Lipid Gene Therapy

Definition

Gene editing therapies (GET) targeting hepatically expressed lipid metabolism genes to achieve durable, potentially one-time reductions in atherogenic lipoproteins. Unlike RNA interference agents (inclisiran, siRNA) that require repeat dosing, GET delivers permanent genomic edits to hepatocytes via lipid nanoparticle (LNP) or GalNAc-conjugated delivery — exploiting the asialoglycoprotein receptor to ensure hepatic specificity.

Key Concepts

Scientific Rationale

• Lipid metabolism genes are ideal LNP-GET targets: (1) expressed exclusively or predominantly in hepatocytes; (2) loss-of-function variants in humans are naturally associated with lower CVD risk, providing genetic validation of the target (sources/gene-editing-acc-2026 — very high)
• Individuals with naturally occurring LOF variants in PCSK9 have lower LDL-C and fewer ASCVD events; those with LOF in ANGPTL3 have lower LDL-C + TG + lower CAD event rates; LPA variation is the primary determinant of Lp(a) levels (a causal risk factor)
• GalNAc conjugation ensures hepatic specificity via high-affinity binding to asialoglycoprotein receptors on hepatocytes — minimizes off-target effects in other organs (sources/gene-editing-acc-2026 — very high)

PCSK9 Gene Editing (Clinical)

• VERVE-102 / Heart-2 (Verve Therapeutics/Eli Lilly; Phase 1; NCT06164730) — NEJM 2026: GalNAc-LNP encapsulating mRNA-encoded adenine base editor (ABE8.8) + gRNA targeting PCSK9 splice site (5' intron 1); A→G base substitution disrupts splicing → read-through to stop codon → permanent hepatic PCSK9 inactivation; single IV infusion; n=35 (HeFH or premature CAD on max oral LLD); 6 dose cohorts 0.3–1.0 mg/kg; data cutoff Feb 27, 2026 (non-prespecified interim analysis) (sources/verve102-pcsk9-nejm-2026 — high; sources/gene-editing-acc-2026 — very high)
  • Safety: No dose-limiting toxicity, no deaths; 20% mild-moderate infusion-related reactions; transient ALT ≤2.4×ULN in 3/35 (peak day 3–4, resolved by day 8); no thrombocytopenia (improvement over predecessor VERVE-101); 1 grade 3 aspiration pneumonitis assessed as unrelated to drug; LNP t½ <20 hours
  • PCSK9 reduction (time-averaged from day 28): −51% at 0.3 mg/kg → −88% at 1.0 mg/kg (range −94 to −78%); 88% suppression consistent with near-complete elimination of hepatic PCSK9 production
  • LDL-C reduction: −9% at 0.3 mg/kg → −62% at 1.0 mg/kg (range −79 to −45%); absolute reduction −78 mg/dL (baseline 128 → achieved 51 mg/dL) at highest dose; comparable to long-term PCSK9 inhibitor therapy (40–60% on ongoing dosing) from a single infusion
  • Durability: Stable PCSK9 and LDL-C reductions to ≥18 months in participants with longest follow-up (n=15 with ≥1 year); day-28 values consistent with time-averaged values → edit persists through hepatocyte turnover (hepatocyte lifespan 200–300 days)
  • Projected ASCVD benefit: 78 mg/dL sustained LDL-C reduction over 20 years predicted to reduce ASCVD risk >50% in most hypercholesterolaemia patients
• YOLT-101 (Yoltech Therapeutics; Phase 1; NCT06461702): LNP-based PCSK9 gene silencing; showed >50% LDL-C reduction in small cohort with hypercholesterolemia (sources/gene-editing-acc-2026 — very high)
• ART002 (Accuredit Therapeutics; Phase 1): LNP-based PCSK9 gene editing; >50% LDL-C reduction in separate small cohort (sources/gene-editing-acc-2026 — very high)

ANGPTL3 Gene Editing (Clinical)

