Understanding FDA Regulations on Peptides in 2026

Medical Disclaimer: This article is for informational and educational purposes only and does not constitute legal advice. FDA regulations change, and this article reflects the landscape as understood in early 2026. Consult a qualified healthcare attorney or regulatory specialist for guidance specific to your situation.

The FDA regulates peptides through multiple overlapping frameworks, and understanding which framework applies to any given compound is essential for both patients and clinicians. The rules are not always intuitive — the same compound can be legal from one source and illegal from another, permitted one year and restricted the next, and available by prescription while simultaneously being sold illegally online.

This guide explains the four primary regulatory frameworks, how the compounding pharmacy permission system works, the GLP-1 shortage episode and its regulatory aftermath, what Category 2 means for research peptides, and how patients and providers can verify current status before making any decisions.

Why Peptide Regulation Is More Complex Than Drug Regulation Generally

Most drugs fit a simple model: approved drugs are available by prescription, unapproved drugs are not. Peptides don’t fit this cleanly because:

Many peptides have pharmaceutical-grade analogs that exist alongside research versions — semaglutide is an FDA-approved drug, but peptide synthesis technology means it can also be produced outside pharmaceutical channels.
Compounding pharmacy law creates a legally permitted middle category — between “FDA-approved drug” and “unapproved compound,” there is a congressionally authorized space for compounding pharmacies to prepare medications using certain bulk substances.
Some peptides straddle multiple regulatory categories — collagen peptides are dietary supplements; BPC-157 is a research chemical; sermorelin is an approved drug and a compoundable substance.
The compounding permitted lists change over time — what was legal to compound in 2022 may not be legal in 2026, and vice versa.

Understanding this complexity is not optional for anyone navigating the peptide landscape seriously.

Framework 1: FDA Drug Approval — The Highest Evidentiary Standard

The Approval Pathway

The FDA approves drugs through the New Drug Application (NDA) or Biologics License Application (BLA) process. For a peptide to reach approval, it must complete:

Pre-clinical studies: In vitro and animal toxicology, pharmacokinetics, dose-range finding
Phase 1 clinical trials: Human safety, tolerability, pharmacokinetics (typically 20-100 subjects)
Phase 2: Efficacy signals, dose optimization, side effect characterization (typically 100-500 subjects)
Phase 3: Large, randomized, controlled efficacy and safety trials (typically 500-5,000+ subjects, multi-site, often multi-year)
FDA review: The agency reviews the complete clinical package, including all adverse events, and may convene an advisory committee

This process typically costs $1-2 billion per approved drug and takes 10-15 years from initial research to approval. It is the reason FDA-approved peptide drugs have well-characterized safety and efficacy profiles — and the reason the investment is rarely made for compounds without a clear commercial pathway (which affects which peptides ever reach approval).

FDA-Approved Peptide Drugs (Current, 2026)

Compound	Brand Names	Indication	Year Approved
Semaglutide	Ozempic, Wegovy, Rybelsus	T2D, obesity	2017/2021/2019
Tirzepatide	Mounjaro, Zepbound	T2D, obesity	2022/2023
Liraglutide	Victoza, Saxenda	T2D, obesity	2010/2014
Sermorelin	Geref (discontinued branded)	Pediatric GHD	1997
Tesamorelin	Egrifta	HIV lipodystrophy	2010
Bremelanotide	Vyleesi	HSDD in premenopausal women	2019
Oxytocin	Pitocin	Labor induction, PPH	Long-established
Insulin analogs	Multiple	Diabetes	Multiple
Vasopressin	Vasostrict	Vasodilatory shock	2014
Leuprolide	Lupron	Prostate cancer, endometriosis	1985
Octreotide	Sandostatin	Acromegaly, carcinoid	1988

Approved medications can be prescribed by licensed physicians for any indication, including off-label use. Off-label prescribing is legal: a physician may prescribe an approved drug for a use not in its labeling, based on clinical judgment and available evidence. This is how sermorelin is used for adult growth hormone deficiency in men and women, even though its original approval was for pediatric GHD.

What FDA Approval Means for Patients

When you use an FDA-approved peptide drug:

The safety profile has been characterized in thousands of trial participants
The prescribing information (package insert) documents adverse events with frequency data, contraindications, drug interactions, and pregnancy/lactation risk
The manufacturing process is verified under Current Good Manufacturing Practice (cGMP) standards
Post-marketing surveillance actively monitors for new safety signals

This doesn’t mean approved drugs are risk-free — all have side effect profiles. It means the risks are known and documented, which enables informed decision-making.

Framework 2: Compounding Pharmacy Law (503A and 503B)

This is the most consequential and frequently misunderstood regulatory framework for peptide therapy.

The Foundation: FDCA Section 503A and 503B

Congress created a legal exception to FDA drug approval requirements for compounding pharmacies through the Drug Quality and Security Act of 2013, codified in Sections 503A and 503B of the Federal Food, Drug, and Cosmetic Act. Understanding why the DQSA was passed requires understanding the regulatory failure that preceded it. In September–October 2012, the New England Compounding Center (NECC), a compounding pharmacy in Framingham, Massachusetts, distributed contaminated methylprednisolone acetate (MPA) for epidural and intrathecal injection to pain management clinics across 20 states. The contaminant was Exserohilum rostratum, a dematiaceous (melanized) mold that is a common plant pathogen rarely implicated in human disease — precisely because it is poorly adapted to human body temperature and should be easily cleared by healthy immune systems. The critical exception is direct inoculation into the central nervous system, bypassing systemic immunity entirely: Exserohilum injected intrathecally or epidurally → organisms enter cerebrospinal fluid and meningeal vasculature → the mold’s melanin-containing cell wall is central to its pathogenicity in this setting. Fungal melanin (synthesized via the DHN-melanin pathway: acetyl-CoA → polyketide intermediates → 1,8-dihydroxynaphthalene → melanin polymer incorporated into the cell wall) physically quenches reactive oxygen species (superoxide, hydrogen peroxide, hydroxyl radical) generated by NADPH oxidase in the oxidative burst of responding neutrophils and macrophages → phagocyte killing is dramatically impaired → organisms survive phagocytosis and persist in granulomas → progressive fungal meningitis and basilar arachnoiditis. The cascade: 753 confirmed cases of fungal meningitis, spinal or paraspinal infections, or stroke in patients who received contaminated injections; 64 deaths. The “stroke” outcome in the outbreak data — predominantly posterior circulation ischemic strokes — has a specific anatomical and vascular explanation that illustrates how contamination pathology extends beyond simple infection. Contaminated methylprednisolone injected epidurally or intrathecally disperses through the subarachnoid space and accumulates in the basal cisterns: the CSF-filled anatomical compartments at the base of the brain that surround the basilar artery, posterior cerebral arteries, superior cerebellar arteries, and their penetrating branches. Exserohilum organisms and the accompanying inflammatory response produce basilar arachnoiditis — progressive granulomatous inflammation within the arachnoid membrane and subarachnoid space of the basal cisterns — which directly contacts the adventitia of vessels in that territory. Inflammatory cells (neutrophils, macrophages, giant cells) infiltrate vessel walls → fungal vasculitis → three overlapping ischemic mechanisms: (1) In situ thrombosis: endothelial damage from arterial wall inflammation → tissue factor upregulation on activated endothelial cells → extrinsic coagulation cascade initiation → platelet-fibrin thrombus within basilar artery branches; (2) Vasospasm: cytokine-driven smooth muscle contraction (IL-1β, TNF-α) → critical stenosis of small penetrating arteries supplying pons, thalamus, and cerebellum; (3) Thromboembolism: fragments from infected mural thrombus → distal occlusion of penetrating branches. The resulting strokes are anatomically concentrated in pontine, cerebellar, and thalamic territories — precisely the brain regions served by vessels embedded within the basal cisterns where fungal exudate concentrated. This posterior-circulation pattern distinguishes NECC-associated stroke from the anterior-circulation strokes more commonly seen in bacterial meningitis, and reflects the anatomy of the injection route rather than any unique feature of Exserohilum neurotropism (Pappas et al., N Engl J Med 2013, PMID: 23249954). It was the deadliest compounding pharmacy contamination event in US history. The DQSA 2013 was the direct legislative response: it created the 503B outsourcing facility category with FDA registration and cGMP manufacturing requirements, established the FDA’s authority to maintain bulk drug substance permitted/prohibited lists (the Category 1/2 system), and gave the FDA clearer enforcement authority over compounders operating at manufacturing scale. The microbial contamination that caused the outbreak — fungal spores introduced during compounding in a facility without pharmaceutical-grade cleanroom controls — is precisely what cGMP sterility testing, environmental monitoring, and manufacturing validation requirements are designed to prevent.

The underlying principle: Patients sometimes need medications that are not available as commercially manufactured products — a different dose, a different formulation, a combination, or a substance that lacks commercial backing despite clinical utility. Compounding pharmacies serve these needs.

The constraint: To prevent compounding from becoming a loophole for marketing unapproved drugs, Congress required the FDA to specify which bulk drug substances can and cannot be used in compounding.

Section 503A: Traditional Compounding Pharmacies

503A pharmacies compound medications for specific, identified patients based on individual prescriptions from licensed practitioners. They must be licensed by state pharmacy boards and operate within state pharmacy law.

A 503A pharmacy may use a bulk drug substance in compounding only if:

The substance appears on the FDA’s Category 1 list (nominated substances the FDA has evaluated and determined are appropriate for compounding), OR
The substance is under active FDA evaluation (nominated but not yet resolved), OR
The substance is the active ingredient in an FDA-approved drug that is either in shortage or is being compounded with a clinical difference (different route, dose, etc.)

