Detection of amyloid abnormality higher in CSF vs. PET in Alzheimer’s

Evidence of amyloid abnormality across the Alzheimer’s disease spectrum is more prevalent with cerebrospinal fluid testing, according to a cross-sectional, individual-participant pooled study published in JAMA Neurology.

“With emerging disease-modifying anti-amyloid therapies, estimating the prevalence of amyloid abnormality in persons across the AD clinical spectrum is important to reduce screening failure rates and improve recruitment efficiency,” Willemijn J. Jansen, PhD, of the Alzheimer Center Limburg and the department of psychiatry and neuropsychology at Maastricht University in the Netherlands, and colleagues wrote.

Researchers sought to determine this in people presenting with various stages of cognitive function.

The study included 19,097 participants (mean age, 69.1 years; 53.1% women) culled from 85 Amyloid Biomarker Study cohorts, of which 10,139 had a PET scan and 8,958 had a CSF measurement.

Participants were determined to have either normal cognition, subjective cognitive decline, mild cognitive impairment or clinical AD dementia. Cognition level was defined by scoring on cognitive tests, with cognitive complaints providing a framework to judge decline. Mild cognitive impairment and clinical AD dementia were subsequently diagnosed according to published criteria.

Data collection was performed from Jan. 1, 2013, to Dec. 31, 2020, with main exposures being AD biomarkers detected through PET scans or in CSF.

Frequency of amyloid abnormality was estimated by age, sex, cognitive status, biomarker modality, apolipoprotein E (APOE) carrier status, educational level, geographical location and dementia severity.

Results showed that adjusted CSF cohort-provided cutoffs produced a 10% higher amyloid abnormality prevalence compared with PET-based estimates in participants with normal cognition (mean difference, 9%; 95% CI, 3-15), subjective cognitive decline (9%; 95% CI, 3-15) and mild cognitive impairment (10%; 95% CI, 3-17). However, the estimates were similar for participants with clinical AD dementia (mean difference, 4%; 95% CI, 2 to 9).

“These updated estimates suggest that preclinical and prodromal AD are more prevalent than previously estimated,” Jansen and colleagues wrote. “The findings may be useful in health care planning, providing potential eligible patient population sizes for anti-amyloid therapies and in recruitment strategies for clinical trials.”

In an editorial, Christina B. Young, PhD, and Elizabeth C. Mormino, PhD, of the department of neurology and neurological Sciences at Stanford University School of Medicine, praised the study authors for extending previous research with critical new findings. Noting that researchers were able to arrive at a statistically significant conclusion by including an increased cohort sample to properly track abnormal amyloid presence across a varied cross-section of cognitive function, Young and Mormino wrote: “These patterns imply that at least during the early stages of AD and before dementia onset, CSF may be a more sensitive marker of amyloid accumulation than PET.”