Which Infection Actually Started the Arthritis
Reactive arthritis after a chlamydial infection raises a question that doesn't have one universal answer — whether antibiotics change the arthritis course turns on which organism triggered it and on how long the arthritis has already been running, and the trials answering each question enrolled different patients.
Dominic R., a 27-year-old software developer, presented three weeks ago with urethral discharge treated presumptively as chlamydial urethritis at an urgent care clinic, confirmed the following week by NAAT testing, and treated with a single dose of azithromycin at that visit. Ten days later he developed asymmetric swelling and pain in his right knee and left ankle, along with conjunctival redness his partner noticed before he did. He is HLA-B27 positive, discovered incidentally on a workup years ago for back pain that never amounted to much, and the combination now reasonably meets criteria for reactive arthritis triggered by his chlamydial infection.
His question in clinic today, asked plainly, is whether more antibiotics would help the joints, since the original infection was already treated. Two separate things determine the answer. The first is the organism. Much of the standard teaching — that reactive arthritis is post-infectious rather than ongoing infection, and that further antibiotics don't change the joint course — is drawn substantially from enteric-triggered cases (Salmonella, Shigella, Campylobacter, Yersinia), where trials including Yli-Kerttula's ciprofloxacin study found no benefit. Chlamydial disease is mechanistically different: Chlamydia trachomatis persists in a viable but not readily culturable form within synovial tissue itself, and Lauhio et al.'s placebo-controlled trial of a three-month tetracycline course found a significantly shortened illness duration in the chlamydia-triggered patients and none in the enteric ones. It enrolled patients with acute reactive arthritis, which is where Dominic's three weeks place him.
The second is how long the arthritis has already been running, and it disqualifies the strongest-looking evidence available. The CACTAR trial (Carter et al., Arthritis & Rheumatism, 2010) tested six months of rifampin plus doxycycline or azithromycin, reporting response in 63% against 20% on placebo — a striking result, but obtained in patients whose arthritis had already persisted at least six months, mean duration over ten years. Dominic is three weeks in and has never had a joint complaint before. Reaching for it today would borrow a chronic-disease result for a man who has not yet had the chance to become that patient. The honest counterweight on his own side of the line is Putschky et al.'s comparison of ten-day against four-month doxycycline in chlamydial reactive arthritis, which found no advantage to the longer course. Dominic had never heard of reactive arthritis before this week; his only point of reference is a college roommate's knee swelling after food poisoning years ago, gone within days. He keeps returning to that comparison, and it is the wrong one twice over — wrong organism, and a course that had already declared itself by the time his own has barely started.
Rheumatology clinic, three weeks after the original infection
I'd give him antibiotics rather than NSAIDs alone, and I want to be precise about which antibiotics, because the obvious answer is the wrong one. Lauhio's placebo-controlled trial gave three months of a tetracycline in acute reactive arthritis and shortened illness duration in the chlamydia-triggered patients specifically. That is Dominic's trial. Doxycycline for three months.
What I am deliberately not proposing is CACTAR's rifampin combination. Its patients had carried this disease for a mean of ten years. He has had it for twenty-one days.
My hesitation is that the broader reactive arthritis teaching I know well says the opposite — that this is a post-infectious immune phenomenon, not ongoing infection, and that further antibiotics don't reliably change the joint course. And even staying inside the chlamydial literature, Putschky compared ten days of doxycycline against four months and found nothing to show for the extra fifteen weeks. So I'm not sure the duration you're proposing is buying him anything.
I'm not doubting Lauhio exists. I'm pointing out that its benefit came from a subgroup analysis in a small trial, and that the one study designed to ask your exact duration question came back flat.
You're both right about your own evidence, and you're arguing past each other because there are two axes here, not one. On the organism axis the rheumatologist is correct: the negative trials were done in enteric disease, where the pathogen is genuinely cleared and there is nothing left to treat, whereas Chlamydia persists viable and non-culturable in synovium. On the duration axis the primary care physician is correct: Putschky is the better-designed question and it came back flat, so a three-month course is a defensible bet, not an established one.
What settles it for me is that CACTAR is off the table for him today and stays off the table until he has been sick for six months — which is precisely the thing nobody can know yet. So the decision that matters isn't which antibiotic. It's setting the date at which he stops being Lauhio's patient and starts being Carter's.
Agreed: a three-month doxycycline course started alongside scheduled naproxen, with a six-month reassessment date entered in the chart today — the infectious disease physician's two-axis reframing turned what looked like a disagreement about antibiotics into an agreement about a calendar, since the drug that would actually be best supported for Dominic depends on a fact nobody has yet.
Not agreed: whether the three months of doxycycline is worth taking at all. The primary care physician would have given ten days and NSAIDs on Putschky's strength; the rheumatologist would not withhold the one positive trial conducted in Dominic's own acute population. Both accepted the six-month date without either conceding the point, and Dominic was told plainly that the longer course is a reasonable bet rather than a proven benefit.
He was also told directly why his roommate's case, triggered by food poisoning, would reasonably have been managed without antibiotics at all — the distinction was made explicit to him rather than left as an unexplained difference in care.