Gene Therapy--The Next Generation

Vectors in development seek to benefit from earlier trials The gene therapy field resembles a toolbox containing instruments researchers haven't quite mastered, and the number of devices--viral and nonviral vectors--in this toolbox keeps increasing. "There are all these different tools out there," notes A. Dusty Miller, researcher at Fred Hutchinson Cancer Research Center in Seattle. "People are still trying to figure out what tools to use for what diseases." Photo: Richard Lobell Photography BIOBYPASS: During a normal open-heart surgery, a surgeon injects an adenovirus carrying a molecule promoting endothelial cell growth into parts of the heart where a normal bypass can't be effective. Insulin syringes (inset) contain the vectors. Miller and other gene therapy investigators will report on the strengths and weaknesses of a host of vectors at the American Society of Gene Therapy's inaugural meeting in Seattle May 27-31. James M. Wilson, director of the Institute for Human Gene Therapy at the University of Pennsylvania Medical Center in Philadelphia, notes that all vectors have one goal in common: delivering therapeutic material into cells. However, most similarities end there: The earliest retroviral vectors, based on the murine leukemia virus, carried their genetic passengers into dividing cells--but most of the body's cells are nondividing. Adenoviruses, the same kind of viruses that cause the common cold, had a higher rate of delivery--but the immune system soon kicked the foreign material out of the body. Lentiviruses, such as HIV, promise to incorporate their passenger genes into nondividing cells--but some scientists are skeptical of using a deadly virus for therapeutic purposes. Adeno-associated viruses (AAVs), which cause no known disease in humans, hold equal promise for long-term expression--but they have been difficult to produce in mass quantities. Nonviral vectors cause a relatively small immune response and can deliver larger chunks of DNA than most viral vectors. But they are limited by a low rate of gene transfer and poor long-term expression--especially when delivered in vivo. Most nonviral plasmid DNA vectors give only transient gene expression because they do not integrate. Gene therapy researchers must find ways to improve each vector, as well as match vector characteristics with diseases they will most successfully target. Wilson emphasizes that there likely won't be one universal vector but rather a portfolio of specialized ones--some for dividing cells and some for nondividing; some for instances when doctors want short-term expression and some for situations when doctors want long-term expression; then some that deliver large chunks of DNA and some that carry smaller pieces. At this point, researchers have clues about which vectors could be effective against certain diseases, but no definitive answers. SPECIAL DELIVERY: Jefferson Medical College's David Andrews calls one gene therapy approach to treat brain tumors "an ingenious model" because the vector only carries the therapeutic gene into dividing cancer cells. All these vehicles will require further refinement to be effective, Wilson thinks. For example, immune response, which ultimately renders the vector useless, has plagued earlier tests of adenoviral vectors aimed at treating cystic fibrosis. Researchers are countering by removing pieces of DNA that aren't necessary for delivery in hopes that the body won't reject the invaders and their passengers. During Phase I gene therapy trials for cystic fibrosis, Wilson and colleagues tried three different vectors before settling on one pared-down version. That vector has triggered significantly less immune response than the earlier two, he notes. Wilson hopes to get this third vector into the next stage of clinical trials soon: "We hope by July or so to initiate Phase II, with aerosol and repeat administration." The adenovirus's biggest weakness--its lack of staying power--can actually be turned into an advantage if physicians aim the vector at a different target, notes Ronald G. Crystal, Bruce Webster professor of medicine at Cornell University Medical College, New York City, and chief of pulmonary and critical care medicine at New York Hospital-Cornell Medical Center. "They give great expression--but the problem is, they go away after a week or two," Crystal says of adenoviruses. After participating in trials for cystic fibrosis that, like Wilson's, resulted in short-term expression and no long-term genetic incorporation, Crystal asked: "What are the applications where you only need expression for a week?" Angiogenesis, the growth of blood vessels, provided one answer to that question. The process only requires gene expression to start it. Other existing genes control it. In December, Crystal and colleagues began trials of a technique he calls "biobypass." During a normal open-heart surgery, a surgeon uses an insulin syringe to inject an adenovirus carrying a molecule that promotes endothelial cell growth into parts of the heart where a normal bypass can't be effective. "Single molecules can initiate the entire process." Crystal thinks biobypass serves as a good, short-term application for gene therapy, since he suspects genetic treatment for hereditary diseases are years away, and gene therapy for AIDS and cancer are even farther down the road. Crystal remains optimistic--as well as pragmatic. "Each vector has its own usefulness," he says. Despite the criticism that retroviral vectors may be limited