Biotech Values

[pharmaclown35]

pharmawire article on TauRx's Rember.

also a ref to this article here:

http://www.memorycenternj.com/index.php/uncategorized/taurxrember-is-it-for-real/


18-Sep-08 11:45 TauRx's Rember looks promising, but data still has to be replicated in larger studies

Story * Rember, a potential first-in-class tau aggregator, is generating interest as an Alzheimer’s treatment though larger studies are called for
* Statistical analysis was designed to test the viability of treatment and lowest effective dose.
TauRx Therapeutics' recent Phase II 24-week primary efficacy data announcement has garnered considerable interest and controversy, CEO Claude Wischik said.
Addressing market misconceptions over its lead product Rember, a new formulation of methylene blue for the treatment of Alzheimer's, Wischik said the trial was not a post-hoc subgroup analysis and still needs to be replicated in a larger number of subjects.
Rember has also been used to treat urinary tract infections.
Recent market reports and some physicians have called into question the hype surrounding TauRx's Rember, and whether that can be replicated in larger studies. If successful, the drug could be a potential first-in-class tau aggregator for the treatment of Alzheimer's disease (AD).
The Alzheimer's space over the last year has been rife with failure, such as Neurochem's Alzhemed, Myriad's Flurizan, and more recently Wyeth and Elan's bapineuzumab, which failed to meet its primary endpoint.
"There is a lot of excitement about this agent, but I’m a bit more cautious," said Dr Gregory Jicha, assistant professor of neurology, University of Kentucky, Sanders-Brown Center on Aging.
The study was done in a very limited number of study sites, as opposed to an international study. "We’re on a rollercoaster ride with some of these agents, just like Alzhemed and Flurizan, which were very promising in the beginning," Jicha said.
Wischik said the primary efficacy end-point at 24-weeks was prespecified not just as a general intention to analyse ADAS-cog change from baseline at 24-weeks, but was very precisely defined in terms of a specific statistical model, with predefined methodology and covariance structure.
The prespecified analysis preserved multiplicity by using the Westfall correction, which increases p-values to take account of simultaneous testing of different dose groups, he said.
TRx-014-001 was an exploratory dose-range finding study whose purpose was to discover whether tau aggregation inhibitor therapy was viable and what was the lowest effective dose, said Wischik. "It was the largest disease modifying Phase II trial for the disease modifying AD indication conducted to date."
Dr Andrew Medhurst, GSK's programme leader for H3 antagonists, said the Rember data was certainly an interesting outcome. "It’s going to take several trials before we really understand the picture with different antibodies and approaches," he added.
"It is an intriguing finding, but follow up studies are required to either confirm or dismiss the claims," said Medhurst.
Wischik said the recent results now need to be confirmed in a larger study, but it is inevitable for a completely novel treatment approach which tests a range of doses, that the groups which prove in retrospect to be of greatest interest are necessarily small.
Even so, there were 321 subjects randomised, 227 entering the 50-week phase of the study, and 169 entering the 84-week phase of the study. The study showed that 60mg tid is the lowest dose needed to achieve control of disease progression in AD.
Another question is what Rember's lowest effective dose is. Having demonstrated efficacy against placebo at the 60mg dose as the lowest effective dose in the ITT analysis, testing of higher doses ceases in a formal sense, said Wischik, although the algorithm provides corrected p-values for all doses tested.
To take account of testing of multiple doses in both mild and moderate subjects, a Bonferoni correction was used, he explained, so that the significance threshold was lowered to p = 0.025. The demonstration of efficacy of the 60mg dose in moderate subjects met this criterion after imposing the Westfall correction for multiplicity at p = 0.021. The treatment effect therefore met all requirements for significance as prespecified, he said.
Wischik also countered market criticism that efficacy was demonstrated at 24 weeks only in a small number of subjects. Although it is true that only 28% of the subjects in the study were moderate at baseline, the numbers of moderate subjects available provided a power of 81% for demonstration of the effect size of -5.5 ADAS-cog units in moderate subjects that was observed. "The study was adequately powered to detect the treatment effect in moderate subjects even after allowing for the Westfall or Bonferoni corrections for multiplicity."
There were 181 mild subjects entering the 24-50 week phase of the study, of whom 41 received the 60mg dose, 28 received the 30mg dose, and 51 were in the control arm. This provided adequate numbers to confirm in mild subjects the efficacy first seen at 24 weeks in moderate subjects at the 60mg dose. Efficacy was also demonstrated after pooling mild and moderate subjects, and separately in the severity subgroups, he said.
