Alzheimer's
disease (AD) is a major health and societal problem and is the
predominant cause of dementia. Age is the major risk factor for AD, and
with the predicted demographic change to an increasingly elderly
population, the prevalence of the disease will reach a staggering 100
million by 2050. Acetylcholinesterase inhibitors such as donepezil,
rivastigmine and galantamine offer a modest symptomatic relief for a
short period of time in mild-to-moderate AD, and memantine, a
low-affinity N-methyl d-aspartate
receptor antagonist, is licensed for use in moderate-to-severe AD.
However, none of the current therapies is able to alter the course of
the disease.
The health care burden of dementia is equivalent to the gross domestic product of several mid-size countries (Wimo and Prince (2011) and unless treatments are found for AD that delay either disease onset or slow down disease progression, the cost to health care services around the world will be punitive.
The health care burden of dementia is equivalent to the gross domestic product of several mid-size countries (Wimo and Prince (2011) and unless treatments are found for AD that delay either disease onset or slow down disease progression, the cost to health care services around the world will be punitive.
Compared
with other diseases of the mind, such as schizophrenia, there is, in
fact, a very solid body of evidence that provide insights into some of
the key pathological processes in the disease. The autosomal dominant
mutations to the amyloid precursor protein (APP), presenilin 1 and
presenilin 2 that cause early-onset AD clearly point to the key role
that Abeta deposition plays. The similarities of the intracellular tau
abnormalities seen in AD and the autosomal dominant forms of
fronto–temporal lobar degeneration mediated by mutations to the tau gene
strongly support the case that tau is implicated in neuronal death, and
is likely downstream of Abeta deposition. These observations have been
formulated in the ‘amyloid cascade hypothesis’, which while having its
opponents, nevertheless incorporates much of the key data in a logical
and compelling manner (Hardy et al. 1998; Hardy and Selkoe 2002; Karran et al. 2011).
And yet, and yet. The AD research field has recently been rocked by
three phase 3 failures: semagecestat, a gamma secretase inhibitor that
inhibits the production of the Abeta peptide failed its primary end
points in 2011; and now, within a few weeks of each other, the news that
both the anti-Abeta monoclonal antibodies, bapineuzumab (Pfizer/Janssen
Alzheimer Immunotherapy) and solanezumab (Lilly), that were being
tested in phase 3 trials have failed to meet their primary end points;
an improvement in cognition and activities of daily living. These last
three disappointments come after the other phase 3 failures of
tramiprosate in 2007 (Aisen et al. 2011) and tarenflurbil in 2008 (Green et al. 2009)
that were also ‘amyloidocentric’ – that is, they targeted the Abeta
peptide so as to reduce (via different mechanisms) its deposition in
brain parenchyma (Table 1).
However, the field acknowledges that the pre-clinical data and human
biomarker data for these latter two molecules were not compelling (Golde
et al. 2011).
Drug name and proposed mechanism of action | Phase 2 results | Phase 3 results |
---|---|---|
Tramiprosate Abeta aggregation inhibitor | 58 mild-to-moderate AD patients randomized to four groups: placebo, 50, 100, 150 mg/kg tramiprosate bid for 3 months. Drug mediated a significant lowering of Abeta 42 in CSF samples (Aisen et al. 2006) | 1052 mild-to-moderate AD patients randomized to three groups: placebo, 100, 150 mg/kg bid for 78 weeks. No significant effects on primary outcome measures of ADAS-cog and CDR-sum of boxes (Aisen et al. 2011) |
Tarenflurbil Gamma-secretase modulator | 210 mild-to-moderate AD patients randomized to placebo, 400, 800 mg bid tarenflurbil for 12 months. Some evidence for an improvement in activities in daily living at the 800 mg bid dose (Wilcock et al. 2008) | 1684 mild AD patients randomized to placebo, 800 mg bid tarenflurbil for 18 months. No significant effects on primary outcome measures of ADAS-cog and ADCS-activities of daily living (Green et al. 2009) |
Semagecestat Gamma secretase inhibitor | 51 mild-to-moderate AD patients randomized to placebo, 100, 140 mg od semagecestat following dose escalation for a total duration of 18 weeks. Significant reduction in plasma Abeta 40 peptide (Fleisher et al. 2008) | Trial data currently unpublished. 2600 mild-to-moderate AD patients randomized to placebo, 100, 140 mg semagecestat od for 76 weeks in two trials (ClinicalTrials.gov identifiers NCT00594568; NTC00762411) enroled. Trials were halted after interim analysis showed increased incidence of skin cancer and worsening of cognition and activities of daily living |
Bapineuzumab humanized monoclonal antibody directed at amino acids 1–5 of Abeta peptide.