because they only work ex vivo and only with dividing cells, uses for them may still exist. Scientists at the National Human Genome Research Institute (NHGRI) and at Johns Hopkins University in Baltimore demonstrated that the vector could deliver a molecule to help rhesus monkeys fight the effects of SIV--the simian equivalent to HIV (R.E. Donahue et al., Nature Medicine, 4:181-6, 1998). "The viruses we have in hand are able to infect and have persistence," remarks Richard A. Morgan, senior investigator at NHGRI's Clinical Gene Therapy Branch. The vector effectively delivered compliments, or "antisense" copies, of the regulatory proteins Tat and Rev. This antisense molecule degrades the SIV RNA and thwarts the ability of the RNA to make Tat and Rev. The monkeys in the study still became infected with SIV, but less virus appeared in their blood and their disease progressed slower. Modifying T cells with antisense molecules differs from one ideal approach to HIV--targeting hematopoietic stem cells, so that they give rise to cells resistant to HIV. "Right now you can do that, but the efficiency is lousy," Morgan says. Until stem cell therapy improves, the antisense approach could serve as an interim measure against HIV-- absorbing the effects of the infection rather than stopping it. However, Morgan cautions that this treatment may be effective only in people in the earlier stages of HIV. Once the immune system becomes too ravaged by the virus, it may be unable to bounce back. Plus, the treatment needs to withstand the transition from animal model to human clinical trial-- a step that has resulted in dashed hope for many gene therapy treatments. Three of the most promising viral vectors--lentivirus, AAV, and herpes--have yet to make that transition. But investigators say existing animal models for these new systems look promising. Luigi Naldini, investigator at Cell Genesys Inc. of Foster City, Calif., reports that a pared-down version of the HIV virus, a member of the retroviral family known as lentivirus, seems to avoid some of the problems that limited first-wave retroviral vectors. "The clear advantage is direct administration and long-term expression." The virus effectively enters nondividing cells, unlike earlier retroviruses, meaning it could be targeted at the brain and liver. But first, researchers must learn how to engineer the virus to be more specific. "You could use an envelope protein from the hepatitis virus to target the liver," Naldini posits. "The other way to target this vector is to control transcription." Scientists have already engineered the vector to be less dangerous than a normal HIV virus by removing six genes that make it deadly. Naldini thinks the vector is safe and anticipates a start to human trials late this year or in early 1999. Bruce A. Bunnell, assistant professor of pediatrics in the division of molecular medicine at Ohio State University in Columbus, agrees that the specificity issue must be resolved. He also thinks the issue of perception must be addressed, since people may be wary of therapeutically using what they see as a deadly virus. Wilson and colleagues evaluated lentiviral vectors for cystic fibrosis and concluded that the system worked well for nondifferentiated cells, while it appeared not to effectively enter and incorporate into differentiated cells (M.J. Goldman et al., Human Gene Therapy, 10:2261-8, 1997). He holds even more optimism for vectors of adeno-associated viruses. "There's a big push here at Penn to get AAV into people on multiple fronts," Wilson reports. "One barrier to that was we couldn't produce enough [AAV]. But we've overcome that with new approaches." Wilson hopes to begin human clinical trials with AAV for hemophilia in late 1998 or early 1999. AAV has two properties that make it a better candidate for treating chronic disease than adenoviral vectors: "It has this ability to efficiently target nondividing cells, which are most of the cells we need to target--muscle, brain, liver, lung--and insert its genome into the genome of the recipient. Because it does that, it's incredibly stable," Wilson explains. "In contrast to anything I've seen, this thing has a way of evading and avoiding an immune response to the cell that it fixes." PREMATURE TRIALS : Thomas Jefferson University Hospital's Andrew Freese notes that many gene therapy trials have failed because the pathogenesis of the diseases they aimed to treat weren't clearly understood. ANTISENSE APPROACH: Experiments in animals show that retroviral vectors can fight equivalent diseases to HIV by delivering complimentary copies of regulatory proteins, explains NHGRI's Richard Morgan. Even the most promising vectors alter 20 percent or less of the targeted cells, notes Andrew Freese, associate professor of neurosurgery at Thomas Jefferson University and director of neurosurgery research at the Philadelphia-based Thomas Jefferson University Hospital. Freese was a member of a team that designed a herpes-based vector years ago. Researchers are now considering the virus as a candidate for gene therapy, based on its ability to deliver chunks of DNA up to 10 times the size of other vectors. However, Freese and his colleagues didn't go clinical with their herpes vector because the vector caused a toxic reaction. "I personally think that some of the gene therapy trials for disease here in the body have been premature, and I think the track record has borne out that belief," he says. "There have been some real problems--immune response problems, long-term