How clinically meaningful are the cognitive changes seen with lecanemab and donanemab in real‑world patients?
Executive summary
The randomized trials of lecanemab and donanemab showed statistically significant slowing of cognitive and functional decline over 18 months, but the absolute differences were small versus placebo and below many conventional minimal clinically important differences on common scales (e.g., CDR‑SB and ADAS‑Cog) [1] [2] [3]. Early real‑world reports and commentaries underscore that these are promising, first‑in‑class disease‑modifying agents, yet their true clinical meaningfulness for individual patients in routine practice remains uncertain because long‑term benefit, generalizability, selection biases, and harms influence net effect outside trials [4] [5] [6].
1. What the controlled trials actually measured and found
Phase‑3 randomized trials met their primary endpoints: lecanemab’s CLARITY‑AD and donanemab’s TRAILBLAZER‑ALZ2 showed reduced brain amyloid and statistically significant, modest delays in decline on composite clinical scales at 18 months, reported as roughly 22–29% slowing of progression on selected endpoints [1] [4] [7]. Trial reports emphasize relative reductions and biomarker clearance—amyloid PET fell dramatically—while clinical scales moved in parallel with small between‑group separations over time [1] [6].
2. How big is “modest” in concrete terms — clinical thresholds
The absolute differences translate to fractions of points on commonly used scales: regulatory summaries and academic centers note changes equivalent to about 0.45–0.67 points on the 18‑point CDR‑Sum‑of‑Boxes (CDR‑SB) or other small ADAS‑Cog shifts, amounts many clinicians view as below classic minimal clinically important differences used historically to judge meaningful benefit [8] [2] [3]. Critics caution that relative percentages can overstate impact because small absolute gains on broad scales may not produce noticeable improvement for patients or caregivers over 18 months [3] [6].
3. Early real‑world experience: limited, mixed, and potentially biased
Real‑world datasets are just emerging; single‑center observational series report feasibility and some stability in cognitive scores over the first year for selected patients, but these series are small, uncontrolled, and enriched for people who pass stringent screening (amyloid PET, APOE genotyping, MRI surveillance), limiting generalizability [5] [9]. Commentaries and meta‑analyses find consistent group‑level benefits across RCTs but stress that registry‑based, longer‑term real‑world follow‑up is necessary to confirm durability and effect magnitude outside trial populations [4] [10].
4. Harms, selection, cost and logistics shape real‑world benefit
Both drugs carry meaningful safety concerns—amyloid‑related imaging abnormalities (ARIA), including brain swelling and hemorrhage—risks that require MRI monitoring and may exclude high‑risk patients (APOE ε4 carriers often at greater ARIA risk) and thereby concentrate benefits in a narrower clinical group [1] [11]. The therapies are costly and resource‑intensive (infusions, scans, specialist visits), which creates access barriers and may bias which patients receive them, affecting observed real‑world effectiveness and equity [11] [8].
5. Competing interpretations and implicit agendas in the literature
Academic re‑analyses and reviews reach divergent tones: some meta‑analyses and systematic reviews conclude statistically significant cognitive and functional improvement and endorse clinical relevance [12] [13], while methodologic critiques emphasize small absolute effect sizes and uncertain long‑term clinical translation [3] [6]. Industry sponsors funded the pivotal trials—information disclosed in trial reports—so independent long‑term registry data and head‑to‑head pragmatic studies will be important to limit sponsor‑related framing [1] [4].
6. Bottom line: what “clinically meaningful” looks like today
For populations matching the trials (early symptomatic AD, confirmed amyloid, close monitoring), lecanemab and donanemab produce a reproducible, modest slowing of decline that is plausibly meaningful at the group level and represents the first therapies to alter an underlying pathology (amyloid) [4] [14]. For individual patients, however, the average benefit is small in absolute terms over 18 months, may not translate into perceptible improvement for many, and must be weighed against ARIA risk, monitoring burden, cost, and patient priorities; thus clinical meaningfulness in routine practice is conditional and remains to be proven by longer, real‑world follow‑up [3] [5] [11].