What is the best test to check for hypercortisolism (Cushing's syndrome)?

Best Test for Hypercortisolism Screening

No single "best" test exists—use at least 2-3 screening tests from the triad of late-night salivary cortisol (LNSC), overnight 1-mg dexamethasone suppression test (DST), or 24-hour urinary free cortisol (UFC), with test selection based on clinical suspicion level and patient-specific factors. 1

Initial Screening Strategy Based on Clinical Suspicion

Low clinical suspicion:

• Start with LNSC (≥2 tests on consecutive days) as it has the highest specificity among screening tests 1
• LNSC is particularly useful because it detects loss of normal circadian cortisol rhythm, which occurs early in hypercortisolism 1

Intermediate to high clinical suspicion:

• Perform 2-3 different screening tests simultaneously (combination of LNSC, UFC, and DST) 1
• This approach maximizes diagnostic accuracy and accounts for test-specific limitations 1

Test Performance Characteristics

Sensitivity (all >90%):

• DST and LNSC have the highest sensitivity 1
• UFC has the lowest sensitivity among the three 1

Specificity:

• LNSC is the most specific 1
• DST and UFC are the least specific 1

Patient-Specific Test Selection

Prefer DST when:

• Evaluating adrenal incidentalomas (cortisol <1.8 μg/dL excludes autonomous cortisol secretion; >5 μg/dL identifies overt hypercortisolism) 1, 2
• Patient works night shifts or has disrupted circadian rhythm 1
• Quick outpatient screening is needed 3

Avoid DST when:

• Patient takes oral estrogen-containing contraceptives (increases cortisol-binding globulin, causing false positives) 1
• Patient takes CYP3A4 inducers (phenobarbital, carbamazepine, St. John's wort) or inhibitors (fluoxetine, cimetidine, diltiazem) 1
• Consider measuring dexamethasone levels simultaneously with cortisol (>1.8 ng/mL confirms adequate absorption) to reduce false positives by 20% 1, 2, 4

Prefer LNSC when:

• Screening for cyclic Cushing's syndrome (multiple sequential tests over time capture episodic hypercortisolism) 1, 5
• Patient compliance with urine collection is poor 1
• Sample at usual bedtime rather than strict midnight to reduce false positives 1

Avoid LNSC when:

• Patient works night shifts or has irregular sleep schedules 1
• Evaluating adrenal tumors (lower specificity in this population) 1
• Mass spectrometry is unavailable and topical hydrocortisone contamination is possible 1

Prefer UFC when:

• Need to assess overall cortisol production independent of cortisol-binding globulin 1
• Monitoring treatment response 3
• Must obtain 2-3 collections to account for day-to-day variability 1, 5

Critical Diagnostic Cutoffs

DST (overnight 1-mg):

• Normal: cortisol <1.8 μg/dL (50 nmol/L) at 8 AM 1, 2
• Higher cutoffs (e.g., 5 μg/dL) reduce sensitivity but are useful for adrenal incidentalomas 1

LNSC:

• Abnormal: above laboratory-specific upper limit of normal 1
• Collect at usual bedtime (not strict midnight) 1
• Requires ≥2-3 tests 1

UFC:

• Abnormal: above laboratory-specific upper limit (assay-dependent) 1
• Requires 2-3 complete 24-hour collections 1

Common Pitfalls to Avoid

Do not rely on a single normal test result to exclude hypercortisolism—repeat 1-2 screening tests if clinical suspicion remains, as cyclic Cushing's syndrome can show intermittent normal results 1, 5

Do not use inferior petrosal sinus sampling (IPSS) to diagnose hypercortisolism—it is only for localizing the source of ACTH-dependent disease after hypercortisolism is confirmed 1

Do not perform testing during acute illness, severe depression, or uncontrolled diabetes—these pseudo-Cushing's states cause false positives; treat underlying condition and retest in 3-6 months 1

For cyclic disease, confirm active hypercortisolism immediately before any dynamic or localization testing (including IPSS) to avoid false-negative results 1, 5

Algorithm for Abnormal Screening Results

If initial screening is abnormal:

• Repeat 1-2 different screening tests 1
• If persistently abnormal, proceed to measure ACTH to determine ACTH-dependent vs. ACTH-independent disease 1
• Consider Dex-CRH test if pseudo-Cushing's state is suspected (though ovine CRH is unavailable in many countries) 1

Mass spectrometry is preferred over immunoassay when available to avoid cross-reactivity with cortisol metabolites, though current guidelines do not mandate it 1, 3