23 November 2011. At the 4th International Conference on Clinical Trials on Alzheimer's Disease (CTAD) held 3-5 November 2011 in San Diego, California, attendees got an earful about an arguably neglected topic in Alzheimer’s research, that is, the use of event-related potentials (ERP) as biomarkers of disease. Part 1 of this series covered how these measures of cognitive function taken by electroencephalography (EEG) were reported to have predictive value in staging AD. But CTAD also featured talks on another application of ERPs that appeals to scientists. It is their use in determining whether a drug is hitting the desired target. “Instead of testing a drug in 100 people to see if after 18 months there are any clinical effects, you can treat a smaller number of patients and look for small differences in ERPs that would indicate target engagement,” said Lon Schneider of the University of Southern California, Los Angeles. EEG potentials have been used in drug development for anti-epileptic and psychiatric drugs for many years to see whether a drug is affecting central nervous system function, but they have not been widely used in AD. Part of the motivation for revisiting ERPs in AD drug discovery comes from recent failures in AD clinical trials. “People are going back and trying to re-evaluate the tools they are using,” Schneider told ARF.

For ERPs to help drug development, they need to be characterized in animal models of disease, where drugs are initially tested. In his talk at CTAD, Steven Leiser, Lundbeck Research, Paramus, New Jersey, described his work in determining whether there is a rodent equivalent of P300. “Are rats that smart?” he asked the audience. Backtranslating a common ERP paradigm from human P300 research into rodents, Leiser developed a test in which rats had to discriminate a target tone from more frequent distracter tones; they then had five seconds to poke a niche with their noses. If they got it right, they received a food reward. The rats learned to perform this task by day six. Leiser and colleagues took EEG recordings during the task and then studied the ERPs elicited by the target tone. By day six, the researchers could clearly detect an ERP component in the rats’ recordings whose timing correlated with that of the human P300. The amplitude of the rat P300 correlated with how well they performed the tone recognition task, Leiser reported.

In addition, the rat P300 recapitulated some of the pharmacological attributes of the human counterpart. When the experimental anticholinergic agent scopolamine is given to healthy individuals, it impairs their memory and reduces their P300 response in a way that mimics, for experimental purposes, some effects of AD. Donepezil, an acetylcholinesterase inhibitor drug that many AD patients take, counters scopolamine’s actions. Sure enough, when Leiser gave scopolamine to the rats, the agent disrupted the rat P300, and subsequent donepezil administration rescued it.

Changing direction back to the usual trajectory of translational research—that is, from rats to humans—Niclas Brynne of AstraZeneca provided an example of how his group deploys ERPs to gauge a drug’s efficacy in a human drug trial. He presented results from a Phase 2 multicenter, randomized, double-blind, placebo-controlled crossover study to evaluate the pharmacodynamic effects of single and multiple oral doses of an agonist of the α4β2 subtype of the neuronal nicotinic acetylcholine receptor. The trial compared the effects of the drug, called AZD1446, versus placebo and a single dose of donepezil on quantified EEG (qEEG)—these are measurements of spontaneous brain waves—and ERPs in 40 patients with early AD (MMSE scores of 18 to 24) over the course of nine weeks. (For a status update on this drug, see ClinicalTrials.gov.)

In AD patients, the power of α-type brain waves detected by qEEG is reduced compared to that in cognitively normal, healthy people (see ARF related news story). At CTAD, Brynne presented data suggesting that both AZD1446 and donepezil ameliorate these changes in α rhythms and in a slow wave index seen in AD patients. The drugs also reversed the patients’ defects in P300 latency and amplitude. These findings “support target engagement by AZD1446 with possible positive effects on cognition,” Brynne said at CTAD.

Although ERPs are not in the mainstream of AD drug development, their moment on center stage at CTAD may signal a change in the field. “We have shown in the symposium that ERPs can provide value, particularly when we are in a situation in which new drugs are not proving successful,” Schneider told ARF. “Many drugs are failing in Phase 2 and 3, so we are asking ‘Could we have predicted those failures earlier using different markers?’”—Laura Bonetta.

This concludes a two-part series. See also Part 1.