2013

The Perception of Innovation: The NIR Innovation Survey

Posted 12/1/13, an excerpt from NP December's review of NeuroInnovation

In early November, NIR conducted a survey of its subscribers and a number of others active in the neuroscience area, asking them to nominate the most important innovations of the past three years. The categories for nomination were left quite open, including therapeutic agents, diagnostics, development tools, and other facilitative approaches to partnering and funding. A total of 83 nominations were received, which while just a small proportion of those polled, provided an interesting window into just what innovations are seen as of value by those directly engaged in R&D. The most glaring pattern was the lack of consensus--the nominations for 'most innovative' were all over the conceptual map. The one element that earned the most nominations is the category of RAADs, rapid-acting antidepressants, and the resurrection of the ketamine model for treating depression, and the rise of alternative mechanistic routes to the same goal, has at least a couple of things going for it. First of all, it addresses a phenomenon that is tangible, even if fraught with the risk of placebo response: People within and outside the pharma industry have no problem understanding the idea of immediate depression relief, in contrast to the two to three week window generally cited for the array of existing drug options. Secondly, the early clinical data, for ketamine, Naurex's GlyX-13, and AstraZeneca's AZD-6765, has been positive, though the caveat of caution regarding small-sample, early-stage data applies strongly here. But if the early data proves replicable, this category of antidepressant could offer the pharma industry access back into a depression market where pharma fortunes were made, but have receded in the face of generic drug competition.

In terms of actual therapeutic agents, the only other program to receive much note is Acadia Pharmaceuticals' pimavanserin, the innovative antipsychotic for Parkinson's (and hopefully, Alzheimer's and beyond in the long run). Strong Phase III data in an underserved population distinguishes pimavanserin, as does the prospect of single-trial approval by the FDA. To us, a somewhat surprising omission was in terms of any frequency in the mention of nicotinic alpha 7 modulators--which had seemed so high-profile during 2012. It does seem that there is some skepticism about the likelihood that the next drugs to report (from Targacept and AbbVie) will replicate EnVivo's success, as opposed to Roche's failure several years back.

Otherwise, the emphasis seemed to be on tools, but with little agreement. The advent of centralized ratings for CNS drug trials received some support, but the data underlying the marketing claims for this expensive process is mixed, there have been instances where centralized raters provided better drug-placebo separation, but others where the converse was true. The imaging of amyloid and tau in the service of assessing Alzheimer's patient populations, and the binding and eventual effect of therapeutic agents, received several votes. There is no doubt that this is a technical achievement that provides subtyping and binding data in the AD population, it remains to be established that this subtyping/binding data is genuinely relevant to AD and predictive of therapeutic outcome. We were surprised to not see more frequent mention of novel clinical trial methodology aimed at reducing the signal-obscuring effect of placebo response, such as Maurizio Fava's SPCD trial design and other 'run-in' models: Given the devastating effect of placebo responders upon the elucidation of genuine treatment effect, one would have to consider this an important step forward, even if it adds expense and complexity to the trial protocol.

The ambiguity that pervades the classification and diagnosis of many CNS disorders is at the root of votes for RDOCs, the NIH-sponsored program looking at devising functional clinical endpoints that cross traditional diagnostic category borders, and the ongoing work on defining endpoints for assessing products that seek to treat prodromal Alzheimer's. There was not a lot of enthusiasm for new models of CNS disease, though iPS cells, functional MRI, and optogenetics all received some votes.

 

PTSD: 'Those who learn too well from history are doomed to repeat it'

Posted 10/9/13, an excerpt from NP October's review of Anxiety and PTSD

PTSD (Post-Traumatic Stress Disorder) is an anxiety disorder, but it stands on its own for several reasons, and is thus addressed separately in this review. First of all, PTSD epitomizes the concept of environmental induction--an anxiety disorder directly triggered by an external event or events, and to some degree, is a result of overlearning from events, or learning gone haywire. It is hardly rare, with lifetime prevalence estimates of 6.8% in the United States, which would yield over twenty million individuals who have had PTSD symptoms at some point. It is also a diagnosis that remains controversial in terms of its definition: While the symptoms of panic, 'flashbacks', depression, anxiety, and insomnia may be generally agreed upon, there is much debate about where one draws the line about contiguity to the traumatic event(s). There is no disagreement that direct exposure to extreme physical risk/stress; such as rape, combat, sexual/physical abuse, can lead to PTSD, but there are many questions as to how proximal such events have to be. Does one have to be the direct target of the trauma? Can PTSD be incurred by those who witness the event but do not suffer the direct physical consequences? Proximity can be interpreted quite literally, as was the case in assessing and treating the longterm aftereffects of the 9/11 attacks in Manhattan. Compensation for treatment expenses has been restricted to people who lived within a specific geographical area around the World Trade Center and the resultant toxic dust plume. For example, Manhattan residents who lived and worked north of Houston Street were arbitrarily defined as not subject to 9/11-induced PTSD, and not eligible for compensation. Admittedly, a line had to be drawn somewhere, given that claims were filed by those who claimed vicarious traumatization from having witnessed the collapse of the twin towers on television. But these are arbitrary lines, because there is no way to objectively measure the extent, or source, of psychological injury, and the implication that 'NoHo' conferred some type of immunity to PTSD is absurd on its face.

It is clear that there are vulnerability factors at work: One estimate is that two-thirds of people in the First World encounter potentially traumatizing circumstances at some point, but with the aforementioned 7% prevalence rate, that would indicate that just one out of every nine individuals will develop PTSD in response to such exposure. Not all first responders who pull body parts from the crumpled wreckage of buildings or automobiles develop clinical PTSD, and yet there are those who do develop PTSD with far more attenuated exposure, there are individuals who are genuinely susceptible to 'vicarious traumatization.' The antecedents of such vulnerability are unclear. Early childhood trauma exposure, often associated with sexual and/or physical abuse, appears to be one major source of such vulnerability, in that such early exposure appears to fuel stress-related neuroanatomical changes. These changes, that include HPA-axis hyperactivation and a locus of amygdala stimulation, 'prime the pump' for subsequent retraumatization. There are likely genetic contributions to such vulnerability, with recent work indicating that individuals with variants of the Oprl1 (nociceptin) gene are more likely to develop PTSD in the face of highly stressful experiences. Such priming may also include epigenetic changes.

While one can ferret out predispositional features, PTSD differs from other psychiatric disorders in one simple aspect: It is the only sudden-onset psychiatric disorder. One can argue that there are cases of acute psychosis, either schizophreniform or bipolar, that seem to emerge with little warning, but one can usually trace these acute exacerbations back to a period of deteriorating prodromal thinking or moodiness. When it comes to PTSD, particularly that seen as a result of trauma experienced in adulthood, such as combat-induced PTSD, there is often no phenotypic harbinger of what will eventuate. This takes a particular toll upon relationships for those with combat PTSD, because for some loved ones waiting back home, the man or woman who returns, after being traumatized in the battlefield, may be quite different from the loved one to whom they had said goodbye.

The diagnosis of PTSD was only codified (via inclusion in the DSM-III) in 1980, as the longterm toll taken by the Vietnam War could no longer be overlooked, and clinicians like psychiatrist Judith Herman began to focus attention upon the previously unacknowledged frequency of sexual abuse. After what seemed to be an initial explosion of trauma patients, as abuse victims came out of the shadows and clinicians started asking the 'right' questions, there appears to have been a subsequent and significant decrease in the incidence of childhood sexual abuse. Over the past two decades, there has been a 62% reduction, according to one well-regarded academic report. That reduction is likely attributable to greater clinical and law enforcement awareness and responsivity. But the legacy of Vietnam has now been recapitulated by more than a decade of war in the Middle East and Asia, and the public profile of PTSD has been elevated by the now-more-systematic and sophisticated assessment of symptomatic veterans. The Veterans Administration reported a year ago that 30% of Iraq/Afghanistan veterans seen in the VA system have received PTSD diagnoses, while estimating--or guessing--that the prevalence amongst all Iraq/Afghanistan veterans, including those who have not sought VA treatment, is less, 20%. This lower estimate is viewed with skepticism by some outside observers, who believe the overall rate is in the 30-35% range. With more than 1.6 million US veterans of Iraq/Afghanistan, this would predict a PTSD population of over half a million individuals comprised of just the veterans of those two wars. The scale of impact has finally led to a substantial allocation of funds for research into its treatment, courtesy of the Department of Defense and VA. In September 2012, those agencies announced grants, totalling up to $45.3 million, to develop a consortium of researchers pursuing biomarkers and treatments pertinent to PTSD.

There is a school of thought that rapid pharmacologic intervention-chemoprophylaxis-can reduce the likelihood of enduring PTSD symptoms, either interfering with the autonomic hyperactivation that contributes to the initial 'imprinting' of the trauma 'circuit', or undoing the memory consolidation process that writes that cognitive circuit into neurobiological stone, with propanolol and morphine among the drugs tried in that context. However, the data to this point is mixed. The Emory group that has worked on the Oprl1 gene hypothesis has also shown that a Scripps Research Institute compound that acts as a nociceptin receptor agonist prevented the development of PTSD symptoms in traumatized mice--if given before or just after the traumatic event. If that compound reaches clinical development, PTSD would be a prime target.

The treatment of PTSD always includes some form of psychotherapy, often accompanied by pharmacotherapy that is symptom oriented, aimed at alleviating the dysphoria, anxiety, and/or insomnia that are so salient--attenuating the autonomic hyperactivation endemic to PTSD. For the most part, the pharma industry (as opposed to federally supported academic programs) has emphasized PTSD as a label-extension for drugs that have often had lackluster records in traditional indications (e.g. iloperidone) or were approaching the end of their patent-life (e.g. paroxetine, gabapentin, topiramate, quetiapine). The exceptions were GSK's trial of their NK-1 antagonist orvepitant, a trial that failed and was terminated early, and Marinus Pharmaceuticals' Phase II trial of ganaxolone in PTSD, a 120pt trial that should be completed early next year. Vasopressin-1a modulation of the stress response system is being pursued by Azevan Pharmaceuticals (with PTSD prioritized) and Roche (where autism has been emphasized).

Other studies looking at ameliorating the symptomatic impact of traumatic events include a trial utilizing IV hydrocortisone being done in Israel and a 102 pt VA-sponsored study of glucocorticoids.
Anti-seizure medications have received some attention as PTSD treatment candidates, including the aforementioned Marinus trial of ganaxolone, a completed study of levetiracetam, and a Yale trial that will use zonisamide in the treatment of patients carrying a dual diagnosis of both PTSD and alcohol abuse. There are some relatively exotic pharmacologic approaches also under development, and obviously, not by the pharma industry. In Israel, no stranger to traumatizing events, there is a 60pt trial utilizing intranasal oxytocin, in the hope that the reduction in social anxiety and increase in pro-affiliative motivation seen with oxytocin will serve as a counterweight to the chronic anxiety and social wariness often seen in PTSD. MGH is running an 80pt trial aimed at preventing the development of PTSD symptoms using intranasal oxytocin, but as is noted in the full review of anxiety, oxytocin may intensify the encoding of negative memories, thus it is not a nonselective anxiolytic.

Perhaps the most unorthodox tactic is the use of MDMA ('Ecstasy') as an adjunct to therapy, which followed a history of anecdotal reports where practitioners (before the drug had been banned) had reported treatment-enhancement, and given a sparse legacy of case histories supporting the use of psychedelic substances in promoting transformations of cognitive and interpersonal 'templates' in trauma and terminal illness contexts. One 20pt study was published in the Journal of Psychopharmacology, and claimed to show durable reduction in PTSD symptoms, though there were numerous anomalies in the construction and control of the trial, leaving the conclusions open to debate. A 24pt followup trial sponsored by the 'Multidisciplinary Association for Psychedelic Studies' is still enrolling patients. Even if the study shows efficacy, the recalcitrance of the US government around the analgesic use of marijuana would predict that the current generation of PTSD patients should not expect access to such unorthodox options in the foreseeable future. The situation would probably be somewhat different in less dogmatic, more pragmatic societies, like Israel.

The variety of cognitive and/or behavioral therapies in use and under investigation is broad--including CBT, exposure therapy, mindfulness training, meditation, and yoga, with no definitive assessment as to relative merits. As is discussed within the anxiety review itself, there are also some attempts to enhance the effects of psychotherapy via drugs that aid in the consolidation of memory, ostensibly strengthening the impact of the new learning provided by cognitive therapy. D-cycloserine, seen as a way to enhance the consolidation of new learning, is being tested by Weill Cornell as an adjunct to CBT, where a 124 pt trial is expected to be completed next year. The alpha2a-adrenergic antagonist yohimbine has been used to facilitate the retrieval of traumatic memories in order to make them available for extinction via CBT, but given yohimbine's anxiogenic properties, that is unlikely to achieve widespread utilization, particularly since consolidation-enhancing therapies may depend on the efficacy of the CBT intervention to dictate whether the effect is one of ameliorating, as opposed to exacerbating, the PTSD symptoms.

For the most part, the development of pharmacotherapies addressing PTSD will continue to be primarily a federally funded cottage industry of micro-biotechs and academics until a well-designed study shows a clear road ahead for a mechanism not already fraught with bureaucratic entanglements. Of the companies surveyed in NeuroLicensing 2013, just AstraZeneca, Lilly, and Shire Pharmaceuticals cited any specific interest in PTSD as a therapeutic focus, and all of them frame it as an indication to be pursued on an opportunistic basis. For others, PTSD may be seen as too small a population--even though it comprises millions of individuals--or governed by unpredictable environmental events that might skip a generation, leaving the demand for treatment ephemeral. By that somewhat cynical NPV calculus, the market potential for a PTSD drug might dip if we were to go through an interregenum like the fifteen years between the end of the Vietnam War and the first Gulf War. Given the geopolitical realities, while that would be a nice problem to have, it appears improbable.

Revisiting the Goal and the Model

Posted 9/17/13, an excerpt from NP September's review of Alzheimer's programs

It may be that the concept of 'cure' for Alzheimer's will prove as superfluous as it is elusive. Given the crushing societal weight of the growing dementia burden, any drug that provides even a couple of years of extended functionality will be useful and cost-effective. If quality of life can be maintained until its natural end, it will not matter if the underlying Alzheimer's pathology has been only held at bay. Considering the tens of millions of adults who have Alzheimer's, or who teeter on the slippery edge of cognitive impairment, the market for a therapeutic that safely slows Alzheimer's march to disability would dwarf anything else in the pharmaceutical industry. Even though the therapeutic state of the art has advanced rather little in comparison to the enormous investments made, the reward for success remains outsized.

With all due respect to Sanofi's decision to walk away from Alzheimer's because the science is not yet mature, Alzheimer's is potentially the richest potential treasure trove in 21st Century medicine, and the industry cannot wait around for someone else to figure out how to most effectively address it. This is not a context where one has the luxury of proving the mechanistic premise first, because there is not the time for the risk-reduction only provided by certainty. This has several principal ramifications:

Moderating Goals: One does not have to flatten the slope of decline, just moderate it. The drastic pharmacoeconomic scenarios presented by Alzheimer's would be ameliorated if a compensatory drug could be shown to improve functioning long enough to reduce the eventual time spent in custodial care. Even if the disease process is not altered per se, the duration and burden of the disease in its severe form could be. Thus it is necessary to be flexible about therapeutic goals for the time being, because even as we shoot at the disease-process, we do not yet know for certain where one should even aim. Some functional enhancement tactics may be well worth the investment, even if they do not turn out to modify the disease process significantly, and the nicotinic alpha 7 agonists stand out here at present--while also raising the possibility of modulating the course of the illness. Modifying the disease itself remains the elusive prize, but seeking to block it at its etiological antecedent is the riskiest approach, since it is likely to be an all-or-none process. Downstream targets, with metal-binding and neuroinflammation standing out in particular at present, are not as dependent upon precisely delineating the optimal upstream target.

Diversifying Tactics: As almost everyone now realizes, a comprehensive Alzheimer's strategy cannot be limited to anti-amyloid programs alone. Even if amyloid-targeting mechanisms have a role to play in such disease-modification, the preliminary data suggests that substantial subgroups of the Alzheimer's population may not be sufficiently or safely responsive to anti-amyloid therapies. There are other targets equally deserving of proof-of-principle testing: Tau, metal-binding, inflammation are the three that stand out at this moment in time. The only prudent strategy is to develop a spectrum of approaches, ranging from upstream bets-on-a-mechanism to downstream correctives which less dramatically, but with clinical relevance, sustain function and independence. The resources devoted to just the bapineuzumab and solanezumab Phase III programs could have funded a dozen or more well-constructed and scaled Phase IIb programs for compounds from a variety of mechanistic tactics. The results would have given us far more information about what does and does not work in Alzheimer's than did these misguided. It may seem to be contradictory in the context of the time pressure noted above, but Phase IIb is not optional. Jumping into a four year $400 million Phase III program without sufficient justification from Phase II ultimately is fare more wasteful of time, money, and finite corporate tolerance.

