23 April 2009. While scientists’ understanding of the pathobiology of neurodegenerative diseases has grown in leaps and bounds over the last 20 years, those advances have not readily translated into cures. Treatments for Alzheimer and Parkinson diseases, the most common neurodegenerative diseases by far, are still rooted in softening symptoms rather than tackling the underlying disease progression. Rarer neurodegenerative diseases, such as amyotrophic lateral sclerosis (ALS) and Huntington disease (HD), to name just two, have not fared any better. What can be done to accelerate the discovery process and get drugs to the clinic? Drug Development for Neurodegenerative Diseases, a Marcus Evans conference held 7-8 April in Boston, attracted several dozen attendees, mostly from industry, to review, rethink, and re-strategize. The conference featured a mix of presentations from academia, big pharma, diagnostic companies, and contract resource organizations. Much of the program would be familiar to Alzforum readers—Neil Buckholtz from the National Institutes of Health, Bethesda, Maryland, gave a concise presentation of ADNI (see ARF related news story), Serge Gauthier, McGill University, Montreal, Canada, reviewed the challenges of diagnosing and monitoring prodromal AD, and Mark Shearman, Merck Research Laboratories, Boston, gave a roundup of the current drug discovery strategies for AD. But some news was on offer as well. “It was good exposure to a range of issues in neurodegenerative diseases,” summarized Barry Greenberg, who now directs neuroscience drug discovery and development at Toronto Western Research Institute, Canada.
Clinical and Preclinical Updates
Ann Saunders from GlaxoSmithKline updated some of the pharmacogenomic data on rosiglitazone, the insulin sensitizer that has been put through its paces in several clinical trials for AD. In Phase 2 trials, that drug failed to improve cognition (ADASCog) or global scores (CIBIC+) in the treatment population as a whole, but appeared to perhaps help those patients who did not carry the ApoE4 allele (see ARF related news story and Risner et al., 2006). Saunders said that those results led to extensive dialog with the U.S. Food and Drug Administration and the European Medicines Agency (EMEA) on how to go forward with Phase 3 trials. Those discussions have led to three trials, two one-year trials where rosiglitazone is being used as an adjunct therapy to either any cholinesterase inhibitor (REFLECT-3) or just donepezil (REFLECT-2), and a 24-week monotherapy trial (REFLECT-1). In all cases, participants are stratified on the basis of ApoE status; i.e., they must undergo ApoE genotyping in advance. The trials have recruited from approximately 5,000 patients, and about 3,000 have entered the trials, which are ending now, said Saunders. She said GSK plans to present some data from them at ICAD in Vienna this July.
Bruce Morimoto of Allon Therapeutics, Vancouver, British Columbia, summarized the standing of AL-108, otherwise known as NAP, a fragment of the neuroprotective protein ADNP. AL-108 has been tested in a Phase 2 trial for amnestic mild cognitive impairment. Participants tolerated the drug well, and the trial appeared to suggest benefit at eight and 16 weeks based on a 12-second delayed match to sample (DMTS) cognitive test (for further details of the trial, see ARF related news story). Morimoto suggested that further testing in a Phase 2b trial is warranted.
Conference attendees debated what the mechanism of action of NAP might be. There are indications that it can prevent both tau and Aβ toxicity, though the rapid improvement in the DMTS test, if it holds up in larger trials, is not compatible with a disease-modifying effect, Morimoto acknowledged. (Typically, decline-over-time curves predict an initial uptick followed by decline as steep as in the untreated group for symptomatic drugs, but no uptick followed by less steep decline for disease-modifying drugs.) Morimoto suggested that AL-108 might have both symptomatic and disease-modifying activity. Attendees also asked for data showing that the drug targets and stabilizes microtubules directly, which has been suggested as a mechanism of action. At the doses given, the stoichiometry for such a mechanism simply would not work, Greenberg pointed out. Morimoto agreed and suggested NAP may be setting off a protective cascade, or that it may stimulate microtubule initiation. Since there are few microtubule initiation sites in any one cell, the stoichiometry would work out at the doses administered, Morimoto said. Another explanation that would fit the stoichiometry is that the drug might act like a capping protein to stabilize the end of microtubules. “It was very nice to get this update since this drug and Dimebon are the only ones that have been reported in Phase 2 to improve patients over background rather than just slow the rate of decline; those certainly need to be expanded into Phase 3,” said Greenberg.
