22 January 2013. When asked why he robbed banks, Willie Sutton allegedly answered, “Because that is where the money is.” Surgeons hoping to treat neurodegeneration answer the same, except their bank is, of course, the human brain. “That is where the disease is,” said Ronald Crystal of Weill Cornell Medical College in New York City. At least in experimental settings, a growing number of neurosurgeons are inserting needles, catheters, and capsules into the brain or spinal cord in hopes of alleviating Alzheimer’s and other neurodegenerative conditions. The techniques allow them to circumvent the blood-brain barrier and place therapeutics near the desired site of action.
Even as surgery, in the form of deep-brain stimulation, has become a standard treatment for Parkinson’s disease (see ARF related news story), invasive therapies for degenerative brain diseases more broadly have been slow in coming, said Nicholas Boulis, a neurosurgeon at Emory University in Atlanta, Georgia. “Has it moved as fast as I had hoped? No. But I think there has been a solid effort,” he said. As a result, preclinical experiments and some clinical trials are underway.
There are three main direct delivery routes to the brain, said Eric Grigsby of the Napa Pain Institute in California: right into the parenchymal tissue via the cerebrospinal fluid (CSF) bathing the spinal cord, or into the CSF that fills the brain’s ventricles. In 1998, an attempt to use the latter method dealt an early blow to surgical treatments for neurodegeneration. Researchers infused nerve growth factor (NGF) directly into the brain ventricles of people with AD, but recipients suffered back pain and weight loss (Eriksdotter Jönhagen et al., 1998). “In retrospect, the dose was too high,” said first author Maria Eriksdotter of the Karolinska Institute in Stockholm, Sweden. “We did not realize at the time that NGF was that potent.” Lower doses only partially alleviated the side effects, however, leading Eriksdotter to conclude that the researchers had not targeted the therapeutic protein correctly. CSF bathes the entire brain and spinal cord, so the added NGF affected all neurons, causing the side effects.
The 1998 study was a defeat for surgical approaches overall. It slowed research, Eriksdotter said. Today, however, many researchers are back in the fight. Some companies think a simple intrathecal injection to the spinal fluid will be sufficient for a drug to reach the surface, and perhaps even inner workings, of the brain. Alternatively, the medical device company Medtronic, Inc., of Minneapolis, Minnesota, hopes to use its implantable pumps to precisely place antibodies and other medicines. Other researchers prefer biological vehicles, such as cells or genes, lodged exactly where the therapeutic need is greatest.
The very nature of degenerative disease makes these approaches an uphill battle. “The challenge for a surgical option in Alzheimer’s has always been that the disease spreads throughout the brain,” noted Constantine Lyketsos at Johns Hopkins University in Baltimore, Maryland. A needle or catheter cannot target all over. The trick, Lyketsos said, is to treat disease very early, at the point where it starts.
Pursuing surgical options is particularly important because current AD treatments only offer symptom relief, added Paul Aisen of the University of California in San Diego (UCSD). “We need to evaluate all plausible approaches until we have made a big impact on this epidemic,” he said. “We need disease modification—if it comes through brain surgery, that is not too high a price to pay.”
Another challenge in adding surgery to the AD toolkit is that dementia specialists may be less familiar with surgery than are movement disorder experts. That makes education an important component of developing treatments that require operations, said Lisa Shafer of Medtronic. “Once we say it is somewhat like a shunt, their eyes open up and they say, ‘Oh, okay,’” she explained of the pump system the company is developing.
Even simpler than a shunt or pump would be a shot to the spinal cord every few months to administer medication. Researchers at NeuroPhage Pharmaceuticals in Cambridge, Massachusetts, believe this route will allow enough of their amyloid-busting phage treatment to reach the brain to treat not only Alzheimer’s, but potentially other conditions where misfolded proteins form amyloids (see ARF related news story; ARF news story). Similarly, Isis Pharmaceuticals, Inc., of Carlsbad, California, aims to reach the brain via spinal cord injections with antisense oligonucleotide therapies for conditions such as Huntington’s disease.
With intrathecal delivery, some scientists think that a drug in the lumbar spine will travel up the spinal cord and infiltrate the brain tissue. “I am excited about this route,” said Kimberley S. Gannon of NeuroPhage. “It is not very well known or well accepted that you can use an intrathecal bolus to get good brain penetration.”
Both intrathecal injections and implanted pumps are standard to provide morphine for pain, medication for the abnormal muscle contractions of spasticity, or chemotherapy for local tumors. But the success of intrathecal delivery depends on the fact that the medication—perhaps a half or full milliliter is injected—stays put in the lumbar spinal cord. If morphine were to reach the brain, for example, it would paralyze the person’s breathing. Can medication provided to the lumbar area really reach the head, in sufficient quantities to do some good?
