Researchers Get Dose of Reality
As Logistics Stymie Gene Therapy
By ROBERT LANGRETH and STEPHEN D. MOORE
Staff Reporters of THE WALL STREET JOURNAL
Eight years ago, Swiss drug maker Sandoz AG was dazzled by a tiny U.S.
biotechnology firm doing promising work in the new field of gene therapy.
Genetic Therapy Inc. and government scientist French Anderson, its
co-founder, claimed that injecting healthy genes into the body to replace
damaged or defective ones would revolutionize medicine by letting doctors
attack the causes of diseases such as cancer and cystic fibrosis, rather
than just the symptoms.
Sandoz, which would merge with Swiss rival Ciba-Geigy AG in 1996 to form
Novartis AG, bought into this bold new vision. It began amassing GTI shares
in 1991; in 1995, it bought the Gaithersburg, Md., company outright,
bringing its total investment in GTI to more than $300 million. By 1997,
Sandoz, looking for new technologies to fuel its profits, also had invested
$548 million to acquire SyStemix Inc., which was developing a related
process.
These days, however, gene therapy looks unlikely to pay off in a big way
anytime soon. And Novartis is scrambling to salvage its nearly $1 billion
investment in the technology. All but one of the human gene-therapy testing
programs once under way at GTI and SyStemix have been scrapped or
temporarily halted. And none of the hundreds of human studies conducted
elsewhere around the globe so far have offered clear proof that gene therapy
works in any significant way to fight disease.
"This has been an overhyped area that has failed to reach its promise for
very obvious technical reasons," says George Poste, a top research executive
at British pharmaceutical maker SmithKline Beecham PLC.
'We Were Naive'
Even Dr. Anderson has tempered his enthusiasm. "We were naive to think we
could revolutionize medicine in 10 years," says the researcher, who now
doubts that major gene-therapy drugs will reach the market before 2004 or
2005.
Indeed, the sudden death last month of a patient undergoing experimental
treatment for a rare liver disorder at the University of Pennsylvania -- the
first death apparently linked to gene therapy -- has raised questions about
whether scientists have been in too big a hurry to test such therapy on
humans. As a precaution, federal regulators since have ordered U.S. drug
maker Schering-Plough Corp., which wasn't involved in the University of
Pennsylvania research project, to temporarily stop enrolling new patients in
certain gene-therapy trials for cancer.
Back in 1990, though, gene therapy was generating enormous excitement. Dr.
Anderson and his colleagues at the National Institutes of Health treated a
young girl with a rare and potentially fatal immune disorder called ADA
deficiency, or "boy-in-the-bubble" disease. Dozens of hospitals and medical
schools sought government approval for human trials for other illnesses.
Boosters went so far as to predict that the first gene therapies would hit
the market as early as 1996.
Sandoz, one of the pharmaceutical industry's most aggressive investors in
the new field, had high hopes of using gene therapy to treat the deadliest
form of brain cancer, glioblastoma, and other cancers. But in 1998, Novartis
quietly abandoned its flagship gene-therapy project after a large-scale
human test failed to prove that the therapy was effective. And, earlier this
year, the drug maker undertook a painful overhaul of its entire gene-therapy
research program. Now, Michael Perry, who has run both GTI and SyStemix
since their operations were combined last fall, says no new human tests are
planned until a host of technical issues are resolved.
A Crucial 24 Months
"Novartis is not proud of it right now," Dr. Anderson says of the program.
"Something has got to show up on the radar screen in the next 24 months or
some really serious decisions are going to be made" says the researcher, who
serves on GTI's board and is now a University of Southern California medical
school professor.
Even so, Novartis Chief Executive Daniel Vasella says his company remains
committed to the field. "There isn't a single major drug on the market that
hasn't been close to the grave at some point," he says. If gene therapy
ultimately fails to produce a major product, "I'd still say fair enough.