503A pharmacies cannot distribute compounded preparations across state lines in large quantities (this would constitute manufacturing, not compounding). They also cannot compound substances on the Category 2 list (substances the FDA has determined are not appropriate for compounding).

Section 503B: Outsourcing Facilities

503B outsourcing facilities are a newer category that bridges the gap between traditional compounding and drug manufacturing. They:

Compound in larger quantities for distribution to hospitals and healthcare providers (not patient-specific)
Must register with the FDA (unlike 503A pharmacies, which are state-regulated)
Must comply with cGMP manufacturing standards (similar to pharmaceutical manufacturers). For sterile injectable peptides, this means: (1) cleanroom environment — filling operations must occur in ISO Grade A (EU Grade A) unidirectional-airflow zones (≤100 CFU/m³ airborne viable contamination), within an ISO Grade B or C background cleanroom, with validated environmental monitoring of both viable organisms and non-viable particulates; (2) sterility testing per USP <71> — membrane filtration method for parenteral products, sampling representative units from each batch, incubation in fluid thioglycollate medium (for anaerobes) and soybean-casein digest medium (for aerobes and fungi) for 14 days; (3) endotoxin testing per USP <85> — Limulus Amebocyte Lysate (LAL) assay must be performed on each batch before release, with a passing specification of ≤5 EU/kg/hr for intravenous routes or product-specific limits for subcutaneous products; (4) potency/identity assay — the active peptide must be measured by HPLC, UV spectrophotometry, or validated bioassay to confirm labeled concentration within specification (typically ±10%), and identity confirmed by mass spectrometry or amino acid analysis; (5) impurity profiling — related substances (degradation products, synthesis byproducts) tested by HPLC-UV or HPLC-MS, with specified limits for total and individual impurities; (6) stability data — real-time and accelerated stability studies supporting the labeled storage conditions and expiry date. A 503A compounding pharmacy is subject to USP Chapter <797> (pharmaceutical compounding — sterile preparations) standards, which define comparable physical infrastructure requirements — a Primary Engineering Control (PEC) at ISO 5 classification with a surrounding ISO 7 buffer zone, plus periodic sterility and environmental monitoring — but without the documentation and release-testing framework that gives 503B products their quality assurance floor. The critical gap is not primarily in cleanroom classification but in release documentation: 503A pharmacies operate under no requirement for formal batch manufacturing records (BMRs), no per-batch potency/identity assay (HPLC or mass spectrometry confirmation that labeled concentration matches actual peptide content within specification), and no formal lot-release sign-off by a qualified person reviewing an assembled test data package before product ships. Beyond-use dating (BUD) under USP <797> is assigned by default category limits — up to 45 days refrigerated for sterility-tested Category 2 compounded sterile preparations (CSPs) — rather than being derived from product-specific real-time or accelerated stability studies as required under 503B cGMP. In practice, this means a 503A-prepared peptide may meet physical environmental standards during the filling operation yet be dispensed without confirmed potency, without a characterized impurity profile, and without a traceable quality record linking each vial to documented manufacturing and testing data. That documentation gap — not the cleanroom hardware — is the meaningful quality difference that explains why 503B-sourced peptides carry lower contamination and potency-accuracy risk than 503A-sourced ones, which in turn carry far lower risk than gray-market products with no sterility or quality requirements at all
Have their own separate bulk drug substances list (which sometimes differs from 503A)
May distribute across state lines

503B facilities provide the pharmaceutical industry with a middle path: cGMP manufacturing quality for compounds that haven’t gone through full drug approval. For patients, this means that 503B-sourced compounds have higher quality assurance than 503A compounds, which in turn have far higher quality assurance than gray-market research chemicals.

503A vs 503B: Side-by-Side Comparison

Feature	503A (Traditional)	503B (Outsourcing)
Patient-specific prescription required	Yes	No
FDA registration	No (state-licensed)	Yes
Manufacturing standard	State board / USP standards	cGMP (pharmaceutical grade)
Distribution	Within state; limited interstate	Can distribute nationally
Batch/scale	Small, patient-specific	Larger, for providers/hospitals
Endotoxin/sterility testing	USP standards	cGMP standards (more stringent)
Bulk substance list	Category 1 / 503A list	Separate 503B list

The Category 1 / Category 2 System: How the FDA Decides

The Evaluation Process

The FDA evaluates bulk drug substances through a nomination and review process. Anyone — a physician, pharmacy, patient organization, or company — can nominate a substance for evaluation. The FDA then assesses:

Whether there is a clinical need for compounded versions (is there an FDA-approved equivalent that adequately serves the patient population?)
Whether the substance can be compounded into a safe preparation that meets USP standards
Whether the physical and chemical properties make it suitable for compounding
Whether there are safety concerns that would make compounding inappropriate

Substances are then placed in one of three categories:

Category 1: Appropriate for compounding (can be used in 503A/503B preparations)
Category 2: Not appropriate for compounding (cannot be used in 503A/503B preparations)
Under evaluation: Nominated but not yet resolved; may be used pending final determination (with some exceptions)

What Category 2 Actually Means

A Category 2 designation does not necessarily mean a substance is dangerous. The FDA’s reasons for Category 2 placement vary and may include:

Lack of evidence of clinical utility that would justify compounding
Safety concerns not resolvable through compounding standards — this category is narrower than it sounds, but consequential. Compounding pharmacy standards (USP <797> for sterile preparations, USP <795> for non-sterile, 503B outsourcing facility cGMP) govern manufacturing controls: environmental monitoring, endotoxin limits, sterility testing, container closure integrity, and finished-product concentration verification. They do not — and structurally cannot — substitute for pre-market toxicology studies. Three types of safety concern fall into this gap: (1) Carcinogenicity assessment: When a peptide’s mechanism involves sustained pro-angiogenic signaling (e.g., VEGFR2 upregulation, chronic HIF-1α stabilization), ICH S1A/S1B guidelines may require a 104-week rodent carcinogenicity bioassay before the compound can be considered adequately characterized. These studies cost millions of dollars and require commercial sponsorship — they are categorically outside the scope of any compounding pharmacy’s regulatory obligations. (2) Immunogenicity from aggregate species: Peptide solutions can form soluble oligomeric aggregates that retain biological activity but generate anti-drug antibodies in a subset of patients — through a mechanism that is structurally distinct from the tolerance-inducing processing of monomeric peptide. Three convergent immunological processes explain why the aggregated form is immunogenic while the identical monomeric sequence is not. First, innate pattern recognition activation: soluble oligomeric protein aggregates present repetitive surface epitopes that mimic the geometric signature of pathogen-associated molecular patterns (PAMPs) — a structural feature absent from monomeric peptide. TLR4 and TLR2 on dendritic cells and macrophages recognize these repetitive aggregate surfaces → MyD88/TRAF6/IKKβ → NF-κB nuclear translocation → CD80 and CD86 (B7-1/B7-2) co-stimulatory ligand upregulation on antigen-presenting cell surfaces. Without this innate activation, the same monomeric peptide engages CD4⁺ T cells without co-stimulation → T cell anergy or regulatory T cell induction (tolerance) rather than effector T cell priming. Second, scavenger receptor-mediated enhanced uptake: aggregates are preferentially endocytosed by dendritic cells and macrophages via scavenger receptor-A (SR-A/CD204) and CD36 — receptors specialized for recognizing hydrophobically-exposed protein surfaces characteristic of non-native (aggregated) conformations — at rates substantially exceeding uptake of monomeric peptides; higher endosomal peptide load → greater loading of MHC-II antigen presentation pockets → higher surface pMHC-II density → lower T cell activation threshold at aggregate concentrations that monomeric forms cannot achieve. Third, T-independent B cell receptor crosslinking: aggregates carrying multiple copies of the same peptide sequence can simultaneously engage and crosslink multiple B cell surface immunoglobulins on a single B cell → BCR clustering → BLNK/Syk/PI3Kδ downstream signaling → B cell activation without conventional T cell help → class-switched anti-peptide IgG production independent of germinal center reactions. Anti-drug antibodies (ADA) that bind the peptide’s receptor-engaging pharmacophore neutralize therapeutic efficacy by competitive inhibition; ADA against structural modifications (fatty acid chains in semaglutide, PEG linkers) additionally disrupt albumin binding and eliminate the engineered extended half-life. Compounding standards’ reliance on reverse-phase HPLC (which separates by hydrophobicity and cannot distinguish aggregate from monomer at equivalent molecular weight) misses this immunogenicity risk entirely — an analytically “pure” HPLC chromatogram can co-occur with a preparation containing immunogenic oligomeric species. Reverse-phase HPLC purity testing — the standard compounding quality metric — separates by hydrophobicity and cannot reliably detect aggregates or endotoxin that co-elutes with the peptide peak; dedicated size-exclusion chromatography (SEC) and limulus amebocyte lysate (LAL) assays are not universally required by compounding standards. The mechanistic reason for this analytical blind spot is rarely explained but is straightforward. RP-HPLC uses a deliberately denaturing mobile phase — typically 0.1% trifluoroacetic acid (TFA) at pH ~2 paired with an increasing acetonitrile gradient — because low pH disrupts hydrogen bonds stabilizing protein secondary structure and acetonitrile at high concentrations breaks hydrophobic contacts stabilizing tertiary structure. Under these conditions, soluble peptide oligomers and aggregates denature and partially unfold, exposing the hydrophobic interior residues that drove their self-association in the first place. These newly exposed hydrophobic surfaces partition to the C18 or C8 stationary phase with essentially the same retention time as correctly folded monomer — both species elute together in the same chromatographic peak. A vial containing 90% monomer and 10% immunogenic oligomers produces an RP-HPLC trace indistinguishable from a 100% monomer preparation; the UV absorbance at 215 nm integrates total peptide mass regardless of oligomerization state. Size-exclusion chromatography (SEC) operates on a mechanistically orthogonal principle — hydrodynamic radius (Stokes radius) under near-native aqueous conditions. SEC mobile phases use phosphate, acetate, or histidine buffers at near-physiological pH (6.0–7.4) without organic denaturing agents, preserving non-covalent oligomeric associations. The porous SEC resin (e.g., Superdex 200, TSKgel G2000SWxL) contains particles with a defined pore-size distribution: monomers are small enough to diffuse into pore spaces and travel a longer total path length through the column, eluting late; oligomers and aggregates with larger Stokes radii are sterically excluded from pore volume and travel only the interstitial column volume, eluting as distinct shoulders or peaks ahead of the monomer. This earlier-eluting aggregate population is invisible to RP-HPLC but resolves clearly in SEC, which is why ICH Q6B guidance for biotechnology products specifies SEC as the primary analytical method for aggregate characterization — and why the compounding sector’s reliance on RP-HPLC as its standard quality metric structurally cannot detect the aggregate species responsible for the TLR4/scavenger receptor/BCR crosslinking immunogenicity mechanism described above. (3) Narrow therapeutic index: USP concentration specifications permit ±10% variation from labeled potency. For peptides with steep dose-response curves, a batch at 110% labeled potency followed by a batch at 90% creates a 22% effective dose swing — potentially unacceptable if the safety margin is narrow. Compounding standards cannot require sponsors to define and disclose a therapeutic index they have not studied in clinical trials.
The compound falls within the statutory definition of a biological product requiring BLA rather than NDA (some peptides are classified as biologics). The FDA’s classification of a compound as a drug (FDCA Section 505, NDA pathway) vs. a biological product (PHS Act Section 351, BLA pathway) turns primarily on production method and molecular origin, not sequence length alone. The BPCIA (Biologics Price Competition and Innovation Act, 2010) and the FDA’s 2020 implementation guidance established the operative rule: chemically synthesized polypeptides produced by solid-phase peptide synthesis (SPPS) are drugs regulated under FDCA Section 505 → eligible for the 503A/503B compounding pathway; proteins produced by recombinant DNA expression systems (E. coli, yeast, CHO cells) are biological products under PHS Act Section 351 → a separate regulatory framework with no equivalent 503A/503B compounding pathway. This distinction has concrete compounding implications: a SPPS-synthesized peptide that appears on the Category 1 permitted compounding list can be lawfully compounded by a 503A pharmacy; if the FDA reclassifies the same compound as a biologic (because it is now commercially produced by recombinant expression rather than chemical synthesis, or because it meets the PHS Act’s “virus, therapeutic serum, toxin, antitoxin, vaccine, blood component, or analogous product” definition), its 503A compounding eligibility evaporates entirely — even without a specific Category 2 placement. Insulin analogs and some cytokine fragments have navigated this boundary. For practical patient purposes: if a compound’s commercial version shifts from SPPS to recombinant manufacturing, verify that its compounding status has not been implicitly revised through reclassification before assuming continued 503A access.
Adequate commercial alternatives exist

