Plasma levels of phosphorylated tau at threonine 181 (p-tau181) may provide a means of monitoring disease progression for patients with Alzheimer’s disease (AD), new research suggests.
In a study of more than 1000 participants, changes over time in levels of p-tau181 were associated with prospective neurodegeneration and cognitive decline characteristic of AD. These results have implications for investigative trials as well as clinical practice, the investigators note.
Like p-tau181, neurofilament light (NfL) is associated with imaging markers of degeneration and cognitive decline; but in contrast to the findings related to p-tau181, the associations between NfL and these outcomes are not specific to AD. Using both biomarkers could improve prediction of outcomes and patient monitoring, according to the researchers.
“These findings demonstrate that p-tau181 and NfL in blood have individual and complementary potential roles in the diagnosis and the monitoring of neurodegenerative disease,” coinvestigator Michael Schöll, PhD, senior lecturer in psychiatry and neurochemistry at the University of Gothenburg, Gothenburg, Sweden, told Medscape Medical News.
“With the reservation that we did not assess domain-specific cognitive impairment, p-tau181 was also more strongly associated with cognitive decline than was NfL,” Schöll added.
The findings were published online January 11 in JAMA Neurology.
Monitoring a patient’s neurodegenerative changes is important for tracking AD progression. Although clinicians can detect amyloid beta and tau pathology using PET and CSF biomarkers, the widespread use of the latter has been hampered by cost and limited availability of necessary equipment. The use of blood-based biomarkers is not limited in these ways, and so they could aid in diagnosis and patient monitoring.
Previous studies have suggested that p-tau181 is a marker of AD status.
In the current study, investigators examined whether baseline and longitudinal levels of p-tau181 in plasma were associated with progressive neurodegeneration related to the disease. They analyzed data from the Alzheimer’s Disease Neuroimaging Initiative, a multicenter study designed to identify biomarkers for the detection and tracking of AD.
The researchers selected data for cognitively unimpaired and cognitively impaired participants who participated in the initiative between February 1, 2007, and June 6, 2016. Participants were eligible for inclusion if plasma p-tau181 and NfL data were available for them and if they had undergone at least one [18F]fluorodeoxyglucose (FDG) PET scan or structural T1 MRI at the same study visit. Most had also undergone imaging with [18F]florbetapir, which detects amyloid beta.
A single-molecule array was used to analyze concentrations of p-tau181 and NfL in participants’ blood samples. Outliers for p-tau181 and NfL concentrations were excluded from further analysis. Using participants’ FDG PET scans, the investigators measured glucose hypometabolism characteristic of AD. They used T1-weighted MRI scans to measure gray matter volume.
Cognitively unimpaired participants responded to the Preclinical Alzheimer Cognitive Composite, a measure designed to detect early cognitive changes in cognitively normal patients with Alzheimer’s disease pathology. Cognitively impaired participants underwent the Alzheimer Disease Assessment Scale–Cognitive Subscale with 13 tasks to assess the severity of cognitive impairment.
The researchers included 1113 participants (54% men; 89% non-Hispanic Whites; mean age, 74 years) in their analysis. In all, 378 participants were cognitively unimpaired, and 735 were cognitively impaired. Of the latter group, 73% had mild cognitive impairment, and 27% had AD dementia.
Results showed that higher plasma p-tau181 levels at baseline were associated with more rapid progression of hypometabolism and atrophy in areas vulnerable to AD among cognitively impaired participants (FDG PET standardized uptake value ratio change, r = –0.28; P < .001; gray matter volume change, r = –0.28; P < .001).
The association with atrophy progression in cognitively impaired participants was stronger for p-tau181 than for NfL.
Plasma p-tau181 levels at baseline also predicted atrophy in temporoparietal regions vulnerable to AD among cognitively unimpaired participants (r = –0.11; P = .03). NfL, however, was associated with progressive atrophy in frontal regions among cognitively unimpaired participants.
At baseline, plasma p-tau181 levels were associated with prospective cognitive decline in both the cognitively unimpaired group (r = −0.12; P = .04) and the cognitively impaired group (r = 0.35; P < .001). However, plasma NfL was linked to cognitive decline only among those who were cognitively impaired (r = 0.26; P < .001).