expression problems." Freese says those problems resulted from trying therapy for diseases in which the pathogenesis was poorly understood, such as cancer and cystic fibrosis. They also failed because, he thinks, diseases like cancer call for a high percentage of delivery, which simply doesn't exist yet. Freese's strategy is to deliver genetic material against diseases that are well understood, that only require a low percentage of genetic material to be expressed, and whose success can be easily measured. He and colleagues are conducting Phase I trials for treatment of Canavan's disease, a neurological disorder characterized by the degeneration of the myelin sheath, the protective insulation for the brain's nerve cells. A genetic flaw that triggers the overproduction of a toxic chemical in the brain causes Canavan's disease. Freese thinks the odds of success are much better than those for treating brain tumors (see story below). "Instead of destroying cells in the brain--which requires high efficiency--we're going to try to restore function to cells in the brain, which allows for lower efficiency." The treatment for the fatal disease uses a liposome polymer complex as a nonviral delivery system. That vector is much smaller than adenoviruses and causes a far less severe immune reaction. In some instances, it's difficult to choose which vector will work best, Freese admits. For example, he and his colleagues are also looking for ways to restore dopamine production to the brain; the loss of that function is characteristic of Parkinson's disease. "We've developed a number of different vector systems: adenovirus, herpes virus, AAV, nonviral systems, all based on the idea of increasing the production of dopamine in the target tissue." R. Michael Blaese, chief of NHGRI's Clinical Gene Therapy Branch, notes that this uncertainty about how best to use new vectors--and how best to improve existing ones--characterizes a young and developing field. The vector in the first human clinical gene therapy trial in 1990, designed to treat adenosine deaminase (ADA) deficiency, a severe immune system disorder, was based on a simple Moloney murine leukemia virus (R.M. Blaese et al., Science, 270:475-9, 1995). Since then, Blaese and others have improved the vector's efficiency by paring it down and combining various elements taken from different retroviruses to take advantage of favorable characteristics of each (M. Onodera et al., Journal of Virology, 72:1769-74, 1998). "Onodera's work combined parts of several different retroviruses to optimize the final vector for use in a specific set of blood immune system cells," Blaese comments. He also notes that criticism of early trials as unsuccessful may be harsh; one of the trial's two subjects, while not cured, has better immune function, at least partially attributable to gene therapy. "To hold out because something is not curative is bunk," he declares. Blaese also echoes many of his colleagues who say that the concept of gene therapy is too elegant not too work. "In the human genome you've got the information for every gene," he concludes. "You've got to find ways to use that information." The first Phase III human gene therapy clinical trial may provide new information, but it likely won't cure a deadly form of brain cancer, scientists familiar with the trial say. The trial, for treatment of glioblastoma multiforme, recently closed enrollment with 246 subjects. The course of treatment entails several steps. First, surgeons remove the brain tumor. Then they inject a retroviral vector containing a herpes gene into the tumor cavity. The vector doesn't kill cancer cells. But it does make them susceptible to ganciclovir, which the patients receive intravenously twice a day for two weeks, beginning two weeks after the surgery. That drug kills the tumor cells that absorb it by disabling their DNA. The patients finally receive radiation two to three weeks after surgery. One group of patients undergoes surgery and radiation, while another receives the gene therapy and the ganciclovir in addition to surgery and radiation. David Andrews, associate professor of neurosurgery at Jefferson Medical College of Thomas Jefferson University in Philadelphia, calls the treatment "an ingenious model" because the retroviral vector can only carry the herpes gene into the dividing cancer cells. It cannot enter the brain's healthy quiescent cells. However, Andrews's colleague at Jefferson, Andrew Freese, director of neurosurgery research at the hospital, has some doubts. "I do not think that gene therapy is going to work particularly well for malignant brain tumors," Freese counters. "There is no vector system that transfers genes into the brain with a high efficiency." Freese also doubts some scientists' assertion that the dying tumor cells that the therapy does reach will somehow kill off their neighboring cancer cells the vector fails to enter. Gloria Stone, spokesperson for the firm testing the treatment, Novartis Pharmaceuticals Inc. of East Hanover, N.J., notes that during Phase II tests the therapy seemed to slow down disease progression. However, she acknowledges that those treatments didn't completely eliminate the fatal cancer. "What's important is how much it's helping advance the field," she says of the trial. Andrews concurs, adding that he and Jefferson colleagues have begun a Phase I trial for a glioblastoma treatment using a different strategy. -P.S.

Gene Therapy--The Next Generation

Interested in reading more?

Become a Member of