PCSK9 inhibitors: clinical evidence and implementation

Authors, Journal, Affiliations, Type, DOI

• Author: Marc S. Sabatine
• Journal: Nature Reviews Cardiology, Vol 16, March 2019 (published online November 2018)
• Affiliation: TIMI Study Group, Division of Cardiovascular Medicine, Brigham and Women's Hospital and Harvard Medical School, Boston, MA
• Type: Review article
• DOI: https://doi.org/10.1038/s41569-018-0107-8

Overview

This comprehensive review by the PI of the FOURIER trial traces the genetic origins of PCSK9 discovery through its role in LDL receptor degradation, the pharmacology of approved monoclonal antibodies (evolocumab, alirocumab), and the results of three major cardiovascular outcome trials. Both approved PCSK9 mAbs reduce LDL-C by approximately 60% on top of statin therapy and significantly reduce major adverse cardiovascular events with no offsetting safety concerns. The evidence supports targeting LDL-C to ≤20 mg/dL, far below current guideline recommendations of 70 mg/dL, and emerging siRNA therapy (inclisiran) promises twice-yearly dosing with comparable efficacy.

Keywords

PCSK9, LDL cholesterol, LDL receptor, familial hypercholesterolemia, evolocumab, alirocumab, cardiovascular outcomes, FOURIER, ODYSSEY, inclisiran, siRNA, Mendelian randomization

Key Takeaways

PCSK9 Biology and Genetic Evidence

• The PCSK9 gene was identified via linkage analysis in families with autosomal dominant hypercholesterolaemia without LDLR or APOB mutations; gain-of-function mutations cause FH via accelerated LDL receptor destruction
• PCSK9 is secreted by hepatocytes and binds the EGF-A domain of the LDL receptor; the PCSK9–LDL receptor complex is internalized and directed to the lysosome for destruction rather than receptor recycling (LDL receptors normally recycle ~100 times)
• Cohen et al. identified individuals with nonsense PCSK9 mutations with LDL-C ~40% lower; subsequent ARIC analysis showed PCSK9 loss-of-function variants lowering LDL-C by 28% were associated with an 88% reduction in CHD risk
• Mendelian randomization (n=112,772; 14 studies; 14,120 CV events): PCSK9 and HMGCR variants yield identical OR of 0.81 per 10 mg/dL LDL-C reduction — confirming that PCSK9 inhibition should yield the same clinical benefit per unit LDL-C lowering as statins
• The concept of cumulative LDL-C burden (LDL-C × years) explains why lifelong exposure to lower LDL-C (as in loss-of-function variant carriers) confers greater protection than any clinical trial can demonstrate over 2–5 years

Evolocumab (Monoclonal Antibody)

• Fully human mAb; subcutaneous; 140 mg Q2 weeks or 420 mg monthly — both produce similar ~60% LDL-C reduction
• Effective whether given as monotherapy, added to statin, in statin-intolerant patients, or in HeFH (Phase III: MENDEL-2, LAPLACE-2, GAUSS-2/3, RUTHERFORD-2)
• HoFH response is genotype-dependent: two defective LDLR alleles → 47% reduction; one defective + one negative allele → 25%; two negative alleles → no response; APOB mutation → 47%; compound heterozygote with PCSK9 GoF + LDLR negative → 6% reduction
• Reduces Lp(a) by ~25% (vs statins which may increase Lp(a)); no effect on CRP; reduces APOB by ~50% (similar to high-intensity statin)
• No neutralizing antibody development (fully human)

Alirocumab (Monoclonal Antibody)

• Fully human mAb; subcutaneous; dose-titrated (75 mg Q2W → 150 mg Q2W if LDL-C ≥70 mg/dL) or fixed 150 mg Q2W or 300 mg Q4W
• Titrated regimen: 45–50% LDL-C reduction; maximum 150 mg Q2W achieves ~60% reduction
• HeFH: 40–60% LDL-C reduction; HoFH: 7% rise to 64% decrease depending on genotype
• Phase III program included ODYSSEY MONO, COMBO I/II, ALTERNATIVE trials

PCSK9 Inhibitors vs Statins — Comparison

FOURIER Trial (Evolocumab)