For patients and clinicians, the practical meaning is clear: Category 2 substances cannot be dispensed by US licensed compounding pharmacies. Any source claiming to provide Category 2 compounds through a “compounding pharmacy” is either mistaken or operating outside the law.

Key Peptides and Their Current Status (Early 2026)

Available through licensed US compounding pharmacies (Category 1 / permitted):

Sermorelin (GHRH analog)
Ipamorelin (GHRP-2/3 analog)
CJC-1295 without DAC
Tesamorelin (also has FDA approval)
Thymosin Alpha-1 (Thymalfasin) — available through some 503B facilities. The contrast between Tα1’s compoundable status and Thymosin Beta-4/TB-500’s Category 2 prohibition is counterintuitive given their shared “thymosin” family name, but it reflects completely distinct molecular mechanisms with opposite carcinogenicity risk profiles rather than any classification by family membership. Thymosin Alpha-1 is a 28-amino acid N-terminally acetylated peptide whose primary mechanism is innate immune agonism via TLR9: Tα1 binds TLR9 (Toll-like receptor 9, which normally senses unmethylated CpG-rich bacterial/viral DNA) on plasmacytoid dendritic cells and B cells → MyD88 adaptor recruitment → IRAK4 (IL-1 receptor-associated kinase 4) → TRAF6 → TAK1 → IKKα/IRF7 → interferon regulatory factor 7 translocates to nucleus → IFN-α and IFN-β transcription → type I interferons drive Th1 immune skewing (via STAT1/STAT2/IRF9-driven interferon-stimulated gene expression), NK cell activation, and dendritic cell maturation → antiviral and tumor-immune surveillance amplification. Separately, Tα1 upregulates MHC class I surface expression on antigen-presenting cells through a JAK/STAT1-dependent pathway — directly relevant to its FDA-approved clinical indications in Italy, Japan, and other markets as Zadaxin (thymalfasin) for hepatitis B, hepatitis C, and immune-compromised patient support. Crucially, none of the steps in the TLR9/MyD88/IRAK4/IRF7/IFN-α cascade produces pro-angiogenic output: TLR9 signaling does not upregulate VEGF-A, does not transactivate VEGFR2, and does not stabilize HIF-1α. The FDA’s Category 2 designation for TB-500/Tβ4 is mechanistically rooted in its ILK/G-actin/MRTF-A/HIF-1α/VEGF-A pro-angiogenic pathway that could theoretically support angiogenesis in dormant micrometastases or pre-neoplastic lesions — a risk requiring ICH S1A/S1B rodent carcinogenicity bioassays to rule out. That carcinogenicity pathway is absent from Tα1’s biology. Additionally, Tα1 has published Phase 2/3 human clinical trial data from its international approval processes — evidence of clinical utility and human safety characterization that Category 1 compounding criteria explicitly weigh. The regulatory divergence therefore reflects mechanism-based safety assessment applied to distinct molecular targets, not a classification error or family-name grouping
Oxytocin (with restrictions)
PT-141/Bremelanotide — also FDA-approved as Vyleesi

Not currently available through US licensed compounding pharmacies (Category 2 / prohibited):

BPC-157 (body protection compound)
TB-500 / Thymosin Beta-4 (and most fragments)
Selank
Semax
Epithalon
Most unnamed research peptides

Important caveat: This list reflects the general regulatory landscape in early 2026. The FDA’s permitted lists are updated periodically. For current, authoritative status, consult the FDA’s published bulk drug substances lists directly, as individual substances’ statuses can change.

The BPC-157 Category 2 Designation: What Happened and Why It Matters

BPC-157 illustrates the Category 2 classification process in practical terms. The compound was used in 503A compounding for several years before the FDA completed its evaluation and placed it on the Category 2 list in 2024.

The FDA’s published rationale cited: insufficient evidence of clinical utility in humans; the absence of published human clinical trial data supporting safety and efficacy; and regulatory concerns about the compound’s status as a potential unapproved drug — specifically, the pro-angiogenic mechanism at the core of BPC-157’s pharmacology raises a carcinogenicity safety concern that cannot be dismissed without formal toxicology studies that do not exist.

