Chapter 29: The Peptide Drugs Already in Your Pharmacy

"The most profound technologies are those that disappear. They weave themselves into the fabric of everyday life until they are indistinguishable from it." — Mark Weiser, The Computer for the 21st Century (1991)

Overview

Before you read another word, do this. Out loud if you are alone, silently if you are not: name three peptide drugs.

Most readers get two. Insulin, the most famous medicine of the twentieth century. Then semaglutide, or tirzepatide, or "Ozempic" — the most famous medicine of this one. Then the list stops. Readers who have made it to Chapter 29 will do better than average and may reach for oxytocin or octreotide. Almost nobody, cold, says teriparatide. Or desmopressin, leuprolide, vancomycin, cyclosporine, linaclotide, cosyntropin, vasopressin.

Which is remarkable, because more than eighty peptide drugs are approved and in routine daily use in major jurisdictions. They are not fringe. They are not experimental. Most are not new — several are older than the people taking them. They are dispensed by ordinary pharmacies, stocked in ordinary hospital carts, and available as ordinary generics. One of them made organ transplantation practical. One is on the World Health Organization's list of essential medicines. One is the reason a parent of a diabetic child can leave the house without panicking.

This is the quiet chapter. No scandal, no frontier, no molecule anybody is selling you on Instagram. It is an inventory — and the inventory does two things.

First, it demolishes the idea that peptide medicine is speculative. When someone calls peptides "an emerging class," they are describing a class that has been shipping approved products for five decades. Second — the harder point, and the one §29.10 is built around — the inventory has a shape. Look at what all eighty-plus of these drugs actually do and you find four jobs: replace a missing hormone, block an overactive receptor, act locally without being absorbed, or ask a diagnostic question. What you will not find anywhere on the list is a single approved peptide whose job is to broadly support, optimize, or modulate a system that is working fine.

That absence is not an accident, and by the end of this chapter it will be one of the most useful priors you own.

In this chapter, you will learn to:

• Name ten or more approved peptide drugs you could not have named an hour ago
• Explain why chronically high parathyroid hormone dissolves bone while once-daily exposure builds it
• Show how one selectivity change turned a vasopressin analog into three different medicines
• Describe honestly what happened to calcitonin, and why it is a story about science working
• Identify the approved peptides that nobody, including most physicians, calls peptides
• Explain why a peptide designed not to be absorbed is a triumph rather than a failure
• Read the whole pharmacopeia as a single argument about what peptides are good for

Learning Paths

All five paths should read §29.8 and §29.10. Those two carry the chapter's argument, and §29.10 is a load-bearing prior for everything in Part VI.

💊 GLP-1 — §29.6 (glucagon) is insulin's mirror image and completes Chapter 11's picture; §29.2 is the cleanest illustration in the book that how often a peptide arrives can matter more than how much. 🏋️ Performance — read §29.10 twice. The absence it identifies is the strongest available argument against the "peptides for optimization" framing, and it is built entirely out of the pharmaceutical industry's own approvals. 🔬 Science — §29.3 and §29.8 are the chemistry. Receptor subtype selectivity and cyclization/N-methylation are the two engineering moves Chapter 33 formalizes. 💄 Cosmetic — §29.4 is your section: what an honest downgrade looks like when the industry doing the downgrading is a regulated one, and the benchmark against which Chapter 30's claims should be measured. 🏥 Clinical — §29.5 and §29.7 will be familiar; read them for the reframing. You have ordered cosyntropin and hung vasopressin without ever thinking "peptide," and that habit is what this chapter is trying to interrupt.

29.1 Name three peptide drugs

The exercise at the top of this chapter is a diagnostic, and what it diagnoses is a distortion in how the category is stored in most people's heads.

The word peptide has, over the past decade, been captured by two populations at opposite ends of the credibility spectrum. At one end, the metabolic blockbusters — semaglutide, tirzepatide, their relatives — so culturally enormous they have become what "peptide drug" means. At the other, the gray market: BPC-157, TB-500, the peptide clinic email, the vial with a research-use label. Both are loud. Between them lies a very large, very quiet middle, and that is where most approved peptide medicine lives.

Consider what a mid-sized hospital pharmacy stocks on an ordinary Tuesday. Vancomycin, in the antibiotic room. Vasopressin, in the critical care fridge. Oxytocin, on labor and delivery. Octreotide, for variceal bleeding and the neuroendocrine tumor patients. Cosyntropin, for the morning cortisol workup. Desmopressin, in three formulations for three different reasons. Leuprolide, for prostate cancer and endometriosis and central precocious puberty. Glucagon, in the code cart. Cyclosporine on the transplant service. Bacitracin in the ointment drawer. Insulin everywhere.

Every one of those is a peptide. Not "peptide-like," not "peptide-derived" — a chain of amino acids joined by peptide bonds, exactly as Chapter 1 described. And in a typical clinical day none will be called a peptide by anybody, because clinicians organize drugs by what they do, not by what they are made of. An antibiotic is an antibiotic. A vasopressor is a vasopressor. Nobody says "hang the peptide."

That is a perfectly reasonable way to run a hospital and a terrible way to reason about a category.

HOW THE CATEGORY IS ACTUALLY STORED IN MOST HEADS

   LOUD                          SILENT                          LOUD
   ┌──────────────┐   ┌────────────────────────────┐   ┌──────────────────┐
   │  BLOCKBUSTERS│   │      THE QUIET MIDDLE      │   │   GRAY MARKET    │
   │              │   │                            │   │                  │
   │ insulin      │   │ teriparatide  desmopressin │   │ BPC-157          │
   │ semaglutide  │   │ octreotide    leuprolide   │   │ TB-500           │
   │ tirzepatide  │   │ vancomycin    cyclosporine │   │ CJC-1295         │
   │              │   │ vasopressin   terlipressin │   │ "peptide therapy"│
   │              │   │ glucagon      cosyntropin  │   │                  │
   │              │   │ icatibant     linaclotide  │   │                  │
   │              │   │ calcitonin    secretin     │   │                  │
   │              │   │ bacitracin    polymyxins   │   │                  │
   │              │   │ daptomycin    oxytocin     │   │                  │
   │  ~5 molecules│   │      80+ molecules         │   │  a dozen names   │
   └──────────────┘   └────────────────────────────┘   └──────────────────┘
        ↑                          ↑                            ↑
   what people      what the category actually IS      what people
   think of as                                          worry about
   "peptide drugs"

That diagram is the whole chapter in one picture. The middle column is most of approved peptide medicine and almost none of the public conversation about peptides. Everything that follows moves a dozen molecules from the middle column into your head, on the theory that a category you can inventory is a category you can reason about.

A note on the count. You will see the number of approved peptide drugs given with false precision. Treat all such figures as soft. The number depends entirely on where you draw the peptide/protein boundary (Chapter 1 §1.5), and that boundary is a convention rather than a chemical fact. Count insulin analogs individually or as a family and the total moves by several. Include recombinant proteins under 100 residues, or every glycopeptide and lipopeptide antibiotic, and it jumps. Count only synthetic peptides under 50 residues with a published sequence and it falls. This is why this book says more than eighty: the honest claim is about magnitude. Anyone who gives you an exact number without stating their inclusion rule has not thought about the question.

🔍 Check Your Understanding

1. Before this chapter, how many approved peptide drugs could you name? Write the number down; you will want it in §29.10.
2. Why do exact counts vary so widely between sources, and what does that variation tell you about the word peptide itself?
3. A clinician says "I don't really use peptide drugs." Name three they almost certainly used this month.

29.2 Teriparatide and abaloparatide: the drugs that should not work

This is the best story in the chapter, and worth reading slowly.

Parathyroid hormone (PTH) is the body's principal regulator of blood calcium. When calcium falls, the parathyroid glands release PTH, and PTH raises calcium three ways: more renal reabsorption, more vitamin D activation, and — the relevant one — mobilization of calcium out of bone.

