Study characteristics

Patient sampling

Forty‐five older adults who were referred from local memory clinics in Australia. No further details of patient sampling and recruitment were reported.

Participants had to be at least 60 years old and to have had at least 7 years of formal schooling. They were also required to communicate fluently in English.

Exclusion criteria included the presence of dementia, a score lower than 23 on the MMSE, the presence of other conditions that may impair their cognition and independence, including other neurological (stroke, multiple sclerosis, epilepsy, moderate‐severe traumatic brain injury), psychiatric (psychotic symptoms, bipolar disorder), or substance use conditions (e.g. drug and alcohol dependence, use of acetylcholinesterase inhibitors and memantine.

Patient characteristics and setting

45 MCI participants diagnosed by the Petersen 2004 and Winblad 2004 criteria.

Demographic data were reported for 45 MCI participants (with SUVR); demographic data were not available for those classified by visual assessment.

¹⁸F‐florbetaben positive: 24, ¹⁸F‐florbetaben negative: 21

Gender: not described

Age Mean (SD): ¹⁸F‐florbetaben positive: 73.5 (6.9), ¹⁸F‐florbetaben negative:71.8 (6.1) years

APOE ϵ4 carrier: not reported

MMSE Mean (SD) ¹⁸F‐florbetaben positive: 26.7 (1.9), ¹⁸F‐florbetaben negative: 27.9 (1.4)

Years of education Mean (SD): ¹⁸F‐florbetaben positive: 13.8 (4.2), ¹⁸F‐florbetaben negative: 13.5 (3.0)

Sources of referral: not reported

Setting: secondary care (memory clinic)

Index tests

Administration instructions and tracer dosis

¹⁸F‐florbetaben was injected intravenously over 38 ± 17 s, with a mean specific activity at the time of injection of 60 ± 29 GBq/μmol. Each participant received on average 286 ± 19 MBq of ¹⁸F‐florbetaben.

PET imaging was conducted using a 3D GSO Phillips Allegro PET camera. A 2‐min transmission scan using a rotating 137Cs source was done for attenuation correction immediately prior to scanning. Images obtained between 90 to 110 min post injection were analysed. Images were reconstructed using a 3D RAMLA.

Image interpretation

PET images were processed with a semiautomatic volume of interest (VOI) method. This method used a preset template of narrow cortical VOI that was applied to either the spatially normalized ¹⁸F‐florbetaben scan or via placement on the subject’s spatially normalized coregistered MRI by a single operator who was blinded to the subject’s clinical status. Minor manual adjustments on the MRI were made to ensure that overlap with white matter and cerebrospinal fluid was minimized.

Spatial normalization and coregistration of the PET and MRI images was performed using SPM8.

Mean radioactivity values were obtained from VOI for cortical, subcortical, and cerebellar regions. The cerebellar cortical VOI was placed taking care to avoid cerebellar white matter. No correction for partial volume effect was applied to the PET data.

• Visual assessment

Baseline ¹⁸F‐florbetaben PET images underwent visual assessment by five independent nuclear medicine physicians blinded to clinical data. Readers had limited or no prior experience with amyloid PET visual interpretation and were trained on an electronic training tool.

The image assessment was performed on axial slices, in a grey scale. The regional cortical tracer uptake (RCTU) scoring system was used to assess the beta‐amyloid deposition in the following four regions: frontal cortex, posterior cingulate/precuneus, lateral temporal cortex, and parietal cortex. A RCTU value of 1, 2, or 3 was assigned if either no, moderate, or pronounced tracer uptake was observed respectively as described below:

1 No tracer uptake: Tracer uptake (i.e. signal intensity) in grey matter in the region is lower than in white matter

2 Moderate tracer uptake: Smaller area(s) of tracer uptake equal to or higher than that present in white matter;

‐ extending beyond the white matter rim to the outer cortical margin.

‐ involving the majority of the slices within the respective region.

3 Pronounced tracer uptake: A large confluent area of tracer uptake equal to or higher than that present in white matter;

‐ extending beyond the white matter rim to the outer cortical margin.

‐ and involving the entire region including the majority of the slices within the respective region.

The RCTU scores for the four brain regions were condensed into a binary brain amyloid plaque load (BAPL) score with positive or negative interpretation as described below:

¹⁸F‐florbetaben PET scan negative: scan without β‐amyloid deposition (RCTU score 1 in each of the 4 brain regions 1, 2, 3, and 4).

¹⁸F‐florbetaben PET scan positive: scan with moderate β‐amyloid deposition (RCTU score 2 in any or all of the 4 brain regions 1, 2, 3, and 4 and no score 3 in these 4 regions) or scan with pronounced β‐amyloid deposition (RCTU score 3 at least in one of the brain regions 1, 2, 3, and 4).

The majority read approach established the final result.

