Anti-Amyloid Therapies in Early Alzheimer Disease: What the Evidence Shows
Synthesising findings from 10 pivotal trials and meta-analyses on lecanemab and donanemab — examining whether statistically significant cognitive slowing translates into clinically meaningful benefit. Evidence through 2025.
This is an original evidence synthesis. See References for source studies.
One-Minute Takeaway
One-Minute Takeaway
- Anti-amyloid therapies in early Alzheimer disease — lecanemab and donanemab — consistently slow cognitive and functional decline by approximately 25–36% over 18 months across multiple phase 3 RCTs, with robust amyloid plaque clearance on PET (CLARITY-AD, TRAILBLAZER-ALZ 2).
- The absolute treatment–placebo difference on CDR-SB ranged from 0.45 to 0.70 points at 18 months — below the commonly cited 1.0-point minimal clinically important difference threshold, generating active debate about perceptibility at the individual patient level (Avgerinos et al. 2024; Espay et al. 2024).
- Open-label extension data from CLARITY-AD show the treatment–control difference on CDR-SB continues to widen through 48 months (1.75-point separation), with delayed-start participants not catching up to early-start participants, supporting the value of early treatment initiation.
- ARIA remains the principal safety concern, with ARIA-E occurring in 12–24% and ARIA-H in 17–31% of treated participants; risk is modulated by APOE ε4 status, and the EMA initially declined donanemab authorisation citing an unfavourable benefit–risk balance before reversing the decision in July 2025.
- Whether the observed slowing of decline is appropriately characterised using the MCID construct applied to group-level trial differences — a use the FDA itself cautions against — remains a fundamental unresolved methodological question.
Why This Topic Matters
Alzheimer disease affects over 55 million people globally and remains the leading cause of dementia. For nearly two decades after memantine’s approval in 2003, no new mechanistic therapy reached clinical practice. The FDA approvals of lecanemab (January 2023, traditional approval July 2023) and donanemab (July 2024) represented the first disease-targeted interventions that modify one of Alzheimer’s defining pathological features — cerebral amyloid plaque deposition.
Yet regulatory approval has not resolved a fundamental clinical question: does the magnitude of cognitive slowing observed in pivotal trials translate into benefit that patients, families, and clinicians can perceive? This question has produced divergent regulatory decisions — the FDA granted traditional approval to both agents, while the EMA initially rejected donanemab in 2025 before reversing its position — and continues to shape prescribing decisions, reimbursement policy, and patient counselling worldwide.
This evidence synthesis examines what the trial data actually show, how different analytical frameworks interpret those data, and where genuinely unresolved uncertainty remains.
What Evidence Was Reviewed
This synthesis draws on the two pivotal phase 3 RCTs (CLARITY-AD and TRAILBLAZER-ALZ 2), their open-label extensions, the foundational phase 2 trials, recent systematic reviews and meta-analyses examining clinical meaningfulness, and regulatory analyses from the FDA and EMA. Studies were selected for their relevance to the central question of whether anti-amyloid therapies produce clinically perceptible cognitive benefit.
