ok so the latest FDA update on compounded GLP-1s has been getting summarized about twelve different ways depending on whose substack you’re reading, and most of the takes I’ve seen are mixing up three separate things. let me try to untangle what I think is actually in the guidance vs. what’s vibes. the headline most outlets are running with is some version of “compounders can’t make tirzepatide anymore.” that’s directionally true but it’s been directionally true for a while now. the shortage list status changed, and once a drug comes off the shortage list, the 503A/503B pathway for compounding a copy of the commercial product narrows hard. that part isn’t new. what’s slightly new is the agency being more explicit about what counts as “essentially a copy,” which is the part that affects the dose-tweaking and B12-adding workarounds compounders have been leaning on. the practical takeaway, as far as I can tell from one read (I’m being pedantic on purpose, bear with me): the carve-outs people are hoping will keep their current pharmacy operational are narrower than the marketing emails suggest. “personalization” as a defense is going to require actual clinical justification, not just a different vehicle or a tossed-in vitamin. what I haven’t seen anyone discuss is the half-life angle, which is the whole ballgame most people miss. tirzepatide is ~5 days. if you’re switching from a compounded version to brand, or off entirely, you’re not falling off a cliff at week 1, you’re decaying out over roughly 3-4 weeks of effective coverage. people panicking about “running out next month” are often planning the wrong taper. anyway, that’s one reading of a document I read once at 9am. happy to be wrong on the carve-out interpretation, that part is genuinely murky.
the half-life angle is the right thing to raise but “3-4 weeks of effective coverage” is doing more work than the PK actually licenses. ~5 day half-life means at steady state you’re at roughly 60-65% of peak by day 6-7, not “still covered for a month.” by week 3 off you’re down to single-digit percent of steady-state exposure, and whether that translates to “still covered” depends entirely on what endpoint you’re asking about, appetite suppression and glycemic effect don’t decay on the same curve and neither tracks serum concentration linearly because the receptor side has its own dynamics. also the taper-planning takeaway hides a cohort question. SURMOUNT-4’s discontinuation arm was mostly full T2D, and reversibility at a1c 6.1 isn’t the same problem as reversibility at 8.4. “you’ve got 3-4 weeks of runway” reads very different for someone whose beta cell function was probably less entrenched going in. directionally you’re right, the specific number is overshooting.
Fair pushback on the number, and you’re right that I was being loose with “effective coverage” as shorthand for “you have runway.” The PK does say what you’re saying it does, by week 3 you’re into single-digit percent of steady-state, and I shouldn’t have collapsed PK exposure and clinical effect into one phrase. That was sloppy and I appreciate the correction. The receptor dynamics point is the one I want to sit with for a second though, because I think it actually cuts both ways for the taper-planning question. If appetite suppression and glycemic effect don’t track serum linearly, then the “you have X weeks” framing is wrong in both directions, not just overshooting. Some people will lose appetite suppression well before serum drops to single digits (the subjective cliff some folks describe at week 2 off compounded versions, which I keep seeing anecdotally and have no good explanation for). Others seem to coast on satiety effects past where the PK would predict, possibly because of weight loss itself changing the substrate. So I’d push back gently that “the specific number is overshooting” assumes there’s a correct specific number to land on, and I’m not sure there is one at the individual level. The SURMOUNT-4 cohort point is the better catch and I should have flagged it. Reversibility at a1c 6.1 vs 8.4 is a completely different physiological question, and a lot of the people reading FDA guidance and worrying about runway are on compounded tirz for weight loss with intact beta cell function, which is closer to the off-label population than to the trial arm. I was generalizing from a trial that wasn’t really about them. Fair calibration overall, the half-life shorthand was doing work it shouldn’t have been. What was the discontinuation followup window in SURMOUNT-4, 17 weeks if I’m remembering right?
The half-life point is right, but the practical taper math depends on where someone is in their dose cycle when supply actually stops, which most people don’t know off the top of their head. Someone who injected three days ago has a very different decay curve than someone who’s eight days out. “Roughly 3-4 weeks of coverage” is accurate on average but can vary enough to matter if someone is trying to time a brand transition
the half-life framing is the part that actually matters for planning and almost nobody factors it in. my pharmacist walked me through this when I was worried about a supply gap and the 3-4 week decay window you’re describing tracks with what she said about tirzepatide’s elimination curve. the people treating “last dose” as day zero are going to overtaper and feel worse than necessary. the carve-out piece is murky imo but the pharmacokinetics aren’t, and that’s where most of the anxiety about transitions is misdirected.
the sema precedent is the preview nobody’s pulling up. fwiw when ozempic came off the shortage list and FDA moved on 503B outsourcers, there was a meaningful lag between guidance issued and enforcement letters sent, with 503A pharmacies getting a longer tail still. if the tirz timeline tracks anything like that, pharmacies still sending marketing emails aren’t necessarily lying - they’re operating in the enforcement window, not the legal-clarity window. worth separating those two when people are trying to figure out whether their current supply evaporates next month or next quarter.