• BE3-Angptl3 (Chadwick et al.; murine proof-of-concept): 3rd-generation base editor + gRNA targeting murine Angptl3 Gln135; median 35% editing rate; no detectable off-target mutagenesis at 10 top predicted sites; −49% ANGPTL3, −31% TG, −19% LDL-C vs non-targeting gRNA control — validated durable base editing for ANGPTL3 in animal models (sources/angptl3-inhibition-tcm-2024 — high)
• CTX310 (CRISPR Therapeutics; Phase 1): LNP-encapsulated CRISPR-Cas9 editor targeting ANGPTL3; 15 patients with severe hyperlipidemias; few adverse events; at highest dose: mean LDL-C −49%, TG −55%; reductions in ANGPTL3 protein levels confirmed (sources/gene-editing-acc-2026 — very high)
• CTX310 — primary Phase 1 report (NEJM 2025;393:2119-30): first-in-human, single ascending IV dose across 0.1–0.8 mg/kg (estimated lean body weight) in 15 adults with dyslipidemia refractory to maximally tolerated therapy. No dose-limiting toxicities, no CTX310-related serious adverse events. Dose–response was clear only above 0.6 mg/kg: ANGPTL3 changed +9.6% (0.1), +9.4% (0.3), −32.7% (0.6), −79.7% (0.7) and −73.2% (0.8) at 30 days; the 0.8 mg/kg cohort (n=4) reached −48.9% LDL-C and −55.2% TG at 60 days, with individual LDL-C responses ranging from −86.6% to −19.5%. Background lipid therapy was unchanged throughout, isolating the effect to the edit (sources/CRISPR-Cas9-NEJM-2025 — high)
• This is the first peer-reviewed clinical demonstration of in vivo CRISPR-Cas9 editing of a lipid gene. The authors infer that extensive hepatocyte editing was probably achieved, and note that simultaneous substantial LDL-C and TG reduction is something no existing therapy delivers (sources/CRISPR-Cas9-NEJM-2025 — high)
• Safety profile of CTX310 specifically: infusion-related reactions in 20% (grade 2 back pain and nausea, managed by pausing the infusion), one allergic rash, and one transient grade 2 ALT/AST rise 3–5× baseline that peaked day 4 and normalised by day 14 without bilirubin or ALP elevation. One death occurred 179 days after the 0.1 mg/kg dose, reported as unrelated. Two serious adverse events occurred, neither attributed to the drug (sources/CRISPR-Cas9-NEJM-2025 — high)
• Dose–response discordance: the cohort with the deepest ANGPTL3 knockdown (0.7 mg/kg) had less lipid lowering than 0.8 mg/kg. With n=2 versus n=4 this is not definitive, but it is the first clinical evidence that plasma ANGPTL3 level may not be a complete surrogate for the downstream lipid effect (sources/CRISPR-Cas9-NEJM-2025 — high)
• VERVE-201 / Pulse-1 (Verve Therapeutics; Phase 1b; NCT06451770): mRNA-encoded adenine base editor + gRNA via GalNAc-LNP targeting ANGPTL3; for patients with homozygous FH and refractory hypercholesterolaemia (ANGPTL3 acts independent of LDL receptor — key advantage for HoFH where LDLR is defective) (sources/gene-editing-acc-2026 — very high)
• Mechanistic note: ANGPTL3 blocks lipolysis of TG-rich lipoproteins via LPL inhibition AND raises LDL-C through endothelial lipase inhibition of VLDL clearance; ANGPTL3 LOF lowers LDL-C through a mechanism independent of the LDL receptor — enabling treatment of HoFH (where PCSK9 editing is less effective due to absent/non-functional LDLR) (sources/angptl3-inhibition-tcm-2024 — high)

LPA / Apolipoprotein(a) Gene Editing (Preclinical → Early Clinical)

• CTX320 (CRISPR Therapeutics): Dose-dependent LPA editing in primary human hepatocytes in vitro; Lp(a) −94% in non-human primates in vivo at 7 months; Phase 1 in development (sources/gene-editing-acc-2026 — very high)
• VERVE-301 (Verve Therapeutics): Advanced to clinical development targeting LPA (sources/gene-editing-acc-2026 — very high)
• CRISPR Therapeutics also initiating Phase 1 for Lp(a) elevation + established CVD
• LPA encodes apolipoprotein(a), the unique protein component of Lp(a); variation in LPA copy number is the primary determinant of plasma Lp(a) — a strongly atherogenic and thrombogenic lipoprotein not reduced by statins or PCSK9 inhibitors

APOC3 Gene Editing (Early Clinical)

• CorrectSequence Therapeutics: Base editing trial for APOC3-directed GET in familial chylomicronemia syndrome (a monogenic severe hypertriglyceridemia due to LOF in lipoprotein lipase/related genes where APOC3 inhibition reduces TG-rich lipoprotein levels) (sources/gene-editing-acc-2026 — very high)

Angiotensinogen (AGT) Gene Editing — Refractory Hypertension

• CTX340 (CRISPR Therapeutics): LNP-based GET targeting hepatically expressed AGT — master regulator of the renin-angiotensin-aldosterone system; advanced to clinical development for refractory hypertension (sources/gene-editing-acc-2026 — very high)
• First GET targeting the RAAS rather than a lipid target

Primary Prevention Implications

• PCSK9, ANGPTL3, and LPA GET represent potential primary prevention tools — somatic editing in high-risk but asymptomatic individuals before disease manifestation could provide lifelong protection from ASCVD with a single intervention
• Proof-of-concept in animal models + early human Phase 1 data support profound and sustained lipid lowering
• Extending to primary prevention requires more stringent safety evaluation and long-term registries (sources/gene-editing-acc-2026 — very high)

