Bisphosphonate, Denosumab, or Neither: Bone Protection in a Premenopausal Woman on Chronic Steroids
Her DEXA shows real bone loss after three years on low-dose prednisone, and she's not done wanting children. Every option on the table protects her bones at some cost to a plan she hasn't given up on.
K.N., a 30-year-old woman, a nurse who has told her rheumatologist at every visit for the past year that she and her husband hope to start a family “within the next couple of years, once things settle,” has been on prednisone 5mg daily for three years for SLE arthritis inadequately controlled on hydroxychloroquine alone, and today's DEXA scan shows a T-score of -1.6 at the lumbar spine, meeting criteria for osteopenia with a FRAX-calculated ten-year major fracture risk that crosses the threshold most guidelines use to recommend starting bone-protective therapy. She has no prior fracture, no family history of osteoporosis, and normal renal function; her only other medication is a prenatal vitamin she started “in case,” which she mentions somewhat sheepishly. She works night shifts on a busy medical floor, twelve hours at a time on her feet, so the fall a bone-protective strategy exists to survive is an ordinary feature of her working week rather than a hypothetical.
Three years at 5mg daily is a real, sustained cumulative glucocorticoid exposure — not a brief burst that will simply reverse once therapy stops, since bone loss from chronic low-dose steroids is a genuine, biologically real accrual, not just a risk-calculator abstraction. Bisphosphonates are the guideline-first-line option, but they bind to bone mineral and are retained in the skeleton for years after the last dose, which matters precisely because she is not done planning a pregnancy. The alternative is not free of the same problem in reverse. Denosumab's rebound risk is real but its evidence is narrow: Cummings and colleagues' post-hoc analysis of FREEDOM and its extension, and Cosman's later multiple-fracture analysis, both describe postmenopausal women with established osteoporosis, in whom a prior vertebral fracture was the single strongest predictor of fracturing after stopping. K.N. is premenopausal, osteopenic rather than osteoporotic, and has never fractured anything, so she sits outside that population on all three counts — the rebound literature describes a risk she has the weakest claim on of anyone it was measured in. What it does describe accurately is the timing: bone turnover overshoots baseline within about nine months of the last dose and stays elevated for up to thirty.
Bone health review, three years into chronic steroid exposure
Her DEXA and FRAX score cross the treatment threshold today, after three real years of steroid exposure — this isn't a theoretical future risk, it's a present, measured one. I'd start a bisphosphonate now rather than defer treatment of a documented finding for a pregnancy that's still a couple of years off.
I'd slow down on the bisphosphonate specifically. These drugs bind to bone mineral and stay there for years after the last dose — that's not disputed pharmacology. No confirmed teratogenic outcome exists in the literature, but she's telling us at every visit that she wants to conceive within roughly this window, and starting a drug with genuine multi-year skeletal retention right before that isn't a decision to make lightly.
I want to be precise: I'm not claiming harm has been demonstrated. I'm saying the retention itself is real and relevant to someone in her specific position, which is different from a patient with no near-term pregnancy plans.
Denosumab solves the retention problem you're both circling — it doesn't bind bone mineral, and its effect resolves after stopping. But it comes with its own cost: the rebound vertebral fracture risk described in the FREEDOM post-hoc analyses if it's stopped without a bisphosphonate to catch her afterward — exactly the scenario she'd be in once she's ready to conceive.
That's not a reason to avoid it, it's a reason to plan the exit before we plan the start — and I want to say plainly that the exit we keep describing doesn't survive contact with her actual plan. The rebound window runs to about thirty months, so a bisphosphonate bridge given at the moment she stops denosumab for pregnancy is a bisphosphonate given at conception, which is the retention problem we chose denosumab to avoid, arriving on a worse schedule. The version that works is a bridge given well before she starts trying, not at the point she stops, which means her conception date has to be planned around the drug rather than the drug around her conception date.
Agreed: start vitamin D and calcium immediately, and start denosumab rather than a bisphosphonate, specifically to avoid multi-year skeletal retention during her planned conception window — with an explicit, documented plan that the bisphosphonate bridge is given well in advance of her stopping to conceive rather than at the moment she stops — since the rebound window runs to roughly thirty months, a bridge timed to her last denosumab dose would put a retained bisphosphonate into the pregnancy it was meant to keep clear.
Not agreed: whether denosumab's rebound-fracture exit risk is genuinely better-managed than a bisphosphonate's retention risk would have been, or whether the group has simply traded one real problem for a different one that will need to be actively managed years from now rather than today. The rheumatologist's original preference for starting a bisphosphonate immediately wasn't disproven, only set aside for this specific patient's timeline.