Caplacizumab Three Days After a Molar Extraction: Weighing the Bleeding It Adds
A woman presenting with acute immune TTP has a healing extraction socket that makes the drug most likely to save her platelets also the drug most likely to make her mouth bleed.
L.F., a 38-year-old woman, had a molar extracted three days ago for an abscessed tooth and was recovering unremarkably until this morning, when she woke up confused, was brought to the emergency department by her sister, and was found to have a platelet count of 11,000/µL, a hemoglobin of 7.8 g/dL with schistocytes on the smear, and an LDH markedly elevated. The combination — microangiopathic hemolysis, severe thrombocytopenia, and new neurologic change, with a creatinine that is only mildly elevated rather than the more pronounced renal injury typical of hemolytic uremic syndrome — fits the clinical picture of thrombotic thrombocytopenic purpura closely enough that plasma exchange was started empirically within hours, before an ADAMTS13 activity level could return to confirm the diagnosis with certainty.
Her extraction socket, three days out, is still actively healing — not frankly bleeding, but not the fully closed tissue a week further out would show either. That detail matters directly to the next decision: caplacizumab, added to plasma exchange and immunosuppression, shortens time to platelet normalization and reduces TTP exacerbations and death in the trial that established it, at a real cost — roughly two-thirds of caplacizumab-treated patients had some bleeding event, overwhelmingly mucocutaneous, against under half on placebo. A healing oral wound is exactly the kind of site where that mucocutaneous bleeding tendency would show up first, and the question in front of the team is not whether caplacizumab works — it does — but whether starting it today, on top of a platelet count already at 11,000 and a mouth that isn't finished healing, trades a faster platelet recovery for a bleeding complication nobody would have chosen to invite.
Her presentation also clears the ADAMTS13-independent clinical criteria clinicians use to justify treating before the confirmatory assay returns: the PLASMIC score, built from exactly the variables in front of the team tonight — platelet count, hemolysis markers, absence of another obvious cause, and a creatinine that stays low relative to the degree of hemolysis — places her at high probability for severe ADAMTS13 deficiency, which is precisely why plasma exchange started within hours rather than waiting on a result that takes days.
In the ICU, before the first plasma exchange finishes
I want caplacizumab on board today. HERCULES dosed it before or right at the end of the first exchange, specifically before ADAMTS13 confirmation, because the benefit — faster platelet recovery, fewer exacerbations, lower mortality — is front-loaded and time-sensitive. Her picture already meets the clinical criteria strongly; waiting for a lab result that takes days defeats the point of an early-acting drug.
I don't disagree with starting the drug, but I want the extraction site named specifically, not folded into a general bleeding-risk conversation. It's day three — granulation tissue is present but the socket isn't closed. In the trial, mucocutaneous bleeding was the dominant complication, and this is exactly the kind of site that shows up first when platelet function and coagulation are both compromised.
I'd push back on treating this as a reason to delay, though — it's a reason to watch that specific site closely, pack it proactively if needed, not a reason to withhold a drug whose whole value is acting before the disease does more damage.
The asymmetry here is the thing worth stating plainly: untreated TTP has a mortality rate that used to run near 90% before plasma exchange existed, and remains real even with it. A mucocutaneous bleed from an extraction site, even a bothersome one, is not in the same category of danger. Start caplacizumab today alongside the exchange and immunosuppression, have the surgical team pack and monitor the socket proactively rather than reactively, and treat any bleeding there as an expected, manageable side effect of doing the right thing quickly — not as a reason we should have waited.
Agreed within the hour: caplacizumab started alongside the ongoing plasma exchange and prednisone, with the oral surgery team packing the extraction site proactively and checking it twice daily rather than waiting for visible bleeding to prompt a check.
It is managed locally as an expected, tracked side effect — not a reason to interrupt caplacizumab, per the team's explicit agreement today.
That would be treated as a signal to reassess the whole picture, including whether a second diagnosis is being missed — not something either position anticipated tonight.