The microtubule theme continued with Robert Stein of KineMed, Emeryville, California. KineMed uses stable isotopes, such as deuterium, to measure the kinetics of biological pathways. That work led to the identification of rapid turnover of microtubules in the sciatic nerve as a prelude to disease onset in a G93A mutant superoxide dismutase 1 (mSOD1) model of ALS. KineMed showed that stabilizing microtubules is protective in this ALS model (see Fanara et al., 2007) and they currently have a candidate neuroprotective in preclinical development. That compound, KM-ID05, slows microtubule turnover, improves axonal transport, helps maintain stride length, and delays mortality in these animals, even when given after onset, reported Stein. It is impossible to routinely measure microtubule turnover in humans, since that requires postmortem analysis of deuterium incorporation into proteins, but Stein showed that microtubule-dependent fast axonal transport might serve as a surrogate. Some molecules transported by fast axonal transport, such as neuregulin-1, end up in the CSF, and it is possible to measure that turnover, he said. KM-ID05 does restore normal fast axonal transport in the mSOD1 model.
KineMed are also using their stable isotope incorporation, this time into DNA, to measure neurogenesis, claiming that it is a more reliable method than incorporation of the nucleoside analog BrdU. Using this method as a screening tool, they found a compound that boosts neurogenesis and promotes neuronal survival. In mouse models of AD, this compound outperforms donepezil in improving learning and memory deficits, including novel object recognition and spatial memory, Stein claimed.
Another class of compound that may prove useful in rescuing memory deficits are histone deacetylase inhibitors (for a recent example, see Ricobaraza et al., 2009). Holger Patzke of EnVivo Pharmaceuticals, Watertown, Massachusetts, reviewed his company’s strategy for preclinical screening in fruit flies. That work yielded histone deacetylase candidates that enhance short- and long-term memory in mice (see ARF related news story). The toxicology and safety profile in mice looks good, said Patzke. EnVivo is currently conducting a Phase 1 trial in healthy human volunteers, which is expected to be completed in 2009.
Express Delivery—Getting Drugs Directly Into the Brain
One formidable obstacle in the path of anyone trying to develop a drug for neurodegenerative diseases is the blood-brain barrier. Many potential therapeutics fail because they don’t reach the brain. Linda Van Eldik runs the Center for Drug Discovery and Chemical Biology at Northwestern University, Chicago, Illinois, where “fail fast, fail early” is the mantra. Van Eldik stressed an integrative approach to drug discovery where medicinal chemistry and preclinical biology combine to weed out unsuitable compounds in the drug discovery pathway. “Small improvements early in the process pay dividends later,” she said. This approach has led to development of two types of molecule, minokines and minozacs, that may prove useful in AD. Both reduce Aβ-induced production of pro-inflammatory cytokines (see ARF related news story). Van Eldik showed that the latter compounds show promise in animal models of traumatic brain injury and multiple sclerosis as well.
If a drug cannot be made to cross the blood-brain barrier, however, an alternative approach might be to deliver it directly into the brain. “I think this is one of the most exciting developments,” said Greenberg. “There may actually be a paradigm shift in the way we discover and deliver drugs to people.”
The shift might be typified by technology being pursued by Medtronic, Minneapolis, Minnesota. This device company is best known for making heart pacemakers, but its neuromodulation division is working on mini-pumps that deliver drugs directly into the cerebrospinal fluid. Medtronic’s Gregory Stewart reported preclinical data on direct delivery of an Aβ antibody (6E10) into the CSF of Tg2576 transgenic mice. That method (see ARF related news story) seems to avoid the microhemorrhaging that has plagued other antibody approaches to clearing Aβ. Speaking with this reporter after his talk, Stewart noted that while it is difficult to get drugs into the brain across the blood vessels, it is much easier to get molecules out of the brain since the CSF drains through the arachnoid granulations, which are basically one-way valves. This helps ensure that the ~500 milliliters of CSF an average person makes every day do not distend the 150 milliliters of fluid space in the brain, and it may also explain how antibody-Aβ complexes can be removed from the CSF without damaging blood vessels.