• Quantitative assessment (SUVR):

PET images were processed with a semiautomatic volume of interest (VOI) method. This method used a preset template of narrow cortical VOI that was applied to either the spatially normalized ¹⁸F‐florbetaben scan or via placement on the subject’s spatially normalised coregistered MRI by a single operator who was blinded to the subject’s clinical status. Minor manual adjustments on the MRI were made to ensure that overlap with white matter and cerebrospinal fluid was minimized. Spatial normalization and coregistration of the PET and MRI images was performed using SPM8. Mean radioactivity values were obtained from VOI for cortical, subcortical, and cerebellar regions. The cerebellar cortical VOI were placed taking care to avoid cerebellar white matter. No correction for partial volume effect was applied to the PET data.

The standardised uptake value (SUV), defined as the decay‐corrected brain radioactivity concentration normalised for injected dose and body weight, was calculated for all regions. These were then used to derive the SUV ratio (SUVR), which was referenced to cerebellar cortex. Neocortical Aβ burden was expressed as the average SUVR of the area‐weighted mean for the following cortical ROIs: frontal (consisting of dorsolateral prefrontal, ventrolateral prefrontal, and orbitofrontal regions), superior parietal, lateral temporal, lateral occipital, and anterior and posterior cingulate. The prespecified SUVR used was ≥ 1.45.

Target condition and reference standard(s)

Target condition as ADD with NINCDS‐ADRDA criteria were established with access to all study results and personal medical records

Flow and timing

Duration of follow‐up: 4 years

Number included in analysis: 45 participants

Visual Assessment: 25 ¹⁸F‐florbetaben (+) and 20 ¹⁸F‐florbetaben (‐)

Progression from MCI to ADD by visual assessment:

¹⁸F‐florbetaben (+): 21 MCI to ADD and 4 MCI‐MCI; ¹⁸F‐florbetaben (‐): 0 MCI to ADD and 20 MCI‐MCI

TP = 21; FP = 4; FN = 0; TN = 20

Progression from MCI to any other form of dementia (non‐ADD) by visual assessment:

¹⁸F‐florbetaben (+): 0 MCI to non‐ADD and 25 MCI‐MCI; ¹⁸F‐florbetaben (‐): 5 MCI to non‐ADD and 15 MCI‐MCI

TP = 0; FP = 25; FN = 5; TN = 15

Progression from MCI to any form of dementia by visual assessment:

¹⁸F‐florbetaben (+): 21 MCI to any dementia and 4 MCI‐MCI; ¹⁸F‐florbetaben (‐): 5 MCI to any dementia and 15 MCI‐MCI

TP = 21; FP = 4; FN = 5; TN = 15

SUVR: 24 ¹⁸F‐florbetaben (+) and 21 ¹⁸F‐florbetaben (‐)

Progression from MCI to ADD by SUVR:

¹⁸F‐florbetaben (+): 21 MCI to ADD and 3 MCI‐MCI; ¹⁸F‐florbetaben (‐): 0 MCI to ADD and 21 MCI‐MCI

TP = 21; FP = 3; FN= 0; TN = 21

Progression from MCI to any other form of dementia (non‐ADD) by SUVR:

¹⁸F‐florbetaben (+): 0 MCI to non‐ADD and 24 MCI‐MCI or ADD; ¹⁸F‐florbetaben (‐): 5 MCI to non‐ADD and 16 MCI‐MCI or ADD

TP = 0; FP = 24; FN = 5; TN = 16

Progression from MCI to any form of dementia by SUVR:

¹⁸F‐florbetaben (+): 21 MCI to any dementia and 3 MCI‐MCI; ¹⁸F‐florbetaben (‐): 5 MCI to any dementia and 16 MCI‐MCI

TP = 21; FP = 3; FN = 5; TN = 16

Comparative

Notes

Methodological quality

Item

Authors' judgement

Risk of bias

Applicability concerns

DOMAIN 1: Patient Selection

Was a consecutive or random sample of patients enrolled?

Unclear

Was a case‐control design avoided?

Yes

Did the study avoid inappropriate exclusions?

Unclear

Unclear

Low

DOMAIN 2: Index Test All tests

Were the index test results interpreted without knowledge of the results of the reference standard?

Yes

If a threshold was used, was it pre‐specified?

Yes

Was the PET scan interpretation done by a trained reader physician?

Yes

Was there a clear definition of a positive result?

Yes

Low

Low

DOMAIN 3: Reference Standard

Is the reference standards likely to correctly classify the target condition?

Yes

Were the reference standard results interpreted without knowledge of the results of the index tests?

No

High

High

DOMAIN 4: Flow and Timing

Was there an appropriate interval between index test and reference standard?

Yes

Did all patients receive the same reference standard?

Yes

Were all patients included in the analysis?

Yes

Was the study free of commercial funding?

No

High