That third route is exactly as destructive as it sounds. In primary hyperparathyroidism, where an adenoma pumps out PTH continuously, the classic picture includes bone loss, fragility, and in advanced untreated disease the skeletal lesions older textbooks described with grim vividness. Chronically elevated PTH takes bone apart. This has been understood for the better part of a century.

So: a drug company decides to treat osteoporosis — a disease of insufficient bone — by injecting patients with parathyroid hormone.

On its face this is one of the worst ideas in the history of endocrinology, the therapeutic equivalent of treating dehydration with a diuretic. And it works. Teriparatide, which is PTH(1–34), the first thirty-four amino acids of human parathyroid hormone, is an approved anabolic bone agent that increases bone mineral density and reduces fracture risk in severe osteoporosis.

The entire difference is temporal.

🧬 The Molecule — teriparatide is a fragment, and the fragment is the active end

Full-length human parathyroid hormone is 84 amino acids. Teriparatide is residues 1 through 34.

Not an arbitrary truncation. The N-terminal region contains everything required to bind and activate the PTH type 1 receptor; the rest contributes to other functions and to clearance but not to receptor activation. Cut at residue 34 and you retain full agonist activity at roughly 40% of the mass — which, per Chapter 1 §1.6, moves the molecule into the synthesizable peptide range rather than the recombinant-protein range.

The name tells you this if you can read it. "Teri-para-tide": the -tide stem marks a peptide (Chapter 1 §1.8) and the middle syllables point at the parent hormone.

Abaloparatide is related but distinct: an analog of parathyroid hormone-related protein, PTHrP(1–34), with substitutions intended to alter its receptor-binding behavior. PTHrP is a separate gene product that happens to act at the same receptor — a reminder from Chapter 2 that receptors are promiscuous about ligands in ways evolution finds useful and chemists find exploitable.

Same hormone, opposite effect, and the only variable is the clock

Chapter 3 §3.5 introduced pulsatility: many peptide hormones carry information in their temporal pattern, and a flat, sustained exposure to a hormone the body normally releases in pulses is not a stronger version of the physiological signal but a categorically different one. Teriparatide is that principle producing an approved drug.

THE SAME HORMONE, TWO TEMPORAL PATTERNS, OPPOSITE SKELETAL OUTCOMES

  CONTINUOUS EXPOSURE (hyperparathyroidism, or a continuous infusion)

  PTH  ████████████████████████████████████████████████████████  sustained
       └──────────────────── days to years ────────────────────┘

       net effect on bone:  RESORPTION  →  bone is disassembled


  INTERMITTENT EXPOSURE (once-daily injection)

  PTH   ▲         ▲         ▲         ▲         ▲         ▲
       ╱ ╲       ╱ ╲       ╱ ╲       ╱ ╲       ╱ ╲       ╱ ╲       spike,
      ╱   ╲_____╱   ╲_____╱   ╲_____╱   ╲_____╱   ╲_____╱   ╲___   then gone
       └─ each pulse rises and clears within hours ─┘

       net effect on bone:  FORMATION  →  bone mineral density INCREASES


  THE VARIABLE THAT CHANGED:  not the molecule
                              not the receptor
                              → the SHAPE OF THE EXPOSURE OVER TIME

The mechanistic account — and the mechanism is genuinely not settled — is roughly that a brief PTH pulse favors the bone-forming side of the remodeling cycle, while sustained exposure lets the bone-resorbing side dominate. Osteoblasts and osteoclasts respond on different timescales.

What is not in doubt is the observation. Continuous PTH removes bone; intermittent PTH adds it. The molecule is identical. The receptor is identical. The clock is the only thing that changed.

Why this matters beyond osteoporosis. Everywhere else in this book, when someone proposes that a peptide's benefits follow from getting more of it into the body, Chapter 3's answer is that physiology may not work that way. Teriparatide is the constructive version of that argument: temporal pattern is not merely a caveat to worry about but a design parameter that can be engineered. Somebody looked at a hormone with a famously destructive effect on bone, noticed the destruction required continuity, and built a therapy out of the discontinuity.

Anabolic versus antiresorptive

That distinction explains teriparatide's clinical position. It is not first-line for the average patient with mild density loss. It is for people whose skeletons are in serious trouble. It is given by daily injection, it is expensive relative to a generic bisphosphonate, and it carries duration-of-use considerations most osteoporosis drugs do not.

📊 Evidence Rating

Claim: Teriparatide, given by daily subcutaneous injection, increases bone mineral density and reduces vertebral fracture risk in postmenopausal women with severe osteoporosis or high assessed fracture risk. Rating: ✅ Strong clinical evidence Reason: Randomized, placebo-controlled fracture-endpoint data supported approval in the early 2000s; continuous clinical use since, an extensive post-marketing record, and inclusion in major osteoporosis guidelines as an anabolic option for high-risk patients. What would change it: Fracture-endpoint data showing no advantage over available antiresorptive therapy in the high-risk populations where it is positioned, or a durable safety signal reversing the benefit-risk balance. (Current as of this writing; a mature evidence base, unlikely to move.)

The osteosarcoma warning, and a translation problem you have already met

In preclinical toxicology, rats given teriparatide for most of their lifespan at high exposures developed osteosarcoma — a bone cancer — at rates well above control. That was serious enough that on approval the drug carried a boxed warning, plus recommendations limiting cumulative duration of use.

Then a long, unglamorous thing happened: the drug was used in enormous numbers of patients for roughly two decades while surveillance systems watched for human osteosarcoma cases. The accumulated post-marketing data did not show the expected excess. On that evidence, the boxed warning was subsequently removed and the labeling revised, with duration-of-use language relaxed.

If that sequence feels familiar, it should. It is structurally the same as Chapter 8's thyroid C-cell story, where rodent studies of GLP-1 receptor agonists produced a tumor finding that drove a boxed warning and where the central question has been whether rodent C-cell biology translates to human thyroid biology at all. Same shape: a real animal finding, a species with different relevant biology, a precautionary label, and years of human data slowly answering the question the animal study could not.

🩺 Safety and Risk — what the teriparatide labeling history should and should not teach you

What it legitimately teaches: animal carcinogenicity findings are hypotheses about humans, not conclusions about them. Rodent skeletal biology differs from human — including lifelong growth plate activity in rats — in ways that make a bone-tumor finding hard to extrapolate. Regulators responded appropriately in both directions: warn while uncertain, revise when the human data arrived.

What it absolutely does not teach: that animal safety signals are generally overblown, or that a boxed warning is a formality that eventually gets removed. Plenty of animal findings have translated. This warning came off because two decades of human surveillance addressed the specific question — not because someone decided rats do not matter. The label followed the evidence, which is a narrower and far more useful lesson than "warnings are noise."

What it does not touch at all: whether teriparatide is appropriate for any given person. Those are Chapter 39 conversations, not book conversations.

29.3 Desmopressin: one selectivity change, three unrelated medicines

Chapter 1 §1.7 made a point waiting to be cashed in: oxytocin and vasopressin are both nine amino acids long, differ at two positions, and do substantially different things. Desmopressin is what happens when a chemist takes that seriously and goes looking for the next increment of resolution.

Vasopressin — also called antidiuretic hormone — acts at more than one receptor subtype, and the subtypes do different jobs.

• V1 receptors, on vascular smooth muscle, mediate vasoconstriction.
• V2 receptors, in the kidney's collecting ducts, mediate water reabsorption, concentrating urine and retaining free water. This is the "antidiuretic" part.

Native vasopressin hits both. For a patient whose problem is that they produce no antidiuretic hormone and therefore pass enormous volumes of dilute urine, the V1 activity is not a feature but an unwanted side effect. Native vasopressin also has a very short half-life, which for a condition requiring continuous coverage is a second problem.