| Study | Design | n | Population | Key Finding | Quality |
|---|---|---|---|---|---|
| CLARITY-AD (van Dyck et al. 2023) | Phase 3 RCT | 1795 | Early AD (MCI/mild dementia), amyloid-confirmed | Lecanemab reduced CDR-SB decline by 27% (−0.45; 95% CI −0.67 to −0.23) | High |
| CLARITY-AD OLE 36-month (2024) | Open-label extension | 1734 | Core study continuers | CDR-SB separation vs ADNI control widened to 1.01 points at 36 months | Moderate |
| CLARITY-AD OLE 48-month (2025) | Open-label extension | 478 | Core study continuers (4-year treatment) | CDR-SB separation vs ADNI control widened to 1.75 points at 48 months | Moderate |
| TRAILBLAZER-ALZ 2 (Sims et al. 2023) | Phase 3 RCT | 1736 | Early AD, amyloid+tau confirmed, stratified by tau | Donanemab slowed iADRS decline by 35% (low/medium tau) and 22% (combined) | High |
| TRAILBLAZER-ALZ 2 LTE (2025) | Open-label extension | NR | Core study continuers (3-year) | Early-start participants had 27% lower risk of stage progression vs delayed-start (HR 0.73) | Moderate |
| TRAILBLAZER-ALZ (Mintun et al. 2021) | Phase 2 RCT | 272 | Early symptomatic AD | Donanemab slowed iADRS decline by 32% at 76 weeks | Moderate |
| Avgerinos et al. 2024 | Systematic review & meta-analysis | NR | All phase 3 anti-Aβ MAb RCTs | CDR-SB effect sizes small (SMD −0.19 to −0.26); below MCID thresholds | High |
| Cummings et al. 2025 (MCID Perspective) | Expert analysis | NR | Clinical trial methodology review | Applying MCID to group-level differences is inappropriate; within-person analyses preferred | Moderate |
| Espay, Kepp & Herrup 2024 | Critical commentary | NR | Lecanemab & donanemab trial data review | ~0.5-point CDR-SB and ~3-point iADRS group differences likely imperceptible to patients | Moderate |
| Zimmer et al. 2025 | Secondary analysis of 2 RCTs + OL | 3030 | Pooled donanemab safety population | ARIA-E 20–24%, ARIA-H 27–31%; most mild-moderate; risk linked to APOE ε4 | High |
Key Findings: Anti-Amyloid Therapies and Cognitive Decline
Magnitude and Consistency of Cognitive Slowing
Both lecanemab and donanemab produced statistically significant slowing of cognitive and functional decline across all primary and secondary endpoints in their respective phase 3 trials. In CLARITY-AD, lecanemab reduced CDR-SB worsening by 0.45 points (95% CI −0.67 to −0.23) relative to placebo over 18 months, representing a 27% relative slowing. Donanemab showed a somewhat larger absolute benefit in TRAILBLAZER-ALZ 2, with CDR-SB differences of 0.67 points in the low/medium tau group (36% relative reduction) and 0.70 points in the combined population (29% relative reduction).
A meta-analysis by Avgerinos and colleagues examining all phase 3 anti-amyloid antibody trials confirmed these effects were consistent across the drug class, though effect sizes fell in the small range (standardised mean difference −0.19 for lecanemab, −0.26 for donanemab on CDR-SB). Notably, amyloid reduction on PET was substantial — approximately 55–88 centiloid reductions — and was moderately correlated with clinical benefit, providing mechanistic coherence even as the clinical magnitude remained a topic of debate.
The Clinical Meaningfulness Debate
The central controversy concerns whether the observed group-level differences are perceptible to individual patients and families. Established MCID thresholds for the CDR-SB — generally set at 1.0 point for MCI and 1.6 points for mild dementia — were derived to detect within-person meaningful change. The 0.45–0.70-point treatment differences in the pivotal trials fall below these thresholds, leading some analysts to argue these effects are imperceptible at the individual level.
However, this application of MCID to group-level trial differences has been contested. Cummings and colleagues argued that using individual-level MCID thresholds to judge between-group differences in clinical trials is methodologically inappropriate, as it would set a standard nearly impossible to achieve with any disease-slowing therapy in an 18-month trial. The FDA itself has stated that treatment effects are distinct from meaningful within-person change. An alternative approach — responder analysis — reveals that a meaningfully greater proportion of treated participants remained clinically stable compared to placebo, with 47% of donanemab-treated participants showing no CDR-SB decline at 12 months versus 29% on placebo in the low/medium tau population.
