Alzheimer Disease
What Do You Need to Know?
- Aβ42 & Tau: amyloid plaques (Aβ42) are the pathologic hallmark; neurofibrillary tangles (hyperphosphorylated tau) correlate with clinical severity
- Biomarker framework (A/T/N): Amyloid (CSF Aβ42↓, amyloid PET+), Tau (CSF p-tau↑, tau PET+), Neurodegeneration (CSF t-tau↑, FDG-PET↓, MRI atrophy)
- Genetics: Early-onset AD — PSEN1 (chr 14, most common familial), APP (chr 21), PSEN2 (chr 1); Late-onset — APOE ε4 (strongest genetic risk factor, NOT deterministic)
- Imaging signature: medial temporal/hippocampal atrophy on MRI; temporoparietal hypometabolism on FDG-PET; posterior cingulate involved early
- Braak NFT staging: transentorhinal (I–II) → limbic (III–IV) → neocortical (V–VI); correlates with clinical progression
- Treatment: cholinesterase inhibitors (mild–moderate), memantine (moderate–severe), anti-amyloid antibodies (lecanemab, donanemab) — monitor for ARIA
- Atypical variants: posterior cortical atrophy (visual), logopenic PPA (language), frontal variant — all have AD pathology with different cortical targets
- ATN biomarker triad: A+ (amyloid PET or CSF Aβ42/40↓) + T+ (CSF p-tau181/p-tau217 or tau PET) + N (MRI hippocampal atrophy or FDG-PET temporoparietal hypometabolism) — A+T+ defines biologically diagnosed AD; A− rules AD off the continuum regardless of T/N
- Anti-amyloid mAbs (lecanemab, donanemab): FDA-approved for early symptomatic AD with amyloid+ biomarkers; APOE ε4 genotyping SHOULD be performed before treatment per FDA labels to counsel ARIA risk (ε4 homozygotes have the highest ARIA-E [~40% with donanemab] and ARIA-H risk) — not an absolute prerequisite; if not performed, treatment is not automatically prohibited but ARIA risk counseling is incomplete. Aducanumab withdrawn (Jan 2024)
- PSEN1 is the most common autosomal-dominant EOAD gene (chr 14, not APP); APP duplications → AD + CAA; trisomy 21 (Down syndrome) → near-universal AD pathology by age 40
- APOE ε4 is a RISK FACTOR, not deterministic — heterozygote 3–4×, homozygote ~12–15× (Fortea 2024 suggests ε4/ε4 may be a genetic form of AD); ε2 protective for AD but ↑ CAA hemorrhage risk
- Logopenic PPA = AD pathology in 70–90% (NOT FTD); impaired sentence repetition + word-finding pauses with preserved grammar/comprehension
- PCA (Benson syndrome) = visual variant of AD: Balint triad (simultanagnosia + optic ataxia + oculomotor apraxia) ± Gerstmann (left angular gyrus: finger agnosia + acalculia + agraphia + L/R confusion); memory preserved early
- Behavioral/dysexecutive (frontal) AD mimics bvFTD but is A+T+ on biomarkers — classic pitfall; biomarkers are the discriminator
- CSF t-tau markedly elevated (>10×) → think CJD, not AD; CSF Aβ42 is LOW in AD (sequestered in plaques) while p-tau is HIGH
- Neuritic plaques (CERAD) + Braak NFT staging required for definite neuropathologic AD diagnosis — diffuse plaques do NOT count; NFT burden (not amyloid load) correlates with clinical severity
- Always exclude reversible mimics before diagnosing neurodegenerative dementia: B12, TSH, structural disease (NPH, subdural hematoma), pseudodementia (depression). Add RPR/VDRL and HIV when history, age, tempo, exam, or risk factors support them — per AAN 2001 (reaffirmed), routine syphilis screening is not justified without risk factors or suggestive evidence
Clinical phenotype
- Amnestic progressive episodic memory loss + anomia + getting lost in familiar places → Classic (typical) AD
- Simultanagnosia + optic ataxia + oculomotor apraxia (Balint triad) → Posterior cortical atrophy (PCA / Benson syndrome) — visual variant of AD
- Finger agnosia + acalculia + agraphia + L/R confusion (Gerstmann syndrome) → Left angular gyrus involvement (often PCA / AD)
- Word-finding pauses + impaired sentence repetition with preserved grammar & single-word comprehension → Logopenic PPA (lvPPA) — usually AD pathology
- Frontal/behavioral mimic of bvFTD with positive AD biomarkers → Behavioral/dysexecutive (frontal variant) AD
- Preserved insight early + sundowning + Capgras delusion → Alzheimer disease
- Apperceptive prosopagnosia + alexia + environmental disorientation → PCA (visual variant AD)
Imaging / biomarkers
- Hippocampal/medial temporal lobe atrophy on coronal MRI (Scheltens MTA scale ≥2) → Alzheimer disease