BPC-157’s best-characterized mechanism of action is VEGFR2 (KDR/Flk-1) transactivation — activating the vascular endothelial growth factor receptor independently of VEGF-A ligand. VEGFR2 signaling is the primary driver of pathological tumor angiogenesis: solid tumors beyond approximately 1–2mm diameter require neovascularization to survive (the “angiogenic switch”), and most tumors achieve this by secreting VEGF-A → VEGFR2 on adjacent endothelial cells → receptor tyrosine kinase activation → sprouting angiogenesis → tumor blood supply. This is why anti-VEGFR2 antibodies (ramucirumab) and VEGF-A/VEGFR inhibitors (bevacizumab, axitinib, sunitinib) are approved cancer therapies: blocking VEGFR2 starves tumors of angiogenic support. A compound that activates VEGFR2 carries the theoretical converse risk: if pre-cancerous microlesions or dormant micrometastases are present in a patient, exogenous VEGFR2 stimulation could provide the angiogenic signal enabling those cells to progress from dormant microscopic clusters to vascularized, clinically detectable tumors. The same concern has been raised for TB-500 / thymosin beta-4, which carries a parallel but mechanistically distinct pro-angiogenic risk. TB-500 (the synthetic fragment Ac-SDKP, or the full thymosin β4 sequence) promotes angiogenesis not by transactivating VEGFR2 directly as BPC-157 does, but by upregulating VEGF-A gene expression in fibroblasts and injured tissue via an ILK (integrin-linked kinase)/Akt/HIF-1α pathway. TB-500 binds monomeric G-actin via its WH2 (Wiskott-Aldrich syndrome homology 2) domain → sequesters G-actin → shifts the G-actin:F-actin equilibrium → released MRTF-A (myocardin-related transcription factor A), which normally remains cytoplasmic when bound to G-actin, translocates to the nucleus → MRTF-A/SRF (serum response factor) complex binds CArG elements in the promoters of actin-cytoskeletal and motility genes. Separately, the reduced G-actin pool activates ILK at focal adhesions → ILK phosphorylates Akt → Akt stabilizes HIF-1α by phosphorylating and inhibiting GSK-3β (which would otherwise phosphorylate HIF-1α targeting it for VHL-mediated ubiquitination and proteasomal degradation) → stabilized HIF-1α dimerizes with ARNT/HIF-1β → binds HREs (hypoxia response elements) in the VEGF-A promoter → transcriptional upregulation of VEGF-A → secreted VEGF-A then engages VEGFR2 on adjacent endothelial cells via the normal ligand-receptor route. TB-500 therefore activates VEGFR2 indirectly through VEGF-A induction rather than receptor transactivation — the downstream angiogenic signaling cascade (VEGFR2/PLCγ/PKC, VEGFR2/PI3K/Akt/eNOS, VEGFR2/Grb2/Ras/ERK) is effectively identical in both cases. The FDA’s carcinogenicity concern is the same: whether VEGFR2 angiogenic signaling is initiated by BPC-157 transactivating the receptor directly or by TB-500 upregulating the endogenous VEGF-A ligand, the pro-angiogenic output reaching pre-neoplastic microlesions is mechanistically equivalent; TB-500 is Category 2 for this reason. Ruling this out pharmacologically requires ICH S1A/S1B rodent carcinogenicity bioassays — 104-week chronic studies in two rodent species — plus genotoxicity panels (Ames test, in vitro chromosomal aberration, in vivo micronucleus). These studies cost several million dollars and require a commercial sponsor with an economic rationale to fund them. Because BPC-157 is a synthetic pentadecapeptide with no patent protection and no commercial proprietary claim, no pharmaceutical company has economic incentive to fund carcinogenicity studies — the data required to exit Category 2 cannot be generated through the mechanism that normally generates it (commercial drug development). This is the structural catch-22: Category 2 designation requires safety data; safety data requires commercial sponsorship; commercial sponsorship requires IP protection; IP protection doesn’t exist for an off-patent compound in the public domain.

The result: Licensed US compounding pharmacies may no longer dispense BPC-157. Patients who were obtaining it through compounding pharmacies lost that access channel. The compound is now available in the US only through:

Gray-market research chemical suppliers (not subject to pharmaceutical quality standards; purchasing for human use is outside FDA-sanctioned channels)
Participation in FDA-authorized clinical trials with an IND (Investigational New Drug application)
International sources (where different regulations may apply, but importation for personal use enters a separate legal gray zone)

The catch-22 this creates: BPC-157 lacks human clinical trial data — in part because conducting trials is expensive, and the compound is in the public domain with no patent protection, eliminating commercial incentive to fund trials. Category 2 designation removes it from the compounding pathway. To exit Category 2, evidence of clinical utility and safety would need to be generated — but the regulatory and commercial barriers make generating that evidence extremely difficult without substantial academic or public funding.

This is not unique to BPC-157. It characterizes the regulatory position of many research peptides that have interesting preclinical profiles but no clear commercial path to approval.

The GLP-1 Shortage Episode: A Case Study in Regulatory Dynamics

The GLP-1 shortage and its resolution illustrates how quickly the peptide regulatory landscape can shift.

2022-2025: Shortage Compounding Permitted

Starting around 2022, semaglutide and later tirzepatide appeared on the FDA’s official drug shortage list, driven by unprecedented demand following their approval and positive trial outcomes. When a drug appears on this shortage list, 503A and 503B facilities gain temporary permission to compound it — and understanding the exact statutory mechanism explains both why the permission existed and why its termination was abrupt, nonnegotiable, and legally predetermined.

The statutory mechanism for shortage-triggered compounding permission. Under FD&C Act Section 503A(b)(1)(A)(i), a licensed 503A compounding pharmacy is ordinarily prohibited from compounding a drug product that is “essentially a copy” of an FDA-approved drug product that is commercially available — this is the drug-copy restriction, designed to prevent compounders from manufacturing inexpensive substitutes for approved drugs without the safety and efficacy data underlying the approved drug’s labeling. Section 503A(b)(1)(A)(i) includes an explicit statutory exception: the drug-copy restriction does not apply when “there is a drug shortage, as defined in section 506C(a) of this Act, in effect for the drug at issue.” Section 503B(a)(2) contains a structurally identical shortage exception for outsourcing facilities. The FDA certifies a shortage by listing the drug on its shortage database — which it does when manufacturers notify the agency under Section 506C of a permanent discontinuance or interruption likely to cause a shortage. Once the FDA certifies a shortage, the 503A/503B drug-copy restriction automatically falls away: the compounding permission is not a separately granted right but rather the default state of the statute whenever commercial availability is insufficient. No new rule, guidance, or administrative action is required to enable the compounding — the shortage listing is the permission.

This architecture makes the mechanism of permission termination equally mechanical and equally abrupt. When manufacturers submit data to the FDA demonstrating restored supply — production volume certificates, distributor inventory records, pharmacy-level availability confirmations — the FDA removes the drug from its shortage database. At that moment, the drug-copy restriction re-activates automatically, without any new rulemaking, without a notice-and-comment period, and without any administrative appeal pathway available to compounders. There is no grace period built into the statute. The FDA’s removal of semaglutide from the shortage list in early 2025 and tirzepatide in mid-2025 terminated the 503A/503B compounding permissions through this self-executing mechanism. Compounders who had invested in manufacturing infrastructure and patient pipelines built around GLP-1 compounding had, structurally, always been operating under permissions they could not extend, negotiate, or appeal — the statute’s design explicitly limits shortage-based compounding to genuine shortage periods, with commercial availability restoration as the automatic endpoint.

Hundreds of compounding pharmacies entered the market. Compounded semaglutide and tirzepatide became widely available at prices significantly below branded products (Ozempic: ~$900/month; compounded: ~$200-400/month). Millions of patients who couldn’t access or afford the branded products used compounded versions.

Quality concerns during the shortage: The FDA issued multiple warnings during this period about:

Compounded products using semaglutide sodium salt or acetate salt rather than the base form used in approved drugs — the salt forms have different pharmacokinetics and unknown safety profiles
Products with inaccurate dosing or labeling
Products from facilities that were not appropriately licensed
Products from facilities that failed inspection

The FDA’s concerns about salt forms are pharmacologically significant: the bioavailability and activity of a salt form of a peptide can differ meaningfully from the base compound. Patients who unknowingly received salt-form compounded semaglutide were receiving an inadequately characterized product.

The physical chemistry explains why. In the branded formulations (Ozempic, Wegovy), semaglutide is present as the free-base peptide — the molecule without an ionically bound counter-ion. When manufacturers or compounders produce semaglutide as the sodium salt or acetate salt, they introduce a counter-ion that changes the compound’s physical behavior in multiple pharmacologically relevant ways. First, the counter-ion alters the hydration shell surrounding the molecule in aqueous solution, affecting whether semaglutide molecules exist as monomers or self-associate into dimers and higher-order aggregates. Aggregation state affects the molecule’s ability to bind its long-chain C18 fatty acid modification to albumin — the primary plasma protein responsible for semaglutide’s extended half-life. The fatty acid/albumin interaction is the pharmacokinetic basis for semaglutide’s once-weekly dosing — and understanding the precise mechanism explains why disrupting it collapses the half-life. Semaglutide carries a C18 fatty diacid (octadecanedioic acid) connected via a bifunctional linker (two mini-PEG/OEG units plus a γ-glutamic acid spacer) to the Lys34 side-chain ε-amino group. This C18 chain inserts non-covalently into fatty acid binding site IIIA on human serum albumin (HSA) — a hydrophobic pocket lined by Ile327, Ile388, Val343, Leu453, and aromatic residues — with secondary engagement at FA sites I and II. At therapeutic plasma concentrations (~50–100 nM), the free/bound equilibrium strongly favors albumin-bound semaglutide (>99% bound). This high-affinity, reversible albumin association extends half-life through two parallel mechanisms: (1) Renal bypass — albumin at ~67 kDa substantially exceeds the glomerular filtration size cutoff for free peptides (~8 kDa); albumin-semaglutide complexes cannot be filtered by the renal glomerulus, eliminating what would otherwise be the primary clearance route for an unmodified ~4 kDa peptide; (2) Reversible plasma depot — only the tiny free (unbound) semaglutide fraction is available for GLP-1R engagement, DPP-IV cleavage, and proteolytic clearance; as free drug is consumed, the albumin-bound reservoir continuously releases new free molecules, sustaining plasma concentrations across the 7-day dosing interval (PMID:17409486). If aggregation (promoted by the salt form’s altered hydration shell) buries semaglutide’s C18 fatty diacid chains within hydrophobic intermolecular interfaces, those chains are unavailable for HSA site IIIA engagement, shifting the equilibrium dramatically toward unbound → unbound peptide is renally cleared within hours → the entire pharmacokinetic basis for once-weekly dosing is eliminated. If aggregation reduces albumin binding efficiency, the ratio of free (rapidly eliminated) to albumin-bound (slowly released) semaglutide shifts, compressing the effective half-life. Second, the salt form’s different ionization equilibria mean the injection-site pH of the reconstituted product differs from the branded formulation. This affects local tissue response — the rate of precipitation at the injection site, the local inflammatory response, and the kinetics of subcutaneous absorption. Third, regulatory characterization data (Cmax, Tmax, AUC) generated for branded semaglutide was established for the free-base form; a salt form is pharmacologically a different entity that would require its own bioequivalence data to claim equivalent pharmacokinetics. None of the compounded salt forms generated such data, meaning patients received a compound with an unknown pharmacokinetic profile relative to the approved drug.

