What is the diagnostic approach for a female patient of reproductive age presenting with symptoms suggestive of Polycystic Ovary Syndrome (PCOS), such as irregular menstrual cycles, weight gain, acne, or excess hair growth?

Diagnostic Testing for Polycystic Ovary Syndrome (PCOS)

For adult women with suspected PCOS, measure total testosterone or free testosterone using liquid chromatography-tandem mass spectrometry (LC-MS/MS), assess menstrual history for oligo/anovulation, and perform transvaginal ultrasound to count follicles per ovary (≥20 follicles) or measure ovarian volume (≥10 mL), as diagnosis requires any two of these three Rotterdam criteria: hyperandrogenism, irregular cycles, or polycystic ovarian morphology. 1

Core Diagnostic Algorithm

Step 1: Clinical Assessment and First-Line Laboratory Tests

Androgen Testing (Primary Diagnostic Marker)

• Measure total testosterone (TT) and free testosterone (FT) using LC-MS/MS as the mandatory first-line test, with TT showing 74% sensitivity and 86% specificity, and FT demonstrating superior 89% sensitivity with 83% specificity 1
• LC-MS/MS is essential because direct immunoassays have significantly lower specificity (78% vs 92%), leading to false positives 1
• If TT or FT are normal but clinical suspicion remains high, measure androstenedione (A4) with 75% sensitivity and 71% specificity, and DHEAS with 75% sensitivity and 67% specificity as second-line tests 1

Menstrual History

• Document menstrual cycle patterns, as oligo/anovulation (cycles >35 days or <8 cycles per year) fulfills one Rotterdam criterion 2
• Measure progesterone in mid-luteal phase to confirm anovulation, with levels <6 nmol/L indicating anovulation 1

Exclusion of Other Conditions

• Measure thyroid-stimulating hormone (TSH) to rule out thyroid disease causing menstrual irregularity 1, 3
• Measure prolactin using morning resting serum levels to exclude hyperprolactinemia, with levels >20 μg/L considered abnormal 1, 3
• Confirm prolactin elevation with 2-3 samples at 20-60 minute intervals via indwelling cannula to exclude stress-related spurious elevation 1

Step 2: Ultrasound Assessment (When Needed)

Critical Decision Point: If the patient has both irregular menstrual cycles AND clinical/biochemical hyperandrogenism, ultrasound is not necessary for PCOS diagnosis, though it will identify the complete PCOS phenotype 1

Ultrasound Criteria (When Performed)

• Use transvaginal ultrasound with ≥8 MHz transducer frequency for optimal resolution in adults 1
• Follicle number per ovary (FNPO) ≥20 follicles is the gold standard ultrasonographic marker with 84% sensitivity and 91% specificity (AUC: 0.905) 4
• Ovarian volume ≥10 mL serves as an alternative criterion with 81% sensitivity and 81% specificity 4, 1
• Follicle number per single cross-section (FNPS) is a secondary alternative with 81% sensitivity and 83% specificity 4

Ultrasound Contraindications

• Do not use ultrasound for PCOS diagnosis in those with gynecological age <8 years (<8 years after menarche) due to high incidence of multifollicular ovaries in this life stage 1
• In adolescents, diagnosis requires both hyperandrogenism AND irregular cycles; ovarian morphology should not be included due to poor specificity 2

Step 3: Metabolic Screening (Mandatory for All PCOS Patients)

Glucose Metabolism

• Perform two-hour oral glucose tolerance test with 75g glucose load to screen for glucose intolerance and type 2 diabetes 1
• Measure fasting glucose and insulin levels, calculating glucose/insulin ratio with >4 suggesting reduced insulin sensitivity 1, 3

Lipid Assessment

• Order fasting lipid panel including total cholesterol, LDL, HDL, and triglycerides, with target LDL <100 mg/dL, HDL >35 mg/dL, and triglycerides <150 mg/dL 1

Anthropometric Measurements

• Calculate body mass index (BMI), with BMI >25 considered obese 1
• Measure waist-hip ratio (WHR) to identify central obesity, with WHR >0.9 indicating truncal obesity 1

Step 4: Additional Hormonal Assessment (Optional)

• Measure LH and FSH between days 3-6 of menstrual cycle, with LH/FSH ratio >2 suggesting PCOS, though this is abnormal in only 35-44% of women with PCOS, making it a poor standalone diagnostic marker 1, 3
• Measure DHEAS if adrenal hyperandrogenism is suspected, especially with values >3800 ng/ml in women aged 20-29, or if levels >10.0 nmol/L to rule out adrenal/ovarian tumors 1, 3

Anti-Müllerian Hormone (AMH) Considerations

• AMH ≥35 pmol/L (5 ng/mL) shows 92% sensitivity and 97% specificity for PCOS 1
• Do not use AMH as an alternative for detecting polycystic ovarian morphology or as a single test for PCOS diagnosis due to lack of standardization, no validated cut-offs, and significant overlap between women with and without PCOS 1
• AMH is a potential future surrogate marker but requires additional validation in different populations 1

Critical Diagnostic Pitfalls

Normal Laboratory Values Do Not Exclude PCOS

• Total testosterone is abnormal in only 70% of women with confirmed PCOS, meaning 30% have normal testosterone levels despite having the condition 1
• A woman can be diagnosed with PCOS based solely on clinical hyperandrogenism (hirsutism, acne, or alopecia) plus irregular menstrual cycles, without any abnormal laboratory values, as the Rotterdam criteria requires only two of three features 1
• Normal ovarian ultrasound does not exclude all reproductive disorders, as women can have ovulatory dysfunction or hyperandrogenism with normal ovarian morphology 3

Screening for Serious Conditions

• Screen for Cushing's syndrome if patient has buffalo hump, moon facies, hypertension, abdominal striae, central fat distribution, easy bruising, or proximal myopathies 1
• Consider androgen-secreting tumors if rapid onset of symptoms, severe hirsutism, or very high testosterone levels 1
• Screen for congenital adrenal hyperplasia if elevated DHEAS levels 1

Diagnostic Accuracy by Criteria

• Rotterdam criteria show higher sensitivity (77-89%) but lower specificity (83%) compared to NIH criteria (sensitivity 51%, specificity 94%) 1
• The Rotterdam criteria generates four phenotypes with heterogeneous clinical features 2