Desmopressin was engineered against exactly those two liabilities. Two modifications — removal of the amino group at position 1, and substitution of the D-isomer of arginine at position 8 for the L-isomer — produced a molecule that strongly favors V2 over V1 and lasts substantially longer than the parent hormone. The D-amino acid does double duty: it contributes to the selectivity shift and makes the molecule a poorer substrate for the peptidases that clear the natural hormone, the resistance strategy Chapter 4 §4.3 described and Chapter 33 formalizes.

Two modifications, two problems solved. And then something the chemists were not necessarily aiming at: three approved indications, in three unrelated fields of medicine.

ONE MOLECULE, THREE INDICATIONS, ONE RECEPTOR

                        DESMOPRESSIN
                    (V2-selective, long-acting
                     vasopressin analog)
                             │
                             ▼
                     V2 RECEPTOR ACTIVATION
                             │
              ┌──────────────┴───────────────┐
              ▼                              ▼
     RENAL COLLECTING DUCT            VASCULAR ENDOTHELIUM
     water reabsorption ↑             releases stored von Willebrand
     urine concentrated               factor and factor VIII
              │                              │
      ┌───────┴────────┐                     ▼
      ▼                ▼            certain bleeding disorders
  central          primary          (hemostatic support)
  diabetes         nocturnal
  insipidus        enuresis

   ENDOCRINOLOGY   ·   PEDIATRICS   ·   HEMATOLOGY
   Three specialties. Three clinics. One receptor.

Central diabetes insipidus — increasingly called arginine vasopressin deficiency — is the straightforward one. The pituitary is not releasing enough antidiuretic hormone, so the kidney cannot concentrate urine, so the patient passes liters of dilute urine and is perpetually thirsty. Desmopressin replaces the missing signal: hormone replacement in its purest form, and about as close to a solved problem as endocrinology offers.

Primary nocturnal enuresis — bedwetting in children — is the one that surprises people. The rationale is partly that some affected children have a blunted overnight rise in vasopressin and therefore produce more urine overnight than their bladder can hold. It is an approved indication with genuine randomized evidence behind it, sitting alongside behavioral approaches such as enuresis alarms.

Certain bleeding disorders looks like it belongs to a different molecule entirely. V2 stimulation does something outside the kidney: it triggers vascular endothelial cells to release their stored pools of von Willebrand factor and factor VIII, both central to clot formation. In patients with mild hemophilia A or type 1 von Willebrand disease, where the problem is insufficient rather than absent factor, mobilizing those stores can raise circulating levels enough to support hemostasis.

Notice what did not happen. Nobody designed three drugs. Somebody designed one molecule with better receptor subtype selectivity, and the V2 receptor turned out to have a job in two organ systems. This is the engineering logic of Chapter 33 in miniature: receptor subtype selectivity is not a refinement, it is a lever. It converts a hormone with a broad physiological footprint into a tool acting on one branch of it.

📊 Evidence Rating — two claims, one molecule (rule 6 in action)

Claim A: Desmopressin corrects polyuria and restores urinary concentrating ability in patients with central diabetes insipidus. Rating: ✅ Strong clinical evidence Reason: Direct physiological replacement of a deficient hormone with an unambiguous, immediately measurable endpoint (urine output and osmolality); decades of use, approved across major jurisdictions, standard of care with no serious competitor. What would change it: Essentially nothing short of a fundamental error in our understanding of renal water handling. Among the most secure ratings in the book.

Claim B: Desmopressin reduces the number of wet nights in children with primary nocturnal enuresis. Rating: ✅ Strong clinical evidence Reason: Randomized controlled trial evidence supports a reduction in wet nights during treatment; the indication is approved and long established in pediatric practice. What would change it: This rating is deliberately narrow — it covers reduction in wet nights during treatment, not cure and not durable benefit after stopping, for which the evidence is weaker and relapse on discontinuation is well recognized. A trial showing no separation from placebo during treatment would change it; a trial showing high relapse would not, because that is already inside the claim's boundary.

Why two ratings and not one. Same molecule, same receptor, same mechanism — different population, endpoint, evidence base, and durability. A source that told you "desmopressin: ✅" would have compressed away the fact that one of these is a permanent physiological correction and the other a symptomatic treatment that frequently stops working when stopped.

The risk that matters here is hyponatremia. A drug whose job is to make you retain free water can make you retain too much, diluting blood sodium. Severe hyponatremia is dangerous and can cause seizures. That is why fluid intake is a genuine clinical issue with these agents, why some formulations carry prominent warnings, and why sodium monitoring is part of how the drug is used. That is a statement of the risk, not a management plan — the plan belongs to a clinician who knows the patient.

29.4 Calcitonin: what it looks like when a drug class honestly shrinks

Almost every story about a medicine losing ground is a scandal story. Data were hidden. A regulator was captured. A company knew and did not say. Those stories are real, and they are disproportionately the ones that get written.

This is not one of those. Calcitonin's clinical role contracted substantially because the evidence base improved and better alternatives arrived. Nobody hid anything. Nobody was punished. The scientific process worked more or less as advertised, slowly and boringly, and a widely used drug moved to the margins. It is the clearest example in this book of the healthy version of that process, and it almost never gets told because there is no villain in it.

The hormone

Calcitonin is a 32-amino-acid peptide hormone produced by the thyroid C cells — the same cells at the center of Chapter 8's rodent tumor discussion, a coincidence worth noticing. Its classical role is to lower blood calcium, largely by inhibiting osteoclasts. In the standard endocrine diagram it is parathyroid hormone's opposite number: PTH raises calcium, calcitonin lowers it.

The reasoning that follows is almost irresistible. Osteoporosis is a disease of excessive bone resorption. Calcitonin inhibits bone resorption. Therefore calcitonin should treat osteoporosis.

The therapeutic agent is usually salmon calcitonin, not the human sequence, because salmon calcitonin is substantially more potent at the human calcitonin receptor than human calcitonin is — a real and slightly funny piece of comparative endocrinology. It was available injectably and, notably, as a nasal spray, one of the earlier successful non-injectable peptide formulations. Calcitonin was used for postmenopausal osteoporosis, for hypercalcemia of malignancy, and for Paget's disease. In its heyday it was a familiar prescription.

What went wrong, and it is not what you think

Nothing went wrong, exactly. Several unrelated things went better elsewhere, and one thing turned out weaker than hoped.

Better competitors arrived. Bisphosphonates became widely available, inexpensive, and supported by large fracture-endpoint trials; later, denosumab and the anabolic agents from §29.2 added options. In a field with several well-evidenced choices, a drug with a weaker evidence base does not need to be disproven to lose its place. It only needs to be outperformed.

The fracture evidence never got strong. Calcitonin reliably does things that are easy to measure — it inhibits osteoclast activity, moves bone turnover markers, and has a genuine analgesic effect in acute vertebral fracture that some clinicians still value. What it never convincingly demonstrated was a robust reduction in fracture risk comparable to the bisphosphonate trials.

🔬 Read the Study — a dose-response problem

The fracture trial most often cited for nasal salmon calcitonin in postmenopausal osteoporosis is the one known by the acronym PROOF (Prevent Recurrence Of Osteoporotic Fractures), a multi-year randomized placebo-controlled study with several active dose arms.

What it reported: a statistically significant reduction in new vertebral fractures in one dose arm compared with placebo.

Why that did not settle the question. The result was not consistent across dose arms. The middle dose separated from placebo; the highest dose tested did not, or did so less clearly.

A real drug effect usually shows a dose-response relationship. More drug should produce more effect, at least to a plateau or a toxicity ceiling. When a middle dose beats placebo and a higher dose does not, one of a few things is true: an unusual biological explanation exists (possible, but it must be argued rather than assumed); substantial dropout differed across arms and distorted the comparison (PROOF had considerable attrition, widely discussed); or the positive arm is a chance finding among several comparisons.

