An emergency physician named Vikas Patel described a patient who had stopped her statin. Coronary artery calcium score of 280. LDL up nearly 100 points since discontinuation. Father dead of a myocardial infarction at 58. She stopped the medication because she'd read online it was causing her brain fog — a side effect that, as we'll get to, a 123,940-person meta-analysis had just concluded is not pharmacologically supported.

She replaced it with BPC-157. Three injections a week into her thigh, from a gray-market website labeled "for research use only." She'd heard about it on a podcast. She was certain it was helping her knee.

Patel wrote that she represents a pattern he now sees weekly. I read that and recognized it immediately — not as my patient, but as a patient type that is becoming impossible to miss in any clinical practice that intersects with metabolic health, longevity, or weight. The specific anecdote belongs to him. The broader pattern belongs to all of us.

And what it represents is not primarily a peptide problem. It is a trust problem — one that sits at the intersection of institutional failure, commercial opportunism, and a media ecosystem that has learned to monetize medical skepticism. I want to spend some time here pulling those threads apart, because I think the clinical and cultural stakes are higher than most of the discourse around peptides acknowledges.

But before we get to the trust crisis, we need to start somewhere most peptide discourse skips: what these molecules actually are, what they do inside a body, and why the regulatory framework around them is built the way it is. Because the categorical confusion — the idea that a gray-market injectable peptide is roughly equivalent to a fish oil capsule — is not a minor misunderstanding. It is the foundation on which the entire market rests.

The Molecule in the Vial

A peptide is a chain of amino acids linked by peptide bonds. Your body uses 20 standard amino acids — each with a slightly different chemical side chain — as building blocks for virtually every structural and functional molecule in your biology. String two together and you have a dipeptide. String ten and you have a decapeptide. String fifty and you are crossing the boundary between peptide and protein.

Length matters here — regulatory and biologically — in ways we'll get to. But the more important distinction is what peptides do. The human body produces thousands of endogenous peptides, and the majority of them are not structural materials. They are information. They carry specific instructions to specific cells. Insulin, a 51-amino-acid peptide, instructs fat and muscle cells to take up glucose from the bloodstream. GLP-1, produced in the gut after eating, signals the pancreas to release insulin, the stomach to slow gastric emptying, and the hypothalamus to suppress appetite. Vasopressin, just 9 amino acids, tells the kidneys to retain water. Oxytocin, also 9 amino acids and structurally near-identical, modulates social bonding and uterine contraction. Two molecules, nearly the same sequence, utterly different jobs — because the precise arrangement of amino acids determines what receptor a peptide binds, and what that receptor then does.

This receptor-binding specificity is the key to understanding both the promise and the hazard of synthetic peptides. When a peptide binds its receptor, it doesn't just trigger a single reaction. It initiates a cascade — a conformational change in the receptor that activates intracellular signaling proteins, which phosphorylate downstream targets, which alter gene transcription, which, in turn, change cell behavior. These cascades are nonlinear, often amplified, and their effects can persist long after the peptide has been cleared from the bloodstream. You are not swallowing a compound that participates in biochemistry. You are injecting one that commands it. And that distinction changes everything about how we should think about safety.

This is also what distinguishes peptides from most supplements at the mechanistic level. Magnesium is a cofactor — it supports enzymatic reactions but does not activate receptors or initiate signaling cascades on its own. Vitamin D is a prohormone precursor that must be converted before it becomes active. Omega-3 fatty acids modulate membrane fluidity and inflammation through relatively diffuse mechanisms. Peptides, by contrast, are precision instruments. The same precision that makes GLP-1 receptor agonists so therapeutically powerful is the same precision that makes an untested synthetic peptide — injected directly into systemic circulation, with unknown receptor interactions — a genuinely different category of biological intervention than anything you'd find in the supplement aisle.

The 40-Amino-Acid Line — And Whether It Makes Scientific Sense

Federal law draws a regulatory line at roughly 40 amino acids. Below it: a compound is typically classified as a drug under the Federal Food, Drug, and Cosmetic Act, which means a licensed compounding pharmacy can legally prepare it for an individual patient with a prescription. Above it: a compound is more likely to be classified as a biologic under the Public Health Service Act, regulated through FDA's Center for Biologics Evaluation and Research, and it cannot be compounded under the traditional pharmacy framework at all. This is the architectural reason why BPC-157 — a 15-amino-acid synthetic peptide — can be prepared at a compounding pharmacy, while somatropin, a 191-amino-acid growth hormone, cannot.

