Diagnosing dementia post-mortem

[combatwombat]

---

Members do not see this ad.

I was hoping someone here could answer a burning question for me: must a pathologist ask the clinician whether a patient had dementia during life in order to diagnose a neurodegenerative disease post-mortem?

I think this is true, as I know healthy older folks can have Alzheimer's pathology, so this would seem to make sense. If someone with experience could weigh in it would be greatly appreciated.

 

[yaah]

The brain can degenerate, and whether the person has outward symptoms or not, the brain has still degenerated. Brains can show early stages of neurodegenerative diseases before there are clear clinical symptoms.

An accurate diagnosis of a specific neurodegenerative disorder likely depends on full integration of pathologic findings with clinical history, however. There is a lot of overlap among different diseases so that is why you need clinical info. Certain diseases have characteristic findings though.

 

[EvilTaz]

The clinical history is important, but at the end of the day, you are stuck with what's on glass to make some statement. And just because a diagnosis of dementia was not made clinically does not mean the patient does not have cognitive problems that haven't been tested.

The criteria for most of the adult neurodegenerative diseases, including Alzheimer's disease, are published.

The NIA-Reagan Working Group. Consensus recommendations for the postmortem diagnosis of Alzheimer's disease. Neurobiology of Aging. 1997; 18(suppl 4): s1-s2.

The most important change from previous criteria, I believe, is that the pathologist is to provide a probabilistic assessment that the pathology you see (severity of neuritic plaques or stage of neurofibrillary tangle pathology) accounts for dementia. So you say there is a high, intermediate or low likelihood.

I agree that in the absence of clinical dementia, it's strange to say there is a high likelihood that the pathology accounts for dementia in this patient without dementia. There have been validations of the criteria, but I can't remember how they do in cases without clinical dementia.

 

[alois]

This is an unfortunately muddy issue at the interface of neuropathology and neurology, partially caused by the sharing of diagnostic names. One way in which neuropathologists have tried to extract themselves from this bind is to shift to pathologically-based terminology. For example, while neurologist talk about patients having frontotemporal dementia, neuropathologists diagnose frontotemporal degeneration. Unfortunately, Alzheimer disease is used for both the clinical and neuropathological diagnoses. The NIA-Reagan criteria cited above are intended to determine the likelihood that a demented person is demented because of Alzheimer disease (as opposed to another cause), and were developed primarily for research purposes. Additionally, in the absence of formal mental status testing, it can be difficult to know about actual cognitive functioning (particularly with people who have developed adaptive systems to deal with deficits in memory, etc). It may also be necessary to have serial cognitive testing, as change from baseline rather than some arbitrary level of deficit is what is most likely to reflect the presence of pathologic changes in a given individual's brain.

Most neuropathologists diagnose what they see -- if there are plaques and tangles, they describe the burden and distribution. The adequacy and/or availability of clinical information doesn't change the physical reality of what is present on the slides. We are comfortable diagnosing coronary artery disease and not worrying about the presence or absence of angina, or diagnosing diverticulitis and not worrying about the presence or absence of abdominal pain.

This issue becomes even murkier when plaques and tangles are found in a surgical neuropathology specimen, such as cortex adjacent to a tumor resection or brain removed at decompression after large strokes. In this setting, when it is not possible to examine all of the anatomic regions required by the various diagnostic criteria, it is even harder to give a firm interpretation to the presence of the lesions. But if they are seen, then they should be included in the diagnostic report.