Combined 64-channel EEG and PET evaluation of lecanemab efficacy: Two case reports

Abstract

Lecanemab, an anti-amyloid-β (Aβ) monoclonal antibody, has shown potential in slowing Alzheimer's disease (AD) progression. We report divergent responses in two AD patients with similar backgrounds following lecanemab treatment. Subject 1 showed worsened daily functioning, increased Aβ/tau deposition, and elevated electroencephalogram (EEG) slow/fast wave ratios (>30%) in right fronto-occipital regions (Fp2/F8/O1). Subject 2 improved cognitively, with modest Aβ reduction, marked tau suppression, and mixed EEG frontal declines (F3/F7/O1/Pz) and parietal/occipital gains (P3/P4/T3/T6). EEG alterations corresponded with positron emission tomography findings, suggesting its potential for therapeutic monitoring. Multimodal analysis revealed region-specific lecanemab effects, supporting EEG's role in evaluating treatment efficacy.

Introduction

Alzheimer's disease (AD), the most common neurodegenerative cause of dementia, is pathologically characterized by extracellular amyloid-β (Aβ) plaques and intracellular hyperphosphorylated tau neurofibrillary tangles.^1,2 Despite decades of research, therapeutic advancements remain limited, with recent efforts focusing on Aβ-targeted strategies. Lecanemab targets soluble and insoluble Aβ aggregates, binding with high affinity to soluble Aβ oligomers and protofibrils and promoting their clearance. Lecanemab has emerged as a promising disease-modifying therapy, potentially delaying AD progression through Aβ-tau interplay modulation.^3,4

Current therapeutic assessments rely primarily on positron emission tomography (PET) with Aβ- and tau-specific tracers (e.g., ¹⁸F-AV45, ¹⁸F-AV1451) and cognitive scales, often neglecting the utility of non-invasive neurophysiological tools.^5,6 Electroencephalography (EEG), with millisecond-level temporal resolution, offers real-time insights into cortical network dynamics, which is an essential yet underutilized domain for assessing spatially specific treatment effects. ⁷

This case report presents a multimodal biomarker investigation in two apolipoprotein E (APOE) ε3/ε4 AD patients (Subjects 1 and 2) undergoing lecanemab therapy. Longitudinal assessments included high-density 64-channel EEG, Aβ and tau PET (¹⁸F-AV45 and ¹⁸F-AV1451), and standardized neuropsychological evaluations conducted pre- and post-treatment. It is the first time integrated PET, EEG, and neuropsychological evaluations for monitoring AD progression and evaluating the therapeutic efficacy of lecanemab.

Case presentation

Two APOE ε3/ε4 AD patients (2024 diagnostic criteria: CSF/Aβ PET-confirmed) with equivalent education were longitudinally evaluated before and after two months of lecanemab treatment. Baseline characteristics and clinical outcomes pre- and post-treatment for Subject 1 and Subject 2 are summarized in Table 1.

Table 1.

Baseline characteristics and clinical outcomes pre- and post-treatment for subject 1 and subject 2.

	Subject 1	Subject 2
Sex	Female	Male
Age	67	73
Education background	bachelor	bachelor
APOE Genotype	ε3/ε4	ε3/ε4
CSF (Aβ1–42)	450.29 pg/ml	480.7 pg/ml
CSF (Aβ1–40)	8945.17 pg/ml	7054.55 pg/ml
CSF (t-tau)	1217.47 pg/ml	487.71 pg/ml
CSF (p-tau181)	131.88 pg/ml	110.69 pg/ml
CSF (Aβ1–42/Aβ1–40)	0.05	0.07
MoCA Score	6 to 8	20 to 20
MMSE Score	12 to 11	24 to 25
ADL	80 to 60	80 to 90
ZBI	51 to 75	40 to 27
Slow/fast wave ratios in EEG	2.7 to 2.7	1.5 to 1.7

CSF: cerebrospinal fluid; MoCA: Montreal Cognitive Assessment; MMSE: Mini-Mental State Examination; ADL: Activities of Daily Living; ZBI: Zarit Burden Interview.

Subject 1 (Sub1)

A 67-year-old female with progressive AD (CSF-positive since 2022), comorbid sleep disturbances, and depression on a regimen of sertraline 100 mg daily. Baseline assessments—including ¹⁸F-AV45/¹⁸F-AV1451 PET-CT, electroencephalography (EEG), Montreal Cognitive Assessment (MoCA, 6/30), Mini-Mental State Examination (MMSE, 12/30), Activities of Daily Living (ADL, 80/100), and Zarit Burden Interview (ZBI, 51/110)—indicated advanced neurodegeneration. Following four doses of lecanemab, she experienced relative stability in cognitive function (MMSE 11/30, MoCA 8/30), worsened daily functioning (ADL 60/100), increased caregiver burden (ZBI 75/110), and secondary suicidal ideation attributable to the protracted illness burden, history of chronic depression, and unauthorized discontinuation of the antidepressant sertraline. It prompted the initiation of psychiatric interventions, including dosage adjustment, psychological counseling, and intensified family supervision. Subsequently, the patient's mood improved, with stabilization achieved within approximately two weeks and concomitant improvement in sleep.