Amyloid plaque clearance mediated by microglial activation
| 234 mild-to-moderate AD patients, randomized to placebo, 0.15, 0.5, 1.0 or 2.0 mg/kg bapineuzumab IV infusions every 13 weeks for 78 weeks. Some evidence for an improvement in cognitive and functional end points in study completers and ApoE4 non-carriers (Salloway et al. 2009) | Trial data currently unpublished. 4500 mild-to-moderate AD patients randomized to placebo and 0.5 mg/kg IV every 13 weeks for 18 months in ApoE4 carriers, and randomized to placebo and 0.5 and 1.0 mg/kg IV every 13 weeks for 18 months in ApoE4 non-carriers in four trials (Clinical Trials.gov identifiers lNCT00575055; NCT00574132; NCT00676143; NCT00667810.) Trials were halted after completion of two trials demonstrated a failure to meet primary outcome measures of cognition and activities of daily living |
Solanezumab Humanized monoclonal antibody directed at amino acids 16–24 of Abeta peptide
Amyloid plaque clearance mediated via peripheral sink mechanism
| 52 mild-to-moderate AD patients were randomized to placebo, 100 mg every 4 weeks, 100 mg weekly, 400 mg every 4 weeks, 400 mg weekly IV solanezumab for 12 weeks. There was a significant dose-dependent increase in Abeta 42 peptide in CSF (Farlow et al. 2012) | Trial data currently unpublished. 2000 mild-to-moderate AD patients randomized to placebo and 400mg solanezumab monthly IV for 18 months (Clinical Trials.gov identifiers NCT00905372; NCT00904683). Trials failed to meet their primary outcome measures of ADAS-cog and ADCS-activities of daily living. A secondary analysis of mild AD patients pooled from both trials showed a significant effect on cognition |
The most
recent findings will stimulate a number of researchers in the field who
will claim that the amyloid cascade has been tested in the clinic and we
can accept the null hypothesis. More worryingly for the future
development of medicines, many senior executives in pharmaceutical
companies will be drawing breath and considering whether they can
continue to pour money into a hugely expensive endeavour that has thus
far delivered no return on investment. In particular, those companies
that have therapeutic approaches in earlier phases of clinical
assessment (Lobello et al. 2012)
that are similar in pharmacological mechanism to those that have failed
will be urgently reviewing their future development strategies. In ‘big
pharma’, there is a continual internal competition for resources and it
could well be that neuroscience research will suffer as companies
divert their funding to other debilitating diseases that require new
therapies – such as cancer, diabetes and arthritis – which have all
produced a flow of phase 3 clinical successes providing excellent new
medicines for patients.
It is helpful to
reflect on some important aspects of drug discovery and development.
