Could Teva's New Drug Protect Against Accidental Gluten?
Maybe, eventually. Teva reported that a single injection of an anti-IL-15 antibody partly protected the intestinal lining of adults who ate gluten daily for six weeks. On biopsy, the treated group lost about half as much villous height relative to crypt depth as placebo. That clears a bar the previous anti-IL-15 drug missed. But it's sponsor-reported topline data from a small study, not a published trial, not an approval, and not a reason to relax the diet.
The reason a result like this matters is that the gluten-free diet doesn't finish the job as often as you'd hope. In a Spanish study that checked gluten exposure through stool testing instead of trusting self-report, half still had villous atrophy two years in. Most of the people with persistent atrophy were asymptomatic and had negative serology.
Feeling fine is not evidence of a healed gut. Standard celiac antibody tests catch persistent villous atrophy only about half the time or less, so a clean tTG doesn't rule out ongoing damage. And that damage isn't cosmetic. In a 694-person cohort, persistent atrophy was linked to more celiac complications and higher mortality than mucosal healing. Inadvertent gluten is the usual suspect, though not the whole story.
So the question isn't whether the diet still matters. It's whether anything can back it up.
The Drug Class That Failed The First Time
IL-15 has been a logical target for years. In active celiac disease it's overexpressed in human biopsies, it helps drive the killer lymphocytes that destroy the cells lining the gut, and it can disable the regulatory brakes that would otherwise restrain them. This is grounded in human tissue evidence, not just mouse work.
Which is why it stung when the first antibody against it didn't deliver. AMG 714 randomized 64 adults already on a gluten-free diet, dosed every two weeks for 10 weeks alongside a daily gluten challenge. The primary endpoint was the same villous-height-to-crypt-depth ratio Teva used, and it showed no difference from placebo at either dose. The higher dose did blunt the rise in intraepithelial lymphocytes, and diarrhea scores improved, but those sat under a failed primary endpoint and weren't corrected for multiple testing. The architecture of the gut wasn't protected. A companion trial of the same antibody in refractory celiac disease missed its primary endpoint too.
Teva's result doesn't explain why AMG 714 failed. It argues the target wasn't dead. Same pathway, different antibody, and this time the biopsy endpoint moved.
What The 50-Person Trial Showed
Participants started with minimal intestinal damage on biopsy while on a gluten-free diet. Two weeks after a single injection under the skin of TEV '408, they began eating gluten every day for six weeks. At week 8, biopsies showed the treated group had lost about half as much villous height relative to crypt depth as placebo. The result cleared the usual statistical threshold, though the uncertainty around how big the benefit is was wide. The inflammation finding was cleaner: lymphocyte density in the gut lining stayed essentially flat on the drug while it climbed sharply on placebo. Teva also reports lower symptom scores, but released no numbers and no statistical test, so that's a claim awaiting data.
For scale: that's roughly the biopsy protection ZED1227, a transglutaminase 2 inhibitor, achieved on the same measure in a larger randomized gluten-challenge trial. ZED1227 still has no approval behind it.
| Evidence | What it showed | Where that leaves you |
|---|---|---|
| Gluten-free diet | Half still had villous atrophy at two years in one closely monitored cohort | Still the only proven treatment, and still worth doing rigorously |
| AMG 714, anti-IL-15 | Failed its biopsy endpoint under gluten challenge and its endpoint in refractory disease; partly blunted inflammatory cell buildup at the higher dose | Target plausible, approach unproven |
| TEV '408, Phase 2a topline | About half the villous-height loss versus placebo over six weeks of gluten; inflammatory cell buildup largely prevented after one dose | Investigational only, not prescribable |
How Much Weight This Can Carry
Fifty people, one dose, week-8 efficacy data, numbers released by the company that owns the drug. No peer review, no meeting presentation, no posted registry results. The registered protocol runs safety follow-up through week 28, so the no-safety-signal statement is interim, not a finished safety record. Fast Track designation, granted in May 2025, speeds development and review. It isn't approval and says nothing about whether the drug works.
There's a deeper gap, and it applies to every celiac drug tested this way. A scripted daily gluten challenge is not what happens to you at a restaurant. Latiglutenase is the cautionary case: challenge studies suggested mucosal protection, but a larger trial in symptomatic people living on a gluten-free diet found no advantage over placebo on histology or symptoms. Whether biopsy protection turns into durable healing and fewer symptoms during ordinary life is the open question for TEV '408.
None of which makes the finding small. If your gut hasn't healed despite careful eating, a medicine that reliably reduced accidental-gluten injury could matter a lot. Teva hasn't shown that. It has shown an early biopsy signal under controlled exposure, which is enough reason to watch what comes next.
What Would Settle It
Full Phase 2a data with quantified symptom results and complete safety follow-up, presented somewhere other than a press release. Then a larger trial showing protection that lasts during ordinary gluten-free eating rather than a scripted challenge. Until then TEV '408 belongs to clinical trials, and the practical work stays what it's been: rigorous gluten avoidance, a dietitian when the diet isn't holding, and follow-up that doesn't rely on symptoms or tTG alone. The encouraging part is narrow. After AMG 714 failed two earlier anti-IL-15 celiac trials, Teva reported protection on the biopsy endpoint.