Subject 2 (Sub2)

A 73-year-old male with mild AD (baseline MoCA 20/30, MMSE 24/30, ADL 80/100, ZBI 40/110). After four doses of lecanemab, the patient showed improved cognitive function (particularly in memory recall: 3-meal recall accuracy), executive functioning (independent public transit use), and daily functioning, with scores of MMSE 25/30, ADL 90/100, and ZBI 27/110, while the MoCA score of 20/30 remained stable.

Resting-state EEG recordings were acquired from participants in an eyes-closed condition using a 64-channel BrainCap (Brain Products) with impedances maintained below 5 kΩ, sampled at 5000 Hz (FCz reference). Potential confounders, including medication and vigilance state, were controlled by maintaining stable pre- and post-treatment regimens, conducting EEG sessions in the morning while patients were awake, and using software to reject artifact-heavy epochs. Subsequent preprocessing in MATLAB/EEGLAB. Source localization was then performed using Brainstorm, employing a three-layer boundary element model with 3003 unconstrained cortical dipoles, where lead-field computation and minimum-norm estimation (SNR = 3 regularization) preceded dimensionality reduction via principal component analysis. Finally, spectral and connectivity features were extracted: power spectral density (Welch's method) and debiased weighted phase lag index (dwPLI) were computed in theta (5–7 Hz) and alpha (8–12 Hz) bands across 465 region-pairs derived from an established fMRI-derived parcellation. Baseline EEG slow/fast wave ratios (Sub1: 2.7; Sub2: 1.5), consistent with severity reported in the literature, supported disease stratification.^8,9 Regional EEG analysis using the 10–20 system revealed divergent treatment responses. A significant change threshold of ±30%—three times the intrinsic EEG noise level (±10%) defined by test-retest variability in healthy volunteers—was applied to enhance the reliability of efficacy evaluations. Sub1 showed >30% increases in slow/fast wave ratios in right fronto-occipital leads (Fp2, F8, O1), while Sub2 exhibited reductions in frontal leads (F3, F7, O1, Pz) and gains in parietal/occipital areas (P3, P4, T3, T6), suggesting early fronto-occipital modulation by lecanemab (Figure 1). The EEG slow/fast wave ratios pre- and post-lecanemab in two cases are summarized in Table 2.

Figure 1.

EEG changes in subject 1 and subject 2 following lecanemab treatment. Pre-treatment power spectra across frequency bands from the 64-channel EEG are shown for Subject 1 (a) and Subject 2 (c), highlighting distinct baseline distributions. Post-treatment spectra are depicted for Subject 1 (b) and Subject 2 (d). Following lecanemab therapy, Subject 1 (e) exhibited >30% increases in slow/fast wave ratios in right fronto-occipital leads (Fp2/F8/O1), whereas Subject 2 (f) demonstrated mixed changes, with decreases in frontal regions (F3/F7/O1/Pz) and increases in parietal/occipital areas (P3/P4/T3/T6). No statistical comparisons were performed due to the two-case nature of this report, with data shown descriptively.

Figure 2.

Aβ/tau PET-CT image of subject 2. Aβ/tau PET-CT image revealed the modest Aβ (¹⁸F-AV45) reduction (TLG −3.7%) and marked tau (¹⁸F-AV1451) suppression (TLG −32.1%) compared with baseline. The green markers represent the situation before the treatment, while the yellow ones indicate the condition after the treatment. No statistical comparisons were performed due to the two-case nature of this report, with data shown descriptively.

Table 2.

EEG slow/fast wave ratios pre- and post-lecanemab in subject 1 and subject 2.

Electrode	Subject 1				Subject 2
	Pre-lecanemab	Post-lecanemab	Percentage a		Pre-lecanemab	Post-lecanemab	Percentage a
Fp1	1.4188	1.5447	9%		0.8621	0.8036	−7%
Fp2	1.2862	1.6915	32%		0.9643	0.807	−16%
F3	0.98	1.0576	8%		1.4615	0.7701	−47%
F4	1.0315	1.154	12%		0.7671	0.7975	4%
F7	1.6293	1.5899	−2%		0.9552	0.5786	−39%
F8	1.296	1.7925	38%		1	0.7606	−24%
Fz	0.9182	1.0115	10%		1.0411	1.1441	10%
C3	1.0821	1.006	−7%		0.8718	0.7593	−13%
C4	1.1351	1.0494	−8%		0.5714	0.7368	29%
Cz	0.683	0.5601	−18%		1.1538	1.1078	−4%
P3	0.9623	1.2275	28%		0.566	0.7778	37%
P4	1.2338	1.058	−14%		0.6667	1.0256	54%
T5	0.7622	0.9037	19%		0.6623	0.7363	11%
T6	0.8855	0.6585	−26%		0.6825	0.9286	36%
Pz	1.2582	1.2712	1%		1.9677	1.2899	−34%
T3	1.1578	1.2468	8%		0.6047	0.8919	48%
T4	0.914	0.8761	−4%		0.6944	0.7903	14%
O1	0.8415	1.1277	34%		0.7867	0.47	−40%
O2	0.9214	1.0138	10%		0.7313	0.5886	−20%