- Bilateral temporoparietal + posterior cingulate/precuneus hypometabolism on FDG-PET with frontal sparing → Alzheimer disease
- Diffuse cortical amyloid uptake on PET (Pittsburgh compound B / florbetapir / flutemetamol / florbetaben) → Amyloid+ (AD continuum)
- Flortaucipir (tau PET) uptake in posterior cingulate / parietal / temporal cortex → AD-type 3R/4R PHF tau
- CSF Aβ42 LOW, Aβ42/Aβ40 ratio LOW, p-tau181/p-tau217 HIGH → Alzheimer disease
- Plasma p-tau217 elevated (Core 1 biomarker, 2024 NIA-AA) → Alzheimer disease
- FLAIR sulcal/parenchymal hyperintensity after anti-amyloid mAb → ARIA-E (vasogenic edema)
- T2*/SWI microbleeds or cortical superficial siderosis after anti-amyloid mAb → ARIA-H
- Occipitoparietal atrophy / hypometabolism with preserved hippocampi → PCA (visual variant AD)
- Lobar microbleeds + cortical superficial siderosis + lobar ICH in elderly (Boston criteria v2.0) → Cerebral amyloid angiopathy (CAA)
Pathology / genetics
- Extracellular Aβ42 neuritic plaques + intraneuronal NFTs (paired helical filaments, mixed 3R+4R hyperphosphorylated tau) + neuropil threads → Alzheimer disease
- Braak NFT staging: entorhinal/transentorhinal (I–II) → limbic (III–IV) → neocortical (V–VI) → AD tau spread (bottom-up)
- Thal amyloid phases 1→5: neocortex → allocortex → diencephalon → brainstem → cerebellum → AD amyloid spread (top-down)
- Granulovacuolar degeneration + Hirano bodies in hippocampal CA1 → Alzheimer disease (non-specific but classic)
- Aβ40 deposition in cortical/leptomeningeal vessel walls → lobar ICH + cortical siderosis → Cerebral amyloid angiopathy (CAA)
- APP duplication / mutation on chr 21 (also trisomy 21 / Down syndrome) → Early-onset AD by 40s ± CAA
- PSEN1 (chr 14) — most common autosomal-dominant EOAD, near-100% penetrance, may include seizures / myoclonus / spastic paraparesis → Familial AD
- PSEN2 (chr 1) — rare, Volga German families, later onset, incomplete penetrance → Familial AD
- APOE ε4 heterozygote 3–4×, ε4/ε4 homozygote ~12–15× risk → Strongest sporadic AD risk factor (impairs Aβ clearance)
- APOE ε2 → Protective for AD but increases CAA-related lobar hemorrhage risk
- TREM2 R47H variant, ABCA7 variants → Mild sporadic AD risk modifiers
- NIA-AA "ABC" score: A = Thal amyloid phase, B = Braak NFT stage, C = CERAD neuritic plaque score → Definite neuropathologic AD diagnosis
Pathophysiology & Amyloid Cascade Hypothesis
APP Processing
- Amyloid precursor protein (APP): transmembrane protein encoded on chromosome 21; normal function includes synaptic plasticity and neuronal survival
- Non-amyloidogenic pathway: α-secretase cleaves APP within the Aβ domain → soluble APPα (neuroprotective) → no Aβ produced
- Amyloidogenic pathway: β-secretase (BACE1) then γ-secretase cleave APP → Aβ peptides released
- Aβ40 vs. Aβ42: Aβ42 is more hydrophobic, more prone to aggregation, and is the predominant species in amyloid plaques
- γ-secretase complex contains presenilin 1 or 2 (PSEN1/PSEN2) — mutations increase Aβ42/Aβ40 ratio
Amyloid Cascade Hypothesis
- Aβ42 monomers → oligomers (most toxic species) → protofibrils → fibrils → insoluble amyloid plaques
- Soluble Aβ oligomers impair synaptic function, induce oxidative stress, trigger neuroinflammation (microglial activation)
- Amyloid deposition → downstream tau hyperphosphorylation → neurofibrillary tangles (NFTs) → neuronal death
- Key distinction: amyloid plaques correlate poorly with clinical severity; NFT burden correlates with cognitive decline
Tau Pathology
- Tau is a microtubule-associated protein; normal function = stabilizes axonal microtubules
- Hyperphosphorylation → tau detaches from microtubules → forms paired helical filaments (PHFs) → NFTs
- NFTs spread in a stereotyped pattern (Braak staging) from entorhinal cortex → hippocampus → neocortex
- Prion-like propagation: tau spreads trans-synaptically along connected neural networks
- Aβ plaques are necessary but not sufficient for AD diagnosis — cognitively normal elderly can have extensive amyloid
- NFT burden (not amyloid plaque load) best correlates with clinical severity and cognitive decline
- γ-secretase contains presenilin — this links PSEN1/PSEN2 mutations directly to increased Aβ42 production
- Down syndrome (trisomy 21) → extra copy of APP gene → virtually all develop AD pathology by age 40