“If you give a drug in the spinal space, it will tend to stay in the spinal space,” said Tony Yaksh of UCSD. That is because the average adult’s 150 milliliters of CSF is not rapidly flowing like a river, he said, but sluggishly sloshing like a backwater bayou. Moreover, any fluid moving towards the brain passes right by the cisterna magna, where CSF drains into the bloodstream at the base of the skull, so drugs would have to avoid that outlet before reaching their target site.
That does not mean that reaching the brain intrathecally is impossible, say proponents of the route. Researchers have found that delivering a quick bolus of drug—say, 10 milliliters over a few minutes—allows some medicine to reach the area surrounding the brain. The fast influx of fluid provides the pressure needed to get the drug moving, and the larger volume boosts the odds of reaching the cranium, Yaksh said. Hydrophilic drugs will have an advantage in reaching the head, whereas lipophilic molecules will stick to lipid membranes along the way, said Tobias Mueller-Bertram of UCSD. Drug developers may also have to worry about side effects from nonspecific drug binding. “It will have an effect on the spinal cord, it will have an effect on the brainstem, it will have an effect on the cortex,” Mueller-Bertram said of drugs delivered via a large intrathecal bolus.
“A surprising amount does get to the brain,” said Andres Lozano of Toronto Western Hospital in Canada. There, it hits another snag: the brain-CSF barrier. It is similar to the blood-brain barrier between blood vessel walls and the brain’s parenchymal space, but less strict. The ependymal cells in the brain-CSF barrier look a bit like beer cans lined up six-pack style, Lozano said. The problem is that drugs may not easily squeeze between the cans. If they make it, they have reached the cortex—but not all pathology happens on the brain’s surface. Even medicines provided directly to the brain may not travel far within its parenchyma, Yaksh said. “There are lots of reasons to be circumspect,” he said.
Drugs might be able to travel the canals of interstitial space that make up 20 percent of the brain, Lozano suggested. Other scientists refute the idea that drugs can diffuse into the parenchyma. “Intrathecal drug delivery to the brain only delivers drug to the surface of the brain,” wrote William Pardridge of the University of California, Los Angeles, in an e-mail to Alzforum. “Transcranial drug delivery to the parenchyma is not effective, owing to the inherent physical limitations of diffusion.” He has argued that drugs in the CSF will be lost into the bloodstream well before they would reach a depth of more than a millimeter or two into the brain’s surface (Pardridge, 2011; Pardridge, 2012).
On this controversial issue, advocates for intrathecal drug delivery beg to differ. For example, in a recent NeuroPhage/MGH study using positron emission tomography to measure drug distribution, the authors claimed that half of the phage drug provided intrathecally to five macaques reached the head within 30 minutes, and 7 percent penetrated the brain tissue between 30 and 60 minutes after injection (Papisov et al., 2012). NeuroPhage has also used quantitative polymerase chain reaction (qPCR) and immunohistochemistry to analyze phage penetration, Gannon told Alzforum. “We did get good brain distribution [by qPCR] … we saw very beautiful cortical and subcortical staining,” she said.
This diffuse delivery system is a good match for the diffuse pathology that occurs in Alzheimer’s, Gannon claimed. It also avoids potential problems with systemic delivery, such as an immune response to the phage or the difficulty the large, 950-nanometer DNA structure would have in traversing the blood-brain barrier. Though the thought of injection into the lower spine might make a patient cringe, intrathecal treatment is less invasive than what NeuroPhage initially considered—multiple injections directly into the brain via holes drilled in the skull.
CSF provides an inhospitable delivery environment in yet another way: The fluid turns over every eight hours, giving medicines limited time in which to enter the tissue. That window might be wide enough for bolus treatments with long-term effects. In the case of the phage, plaques take a long time to form; hence, periodic treatment might be sufficient to keep the brain clear, Gannon believes. Mice retained plaque reduction for at least six weeks post-treatment, and Gannon predicted people might be able to go several months between shots. The company will conduct toxicity studies this year, and hopes to start a Phase 1 trial in 2014.
Like NeuroPhage, Isis realized early on that its treatment would bounce off the blood-brain barrier. “For us to get our drugs into the central nervous system, we need to administer them through some invasive procedure,” said the company’s senior vice president Frank Bennett. Not all of the company’s targets are in the brain. For spinal muscular atrophy (SMA; see next paragraph) and amyotrophic lateral sclerosis (ALS), the spinal cord is the site of the action. For ALS, Isis has completed a Phase 1 safety study of antisense RNA to superoxide dismutase 1 (see ARF related news story). It is now revamping the oligonucleotide to make it more potent and will perform a new Phase 1 study with the improved oligo, Bennett said. No date for that trial is set, but Bennett expects the trial to progress quickly, since the company has gained experience in intrathecal dosing and FDA review. Isis researchers are already working with ALS scientists on a potential antisense therapy for ALS based on C9ORF72 expansions as well (see ARF related news story).
