The Gene Therapy That Sounds Too Good to Be True (Because It Uses Sound)
A tiny California startup says it can deliver gene therapy with ultrasound and bubbles instead of viruses. The animal data is so good that top scientists "find it hard to believe." Here's why the Duchenne muscular dystrophy field is both electrified and deeply skeptical.
What if you could deliver gene therapy with the same ultrasound wand your doctor uses to check on a baby? No viruses. No immune meltdowns. Just sound waves, tiny bubbles, and a full-length copy of one of the largest genes in the human genome.
That's essentially what SonoThera, a South San Francisco startup, claims it can do. And the early animal data is so impressive that top scientists in the field aren't celebrating. They're squinting.
"I find it hard to believe," said Eric Olson of UT Southwestern. Jeffrey Chamberlain, a veteran Duchenne muscular dystrophy researcher at the University of Washington, put it more bluntly: "It seems a bit too good to be true."
Welcome to the most fascinating tension in biotech right now: data that could change everything, if it holds up.
The Problem With How We Do Gene Therapy Today
Duchenne muscular dystrophy is caused by mutations in the dystrophin gene, which encodes a protein that acts like scaffolding for muscle fibers. Without it, muscles progressively weaken and waste away. It primarily affects boys, and most patients lose the ability to walk by their early teens.
The obvious fix is gene replacement: give patients a working copy of the dystrophin gene. But the dystrophin gene is enormous, roughly 11 kilobases of coding sequence. And the go-to delivery vehicle for gene therapy, a virus called adeno-associated virus (AAV), can only carry about 4.7 kilobases. Think of it like trying to ship a couch through a mail slot.
So companies like Sarepta Therapeutics engineered a workaround: micro-dystrophin, a stripped-down version of the protein that fits inside AAV. Sarepta's Elevidys became the first approved gene therapy for Duchenne. It was a milestone. It was also imperfect.
AAV gene therapy comes with serious baggage. The virus triggers immune responses that can cause liver damage, blood clotting disorders, and inflammation of the heart. At least two non-ambulatory teenage boys treated with Sarepta's AAV platform died from acute liver failure. The therapy can only be given once, because the immune system remembers the virus and attacks it on sight. And that micro-dystrophin? It's a compromise. It doesn't fully replicate what normal dystrophin does.
Get tomorrow's biotech intelligence before your competitors.
Join thousands of biotech professionals who start their day with our free, daily briefing.
The field has been stuck in a frustrating loop: the delivery vehicle works well enough to get approved, but not well enough to stop searching for something better.
Enter the Bubble Machine
SonoThera's approach throws out the viral playbook entirely. Instead of stuffing genetic material inside a virus, the company uses microbubbles and ultrasound to push DNA directly into cells.
The science is called sonoporation, and it works like this: tiny gas-filled bubbles (about the size of red blood cells) are mixed with the genetic payload and injected intravenously. A doctor then aims a standard ultrasound probe at the target tissue. The sound waves make the microbubbles vibrate, expand, and eventually pop. That popping creates brief, microscopic holes in blood vessel walls and cell membranes, like punching temporary doors into a building. The DNA slips through before the holes seal shut.
It's beautifully simple in concept. The ultrasound devices are already FDA-cleared for diagnostic imaging. The microbubbles are based on contrast agents that have been used in cardiology for decades; SonoThera co-founder Dr. Steve Feinstein actually invented the first two FDA-approved ultrasound contrast agents, with work beginning in the 1980s and the agents developed and approved over the 1980s and 1990s.
But the real kicker is what this approach theoretically solves. Because there's no virus involved, there's no size limit on the genetic cargo. SonoThera can deliver the full-length dystrophin gene, all 14 kilobases of it. No compromises, no micro-dystrophin shortcuts. And because the immune system doesn't mount a lasting response to bubbles and sound waves the way it does to viruses, patients could potentially receive repeat doses over time.
Animal Data That Made Experts Do a Double Take
SonoThera has been rolling out preclinical results at major conferences throughout 2025 and 2026, including the European Society of Gene & Cell Therapy (ESGCT), the Muscular Dystrophy Association (MDA) conference, and the American Society of Gene & Cell Therapy (ASGCT) annual meeting.
The headlines from those presentations are remarkable. In mouse models of Duchenne, SonoThera's RIPPLE platform (which stands for Remote Induction of Pulsed Pressure Lateral to Energy) delivered full-length human dystrophin to skeletal muscle, heart tissue, and the diaphragm. That trifecta matters because Duchenne attacks all three.
The mice showed what the company calls "phenotypic correction," meaning actual improvement in disease symptoms, not just protein expression on a lab slide. When the team repeated doses using a reporter gene (a glowing protein that lets scientists track delivery), expression got stronger with each treatment, suggesting redosing actually works.
