Peter Attia MD407 ‒ Preventing cardiovascular and Alzheimer’s disease: lowering LDL early, APOE4, and more
CHAPTERS
- 0:01 – 2:17
Michael Davidson’s origin story: family tragedy → lipidology and prevention career
Peter introduces Michael Davidson, a cardiologist and lipidologist at the University of Chicago. Davidson explains how his father’s fatal heart attack at 47 shaped his lifelong focus on lipids, trials, and primary prevention.
- •Runs a prevention-focused lipid clinic; long career in clinical trials
- •Family history (father died at 47) as motivation for early prevention
- •Early work: niacin trials, omega-3 research, and statin trials
- •Transition into biotech startups later in his career
- 2:17 – 7:55
Primordial prevention: why lowering LDL early matters more than 10-year risk
The discussion reframes prevention away from short time horizons and toward lifetime exposure. Davidson argues that preventing plaque formation early is easier than trying to reverse advanced disease later.
- •Primordial prevention: stop plaque before it forms
- •Lifetime LDL exposure concept (“eight gram rule”)
- •Keeping LDL <80 mg/dL across life as a practical target
- •Using genetics, CAC, plaque imaging, Lp(a), CRP to personalize early action
- 7:55 – 15:36
Causality framework: treating LDL like smoking, blood pressure, and diabetes
Peter and Michael argue LDL should be treated as a causal driver independent of near-term risk, similar to smoking cessation or hypertension treatment. They explore why LDL-lowering remains more controversial than other risk-factor interventions.
- •Alan Sniderman framing: treat causal drivers, not just 10-year risk
- •Analogy: you don’t ‘wait’ to stop smoking once risk becomes high
- •LDL is treated differently than BP/diabetes despite similar causality logic
- •Perceived lifestyle-fixability and guideline conservatism slow adoption
- 15:36 – 20:45
CETP biology 101: what CETP does and why HDL-raising once looked like the holy grail
They introduce CETP (cholesteryl ester transfer protein) and why inhibiting it raises HDL while lowering LDL. The conversation reviews evolutionary speculation and the historic belief that raising HDL would directly reduce events.
- •CETP transfers cholesterol esters from HDL to apoB particles; inhibition raises HDL and lowers LDL
- •Animal observations: CETP presence aligns with more atherogenic profiles
- •Evolutionary hypothesis: CETP as a cholesterol-conservation mechanism
- •Mechanistic story for HDL: reverse cholesterol transport, ‘garbage truck’ functions
- 20:45 – 28:29
Pfizer’s torcetrapib failure: off-target toxicity (aldosterone/BP) sinks a blockbuster
Torcetrapib—the first high-profile CETP inhibitor—was stopped after increased mortality. Davidson explains why the failure was viewed primarily as off-target adrenal effects rather than proof the target was wrong.
- •Torcetrapib raised HDL ~75% but lowered LDL only ~10–15%
- •Trial halted for increased cardiovascular and non-cardiovascular mortality
- •Mechanism: off-target adrenal stimulation → aldosterone/steroid effects and BP rise
- •Lesson: molecule-specific toxicity, not necessarily target invalidation
- 28:29 – 34:12
The CETP inhibitor ‘graveyard’: dalcetrapib and evacetrapib miss outcomes (and why timing mattered)
Next-generation CETP programs failed to show MACE benefit, deepening skepticism. Davidson argues at least one program (evacetrapib) may have been stopped too early given modest LDL reductions and the time needed for curves to separate.
- •Roche dalcetrapib: weak CETP inhibition, minimal LDL effect, no MACE benefit
- •Lilly evacetrapib: large HDL rise + modest LDL drop; stopped at ~2 years for futility
- •Comparison to ezetimibe ACS trial: benefits may emerge only after >2 years
- •Class signal: CETP inhibitors repeatedly showed diabetes-risk reduction
- 34:12 – 37:14
Merck’s REVEAL trial rescues the class: LDL lowering predicts event reduction
Merck’s large REVEAL outcomes trial demonstrated that CETP inhibition could reduce events in line with expected benefit from absolute LDL lowering. The drug didn’t proceed commercially due to extreme lipophilicity and long tissue persistence.
