Targacept: Commentary NeuroInvestor
(9/16/08 comment on Phase IIb Alzheimer's data)
Targacept/AstraZeneca's TC-1734/AZD-3480 underwhelmed with its Phase IIb Alzheimer's results, topline data released by partner AstraZeneca. It did not definitively fail, even though it missed its primary endpoint, but even when looking at the silver lining, it does appear expectations for this drug in dementia will have to be ramped down. Here are some of the interesting findings and our conjectures:
1) That confounding placebo effect. The placebo group did not show the modest decline one would expect over just three months, indeed they improved. And of course, the safety/tolerability profile of the placebo was outstanding. In retrospect, AZ may/should kick themselves for not running a six month trial, where such an effect is likely to diminish.
2) The benefit of AZD3480 did not outshine Aricept in any consistent way, though it did outperform Aricept on the ADS-CGIC. That is a caregiver/prescriber assessment of patient outcome, which to some degree represents a shot at capturing real-life functional gains that do not translate well into cognitive test scores, but this kind of data is also more subjective and 'squishy'.
3) The report of benefit on the blunt-instrument MMSE is not entirely reassuring, we would have hoped to see some trend on the more detailed CDR, and perhaps something will yet come from more detailed analysis of the several domains assessed.
4) Aricept did not perform any better, and indeed was worse on the ADS-CGIC. It is a reminder that even this lucrative gold standard drug doesn't present all that high a hurdle to match, and matching it is a consolation prize at best.
5) The patient population in this large (567 patients were dosed, at 84 sites!) trial was from Europe and Canada. We will be curious to see what proportion was enrolled in Eastern Europe, where the quality control and experience that can be assumed in clinical trial execution is not as assured, but rapid enrollment is. Just as we have some qualms about Medivation's reliance upon 181 Russian patients, we wonder if there was any divergence associated with those trial sites.
6) There is a modest signal of effect here, but based on what is known thus far, one would have to project a drug which is as effective, but probably not superior to, Aricept in terms of efficacy, but has better tolerability. Whether that will be enough to persuade AstraZeneca to go into Phase III (which would have to be at least six months of treatment) remains to be seen.
7) This raises the stakes on the schizophreniform cognition Phase IIb results, which will be out in December, the same month that AstraZeneca has to decide whether or not they are going to go into Phase III for AD or schizophrenia, and pay the requisite milestone to Targacept. ADHD data will be out in early 2009, but AZ will have to make its decision before then.
8) While comparing studies is a dicey proposition, it is worth noting that Memory Pharma's nicotinic alpha 7 drug (MEM3454), which has a different target than AZD3840, which is an alpha4beta2 modulator, produced a statistically significant effect on the CDR in just eight weeks. Interestingly, both drugs showed an inverted-U response curve, with the mid-dose performing best. This leaves open the question of which NNR target is optimal for AD, and what about the pharmacokinetics of these drug/receptor interactions lead to this dimunition of effect with higher doses.