The company’s most advanced neurological treatment is designed to amp up production of "survival of motor neuron," the protein missing in SMA. In a Phase 1 safety trial, 24 children with SMA tolerated a single dose delivered via injection into their lower spines quite well, Bennett said. Isis will present the data from this study at the American Academy of Neurology meeting in San Diego in March. Isis is now conducting a Phase 1b/2a trial, again in 24 children, to try a series of two or three doses for each. Besides confirming safety, trial researchers are also testing motor and muscle function, Bennett said. He expects results by the end of 2013.
In addition, Isis is in late preclinical stages with a treatment for Huntington’s (see ARF related news story on Kordasiewicz et al., 2012). AD and PD oligos are in even earlier stages, Bennett said, but declined to make details public. Reaching the deeper brain areas affected in HD, AD, and PD may require a higher dose of antisense, Bennett said, and Isis is still optimizing delivery for those areas. “I suspect you will see more drugs being administered by intrathecal dosing,” said Bennett, proposing antibodies and enzymes as other possible intrathecal medicines.
Antisense oligos last a long time, and the dose interval will depend on the specific oligo, Bennett said. He predicted the SMA dosing schedule might work out to every six months, and for HD or ALS, perhaps every three months. For infrequent treatments such as Isis and NeuroPhage envision, recipients would not need an implanted pump of the kind routinely used for other chronic conditions. A neurologist, or an anesthesiologist specializing in pain, would have the skill to administer the shot in a procedure much like a lumbar puncture, said Mark Wallace of UCSD. “Most physicians would not feel comfortable doing spinal injections,” he said. Older people may have spine degeneration, making the injection more difficult, he noted. On the upside, the elderly are less susceptible to the headaches that are a side effect of lumbar puncture. The procedure takes 10 minutes, Bennett said.
Pump It Up
Medtronic, in contrast, believes that direct delivery to the brain could open a new market for its implanted pumps. Shafer’s team is working to develop a delivery procedure for anti-amyloid antibodies. Delivered intravenously, this passive immunotherapy has not been reliably effective (see ARF related news story
), but less than 1 percent of the antibody gets to the brain, Shafer noted (Banks et al., 2002). The right pump, she hopes, could boost the local dose.
SynchroMed II pump. Image courtesy of Medtronic
Medtronic’s SynchroMed II pump is already in use to deliver medicines for pain and spasticity to the intrathecal space. The pump measures about the diameter of a drink coaster and is two centimeters thick. Surgeons implant the device, containing pump and reservoir, in the abdomen. A catheter sticks out of one side to deliver the drug to the spine. Doctors and nurses can program the pump with a remote control to mete out a customized dose of medication. They refill it by inserting a syringe through the skin and a self-sealing, silicone membrane on the device. Since 1992, more than 60,000 people have received the pumps, according to Medtronic. The machinery lasts up to seven years before the battery runs down and the pump must be replaced. Company researchers would like to modify the device so the pump itself stays in the abdomen, but the catheter reaches the brain’s ventricles.
In experiments with AD model mice, infusion of antibody to the ventricles reduced plaques at lower antibody concentrations than needed by intravenous delivery. It also resulted in fewer microbleeds (see ARF related news story on Thakker et al., 2009). Shafer thinks the antibody leaches amyloid away from the brain via the CSF drainage system.
Since the rodent work, Medtronic has tested the therapy in 12 aged macaques, which frequently exhibit amyloid-β plaques. For these studies, they used a new antibody to the amino terminus of amyloid developed in conjunction with Antitope of Cambridge, U.K., and Paradise Valley, Arizona. After a pilot with two monkeys, the researchers treated six more with their antibody for three months via pump and catheter, while three received only the vehicle solution from the pump and one underwent a sham operation. The researchers could only measure plaque load after the macaques died. Then, they observed that the amount of plaque was 85 percent lower in the cerebral cortex of treated animals than controls. The team is currently analyzing data from a larger cohort. The researchers did not do a head-to-head comparison of intravenous and pump delivery of their antibody.
Shafer told Alzforum the treatment is nearly ready for clinical trials, but Medtronic is still seeking a partner with neurology expertise, since the company focuses on devices and Antitope on antibody engineering. Between the failures of other antibody trials and pharmaceutical companies’ reluctance to try a new delivery route, it has not yet managed to hook a collaborator.
