Log in with your email address username.

×

Do we need a new definition for Alzheimer’s disease?

 

A high-profile US research panel has proposed a new way of defining Alzheimer’s disease that depends on biological markers, rather than a symptom-based clinical diagnosis.

The researchers from the National Institute of Aging and the Alzheimer’s Association say their proposed new definition would for the moment be used only for research purposes, to ensure that the right people are tested in Alzheimer’s trials. But some specialists fear that once the biological marker genie is out of the bottle, it will be very hard to put it back in.

A biological definition of Alzheimer’s has only recently become possible thanks to new technologies that allow measurement and imaging of what is considered the key pathology of the disease – the abnormal buildup in the brain of two proteins. Clumps of amyloid-beta are thought to disrupt communications between neurons and lead to their death, while tau protein causes neurofibrillary tangles in the brain which also disrupt communications.

In the past few years, PET scans and cerebrospinal fluid taps have been developed to detect these protein buildups, allowing for a clearer picture of what is happening inside the brain of someone sliding into dementia. The new paper, published in the journal Alzheimer’s and Dementia, bases its definition on what the researchers call an amyloid-tau-neurodegeneration (ATN) classification system, whereby if someone has biomarker evidence of both amyloid and tau, they will be diagnosed as having Alzheimer’s disease. A person with amyloid deposition but no tau pathology would be considered to have “Alzheimer’s pathologic change”. These two classifications would be situated on an Alzheimer’s continuum of varying severity, independent of outside symptoms. Clinical symptoms and evidence of neurodegeneration – brain shrinkage on imaging for example – could contribute to determining disease severity, but not to the presence of the disease itself.

The researchers say the need for a biological definition of Alzheimer’s is partly to do with the way dementia and Alzheimer’s have become conflated. Around 30% of people currently diagnosed with Alzheimer’s don’t seem to have amyloid plaques or tau tangles, which means they would not be suitable for the extremely expensive trials targeting those pathologies.

But some experts point to problems with this biological definition of Alzheimer’s. One is that just as not all people with dementia have amyloid plaques and tau tangles, not everyone with amyloid and tau buildup has clinical symptoms of dementia. Over recent years, there has in fact been some pushback against what is known as the “amyloid hypothesis” of Alzheimer’s, which posits that these amyloid clumps are the drivers of the disease. A key reason for the pushback is that in numerous trials over the past two decades, drugs targeting amyloid in the brain have comprehensively failed at the phase 3 level. Amyloid naysayers suggest that amyloid buildup may be more of a consequence than a cause of Alzheimer’s, and that the key problem may be further upstream.  One proposal is that Alzheimer’s may be linked to inflammation in the brain, and that the amyloid buildup may in fact be a response to this inflammation.

But there are other issues as well. If Alzheimer’s disease gets redefined for research purposes, a redefinition for clinical purposes can’t be far behind. Once people know that there are tests for Alzheimer’s biomarkers, they are likely to ask for them. At the moment, the costs for these tests are prohibitive and are not reimbursable. But if costs were to come down, as they likely will eventually, the upshot may be a dramatic increase in the diagnosis of Alzheimer’s disease, since Alzheimer’s pathology can appear years and perhaps decades before clinical symptoms. With no disease-modifying treatment for Alzheimer’s disease currently available and no evidence that early intervention is beneficial, it may seem futile to expand diagnosis beyond those for whom there are clear clinical symptoms that need to be dealt with.

And there’s the fundamental problem that we simply don’t know enough about what these biomarkers signify, in terms of clinical development of the disease.

“You have a 72-year-old woman who is cognitively normal but has a positive amyloid PET scan. What does that mean? What do we tell her?” asks Dr Ron Peterson, a Mayo Clinic Alzheimer’s expert, in an interview with the publication MedPage Today.

“Right now, according to this new definition, you’re going to say, ‘Well, you’re on the Alzheimer’s continuum. You have Alzheimer’s pathology changes. What does that mean? The bottom line is, we don’t know.”

You can read the full paper here.

How common is misdiagnosis?

 

In the news this week has been the horrific case of Magdalena Malec, 31, who, despite showing all the classic signs of sepsis, went undiagnosed as she lay in a UK hospital recovering from an operation. Admitted following an ectopic pregnancy, Ms Malec had a raised temperature and a number of other red flags for sepsis, following her surgery. She later developed extensive limb ischaemia and eventually had to have both of her legs, her right arm and the fingers of her left hand amputated, and she also needed a kidney transplant.

“The catastrophic chain of events which led to Magdalena’s near death and horrendous injuries were completely avoidable if the hospital had followed its own sepsis protocol,” her lawyer said. The hospital has since apologised.

Cases like that of Magdalena Malec put a human face on the very real problem of diagnostic error. More publicity tends to surround cases of wrong-site surgery or drug errors, but the research shows that misdiagnosis is a far more common – and more intractable – issue.

One recent attempt at putting a figure to the problem found that patients were misdiagnosed around 10-15% of the time. And an Australian systematic review of diagnostic error in older patients found rates of over 10% for COPD, dementia, Parkinson’s, heart failure, stroke and MI.

But it’s not easy to nail down figures, as many hospitals and practices don’t keep count and there can be strong motivations for doctors not to report a wrong diagnosis. On top of that, misdiagnosed patients may also follow up with another doctor who makes the correct diagnosis, with the first doctor never realising their mistake.

There are various methods researchers use to try and quantify diagnostic error. The gold standard is said to be autopsy studies, which consistently identify diagnostic discrepancies in 10-20% of cases. Of course, not all deaths are subject to autopsy, and those that are may already be cases where the diagnosis is unclear.

Another, if subjective, method is patient surveys, which show that around a third of patients have been the subject of a misdiagnosis or have had a family member of close friend who has been misdiagnosed. ‘Secret shopper’ studies have also been carried out, where real or simulated patients with classic symptoms of a condition present to a doctor or hospital. In these studies, doctors misdiagnose the patient in around 13% of cases.

Second reviews offer yet another opportunity to quantify diagnostic error. These have shown that 10-30% of breast cancers are missed on mammography, and 1-2% of cancers overall are misread on biopsy samples.

Misdiagnosis and delayed diagnosis are also far and away the greatest cause of malpractice suits, which point to how common they are. Of course, not every misdiagnosis is actionable: under Australian law, diagnostic error is only considered negligent if it falls short of Australian medical standards and is the result of the doctor failing to take “reasonable care” in diagnosing the condition.

Misdiagnosis is “the hidden part of the iceberg of medical errors that dwarfs other kinds of mistakes,” says Dr David Newman-Toker of the Johns Hopkins School of Medicine, who has extensively studied the problem.

In a survey, the top reasons doctors gave for cancer misdiagnosis were “fragmented or missing information across medical information systems”, along with “inadequate diagnostic resources”. But Dr Newman-Toker says they more typically result from flawed ways of thinking, sometimes coupled with negligence.

He says drug errors and wrong-site surgery are not only less common but more amenable to solutions such as color-coded labelling or preoperative checklists. But there is no such easy or obvious fix for diagnostic error, he says.