FDA Warning Letter Breakdown: Sterility Failures, Subpotent GLP-1 Drugs, and a Compounding Pharmacy Operating Without Control
When compounded semaglutide and tirzepatide fail sterility and potency testing, and get distributed anyway, the implications extend far beyond one pharmacy.
This breakdown is available for paid subscribers. Only paid subscribers get regular full access to our breakdowns and other analyses. If you’re not already a paid subscriber, you can upgrade here.
Want to stay out of our warning letter breakdowns? Contact us to access our global network of 3,750+ consultants and 325+ former FDA employees. We run audits, mock inspections, and remediation for 17 of the top 25 life science firms.
On January 16, the FDA’s CDER issued a warning letter to a compounding pharmacy following an inspection conducted from May 12 through June 9, 2025, at the firm’s facility in Kennett Square, Pennsylvania.
The firm compounds sterile drug products, including semaglutide and tirzepatide injectables—the active ingredients in branded GLP-1 drugs like Ozempic, Wegovy, Mounjaro, and Zepbound.
We need to break this one down because it lands in the middle of one of the most high-profile regulatory and legal battles in recent FDA history: the fight over compounded GLP-1 drugs.
With major players in this space actively pursuing legal and regulatory strategies to limit compounding of their blockbuster weight-loss and diabetes medications, and with Congress engaged in ongoing debate over the FDA’s authority to restrict compounding when drug shortages resolve, this letter gives every stakeholder something concrete to point to.
But set the politics aside for a moment. What the FDA documented here is alarming on its own terms.
This doesn’t appear to be a minor paperwork gap or a single missed test. The FDA describes a facility where drug products intended to be sterile were produced in environments that may not have provided adequate contamination protection, where products that failed sterility and potency testing were released and distributed, where out-of-specification investigation procedures were not followed in a single documented case out of 23 reviewed, and where patients were not notified when the drugs they received were found to be subpotent or failed sterility testing.
For anyone following the compounding debate, this letter is Exhibit A for why the FDA’s concerns about quality in the compounding space are not abstract.
Why this letter matters beyond the walls of one pharmacy
The compounding of GLP-1 receptor agonists has exploded alongside the surging demand for semaglutide and tirzepatide. With branded products commanding having experienced intermittent supply constraints, compounding pharmacies stepped in to fill the gap—operating under Section 503A of the Federal Food, Drug, and Cosmetic Act, which exempts qualifying compounders from certain FDA requirements, including CGMP compliance, labeling requirements, and premarket approval.
That exemption framework is built on an assumption: that compounding pharmacies are operating with adequate quality controls. When a facility demonstrates the kinds of breakdowns documented in this letter, it raises questions about the oversight model within these facilities.
This letter in a narrow respect validates, at least anecdodately, the core argument firms like Novo Nordisk and Eli Lilly are making: that compounded versions of their drugs, produced outside a more robust CGMP framework, carry real and documented patient safety risks.
Let’s quickly break it down.
Insanitary conditions and an environment that may not protect sterile products
The FDA found that drug products intended to be sterile were prepared, packed, or held under insanitary conditions, rendering them adulterated under Section 501(a)(2)(A) of the FDCA.
Two core findings seem to have driven this:
Inadequate smoke studies. The firm failed to perform adequate smoke studies under dynamic conditions to demonstrate unidirectional airflow within the ISO 5 area. Smoke studies are supposed to confirm that the air flowing over critical operations moves in a single, predictable direction: sweeping contaminants away from exposed product. The FDA found that the firm’s studies didn’t adequately demonstrate this. In fact, visual aids provided by the firm in its response actually depicted turbulence and a lack of first air in the central work area near the deck of the biological safety cabinet—exactly where most of the sterile drug products are compounded. The FDA noted that the firm’s post-inspection response described new smoke studies and corrective actions but failed to include the actual protocols or completed studies as supporting documentation.
Porous, particle-generating, and difficult-to-clean surfaces. The FDA identified production areas with surfaces that are difficult to clean, porous, or particle-generating. The firm in its response described a cleanroom remodel, but the FDA found the supporting documentation insufficient: no specification documentation for ceiling tiles, no demonstration that materials are suitable for pharmaceutical use, no updated cleanroom certification, and no environmental monitoring data for the remodeled suite. The agency also noted it remains unclear whether ceiling tiles were properly caulked or whether gaps exist.
For any facility producing injectable drug products, the cleanroom environment is the last line of defense against contamination. When that environment cannot be demonstrated to be under control, every product manufactured in it is suspect.
Products that failed testing and were distributed to patients anyway
The FDA also documented that the firm released and distributed drug products with strengths that differed from their labeled amounts and, in at least one case, released a product that failed sterility testing.
The specific findings:
Subpotent products distributed. Fluorescein 2% Ophthalmic Solution was found at potencies of 85.9% and 76.1% (well below the labeled strength). Semaglutide 2.5mg/mL Injection tested at 79.9% potency. Tirzepatide 10mg/mL Injection tested at 86.8% and 89.1% potency. These products were distributed to patients.
A sterility failure distributed. A specific lot of Tirzepatide 17mg/mL / Glycine 5mg/mL / Methylcobalamin 5mg/mL Injection failed sterility testing and was distributed.
