Cardiology · Medical Literature
In Vivo Base Editing of PCSK9 with VERVE-102 for Hypercholesterolemia
Abstract
New England Journal of Medicine, Ahead of Print.
Clinical relevance
In adults with heterozygous familial hypercholesterolemia (HeFH) or premature coronary artery disease (CAD) on maximally tolerated lipid-lowering therapy, does a single intravenous infusion of VERVE-102 — an in vivo PCSK9-targeting adenine base editor delivered via a GalNAc-lipid nanoparticle — produce safe, durable, and dose-dependent reductions in circulating PCSK9 protein and LDL cholesterol? __SNAPSHOT__ {"studyDesign":"Phase 1b, open-label, single-ascending-dose study (interim analysis; trial designation: Heart-2; NCT06164730)","setting":"Multi-site international study; specific country/institution mix not fully reported in abstract — includes sites in the United Kingdom, Canada, and other regions based on trial registration","sampleSize":"n=35 (interim analysis; 6 dose cohorts: 0.3, 0.45, 0.6, 0.7, 0.8, and 1.0 mg/kg total RNA); full planned enrollment up to 85 participants","population":"Adults with heterozygous familial hypercholesterolemia (HeFH; 57% HeFH alone, 26% HeFH with premature CAD, 17% premature CAD alone) on maximally tolerated statin therapy with or without ezetimibe. Mean age 52 years; 69% male. Baseline lipid-lowering: high-intensity statin 71%, moderate/low-intensity statin 20%, no statin 9%; ezetimibe 43%.","intervention":"VERVE-102: a single intravenous infusion of an mRNA encoding an adenine base editor protein combined with a guide RNA targeting hepatic PCSK9, encapsulated in a GalNAc-containing lipid nanoparticle, administered at one of six ascending doses (0.3–1.0 mg total RNA/kg body weight) over 2–4 hours","comparator":"Not applicable (single-arm, open-label dose-escalation study; no control or placebo arm)","primaryOutcome":"Safety and tolerability (phase 1 primary endpoint); pharmacodynamic endpoints include change from baseline in circulating PCSK9 protein and LDL cholesterol levels","followUp":"All 35 participants followed for at least 28 days; subset of 15 participants followed for ≥12 months (maximum approximately 18 months) as of the February 27, 2026 data cut-off. Median follow-up approximately 9 months. Participants to be enrolled in long-term follow-up study for up to 15 years.","fundingSource":"Industry (Verve Therapeutics, a wholly owned subsidiary of Eli Lilly)"}
Key findings
Across all six dose cohorts, a single infusion of VERVE-102 produced dose-dependent reductions in circulating PCSK9 protein and LDL cholesterol. At the lowest dose (0.3 mg/kg), mean PCSK9 reduction was 51% with a corresponding mean LDL-C reduction of 9% by day 28. At the highest dose (1.0 mg/kg), mean PCSK9 reduction reached 88% with a mean time-averaged LDL-C reduction of 62% and a mean absolute LDL-C reduction of 78 mg/dL. Pearson's correlation coefficient between total RNA administered and LDL-C reduction was -0.68, indicating a strong dose-response relationship (calculated from reported interim data). Among the 15 participants with ≥12 months of follow-up, reductions in PCSK9 and LDL-C established at day 28 appeared broadly durable across the observation window. No dose-limiting toxicities were reported. No treatment-related serious adverse events were reported; one serious adverse event (grade 3 aspiration pneumonitis in a patient with pre-existing GERD/hiatal hernia) was deemed unrelated to treatment and resolved with overnight observation. Infusion-related reactions occurred in approximately 20% of patients (grade 1–2). No clinically significant changes in ALT, AST, bilirubin, or platelets were observed at any dose. Formal p-values, confidence intervals, and NNT/NNH are not reported in the available abstract.