The general lesson: in a trial with multiple dose arms, do not just look for the arm with the p-value. Look at whether the arms line up. A tidy dose-response is one of the strongest internal signals that an effect is real; a scattered one is a reason to want replication before believing it — and replication at the necessary scale never came for calcitonin.

Regulatory reviews raised questions about long-term use. In the early 2010s, regulators on both sides of the Atlantic revisited calcitonin. A European review concluded the benefit-risk balance did not favor use for osteoporosis, citing a small increase in cancer diagnoses observed with long-term use across pooled trial data, and restricted the approved uses accordingly. In the United States, an advisory committee reviewing the osteoporosis indication expressed substantial doubt about it. Details differ by jurisdiction and formulation; the direction of travel has been consistent.

Put those three together and you have a contraction. Not a withdrawal, not a scandal — a contraction. Calcitonin remains available and retains a role in hypercalcemia, in Paget's disease, and in acute vertebral fracture pain. But the broad osteoporosis role it once had is gone, for the most respectable reason a treatment can lose ground: we found out more, and what we found out did not support the original enthusiasm at the original scale.

📊 Evidence Rating

Claim: Calcitonin (salmon calcitonin, nasal or injectable) meaningfully reduces osteoporotic fracture risk in postmenopausal women. Rating: ⚠️ → ❌ (a downgrade over time; ❌ as of this writing) Reason: The single supportive fracture-endpoint trial showed an inconsistent dose-response and substantial attrition; no adequately powered replication established the effect; comparative evidence favors bisphosphonates and other agents; regulatory reviews concluded the benefit-risk balance did not support the osteoporosis indication. The ❌ describes the evidence for this claim — it is not a statement that the molecule does nothing. What would change it: An adequately powered, well-retained randomized fracture-endpoint trial showing a consistent dose-response effect. Nobody is likely to run it, which is part of why this rating will probably stay put.

Ratings this does NOT cover: calcitonin for hypercalcemia (different claim, population, and endpoint, and better supported for short-term calcium lowering); for Paget's disease; for acute vertebral fracture pain. Rule 6 — the ❌ above travels with its claim, not with the molecule.

Why this section is in the book. Because you will be told, repeatedly and by people with something to sell, that mainstream medicine never revises itself. Calcitonin is a counterexample in plain sight — a treatment mainstream medicine promoted, studied further, and then substantially demoted on its own accumulating data, with no external pressure and no reputational payoff for anyone. That is what a functioning evidentiary process looks like from inside. It is undramatic, it takes decades, and it is the thing the peptide gray market conspicuously lacks: a mechanism for a claim to lose.

29.5 Vasopressin and terlipressin: the peptides in the critical care fridge

Return to vasopressin, but consider the other receptor. In §29.3 the V1 vasoconstrictor activity was a nuisance to be engineered away. In the intensive care unit it is the entire point.

Vasodilatory shock — most commonly septic shock — is a state in which blood vessels lose tone. Cardiac output may be adequate or high, but the vasculature is so dilated that perfusion pressure collapses. The standard response is catecholamine vasopressors, principally norepinephrine. But some patients require escalating doses with diminishing returns, and high catecholamine exposure has its own costs.

Vasopressin acts through an entirely different receptor system. Its V1-mediated vasoconstriction does not depend on adrenergic signaling, so it can add vascular tone through a parallel pathway. There is also a physiological rationale: endogenous vasopressin levels are often inappropriately low in prolonged vasodilatory shock.

Randomized trials — including the large multicenter trial known by the acronym VASST and the later VANISH trial — examined vasopressin's role alongside or in place of norepinephrine in septic shock. The picture that emerged is worth describing honestly rather than tidily: vasopressin is useful as an adjunct that reduces catecholamine requirements, its effect on mortality has not been convincingly demonstrated as a general matter, and major sepsis guidelines position it as an agent to add to norepinephrine rather than a first-line replacement. A real and defensible role, and a more modest one than early enthusiasm anticipated.

A regulatory footnote says something about how old this molecule is. Vasopressin was used in American hospitals for decades as an unapproved marketed drug — a legacy of the era before the modern approval framework — and only went through formal FDA approval in the mid-2010s, under the initiative to bring long-marketed unapproved drugs into the system. A peptide can be a fixture of routine hospital practice for longer than most physicians have been alive and still be, technically, a recent approval.

Terlipressin is a vasopressin analog with a different profile: a prodrug arrangement releasing the active moiety gradually, giving longer effective duration with relative V1 preference. Long used in Europe for esophageal variceal bleeding, its most notable modern indication is hepatorenal syndrome — kidney failure in advanced liver disease, driven substantially by splanchnic vasodilation and the resulting fall in effective renal perfusion. Terlipressin constricts the splanchnic circulation and can thereby improve renal function where alternatives are limited and prognosis is poor. It received US approval for hepatorenal syndrome in the early 2020s on the strength of a randomized trial program, with labeling reflecting real safety considerations including respiratory failure risk.

💊 In the Clinic — the peptide nobody calls a peptide, running through a central line

Picture the drug board in an intensive care unit at three in the morning. Norepinephrine at some rate. Vasopressin as a fixed-rate adjunct. Maybe an insulin infusion. Vancomycin due at four. Octreotide running for a variceal bleed. Cyclosporine on the transplant patient's list.

Most of those are peptides, and in a twelve-hour shift not one person will say so. They will say pressor, antibiotic, somatostatin analog, immunosuppressant — functional categories, the right vocabulary for the job.

This explains the gap the whole chapter is about. The medical system does not organize drugs by molecular class, so the enormous installed base of peptide medicine is invisible even to the people administering it. Meanwhile the consumer wellness market does organize by molecular class, precisely because "peptide" is a useful marketing category — so the word ends up meaning, in public discourse, almost exactly the set of compounds it should not mean.

A clinician who has never thought of vasopressin as a peptide is not making an error. But a reader trying to evaluate the category needs the inventory the clinician's vocabulary hides.

29.6 Glucagon: the rescue peptide, and a problem where the user was the constraint

Chapter 7 traced the proglucagon precursor — a single gene product cut by different enzymes in different tissues to yield different active peptides, GLP-1 among them. Glucagon is the other major product of that same precursor: a 29-amino-acid peptide, released by pancreatic alpha cells, whose job is to raise blood glucose.

Insulin (Chapter 11) moves glucose out of the blood; glucagon signals the liver to mobilize stored glycogen and release glucose into it. Between them they hold blood sugar inside a narrow band, and the failure mode that matters here is what happens when insulin therapy overshoots.

Severe hypoglycemia — blood glucose low enough to cause confusion, seizure, or loss of consciousness — is the acute emergency of insulin treatment (Chapter 11 §11.4). It frightens parents of children with type 1 diabetes, and it ends careers, driving licenses, and independent living for adults who experience it repeatedly. Its defining feature is that the person experiencing it cannot treat it. They cannot reliably swallow juice. They may not be conscious. Someone else has to act.

Injectable glucagon has been available as rescue therapy for decades and it works. The problem was never the molecule.

The problem was the kit

For most of that history, emergency glucagon came as a kit built around a chemical constraint: glucagon is poorly soluble and unstable in aqueous solution under the relevant conditions, so it was supplied as a lyophilized powder in one vial and a diluent in a syringe, combined at the moment of use. Reconstitution. Multiple steps, in sequence, correctly.

Now consider who performs those steps. Not a nurse. The realistic user is a parent at 2 a.m., a spouse, a teacher, a stranger at a restaurant — someone who has possibly never done this, is frightened, is looking at an unresponsive person, and may be reading the instruction card by phone flashlight. Studies of emergency kit use consistently found what anyone could have predicted: untrained and even trained users frequently failed to deliver a complete dose, and the failures clustered at the reconstitution steps.

This is a delivery-engineering problem where the constraint was not the molecule. It was the user.