Collaborative Risk-Sharing: For companies who want to go for the gold ring of arresting Alzheimer's at its biological source, forming consortia around a specific mechanism may be the only way to achieve that goal without risking too much of their credibility and capital. The intra-industry and public-private cooperative efforts that have been productive in biomarker projects like ADNI, exemplify the ethos from which therapeutic development must draw, albeit with a smaller array of participants, since drug-development-by-committee would have its own pitfalls. One thing that was done right in the bapineuzumab pivotal program was that three companies, Pfizer, JNJ, and Elan, ended up collaborating, and this allowed them to move on from bapineuzumab's failure. Lilly would be well-advised to consider something similar for solanezumab, as might Merck regarding MK-8931, because a total failure in programs of that scale could have lasting and devastating ramifications for the CNS aspirations of any Big Pharma.

Alzheimer's Phase III programs are so expensive and time-consuming that the old model, where every company had its own inhouse entrant in any mechanistic category, is simply not affordable. That model spurs companies to sustain redundant and even inferior molecules for the sake of owning one. Instead, two or three companies that combine forces around a single GSM, one tau therapeutic, or one mAB against beta-amyloid, can not only absorb setbacks more easily, but will then have resources freed up for other approaches and the pursuit of other CNS indications. A replication of past history, where the industry expended its resources on a herd of me-too SSRI and atypical antipsychotic drugs, is not going to be economically viable in Alzheimer's. Collaboration means relinquishing hubris in the service of sharing both the risk and the reward, and one-third or one-half the revenue that will come from a successful Alzheimer's drug will be--or should be--more than sufficient for any pharma company.

 

A Tale of Two Sectors

Posted 7/19/13, an excerpt from NP July/August's overview of the neuro sector

"It was the best of times, it was the worst of times, it was the age of wisdom, it was the age of foolishness, it was the epoch of belief, it was the epoch of incredulity, it was the season of Light, it was the season of Darkness, it was the spring of hope, it was the winter of despair, we had everything before us, we had nothing before us, we were all going direct to Heaven, we were all going direct the other way." --Dickens, Charles A Tale of Two Cities 1859
So how has 2013 gone so far? It depends upon who you ask. It has been the most blatant of bimodal distributions, this first half of 2013 for neurotherapeutics. On the one hand, for investors in Acadia Pharmaceuticals, 1H:2013 was a fever dream of glad tidings piled upon good news. Similarly, Biogen-Idec had a generally scintillating first half. Acorda reached new highs, Lundbeck had Selincro reach the EU market and has Brintellix poised for regulatory approval. Indeed, the NIR Index of neuro company valuations reached an end-of-month high of 346.11, the second highest ever recorded since the inception of the index in 1995. However, this reflects the fact that this unweighted index was skewed by the presence of so many mid to relatively large-scale companies who did indeed greatly enjoy the past six months.
Unfortunately, as is usually the case when discussing the macro view of any economic phenomenon, this did not tell the whole story, and there were numerous smaller companies for which 1H:13 was a descent into Hell, particularly those who had their lead program fail in the clinic. Ceregene (CERE-120) and Seaside Therapeutics (STX209) failed in the clinic, while NeurogesX shut its doors, selling its assets to Acorda. Addex fired almost everyone, and will wait for JNJ to produce something with their partnered program, Allon Therapeutics is being dismantled, and NeuroSearch, having sold their only real asset to Teva, is trying to get someone to haul away their furniture. Cortex Pharmaceuticals now exists solely as an insurgent BOD who deposed what was left of the previous management and now reins over an IP trove now bereft of its two most important academic licenses, their intentions completely opaque.
And now--GSK's head of neuroscience has resigned in disgrace. Neuroscience had few friends left at the top of GSK's management, and Jingwu Zang had put himself in a vital and singular position of advocacy.It is now not so farfetched to think that GSK might sell its Shanghai CNS R&D center, beheaded as it is, shuttering CNS at GSK. Until 2016, when GSK's BOD will look at the transformed neuro sector and ask--'why don't we have something like that?' We earnestly hope we are wrong on this, but our guess is, things will get worse at GSK before they eventually get better.

Addendum: With the China scandal now expanding in scope, well beyond CNS, this signals that things are indeed going to get worse.

 

Rapid-Antidepressant Candidates

Posted 1/18/13, an excerpt from NP December' s review of Depression

The Ketamine Model
In 2006, a NIMH-sponsored study received enormous publicity when it reported that the addition of IV ketamine to the pharmacotherapy of depressed inpatients yielded a therapeutic response within two hours. A Yale group had reported similar results several years earlier, but it was the NIMH report that catalyzed attention in the area. Other trials have indicated a durable effect, extending up to two weeks for most patients, approaching a month for a minority. But this comes with a plethora of complicating factors: Ketamine is a NMDA-antagonist similar to PCP, and like most NMDA-blockers, it can precipitate psychotomimesis, dissociation, and other adverse cognitive effects. Indeed, ketamine is abused as a street drug, and the ketamine story raised the question for some observers as to whether the effect reflected a genuine antidepressant impact, or instead was based on subjectively apparent changes in mental activity. The experience of 'surfing' the edge of psychosis would, at the very least, be a distinct and perhaps refreshing contrast to dysphoric bleakness. This is in fact a conundrum for many, if not all, trials testing drugs with this kind of rapid advent of effect--'something' occurs subjectively that is quite different from placebo: the question is whether it is specifically anti-depressant, and whether such a trial is truly double-blinded (some are now utilizing an IV benzodiazepine as the control group). The ketamine data thus far comes from small trials often utilizing just one site; larger-scale, better designed trials are underway. But there are enterprising psychiatrists and even academic centers--such as UCSD--already offering off-label ketamine treatment programs, and the consensus, albeit far from unanimous, is that this is a mechanistic approach of great potential value for the treatment of depression per se-- if it can be harnessed safely.

Harnessing Ketamine
There are several approaches to this balancing act, beginning with some tactical variants. The first tactic is to titrate the ketamine dose quite slowly, similar to what is done in utilizing memantine, a weak NMDA antagonist, in Alzheimer's. One clinical trial of ketamine in cancer patients suffering from neuropathic pain used a slowly increasing rate of ketamine infusion, and reported that over a several day period of infusion, the dose could be (on average) quintupled, but only 7% of patients showed sedation, none were reported to experience psychotomimesis. Frequency of repeat dosing serves some of the same titrating goal, JNJ is running a TRD trial seeking to assess the optimal frequency of multiple dosing for ketamine.
Adding an adjunctive drug that modulates glutamatergic activity via other sites/pathways is another ketamine tactic. A Yale group pre-administered a glycine transporter inhibitor (Org25935) along with ketamine, and reported significant reduction in psychotomimetic symptoms. While they reported this as indicating the potential of GlyTi for treating schizophrenia, using such a drug as an adjunct to ketamine in depression could be considered. AstraZeneca conducted a trial of ketamine accompanied by a mGluR2/3 PAM, something Lilly and JNJ are also trying. Lilly is running trials (in healthy volunteers) of both a mGluR2 agonist (LY2979165) and a mGluR2/3 agonist (LY2140023/pomeglumetad) as adjuncts to ketamine. JNJ is running a trial using Addex's mGluR2 PAM (ADX71149) as an adjunct to ketamine's enantiomer, S-ketamine. The hope is that this may have advantages over racemic ketamine, and that the adjunct will modulate the effect on the glutamate system. Another suggestion regarding utility in schizophrenia was reported by a group that pre-administered an experimental GSK3 inhibitor, with the reduction in psychotomimesis cited as indicating an approach in TRD. A Washington University group is conducting a trial where ketamine is infused for over four days, accompanied by the alpha-2 adrenergic agonist clonidine. We have also seen a group claim that adding lamotrigine provides a safety buffer for ketamine.Given the added burden of IV drug delivery, a Mt. Sinai group is conducting a 20pt trial of ketamine that is administered intranasally. The IP on intranasal ketamine is jointly held by NIMH, Yale, and JNJ's Husseini Manji. Based on the consensus that it has established its therapeutic benefit better than any of the competitors thus far, considerable work on ketamine continues.

Alternatives to Ketamine: GLYX-13
With the challenges presented by ketamine with or without an adjunct, a cleaner clinical and regulatory path would probably be available for a drug that offers ketamine's antidepressant advantages, but with a better safety profile.

It appears that Naurex may have solved this puzzle, though they are still early in their clinical testing process. The backstory for Naurex is that, more than a decade ago, Northwestern's Joseph Moskal developed a group of small peptide, NMDA-complex partial agonists. However, his first attempt to spin this program off commercially, Nyxis Neurotherapies, demised in 2004. The program returned to his lab, where further work was done on the lead compound's mechanism. It appears to function as a partial agonist which binds to a novel site near, but not on, the glycine receptor. Originally intended to address neuropathic pain and/or MCI, further animal work led to depression being defined as the primary indication, and the program was resurrected as Naurex, with depression the priority indication. IV GLYX-13 successfully completed Phase I, and the therapeutic index for single dosing has been estimated to be 1000, affording a substantial buffer vis-a-vis psychotomimesis. Whether that margin is maintained in the course of repeated dosing remains to be seen. Naurex has stated that patients did not experience any acute, subjective shift that would signal drug effect, unlike ketamine and AZD6765. Naurex completed their Phase IIa trial back in June, and did not release the results publicly until December's ACNP meeting. However, they had reported that they would indeed go into Phase IIb, and that they would be supported by their present investors in doing so, a roster including the venture arms of Lundbeck, Shire, and Takeda. This gave away the gist of the story, since it would not have occurred had the data been negative.

The Phase IIa data was just released, and in spite of the fact that this was a 116pt dose-ranging study in patients with at least one antidepressant treatment failure (and mean baseline HAM-D scores of 25-26), limited by the small cohort sizes for each of the active treatment groups, GLYX-13 exceeded expectations. The HAM-D results showed clinically meaningful and durable improvement in depressive symptoms within a day of its administration, while results from the Bech-6 Inventory showed an improvement at two hours, at the 5mg/kg dose. 10 mg/kg produced a the largest effect on the HDRS-17, with 3.9 and 4.9 point changes at Day 1 and Day 7, respectively. At seven days, this represented a effect size of .58, and reached statistical significance, in spite of the small size of the dose-cohort. The therapeutic effect largely dissipated by two weeks. The side effect profile was benign, with a trend towards dizziness, experienced by 15% of the patients receiving 5mg/kg. There was also a trend towards mild QTC elongation, seen in 15% of the 5mg/kg group, which while probably not of clinical significance in the context of a time-limited, intermittent dosing regimen, will bear watching with repeated administration. Most importantly, none of the doses produced psychotomimetic effects in any patients.

Like ketamine, GLYX-13 is an IV drug, which would be untenable as a first-line option, but would make it potentially applicable to TRD, particularly more refractory patients. If these results are borne out by multiple-dose, larger-scale testing, with both the 5mg/kg and 10mg/kg doses being tried in an adaptive dosing model, GLYX-13 would offer a radical transformation in the treatment paradigm for depression, one where patients would have an option, most commonly after a first or second treatment failure with a generic SSRI, to try an alternative for which the outcome is very quickly apparent.

Some caveats are in order. After all, NIR has not infrequently commented on the pitfalls of an overenthused embrace of Phase IIb results (e.g. TC-5214, Dimebon); with the results for GLYX-13 being even earlier-stage than those predecessors, and based on just a single IV dose, there is risk of overstating the case here. However, some of the obvious pitfalls have been avoided: The trial was run at twelve sites in the US, and the data were vetted by an external rating service. The population-source factors and data validity questions that undermined the validity of the Phase IIb data for such programs as Dimebon and TC-5214 are not factors here. But the pilot study now requires confirmation from a multiple-dose trial. The possibility that GLYX-13's safety/efficacy profile might be dose/frequency-contingent cannot be ruled out by animal data alone, as reassuring as that data has been. The initiation of repeat dosing in the Phase IIb trial will be a 'moment of truth' for the project.

Alternatives to Ketamine: AZD6765
AstraZeneca has been developing its own IV NMDA-complex targeting compound, AZD6765. Previously developed and eventually terminated as a stroke treatment, their first pilot Phase II in depression was completed back in 2007. Hints of efficacy in Phase IIa were reported from a 22pt pilot study, but in fact these 'hints' were fleeting. A 150mg single dose produced HAM-D changes from placebo beginning at 80 minutes post-infusion, ending some 30 minutes later, returning at the two day point. Only 32% of patients were responders, compared to 15% of those on placebo. This led to a 152pt Phase IIb trial in TRD (two trial failures, mean HAM-D baseline of 28) patients, and the best performance was at 100mg per infusion (three infusions per week) with a MADRS score change of 5.5 points at three weeks, a HAM-D delta of 3.8 points. Like GLYX-13, this drug appears to have an inverted U-shaped dose-response curve. Some patients were described as displaying "transient" responses following each infusion, which appears to have included brief dissociative and euphoric episodes, but durable effect was not seen until two weeks post-infusion, and it peaked at three weeks. While they reported that benefit was sustained for another three weeks, this magnitude of change falls short of what is achieved by current SSRI/SNRI drugs. Another substantial (282pt) Phase IIb trial is now underway, results expected in 2013. But the question is begged: If it takes two weeks and multiple IV infusions for AZD6765 to achieve a magnitude of therapeutic benefit less than is obtained from generic oral therapies, where is the advantage for AZD6765? It certainly is reasonable to speculate that the mechanistic contrast might provide utility in patients refractory to these oral drugs, but with the lack of improvement in speed of relief, this falls short of what might be hoped for in a revolutionary alternative.

In terms of safety/tolerability for AZD6765, 'CNS type' adverse effects have been reported, with dizziness the most frequent (49%), though it tended to resolve within four hours. 15.7% experienced somnolence. Psychotomimesis was not reported in the 100mg group in the single-dose study, but psychotomimetic effects were observed in a small number of patients in the three-week trial, and three patients were reported to have become 'euphoric.' Overall, these results are quite different from those reported for ketamine, in that the onset of effect was delayed, though it appears to have been durable, once achieved via multiple infusions. There is thus something of a divergence between the profile for this drug and the profile for GLYX-13, and indeed compared to ketamine itself.

The Mechanism
The effects shown by ketamine in depression, which Naurex and AstraZeneca seek to emulate, have raised questions as to how this mechanism exerts such rapid antidepressant effects. GLYX-13, like ketamine, produces an upregulation of NMDA NR2B receptor subunits. A Yale group reported in Science that ketamine produced prefrontal activation of mTOR signaling, which in turn led to the rapid formation of new synapses. The authors suggested that the mTOR signaling pathway might offer an alternate target, avoiding the psychotomimetic risks posed by ketamine itself. But this leaves open the question of how this kind of impact on synaptic structure and connectivity can produce such rapid subjective effects, the acute impact is more likely due to the indirectly mediated release of glutamate. An acute effect likely explains the transient dissociative effects seen after each infusion of AZD6765, but there is some mystery about the two week gap preceding durable impact.

Adding another glutamatergic twist, an NIH group has data suggesting that AMPA receptor upregulation accounts for ketamine's antidepressant effect; one hypothesis is that this blockade of NMDA-gated ion channels causes a compensatory increase in AMPA receptors. Interestingly, in animal models, administering an AMPA-antagonist blocks the effect of GLYX-13. It has been reported that there is an acute serotonergic effect for ketamine, and nicotinic receptor effects have been postulated for the AstraZeneca drug.

The State of the Race
With GLYX-13 having completed only a single-dose Phase IIa, the comparison of these programs is very asymmetric. But based on the data thus far, while there may be a acute transient dissociative effect for some patients, AZD6765 does not have the speed of antidepressant impact we would see as congruent with the concept of a RAAD. Such speed is indicated by GLYX-13's pilot results, where there was some change within two hours, relief within 24 hours, and it is seen with ketamine. Looking at the limited clinical and preclinical safety data available, it appears that GLYX-13 also has a therapeutic index superior to both AZD6765 and ketamine, but this is tentative, pending the completion of multiple-dose clinical testing. A 300pt Phase IIb multiple-dose trial for GLYX-13 has begun enrollment, and is projected to finish late in 2013. The design of that trial includes an interesting component, in that it will adapt both dose and frequency to each patient, in a relapse-prevention model.

