PCOS Treatment Choice: Metformin, Combined OCP, or Spironolactone by Phenotype
A newly diagnosed patient with three genuinely distinct problems — irregular cycles, biochemical insulin resistance, and hirsutism she treats as her real complaint — and three first-line drugs, each of which answers a different one of them best.
Priya M., a 24-year-old woman, manages overnight inventory for a regional bakery distributor, a job that has had her counting pallets at five in the morning three days a week for the past two years — roughly the same stretch of time over which she stopped being able to predict her own cycle at all, some months skipping entirely, others arriving twice. She started shaving along her jawline eight months ago, past the point where plucking kept up, and it's that detail, not the calendar, that finally brought her in. She isn't trying to conceive and doesn't plan to for at least a few years, but her sister's second-trimester loss last year has made her wary of anything that might complicate fertility later without a clear reason.
Her labs satisfy all three of the Rotterdam criteria carried forward by the 2023 guideline — irregular ovulation, biochemical hyperandrogenism (free testosterone 6.2 pg/mL, roughly double the upper limit of normal), and an antral follicle count of 24 on the right ovary alone. The value that actually decides which drug fits her, though, is a fasting insulin of 22 µU/mL against a normal fasting glucose of 91 mg/dL — insulin resistance without diabetes, a genuinely different phenotype from a patient whose hyperandrogenism runs equally hot but whose insulin sensitivity tests normal. The 2023 International PCOS Guideline (Teede et al.) routes the irregular-cycle-plus-hyperandrogenism presentation to combined oral contraceptive therapy first, reserves metformin primarily for the metabolic phenotype she also happens to carry, and treats antiandrogen therapy as a slower-acting add-on for hirsutism specifically — three separate recommendations that, for most patients, point in the same rough direction. Priya sits at the rare intersection where they genuinely diverge.
Her family history adds real weight to the metabolic reading specifically: her mother was diagnosed with type 2 diabetes at 44, and Priya's own lipid panel, drawn at today's visit, already shows a fasting triglyceride level at the upper edge of normal — not abnormal enough to name as its own diagnosis yet, but consistent with the same insulin-resistant picture her fasting insulin already suggests, and not a coincidence given her mother's history. She mentioned, almost in passing, that she'd seen a headline this year about PCOS being renamed “polyendocrine metabolic ovarian syndrome” and asked whether that changed anything about her own care. It doesn't change today's decision. What does is that the guideline's routing depends on which of her three findings is treated as primary, and the fasting insulin is the only one of them that her mother's diabetes and her own triglycerides independently corroborate.
Clinic visit, new diagnosis
Start with metformin. Her fasting insulin is doing the actual explanatory work here — 22 against a normal glucose is a real, documented insulin-resistant phenotype, and that resistance is plausibly driving both the anovulation and the androgen excess through the same hyperinsulinemia-stimulates-ovarian-androgen- production pathway. It's also the only one of the three options that asks nothing of her regarding contraception or VTE risk, which matters given she isn't sure when she'll want to revisit that conversation.
If her fasting insulin had come back normal, I wouldn't be arguing this at all — metformin's benefit in PCOS concentrates specifically in the insulin-resistant subgroup, not the syndrome broadly.
The metabolic argument is real, but it isn't what she came in for, and it isn't what the guideline actually recommends first for her presenting phenotype. The 2023 International PCOS Guideline names combined OCP as first-line specifically for the irregular-cycle-plus-hyperandrogenism picture — which is exactly her chief complaint — and metformin's own effect on hirsutism and cycle regularity is real but slower and more modest by direct comparison in the trial literature the guideline is built on.
I'm not disputing that her insulin is elevated. I'm disputing that treating the lab value should come before treating the two things she's actually experiencing — and OCP does that while also giving her reliable contraception she doesn't currently have, for a decision she's said she isn't ready to revisit.
Both of you are treating problems she didn't lead with. She led with the razor. Neither metformin nor OCP touches hirsutism quickly — OCP's androgen-lowering effect on hair growth takes months to become visible, and metformin's is smaller still. Spironolactone is the direct antiandrogen answer to the actual complaint that got her into this room. The complication is that it's a real teratogen, so it can't be started responsibly without the same contraceptive backbone the primary care physician is already arguing for — which argues for OCP now, spironolactone layered on once that's established, rather than either alone.
Agreed: combined OCP started today, with spironolactone explicitly named as the planned addition at the 6-month mark if hirsutism hasn't meaningfully responded. The insulin resistance was documented in the chart and flagged for revisiting, not dismissed.
Not agreed: whether metformin should have started today as well, in parallel, rather than waiting. The reproductive endocrinologist who opened the debate held that a documented insulin-resistant phenotype is reason enough to treat it now, independent of which symptom gets addressed first; the primary care physician's view was that starting three mechanisms across two visits, rather than one now and more only if needed, keeps the plan legible to a patient who is still absorbing a new diagnosis. Neither position was overruled — metformin stays on the table for the follow-up visit rather than closed off.