I managed to find time last November to attend the second CNS Summit Meeting. It seems difficult to find time for these events, but I am always pleased to have the opportunity to attend. The theme for this event was development and innovation in CNS drug development, a topic close to my heart. Plus the Boca Raton Resort & Club is no bad place to hang out. We started with an excellent ‘Call to arms’ from Patrick Kennedy, an inspirational speaker (and if you closed your eyes you could hear the same rich speaking tones of his uncle John). His topic was ‘The Importance of Collaboration and Innovation in Developing New CNS Treatments’. This is just one of the drives for innovation, a keenness to establish collaborative arrangements in a pre-competitive space. The theme is one of sharing, especially of data collected for validation purposes. I’m convinced this is an idea whose time has come. Working in psychiatric and neurological drug development is a considerable challenge. Luca Santarelli, Sr. Vice President at Roche, has a magnificent slide on which he uses a Dantean metaphor comparing different therapeutic areas – with CNS development mapping to ‘hell’.
I get the opportunity to work in a number of CNS indications, but Alzheimer’s disease (AD) still occupies more than 50% of my professional time. AD seems to me the quintessential example of the challenging CNS disease. We need desperately to find a cure, or even some symptomatic relief, as the number of cases in many countries is set to treble over the next few years. Not too long ago we had a number of compounds in late stage development, most of which were focused on amyloid plaques, one of the hallmark pathologies of this dreadful disease. Things are not so rosy now, with lots of high profile failures in the interim, though we still have Bapineuzumab and Solanezumab to read out.
Remedying CNS disease usually means finding a compound for diseases for which the cause, or causes are seldom known. We have become adept at screening for disease and have often met with success in identifying the pathological markers and have had some success with developing biomarkers. We’ve also become very aspirational with respect to compound development, genetic screening, neuroimaging, etc., and have ploughed a good deal of cash into the enterprise and into the aforementioned collaborative initiatives. Yet there is still a good deal to do be done, especially with respect to the measurement instrumentation we employ.
For example , take cognition, a field of key interest for me. As is the case for many CNS indications, when we need a measure we tend to borrow from clinical psychology and end up employing tests that were rarely developed for repeated assessment of the kind we need to conduct to detect drug effects. A consequence of this approach is that we have employed measures such as the Mini-Mental States Examination (MMSE) for recruiting patients and the ADAS-cog for measuring efficacy. However neither test was designed for the purposes to which we have put them.
The inadequacies of the MMSE were recently very publicly highlighted in a BBC drama ‘Five Days’, which began with a caregiver tutoring her AD suffering mother on the content of the scale! The inadequacies of the ADAS-cog have also been evident for some time. This is not the place to rehearse the various issues attached to its use, but we might profitably consider a key issue, the presence of ‘ceiling’ effects. Here I am referring to tests on which study participants tend to perform perfectly and which are consequently incapable of showing improvement. Previous studies have reported that performance on nine of the 12 most popularly used subtests reaches ceiling in about 80% of all patients. The three tests that are not prone to this effect all assess memory. A popular comment in defence of the continued use of the ADAS-cog is that a genuinely efficacious drug would be capable of demonstrating improvement. This may or may not be true, but it still seems troubling that such an innovative, creative and progressive group of professionals such as CNS drug developers would be willing to make do with the status quo. Incidentally, the ADAS-cog can be a confusing instrument to administer and consequently recording scores can be a source of error. We could also bring a little profitable innovation by parsing early returns from site for any conspicuous errors. This could be an automatic process based on known data characteristics, such as the correlations between different subtests.
I began considering the content of this blog a couple of weeks ago when it was still appropriate to wish email correspondents a ‘Happy New Year’. New Year is a time to be resolute and it seems to me that this might be the time to start work on improving the quality of our measurement instruments so that when we do finally explain the causes and develop cures, we can characterize the effects of our drugs using first rate, best of class, validated tools.
John Harrison, PhD
Scientific Consultant, CRF Health