Home >> ALL ISSUES >> 2021 Issues >> CSF biomarkers in ‘a new era’ for Alzheimer’s

CSF biomarkers in ‘a new era’ for Alzheimer’s

image_pdfCreate PDF

Karen Titus

August 2021—The FDA’s approval in June of a new drug to treat Alzheimer’s disease has incited motley responses, a sort of Grand Tour of emotions and reactions. As with any stormy action, the aftermath becomes clear only over time, during the hard work of cleaning up. What comes next?

The arrival of aducanumab (Adu­helm) will likely speed the development of other therapeutics, as well as accompanying biomarkers. “Many of us feel we’re entering a new era of treating Alzheimer’s disease,” says William Hu, MD, PhD, associate professor and chief of cognitive neurology, Rutgers-Robert Wood Johnson Medical School and Institute for Health, Health Care Policy, and Aging Research—both part of Rutgers Biomedical and Health Sciences.

It has also turned up the volume on discussions about inequities in health care. As a recent New England Journal of Medicine editorial notes (Levey AI. 2021;​384[18]:1762–1763), Black and Hispanic or Latino people are less likely than other groups to receive timely diagnoses and treatment despite the outsized impact of Alzheimer’s disease on these groups. Researchers are racing to recruit patients who are typically underrepresented in clinical trials, including Blacks and Asians. (And, as Abigail Adams once requested regarding another crucial matter, “Remember the Ladies.”)

Moreover, varying tau levels in different patient groups (as the editorial points out) raise worries of false-negative diagnoses and underestimated staging of the disease in Black patients. Then there’s the question of access to the basic diagnostic tools of CSF biomarkers and PET imaging.

All these bits are converging like the riot of cast members in the final pages of “Cymbeline.” But absent Shakespeare’s deft hand, no Jupiter will descend astride an eagle, helping to reunite the players and smooth over the plots. Instead, labs will have to make sense of all this themselves.

In ordinary times, more treatments and more patients would mean more testing.

Alicia Algeciras-Schimnich, PhD, and her colleagues at Mayo Clinical Laboratories are ready for both testing as well as questions about it. Her lab (she’s the director of the clinical immunoassay laboratory and medical director, BioPharma Diagnostics) has been offering Alzheimer’s disease biomarker testing for a little over a year and a half. Three CSF-based tests are packaged as an evaluation: Aβ42, T-tau (total tau), and P-tau (phosphorylated tau). They then use the P-tau/Aβ42 ratio to interpret whether changes are consistent with the presence of pathological changes associated with Alzheimer’s disease.

The neurology group at Mayo Clinic has been a regular user of the tests, says Dr. Algeciras-Schimnich, who is also a professor of laboratory medicine and pathology. While they’re early adopters, she expects aducanumab’s approval will spark more interest and inquiry from other groups of clinicians.

Will that translate into more CSF biomarker testing? “That is a core question,” says Leslie Shaw, PhD, director of the toxicology laboratory, director of the biomarker research laboratory, and co-director of the Alzheimer’s Disease Neuroimaging Initiative Biomarker Core laboratory, Department of Pathology and Laboratory Medicine, University of Pennsylvania Perelman School of Medicine.

“It’s a little too soon to tell,” says Dr. Algeciras-Schimnich. “We were expecting demand to go up for our tests, and it is going up some, but with the mixed reviews of the drug approval, perhaps some clinicians are a little more hesitant, or waiting a little longer, to start using the testing.”

The real demand, at least in the near term, may be for knowledge.

It’s worth stating, Dr. Shaw says, that Alzheimer’s is a complicated disease. Some patients have more neurovascular disease than others; others have Lewy body dementia. Lewy bodies are the hallmarks of Parkinsonian diseases, but 40 percent of patients with Alzheimer’s have some degree of Lewy body pathology. Perhaps 30 percent of the total sporadic cases of AD are “pure” amyloid plaque and tau tangle pathology. The bulk of cases, however, have varying degrees of neurovascular pathology, Lewy body pathology, TDP-43 pathology, and other less frequent findings, such as hippocampal sclerosis. For now, only a brain autopsy reveals the true picture of the mixed pathology, Dr. Shaw says.