^a Percentage: (Post-lecanemab slow/fast wave ratio – Pre-lecanemab slow/fast wave ratio)/ Pre-lecanemab slow/fast wave ratio.

EEG-guided PET quantification (Using 3D Slicer-defined regions of interest) and total lesion glycolysis (TLG, TLG = volume × mean SUV) showed: Sub1: Paradoxical increases in Aβ burden (TLG +18.4%, volume: 129.9 to 153.9 cm³) and tau deposition (TLG +18.7%, volume: 155.3 to 184.2 cm³), correlating with clinical changes.

Sub2: Modest Aβ reduction (TLG −3.7%) with marked tau suppression (TLG −32.1%), corresponding with clinical improvement (Figure 2). Although two-month Aβ clearance was limited, Sub2's tau and clinical response supports lecanemab's potential as a disease-modifying therapy.

Discussion

This longitudinal case study highlights the differential therapeutic responses to lecanemab in two APOE ε3/ε4 AD patients stratified by disease severity. The findings indicate the reference value of baseline neuropathological burden, stage-specific therapeutic windows, and the integration of multimodal biomarkers in evaluating anti-Aβ efficacy. Despite similar genetic risk and identical treatment protocols, the contrasting clinical trajectories—observed clinical decline in Subject 1 (Sub1, advanced AD) and functional improvement in Subject 2 (Sub2, mild AD)—emphasize the critical influence of disease stage on treatment outcomes. These observations support the efficacy of Aβ-targeted therapies in patients with early AD.

Mechanistic basis of response heterogeneity

The observed variability in response is attributable to differences in neuropathological staging, as assessed through multimodal biomarkers:

Sub1 (advanced AD)

Severe baseline cognitive impairment (MoCA 6/30, MMSE 12/30) and comorbid neuropsychiatric conditions (treatment-resistant depression, disrupted sleep architecture) were consistent with hypotheses suggesting irreversible synaptic loss in advanced AD. Post-treatment increases in Aβ and tau burden (TLG +18.4% and +18.7%, respectively), alongside clinical decline (ADL −20), may reflect short-term fluctuations in lesion volume. Long-term follow-up is needed to exclude pseudo-progression (such as microglial activation).^10,11 Notably, right fronto-occipital EEG slow/fast wave ratios elevation (>30% at Fp2/F8/O1) may reflect maladaptive cortical hyperexcitation, a compensatory mechanism associated with accelerated cognitive decline in advanced AD cohorts. ⁸ Regarding the suicidal ideation observed during the course of treatment, a multidisciplinary evaluation conducted by psychiatrists and neurologists determined that it was attributable to unauthorized medication discontinuation and disease progression, rather than being linked to lecanemab therapy. Following the implementation of active therapeutic interventions and enhanced familial support, the patient's affective symptoms demonstrated significant improvement, with resolution of suicidal ideation. The observed clinical deterioration during treatment may be attributed to a combination of factors: a long disease history marked by short-term memory impairment dating back over seven years, with CSF biomarker confirmation more than three years ago indicating accelerated progression prior to lecanemab initiation; the typically limited therapeutic response to lecanemab in moderate-to-severe AD; and comorbid major depressive disorder with disrupted sleep architecture, which not only reflects disease progression but also worsens cognitive function, accelerates functional decline, and increases caregiver burden. The convergence of these pathological processes likely contributed to the rapid clinical decompensation.