An alternative framing — the “time saved” analysis — provides a more intuitive expression of benefit. Slope modelling from CLARITY-AD estimated that lecanemab-treated participants took approximately 25.5 months to reach the same level of decline that placebo participants reached at 18 months, representing roughly 7.5 months of preserved function. TRAILBLAZER-ALZ 2 yielded similar estimates: approximately 4.4–7.5 months of delayed progression on iADRS and CDR-SB, respectively, for the low/medium tau population.
Whether a delay of 4.5 to 7.5 months constitutes meaningful benefit depends substantially on individual patient context: stage of disease, personal goals, and family circumstances. For some individuals, preserving functional independence for an additional half-year may be highly valued; for others, the burden of biweekly infusions, safety monitoring, and ARIA risk may outweigh this time advantage.
TRAILBLAZER-ALZ 2 uniquely stratified participants by baseline tau PET, revealing that those with lower tau burden — who were earlier in their disease trajectory — experienced substantially greater treatment benefit. The primary analysis population with low/medium tau showed 35% slowing of iADRS decline, compared with a smaller benefit in the high-tau subgroup. This pattern was consistent across CDR-SB, ADAS-Cog, and functional measures.
This differential response by tau burden reinforces the biological rationale for treating as early as possible in the disease course, when downstream tau-mediated neurodegeneration is less advanced. In the Clarity-AD open-label extension, participants with low or no tau PET showed the most favourable long-term outcomes, with 69% having no CDR-SB decline and 56% showing actual improvement at 48 months — findings that may reshape how clinicians conceptualise treatment windows.
APOE ε4 homozygotes — who comprise approximately 15% of clinical trial populations — represent a complex subgroup for anti-amyloid therapy. In CLARITY-AD, pre-specified exploratory analysis of ε4 homozygotes showed that lecanemab numerically favoured treatment on ADAS-Cog14 and ADCS-MCI-ADL but did not demonstrate benefit on CDR-SB or ADCOMS. These subgroup findings are limited by sample size but raise the question of whether the clinical benefit profile may differ in this genotype.
Simultaneously, APOE ε4 carriers — particularly homozygotes — face substantially elevated ARIA risk. Pooled safety data from the donanemab programme showed ARIA-E rates of approximately 40% in ε4 homozygotes versus 19% in non-carriers. Current appropriate use recommendations from both manufacturers advise caution, and the EMA’s lecanemab approval included APOE ε4-related restrictions. Both drug labels recommend APOE genotyping prior to treatment initiation so that clinicians and patients can engage in fully informed shared decision-making about the benefit–risk balance.
CLARITY-AD included exploratory health-related quality-of-life assessments using the EQ-5D-5L (patient-rated) and QOL-AD scales. At 18 months, lecanemab was associated with 49% less decline on EQ-5D-5L and 56% less decline on QOL-AD as rated by participants themselves. Caregiver burden, measured by the Zarit Burden Interview, showed 38% less increase in the lecanemab group.
These quality-of-life results provide an additional lens through which to evaluate clinical meaningfulness beyond traditional cognitive scales, though it should be noted that these were exploratory endpoints and the quality-of-life instruments may capture different aspects of the disease experience than cognitive assessments. Patient and caregiver perspectives on what constitutes meaningful benefit may diverge from purely scale-based definitions.
Long-Term Trajectory: Open-Label Extension Data
Perhaps the most compelling evidence for clinical meaningfulness comes from longitudinal follow-up. The Clarity-AD open-label extension demonstrated that the treatment–control CDR-SB difference continued to widen well beyond the 18-month trial period: from 0.45 points at 18 months, to 1.01 points at 36 months, to 1.75 points at 48 months when compared against a matched ADNI observational cohort. This expanding trajectory suggests the 18-month trial captures only the beginning of cumulative benefit.