That deserves naming, because Chapter 4 cataloged a completely different set of delivery challenges: proteolysis, absorption, half-life, membrane permeability, the blood-brain barrier. Every one is a contest between a peptide and a body. This one is a contest between a peptide and a panicking person's fine motor control, and no technique in Chapter 4's toolkit touches it.

TWO KINDS OF PEPTIDE DELIVERY PROBLEM

  CHAPTER 4 PROBLEMS                    THE GLUCAGON PROBLEM
  (molecule vs. body)                   (product vs. user)

  proteases digest it        ─────►     the user has never done this before
  it is too big to absorb    ─────►     the user is frightened
  the kidney clears it fast  ─────►     the patient is unconscious
  it cannot cross membranes  ─────►     it is the middle of the night
  the brain is walled off    ─────►     seven steps and no second chance

  SOLVED BY: chemistry                  SOLVED BY: formulation + industrial design

Since 2019, that second column has been genuinely addressed. Nasal glucagon — a dry powder delivered into the nostril, absorbed across the nasal mucosa and notably not requiring the recipient to inhale, which matters when the recipient is unconscious — reached the market, as did ready-to-use liquid glucagon in autoinjector and prefilled syringe formats and a glucagon analog reformulated for aqueous stability. The common thread is the elimination of reconstitution. One action. No mixing. Nothing to get wrong in sequence.

Comparative studies against the old kits found large improvements in the proportion of untrained users who successfully delivered a full dose, and in how long it took them. The molecule did not change. The number of ways to fail did.

Two lessons that generalize. First, nasal delivery of a peptide is achievable — but notice the conditions: a relatively small peptide, a single rescue event, systemic exposure attainable across nasal mucosa, and reliability mattering more than precision. Calcitonin nasal spray (§29.4) worked for related reasons. None of that generalizes to "peptides can be given nasally," and Part III will meet products that assume it does. Second: "is this drug usable by the person who has to use it?" is a real pharmaceutical question, distinct from efficacy and from safety, and here it produced more patient benefit than any molecular improvement could have.

📊 Evidence Rating

Claim: Glucagon administered by a bystander raises blood glucose and reverses severe hypoglycemia in a person with insulin-treated diabetes who is unable to self-treat. Rating: ✅ Strong clinical evidence Reason: Direct, rapid, physiologically unambiguous effect with an objectively measurable endpoint; approved and in routine emergency use for decades; modern nasal and ready-to-use liquid formulations showed non-inferior glucose recovery to injectable glucagon while substantially improving successful administration by untrained users. What would change it: Nothing about glucose recovery, which is about as settled as pharmacology gets. The live questions are downstream — comparative real-world success across formulations, and outcomes in populations with depleted hepatic glycogen, in whom the mechanism is inherently limited because there is less stored glycogen to mobilize. That is a boundary on the claim, not a challenge to it.

29.7 The diagnostic peptides: molecules used to ask a question

Here is a category this book has not yet named: peptides given not to treat anything, but to ask the body a question and see what it says.

The logic of a stimulation test is elegant. You suspect a gland is underperforming. Baseline hormone levels are often ambiguous, varying with time of day, stress, illness, and medications. So instead of measuring the resting state, you deliver the stimulus that should provoke a response and measure whether it comes. A gland that responds normally works. A gland that does not has told you something a baseline measurement could not.

Cosyntropin — also called tetracosactide — is the standard example: a synthetic fragment of adrenocorticotropic hormone, the first 24 amino acids of ACTH, which is the pituitary hormone telling the adrenal cortex to make cortisol. Native ACTH is 39 residues; as with teriparatide, the N-terminal portion carries the receptor-activating business, so a 24-residue fragment does the job.

Give cosyntropin, wait a defined interval, measure cortisol. If cortisol rises appropriately, the adrenal glands can respond; if not, they cannot — and the test has separated adrenal insufficiency from the many other causes of fatigue, low blood pressure, and electrolyte abnormalities. This is one of the most-used endocrine tests in medicine, and it is a peptide drug.

Two more, both historically important. GnRH (gonadorelin) and TRH (protirelin) were used in pituitary stimulation testing — GnRH to probe gonadotropin release, TRH to probe thyrotropin and prolactin. Both have largely been displaced by better hormone assays; when you can measure the downstream hormone sensitively enough, you often no longer need to provoke it. A pattern worth noticing: diagnostic peptides get retired by improvements in measurement, not by being disproven. A different mechanism of obsolescence than §29.4's, and an equally healthy one.

Secretin is the loveliest one, and it closes a loop this book opened long ago. Chapter 3's first case study covered Bayliss and Starling's 1902 experiment on the dog pancreas — the work establishing that a chemical messenger released from the intestine could travel through the bloodstream and act on a distant organ, and producing the concept of the hormone. The messenger they identified was secretin. It is the first hormone ever discovered, the molecule the entire concept of endocrinology was built on.

Secretin is a 27-amino-acid peptide, and today an approved diagnostic agent used to stimulate pancreatic secretion in the assessment of exocrine function and to aid visualization during certain endoscopic procedures. The founding molecule of endocrine science is now a workaday hospital reagent. There is something both deflating and rather wonderful about that trajectory.

🔍 Check Your Understanding

1. Explain in one sentence why a stimulation test can be more informative than a single baseline hormone measurement.
2. Cosyntropin is ACTH(1–24); teriparatide is PTH(1–34). What general principle do both illustrate, and what practical advantage does it give a manufacturer?
3. GnRH and TRH stimulation tests have largely been retired. Was that because the peptides stopped working? What actually changed?
4. What is the endpoint for a claim about a diagnostic agent, and how does it differ from a therapeutic endpoint?

29.8 Cyclosporine, vancomycin, and the peptides nobody calls peptides

This is the section that should genuinely surprise you.

Everything so far has been a peptide that at least sounds like one — hormones, analogs, fragments, things ending in -tide or -pressin. Now consider a set of drugs that no one, anywhere, in any clinical context, describes as peptides, and that are peptides.

Cyclosporine: the immunosuppressant you can swallow

Cyclosporine is a cyclic undecapeptide. Eleven residues joined in a closed ring, with no free N-terminus and no free C-terminus. It was isolated from a soil fungus and reached the market in the early 1980s, and its arrival is one of the genuine turning points in modern medicine: by making rejection controllable it converted organ transplantation from a heroic gamble into an established therapy. If you know someone alive with someone else's kidney, the causal chain runs through this molecule and its successors.

And here is the part that should stop you: it is taken orally.

Everything in Chapters 1 and 4 says that should not work. Swallow a peptide and it meets pepsin, then pancreatic proteases, then the peptidases of the intestinal brush border, and what survives is generally too large and too polar to cross the intestinal wall. That default is correct, and it is why almost every peptide drug in this chapter is an injection.

Cyclosporine violates it because it is not built like an ordinary peptide. Three features do the work:

It is cyclic. The chain is closed into a ring, so exopeptidases — the enzymes that chew inward from a chain's ends — have nothing to grab.

It contains a D-amino acid. Human proteases evolved to recognize the L-amino acids human proteins are made of. A D-residue is a bond the enzyme's active site does not read properly — the same trick desmopressin uses at position 8 (§29.3).

It is heavily N-methylated. Several of its peptide bonds carry a methyl group on the backbone nitrogen. This obstructs protease recognition and — critically — removes hydrogen-bond donors from the backbone. A backbone with fewer available hydrogen bonds is held less firmly by surrounding water, which makes the molecule far more membrane-permeable than its size predicts. Cyclosporine also contains unusual residues outside the standard twenty of Chapter 1 §1.2.

Add those together and you get a peptide that resists digestion and crosses membranes: orally bioavailable, protease-resistant, and biologically active inside cells, where its target lives. Its mechanism involves binding an intracellular protein and, as a complex, inhibiting calcineurin and therefore T-cell activation — an intracellular target, which Chapter 1 §1.6 listed as essentially off-limits to peptides.

