The Drug That Stops Working the Same Way It Started
A single patient whose prokinetic stopped working the same way it started — not because his disease changed, but because the drug's own mechanism predicted this. The disagreement is about how much to ask of a 66-year-old managing an irregular dosing schedule alone.
W.H., a 66-year-old retired postal carrier, has been slow to eat a full meal since a partial gastrectomy for a bleeding ulcer eighteen months ago left him with vagal injury and a stomach that empties on its own unpredictable schedule. Erythromycin, started six weeks ago at a low prokinetic dose, gave him a genuinely good first two weeks — real symptomatic improvement, meals he could finish — before the effect began fading by week three and had largely disappeared by week five, even as he stayed on the identical dose.
This isn't a failed drug; it's the drug's own mechanism catching up with continuous use. Erythromycin's prokinetic effect comes from motilin receptor agonism at doses well below its antibiotic threshold, but motilin receptors downregulate with sustained continuous exposure the way most receptors do under constant agonist pressure — tachyphylaxis, not tolerance in the addiction sense. It is a well-described property of this drug class rather than a sign gastroparesis itself has worsened, and it is the reason the 2025 AGA gastroparesis guideline, which conditionally recommends erythromycin, pairs that recommendation with an explicit caution that its prokinetic effect is limited by rapid tachyphylaxis on continuous dosing. His symptom diary actually supports this reading directly: the fade tracks his dosing calendar almost exactly, not any change in his diet or activity. Repeat gastric emptying scintigraphy, obtained specifically to rule out disease progression as the alternative explanation, showed emptying times essentially unchanged from his original postsurgical baseline — further evidence the fade is pharmacologic, not anatomic, and squarely a property of the drug rather than a worsening of the vagal injury underlying his gastroparesis.
The real decision in front of the group isn't whether to keep prescribing erythromycin — everyone agrees the mechanism argues for a change — it's how. Continuous dosing predictably burns out the same receptor population it depends on; intermittent or pulsed dosing, giving the receptor population time to recover between exposures, is the pharmacologically coherent fix, but it asks a 66-year-old man managing his own meal schedule to adopt an irregular regimen rather than a simple twice-daily habit.
Fixing the schedule, not the diagnosis
His symptom diary is doing the diagnostic work for us here — the fade tracks his dosing calendar almost exactly, not any change in diet or activity, which is exactly the signature of motilin receptor downregulation under continuous agonist exposure rather than disease progression.
The fix follows directly from the mechanism: intermittent or pulsed dosing gives the receptor population recovery time between exposures instead of holding it under constant pressure. I'd move him to a structured pulsed schedule now rather than continue a continuous regimen that's already stopped doing its job.
The pharmacology isn't in question — I've seen this exact fade pattern before and it does track continuous dosing the way you're describing.
What I'd weigh against it is that he manages his own regimen alone, and a simple twice-daily habit is a lot easier to sustain than an irregular pulsed schedule that asks him to track a pattern rather than a routine. A drug he takes correctly and inconsistently helps him less than one that's mechanistically ideal but that he stops following after a few confused weeks.
That's a real concern, and I don't want to hand him a schedule so irregular it becomes its own burden. I'd suggest a middle path: erythromycin dosed for one week on, one week off, rather than a more complex intermittent pattern — simple enough to anchor to his existing walking routine, while still giving the receptor population real recovery time each cycle.
Erythromycin restructured to a one-week-on, one-week-off pulsed schedule, anchored to his existing daily walking routine to make the cycle easier to track, with a symptom diary continued to confirm the effect returns during "on" weeks.
Agreed on the plan; the adherence concern was noted rather than dismissed, with an explicit follow-up in four weeks specifically to check whether the schedule itself, not just the pharmacology, is actually working for him.