Genuine paradigm-shifts in the neurotherapeutics field have been an exceptionally rare breed over the past fifteen years. Outside of MS, the list consists of the use of second-generation antipsychotics like Abilify/aripiprazole in TRD, and the development of Provigil/modafinil for Excessive Daytime Sleepiness. GLYX-13 and AZD6765, albeit with limited data for the former, and mixed data for the latter, and the ketamine permutations noted earlier, potentially represent the most dramatic shift in the treatment paradigm for depression since the advent of the SSRIs. As noted above, the speed of impact for AZD6765 could well take it out of the RAAD class. It should be noted that there are some observers who are concerned about the potential safety ramifications of chronic use for a drugs operating via the NMDA complex, a question that will have to be answered to the satisfaction of the FDA in order for one of these compounds to be approved for anything more than acute use. It can be argued that this would be a viable niche in itself, providing depressed patients with subjectively tangible symptom relief while waiting for a SSRI or SNRI to 'kick in.'

If the Phase IIb trials for any of these drugs confirms rapid relief of depressive symptoms, they could be advanced in either, or both, of two tactical avenues. An IV antidepressant could be conceptualized as a 'first response' option that bridges the first few weeks of depression therapy, until a slower-onset oral option takes hold. More likely, it would be at best a second, most likely a third-tier choice, since most payors will insist on at least one trial of an oral generic, and many patients will prefer the familiar oral option. But the TRD population is very large, and if even a modest percentage of those patient opt for an IV alternative, this could be commercially successful. It should be noted that defining IP territory for a ketamine therapy in TRD will be a challenge, since it is already in use as such off-label. Defining and patenting treatment via S-ketamine or a combination therapy might be viable. This could pose a pricing challenge for branded IV drugs, having to compete with a generic IV ketamine offered in locales by clinics with an entrepeneurial bent and a strong trust in liability waivers.

NRX-1074: An Oral Option?
While there is no oral analog for AZD6765, Naurex has an orally bioavailable version of GLYX-13. This is NRX-1074, which is considered more potent, and is approaching IND-readiness by year-end. If it emulates GLYX-13's impact and safety, NRX-1074 would widen the scope of Naurex's market across the full range of depression, and its differentiation in terms of rapidity and duration of effect could make it a best-in-class competitor to the generics, a potential first-tier depression choice, and the biggest psychiatric blockbuster since Abilify.

 

 

Pimavanserin and Persistence

from NeuroPerspective January 2013

Best News of 2012
The Best News is that NeuroPerspective had multiple deserving candidates for Best News, a category that has not been oversupplied in recent years:
Positive Phase IIa data for Naurex's GLYX-13 as a fast-acting antidepressant.
Positive PhIIb data from EnVivo for EVP-6124 in Alzheimer's.
Positive PhIII data from Acadia Pharmaceuticals on pimavanserin in Parkinsonian Psychosis, leading to financing proceeds of $183 million.
FDA approval for Alexza's Adasuve as a inhalable treatment for agitation due to schizophrenia/bipolar illness.
Teva Pharma is going to devote $10 billion to building its inhouse drug-development capabilities and portfolio, with CNS one of the two major areas (respiratory disorders is the other).

Worst News of 2012
Bapi and Sola. For a billion dollars, what did we end up with? No signal for bapi; just a faint signal for sola, seen by some as a datamining artifact; nothing definitive regarding the amyloid hypothesis; and some appalling denial and spin from people who should know better.
Merck Serono, whose neurodegeneration programs were gutted and left to flounder in spin-off limbo by corporate parent Merck KGa.
The failure of the AstraZeneca/Targacept program for TC-5214.
The failure of the Allon Therapeutics PhII trial for davunetide in PSP.
Disappointing clinical results for Catalyst, Neuraltus.

The Blockbuster Concept is Not Dead
Biogen-Idec's BG-12

Regulatory Thinking That Earns the Stereotype
The rejection of Pfizer's tafamidis for an untreated amyloidosis after the Advisory Committee recommended approval on the basis of biomarkers to be confirmed by Phase IV. That could have been both humane and forward -looking. Instead, the decision was neither one.

It was the Advisory Committee that suffered the cognitive occlusion on the NDA meeting on Zogenix's Zohydro, the first acetaminophen-free formulation of hydrocodone. They voted 11-2 against approval, largely because it could be abused--not that it was more likely to be abused than the available opioids, but they were apparently intent on signalling their dissatisfaction with the current REMS plan for opioid analgesics. These are different issues, and shooting down a new analgesic because it has the same abuse liability as the others does not address the REMS issue. If a statement needs to be made, make it in a way that does not unfairly penalize a company for happening to be in the right place but at the wrong time.

2012

Pimavanserin and Persistence

from NeuroGram December 2012

Acadia Pharmaceuticals is one of the very few survivors from the generation of neuropharm companies born in the early-mid 1990s. As is the case for many small firms, their prospects, were eventually and essentially condensed into one project, pimavanserin, a 5HT-2a inverse agonist with antipsychotic effects. Because it does not have direct dopamine-antagonist effects, it was envisioned as a treatment for Parkinsonian psychosis, where motor functioning is further impaired by DA-blockers. Acadia’s other research, indeed their very existence, became contingent upon pimavanserin. External factors played a role in this as well: Biovail, during its all-too-brief renaissance of neuro inlicensing, partnered pimavanserin when it was in its first Phase III testing for Parkinsonian psychosis, and sustained that partnership even after the first Phase III trial failed. Unfortunately, Valeant’s acquisition of Biovail put an end to that partnership and the label-expansion agenda behind it.
Acadia’s revamped, single-dose (40mg) Phase III program has now reported its first data, and pimavanserin provided a robust effect on Parkinsonian psychosis, enhanced performance on secondary endpoints, and did not cause any worsening of motor symptoms. Acadia could not have asked for better results from this Phase III trial, in what was essentially a do-or-die moment for the company. Another Phase III must be carried out, but we are very optimistic that the design and enrollment criteria that worked well for this trial will work again. Acadia will need a partner to fully maximize pimavanserin’s potential, which is could first expand into psychosis associated with Alzheimer’s, but might be fruitfully considered as a dose-sparing adjunct in the treatment of schizophrenia, reducing the dose-related side effects associated with all current antipsychotic drugs, albeit in different profiles and proportions.
There are a few lessons here, most of which relate to what we will call adaptive persistence. Acadia has been steadfastly focused on pimavanserin for the past four or more years. Give the late and lamented Biovail credit as well: They could have simply walked away from pimavanserin and Acadia after the first Phase III failed, and might have been applauded by those who believe pharma doesn’t kill its floundering projects fast enough. But they saw that the issue was dosing, revamped the partnership, and when Valeant then did walk away, they had to provide some funding to Acadia, money that was crucial to the program and Acadia’s survival. This does not by any means prove that all trial failures require or deserve resurrection, but there is a place for adaptive, rational persistence. Now, Acadia has an eminently partnerable success on its hands.

 

Alexza: Wrestling with Agitation

from NeuroGram December 2012

While the cognitive and negative symptoms of schizophrenia are now a major chronic care issue, the positive symptoms often return as the acute-care focus when patients, due to medication noncompliance and/or the fluctuating intensity of their illness, become agitated, with danger to self or others. Alexza's Adasuve/loxapine inhalant delivery system has shown efficacy in Phase III testing, for both schizophrenic and bipolar agitation. It offers an alternative to IM or oral antipsychotic administration (oral medications tend to be too slow, and patients often noncompliant) for agitated patients, which may seem a small issue to outside observers, but is not. During this writer's early psychiatric hospital experience/training, I participated 800-1000 times in the physical immobilization of psychotic, agitated patients, often so that they could be given IM medications. It was and is a primitive process, three or four staff members tackling the patient; hopefully avoiding being punched, kicked, or bitten. Even the IM route lags in its effect for 20-40 minutes, during which time the patient often must remain physically restrained, because recipients of a forcible injection in the buttock tend to not perceive this as compassionate medical care. Forcible IM establishes a climate of coercion and control which is traumatizing, both for the patient being medicated, and for other patients who witness the process. At times, the agitation which required physical restraint spreads amongst other patients, as if contagious. This would be avoided with an inhalant alternative; compliance would undoubtedly be far higher, and its onset of effect much faster. Adasuve produces a very rapid onset of effect (cMax in two minutes), the trials used 10 minutes as the first timepoint, wherein Adasuve produced significant relief of agitation. An inhalant antipsychotic is also an intervention that can be offered earlier in the escalation cycle than IM meds, 'nipping it in the bud' before it approaches a dangerous crescendo.

The question is not whether this is a potentially useful tool for dealing with acutely psychotic, agitated patients; it is. The question is whether the American inpatient psychiatry system, which has been gutted by managed care, still exists in large enough scale to provide an adequate market for the product. But the fact is that the highly agitated, psychotic patients who are the key market for Adasuve are still those most likely to be seen in an ER or inpatient setting. From a commercial perspective, the shrinkage of the inpatient psychiatric world means that marketing Adasuve to psych units and emergency rooms would be manageable for a small sales force. There will be some price point sensitivity for Adasuve relative to IM haloperidol, though IM Geodon is also premium-priced. But there are hidden costs to IM med use in terms of staff time, both for the medication administration itself and the incident reports that have to be completed for each involuntary restraint/medication. Our  2009 estimate of market penetration yielded annual sales potential in the $200-225 million range. With the subsequent REMS restrictions in terms of screening patients who might be vulnerable to bronchospasm, and even with EU approval now appearing probable, the peak annual sales estimate is revised to $200 million. For a company of Alexza's size, this represents a very important 'win.'

Elan Corporation 1969-2013

from NeuroPerspective, September 2012

Elan has announced that it will spin off Neotope Biosciences as a repository of its preclinical/discovery research, largely devoted to targeting rogue proteins (amyloid, tau, synuclein) in the treatment of neurodegeneration, plus early-stage diabetes and oncology programs. Elan's longterm CSO, Dale Schenk, will be the CEO; Elan intends to fund the to-be-public company with $120-130 million (two years of cash), retaining an equity stake. Parenthetically, it should be noted that Elan states Neotope's programs are ten years from the Market, and will require $50-60 million per year. Over ten years, that would thus total $600 million, which means Neotope would have to raise $470-480 million from VC's, stock offerings, and partnerships. That will be a tall order.

Neotope's spin-off will leave Elan comprised of the following:
1) The 50% of Tysabri not already owned by Biogen-Idec.
2) ELND-005, a former Alzheimer's therapeutic licensed from Transition Therapeutics. Elan tried to outlicense this program for over a year, without success, and now plans a trial in bipolar disorder.
3) A stake in the Janssen/Pfizer vaccine portfolio for Alzheimer's. With bapineuzumab's demise, the perceived value of this program has never been lower.
4) Equity stakes in Alkermes and Proteostasis Therapeutics. Elan recently terminated their partnership with Proteostasis. 5) A 5-year, $10 million commitment to a Cambridge University collaboration on Alzheimer's/Parkinson's. From a succinct, market-valuation perspective: The 50% of Tysabri not already owned by Biogen-Idec.

Last winter, Elan announced that Kelly Martin's planned departure from the CEO role would be delayed until bapineuzumab's results were in. Now that they are, there is no mention of CEO succession. With Martin carving off almost all non-Tysabri assets that have any burn or perceived value attached to them, it is hard to avoid the conclusion that they are preparing to sell Elan. Biogen-Idec would not want those assets, they have a substantial neurodegeneration platform of their own. Neotope's pipeline would be, at best, redundant and a distraction from the main event, getting maximum value for demi-Tysabri. Biogen-Idec does not need to hurry to acquire Elan's half of Tysabri, particularly for what Elan thinks it is worth, because of BG-12. It is BG-12 which is the huge variable here, because even with the jcv test, we believe Tysabri is going to be a depreciating asset. Elan cannot shop Tysabri around to other buyers, since Biogen-Idec has preemptive buyout rights, and Elan has a lame duck CEO. Elan shareholders anticipating a large return on Tysabri may have to ratchet back their expectations significantly, because unless one expects that BG-12 will underperform and Tysabri will iexpand its market share, Elan is under more time pressure than is Biogen-Idec. Neotope will be spun off by the end of 2012, and we believe that Elan aspires to no longer exist as a independent entity by the end of 2013. This does not necessarily mean Biogen-Idec will cooperate.

 

Merck Serono Restructuring Update

from NeuroPerspective, September 2012

Despite the efforts of Geneva based staff to block layoffs ('Aux Barricades', NP June and NP July/August), Merck KGa is steaming ahead with their plan to close Serono's legacy site in Geneva. The loss of that R&D hub also means significant and unfortunate downsizing of the Merck Serono CNS agenda. The concept that will dictate their CNS strategy going forward is that any program retained in their portfolio must have have some potential relevance to Multiple Sclerosis, a core competency for Merck Serono. How this will be operationalized remains to be seen: Programs that, for example, address neuroinflammation or regeneration/remyelination in MS might eventually be developed for secondary indications wherein those pathophysiological features are also salient. But otherwise, such assets will be either spun out, sold, or simply terminated. The first CNS spin-out is that of a preclinical mGluR3 and 4 program in Parkinson's being run by Francois Conquet, formerly head of mGluR specialist Addex. Conquet will be CEO of this new company, Prexton Therapeutics. Merck Serono is providing $2.6 million in seed funding (they have allocated $30 million for such R&D spinouts). Merck Serono has two active CNS partnerships, with Domain Therapeutics (also mGluR4) and with Affectis (P2X7): These partnerships will have a future at Merck Serono only to the extent to which they can be framed as MS-relevant, which is more likely for the latter than the former.

 

Silver Lining for Sola?

from NeuroGram 8/25/12

The headlines for the Phase III topline results for Lilly's solanezumab were pretty much the same as for bapineuzemab: Primary endpoints were missed, and by that definition, the Phase III program was a failure. But there was a very ambiguous subplot which leaves the sola saga looking somewhat different from bapi, whose development as an IV drug was summarily terminated by JNJ/Pfizer. Lilly reported that a secondary analysis of pooled data for mild-moderate patienrs showed cognitive benefit, and a secondary analysis of mild patients showed that the reported benefit was confined to that group. However, a primary endpoint of cognition in mild AD patients had been implemented in the second Phase III trial, and the results there were negative. Lilly will not unveil details until October, which leaves some important questions yet-to-be-answered. For example, what were the actual p values for these analyses? If the pooled data for mild patients from both trials was significant at, say, p=.049, but the primary endpoint for cognition in the mild AD patients in the second trial 'failed' at p=.051, the inconsistency in 'significance' might just reflect the effect of having a smaller sample size to work with. This could be another instance where the binary concept of 'statistical significance' obscures rather than illuminates. Just as importantly, what was the magnitude of the reported cognitive benefit? From the little information provided thus far, it does not sound like it was accompanied by any measurable functional benefit, and perhaps these cognitive fluctuations are just statistical noise. But it does leave open the door for the hypothesis that in treating Alzheimer's, earlier is better, and that the differences between sola and bapi (dosing, epitope binding site) that had generally been interpreted as being in bapi's favor, perhaps worked for solanezumab.

One thing we do know for sure: A couple of Wall Street analysts who speculated that Lilly could potentially file a NDA on the basis of these data must be asked--'What Have You Been Smoking (WHYBS)?' Had Lilly hit a primary endpoint in one but not both of the trials, perhaps one could assemble a scenario for filing, but not here. Had solanezumab struck out as definitively as bapineuzemab did, even in the earliest-stage AD patient cohort, this would have been a heavy blow to the amyloid hypothesis. For now, that judgment hangs in abeyance, at least until a more complete disclosure of the results, and almost certainly beyond. The upcoming September issue of NeuroPerspective will feature our annual review of Alzheimer's, where we examine all of the recent developments in the AD field in context.

Viva EnVivo: EVP-6124 and Alzheimer's

from NeuroGram 7/19/12

NIR has been waiting for the results of two particular clinical trials to be publicly released, since these programs have achieved the kind of success that should start to catalyze a shift in sentiment about prospects for CNS drug development.

One of the two trials finally has had data presented. It is EnVivo Pharma's EVP-6124, a nicotinic alpha7 agonist. In this 409pt Phase IIb trial, the highest dose used (of three), 2.0mg once-daily (oral), hit both primary endpoints in this mild-moderate AD population. Some patients were already on a regimen including a cholinesterase inhibitor, giving the opportunity to assess how EVP-6124 might perform in a combination therapy context.

These Phase IIb results are the best Alzheimer's data we recall from a robustly-scaled efficacy trial. At 23 weeks of treatment, cognition was improved on the ADAS-Cog-13 (p=.0189) and clinical functioning on the CDR-SB was improved as well, (p=.0253). The effect size on the ADAS-Cog-13 was .39, considerably better than the effect size for currently approved AD drugs (.15-.28)--which means that the magnitude of clinical benefit considerably exceeds that afforded by current drugs. Secondary cognitive endpoints were reached, and with room to spare. Positive trends were seen on the MMSE and ADL's.