Cyclosporine is the standing existence proof that the peptide limitations in this book are limitations of typical peptides, not laws. Chapter 33 covers how the industry has spent two decades trying to reproduce this feat on purpose, in the field called macrocyclic peptide design, with real but hard-won success. Nature got there first, in a fungus.

Vancomycin and the antibiotic peptides

Vancomycin is a glycopeptide antibiotic — a heavily modified, cross-linked peptide core with sugars attached. In clinical use since the 1950s, a mainstay against methicillin-resistant Staphylococcus aureus, and on the World Health Organization's Model List of Essential Medicines. It is given intravenously for systemic infection precisely because, being a peptide, it is not meaningfully absorbed from the gut — which has a corollary we return to in §29.9, since oral vancomycin is used because it stays in the gut, for Clostridioides difficile infection.

It is not alone. Chapter 25 covered antimicrobial peptides as a class, and several members are old, approved, and routine:

Notice the recurring structural theme: cyclic. Over and over. Nature's antimicrobial peptides, like cyclosporine, are frequently cyclized, frequently contain non-standard residues, and are far more stable than a linear peptide of the same length. Microorganisms solved the peptide-stability problem long ago, using the chemistry pharmaceutical scientists later reinvented.

Two more, for completeness. Octreotide (Chapter 27) is an eight-residue cyclic somatostatin analog containing D-amino acids — cyclosporine's design language applied to a hormone analog — and a workhorse in acromegaly, neuroendocrine tumors, and variceal bleeding. Oxytocin (Chapter 21) is a nine-residue cyclic peptide, one of the most widely administered drugs on earth by volume.

⚠️ Hype Check — "peptides are the next generation of medicine"

The claim, in its usual form:

"Peptides are the next frontier in medicine — a revolutionary new class of therapeutics that's only now becoming possible. We're at the very beginning of the peptide era."

What's true in it. There is real and accelerating activity in peptide drug development. Advances in synthesis, half-life extension, macrocyclization, and oral delivery have opened possibilities that did not exist twenty years ago. GLP-1 receptor agonists have been genuinely transformative. The pipeline is full.

Where it fails. Insulin has been in clinical use since 1922. Vancomycin since the 1950s. Synthetic oxytocin since the 1950s. Cyclosporine since the early 1980s, and it changed transplantation permanently. Desmopressin, calcitonin, leuprolide, octreotide, and cosyntropin are all decades old. Peptide medicine is not an emerging field. It is a mature field having a good decade.

Why does that matter? Because "emerging field" carries an implicit permission structure. If a field is brand new, the absence of long-term data is expected, the absence of large trials is understandable, and the enthusiast running ahead of the evidence is a pioneer rather than someone making things up. "We're at the frontier" is the most effective available excuse for not having evidence.

But the field is not at the frontier. It has an eighty-plus-drug pharmacopeia, a fifty-year track record, established regulatory pathways, and well-understood manufacturing. So the right question about any unapproved peptide is not "why hasn't this new field caught up?" It is: this field routinely gets peptides approved and has for decades — so what specifically happened with this one?

That question has informative answers. Sometimes it is in trials right now, properly (🔬). Sometimes trials were run and did not work. Sometimes nobody has ever run one, despite the compound being cheap to make and thirty years old — which is itself a data point.

Verdict: a category error functioning as an alibi. The maturity of peptide medicine is not a point in favor of unapproved peptides — it is the strongest argument against excusing their absence of evidence.

29.9 Icatibant and linaclotide: what modern peptide approvals look like

The older drugs in this chapter were mostly hormones or natural products. The newer ones show what happens when peptide design is deliberate from the start.

Icatibant: blocking, not replacing

Hereditary angioedema is a rare genetic disorder of episodic deep tissue swelling — face, limbs, abdomen, and dangerously, the airway. The underlying defect involves the complement and contact systems, and the immediate mediator is bradykinin, a nine-residue peptide that increases vascular permeability. During an attack, excess bradykinin acts at the bradykinin B2 receptor and fluid leaks into tissue.

Icatibant is a synthetic decapeptide that antagonizes the bradykinin B2 receptor. Not a replacement hormone, not a natural product — a purpose-built blocker designed to occupy a receptor and prevent the natural ligand from acting there. It contains several non-proteinogenic amino acids, residues outside Chapter 1's standard twenty, which give it the stability to function as a drug.

Most peptides in this book are agonists that activate a receptor, usually mimicking something the body makes. Peptide antagonists are less common and harder to design, because blocking a receptor requires binding tightly while failing to trigger the conformational change activation requires. Icatibant does it, and it is one of the cleaner demonstrations that peptides are not confined to the "replace a missing signal" role.

Linaclotide: engineered not to be absorbed

Linaclotide is a 14-amino-acid peptide, stabilized by three disulfide bonds, taken orally and designed not to be absorbed.

Read that again, because it inverts everything.

Chapter 4 §4.4 established systemic bioavailability as the central problem of peptide delivery. Get the molecule into the bloodstream intact; everything else is downstream. The entire toolkit — lipidation, PEGylation, protease-resistant substitutions, absorption enhancers, oral semaglutide's roughly 1% achievement treated as a triumph — exists to raise a number peptides are naturally terrible at.

Linaclotide's target is the guanylate cyclase-C receptor on the luminal surface of intestinal epithelial cells. It faces the inside of the gut. Activating it increases intestinal fluid secretion and accelerates transit — exactly what you want in constipation-predominant irritable bowel syndrome and chronic idiopathic constipation. Linaclotide is approved for both. The drug needs to reach the gut lumen and to not reach anything else, since systemic absorption would contribute nothing therapeutically and could only add off-target exposure.

BIOAVAILABILITY: USUALLY THE GOAL, SOMETIMES THE ENEMY

  SEMAGLUTIDE, ORAL                        LINACLOTIDE, ORAL
  target: pancreas, brain, gut, elsewhere  target: the inside surface of the gut
  requires: systemic circulation           requires: gut lumen only

  ~1% systemic bioavailability             minimal systemic absorption
  = a hard-won engineering triumph         = the specification, met

  the other 99% is loss                    absorption would be a defect

Linaclotide's peptide nature is not a liability it overcame. It is the feature. A small molecule with the same receptor activity would need active engineering to prevent absorption. Linaclotide gets confinement for free, because being a peptide in the gut is a strong guarantee of staying there. The compartment does the targeting.

Plecanatide is a related agonist at the same receptor, structurally closer to the natural human ligand uroguanylin, likewise approved for chronic idiopathic constipation and IBS-C and likewise minimally absorbed. Oral vancomycin for C. difficile (§29.8) runs the same logic from the opposite direction: a drug whose lack of absorption is a nuisance when you want systemic antibiotic activity becomes exactly right when the infection is in the colon.

The generalizable idea, one of the most useful in Part V: a peptide's confinement to a compartment is a limitation only relative to a target outside that compartment. Move the target and the same property becomes an advantage. Ask, of any peptide claim, not "can this be absorbed?" but "where does this need to be, and does its behavior put it there?"

📊 Evidence Rating — two modern approvals

Claim A: Linaclotide improves abdominal pain and bowel symptoms in adults with constipation-predominant irritable bowel syndrome, and improves bowel frequency in chronic idiopathic constipation. Rating: ✅ Strong clinical evidence Reason: Multiple randomized, placebo-controlled, adequately powered phase 3 trials with prespecified symptom endpoints supported approval for both indications; extensive post-marketing record; appears in gastroenterology treatment guidelines. What would change it: Little, for the approved endpoints. The honest boundary is magnitude rather than existence — the effect is real and statistically robust, and it is symptomatic improvement rather than cure, with diarrhea the predictable common adverse effect of a drug whose mechanism is increased intestinal fluid secretion.