That's our business. If you don't spend big money and take big risks, you
shouldn't be in the pharmaceutical industry," adds Mr. Vasella, who notes
that gene therapy accounts for a small fraction of Novartis's annual
health-care research budget of more than $2 billion.
From the beginning, the therapy's main difficulty has been a logistical one:
how to deliver enough healthy genes to the appropriate site and get them to
stay there long enough to cure or alleviate a disease. For example,
researchers estimate that to cure cystic fibrosis, a hereditary respiratory
disorder, new genes would have to be permanently placed in a billion
defective cells in the lung's airways. Researchers often use genetically
engineered viruses as vehicles for such genes. For example, an adenovirus
was used as the carrier in the University of Pennsylvania trial in which the
patient died. But the technique has its drawbacks: Sometimes a patient's
immune system, wary of foreign invaders, tries to fight off the viral
vehicles. To be sure, gene-therapy researchers recently have made progress
in some areas, such as treating heart disease or hemophilia. But overall,
the therapy's short-term applications seem narrower than they did a few
years ago.
Confident of a Breakthrough
After Sandoz and GTI began working together in November 1991, GTI scientists
were confident of achieving a breakthrough. The feeling was: "We are going
to do something very important here," says Alan McClelland, a former GTI
research manager now at rival Avigen Inc.
GTI's showcase gene-therapy effort, aimed at treating glioblastoma, was
conceived by colleagues of Dr. Anderson at NIH. It involved inserting
so-called suicide genes into the DNA of tumor cells to make them sensitive
to the antiviral drug ganciclovir; doctors could then give patients the
antiviral to kill the tumor cells.
The first human test was conducted in the early 1990s at the NIH, with GTI
as a partner. Initial results seemed highly encouraging. The first patient's
tumor shrank dramatically following the treatment. Although distant tumors
eventually killed him eight months later, the original tumor never came
back. A second patient showed no response. A third patient's tumor also
shrank significantly, though only temporarily. Overall, the tumors of six of
the 15 patients studied showed some shrinking in response to the therapy; in
one case, the tumor disappeared entirely and the patient seemed to fully
recover, a rare outcome in glioblastoma.
GTI executives didn't wait for all the data to come in before they began
planning a second test on a few dozen patients at the University of
Washington and elsewhere. But even as these new tests were launched, Edward
Oldfield, an NIH neurosurgeon, was analyzing the results of the first trial,
which he had led, and he was growing increasingly skeptical of the therapy.
It seemed the new genes penetrated only a tiny fraction of tumor cells in
the immediate vicinity of the injection. While this killed nearby cancer
cells, a phenomenon dubbed "the bystander effect," it seemed to have no
impact on cells farther afield. And Dr. Oldfield wasn't convinced that the
one patient's recovery was fully a result of the gene treatment.
"We didn't think it had enough activity to be meaningful. Based on what we
had seen, I saw no hope of distributing the gene where it needed to go,"
says Dr. Oldfield, a surgical chief at the National Institute of
Neurological Diseases and Stroke.
By 1994, Sandoz had decided to expand its collaboration with GTI to include
the glioblastoma work. It started its own small human test in Europe. The
initial results were mixed, and oncologist Athos Gianella-Borradori, the
Sandoz official helping coordinate the trial, says he had doubts the
treatment would work. He says he told his superiors at Sandoz and officials
at GTI of his reservations, aware they might be interpreted "as a betrayal
of our cause."
He had another concern: By late 1995, several prominent gene-therapy trials
elsewhere had failed, leading some to question the validity of the entire
approach. Dr. Gianella-Borradori, who left Sandoz a few years ago and now
works at Bavarian Nordic GmbH in Germany, says he didn't want another such
failure.
That same year, an NIH committee set up to review gene therapy concluded
that problems remained "in all basic aspects" of the technology. Top
researchers were reporting disappointing results. Scientists at the
University of North Carolina in Chapel Hill disclosed that a trial for
cystic fibrosis produced side effects and didn't alleviate the disease.