Critically, participants who initiated lecanemab during the randomised phase (early-start group) maintained a separation from those who began during the open-label phase (delayed-start group) through 48 months, with the delayed-start cohort never fully catching up. Similarly, TRAILBLAZER-ALZ 2 extension data showed early-start participants had a 27% lower hazard of progressing to the next clinical disease stage compared with delayed-start participants (HR 0.73). These findings provide evidence that earlier treatment initiation may confer enduring benefit — though the absence of a concurrent placebo control in the extension phases limits the strength of this inference.
Quality & Consistency of Evidence
What the Evidence Does Not Show
The current evidence base, while growing, leaves several clinically important questions unanswered.
Real-world effectiveness versus trial efficacy: Pivotal trials enrolled highly selected populations confirmed by amyloid PET or CSF biomarkers, with extensive monitoring. Whether the same magnitude of benefit will be observed in routine clinical practice — including patients with more comorbidities and potentially less rigorous MRI surveillance — remains unknown. CMS-mandated registry data will be critical for answering this question.
Optimal treatment duration and stopping criteria: Donanemab’s fixed-duration dosing paradigm (treatment until amyloid clearance) showed sustained benefit after discontinuation in the pivotal trial. Lecanemab’s FDA-approved maintenance dosing (reduced-frequency infusions after 18 months of full treatment) adds a new dimension. However, data on amyloid re-accumulation kinetics after treatment cessation and the clinical consequences of re-accumulation are limited to 3–4 years of follow-up.
Differential benefit across demographic and genetic subgroups: Trial populations underrepresented several key groups, including Black and Hispanic individuals, patients with significant cerebrovascular disease, and those taking anticoagulants (who were excluded). The extent to which trial results generalise to these populations is unknown.
Interaction with concomitant tau pathology: While tau stratification in TRAILBLAZER-ALZ 2 demonstrated that lower tau burden predicts greater anti-amyloid benefit, no data exist on whether combining anti-amyloid therapy with emerging anti-tau interventions could produce additive or synergistic clinical effects.
Practical Implications
Based on the current evidence, the following considerations may inform clinical decision-making when evaluating anti-amyloid therapy for eligible patients with early Alzheimer disease.
Consistent data across both pivotal trials and their extensions indicate that treatment benefit is greater in individuals with lower tau burden and earlier clinical stage. The delayed-start analyses in both programmes show that earlier initiation is associated with enduring advantage that delayed treatment does not fully recover.
APOE ε4 status modulates both ARIA risk and potentially clinical benefit profile. Both approved drug labels recommend genotyping before treatment to allow a fully informed discussion of individual benefit–risk balance, particularly for homozygous carriers who face the highest ARIA rates.
The available evidence supports framing treatment effects as a slowing of decline rather than stabilisation or improvement. Communicating benefit as “approximately 5–7 months of preserved function over 18 months” or as a reduction in the proportion of patients progressing may be more intuitive for patients than percentage-slowing metrics.
ARIA surveillance with serial MRI during the first 6 months of treatment — when most events occur — is a non-negotiable component of safe prescribing. Current data indicate that ARIA risk stabilises after approximately 6 months and does not appear to increase with long-term treatment. Anticoagulant use remains a contraindication pending further safety data.
Evidence Grade & Bottom Line
Evidence Grade: Moderate
Two large, well-conducted phase 3 RCTs with concordant results provide strong evidence that anti-amyloid therapies slow cognitive and functional decline in early Alzheimer disease. The grade is capped at moderate rather than strong because (1) the clinical perceptibility of the 18-month treatment effect remains genuinely debated, (2) long-term data rely on open-label extensions without concurrent placebo controls, and (3) the generalisability to real-world clinical populations is not yet confirmed.
- Lecanemab and donanemab consistently slow cognitive and functional decline by 25–36% over 18 months in biomarker-confirmed early Alzheimer disease.
- Absolute group-level differences on CDR-SB (0.45–0.70 points) fall below traditional MCID thresholds, but applying MCID to between-group trial comparisons may be methodologically inappropriate.
- Open-label extension data through 48 months show a progressively widening treatment effect, with earlier treatment initiation associated with sustained advantage.