How was this line drawn? The honest answer is: through a combination of legislative history, administrative pragmatism, and the limits of what Congress understood about molecular biology when writing the Biologics Price Competition and Innovation Act of 2009. The BPCIA defined biological products to include proteins — and deliberately excluded "chemically synthesized polypeptides" from that definition. The intent was to preserve regulatory flexibility for smaller synthetic peptides while creating a more rigorous approval pathway for larger, more complex biologic drugs. The ~40 amino acid threshold is not a precise statutory number; it is a rough guideline that FDA has operationalized in guidance documents and applied in practice. It was drawn at a size where synthetic manufacturing reliability, structural complexity, and immunogenic risk begin to meaningfully increase — but it was drawn there for administrative manageability, not because someone ran a controlled study and found that amino acid 41 was where the safety cliff began.

Is the size-based distinction scientifically justified? Partially. Larger peptides and proteins are generally more immunogenic — they present more antigenic surface area, are more likely to be recognized by the immune system as foreign, and carry a higher risk of provoking antibody responses that could lead to allergic reactions, serum sickness, or antibody-mediated neutralization of endogenous proteins with similar sequences. Larger molecules are also more structurally complex, more sensitive to manufacturing variability, and harder to synthesize at consistent purity. These are real pharmacological and manufacturing arguments for treating large peptides differently.

But the argument breaks down at the edges — which is exactly where most gray-market peptides sit. BPC-157 is 15 amino acids. Its small size does not make it biologically inert. It does not prevent it from binding to receptors, activating signaling pathways, promoting angiogenesis, or generating free radicals. The regulatory threshold was built around manufacturing logic and immunogenicity risk, not around the question of whether a small synthetic peptide injected subcutaneously at pharmacologically active doses can do meaningful — and potentially harmful — things in human tissue. That question has never been systematically answered for most of these compounds. The size cutoff does not answer it. It simply places them in a regulatory bucket that allows compounding. What it does not provide is a safety guarantee.

Pharmacokinetics and Pharmacodynamics: Why Injection Changes Everything

Pharmacokinetics describes what the body does to a drug: how it's absorbed, where it distributes, how it's metabolized, and how it's eliminated. Pharmacodynamics describes what a drug does to the body: how it binds receptors, what signals it triggers, and how long those effects last. Together, they explain why the route of administration isn't a minor detail — it's the difference between a compound that barely reaches systemic circulation and one that floods it.

Absorption. Orally administered peptides have notoriously poor bioavailability. The GI tract is a proteolytic environment — digestive enzymes specifically cleave peptide bonds, breaking proteins into individual amino acids for absorption. Most biologically active peptides are hydrolyzed in the stomach and small intestine before they can be absorbed intact. Those that survive degradation face a second barrier in first-pass hepatic metabolism, where the liver extracts and inactivates a significant fraction of anything absorbed from the gut. This is why insulin cannot be taken as a pill. It is why GLP-1 itself has a half-life of less than two minutes in circulation before DPP-4 enzymes cleave it, which is precisely why pharmaceutical GLP-1 agonists like semaglutide require extensive chemical modification to survive long enough to be therapeutically useful. When a gray-market peptide is sold in capsule form with claims of systemic bioactivity, the pharmacology is simply not compatible with those claims. If it's doing anything, it's not doing it because the peptide reached your bloodstream intact.

Injectable peptides are a completely different story. Subcutaneous injection — the most common route for gray-market peptide use — bypasses the GI tract entirely. A subcutaneously injected compound is absorbed through the lymphatic system and small capillaries in the subcutaneous tissue, enters the systemic circulation largely intact, and achieves bioavailability of 80-100% for most small peptides. Intravenous injection achieves 100% bioavailability by definition. This means that when someone injects BPC-157, a pharmacologically active dose of that compound is reaching tissues throughout the body — not a trace, not a metabolite, but the actual molecule, intact, binding to whatever receptors it has affinity for.

