Cerebrospinal Fluid Testing in Dementia Diagnosis

As the population ages, an increasing number of individuals will present for evaluation of cognitive concerns.¹ Individuals presenting with memory impairment or other cognitive changes require a comprehensive evaluation including a detailed clinical history, physical and neurologic examinations, cognitive testing, routine laboratory tests, and structural brain imaging. Amyloid positron emission tomography (PET), blood tests, and cerebrospinal fluid (CSF) tests are available to help determine the etiologic cause of cognitive impairment.² Biomarker testing is indicated if a more certain diagnosis is desired and expected to affect the patient’s diagnosis, management, or decision-making; testing is not recommended if the patient does not want to know this information or if it is not expected to benefit the patient. For patients in whom biomarker testing is indicated, if Alzheimer disease (AD) is the leading diagnostic consideration and non-AD neurologic conditions are less likely or the individual is a potential candidate for amyloid-lowering treatments, obtaining an amyloid PET scan may be the best option. Blood tests that detect evidence of AD pathology are useful in some individuals, especially those who are not candidates for amyloid-lowering treatments or who have concerns about amyloid PET or CSF tests.³ As has been described in expert recommendations, CSF testing may be the optimal modality for individuals presenting with atypical dementia syndromes, such as rapidly progressive dementia.⁴

There is a broad differential diagnosis for atypical dementia syndromes that includes etiologies for which CSF tests are informative and in some cases diagnostic (see Box).^5,6 Available CSF tests include routine laboratory studies (eg, cell counts, cell differential, protein, glucose); tests for autoimmune conditions (eg, immunoglobulin profile, oligoclonal bands, autoantibody panels); tests for AD pathology (eg, AΒ42, AΒ40, p-tau181, total tau); tests for Creutzfeldt-Jacob disease (eg, prion protein, real-time quaking-induced conversion, total tau, 14-3-3 protein); and tests for α-synucleinopathies (eg, seed amplification assays).^6,7 Some of these tests (eg, immunoglobulin profile, oligoclonal bands, autoantibody panels) can be run on blood samples and provide complementary data to CSF tests.⁶ Other tests (eg, tests for Creutzfeldt-Jacob disease, α-synucleinopathies) can only be performed with CSF.

Because AD is the most common cause of dementia in adults ≥ 65 years, performing CSF testing for AD can be informative in many older individuals with cognitive impairment.^8,9 If the CSF test is positive, it indicates that some degree of AD brain pathology (ie, amyloid plaques, tau neurofibrillary tangles) is present and could potentially be a cause or contributor to cognitive impairment. Many individuals have copathologies and comorbidities (eg, obstructive sleep apnea and depression) or take medications that can cause or contribute to cognitive impairment, so a positive CSF test for AD by itself does not establish that cognitive impairment is caused by AD.⁴ If the CSF test for AD is negative, in most cases this rules out AD as a cause or contributor to cognitive impairment, and clinicians can focus on the evaluation of alternative etiologies. CSF tests should be interpreted within the context of the entire evaluation, and the presence and effects of other conditions should be carefully assessed regardless of the result of the CSF test for AD.

Most CSF tests for AD include a panel of analytes, and interpretation of these results may be confusing. AD pathology is associated with lower levels of CSF AΒ42, because soluble AΒ42 is sequestered into amyloid plaques. AD pathology also is associated with higher levels of CSF p-tau181 and total tau, which likely reflect a pathologic response to amyloid plaques. For the 3 Food and Drug Administration–approved CSF tests for AD, a ratio of analytes (AΒ42/AΒ40, p-tau181/AΒ42, or total tau/AΒ42) most accurately determines amyloid status.^10,11 The provider must assess amyloid status based on the ratio of analytes and not on AΒ42 by itself, which is substantially less accurate than the ratios.^4,12,13 In addition, the clinician should consider whether the result is a clear positive, clear negative, or an intermediate result.¹⁴ For individuals with intermediate results, it may be reasonable to follow the CSF test with an amyloid PET scan, especially if the individual is a potential candidate for amyloid-lowering treatments. If CSF biomarker results are clearly positive or negative, there is currently no evidence that repeating CSF testing within 1-3 years has utility, especially because CSF biomarkers change very slowly. If the patient has received amyloid-lowering treatments, repeat CSF biomarker testing could potentially be helpful in deciding whether to continue or stop treatment, although currently there are not recommendations about when to perform this testing or how to interpret the results.

When communicating the results of CSF tests, the practitioner must emphasize that there are many causes of dementia and that biomarker tests help determine the specific cause of symptoms. Because many individuals and caregivers do not understand that AD is one of many causes of dementia, they may be confused if AD test results are negative. It is important to explain that the cognitive impairment is real but is likely caused by a condition other than AD. In addition, individuals with positive AD test results may have other conditions causing or contributing to cognitive impairment, and these other conditions require evaluation and management.

Limitations of CSF testing include that most individuals have never undergone a lumbar puncture (LP) and may perceive it to be associated with high risk.¹⁵ Risks of LP are typically mild, of short duration, and manageable but increase the burden of the procedure for individuals, caregivers, and providers. Mild back soreness is relatively common, and there is a risk of post-LP headache that typically self-resolves but may require a blood patch in some patients.¹⁶ Individuals on anticoagulation or who have structural abnormalities in the lumbar spine region may not be able to have an LP. Furthermore, many clinicians are not trained to perform LPs, limiting access. Reimbursement for performing LPs may not cover provider costs, decreasing the incentive to train more LP providers.¹⁷ Despite these drawbacks, for some individuals with cognitive impairment, CSF testing is the optimal modality to provide essential information leading to an accurate diagnosis. The following box contains key points associated with CSF testing for potential use in diagnosis of dementia.