Then came the primate data. In non-human primates, SonoThera reported durable dystrophin protein expression up to 50% of normal endogenous levels in skeletal muscle, with robust expression also observed in cardiac and diaphragm muscle. For context, many gene therapy researchers would be thrilled with 20%. Hitting 50% with a non-viral approach borders on the unbelievable.
Which is exactly the word experts used.
Why the Skeptics Have a Point
The history of gene therapy is littered with gorgeous animal data that fell apart in humans. Mice are not people. Even non-human primates are not people. The translation gap is real, and it's especially treacherous in Duchenne, where scaling ultrasound coverage to the entire muscle mass of a growing boy presents challenges that don't exist in a 25-gram mouse.
How many ultrasound sessions would it take to treat a human patient's worth of muscle? How long would each session last? Would the transient membrane pores cause problems in larger, more complex tissues over repeated treatments? These are open questions, and SonoThera has not yet dosed a single human patient.
The quantitative details in SonoThera's public presentations also leave room for skepticism. Most of the reported metrics are qualitative: "robust," "widespread," "durable." The 50% endogenous expression figure in primates is striking, but detailed functional readouts (grip strength improvements, blood biomarker changes, fibrosis scores) haven't been fully enumerated in public abstracts. Scientists want to see the receipts, not just the claims.
The Money Believes
Investors, at least, are betting this is real. SonoThera has raised roughly $186 million across multiple rounds. The $60.75 million Series A in 2022 was led by ARCH Venture Partners, with participation from Johnson & Johnson's venture arm, Illumina Ventures, and Eli Lilly. The more recent $125 million Series B, led by Vida Ventures, brought in ARK Invest, CureDuchenne Ventures, Leaps by Bayer, Otsuka Pharmaceutical's venture arm, and RA Capital, among others.
That Series B is specifically earmarked to push the RIPPLE platform into clinical trials for Duchenne and autosomal dominant polycystic kidney disease (ADPKD), a genetic kidney disorder. The investor roster reads like a who's who of smart biotech money, and the presence of CureDuchenne Ventures signals that patient advocacy groups see enough promise to put capital behind it.
The company was co-founded by CEO Dr. Kenneth Greenberg (a gene therapy scientist and serial entrepreneur), Dr. Michael Davidson (a cardiologist who also runs NewAmsterdam Pharma and co-founded Corvidia Therapeutics, later acquired by Novo Nordisk), and Dr. Feinstein, the microbubble pioneer.
A Crowded Race With No Clear Winner
SonoThera isn't operating in a vacuum. The Duchenne gene therapy landscape is intensely competitive.
On the viral side, Solid Biosciences' SGT-003 uses a next-generation AAV capsid designed to target muscle more efficiently, with early trial data showing widespread micro-dystrophin expression. Regenxbio's RGX-202 reported that all four patients treated at the pivotal dose surpassed their expected disease trajectory on a key motor function scale at 18 months. CRISPR-based approaches from companies like Precision BioSciences are also inching toward human trials, aiming for permanent genomic correction.
On the non-viral side, Avidity Biosciences' del-zota uses an antibody-oligonucleotide conjugate to deliver exon-skipping therapy directly to muscle, with one-year data suggesting reversal of disease progression. And existing exon-skipping drugs like eteplirsen and viltolarsen, while modest in effect, are already on the market.
Each approach has trade-offs. AAV therapies are potent but toxic, one-shot, and cargo-limited. Exon-skipping drugs are safe and repeatable but only partially effective and mutation-specific. CRISPR is potentially curative but still very early and faces its own delivery challenges.
SonoThera's pitch is that it threads the needle: full-length protein, non-viral safety, and repeat dosing capability. If the human data match the animal data, it would be genuinely paradigm-shifting.
The Billion-Dollar "If"
That "if" is doing a lot of heavy lifting. The gap between stunning preclinical results and a working human therapy has swallowed countless biotech dreams. SonoThera's technology is elegant, its data are eye-popping, and its investor syndicate is blue-chip.
But as Jeffrey Chamberlain reminded us, extraordinary claims need extraordinary proof. The next chapter, the one where RIPPLE meets actual patients, will determine whether this is the breakthrough Duchenne families have been waiting for, or the latest beautiful idea that couldn't survive first contact with human biology.
The ultrasound is on. Now we wait to hear what it tells us.
Clinical & Regulatory
4 min read
The FDA Liked the Drug. It Hated the Factory.
The FDA rejected Lantheus' cancer imaging kit, and the reason had nothing to do with safety, efficacy, or clinical data. It's the latest in a growing pattern of biotech approvals killed by factory problems, not science failures.
