The Gene Therapy That Killed Four Kids Is Getting a Second Chance

What went wrong? It turns out XLMTM patients carry a hidden vulnerability. Many have undiagnosed cholestatic liver disease, a condition where bile doesn't flow properly. Nobody knew this beforehand; it didn't show up in animal testing. When you flood those already-fragile livers with massive quantities of viral vector, the result is catastrophic.

Of 24 patients treated in ASPIRO, four died. Five more survivors developed milder liver problems. The survivors who avoided liver issues? Many gained life-changing improvements. The therapy worked. It just killed some of the kids it was trying to save.

The Do-Over: ASP2957 and the VALOR Trial

Astellas formally terminated AT132 in October 2024. But they weren't walking away from XLMTM. They were switching vehicles.

The new candidate, ASP2957, is built on fundamentally different technology. Instead of the old AAV8 viral shell that spread throughout the body (liver included), ASP2957 uses a next-generation engineered capsid called MyoAAV3.8. Think of it like a delivery truck with a GPS locked onto muscle tissue. It enters muscle cells through integrin receptors while largely avoiding the liver.

The practical result: Astellas can use a dose roughly 100 times lower than what AT132 required. Less virus in the body means dramatically less liver exposure. It's the difference between watering your garden with a hose versus flooding your entire yard to reach one flower bed.

Astellas licensed this technology from Kate Therapeutics (now part of Novartis) in 2023, spending the intervening years running preclinical work to prove the concept held up.

Built for Paranoia

The VALOR trial design reads like it was written by someone who lost sleep over what happened before. Every safety lesson from ASPIRO is baked into the protocol.

Patient selection is ruthlessly narrow. Only boys aged three or younger who need a ventilator for 20-plus hours daily. And critically: anyone with a history of liver dysfunction is excluded. That cholestatic vulnerability that killed patients last time? Now it's a hard disqualifier.

Dosing is glacially cautious. Sentinel dosing means one patient goes first, then everyone waits eight full weeks before the next patient receives treatment. If something goes wrong, it shows up before others are exposed. The trial plans to enroll up to nine patients across two dose levels.

Prophylactic immunosuppression is standard from day one, designed to prevent the immune system from attacking the viral vector and triggering inflammatory cascades.

Astellas is also running a companion study called EXCEL, tracking XLMTM patients to map how their livers behave naturally over time. They're building the baseline data that didn't exist before.

The Bigger Picture for Gene Therapy

This isn't just about one company redeeming itself. It's a test case for the entire field.

Gene therapy's first wave promised miracles but delivered some nightmares alongside the breakthroughs. Dose-related toxicity, immune reactions, and organ damage have haunted multiple programs across different companies. The industry's response has been a shift toward tissue-targeted vectors that go where they're needed and avoid where they're not.

ASP2957 represents that next generation in action. If it works safely, it validates the thesis that engineered capsids can solve the toxicity problem that has plagued AAV gene therapies. If it fails, it raises uncomfortable questions about whether some patient populations are simply too fragile for viral vector approaches.

What Families Are Weighing Right Now

For XLMTM families, this is an agonizing calculation. The disease will almost certainly kill their child. The last attempt at treatment killed four children. The new version is supposed to be safer, but "supposed to be" carries a different weight when you've watched the previous trial go sideways.

No public statements from advocacy groups or patient families have emerged yet in response to the VALOR launch. That silence might reflect cautious optimism, or it might reflect the kind of guarded hope that only comes after being burned.

Nine slots. Nine families who will decide whether a 100-fold dose reduction and a muscle-targeted vector are enough to tip the risk-benefit calculation. For a disease with zero approved therapies and near-certain early death, the math may ultimately favor courage over caution.

But nobody should pretend this is simple.

Amgen Just Bet on a Computer That Doesn't Fully Exist Yet

Quantinuum, a quantum computing company backed by Amgen and valued at $10 billion, just filed for what could be the largest quantum IPO ever. The real story: Amgen is betting quantum computers can fundamentally redesign how drugs get discovered.