- ARIA requires structured MRI monitoring; risk is elevated in APOE ε4 carriers and is greatest during the first three infusions.
- Real-world effectiveness data, optimal treatment duration, and benefit in underrepresented populations remain critical unresolved questions that ongoing registries and extension studies may answer.
Article Information
For Educational Purposes Only. This is an original evidence synthesis informed by the studies listed below. It does not replace clinical judgement. Drug dosages should be verified against current prescribing information.
References
- Avgerinos KI, et al. Critical assessment of anti-amyloid-β monoclonal antibodies effects in Alzheimer’s disease: a systematic review and meta-analysis highlighting target engagement and clinical meaningfulness. Sci Rep. 2024;14:25095. DOI: 10.1038/s41598-024-75204-8
- van Dyck CH, Swanson CJ, Aisen P, et al. Lecanemab in early Alzheimer’s disease. N Engl J Med. 2023;388(1):9–21. DOI: 10.1056/NEJMoa2212948
- Sims JR, Zimmer JA, Evans CD, et al. Donanemab in early symptomatic Alzheimer disease: the TRAILBLAZER-ALZ 2 randomized clinical trial. JAMA. 2023;330(6):512–527. DOI: 10.1001/jama.2023.13239
- Cummings J, et al. Perspective: Minimal clinically important difference (MCID) and Alzheimer’s disease clinical trials. Alzheimers Dement (N Y). 2025;11(1):e70059. DOI: 10.1002/trc2.70059
- Espay AJ, Kepp KP, Herrup K. Lecanemab and donanemab as therapies for Alzheimer’s disease: an illustrated perspective on the data. eNeuro. 2024;11(7):ENEURO.0319-23.2024. DOI: 10.1523/ENEURO.0319-23.2024
- Zimmer JA, Ardayfio PA, Wang J, et al. Amyloid-related imaging abnormalities with donanemab in early symptomatic Alzheimer disease: secondary analysis of the TRAILBLAZER-ALZ and ALZ 2 randomized clinical trials. JAMA Neurol. 2025;82(5):510–522. DOI: 10.1001/jamaneurol.2025.0200
- Mintun MA, Lo AC, Duggan Evans C, et al. Donanemab in early Alzheimer’s disease. N Engl J Med. 2021;384(18):1691–1704. DOI: 10.1056/NEJMoa2100708
- Cummings J, Zhou Y, Lee G, et al. Maximizing the benefit and managing the risk of anti-amyloid monoclonal antibody therapy for Alzheimer’s disease. Alzheimers Dement. 2025;21(4):e14619. DOI: 10.1002/alz.14619
- De Strooper B, Haass C, Hardy J, Zetterberg H. The regulatory rollercoaster continues — EMA refuses donanemab. Lancet. 2025;405(10492):1810–1812. DOI: 10.1016/S0140-6736(25)00833-5
- Honig LS, Sabbagh MN, van Dyck CH, et al. Updated safety results from phase 3 lecanemab study in early Alzheimer’s disease. Alzheimers Res Ther. 2024;16(1):105. DOI: 10.1186/s13195-024-01507-7
- Lansdall CJ, Saunders S, Geber A, et al. The search for clarity regarding “clinically meaningful outcomes” in Alzheimer disease clinical trials: CLARITY-AD and beyond. Alzheimers Res Ther. 2024;16(1):34. DOI: 10.1186/s13195-024-01412-z
- Dickson SP, Wessels AM, Dowsett SA, et al. ‘Time saved’ as a demonstration of clinical meaningfulness and illustrated using the donanemab TRAILBLAZER-ALZ study findings. J Prev Alzheimers Dis. 2023;10:595–599. DOI: 10.14283/jpad.2023.56
- Hardy J. Alzheimer’s disease: treatment challenges for the future. J Neurochem. 2025;169:e70024. DOI: 10.1111/jnc.70024