Stewart said that in terms of direct delivery of drugs to the brain, the organ’s parenchyma is the most challenging target. Currently, convection-enhanced delivery is being explored; this basically means delivering therapeutic solutions at high enough pressure to push them through the parenchyma. Medtronic is currently exploring how to scale this methodology up for human use. However, Stewart cautioned that the size of the brain changes by an order of magnitude going from rat (2 g), to rhesus monkey (90 g), and again to human (1,500 g), making parenchymal delivery the “brave new world” for human studies.
Another brave new approach is using viral or cell delivery systems. Lamya Shihabuddin from Genzyme Corporation in Cambridge, Massachusetts, reviewed an adenovirus delivery strategy for treating Parkinson disease patients. The company is using this strategy to boost levels of the enzyme amino acid decarboxylase (AADC) in the striatum. Loss of this enzyme in later stages of the disease means that the drug L-dopa cannot be efficiently converted to dopamine. To date, the strategy has been tested in 10 patients, in whom it seemed well tolerated, Shihabuddin said. PET imaging showed that AADC activity resumed and the patients showed signs of improvement, though this must be taken cautiously, said Shihabuddin, since this is an open-label study with small sample size and placebo effects in PD have been known to be quite large. She said a Phase 2, multi-center study that will be sham-controlled is planned and will enroll 60 subjects (see Cohen comment on surgery control in PD trials).
Stem cell therapy is yet another strategy for compensating for neurodegeneration, though the message from Paul Stroemer of ReNeuron in Guildford, U.K., was that using stem cells is going to be a lot more difficult than people think. “Stem cells for research and stem cells as a product are very different entities,” he said, adding that the main challenge is the cells’ limited ability to grow and their instability. Nevertheless, ReNeuron has developed stem cell lines incorporating an inducible cMyc construct. cMyc is basically a “stemness” gene, said Stroemer, since it drives self-renewal of the cells without affecting their multipotency.
Using “good tissue practices” (the stem cell equivalent of the good laboratory practice standard), ReNeuron has developed a stem cell line (VME0B06) that can differentiate into neurons, astrocytes, and glia. The scientists injected the cells into the striatum and substantia nigra of a rat model of PD. Over six months the animals showed improved motor function compared to sham controls, but interestingly, none of the transplanted cells differentiated into tyrosine hydroxylase-expressing cells—hence, they are not producing any dopamine. Stroemer said that stem cells are unlikely to work as a patch to fill in for missing neurons, but are more likely to simply add trophic support to the neurons that are still there by releasing growth factors and encouraging angiogenesis, for example. The company is developing a second cell line for treatment of stroke, currently in a Phase 1 clinical trial where two to 20 million cells will be injected into the putamen using a minimally invasive procedure.
Strategies for Discovery
Other speakers described tools and strategies that could be useful to people pursuing clinical trials or screening for possible targets. Samuel LaBrie from Rules-Based Medicine in Austin, Texas, a diagnostics company, reported that his company is expanding to work on focused biomarker panels in AD. In cooperation with the Stanley Medical Research Institute of Chevy Chase, Maryland, LaBrie and colleagues are looking prospectively at blood samples taken prior to the emergence of schizophrenia to see if they can identify predictive markers. Matt Lear from Asterand in Detroit, Michigan, described how that company’s tissue bank work tries to bring donors and users together, with antemortem consent and the approval of institutional review boards to make the process as smooth as possible. Gail Adinamis of Clinical Resource Network, LLC, Deerfield, Illinois, showed how her company can improve clinical trial enrollment and retention by providing a variety of in-home testing and diagnostic services. This may be particularly useful for AD trials, given the often high dropout rates and the difficulties in getting more functionally impaired patients into the clinic.—Tom Fagan.