To put this plainly, patients are presumed to have received injectable medications that the pharmacy’s own testing showed did not meet specifications for sterility or potency.
In the case of the semaglutide injection at 79.9% potency, a patient expecting a 2.5mg/mL dose was receiving something materially weaker. In the case of the tirzepatide product that failed sterility testing, a patient received an injectable that the pharmacy’s own lab determined was not sterile.
Between July 2024 and May 2025, the firm distributed sterile drug products with multiple lots failing sterility specifications and approximately multiple additional lots failing potency specifications. The scope of the problem is not a one-off event.
23 OOS investigations reviewed—zero followed the firm’s own SOP
The FDA says it reviewed 23 variance/OOS investigation reports from November 2024 through May 2025.
None of them followed the firm’s own standard operating procedure for investigating OOS results.
None included documentation of contacting the affected patients.
Nineteen of the 23 remained open at the time of the inspection.
This finding is devastating. The firm’s quality system was not functioning as a quality system. SOPs existed on paper, but the evidence shows they were not being implemented in practice. Not in isolated cases, but across every investigation the FDA reviewed during a seven-month window.
The firm’s own SOP states that OOS results must be immediately reported to Quality and the pharmacist-in-charge, and that investigation, root cause analysis, recall, and customer notification procedures “shall be followed when applicable.” The FDA pointedly notes that the “when applicable” language effectively makes investigation and notification optional rather than mandatory. It also notes that the SOP lacks defined timeframes for patient notification and investigation following OOS events.
The firm acknowledged that releasing OOS compounded preparations “reflects a deviation from internal quality assurance procedures and obligation to notify patients and prescribers.” But the FDA found that no corrective actions were provided to address the failure to notify patients and prescribers when OOS results were identified.
An “at risk” release problem
Another section of the warning letter addresses contradictory provisions within the firm’s batch release SOP.
One section appears to require final sterility test results before release.
Another section permits release prior to receiving final sterility test results.
The firm provided no scientific justification for releasing ophthalmic drug products before sterility results are available.
Further, the FDA found that a related SOP permits “at risk” release of any sterile drug product (not limited to ophthalmic preparations) if authorized by a patient or prescriber. The procedure doesn’t specify whether patients or prescribers would be informed if a product subsequently fails sterility testing.
Given the firm’s documented history of sterility failures where patients were not notified, the FDA’s concern here seems to be well-founded. An “at risk” release framework only works if the “risk” part is actually communicated and managed. The evidence suggests it was not here.
The FDA’s response assessment: corrective actions found deficient
The firm submitted two responses to the Form FDA 483 and initiated a voluntary recall of all drug products within expiry that had OOS results. The firm also ceased production of all sterile GLP-1 products and subsequently notified the FDA of its intent to resume production.
But the FDA found multiple corrective actions deficient or unverifiable, including:
The smoke studies (no supporting documentation provided).
The cleanroom remodel (insufficient specification and certification data).
The OOS investigation process (no corrective actions for failure to notify patients)
The root cause of sterility and potency failures (not identified or discussed in the firm’s responses).
The FDA strongly recommended that the firm engage a third-party consultant with relevant sterile drug manufacturing expertise to conduct a comprehensive assessment of operations.
A few key takeaways
Some important things to consider in light of this letter that has import into compounding and drug production in general:
This letter documents systemic failure, not isolated incidents. Every OOS investigation the FDA reviewed here failed to follow the firm’s own procedures. Products that failed testing were distributed. Patients were not notified. The cleanroom environment could not be demonstrated to be under control. This is what a dysfunctional quality system looks like in practice.
The GLP-1 context makes this nationally significant. Compounded semaglutide and tirzepatide are at the center of a major regulatory and legal battle. This letter gives some concrete, documented evidence of the patient safety risks that critics of compounding have been warning about. We wouldn’t be surprised if this letter is cited in legislative testimony, legal filings, and regulatory proceedings.
“At risk” release without functional follow-through is not a quality framework. Releasing sterile products before final testing is complete can be defensible—but only with rigorous communication, tracking, and notification systems. Without those controls, “at risk” release is just release.
SOPs that aren’t followed are in some ways worse than no SOPs at all. The existence of written procedures creates an expectation of compliance. When an FDA investigator finds that zero out of 23 investigations followed the firm’s own SOP, it demonstrates not just a failure of execution but a failure of management oversight.
The FDA is watching the compounding space closely. This warning letter, combined with the broader regulatory environment around GLP-1 compounding, signals that the FDA intends to hold compounding pharmacies accountable (particularly those producing high-profile drug products where patient safety concerns are most visible.
Ask yourself:
If you pulled your last 20 OOS investigation reports right now, how many actually followed your own SOP step-by-step? The problem isn’t having a bad SOP, it’s having a perfectly adequate SOP that nobody is executing. If your answer to this question requires hedging, you might have the same problem seen here.
Can you show, for every sterile product lot you’ve released in the last 12 months, that final sterility and potency results were received and reviewed before the product reached a patient — and if not, can you show exactly who was notified of the “at risk” status and what happened when results came back OOS? Again, this firm had an “at risk” release provision that the FDA found functionally meaningless because there was no documented follow-through.
