Rapidly Progressive GPA: Rituximab, Cyclophosphamide, or Both
A single patient, three weeks into a rapidly progressive renal decline from newly diagnosed GPA. No trial cleanly favors one induction agent over the other for a case this severe — that absence of a clean answer is the actual disagreement.
Walter H. spent thirty-four years delivering mail on the same suburban route before retiring two years ago, and had never had a serious illness until a persistent sinus congestion and occasional nosebleeds started about two months ago — the kind of thing he assumed was allergies until his creatinine, checked incidentally at a routine physical, came back at 1.1, unremarkable, and then 2.8 three weeks later when his primary care doctor rechecked it after he mentioned feeling unusually tired. PR3-ANCA came back strongly positive; a chest CT ordered the same day showed bilateral non-cavitating pulmonary nodules he had no symptoms from at all. The kidney biopsy, done within the week, confirmed pauci-immune crescentic glomerulonephritis with 45% of glomeruli showing crescents — active, ongoing injury, not old scarring.
A creatinine that nearly tripled in three weeks with active urine sediment and biopsy-proven crescents is rapidly progressive glomerulonephritis by any definition, and severe renal involvement is where the induction-agent literature is least settled. RAVE found rituximab noninferior to cyclophosphamide across GPA and microscopic polyangiitis broadly — genuinely reassuring, but its own severe-renal subgroup was underpowered to detect a difference either way. RITUXVAS tested something more targeted for exactly his situation: rituximab combined with a reduced course of cyclophosphamide, specifically in severe renal AAV — and found outcomes essentially comparable to cyclophosphamide alone, not clearly superior. That's an honest, uncomfortable finding for anyone hoping the combination would resolve the question outright; it doesn't prove combination therapy adds nothing, but it doesn't prove it adds enough to justify itself either, in a trial built specifically to answer that. Layered onto that open question is a second, related one the team will also have to settle regardless of which induction agent wins the argument: PEXIVAS found reduced-dose glucocorticoids noninferior to the traditional high-dose taper, with meaningfully fewer serious infections along the way — a separate finding that shapes his regimen no matter which induction agent is chosen, and one easy to lose sight of while the induction-agent debate takes the room's full attention.
On the renal consult service, three weeks into decline
RAVE settled the general question — rituximab is noninferior to cyclophosphamide for new PR3-ANCA disease, and he's sixty-seven with decades of bladder and marrow surveillance ahead of him if we choose cyclophosphamide instead. I'd start rituximab and reserve cyclophosphamide for genuine non-response.
RAVE's overall result is real, but its severe-renal subgroup — patients closer to where he actually is, 45% crescents and a creatinine that nearly tripled in three weeks — wasn't powered to confirm equivalence. I'm not disputing rituximab's general noninferiority; I'm saying we don't actually know it holds at this severity, and cyclophosphamide's track record specifically in RPGN is longer and deeper.
Extending RAVE's overall noninferiority finding to his subgroup treats an underpowered analysis as though it were a confirmed negative result — it isn't; it's genuinely unanswered.
I want to be honest about what RITUXVAS actually found, because it would be easy to oversell this: rituximab plus reduced-dose cyclophosphamide, tested specifically in severe renal AAV, came out comparable to cyclophosphamide alone — not superior. That's not a ringing endorsement. But it does mean combining the two doesn't demonstrate added harm in exactly his situation, and a reduced cyclophosphamide course alongside rituximab may lower his lifetime alkylator exposure relative to a full course.
Given neither monotherapy option has a clean, decisive trial behind it at his severity, I'd rather use the regimen that's been directly tested here, with its real limitations disclosed, than extrapolate confidently from either RAVE's overall result or cyclophosphamide's general track record.
Agreed: combination rituximab plus reduced-dose cyclophosphamide per the RITUXVAS protocol, with reduced-dose glucocorticoids per PEXIVAS. Creatinine and urine sediment rechecked weekly for the first month.
The nephrologist's severe-renal-subgroup concern was accepted as the deciding factor over rituximab monotherapy — not because cyclophosphamide alone was chosen, but because it argued against relying on RAVE's underpowered subgroup for a decision this consequential.