Additional analyses showed that p-tau181, unlike NfL, was associated with hypometabolism and atrophy only in participants with amyloid beta, regardless of cognitive status.
Between 25% and 45% of the association between baseline p-tau181 level and cognitive decline was mediated by baseline imaging markers of neurodegeneration. This finding suggests that another factor, such as regional tau pathology, might have an independent and direct effect on cognition, Schöll noted.
Furthermore, changes over time in p-tau181 levels were associated with cognitive decline in the cognitively unimpaired (r = –0.24; P < .001) and cognitively impaired (r = 0.34; P < .001) participants. Longitudinal changes in this biomarker also were associated with a prospective decrease in glucose metabolism in cognitively unimpaired (r = –0.05; P = .48) and cognitively impaired (r = –0.27; P < .001) participants, but the association was only significant in the latter group.
Changes over time in p-tau181 levels were linked to prospective decreases in gray matter volume in brain regions highly characteristic of AD in those who were cognitively unimpaired (r = –0.19; P < .001) and those who were cognitively impaired (r = –0.31, p < .001). However, these associations were obtained only in patients with amyloid beta.
Schöll noted that blood-based biomarkers that are sensitive to AD could greatly expand patients’ access to a diagnostic workup and could improve screening for clinical trials.
“While the final validation of the existence and the monitoring of potential changes of neuropathology in vivo is likely to be conducted using neuroimaging modalities such as PET, our results suggest that at least a part of these examinations could be replaced by regular blood tests,” Schöll said.
Lead author Alexis Moscoso, PhD, postdoctoral researcher in psychiatry and neurochemistry at the University of Gothenburg, reported that the researchers will continue validating blood-based biomarkers, especially against established and well-validated neuroimaging methods.
“We are also hoping to be able to compare existing and novel blood-based Alzheimer’s disease biomarkers head to head to establish the individual roles each of these play in the research and diagnosis of Alzheimer’s disease,” Moscoso told Medscape Medical News.
Commenting on the findings for Medscape Medical News, David S. Knopman, MD, professor of neurology at Mayo Clinic, Rochester, Minnesota, said that this is “an outstanding study” because of its large number of participants and because the investigators are “world leaders in the technology of measuring plasma p-tau and NfL.”
Knopman, who was not involved with the research, noted that the study had no substantive weaknesses.
“The biggest advantages of a blood-based biomarker over CSF- and PET-based biomarkers of Alzheimer disease are the obvious ones of accessibility, cost, portability, and ease of repeatability,” he said.
“As CSF and PET exams are largely limited to major medical centers, valid blood-based biomarkers of Alzheimer disease that are reasonably specific make large-scale epidemiological studies that investigate dementia etiologies in rural or urban and diverse communities feasible,” he added.
Whereas p-tau181 appears to be specific for plaque and tangle disease, NfL is a nonspecific marker of neurodegeneration.
“Each has a role that could be valuable, depending on the circumstance,” said Knopman. “Plasma NfL has already proved itself useful in frontotemporal degeneration and chronic traumatic encephalopathy, for example.”
He noted that future studies should examine how closely p-tau181 and NfL align with more granular and direct measures of AD-related brain pathologies.
“There has got to be some loss of fidelity in detecting abnormality in going from brain tissue to blood, which might siphon off some time-related and severity-related information,” said Knopman.
“The exact role that plasma p-tau and NfL will play remains to be seen, because the diagnostic information that these biomarkers provide is contingent on the existence of interventions that require specific or nonspecific information about progressive neurodegeneration due to Alzheimer disease,” he added.
The study was funded by grants from the Spanish Instituto de Salud Carlos III, the Brightfocus Foundation, the Swedish Alzheimer Foundation, and the Swedish Brain Foundation. Schöll reported serving on a scientific advisory board for Servier on matters unrelated to this study. Moscoso and Knopman have reported no relevant financial relationships.
JAMA Neurology. Published online January 11, 2021. Full text