Clinical Stages & Diagnostic Criteria
Clinical Stages of Alzheimer Disease
| Stage | Cognition | Function | Key Features |
|---|
| Preclinical AD | Normal | Normal | Biomarker-positive only (amyloid PET+ or CSF Aβ42↓); no symptoms; research framework only |
| MCI due to AD | Impaired (1–1.5 SD below mean) | Preserved (or minimal impairment) | Amnestic MCI most common; episodic memory loss (hippocampal); ~10–15% convert to AD dementia per year |
| Mild AD Dementia | Impaired | Impaired IADLs | Repetitive questions, getting lost, difficulty with finances/medications; word-finding difficulty |
| Moderate AD Dementia | Significantly impaired | Needs assistance with BADLs | Cannot dress/bathe independently; behavioral symptoms (agitation, wandering, sundowning); delusions |
| Severe AD Dementia | Profoundly impaired | Fully dependent | Loss of speech, incontinence, dysphagia, bed-bound; death from aspiration pneumonia, infection, or inanition |
NIA-AA Diagnostic Framework (2018)
- Defines AD biologically using the A/T/N system — independent of clinical symptoms
- A (Amyloid): CSF Aβ42↓ (or Aβ42/40 ratio↓), amyloid PET positive
- T (Tau pathology): CSF p-tau↑ (phosphorylated tau-181 or -217), tau PET positive
- N (Neurodegeneration): CSF t-tau↑, FDG-PET temporoparietal hypometabolism, MRI hippocampal/cortical atrophy
- A+T+N+ = full AD pathological change; A+T−N− = Alzheimer pathologic change (preclinical)
- A− = NOT on the Alzheimer continuum regardless of T or N status
NIA-AA Revised Criteria (2024) — Jack et al., Alzheimers Dement
- ATN expanded to A / T / N + I (inflammation) + V (vascular): reflects the multifactorial nature of late-life cognitive decline (neuroinflammation via GFAP/microglial markers; vascular contribution via white matter hyperintensities, infarcts, CAA)
- Biomarkers reclassified into two tiers:
- Core 1 (establish AD pathology, can be used for diagnosis): amyloid PET, CSF Aβ42/40 ratio, and plasma p-tau217-based assays (validated for AD pathology — interpret with assay-specific intended use; the FDA-cleared Lumipulse G pTau217/Aβ42 Plasma Ratio is indicated to aid diagnosis in cognitively impaired adults in specialty care, not as a stand-alone screen)
- Core 2 (stage and prognosticate): tau PET, other fluid p-tau species (p-tau181, p-tau231), MTBR-tau243
- Biological diagnosis of AD can now be made on the basis of a positive Core 1 biomarker alone — clinical symptoms determine staging, not diagnosis
- Plasma p-tau217 (alone or as a ratio with Aβ42/40) is the first blood-based Core 1 biomarker family in the NIA-AA framework; the Fujirebio Lumipulse G pTau217/Aβ42 Plasma Ratio is the first FDA-cleared blood test for AD pathology — intended to aid diagnosis in cognitively impaired adults in specialty care, not as a stand-alone screen. PrecivityAD2 / ALZpath / Quanterix are clinically used and validated as LDTs but are NOT FDA-cleared
| A/T/N Profile | Classification |
|---|
| A+T+N+ | Alzheimer disease (full pathological change + neurodegeneration) |
| A+T+N− | Alzheimer disease (pathological change without neurodegeneration yet) |
| A+T−N− | Alzheimer pathologic change (amyloid only; preclinical) |
| A+T−N+ | Alzheimer pathologic change + non-AD neurodegeneration |
| A−T+N+ | Non-AD pathologic change (e.g., PART, primary tauopathy) |
| A−T−N+ | Non-AD neurodegeneration (suspected non-Alzheimer pathology — SNAP) |
| A−T−N− | Normal biomarkers |
- Amnestic MCI (episodic memory loss with preserved function) is the classic prodrome of AD; converts to AD dementia at ~10–15% per year
- A/T/N framework: you MUST be A+ to be on the Alzheimer continuum — A− with T+ and/or N+ is NOT AD
- PART (Primary Age-Related Tauopathy): NFTs limited to medial temporal lobe WITHOUT amyloid — A−T+; distinct from AD
Biomarkers
CSF Biomarkers
| Biomarker | Direction in AD | What It Reflects | Notes |
|---|
| Aβ42 | ↓ Decreased | Amyloid sequestration in plaques (less free in CSF) | Aβ42/40 ratio improves specificity over Aβ42 alone |
| p-tau (181 or 217) | ↑ Increased | Tau phosphorylation (AD-specific tau pathology) | p-tau217 has highest accuracy for differentiating AD from non-AD dementias |
| t-tau (total tau) | ↑ Increased | Neuronal injury/neurodegeneration (non-specific) | Also elevated in CJD (markedly), stroke, TBI; not AD-specific |
