Case Study 1 — Forty Years of Nobody Caring

How gut hormone physiology became a hundred-billion-dollar drug class

Type: Real, public, historical · Tier 1 institutional facts, Tier 2 timeline detail · Relevance: §7.1, §7.8 · Deliberate contrast with Chapter 6

Background: a field with no application

In 1902, Bayliss and Starling identified secretin and coined the word "hormone" (Chapter 3, Case Study 1). The first hormone ever described was a gut hormone.

For most of the following century, gastrointestinal endocrinology was a respectable, productive, and commercially uninteresting field. Investigators characterized gut peptides — secretin, cholecystokinin, gastrin, motilin, and eventually the incretins — worked out what triggered their release and what they did, and published in specialist journals.

There was no drug at the end of it. Nobody working on the incretin effect in 1975 was working toward a weight-loss medication, because nobody thought a gut hormone could become one. The half-life problem alone (§7.6) made the idea look unserious.

The work proceeded anyway, funded largely by public research agencies and academic institutions, because understanding how digestion coordinates with metabolism is a legitimate scientific question whether or not anyone can sell the answer.

The sequence

Compressed, and with the caveat that any such timeline flattens a much messier reality in which multiple groups worked in parallel and priority disputes exist:

Early twentieth century — the "incretin" hypothesis is proposed: some intestinal factor stimulates the pancreas. It is untestable with available methods and largely set aside.

1960s — reliable insulin assays make the oral-versus-intravenous comparison possible. The incretin effect is demonstrated. The hypothesis is vindicated after roughly half a century.

1970s — GIP is identified and characterized. Initially named for an inhibitory effect on gastric secretion, later renamed for its insulinotropic action — a name change that records a shift in understanding.

1980s — the proglucagon gene is sequenced, revealing that it encodes more than glucagon. GLP-1 is identified as one of its products, and its insulinotropic activity is characterized. The critical realization follows: GLP-1's effect is glucose-dependent.

Late 1980s–1990s — GLP-1 infusion studies in humans confirm the physiology, including in people with type 2 diabetes, where GLP-1's effect is preserved while GIP's is blunted. The therapeutic implication is now obvious and the delivery problem is now clearly the obstacle.

1990s — exendin-4 is characterized from Gila monster venom and recognized as a DPP-4-resistant GLP-1 receptor agonist (Chapter 4, Case Study 1).

2005 — exenatide is approved: the first GLP-1 receptor agonist.

2010s — liraglutide, then semaglutide. Doses and durations increase; weight effects become the dominant clinical story.

2020s — tirzepatide; cardiovascular outcome data; a market of a scale nobody predicted.

Roughly a century from hypothesis to blockbuster. Roughly forty years from the demonstrated incretin effect to a drug class reshaping metabolic medicine.

🔬 Read the Study — the GLP-1 infusion studies

text FIGURE 7.CS1 — "Proving it works before you can deliver it" [real published work] THE STUDY Human physiological studies, late 1980s–1990s: continuous intravenous infusion of native GLP-1 in healthy volunteers and in people with type 2 diabetes, with measurement of insulin, glucagon, glucose, gastric emptying, and food intake. Small, mechanistic, academically funded. THE QUESTION Does GLP-1 produce the predicted metabolic effects in humans, and are they preserved in type 2 diabetes? WHAT IT SHOWS Yes. Infused GLP-1 lowers glucose, enhances insulin secretion glucose-dependently, suppresses glucagon, slows gastric emptying, and reduces food intake — and unlike GIP, its insulinotropic effect is substantially preserved in type 2 diabetes. THE DOESN'T It does not establish that any of this is achievable outside a hospital. Continuous IV infusion is not a therapy. It also does not establish long-term efficacy, safety, or effects on any clinical outcome — these are short mechanistic studies with surrogate endpoints. THE VERDICT Definitive for the PHYSIOLOGY, silent on the THERAPY. Exactly the state of affairs that §7.6's ❌ describes. THE LESSON Establishing that a mechanism operates in humans is a real and necessary achievement — and it is Chapter 2's step 1 through 4, not step 6. The gap between these studies and exenatide's approval was fifteen years, and the obstacle was delivery, not biology.

Why this case belongs in Chapter 7

Because it is the opposite of Chapter 6's pipeline.

Chapter 6 described information degrading as it travels from a real result toward a market. This case describes a real result traveling toward a market over four decades, with the qualifiers intact, through peer review, physiological confirmation in humans, a delivery problem honestly acknowledged and eventually solved, formal clinical development, and regulatory review.

Same starting material — a genuine preclinical and physiological finding. Completely different journey.

The difference is not that the scientists were more virtuous. It is structural: the pathway from basic physiology through pharmaceutical development to approval has checkpoints that require the qualifiers to survive. A regulator asks which population, which endpoint, which duration. A trial protocol has to specify them in advance. A label has to state them.

That process is slow, expensive, and frequently criticized — sometimes rightly. It is also the only mechanism anyone has built that reliably transmits qualifications from a laboratory to a patient.

And this is the honest counterweight to Chapter 6. A reader who finishes Chapter 6 believing that all health information is compromised has over-learned it. The system that produced semaglutide worked — slowly, expensively, publicly funded at the front end, and with the crucial information preserved.

The uncomfortable parts

Three, because a triumph narrative would be its own distortion.

Nobody could have picked this in advance. The forty years of incretin physiology were not a targeted investment that paid off. They were one of many lines of basic research, most of which do not produce a drug class. You cannot use this case to argue for funding any particular research program, only for funding basic research in general — and that argument is available to every project, which makes it weak as a decision procedure. Chapter 3's discussion question on this is not rhetorical.

The people who did the foundational work are not the ones who profited. This is the normal structure of pharmaceutical development and it is not a scandal, but it is worth noticing that the public paid for the physiology and pays again for the drug. Chapter 12 takes this up.

And the timeline's slowness is a real cost. Between the confirmed human physiology in the early 1990s and a drug that produced substantial weight loss, roughly thirty years elapsed, during which people had the condition and did not have the treatment. Chapter 5's discussion of how much evidence is enough before approval is not academic; the cost of caution is paid by people who wait.

Both things are true, and neither cancels the other. The process transmitted the qualifiers, and it took a very long time, and the second is not simply a defect to be engineered away.

Discussion questions

1. Compare this case's path directly to Chapter 6's five-stage pipeline. At which stages does the pharmaceutical development pathway force qualifiers to survive, and what is the mechanism in each case?

2. The GLP-1 infusion studies established the physiology fifteen years before a drug existed. Was that work valuable during those fifteen years? How would you have argued for continuing to fund it in 1995?

3. Nobody working on gut hormones in 1975 was working toward a weight-loss drug. Does this case support or undermine the argument for funding basic research without regard to application? Be precise about what it does and does not show.

4. GIP was characterized before GLP-1 and then largely set aside for decades — and has now returned as part of the most effective agent in the class. What does that suggest about how confident to be that a currently uninteresting finding will stay uninteresting?

5. The process took a century from hypothesis to blockbuster and thirty years from confirmed human physiology to a highly effective drug. Identify one place in that timeline where you would have accepted more risk to move faster, and state what you would have been risking.

6. Compare with Chapter 6. Both chapters describe a real finding traveling toward the public. Write one paragraph on what actually differs — not "one is science and one is marketing," but the specific structural features that produce the different outcomes.