With respect to semagecestat, it is noteworthy that a detailed
biochemical analysis of the activity and selectivity of that compound
(Chavez-Gutierrez et al. 2012)
was published after the compound had been through its entire
development and failure to show efficacy in AD patients. This study
demonstrated that in a controlled biochemical, cell-free assay,
semagecestat was slightly more potent as an inhibitor of notch
processing compared with its activity as an inhibitor of Abeta
production, in contradiction to data produced in less well-controlled
cell-based assays that suggested the opposite. The liability of notch
inhibition is well understood and likely contributed to some of the side
effects seen in the phase 3 trials – notably, an increased incidence of
skin cancer. This simple fact exemplifies that the nature of the target
is hugely complex and that much scientific progress can be made in
parallel to a very lengthy drug development process. This may prompt the
question: why go ahead until you are absolutely sure of the profile of
your drug? But, as those who work at the coalface of drug development
know, if you want to wait for the perfect compound, then you will wait
forever – as will the patients you seek to treat. Another key element to
draw from the semagecestat story is that the clinical trial was
informative – while the final data have not yet been published, the
completely unexpected worsening of cognition mediated by the drug most
probably closes that particular avenue down for drug hunters. With
respect to bapineuzumab, here the failure, while again disappointing,
may eventually be shown to be informative. First, did the antibody reach
its target and exert its desired pharmacological effect – the clearance
of parenchymal plaque? The dose-limiting side effects of bapineuzumab
were vasogenic oedema and microhaemorrhage, mediated most probably by
antibody binding to vascular amyloid (Racke et al. 2005);
but what about parenchymal plaque? There were preliminary data from the
phase 2 clinical trials demonstrating that bapineuzumab treatment did
reduce amyloid in brain parenchyma as evidenced by reductions in the
binding of 11C-PIB, an amyloid-selective positron emission tomography (PET) ligand (Klunk et al. 2004; Rinne et al. 2010).
However, if this is not corroborated in the phase 3 trials, the trial
will not have established a proof of mechanism and thus it would be
difficult to conclude very much with respect to the viability of the
approach. If the treatment was able robustly to remove amyloid plaques
from the brains of patients, but with no hint of cognitive benefits,
then one interpretation is that by this stage in the disease process
amyloid plaques per se do not drive disease progression. Also,
we have yet to discover whether the earlier findings that the treatment
may have reduced the release of phospho-tau protein into the
cerebrospinal fluid (CSF) – widely believed to be a marker of neuronal
damage – have been confirmed (Blennow et al. 2010, 2012).
If tau release has been significantly and robustly reduced, but in the
absence of cognitive benefit, then the use of this particular biomarker
will rightly be called into question.
With
respect to solanezumab, again the failure to meet its primary end
points is very disappointing, but early reports suggest that in a
secondary analysis, there is a statistically significant improvement in
cognition in mild AD patients. Other questions of course immediately
come to mind: what was the magnitude of the improvement; was the
improvement in cognition reflected in improved activities of daily
living; was amyloid plaque reduced in the brain as assessed by amyloid
PET brain imaging; was CSF tau lowered? Thus, if the biomarkers and end
points measured in this trial point in a consistent manner to an effect
in mild AD – for example, a reduction in amyloid in the brain as
measured by an amyloid PET ligand; a reduction in CSF tau; the
restoration of normal Abeta 42 levels in CSF, correlated with a
preservation of cognitive function – then this would strongly suggest
that the field has taken the first step down the path of finding an
effective therapeutic for AD.
Finally, a
full analysis and comparison of the two antibody trials should be very
helpful to the field, as these monoclonal antibodies are quite
dissimilar in affinity, proposed mechanism of action and were
administered using different regimens. The affinity of solanezumab for
its epitope in the mid-domain of the Abeta peptide is in the picomolar
range (DeMattos et al. 2001), that of bapineuzumab for the N-terminal epitope in the Abeta peptide is in the nanomolar range (Bard et al. 2000).
At its most fundamental level, this will mean that solanezumab will
capture Abeta and sequester it for a significantly longer period of time
than will bapineuzumab. Solanezumab does not bind to deposited amyloid
plaque, either in the vasculature or the parenchyma, only to soluble
Abeta (Racke et al. 2005).
This explains why solanezumab did not cause an increased incidence of
vasogenic oedema and microhaemorrhage. It is designed to mediate
clearance of Abeta from the brain according to the peripheral sink
hypothesis, by shifting the equilibrium of Abeta in favour of transit
out of the brain of soluble Abeta (DeMattos et al. 2002).
Recent data from the phase 2 studies indeed confirm that levels of the
uncomplexed Abeta 40-amino acid peptide were decreased in the CSF of
patients in a solanezumab dose-related fashion, with a concomitant
increase in the levels of uncomplexed Abeta 42-amino acid peptide – this
being the type of profile expected if the more insoluble and
plaque-prevalent Abeta 42 peptide was being mobilized from an insoluble
to a soluble compartment (Farlow et al. 2012).
