One Shot, Lifelong Cholesterol Control: What Lilly's VERVE-102 Data Mean for the Future of Cardiovascular Medicine

Cardiovascular disease kills more people every year than any other cause. The pharmacological response to that reality has been, for decades, a daily pill. Statins, taken every morning for the rest of a patient's life, have saved millions of lives and remain the backbone of LDL cholesterol management. But they come with a fundamental limitation that no amount of clinical success has resolved: patients stop taking them. Approximately half of all patients prescribed lipid-lowering therapy discontinue it within a year. The biology of the disease does not pause when the pill bottle runs out.

On May 25, 2026, Eli Lilly announced positive Phase 1b results from the Heart-2 clinical trial of VERVE-102, an investigational in vivo base editing medicine developed by Verve Therapeutics under a collaboration agreement with Lilly. The data represent something the cardiovascular field has been working toward for years: a single infusion that permanently silences the PCSK9 gene in the liver, producing durable reductions in LDL cholesterol without requiring the patient to do anything else. If the results hold through later-stage trials, the implications for how cardiovascular disease is managed could be profound.

What Base Editing Actually Does

VERVE-102 is not a gene therapy in the traditional sense, and the distinction matters. Conventional gene therapies typically work by delivering a functional copy of a missing or defective gene. Base editing, by contrast, makes a precise chemical change to a single letter of the DNA code, converting one nucleotide base to another without cutting the double helix. In the case of VERVE-102, an adenine base editor targets the PCSK9 gene in liver cells, introducing a change that mimics naturally occurring loss-of-function variants observed in people who have lifelong low LDL cholesterol and dramatically reduced rates of heart disease. The edit is permanent. The liver continues producing the modified gene product indefinitely, and PCSK9 protein levels fall accordingly.

The delivery mechanism is equally important. VERVE-102 uses a proprietary GalNAc-LNP, a lipid nanoparticle engineered to target liver cells specifically via two distinct receptor pathways. That liver specificity is what makes the approach clinically viable. Earlier attempts at in vivo gene editing for cardiovascular targets were limited by delivery challenges and off-target safety concerns. The GalNAc-LNP design is intended to address both, and the Heart-2 safety data so far are consistent with that goal.

What the Heart-2 Data Show

The Phase 1b Heart-2 trial enrolled 14 adults with heterozygous familial hypercholesterolemia or premature coronary artery disease, two populations defined by their need for deep and durable LDL reduction. Participants received a single intravenous infusion of VERVE-102 across three weight-based dose cohorts: 0.3 mg/kg, 0.45 mg/kg, and 0.6 mg/kg.

The efficacy signal was dose-dependent and clinically meaningful. At the 0.6 mg/kg dose, the mean time-averaged reduction in LDL cholesterol from baseline was 53 percent, with a maximum reduction of 69 percent observed in one participant. Across all 14 participants, PCSK9 protein levels fell in parallel with LDL, confirming that the gene editing mechanism is working as intended. When the data were analyzed by total RNA dose rather than weight-based dose, the dose-response relationship became even cleaner: participants who received 50 to 60 mg of total RNA achieved a mean LDL reduction of 59 percent, with each of the three participants in that range achieving greater than 50 percent reduction.

The safety profile was reassuring. There were no treatment-related serious adverse events, no dose-limiting toxicities, and no clinically significant changes in liver enzymes or platelets at any dose level. One participant experienced a Grade 2 infusion-related reaction that resolved with acetaminophen. For a first-in-class genetic medicine delivered intravenously, that profile is notable. The liver enzyme signal, in particular, has been a concern with other LNP-delivered genetic medicines, and the absence of meaningful ALT or AST elevations in Heart-2 is a meaningful data point for the field.

The FDA has granted VERVE-102 Fast Track designation for LDL reduction in patients with hyperlipidemia and high lifetime cardiovascular risk. Lilly plans to initiate a Phase 2 clinical study by the end of 2026.

The Adherence Problem That No Pill Has Solved

To understand why a one-time intervention for LDL control matters, it helps to sit with the scale of the adherence problem. Statins are among the most prescribed drugs in the world. They are inexpensive, generally well-tolerated, and supported by decades of outcomes data. And yet real-world adherence remains stubbornly poor. The reasons are familiar: side effects, pill fatigue, cost, the difficulty of maintaining daily habits for a disease that produces no symptoms until it kills you. PCSK9 inhibitors, the injectable biologics that followed statins, achieve deeper LDL reductions but require injections every two to four weeks and carry significant cost barriers that limit their use to a fraction of the patients who could benefit.

A single infusion that permanently lowers LDL by 50 to 70 percent sidesteps the adherence problem entirely. The patient does not need to remember a pill. The effect does not wear off when the prescription lapses. For patients with familial hypercholesterolemia, who face a lifetime of elevated cardiovascular risk beginning in childhood, the prospect of a one-time intervention that resets their LDL trajectory is qualitatively different from anything currently available.

What Comes Next and What Remains Uncertain

The Heart-2 data are early. Fourteen patients across three dose cohorts, with a data cutoff that captures relatively short follow-up, cannot establish long-term durability, rare safety signals, or the effect size needed to power a cardiovascular outcomes trial. The Phase 2 study that Lilly plans to initiate by year-end will need to demonstrate that the LDL reductions seen in Heart-2 are reproducible at scale, that the safety profile holds across a broader and more diverse population, and that the editing effect persists over years rather than months.

Durability is the central question. The base editing mechanism is designed to be permanent, and preclinical data from VERVE-101, an earlier program targeting the same pathway, have shown sustained LDL reductions out to nearly two years in animal models. But translating that durability to humans requires clinical confirmation, and the field has learned repeatedly that preclinical signals do not always survive contact with human biology at scale.

There is also the question of what happens to liver cells that have been edited. Hepatocytes turn over slowly, but they do turn over. Whether the edited cell population maintains its proportion over decades, or whether unedited cells gradually repopulate the liver and dilute the effect, is a question that only long-term follow-up can answer. The cardiovascular field will be watching the durability data from Heart-2 with particular attention.

A Glimpse of a Different Future

The broader significance of VERVE-102 is not just about cholesterol. It is about what becomes possible when genetic medicine matures enough to address common, chronic diseases rather than only rare, monogenic conditions. The PCSK9 gene is one of the best-validated targets in cardiovascular medicine. The biology is clear, the clinical benefit of lowering PCSK9 is established, and the patient population is enormous. If base editing can deliver durable LDL reduction from a single infusion in that population, it establishes a template for applying the same approach to other cardiovascular targets, including ANGPTL3 and LPA, which Verve is already pursuing in parallel programs.

The cardiovascular field has been waiting for a treatment paradigm that matches the permanence of the disease. Atherosclerosis does not take days off. The daily pill has been the best available answer for decades. The Heart-2 data suggest that a better answer may be coming, and that the question is no longer whether base editing can work in humans, but how well it works and for how long.