Chapter 10 — Quiz

Twenty-two questions.

Multiple choice

1. FLOW studied semaglutide in adults with: a) obesity alone b) type 2 diabetes and chronic kidney disease c) heart failure d) MASH

2. FLOW was: a) stopped early for futility b) stopped early for efficacy c) completed as planned d) never started

3. Trials stopped early for efficacy tend to: a) underestimate the effect b) overestimate the effect c) estimate it precisely d) be uninterpretable

4. MASH was previously called: a) NAFLD b) NASH c) cirrhosis d) hepatitis C

5. The principal problem with a liver biopsy endpoint is that it is: a) expensive b) invasive, subject to sampling error and reader variability, and still a surrogate c) unavailable d) not reproducible at all

6. HFpEF accounts for approximately what share of heart failure cases? a) 10% b) a quarter c) about half d) 90%

7. The HFpEF trials in this chapter used primarily: a) mortality endpoints b) hospitalization endpoints c) symptom and physical function endpoints d) biopsy endpoints

8. Sleep apnea benefit is best explained by: a) a direct airway drug effect b) reduced soft tissue around a collapsible airway — a mechanical, weight-mediated effect c) anti-inflammatory action d) improved sleep architecture

9. The apnea-hypopnea index measures: a) oxygen saturation b) apnea and hypopnea events per hour of sleep c) total sleep time d) snoring volume

10. Semaglutide for Alzheimer's disease is rated: a) ✅ b) ⚠️ c) ❌ d) 🔬

11. The reason for that rating is that supporting evidence is currently: a) contradictory b) mechanistic, preclinical, and observational rather than randomized clinical c) fabricated d) sufficient

12. Confounding by indication means: a) the drug has multiple indications b) people who receive a treatment differ systematically from those who do not, in ways related to the outcome c) the indication was mislabeled d) two drugs were compared

13. The addiction hypothesis originated from: a) a pharmaceutical target program b) animal studies c) unprompted patient reports d) a regulator

14. GLP-1 agonists for addiction are rated: a) ✅ b) ⚠️ c) ❌ d) 🔬

15. "Anti-inflammatory" is described as a weak mechanistic claim because: a) inflammation is not real b) it can explain benefit in any disease, and therefore explains benefit in none until specified c) markers cannot be measured d) it applies only to arthritis

16. Upward compression is: a) treating ✅ indications as licensing 🔬 ones b) treating 🔬 indications as discrediting ✅ ones c) rounding a number up d) combining two doses

17. A single overall rating for this drug class would: a) be efficient and accurate b) have to be wrong about at least four of the ten claims c) be impossible to compute d) match the label

Short answer

18. State the four candidate explanations for the breadth of effects, and what would distinguish each.

19. Explain why early stopping is ethically motivated and statistically costly.

20. State the difference between ⚠️ and 🔬 as this book uses them.

Applying the rating discipline

21. Sleep apnea is ✅ and its mechanism is probably "just weight loss." Explain why the mechanism does not weaken the rating.

22. Explain what upward and downward compression have in common, and name the rule that prevents both.

Answer key **1.** b. **2.** b. **3.** b. **4.** b. **5.** b. **6.** c. **7.** c. **8.** b. **9.** b. **10.** d. **11.** b. **12.** b. **13.** c. **14.** d. **15.** b. **16.** a. **17.** b. **18.** **① Weight loss is upstream of many diseases** — distinguished by whether comparable weight loss achieved by other means produces comparable benefit. **② Direct tissue effects at receptors in heart, kidney, vasculature, and immune cells** — distinguished by whether benefit appears before meaningful weight loss or in people who lose little. **③ Anti-inflammatory action as a common thread** — distinguished by whether benefit tracks inflammatory markers better than it tracks weight, which requires mediation analysis and is therefore weak evidence. **④ Some of these will not replicate** — distinguished by time, and by larger, independent, longer trials. All four are probably operating in proportions nobody can specify. **19.** **Ethically motivated:** if an interim analysis shows a treatment is clearly working, continuing to assign half the participants to placebo becomes difficult to justify. **Statistically costly:** the trials that cross an interim boundary are enriched for results where random variation happened to favor the treatment — the same selection logic as Phase 2 optimism. So a truncated trial's point estimate is likely larger than the truth and its interval wider than a completed trial's would have been. The direction remains trustworthy; the magnitude should be held loosely. **20.** **⚠️ Promising but preliminary** requires **real human data that does not yet settle the question** — Phase I/II trials, small or short RCTs, mixed results, or surrogate endpoints only. **🔬 Frontier** is for **early-stage science proceeding properly** where mechanism may be established but clinical translation is unproven and it is simply too soon to rate. The Alzheimer's and addiction programs sit at 🔬 because their supporting evidence is mechanistic, preclinical, and observational rather than randomized clinical — however well-motivated the rationale. **21.** Because the rating attaches to a **claim about an outcome**, not to a mechanism. The claim is "tirzepatide reduces apnea-hypopnea index in adults with moderate-to-severe obstructive sleep apnea and obesity," and randomized trials with a standard objective endpoint established it. **How** it works does not bear on **whether** it works. The mechanism does have two legitimate consequences, and neither is a weakening. First, it makes a prediction: comparable weight loss by other means should produce comparable benefit, which is testable. Second, it constrains extrapolation — the drug should not be expected to help sleep apnea that is not weight-related. **Naming the mechanism sharpens the claim rather than diminishing it**, and Chapter 2's warning runs the other way: mechanism cannot upgrade a rating, and it cannot downgrade one either. **22.** Both attach a rating to a **molecule** rather than to a **claim**. Upward compression treats established indications as licensing unestablished ones ("proven for the heart, so probably right for the brain"). Downward compression treats unestablished indications as discrediting established ones ("they claim it treats everything, so I don't believe the heart result"). The claims rest on entirely different evidence bases and neither inference is available. **The rule that prevents both is rule 6: one molecule, many ratings** — a rating attaches to a claim, with a population and an endpoint, never to a molecule in the abstract.