An Unplanned Pregnancy on Mycophenolate: Switching Fast Without Losing Control
She found out she's pregnant on a drug with a well-documented birth-defect pattern, and there's no version of today's plan that doesn't involve some gap in her disease control while the switch happens.
L.V., a 26-year-old woman who manages a restaurant, called the clinic this morning after a home pregnancy test came back positive, unexpectedly — she and her partner had not been trying to conceive, and she is, by her own estimate and a same-day beta-hCG-based dating, roughly six weeks along. She has SLE with prior lupus nephritis in remission for two years, currently maintained on mycophenolate mofetil 2000mg daily, hydroxychloroquine, and prednisone 5mg daily, and has had no flare in fourteen months. She is anti-Ro/SSA antibody positive on prior testing, a detail from her chart that becomes clinically relevant today in a way it hasn't been before. She came in directly from a lunch shift, still in her work apron, having called as soon as she saw the test rather than waiting for a scheduled visit. She has never been tested for thiopurine methyltransferase activity, and the drug she is about to be switched onto is the one that test exists for.
Mycophenolate's teratogenic risk is not a theoretical caveat — it is a specifically described syndrome, with a recognizable pattern of ear and craniofacial malformation, cleft lip and palate, and congenital heart defects documented across case series and pregnancy registries, and current guidance is unambiguous that it should be stopped before conception is attempted, not managed through pregnancy. She is already six weeks in, past the point of a planned preconception switch, which means today's visit is about damage control and disease-control continuity happening at the same time, not a leisurely medication-optimization conversation. Her anti-Ro/SSA positivity carries its own stake in this pregnancy, and the numbers attached to it are not the ones usually quoted. Izmirly's PATCH trial, which found hydroxychloroquine roughly halved recurrent congenital heart block, enrolled only women who had already had a pregnancy complicated by it, against a historical recurrence rate near eighteen percent. This is L.V.'s first such pregnancy, so her baseline is the far lower first-occurrence risk of about two percent, and PATCH is therefore an argument for continuing hydroxychloroquine rather than a measure of what she personally stands to gain — a distinction worth keeping straight when the drug is being continued anyway on three other grounds.
Urgent same-day visit, six weeks pregnant
Mycophenolate has to stop today. This isn't a theoretical caveat — there's a specifically described pattern of ear, craniofacial, and cardiac malformation associated with continued exposure, documented across case series and pregnancy registries. However far along she already is, continuing it another day is the one thing on today's list that isn't negotiable.
It has to stop, and I'm not disputing that. Where I part company is on what replaces it. Azathioprine is the reflex answer and it takes several weeks to reach effect — which means you are proposing a gap of exactly that length in a woman whose nephritis is the thing that has hurt her most, at the point in gestation when a flare does the most damage. The ACR reproductive health guideline lists calcineurin inhibitors alongside azathioprine as pregnancy-compatible, and tacrolimus works in days rather than weeks. If the problem you are actually solving is the gap, azathioprine is the drug that creates it.
I'd want a specific plan for that gap — close monitoring of urine protein and renal function, and a low threshold to adjust prednisone temporarily — not just a stop order and a new prescription.
And send thiopurine methyltransferase and NUDT15 with today's labs before the first azathioprine dose. A poor metabolizer started at a standard dose can be profoundly myelosuppressed inside two weeks, and doing that to a pregnant woman while we are already braced for a nephritis flare is not a risk worth accepting to save three days.
Tacrolimus isn't free either — it needs level monitoring, and her blood pressure and renal function both become variables in a pregnancy where preeclampsia is already on the list. I'd still start azathioprine and cover the gap with prednisone rather than take on a second thing to titrate this afternoon, but I'll grant that's a preference about which uncertainty I'd rather manage, not a claim that yours is wrong. What can't wait for either of us to settle it is this: she's anti-Ro/SSA positive, which carries its own independent risk of neonatal lupus and, in a small but serious minority, fetal congenital heart block.
Hydroxychloroquine has real evidence behind it there — PATCH halved recurrence in women who had already lost a pregnancy to heart block, which isn't her situation and where her own absolute risk starts far lower — and it needs to keep going regardless of which immunosuppressant she's on. I'd also get maternal-fetal medicine involved today for serial fetal echocardiogram monitoring starting around 16 weeks, on its own timeline, independent of the medication switch.
Agreed: stop mycophenolate today, send TPMT and NUDT15 with today's labs before the first azathioprine dose, start azathioprine with close renal-function monitoring during the transition gap, continue hydroxychloroquine and prednisone unchanged, start low-dose aspirin for preeclampsia prophylaxis, and refer to maternal-fetal medicine today for serial fetal echocardiogram monitoring given her anti-Ro/SSA status.
Not agreed, and left open rather than resolved: whether azathioprine was the right replacement at all, or whether tacrolimus's onset in days rather than weeks should have outweighed the extra monitoring burden in a woman whose worst organ is the one exposed during the gap. The maternal-fetal medicine specialist's position was not withdrawn, only outweighed by a preference for managing one new variable instead of two on the same afternoon — and if her urine protein moves during the transition, that is the decision that gets revisited first. Both agreed weekly monitoring for the first month was the right starting cadence regardless.