One finding was particularly provocative, albeit preliminary: "Patients on the 2.0 mg dose saw an improvement in cognition (not just maintenance of current cognitive function) over the 23 weeks of dosing and were still separating from the placebo group at the end of the trial, as assessed by the ADAS-Cog-13." At six months, the mean results for high dose EVP-6124 patients demonstrated cognitive functioning that exceeded baseline; only further testing will show to what degree that improvement continues over time. The fact that, in this trial, EVP-6124's impact on cognition outstripped the negative effect of the degenerative process begs a question: Does EVP-6124 have a disease-modifying effect? Longer-duration studies will be needed to ascertain this, but that possibility is suggested by the six-month data.

There were GI side effects in less than 10% of EVP-6124 patients, and overall the safety and tolerability of this compound appears good. This substantial Phase IIb trial  shows that this once-daily drug provides robust improvement on a wide variety of cognitive and other functions, and holds out the tantalizing possibility that the disease process  itself was being modified, progression slowed.

EVP-6124's results suggest that this is an alternative mechanism that may achieve some, and perhaps all, of the primary goals of Alzheimer's treatment. At the very least, there is compelling evidence of symptomatic improvement, and given the  value of extending a patient's capacity for independent functioning, that in itself is very important. If this drug also has some impact on the progression of the underlying disease, that would dramatically change the current calculus of Alzheimer's therapies-in-development.  Given that the current crop of putative disease-modifiers have any number of delivery, safety, efficacy, and production  issues, this is big news. With so much attention currently being  consumed by bapi and sola, and by the highly preliminary pilot data for Gammagard, EVP-6124 may not immediately receive the media attention it deserves , but it should not take too long for the Alzheimer's field to take notice--we know that Big Pharma has.

MidYear Status Report for CNS

from NeuroPerspective July/August 2012

• Best News
  We can't tell you. Not yet. Sorry. Really.
• Next Best News
  Successful depression trials, from Sunovion, Lundbeck/Takeda, Forest/Pierre Fabré
• Best First Half
  Lundbeck, whose patience and persistence (in conjunction with Takeda's) paid off with Lu A21004 in depression, and with Saegis' 5HT-6 drug, Lu AE58054, in Alzheimer's. Selincro (with BioTie) increasingly looks to have some clinically meaningful utility in alcoholism.
• Worst First Half
  Targacept, in the wake of TC-5214's Phase III failure, a data snafu, and the departure of their founding CEO. Bad things can happen to good companies, and Targacept will rebound, but this has been painful to watch.
• The Blockbuster Concept is Not Dead
  Biogen-Idec's BG-12
• Smart People Doing Dumb Things, Again
  Vanda licensing a NK-1 antagonist (LY686017) from Lilly. No one has been able to sort out NK-1 other than for chemotherapy-induced vomiting (Merck's aprepitant). But Vanda plans to bring the same magic to NK-1 that they have brought to melatonin agonism.
• Government Filling the Gap
  Not completely of course, but the EU's Innovative Medicines Initiative, which is part of their seventh 'Framework' Program, sets the stage for more efficient leveraging of resources in the service of expanding the precompetitive research space.
  The 'NIH Blueprint for Neurotherapeutics' represents a commendable step towards bridging the funding gap leading up to clinical testing, though it does not, nor does it seek to, address the precompetitive space issue. Perhaps ironically, its most important component is the way that it offers protection to commercial IP rights.
• Most Memorable Backpedal of the First Half
  "Bapineuzumab wasn't our idea. It came with Wyeth."
  --Pfizer executive in private conversation with NIR

And Now a Word From the Egyptian Military

from NeuroPerspective July/August 2012

May 15th, Russell 'Rusty' Katz, the head of the FDA's Neurology Division, caused something of a stir at a NIH sponsored Alzheimer's summit when he stated:

1. A single trial of a disease-modifying drug could suffice for approval in AD.
2. approval on the basis of a biomarker could occur so long as the biomarker is "reasonably likely to predict clinical benefit."

Given the costs and immense timeframes associated with Alzheimer's efficacy trials of the disease-modifying sort, both of these would be welcome evidence of flexibility and responsivity on the part of the FDA's Neurology Division.

On May 24th, an Advisory Panel for the Neurology Division voted 13-4 that the biomarker data for tafamidis in Transthyretin Familial Amyloid Polyneuropathy (TTR) would be sufficient for approval, provided confirmatory post-approval testing be conducted. NP noted on Neurogram: "This could be a huge precedent. Given the immense challenges, particularly the time required, associated with confirming impact on disease progression, the process of having to do so prior to drug approval mandates enormously expensive and slow Phase III testing. NIR believes strongly that changes in specific functional measures and biomarkers may prove to be equally useful predictors in other neurodegenerative disorders, including Alzheimer's and Parkinson's....the FDA's AC move in this direction is a sign of where the regulatory process must move eventually, and the sooner, the better." We added: " If they do not accept the AC's recommendation, that would be an indication that their public statements regarding improvements in the regulatory environment constitute insincere posturing. We will know soon enough."

The situation before the FDA was this: A test-case of a small, clearly identifiable population, with biomarkers likely to be clinically relevant. A devastating orphan disease, without effective treatment (other than liver transplant), with biomarker data that the Advisory Panel strongly endorsed as sufficient for conditional approval. One could not have devised a more ideal time for Katz, or Temple, or Hamburg, to deliver a message, to step up to the plate and show that Katz's encouraging words of May 15 were not just platitudes to anesthetize the audience-at-hand.

The FDA turned down Pfizer's application for tafamidis. Message delivered, circa 1999. So now we know. It is business as usual. The company has been instructed to run another clinical trial, which based on past experience, will take 2-3 years.

The FDA had the chance to break from tradition here, with all necessary justification afforded by its AC. One might wonder why the FDA bothers with Advisory Committees, since they are so willing to ignore their advice. It almost seems like the AC's purpose is to give cover to the Agency in going in the direction it wants to go, and if the AC recommends a different course, they disregard them.

Our suspicion is that the Agency was afraid of the very thing that we welcomed: A Precedent. For us it looked like an opportunity, to them, a slippery slope. They did not want the bar to suddenly drop across the board in favor of surrogate endpoints, and they decided it would be too much work to hash out just where the benchmark for consideration should be placed. For the FDA, it is always easier to say no; they can always fall back on the old standby rationalization that they are protecting the public from ineffective and/or unsafe therapeutics. In the 'gotcha!,' gaffe-obsessed world of Washington D.C., passivity is seen as a virtue, regardless of the costs incurred outside the Beltway.

TTR patients, and the public in general, are not safer or better protected due to this decision. The Agency's agenda was to insulate itself and its Standard Operating Procedure. We are once again reminded of a painful reality, that talk is cheap: Hearing Rusty Katz talk about regulatory reform is like listening to Egyptian generals commend the concept of democracy.

Triage Time at Targacept

from NeuroPerspective May 2012

It has been a dismal six months for Targacept, which had been such a sustained success story over time. The failure of all four Astra Zeneca-run TC-5214 Phase III trials in depression constituted the major blow, and it is impossible to tell, without a level of access to the raw data no one outside the company and their CRO will ever have, whether those failures were due to some trial design flaw on the part of AstraZeneca; trial execution problems on the part of the CRO or its clinical sites; or continued evidence of the risks attached to drug development in depression.

But Targacept also caused some unnecessary pain for itself. Seeking to extend the reach of their nicotinic platform beyond CNS, they ran pilot trials of TC-6987 in both diabetes and asthma. The diabetes trial came up empty, but the initial report on asthma as an adjunct to corticosteroid treatment was of success on both co-primary endpoints. But just a couple of weeks later, Targacept reported that, when a data analysis error was corrected, the trial hit on just one of the two endpoints. When this kind of error happens when a company is on a downturn, it feeds the murmur in the crowd---'do they know what they are doing?' After all, there are enough problems that cannot be foreseen, this one was eminently avoidable.

Having said that, the revised dataset deserves some clarification. In this hint-of-concept pilot study, Targacept did a one-tail test of significance set for p=.10. In other words, the presumption was that any change would go in one direction, and that a 90% likelihood that such a change was not a fluke would be sufficient evidence of a clinical signal. Of course, neither of these contingencies would be acceptable in a larger Phase IIb or III study, but this was a pilot study. The two endpoints consisted of changes in expiratory volume from baseline to just before the dosing on the 28th day of treatment, and two hours after dosing on that day. With the revision, there continued to be a pattern of improved respiration, but the revised results for the pre-dosing measure were significant at p=.142, rather than .09. It should be mentioned that the p value for the post-dosing measure improved almost the same amount, from .07 to .052.

In other words, the pre-dosing measure still showed improvement, but to a degree only 86% likely to be attributable to the drug, compared with 91% likelihood based on the previous analysis. This is not a binary shift from success to failure, it is a slight shift from 'very likely to be real' to being 'almost as likely.' And the magnitude of change post-dosing, which may perhaps speak to an acute rather than durable drug effect, was even more pronounced. What this means in terms of clinical meaningfulness--we will leave to the pulmonologists. But it is clear that the damage here has not been done to the prospects for TC-6987 as an asthma drug, whose promise is arguably enhanced by the revised analysis, but to Targacept's reputation for excellence in execution.

Now Targacept is sorting out its strategy for recovering its credibility and momentum. They continue to have immense resources and expertise on hand, but also have major clinical trial cost obligations, and thus are engaging in painful cost-cutting. Targacept is now looking at widening their scope of activity beyond CNS, both inhouse and via the acquisition of external assets for development, the latter being a step that Targacept has eschewed to this point. They remain fully and solely committed to the nicotinic cause, which has stirred some debate amongst its stakeholders. It does not appear that the Company will turn to inlicensing to diversify their CNS assets; their belief, which has yet to be confirmed or disconfirmed, is that the nicotinic repertoire they have developed is better than anything available externally. Perhaps it is, but given the inexpensive assets out there, and the pitfalls of predicting outcomes, it might be wished that Targacept felt a little less certain about this.

Amyvid and Clinical Irrelevance

from NeuroPerspective May 2012

The FDA finally gave marketing approval to Lilly's (via Avid) Amyvid/florbetapir radio-labeled imaging agent, for use in PET-scan identification of beta-amyloid plaque. Approval had been delayed due to concerns regarding interrater variability/reliability. But reliability is not, in our view, the biggest issue here. The overriding questions are validity and clinical utility.

The validity issue harkens back to the longrunning debate regarding beta-amyloid and Alzheimer's: Is it a cause, a result, or neither? And if it is a causal factor, in what form does it exert these pathological effects? The sentiment in the field has distinctly shifted to a focus on soluble, non-plaque forms of beta-amyloid as a target for intervention. Which begs the question--what do positive or negative plaque findings from a florbetapir-assisted PET-scan really mean? If a patient has significant cognitive decline, negative findings may reduce the likelihood that the cause is Alzheimer's per se, since even if it is soluble AB that 'caused' the deterioration, one would expect some buildup of insoluble AB as well. Yet the potential for false positives remains troubling, where individuals without cognitive decline have been found to be riddled with plaque post-mortem.

But setting those questions aside for the moment, what about clinical utility? Since one would/should not interpret the presence of plaque as indicating AD in the absence of cognitive decline (and one would hope, a clinician would not order a PET-scan on someone who is asymptomatic), the Amyvid-scan would be ostensibly used in patients wherein cognitive decline has already been identified. In which case, other (and cheaper) evaluative procedures are available to rule out other potential causes of decline (e.g. depressive pseudodementia). And if these other etiologies do not appear relevant, the question is then--How would the results of the PET-scan inform treatment? If plaque turns out salient, one has the option of using typical dementia drugs like Namenda and Aricept. If plaque is not seen, leaving open the possibility that soluble beta-amyloid is having a pathological effect, or that it is a vascular dementia, one again has the option of using typical dementia drugs like Namenda and Aricept.

Amyvid offers an expensive, out-of-pocket avenue to not informing treatment decisions. The wholesale price of Amyvid will be $1600, which we suspect, will boost the price of Amyvid-potentiated PET-scans towards the upper end of the $3000-5000 range often cited for PET-scans. Medicare will not, at least for now, cover these costs.

Well-off early adopters may embrace these scans as an additional diagnostic tool, which they are. But for most families, it will be less-than-convincing to be told: "We'd like to order a scan for your (mother, father, other elderly family member). It involves injecting a radioactive substance intravenously, and will cost you or them $5000. It will add some tentative clarity to the diagnosis, but it will not affect which of our very limited treatment choices will be used." 'It hurts less than a lumbar puncture' is not a marketing motto with a future.

As has been the case with other AD-relevant imaging agents, Amyvid can confirm target-engagement for amyloid-targeting therapeutics. It is still unclear whether reducing amyloid plaque is good, bad, or irrelevant. We certainly do not concur with Public Citizen's Sidney Wolfe, whose predictably histrionic comment was that Amyvid is "dangerous" and should not have been approved. That's silly and hyperbolic. It's not. At least for now, it's just superfluous.

Executive Functions Modulating Allosteric Fiscal Output: Astra Zeneca's Biggest Clinical Trial

from NeuroPerspective March 2012

NIR has become a keen observer of, and an occasional participant in, discussions regarding the path forward for Big Pharma in neurotherapeutics. Those who have advocated for the transition of Big Pharma to an outsourcing model, at least up to the point of late-stage drug development and commercialization, now have a test case on the grand scale to monitor: Astra Zeneca. The experiment springs from painful necessity: Forbes' Matthew Herper recently cited a cost of $12 billion for every new drug developed by Astra Zeneca over the past fifteen years, and while the methodology of that estimate might be debated, the conclusion is not: The status quo has not worked, and is not sustainable. Recognizing this, AstraZeneca has chosen a radical course, eliminating almost all of its inhouse neuroscience group, keeping just 40 to 50 R&D staff in the US and UK to maintain its iMED (Innovative Medicines) neuroscience programs on a near-virtual basis. The intention is to create a semi-autonomous (they will be authorized to establish licensing and collaborative relationships on their own, up to an undisclosed cost-threshold), entrepeneurial, biotech-type enterprise within Astra Zeneca. Other companies have developed small scale initiatives of this sort (e.g. Lilly's Chorus), but this is a daring move, on a scale not previously seen in the pharma industry.

We see this as a 'Frontal Lobe' model for Big Pharma R&D, wherein the Pharma core group organizes and integrates outsourced drug discovery and development, a corporate corollary to the role that the brain's executive functions serve in planning, coordinating, and monitoring the ongoing flow of neural activity.

Despite our longstanding conviction that small company nimbleness has been sorely absent from the Big Pharma model, our initial reaction was skeptical: Can 40-50 people effectively provide these functions for major endeavors in neuroscience? Would this small group be expected to function as generalists, and to carry out a spectrum of roles for which they might be ill-suited? Would the size of the inhouse staff component be correlated with systemic clout? And last but not least, would the three-year timeframe cited for many of AZ's iMED units mean that the neuroscience group would feel compelled to go with familiar approaches and indications, lest they fail to meet that timeframe, and be eventually guillotined in turn?

Only time will provide the answers to some of these questions, but our initial concerns were assuaged by a lengthy conversation with Menelas Pangalos PhD, former head of neuroscience for both Wyeth and Pfizer, now the EVP for AZ's Innovative Medicines programs. The emerging picture of the iMED in neuroscience is as follows:

First of all, the 40-50 person brain trust will not be expected to function in generalist fashion: Five to ten people within the group will take on the 'speed-reading' task of screening thousands of neuro-oriented programs for licensing opportunities, filtering candidates for selective referral to members of the group whose specialization is pertinent to those opportunities. The autonomy of this group is highly unusual, and AZ is aware that the large-company mindset characteristic of many Big Pharma employees will not be suitable for the Neuroscience iMED. Recent and future hiring has been and will be geared towards those who might otherwise have found a more comfortable fit in the biotech world. The group will have access to, and be able to choose between, the utilization of chemistry/biology/regulatory resources within and outside of AZ. Indeed, in contrast to the CNS shrinkage so pervasive in the industry of late, Pangalos expects that the range of programs will not shrink, as it might well have in response to anxiety-laden haste, but will in fact expand, even within psychiatry.

Perhaps most importantly, AZ is aware that a three-year timeframe will not be sufficient for quality POC programs to germinate in neuroscience. Indeed, just starting to wire together these new networks will take all of 2012. NIR has been arguing for a shift in pharma timelines (e.g. 'Kill the Quarterly Call'), and this restructuring makes such a shift more possible, albeit not inevitable. The same resources previously allotted by AZ to neuroscience will be available, but instead of being largely consumed by internal fixed costs, they can be more tactically and efficiently allocated. Additionally, and this is a vital element that has not been emphasized in public: Programs that are slow to develop will thus not steadily hemorrhage massive amounts of money, neuroscience expenditures will ebb and flow along with the pace of development in those programs.

Superimposing another neural metaphor on this business analysis: AZ's neuroscience spending will become allosteric, activity-dependent. This reduction in fixed costs will allow AZ to be more patient, not less, as is necessitated by the lengthy timelines inherent to the neuro area.

