One CRISPR Infusion Held Cholesterol Down for a Year

A single infusion of CTX310 cut LDL cholesterol 52.5% at one year in the four patients on the highest dose of a 15-person CRISPR trial. Triglycerides fell 47.8% in the same group, and investigators reported no treatment-related serious adverse events.

The deeper issue is whether a liver gene edit still works after the organ replaces itself. Steven E. Nissen, M.D., of Cleveland Clinic, said adult liver cells turn over fast enough that an uncopied edit could fade. In this trial, the lipid curves stayed flat.

Four Patients at the Top Dose Still Had Half the LDL

CTX310 is a lipid-nanoparticle shot of CRISPR-Cas9 messenger RNA and a guide RNA aimed at ANGPTL3 in the liver. Switching that gene off is meant to raise the activity of enzymes that clear fats from blood. Cleveland Clinic cardiologist Luke J. Laffin, M.D., led the one-year Cleveland Clinic follow-up with Stephen J. Nicholls, M.B., B.S., Ph.D., of the Victorian Heart Institute, as co-first author. CRISPR Therapeutics funded the work.

Phase 1a enrolled 15 adults from May 2024 through August 2025 with uncontrolled high LDL, moderate or severe high triglycerides, or mixed dyslipidemia, all on the strongest lipid drugs they could tolerate. Median age was 53, and 87% were men. Forty percent had clinical atherosclerotic disease, 40% had familial hypercholesterolemia, 13% had severe high triglycerides, and 40% had mixed dyslipidemia. Mean LDL was 155 mg/dL. Median triglycerides were 192 mg/dL.

Patients received one of five doses, 0.1, 0.3, 0.6, 0.7, or 0.8 mg/kg of estimated lean body weight, after steroids and antihistamines. At 60 days, the 0.8 mg/kg group had already shown a 48.9% LDL drop and a 55.2% triglyceride drop. The new question was whether those shifts would last a full year.

Among the four people at 0.8 mg/kg, they did. ANGPTL3 protein was 78.6% lower than baseline (from 89.0% to 63.1% lower). LDL was 52.5% lower (from 84.2% to 24.4% lower). Triglycerides were 47.8% lower (from 77.6% to 14.7% lower). Other atherogenic markers moved with them.

HIGHEST DOSE AT 12 MONTHS

Those means come from four people. Laffin warned against reading the spread as a finished efficacy map. “When you have small numbers, like four patients, there’s going to be some variability,” he said. The averages, he added, are a starting point for larger studies that separate severe high triglycerides from familial hypercholesterolemia and mixed dyslipidemia.

The Liver Kept Copying the CRISPR Edit

Gene editors sell a one-time rewrite. The liver does not stand still while they wait. Nissen, senior investigator and chief academic officer of Cleveland Clinic’s Heart, Vascular and Thoracic Institute, put the biology in plain terms after the 12-month visit.

When you edit a gene, it raises the question of whether the effect is temporary or permanent because of cell turnover. Hepatocytes have a turnover rate of nearly 20% per year, and the average age of adult liver cells is less than three years regardless of a person’s age. So what if some of the cells that were not altered are replicating? Does that mean the effect of the gene editing goes away or is diminished? We really need to know if this is a permanent alteration that can enable once-in-a-lifetime treatment. The good news is that the levels of ANGPTL3 and lipids remained low and flat through the full year, which suggests we’re editing the genes so efficiently that when the hepatocytes replicate, the genetic change we’ve induced is replicated along with them.

Steven E. Nissen, M.D., Cleveland Clinic

If unedited cells had taken over, ANGPTL3 and LDL would have drifted back toward baseline. They did not. Laffin called that pattern encouraging for other CRISPR programs, because it suggests a one-and-done shot can survive the organ it is meant to change. That is a larger claim than a 12-month cholesterol chart, and it still rests on 15 people.

Laffin said the practical problem in high-risk patients is often staying on therapy, not starting it, because daily pills demand lifelong adherence. A copied liver edit is the proposed answer to that gap. It is also a bet that cannot be walked back with a missed dose.

People Born Without ANGPTL3 Already Live This Biology

CTX310 is trying to copy a human mutation. ANGPTL3 is a liver protein that blocks lipoprotein lipase and endothelial lipase, two enzymes that break down circulating fats. People who inherit loss-of-function copies have low LDL, low triglycerides, and less coronary disease, without a clear toxic syndrome attached to the missing protein.

