Tirzepatide produces roughly 21% mean weight loss in adults with obesity after 72 weeks of treatment (Jastreboff et al., 2022, NEJM). That is a pharmacological achievement without precedent. It is also — for now — a weekly subcutaneous injection that costs over $1,000 per month in the United States, and that most patients must continue indefinitely to maintain the benefit. Stop the drug, and weight returns. Across several extension studies, roughly two-thirds of the lost weight comes back within a year of discontinuation (Rubino et al., 2022, JAMA).
This is the central tension of modern metabolic medicine. We finally have drugs that work, but the business model, the adherence data, and the biology all point in the same uncomfortable direction: these therapies require lifelong commitment from patients, payers, and health systems. A natural question follows. If a single intervention could deliver a durable fraction of the benefit — no weekly injection, no refrigeration, no prior authorization every year — would we take it?
Verve Therapeutics has already shown the answer is technically possible for at least one metabolic target. In 2023–2024, VERVE-102 demonstrated that a single intravenous dose of a base editor, delivered by lipid nanoparticle, could durably reduce LDL cholesterol by up to 69% in patients with familial hypercholesterolemia. The edit is designed to be permanent. If the same strategy could be pointed at an obesity target, the calculus of metabolic medicine would change fundamentally. This article asks whether that is plausible, what targets exist, and what stands in the way.
The Verve Precedent: A Proof of Principle
Before we can talk about a one-shot obesity therapy, we have to be specific about what Verve actually did — because the architecture of that program is the template.
VERVE-102 is an adenine base editor. It uses a modified Cas9 that cannot cut DNA, fused to a deaminase that converts a single adenine to a guanine at a precise location — in this case, introducing a stop codon in the PCSK9 gene in liver hepatocytes. The editor is delivered as mRNA packaged inside a lipid nanoparticle targeted to the liver via its natural apoE-LDLR tropism. One intravenous infusion. No viral vector. No permanent genomic integration of DNA. Just a transient exposure to an editor that makes a permanent change.
The clinical results — first presented at AHA 2023 and expanded through 2024 — showed dose-dependent PCSK9 protein reductions of up to 83% and LDL-C reductions of 53% to 69% in treated patients with heterozygous familial hypercholesterolemia. Follow-up has remained stable out to 12 months and beyond. Serious adverse events have been limited, though the field is watching long-term safety data carefully.
What makes VERVE-102 matter for this discussion is not the LDL reduction. It is that a single IV infusion produced a durable pharmacodynamic effect that mimics the phenotype of humans with natural loss-of-function mutations in PCSK9. Verve turned human genetics into a one-shot drug.
What Would a One-Shot GLP-1 Equivalent Look Like?
Applying the Verve template to obesity requires a target gene where durable loss of function produces metabolic benefit without unacceptable side effects. The target doesn't have to be GLP1R — GLP-1 is the peptide, not the gene you want to edit. The point is to find a gene whose modulation produces the downstream metabolic phenotype that GLP-1 drugs create. Several candidates have emerged from large-scale human genetics studies.
| Candidate gene | Pathway | Human genetic evidence | Phenotype of LoF carriers | Therapeutic direction | Stage |
|---|---|---|---|---|---|
| INHBE | Inhibin βE, liver-secreted hepatokine | Akbari et al., 2022, Nature — LoF in ~1:600 | Lower waist-to-hip ratio, lower T2D risk | Knockdown | Preclinical / target validation |
| GPR75 | Orphan GPCR | Akbari et al., 2021, Science — LoF in ~1:3,000 | Lower BMI (~1.8 kg lower) | Knockdown | Regeneron preclinical |
| GIPR | GIP receptor | Killian et al., 2018 — partial LoF variants | Lower BMI, improved lipids | Knockdown (debated) | Competing with agonism |
| MC4R | Melanocortin-4 receptor | Well-established; LoF causes obesity | Obesity in LoF; GoF protective (rare) | Activation (setmelanotide model) | Harder — activation by edit is difficult |
| ANGPTL3 | Lipid metabolism | LoF → low lipids, lower cardiovascular risk | Favorable lipids | Knockdown | Verve VERVE-201 |
| LEPR | Leptin receptor | LoF causes severe obesity | Obesity | Activation | Very difficult |
Two patterns jump out. First, the most tractable edits are loss-of-function: base editors are much better at knocking genes out than at activating them. Second, genes where "less is better" — INHBE, GPR75, ANGPTL3 — are the natural first targets. MC4R and LEPR are the opposite: you want more signaling, which base editing cannot easily deliver.
INHBE: The Target That Could Change Everything
Of the candidates, INHBE is the one to watch. It encodes inhibin βE, a hepatokine secreted by the liver in response to nutrient intake. In the 2022 paper in Nature, a Regeneron-led team led by Akbari analyzed exome sequencing data from more than 600,000 individuals across multiple biobanks and found that rare predicted loss-of-function variants in INHBE were associated with significantly lower waist-to-hip ratio adjusted for BMI — a marker of healthier fat distribution — and lower risk of type 2 diabetes.
