A GLP-1 drug dripped into the bloodstream during open-heart surgery did nothing to protect the lungs three months later. The lungs of patients who got the drug declined just as much as the lungs of patients who got salt water.

That is the result from a predefined substudy of the GLORIOUS trial, published in Acta Anaesthesiologica Scandinavica ↗. It is a clean negative answer to a hopeful question, and the question was a reasonable one.

The hypothesis was not crazy

Open-heart surgery is hard on the lungs. When a patient is put on a heart-lung machine (extracorporeal circulation, the pump that takes over while the surgeon works on a stopped heart), the blood runs through plastic tubing, oxygen delivery to tissue gets interrupted and then restored, and the resulting inflammation can leave the lungs measurably worse for months. This is not rare and it is not subtle.

GLP-1 receptor agonists, the drug class behind Ozempic and Wegovy, have a long list of claimed protective effects beyond blood sugar and weight. They calm inflammation in animal models. They have shown signals in the heart and the kidney. So the GLORIOUS investigators ran a continuous infusion of exenatide ↗, one of the first GLP-1 drugs ever approved, into 878 adults during their bypass surgery and the early hours after, then asked whether their lungs held up better than the placebo group's at three months. The patients were having non-emergency coronary bypass grafting, surgical aortic valve replacement, or both.

The lungs dropped, the drug did not catch them

They measured two things. The first was diffusing capacity, which is how well the lungs move gas across into the blood. In plain terms, it is a number for how efficiently you absorb oxygen with each breath. It fell from 80 percent of predicted before surgery to 72 percent three months after, a drop of 7.7 percentage points (95 percent confidence interval 6.2 to 9.1). The second was a breathing-mechanics ratio (FEV1/FVC, roughly how much of a full breath you can push out in the first second), which slipped from 0.75 to 0.73.

Both declines were real and both were statistically solid. Neither budged with the drug. The difference in decline between the exenatide and placebo groups landed at every comparison above a p-value of 0.3, which is statistics for "no signal here." Splitting the patients into subgroups did not rescue a hidden benefit either.

A note on the dose, because it matters for how much weight to put on this. This was exenatide given intravenously during surgery, not the weekly injection people take for diabetes or weight loss. The infusion route was chosen to flood the receptors during the exact window when the lungs take their hit. If the protection were going to show up anywhere, this was the design built to find it.

The pattern is worth naming

GLP-1 organ-protection claims have a habit of shrinking when someone runs the controlled test. Earlier this month a separate analysis found GLP-1 drugs cut kidney events by about 20 percent but left heart failure untouched ↗. The benefits that survive scrutiny tend to be the ones tied to the drugs' core metabolic job. The ones that read like a halo effect, lungs after bypass among them, keep coming back null.

That is not an argument against the class. Exenatide and its successors do what they were approved to do. It is an argument for treating the longer list of maybes as maybes until a trial like this one settles them. The GLP-1R target on peptidemodel hosts the marketed agonists; the protective stories attached to it are a different evidence tier from the weight and glucose data, and a result like GLORIOUS is how that tier gets sorted.

The honest read is short. The lungs take a hit from the pump. As of this trial, no GLP-1 drug stops it.