In a 2017 analysis of up to 21,980 people with coronary disease and 158,200 controls, about 1 in 309 people carried a heterozygous loss-of-function mutation. Carriers had 17% lower triglycerides and 12% lower LDL, and 34% lower odds of coronary disease, than people without the mutation. Three people with complete ANGPTL3 deficiency had no coronary plaque on CT angiography, unlike first-degree relatives who still made the protein.

That natural experiment is why this target exists. It is also why a skeptic line that “the body needs cholesterol” does not land cleanly here. Cells still make cholesterol. The edit is aimed at a brake on fat clearance, not at cholesterol synthesis. Later genetic papers using common variants have been less sure that modest ANGPTL3 shifts cut events, which is a real tension, and it is why a 15-person lipid study cannot stand in for a outcomes trial.

The same biology already supports a marketed antibody, evinacumab, approved in 2021 for homozygous familial hypercholesterolemia. In volunteers, that antibody cut fasting triglycerides by as much as 76% and LDL by as much as 23%. In homozygous FH, LDL fell about 43%. Those patients still need repeat dosing. CTX310 is the attempt to bake the same pathway into liver DNA once.

What CTX310 Changes That Daily Pills Do Not

Statins, ezetimibe, and PCSK9 shots already lower LDL. Inclisiran, a twice-yearly RNA shot, already stretches the calendar. CTX310 is not competing on a 12-week LDL chart. It is competing on two other axes: a dual LDL-and-triglyceride hit, and a dose you do not take again.

PCSK9 drugs, and PCSK9 gene editors, mainly move LDL. ANGPTL3 sits upstream of both LDL and triglyceride-rich particles, which is why the 0.8 mg/kg group also saw triglyceride-rich lipoprotein cholesterol fall 71.7%. That dual profile is the product difference versus a PCSK9 rewrite. It is also why mixed dyslipidemia, not only familial hypercholesterolemia, was in this first cohort.

A rival path is already in print. Verve Therapeutics, now part of Eli Lilly, reported a PCSK9 base-editing trial of 35 people in which the top VERVE-102 dose cut PCSK9 protein 88% and LDL 62%, with follow-up of at least a year in 15 participants. That program uses a base editor, which changes a single DNA letter, rather than Cas9 cutting both strands. An earlier Verve construct, VERVE-101, was dropped after liver-enzyme rises and low platelets. Verve is also building VERVE-201, a base editor against ANGPTL3.

So the field now has two one-shot ideas in the same patients: cut PCSK9, or cut ANGPTL3. The first is a proven LDL target with existing antibody and RNA drugs. The second is a dual-lipid target whose antibody is already used in the rarest FH. CTX310’s 12-month package is CRISPR Therapeutics’ bid to own the second door with a nuclease, not a base editor.

Dose still decides whether that door opens. The dose-by-dose 12-month lipid changes presented at the European Society of Cardiology meeting show a steep protein knockdown at mid doses, and a much patchier LDL effect until 0.8 mg/kg.

ALL DOSES, MEAN CHANGE AT 12 MONTHS

The 0.7 mg/kg pair knocked ANGPTL3 down almost as far as the top dose, yet LDL barely moved. The 0.6 mg/kg group cut triglycerides in half with almost no LDL change. That is what a mixed Phase 1a looks like, and it is why Phase 1b is splitting patients by diagnosis instead of pooling them.

Fifteen People, One Sudden Death, and a Wide Range

Safety was the primary endpoint. No dose-limiting toxic effect related to CTX310 was reported. CRISPR Therapeutics said on August 28, 2026, that there were no treatment-related serious adverse events, no grade 3 or higher liver-enzyme changes, and no new treatment-related adverse events since the 60-day update.

Two serious events had already been logged in the first paper. One patient had a spinal disk herniation seven months after the 0.3 mg/kg dose. The other, a man in his 50s with familial hypercholesterolemia, prior stents, and bypass surgery, died suddenly 179 days after the 0.1 mg/kg dose after collapsing while walking. Investigators judged both events unrelated. One of the two serious events was later called a preexisting condition. No new serious events appeared in the extra follow-up.

SAFETY THROUGH ONE YEAR

• Related serious events: None of the 15 patients had a serious adverse event judged related to CTX310.
• Infusion reactions: Three patients (20%) had infusion-related reactions, including fever, nausea, and back pain at the 0.6 and 0.8 mg/kg doses.
• Liver enzymes: One patient (7%) with high enzymes at baseline had a transient rise to three to five times baseline on the highest dose, peaking on day 4 and back by day 14, without a bilirubin or clotting-time spike.
• Any related event: Seven patients (47%) had an adverse event deemed related to the drug; 14 (93%) had any investigator-reported event.

