FDA's plan to keep semaglutide and tirzepatide off the outsourcing list is still a proposal. The dose studies nominators cited did not move it.

Why this note exists

pepmg's tirzepatide note covered how the shortage-based compounding windows closed in 2025. This is the next step in the same file: FDA's formal answer to requests that outsourcing facilities be allowed to keep compounding these drugs from bulk powder without a shortage. It was published in May, drew a comment-period extension, and is still open.

The notice matters for a narrower reason too. Microdosing, custom strengths, sublingual tablets and propylene glycol-free injections were all among the uses the nominators proposed, and FDA went through the evidence offered for each, document by document.[1] This note checks those documents against FDA's reading of them.

The regulatory record

What the notice does, and what it does not

An outsourcing facility, a compounder subject to FDA inspection and manufacturing-quality rules that may supply clinics with office stock rather than filling individual prescriptions, may compound from a bulk drug substance only if the substance is on the 503B Bulks List or the drug is on FDA's shortage list "at the time of compounding, distribution, and dispensing."[1] FDA's notice proposes not to put semaglutide, tirzepatide or liraglutide on that list.

Its test has two threshold questions: does some attribute of the approved product make it medically unsuitable for certain patients, and must the compounded drug be made from bulk powder rather than from the approved product. FDA says it does not count supply problems, convenience, or the cost of an approved drug as clinical need.[1] All three nominations failed the first question, so FDA "did not proceed further"; it never evaluated whether compounded versions are safe, high quality or effective.[1]

The documents are consistent on that status. The comment deadline moved from June 30 to July 30, 2026 after a request for more time.[1][2] A Federal Register query for FDA documents mentioning semaglutide on October 11 returned the proposal and its extension and no later 503B notice.[3] FDA's 503B list page names neither included nor excluded GLP-1 drugs.[4] FDA says it will answer Novo Nordisk's two citizen petitions at the same time as the final notices.[1]

Two limits matter. The notice is about 503B outsourcing facilities. Pharmacy compounding under section 503A runs under separate conditions; FDA's GLP-1 page describes them, including its policy on "essentially copies" of approved drugs.[6] The three were not in the interim policy's holding group either: FDA's interim list of 503B nominations, updated March 21, 2025, does not place any of the three in category 1, the substances FDA does not intend to act against outsourcing facilities for using while it evaluates them.[5][16]

The higher-dose argument

The trial a nominator cited: little added glucose control, more weight loss, more dropouts

Nominators proposed injectable semaglutide at up to 16 mg/mL and said "[c]ertain patients benefit from higher doses than are commercially available."[1] FDA's first answer was that the market has moved: semaglutide is now approved as Wegovy HD, a 7.2 mg/0.75 mL pen, so the proposed 4, 5 and 8 mg/mL concentrations no longer exceed an approved product; only the 16 mg/mL proposal does.[1][7] Its second answer was a trial a nominator had cited, NCT05486065, whose results have since been published.[1]

That phase 2 trial randomized 245 adults with type 2 diabetes and a BMI of at least 27 to weekly semaglutide 2, 8 or 16 mg or placebo for 40 weeks.[8] On its primary endpoint, HbA1c, analyzed by the treatment policy estimand, which counts everyone randomized whether or not they stayed on the drug, 16 mg beat 2 mg by 0.3 percentage points, with a 95% confidence interval of -0.7 to 0.2 (P = 0.245).[8] Mean HbA1c fell 1.8 points with both 2 mg and 8 mg.[8]

Aroda et al., Diabetes Care 2025; adults with type 2 diabetes and overweight or obesity on metformin; doses forced up every four weeks with no dose changes allowed.[8]

The paper also reports results FDA's summary leaves out. In the hypothetical estimand, which estimates the effect if everyone had stayed on treatment without rescue medication, 16 mg lowered HbA1c 0.5 points more than 2 mg (95% CI -1.0 to -0.1) and weight 4.5 kg more.[8] The authors call the added glucose effect "modest" and the weight dose-response "clear." They also did not adjust for multiple comparisons, and Novo Nordisk funded the trial and employed three of its six named authors.[8] FDA's point is narrower: the paper does not show that the approved strengths are medically unsuitable for anyone.[1]

The tolerability cost was real. Four of 60 people on 2 mg stopped treatment because of adverse events, against 14 of 60 on 8 mg and 11 of 62 on 16 mg, mostly gastrointestinal. Dysesthesia, abnormal skin sensation, was reported by 8% on 8 mg and 18% on 16 mg, against none on 2 mg.[8] The authors attribute part of the dropout to the forced monthly escalation.[8]