2025: Shortage Resolution and Restriction

As manufacturers addressed supply constraints, the FDA removed semaglutide and tirzepatide from shortage lists through 2025. This triggered the expiration of the shortage-based compounding permission. The FDA communicated that 503A and 503B compounding of these compounds was no longer permitted under most circumstances.

This created legal and commercial disruption for compounding pharmacies that had built significant business around GLP-1 compounds, and left some patients without access to the lower-cost compounded versions they had been using.

The regulatory lesson: Compounding permissions tied to shortage status can change quickly. Patients building long-term treatment plans around compounded GLP-1 drugs should understand this structural instability and discuss contingency planning with their providers.

Framework 3: Dietary Supplements

Collagen peptides, whey protein hydrolysates, certain amino acid sequences, and similar food-derived peptides are regulated as dietary supplements under the Dietary Supplement Health and Education Act of 1994 (DSHEA).

Key features of dietary supplement regulation:

No pre-market approval required: Manufacturers don’t need to demonstrate safety or efficacy to the FDA before selling
Manufacturer responsibility: The manufacturer is responsible for ensuring the supplement is safe — but the FDA cannot require proof before market entry
No efficacy claims on labels: Supplements can make “structure/function” claims (“supports joint health,” “maintains bone density”) but cannot claim to diagnose, treat, cure, or prevent diseases — making that claim converts the product into an unapproved new drug. The statutory mechanism: FD&C Act Section 201(g)(1)(B) defines a “drug” as any article “intended for use in the diagnosis, cure, mitigation, treatment, or prevention of disease in man or other animals.” Intended use is determined from the totality of evidence — label text, advertising, website claims, testimonials — not only the words printed on the bottle. A collagen peptide supplement whose label states “clinically proven to reduce osteoarthritis pain” has made a disease claim, satisfies the Section 201(g) drug definition, and is immediately subject to FDCA Section 505’s pre-market approval requirement — an NDA requiring large-scale clinical trials the manufacturer almost certainly has not conducted. DSHEA Section 403(r)(6) carves out the permitted category: claims describing the product’s effect on the body’s normal “structure or function” without referencing any named disease state (“supports healthy joint cartilage,” “promotes skin elasticity”) are lawful structure/function claims, provided the manufacturer submits a 30-day pre-market notification to the FDA and the finished label carries the mandatory disclaimer required by 21 CFR 101.93: “This statement has not been evaluated by the Food and Drug Administration. This product is not intended to diagnose, treat, cure, or prevent any disease.” The boundary is not always obvious — “reduces inflammation” is borderline because inflammation is associated with inflammatory diseases; “supports a healthy inflammatory response” achieves the structure/function framing without disease reference. The FDA’s Division of Dietary Supplement Programs enforces this boundary primarily through warning letters; making a disease claim converts the product to an unapproved new drug and subjects the manufacturer to FDCA Section 301(a) violations (introduction of adulterated or misbranded drugs into interstate commerce), with remedies including product seizure, injunctive relief, and criminal prosecution for knowing violations
FDA enforcement authority: The FDA can take post-market action against adulterated or misbranded supplements, but the burden of proof is on the FDA — and the legal standards for doing so are more demanding than drug enforcement. Under 21 U.S.C. § 342(f)(1)(A), to pursue a dietary supplement as adulterated, the FDA must demonstrate that it presents “a significant or unreasonable risk of illness or injury under conditions of use recommended or suggested in labeling, or if no conditions of use are suggested or recommended in the labeling, under ordinary conditions of use.” This risk-based standard requires affirmative FDA evidence, not merely absence of safety proof from the manufacturer. A separate hook exists for New Dietary Ingredient (NDI) notifications under 21 U.S.C. § 350b: any dietary ingredient not marketed in the US in a dietary supplement before October 15, 1994 is an NDI requiring a 75-day pre-market notification to the FDA — but even this notification process does not require FDA approval, only acknowledgment. When manufacturers fail to submit required NDI notifications for novel peptide sequences, the FDA can declare the product adulterated on that basis alone, without proving harm. Once adulteration is established under either standard, the enforcement tools available to the FDA are seizure of product (21 U.S.C. § 334), injunctions against continued distribution (21 U.S.C. § 332), and criminal prosecution for knowing violations (21 U.S.C. § 333). The practical enforcement gap: these actions are resource-intensive, require case-by-case evidence building, and generate retroactive remedies rather than pre-market gatekeeping — meaning a product can reach millions of consumers before enforcement begins. The NDI requirement explains why gray-market peptide vendors typically prefer the “research chemical” framing over “dietary supplement” framing: research chemicals intended for laboratory use are not dietary supplements and face no NDI notification requirement, no structure/function claim restrictions, and no cGMP manufacturing requirements under 21 CFR Part 111 (which apply to supplement manufacturers). The research use exemption is not a right to sell for human use — but as a framing choice it exposes vendors to fewer affirmative regulatory hooks than claiming supplement status would, while providing marginally more legal shelter for novel synthetic sequences

This permissive framework is why collagen peptide supplements are widely available in grocery stores, while pharmacological-dose research peptides require prescription or operate outside regulated channels.

Oral collagen peptides represent the largest commercially available peptide supplement category. The evidence base for skin and joint health claims is meaningful — multiple randomized controlled trials support fibroblast signaling effects from Pro-Hyp dipeptides derived from hydrolysis — and these products are appropriately regulated as supplements because they are derived from food proteins, not synthetic pharmaceutical compounds. The molecular basis for this activity is now reasonably well-characterized. Enzymatic hydrolysis of ingested collagen protein yields short peptides including Pro-Hyp (proline-hydroxyproline) and Hyp-Gly (hydroxyproline-glycine) as the predominant bioactive fragments. These dipeptides reach intestinal epithelial cells largely intact because proline and hydroxyproline are imino acids — their nitrogen is incorporated into a five-membered pyrrolidine ring, making it a cyclic secondary amine rather than the free primary α-amino group that brush border aminopeptidase N (APN/CD13) requires for substrate binding. APN is a zinc metalloprotease that cleaves free amino acids from the N-terminus of dietary peptides; its catalytic mechanism depends on the N-terminal primary α-amine coordinating the active-site zinc ion to orient the scissile bond for nucleophilic attack by a zinc-activated water molecule. When Pro or 4-Hyp is at the N-terminal position, the cyclic secondary amine cannot donate the lone-pair geometry needed for zinc coordination → the peptide is not cleaved. This structural resistance allows Pro-Hyp and Hyp-Gly to reach the brush border intact, where they are transported across the membrane via the PepT1 (SLC15A1) di-/tripeptide transporter — a low-affinity, high-capacity proton-coupled symporter that exploits the inward proton electrochemical gradient (maintained by NHE3’s Na⁺-driven H⁺ export) to drive dipeptide uptake against their concentration gradient. This defined transport selectivity is part of what distinguishes collagen peptides from most supplement claims — the absorption mechanism is specific and characterized, not inferred. Once in systemic circulation, Pro-Hyp reaches dermal and cartilaginous fibroblasts, where it activates intracellular signaling through a receptor-mediated pathway whose surface receptor has not yet been conclusively identified but whose downstream effectors are established: ERK1/2 MAPK activation → phosphorylation of the AP-1 transcription factor complex (c-Fos/c-Jun heterodimers) → binding to AP-1 response elements in the COL1A1 and COL1A2 gene promoters → upregulated Type I procollagen synthesis and secretion. Additionally, Pro-Hyp stimulates fibroblast TGF-β1 autocrine signaling through a mechanism that begins downstream of the ERK/AP-1 axis already active. Pro-Hyp-driven ERK1/2 → phospho-AP-1 complex → AP-1 response elements drive transcription not only of COL1A1/COL1A2 but also of CTGF (connective tissue growth factor) and LTBP-1 (latent TGF-β binding protein-1); LTBP-1 is a matrix protein that forms a disulfide-bonded complex with pro-TGF-β1 and targets it for secretion and extracellular matrix deposition in a latent, inactive form. Once the latent TGF-β1/LTBP-1 complex is deposited in the fibroblast pericellular matrix, it requires activation to become biologically signaling-competent: membrane-bound integrins αvβ6 and αvβ8 (expressed on activated fibroblasts) bind the RGD motif in the TGF-β1 prodomain (LAP, latency-associated peptide) → integrin-generated cytoskeletal tension mechanically unfolds LAP → active TGF-β1 is released from the latent complex into the immediate pericellular space. Locally released active TGF-β1 then binds the TGF-βR I/II heterodimer → SMAD2/3 phosphorylation → SMAD2/3 associates with SMAD4 → nuclear translocation → binding to SMAD-responsive elements (SBEs) in the COL1A1 promoter → further collagen gene induction. This mechanism explains why Pro-Hyp’s SMAD-mediated collagen induction is autocrine (the TGF-β1 is produced and activated by the same cell that receives the signal) and integrin-dependent (requiring the αvβ6/αvβ8-mediated latent TGF-β activation step), and it explains why Pro-Hyp signals through a self-amplifying collagen synthesis loop: the Pro-Hyp-driven ERK output produces not only direct AP-1/COL1A1 transcription but also the LTBP-1 secretion machinery that feeds the parallel SMAD arm (PMID: 20565587). This mechanistic specificity — a defined peptide fragment, a characterized transport mechanism, a receptor-downstream MAPK + SMAD signaling axis, and a gene-level endpoint — explains why collagen peptide evidence is more credible than most dietary supplement categories and is relevant context for understanding which supplement-category peptide products have genuine mechanistic grounding.