Whether a small, core group can effectively scan, aggregate, organize, and oversee a complex array of neurotherapeutics programs, while maintaining its systemic clout and access to resources within a huge corporate entity, remains to be established over time. But Astra Zeneca's radical restructuring and reconceptualization of neuro R&D is a bold experiment. By 2017, we should have some clarity as to whether it is working.

Identity Crisis

from NeuroPerspective February 2012

While Big Pharma corporate presentations tend to offer little that is new or persuasive in terms of overt informational content, what they provide as covert barometers of company attitudes and priorities can be unintentionally revealing. Over the past two or three years, it has not been unusual for Big Pharma CEOs to downplay CNS components in their public presentations of their Company and its prospects. After all, Big Pharma successes in CNS have been anything but common in recent years, and they naturally want to put their best organizational foot forward in public venues. At the most recent JP Morgan conclave, NIR had the opportunity to watch the current iteration of CEO/CFO-speak from most of the major pharmas, and a few of these corporate Rorschach tests were unusually evocative.

Our first metric in evaluating these presentations is a crude one: How long does it take for the CEO/CFO to acknowledge the existence of the brain? If the anatomical terrain were straightforwardly reflected in the emphasis given, one might sometimes think that the 2012 human body consists of a pancreas and assorted tumors. But this year, we encountered a more awkward phenomenon, a Big Pharma CEO who appears to wish that they worked someplace other than in the pharmaceutical industry: GSK' CFO Simon Bingemans, in his twenty-minute presentation, uttered neither the word 'drug' nor 'pharmaceuticals.' The closest that he came to mentioning this category, which constitutes 67% of GSK's business, were a few references to "vaccines." But otherwise, the entire presentation featured a detailed and not-compelling-in-the-least tour of GSK's new 'Financial Architecture', which offered repeated references to 'improved efficiencies,' debt load, and 'working capital.' Had we been blindfolded, there would have been virtually no clue as to what might constitute the contents of GSK's pipeline: Video games for emerging markets? Athletic shoes that light up in the dark? Cat food for the underserved 'pet segment'?

The other presentations were highly variable in their acknowledgement of CNS, but did tend to acknowledge that pharmaceuticals are pertinent to their core business. Sanofi's Chris Viehbacher loaded his presentation with a fiscal emphasis which could have applied to anything, but he also discussed diabetes, emerging markets, and the aforementioned "pet segment." Last year, Roche's Rolf Hunziker provided a stirring call-to-arms for the neuro area, declaring that "CNS is the new oncology." The new CFO, Alan Hippe, instead focused primarily on the topics of improved efficiency, personalized medicine, and biosimilars. He was not visibly enamored of CNS as the 'new oncology', he cited a fondness for the old oncology, due to its higher success rates. Pfizer emphasized its pharma focus, with a brief nod to the brain. Bristol Myers Squibb has already divested all non-pharma components, but CNS did not float into view during their talk. Lilly's John Lechleiter launched into a refreshingly full description of Lilly's pharmaceutical business, and noted CNS interests in the process; he even had the courage to spend a minute on solaneuzumab. The most focused and single-minded of all the talks was given by Astra Zeneca's CFO Simon Lowth, who provided a 20 minute infomercial about AZ's emphasis upon partnering going forward, and all the reasons that why one should partner with AZ, thankfully within the clearly defined context of being a drug company.

It's not that the absence of mention confirms a lack of pharma substance, companies may have CNS strengths but not emphasize them in public, just as others may overstate the quality of their pharma components and CNS capabilities. But it does say something about the filter through which these executives view and appraise the companies that they lead, and the spirit with which they inspire--or not--those companies. NIR holds to the old-fashioned view that pharmaceuticals are not just another type of widget, and that this belief system is necessary in order to transcend and master the particular challenges that are endemic to this calling. We hope that GSK's CFO had simply lost a couple pages from his presentation, hence the eye-widening chasm of omission. If not, he'd better set to the task of writing them, or hiring someone else to do so; this is not an endeavor where the rallying-cry is 'may the best spreadsheet win.'

Postscript: AstraZeneca's emphasis upon inlicensing now takes on a different meaning, with their announcement that they are essentially shutting down all neuroscience R&D, maintaining a skeleton crew of 40-50 scientists who will basically provide the interface between AZ and their outsourced CNS drug development program.

The Cycle Begins to Turn: 2012

from NeuroPerspective January 2012

Those readers who saw NIR's recent presentations at CNS Summit (Boca Raton) or Windhover TA (Boston) in recent weeks know that we have made the case for 2012 seeing the tidal cycle begin to shift back to one of credibility and ascendancy for the CNS therapeutics industry. We appreciate the concern expressed by some that a certain speaker's lithium level might have ebbed into the subtherapeutic range--but no, that's not the case. Certainly, this is not the viewpoint of those who focus on the macroeconomic climate, riven as it is with doubts as to whether the US Executive and Legislative Branches might ever work together for the common good (highly unlikely for the next year), and fears for the future of the Euro Zone and the smorgasbord of economies and attitudes awkwardly melded therein. Many will justify sitting on their venture wallets for the foreseeable future, given that taking risks now might be seen as foolhardy, albeit timely.

But it's not going to be about them in 2012. Just as clinical failures greased the skids on the way down, clinical successes will be the winch by which the sector extricates itself from the quicksand, trial by trial. There are a number of such clinical programs, with profiles of varying height, that will report data during 2012. First of all are the fab 'mab' twins, Bap and Sola. For a number of reasons, bad news for these highly expensive trials has already been largely built into expectations, and a show of hands of those expecting positive results primarily yields scattered waves from those who need success more than they expect it. Any hint of efficacy, even in a subpopulation, would now count as a pleasant surprise. But moving beyond that cohort, there are several other trials of promise. On the less speculative end, we expect CeNeRx's TriRima to show well in Phase IIb as a selective MAO-A inhibitor for TRD, that data coming during 2Q:12. Moving away from relative predictability, EnVivo will have Alzheimer's data for EVP-6124 during that same quarter. On the more risk-laden end of the spectrum, Allon Therapeutics will have davunetide data in Progressive Supranuclear Palsy late in 2012, and if positive, one might well speculate about single-trial regulatory consideration. Naurex will report Phase II data for GlyX-13 in depression by midyear. Phase III data in Fragile X could come from both Novartis and Seaside Therapeutics. Important clinical results will also come from Biogen-Idec/Knopp, Catalyst Pharmaceuticals, Lundbeck/Takeda, Targacept/AstraZeneca, and Acadia Pharmaceuticals.

It is not that all of the clinical events coming up in the next 12-18 months will pan out positively. After all, December was a remarkably cruel month, marked as it was by the failure of the second TC-5214 Phase III, and the less surprising failures in high-risk trials from Trophos (ALS) and Sygnis (stroke). These unfortunate dénouements serve as reminders that nothing can be taken for granted, particularly in CNS. But some of these programs will show clinical benefit, and the negative psychology beclouding the neurotherapeutics area will begin to clear. In 2012. Really.

 

2011

Ivy League Outsourcing

From NeuroPerspective November 2011

This roster would gladden the heart of the fiercest Tiger Mom: Harvard, Columbia, Johns Hopkins, U.Penn, UCSF, and Vanderbilt; they are just some of the academic institutions to which companies like Pfizer, AstraZeneca, JNJ, and Genentech are entrusting a slice from their shrunken research budgets. And there are the freestanding research institutes, like Salk Institute, Sanford-Burnham, and Gladstone Institute, which have attracted collaborations from JNJ, Lundbeck, and Sanofi. Much, albeit not all, of Big Pharma has turned to academic outsourcing with a fervor suited to new converts, the dilated-pupil certainty that academia will secure those elusive new molecular keys to the kingdom-and more cheaply. But this is outsourcing, and our guess is that some of the cultural mismatches that arise when Bangalore tries to imitate Tulsa are going to crop up here as well. The Culture of Industry fundamentally differs from the Culture of Academia: The security of tenure, the pace of academic endeavor, the goals of recognition via publication, the access to postdoc indentured servants, and the belief that not all knowledge must be applied; these are not going to coexist smoothly, at least not often, with the desperate neediness of an industry where no job is safe, and MBA's are beating the drum to which the oarsmen must hew their pace. Either Academia will forfeit the curiosity that is its strength, or Industry will grow impatient as the results that they demand do not arrive fast enough. Either way, our guess is that a couple of years from now, many of these joint ventures will begin to unravel, because underfunding academics will not work any better than squeezing one's own scientists, and neither of the parties will feel fulfilled by the compromises required by the clash of their respective cultures.

The Illusion of Certainty

From NeuroPerspective October 2011

More often than not these days, when Big Pharma peels back the veil from their licensing wishlists, one of the first criteria cited goes something like this; 'Program must address validated molecular target.' Seems perfectly reasonable, since knowing the target allows for the anticipation of potential issues, including but not restricted to adverse events and drug interactions. This provides a more rational foundation than a program that is driven by empirical observation, where a molecule is seen to exert an effect, albeit via a mechanism yet-to-be-defined. The problem is: The science has not yet caught up with this worthy aspiration. The only truly validated disease-modifying targets are those being successfully addressed in Multiple Sclerosis. The symptomatic targets that have been validated are largely those that were explicated after the fact, for drug classes that are now long-in-the-tooth. And as for novel, validated symptomatic targets, there are a couple which now appear to be establishing themselves in the area of schizophreniform cognition, such as nicotinic alpha 7 and the glycine transporter, but otherwise, it's a short list. The hope of quickening the process via biomarker validation is at present a mirage, because truly validated biomarkers are equally nonexistent. It requires a leap of biomarker faith or mechanistic ideology to deem a novel drug as addressing a 'validated molecular target.' And the less chic alternative, animal phenotypic validation, while in some ways benefiting from the face-validity of overt behavior, rests on its own bed of quicksand, the predictive uncertainties attendant to animal models for most disorders. That may be less irrational than predicating programs on mechanistic hypotheses, but it still falls short of the goal. Over the next few years, validated targets and biomarkers will emerge, confirmed by posthoc observations of the human phenotype in response to novel drugs. But to the degree to which the goal is held up as the entrance criterion, the pace at which this can be achieved will in fact be slowed.

Autism: Enhancing Social Relatedness

From NeuroPerspective's review of Autism therapeutics, September 2011

Oxytocin/Carbetocin As was noted earlier, oxytocin is a peptide which appears to be an important, endogenous upregulator of affiliative/social drive, and/or a downregulator of social anxiety. The great appeal is that this activity appears to target one of the core areas of deficit in ASD: the orientation towards, and tolerance of, human interaction. There have been a number of pilot studies indicating that oxytocin can be of benefit in autism; improving eye gaze, social awareness, and interpersonal responsivity. One imaging study indicated an increase in activity in the fusiform gyrus, wherein the recognition of human faces appears to be centered, perhaps increasing the valence of human vs. nonhuman object responsivity.

Some question whether oxytocin's effect reflects a specific impact upon affiliative 'circuits', as opposed to more general anxiolysis. However, the latter cannot be the whole story, because if it were, SSRIs and benzodiazepines, which are effective anxiolytics, would enhance affiliation/interaction in ASD patients, and they do not.

The plasma half-life of oxytocin is only three minutes, but it is believed that the duration of CNS activation considerably exceeds that, perhaps tenfold. It is possible that one only needs an initial, facilitative effect that would allow someone to enter into interaction, after which both behavioral therapy and the intrinsic reward of human interaction would hopefully sustain that behavioral change. A federally funded trial of intranasal oxytocin is underway, with the goal of enrolling 34 adults. Data is expected in late 2012. Oxytocin, which is used as Pitocin in obstetric settings, has safety issues with chronic use, its effect on the vasopressin 2 receptor can lead to hyponatremia.

To work around oxytocin's deficits as a therapeutic payload, other approaches have been explored. Carbetocin is a synthetic form of oxytocin that is used via IV to address peripartum hemorrhage outside the US. It has the advantage of a 40 minute peripheral halflife, which they hope will translate into extended CNS effect--the goal would be to dose twice during the day. Cypress Biosciences licensed an intranasal formulation of carbetocin from Marina Biotech (formerly Nastech), and had planned to go into a pilot study in autism during 1Q:11. Cypress' hostile takeover derailed those plans, and the carbetocin program is being outlicensed to a startup founded by Cypress' CSO, Kyalin Biosciences. Carbetocin has had considerable shortterm Phase I testing, now Kyalin is raising the funding needed to run a Phase 1b trial of carbetocin in healthy subjects, which would appraise the effect of carbetocin on social interaction measures, as well as assessing dosing and duration of effect. That trial would take 8-10 months, the next step would be a Phase IIa trial in autism. The key endpoints would be measures of the interactive behaviors (e.g. eye contact) which represent a core deficit of autism, one not currently addressed via pharmacotherapy.

There is another key question: How does oxytocin, which does not appear to have BBB access, regardless of its mode of administration, exert CNS effects? Small molecule oxytocin-receptor agonists have received some attention, Wyeth developed WAY267464, a promising first-generation prototype, but that program demised at Pfizer.

Besides targeting oxytocin receptors, alternative tactics could involve increasing the release of endogenous oxytocin. The melanocortin-4 system, which has been heavily studied in the area of feeding/obesity, has been shown (by work at Emory University) to be a route by which oxytocin release can be increased downstream. Of course, given its linkage to key metabolic processes, melanocortin-4 will have its own safety/tolerability issues as a target.

A New and Improved Target in ALS

From NeuroPerspective September 2011

A paper published in Nature may transform the world of research regarding therapeutics for ALS. The investigators (the lead is from Northwestern University) identified ubiquilin-2, a protein involved in protein degradation by the ubiquitin-proteasome system, as a key pathological variable--and hence therapeutic target--in both familial and sporadic forms of ALS. For many years, research into the treatment of ALS has, to a large degree, rested on an unproven extrapolation of findings from rare familial forms to the vast majority of cases that are nonfamilial. SOD1 mutations have been utilized extensively as trangenic models for the disease, while TDP-43 defects have in recent years emerged as a more prevalent and perhaps explanatory pathological culprit. But SOD1 and TDP-43 mutations together appear to explain no more than 25-30% of ALS cases. Now, ubiquilin-2 may have been identified as the missing link underlying ALS in all forms. Ubiquilin-2 has been found to be in protein aggregates from the spinal cords of ALS patients, and in the brains of ALS patients who also show signs of frontotemporal dementia. Ubiquilin-2 co-aggregates with TDP-43, but is even more prone to toxic aggregation than is TDP-43. It also appears to aggregate with several of the other factors believed to play a role in some cases of ALS, such as optineurin, FUS, and ubiquitin. However, these ubiquilin-2 aggregates did not display iimmunoreactivity to SOD1 antibodies, indicating that SOD1 is not a component of those aggregates.

If confirmed, this would mean that therapeutic candidates that have been vetted by their activity in SOD1 transgenic models have been assessed in an invalid model. NIR has discussed this as a potential Achilles Heel of ALS research for several years now, and it may be that much ALS research has been based on a foundation of sand. The question remains open, as it does for the role of protein aggregates like beta-amyloid and alpha-synuclein in Alzheimer's and Parkinson's, as to where this target stands in the causal sequence of the pathophysiology. But at the very least, a transgenic model that utilizes ubiquilin-2 mutations may offer a more valid preclinical screen for ALS therapeutics, and interventions that address the clearance of ubiquilin-2 linked protein aggregates could become a promising new avenue for treatment.

Erratum and Apology: Avanir

From NeuroPerspective July/August 2011

In the July/August issue of NeuroPerspective, we were critical of Avanir Pharmaceuticals and their marketing of Nuedexta for Pseudobulbar Affect (PBA) in conditions beyond the Multiple Sclerosis and ALS populations for which it was tested in Phase III. We were incorrect in doing so. In fact, in spite of the relatively narrow breadth of populations assessed in Phase III, the FDA did not restrict the Nuedexta label to those two indications. Their reasoning (which was more progressive and reasonable than we would have expected given the Neurology Division's history) was that the extrapolation of benefit to PBA in other disorders would be acceptable, that MS and ALS are sufficiently different in their pathophysiology that the impact on this symptom-in-common does not depend upon population-specific underpinnings. Which means that the FDA will not take umbrage at a broader marketing of the drug, so long as the target symptom is PBA. Our critique was thus in error; our apologies to Avanir and its CEO for having stated that they were placing the Company in jeopardy with the FDA--such is not the case.

