Antioxidants for Chronic Pancreatitis Pain: Which Trial Actually Describes This Patient?
Two real randomized trials of the same antioxidant combination for chronic pancreatitis pain reached opposite conclusions — one patient's own lab work is the closest thing the team has to deciding which trial's population she actually belongs to.
Priya S. moved to the United States eleven years ago and has spent most of that time working double shifts at a family-owned restaurant, a schedule that left little room for the kind of regular meals she describes eating growing up. Her chronic, idiopathic pancreatitis was diagnosed two years ago after a third admission for pain with no gallstones and no alcohol history, and her gastroenterologist has since been managing a pain pattern that flares roughly weekly, currently on acetaminophen and gabapentin with partial benefit. At her last visit, a broader nutritional panel — ordered after she mentioned brittle nails and unusually slow wound healing — came back showing a real, measured deficiency: selenium below the reference range, vitamin C at the low end of normal, and a dietary history consistent with years of irregular, restaurant-hour eating rather than any single cause.
That finding lands her squarely inside a real disagreement in the literature. A 2009 randomized trial in 127 patients (Bhardwaj et al., Gastroenterology) found that combined antioxidant supplementation — selenium, vitamin C, beta-carotene, vitamin E, and methionine — significantly reduced pain days compared with placebo in patients with chronic pancreatitis, many of whom had documented baseline micronutrient deficiency. A later UK trial, ANTICIPATE (Siriwardena et al., Gastroenterology, 2012), tested a closely related preparation and found no significant pain benefit at all. It is worth being precise about what that second trial is, because the obvious reading of it is wrong: ANTICIPATE randomized 70 patients at a single Manchester center, so it is the smaller and narrower of the two, not a larger study overturning an earlier one, and its formulation omitted beta-carotene. Its population was predominantly alcohol-related, and more than half had already failed endoscopic or surgical intervention. Priya's own deficient labs put her closer to one of those two populations than the other, without settling which trial's result actually applies to her.
Her physical exam adds a third data point worth reading alongside the labs rather than after them: her nail beds show early koilonychia, spoon-shaped changes classically linked to iron and, less specifically, to broader micronutrient depletion, and a healed cut on her forearm from several weeks ago is still faintly visible in a way her gastroenterologist notes is slower than expected for someone her age with no diabetes or vascular disease. None of these findings individually proves a nutritional cause is driving her pain, but together they describe a genuinely deficient physiology, not a borderline lab value that might just be noise — the same kind of real, converging picture that would have qualified a patient for Bhardwaj's original trial rather than excluded her from it.
Clinic, reviewing the nutritional panel
I'd start the antioxidant combination. Bhardwaj's trial found a real, statistically significant reduction in pain days with selenium, vitamin C, beta-carotene, and methionine against placebo — this isn't an uncontrolled case series, it's a randomized trial directly on this question, and she's already run through the lower-risk analgesic options with only partial benefit.
That trial exists, but it isn't the only one, and it isn't the last word. ANTICIPATE, published three years later, tested a closely related antioxidant preparation and found no significant pain benefit at all. I want to be careful about how much that buys me, though, because the reflex here is to say the newer trial wins and it doesn't earn that: ANTICIPATE randomized 70 patients at one center against Bhardwaj's 127, so this is a later and narrower study, not a bigger and better-powered one. What it does establish is that the result does not reproduce on its own, which is enough to stop us citing whichever trial happens to support what we were already inclined to try.
The general rule about weighting the bigger, later trial is usually right, but it assumes the two trials were sampling the same population — and here they plausibly weren't. Bhardwaj's cohort had a documented high rate of baseline micronutrient deficiency; ANTICIPATE's population wasn't selected for deficiency at all — it was predominantly alcohol-related, still drinking and smoking in many cases, and over half had already failed an endoscopic or surgical intervention — and a nutritionally replete population has much less room for an antioxidant to move the needle regardless of whether the underlying mechanism is real.
Priya's own labs are a real, measured deficiency, not an assumption — that doesn't resolve which trial is "right" in general, but it's a genuine, patient-specific reason to think she may sit closer to Bhardwaj's population than to ANTICIPATE's null result.
Agreed: an eight-week trial of the antioxidant combination at Bhardwaj's studied doses, gabapentin unchanged, with repeat nutritional labs and a pain diary at follow-up.
Not agreed: whether a partial pain improvement without labs normalizing should count as the therapy working. The dietitian wants to see both together before calling it a genuine effect; the gastroenterologist is willing to credit pain improvement alone, given how contested the underlying trial evidence already is either way.