An asthma antibody clears food-pipe inflammation on biopsy but eases swallowing less

In Short. In a year-long trial, an asthma drug cleared inflammation of the food pipe in about two-thirds of patients. Swallowing improved by less, because patients on dummy injections also reported large gains.

EOSINOPHILIC ESOPHAGITIS, or EoE, is an allergic disease of the esophagus, the tube that carries food from the mouth to the stomach. White blood cells called eosinophils gather in its lining and inflame it, so that swallowing becomes difficult and food can get stuck on the way down. Over time the inflammation can narrow the tube. Doctors diagnose the disease by passing a small camera on a tube into the esophagus, an endoscopy, and taking small samples of tissue. A count of 15 or more eosinophils in a single microscope field is the usual cutoff. In America about 1 in 700 people has the disease, and its standard treatments are special diets, acid-suppressing pills, swallowed steroids and dupilumab, an injected antibody.

Tezepelumab, an injected antibody from AstraZeneca and Amgen that is already approved for severe asthma, works one step earlier. It blocks TSLP, a signal that the lining releases when it is irritated and that sets off allergic inflammation. On October 20th the two companies released full results from CROSSING, a trial of the drug in EoE that lasted a year.

The trial enrolled 368 people aged 12 to 80, each of whom had at least two episodes of difficulty swallowing a week. Patients were assigned by chance, or randomized, to one of three groups, so that the groups would differ only in their treatment. Some 122 people received 210mg of the drug every four weeks, 122 received 420mg and 121 received dummy injections, the placebo. Neither patients nor doctors knew who was in which group.

Before the trial began, the companies set two main tests, the co-primary endpoints, and both had to succeed at 24 weeks. The first was a biopsy test. A patient counted as being in histologic remission if no microscope field held more than 6 eosinophils. The second was a swallowing diary, in which patients answered four questions each day about how often food stuck and how bad it was, giving a score from 0 to 84 over two weeks. Higher scores mean worse symptoms.

On the biopsy test the drug won clearly. After 24 weeks 64.8% of patients on the lower dose (of 122) and 68.9% on the higher dose (of 122) were in remission, against 3.5% on placebo (of 121). The diary told a different story. Scores fell by 18.8 points on the lower dose and 22.0 on the higher dose, but they also fell by 14.7 points on placebo. The lower dose therefore beat placebo by just 4.1 points out of 84.

Whether such a gap could arise by chance is judged with the p-value, the probability of seeing a difference at least this large if the drug did nothing. For the lower dose it was 0.049, just under the usual cutoff of 0.05, and the range of plausible true differences reached as low as zero. For the higher dose the gap was 7.3 points, with a p-value below 0.001. By week 52 the gaps had widened to 8.9 and 10.8 points.

The biopsy and the diary can disagree because they measure different things. A pathologist counts cells under a microscope, and on placebo that count rarely fell far enough to qualify. A diary records how patients feel, and that can improve for reasons other than the drug. In this trial patients in every group were allowed to stay on their usual acid-suppressing pills and swallowed steroids, provided the doses were stable. The companies did not say why placebo patients improved so much.

The companies' releases leave several questions open. They name the most common side effects (colds, headache and diarrhea) and describe safety as consistent with the drug's approved uses, but they have not released how often side effects occurred. They have not released how many patients dropped out of each group. Adolescents were not reported separately. The trial did not compare tezepelumab with dupilumab or with swallowed steroids. In two separate trials of dupilumab, summarized in a 2026 review by Gautam and Ruffner, about three in five patients reached the same biopsy target after 24 weeks, against about one in twenty on placebo, but trials with different patients cannot settle which drug is better. The results were presented as a conference talk and have not yet been reviewed by other scientists. It is not yet known whether biopsy remission prevents the narrowing of the esophagus that can come later.

The companies plan to share the data with regulators. A full journal paper, with figures on side effects and dropouts, will show more.