• n=27,564 patients with clinically evident ASCVD (prior MI, nonhaemorrhagic stroke, or symptomatic PAD); LDL-C ≥70 mg/dL on optimised statin therapy; median follow-up 2.2 years
• Evolocumab lowered LDL-C by 59% at 48 weeks (median absolute LDL-C 30 mg/dL; reduction of 56 mg/dL from baseline of 92 mg/dL)
• Primary endpoint (CV death/MI/stroke/coronary revascularisation/UA hospitalisation): HR 0.85 (95% CI 0.79–0.92)
• Main secondary endpoint (CV death/MI/stroke): HR 0.80 (95% CI 0.73–0.88)
• Components: fatal/nonfatal MI −21%; fatal/nonfatal stroke −21%; coronary revascularisation −27% (urgent) vs −13% (elective)
• No effect on cardiovascular death or UA hospitalisation (likely attributable to short follow-up per statin trial data)
• Landmark analysis: risk reduction in CV death/MI/stroke was 16% in year 1 and 25% after year 1 — consistent with time-dependent translation of LDL-C lowering
• Benefit consistent in patients with baseline LDL-C <70 mg/dL: achieved LDL-C 21 mg/dL → CV death/MI/stroke HR 0.70 (95% CI 0.48–1.01) — 30% reduction
• Safety: no excess myalgias, cataracts, new-onset DM (including prediabetes subgroup), neurocognitive events, aminotransferase elevation, or haemorrhagic stroke
• EBBINGHAUS substudy (n=1,974): no neurocognitive differences on Cambridge Neuropsychological Test; no association between LDL-C levels and cognitive changes
• Only excess: injection-site reactions (0.2%/year excess); no neutralizing antibodies
• PAD subgroup (n=3,642; 13.2%): CV death/MI/stroke 13.0% vs 7.6% (vs non-PAD); evolocumab → 27% relative risk reduction (similar to non-PAD); greater absolute benefit (3.5% vs 1.4%); 42% reduction in major adverse limb events (acute limb ischaemia/major amputation/urgent peripheral revascularisation; HR 0.58; P=0.009)
• Very high-risk subgroups (recent MI, multiple prior MIs, multivessel CAD): baseline risk 40–100% higher; relative risk reductions twice that of lower-risk patients (21–30% vs 11–16%); absolute risk reduction >3% at 3 years vs ~1%

SPIRE-1 and SPIRE-2 Trials (Bococizumab) — Terminated

• n=27,438 combined; bococizumab 150 mg Q2W vs placebo
• Terminated prematurely: bococizumab is humanized (not fully human) → neutralizing antibodies developed → LDL-C reduction attenuated from 59% at 14 weeks to 38% at 2 years
• SPIRE-2 (longer follow-up, higher baseline LDL-C): HR 0.79 (P=0.02) for primary composite; directionally consistent individual components
• Injection-site reactions >8× more frequent with bococizumab vs placebo
• Lesson: fully human mAb architecture is essential for sustained efficacy

ODYSSEY Outcomes Trial (Alirocumab)

• n=18,924 patients; enrolled 1–12 months after ACS hospitalisation; LDL-C ≥70 mg/dL on high-intensity statin; median follow-up 2.8 years
• Alirocumab titrated to achieve LDL-C 25–50 mg/dL; stopped if persistently <15 mg/dL
• LDL-C reduction: 57% at 4 weeks; attenuated to 36% at end of study (due to protocol-mediated down-titration)
• Primary endpoint (CHD death/MI/ischaemic stroke/UA hospitalisation): HR 0.85 (95% CI 0.78–0.93; P=0.003)
• Nominal 15% all-cause mortality reduction (HR 0.85, 95% CI 0.73–0.98) — not statistically significant in testing hierarchy (tested after CHD death and CV death, both NS)
• Non-monotonic pattern in LDL-C subgroups: patients with lower baseline LDL-C had less benefit — but this was attributable to protocol-mediated alirocumab dose reduction in that subgroup
• Safety profile comparable to FOURIER; injection-site reactions ~0.6%/year excess (higher than FOURIER)

Putting PCSK9 Inhibition in Perspective

• Relative cardiovascular risk reduction is a function of: (1) absolute amount of LDL-C lowering and (2) duration of therapy
• Meta-analysis of statin and non-statin trials: relative risk per 1 mmol/L LDL-C reduction = 0.77 for statins; 0.75 for non-statin LDL-C lowering — absolute LDL-C reduction accounts for 98% of between-study variability in risk reduction
• Time dependence: in year 1 of statin therapy, RRR per mmol/L = 9%; grows to 22–24% in subsequent years; 20% overall (based on median ~5-year statin trials)
• In FOURIER/ODYSSEY Outcomes (median 2.2–2.8 years), clinical benefits are consistent with what the time-dependent LDL-C reduction model would predict — not inferior
• Absolute risk reductions in PCSK9i trials (~1.5–2.0% over 2–3 years) extrapolate to ~3% at 5 years — comparable to statin trials (CARE, LIPID, HPS, TNT: ~2–4% at 5 years)