The efficacy of the 60mg dose was again confirmed at 84 weeks, in that subjects treated with Rember at the 60mg dose had not declined significantly from their baseline scores.
The analysis at the 24 week interval did not invalidate subsequent analyses. Although senior TauRx management had access to this analysis, due to strategic planning, the conduct of the study remained entirely double-blind at the operational level, for the investigators, patients and study monitors in the trial over the duration of 84 weeks, said Wischik.
The company did not pool placebo and 100mg subjects after the 24-weeks analysis. Subjects originally randomised to placebo were switched to 100mg bd after 24 weeks, and then were re-randomised to either 60mg tid or 100mg tid after 50 weeks. The switch to a presumptively active dose after 24 weeks was required because of ethical objections raised at the trial design stage against keeping subjects on placebo for longer than 24 weeks.
It was only because the 100mg dose was found to lack clinical efficacy for reasons of formulation that permitted subjects randomised to the placebo/100mg arm to be used as the control arm at the 50-week time point. This analysis therefore respected the original randomisation, and did not include any post-hoc pooling of subjects, stressed Wischik.
It was not known prior to initiation of the study that the 100mg dose would lack clinical efficacy, he said. This only emerged after the 24 week analysis, and the study was allowed to continue as designed. This lack of efficacy first seen after 24 weeks was further confirmed in subjects transferred from placebo to the 100mg bd dose after 24 weeks.
Calculation of an inferred placebo decline over weeks 24 to 50, by subtracting the small effect observed for the 100mg dose over the first 24 weeks in a mixed effects repeated measures analysis, revealed that there was essentially no difference between the expected placebo decline over 50 weeks and that actually observed in the placebo/100mg arm used as control for the 50 week analysis.
Therefore, the fortuitous failure of the highest dose 100mg capsule provided a useful internal minimal efficacy comparator arm for the purpose of the 50 week analysis. This made the study much more powerful over 50 weeks than had originally been intended, said Wischik.
The original plan had been for the 50 week analysis to have been conducted with reference to historical controls, despite the well-known weaknesses of this approach. This was because there was at the time no ethically acceptable alternative to demonstrating a long term disease modifying effect in a monotherapy design.
Subjects originally randomised to the 100mg dose originally remained on this dose throughout the study and were not pooled with any other group. They remained a separate study arm, despite the fact that the formulation defect prevented the release of the active form of Rember. This study arm was not used for main analyses at either 50 weeks or 84 weeks.
The possibility of an apparent effect on disease progression by biased retention of non-declining subjects was also examined, he said.
There was no evidence of biased withdrawal to 24 weeks, but there was evidence of biased withdrawal occurring between weeks 24 and 50. This occurred selectively in the minimal efficacy treatment arms (placebo/100mg bd control arm and the 100mg tid treatment arm).
The effect of this biased retention was to reduce the apparent rate of decline in the control arm, leading to an underestimate of the effect size. Analyses conduced with and without correction for biased withdrawal both showed that the effect of treatment with 60mg tid remained significant. Therefore, the effect of treatment with Rember at the 60mg dose in retarding rate of disease progression over 50 weeks cannot be explained by biased retention. There was no evidence of biased retention in the 60mg tid treatment arm over 50 weeks.
The most striking outcome of the trial was demonstration of the treatment effect precisely in those brain regions known to be affected by tau pathology in prior clinico-pathological studies, said Wischik. In particular, the demonstration of treatment efficacy in the medial temporal lobe structures, the regions affected earliest and most severely by neurofibrillary pathology, supports the possibility of early intervention in disease progression from Braak stage 2 onwards, when these brain regions are first affected by the disease.
The trial was conducted in 17 trial centers, 16 of them in the UK, and one in Singapore. All of the trial planning, management and oversight was conducted from UK.
There is an extensive and distinguished publication record over 24 years from Wischik's research team, advocating the tau pathology of AD as the most appropriate therapeutic target for prevention and treatment of AD.
The company's recent clinical trial is the result of 24 years of research aiming to discover a therapeutic approach based on the neurofibrillary tangles originally discovered by Alois Alzheimer.

by Kimberly Ha in New York