The development of anti-amyloid agents will help clinicians treat Alz­heimer’s, Dr. Hu says, “but that piece is just one of many” and shouldn’t be used in patients who have underlying frontotemporal lobar degeneration or Lewy body disease.

New drugs will launch more questions about testing context, Dr. Hu predicts. “When we’re looking at spinal fluid changes, does the biological profile fit the clinical picture? And even if someone does have pathologic Alzheimer’s disease, is it appropriate to treat the person with a drug, given what else might be going on?

“For those of us who’ve been dealing with these scenarios, we are seeing very exciting developments,” says Dr. Hu, who, in addition to being a practicing cognitive neurologist (focused on mild and more severe dementing illnesses), leads research to establish novel, validated fluid biomarkers to reflect the underlying pathology. “And as we do more diagnostic work, we will probably also find more caveats, in terms of what medical and personal factors may influence measurement. That will really, hopefully, advance the understanding of CSF biomarkers.”

Here, too, there is much to learn, though testing methods are, compared with other aspects of Alz­heimer’s disease, more straightforward (more “Macbeth,” less ridiculous romance), thanks to the experience of European institutions and the network of National Institute on Aging-sponsored Alzheimer’s research centers in the United States.

Aducanumab’s label doesn’t address methods for detecting plaque burden, Dr. Shaw says. Currently there are two highly automated methods for reliably testing CSF Aβ-tau and P-tau, he says, adding, “They work very, very well.” The companies have been engaged with the FDA regarding approval—Dr. Shaw says it’s “always an ongoing negotiation”—while the European Medicines Agency has already approved them.

Meanwhile, a tremendous amount of work is being done at the dozens of NIA research centers. “That’s where a lot of the focus tends to be in terms of biomarker studies,” Dr. Shaw says.

Put another way, methodology and automation are the lowest speed bumps when it comes to using CSF biomarkers. “That’s solved for CSF AD biomarkers,” Dr. Shaw says.

Dr. Algeciras-Schimnich’s experience with CSF biomarkers offers a peek at how they might be used should demand for testing rise. First, she says, “It’s important to know these are not diagnostic tools.” Rather, they’re used in combination with other parameters to determine whether a patient’s mild cognitive impairment is due to Alzheimer’s disease.

In terms of test interpretation, she says, “We are looking for a decrease in the Aβ42 levels and an increase in phosphorylated tau.”

Before that, however, they’re looking for the right collection tube. Dr. Algeciras-Schimnich calls it the “critical preanalytical component of this testing,” especially related to Aβ42. The sample needs to be collected in so-called low-bind tubes, made of polypropylene, which prevents further absorption of Aβ42. Low Aβ42 is consistent with Alzheimer’s disease pathological changes, indicating presence of Aβ plaque in the brain. “We want to avoid a false-positive result due to sample handling.

Dr. Algeciras-Schimnich

“When we first implemented it,” she continues, “our compliance for getting the right tube to us was quite low—around 60 percent. We’re now about 95 percent. But we had to do a lot of education” with clinical colleagues and within the lab, to make sure everyone “understood why it was so important, why we were being so restrictive on the type of tube.”

Dr. Hu says solving the problem of “sticky proteins” took effort in his lab as well. The standard LP tray contains collection tubes made of polystyrene, itself a sticky substance (hence its use in ELISA plates). “If you put a sticky protein in a sticky tube,” he says, “you start to have a lot of issues in terms of the eventual analytic results.”

“We’ve known for some time that polypropylene tubes have to be used when collecting spinal fluid,” Dr. Hu adds, though here difficulties arise—most commercially available trays contain polystyrene tubes. “It’s a logistics issue. We actually engineer and customize our own trays to bypass that issue, but that has not been pushed at a national level.”