PET Biomarkers
| PET Tracer | Target | AD Pattern | Key Points |
|---|
| Amyloid PET (florbetapir, florbetaben, flutemetamol, Pittsburgh compound B) | Fibrillar Aβ plaques | Diffuse cortical uptake (frontal, parietal, temporal, posterior cingulate/precuneus) | Prevalence in cognitively normal elderly is age-dependent, rising with each decade — approximately 10–40% of cognitively normal elderly (lower at age 65, higher at age 85+); a negative amyloid PET makes biologically defined AD very unlikely |
| Tau PET (flortaucipir / AV-1451) | Paired helical filament tau (mixed 3R/4R) | Medial temporal → lateral temporal → parietal → frontal (mirrors Braak staging) | More closely correlates with cognitive symptoms than amyloid PET; FDA-approved 2020. Flortaucipir selectively binds AD-type 3R/4R PHF tau; poor binding to pure 3R tauopathies (Pick disease) or pure 4R tauopathies (PSP, CBD) — thus a positive scan supports AD-type tau specifically |
| FDG-PET | Glucose metabolism (neuronal activity) | Temporoparietal & posterior cingulate/precuneus hypometabolism | Frontal sparing early; primary sensorimotor and visual cortex spared; pattern distinguishes AD from FTD |
Plasma Biomarkers
- Plasma p-tau217: most promising blood-based biomarker; high accuracy for detecting AD pathology; comparable to CSF p-tau
- Plasma p-tau181: elevated in AD; less accurate than p-tau217 but still useful for screening
- Plasma Aβ42/40 ratio: decreased in AD; modest accuracy alone; improves when combined with p-tau
- Plasma GFAP (glial fibrillary acidic protein): reflects astrocyte activation; elevated early in AD continuum
- NfL (neurofilament light chain): non-specific marker of neurodegeneration; elevated in AD but also in FTD, ALS, MS
- CSF Aβ42 is LOW in AD (trapped in plaques) — do not confuse with tau which is HIGH
- Negative amyloid PET has high negative predictive value — a negative scan makes biologically defined AD very unlikely (clinical syndrome should be reassessed for non-AD etiologies)
- CSF t-tau markedly elevated (>10×) → think CJD, not AD (AD has modest elevation)
- p-tau217 is emerging as the best single blood-based biomarker for AD detection
- FDG-PET pattern: AD = temporoparietal hypometabolism; FTD = frontal/anterior temporal hypometabolism — key differentiator
Genetics
Early-Onset Familial AD (Autosomal Dominant)
| Gene | Chromosome | Protein | Key Features |
|---|
| PSEN1 | 14 | Presenilin 1 (γ-secretase component) | Most common cause of early-onset familial AD; onset 30–60 yr (mean ~45); >300 mutations; nearly 100% penetrance; may present with seizures, myoclonus, spastic paraparesis |
| APP | 21 | Amyloid precursor protein | Mutations near secretase cleavage sites → ↑Aβ42 production; onset 40–65 yr; duplications cause AD + cerebral amyloid angiopathy (CAA) |
| PSEN2 | 1 | Presenilin 2 (γ-secretase component) | Rarest of the three; later onset (40–75 yr); incomplete penetrance; originally described in Volga German families |
Late-Onset (Sporadic) AD — Genetic Risk Factors
| Gene/Allele | Chromosome | Effect | Key Details |
|---|
| APOE ε4 | 19 | Risk factor (↑) | Strongest genetic risk factor for late-onset AD; 1 copy → 3–4× risk; 2 copies (ε4/ε4) → ~12–15× risk; recent data (Fortea et al., Nat Med 2024) suggest ε4/ε4 may represent a genetic form of AD with near-complete penetrance; impairs Aβ clearance; heterozygous ε4 remains a risk factor, NOT deterministic |
| APOE ε2 | 19 | Protective (↓) for AD; ↑ for CAA hemorrhage | ~40% reduced risk for AD compared to ε3; delays onset; ε2/ε2 is most protective for AD; associated with type III hyperlipoproteinemia. Note: ε2 paradoxically increases risk of CAA-related lobar hemorrhage (vasculopathic effect on vessel walls) |
| APOE ε3 | 19 | Neutral (reference) | Most common allele (~78% of population); considered baseline risk |
| TREM2 | 6 | Risk factor | Microglial receptor; R47H variant → 2–4× increased AD risk; impairs microglial clearance of Aβ |
Down Syndrome & AD
- Trisomy 21: three copies of chromosome 21 → three copies of APP gene → lifelong Aβ overproduction
- Virtually all individuals with Down syndrome develop AD neuropathology by age 40
- Clinical dementia onset typically in the 50s; accelerated course
- Higher risk of early-onset seizures with AD