Claim B: Icatibant relieves acute attacks of hereditary angioedema in adults. Rating: ✅ Strong clinical evidence Reason: Randomized controlled trial evidence in acute attacks supported approval; the mechanism targets the established mediator of the swelling directly; established in hereditary angioedema guidelines as an acute-attack option. What would change it: Very little for the acute-attack claim. Note what this does not cover: icatibant has been studied in other conditions involving bradykinin-mediated swelling, with results that did not replicate this success. A ✅ in a rare disease with a precisely identified mediator does not travel to a different disease with a superficially similar mechanism. That is rule 3 — never upgrade with mechanism — applied to a molecule that already earned a ✅ elsewhere.

29.10 What the pharmacopeia proves

You have now met roughly twenty approved peptide drugs, most of which you could not have named at the start of the chapter. Go back and look at the number you wrote down in §29.1.

The first conclusion is easy: peptide medicine is mature. Not emerging, not speculative, not a frontier. It has an installed base of more than eighty approved drugs, generics, essential-medicines listings, established manufacturing, well-worn regulatory pathways, and a track record going back a century if you start at insulin. Anyone describing this field as new is either misinformed or using the newness to excuse something.

The second conclusion is sharper, and it is why this chapter exists. Look at what these drugs actually do. Not what they are made of — what job they perform.

EVERY APPROVED PEPTIDE DRUG IN THIS CHAPTER, SORTED BY WHAT IT DOES

 ┌─ 1. REPLACE A MISSING OR INSUFFICIENT SIGNAL ────────────────────────────┐
 │   insulin ............... the pancreas does not make enough              │
 │   desmopressin .......... the pituitary does not release enough ADH      │
 │   glucagon .............. the counter-regulatory response failed         │
 │   teriparatide .......... exploits a pattern the body no longer produces │
 │   → all require a DEFICIT. No deficit, no indication.                    │
 └──────────────────────────────────────────────────────────────────────────┘
 ┌─ 2. SUPPRESS OR BLOCK AN OVERACTIVE SIGNAL ──────────────────────────────┐
 │   icatibant ............. blocks bradykinin B2 in HAE attacks            │
 │   leuprolide/degarelix .. suppresses the gonadal axis in cancer          │
 │   octreotide ............ suppresses excess growth hormone / secretion   │
 │   → all require an EXCESS or a pathological drive.                       │
 └──────────────────────────────────────────────────────────────────────────┘
 ┌─ 3. ACT LOCALLY, IN A DEFINED COMPARTMENT ───────────────────────────────┐
 │   linaclotide, plecanatide ... the gut lumen                             │
 │   oral vancomycin ............ the colon                                 │
 │   bacitracin ................. the skin surface                          │
 │   → the compartment is doing the targeting.                              │
 └──────────────────────────────────────────────────────────────────────────┘
 ┌─ 4. KILL A PATHOGEN ─────────────────────────────────────────────────────┐
 │   vancomycin, daptomycin, polymyxins, bacitracin                         │
 │   → an external invader, not a human system.                             │
 └──────────────────────────────────────────────────────────────────────────┘
 ┌─ 5. ASK A DIAGNOSTIC QUESTION ───────────────────────────────────────────┐
 │   cosyntropin, secretin, historical GnRH/TRH testing                     │
 │   → no therapeutic intent at all.                                        │
 └──────────────────────────────────────────────────────────────────────────┘

 ┌─ 6. BROADLY SUPPORT, OPTIMIZE, ENHANCE, OR MODULATE A HEALTHY SYSTEM ────┐
 │                                                                          │
 │                              — empty —                                   │
 │                                                                          │
 └──────────────────────────────────────────────────────────────────────────┘

That sixth box is the most important thing in this chapter.

There is no approved peptide drug whose indication is optimization. None whose label reads "for general recovery." None approved to broadly support the immune system, to enhance healing in a person who is healing normally, or to improve function in an organ that is functioning. Every approved peptide either corrects an identified deficiency, opposes an identified excess, works inside a defined compartment, kills something, or asks a question.

Now be scrupulous about the objection, because it is a fair one.

Objection: "That is circular. Approval requires demonstrating benefit on a clinical endpoint in a defined population, and 'optimization' is not a clinical endpoint, so of course nothing is approved for it. The absence reflects regulatory categories, not biology."

Partly right, and the right part is worth conceding fully. Regulatory frameworks are built around disease. A compound making healthy people marginally better at something would have genuine trouble defining an endpoint, an eligible population, and an acceptable risk-benefit ratio. Some of box six's emptiness is structural.

But not all of it. Three independent observations:

The incentive is enormous and unclaimed. A drug for healthy people is the largest market in pharmaceuticals — numerous, generally solvent, not sick enough to be frightened of side effects. Companies have tried repeatedly to enter it, and when one succeeds it is gigantic. The absence of peptide entries here is not for lack of commercial motivation.

Adjacent regulatory categories exist and are not full either. Preventive, risk-reduction, and functional-improvement indications are real and approvable; drugs hold them. Statins carry a risk-reduction indication in people without established disease. The framework can accommodate something short of treating active illness. Yet peptide entries in those categories remain conspicuously thin.

Most tellingly, the pattern holds everywhere we have looked. Chapter 27 examined peptides in oncology and found precisely this shape: peptides succeed as receptor-targeted agents against a specific identified abnormality — a receptor a tumor overexpresses, an axis that must be shut down — and fail as general supportive agents. Chapter 37 assembles the master table of every compound in this book and finds the same thing at full scale. When a pattern holds independently across the approved pharmacopeia, across oncology, and across the entire book's compound list, "it's just a regulatory artifact" stops being sufficient.

The physiological account, argued from first principles in Chapter 3 and demonstrated here with eighty-plus worked examples: peptide signaling systems are homeostatic and heavily regulated. They exist to keep variables inside ranges. Adding signal to a system already in range is not obviously beneficial and is frequently met by counter-regulation — receptor downregulation, suppression of endogenous production, altered clearance. These systems are built to resist exactly the push "optimization" proposes.

The prior this gives you

When you encounter a claim that a peptide will support, optimize, enhance, restore, or modulate something in a person who does not have an identified deficiency, excess, pathogen, or localized pathology — you are looking at a claim of a type that has never once succeeded in obtaining approval, despite fifty years of development, enormous commercial incentive, and more than eighty approved drugs in the class.

That is not proof the claim is false. Priors are not proofs, and this book has been careful to say that a ❌ is a statement about evidence rather than a verdict on a molecule. Some claim of this type may someday succeed, and if it does, this paragraph will need revising.

But it is an extremely strong prior, and you built it yourself out of public facts — a list of approved drugs and their indications, available to anyone who bothers to look. You do not need to know anything about a specific compound to apply it. You need only ask which of the six boxes the claim belongs in, and notice when the answer is the empty one. Part VI is where you will use it most.

📋 Your Evidence Dossier

This chapter fills Field 7 — Regulatory Status — at breadth rather than at depth.

Chapter 28 had you fill in Field 7 for each dossier peptide: approved or not, where, for what, under what conditions. That was a vertical exercise, one molecule at a time. This chapter asks for the horizontal version — a scan across the whole approved pharmacopeia, looking for molecules you already knew about without knowing they were peptides.

Step 1 — The recall test, scored

Go back to §29.1. You wrote down how many approved peptide drugs you could name before reading this chapter. Now write down how many you can name after. Record both, with today's date.

The gap is the point. Most readers move from two or three to fifteen or more in one sitting, and essentially none of the new entries are new drugs — they are drugs the reader already knew existed and had filed under a different heading. You did not learn about new medicines. You recategorized familiar ones. That kind of learning changes judgments rather than adding facts.

Step 2 — Add at least two "I didn't know that was a peptide" entries

Choose at least two approved peptide drugs from this chapter that you had encountered before — taken, been prescribed, seen in a family member's cabinet, or handled professionally — but never thought of as peptides. Good candidates depending on your history: vancomycin, cyclosporine, oxytocin, desmopressin, leuprolide, octreotide, bacitracin, calcitonin, glucagon, linaclotide, cosyntropin.

