Gene Therapy Targets Canavan Disease

The Canavan trial signals a new phase in a 10-year offensive that gene therapy researchers have waged against neurodegenerative disorders.

Sep 17, 2001
Douglas Steinberg
Days after learning that their project would survive, thanks to a newly awarded $1.8 million National Institutes of Health grant, researchers at Jefferson Medical College in Philadelphia made medical history on June 5. During a three-hour operation, neurosurgeon Andrew Freese cut six small holes into a girl's skull. Through hair-thin catheters, he then infused areas of her brain with 90 billion virus particles that are expected to infect neurons and express a normal human gene that she lacks. By all accounts, Lindsay Karlin, a 6-year-old afflicted with Canavan disease, thereby became the first person to have recombinant viruses injected into her brain to treat an illness other than cancer.

The Canavan trial signals a new phase in a 10-year offensive that gene therapy researchers have waged against neurodegenerative disorders. Previously limited mostly to cell-culture and animal experiments, the scientists are now poised or starting to take their protocols and reagents to the clinic.

Courtesy of Larisa Bilaniuk, Department of Radiology, Children's Hospital of Pennsylvania

A sagittal nuclear magnetic resonance image taken from a normal patient compared to a patient with Canavan disease. Brain tissue atrophy is noticeable in the Canavan patient, along with an enlargement of the ventricular spaces, and thinning of the corpus callosum and other fiber tracts.

In April, Mark H. Tuszynski, a neuroscientist at the University of California, San Diego, initiated an eight-subject trial in which he infects cultured fibroblasts with a recombinant virus and then injects the fibroblasts into Alzheimer's brains. Last summer, NIH's Recombinant DNA Advisory Committee (RAC) held a public meeting on a gene-therapy protocol for Parkinson's disease. In October 2000, a workshop at the National Institute of Neurological Disorders and Stroke (NINDS) hailed advances in gene therapy research on the lysosomal storage diseases, which include Tay-Sachs. And in July, five NIH institutes issued a Request for Applications (RFA-NS-02-007) to accelerate clinical use of gene-transfer methods developed by basic scientists.

Giovanna M. Spinella, NINDS program director in neurogenetics and neurodevelopment, predicts that within the next three to five years, additional clinical trials applying gene therapy to neurodegenerative diseases will occur. For these later efforts, the Canavan trial, if successful, will prove that viral-based gene therapy can, in principle, prevent brain deterioration in humans. The trial could also reveal pitfalls associated with gene therapy targeted at the brain.

Why Canavan Disease?

Canavan disease afflicts an estimated 200 to 300 Americans, mainly from an Ashkenazi Jewish background. The disease leaves children in a vegetative state, killing most by age 10. Canavan is caused by an enzyme deficiency stemming from mutations in the gene encoding aspartoacylase (ASPA).1 ASPA normally hydrolyzes N-acetyl-aspartate (NAA), an acid thought to play roles in lipogenesis and osmotic regulation. Without ASPA, NAA accumulates in the brain. Though Canavan pathology isn't well understood, excess NAA appears to damage, if not destroy, neurons and the oligodendrocytes that make the myelin sheaths around neurons' axons. Demyelination, in turn, hinders the conduction of nerve signals.

The roots of the current Canavan trial date back to 1995, when Lindsay's parents approached Matthew J. During, then at Yale University and now director of Jefferson's CNS Gene Therapy Center, about treating their daughter. Lindsay had just been diagnosed with Canavan, and her family was impressed with a gene therapy study on Parkinsonian rats conducted by During's lab.

During offers multiple reasons for his decision to tackle this obscure condition: Unlike many genetic disorders, Canavan affects only the central nervous system; the involved gene had just been cloned and its protein product had a straightforward function; and gene-therapy efficacy could be gauged by measuring NAA levels in the brain through spectroscopy. In addition, During recalls, "There were no alternative therapies, and [the disease] was all progressively downhill. There was also a potential window of intervention--the children were essentially normal at birth."

Treated Children Inch Forward

The first gene-therapy trial on Canavan children occurred in During's native New Zealand in 1996. This trial minimized risk at the expense of efficacy by encasing the ASPA gene in liposomes, not viruses, and by injecting them into the brain's accessible ventricles, not into neural tissue. Lindsay, one of two subjects, showed limited improvement.2 Nevertheless, the trial provoked ethical and procedural controversies in New Zealand and the United States.3

In 1998, the team conducted a similar trial on 14 children at Yale. The subjects' responses varied greatly, partly because of their widely ranging ages, according to Paola Leone, who entered the gene-therapy field as During's postdoc and is the current study's principal research investigator. But she adds that children who experienced falling NAA levels also improved clinically--for one, they could see better--and their brains exhibited mild increases in myelination.