Contradictions / Open Questions

• No long-term safety data: All clinical GET trials for lipid targets are Phase 1 (PCSK9, ANGPTL3) or preclinical (LPA/APOC3); no CV outcomes trial exists; off-target editing causing malignancy may emerge over decades — FDA requires 15-year minimum follow-up for gene therapies; VERVE-102 Heart-2 participants will enroll in a 15-year long-term follow-up study (sources/gene-editing-acc-2026 — very high; sources/verve102-pcsk9-nejm-2026 — high)
• MAGNITUDE hepatotoxicity precedent: The MAGNITUDE Phase 3 pause (TTR target, also LNP-delivered CRISPR) for severe hepatotoxicity including one death raises the question of whether LNP-CRISPR hepatotoxicity is target-specific (TTR) or a class effect that could manifest with PCSK9/ANGPTL3/LPA editing at larger scale (sources/gene-editing-acc-2026 — very high)
• PCSK9 editing vs. inclisiran/evolocumab — durability vs. safety trade-off uncharacterised: Inclisiran (GalNAc-siRNA) achieves ~50% LDL-C reduction with twice-yearly SC dosing without permanent genomic change; PCSK9 inhibitors (evolocumab/alirocumab) have >5-year CV outcomes data. Whether the one-time convenience of permanent PCSK9 base editing justifies the unknown long-term genomic risks versus well-characterised existing agents remains unanswered (sources/gene-editing-acc-2026 — very high)
• HoFH and LDLR-dependent strategies: PCSK9 editing (which works by sparing LDL receptor from degradation) is less effective in HoFH patients with absent/severely dysfunctional LDLR; ANGPTL3 editing (LDLR-independent) is therefore the rationale for HoFH-focused trials — comparative efficacy not yet established (sources/gene-editing-acc-2026 — very high)
• Germline transmission risk: Somatic editing of hepatocytes should not affect germline, but rigorous exclusion is required; current trials preferentially enroll adults past reproductive age or use contraception protocols (sources/gene-editing-acc-2026 — very high)
• Plasma target protein may be an incomplete surrogate (CTX310): The 0.7 mg/kg cohort achieved the deepest ANGPTL3 knockdown (−79.7%) but only −21.0% LDL-C and −19.2% TG, whereas 0.8 mg/kg (−73.2% ANGPTL3) roughly doubled the lipid effect. With n=2 and n=4 this is not conclusive, but if confirmed it means editing efficiency and downstream lipid benefit are decoupled — with implications for using target-protein reduction as the surrogate for dose selection and for regulatory approval (sources/CRISPR-Cas9-NEJM-2025 — high)
• Durability of a permanent edit remains unproven in humans: CTX310's premise is a lifelong loss-of-function mutation from a single infusion, yet the primary report covers only 60–90 days. Hepatocyte turnover, immune clearance of edited cells, or incomplete editing of the hepatocyte pool could all erode the effect over years, and none can be assessed at this follow-up. This is the key evidentiary gap separating GET from siRNA, where reversibility is a safety advantage (sources/CRISPR-Cas9-NEJM-2025 — high)
• No head-to-head comparison within the GET class: CTX310 (Cas9 nuclease, LNP) and VERVE-201 (adenine base editor, GalNAc-LNP) both target ANGPTL3, and both CTX310 and evinacumab achieve comparable lipid reductions — but no trial compares modalities. Whether base editing's avoidance of double-strand breaks translates into a safety advantage is untested (sources/CRISPR-Cas9-NEJM-2025 — high; sources/gene-editing-acc-2026 — very high)

Connections

• Related to concepts/CRISPR-Cas9-in-Channelopathies — editing platform
• Related to concepts/Gene-Editing-Risk-Benefit-Framework — patient selection; primary prevention expansion
• Related to concepts/AAV-Gene-Delivery — contrast: lipid targets use LNP not AAV
• Related to concepts/Gene-Silencing-Therapy — contrast: GET permanent vs. siRNA/ASO repeated dosing
• Related to concepts/Dyslipidemia-Management — pharmacological context
• Related to entities/Familial-Hypercholesterolemia — primary therapeutic target
• Related to entities/ATTR-Amyloidosis — parallel LNP-GET application; MAGNITUDE hepatotoxicity precedent
• Related to entities/CTX310 — first in vivo CRISPR-Cas9 lipid gene-editing product with primary clinical data
• Related to entities/CRISPR-Therapeutics — developer of CTX310, CTX320 and CTX340

Sources

• sources/verve102-pcsk9-nejm-2026
• sources/gene-editing-acc-2026
• sources/angptl3-inhibition-tcm-2024
• sources/CRISPR-Cas9-NEJM-2025