Amyloid PET

Amyloid PET shows beta amyloid plaques in the brain in living patients. A radiolabelled tracer binds to fibrillar beta amyloid in the cortex, and the amount of binding shows whether plaques are present. It is used in patients with cognitive impairment to help diagnose Alzheimer disease or to exclude it.

Three fluorine-18 tracers are approved: florbetapir (about 370 MBq, imaging at 30 to 50 minutes), florbetaben (about 300 MBq, imaging at 90 minutes) and flutemetamol (about 185 MBq, imaging at 90 minutes). A carbon-11 amyloid tracer is used in research only.

The acquisition takes 10 to 20 minutes with a low dose CT or MRI for attenuation correction.

The effective dose is about 6 to 7 mSv. Alzheimer disease affects millions of people, and pathological changes begin 15 to 20 years before symptoms.

Clinical diagnosis is wrong in about 20 to 30% of patients when checked against pathology, especially in early or atypical disease. Amyloid PET has a sensitivity of 90 to 95% and a specificity of 85 to 90% against pathology.

Its main value is a negative result, which makes Alzheimer disease very unlikely as the cause of symptoms. A positive result shows plaques are present but does not prove they cause the symptoms.

Indications.

• Persistent or progressive unexplained mild cognitive impairment.
• Atypical clinical presentation of dementia, or early onset (under 65 years) where the cause is uncertain.
• Dementia with mixed clinical features where Alzheimer disease is in the differential.
• Selection of patients for anti-amyloid antibody treatment (such as lecanemab or donanemab) and monitoring of plaque clearance.
• Research studies.

Not for routine use in typical Alzheimer disease, in people without symptoms, in people with only a family history or a genetic risk (such as apolipoprotein E4), or to determine severity of dementia. Not as a replacement for clinical assessment.

Contraindications and precautions.

• Pregnancy: avoid unless essential. Breast feeding: short interruption as the local rule requires.
• Allergy to the tracer is very rare.
• The patient must lie still for 10 to 20 minutes. Sedation is rarely required.
• No fasting is needed. Routine medicines are continued.
• Counselling before the scan: the result may show amyloid in a person who will not develop dementia, and a positive result has implications for the patient and family.
• Capacity and consent should be checked, since the result may be sensitive.
• Brain scans or MRI should be done first to exclude other treatable causes.
• Insurance, driving and employment implications of a positive result should be discussed in advance.

Preparation and technique.

The patient is given the tracer through a venous line and rests for the uptake period (30 to 90 minutes, depending on the tracer). The patient then lies in the scanner with the head in a holder, and a low dose CT (or MRI) is taken for attenuation correction.

PET images of the brain are taken for 10 to 20 minutes. Images are reconstructed and displayed in transaxial, coronal and sagittal planes with a grey scale or colour scale, with the colour scale adjusted as the manufacturer's approved method requires.

Reading is visual by a trained reader, using the approved method for the tracer. Semi-quantitative values (uptake ratio to cerebellum, or a calibrated scale) are used in research and treatment monitoring.

Normal appearances.

In a negative scan, tracer is retained in the white matter, and the grey matter is not labelled, so there is a clear contrast between grey and white matter, with a pattern similar to a healthy brain.

Uptake is low in the cortex and high in the pons and cerebellar white matter (non-specific binding).

Interpretation.

• Positive scan: tracer in the cortical grey matter in at least one cortical region (frontal, lateral temporal, parietal, posterior cingulate and precuneus), with loss of the grey white contrast. It indicates moderate to frequent plaques.
• Negative scan: no cortical grey matter tracer, with white matter uptake only. It makes Alzheimer disease very unlikely as the cause of cognitive impairment, and other causes should be sought.
• Equivocal scans are read together with semi-quantitative values and repeat reading.
• Positive result in a cognitively impaired patient: consistent with Alzheimer disease, but plaques are also seen in dementia with Lewy bodies, cerebral amyloid angiopathy and in cognitively normal older people (10 to 30%, increasing with age and apolipoprotein E4).
• Positive result in a patient with atypical features: a mixed pathology is possible, and the clinical picture decides.
• Treatment monitoring: a fall in uptake after anti-amyloid antibody treatment indicates plaque removal.
• Amyloid angiopathy: occipital predominance with microhaemorrhages on MRI.
• Results are reported as positive, negative or equivocal, with the quantitative values.

Critical values and findings needing immediate action.

• The result rarely needs emergency action.
• Incidental findings: a mass lesion, large infarct, haemorrhage or hydrocephalus on the CT or MRI requires urgent neurology or neurosurgical review.
• Positive result in a patient being considered for anti-amyloid treatment: baseline MRI for microhaemorrhages and apolipoprotein E testing before treatment.
• Psychological distress after disclosure: access to counselling and support.
• Safeguarding and capacity issues in patients who cannot consent: handled by the clinical team.

Limitations and artefacts.

• A positive scan does not prove that amyloid causes the symptoms. It is common in cognitively normal older people.
• A positive scan does not separate Alzheimer disease from dementia with Lewy bodies or amyloid angiopathy, or from mixed pathology.
• A negative scan does not rule out other neurodegenerative disease, vascular disease, or tauopathies.
• Cortical atrophy, white matter hyperintensities and misregistration cause errors in reading.
• Scanner differences, reconstruction methods and reader training affect the result.
• Tracers differ in white matter binding and image appearance, and each tracer needs its own reading method.
• Cost and availability are major limits.
• Scans do not measure tau or neurodegeneration, which correlate better with symptoms.
• Early disease with low plaque load can be missed by visual reading.
• Ethical issues need attention: information about risk in people without symptoms.

Special groups.

• Younger patients (under 65): the strongest case for use, because atypical presentations are common.
• Older patients: positive results are common and less specific.
• Patients with Down syndrome: high rates of amyloid, and the test has a research role.
• Patients with suspected amyloid angiopathy: combined with MRI.
• Pregnant patients: avoid.
• Regions: dementia is under diagnosed and cause specific testing is unavailable. Amyloid PET is available in very few centres, so history, cognitive tests, laboratory tests for reversible causes, CT or MRI and clinical follow up are used.

Follow up tests.

• MRI of the brain with volumetric, vascular and haemorrhage sensitive sequences.
• Fluorodeoxyglucose PET for neurodegeneration patterns.
• Cerebrospinal fluid amyloid and tau markers, or plasma markers (phosphorylated tau, amyloid ratio).
• Tau PET in selected centres.
• Neuropsychological tests.
• Blood tests for reversible causes (thyroid function, vitamin B12, calcium, glucose, renal and liver function, syphilis and HIV tests where relevant).
• Apolipoprotein E genotyping before anti-amyloid antibody treatment.
• Referral to a memory clinic, neurology, geriatrics or psychiatry.