Three features of INHBE make it especially interesting as a base-editing target:
- Liver-restricted expression. INHBE is almost exclusively expressed in hepatocytes. That matches the exact delivery envelope that LNP-delivered base editors already solve.
- Secreted protein product. The therapeutic effect is mediated by a circulating factor, so reducing hepatic output should translate cleanly to a systemic phenotype — similar to how PCSK9 works.
- Human LoF carriers appear healthy. The people the edit would mimic already exist and have no overt phenotype beyond favorable metabolism. That is as close as human genetics gets to a safety pre-trial.
As of early 2026, Regeneron has advanced antibody and siRNA approaches against inhibin βE. Verve has not publicly named an INHBE program but the target fits its platform so precisely that independent observers have called it the logical third program after PCSK9 and ANGPTL3. Several other undisclosed biotechs are working the same target.
None of this means an INHBE base editor will produce 21% weight loss. The phenotype in human LoF carriers is more modest — favorable fat distribution and metabolic markers, not Wegovy-scale weight loss. But the same was true of PCSK9 variants and statin-level LDL reduction, and the drug approaches still delivered clinically meaningful benefit. The question is whether a durable, moderate metabolic improvement from a single edit beats a weekly injection delivering a larger effect that is lost on discontinuation. For many patients and payers, the answer will depend on price and risk.
The Safety Problem
Here is where the conversation has to slow down. GLP-1 peptides are reversible. If a patient develops pancreatitis, severe nausea, or unexpected side effects, the drug is stopped and washes out within days or weeks. Base editing has no washout. The edit is the drug.
This produces specific safety concerns unique to durable genome editing:
- Off-target editing. Base editors can occasionally edit bystander bases or unintended genomic locations. Verve and the broader field have invested heavily in off-target profiling and editor engineering (e.g., ABE8.8), but no method guarantees zero off-target events across billions of edited cells.
- Unpredictable phenotype. Even a perfect on-target edit could produce effects that only manifest years later. Human LoF carriers are a good proxy, but rare carriers don't capture every possible interaction with environment, aging, or other diseases.
- No dose titration. A drug that produces 70% knockdown in week one cannot be dialed down to 50% if the patient can't tolerate it.
- Reproductive concerns. Although somatic liver edits don't affect germline DNA, regulators will expect data on exposure during pregnancy and long-term effects on patients who haven't completed childbearing.
- Irreversibility under evolving knowledge. What if, ten years from now, we learn that full INHBE knockdown produces some previously unknown liability? A drug can be discontinued. An edit cannot be un-made with current technology.
None of these concerns is disqualifying. All of them are solvable in principle. But they mean the first obesity base editors will almost certainly be reserved for patients with severe disease and failed alternatives — not the millions of people currently picking up Wegovy at their local pharmacy.
The Economics
Consider a simplified cost comparison. Assume a patient begins chronic GLP-1 therapy at age 45 and lives to 80.
- Chronic GLP-1 peptide (tirzepatide at list price): ~$1,000 per month × 12 × 35 years = $420,000 nominal. Discounted at 3% to present value, roughly $260,000. Reality is more complex due to insurance, rebates, and compounding, but this is the ballpark payers face.
- One-shot base editor (hypothetical): A single infusion priced between $500,000 and $1,500,000. High upfront cost but zero ongoing medication cost. Casgevy, Luxturna, and Zolgensma have established that payers will absorb multi-million-dollar one-time gene therapies if the durability is real.
| Platform | Upfront cost | Ongoing cost | Durability | Reversibility | Best suited for |
|---|---|---|---|---|---|
| Statins / metformin | Cheap | ~$50/year | While taken | Yes | Mild-to-moderate disease |
| GLP-1 / GIP peptides | Moderate | $8–15k/year | While taken | Yes | Moderate-to-severe disease, patient preference |
| siRNA (e.g., inclisiran) | Moderate | ~$6k/year | ~6 months | Yes (washout slow) | Chronic adherence concerns |
| Antibody (e.g., evolocumab) | High | $5–7k/year | Weeks | Yes | Chronic, well-tolerated |
| Base editor (VERVE-style) | Very high (~$0.5–1.5M) | ~Zero | Potentially lifetime | No | Severe disease, failed alternatives |
| AAV gene therapy | Very high ($1–3M+) | ~Zero | Years to lifetime | No | Rare, severe disease |
The math favors durable interventions only if the edit really lasts and only if the patient would otherwise have paid for decades of chronic therapy. The break-even calculation depends heavily on discount rate assumptions, patient age at treatment, and whether the drug is priced based on cost-plus or value-based formulas.
