CVID and Refractory ITP: Immunosuppressing an Immunodeficient Patient
A hairdresser with common variable immunodeficiency develops steroid-refractory immune thrombocytopenia. Treating the platelet count means adding an immunosuppressant to a patient whose entire diagnosis is that her immune system already can't protect her — the paradox the case turns on.
Renata S., 37, has cut hair at the same strip-mall salon for nineteen years, long enough that she now does the highlights for clients whose mothers she also did highlights for when they were young. She was diagnosed with common variable immunodeficiency at thirty-one after her third pneumonia in two years, and has been on monthly immunoglobulin replacement for the six years since — stable, working full shifts on her feet all day, infection-free on that regimen until bruising on her shins three months ago turned out to be a platelet count of 8,000. She has no family history of bleeding disorders and no other autoimmune diagnoses; this is her first clear autoimmune complication since the original CVID diagnosis itself, and she has never required a blood transfusion despite the low count, a detail her hematologist notes as meaningfully reassuring given how far below normal her platelets have run.
Four weeks of high-dose prednisone brought her platelets to 24,000 and no higher, with the cushingoid weight gain and mood swings that made her ask to stop it before her hematologist was ready to call it a failure outright. The paradox her team keeps returning to is structural, not incidental: CVID is, by definition, a failure of humoral immunity, and the two most effective next steps for refractory immune thrombocytopenia — splenectomy and rituximab — both act on the same immune system that's already failing her. Gobert et al., 2011, British Journal of Haematology, followed thirty-three patients with exactly her combination — CVID plus immune thrombocytopenia or hemolytic anemia — treated with rituximab: an 85% initial response rate, 74% complete, though ten of the initial responders relapsed over a mean follow-up of thirty-nine months, and retreatment with a second rituximab course was successful in seven of the nine who tried it. Severe infections occurred afterward in eight of the thirty-three — just under a quarter of the cohort — and half of those eight were not on immunoglobulin replacement at the time. She already is, which is exactly the variable her team keeps returning to as the one genuinely in their control.
Treating the platelets without unmaking the antibody coverage
Steroids have failed at a dose she can't tolerate longer anyway. Gobert and colleagues followed thirty-three patients with CVID-associated cytopenias treated with rituximab and found an 85% initial response, 74% complete — real numbers in her exact population, not an extrapolation from primary ITP. Splenectomy is the other option at this point in most refractory ITP, but in CVID it carries a documented worse long-term infection burden than rituximab does. I want to move to rituximab.
You're right that rituximab is the better-supported next step here, and I'm not arguing against it. What I want confirmed first is that her immunoglobulin coverage is actually adequate, not just present. That same Gobert cohort had severe infections in eight of thirty-three patients after rituximab — and four of those eight, half of them, were not on immunoglobulin replacement at all.
The study never reports how many of the thirty-three were on replacement — only that half the severe infections landed in patients who weren’t. That is the number I want settled on our side of the ledger before we infuse, because it is the one variable in this whole discussion we can actually set.
One thing worth naming before we start the infusion: her trough of 780 covers antibody protection, but rituximab depletes the B-cell compartment that also cooperates with T cells and innate immunity in ways immunoglobulin doesn't fully replace. That's not a reason to withhold it — the data supports using it here — but it's a reason to increase her infusion frequency and monitoring for the months immediately after, rather than treat adequate trough as the whole safety picture.
Rituximab was given on the standard four-weekly protocol with immunoglobulin infusions moved to every three weeks through that period. Platelets climbed above 100,000 by week six and have held there since, off prednisone entirely.
Agreed by all three: infection surveillance stays heightened for at least six months post-rituximab, with a lower threshold for repeat Ig-level checks and for treating any fever empirically while B-cell recovery is confirmed — the one point where structural agreement outlasted the disagreement about how much weight the theoretical concern deserved.