Practice implications
IMPORTANT LIMITATION: Only the abstract is available; the full dataset, pre-specified statistical analyses, and complete adverse event tables are accessible solely to subscribers of the full NEJM publication. Conclusions are therefore drawn from the abstract and corroborating press release/congress data. What this study established: VERVE-102 produced large, dose-dependent reductions in PCSK9 and LDL-C with a single infusion, and early safety data are reassuring — particularly when contrasted with its predecessor VERVE-101, which caused grade 3 ALT elevations and thrombocytopenia at higher doses using a different LNP formulation. LDL-C is an FDA-accepted surrogate endpoint in this population, lending pharmacodynamic credibility to the findings. The durability signal at up to 18 months is mechanistically plausible given permanent genomic editing. What this study did NOT establish: Cardiovascular event reduction, hard clinical outcomes, long-term safety beyond 18 months, off-target editing rates in vivo, or safety in higher-risk subgroups (e.g., homozygous FH, severe renal/hepatic disease, elderly patients, pregnancy). The study was not powered for efficacy comparisons. Whether the achieved LDL-C reductions are truly lifelong remains unknown. The 9% LDL-C reduction at the lowest dose raises questions about optimal dosing for patients with LDLR-deficient HeFH who may rely more on ASGPR-mediated uptake. Generalizability: Findings apply only to adults with HeFH or premature CAD already on maximally tolerated statins ± ezetimibe in a selected, predominantly male (69%), middle-aged cohort. Patients with HoFH, severe comorbidities, or those not on background statin therapy are not represented. Practice implications: Results do not yet support clinical use. Phase 2 data with cardiovascular outcomes or expanded safety follow-up are required before any practice change is appropriate. FDA Fast Track designation has been granted. __LIMITATIONS__ Interim analysis of an ongoing phase 1 open-label single-arm study — no control group, no randomization, no blinding; subject to reporting bias and placebo effect on subjective endpoints.,Very small sample size (n=35 at interim); individual cohorts contain only 3–9 patients, severely limiting statistical precision and adverse event detection.,Absence of a comparator arm precludes attribution of LDL-C changes solely to VERVE-102 in a population also receiving background lipid-lowering therapy.,Surrogate endpoint (LDL-C): no cardiovascular event data; clinical benefit remains unproven.,Median follow-up only ~9 months; only 15 participants have ≥12 months of data; long-term durability, off-target editing, and late adverse events (including oncogenicity risk from off-target base editing) are unknown.,Industry sponsorship (Verve Therapeutics/Eli Lilly) with multiple authors as company employees, raising potential for reporting and interpretation bias.,Open-label design and heterogeneous baseline lipid-lowering therapy complicate clean pharmacodynamic interpretation.,Predominantly male cohort (69%); limited representation of older adults, women, and patients with non-LDLR-mediated HeFH.,Formal p-values, confidence intervals, and pre-specified primary statistical analyses are not reported in the available abstract.,Off-target genomic editing rates, bystander editing, and immunogenicity data are not reported in the abstract.,Context of prior VERVE-101 program pause (grade 3 ALT elevation and thrombocytopenia) underscores that LNP-related hepatotoxicity is a class concern requiring sustained vigilance. __BOTTOM_LINE_CATEGORY__ promising_but_preliminary __BOTTOM_LINE__ A single infusion of VERVE-102, an in vivo PCSK9-targeting adenine base editor delivered via a GalNAc-LNP, produced dose-dependent reductions in LDL-C of up to 62% (mean absolute 78 mg/dL) with early durability at up to 18 months in a small phase 1 cohort of patients with HeFH or premature CAD, with no dose-limiting toxicities or treatment-related serious adverse events reported. These are early-stage, interim, uncontrolled findings in a highly selected population, and cardiovascular outcomes, long-term safety (including off-target editing), and durability beyond 18 months remain unestablished. Clinicians should recognize this as a scientifically compelling but very preliminary proof-of-concept signal that requires phase 2/3 randomized trial confirmation before any change in practice is warranted.