18F-florbetapir PET scan for the early diagnosis of Alzheimer’s disease dementia and other dementias in people with mild cognitive impairment
Review question: In people with mild cognitive impairment (MCI), does using a 18F PET scan with florbetapir predict the progression to Alzheimer’s disease dementia (ADD) and other dementias?
Due to global ageing, the number of people with dementia is expected to increase dramatically in the next few decades. Diagnosing dementia at an early stage is desirable, but there is no widespread agreement on the best approach. A range of simple pen and paper tests used by healthcare professionals can assess people with poor memory or cognitive impairment. Whether or not using special PET scans that detect amyloid —one of the hallmarks of Alzheimer’s disease— improves our ability to predict the progression from MCI to ADD or other forms of dementia remains unclear. Since these tests are expensive, it is important that they provide additional benefits.
We aimed to evaluate the accuracy of the 18F-florbetapir PET scan in identifying those people with MCI who clinically progress to ADD, other types of dementia, or any form of dementia over a period of time.
The evidence is current to May 2017. We found three studies including 453 participants with MCI. Two studies evaluated the progression from MCI to ADD and one study evaluated the progression from MCI to any form of dementia.
Regarding the two studies that evaluated the progression from MCI to ADD, one study had 401 participants with a follow-up of 1.6 years and the mean age was 72 years. The other study had 47 participants with a follow-up of three years, and the mean age was 72 years.
The other study that looked at any form of dementia included 5 participants over 90 years old.
Two of the studies were funded by the test manufacturer.
Quality of the evidence
The main limitation of this review was that our findings were based on only three studies, with insufficient detail on how the people were selected, whether the information from the scan was assessed separately from the final diagnosis. The studies were considered to be at high risk of bias due to potential conflicts of interest detected.
In this review, we found the following results based on the three studies.
At a follow-up of 1.6 years, using visual assessment, the scan correctly classified 89% of the participants who progressed to ADD but only 58% of the participants who did not progress to ADD. This means that in a group of 100 people with MCI, 15% of whom will develop ADD, we would expect 13 of 15 people to have a positive result and the other 2 participants to be falsely negative. Also 49 people who will not develop ADD would have a negative result, but 36 people who will not develop ADD would have a positive result (false positives).
In the study that followed up people for three years and used visual assessment, the scan correctly classified 67% of people who progressed to ADD and 71% who did not progress to ADD. This means that in a group of 100 people with MCI, 19 of whom will develop ADD, we would expect 13 people to have a positive result of the scan and 6 people to have a falsely negative result. In addition, 58 of 81 participants who will not progress to ADD would have a negative result, but 23 people who will not develop ADD would have a positive result (false positives). The small number of participants evaluated at three years lowered our confidence on these estimates of accuracy.
Regarding progression to any form of dementia, the extremely small number of participants meant that we were unable to provide meaningful estimates of accuracy.
We conclude that 18F-florbetapir PET scans cannot be recommended for routine use in clinical practice to predict the progression from MCI to ADD or any form of dementia based on the currently available data. More studies are needed to demonstrate its usefulness.