FDA Warning Letter to Empower: Are Add-In GLP-1s Just Copies?

In the United States, the Food and Drug Administration (FDA) issued warning letter 738238 on September 18, 2026 to Empower Clinic Services, LLC, doing business as Empower Pharmacy, in Houston, Texas, addressed to its founder and CEO, Arta Shaun Noorian¹. It is among the first times the agency has set out in writing where individualized compounding ends and mass copying of an approved drug begins⁵⁶. For prescribers of GLP-1 products with added vitamins, the letter matters for its reasoning, not just for the pharmacy named.

What FDA letter 738238 says

Signed by the Office of Compounding Quality and Compliance at the Center for Drug Evaluation and Research (CDER), the letter follows an inspection of the Houston facility from November 3 to 14, 2025 and a Form FDA 483 issued November 14, 2025. The agency acknowledges company responses dated December 8, 2025, March 18, 2026 and April 30, 2026¹.

According to the letter, for each product the firm filled more than a redacted number of orders in each month from July through October 2025. The FDA concludes that all three products "appear to be essentially copies" of the approved semaglutide and tirzepatide products¹.

The "copy" theory: why an added vitamin is not enough

Section 503A of the federal law exempts pharmacy compounders from current good manufacturing practice, adequate directions for use and FDA approval, provided conditions are met. One is that a product may not be compounded "regularly or in inordinate amounts" if it is essentially a copy of a commercially available product. A change made for an identified patient that produces a "significant difference," as determined by the prescriber, avoids that label¹.

The FDA had signaled its view on April 1, 2026. Its clarification page says it may consider a combination of semaglutide with another ingredient, such as vitamin B12 (cyanocobalamin), to be a copy if it is given the same way as the commercial product and the amounts of both components are within 10% of the commercial strengths; and that it does not intend to act against a 503A compounder that fills four or fewer prescriptions of the same product in a calendar month². The law firm Foley & Lardner stresses that the statement is a policy clarification without the force of law and that the FDA has not said what documentation a prescriber must keep⁸.

What is new in the Empower letter is the use of volume as evidence. Frier Levitt, a law firm focused on the sector, describes it as the first time the FDA has used the "pretextual" framing for GLP-1 combination products in a warning letter³. The FDA's letter sets two other findings beside volume: orders with no prescriber determination of significant difference, and determinations that "appear to be repeated verbatim" across many records, suggesting they were pre-generated¹. It also questions prescriptions generated through third-party platforms with pre-selected menu options¹.

What a "significant difference" must be able to carry

The American Med Spa Association (AmSpa) puts it this way: the medical necessity exception "is patient specific and has defined elements that need to be met" and "is not a blanket authorization to compound GLP-1s." It adds that cost alone is generally not sufficient and that adding B12 or niacinamide does not automatically create a qualifying difference⁴. Frier Levitt reaches the same conclusion: the added ingredient must produce a genuine, patient-specific difference, and the prescriber must document it³.

• Justification text repeated almost unchanged across many patients is, per the letter, a red flag¹⁴.
• Pre-selected menus on telehealth platforms are under scrutiny¹³.
• Regulatory consultants summarize the message as: 503A eligibility "must be demonstrated not assumed"⁷.

The sterility findings: smoke studies and media fills

The letter is not only about the copy theory. It cites insanitary conditions in sterile production (section 501(a)(2)(A)): the firm did not perform adequate dynamic smoke studies to demonstrate unidirectional airflow in the ISO 5 area, and its media fills were not run under the most challenging conditions. A new smoke study (revision 005, executed March 30 to April 3, 2026) remained in post-execution review and no final report was provided¹.

On the January 2026 media fill, the FDA notes that the vial count in the summary does not match the records, that some vials have no records, that multiple entries read an undefined "N/A," and that results were transcribed onto another form without a clear audit trail. It concludes that the claim that no contamination was found "cannot be supported" by those records¹. It also cites CGMP violations under 21 CFR 211, lack of approval (sections 505(a) and 301(d)) and misbranding for lack of adequate directions for use (502(f)(1))¹.

Sterility is not the only clinical concern with compounded vials. Poison-center series describe administration errors with compounded semaglutide, including 10-fold dosing errors, with nausea, vomiting and abdominal pain¹². At a California poison center, 80.0% of 1,047 GLP-1 receptor agonist exposures from 2017 to 2023 were unintentional therapeutic errors¹³, and at a Colorado center they were 69.2% of 237 cases¹⁴. These studies do not by themselves distinguish compounded from brand-name product, and they do not concern Empower.

Context: the April policy, the 503B proposal and Lilly's suit

The end of regulatory leniency was gradual. The FDA declared the tirzepatide shortage resolved on December 19, 2024 and the semaglutide shortage on February 21, 2025; enforcement discretion for 503A pharmacies ended March 5, 2025 for tirzepatide and April 24, 2025 for semaglutide². On April 30, 2026 the agency proposed excluding semaglutide, tirzepatide and liraglutide from the 503B bulks list, with comments through June 29, 2026; Commissioner Marty Makary said "outsourcing facilities cannot lawfully compound using bulk drug substances unless there is a clear clinical need"⁹.

Eli Lilly sued Empower (and, separately, Strive) over marketing compounded tirzepatide as a "personalized" alternative when, Lilly says, these are mass-produced single formulations⁵¹⁰. In May 2026 a federal judge in Houston dismissed Lilly's Lanham Act claims against Empower, finding it had not shown lost sales or irreparable harm to its brand¹¹. That ruling did not touch the FDA's regulatory question. Our October 27 article on this site covers a compounded tirzepatide vial recall and a counterfeit Ozempic indictment; this one deals with a different front: the copy theory.