In Vivo Base Editing of PCSK9 with VERVE-102 for Hypercholesterolemia
Authors, Journal, Affiliations, Type, DOI
- Scott B. Vafai, Jörg Täubel, Thomas Ashdown, Riyaz S. Patel, Sadaf Diamondali, Jaimini Cegla, Handrean Soran, Bilal Bashir, Alexander Abitbol, Daniel Gaudet, Alex Lauzière, Liam R. Brunham, David E. Newby, Stephen J. Nicholls, Russell S. Scott, Jane Kerr, Jean-Claude Tardif, Catherine Lunken, Steve E. Humphries, Verena Karsten, Patrick D. Tyler, Xinyan Zhang, Nidal Huniti, Patrick A. Flight, Chelsey L. Jensen, Rick Falzone, Joseph C. Biedenkapp, Troy Lister, Leslie E. Stolz, Amit V. Khera, Sekar Kathiresan
- The New England Journal of Medicine (published online May 25, 2026)
- Verve Therapeutics (Eli Lilly subsidiary), Boston; study sites in Australia, Canada, New Zealand, United Kingdom
- Phase 1, open-label, single-ascending-dose study (Heart-2; NCT06164730); non-prespecified interim analysis
- DOI: 10.1056/NEJMoa2601283
Overview
VERVE-102 is a GalNAc-lipid nanoparticle (LNP) carrying mRNA encoding an adenine base editor (ABE8.8) and a guide RNA targeting the 5' splice site of PCSK9 intron 1. A single adenosine-to-guanosine substitution at this site disrupts splicing, enabling read-through to a stop codon that permanently prevents hepatic PCSK9 protein translation — mimicking the cardioprotective effect of naturally occurring PCSK9 loss-of-function variants. In 35 participants with heterozygous familial hypercholesterolaemia or premature coronary artery disease on maximal oral lipid-lowering therapy, VERVE-102 achieved dose-dependent and durable reductions in plasma PCSK9 (51–88%) and LDL-C (9–62%), with a mean absolute LDL-C reduction of 78 mg/dL at the 1.0 mg/kg dose. No dose-limiting toxicities or deaths occurred; the GalNAc-LNP formulation avoided the thrombocytopenia seen with predecessor VERVE-101.
Keywords
Base editing, PCSK9, LDL cholesterol, familial hypercholesterolaemia, lipid nanoparticle, GalNAc, adenine base editor, gene therapy, in vivo editing, Heart-2 study
Key Takeaways
Background and Rationale
- PCSK9 loss-of-function variants confer lifelong lower LDL-C and up to 88% lower lifetime CHD risk — providing genetic validation for PCSK9 as a therapeutic target
- Mendelian randomization shows the cardiovascular benefit of LDL-C lowering accrues with cumulative duration of exposure; a 5-year statin-driven LDL-C reduction of 39 mg/dL reduces MACE ~22%, while carriers of equivalent PCSK9 genetic variants have up to 88% lower CHD risk — confirming that early and durable LDL-C lowering yields disproportionate benefit
- Real-world discontinuation of lipid-lowering therapies (including PCSK9 inhibitors) reaches 30–50% within 12 months after initiation, and treatment non-persistence is associated with worse cardiovascular outcomes — motivating a single-dose permanent intervention
Mechanism of VERVE-102
- Delivery: mRNA encoding ABE8.8 adenine base editor + guide RNA encapsulated in GalNAc-LNP; GalNAc moiety binds hepatocyte asialoglycoprotein receptor (ASGPR); LNP also enters hepatocytes via ApoE-mediated LDL-receptor endocytosis; endosomal pH changes destabilize the LNP, releasing mRNA and guide RNA to the cytosol
- Editing: ABE-gRNA complex translocates to the nucleus; guide RNA aligns with the 5' end of PCSK9 intron 1; adenosine deaminase domain converts adenosine to inosine (read as guanosine) on the target strand; nicking of the opposite strand triggers repair, completing the A·T → G·C substitution; the resulting splice site disruption causes read-through to a stop codon → permanent loss of PCSK9 protein expression