- •REVEAL: ~30,000 patients, ~4 years, baseline LDL ~60 mg/dL
- •~11 mg/dL absolute LDL reduction → ~9% relative risk reduction as predicted
- •Validated principle: benefit tracks absolute LDL/non-HDL lowering
- •Drug abandoned due to fat deposition, very long washout, environmental concerns
- 37:14 – 40:25
LDL is causal: why outcome effects are predictable across mechanisms
Peter and Michael connect Mendelian randomization, epidemiology, and RCTs showing consistent event reduction per unit LDL lowering. They reject ‘pleiotropic effects’ as the main explanation for statin benefit.
- •Rule of thumb: ~22% RR reduction per 1 mmol/L (≈38.8 mg/dL) LDL drop
- •Relationship is largely linear and mechanism-agnostic
- •Key is absolute LDL lowering and sufficient duration for separation
- •Implication: CETP benefit should be predictable if LDL lowering is robust
- 40:25 – 49:54
Obicetrapib is revived at NewAmsterdam: potency, TULIP data, and abandoning the statin-intolerance route
Davidson explains how obicetrapib was acquired (post-Amgen shelving) and why the company pursued a full LDL-lowering development program. Early phase 2 data showed far greater potency than prior CETP inhibitors.
- •Amgen shelved obicetrapib after deprioritization; NewAmsterdam resurrected it
- •TULIP phase 2: ~40–50% LDL lowering at low doses; HDL up ~150%
- •Initial plan: statin intolerance indication; FDA skepticism forced a broader program
- •Mendelian randomization clarified: HDL raising isn’t the causal lever—LDL lowering is
- 49:54 – 56:07
Phase 2/3 program and PREVAIL: ROSE, BROOKLYN, BROADWAY, TANDEM, and the approval pathway
They review the clinical program in FH and ASCVD populations, including combination strategies and early signals on MACE. Davidson outlines why Europe may approve based on LDL lowering while the U.S. is likely to require outcomes (PREVAIL).
- •ROSE/ROSE2: strong LDL reductions as mono/combination therapy; combo with ezetimibe improves potency
- •BROOKLYN (FH): ~40% additional LDL lowering on top of intensive background therapy
- •BROADWAY (ASCVD): ~35% LDL lowering and an encouraging (non-powered) MACE signal over ~1 year
- •PREVAIL: global, event-driven outcomes trial; Europe vs U.S. regulatory differences
- 56:07 – 1:04:42
Biomarker nuances: LDL-C vs LDL-P vs apoB discordance, plus strong Lp(a) lowering
Peter probes discordance seen with CETP inhibition across LDL-C, LDL-P, and apoB. Davidson discusses possible explanations and highlights a notable Lp(a)-lowering effect that could differentiate obicetrapib clinically.
- •Observed discordance: apoB lowering can appear smaller than LDL-C/LDL-P lowering
- •LDL-P assay used: LipoScience/LabCorp NMR; small LDL particles drop dramatically
- •Hypotheses: particle-size effects and mass-based apoB measurement quirks
- •Lp(a) lowering around ~50%—larger than typical PCSK9 inhibitor effect
- 1:04:42 – 1:12:37
Where it fits in practice: combo therapy over high-dose statins, and the reality of U.S. drug pricing
They argue for moving away from maximal-dose statins toward combination therapy that achieves targets with fewer liabilities. The discussion then turns to PBM-driven pricing dynamics that can determine access more than list price.
- •Case against high-dose statins: diminishing LDL benefit, higher side-effect/diabetes signal
- •Obicetrapib as a strong add-on to statins; combo with ezetimibe to reach PCSK9-like LDL lowering
- •Potential advantages: lowers small LDL particles, reduces diabetes risk, lowers Lp(a)
- •Pricing/access constraints: PBM rebate incentives can punish ‘too low’ pricing
- 1:12:37 – 1:29:12
Alzheimer’s rationale: APOE4 + CETP genetics, animal models, and why HDL may matter in the brain
Davidson explains the Alzheimer’s thesis: CETP loss-of-function appears protective in aging cohorts, especially among APOE4 carriers, and CETP expression worsens pathology in mouse models. They discuss how brain lipid handling differs from peripheral lipid metabolism.
