New Bruising at 79: Choosing Between Two Imperfect Ways to Silence an Inhibitor
An otherwise-independent 79-year-old with a new, spontaneous factor VIII inhibitor needs immunosuppression to survive it — and the regimen with the better remission odds in the registry data is also the one most likely to cost him that independence.
H.G., a 79-year-old retired librarian, still lives alone, drives herself to a weekly book club, and came to the emergency department only because her daughter noticed a large, unexplained bruise across her thigh during a video call and insisted she be seen. She has no personal or family bleeding history, no prior surgeries with bleeding complications, and takes no anticoagulants — the kind of presentation that would ordinarily prompt a search for a hidden trauma before anything else, except her labs showed a markedly prolonged activated partial thromboplastin time that a mixing study failed to correct, and a factor VIII activity under 2% with an inhibitor titer of 12 Bethesda units. Acquired hemophilia A, an autoantibody against her own factor VIII with no identifiable trigger, is rare enough that most clinicians see it once or never, and most common in exactly her demographic — elderly patients with no prior bleeding history at all.
The bleeding itself is being managed with a bypassing agent, but the actual disease requires immunosuppression to eliminate the autoantibody, and the two standard regimens carry meaningfully different tradeoffs in a 79-year-old who was, until this week, entirely independent. The registry data most often cited here, the European Acquired Haemophilia cohort, found stable complete remission in 70% on steroids plus cyclophosphamide against 59% on rituximab-based regimens — but that gap has since been tested directly. CREHA, the randomized trial that followed, found no significant difference overall, 67% against 62%, with cyclophosphamide's advantage concentrated in a poor-prognosis group defined by factor VIII under 1 IU/dL together with an inhibitor titer above 20 Bethesda units. Cyclophosphamide remains cytotoxic, carries real infection and secondary malignancy risk, and in a frail older patient can be the thing that ends independent living rather than the bleeding disorder.
Her titer of 12 Bethesda units sits below the 20-unit threshold that defined CREHA's poor-prognosis group, which is the one place cyclophosphamide's advantage was statistically real rather than numerical — so the strongest published argument for the more toxic regimen is an argument about patients she does not resemble.
Deciding immunosuppression, with the bypassing agent already running
She's actively bleeding until this autoantibody is suppressed, and the EACH2 registry found stable remission in 70% on steroids plus cyclophosphamide against 59% on rituximab-based regimens. For a disease where remission is the actual endpoint that stops the bleeding risk, I'd start there.
I'd want to name the other side of that registry number directly: cyclophosphamide's infection and secondary malignancy risk concentrates specifically in frail older patients, and for a woman who was independently driving herself to book club last week, a cytotoxic complication could end that independence faster than the bleeding disorder itself would. I'd hold that argument loosely, though — EACH2's own adverse-event rates were 41% for the cyclophosphamide arm against 37% for rituximab, which is not the gulf my own instinct assumes. The concern is the character of the harm in her age group, not its raw frequency.
A registry remission-rate difference is real, but it isn't the only outcome that matters for a 79-year-old — treatment-attributable harm belongs in the same accounting.
I'd push on the premise a little, though — she has no comorbidities, no cognitive impairment, and was fully independent until this week. That's not the frail, dependent 79-year-old the age-risk argument usually pictures. And CREHA is the more relevant comparison now anyway — randomized, no significant overall difference, with the cyclophosphamide advantage sitting in a titer band above hers. I'd want a formal frailty assessment today before deciding which trial population she actually resembles, rather than letting her age alone answer a question her actual physiologic reserve might answer differently.
If that assessment comes back reassuring, I'd still lean cyclophosphamide for the better remission odds. If it doesn't, rituximab becomes the clearer choice, and that's a real answer either way, not a stalemate.
Agreed: a formal geriatric frailty assessment ordered today before committing to either regimen, with prednisone started immediately regardless, since steroids alone are appropriate first-line therapy while the assessment result and second agent are finalized.
Cyclophosphamide is added, favoring the regimen with the better remission odds in the registry data.
Rituximab is added instead, prioritizing her independence over the marginal remission-rate difference.