Framework 4: Research Chemicals and Gray-Market Vendors

Many peptides are sold online as “research chemicals” with disclaimers stating they are “for laboratory research use only — not for human consumption.” This framework deserves careful analysis.

The Legal Basis (and Its Limits)

Selling compounds for legitimate laboratory research — in vitro studies, animal experiments, academic research — is legal. Biochemical supply companies routinely sell peptides and other compounds to researchers.

The legal problem arises when:

The vendor’s actual target market is individuals intending to self-administer
The vendor’s marketing (even if indirect) implies human use
The product is sold in forms (with syringe supplies, dosing instructions, before/after testimonials) consistent with human use

Under these circumstances, the “research use only” disclaimer may not provide legal protection. The FDA can argue the vendor is effectively marketing an unapproved drug for human use — and the legal doctrine enabling this argument is well-established: the “intended use” doctrine under the Federal Food, Drug, and Cosmetic Act. Under 21 CFR 201.128, a product’s classification as a drug is determined by its intended use, which the FDA establishes from “any relevant source of evidence available including, but not limited to, labeling, promotional material, advertising, and any oral or written statements by such persons or their representatives.” The critical word is “any” — the FDA is not limited to what appears on the bottle or in the disclaimer. If a vendor’s website contains before/after testimonials from users describing injury recovery, dosing guides with syringe volumes and injection frequency, “cycle stacking” protocols, or forum post links describing human therapeutic effects, the totality of this evidence establishes human-use intent regardless of a “for research use only” disclaimer printed elsewhere. The statutory hook: FD&C Act Section 201(g)(1)(B) defines a “drug” as an article “intended for use in the diagnosis, cure, mitigation, treatment, or prevention of disease in man or other animals” — a definition that turns entirely on intended use, not on what the substance is chemically. A vendor selling a pure peptide with an RUO disclaimer but marketing through contexts that establish human therapeutic intent is selling an unapproved new drug under Section 505 of the FD&C Act, subjecting them to enforcement under Section 301(a) (introduction of misbranded or unapproved drugs into interstate commerce) and potentially Section 301(d) (manufacturing without FDA approval). FDA warning letters to research chemical vendors routinely cite vendor websites, affiliated social media accounts, and Amazon storefronts as evidence of intended human use — the physical product’s disclaimer label is legally inadequate to rebut this contextual evidence.

The FDA’s enforcement posture: The agency has issued warning letters to research chemical vendors and has taken enforcement actions in some cases. However, enforcement is resource-intensive and inconsistent. Many gray-market peptide vendors operate openly despite technically violating FDA regulations.

What Buyers Should Understand

From a patient’s perspective, obtaining peptides from gray-market vendors:

Is outside FDA-sanctioned channels regardless of the “research use only” label
Carries no quality assurance (no mandatory purity testing, no sterility testing, no endotoxin testing)
Personal importation is technically unlawful, though rarely prosecuted — characterizing it as “legally ambiguous” understates the actual regulatory posture: under FD&C Act § 381(a), the FDA may refuse admission of any drug that “appears” to violate the Act, including unapproved new drugs. Importation of an unapproved peptide research chemical for personal use constitutes introduction of an unapproved new drug in violation of Section 301(a) (prohibiting adulteration or misbranding of any drug in interstate commerce). The apparent ambiguity derives from the FDA Personal Importation Policy (Regulatory Procedures Manual, Chapter 9-71), under which the agency exercises enforcement discretion — not a legal right — for personal-use importation when: (1) the product is intended for a serious condition for which no effective domestic treatment exists; (2) there is no commercial promotion to US residents; (3) the patient acknowledges personal use and quantity is consistent with personal use (generally ≤ 3-month supply); and (4) the product does not present unreasonable risk. The critical legal distinction: the Personal Importation Policy is an internal enforcement-prioritization document that can be modified or withdrawn at any time without notice-and-comment rulemaking. Customs seizure of a personal-use peptide shipment is lawful under existing statute regardless of the Policy; the Policy creates an expectation that FDA will not pursue enforcement, not a legal right to import. Customs and Border Protection independently enforces 19 U.S.C. § 1595a(c) (seizure of merchandise imported contrary to law), parallel to FDA’s refusal-of-admission authority. Patients relying on international sourcing should understand that the non-prosecution expectation rests on administrative discretion rather than any statutory protection, and that enforcement emphasis can shift without prior notice
Does not constitute “medical supervision” even if a vendor offers access to a physician

The risk is not primarily legal for individual buyers in practice — FDA enforcement against individuals purchasing for personal use is extremely rare. The risk is primarily safety: product quality, accurate dosing, sterility, and endotoxin are all unguaranteed.

The endotoxin risk deserves specific explanation because it is the most frequently misunderstood quality hazard of injectable gray-market peptides. Endotoxin (lipopolysaccharide, LPS) is a structural component of the outer membrane of Gram-negative bacteria — the most common contaminants of non-pharmaceutical manufacturing environments. Unlike live bacteria, endotoxin is a heat-stable pyrogen: autoclaving at 121°C kills organisms but does not destroy LPS. A vial can be bacteriologically sterile while containing dangerous endotoxin from bacteria killed during any sterilization step applied post-manufacturing. When injected subcutaneously or intramuscularly, LPS binds TLR-4/MD-2 (Toll-like receptor 4 and its essential co-receptor MD-2) on resident macrophages and dendritic cells → MyD88 and TRIF adaptor proteins are recruited → IRAK4 (IL-1 receptor-associated kinase 4) and TRAF6 → TAK1 → IKKβ → NF-κB (p65/p50) nuclear translocation → rapid transcription of TNF-α, IL-1β, IL-6, and IL-12 within 30–60 minutes. At low doses this produces fever, rigors, and severe injection-site reactions; at higher doses or in susceptible individuals the systemic cytokine surge can progress to septic shock physiology — vasodilation, hypotension, multi-organ failure — despite the absence of any viable bacterial infection. Pharmaceutical-grade injectable preparations must pass the Limulus Amebocyte Lysate (LAL) test (USP <85>) before release, with regulatory endotoxin limits set per route of administration (typically 5 EU/kg/hour for intravenous routes; sub-cutaneous limits are higher but not unlimited). Gray-market research chemical preparations face no such testing requirement and routinely report no endotoxin data. This pyrogen risk — not bacterial contamination, which is understood, but endotoxin from killed organisms that standard sterilization leaves intact — is the most concrete and immediately serious safety hazard of injectable gray-market peptides.

International Regulation: A Brief Comparison

For context, the US regulatory framework for peptides is more restrictive than some other countries but less permissive than others:

Canada: Health Canada regulates compounding under different rules. Some compounds restricted in the US are available through Canadian pharmacies, though importation from Canada into the US for personal use enters a gray zone.

United Kingdom: The MHRA regulates medicines, and compounding (called “specials”) is available under physician authorization. Several peptides available in the UK differ from US availability.

Australia: The TGA regulates therapeutic goods, including a special access pathway for unlicensed medicines under physician authorization (Special Access Scheme Category B). This has enabled Australian physicians to prescribe some research peptides through legitimate channels.

European Union: EMA regulates medicines, and member states have varying approaches to compounding. Some peptides are available in specific EU countries through legitimate pharmacy channels.

Practical implication: International differences in regulatory status mean that patients sometimes seek compounds from overseas sources. This intersects with US customs law, FDA importation policy, and the same quality-assurance concerns that apply to gray-market domestic sources. The FDA generally allows personal importation of small quantities for personal use (the “personal importation policy”) but reserves the right to intercept, and this policy is discretionary, not a legal right.

Telehealth and Peptide Prescribing: The Regulatory Layer

Telehealth peptide prescribing introduces a specific regulatory dimension: licensing and prescribing authority.

A physician prescribing a controlled substance via telehealth must comply with the Ryan Haight Online Pharmacy Consumer Protection Act, which generally requires an in-person evaluation before prescribing. Most peptides are not controlled substances (no DEA scheduling), so the Ryan Haight Act doesn’t apply — but telehealth prescribing of non-controlled substances is still subject to:

State medical licensing: The physician must hold a license in the state where the patient is located. A California-licensed physician cannot legally prescribe to a New York patient without New York licensure (with limited exceptions during declared public health emergencies).
Standard of care: Telemedicine prescribing is held to the same standard of care as in-person prescribing. A telehealth provider who prescribes GH secretagogues without obtaining baseline IGF-1, or who prescribes GLP-1 drugs without reviewing contraindications, may be practicing below the standard of care regardless of medium.
Pharmacy licensing: The compounding pharmacy filling the prescription must be licensed in the state where the patient is located (for 503A pharmacies).

Telehealth peptide prescribing is legal and can be appropriate — but the proliferation of wellness-focused telehealth platforms with minimal evaluation standards has created significant compliance and quality-of-care concerns.