- •Genetic signals: CETP loss-of-function enriched in longevity cohorts; mitigates APOE4 Alzheimer’s risk
- •Mouse models: adding CETP worsens amyloid-related cognitive decline; CETP inhibition improves behavior
- •Brain lipid biology: blood–brain barrier separates apoB lipoproteins from brain cholesterol handling
- •ApoE4 mechanism: impaired astrocyte cholesterol efflux/lipidation → inflammatory cascade → amyloid/tau
- 1:29:12 – 1:50:32
Biomarkers and prevention strategy: CSF sterols, p-tau 217 signals, and the challenge of proving clinical benefit
They summarize early target-engagement findings (CSF sterols/antioxidants) and describe a BROADWAY substudy showing favorable shifts in Alzheimer’s blood biomarkers, especially in APOE4 homozygotes. The chapter closes with skepticism about biomarker-only inference and why prevention trials must start before MCI.
- •Pilot: reduced CSF 24- and 27-hydroxycholesterol; antioxidant changes; exploratory cognitive observations
- •BROADWAY biomarker analysis: p-tau 217 and other markers improved most in older APOE4 (especially 4/4)
- •Debate: p-tau 217 as potential surrogate endpoint vs risk of being misled (HDL ‘lesson’)
- •Strategic pivot: intervene pre-MCI; plan prevention-style trials rather than dementia-stage treatment
- 1:50:32 – 1:57:48
Fish oil update: EPA vs DHA trial lessons and the bottleneck of delivering DHA to the brain
Davidson reviews why REDUCE-IT (EPA-only) showed benefit while STRENGTH (EPA+DHA) did not, and discusses placebo issues (mineral oil vs corn oil). They then pivot to the emerging science of DHA transport into the brain and why conventional supplementation may not address APOE4 brain DHA deficits.
- •STRENGTH (EPA+DHA) negative overall; possible signal in secondary prevention subgroup
- •REDUCE-IT benefit likely real though magnitude debated; mineral oil placebo may have worsened inflammation/statin absorption
- •Triglyceride lowering alone doesn’t guarantee outcomes (e.g., pemafibrate PROMINENT)
- •New focus: Mfsd2a transporter prefers lyso-PC DHA—potential path to improve brain delivery
- 1:57:48 – 2:07:24
Two open problems: AI to reform clinical trials and the statin–diabetes signal
They explore how AI might reduce trial cost/time by analyzing adherence and treatment exposure rather than intention-to-treat alone, noting regulatory constraints. They also discuss uncertainty around mechanisms behind statin-associated insulin resistance and how combination therapy could mitigate risk.
- •AI idea: model ‘on-treatment’ efficacy and reduce sample sizes—blocked by current regulatory norms
- •Clinical trials are getting harder as background therapies improve and event rates fall
- •Statin–diabetes signal: small but real; mechanisms unclear and likely dose/age/weight dependent
- •Preference for lower statin doses plus add-on therapy to preserve benefit and reduce harm
- 2:07:24 – 2:19:15
Drug development realities and closing reflections: biotech funding, klotho, APOE4 guidance, and avoiding over-testing
Davidson details the capital intensity of drug development and the pressures shaping biotech and pharma pipelines. They close with klotho’s promise and the practical question of what APOE4 carriers can do today—balancing motivation with the risks of chasing noisy biomarkers.
- •Cost to approval: billion+ for one asset (plus prior sunk costs); industry-wide average even higher including failures
- •Big pharma patent cliffs and acquisition dependence keep biotech relevant despite hard funding cycles
- •Klotho program: strong genetics/animal data, uncertain MOA; pushing toward first-in-human studies
- •For APOE4: prioritize fundamentals (exercise, metabolic health, BP, lipids); avoid anxiety-driven over-testing; multiple therapies will likely be needed
