Antiphospholipid Syndrome: Anticoagulant Choice After an Arterial Stroke
A 34-year-old veterinary technician, fourteen months past an ischemic stroke and confirmed triple-positive for antiphospholipid antibodies, wants to trade her warfarin and regular blood draws for a once-daily anticoagulant. The evidence on that trade is specific to her exact antibody profile — and unusually clear. What isn't yet resolved is how to make her ongoing monitoring less of a burden.
K.S. is a 34-year-old woman working as a veterinary technician at a large-animal practice, splitting her week between clinic rounds and hour-plus drives to farm calls across three counties — work she loves, but that leaves her, in her own words, "bad at remembering to stop for anything, including blood draws." She was diagnosed eight years ago with Hashimoto's thyroiditis, well controlled on levothyroxine, and had no other health problems before the event that brought her into this case.
Fourteen months ago, at 33, she had sudden right-sided weakness and slurred speech while loading a horse trailer — an ischemic stroke with no source found on echocardiogram, carotid imaging, or rhythm monitoring. The workup that followed found a lupus anticoagulant, an anticardiolipin antibody, and an anti-β2-glycoprotein I antibody all positive on her initial labs, and — the finding that actually confirmed the diagnosis — all three still positive on repeat testing twelve weeks later, meeting criteria for primary, triple-positive antiphospholipid syndrome. She was started on warfarin during that hospitalization, and after a two-month dose-titration period, her INR has held steady in the 2.0–3.0 target range for the twelve consecutive months since.
That stability is exactly what brought her back to clinic today, not a new problem. She's asking to switch to a once-daily anticoagulant that wouldn't require the monthly draws currently built around her clinic schedule — a request that would be entirely reasonable for most patients on long-term anticoagulation, and one this team has granted routinely for other indications. What's different about her is that her triple-positive result, combined with the fact that her qualifying event was arterial rather than venous, places her inside the exact population a major trial was built to test — and that trial found the drug she's asking about caused more strokes, not fewer, against warfarin in patients who look like her. Whether that evidence applies to her isn't really the open question; it does, about as directly as a trial result applies to an individual patient. The open question is what, short of changing the drug, can honestly be done about the burden that brought her back in the first place.
Fourteen months in, revisiting the anticoagulant
She's not asking this lightly. She's driven back into clinic every four weeks for fourteen months, and she asked me directly today whether a once-daily pill could replace that — the same switch we make routinely for atrial fibrillation, for a first unprovoked clot. I want to be honest that her case isn't that case. Her triple-positive result and the fact that her event was a stroke put her squarely inside the population the TRAPS trial enrolled to test exactly this question, and TRAPS was stopped early — eleven thromboembolic, bleeding, or vascular-death events on rivaroxaban against two on warfarin, a hazard ratio of 6.7. I'm not going to argue for the DOAC. What I do want to argue for is not leaving her with nothing — she came in with a real problem, and "no" to the drug switch can't be the whole answer.
A separate small trial of apixaban in thrombotic APS — broader entry criteria than TRAPS, not restricted to triple-positive patients — was stopped early for the same reason: six strokes on apixaban against zero on warfarin. This isn't a rivaroxaban-specific finding; it's a class finding, which is exactly what the current guidance concludes.
Agreed on the drug — not agreed on where to look for relief. Her classification isn't provisional or borderline; triple-positivity confirmed twelve weeks apart, with an arterial qualifying event, is precisely the profile the International Society on Thrombosis and Haemostasis names directly — arterial thrombosis or triple-positivity, either one — as the group DOACs should be withheld from, independent of which factor Xa inhibitor is proposed. A pooled analysis across the DOAC trials in this population found ischemic stroke in roughly 9% of DOAC-treated patients against essentially none on warfarin. That's the population she belongs to, and it's a population where the cost of a missed or falsely reassuring reading is a second stroke, not a bruise. I'm not comfortable being the one who introduces a monitoring method that hasn't been validated in anyone who actually looks like her, however well it's worked in other anticoagulated groups.
I recognize that leaves her carrying the exact burden she came in to ask about relieving, and I don't think that's a satisfying place to stop.
There's a version of this that doesn't require resolving the self-testing question today. Twelve straight months of INR values sitting inside a one-point-wide target range is real pharmacokinetic evidence of a stable maintenance dose and a consistent metabolic and dietary pattern — that's not nothing, and it's evidence about her specifically, not an extrapolation from a different population. Widening her check interval from four weeks to six doesn't introduce any new device or method into a population where it's unvalidated; it simply asks less of a system that's already shown it holds steady. If a value comes back outside range, we tighten straight back to monthly. It costs her less today and asks nothing of anyone that the self-testing question still needs settling before it's tried.
Agreed quickly and without real dissent: K.S. stays on warfarin, standard intensity, target INR 2.0–3.0, unchanged. No DOAC — not rivaroxaban, not apixaban — given her confirmed triple-positive serology and the specific weight the evidence carries for a prior arterial event.
Not agreed: how to make the monitoring itself less of a burden.
The rheumatologist's reservation stands — no case-specific validation yet exists for this population. Any future trial of a home device would need a structured comparison period against clinic draws before being adopted as her standing method.
A single out-of-range value at six weeks reverts her to monthly draws without further argument — the pharmacologist's proposal was explicitly conditional, not a new permanent baseline.
What the team settled on for now: extend her check interval from four weeks to six, the one change nobody objected to, while the self-testing question stays open and is revisited at her next visit. Nobody set a preference on that larger question; the rheumatologist set the standard instead — any monitoring change beyond the interval extension needs its own evidence in a population that looks like her before it earns her trust the way twelve months of clinic draws already have.