- PSEN1 (chr 14) = most common cause of early-onset familial AD — NOT APP
- APOE ε4 is a risk factor, NOT a deterministic gene — it does not cause AD; many carriers never develop disease
- APOE ε4 homozygotes (ε4/ε4): ~12–15× increased risk + higher ARIA rates with anti-amyloid antibodies; Fortea et al. (Nat Med 2024) propose ε4/ε4 may represent a genetic form of AD with near-complete penetrance
- Down syndrome (trisomy 21): extra copy of APP on chr 21 → AD pathology by age 40; clinical dementia in 50s
- Early-onset familial AD = autosomal dominant, <5% of all AD; accounts for only ~1% of total AD cases
- PSEN2 onset can extend into the typical late-onset range, blurring the early/late-onset distinction — do not exclude PSEN2 in apparent late-onset familial cases
Neuroimaging
Structural MRI
- Hippocampal atrophy: earliest and most sensitive structural finding; medial temporal lobe volume loss on coronal T1
- Entorhinal cortex atrophy: precedes hippocampal atrophy; earliest region affected
- Posterior parietal atrophy: precuneus involvement; correlates with episodic memory and visuospatial deficits
- Progressive pattern: medial temporal → lateral temporal → parietal → frontal (mirrors Braak staging)
- Medial temporal atrophy (MTA) score: visual rating scale on coronal MRI; MTA ≥2 supports AD
- Primary motor, sensory, and visual cortices are relatively spared until late
FDG-PET Patterns by Dementia Type
| Dementia | FDG-PET Pattern |
|---|
| AD | Temporoparietal + posterior cingulate/precuneus hypometabolism; frontal sparing early |
| FTD (behavioral variant) | Frontal + anterior temporal hypometabolism |
| DLB | Temporoparietal + occipital hypometabolism (occipital involvement distinguishes from AD) |
| PCA (visual variant AD) | Occipitoparietal hypometabolism |
- AD on FDG-PET: temporoparietal + posterior cingulate hypometabolism with frontal sparing — this pattern is the classic board answer
- DLB vs. AD on FDG-PET: both have temporoparietal hypometabolism, but DLB uniquely includes occipital hypometabolism
- Hippocampal atrophy on coronal MRI is the most tested structural finding in AD
Neuropathology & Staging
Microscopic Hallmarks
| Finding | Composition | Location | Key Details |
|---|
| Neuritic (senile) plaques | Aβ42 core + dystrophic neurites + activated microglia | Extracellular; neocortex > hippocampus | Required for definite AD diagnosis (NIA-AA neuropathologic criteria); correlate poorly with clinical severity |
| Diffuse plaques | Aβ without neuritic changes | Widespread cortex | Can be seen in normal aging; NOT counted for AD diagnosis (CERAD scoring counts neuritic plaques only) |
| Neurofibrillary tangles (NFTs) | Hyperphosphorylated tau (paired helical filaments) | Intraneuronal; entorhinal → hippocampus → neocortex | Correlate with clinical severity; Braak staging based on NFT distribution |
| Neuropil threads | Tau-positive dystrophic neurites | Neuropil surrounding plaques and tangles | Abundant; contribute to tau pathology burden |
| Granulovacuolar degeneration | Intraneuronal vacuoles with dense granules | Hippocampal pyramidal neurons | Not specific to AD but prominent in hippocampus |
| Hirano bodies | Actin-containing eosinophilic rod-shaped inclusions | Hippocampal CA1 | Non-specific; seen in aging and AD |
Braak Neurofibrillary Tangle Staging
| Stage | Region | Clinical Correlate |
|---|
| I–II (Transentorhinal) | Transentorhinal cortex, entorhinal cortex | Preclinical; cognitively normal or subtle memory changes |
| III–IV (Limbic) | Hippocampus (CA1), amygdala, anterior thalamic nucleus (variable) | MCI; early clinical symptoms; episodic memory impairment |
| V–VI (Neocortical) | Association neocortex (temporal, parietal, frontal) | Dementia; moderate to severe AD; widespread cognitive decline |
Thal Amyloid Phases
| Phase | Region of Amyloid Deposition |
|---|
| 1 | Neocortex (frontal, parietal, temporal, occipital) |
| 2 | Allocortex (entorhinal, hippocampus, amygdala, cingulate) |
| 3 | Diencephalon (striatum, basal forebrain, thalamus) |
| 4 | Brainstem |
| 5 | Cerebellum |
- Key contrast: Thal phases (amyloid) spread top-down (neocortex → brainstem → cerebellum); Braak staging (tau) spreads bottom-up (entorhinal → hippocampus → neocortex)
Mixed Pathology in AD (Pure AD is Uncommon in the Elderly)