Bapineuzumab, on the other hand, binds both soluble and deposited
plaque. This mechanistic difference could prove to be very relevant to
data interpretation. For example, if both antibodies were able to remove
approximately equivalent amounts of amyloid plaque from the brain as
assessed by amyloid PET imaging, but only solanezumab has shown a
statistically significant effect on preserving cognitive performance in
mild AD patients, then one might argue that the microglial activation
that is thought to underlie at least some of the plaque-clearing
efficacy of bapineuzumab (Schenk et al. 1999)
might have caused collateral damage that obscured any potential
benefit. Bapineuzumab was administered to patients who were
apolipoprotein E4 (ApoE4) carriers at 0.5 mg/kg, and in ApoE4
non-carriers at 0.5 mg/kg, 1 mg/kg and 2 mg/kg initially, with the
2 mg/kg dose being abandoned, again because of concerns regarding
vasogenic oedema and microhaemorrhage. Bapineuzumab was given via
intravenous infusion every 13 weeks for 18 months. Solanezumab was also
administered via an intravenous infusion every month at a dose of 400 mg
per patient (i.e., approximately 5–6 mg/kg) for 80 weeks. For both
monoclonal antibodies, a comparison of the minimal effective doses
defined in the pre-clinical studies with those used in humans will be
instructive to interpreting the pre-clinical to clinical translation.
Thus, if the final doses selected for human studies fall well short of
the pre-clinical minimum effective dose, allowing for allometric
differences, then a failure to show proof of mechanism in man may not be
too surprising.
Despite the wealth of new
data that these trials provide, the ineluctable truth is that they
failed their primary end points. So, should we retreat from
amyloidocentric therapies? My view is unequivocally ‘no’. We now know
that amyloid deposition precedes the presentation of cognitive deficits
in AD by about 15 years (Rowe et al. 2010). While there are robust data that Abeta deposition is a key component of the disease process, we are much less sure about when and how
Abeta mediates its deleterious effects. Much of the data support the
notion that Abeta deposition triggers the disease – there are less data
to suggest that it drives the disease in a continuous fashion (Karran et al. 2011).
The recent phase 3 clinical trials were of 18 months duration, which
may be just too short a period over which to see a beneficial effect on
the primary end points. The field needs to persevere with
amyloidocentric therapies, but to invest more resource into
understanding the earliest phase of the disease process so as to
identify individuals that are on the cusp of Abeta deposition. If
individuals can be reliably, and relatively inexpensively, identified at
this juncture, then credible anti-amyloid therapeutics have a good
chance of demonstrating efficacy. Indeed, mutations to the APP gene that
have the effect of modestly reducing the production of Abeta peptide
from the APP holoprotein have recently been shown significantly to
protect carriers from AD – a natural, genetic proof of concept (Jonsson et al. 2012).
However, in this case, the reduction in Abeta production occurs from
birth, and whether a much later therapeutic intervention would have the
same effect remains unknown. A longitudinal study that takes repeated
measures of relevant biomarkers in a large cohort of subjects in their
mid-sixties could provide critical data to help identify individuals at
risk of early AD pathology –and potentially well-defined cohorts for
clinical trials. Also, work on other treatment modalities should
continue apace: for example, approaches to ameliorate tau pathology. It
is very likely that no single treatment approach will prove to be fully
efficacious.
In summary, although the
clinical landscape currently looks bleak, now is not the time to falter.
Each failed study provides additional data on which to build better
clinical programmes. And, more importantly, as a society we simply
cannot afford to give up.
Acknowledgement
EK holds stock in Johnson and Johnson.