No Free Lunch

From NeuroPerspective July/August 2011

In American politics these days, one of the fundamental debates has to do with the scale and reach of government, both in terms of what it takes in via taxation, and what it provides in terms of services/benefits. The uproar over taxation greatly exceeds the expressed willingness to cut services/benefits, and while this is a gross oversimplification, to some degree the conundrum comes down to this: People want their governmental services and benefits, but do not want to pay for them. The same can be said for the pharmaceutical industry, where there is no dearth of discourse over the decline of industry pipelines, and the need to replace revenues lost or soon-to-be-lost to generics. But there is a real disconnect between the cited need and the demonstrated willingness to invest in the next generation of drugs. As the abyss has yawned ever more near, the pharma and investment communities have in fact shrunken their level of fiscal engagement. In 2009, licensing upfront payments totaled $921 million, in 2008, $655 million. CNS partnering upfronts during 1H:11 (for drug development, not post hoc marketing or royalty agreements) have totaled just $27.3 million. Funding via stock offerings or VC rounds has totaled $257.1 million during the same period. This pace contrasts with the $821 million and $917 million provided during 2009 and 2010 respectively. While we expect activity to modestly improve in both categories over the second half of this year, the trend indicates a dramatic shortfall in both categories, at a time that the salience of pipeline deficits is increasing. This is almost bizarrely short-sighted and self-defeating.

Machiavelli's Bio Hall of Fame

From NeuroPerspective July/August 2011

Presuming that the Teva acquisition of Cephalon is eventually completed, after the FTC has made its usual inquiries in the service of looking like they are genuinely concerned about maintaining competition, it is worth noting the panoply of innovations and accomplishments that can be ascribed to Cephalon:

1. The Trojan Horse Orphan: Modafinil/Provigil was initially tested and approved for the treatment of narcolepsy, an orphan disorder comprising a patient population in the tens of thousands, with a market potential that likely would have topped out at couple hundred million dollars annually. This foot-in-the-door set the stage for additional innovations (see below) and an eventual billion-dollar plus annual sales pace.
2. 'Doctors Without Borders': No, we are not referring to the commendable organization which provides medical services in some of the world's most desolate and devastated quarters; we refer to Aggressive Off-Label Marketing to physicians of all stripes. Cephalon promoted off-label uses for Provigil, Actiq, and Gabitril, boosting sales even as they carried out clinical trials to assess the validity of the claims. In 2008, they paid a $425 million penalty, a mere speeding ticket compared to the revenues gained.
3. Excessive Daytime Sleepiness: Who knew this was a disorder? Cephalon was the first neuro company that we know of that took a set of disparate symptom syndromes and successfully reframed them as an overarching disorder.
4. Pay-for-delay: Cephalon wrote the book on how to delay generic competition by cutting deals with generic companies, providing legal savings and certainty in exchange for another couple years of marketing exclusivity.
5. Minimally Differentiated Substitution: In the neuro sector, Cephalon was the first company to substitute a chemical relative with little or no clinical advantage (unlike Adderall's replacement by Adderall XR) in the hope of staving off a generic challenge to the forebear. Nuvigil, Provigil's enantiomer, offers no substantive advantage other than its patent life. But once approved, Cephalon jacked up the price of Provigil in order to steer patients to the newer option. We suspect that, once modafinil goes generic, this ploy will be revealed as a failure.
6. Confronting the FDA. When Myotrophin was stymied by the FDA, Cephalon loudly complained, and refused to run another trial. Instead, they rallied masses of patients, prescribers, and pundits to flood Washington D.C. and turn Congress against the FDA...OK, that did not work out so well. Myotrophin is still 'approvable.' But it did provide a teaching moment, where other companies learned that publicly 'dissing' the FDA was, and is, not a prudent tactical choice.
7. Hedging Against Risk: Cephalon created a separate business entity which held Myotrophin as an asset, and when Myotrophin demised, it was the holders of that entity that were burned far more badly than were Cephalon shareholders.
8. Inlicensing as a cheaper route to building a pipeline: Provigil, Actiq, and Gabitril were all inlicensed, indeed Cephalon never developed a neuro NCE other than enantiomer progeny one step removed from their parents.

One unfortunate dynamic is that Cephalon triggered what might be thought of as an allergic reaction at the FDA, sensitizing the Agency to future uses of the very tactics that served Cephalon so well. The FDA now sniffs out Trojan Horses, demanding that trials be run with a range of likely clinical populations, and off-label marketing has diminished, along with the lavish 'educational opportunities' (shrimp, skiing, scuba) that had greased the way. Pay-for-delay is under fire. Molecular tweaking is still attempted, but the pricing power now possessed by generics makes this look like a losing battle. The inlicensing model has become increasingly popular, as more companies eschew the costs and risks of CNS drug R&D. This is not a benign outcome.

All in all, Cephalon is a successful company which got there via clever maneuvers and opportunism rather than innovative science. And they essentially abandoned neuroscience, seeing easier paths in oncology and inflammation. Towards the end, there was a glimmer of a renewed taste for adventure (Mesoblast), but its viability has yet to be established. Ironically, when we first contemplated writing this piece, it was planned to be a congratulatory note of appreciation for a job well done. But it did not turn out that way. It is a sad statement that Cephalon is one of the best examples of a successful neuroscience company, in spite of the fact that its legacy is one that only a confirmed cynic could look at with unabashed pride. Provigil was, and is, a good drug which has provided benefit to millions. Other than that, Cephalon's lasting contributions to the CNS field are far outweighed by the tarnished history of its strategic maneuvers and opportunism.

'Nicotinics'

From NI June 2011

In the iconic American film of 1967, The Graduate, Dustin Hoffman is given one word of career advice: "Plastics." Had a remake been done in 2010, the word-of-choice might well have been "Nicotinics." Nicotinic alpha 7 modulation has become the mechanism which, at least for now, is at the head of the pack seeking a treatment for cognitive dysfunction associated with schizophrenia. What makes it all the more intriguing for longterm followers of the CNS world, is that there are two well-funded, quality-science companies who are neck-and-neck in their head-to-head competition to get there first. During May, the pairing took shape as EnVivo Pharmaceuticals reported positive results from their Phase IIb trial of EVP-6124 in schizophrenia: Statistically significant or positive trends were seen on a range of endpoints, including the MATRICS battery. At the same time, AstraZeneca decided to not exercise their option for Targacept's TC-5619, which had performed well--albeit with some dosing questions left unanswered--in its own Phase IIb. Our impression is that AZ did not want to spend $30 million for a program which may not be fully Phase III ready, and our guess is that they may come to rue that decision.

At this point, Targacept has buttressed its financial position with a $70 million secondary stock offering, while EnVivo has Fidelity as its sole owner. Which means that we may see these two relatively small companies enter pivotal testing in the same timeframe, with high-quality molecules, for a disorder which constitutes a huge 'unmet need.' There are other companies in the nicotinic alpha7 space, including Abbott, Roche, and BMS, but these are the two leading contenders at present. The word 'fun' has seldom come to mind in watching the neurotherapeutics area of late, but watching this competition will indeed be fun.

Ecosystem In Jeopardy

Excerpt from CNS 2011: Therapeutics and Licensing

If there is a pattern emerging for 2011, it is the 'donut hole' between discovery stage and Phase IIb. Partnerships are burgeoning for discovery stage collaborations, many of them with academic researchers. These require very little upfront money, featuring research support in exchange for the option to buy in (cheaply) later. On the other side of the divide, there are the programs with Phase IIb data, where the hint of concept has turned to proof, the dosing is relatively clear, and while the expense of Phase III lies ahead, much derisking has taken place. In a pharma world hungry for commercial candidates, these programs are very popular. But there is the growing body of the preterite, the unloved programs that had the chutzpah to move into early clinical trials, thinking that a pilot study, yielding Phase IIa HOC, would earn them a partnership. Not anymore, or at least, far less often. They are too far along, with too much invested, to outlicense their prized asset for nothing upfront, but they have not yet been derisked, and do not have the resources to carry out a decently-scaled Phase IIa, let alone Phase IIb. VCs won't touch them, they rarely qualify for grant money, and never at the scale they need. Reviewing these opportunities, large and midsize pharma companies who do have resources see them as missing their sweet spots of cheapness and/or risk-mitigation. A few are fortunate enough to be adopted by a New England patriarch of the financial industry, or a wealthy recluse on the Isle of Man. But the group in the donut hole is literally starving to death, slowly and painfully.

By the time the industry realizes this, their fearful negligence will have taken a serious toll on the next generation of drugs that they need, and it will be too late for many of them. Big Pharma will bewail the sorry state of their pipelines, while turning to emerging markets where they can sell the products of yesteryear.

There are some in the industry who realize that the pharma drug development business, CNS and otherwise, is an ecosystem. Particularly with the downsizing and outsourcing that has roiled Big Pharma, the process of moving from discovery to commercialization now involves separate but interdependent parts, and the collapse of any component along the chain jeopardizes everything: If one component withers, the whole system is at risk. But some of the players do not seem to understand that. We have heard licensing professionals from the largest of Big Pharma chortle about being able to squeeze small companies into accepting small upfront payments. When small companies are nickel-and-dimed like that, they become indentured servants of a sort, bound to their licensee, and to VCs who have no exit. This cuts off the flow of capital to other small companies.

If one wants to build a culture of hostile dependency, this is a surefire route, and for those large companies who would like to have the same level of control over their partners that they had over their employees (and without having to fund their pensions), this is one way to do so. But they should be careful of what they wish for, because as a route to boosting creativity and productivity, this does not just fall short, it does quite the opposite. As a Novartis Business Development professional aptly noted in a recent meeting, shortchanging small companies jeopardizes the sustainability of the entire field. When the top of the food chain systematically imposes a starvation regimen on the lower strata of the food chain, in the long run, no one gets what they need.

Teva as White Knight

Online comment May 2011

While our preferred outcome would have been for Cephalon remaining independent and entrepeneurial, its acquisition by Teva Pharmaceuticals for $6.8 billion is infinitely preferable to the attempted takeout by Valeant Pharmaceuticals, who had offered $5.7 billion. Valeant would have stripped all R&D from Cephalon, packaged it for shortterm earnings, and positioned itself for a buyout. Teva, while heavily oriented towards generic pharmaceuticals, does have a branded CNS drug business, featuring Copaxone, and we expect that that much (probably not all, the Mesoblast cell therapy partnership might be too much of a stretch for Teva) of Cephalon's drug development portfolio will survive. Compared to the scorched-earth dismantling that Valeant would have imposed, this is a benign dénouement.

AstraZeneca Opts Out of Targacept's TC-5619

Online comment 5/2/11

Targacept had recently hinted that AZ would not exercise their option for the nicotinic alpha7 modulator TC-5619, and today, AZ confirmed that. This is in the wake of the January PhII data in schizophrenia, about which NI had made this comment in the February issue:

News from Alphaville

Targacept's PhIIb trial of TC-5619 in schizophrenia hit its primary endpoint, a computerized measure of frontal-lobe mediated problem solving/executive function (Groton Mazes). Perhaps more importantly in the long run, it also produced positive signals on global functioning and negative symptoms, as assessed both by a clinician and by patients themselves. It did not produce a signal on the overall cognition battery used (CogState), and Targacept has not yet completed a full analysis of the other six cognitive subdomains tested to see if changes occurred there (parenthetically, it is unlikely that the higher-order thinking required in mazes could improve if there were no changes whatsoever in lower-order domains like attention, memory, and/or processing speed, to name but three). It will be reassuring if positive trends are found in some other CogState subtests. The initial data summary raised a number of intriguing subplots and questions:

1. The first question emerged from the use of the pre-specified hurdle as bettering p=.10 in a one-tailed test, which is more easily reached than the typical two-tail, p=.05 criterion. This is a benchmark used by some pharma companies internally in making go/no-go decisions on furthering development of a program, and the fact is that the Mazes score came very close to p=.05 here.
2. The greatest divergence between drug/placebo came at the four week mark, when the lowest dose (1mg) had been utilized. This begs the question of whether, when it comes to dosing the nicotinic alpha 7 receptor, 'less is more.' Given that these were not independent groups, but rather, the same patients gradually escalating in dosing from 1 to 5 to 25mg (and thus perhaps experiencing some changes in receptor density or sensitization), that question is not definitively answered. Targacept believes that they are in the right dosing 'ballpark;' our suspicion is that ultimately, optimal dosing will be towards the lower end of the range--and the FDA may want them to prove they have identified the minimum effective dose.
3. The most counterintuitive finding was that non-smokers did not show benefit from TC-5619, only the 46% of patients who were smokers improved with the drug. Many observers had expected the opposite, thinking that smoking might obscure drug effects, perhaps due to nicotine saturation of receptor sites. Instead, the results leave open the question of whether smoking sensitizes receptors and allows greater response to alpha 7 binding, or if receptor density is different between the two populations. Targacept worked at having a schizophrenia trial with just 46% smokers (likely by including an Indian patient subgroup more likely to be nonsmokers) and in retrospect, this made achieving an overall positive signal all the more impressive, in that half the patient population consisted of a group that turned out to be nonresponders at this dose-range. With 75-80% of schizophrenics being smokers, this puts the vast majority of schizophrenics into the potential responder population, and powering the next trial will be made somewhat easier, simply by virtue of the proportion of smokers in the general schizophrenia population (offset by the need to hit a more exacting statistical hurdle).

The bottom line is that this TC-5619 trial provided something between hint of concept and proof of concept. There are some questions about breadth of cognitive effect and optimal dosing, but the drug appears to be very safe and well-tolerated. It is possible that the next trial could be powered to be a potentially pivotal Phase IIb/III, even with a dose-ranging component. The next step is for AstraZeneca to decide--they have through 1H:11--if they will pay the required $30 million and exercise their option on TC-5619. If they are at all serious about staying in psychiatric drug development, they will. Perhaps AZ will be influenced by the results--early 2Q-- from EnVivo Pharmaceuticals' Phase IIb schizophrenia trial of EVP-6124. Replicating success and safety there might provide enough additional validation of the mechanism's promise to seal the decision for AZ. Targacept has the resources and motivation to take the program ahead if need be, but we do not expect this will be necessary.

The nicotinic alpha 7 story, from Targacept and perhaps from EnVivo as well, has the potential to be one of the most important CNS therapeutics stories of the year, perhaps shifting the psychology away from what has become an atmosphere of learned helplessness.

NI had spoken to one of AZ's licensing heads last week (not about this specifically) and he had noted that AZ had terminated all inhouse psychiatry R&D, but was still open to licensing in sz and depression. I suspect that they felt that for TC-5619 to have fallen short of POC in a large Phase IIb, with substantial questions about dosing, means that the program will take more work to advance than they are willing to allocate. It is going to be difficult, if not impossible, to find psychiatry programs with that level of maturation and certainty.

Pillage

from NI April 2011

Being asked out by Valeant Pharmaceuticals is like having Dexter ask to friend you on Facebook: Be afraid, be very afraid. Having disemboweled Biovail, Valeant has now trained its sights on Cephalon, the long-ago bellwether of the nascent neurotherapeutics sector. Having been rebuffed in their initial 'friendly' overtures, Valeant has now launched a hostile takeover bid, offering $73 per share; $5.7 billion.

Cephalon's determination to remain independent is unlikely to be as resilient as it would have been under the late Frank Baldino. There is also some analyst disagreement as to whether another salvo at a higher price is likely. One argument against Valeant escalating its bidding (or having someone else compete) is the fact that Cephalon's longrunning hope of supplanting Provigil's $1.2 billion sales pace with Nuvigil ($240 million sales pace) is doomed to failure, leaving them vulnerable in 2012, when Provigil's exclusivity expires.

Beyond Cephalon's valuation is the question of what value would be added for the pharma industry, and CNS sector, if this eventuates. The answer is: None. Cephalon had excised its own CNS R&D over time, save for label-expansion programs. But the recent acquisition of Mesoblast, whose cell therapy portfolio includes CNS components, has signaled a willingness to invest once again in high-risk, high-reward scenarios. That kind of forward thinking is incompatible with Valeant's commercial agenda, wherein Valeant strips all extraneous (and creative) elements, in the hope that once sufficiently bloated, they will in turn be taken out by a larger company desperate for a quick fix. This is a cynical strategy, and we can only hope that it fails.

Captain Kirk's Enterprise

from NI March 2011

Randal J. Kirk is the master salesman of pharma, ever capable of convincing companies that they should buy something mediocre, because otherwise, someone else might get it. Back in 2005, Kirk convinced Shire, panicking because of Adderall XR's shrinking patent life, that if they didn't overpay to partner New River's Vyvanse, a drug offering marginal and overstated advantages, someone else would. New River also managed to avoid the head-to-head comparison with Adderall XR that would have been logical and informative. In 2007, he convinced Shire to buy the rest of the package, New River itself, for $2.6 billion (Kirk receiving 46%).

