An antibiotic raised orforglipron levels. An epilepsy drug lowered them.

A common antibiotic pushed blood levels of orforglipron up. A widely used epilepsy drug pushed them down. Neither result would matter much for the GLP-1 shots people inject today, and that contrast is the point.

Orforglipron is the GLP-1 drug you swallow. Eli Lilly's candidate is a small molecule, not a peptide, which is what lets it survive the gut as a daily pill instead of a weekly injection. That same chemistry hands it a property the injectable peptides never had to worry about: it runs through the ordinary machinery of drug metabolism, so other drugs can move its levels around, and it can in principle move theirs.

A set of six phase 1 studies in healthy volunteers, published July 3 in Clinical Pharmacology and Therapeutics ↗, mapped that traffic. Orforglipron is handled by CYP3A4 (a liver enzyme that metabolizes a large share of all prescription drugs), by the OATP1B transporters that pull compounds into liver cells, and by P-glycoprotein, a pump that shoves drugs back out of cells. The study paired orforglipron with drugs that block or boost each of those routes.

Clarithromycin, an antibiotic that blocks CYP3A4, raised orforglipron exposure. Cyclosporine, an immune-suppressing drug that blocks both CYP3A4 and OATP1B, raised it more. Carbamazepine, an epilepsy and mood drug that revs CYP3A4 up, lowered orforglipron exposure. Quinidine, a heart-rhythm drug that blocks P-glycoprotein, did not meaningfully change it.

Plain version: drugs that slow the liver's cleanup let orforglipron build up. Drugs that speed the cleanup burn it off faster. The efflux pump does not seem to matter much on its own.

The other direction came out reassuringly quiet. Orforglipron had no clinically meaningful effect on midazolam (a standard probe for CYP3A4), digoxin (a P-glycoprotein probe), or atorvastatin. It nudged rosuvastatin up a little, which the authors traced to weak blocking of a separate pump called BCRP rather than anything at the liver-uptake step. A bump in simvastatin they pinned on that drug's own quirks, not on orforglipron. So the pill takes an occasional push from other drugs but rarely pushes back.

Here is why this is worth a paragraph on a peptide site. The injectable GLP-1 peptides, semaglutide ↗ among them, are large molecules cleared by being broken down into amino acids, a route that ignores CYP3A4 and the transporters entirely. That is why the injectable labels carry almost no drug-interaction warnings. Move the same GLP-1 receptor ↗ target onto a small molecule and the interaction sheet reappears, because the molecule now travels the same crowded roads as most other pills.

None of this is a safety alarm. The interactions the study found are the ordinary, manageable kind, the sort a prescriber handles by adjusting a dose or watching a level. Lilly's framing is that the work fills in gaps so those adjustments can be written into future prescribing guidance. Orforglipron already cleared its three late-stage trials ↗ on efficacy. This is the less glamorous homework that decides how easily a daily pill slots into the medicine cabinet of someone already taking three other things.

The trade is clean. Swallowing beats injecting for most people. The cost of swallowing is that the drug has to share a metabolic road with everything else you take, and the map of that road is what this study drew.