamethasone levels along with cortisol improves test interpretability. 1, 3
- Less useful in women taking estrogen-containing oral contraceptives. 3
Interpretation of Screening Results
- If any test is abnormal, repeat 1-2 screening tests to confirm the diagnosis. 2, 3
- If all tests are normal and clinical suspicion is low to moderate, Cushing's syndrome is unlikely. 1, 3
- If tests are normal but clinical suspicion remains high, refer to an endocrinologist for further evaluation. 1
Important Caveats for False Positives
Consider pseudo-Cushing's states that can cause false-positive results: 2, 3
In cases with inconsistent results, consider cyclic Cushing's syndrome. 3
Determining the Etiology
Step 1: Measure Morning Plasma ACTH
Once Cushing's syndrome is confirmed, measure morning plasma ACTH level to differentiate ACTH-dependent from ACTH-independent causes: 2, 3, 4
- Normal or elevated ACTH (>5 ng/L or >1.1 pmol/L) indicates ACTH-dependent Cushing's syndrome (pituitary or ectopic source). 3, 4
- Low or undetectable ACTH indicates ACTH-independent Cushing's syndrome (adrenal source). 3
Step 2A: For ACTH-Dependent Cases
Perform pituitary MRI to detect pituitary adenoma: 2, 4
If pituitary MRI is negative or equivocal, proceed with bilateral inferior petrosal sinus sampling (BIPSS): 1, 2
- Diagnostic criteria: central-to-peripheral ACTH ratio ≥2:1 before stimulation and ≥3:1 after CRH stimulation. 1
- Sensitivity is 100%. 1
CRH stimulation test can support pituitary origin if ≥20% increase in cortisol from baseline occurs. 3, 4
Step 2B: For ACTH-Independent Cases
Proceed with adrenal imaging (CT or MRI) to identify adrenal adenoma or carcinoma. 3
Special Populations
Children and Adolescents
- Screen only if weight gain is inexplicable and combined with either decreased height standard deviation score or decreased height velocity. 3, 4
- Lack of height gain with concurrent weight gain is the most common presentation. 2
- Growth failure sensitively discriminates simple obesity from Cushing's syndrome in prepubertal children but is unreliable in post-pubertal children. 4
Paraneoplastic Cushing's Syndrome in Lung Cancer
- Suspect in patients with lung cancer presenting with moon facies, acne, purple striae, proximal muscle weakness, peripheral edema, hypertension, and metabolic alkalosis with hypokalemia. 1
- Skin hyperpigmentation is usually more prominent with ectopic ACTH. 1
- Weight loss occurs in about 10% of cases associated with small cell lung cancer. 1
Common Pitfalls
- Neither a normal UFC nor a normal LNSC excludes mild Cushing's syndrome—multiple samples and DST are needed for diagnosis. 5
- Do not perform biochemical screening unless pre-test probability is reasonably high, as common symptoms like hypertension, weight gain, or diabetes occur frequently in the general population and do not justify testing alone. 6
- Be aware that severe obesity can cause false-positive results on all screening tests. 3