Distribution. After entering systemic circulation, a compound distributes to tissues based on its physicochemical properties — molecular size, charge, lipophilicity, and protein binding. Volume of distribution (Vd) is the pharmacokinetic parameter that describes this: a high Vd means the compound distributes widely into tissues; a low Vd means it stays largely in the plasma. Small peptides often distribute through total body water, reaching tissues including the liver, kidneys, and, in some cases, the brain. Blood-brain barrier penetration is particularly worth flagging: most hydrophilic peptides do not cross the BBB readily because it lacks the transporters needed to shuttle them in. But some smaller, more lipophilic synthetic peptides — including Semax, which was engineered partly for CNS penetration — do cross, with neurological effects that have not been characterized in controlled human studies. When you inject a compound with unknown CNS distribution properties and no human pharmacokinetic data, you do not know whether it is staying where you want it.

Metabolism and half-life. Unmodified peptides are cleared rapidly — peptidases in blood, liver, and tissues cleave them within minutes to hours. Therapeutic peptides are engineered to resist this: semaglutide uses an albumin-binding fatty acid chain to extend its half-life to approximately seven days. GLP-1 itself is modified in tirzepatide using a different fatty acid linker. CJC-1295 uses a drug-affinity complex technology that binds albumin and extends its half-life from minutes to days. These modifications are not cosmetic. They fundamentally change the pharmacological profile — the duration of receptor occupancy, the accumulation with repeated dosing, and the opportunity for off-target interactions. Gray-market CJC-1295 may or may not have these modifications incorporated correctly. More importantly, no human pharmacokinetic study has characterized its actual half-life, tissue accumulation, or metabolite profile in humans. We don't know how long it stays active after injection, what it builds to with weekly dosing, or what its degradation products do.

Receptor pharmacology and downstream effects. This is where the pharmacodynamic picture becomes most clinically significant — and most uncertain. Peptides bind to specific receptor families: G protein-coupled receptors (GPCRs), receptor tyrosine kinases, nuclear receptors, and others. The binding triggers conformational changes that activate intracellular second messengers — cAMP, calcium, MAP kinase cascades, PI3K/Akt pathways — which, in turn, drive gene expression changes, protein synthesis, cellular proliferation, immune modulation, and metabolic shifts. Critically, these effects can persist for hours or days beyond the compound's lifespan, depending on receptor internalization, downstream signaling persistence, and epigenetic modifications at gene promoters.

Off-target receptor binding is the underappreciated risk. Peptides with structural similarity to endogenous signaling molecules can bind to related receptor subtypes with varying degrees of affinity, producing effects that the user did not intend and that the compound's discoverer did not study. Melanotan II is the starkest example: designed as a melanocortin receptor agonist for tanning, it non-selectively activates MC1R (skin pigmentation), MC3R (energy homeostasis), MC4R (sexual arousal, cardiovascular regulation), and MC2R (adrenal steroidogenesis). Users report desired effects — tanning, libido — alongside unexpected ones: nausea, spontaneous erections, cardiovascular changes, and a case series linking it to melanoma and changing nevi through the same MC1R pathway UV radiation uses to drive melanin production and, in susceptible cells, malignant transformation. None of this was characterized in Phase 3 trials because no Phase 3 trial was ever run.

These Are Not Supplements. This Is Why That Matters.

The supplement regulatory framework was built around compounds with a history of safe use in food and traditional medicine — things humans have been ingesting, in various forms, for generations. It presumes a kind of biological familiarity. The FDA does not require pre-market safety or efficacy testing for supplements because the ingredient history is presumed to establish a safety floor. That architecture was not designed for synthetic peptide chains that activate AMPK, stimulate angiogenesis, or non-selectively bind melanocortin receptors throughout the body.

This is the core of the education gap. When a patient decides to inject BPC-157 three times a week, the risk calculation they are running in their head is almost certainly being benchmarked against a supplement — against something like fish oil or a B-vitamin complex, where the downside of an unproven benefit is wasted money and perhaps mild GI discomfort. The appropriate benchmark for a synthetic peptide injected subcutaneously with full systemic bioavailability, receptor binding activity across multiple tissue systems, unknown human pharmacokinetics, and no completed Phase 1 safety trial is an investigational drug being administered outside of an IND framework — without the informed consent requirements, safety monitoring, or adverse event reporting that framework mandates. Those are not the same risk profile. They are not even close.

BPC-157 is a synthetic compound derived from a gastric protein fragment. MOTS-c was identified in 2015 from mitochondrial DNA that was considered junk for decades. These are genuinely novel chemical entities with no history of human use outside of uncontrolled self-experimentation. The argument that they should face no evidentiary requirement because they are "natural" confuses origin story with pharmacology. Botulinum toxin is natural. So is arsenic. The biology doesn't care what narrative we've built around a compound's source. It cares what the compound does to receptors.

