In September 2026, three in vivo gene editing programs reported progress against the same rare disease, aimed at the same mutation. On September 7, YolTech reported patient data for YOLT-202, an adenine base editor. A day later, Beam Therapeutics' presentation of BEAM-302 data. On September 24, Prime Medicine announced FDA clearance of its IND for PM647, a prime editor. All three are lipid-nanoparticle-delivered editors correcting the PiZ (E342K) mutation in SERPINA1, the primary cause of alpha-1 antitrypsin deficiency (AATD).
Each program reports through company press releases rather than peer-reviewed publications, and the datasets differ in maturity. YolTech's release describes a four-patient investigator-initiated study. A single 45 mg dose raised mean total AAT from roughly 5 µM at baseline to about 20 µM by Week 4, above the 11 µM protective threshold, with up to 57% target-site editing reported on liver biopsy. Beam's ERS presentation covered 29 patients in dose escalation, reporting steady-state mean total AAT of 14.4 µM at 60 mg against 5.0 µM at baseline, with durability reported up to 18 months. Prime's PM647 has cleared its IND but has no human data yet, with initial clinical results expected in 2027.
AATD suits this kind of convergence because its genetics are concentrated. The Z allele is present in about 98% of severe cases, and roughly 96% of people with AATD-associated disease carry the Pi*ZZ genotype (Blanco et al., Int J Chron Obstruct Pulmon Dis, 2017). A single base substitution defines most of the treatable population, which is why two base editors and one prime editor can pursue the same correction in parallel. The scientific race is narrowing toward one target rather than spreading across many.
The population that carries this target is small and largely hidden. Prime Medicine estimates that roughly 100,000 people in the United States carry the PiZZ genotype, and about 200,000 across the US and Europe combined. As many as 95% of individuals with the PiZZ genotype remain undiagnosed, according to Alpha-1 Foundation expert consensus. A genetic-epidemiologic survey across 21 countries identified only about 2.4% of Pi*ZZ individuals (J Allergy Clin Immunol Pract, May 2023). Diagnosis is frequently delayed by five to eight years and misclassified as smoking-related COPD (Int J Chron Obstruct Pulmon Dis, UK registry, 2023).
When several programs pursue the same correction, editing precision stops being the point of differentiation. Delivery and access, rather than editing precision alone, govern the clinical translation of gene editing. In AATD, the binding constraint becomes patient identification. A genotype-defined population that is up to 95% undiagnosed is one that every sponsor now needs to find, confirm, and enroll at the same time, drawing from the same limited pool.
Feasibility built on epidemiology estimates overstates the reachable population when most eligible patients are undiagnosed. Starting feasibility from identified individuals who meet genotype and phenotype criteria, and then mapping site capacity to them, produces a more accurate view than starting from sites and prevalence figures. The gap is upstream of the trial. Patient identification depends on provider recognition and referral pathways, and genetic testing is unevenly integrated into routine care, so eligible patients are found late or not at all.
For an in vivo editing program, three operational capabilities become gating rather than supporting. The first is genotype-first identification that surfaces PiZZ individuals before a site tries to recruit them. The second is confirmatory genetic testing to establish eligibility with precision, since the therapeutic rationale rests entirely on the genotype. The third is long-term follow-up: a one-time correction reported with durability out to 18 months needs recontactable, well-characterized cohorts tracked well beyond the dosing window. Solving identification early reduces the risk of enrollment timelines slipping while a small population is contested across programs.
The next year will test whether identification keeps pace with the editing science. Several developments will indicate how the AATD race resolves.
The editing data from all three programs is early and company-reported, so read it with that caveat. The structural point holds regardless of which modality leads. Once several editors target one genotype, the programs that identify, confirm, and retain patients earliest will set the pace, and the ones that treat patient identification as a downstream task will spend the race waiting for enrollment to catch up.