Infliximab in Remission: Checking a Level Nobody Asked For
A single Crohn's disease patient in stable clinical remission on infliximab, testing whether checking a drug level she has no symptoms to justify is worth the cost of the test itself.
Amara J., a 33-year-old woman who works as a pediatric nurse, has been in steady clinical remission from ileal Crohn's disease for the last sixteen months on maintenance infliximab, with a calprotectin that's held under 50 at every check and a life that, by her own description at today's visit, finally doesn't revolve around her diagnosis the way it did for the first two years after she was diagnosed. She works twelve-hour shifts on a busy pediatric floor, hasn't missed a single scheduled infusion, and came in today for what both she and her chart expected to be a routine, uneventful follow-up — no new symptoms, no concerns to report, nothing prompting today's visit beyond the standing three-month interval.
The question the team is actually facing has nothing to do with anything she's reported. It's whether to draw a trough infliximab level anyway, proactively, despite her being entirely asymptomatic and biochemically quiet — a strategy with real trial support behind it, though not uncomplicated support. TAXIT, the largest randomized trial directly testing proactive against reactive infliximab dosing, adjusted patients in the proactive arm to a target trough level regardless of symptoms, and found a real reduction in subsequent flares and undetectable trough levels in that group — but its primary endpoint, clinical and biochemical remission at one year, did not show a significant difference between the two strategies. TAILORIX (D'Haens et al., Gastroenterology 2018) then tested the same strategy in biologic-naive Crohn's patients and found nothing: its published title is that increasing infliximab dose based on symptoms, biomarkers and serum drug concentrations does not increase clinical, endoscopic, or corticosteroid-free remission. So the case for drawing Amara's level rests on one secondary endpoint from one trial, with a second trial finding no benefit at all — and the sixteen months she has spent asymptomatic on an unchanged dose are themselves evidence that the undetectable-trough problem TAXIT's secondary endpoint was picking up is not one she has been having.
Clinic, a routine follow-up with no complaints
I'd draw a proactive trough level today, even without symptoms. TAXIT found proactive dosing reduced subsequent flares and undetectable trough levels compared to waiting for symptoms to reappear. I know TAILORIX didn't reproduce that, and I'd argue its design is why — escalation capped at two steps, against a comparator arm that was itself escalating on symptoms. She's doing well now, but preventing the next flare before it starts is a genuinely different, arguably more valuable goal than catching it after it happens.
I'd hold off, and I'd put the second trial on the table rather than let it be explained away. TAXIT's own primary endpoint — clinical and biochemical remission at one year — didn't show a significant difference between proactive and reactive dosing, and TAILORIX, which set out to test the same thing, found no benefit on any endpoint it measured. She's asymptomatic and her calprotectin has been consistently under 50. Checking a level in a patient who isn't giving us any clinical reason to adds cost and real potential for a borderline result to prompt a dose change she probably didn't need in the first place.
You're each reading the evidence correctly and weighting a different part of it, and I want to be honest about how thin the pro-testing side actually is. TAXIT's primary endpoint genuinely was null. Its prespecified secondary analysis did find fewer subsequent flares and fewer undetectable trough levels in the proactive arm, and that isn't a lesser finding just because it wasn't the primary one — but it is one secondary endpoint, and TAILORIX found nothing at all. For someone like Amara, sixteen months into hard-won stability with a demanding job she's built her life back around, I think avoiding the next flare specifically is worth more to her than it might be to a patient less invested in staying exactly where she is. So I'd offer the level rather than order it, and be explicit with her that this is a weak-evidence test with a strong-evidence alternative: sixteen months of undetectable-flare stability on an unchanged dose is the single best predictor she has, and it is her own. If she wants the number, a normal result changes nothing and an undetectable one would genuinely be worth knowing. If she'd rather not, the evidence does not oblige her to.
Agreed: a proactive trough infliximab level drawn today, with an explicit understanding documented in the chart that a normal result changes nothing about her current regimen — the test is being run to confirm stability, not to search for a reason to intervene. Not fully agreed: whether this should become a standing every-visit practice going forward or a one-time check given how long she's been stable. Left for the next visit to decide once this result is in hand.