Another trial, at Ohio State University, found that a muscular-dystrophy
treatment didn't improve muscle strength.
Still, GTI and Sandoz forged ahead. Although gene therapy affected only a
limited number of cancer cells directly in the glioblastoma project, Sandoz
was impressed by the fact that a handful of patients in the various trials
appeared to be among the rare long-term survivors of the disease.
Researchers theorized that the new genes had triggered an immune response
against the cancer. A neurosurgeon who led the University of Washington
trial says he was so taken by his results that he flew to Switzerland, to
urge Sandoz to continue the program.
Comparative Study
In 1996, Sandoz, soon to become Novartis, launched a large trial that would
compare the effectiveness of gene therapy combined with standard treatment
to that of standard treatment alone. It would be by far the largest
gene-therapy trial to date, with more than 200 patients.
Novartis development chief Joerg Reinhardt says the company was fully aware
of the high risk of failure. But given the reports of patients who seemed to
respond to treatment, Novartis felt larger brain-cancer trials were in
order. The test began in the summer of 1996 at hospitals in Europe and the
U.S.
Meanwhile, GTI's gene-therapy treatments for other diseases were running
into problems. When Sandoz bought it out in August 1995, GTI was involved in
early-stage human tests of therapies for cystic fibrosis, breast cancer and
several rare genetic diseases. But these efforts were halted, one by one, as
it became clear that existing gene-delivery technology couldn't deliver
enough genes to do the job. Novartis says Sandoz wasn't counting on any of
these preliminary tests to succeed when it bought GTI.
Different Concepts of Time
As project after project stumbled, Novartis grew worried that GTI's approach
was too scattered. There were "too many things going on, and not enough
focus," says Paul Herrling, Novartis's research director. Novartis attempted
to impress its deliberate decision-making style on GTI, but ended up
frustrating some researchers there. A running joke was that at GTI the clock
ticked in days, but at Novartis it ticked in quarters. Dr. Herrling says
there were clashes. "We had no experience with this biotechnology; they had
no experience with drug development."
In another setback, the U.S. Federal Trade Commission, reviewing the
proposed Sandoz-Ciba merger in 1996, demanded -- as a condition for
approving the deal -- that rights to key GTI patents be offered to rival
companies. Sales of gene-therapy drugs, the FTC predicted, could reach $45
billion a year by 2010 -- a figure that today seems highly unrealistic.
By early 1998, preliminary results of the large-scale comparative trials on
brain-cancer patients were rolling in. They didn't look good. Patients who
underwent gene therapy didn't live longer than those receiving standard
treatment. For Novartis, it was the final straw.
Returning a Verdict
In mid-1998, the panel of scientists that reviews Novartis's research each
year took a hard look at gene therapy. Its verdict: Improved gene-delivery
methods were needed in order to give human tests a better chance of success.
Soon thereafter, Novartis combined GTI and SyStemix. Many GTI officials were
shocked when GTI bore the brunt of the cuts, which included most of GTI's
senior managers for drug development. Roughly 90 employees, or about 20% of
the two units' combined staffs, were laid off, the vast majority of them at
GTI. Mr. Perry, a Novartis insider who had been running SyStemix for more
than a year, was chosen to head the combined operation from SyStemix's Palo
Alto, Calif., laboratory. The decision, Mr. Perry says, didn't reflect on
the quality of GTI's researchers.
Dr. Anderson sees it differently. GTI was "charging off in various
directions" and not always conforming to Novartis's corporate goals.
Novartis had tried to rein in the company, but "wasn't successful."
Novartis has since scaled back or eliminated many gene-therapy research
projects. "It is not on my high priority list to go into the clinic now,"
Mr. Perry says. But he says he is "positive" that gene therapy will play a
major role in medicine, and that Novartis's past experience will give it an
edge. "Certainly, there is pressure" to produce a return on Novartis's
investment, he says. "It is a question of when, not if."

Larry Fleming
[log in to unmask]
http://homepages.msn.com/HobbyCt/flemco