Case Study 38.1 — Fifty-One and Thirty-One: A Number That Decides a Market

What this case examines: how a definitional threshold written into regulation — more than 40 amino acids, or 40 and fewer — determined the competitive future of the two most consequential peptide drug families of the last century, and why the consequences show up in price rather than in pharmacology.

A note before starting. This case describes United States regulatory structures as they stand at the time of writing, in 2026. Those structures differ by country and change. Nothing here is legal advice, and nothing here describes what any reader may lawfully do.

The setup

Two molecules, both peptides in the ordinary chemical sense of Chapter 1, both enormously important, both made largely of the same twenty building blocks.

Insulin. Fifty-one amino acids, arranged in two chains held together by disulfide bonds. In clinical use since 1922. On any list of the most important medicines ever developed.

Semaglutide. A peptide backbone of thirty-one residues, based on human GLP-1, modified at two positions and carrying a fatty acid chain hung off a lysine (Chapter 33). Approved in the late 2010s, and by the mid-2020s the center of the largest pharmaceutical story in decades.

Now apply the line. Insulin is over 40 and is a biological product. Semaglutide is under 40 and is a drug.

Nothing about their chemistry, mechanism, or clinical use follows from that. Everything about their follow-on competition does.

Part 1 — Insulin's strange legacy position

Insulin has been sold in the United States since long before the modern biologics framework existed. Because of that history, it was regulated as a drug, under New Drug Applications, for most of the twentieth century — despite being, by any modern definition, a protein produced in living cells.

That legacy classification created a specific and unusual trap.

In principle, a drug can face generic competition. A generic applicant demonstrates that its product contains the same active ingredient and is bioequivalent — that the same amount of drug reaches the bloodstream on the same time course — and, that shown, the agency accepts the original efficacy data. Cheap to do, relative to a full development program, which is why generic small molecules cost a fraction of the originals.

In practice, that route did not work for insulin. Bioequivalence demonstrations were designed for molecules a chemist can make identically. A recombinant protein produced in a particular cell line with a particular process is not reproducible atom for atom by a different manufacturer, and the analytical and clinical work needed to show equivalence looks nothing like a standard generic filing.

So insulin sat in a category whose competition pathway was, for insulin, largely theoretical — and it had no access to the biosimilar pathway, which had been built specifically for molecules like it. Follow-on insulins that did reach the market generally came through routes that did not deliver the price dynamics of true generic competition.

Meanwhile the price of insulin in the United States became one of the most visible cost-of-medicine controversies of the 2010s, and it did so for a molecule whose patents on the original form had expired decades earlier. Patent expiry alone did not produce competition, because the pathway for competitors to arrive was missing.

Part 2 — The transition

Under a statutory transition that took effect in March 2020, insulin and a set of other protein products were deemed to be biological products. They moved from NDAs to the biologics framework.

The molecule did not change. Not one atom. What changed was which application type governed it and, consequently, which follow-on pathway existed: the biosimilar route, with its analytical and clinical comparability requirements and its separate interchangeability determination for pharmacy-level substitution.

It is worth being precise about what this did and did not accomplish.

What it did. It opened a route. Biosimilar and interchangeable insulin products became possible in a way they had not been. Interchangeability, in particular, matters because it bears on whether a pharmacist can substitute without going back to the prescriber — which is the mechanism by which follow-on competition actually reaches patients rather than merely existing on paper.

What it did not do. It did not make follow-on insulin cheap to develop. The biosimilar pathway is abbreviated relative to a full original application and remains enormously more demanding than a generic filing — which means fewer entrants, later, with less price pressure than a crowded generic market produces. Whether the resulting price effects have been large or modest is a genuine and live argument. What is not in dispute is that a reclassification changed the structure of the market.

Part 3 — The other side of the line

Now look at semaglutide and tirzepatide, at 31 and 39 residues.

When their exclusivity ends — and the timing depends on patent estates, not on the classification (§36.9) — the route open to competitors will be the generic route. Abbreviated applications. Bioequivalence rather than comparative clinical programs. A dramatically lower cost of entry, and historically, in crowded generic markets, dramatic price declines.

This is not a prediction that these drugs will become cheap on a particular date. Peptide manufacturing at scale is genuinely difficult (Chapter 32), device patents and formulation patents can extend practical exclusivity beyond the compound patents, and the market may not behave like a small-molecule generic market. But the pathway is the generic pathway, and that is a structurally different situation from insulin's for the fifty years before 2020.

Two residues, in tirzepatide's case, is the margin.

Part 4 — What the case is actually about

It would be easy to read this as a story about a quirk. It is better read as a story about how categories do work that people attribute to other causes.

Ask most people why insulin was expensive in the United States and you will hear about patents, corporate behavior, and the insurance system. All of those are part of the story. But a substantial part of it is that a molecule sat in a regulatory category whose competition mechanism did not function for molecules of its kind, and that this was a consequence of when it happened to be approved rather than of anything anyone decided about insulin specifically.

And ask most people what determines whether a peptide drug gets cheap competitors and you will hear about patents. Patents determine when. The forty-amino-acid line determines what kind of competitor is even possible — which is the question that decides how much cheaper.

The general lesson: a number in a definition is not a technicality. It is a policy instrument, and in this case it is one of the largest in the peptide field.

Discussion Questions

1. The forty-amino-acid line has no chemical meaning — nothing happens at residue 41. Defend the line anyway. What is a regulator trying to accomplish with a bright-line rule, and what would the alternatives cost? Then argue the other side: what does the arbitrariness actually cost, and who bears it?

2. Insulin's price problem persisted for decades after its original patents expired. Using Part 1, explain to someone who assumes patents are the whole story why patent expiry was necessary but not sufficient. What else has to be present for prices to fall?

3. The March 2020 transition changed insulin's classification without changing the molecule. List everything that changed and everything that did not. What does the length of the first list, relative to the second, tell you about how much of drug pricing is determined by law rather than by chemistry?

4. Tirzepatide is 39 residues — one residue from the line. Suppose a future molecule in the same class were designed at 41. What would change about its commercial life, and would a sponsor have any reason to prefer one side over the other? Consider both the development cost and the post-exclusivity period, and note where you are speculating.

5. Chapter 12 argued that access to GLP-1 drugs is the central social question they raise. Using this case, state precisely how the forty-amino-acid line bears on that argument — and how strong the bearing is. Is it a major determinant of the access picture a decade out, a minor one, or unknowable?

6. This case shows regulatory categories producing large economic consequences. Does that make the categories good policy instruments, bad ones, or simply unavoidable ones? In your answer, distinguish between the effects of drawing a line somewhere and the effects of drawing it at 40.