The compliance gaps follow compound-specific patterns that are clinically consequential. For GH secretagogue prescribing (ipamorelin, CJC-1295, sermorelin), minimum standard-of-care evaluation includes: (1) Baseline IGF-1 measurement — GH secretagogues stimulate endogenous GH release; if the patient has undiagnosed pituitary micro-adenoma or baseline high-normal IGF-1, secretagogue-driven GH augmentation can push IGF-1 above the physiological range, producing the same soft tissue, carpal tunnel, and insulin resistance risks as exogenous GH — three mechanistically distinct consequences sharing the GH/IGF-1 axis as their upstream driver but operating through different tissue-specific pathways. Soft tissue swelling reflects IGF-1R/PI3K/Akt/mTOR-driven proteoglycan and procollagen synthesis in connective tissue fibroblasts combined with GH’s direct renal Na⁺-retaining effect (GH binds GHR (growth hormone receptor) expressed on proximal and distal renal tubular epithelial cells → JAK2/STAT5b signaling → STAT5b transcriptional activation of NHE3 (SLC9A3, Na⁺/H⁺ exchanger 3) at the apical tubular membrane → NHE3 exchanges luminal Na⁺ for intracellular H⁺ → net Na⁺ reabsorption from the filtrate; in parallel, GH-stimulated hepatic IGF-1 → IGF-1R on tubular cells → PI3K/Akt → phosphorylation of NHERF-1 (Na⁺/H⁺ exchanger regulatory factor 1) → NHERF-1 scaffolds and stabilizes NHE3 at the apical membrane surface, amplifying NHE3 activity beyond the transcriptional effect alone; increased tubular Na⁺ reabsorption → osmotic gradient → AQP2-mediated water co-retention in the collecting duct — the entire cascade proceeding through GHR/JAK2/STAT5b and IGF-1R/PI3K/Akt rather than through mineralocorticoid receptor/ENaC (the aldosterone-independent mechanism)), producing the characteristic periarticular edema. Carpal tunnel syndrome arises through this same fluid and connective tissue expansion operating within a specifically constrained anatomical compartment: the carpal tunnel is bounded rigidly by the flexor retinaculum anteriorly and the carpal bones on the remaining three sides, making it a nearly fixed-volume space. IGF-1-driven synoviocyte proliferation and flexor tendon sheath hypertrophy — via IGF-1R/IRS-1/PI3K/Akt/mTOR → fibroblast collagen I synthesis and synovial fluid accumulation — progressively increases intracanal pressure, compressing the median nerve against the unyielding retinaculum. Increased pressure slows conduction velocity in the median nerve’s large-diameter Aβ sensory fibers (the threshold for conduction slowing is approximately 30–40 mmHg sustained intracanal pressure) → paresthesias in the first three digits and thenar eminence; sustained compression → Wallerian degeneration of the most pressure-vulnerable large myelinated fibers → thenar motor weakness in chronic cases. Because the mechanism is connective tissue hypertrophy rather than direct neurotoxicity, carpal tunnel syndrome in GH excess is typically reversible with IGF-1-guided dose titration — which is the direct clinical rationale for requiring a baseline IGF-1 before prescribing, since only serial monitoring can identify when secretagogue-driven GH augmentation has elevated IGF-1 above the physiological range where these consequences emerge; prescribing without a baseline excludes the provider’s ability to monitor and titrate safely. (2) Active cancer history screening — GH/IGF-1 promotes tumor cell proliferation via IGF-1R/IRS-1/PI3K/Akt/mTORC1 in many solid tumor types; patients with active or recent cancer are typically excluded from GH-augmenting therapies; an intake form that asks about medical history but does not specifically elicit cancer history or medication list is not a substitute for this evaluation. (3) Diabetes or insulin resistance assessment — GH chronically impairs insulin signaling in skeletal muscle and hepatocytes through the post-receptor GHR/JAK2/STAT5b-driven induction of SOCS-2 and SOCS-3 (suppressors of cytokine signaling), which physically displace IRS-1 from the JAK2 activation scaffold — the same IRS-1 that insulin uses for PI3K/Akt/GLUT4 translocation; this GH-driven IRS-1 competition mechanism is the physiological basis of growth hormone-induced insulin resistance, and it is compounded at the level of free fatty acid flux (GH-stimulated lipolysis → elevated plasma FFA → DAG/PKC-θ → IRS-1 Ser307 phosphorylation → insulin signaling impairment). Patients with T2D or pre-diabetes may have significantly worsened glycemic control on GH secretagogues; HbA1c or fasting glucose baseline is the minimum appropriate screen.

For GLP-1 medication prescribing, the specific evaluation elements that matter are: (1) Personal or family history of medullary thyroid carcinoma (MTC) or MEN2 — GLP-1R is expressed on thyroid parafollicular C-cells (the calcitonin-secreting neuroendocrine cells from which MTC arises). GLP-1R/Gαs/cAMP → PKA → CREB Ser133 phosphorylation → upstream activation of B-Raf/MEK/ERK → cyclin D1 transcription and CDC25A phosphatase stabilization → G1→S cell cycle entry in C-cells — the same ERK/cyclin D1 proliferative cascade documented in GLP-1R-expressing ductal cells. In rodents, thyroid C-cell GLP-1R density is approximately 100–1,000-fold higher than in human C-cells (established by autoradiography and single-cell RNA sequencing studies). At these elevated receptor densities, pharmacological GLP-1R agonism in rats and mice generates a cAMP/ERK signal sufficient to produce dose-dependent C-cell hyperplasia and adenoma within the 2-year rodent carcinogenicity study duration — which is the direct basis of the black box warning. In humans, with far lower C-cell GLP-1R density, the equivalent plasma concentrations of semaglutide or liraglutide produce a proportionally attenuated cAMP signal in C-cells that epidemiological studies of GLP-1R agonist-treated patient populations have not shown to elevate MTC incidence above background. The black box warning is therefore mechanistically precautionary, reflecting rodent receptor biology extrapolation rather than human case series. The specific contraindication for MEN2A and MEN2B patients has a distinct mechanistic rationale: MEN2 results from germline gain-of-function RET proto-oncogene mutations (MEN2A: RET extracellular Cys634Arg/Tyr or related mutations → constitutive RET homodimerization without GDNF ligand; MEN2B: RET Met918Thr kinase domain mutation → elevated intrinsic catalytic activity). Constitutively active RET → sustained MAPK/ERK activity in C-cells — these cells are already biologically primed for enhanced proliferative response to any additional ERK input. The RET-to-ERK relay operates through a defined adaptor chain: constitutively active RET kinase domain undergoes transautophosphorylation at intracellular tyrosines including Tyr905 (activation loop — completing kinase activation), Tyr1015 (recruits PLCγ for IP₃/Ca²⁺ signaling), and Tyr1062 (the multidocking site most relevant to MAPK). Phospho-Tyr1062 recruits SHC1 via its PTB (phosphotyrosine-binding) domain → SHC1 is phosphorylated by RET at Tyr239/240 → phospho-SHC1 recruits GRB2 (via GRB2’s SH2 domain) → GRB2’s two SH3 domains constitutively bind SOS1 (Son of Sevenless 1, a RAS guanine nucleotide exchange factor) → the RET/SHC1/GRB2/SOS1 complex localizes SOS1 to the plasma membrane where Ras is anchored (via farnesylated CAAX motif) → SOS1 catalyzes GDP→GTP exchange on Ras → Ras-GTP binds and allosterically activates CRAF/BRAF → RAF phosphorylates MEK1/2 at Ser218/Ser222 → MEK1/2 phosphorylates ERK1/2 at Thr202/Tyr204 → active ERK1/2 drives CCND1 (cyclin D1) transcription via RSK2→CREB Ser133 phosphorylation at CRE elements and ELK-1/SRF at SRE elements in the cyclin D1 promoter → G1→S progression. In MEN2A/2B C-cells, this Ras/RAF/MEK/ERK cascade operates constitutively at elevated baseline; each additional ERK-activating input — including GLP-1R/PKA/B-Raf — stacks onto a signaling pathway that is already above its normal setpoint, providing the biologically plausible additive proliferative risk that the absolute contraindication reflects. GLP-1R-driven PKA/B-Raf/ERK signaling adds to an already-elevated ERK baseline in RET-mutant C-cells, creating a plausible additive mechanism for which the human GLP-1R density argument provides no comfort. This is why MEN2 personal or family history is an absolute rather than relative contraindication. Practical note: PKA/CREB-mediated CALCA (calcitonin gene) transcription is the same cascade that generates serum calcitonin as an MTC biomarker — baseline calcitonin measurement before GLP-1R agonist initiation and periodic monitoring in patients with relevant family history is the monitoring standard for this reason; this requires specific family history elicitation, not just a symptom checklist. (2) Gastroparesis history — GLP-1 receptor agonism slows gastric emptying through three converging peripheral and central mechanisms: (a) myenteric plexus inhibitory motor neuron activation — GLP-1Rs expressed on inhibitory motor neurons of the myenteric plexus (which innervate longitudinal and circular smooth muscle of the antrum) couple to Gs/cAMP/PKA → enhanced VIP (vasoactive intestinal peptide) and nitric oxide release from nerve terminals → reduced antral circular muscle contraction amplitude and reduced peristaltic propagation frequency, diminishing the propulsive force that drives gastric contents toward the pylorus; (b) vagal afferent ileal brake — GLP-1Rs on nodose ganglion vagal afferent neurons activate upon luminal GLP-1 release from L-cells → increased vagal afferent firing → NTS (nucleus tractus solitarius) → dorsal motor nucleus of the vagus (DMV) efferent cholinergic output to the pyloric sphincter → sustained pyloric tone between antral contractions → increased gastric outflow resistance; (c) direct pyloric smooth muscle effect — GLP-1R/Gs/cAMP/PKA on pyloric smooth muscle → PKA phosphorylation of MLCK (myosin light-chain kinase) reduces its sensitivity to calmodulin → less myosin light-chain phosphorylation → reduced tonic contractile responsiveness, compounding pyloric resistance. In patients with pre-existing gastroparesis (commonly seen in long-duration T1D or T2D with autonomic neuropathy, where enteric cholinergic neuron loss from advanced glycation product-mediated neuroinflammation has already reduced antral propulsive drive, or post-fundoplication where vagal trunk disruption impairs normal gastric accommodation), GLP-1-mediated further inhibition of the same pathways compounds an already-impaired gastric motor pattern — the inhibitory mechanisms described above are additive to existing enteric nerve deficits — explaining why clinically compensated delayed emptying can decompensate into severe refractory nausea, vomiting, and dehydration requiring hospitalization. A wellness telehealth intake that focuses on weight loss goals without asking about dysmotility symptoms or prior GI surgeries misses this contraindication entirely.