The lower-dose argument

"Hyper-responders" and microdosing came with no data, FDA says

One nominator wrote that "5-15% of the population are hyper-responders" who need lower doses. FDA wrote that the nominator gave no source for the figure and that the paper it did cite, the STEP 1 trial, "does not mention 'hyper-responders' or the 5-15% statistic."[1] For a slower or lower regimen, FDA pointed to what the approved products already allow: semaglutide pens at 0.25, 0.5 and 1 mg, Wegovy's label instruction to consider delaying escalation by four weeks, and for tirzepatide, vials and a label that says "consider a lower maintenance dosage."[1]

The Outsourcing Facilities Association cited a 2024 New York Times article on microdosing. FDA's reply: the article "does not include any data or information" showing the approved product could not deliver lower doses, and clinical need does not "turn on information about general 'trends.'"[1]

pepmg checked whether that has changed. A PubMed search on October 11, 2026 for "microdos*" in the title or abstract alongside semaglutide, tirzepatide or liraglutide returned seven records: letters, commentaries, one narrative review and an anti-doping study that gave one volunteer a single oral dose. None is a randomized trial of microdosing.[12] ClinicalTrials.gov the same day listed three registered studies that mention it, none with posted results: a randomized phase 2 trial in people with HIV planned for 30 participants and recruiting, a randomized early phase 1 study planned for 150 and enrolling by invitation, and an observational cohort planned for 2,310 people, including users of compounded products, that had not started recruiting.[13]

Formulation arguments

Sublingual tablets, propylene glycol and a 50 mg pill

For oral and sublingual products, the gap FDA identified is absorption. Approved oral semaglutide is co-formulated with an absorption enhancer, salcaprozate sodium (SNAC), and even so FDA puts its absolute bioavailability at about 0.4% to 1% for the 3, 7 and 14 mg tablets and 1% to 2% for the 1.5, 4 and 9 mg tablets and for oral Wegovy.[1] The 25 mg and 50 mg doses nominators pointed to were tested in PIONEER PLUS, 1,606 adults with type 2 diabetes, and OASIS 1, 667 adults with overweight or obesity, both with a new formulation built to raise bioavailability.[9][10] In OASIS 1, 50 mg daily cut weight 15.1% against 2.4% on placebo over 68 weeks.[10] A 25 mg tablet has since been approved; for the proposed 50 mg, FDA said a better response at a higher strength does not make the approved products unsuitable.[1] FDA also noted the nominations gave no information on the compounded formulation and did not claim it would overcome semaglutide's poor oral absorption; for sublingual and buccal products, the nominators cited a paper setting out a concept for such tablets, which FDA said identified no unsuitability of the approved drugs.[1]

On injections, one nominator said propylene glycol in Ozempic irritates the skin and claimed a study found "95% of patients preferred the formulation without propylene glycol." FDA said the study does not say that.[1] The study, run by Novo Nordisk, found the injection-site experience with a propylene glycol-free formulation, similar to Wegovy's, "almost indistinguishable" from the Ozempic-style pen, with more than 80% of injections causing no or very mild pain.[1][11] Wegovy and Wegovy HD contain no propylene glycol; every approved liraglutide product does.[1]

The loose end

Liraglutide is still on FDA's shortage list

The proposal treats liraglutide like the other two. Unlike semaglutide and tirzepatide, though, liraglutide injection is listed as "Currently in Shortage" in FDA's drug shortage database, first posted July 18, 2023, with manufacturer entries reverified or revised in August and September 2026.[14] FDA's April 2026 compounding page says semaglutide and tirzepatide are on neither the bulks list nor the shortage list.[6] Victoza presentations show limited availability, and Saxenda is listed as available "until discontinuation in January 2027."[14] The notice's liraglutide section does not mention the shortage.[1] Because section 503B separately allows bulk compounding of a drug on the shortage list, a final "no" on the bulks list would not by itself close that route while the shortage lasts, though the law's other 503B conditions would still apply.[1]

Three things this note is not saying

It is not saying the final decision is settled. FDA can finalize the proposal unchanged or modify it after reading the comments.[1]

It is not saying higher doses never help. The phase 2 trial found more weight loss at 16 mg, and FDA now approves a 7.2 mg weekly semaglutide pen.[7][8]

It is not saying any of this touches research-labeled vials. FDA's GLP-1 page, current as of October 1, 2026, says it has warned companies that sold semaglutide, tirzepatide and other GLP-1 drugs "falsely labeled 'for research purposes' or 'not for human consumption.'"[15] The 503B debate is about outsourcing facilities; those products were never part of it.

Questions people are asking