CAN AN ANTIBODY GOBBLE UP COCAINE CRAVINGS?

Biochemist Donald W. Landry took up the cocaine challenge at the height of
the crack epidemic, after he heard then-President George Bush ask in a
speech whether science might be able to devise a vaccine that could somehow
render addictive drugs harmless.

A decade later, he may have done it. Using methods seldom applied in drug
abuse research, he built a "catalytic antibody" that eats cocaine in a lab
rat's bloodstream the way Pac-Man gobbles bad guys in a computer maze.

And this spring, the Gaithersburg biotech company MedImmune Inc. will join
forces with him to refine his technique to make an antibody strong enough
to treat cocaine abuse in humans.

Landry may not have found the "magic bullet" in the war on drugs, but he
appears to be as close as anyone. "If he is successful," said Frank Vocci,
director of treatment research and development at the National Institute on
Drug Abuse, "he would actually have an antibody able to reduce cocaine to
an inactive substance as fast as people put it into their bodies."

Success is by no means assured. Scott Koenig, MedImmune's senior vice
president for research, said that while Landry "has shown conceptually that
it can work," the antibody does not yet function rapidly or efficiently
enough to be used in humans.

Koenig said that MedImmune, by engineering and testing thousands of
variants of Landry's antibody, as well as new candidates, expects to show
whether the technique can be commercially viable.

"We have no way of knowing," Koenig said. "But we'll have an answer in
2000." Once proven, however, the treatment would still require some years
of clinical evaluation before it could reach the market, he said.

The potential is enormous. The President's Office of National Drug Control
Policy (ONDCP), which has given Landry $2.8 million in research grants
since 1994, estimates that there are 5.2 million users of cocaine and its
derivatives in the United States, and 3.3 million addicts.

Americans spend $39 billion per year on cocaine, and cocaine's "social
cost" in law enforcement, prisons, rehabilitation, lost wages, medical care
and family violence is another $66 billion, ONDCP estimates.

Landry's--and MedImmune's--goal is to create an antibody that will mop up
any cocaine that might be in a person's bloodstream for about a month. Add
boosters so the person goes four or five months without a reinforcing high,
and the craving goes away.

"And when that happens, rates of abstinence go way up," Landry said, citing
studies he said showed that heroin treatment with both methadone and
counseling produced abstinence rates of 60 percent to 80 percent, compared
with 10 percent to 30 percent for programs relying on counseling alone.

Still, he cautioned, while the antibody effectively "vaccinates" a person
against cocaine for a month, it is not a "vaccine" that causes a person's
body to become permanently immune.

Also, "a vaccine is different for something you want than for something you
don't want," said Alan I. Leshner, director of the National Institute on
Drug Abuse. "People want cocaine," and they can always wait for the vaccine
to wear off.

When Landry, a biochemist at the Columbia University College of Physicians
and Surgeons in New York, began his research, the development of catalytic
antibodies was in its infancy.

Unlike natural antibodies, which are formed within the human body and
neutralize foreign substances by binding to them, a catalytic antibody is
created artificially in the lab and injected into the bloodstream.

Outside the host, scientists are able to engineer the antibody so that it
will not only bind to a foreign substance, but will encourage, or
"catalyze," a chemical reaction that will cause the substance to break up.

Cocaine is a particularly good target for such a technique, Landry
explained, because its molecules could be cleaved easily to produce two
harmless chemicals. A catalytic antibody that encouraged this cleavage
reaction would, in effect, kill cocaine.

First, Landry's team built a synthetic "analogue" that mimicked the
molecular structure of cocaine as it breaks apart. They injected it as a
foreign substance into lab mice, which created natural antibodies to
counter it.

Then they removed the mice's spleens, isolated the cells that produced the
antibodies and cultured them individually. Some of the cells produced
catalytic antibodies that would, in theory, search out cocaine molecules,
induce them to change chemically and break apart, then move on in search of
a new binding partner--the Pac-Man effect.

This suggested that unlike conventional vaccines, which could be
overwhelmed by an abuser who exhausts the supply of antibodies by taking
more and more cocaine, the catalytic antibody binds and kills again and
again. "It is not one-to-one," Landry said.

For this reason, although some conventional cocaine vaccines have reached
clinical trials, they can all be surmounted, Vocci said. These will likely
be used to blunt the initial "rush" when an abstainer falls off the wagon.

Landry's initial report on his research in 1993 suggested that the
antibodies would not only repeatedly bind and neutralize cocaine, but
theoretically could do it so fast that the cocaine would be broken up
before it reached the brain.

This was a marked difference from the battle plan embodied in methadone and
other "blockers" that screen a drug's effects by cementing themselves to
the same pleasure receptors in the brain that the drug stimulates. Instead,
Landry's antibodies would intercept the cocaine before it could get into
these neurological pathways.

In 1998, he showed that this was possible. In the first "overdose"
experiment, Landry's team injected rats with lab-created catalytic
antibodies, then gave them enough cocaine to kill them. The rats didn't die.

In the second "addiction" experiment, the team trained rats to push a lever
to receive a dose of cocaine. When salt water was substituted, the rats
pushed the lever a few times, but soon lost interest.

Then the team injected the rats with antibodies and tried to dose them with
cocaine again. The rats tried the mixture, then ignored it, as if it were
salt water.

Nevertheless, Landry said, "everything we've accomplished so far is merely
proof of principle." Indeed, both Landry and MedImmune agree that Landry's
best-performing antibody binds to individual molecules much too sluggishly
and doesn't change targets fast enough.

The antibody also probably needs to be "humanized," so a patient's body
will not attack it by making an antibody for the antibody. MedImmune will
try to do all this by testing many sample antibodies--both Landry's and
others made by MedImmune--and "tweaking" them to enhance performance.

"I don't want to say it's a no-brainer," said MedImmune's Koenig. "It's
challenging, but we have the experience to give it a good shot. If it can
be done, we'll do it."