Discussion Guide — Chapter 22

Six prompts, roughly in the order the material supports them. Each has a "what to listen for" note — these describe the quality of reasoning you want, not a target answer.

Prompt 1 — "The NK1 program spent two decades and enormous resources and produced no analgesic and no antidepressant. Was it a waste?"

What to listen for. The first answers will be yes, and the interesting move is what happens next. You want students to arrive at the distinction between an absent evidence base and a present and negative one — that a well-conducted negative result is knowledge that did not exist before, and that the field now knows something permanent about NK1 and human pain and mood. Listen for students who notice that the aprepitant antiemetic indication came out of the same program, and press them on whether that makes it not-a-waste for the wrong reason (a commercial accident) or the right one (the emesis hypothesis was structurally better all along). Weak answers treat "waste" as purely financial. Strong answers separate the epistemic ledger from the commercial one.

Prompt 2 — "Human PET studies confirmed the drugs occupied NK1 receptors in the brain. What did that fact do to the argument?"

What to listen for. Students should be able to name what it eliminates — underdosing, failure to cross the barrier, target miss — and what it leaves open — species differences, redundancy, assay-to- disease mismatch, population heterogeneity, and the possibility that the receptor simply is not causally central in humans. The pedagogical goal is the habit of asking what a control rules out rather than treating it as generic reassurance. If a student says "so the drug worked, it just didn't work," praise that; it is exactly the paradox the chapter is built on, and unpacking it is the lesson.

Prompt 3 — "Substance P and CGRP are found in many of the same sensory neurons. Why did one produce a drug class and the other nothing for pain?"

What to listen for. Do not let the class resolve this too quickly. The surprise should be genuine before it is dissolved. When they get to §22.9's three differences, push on which one they think is doing the most work — there is no consensus answer and the argument is the point. Listen for whether anyone reaches for "better mechanism," and if they do, ask them to specify what would have been visible in 1995 that would have distinguished the two. Usually nothing was. That realization is the chapter's core.

The strongest answers rank the three and defend the ranking. My own view, offered only if the discussion stalls: the provocation study is the decisive one, because indication specificity and target accessibility both improve the odds of detecting an effect, while provocation is the only item that tells you the effect is there to detect.

Prompt 4 — "We understand narcolepsy type 1 better than almost any neurological disease, and we still cannot replace the missing peptide. What does that tell you about the rest of Part IV?"

What to listen for. Students should get to delivery quickly. What you want next is the comparison with insulin (Chapter 11): identical disease logic, opposite therapeutic outcome, and the difference is entirely whether the target tissue sits inside or outside the blood-brain barrier. Then push toward the asymmetry: antagonists were tractable because any small brain-penetrant molecule that occupies the site will do, while agonism requires inducing a conformational change. Listen for students who generalize this into a prediction — that CNS peptide blockade will keep producing drugs faster than CNS peptide replacement. That prediction is defensible and it is the right kind of thing to take out of the chapter.

Prompt 5 — "Someone shows you a supplement whose selling point is that elite soldiers have higher NPY. Walk me through your response — and be fair to the underlying science."

What to listen for. The fairness clause matters. The NPY stress literature is real and coherent, and a student who dismisses it as junk has learned the wrong lesson. You want: acknowledge the animal data and the human associations; identify the design as observational; enumerate the four causal readings; distinguish circulating from central; and land on the delivery question. Chapter 13's three-joint test is the scaffold, and Joint 3 — existence proof — is where it fails hardest and for a physical reason rather than merely an evidentiary one.

The best answers end by saying what would change their mind, specifically and in terms of measurements. If nobody does, ask for it directly; the ability to name your own falsifier is most of what this book is teaching.

Prompt 6 — "Erenumab is a migraine drug that works by targeting a peptide. It is not a peptide. Does it belong in this book?"

What to listen for. This is a deliberately destabilizing question and it works best at the end. The easy answer is yes-because-the-target-is-a-peptide. Push past it. What you want is for students to articulate what "understanding peptides" means if it does not reduce to "understanding peptide drugs": that peptides are the body's signaling vocabulary, and that the therapeutic opportunities they create include being blocked, being mimicked, being used as delivery addresses (Chapter 27), and being raised indirectly (Chapter 28) — of which only one produces a drug that is itself a peptide.

Listen for students who connect this back to Chapter 1's thesis that the word peptide carries no evidentiary weight. The two ideas are the same idea seen from different ends, and a student who sees that has understood the book's spine.