For each, fill in Field 1 (Identity) as you did in Chapter 1 and Field 7 (Regulatory Status) as you did in Chapter 28.

FIELD 7 — REGULATORY STATUS          [the breadth version]
  Approved?              yes / no / in some jurisdictions only
  Where                  FDA / EMA / other; note differences
  Approved for what      the exact indication wording, as specifically as you can find it
  Since when             approximate year of first approval
  Generic available?     yes / no — a proxy for how long and how widely it has been used
  Which functional box   1 replace · 2 block · 3 local · 4 antimicrobial · 5 diagnostic · 6 optimize

That last line is the one that will earn its keep. Assign every entry in your entire dossier — not just the new ones — to one of the six boxes from §29.10.

Step 3 — The sort, and what it will show you

Lay the dossier out and count how many entries fall into each box. You will very likely find your approved peptides distributed across boxes 1 through 5, and your unapproved ones — the gray-market compounds, the clinic offerings, the one a training partner recommended — clustered overwhelmingly in box 6.

Sit with that before deciding what it means, because two readings are available and they are not equally supported.

Reading A: box 6 is where the unmet need is, and the approved pharmacopeia has not gotten there yet.

Reading B: box 6 is where claims go when they do not have to survive an endpoint — because a claim about optimization in a healthy person is difficult to falsify, which is exactly what makes it attractive to market and unattractive to test.

Both come from the same data. This book's position, argued in §29.10 and reinforced in Chapters 27 and 37, is that Reading B carries far more of the weight — but write down which reading you find more plausible today, and date it. Chapter 40 will ask again.

Step 4 — One question to carry forward

For each box-6 entry, write one sentence answering this: what would this compound's approval trial even look like? Who would be enrolled, what measured, over how long?

Some of those questions have real answers. If a compound's advocates can answer and no one has run the trial, that is a fact about funding and incentives, and worth knowing. If nobody can answer — if the claim genuinely cannot be turned into a measurable endpoint in a definable population — you have learned something important about the claim, without needing a single piece of evidence about the molecule.

Conclusion

More than eighty peptide drugs are approved and in routine use. They treat diabetes and osteoporosis and bedwetting and bleeding disorders and shock and kidney failure and constipation and hereditary angioedema and prostate cancer and acromegaly and bacterial infection. They prevent organ rejection, reverse hypoglycemia in an unconscious person, induce labor, and ask whether an adrenal gland works. Several are cheap generics. One is an essential medicine. One made transplantation possible. Most have been available for decades.

Almost none of them come up when people talk about peptides.

Along the way this chapter made smaller points worth keeping. That the same hormone builds bone or dissolves it depending on nothing but temporal pattern — Chapter 3's pulsatility principle turned into an approved therapy. That one change in receptor subtype selectivity gave a single molecule three indications in three specialties. That a drug class can honestly contract because the evidence improved, with no villain and no scandal, and that this process — which the gray market entirely lacks — is what a functioning evidentiary system looks like from inside. That the hardest delivery problem in the glucagon story was not the molecule but the frightened person holding it. That the immunosuppressant which made transplantation practical is a peptide you can swallow. And that a peptide engineered not to be absorbed is not a failure but a specification met.

But the durable thing is the sixth box.

Sort every approved peptide drug by what it does and you get five populated categories and one empty one. Replace a missing signal. Block an overactive one. Act locally. Kill a pathogen. Ask a question. And — nothing at all — broadly support, optimize, or enhance a system that is working.

Fifty years of development. Enormous commercial incentive. More than eighty approved drugs. Zero entries in box six.

Chapter 30 turns to cosmetic peptides, where the claims are almost entirely box-six claims and the regulatory framework differs in ways that matter enormously. The prior you built here does a great deal of work there — without requiring you to evaluate a single molecule on its merits first.

That is what a good prior is for. It tells you where to spend your skepticism.

Key Terms

Anabolic agent (bone) — a drug that stimulates new bone formation rather than merely slowing loss.

Antiresorptive — a drug that slows removal of existing bone by inhibiting osteoclasts.

Parathyroid hormone (PTH) — an 84-residue hormone regulating blood calcium. Continuous elevation removes bone; intermittent exposure adds it.

PTH(1–34) — the N-terminal fragment of PTH, retaining full receptor activity. Teriparatide.

Receptor subtype selectivity — a ligand's preference for one receptor subtype over another; the design change behind desmopressin.

V1 receptor — the vasopressin subtype on vascular smooth muscle; mediates vasoconstriction.

V2 receptor — the vasopressin subtype in the renal collecting duct and vascular endothelium; mediates water reabsorption and release of von Willebrand factor and factor VIII.

Hyponatremia — abnormally low blood sodium; the principal risk of drugs promoting free water retention.

Vasopressor — a drug raising blood pressure by increasing vascular tone.

Vasodilatory shock — circulatory failure from loss of vascular tone, most commonly in sepsis.

Hepatorenal syndrome — kidney failure in advanced liver disease driven substantially by splanchnic vasodilation. Terlipressin's modern indication.

Proglucagon — the precursor cleaved differently in different tissues to yield glucagon, GLP-1, and related peptides (Chapter 7).

Cosyntropin (tetracosactide) — a synthetic 24-residue ACTH fragment used to test adrenal responsiveness.

Stimulation test — giving a provoking agent and measuring the hormonal response, to assess whether a gland can respond.

Glycopeptide — a peptide bearing attached sugar groups. Vancomycin is the dominant example.

Cyclic peptide — a chain closed into a ring, eliminating free termini and conferring resistance to exopeptidases.

N-methylation — a methyl group on a backbone amide nitrogen; reduces hydrogen-bond donors and protease recognition, increasing membrane permeability. Central to cyclosporine's oral activity.

D-amino acid — the mirror-image stereoisomer of the L-amino acids of human proteins; poorly recognized by human proteases, and therefore a stability strategy.

Undecapeptide — a peptide of eleven residues. Cyclosporine is a cyclic undecapeptide.

Guanylate cyclase-C agonist — a drug activating the receptor on the luminal surface of intestinal epithelium. Linaclotide, plecanatide.

Systemic bioavailability — the fraction of a dose reaching circulation intact. Usually the goal; for linaclotide, deliberately minimized.

Hereditary angioedema — a genetic disorder of episodic bradykinin-mediated tissue swelling.

Boxed warning — the most prominent safety warning a label can carry. Teriparatide's was added on rodent data and later removed on human surveillance data.

Indication — the condition, population, and use a drug is approved for. Ratings attach to claims about indications, never to molecules.

Spaced Review

1. (Ch 3 + Ch 29) State Chapter 3 §3.5's pulsatility principle in one sentence, then explain what would happen if teriparatide were delivered by continuous infusion instead of daily injection — and why that prediction is not merely theoretical.

2. (Ch 11 + Ch 29) Glucagon's modern nasal and ready-to-use formulations solved a problem none of Chapter 4's delivery techniques addresses. What was the problem, why is it a different kind of problem, and what does that suggest about evaluating any product whose selling point is convenience of administration?

3. (Ch 28 + Ch 29) Take one approved and one unapproved compound from your dossier and assign each to one of the six functional boxes in §29.10. If the unapproved one lands in box six, write the strongest argument against treating that as damning — then the strongest response to it.

4. (Ch 29) Desmopressin has three approved indications from one receptor; calcitonin lost an indication it once held. Explain how a source assigning a single overall rating to each molecule would mislead a reader in each case, and in opposite directions.

5. (Ch 29) Someone tells you a peptide "isn't approved yet because the field is still so new." Using only this chapter, construct a three-sentence reply. Name at least two approved peptide drugs and the approximate era of their introduction, and end by stating the question to ask instead.