Persistent Parvovirus B19 Arthropathy: NSAIDs or Early Hydroxychloroquine
Two months after a parvovirus B19 outbreak in her classroom, a kindergarten teacher's symmetric hand arthritis hasn't resolved, though repeat rheumatoid factor and anti-CCP remain negative. The disagreement is whether her persistent symptoms have earned symptomatic DMARD therapy, or whether the more useful move is formally re-testing her against rheumatoid arthritis classification criteria rather than treating the uncertainty itself.
Renee F., a 34-year-old kindergarten teacher, has spent the last two months struggling with the same fine-motor tasks that fill most of her workday — gripping a marker at the whiteboard, buttoning small buttons on winter coats for five-year-olds who can't yet do it themselves. Ten weeks ago, a classroom outbreak of fifth disease worked through most of her students; about two weeks after the outbreak started, she developed a bright facial rash of her own, followed within days by symmetric swelling and pain in her wrists and the small joints of both hands. Serology confirmed parvovirus B19 IgM positivity at that visit; she has since seroconverted to IgG, with declining IgM, consistent with a resolving primary infection.
She was previously healthy, with no personal or family history of inflammatory arthritis, and had never had joint symptoms of any kind before this outbreak. Eight weeks out from the onset of her joint symptoms — well past the typical several-week course of parvovirus-associated arthropathy — she still has bilateral wrist and MCP synovitis on exam, with roughly forty-five minutes of morning stiffness. Rheumatoid factor and anti-CCP antibody, checked at symptom onset and again this week, are both negative on both occasions; a second negative result at eight weeks argues against an evolving seropositive rheumatoid arthritis rather than simply reflecting how early in the process anyone is still looking. Plain films of both hands show no erosive change.
The long-term natural history is more reassuring than her eight weeks feel. Speyer and colleagues followed fifty-four adults with recent parvovirus B19 infection for a mean of five years and found none of them left with persistent joint swelling or restricted motion. Against that sits Naides and colleagues' earlier, smaller series of twenty-one adults, in which several of those whose arthritis did turn chronic went on to meet criteria for rheumatoid arthritis outright — the outcome her two negative antibody panels are meant to exclude and, in a patient seronegative throughout, cannot. Which of those two literatures she is read against turns on her own numbers rather than on how long another month feels with a classroom to run: six swollen small joints, eight weeks of symptoms, mildly elevated acute-phase reactants, negative serology twice. Scored formally against classification criteria, that is not a coin flip — it lands one point short of the line, and the missing point is the one her serology was never going to supply.
Rheumatology follow-up, eight weeks in
I'd start hydroxychloroquine today. Eight weeks of significant hand and wrist synovitis that's actually limiting her ability to do her job is a real cost, not something to keep waiting out on the strength of an average outcome. Hydroxychloroquine's safety profile is benign enough that starting it doesn't require deciding she has rheumatoid arthritis — it's symptomatic relief for a synovitis that's already lasted twice as long as most parvovirus-associated arthropathy does.
I understand why eight weeks feels like it's earned something more than NSAIDs, but Speyer and colleagues followed fifty-four adults with recent parvovirus B19 infection for a mean of five years, and not one of them was left with persistent joint swelling or restricted motion. Starting a DMARD now commits her to real monitoring and drug exposure for a process that, by the best long-term data we have, gets there without one.
Your safety argument for hydroxychloroquine is correct on its own terms, but it isn't actually an argument that she needs it — a drug being low-risk doesn't mean starting it early has no cost when the honest odds still favor watching a little longer instead.
I think you're both treating this as a choice between two empiric options when there's a more direct move available: formally score her against the 2010 ACR/EULAR RA classification criteria today, right now, rather than picking a drug and revisiting the diagnosis later. I'll concede the weak point in my own proposal first, because it's real: those criteria are written for synovitis not better explained by another disease, and she has a confirmed parvovirus infection sitting right there as another explanation. So the score isn't a diagnosis. What it is, is the only thing on this table that produces a number we can act on twice — today and again at re-check — instead of two drugs that would each leave the question exactly where it is. Run her numbers: six small joints scores three, symptom duration past six weeks scores one, mildly elevated acute-phase reactants score one, and seronegative on both draws scores zero. Five, against a threshold of six. If that had come out the other way, we would treat her as rheumatoid arthritis and start methotrexate, not hydroxychloroquine. It didn't, so we stay on NSAIDs and set a defined eight-week re-check rather than an open-ended wait.
Agreed: formally score her today against the 2010 ACR/EULAR rheumatoid arthritis classification criteria using her current joint count, symptom duration, negative serology, and mildly elevated acute-phase reactants, rather than starting either hydroxychloroquine or an extended NSAID trial empirically.
Her score today is five against a threshold of six — three for six involved small joints, one for symptom duration beyond six weeks, one for mildly elevated acute-phase reactants, and none for serology negative on both draws — leaving the group without a single answer for today, but with an explicit plan for what happens next.
Start methotrexate promptly as rheumatoid arthritis, rather than continuing to treat it as a parvovirus aftereffect once the classification threshold is actually met.
Continue ibuprofen and reassess again in eight weeks, treating this as still-resolving parvovirus-associated arthropathy per the natural-history data.
Ibuprofen continues in the interim regardless of which branch follows; nobody proposed starting hydroxychloroquine as a bridge, since it would treat the same uncertainty the reclassification plan is designed to actually resolve.