Six Weeks Off Etanercept, and the Serology Still Hasn’t Gotten the Message
Most drug-induced lupus quietly resolves once the drug is gone. Hers hasn’t — and the specific way it hasn’t is exactly what makes anti-TNF-induced lupus different from the classic kind.
Teresa L., a 45-year-old high school librarian, has managed ankylosing spondylitis with etanercept for just over three years, well enough that she had stopped needing the cane she carried in her twenties. Ten weeks ago she developed a malar rash that worsened distinctly with sun exposure, along with new symmetric joint pain in her hands and a fatigue heavy enough that she began leaving work early most days. Serologic testing found a newly positive ANA and, more specifically, a newly positive anti-dsDNA antibody with depressed C3 and C4 complement levels — a pattern that immediately marked this as something other than the classic drug-induced lupus long associated with older agents like hydralazine and procainamide, which typically spares complement and dsDNA reactivity almost entirely. Ramos-Casals and colleagues' large systematic review of TNF-targeted-therapy-induced autoimmune disease specifically described this atypical, complement-consuming, anti-dsDNA-positive pattern as a real and distinct entity within anti-TNF-induced lupus, occasionally including renal involvement not seen in the classic drug-induced form.
Etanercept was discontinued six weeks ago on that basis. Her rash has partially improved, but her joint pain and fatigue persist essentially unchanged, and repeat serology drawn this week still shows an elevated anti-dsDNA titer and complement levels that remain low rather than trending back toward normal. A urinalysis obtained as part of this recheck, sent because of the atypical serologic pattern rather than any urinary symptom she has reported, showed new microscopic hematuria and a protein-to-creatinine ratio modestly above normal — a finding that was not present on a routine urinalysis done eight months ago, before any of this began, and one that was not part of her presentation when etanercept was first stopped. Six weeks is not, by itself, an unusually long time for a drug-induced process to still be resolving, and that sentence is doing most of the work in the argument for waiting. It is worth noticing what it quietly assumes. A process still resolving produces numbers moving slowly in the right direction; hers have not moved at all, and the one genuinely new finding since the drug came out arrived after it.
A resolution window that hasn’t closed — or hasn’t started
Anti-TNF-induced lupus, even this atypical dsDNA-positive pattern Ramos-Casals described, generally resolves over weeks to months once the drug is stopped. Six weeks is still within that window. I'd rather continue observing a process likely to still be resolving than start immunosuppression before it's had the time it usually needs.
I'd weigh the trajectory differently. Her complement hasn't budged and her anti-dsDNA hasn't come down — that's not a process visibly resolving, even slowly. And now there's a new renal finding that wasn't there when the drug was stopped. That combination looks less like a self-limited drug reaction still working itself out, and more like a genuine lupus process the drug may have unmasked rather than simply caused.
I understand six weeks isn't unusually long in the abstract. But 'still within the window' assumes the numbers are moving in the right direction, even slowly. Hers aren't moving at all.
I think the renal finding is exactly the piece that should decide this, and neither of you has it fully characterized yet. A quantified 24-hour protein collection, and a nephrology opinion on whether this looks like true lupus nephritis or a milder reactive finding, would tell us whether we're looking at genuine ongoing disease activity or something that could still resolve without escalating treatment.
Starting hydroxychloroquine now carries very little downside regardless of which of you turns out to be right. Starting steroids for a renal finding we haven't actually characterized is a bigger commitment, and I'd rather have that specific information first before making it.
Hydroxychloroquine was started, and a nephrology referral with a quantified urine protein collection was arranged before committing to corticosteroids — addressing both the Rheumatologist's caution against premature immunosuppression and the Nephrologist's concern about unaddressed ongoing activity.
Not agreed, and stated directly: if the renal workup comes back genuinely reassuring — mild, non-progressive findings rather than a nephritic pattern — the Rheumatologist would still expect the joint and fatigue symptoms to resolve on hydroxychloroquine alone without steroids, while the Nephrologist would want a defined re-check interval regardless of a reassuring first read, given how long the serology has already stayed abnormal. Both agreed the nephrology result, not tonight's positions, should settle which of them turns out to be right.