The Impossible Question
The FDA lifted its clinical hold on two CRISPR gene therapy trials on March 2, 2026, four months after a single ambiguous death forced regulators to answer a question with no clean solution: when a one-time permanent gene edit meets a population already dying from heart failure, how do you distinguish treatment harm from the disease itself?
On October 29, 2025, the agency stopped enrollment in MAGNITUDE and MAGNITUDE-2, trials testing nexiguran ziclumeran (nex-z) in 1,200 patients with cardiac amyloidosis and 60 patients with nerve-destroying polyneuropathy. The trigger was Grade 4 liver toxicity in a man in his 80s enrolled in the cardiac trial. He died during hospitalization, but not from liver failure. Sepsis following surgery for a perforated ulcer killed him.
The regulatory system had to decide whether CRISPR caused the liver damage or whether an elderly man with proteins strangling his heart simply had the kind of catastrophic outcome that happens in advanced disease.
Why Gene Therapy Creates Unique Regulatory Dilemmas
Nex-z uses CRISPR/Cas9 technology to permanently inactivate the TTR gene, which encodes the transthyretin protein. In ATTR amyloidosis, this protein misfolds and accumulates in organs, destroying heart tissue in ATTR-CM patients and peripheral nerves in ATTRv-PN patients. A single 55 mg infusion edits the gene once, forever.
That permanence creates an epistemological problem traditional drug trials don't face. If a patient on a daily pill develops toxicity, you stop the pill, watch what happens, maybe rechallenge at lower doses. The relationship between drug and harm becomes clearer over time.
Gene editing offers no such clarity. You can't withdraw the edit. You can't titrate the dose down. You can't rechallenge. Phase 1 data showed nex-z produced consistent, deep, and long-lasting TTR reduction, exactly what you want if the therapy works, exactly what you fear if it causes harm.
So when an 80-year-old cardiac patient shows severe liver enzyme elevation, dies from surgical sepsis, and you're evaluating a therapy that will be given to 1,200 people with failing hearts, what's the right move?
The Clinical Hold Mechanism in Action
The FDA's answer was a full stop. Clinical holds represent the agency's emergency brake for gene therapies, a regulatory tool that pauses enrollment and dosing while investigators examine whether the intervention caused the adverse event. For four months, Intellia Therapeutics and Regeneron Pharmaceuticals provided data while regulators tried to establish causality in a patient population where bad outcomes are the baseline.
ATTR-CM patients are already dying. The disease is progressive and fatal. Misfolded proteins infiltrate cardiac tissue, causing restrictive cardiomyopathy. The MAGNITUDE trial's primary endpoint measures a composite of cardiovascular-related events, including mortality, because death is the expected trajectory without intervention.
This creates a signal-to-noise problem. When your trial population has advanced heart disease, elevated liver enzymes in an elderly patient could mean the gene edit targeted the wrong cells. Or it could mean an 80-year-old with cardiac amyloidosis developed liver congestion from heart failure, then got septic after emergency surgery, and the CRISPR therapy was incidental.
The FDA's recent loosening of approval standards for common drugs, allowing single-trial approvals for established drug classes, makes the four-month pause over one ambiguous death in a novel gene therapy particularly striking. Regulatory risk tolerance is not uniform. The agency applies hair-trigger sensitivity to permanent genetic interventions while streamlining pathways for familiar molecules.
The Resolution: Enhanced Monitoring, Not Answers
The FDA lifted the hold on MAGNITUDE-2 on January 27, 2026, and on MAGNITUDE on March 2, 2026. The resolution came with new safeguards, not new certainty about what killed the patient.
Both trials now require enhanced liver laboratory monitoring. Investigators received guidance for short-term steroid treatment if elevated liver enzymes appear. Patients with certain liver abnormalities are now excluded from enrollment.
MAGNITUDE added cardiovascular exclusion criteria. Patients with ejection fraction below 25% at screening can't enroll. Neither can patients with recent cardiovascular instability. The trial is now selecting for a slightly healthier subset of a sick population, people whose hearts are failing, but not quite as badly.
The system's answer to "Did CRISPR cause this death?" is effectively "We don't know, but here's how we'll watch more carefully." That's not evasion. It's the honest acknowledgment that causality in progressive disease populations often remains ambiguous even after months of investigation.
What This Reveals About Gene Therapy Development
The pause-and-restart pattern reveals how regulators navigate permanent uncertainty. Gene therapy trials operate in a space where the intervention can't be reversed, the patient population is deteriorating, and adverse events may be indistinguishable from disease progression.
The FDA's clinical hold mechanism worked exactly as designed: detect signal, stop everything, investigate thoroughly, implement additional safeguards, resume with enhanced monitoring. But the four-month timeline for 1,200 ATTR-CM patients meant four more months of cardiac deterioration while regulators examined one elderly man's liver enzymes and surgical complications.
Nex-z has Orphan Drug and Regenerative Medicine Advanced Therapy designations from the FDA, plus Orphan Drug status from the European Commission, regulatory pathways designed to accelerate development for rare diseases. Yet a single ambiguous adverse event triggered a full stop.
The trials are now enrolling again. MAGNITUDE will continue evaluating whether one infusion can prevent cardiovascular death in 1,200 people whose hearts are being destroyed by rogue proteins. MAGNITUDE-2, which increased its target enrollment from 50 to 60 patients, will measure whether the same intervention can slow nerve damage, tracking changes in neuropathy scores and serum TTR levels.
Both trials are betting that regulators can eventually distinguish treatment toxicity from the background noise of progressive organ failure. The four-month pause suggests that bet remains uncertain.