Calling BPC-157 a supplement is like calling semaglutide a wellness product. The labeling is convenient. The pharmacology is the same either way — and pharmacology demands the same evidentiary standard regardless of what we decide to call it.

The Evidence Audit, Briefly Restated

If you've been with this series from the start, you already know the BPC-157 evidence profile in detail. For those who haven't: the short version is that in the over thirty years since Body Protection Compound 157 was discovered in human gastric juice, it has accumulated approximately 30 human subjects across all published research — including a 12-patient retrospective case series with no control group. A Phase 1 trial registered in 2015 (NCT02637284, 42 subjects) was canceled in 2016 with no published results and no public explanation. Zero randomized controlled trials of any kind have ever been completed. A 2025 systematic review in the HSS Journal identified 36 relevant studies — 35 preclinical and 1 clinical.

The most-hyped compound in the gray-market peptide boom has been studied in human beings in numbers that would not meet the evidentiary bar for a conference poster presentation. The hype-to-evidence gap on BPC-157 is not a gap waiting to be filled by future research. Major pharmaceutical companies have had access to this compound for decades and have not pursued it. The single-initiated trial was canceled. The absence is a verdict, rendered repeatedly by every party with the resources to run a trial. As I wrote in Part II: the angiogenic mechanisms that make it exciting in rodents also carry a theoretical pro-oncogenic implication that has never been evaluated in human safety studies. And it is now being self-injected by hundreds of thousands of Americans.

That evidence profile is not the subject of this piece — we've covered it in detail. What I want to sit with here is a different and more unsettling question: Patel's patient knew some version of that evidence. She had done research. She had read about BPC-157. She had considered her options. And she still stopped a medication backed by 170,000 people in controlled trials and replaced it with one backed by thirty. That decision is not a knowledge failure. It is something else entirely.

The Comparison That Should Stop Us Cold

While Patel's patient was discontinuing her statin, an extraordinary piece of evidence was published that she — and most of her peers — likely never heard about. The Cholesterol Treatment Trialists' Collaboration conducted an individual-patient meta-analysis drawing on 19 double-blind randomized trials — 123,940 participants followed for a median of 4.5 years — and systematically evaluated every adverse event listed on statin product labels.

The drug she stopped? Studied in over 170,000 people across cardiovascular outcomes trials, showing a 25% reduction in major adverse cardiovascular events and meaningful reductions in all-cause mortality. The adherence literature in older adults consistently shows that poor statin adherence is associated with significantly higher cardiovascular mortality — the direction and magnitude of that effect, across multiple meta-analyses, is not ambiguous.

The compound she replaced it with? Approximately thirty human subjects across all published research. Zero RCTs. Thirty-five of thirty-six relevant studies preclinical.

Patel put it precisely: this is not a fringe case of a misinformed patient. It is the defining epistemological feature of our current moment in consumer health — the volume of evidence behind a therapy has become inversely correlated with public trust in it. The more we know, the less patients believe us. That observation deserves to sit with us for a moment before we continue.

The Science Is Real. That's the Problem.

I want to be direct about something that gets lost when this debate becomes tribal: the underlying peptide science is genuinely exciting. The mitochondrial peptide revolution — MOTS-c, humanin, SHLP, the compounds Cohen's lab has been naming in Yiddish for a decade — represents a pharmacological frontier that could redefine how we approach metabolic disease, sarcopenia, and aging. As I covered in Part II, MOTS-c has no published human trials of any phase; it is being prescribed for longevity and metabolic health based entirely on mouse data. That fact coexists with the legitimate possibility that it will eventually prove meaningful in humans. The preclinical signal is real.

Pinchas Cohen, the USC gerontologist whose lab identified many of these compounds, told The New Yorker he considers what's happening in the gray market an appropriation of his life's work. "People should not be taking them until they're fully tested," he said. "The history of shortcuts is not a happy one." He knows better than anyone what the science might eventually deliver. His concern is not that it won't deliver — it's that injecting gray-market fragments of unstable, unverified compounds is not a path toward finding out.

