Triple play: PET, MRI + CSF improve Alzheimer’s prediction

Twitter icon
Facebook icon
LinkedIn icon
e-mail icon
Google icon
 - Alzheimer's Disease

Combined analysis of FDG-PET, MRI and cerebrospinal fluid (CSF) markers increased the accuracy of prediction of conversion from mild cognitive impairment (MCI) to Alzheimer’s disease, according to a study published Dec. 11 in Radiology. FDG-PET delivered the greatest prognostic value of the biomarkers evaluated in the study.

Approximately 10-15 percent of patients with MCI convert to Alzheimer’s disease annually. “Although there is no cure for Alzheimer’s disease, there are four symptomatic treatments that might provide some benefits,” P. Murali Doraiswamy, MD, professor of psychiatry at Duke University Medical Center in Durham, N.C., said in a press release. “So developing the right combination of diagnostic tests is critical to make sure we enable an accurate and early diagnosis in patients, so they can evaluate their care options.”

Previous research had demonstrated that structural MRI reveals medial temporal lobe atrophy in patients with MCI and Alzheimer’s disease, and FDG-PET shows hypometabolism in the temporoparietal regions. However, potential markers have not been studied in combination. Jennifer L. Shaffer, MD, from the department of radiology at Duke, and colleagues devised a retrospective study to determine the extent to which combined biomarkers could improve the ability to predict future decline in patients with MCI compared with prediction based solely on clinical parameters.

The study population was comprised of 97 patients with MCI enrolled in the Alzheimer’s Disease Neuroimaging Initiative. Researchers created eight models that analyzed the prognostic value of combinations of MRI, PET and CSF markers with clinical and covariates (age, education apolipoprotein E genotype [ApoE4] and Alzheimer’s Diseases Assessment Scale-Cognitive subscale score).

A total of 43 patients progressed to Alzheimer’s disease within four years.

The misclassification rate for conversion to Alzheimer’s disease based on neuropsychological testing and other clinical data was 41.3 percent, according to Shaffer et al. This rate was reduced to 28.4 percent with the addition of MR, PET and CSF biomarkers.

On an individual basis, FDG-PET was the only biomarker that significantly improved the predictive value of the covariates. This may be because metabolic deficits are greater in magnitude than volumetric changes earlier in the disease, according to the researchers.

“In an ideal world, you’d obtain all information available—regardless of cost or number of tests—for the best prediction of cognitive decline,” Jeffrey R. Petrella, MD, Alzheimer’s Disease Research Lab at Duke, said in the release. “However, there’s a trade-off between adding testing—some of which may add little new information—with the inconvenience, cost and risk to the patient.”

“Though all the tests added some unique information FDG-PET appeared to strike the best balance, adding the most prognostic information for patients with mild cognitive impairment,” Petrella said. However, the researchers noted MRI and CSF markers could add critical information in patients lacking ApoE4 geneotyping.

“Our findings should not be interpreted as confirming predictive utility but merely as initial findings that should be replicated and cross-validated in a prospective study,” Shaffer et al concluded.