Sub2 (mild AD)

Relatively preserved cognitive function (MoCA 20/30, MMSE 24/30) and mild baseline Aβ/tau pathology enabled more effective engagement with therapy. Reductions in frontal EEG slow/fast wave ratios (F3/F7/O1/Pz) were associated with marked tau clearance (TLG −32.1%) and functional recovery (independent public transit use; ADL +10). These outcomes suggest that lecanemab exerts dual therapeutic effects—modest Aβ reduction (TLG −3.7%) and marked tau suppression—particularly in early-stage AD with retained neuroplasticity. This is consistent with the evidence of lecanemab's preferential binding to soluble Aβ protofibrils and its indirect modulation of tau. ³

Spatiotemporal specificity of therapeutic targeting

Pre-treatment EEG analysis revealed slow/fast wave ratios of 2.7 in Subject 1 and 1.5 in Subject 2, indicating differing baseline disease severity. Prior research has demonstrated a positive correlation between elevated slow/fast wave ratios and cognitive impairment, consistent with the higher AD severity in Subject 1.^8,9 Although the overall slow/fast wave ratios remained stable post-treatment, regional analyses (threshold: >30% change) across 19 electrodes (10–20 system) revealed distinct patterns. Subject 1 showed >30% increases at Fp2, F8, and O1, while Subject 2 exhibited increases at P3, P4, T3, and T6, and decreases at F3, F7, O1, and Pz. These findings suggest that the frontal and occipital lobes may serve as early targets of lecanemab's therapeutic effects. Nonetheless, limited by the single-center design (n = 2) and short follow-up duration (2 months), these preliminary findings require validation in larger cohorts to substantiate claims regarding regional specificity of lecanemab's effects.

Multimodal biomarkers: transcending conventional cognitive metrics

This case further underscores the limitations of traditional cognitive scales (MoCA/MMSE) and the need for hierarchical assessment strategies. Performance-based ADL assessments (e.g., meal preparation, navigational competence) revealed treatment effects obscured by ceiling effects in MoCA/MMSE, echoing calls for real-world functional endpoints in AD trials.

Previous studies have established robust relationships between EEG biomarkers and Aβ/tau PET pathology. Specifically, Aβ⁺ patients validated by PET imaging exhibit enhancement of frontal/frontotemporal theta oscillations, and attenuation of mid-beta power in centroparietal regions. ¹² Concurrently, both Aβ and tau pathologies demonstrate significant associations with alpha peak frequency slowing. ¹³ Furthermore, functional connectivity alterations manifest as notably decreased intra-precuneus coherence, while exhibiting significant enhancement between the precuneus and bilateral inferior parietal lobules within low-frequency bands (theta/delta) in Aβ⁺ cohorts. ¹⁴

EEG provides a critical tool to recognize subclinical neurophysiology changes during early AD interventions. In patients with unchanged MoCA/MMSE scores, spectral power alterations—particularly decreased delta/theta (slow-wave) activity and increased alpha/beta (fast-wave) activity—may serve as early indicators of treatment response. The >30% regional slow/fast wave shifts in Subject 2, despite unremarkable Aβ-PET changes, support EEG's sensitivity as a biomarker of therapeutic efficacy.

Given the low Aβ and tau burden in early-stage AD, PET imaging alone risks underdiagnosis. Integrating PET with EEG and neuropsychological profiling may mitigate this limitation. PET provides molecular specificity (e.g., ¹⁸F-AV45, ¹⁸F-AV1451), while EEG offers temporally precise insights into network dynamics. ¹² Together, these modalities may improve early AD detection and reduce diagnostic latency, enabling timely therapeutic intervention.

While this case study offers valuable preliminary insights, its limitations must be emphasized. First, the two-month treatment window is insufficient to infer long-term therapeutic effects or disease-modifying outcomes. Second, the single-center design and small sample size (n = 2) limit the generalizability of the findings. Third, the clinical worsening in Subject 1 could be influenced by comorbid conditions (e.g., depression, sleep disturbances) rather than solely reflecting AD progression. Finally, the use of a flat dose of 400 mg (vs. the recommended 10 mg/kg) may have affected treatment efficacy. These constraints emphasize the exploratory nature of this report and preclude definitive conclusions.

Despite these limitations, this case study provides clinically meaningful insights. Future studies with larger samples, longer follow-up, and standardized dosing are needed to elucidate lecanemab's mechanisms, regional specificity, and EEG's utility in therapeutic monitoring.

Conclusion

In summary, this longitudinal case study offers preliminary insights into differential clinical and neurophysiological responses to lecanemab in two APOE ε3/ε4 AD patients stratified by severity. Our findings indicate that disparate outcomes may be partly attributable to baseline disease stage and neuropathological burden. Combined EEG and Aβ/tau PET imaging revealed spatially distinct cortical modulation after treatment, correlating significantly with clinical trajectories and molecular imaging data. Specifically, EEG-derived slow/fast wave ratios showed initial sensitivity in detecting early region-specific neurophysiological changes, supporting their utility as a dynamic, complementary biomarker alongside conventional PET and cognitive assessments. These results collectively reinforce that Aβ-targeted therapies may be more effective in early AD and underscore the prospect of multimodal biomarker strategies to transcend conventional scale limitations, enabling more personalized treatment monitoring. Larger cohorts and extended follow-up are warranted to validate these findings and elucidate the spatiotemporal dynamics of lecanemab in AD.