The GLP-1 receptor agonist story is the proof of concept for what the legitimate pipeline looks like. Insulin is a peptide. Semaglutide is a peptide. Tirzepatide — which achieves 20-22% body weight reduction in trials, which I have written about and taken myself — is a peptide. These compounds took decades of human trials to reach clinical use. They are now changing the trajectory of metabolic disease. That is the template. What the gray market is doing is not an impatient shortcut toward the same destination. It is a different road, with no map, no safety data, and a vendor who may be sending you lead contamination in the vial alongside whatever the active compound is supposed to be.

What 2026 Has Added to This Story

Since we published Parts I and II, the regulatory and political environment has moved. In February, RFK Jr. announced on Joe Rogan's podcast that approximately 14 of 17 Category 2 peptides would be reclassified — a procedural access decision that, as I argued then, does not change the evidence base by a single data point. That reclassification is now being pressed further. In March 2026, the FDA held a public meeting — requested by the supplement industry's Natural Products Association — to consider broadening the legal definition of dietary ingredients to include peptides and other druglike compounds that have never been part of the food supply. The FDA's own top food official opened the session by emphasizing the administration's commitment to "cutting red tape."

Consumer advocates at the same meeting stated the obvious: loosening the definition will not produce a better-evidenced market. It will produce a larger unregulated one. The practical consequence — already visible — is that gray-market peptide imports from China roughly doubled in 2025. A journalist who ordered several compounds from SwissChems found lead contamination in BPC-157, endotoxins in TB-500, and less than 42% of the advertised dose in CJC-1295. These are not edge cases of a mostly functional system. They are the predictable output of a market operating without manufacturing standards or enforcement.

At a Las Vegas biohacking conference last year, two women received peptide injections at a conference booth and ended up on ventilators. The peptides involved have not been disclosed. The physician who administered them cited an AI-generated report in his defense. As I noted in Part II, the tragedy here is structural: there is no mandatory adverse event reporting, no pharmacovigilance infrastructure, no mechanism to find out when something goes wrong at scale. Reclassification accelerates access without building any of those systems.

We may be approaching a regulatory environment in which a compound with thirty published human subjects can be legally sold in gummies, alongside creatine and fish oil, with no evidentiary requirement and no manufacturing standard enforced. That is not a hypothetical. It is what the supplement industry explicitly requested at the March 2026 FDA meeting — and the agency's opening framing suggested receptiveness to the argument.

The Deeper Problem: A Trust Crisis We Helped Build

The wellness industry's standard defense — the absence of evidence is not evidence of absence — is technically accurate and practically misleading. BPC-157 has existed since 1992. The absence of human trial data is not a gap waiting to be filled; it is a verdict that has been rendered repeatedly by every party capable of running a trial. Major pharmaceutical companies have had access to this compound and have not pursued it. Academic medical centers have not found the preclinical signal compelling enough to fund human work. The single initiated trial was canceled.

But I also need to say something more uncomfortable: patients like this have not arrived at this moment irrationally. They have learned, correctly, that pharmaceutical companies have financial incentives that sometimes distort their messaging. They lived through the opioid crisis. They watched institutions fail them during COVID — on messaging, on trust, on nuance. The skepticism is earned. What has gone wrong is the migration of trust from one set of financially motivated actors to another. The peptide clinic charging $400 per vial for a compound with approximately thirty published human subjects has identical financial incentives to the pharmaceutical company it is positioned as an alternative to. The difference is that the pharmaceutical company had to prove its product works before selling it.

This is the logic trap of our moment: skepticism of institutional medicine has been weaponized not to produce better medicine, but to produce less evidence-constrained commerce. "Medical freedom" in the absence of safety standards is not freedom. It is, as one physician-journalist put it, medical anarchy. The result is a swirl of competing claims that no individual patient has the tools to evaluate — which is precisely the condition the regulatory framework exists to prevent.

The Conversation I Am Now Having

I am not going to win the peptide argument with data alone. I know that. And I am not interested in dismissing the patients asking about these compounds — many of them are intelligent, motivated people who have been underserved by conventional medicine's dismissiveness and its historical failures to adequately address chronic disease, obesity, pain, and aging. Their frustration is legitimate. Their search for solutions is legitimate.

What I can offer is this: I understand why the system has failed your trust. I share some of those concerns — I have written about the pharmaceutical industry's incentive structures and medicine's historical failures at length. I am not here to defend institutional medicine uncritically. I am here to tell you that the compound you are injecting three times a week has less evidence behind it than any medication in your medicine cabinet, including ibuprofen and antihistamines. The medication you stopped for side effects that a 123,940-person trial just confirmed are not pharmacologically plausible has more evidence behind it than almost anything I could prescribe you. Those two facts can coexist with real institutional critique and real pharmaceutical skepticism — they do not require us to pretend evidence doesn't matter.