New LDL-C Targets

• Benefit of LDL-C lowering is consistent down to LDL-C levels <10 mg/dL (FOURIER observational analysis): monotonic dose-response relationship between achieved LDL-C and cardiovascular outcomes
• No association between achieved LDL-C and safety outcomes (transaminases, CK, cancer, cataracts, haemorrhagic stroke, new-onset DM, neurocognitive events) across the full range
• IMPROVE-IT (ezetimibe) and HPS3/TIMI55-REVEAL (anacetrapib) both showed benefit in patients with median LDL-C <70 mg/dL at baseline
• Evidence supports targeting LDL-C to ≤20 mg/dL as a reasonable strategy — far below current guideline target of 70 mg/dL

Future Directions — PCSK9 siRNA (Inclisiran)

• Inclisiran: chemically synthesized siRNA covalently linked to three GalNAc residues → hepatocyte-selective uptake via asialoglycoprotein receptor → engages RISC → cleaves PCSK9 mRNA
• siRNA-RISC complex is catalytic (cleaves many transcripts); single RISC complex effect is durable
• ORION-1 Phase II: 300 mg Q6 months → average ~50% LDL-C reduction over 210 days; no increase in serious adverse events; injection-site reactions 7%
• Major advantage: Q6-month dosing vs Q2-week or monthly for mAbs
• ORION-4 Phase III CVOT (HPS-4/TIMI 65; planned n=15,000; 5-year duration) initiated
• Other approaches: PCSK9 peptide vaccine (murine models showing cholesterol and atherosclerosis burden reduction); CRISPR-Cas9 gene editing of PCSK9 (murine data); potential for single-intervention permanent LDL-C lowering

Limitations of the Document

• Author conflict of interest: Marc Sabatine was the PI of FOURIER; review reflects optimistic interpretation of evolocumab data
• Short follow-up of CVOTs: median 2.2 years (FOURIER) and 2.8 years (ODYSSEY) vs ~5 years for statin landmark trials; cardiovascular mortality benefit may be underestimated
• Bococizumab data incomplete: both SPIRE trials terminated early; SPIRE-1 followed for only 7 months — limited interpretation
• ODYSSEY results preliminary at time of writing: presented at ACC but not yet peer-reviewed/published; data from conference presentation only
• HoFH data from small trials: TESLA Part B n=49 for evolocumab in HoFH; open-label TAUSSIG interim; limited power for genotype-specific conclusions
• Cost and access: PCSK9 inhibitors are expensive; cost-effectiveness analysis referenced but not deeply explored; cost-benefit analysis changes over time with price reductions
• No head-to-head data: no RCT comparing evolocumab vs alirocumab; cross-trial comparisons only
• No CV death benefit in FOURIER: attributed to short follow-up but not proven; requires longer trials
• Inclisiran data at time of writing limited to Phase II (ORION-1); no CVOT outcomes available

Key Concepts Mentioned

• concepts/PCSK9-Inhibitors — primary subject of the review
• concepts/Familial-Hypercholesterolemia — FH genetics and PCSK9 gain-of-function mutations; HoFH dosing
• concepts/Dyslipidemia-Management — LDL-C target debate, PCSK9i place in treatment algorithm
• concepts/Lipid-Gene-Therapy — inclisiran siRNA, future CRISPR-Cas9 approaches
• concepts/ASCVD-Risk-Assessment — high-risk subgroup identification for optimal PCSK9i use
• concepts/PAD-Medical-Therapy — FOURIER PAD subgroup: 42% major adverse limb events reduction
• concepts/Lipoprotein-a — Lp(a) reduced ~25% by PCSK9 mAbs

Key Entities Mentioned

• entities/Evolocumab — FOURIER trial; HoFH genotype-specific response; PAD benefit
• entities/Alirocumab — ODYSSEY Outcomes; titration scheme; nominal mortality signal
• entities/Inclisiran — ORION-1 Phase II siRNA; Q6-month dosing; future CVOT

Wiki Pages Updated

• Created: wiki/sources/pcsk9-inhibitors-nrc-2018.md
• Created: wiki/concepts/PCSK9-Inhibitors.md
• Created: wiki/entities/Evolocumab.md
• Created: wiki/entities/Alirocumab.md
• Created: wiki/entities/Inclisiran.md
• Updated: wiki/concepts/Dyslipidemia-Management.md
• Updated: wiki/concepts/Familial-Hypercholesterolemia.md
• Updated: wiki/wikiindex.md
• Updated: wiki/sourceindex.md
• Updated: log.md