The engineering solutions, coupled with current automated analyzers, have meant superior precision, says Dr. Hu, who has published on the topic. “I was pleasantly floored.” Though he says he worries about the impact of automation on trainees’ jobs, “from a clinical proficiency perspective, it is a dramatic step forward” from manual methods. “One of the challenges of this testing has been the reproducibility across lots and assays.”

Mayo relies heavily on the P-tau/Aβ42 ratio, says Dr. Algeciras-Schimnich, since it’s the more reliable of the tests the lab offers. “In some instances there might be discrepancies in terms of the individual analytes.” When clinical colleagues call her to ask questions about results, that’s usually the reason why. Differences can occur because of changes in CSF dynamic that dilute all the analytes. “So you might see a low Aβ42, but at the same time the T-tau and P-tau will be decreased, indicating some other process might be happening in the patient.” Normal pressure hydrocephalus may be a cause of a decreased Aβ42 but a normal P-tau/Aβ42 ratio.

Dr. Shaw also is a fan of the P-tau/Aβ42 ratio. Differences in opinion are to be expected, even among experts, he says. “But I don’t think anybody would disagree with the findings that are reported in many studies: that the P-tau/Aβ42 ratio is probably the best performer,” compared with Aβ42 alone, and is slightly better than the Aβ42/Aβ40 ratio.

As CSF biomarker use expands, labs will also need to figure out how to report results, something in which European labs have already been immersed. Should it be a black-and-white number? “Some do that,” says Dr. Shaw, “with a stated target range. But others are more detailed about it and will report not only the result but some degree of interpretation,” such as whether there is a gray zone around cut-point values.

His lab doesn’t routinely report results, of course, but “We are doing a lot of studies with our clinical colleagues and are always engaged in discussions around interpretation.” The lab and dozens of other centers were part of a recent study to look at their experiences “and lay it all out,” he says, noting the paper has been submitted for publication and provides an overview of how different centers report and interpret results.

There’s plenty of interpretation over what results mean—not the numerical value, but what biological processes they reflect, says Dr. Hu.

T-tau and P-tau reflect the neurofibrillary tangle component of Alz­heimer’s disease, he says, whereas Aβ42 primarily reflects the neuritic plaque component. The role of Aβ40 is, in theory, to normalize Aβ42, Dr. Hu continues, but at the same time it is highly correlated with tau and phosphorylated tau. So regardless of which ratio is used, “with Aβ42 we are very often using the same disease indicator—not just cerebral amyloidosis, but Alzheimer’s disease, consisting of plaques and tangles.”

“There are a number of different personal interpretations between investigators, in terms of whether you use the ratio or you use an arbitrary cutoff of tau, as well as an arbitrary cutoff for Aβ,” Dr. Hu says. The impact, on the whole, is focused on reclassifying a minority of patients. “But overall they are very consistent in the positive and negative cases. I understand that people have different preferences,” he says, “but these measures are giving very much of the same message.”

The basic clinical question is simple: Is it Alzheimer’s?

“My answer will be, ‘It depends,’” Dr. Hu says. Tests need to be ordered in the right context.

Alzheimer’s disease pathology can occur many years before symptoms appear. “If we start performing lumbar punctures on everybody at age 50, we’re going to find some people with Alzheimer’s biomarker changes,” Dr. Hu says. “The meaning of that right now is a little less clear, which is why we often say we want to reserve this test at least for people who have noticed a subjective memory complaint or a change from baseline.”

In addition, he says, “We need to be cautious when we are looking at other diseases, and then we find an incidental alteration in Alzheimer’s biomarkers.” A prime example: Patients with HIV, with or without dementia, may have decreases in amyloid similar to that seen in Alzheimer’s disease. “But when we look at the amyloid PET imaging, they don’t have plaque deposition in the brain.” That’s one area where there’s a dissociation between amyloid PET and spinal fluid Aβ42 levels, but there may be others, which is why Dr. Hu urges caution in using biomarkers. “We can’t just blindly apply these tools to other conditions.”

CAP TODAY
X