The structural problem: many wellness-focused telehealth peptide prescribers use brief online intake forms that ask about contraindications by name (e.g., “do you have thyroid cancer?”) rather than by mechanism (e.g., eliciting family history of thyroid tumors and performing a thyroid exam reference), creating a documentation artifact without genuine clinical evaluation. See our clinic evaluation guide for specific questions to assess telehealth prescribers.

When Regulatory Status Changes Mid-Treatment: A Patient’s Roadmap

One of the most disorienting aspects of peptide therapy is that regulatory status can change while a patient is actively using a compound. The GLP-1 compounding disruption in 2025 and BPC-157’s Category 2 reclassification are both examples of significant regulatory changes that affected patients mid-treatment. Understanding what to do when this happens — and how to monitor for pending changes before they occur — is practical clinical knowledge.

How Changes Happen: The Rulemaking Timeline

When the FDA moves a substance to Category 2 or removes it from shortage-based compounding permission, it typically:

Issues a proposed rule or guidance in the Federal Register with a public comment period (often 60–90 days)
Reviews public comments from pharmacies, physicians, patient advocates, and industry
Publishes a final rule that takes effect after a specified compliance window (which can be 30 days to several months)

This means patients and providers often have advance notice before a compounding restriction takes effect. The practical strategy: if you are using a compounded substance, check its FDA regulatory status periodically — specifically whether it remains on the Category 1 list and whether any proposed rulemaking affecting it is in progress.

Where to monitor: The FDA publishes updates to bulk drug substance lists on its Human Drug Compounding webpage. The Federal Register (federalregister.gov) contains all proposed and final rules. Some patient advocacy organizations and pharmacy associations monitor these publications and issue alerts.

Options When a Compoundable Substance Moves to Category 2

If a compound you are currently using is reclassified to Category 2:

Option 1: Transition to an FDA-approved equivalent (if one exists). For patients on compounded semaglutide who lost access, transitioning to branded Wegovy or Ozempic — if insurance coverage or financial assistance programs made it accessible — preserved the therapeutic goal. Not all compounds have approved equivalents, but exploring this first is the right clinical step.

Option 2: Participate in an FDA-authorized clinical trial. Investigational use of Category 2 compounds is permitted under an IND (Investigational New Drug application) because the two statutory frameworks — compounding pharmacy law and investigational drug law — operate completely independently and do not overlap. The Category 2 designation is an exercise of the FDA’s authority under the DQSA 2013 to restrict which bulk substances compounding pharmacies may use in preparing and dispensing drugs; it says nothing about whether a compound can be studied in a clinical investigation. The IND pathway is governed by an entirely separate statutory provision: 21 U.S.C. § 355(i) and its implementing regulations at 21 CFR Part 312, which create an explicit exemption from Section 505’s NDA pre-market approval requirement for drugs “intended solely for investigational use in human subjects.” An IND authorizes a registered sponsor (typically an academic institution, hospital, or biotech company, not a compounding pharmacy) to ship and administer an unapproved compound to enrolled subjects. The mechanics: the sponsor submits an IND application to the FDA containing Phase 1 safety data, the proposed investigation protocol, investigator qualifications, and informed consent procedures; the FDA then has 30 days to place a clinical hold if it finds “unreasonable or significant risk of illness or injury” to subjects or other grounds under 21 CFR 312.42. If no clinical hold is issued within 30 days, the investigation may proceed — the statute treats absence of a hold as authorization. Critically, a compound’s Category 2 designation is not itself grounds for a clinical hold: the FDA’s hold authority under 21 CFR 312.42 requires affirmative findings about subject risk, protocol inadequacy, investigator qualifications, or inadequate human subject protections — none of which are automatically triggered by Category 2 compounding status. An academic investigator can therefore submit an IND for BPC-157 following its Category 2 designation, include whatever animal and mechanistic safety data exists, and if the FDA does not place a hold within 30 days, begin human safety studies despite the compound’s simultaneous prohibition on the compounding pharmacy pathway. In practice this means that Category 2 closes the commercial-access route (prescriptions through licensed compounders) while leaving the research-access route (IND-authorized trials) intact — the two routes are parallel statutory channels, not a hierarchy where Category 2 blocks the other. ClinicalTrials.gov lists recruiting trials; enrollment in a well-run trial provides the compound under supervised conditions with proper monitoring — often at no cost to the patient.

Option 3: International access under physician supervision. Some compounds restricted in the US are legally available in other countries. Patients who travel internationally or have providers in other jurisdictions sometimes continue access this way. This is legally complex and falls outside FDA-sanctioned channels for US import.

Option 4: Gray-market sourcing with full risk acknowledgment. Some patients continue sourcing from research chemical vendors after a compound moves to Category 2. This is outside FDA-sanctioned channels, carries no quality assurance, and represents a patient-borne risk decision. A responsible provider should document that the patient understands the regulatory status change and is making an informed personal decision.

Option 5: Therapeutic substitution. Sometimes the clinical goal can be addressed by a different, still-available compound. A patient who loses access to BPC-157 for gut healing might discuss collagen peptides, glutamine, or other evidence-based gut support with their provider.

Preventing Disruption: Building in Regulatory Awareness

Before starting any compounded peptide protocol, ask your provider: “What is the current regulatory status of this compound, and what would we do if the status changes?” The answer to this question reveals whether your provider is tracking the regulatory landscape. A provider who responds with a clear plan — including what alternatives exist — is demonstrating appropriate clinical oversight.

How to Verify Regulatory Status: A Practical Guide

For any peptide you are considering:

Step 1: Check FDA approval status Search Drugs@FDA for the compound name. If it appears with an approval, you have an FDA-approved option to discuss with your provider.

Step 2: Check compounding status The FDA publishes its bulk drug substances lists on FDA.gov under “Human Drug Compounding.” Look for:

The 503A bulk drug substances list (Category 1 = permitted)
The 503B bulk drug substances list
The list of nominated substances under evaluation
The Category 2 list (not permitted for compounding)

These lists are the authoritative source. Vendors or pharmacies claiming permission to compound a Category 2 substance should be treated with significant skepticism.

Step 3: Verify pharmacy credentials For any compounding pharmacy:

Confirm 503A state licensing through the relevant state pharmacy board’s online license lookup
For 503B outsourcing facilities, confirm FDA registration at FDA’s 503B database
Look for PCAB accreditation (Pharmacy Compounding Accreditation Board) — this is voluntary but indicates commitment to quality standards beyond the minimum

Step 4: Check for drug shortage status (for approved drugs) FDA Drug Shortage Database lists current shortages. Shortage status affects whether approved drugs can be compounded.

Step 5: Ask your provider directly A qualified prescriber should be able to answer:

Is this compound FDA-approved, compoundable, or a research chemical?
Where is it coming from, and what quality testing has been done?
What is the current regulatory status, and could it change?

If a provider cannot answer these questions, that is a signal about the quality of oversight you are receiving.

Key Takeaways

Four regulatory frameworks govern peptides: FDA drug approval, compounding pharmacy law (503A/503B), dietary supplement regulation (DSHEA), and the gray-market research chemical category — each carries different quality assurance and legal implications
FDA-approved peptide drugs (semaglutide, sermorelin, insulin analogs) have the most robust regulatory standing; off-label use is legal under physician supervision
503A/503B compounding creates a legally sanctioned middle category — but only for substances on the FDA’s permitted lists; those lists change
Category 2 designation (BPC-157, TB-500, selank, semax, and others) prohibits US compounding pharmacy dispensing; these are available only through gray-market channels domestically
GLP-1 compounding was temporarily permitted during shortage periods but is now significantly restricted as shortages have resolved
Gray-market research chemicals carry no quality guarantees; the “for research use only” disclaimer does not create a legal right to sell unapproved drugs for human use
Verifying regulatory status requires checking FDA’s published bulk drug substances lists and pharmacy credentials — not relying on vendor claims

For current status on specific compounds, our peptide directory maintains regulatory status information in individual monographs, updated as the landscape changes.

Regulatory Sources

U.S. Food and Drug Administration. Compounding Laws and Policies — Section 503A. FDA.gov
U.S. Food and Drug Administration. Bulk Drug Substances That May Be Used in Compounding Under Section 503A of the Federal Food, Drug, and Cosmetic Act. FDA.gov
U.S. Food and Drug Administration. Registered Outsourcing Facilities (503B). FDA.gov
U.S. Food and Drug Administration. Drug Shortages. FDA.gov
Drug Quality and Security Act of 2013, Pub. L. 113-54, amending sections 503A and 503B of the Federal Food, Drug, and Cosmetic Act.

Understanding FDA Regulations on Peptides in 2026

Why Peptide Regulation Is More Complex Than Drug Regulation Generally