Before I write a prescription for chronic insomnia, I want to know what molecule I am handing the patient, at what dose, and on what basis. Those are ordinary requirements. Yet, last week, a federal advisory committee recommended that compounding pharmacies be allowed to manufacture epitalon for insomnia without this information.
FDA career scientists had recommended against loosening compounding restrictions for epitalon. In fact, they recommended against easing restrictions for all seven peptides on the 2-day docket. Meanwhile, the recently reconstituted Pharmacy Compounding Advisory Committee (PCAC) endorsed adding six of the seven to the 503A bulk drugs list, which governs the substances a compounding pharmacy may use to compound medications. A substance on that list can be compounded without FDA approval, and without the manufacturing controls approval imposes.
Placement on the list will still require notice-and-comment rulemaking, which has not begun. A second committee meeting will take up the remaining substances before the end of February 2027.
So, what went into the votes? FDA staff scientists applied a four-part statutory test: physical and chemical characterization, history of compounding use, evidence of effectiveness, and safety. The peptides failed on multiple grounds. For several of them there is no universally accepted chemical definition or formula. Identity cannot be established, and there is no reference standard against which to measure purity. A dose presumes a defined molecule. Without one, a prescription is an instruction to administer an unspecified quantity of an unspecified substance, and the physician’s signature moves the risk onto the patient.
Set that against the standards now being pushed for vaccines. In May 2025, FDA officials called for new randomized trials to establish whether healthy people from 6 months through 64 years should receive COVID-19 vaccines, after more than 4 years of population-wide use. And in late 2025, the director of the Center for Biologics Evaluation and Research told staff that pneumococcal vaccine manufacturers would have to demonstrate that their products generate reductions in pneumonia rather than relying on antibody titers. Serologic surrogates, adequate for decades, were declared insufficient.
Demanding clinical endpoints is a defensible regulatory position, and in some applications the correct one. The problem is that the same department that is demanding new endpoints for vaccines has given no indication it will decline the recommendation from its advisors to loosen restrictions on substances whose molecular formula is unsettled or unknown.
The peptide votes didn’t take place in a vacuum, nor did they come from a panel full of long-established committee members. And just like other health department advisory committees, members appear to be rife with conflicts of interest.
In June 2025, Health and Human Services Secretary Robert F. Kennedy Jr. removed all 17 sitting members of the Advisory Committee on Immunization Practices (ACIP), citing conflicts of interest on a panel whose members were already barred from direct industry funding and required to recuse themselves from conflicted votes. Soon after, he appointed new members, many of whom carried their own conflicts. Before last week’s PCAC meeting, FDA added eight temporary members to the compounding committee. Reporting has identified panel members affiliated with telehealth and men’s health companies that sell peptide services. One member, explaining a yes vote, cited “medical freedom.”
Each substance was reviewed against a stated indication, BPC-157 for ulcerative colitis and epitalon for insomnia among them. But once a substance becomes compoundable, prescribing is not confined to the indication reviewed; medications are often prescribed off-label. A narrow review produces an open market. Compare this to the vaccine actions of the past year, which ran the other direction, narrowing eligibility by age and risk.
I have two objections that deserve answers. The first is statutory. Section 503A (the bulk drug list) is a compounding exemption Congress created. The criteria are deliberately lower than those governing a new drug application or any biologics license. But approaching a bulks-list vote like a vaccine vote compares two things Congress never meant to be compared. FDA requires that a substance be characterized physically and chemically, alongside evidence of safety and effectiveness. Staff scientists applied those criteria to these peptides and found the evidence insufficient. Yet, the committee voted in favor of six of them anyway. The issue is a matter of differing conduct regarding evidentiary standards.
My second objection is a matter of harm reduction. Compounding generally works because the standard compounded drug begins as an approved active ingredient: it has a monograph, a reference standard, and a known impurity profile that is reviewed and regulated. A compounding pharmacy can alter the dose or the vehicle, but not the identity.
That is what is missing here: several of these peptides have no accepted chemical definition and cannot be identified as unique, so the compounding pharmacy buys the same uncharacterized bulk substance from suppliers, and sterile compounding standards do not test for the failure modes that matter in a synthetic peptide. Harm reduction requires a defined product and a surveillance system capable of detecting harm. Compounded preparations carry no equivalent of the safety systems built around vaccines, and the projected telehealth market may lead to far more exposed patients, counted by no one.
The agency named these hazards itself. In 2023, FDA placed 19 peptides into Category 2 or the “unsafe” list, citing immunogenicity, limited human safety data, and contamination risk. No new evidence resolved those concerns.
Turning back to vaccines, the cost of casting doubt on immunization is already visible. Kindergarten measles, mumps, rubella (MMR) vaccine coverage fell from 95.2% in 2019-2022 to 92.5% in 2025, leaving roughly 286,000 children unprotected. The U.S. recorded more measles cases in the first 7 months of 2026 than in all of 2025. Three people died of measles in 2025, the first such deaths in this country since 2015.
The asymmetry in standards, tracking, and measurement is the danger. We can measure what happens when confidence in vaccines erodes, because measles is reportable and the count is published weekly. We will not be able to measure what happens when uncharacterized peptides enter routine practice, because nothing has been built to count them. A regulatory posture that demands new trials where the harm is already tallied, and waives characterization where the harm would go un-tallied, inverts the purpose of drug regulation. Patients on both sides of it will not know they were the experiment.