References
- 2006) A phase II study targeting amyloid-beta with 3APS in mild-to-moderate Alzheimer disease. Neurology 67, 1757–1763. , , , , , and (
- 2011) Tramiprosate in mild-to-moderate Alzheimer's disease - a randomized, double-blind, placebo-controlled, multi-centre study (the Alphase Study). Arch. Med. Sci. 7, 102–111. , , et al. (
- 2000) Peripherally administered antibodies against amyloid beta-peptide enter the central nervous system and reduce pathology in a mouse model of Alzheimer disease. Nat. Med. 6, 916–919. , , et al. (
- 2010) Immunotherapy with bapineuzumab lowers CSF tau protein levels in patients with Alzheimer's disease. Alzheimers. Dement. 6, 2. , , , , and (
- A. A. B. I. (2012) Effect of immunotherapy with bapineuzumab on cerebrospinal fluid biomarker levels in patients with mild to moderate Alzheimer disease. Arch. Neurol. 69, 1002–1010. , , , , , , , and for the,
- 2012) The mechanism of gamma-Secretase dysfunction in familial Alzheimer disease. EMBO J. 31, 2261–2274. , , et al. (
- 2001) Peripheral anti-A beta antibody alters CNS and plasma A beta clearance and decreases brain A beta burden in a mouse model of Alzheimer's disease. Proc. Natl Acad. Sci. USA 98, 8850–8855. , , , , and (
- 2002) Brain to plasma amyloid-beta efflux: a measure of brain amyloid burden in a mouse model of Alzheimer's disease. Science 295, 2264–2267. , , , and (
- 2012) Safety and biomarker effects of solanezumab in patients with Alzheimer's disease. Alzheimers Dement. 8, 261–271. , , et al. (
- 2008) Phase 2 safety trial targeting amyloid beta production with a gamma-secretase inhibitor in Alzheimer disease. Arch. Neurol. 65, 1031–1038. , , et al. (
- 2011) Anti-abeta therapeutics in Alzheimer's disease: the need for a paradigm shift. Neuron 69, 203–213. , and (
- Tarenflurbil Phase 3 Study, G. (2009) Effect of tarenflurbil on cognitive decline and activities of daily living in patients with mild Alzheimer disease: a randomized controlled trial. JAMA 302, 2557–2564. , , , , , , and
- 2002) The amyloid hypothesis of Alzheimer's disease: progress and problems on the road to therapeutics. Science 297, 353–356. and (
- 1998) Genetic dissection of Alzheimer's disease and related dementias: amyloid and its relationship to tau. Nat. Neurosci. 1, 355–358. , , , and (
- 2012) A mutation in APP protects against Alzheimer's disease and age-related cognitive decline. Nature 488, 96–99. , , et al. (
- 2011) The amyloid cascade hypothesis for Alzheimer's disease: an appraisal for the development of therapeutics. Nat. Rev. Drug Discov. 10, 698–712. , and (
- 2004) Imaging brain amyloid in Alzheimer's disease with Pittsburgh compound-B. Ann. Neurol. 55, 306–319.Direct Link:
, , et al. (
- 2012) Targeting Beta amyloid: a clinical review of immunotherapeutic approaches in Alzheimer's disease. Int. J. Alzheimers Dis. 2012, 628070. , , , and (
- 2005) Exacerbation of cerebral amyloid angiopathy-associated microhemorrhage in amyloid precursor protein transgenic mice by immunotherapy is dependent on antibody recognition of deposited forms of amyloid beta. J. Neurosci. 25, 629–636. , , et al. (
- 2010) 11C-PiB PET assessment of change in fibrillar amyloid-beta load in patients with Alzheimer's disease treated with bapineuzumab: a phase 2, double-blind, placebo-controlled, ascending-dose study. Lancet Neurol. 9, 363–372. , , et al. (
- 2010) Amyloid imaging results from the Australian Imaging, Biomarkers and Lifestyle (AIBL) study of aging. Neurobiol. Aging 31, 1275–1283. , , et al. (
- 2009) A phase 2 multiple ascending dose trial of bapineuzumab in mild to moderate Alzheimer disease. Neurology 73, 2061–2070. , , et al. (
- 1999) Immunization with amyloid-beta attenuates Alzheimer-disease-like pathology in the PDAPP mouse. Nature 400, 173–177. , , et al. (
- Tarenflurbil Phase, I. I. S. i. (2008) Efficacy and safety of tarenflurbil in mild to moderate Alzheimer's disease: a randomised phase II trial. Lancet Neurol. 7, 483–493. , , , , , and
- 2011) World Alzheimer report 2010: the global economic impact of dementia, in Alzheimer's Disease International pp. 1–52. Alzheimer's Disease International, London. and (
No comments:
Post a Comment