- Pure AD is the exception, not the rule, in older patients — autopsy series consistently show overlapping pathologies
- AD + cerebrovascular disease: co-occurs in ~30% of AD brains (lacunar infarcts, white matter disease, CAA-related injury)
- AD + Lewy body co-pathology: Lewy body inclusions identified in ~30–50% of AD brains (often limbic-predominant); contributes to fluctuations and visual hallucinations when present
- AD + TDP-43 (LATE — Limbic-predominant Age-related TDP-43 Encephalopathy): TDP-43 inclusions found in ~25–50% of AD brains; LATE can mimic or coexist with AD, accelerates hippocampal atrophy, and contributes to amnestic phenotype in the very old (>80 yr)
- Clinical implication: antemortem biomarkers are amyloid- and tau-specific — concurrent vascular, Lewy, or TDP-43 pathology is typically under-recognized in life and may explain heterogeneous trajectories and incomplete response to anti-amyloid therapy
Cerebral Amyloid Angiopathy (CAA)
- Aβ40 (not Aβ42) deposits in walls of cortical and leptomeningeal blood vessels
- Present in >80% of AD brains; can occur independently of AD
- Clinical consequences: lobar intracerebral hemorrhage, cortical superficial siderosis, microbleeds
- Boston criteria v2.0: lobar hemorrhage + cortical superficial siderosis + lobar microbleeds on MRI
- CAA increases risk of ARIA with anti-amyloid antibody therapy
- Neuritic plaques (NOT diffuse plaques) are required for neuropathologic AD diagnosis — CERAD scoring counts neuritic plaques only
- Braak NFT stages correlate with clinical severity: I–II = preclinical; III–IV = MCI; V–VI = dementia
- Thal (amyloid) = top-down spread (cortex → cerebellum); Braak (tau) = bottom-up spread (entorhinal → neocortex) — opposite patterns
- CAA: Aβ40 in vessel walls (not Aβ42); causes lobar hemorrhages; present in >80% of AD brains
- NIA-AA neuropathologic criteria (ABC score): A = Thal amyloid phase, B = Braak NFT stage, C = CERAD neuritic plaque score
Treatment
Cholinesterase Inhibitors (ChEIs)
| Drug | Mechanism | Indication | Key Side Effects / Notes |
|---|
| Donepezil | Reversible AChE inhibitor | Mild–severe AD (only ChEI approved for all stages) | Nausea, diarrhea, insomnia, vivid dreams, bradycardia; once-daily dosing; available as transdermal patch |
| Rivastigmine | Pseudo-irreversible AChE + BuChE inhibitor | Mild–moderate AD; also approved for PDD | GI side effects; transdermal patch preferred (better tolerability); only ChEI approved for Parkinson disease dementia |
| Galantamine | Reversible AChE inhibitor + allosteric nicotinic receptor modulator | Mild–moderate AD | Dual mechanism; GI side effects; avoid in severe hepatic/renal impairment |
- Cholinergic hypothesis: nucleus basalis of Meynert degeneration → cortical ACh deficit → memory impairment
- Modest symptomatic benefit; do NOT slow disease progression
- Monitor for bradycardia, syncope, GI effects; avoid in sick sinus syndrome
NMDA Receptor Antagonist
- Memantine: uncompetitive NMDA receptor antagonist; reduces glutamate excitotoxicity
- Approved for moderate–severe AD; often combined with donepezil (Namzaric = combination)
- Generally well tolerated; side effects: dizziness, headache, confusion
Anti-Amyloid Monoclonal Antibodies
| Drug | Target | Approval/Status | Key Trial / Notes |
|---|
| Aducanumab (historical interest only) | Aggregated Aβ (plaques + oligomers) | FDA accelerated approval 2021 (controversial); DISCONTINUED by Biogen January 2024 — no longer commercially available | EMERGE/ENGAGE trials; inconsistent efficacy; reduced amyloid on PET; ARIA-E ~35% at the high (10 mg/kg) dose, higher in APOE ε4 carriers |
| Lecanemab | Aβ protofibrils (soluble aggregates) | FDA full approval 2023 | CLARITY AD trial; 27% slowing of decline on CDR-SB at 18 months; ARIA-H ~14% microhemorrhages (~17% when combined with superficial siderosis); ARIA-E ~13%; symptomatic ARIA ~3%; 10 mg/kg IV every 2 weeks. Baseline MRI + monitoring MRI before infusions 3, 5, 7, and 14. After 18 months of induction, maintenance options: continue 10 mg/kg IV q2 weeks, transition to IV q4 weeks, or switch to weekly subcutaneous LEQEMBI IQLIK 360 mg |
