Obinutuzumab or Rituximab for Refractory Lupus Serositis, Neither Studied in This Exact Patient
Both drugs on the table were approved, or are used off-label, for a version of lupus she doesn't have. The disagreement is which mismatch is the smaller one — and whether that's even the right way to be deciding.
D.K., a 38-year-old woman who teaches high school chemistry, has had recurrent pleuritic chest pain and pericarditis attributable to active SLE four times in the past two years, each episode requiring a burst of high-dose prednisone she describes as leaving her “shaky and useless in front of a classroom for a week.” She has already failed sequential trials of both anifrolumab and belimumab, added on top of hydroxychloroquine and methotrexate, with no meaningful reduction in serositis flare frequency; her most recent echocardiogram confirmed a small pericardial effusion without tamponade physiology. She has no renal involvement on any biopsy or urinalysis in six years of follow-up, and no hematologic or neuropsychiatric disease. She has asked, at each of the last two flares, whether there is anything left to try that isn't another steroid burst. Both biologics licensed for her kind of disease have already been tried in sequence and neither held, so everything remaining on the table is being used outside the population it was studied in.
Her prednisone burden over the past two years totals roughly the equivalent of eleven months at 20mg daily when the burst courses are added up — a real, cumulative steroid exposure in a 38-year-old that the group is trying to end, not just manage flare by flare. Obinutuzumab's October 2025 approval rests on REGENCY, and on the phase II NOBILITY trial it was built to verify, both of which enrolled only biopsy-proven active proliferative lupus nephritis; she has never had nephritis, and no biopsy has ever been indicated to look. Rituximab's own pivotal non-renal trial, EXPLORER, missed its primary endpoint entirely, a negative result now widely attributed to unusually effective background glucocorticoid response in both arms rather than genuine lack of drug effect — but a negative trial is still a negative trial, whatever the explanation offered for it afterward. Neither drug, in other words, has positive randomized evidence in the specific population sitting in this room.
Refractory case conference, after two biologics have already failed
We're extrapolating no matter which of these we pick, so let's be honest about that and reason from mechanism. Obinutuzumab, as a type II anti-CD20 antibody, depletes B cells more completely and more durably than rituximab's type I mechanism — REGENCY held CD19 depletion in about ninety-five percent of patients out to seventy-six weeks — that's not a marketing claim, it's the actual pharmacology, and lupus serositis this refractory to two other biologics is about as B-cell-driven a presentation as we're going to see.
Mechanism is a real argument, but it's the only thing obinutuzumab has here — REGENCY was a renal trial, and she's never had nephritis. Rituximab has a decade of real-world and registry use in exactly her situation, refractory non-renal serositis, and EULAR conditionally recommends it for that use despite EXPLORER's negative result, which the field has largely traced to unusually effective glucocorticoid response in the placebo arm rather than a true lack of drug effect.
I want to be precise about what I'm conceding: EXPLORER was negative. I'm not pretending otherwise. I'm arguing that a decade of convergent real-world experience since then is worth more than a plausible mechanism with zero matching outcome data of its own.
You're both reasoning from evidence that doesn't actually cover her, and I don't think either argument wins cleanly. Obinutuzumab is dosed as four initial infusions and then twice yearly; rituximab needs re-dosing roughly every six months on a less predictable schedule. For someone managing a full class load around infusion visits, that's a real, practical difference — not a clinical one, but real.
Given the evidence genuinely doesn't separate them, I'd let that scheduling advantage decide it: obinutuzumab, with an explicit plan to move to rituximab if she doesn't respond within two infusion cycles.
Agreed: start obinutuzumab, given the tiebreak on dosing schedule, with an explicit plan to switch to rituximab if serositis flares recur within two infusion cycles. Hydroxychloroquine and methotrexate continue unchanged.
Not agreed: whether a scheduling tiebreaker was the right way to resolve a decision neither drug's own trial evidence actually settles — the pharmacologist would have given rituximab's decade of real-world experience more weight than a dosing-convenience argument. The two-cycle response check was set as the explicit point where that disagreement gets revisited against real data from her own case.