During these early trials, the researchers knew that viruses delivered genes more efficiently than liposomes did. Their vehicle of choice was nonpathogenic adeno-associated virus (AAV), which mainly infects neurons and can't replicate by itself. The problem was that adenovirus, which can be toxic and immunogenic, was needed to package AAV particles, and adenovirus couldn't easily be purified away from AAV. Only when an adenovirus-free packaging method became available in 19984 did an AAV-based clinical trial become possible. Leone, Freese, and During then began preparing to apply to RAC, FDA, and institutional review boards. Leone had to go to St. Kitts to set up studies on infant primates, which are scarce in the United States.

The final hurdle was finding money for this expensive venture. Private funding, often from families of Canavan children, had underwritten earlier trials but was no longer sufficient. After a close call--support for salaries was due to expire in May--Leone's grant application to NINDS was approved June 1.

By mid-July, two other children had undergone an operation like Lindsay's, and researchers examined them a month later. Leone reports that they all improved on various neurological and psychometric measures. Lindsay, for example, showed an increased interest in her surroundings, and she vocalized and moved her hands more. "All the neurologists and physiotherapists were amazed," Leone recalls.

The crucial test measured concentrations of NAA, Canavan's destructive agent. Magnetic resonance spectroscopy on four brain regions disclosed no change in Lindsay's NAA levels, while 4-year-old Max Randell displayed an NAA decrease in only one region. Encouragingly, NAA levels dropped in all four regions of 5-year-old Jacob Sontag's brain, with two areas declining 21 percent. Still, Jacob's lowest post-surgery NAA concentration, 7.9 millimolar (mM), remained above the normal range of 5.5 to 7.5 mM. Noting that the three-year, 15-subject study has just begun, Leone stresses that these early findings must be interpreted cautiously.

Trial Still Controversial

The current Canavan trial, like previous ones, is controversial. One reason is that gene therapy wasn't tested first on an animal model of the disease. None existed. Such a model, an ASPA knockout mouse, was just developed by Reuben Matalon, a pediatrics professor at the University of Texas Medical Branch in Galveston.5 He and Ronald J. Mandel, a University of Florida neuroscientist, are treating the mice with gene therapy. Matalon and Evan Y. Snyder, an assistant neurology professor at Harvard Medical School, plan to combine stem-cell and gene-therapy treatments. Leone has been working on another animal model obtained from Japanese scientists. Known as the "tremor rat" because of the behavioral effect of its spontaneously arising ASPA deletion, it was discovered in an animal colony.

Another concern about the Canavan trial relates to where the enzyme and substrate are located. Normally, neurons synthesize NAA and release it to surrounding oligodendrocytes, which make ASPA, notes Morris H. Baslow, a neurochemist at the Nathan S. Kline Institute for Psychiatric Research in Orangeburg, N.Y. ASPA then hydrolyzes NAA, presumably on the oligodendrocyte's surface or after NAA is taken up by the cell. With gene therapy, NAA and ASPA are both being made in the same neuron. "You may have a problem because the neuron, which is normally not supposed to have the enzyme, will have it," explains Baslow. Fears that this situation could cause NAA overdepletion might be allayed by the recent discovery of a 3-year-old whose brain lacks NAA.6 Baslow says that the child, while having some physical problems, nevertheless sees, hears, and walks.

A third objection to the Canavan trial is that it supplies localized treatment, but the disease affects the entire brain. No evidence exists that AAV-infected neurons secrete recombinant ASPA so that it can spread elsewhere. This means that ASPA might hydrolyze NAA only where the virus is injected. Because spectroscopy is so time-consuming, researchers aren't testing this possibility by measuring NAA levels throughout patients' brains. Instead, says Leone, the study is using magnetic resonance imaging to monitor the whole brain for signs of decreased atrophy and increased myelination.

The Canavan trial also poses questions common to all gene therapy. How long will the transgene function in people? Researchers aren't sure. They don't even know definitively whether AAV transgenes integrate into a neuron's genome or remain episomal. More pre-clinical studies might help answer these questions. Yet, as the NINDS' Spinella observes, "At some point, you have to do some things in the human in order to know whether or not you're heading in the right direction."

Douglas Steinberg ( is a freelance writer in New York.
1. R. Kaul et al., "Canavan disease: mutations among Jewish and non-Jewish patients," American Journal of Human Genetics, 55[1]:34-41, 1994.

2. P. Leone et al., "Aspartoacylase gene transfer to the mammalian central nervous system with therapeutic implications for Canavan disease," Annals of Neurology, 48[1]: 27-38, 2000.

3. S. Coney, "Gene trial causes ethical storm in New Zealand," Lancet, 347:1759, 1996.

4. D. Grimm et al., "Novel tools for production and purification of recombinant adenoassociated virus vectors," Human Gene Therapy, 9:2745-60, 1998.

5. R. Matalon et al., "Knock-out mouse for Canavan disease: a model for gene transfer to the central nervous system," Journal of Gene Medicine, 2[3]:165-75, 2000.

6. E. Martin et al., "Absence of N-acetylaspartate in the human brain: impact on neurospectroscopy?" Annals of Neurology, 49[4]:518-21, April 2001.