The Regulatory Path
The FDA has begun developing frameworks for what it internally calls "functional cures" in chronic disease — interventions that produce durable biomarker changes rather than treating acute symptoms. The precedent-setting approvals so far (Casgevy, Luxturna, Zolgensma) all involved serious or life-threatening rare diseases where the unmet need was obvious.
Obesity is different. It is common, it is chronic, and effective alternative therapies already exist. The regulatory bar for a permanent edit in a patient who could otherwise take a weekly injection will be high. Expect:
- Long-term follow-up requirements. Likely 15 years or more of pharmacovigilance.
- Restricted initial labels. First approvals may be limited to patients with severe obesity, metabolic complications, and failed or intolerant to GLP-1 therapy.
- Robust off-target reporting. Comprehensive genome-wide off-target analysis will be expected pre- and post-dosing.
- Biomarker endpoints that map to durable clinical benefit. Not just weight loss at one year, but evidence the phenotype sustains at three and five years.
The Coexistence Scenario
The most likely near-term future is not base editing replacing GLP-1 injections. It is peptides serving as the bridge — the drug you use to prove the target, achieve acute benefit, and maintain the patient — while gene editing gradually earns its place at the end of the treatment algorithm for patients whose disease biology or adherence burden make lifelong injection untenable.
Think of it as a staircase. A newly diagnosed patient starts on metformin and lifestyle. If that fails, they escalate to GLP-1 or GIP/GLP-1 dual agonists. If adherence fails or side effects become intolerable, and if a well-validated metabolic base editor is eventually approved, they graduate to a durable one-shot intervention. The peptide era doesn't end — it becomes the on-ramp.
This is precisely the pattern that emerged for PCSK9: statins remain first line, evolocumab and inclisiran occupy the middle, and VERVE-102 is being developed for patients at the severe end of the spectrum. No modality replaced another. They stacked.
Key Takeaways
- Verve's VERVE-102 proved that a single IV base-editor infusion can durably reduce a metabolic biomarker — the template exists.
- No equivalent obesity target has reached the clinic yet, but INHBE, GPR75, and ANGPTL3 are the most credible near-term candidates.
- INHBE is especially attractive because expression is liver-restricted, the protein is secreted, and human LoF carriers exist and appear healthy.
- Loss-of-function edits are tractable for base editing. Gain-of-function edits (MC4R, LEPR) are far harder.
- Safety is the hard wall: no reversibility, no dose titration, long-term follow-up required.
- The economics favor a one-shot approach only for patients who would otherwise pay for decades of chronic peptide therapy.
- The likely future is coexistence: peptides dominate early treatment, gene editing eventually serves severe or intolerant patients.
Frequently Asked Questions
Is base editing currently approved for any obesity or diabetes indication?
No. As of early 2026, the only clinical-stage in vivo base editor targeting metabolic disease is Verve's PCSK9 program for cardiovascular disease. No obesity-targeted base editor has entered clinical trials under any public disclosure.
Why not just edit GLP1R directly?
Because GLP-1 drugs work by activating GLP1R, not by knocking it out. Base editors excel at creating loss-of-function mutations, which would do the opposite of what you want. To mimic GLP-1 agonism with a gene edit, you would need gain-of-function engineering or edits in downstream or regulatory pathways — much harder to achieve precisely.
How durable would a single base-editing treatment actually be?
Hepatocytes turn over slowly but continuously — with a half-life of roughly 200–400 days in humans. This means edited cells are gradually replaced. Verve has reported stable LDL-C reductions out to 12+ months for VERVE-102, suggesting durability is meaningful. True lifetime durability remains to be confirmed with longer follow-up.
What about germline safety?
Somatic liver edits do not affect eggs or sperm and therefore cannot be passed to offspring. Regulators will still require studies to confirm this and to characterize exposure during pregnancy before approving any metabolic base editor for broad use.
Could patients ever "reverse" a base edit?
Not with current technology. Research into base editors that could edit a base back to its original identity is under way, but no such reversal system has been clinically validated. For the foreseeable future, a base edit is permanent.
Will this ever be cheaper than tirzepatide over a lifetime?
Possibly. If a one-time base-editing intervention is priced at $750,000 and produces durable benefit for 30+ years, the lifetime cost to a payer could be lower than 30 years of branded GLP-1 therapy. That comparison ignores the risk premium associated with permanence.
Further Learning
- Base Editing: Rewriting Genetic Errors — technical deep dive on how base editors work.
- How GLP-1 Drugs Work: Ozempic Explained — the peptide mechanism that gene editing would ultimately compete with.
- Semaglutide and Tirzepatide Mechanism Deep Dive — the pharmacology behind current GLP-1/GIP drugs.
- Lipid Nanoparticles for Delivery — the delivery chemistry that makes VERVE-102 possible.
- Casgevy: The First CRISPR Medicine — the regulatory precedent for durable genetic medicine.
