A Harder Look at Treatment Staying Power
Alzheimer’s drug data rarely arrive without a thicket of caveats, and that is exactly why this Leqembi update deserves a calm read rather than a victory lap. BioArctic’s partner Eisai presented a three-year real-world analysis at CTAD 2026 covering 560 early Alzheimer’s patients across 13 U.S. sites. The headline is striking: in a pre-specified comparison, patients treated with Leqembi had a 76% lower rate of progression to moderate or severe dementia through 36 months versus a weighted external ADNI cohort. The reported hazard ratio was 0.24, with a 95% confidence interval of 0.15 to 0.39.
For BioArctic, this matters because Leqembi is not merely a lab asset sitting behind a locked door. The company has no Alzheimer’s development costs for the drug, while retaining rights to sales-milestone payments and global sales royalties through its long-running alliance with Eisai. That makes evidence supporting broader, durable real-world use commercially relevant to BIOA B, even though Eisai handles global development, regulatory work and commercialization.
Persistence Is the Part Investors Should Not Skip
Drug launches live or die in the messy world of appointments, scans, tolerability, caregiver burden and clinician willingness to stick with treatment. A result from a pristine trial is one thing. Seeing patients stay on therapy in regular practice is another.
In LEADER, 87.5% of the 560 patients remained on Leqembi at the time charts were extracted, after an average treatment duration of roughly 19 months. Among the 546 patients whose disease-stage change could be assessed after about 18 months, 81.1% were described as stable or improved from baseline. Most of that group, 75.8%, stayed stable; 5.3% improved.
Those are encouraging figures, plain and simple. They also line up with the study’s estimate that 95.0% of all 560 participants remained in early-stage disease through three years. Still, “estimated” is doing real work in that sentence: the three-year figure relies on statistical modeling of observed real-world data, not every patient being directly followed for a full three years.
Maintenance Options Could Ease a Practical Bottleneck
The maintenance story may be more consequential than it first looks. After 18 months of initial treatment, 218 patients moved to once-every-four-weeks intravenous maintenance. Another 64 transitioned to once-weekly subcutaneous maintenance after that option was approved. All 64 subcutaneous-maintenance patients remained on therapy at chart extraction. That is a small sample, no use pretending otherwise, but it is a useful early signal that a less clinic-intensive approach may fit real life better.
Among monthly IV maintenance patients, 71.1% remained stable after an average 26 months from treatment start. For weekly subcutaneous maintenance, 76.6% remained stable after an average 25.1 months. No ARIA events were reported during the subcutaneous maintenance period, while isolated ARIA-H occurred in 0.5% during IV maintenance. Small groups and limited exposure demand restraint, but convenience plus persistence is a combination investors ought to watch closely.
Safety Looks Familiar, Not Miraculous
The safety readout does not erase the need for monitoring. ARIA, the imaging abnormalities associated with anti-amyloid therapies, remains a serious operational consideration. In LEADER, ARIA-E was observed in 6.6% of patients and ARIA-H in 11.1%. Those rates were close to the independent ALZ-NET registry figures of 6.3% and 10.1%, respectively, among more than 1,700 Leqembi patients at extraction.
A separate exploratory meta-analysis of 16 real-world cohorts across six countries, covering about 5,000 patients, reported pooled total ARIA of 12.4%, with symptomatic ARIA at 0.8% and adverse-event discontinuations at 3.2%. Those figures were lower than the Phase 3 Clarity AD trial comparisons cited by the company. Useful? Yes. A clean apples-to-apples proof that real-world treatment is safer? No. Different patient selection, data collection, follow-up and reporting habits can bend these comparisons in either direction.
Real-world evidence can strengthen a clinical picture, but it does not magically become a placebo-controlled trial because the sample gets bigger.
The Commercial Thread Is Getting Stronger
Leqembi is approved in 53 countries and under regulatory review in six more. IV maintenance is approved in nine countries and regions, while subcutaneous initiation treatment was approved in China and Japan in September 2026. That expanding footprint gives BioArctic’s royalty and milestone economics more possible fuel over time.
The sensible takeaway is not that every uncertainty has vanished. LEADER is retrospective, incorporates chart and electronic-record data, lacks a randomized control arm, and faces the usual confounding and completeness problems. But the data do offer something investors have wanted: evidence that Leqembi can be maintained in everyday care, with outcomes and safety broadly consistent with prior experience. In a disease area littered with disappointment, that is meaningful progress, even if it still needs a skeptical set of eyes.