- Preclinical specificity: Off-target screening in primary human hepatocytes showed editing specific to the PCSK9 target site; biodistribution in non-human primates showed high liver editing and low adrenal/spleen editing; no germline transmission in offspring of mice treated with the orthologous murine equivalent
Study Design and Population
- Phase 1, open-label, single-ascending-dose; 6 dose cohorts: 0.3, 0.45, 0.6, 0.7, 0.8, and 1.0 mg total RNA per kg body weight
- Eligibility: adults 18–70 years with HeFH or premature CAD (≤55 years in men; ≤65 years in women); fasting LDL-C ≥70 mg/dL on maximum tolerated oral therapy (statin ± ezetimibe); PCSK9 inhibitor users excluded
- Premedication: dexamethasone the day before; dexamethasone + H1 + H2 antihistamine on the day of infusion
- Data cutoff: February 27, 2026; 35 participants enrolled (4, 6, 4, 8, 6, 7 per cohort); median follow-up ~9 months; 15 participants had ≥1 year of follow-up
- Participant characteristics: mean age 52 years; 69% male; 83% HeFH (of whom 9 also had premature CAD); 6 (17%) premature CAD alone; mean baseline LDL-C 129 mg/dL; 91% on statin therapy (71% high-intensity)
Safety
- No dose-limiting toxic effects; no deaths; no withdrawals; all 35 participants received full planned dose
- Infusion-related reactions (IRRs): 7/35 (20%); all grade 1 or 2; 5 grade 1 (resolved without treatment); 2 grade 2 (treated with paracetamol or antihistamines); 1 grade 1 IRR led to a 40-minute infusion interruption
- Liver enzyme elevation: 3/35 had ALT elevations ≥2×ULN (maximum 2.4×ULN); all peaked on day 3–4 and resolved below 2×ULN by day 8 — transient and asymptomatic
- Serious adverse event (1): Grade 3 aspiration pneumonitis ~2 weeks post-infusion in 1 patient with GERD/sliding hiatal hernia; hospitalized and discharged the following day; site investigator assessed as unrelated to VERVE-102
- No events higher than grade 3; no thrombocytopenia (contrast with VERVE-101)
- LNP terminal half-life <20 hours (rapid systemic clearance)
Pharmacodynamics — PCSK9 Reduction (time-averaged, day 28 onward)
| Dose (mg/kg) | n | Mean PCSK9 Reduction | Range |
|---|---|---|---|
| 0.3 | 4 | −51% | −73 to −30% |
| 0.45 | 6 | −59% | −92 to −34% |
| 0.6 | 4 | −61% | −84 to −38% |
| 0.7 | 8 | −64% | −93 to −34% |
| 0.8 | 6 | −77% | −87 to −66% |
| 1.0 | 7 | −88% | −94 to −78% |
Pharmacodynamics — LDL-C Reduction (time-averaged, day 28 onward)
| Dose (mg/kg) | n | Mean LDL-C Reduction | Range | Absolute Reduction |
|---|---|---|---|---|
| 0.3 | 4 | −9% | −26 to +26% | — |
| 0.45 | 6 | −44% | −57 to −30% | — |
| 0.6 | 4 | −45% | −54 to −34% | — |
| 0.7 | 8 | −33% | −70 to −14% | — |
| 0.8 | 6 | −51% | −86 to −16% | — |
| 1.0 | 7 | −62% | −79 to −45% | −78 mg/dL (128→51 mg/dL) |
Durability
- PCSK9 and LDL-C reductions were stable from day 28 through the last available follow-up visit (maximum 18 months)
- Day-28 values were consistent with time-averaged values across follow-up, reflecting a stable pharmacodynamic response
- Of 35 participants, 15 had ≥1 year of follow-up — all showing sustained reductions
- Given that mature hepatocytes have an average lifespan of 200–300 days, stable reductions beyond this period indicate the edited gene persists through hepatocyte turnover
- Precedent for stable inactivation with other in vivo gene-editing medicines: documented to 3 years without waning effect (nexiguran ziclumeran for ATTR amyloidosis)