Now, he has convinced Forest Laboratories that they should acquire Clinical Data. Forest is undoubtedly anxious about 2012 and the end of Lexapro's patent protection, and bought the concept that they would not want someone else to get hold of vilazodone/Viibryd. Even though it is a drug with marginal advantages at best, never put through a Phase III with an active comparator, which would have been a logical and informative study. Forest paid $928.6 million net (Kirk receiving around 52%), with some contingent rights that will never come into play. Some observers think someone finally got the best of Kirk. That seems very unlikely to us. Even at this price, Forest overpaid.

Subtraction Through Addition

from NI March 2011

It wasn't a question of whether, but when. Sanofi-Aventis finally bagged Genzyme, for $20.1 billion and some potential added value via contingent rights attached to certain milestones. In theory, those rights ($14/share) could add up to almost another $4 billion. However, since most of that ($10 per share) is attached to Lemtrada/Campath MS sales goals that will likely never be reached, the final bill to Sanofi-Aventis will total US$21.2 billion. Sanofi-Aventis is hoping that adding Genzyme to a company structure that has never fully integrated its two namesake components, gives it the heft in biologics that will prevent stagnation in the next decade. But none of the Pharma mega-mergers has produced anywhere near the bonanza of lean productivity promised. Even the Roche acquisition of Genentech, which sought to preserve the cultural differences that made each company successful, receives mixed reviews. Someday, someone wiser than NI will analyze these huge pharma mergers/acquisitions, and will quantify just what was lost forever in the midst of all the additions, the sum ending up less than the total of its parts. To us, $20.1-21.2 billion could have been used in many other ways, many if not most of which would have provided S-A with a far wider array of options and pipeline assets.

The First Commandment of M&A: A merger or acquisition must provide more innovative CNS programs with the opportunity to show that they work--or that they do not--in human testing. Sanofi-Aventis had already cut back in CNS, and Genzyme has never made more than minuscule efforts in the area. This acquisition likely will reduce CNS program resources. Thus it fails to meet the litmus test of the First Commandment for M&A.

The Second Commandment of M&A: There is no Second Commandment.

Some may ask--Why should the enhancement of CNS drug development be considered the benchmark for all pharma M&A activity? Doesn't that reflect some kind of neurocentric tunnelvision? We would argue that a Big Pharma that avoids the challenges presented by the largest category of 'unmet need' in the pharmaceutical world is guaranteeing that they will not be a full-fledged player. It is like a car company failing to invest in an alternative fuels strategy. What may seem more profitable in the nearterm shirks both corporate and societal responsibility.

2010

Accelerating the Cycle: Returning CNS to the Spotlight

from NI December 2010

It is part of a recurring cycle, but knowing that does not ease its sting: CNS has fallen dramatically out-of-favor in the drug development world, losing much of its cachet and clout in the process. Because of the common perception that CNS is more difficult and risk-laden than other therapeutic areas, many larger companies have sought to redirect a shrunken resource pool away from CNS, with oncology and diabetes for the moment seen as more likely to produce an acceptable ROI. Lip service has been paid to the concept that small companies can work more efficiently, and thus should be entrusted with R&D formerly done inhouse by Big Pharma. However, outsourcing must be accompanied by resources, and in sharp contrast to what might thus be hoped for, upfront deal payments received by small CNS companies are down 70% YTD from 2009. While this is just another example of cyclicity, it makes life for all working in the CNS area more problematic./p>

But even if the flow of resources were to be reinstated to some degree, the neurotherapeutics area cannot afford to go back to its usual modus operandi: It is too redundant, expensive, and for the most part, has failed to generate truly valuable new CNS drugs. Here are some suggestions for accelerating the cyclical rotation of neurotherapeutics back to its place at the forefront of drug development; they are intended to spur discussion, comments and counterpoints are welcome.

Download PDF: Accelerating the Cycle

Prague Spring Turns to Nuclear Winter at Biovail

from NI December 2010

It is admittedly hyperbolic, but the wrecking ball Valeant has taken to Biovail's nascent CNS strategy brought to mind the memory of 1968's 'Prague Spring,' when seven months of daring democratization in Czechoslovakia was eventually crushed by the iron fist of the Soviet Union, which intervened to reinstate the old order. Even these names seem quaint now, but the concept is not: A radical and brave experiment in change, terminated by an anachronism unable to fully grasp the nature of what they seek to extinguish. Previously, Alexza Pharmaceuticals' AZ-004, Acadia Pharmaceuticals' pimavanserin, and Santhera's fipamezole had been returned during the first wave of terminations. The two latest casualties of the purge are Cortex's Respiratory Depression program and MedGenesis' GDNF/Parkinson's project. Now, only one inlicensed program has yet to be expunged. This is the istradefylline program licensed from Kyowa, which survives only because no further clinical development work is planned: Either the FDA accepts Japanese Phase III data (unlikely), or they do not. Valeant is laying off much of Biovail's staff, the entire Business Development group responsible for Biovail's CNS resurgence has been sent into exile. It could not be a more clearcut or complete repudiation of CNS as a development focus. The purge came even as Valeant announced a massive share repurchase program. This is a return to the most craven instincts to be found in the darkest recesses of the pharmaceutical industry, chillingly reminiscent of how the worst extremes of the 'Tea Party' movement evoke the most primitivist elements of the American psyche. Biovail had launched a valiant and potentially valuable stab at recreating itself and a new model for targeted programmatic growth in CNS. They seemed well on the way--until Valeant sent in the tanks. Now, while Valeant management congratulates itself on its regression to and below the mean, we can only wonder what might have been.

Setback for AstraZeneca and Targacept

from NI November 2010

AZD1446, the second alpha4beta2 nicotinic modulator to reach the clinic from the 2005 partnership between AstraZeneca and Targacept, failed to hit its primary endpoint in a 79pt adult ADHD trial, on the Connors Rating Scale. There were some significant and near-significant results on a few secondary measures, but only for non-nicotine users, and no Bonferroni correction for multiple comparisons was utilized, thus there was an increased risk of a 'false positive' finding. AstraZeneca is going to drop the ADHD development program. This is a disappointment for Targacept, who had pushed for the addition of ADHD to the roster of indications being assessed. The compound in theory remains alive as an Alzheimer's drug, and Targacept emphasized these selective improvements as indicating the potential for cognitive benefit in dementia. However, the most functionally relevant endpoint, delayed recall for a memorized list, came in at a nonadjusted p value of .086, and one must wonder whether this is enough to sustain the AD indication. Those doubts are reinforced by negative findings from a small trial in Alzheimer's, where AZD1446 was used as an adjunct to Aricept. Targacept noted that the four week duration of the trial was too short to be definitive regarding AD cognition effects, but they had cited limited procognitive findings on the ADHD trial to support the AD concept, and that trial was only two weeks in duration.

Given the failure of the first partnered alpha4beta2 compound in an Alzheimer's trial, it is difficult to muster optimism that AZD1446 will survive in Alzheimer's either, although AstraZeneca may wait for the results of an AZD1446/Aricept trial using EEG outcome measures, due early in 2011, before making a final call. Nicotinic alpha7 targeting is looming as a more attractive alternative, and perhaps AZ will exercise their option on Targacept's TC-5619. This also highlights the crucial importance that the depression adjunct program, once considered an afterthought, has now assumed for Targacept and their relationship with AZ.

It's Déja Vu All Over Again

Alexza Pharmaceuticals received a Complete Response Letter from the FDA for AZ-004 as an inhalable option for psychotic agitation in the context of schizophrenia or bipolar disorder. One might consider a broadened indication, agitation in the context of regulatory blind-siding, because both the process and content of that CRL were mystifying. Four areas of concern were noted, three of which, related to manufacturing, stability, and 'human factors,' have either been largely addressed or can be, with alacrity. But the fourth issue that the FDA raised, pertaining to safety in individuals with pulmonary dysfunction due to COPD or asthma, was and is the primary focus here. What is strange is that Alexza states definitively that, during the long process of NDA review, the FDA had forwarded scores of questions for clarification, none of which pertained to pulmonary safety concerns. The first that they heard of this issue was in the CRL. It is a reminder that, when it comes to the FDA, it is often true that "What You Don't See Is What You Get."

The content of the concern was also somewhat baffling, because Alexza had done three safety studies, two in COPD/asthma patients, one in healthy adults. Minor shifts in respiratory function seen in the patient groups were resolved easily, either by use of their usual bronchodilator, or were 'self-limiting', they simply went away. The interesting element to us is that these pulmonary-challenged patients were taken off their usual bronchodilation regimen for this safety testing, thus one would expect, a low threshold of reactivity to an inhaled powder, whatever its origin, given the cessation of their (generally) daily use of albuterol, which they use to attenuate such symptoms. The fact that 78% of the schizophrenic/bipolar patients tested were smokers speaks to the likelihood that AZ-004 is the least of their pulmonary problems. The fact that returning to their standard regimen remediated these minor pulmonary symptoms leads to one bottom-line conclusion: If these patients simply utilize their standard pulmonary regimen, it successfully addresses the adverse event, just as it addresses other pulmonary 'disruptors.' In the 800 individuals who have received AZ-004, there has been no evidence of any persisting negative effect on lung function.

Whether or not the process or content of the concern are justified, the CRL raises some important questions for Alexza. Is this a harbinger of continued obstruction of the regulatory path forward for any Staccato-delivered product? How much does it set back the development timetable for other Staccato programs, like AZ-007? As for Valeant/Biovail, the FDA response shuttled AZ-004 from a safe position in their portfolio, ready for marketing, to one where they decided the uncertainty was not worth their time and energy. They have returned the program to Alexza, and we suspect FDA clarification will be necessary for anyone else to step up.

Our reference to déja vu? Alexza's CEO was CEO of Anesta back in 1997, when the FDA overrode its Advisory Committee and refused to approve Anesta's Actiq, the fentanyl 'lollipop.' It took a year of going through the trials and other data already acquired, without doing new trials, but the FDA finally approved Actiq, in late 1998. Anesta was acquired by Cephalon, and Actiq then went on to be a major driver for Cephalon's success. Nothing that we have heard from the CRL changes our belief that AZ-004 is going to be a useful product in psychiatry/ER settings. Whether the FDA might seek to first limit marketing to inpatient/ER contexts remains a possibility, although we see no compelling safety reason for doing so. Alexza's management has rare experience in getting the FDA to reconsider and shift course--our suspicion is that they will achieve that for AZ-004 as well--sometime during 2011.

**Addendum (10/18/10): One afterthought regarding the safety concerns regarding asthmatic/COPD patients receiving AZ-004: The real risk-benefit question is whether treatment with AZ-004 presents greater pulmonary risk than does escalation into a state of agitation sufficient to warrant involuntary IM injection of antipsychotic medication--which requires physical 'takedown' and restraint. Given that physical exertion can trigger an asthma attack, preventing such escalation might actually avoid more serious pulmonary compromise. NI is not prepared to conduct a search of hospital incident reports regarding asthma/COPD adverse events during such restraint situations, but we suspect that such data could provide empirical support for this hypothesis.

The September 2010 issue of NeuroInvestment reviewed Alzheimer's: "This is an incredibly well done issue. Perhaps the best researched and most articulate synthesis of the current state of the field that I have seen. "

--Murali Doraiswamy MD; Head, Biological Psychiatry; Duke University

Alzheimer's Excerpt: Secretase Inhibition

from NI September 2010

Inhibiting beta-secretase and/or gamma-secretase, thereby reducing abnormal APP cleavage, in theory should lower the amount of AB produced. One can try to inhibit either secretase, or indirectly do so by upregulating the production of the alpha-secretase enzyme, since presumably this would allow more of the substrate APP to be normally processed, less of it diverted to pathological forms. BACE and gamma-secretase inhibition have turned out to be easier said than done, particularly in terms of ensuring that normal and necessary activities are not unhealthily circumscribed. Our suspicion is that, when the Alzheimer's story is fully written, that the secretase strategies will be highly expensive footnotes, describing therapeutic dead-ends.

Gamma-Secretase

Gamma-secretase cleavage produces Notch protein, and without Notch, normal cell growth and differentiation of the murine immune system and intestinal tract becomes fatally impaired The Notch issue, where gamma-secretase inhibition appeared to risk unacceptable adverse events, initially seemed to doom the gamma-secretase strategy. Alternate gamma-secretase strategies were developed which avoided interference with Notch production; those compounds which alter the gamma-secretase substrate in order to specifically reduce AB42 production (without impacting Notch) are referred to as gamma secretase 'modulators'.

Gamma-secretase targets have been widely-pursued, reported to both reduce AB42 production and aggregation. Some NSAIDs were reported to do so, which spurred Myriad Genetics to develop an isomer of the NSAID flurbiprofen ('Flurizan') for AD. On a failed 207 pt Phase II trial, a subgroup with milder AD, receiving a high dose, showed improvement on the ADAS-cog during the 12-18 month open-label extension. The data, from a cohort of just 35 patients, was skewed by the fact that more than half of the extension patients discontinued. Myriad claimed that at 24 months there was a dramatic difference for Flurizan patients, proclaiming that Flurizan "halts" Alzheimer's "in its tracks." The $150 million Phase III was scaled to prove or disprove that point, and it was Flurizan that was halted in its tracks, the trial was a complete failure.

Lilly's LY450139/semagacestat reached Phase III in spite of the tiny window between its effect on beta-amyloid production and Notch-inhibition. Indeed, one PhII trial reported 14% of patients experiencing rashes, while 14% had hair color changes. NI had previously projected that 'These toxicity flags herald a drug that will never be commercialized', and indeed, Lilly had to terminate the program in August; interim Phase III data showed that semagacestat patients not only showed a higher incidence of skin cancer, but actually performed more poorly on measures of cognition and ADLs. Pfizer is in Phase I with Wyeth's begacestat/GSI-953, which along with a Phase I compound from Elan (ELND006) and Bristol Myers-Squibb's BMS-708163, reportedly have better separation between their gamma-secretase action and Notch inhibition. The same claims are made for the Phase I programs from Chiesi, and Eisai, as well as the preclinical programs from EnVivo Pharmaceuticals (EVP-0962), Archer Pharmaceuticals, and Neurogenetic Pharmaceuticals. Galapagos has a novel gamma-secretase substrate target, GPR3, which they hope will avoid these problems.

While parsing Notch activity may offer better safety, the harm semagacestat did to cognition/function raises warning flags for all of these programs. Contrarian voices have been lent credence by the semagacestat cognition results: A Tel Aviv University group published a report that beta-amyloid is an essential contributor to synaptic transmission, particularly in the hippocampus. A Harvard researcher had made the case for insufficient gamma-secretase as a key causal element in Alzheimer's. She knocked out nicastrin (an element of gamma-secretase) in adult mice, and even without any amyloid features, the mice showed cognitive decline. Similar work knocking out presenilins, which are other constitutents of gamma-secretase, also produced in vitro evidence of cellular dysfunction. A UCSD researcher published ( in PNAS) a warning that secretase-targeting drugs may worsen Alzheimer's, by creating peptide chains that themselves form new ion channels, whereby neurotoxic calcium influx increases.

As is the case for BACE-inhibitors, companies who have devoted resources to gamma-secretase inhibition must now question the utility of this approach. This is another example of an animal model whose a priori design was hypothesis-driven, wherein the production of a dementia-like destination does not necessarily mean that the route is the same as is taken in human AD.

Beta-secretase

Beta-secretase (BACE) inhibition is a popular route now beclouded with doubts about both its safety and relevance. Papers published in Science and Nature Neuroscience reported that beta-secretase, in conjunction with neuregulin-1, plays an essential role in myelination. An Amgen-sponsored group has also found that knocking out BACE1 led to impaired synaptic plasticity and memory, and another group has published work showing BACE1 inhibition undermines presynaptic function. This raised the question of whether BACE inhibition might have unacceptable effects on normal function.

As will be discussed at length, two groups have now produced data that puts into question the essence of the beta-secretase strategy. They suggest that beta-secretase is a major source of beta-amyloid only in the rare Swedish mutation familial variant. Any company developing a beta-secretase inhibitor should by now be experiencing deep existential qualms regarding the validity of the premise. There have been BACE inhibition programs at Amgen, Evotec, Lilly, Merck, Actelion, Takeda, and Elan, among others. Merck's candidate reportedly reduces amyloid levels in a non-human primate model, but only reaches sufficient plasma levels when administered along with a hepatic enzyme (CYP3A4) inhibitor. CoMentis' CTS-21666 was acquired via CoMentis/Athenagen's merger with Zapaq. The compound was said to be selective for BACE2, which could mean the myelination concerns raised by some regarding BACE inhibition would not apply. This beta-secretase inhibitor elicited a deal ($80 million in cash upfront, $20 million equity buy, US co-promotion rights) with Astellas early in 2008. Phase IIa was expected to start in 2009, but did not, we wonder if the compound has run into problems. Archer Pharmaceuticals' ARC-050 is a BACE inhibitor, currently in preclinical testing. Galapagos and a Belgian academic team identified a GPCR (GPR3) whose overexpression leads to excessive AB production, likely via heightened beta-secretase activity. GPR3 appears to not have any 'normal' function, at least so far as can be determined via knockout models. However, Galapagos notes that elevated GPR3 levels are not seen in all Alzheimer's patients, thus this would be pertinent to just a subgroup. Galapagos is now screening small molecule inhibitor candidates. Boehringer Ingelheim has partnered Vitae's preclinical BACE program.