| Donanemab | N-terminal pyroglutamate Aβ (deposited plaques) | FDA approved 2024 | TRAILBLAZER-ALZ 2; 35% slowing of decline in the low/medium tau subgroup; FDA-labeled eligibility = confirmed amyloid β pathology (amyloid PET or CSF) — tau PET was used in the trial and can inform staging/prognosis but is NOT a labeled prerequisite; treatment may be stopped after amyloid PET clears; ARIA-E ~24% overall, ~40% in APOE ε4 homozygotes; symptomatic ARIA ~6%; MRI before infusions 2, 3, 4, and 7 |
ARIA (Amyloid-Related Imaging Abnormalities)
- ARIA-E: vasogenic edema/sulcal effusions; usually asymptomatic; resolves with dose interruption; on MRI = FLAIR hyperintensity
- ARIA-H: microhemorrhages or superficial siderosis; on MRI = GRE/SWI hypointensities
- Risk factors for ARIA: APOE ε4 carriers (especially homozygotes), pre-existing CAA/microbleeds, higher doses, anticoagulant use
- APOE ε4 homozygotes (ε4/ε4): ARIA-E rate ~40% with donanemab; lecanemab ARIA-E also markedly higher in ε4 homozygotes
- Monitoring schedule:
- Lecanemab: baseline MRI + monitoring MRI before infusions 3, 5, 7, and 14. After 18 months: maintenance options are continued IV q2 weeks, IV q4 weeks, or weekly subcutaneous LEQEMBI IQLIK 360 mg
- Donanemab: baseline MRI + MRIs before infusions 2, 3, 4, and 7
- APOE ε4 genotyping should be performed before therapy per FDA label (to counsel ARIA risk, especially in ε4 homozygotes); it is recommended, not an absolute prerequisite — if not performed, treatment is not automatically prohibited but ARIA counseling is less complete
- Symptomatic ARIA: headache, confusion, visual disturbance; rarely serious (macrohemorrhage, death) — symptomatic ARIA ~3% with lecanemab and ~6% with donanemab
- Donepezil is the only ChEI approved for all stages (mild through severe) of AD
- Rivastigmine is the only ChEI also approved for Parkinson disease dementia (PDD)
- Memantine is indicated for moderate–severe AD — NOT mild AD
- ARIA risk is highest in APOE ε4 homozygotes — per FDA label, APOE ε4 genotyping should be performed before initiating anti-amyloid therapy to counsel ARIA risk; it is recommended, not required, and treatment is not automatically prohibited if it is not performed (counseling is just less complete)
- Lecanemab targets protofibrils (soluble); donanemab targets deposited plaque (pyroglutamate Aβ) — different mechanisms
- ChEIs and memantine provide symptomatic benefit only; anti-amyloid antibodies are the first disease-modifying therapies (lecanemab and donanemab are currently marketed; aducanumab was discontinued by Biogen in January 2024 and is historical only)
Atypical Presentations
Atypical AD Variants
| Variant | Core Deficit | Atrophy Pattern | Key Features |
|---|
| Posterior Cortical Atrophy (PCA) | Visuospatial / visuoperceptual | Occipitoparietal, posterior temporal | Progressive visual dysfunction (simultanagnosia, optic ataxia, oculomotor apraxia = Balint syndrome); alexia; visual agnosia; environmental disorientation; memory relatively preserved early; most common cause is AD pathology; young onset (50s–60s) |
| Logopenic Primary Progressive Aphasia (lvPPA) | Word-finding / sentence repetition | Left posterior temporal & inferior parietal (angular gyrus, posterior superior temporal) | Frequent word-finding pauses; impaired sentence/phrase repetition; phonologic errors; spared grammar and motor speech; underlying pathology is AD in ~70–80% of cases (some series up to 90%) |
| Frontal Variant AD | Executive / behavioral | Frontal lobes | Executive dysfunction, apathy, or disinhibition mimicking bvFTD; younger onset; AD pathology confirmed at autopsy; biomarkers (amyloid PET, CSF) distinguish from FTD |
| Corticobasal Syndrome (CBS) due to AD | Asymmetric apraxia / parkinsonism | Asymmetric frontoparietal | Limb apraxia, alien limb, myoclonus, cortical sensory loss, asymmetric rigidity; ~25% of CBS cases have AD pathology (not CBD) |
Key Differentiating Feature of lvPPA from Other PPAs
| PPA Variant | Fluency | Repetition | Single-Word Comprehension | Grammar | Typical Pathology |
|---|
| Logopenic (lvPPA) | Reduced (word-finding pauses) | Impaired | Preserved | Preserved | AD (~70–80%, up to 90%) |
| Nonfluent/Agrammatic (nfvPPA) | Effortful, halting | Relatively preserved | Preserved | Impaired | Tau (4R, CBD, PSP) |
| Semantic (svPPA) | Fluent | Preserved | Impaired | Preserved | TDP-43 (FTLD) |