Peptides are not the enemy of evidence-based medicine. They are among its most promising frontiers. Insulin proved that. GLP-1 agonists proved it again. The junk-DNA revolution in mitochondrial peptide biology may prove it a third time — through trials, not podcasts.

In the meantime, the clinical work is to stay in the room with patients who are frustrated, curious, and increasingly skeptical of us — and to say clearly: here is what we know, here is what we don't, and here is the difference. That difference is not a technicality. It is how medicine protects the people it serves.

REFERENCES

1. Reith C, et al. (Cholesterol Treatment Trialists' Collaboration). Assessment of adverse effects attributed to statin therapy in product labels: a meta-analysis of double-blind randomised controlled trials. The Lancet. 2026. doi:10.1016/S0140-6736(25)01578-8

2. Cholesterol Treatment Trialists' Collaboration. Efficacy and safety of more intensive lowering of LDL cholesterol: a meta-analysis of data from 170,000 participants in 26 randomised trials. The Lancet. 2010;376(9753):1670–1681. doi:10.1016/S0140-6736(10)61350-5

3. Wood FA, Howard JP, Finegold JA, et al. N-of-1 Trial of a Statin, Placebo, or No Treatment to Assess Side Effects (SAMSON). N Engl J Med. 2020;383:2182–2184. doi:10.1056/NEJMc2031173

4. Ofori-Asenso R, Ilomäki J, Tacey M, et al. Adherence and Persistence Among Statin Users Aged 65 Years and Over: A Systematic Review and Meta-analysis. J Gerontol A Biol Sci Med Sci. 2018;73(6):813-819. doi:10.1093/gerona/glx169

5. BPC-157 Phase I safety trial (Targeting Ulcerative Colitis). ClinicalTrials.gov identifier NCT02637284. Registered 2015; cancelled / unpublished.

6. Staresinic M, et al. Stable Gastric Pentadecapeptide BPC 157 and Striated, Smooth, and Heart Muscle. Surgeon. 2021. PubMed PMID 34324435.

7. FDA. Certain bulk drug substances for use in compounding may present significant safety risks — Category 2. FDA.gov. Updated 2023.

8. Lee C, Yen K, Cohen P. Humanin: a harbinger of mitochondrial-derived peptides? Trends Endocrinol Metab. 2013;24(5):222–228. doi:10.1016/j.tem.2013.01.005

9. Lee C, Kim KH, Cohen P. MOTS-c: A novel mitochondrial-derived peptide regulating muscle and fat metabolism. Free Radic Biol Med. 2016;100:182–187. doi:10.1016/j.freeradbiomed.2016.05.015

10. Khullar D. Why are people injecting themselves with peptides? The New Yorker. 2026 Apr 13. Available from: https://www.newyorker.com/magazine/2026/04/13/why-are-people-injecting-themselves-with-peptide

11. Patel V. My patient would rather take a peptide than a statin. That reveals an uncomfortable truth in medicine. STAT News, First Opinion. April 3, 2026.

12. Perrone M. Makers of dietary supplements push the FDA to allow peptides and other new ingredients. Associated Press / PBS NewsHour. March 29, 2026.

13. Jastreboff AM, Aronne LJ, Ahmad NN, et al. Tirzepatide once weekly for the treatment of obesity. N Engl J Med. 2022;387:205–216. doi:10.1056/NEJMoa2206038

14. Biologics Price Competition and Innovation Act of 2009 (BPCIA). Pub. L. No. 111-148, §7001–7003 (2010).

15. FDA. Guidance for Industry: Questions and Answers on Biosimilar Development and the BPCI Act. FDA.gov. Updated 2022.

16. Lau JL, Dunn MK. Therapeutic peptides: Historical perspectives, current development trends, and future directions. Bioorg Med Chem. 2018;26(10):2700–2707. doi:10.1016/j.bmc.2017.06.052

17. Sato AK, Viswanathan M, Kent RB, Wood CR. Therapeutic peptides: technological advances driving peptides into development. Curr Opin Biotechnol. 2006;17(6):638–642. doi:10.1016/j.copbio.2006.10.002