Exploratory Analysis
- Pearson correlation between total RNA dose administered (mg/kg × body weight) and LDL-C reduction: r = −0.68 — suggests total RNA may be a better predictor of response than weight-based dose alone, warranting further study
Discussion Highlights
- GalNAc-LNP advantage over VERVE-101: The predecessor VERVE-101 (conventional LNP without GalNAc) was associated with thrombocytopenia in prior studies; the GalNAc-LNP used in VERVE-102 did not produce this finding — supporting the concept that LNP formulation is the primary driver of acute adverse events in in vivo gene editing
- Near-complete hepatic PCSK9 suppression at 1.0 mg/kg: 88% PCSK9 reduction consistent with editing in a large proportion of hepatocytes; liver is the principal source of circulating PCSK9
- Efficacy comparable to long-term PCSK9 inhibitor therapy: Sustained PCSK9 mAbs/siRNA achieve 40–60% LDL-C reduction with ongoing dosing; VERVE-102 achieves equivalent or greater LDL-C lowering from a single infusion
- Projected ASCVD benefit: A 78 mg/dL absolute LDL-C reduction maintained over 20 years is predicted to reduce ASCVD risk by >50% in most patients with hypercholesterolaemia
- VERVE-201: The same GalNAc-LNP is being used in a second investigational medicine targeting ANGPTL3 (NCT06451770)
Limitations of the Document
- Phase 1 design: selected population (stable disease), inpatient monitoring for ≥2 days, premedication protocol — real-world safety may differ
- Non-prespecified interim analysis; no formal statistical testing performed
- Not powered for statistical effect size of pharmacodynamic response or cardiovascular outcomes
- Short follow-up for high-dose cohorts (1.0 mg/kg: maximum follow-up at data cut-off was <1 year for most participants); longest follow-up available in lowest dose cohorts
- Long-term safety including off-target editing and potential malignancy risk unknown; FDA mandates 15-year minimum follow-up for gene therapies; all participants expected to enter long-term follow-up study
- Predominantly White population; off-target screening conducted largely in cells from White donors — generalizability to other ancestries is uncertain
- Small sample: n=35 total; individual cohorts as small as n=4
Key Concepts Mentioned
- concepts/Lipid-Gene-Therapy — primary concept for PCSK9 base editing and VERVE-102
- concepts/CRISPR-Cas9-in-Channelopathies — base editing platform (adenine base editor ABE8.8)
- concepts/AAV-Gene-Delivery — contrast: lipid-target gene therapy uses LNP not AAV
- concepts/PCSK9-Inhibitors — pharmacological context; equivalent LDL-C lowering with ongoing therapy vs single-infusion permanent approach
Key Entities Mentioned
- entities/Familial-Hypercholesterolemia — primary therapeutic target population (HeFH + premature CAD)
- entities/ATTR-Amyloidosis — parallel LNP-delivered CRISPR precedent (nexiguran ziclumeran; stable inactivation to 3 years)
Wiki Pages Updated
wiki/sources/verve102-pcsk9-nejm-2026.md— created (this file)wiki/sourceindex.md— added entrywiki/wikiindex.md— updated Lipid-Gene-Therapy descriptionwiki/concepts/Lipid-Gene-Therapy.md— updated VERVE-102 PCSK9 section with full NEJM datawiki/concepts/CRISPR-Cas9-in-Channelopathies.md— updated VERVE-102 referencewiki/concepts/PCSK9-Inhibitors.md— updated Future Directions and Contradictionswiki/concepts/Familial-Hypercholesterolemia.md— added VERVE-102 in emerging therapies