- PCA (visual variant AD): Balint syndrome (simultanagnosia + optic ataxia + oculomotor apraxia) with occipitoparietal atrophy; memory preserved early; most cases are AD pathology
- lvPPA: the only PPA variant where the most common underlying pathology is AD; key finding = impaired sentence repetition with preserved grammar
- ~25% of corticobasal syndrome (CBS) is caused by AD pathology (not CBD) — biomarkers help differentiate
- All atypical AD variants share the same amyloid/tau pathology — they differ only in which cortical region bears the greatest burden
Differential Diagnosis
AD vs. Other Common Dementias
| Feature | AD | DLB | FTD (bvFTD) | Vascular Dementia | NPH |
|---|
| Core Deficit | Episodic memory (hippocampal) | Attention, visuospatial, fluctuations | Behavior, executive function, personality | Executive, processing speed | Gait, urinary, cognitive (triad) |
| Age at Onset | Typically >65 | >50 | 45–65 (younger than AD) | Variable | >60 |
| Memory | Early, prominent encoding deficit | Fluctuating; retrieval > encoding | Relatively preserved early | Retrieval deficit (improves with cues) | Subcortical pattern (slow retrieval) |
| Hallucinations | Late feature | Early, visual (detailed, recurrent) | Uncommon | Uncommon | Uncommon |
| Motor | Normal until late | Parkinsonism; RBD | Normal early | Focal deficits; gait disorder | Magnetic gait (wide-based, shuffling) |
| MRI | Hippocampal atrophy | Relatively preserved hippocampus | Frontal/anterior temporal atrophy | White matter disease; lacunar infarcts | Ventriculomegaly out of proportion to sulcal atrophy |
| FDG-PET | Temporoparietal ↓ | Temporoparietal + occipital ↓ | Frontal/anterior temporal ↓ | Multifocal ↓ | Global ↓ (nonspecific) |
Normal Aging vs. MCI vs. AD Dementia
| Feature | Normal Aging | MCI | AD Dementia |
|---|
| Subjective complaints | Occasional word-finding, name recall | Noticeable decline (self or informant) | Significant, often minimized by patient (anosognosia) |
| Objective testing | Within normal range for age | 1–1.5 SD below mean | ≥2 SD below mean; multiple domains |
| Daily function | Fully independent | Independent (or minimal difficulty with complex tasks) | Impaired IADLs (mild) → BADLs (moderate–severe) |
| Progression | Stable over years | May be stable, improve, or progress to dementia (~10–15%/yr to AD) | Progressive decline; no reversal |
Reversible Causes of Cognitive Decline to Exclude
- Metabolic: hypothyroidism (TSH), B12 deficiency, hepatic/uremic encephalopathy, hyponatremia
- Infectious: HIV, syphilis (RPR/VDRL), Lyme (endemic areas)
- Structural: NPH, subdural hematoma, brain tumor
- Psychiatric: depression (pseudodementia) — key mimic; responds to antidepressants
- Toxic/Drug: anticholinergics, benzodiazepines, opioids, alcohol
- Core/common workup: medication review, depression/delirium/sleep screen, CBC, CMP, TSH, B12, and structural neuroimaging (MRI > CT). Add RPR/VDRL, HIV, inflammatory, autoimmune, CSF, EEG, and genetic tests when history, age, tempo, exam, or risk factors support them (per AAN 2001, reaffirmed; routine RPR/HIV are not required without risk factors)
- Pseudodementia (depression): patients complain about memory (vs. AD patients often unaware); answer "I don't know" (vs. AD patients confabulate); onset is more acute; responds to antidepressant treatment
- DLB vs. AD: relatively preserved hippocampi on MRI + occipital hypometabolism on FDG-PET favors DLB; early visual hallucinations + parkinsonism + fluctuating cognition = DLB
- Always check for reversible causes before diagnosing a neurodegenerative dementia
References
- Jack CR Jr, Bennett DA, Blennow K, et al. NIA-AA Research Framework: toward a biological definition of Alzheimer's disease. Alzheimers Dement. 2018;14(4):535-562.
- Knopman DS, Amieva H, Petersen RC, et al. Alzheimer disease. Nat Rev Dis Primers. 2021;7(1):33.
- van Dyck CH, Swanson CJ, Aisen P, et al. Lecanemab in early Alzheimer's disease (CLARITY AD). N Engl J Med. 2023;388(1):9-21.
- Sims JR, Zimmer JA, Evans CD, et al. Donanemab in early symptomatic Alzheimer disease (TRAILBLAZER-ALZ 2). JAMA. 2023;330(6):512-527.
- Braak H, Braak E. Neuropathological stageing of Alzheimer-related changes. Acta Neuropathol. 1991;82(4):239-259.
🔒
Continue reading — sign in
The full note has more clinical pearls, tables, and board-focused tips. Free account, no fee.
