Monday April 16 2012
It's unclear yet whether the 'new Alzheimer's test' will be useful
The Daily Express says a “breakthrough” new test for Alzheimer’s could “pave the way to early diagnosis years before the devastating symptoms appear”. Its front-page news story says that scientists are hailing the new test as a potential way to identify people who are likely to go on to develop Alzheimer’s, so they could potentially be treated early.
Alzheimer’s disease is a devastating condition, and one that is set to become more common in our ageing population. Alzheimer’s is a particular form of dementia in which a protein called amyloid forms into abnormal deposits called plaques in the brain. These plaques and other protein “tangles” in the neurons are thought to contribute to the symptoms of the condition. However, at present the only way formally to confirm a diagnosis of Alzheimer’s is by identifying plaques in the brain during a post-mortem after the patient has died. This new research aimed to develop a technique for confirming the diagnosis of Alzheimer’s in living patients by injecting them with a special chemical that sticks to plaques and giving them brain scans to see if the chemical becomes deposited in the brain.
What it is important to note about this front-page news is that very few details of this research have been confirmed, as the research has not yet been fully published. This means it is hard to tell whether the technique will prove useful in a medical setting, particularly as there are still only limited options for slowing Alzheimer’s if it is detected early.
What is the basis for these current reports?
This story is based on research that is going to be presented at the American Academy of Neurology annual meeting at the end of April. The research was led by Dr Marwan Sabbagh, director of Banner Sun Health Research Institute in Sun City, Arizona. It was funded by the pharmaceutical manufacturer Bayer Healthcare, Berlin. The full abstract for this presentation is not yet available online, with only a press release published so far.
How can we currently diagnose Alzheimer's?
People with Alzheimer’s disease often experience progressive problems with memory; thinking and reasoning; language and understanding; and mood and behavioural changes. At the moment, a diagnosis of probable Alzheimer’s is made only after all other causes of dementia (for example, vascular dementia, or dementia with Parkinson’s disease) have been ruled out, based on a number of cognitive assessments and other tests, including brain imaging. A diagnosis of Alzheimer’s disease cannot be confirmed during life, as the only way to confirm the condition is by examining the brain after death to look for the characteristic amyloid protein plaques.
What have the researchers done?
There are only very limited details of the study’s methods and results available in the press release, but it would appear that the study looked at the performance of a potential technique for detecting the presence of amyloid plaques in the brains of living patients. These amyloid plaques are abnormal deposits of protein seen in the brains of people with Alzheimer’s disease during post-mortem. It seems that, in this particular study, researchers compared people’s test results during life with findings in their brains after death.
The researchers reportedly enrolled more than 200 volunteers who were nearing death and were willing to have their brains examined after death. This included people with suspected Alzheimer’s disease, and people without known dementia. The researchers carried out brain scans on the volunteers using magnetic resonance imaging (MRI) and also the new technique, which is called a florbetaben PET scan. This involved injecting the participants with a radioactive compound called florbetaben, which binds to amyloid plaques. The PET scan allows the researchers to detect whether florbetaben is concentrating in specific areas of the brain, which would suggest that amyloid plaques were present in those regions and, therefore, that the patient had Alzheimer’s.
The press release states that, at the point of writing, 31 volunteers were reported to have died and had their brains examined by post-mortem. These were compared to the brains of 60 volunteers who didn’t have Alzheimer’s symptoms.
What did the researchers find?
The press release reports two different sets of results. The first analysis looked at amyloid plaques found in the brain at autopsy. The researchers found that the florbetaben PET scan could detect amyloid plaques with a “sensitivity” of 77% and a “specificity” of 94%. This is not further explained, but is likely to mean that the technique picked up plaques in 77% of people in whom they were found after death, and found no plaques in 94% of people who were found to be free from plaques after death. Consequently, this means that the technique missed 23% of people who had plaques, and incorrectly identified 6% of people without plaques as having them.
The second analysis appears to examine the procedures for assessing the florbetaben PET scans that would be suggested for use during clinical practice. This analysis tested the use of florbetaben PET against diagnoses made after death. In this analysis the florbetaben PET scans were reported to have 100% sensitivity – meaning they picked up everybody who would go on to be diagnosed with Alzheimer’s after death. Under the proposed scan assessment technique, the florbetaben PET scans had 92% specificity, meaning that they correctly ruled out Alzheimer’s in 92% of people who were diagnosed as not having Alzheimer’s on autopsy.
What did the researchers conclude?
The lead study author, Marwan Sabbagh, concluded that this test provides an “easy, non-invasive way to assist an Alzheimer’s diagnosis at an early stage”. He said that it also offers exciting possibilities of using florbetaben as a tool in future clinical research studies looking at potential ways to reduce levels of amyloid in the brains of living patients.
Are there any limitations to this study?
It’s not possible to assess the quality of this study because of the limited information available from the press release. It is very early days for this technique, and we don’t yet know whether it will be helpful enough to be used in clinical practice.
Despite the Daily Express’ front-page suggestion that the research is aimed at developing screening techniques to check symptomless people for Alzheimer’s, it seems unlikely that this technique would be used in this way because carrying out brain scans in a large number of people is unlikely to be feasible. It seems the technique, at least from the brief details available, would be more likely to have potential for use as part of the assessment of a person with dementia symptoms in whom other possible causes have been ruled out.
If further research finds that this technology is reliable enough for further testing, studies will also need to be carried out to determine whether its use improves outcomes in people with dementia. Very early diagnosis is only likely to be truly clinically useful if the available interventions are effective at slowing Alzheimer’s disease at this very early stage.
Is conference research reliable?
Scientific research is often presented first at conferences. It gives researchers a chance to speak about their results and discuss them with their peers. However, the results they present are often preliminary, and generally haven’t gone through the same peer-review quality assurance process that is needed for publication in a journal. During these checks, which are applied by most journals during publication, experts in the field will assess the quality and validity of a study’s methods and results, and say whether they think research is good enough to be published. Also, as conference presentations are summarised in very brief “abstracts” for the public, very limited details are usually available on the study’s methods and results. This makes it difficult to judge the study’s strengths and limitations.
Some of the research presented at conferences never makes it to full publication. This could be for a number of reasons, for example, initially promising findings might not be confirmed in further tests or analysis, or the research may not be accepted by peer-reviewers or journal editors. A systematic review by the Cochrane collaboration found that nine years after the release of a conference abstract only just over half of the studies described (52.6%) had been fully published.
Health news stories are sometimes published based only on conference presentations, abstracts and press releases for forthcoming research. While some of these may be newsworthy and provide useful signposts to forthcoming research, they are not based on full reports of the research in question. This approach can be compared to a newspaper writing a film review based on watching the trailer for a film rather than the whole movie. This doesn’t mean that all research presented at conferences isn’t reliable, it just means that it’s best to reserve judgement until the research has been completed and published in a peer-reviewed journal.