16 March 2012. Continued setbacks in Alzheimer’s disease clinical trials have fueled a sense that candidate drugs have been tested in patients whose disease has already progressed too far. For a better shot at their goal, experts say, sponsors should give experimental compounds to barely impaired people on the verge of further decline. Cerebrospinal fluid levels of the two hallmark AD proteins—amyloid-β and tau—can help identify such individuals, and new work published in the March 6 Neurology suggests a third CSF marker may be coming up the ranks. Researchers led by David Holtzman at Washington University School of Medicine in St. Louis, Missouri, report that CSF levels of visinin-like protein-1 (VILIP-1)—alone or with Aβ42—can predict rates of cognitive decline in people with mild dementia as well as or better than CSF tau and tau/Aβ42.
VILIP-1 is a calcium-binding protein that gets released into cerebrospinal fluid from injured neurons (Laterza et al., 2006; Lee et al., 2008). It came up as a top hit in a high-throughput proteomic screen the WashU group conducted to identify potential CSF biomarkers (see ARF conference story). In a prior study, the researchers showed that high CSF concentration of VILIP-1 could distinguish AD patients from cognitively normal seniors and from people with other neurodegenerative disorders. In addition, both VILIP-1 and the VILIP-1/Aβ42 ratio predicted conversion to dementia (i.e., going from a Clinical Dementia Rating of 0 to 0.5) in the cognitively normal group over a two- to three-year follow-up (Tarawneh et al., 2011). However, beyond the binary outcomes of who did or did not develop cognitive impairment, “what you really want to know, as a clinician, is how fast somebody is getting worse, and if you can block that,” Holtzman told ARF.
The 2011 paper did not follow up on those who became mildly impaired to see if VILIP-1 tracked with rates of further decline. This was the focus of the current paper. First author Rawan Tarawneh and colleagues analyzed 60 seniors who began the study with mild or very mild AD, that is, respective baseline CDR scores of 1 or 0.5, and looked at whether VILIP-1 or VILIP-1/Aβ42 could predict how quickly they worsened over the next two or three years.
For each biomarker, the participants fell into three groups determined by 33rd and 66th percentile values. In the CDR 0.5 and CDR 1 cohorts, as well as the combined group, those with CSF VILIP-1 or CSF VILIP-1/Aβ42 in the upper third declined more quickly on global cognitive measures than did participants in the lower two-thirds for each CSF marker. Moreover, high CSF VILIP-1 and VILIP-1/Aβ42 predicted rate of progression comparably, or somewhat better than, high tau or tau/Aβ42 ratios. “It’s at least as good as tau as a marker of neurodegeneration,” Holtzman said of VILIP-1. Holtzman and colleagues had reported earlier that high CSF tau/Aβ42 ratios, high tau levels, or low Aβ42 correlated with faster decline in people with mild dementia (CDR 0.5) (Snider et al., 2009).
VILIP-1 is not found in neurofibrillary tangles, though it does appear in dystrophic neurites near plaques and tangles. “Why would something be elevated that’s not related to tangles?” Holtzman asked. “We know there is degeneration of synapses and neurons in AD. Maybe that’s what CSF VILIP-1 is detecting.”
The research on CSF VILIP-1 to date suggests that “if you are going to enroll people in trials, you would want to use this type of marker to segregate those most likely to decline quickly,” Holtzman said.
He and his colleagues want to confirm the findings in larger cohorts with mild cognitive impairment or very mild dementia—perhaps in Alzheimer's Disease Neuroimaging Initiative (ADNI) samples and/or people with dominantly inherited AD. Ultimately, the scientists would like to collaborate with a company to see if CSF VILIP-1 goes down in response to an experimental AD treatment, Holtzman told ARF.—Esther Landhuis
Tarawneh R, Lee JM, Ladenson JH, Morris JC, Holtzman DM. CSF VILIP-1 predicts rates of cognitive decline in early Alzheimer disease. Neurology. 2012 Mar 6;78(10):709-19. Abstract