Staredown in San Diego

from NI September 2010

Cypress Biosciences is facing a confrontation with Ramius, a minority (9.9%) shareholder. Ramius is unhappy with Cypress' decision to actively expand its CNS licensing and development activities, and when Cypress announced their first deal, with BioLineRx (BL-1020 for schizophrenia), Ramius responded with a buyout offer, which was rejected. One of Cypress' directors resigned due to a vaguely defined disagreement over the revised strategy, and Cypress has doubled down with two more licensings, with Alexza (Staccato nicotine for smoking cessation) and Marina Biotech (intranasal carbetocin/oxytocin for autism). The Marina Biotech licensing, of a synthetic oxytocin (carbetocin), is particularly interesting. There have been a number of pilot studies indicating that oxytocin can be of benefit in autism, improving social awareness and responsivity. One imaging study indicated an increase in activity in the fusiform gyrus, wherein the recognition of human faces appears to be centered, perhaps increasing the valence of human vs. nonhuman object responsivity. Parenthetically, in the context of Oliver Sacks' discussion of his own prosopagnosia (inability to recognize faces), in a recent issue of The New Yorker, it begs the question of whether there is a common anatomical denominator--and perhaps treatment target--for a spectrum of disorders that includes autism, Asperger's, and prosopagnosia.

Whether intranasal carbetocin would emulate those pilot studies and provide benefit in this core autistic deficit will have to be shown in a POC trial expected to start 1Q:11, but Cypress paid only $750,000 upfront for that license. Now the question is whether Ramius will try to emulate a different kind of paradigm, the corporate takeovers engineered by Biotechnology Value Fund and Kevin Tang at Avigen and Endo respectively--and if they try, whether they will be successful. Each of these licensed programs has an appealing rationale and potential--but if Ramius does mount a credible challenge, Cypress will have to make their case for a longterm perspective, prior to obtaining confirmatory data.

1H:2010 Lowlights

from NI July/August 2010

1. Pfizer/Medivation's Dimebon bombed in its first real test of its benefit in Alzheimer's. The biggest divergence between groups was on the MMSE, where a p=.10 trend towards more improvement was seen--for placebo. Pfizer had pretended that this Medivation-run trial did not exist, one can now see why. It is impossible to salvage even a scrap of optimism from this Flurizan-scale debacle. Medivation has tried to preempt Pfizer's cancellation-option by terminating the remaining two monotherapy trials. Will adding Aricept make a difference, as is still being tested? In that scenario, one would have to hope for some type of totally unexpected biochemical synergy. This seems highly unlikely.


2. "He had been stuck for four days when his water ran out. On the sixth day, the 27-year-old mountain climber knew there was only one way he could survive. Using a pocketknife, Ralston cut off his own arm." -- St. Petersberg Times, 5/3/2003
   
   Big Pharma has apparently taken a page from The Survivalist's Handbook, and confused it with a Harvard Business School case study. Virtually every BP with a CNS program worth mentioning has amputated thousands of staff members. They began with the armies of sales reps rendered irrelevant in a de facto price-controlled environment, where access to physicians is being increasingly restricted. But now they are slicing off strategic development planners and R&D departments. Several companies (Lilly, Sanofi-Aventis) deleted CNS from their roster of business units; others deleted inhouse research on psychiatry (AstraZeneca) or shelved entire indications, like GSK's divestiture of depression and pain. We are very aware of the patent cliff, and the concept of shifting some R&D to an outsourcing model, utilizing small company efficiency and focus, is something NI has advocated for years. But this culling is draconian and ill-planned. Within the next two years, we expect that several of these companies will reverse course, reinstating programs and emphases deleted in the heat of panic, looking to rehire some of those now being treated as so much irrelevant flotsam.


3. Biovail Merger with Valeant Pharma: In theory, growing the company via this merger could fund development for Biovail's CNS pipeline. But given the fact that Valeant has been putting its own resources into other areas, like dermatology and emerging markets; exited CNS after its successfully licensed retigabine to GSK; and that it is the Valeant CEO, Michael Pearson, who will oversee daily operations as CEO of the merged company, there is reason to worry for Biovail's CNS pipeline strategy. We would not be surprised to see an inlicensed program dropped, cheering only the short-sighted denizens of Wall Street, who believe that addition is best achieved by subtraction.

Through A Glass Darkly: Transparency II

Last month, NI commented favorably upon the FDA task force recommendations for increased Agency transparency, which include the public release of such crucial communications as Complete Response and Nonapprovable letters. Our view was, and is, that the nondisclosure of such information permits companies to convince their investors, and sometimes themselves, that doomed projects deserve to be sustained. It also precludes companies from learning from each other's mistakes, and thus increases the likelihood that they will repeat them. Since that time, industry organizations like PhRMA and BIO have lined up squarely against the proposal, suggesting that public disclosure will provide a competitive advantage to straggler companies, allowing them to streamline and focus their development tactics by riding the coat-tails of the regulatory lessons learned by their forerunners. PhRMA had the chutzpah to argue that companies should not be compelled to reveal even the fact that they had filed for approval, and/or received a FDA response. BIO suggested that companies should have veto power over what is revealed, which would render the policy impotent.

This is rubbish. Beyond the absurdity of claiming that the very fact of having filed provides aid and comfort to the competitive enemy, there is an underlying philosophical premise here that needs to be sent into retirement. Both collective boon and bane come from increased transparency: Yes, from a short-sighted viewpoint, eliminating some minuscule aspect of the uncertainty attendant to any CNS drug development project might be seen as a concession to the competition. But so long as the rules apply to all, eventually it will even out; everyone benefits from having increased access to information that allows them to not go down blind alleys. Years of trial-and-error, and billions in unnecessary costs, could be eliminated through improved transparency. Indeed, it is this principle that has recently led the FDA and NIH to take the lead in setting up an industry database for Alzheimer's, one that will aggregate clinical data in the service of promoting just that kind of efficiency. Perhaps the main reason that this is being resisted at the regulatory disclosure level is that corporate accountability for performance would be heightened. Indeed, we believe that this is the core fear that feeds the industry's recalcitrance, that they will be held more accountable. One could conveniently and disingenuously argue that a company which is always in the lead will always then find itself carrying a greater relative proportion of the experimental risks and costs. But the reality is that no company has come anywhere near establishing itself as being first on all fronts--it would be a nice problem to have. Critical IP can continue to be protected via the patent process. This is not naive altruism: Transparency is just one of several avenues to reducing the time and expense which are turning pharmaceuticals into luxuries that stressed and/or underdeveloped economies cannot afford. If the industry tries to be the immovable obstacle, it will be crushed by the irresistible force of societal demands for less expensive healthcare. Given the economic realities, trying to ensure that everyone must repeat the same expensive mistakes in order to 'level the playing field', or camouflaging them in order to continue to tap investors for funding, is neither fiscally or ethically tenable.

Gilenia Set to Rock the MS World

from July/August 2010

• Gilenia is effective in reducing relapse: 25-0
• Gilenia delays disability: 24-1
• Gilenia should be a first-line therapy: 21-3
• Gilenia is safe (enough) for its intended use: 25-0
• Gilenia should be tested at a lower dose post-approval: 20-5

The FDA Advisory Committee votes for Novartis' MS drug Gilenia/FTY720/fingolimod were less ambivalent, and portend more for Gilenia's prospects, than NI had anticipated. We had expected that efficacy would be deemed established, but there was somewhat less angst about Gilenia's side effect/AE profile than we had forecast, given the FDA's past experience with Biogen-Idec/Elan's Tysabri. The FDA will undoubtedly still insist on a REMS program that will pinpoint any emerging adverse events earlier rather than later, and baseline cardio, pulmonary, and vision testing will likely be required. But two of the aforementioned recommendations stand out in their probable impact on the MS market.

First, had the AC defined the minimum effective dose as needing to be established before approval, that trial would have delayed approval by years. Novartis has already stated that they will run a major Phase IV, doing so post-approval will allow Gilenia a vital head start as the first oral immunotherapy for MS. This is an area where the bar may be raised considerably for the second oral drug to be considered. The prospects for the other several immunotherapies in late clinical development (e.g. Campath, cladribine, laquinimod, BG-12, daclizumab) could well be colored by Gilenia's advent and performance, because the FDA could tighten up on later entrants IF Gilenia proves relatively safe and predictable. On the other hand, it has enough question marks in terms of longterm safety that the FDA could be receptive to anything that offers the prospect of an improved risk-benefit calculation. MediciNova's ibudilast is much earlier in development, and recent evidence that it acts via MIF (macrophage migration inhibiting factor) leaves open a small possibility that this anti-inflammatory tactic might turn out to have unexpected immune system implications.

For now, with Tysabri generally seen as a second-tier, post-ABC therapeutic step, to have Gilenia endorsed as a first-line therapy option puts it in an entirely different risk category from a regulatory and prescriber point-of view. The fact that it is orally-administered (even if the first dose is supposed to be done at the MD's office) will overcome a significant source of patient resistance.

There is always the possibility that the FDA could err on the side of greater caution in its final response to the NDA, but there is no clinical basis for that, and no discernible regulatory agenda in that direction. While Merck Serono has refiled the NDA for cladribine, we do not expect that cladribine is now going to have an accelerated path forward in the wake of having its initial filing refused by the FDA. Gilenia will be the first-to-market of the oral alternatives. What does this mean for the MS market in the United States?

Novartis can expect rapid user uptake, as patients who have been on-the-fence about Tysabri jump off, and on to Gilenia's bandwagon-to-be. Barring some unexpected AE that arises as use expands, as has now been the case with Tysabri and PML, Gilenia's market potential will be on the high side of the $1 billion to $2.3 billion range we have seen from various analyst estimates.

As the major US player for MS, via both Avonex and Tysabri, no one is more vulnerable to Gilenia than Biogen-Idec. Patients who ordinarily would have embarked on a first course of Avonex may well start with Gilenia instead, those whose response to a beta-interferon or Copaxone is flagging, for whom Tysabri would have been the next step, will look at Gilenia first. The value of a PML predictive test will be been diminished, because, one can entirely avoid PML risk by taking the Gilenia route. Ris-averse prescribers and patients may choose to wait for a larger mass of clinical safety data to be developed for Gilenia, but that end of the user spectrum is the least likely to utilize Tysabri, particularly as the number of PML cases continues its inevitable rise (now at 55). The question to be answered by sometime in 1Q:11 will not be whether Tysbari has regained sales growth momentum, it will be whether one any decline in utilization has yet become apparent.

This was the nightmare scenario for Biogen-Idec, and is hardly better for Genzyme, Teva, or Merck Serono. Barring a surprise from the regulatory or safety side, it will come to pass.

Cortex Signs RD Deal with Biovail

(ADDED 3/26/10)

Cortex Pharmaceuticals and its AMPA-modulating platform have long been of interest due to their potential applicability to a range of psychiatric and neurological disorders--but the last few years have been extremely difficult due to chronic, dire fiscal straits. However, Cortex yesterday sold its respiratory depression program, and four related compounds (two of which are preclinical) to Biovail, which continues to be highly opportunistic and creative in its MAP activities. The $10 million ( there is another $15 million in potential milestones) will allow Cortex to finally conduct the Phase II ADHD trial using CX-1739 that has been their elusive goal for a couple of years. They can also complete their sleep apnea study, although that is much more of a wild card at this point.

Cortex now has a degree of control over their destiny that has been missing since the FDA hamstrung their original ADHD program, which used a predecessor molecule. More importantly, this allows them to return to the psychiatric/neurological focus which has always been their inherent raison d'etre, while Biovail can explore the somewhat more arcane potential of the respiratory depression indication. We like the deal a lot, but it should be noted that NI Research played a role in initiating the discussions that eventually led to this successful resolution--so we are not completely objective on that score. Having said that--we doubt that even the most dispassionately objective observer could quarrel with the conclusion that this provides Cortex Pharmaceuticals with an opportunity for corporate resurrection that many had given up for lost.

There Is No Joy in Mudville--Mighty Dimebon has Struck Out

Revised 4/1/10

Pfizer/Medivation's Dimebon bombed. Completely. In this first real test of its benefit in Alzheimer's, the biggest divergence between groups was on the Mini-Mental State Exam--where a p=.10 trend towards more improvement was seen--for placebo. Pfizer had kept this Medivation-run trial at arm's length (essentially pretending it did not exist), and now one can see why. It is impossible to salvage even a scrap of optimism from this Flurizan-scale debacle. Even if one could parse out some subgroup which obtained significant benefit from the drug, that must be offset by a subgroup which is made significantly worse. This would change the risk-benefit calculation dramatically, but we believe it is a moot point; Dimebon probably neither helped nor hurt any patient subgroup. Would adding Aricept make a difference, as is being currently tested? In that scenario, one would have to hope for some type of totally unexpected biochemical synergy. Which seems highly unlikely.

Pfizer's neuroscience program is in flux, following the news that their new neuroscience head, who had came over from Wyeth, is now departing to run drug research at AstraZeneca--not reassuring for Pfizer observers. They will be searching for a permanent replacement, which could delay a decision on whether Dimebon is wheat, or chaff. There is a fair chance that Pfizer could decide that the potential for even bigger disappointment may outweigh the much attenuated hope that Dimebon can replace Aricept, and terminate the partnership. Indeed, the best chance the Dimebon partnership has for continuation could rest in the likelihood that there is no one currently in a position of sufficient power within Pfizer's neuroscience area to actually pull the plug.

Post-Mortem

Just as a botched hospital surgery leads to a QA review to sort out what went so horribly wrong, it is worth giving some consideration to where this process went astray; it is most assuredly not just in the vagaries of the clinical trial process. Several systemic flaws are on display here:

1. Pedigree: The cachét afforded by a major partner can hinder objective assessment. Back in the day, when Dimebon was Medivation's Russian Phase II baby, we were rather harsh in our assessment: (NI May 2008): "Having already been the basis for a major financing, a hefty increase in company valuation, and a Phase III program predicated upon using this trial as one of two pivotal trials, Medivation just had two presentations given on Dimebon with still more data mined from the same arcane lode....The limitations of, and questions about, that trial's adequacy are not resolved by upping the sheer volume of data garnered from that small, geographically limited trial."
   But then Pfizer paid $225 million up front, and the Karolinska Institute confirmed the claim of mitochondrial activity. We backed off somewhat, for while worried by the weirdness of the dataset, it was hard to not feel intimidated by the fact that centers of bastions of scientific prestige had vouched for Dimebon in some fashion. Surely, they had to know what they were talking about.
2. If it seems too good to be true, it probably isn't true: Medivation, Pfizer, their outside advisors, the investors who gave Medivation millions; they all fell in love with the idea that they had found that special something. Pfizer's braintrust finetooth-combed the Russian data and found it worthy. But no matter how clean it looked in retrospect, it was still a small (183pt), short (three month), narrow population trial carried out in a context where oversight is not all that it can be (Russia).The recent history of the Alzheimer's field (Flurizan, Alzhemed, bapineuzumab) clearly highlights the risks of zooming into Phase III, but once these players were on that path, they could not objectively appraise their own strategy. It was cognitive dissonance theory played out at high cost: Once people behave in accordance with a belief system, that belief system becomes all the more entrenched.
3. Hubris: Part of this lost objectivity reflects a hubris that is endemic to much, albeit not all, of the pharma industry. It may actually be necessary, in order to take on the kinds of risks that are involved, when so many unknowns are being juggled at once. They believe that they are more clever, more opportunistic--they think they know when one can bend the rules. NI has no doubt that those atop Pfizer and Medivation have forgotten far more neuroscience than we will ever know. But this pride is both a useful impetus and an inevitable peril. Outsiders less enamored may recognize this before those who are both more versed and immersed.
4. Phase II is not optional: Belaboring the obvious--all of the above could have been corrected-for, had only Pfizer first run a reasonable Phase IIb trial. Both Medivation and Pfizer would have saved many millions.
5. The Syllogism of Mechanism: Mitochondrial protection is (in animals) a means of neuroprotection. In the lab, Dimebon is a mitochondrial protectant. Ergo, Dimebon must be neuroprotective in humans. Not necessarily--not at these dose-levels.

This can only further chill a CNS drug development environment that was already subzero, where companies and investors will wonder; if Pfizer couldn't see this coming, why would we dive into the same sinkhole? The answer: Alzheimer's is still the juiciest plum yet to be harvested in the pharmaceutical orchard.