The LDL range at the top dose, from 84.2% lower to 24.4% lower, is wide enough that a single extra nonresponder would move the mean. Off-target cuts and later cancer risk are the reasons the U.S. Food and Drug Administration wants 15 years of watch, Nissen said, not because anyone expects a new LDL reading next spring.

Phase 1b Splits the Next Cohort by Lipid Disorder

Phase 1a is complete. Phase 1b (NCT07491172) is already giving a flat regimen aimed at the 0.8 mg/kg exposure, in separate groups defined by lipid disorder rather than a mixed queue. Laffin said that design is the one that can tell whether people with severe high triglycerides, heterozygous FH, and mixed dyslipidemia respond the same way.

THE CTX310 CALENDAR

1. May 2024: Enrollment opens in the Phase 1a ascending-dose trial (ACTRN12623000809639).
2. August 2025: The last of 15 adults receives a single CTX310 infusion.
3. November 8, 2025: Sixty-day data go to the American Heart Association meeting and the New England Journal of Medicine.
4. August 28, 2026: Twelve-month durability is presented at the ESC meeting and published as a research letter dated August 28, 2026.

Presented at #ESCCongress:

In a phase 1a trial, one-time CRISPR-Cas9 gene editing with CTX310 was associated with no treatment-related dose-limiting toxic effects and with sustained reductions in ANGPTL3 and atherogenic lipids at 1 year. Full trial results:… pic.twitter.com/HDxseqSADG

— NEJM (@NEJM) August 28, 2026

Laffin’s group has already heard from patients who want into the next cohorts. The original 15 stay in a 15-year safety watch, in line with FDA guidance for gene-editing drugs. That clock, not the 52.5% mean, is what will decide whether a copied ANGPTL3 edit becomes a one-time heart drug or a carefully watched experiment that lowered numbers in four people.

Frequently Asked Questions

How Long Does a CTX310 Infusion Take?

The Phase 1a protocol ran the intravenous infusion over 1 to 4.5 hours after corticosteroid and antihistamine premedication. Doses were set by estimated lean body weight, not total body weight, which is why the milligram-per-kilogram figures are not a simple scale-up from a bathroom scale.

What Is Familial Combined Hypolipidemia?

It is the inherited state of complete ANGPTL3 loss, first tied to the gene by exome sequencing in families with very low LDL, HDL, and triglycerides. The three people with that full deficiency who underwent CT angiography in the 2017 genetics study had no coronary plaque, which is the human knockout CTX310 is trying to mimic in liver cells only.

Where Was the Phase 1a Trial Run?

Investigators listed six sites in New Zealand, Australia, and the United Kingdom, including New Zealand Clinical Research in Christchurch, Aotearoa Clinical Trials in Auckland, Royal Adelaide Hospital, and the Victorian Heart Institute in Melbourne. Cleveland Clinic’s coordinating center in Cleveland ran the analysis. No U.S. dosing site was named in the 12-month letter.

Can Doctors Undo the ANGPTL3 Edit If Something Goes Wrong?

The trial did not test a reversal dose. CTX310 is built to create a lasting loss-of-function mutation in hepatocytes, and the 12-month curves are being read as evidence that the change is copied when those cells divide. If a late harm appears, the FDA-mandated 15-year follow-up is the detection plan, not an off switch.

Does CTX310 Lower HDL Cholesterol Too?

Yes. At 0.8 mg/kg, HDL cholesterol was 23.5% lower at one year, with a range from 44.9% lower to no change. That matches the natural ANGPTL3-deficiency pattern, in which all major lipid fractions fall, and it is one reason larger trials will have to track whether the HDL drop offsets any gain from lower LDL and triglycerides.

Disclaimer: This article is news reporting on a published Phase 1a clinical trial and related genetic studies. It is for information only and is not medical advice, a treatment recommendation, or a prediction of individual results. Readers should not start, stop, or change cholesterol, triglyceride, or gene-editing care based on this piece; those decisions belong with a licensed physician who knows the patient’s history, current medicines, and lipid disorder. Trial figures, safety findings, and follow-up plans reflect the papers and institutional statements cited here and may change as Phase 1b and the 15-year watch continue.