Take This Gene and Call Me in the Morning
Delivery to the intrathecal space or ventricles gets treatment to the general neighborhood, but some surgeons prefer to drop off their therapy at the precise address where it will do the most good. For Alzheimer’s, one such address is the nucleus basalis of Meynert, the heart of the brain’s cholinergic system. When these neurons get sick, they produce insufficient acetylcholine, and memory worsens. “We need to get the NGF into the specific region of the brain where we want it to act,” said Aisen. “The symptoms of AD are directly related to the cholinergic neurons.” Promoting cell survival in the nucleus basalis ought to strengthen the entire circuit, he said. Aisen directs the Alzheimer’s Disease Cooperative Study, which is spearheading a trial of local NGF gene therapy from Ceregene, Inc., in San Diego, California.
Ceregene has moved gene therapy for Parkinson’s as well as Alzheimer’s into Phase 2. Their vehicle is an adeno-associated virus toting growth factor DNA. To place the gene by intracranial injection, surgeons insert a fine needle that Ceregene designed to keep the virus from sticking to it and to minimize tissue damage. Then they slowly infuse the viral suspension. The therapeutic DNA likely settles into permanent, extrachromosomal structures in the neurons, said Jeffrey Ostrove, president of Ceregene. With the excess genes, the neurons produce and secrete the growth factors, which they can then pick up on their own surface receptors like an autocrine hormone.
After the NGF treatment for AD passed a Phase 1 trial (reviewed in Mandel, 2010), study leaders now have nearly completed recruitment of 50 participants for the current Phase 2, sham surgery-controlled study of people with mild to moderate AD. There has been no shortage of interest from potential enrollees or clinics, Ostrove said. Patients are willing to take a chance on the therapy, and among clinics, every center Ceregene asked to participate has signed on. Trial organizers had to turn away some clinics that volunteered. Ostrove expects to see results in late 2014.
In the case of Parkinson’s disease, the viral particles carry the gene for neurturin, a cousin of GDNF whose physiological role is to prevent cell death. Surgeons deliver this therapeutic to the putamen. In an open-label Phase 1 trial, 12 people safely underwent the treatment and several improved their movement scores on the Unified Parkinson’s Disease Rating Scale (Marks et al., 2008). “It was quite robust,” Ostrove said. In a randomized, double-blinded Phase 2 trial with 58 subjects, people who received the gene therapy versus a sham surgery had similar results after one year (Marks et al., 2010), but the treatment group started to diverge from the sham control set after a few months more. Participants in Parkinson’s trials often exhibit a large placebo effect, because the hope of improvement boosts the dopamine their brains need. The company is currently running a second, 60-person Phase 2 trial, and plans to break the blinding in early 2013, Ostrove said.
A Capsule You Don’t Swallow
Gene therapy is basically a permanent drug delivery system, Cornell’s Crystal said. Whereas Ceregene delivers the instructions for needed growth factors, NsGene Inc., based in Providence, Rhode Island, and Ballerup, Denmark, provides the cells that make those factors. The idea is to implant genetically engineered retinal pigmented epithelial cells inside polymer capsules, whose permeable membranes let nutrients in and therapeutic proteins—NGF for AD and GDNF for PD—out.
The NsGene capsule is tethered to the skull for easy removal. Image courtesy of NsGene, modified from Wahlberg et al., 2012
Surgeons implant the capsules in the nucleus basalis of Meynert or putamen with stereotactic surgery. A tether runs from the capsule to the entry hole in the skull, enabling easy removal. That is an advantage if there is any problem with the treatment, said Lars Wahlberg, president of NsGene. The tether stays under the skin for the duration of the implant.
In the first Phase 1 trial for AD, surgeons working with Wahlberg and Eriksdotter placed capsules in the brains of six people for a year (Eriksdotter-Jönhagen et al., 2012; Wahlberg et al., 2012). All tolerated the implant well, and two improved on the Mini-Mental State Examination (MMSE) and the cognitive portion of the Alzheimer’s Disease Assessment Scale (ADAS-cog).
However, Eriksdotter was disappointed to find that after one year, many of the cells in the capsules were producing much less NGF than at the start; they were, in fact, dying. The researchers tweaked the capsules and tried again in four more patients, with higher NGF output after six months, Wahlberg said. He hopes the PD capsules, currently undergoing toxicity testing, will be in clinical trials by the end of 2013.
Any successful invasive treatment might change what kinds of doctors whom patients would seek out for their care, noted Grigsby in Napa. Most neurologists currently focus on diagnosis, imaging, and prescribing pills, and, where appropriate, physical therapy. It is primarily pain specialists and some neurologists who possess the know-how to install and manage drug delivery pumps for the nervous system. They would be too few in number if many Alzheimer’s patients suddenly wanted a pump. “That will be rate limiting if one of these treatments breaks through,” he said, adding, “it would be a good problem to have.”—Amber Dance.