C Plenary Lecture I
Avram Goldstein, M.D.
Professor Emeritus of Pharmacology, Stanford University, and Founder and Director, Addiction Research Foundation
My talk this morning will not deal directly with the main theme of the workshop—barriers to entry of young scientists into drug abuse research. Rather, it may contribute indirectly by presenting my overall view of the field, its problems, its accomplishments, and its opportunities.
This is billed as a plenary lecture. The dictionary definition of plenary, from the Latin, is ''full"; but it doesn't say full of what! Full of enthusiasm, I hope, about the new possibilities in this rewarding field of research.
The workshop is about "strategies to raise the profile of substance abuse and alcoholism research." But what, pray tell, is a substance? And whatever a substance may be, the phrase "substance abuse and alcoholism," implies that alcohol is not a substance. These peculiar circumlocutions and euphemisms are political, not scientific, in origin.
Drugs, you see, are bad, and their use is illegal. Alcohol and nicotine are legal, so they can't be drugs. This peculiar distinction resulted in the establishment of two separate institutes, the National Institute on Drug Abuse and the National Institute on Alcoholism and Alcohol Abuse, the latter with its own constituency of alcoholics and recovered alcoholics—not, perish the thought, alcohol addicts and ex-addicts.
Rather than call drug abuse what it is, some well-meaning lexicographer thought to cover the whole field with a new term, "substance abuse," which
could embrace all the addictive drugs without offense to the standard hypocrisy. The problem here is not merely semantic; it reflects a profound gulf between the science and the perceptions of politicians and the public. One way we, as scientists, can raise the profile of drug abuse research among the politicians and public who fund our work is to educate them about the neurobiologic underpinnings of this curious medical, behavioral, societal disease.
I think an essential first step is to call a drug a drug. There are seven families of addictive drugs, which comprise the drug abuse problem. These are, in descending order of societal importance: alcohol, nicotine, cocaine and amphetamines, heroin and other opiates, hallucinogens, cannabis, and caffeine.
The field of drug abuse—substance abuse, whatever you call it—must seem a bewildering morass to the newcomer, especially to the young scientist trained in the cutting-edge techniques of molecular biology. Raising the profile of drug abuse research among scientists means identifying those critical questions that can be attacked by the most up-to-date techniques of molecular genetics and neurobiology as well as by carefully designed, controlled experiments on animal and human behavior.
Drug self-administration is a behavior—a remarkably specific behavior, for just these seven chemical families of drugs are self-administered, and laboratory animals self-administer the same drugs as do humans. All behaviors are determined in part by genetics, in part by environment. And all behaviors are rooted in neurobiology, in the anatomic structures and circuities, and in the neurochemical processes of the brain. There was a time, not so many years ago, when animal and human behaviors could only be studied as phenomena in their own right. But molecular neurobiology has changed that dramatically, so that the behaviors we identify as drug addiction are beginning to be understood at the most basic level. Nevertheless, the fullest understanding of a behavioral disease will require multidisciplinary approaches, and the tendency of some molecular biologists to disdain less reductionist fields of research can be a hindrance to progress.
MULTIPLICITY OF DRUG EFFECTS
Any addictive drug has multiple effects, both psychopharmacologic and toxicologic, many of them unrelated to the addictive property. These comprise the ensemble of actions that make each drug unique, with its own special dangers for the user and for society. These special pharmacologic properties of each drug should be (but rarely are) taken into account in debates and decisions about our national drug policies.
The multiple effects of any psychoactive drug pose a unique difficulty in elucidating the neurochemical basis of addictive behavior. In studying other brain diseases, it is difficult enough to sort out which neurotransmitter or receptor
abnormalities are causes and which are consequences of the disease. Any change in one brain system leads to consequent changes—some adaptive, some maladaptive—in other neural systems. The problem is compounded in the study of an addictive drug, for most of its effects on the brain will be completely unrelated to our main interest—the self-administration behavior itself.
For example, much research has focused on the prominent abstinence syndrome that results from stopping an opiate after prolonged chronic administration. Yet it turns out, as Aghajanian and his colleagues showed, that this is mediated at a different anatomic site (the locus coeruleus) from the mesolimbic pathway that mediates drug reinforcement. Another example: The disruptive effects of alcohol on psychomotor performance are certainly important for society—note this drug's causal involvement in half of all homicides and traffic fatalities—but they may have nothing to do with the addictive properties of alcohol. Multiple drug effects present major challenges to research design, for anything one chooses to measure in the brain after administration of an addictive drug is likely to be irrelevant to the addictive property of that drug. Such experiments have tended to measure whatever the current fad dictates—years ago it was biogenic amines, later it was neuropeptides, then NMDA receptors, then expression of early genes, now nitric oxide, and so on. Of course, we want to learn about any neurochemical change induced by an addictive drug. But what question about addiction does such a finding answer? And is the measured change a cause or a consequence of the addicted state? Subtractive cloning suggests an approach that is too little exploited—the search for differential neurochemical changes that occur only with self-administration but not with passive exposure to the same drug.
Research on the addictive process itself is a conceptual challenge. Addiction proceeds through three distinct phases, requiring study by different means, and offering different possibilities for intervention. We need to learn what neurobiology drives each phase: (1) the initiation of drug addiction, that is, the first self-administration and the pattern of subsequent use of the drug; (2) the full-blown active disease, that is, the compulsive persistence of self-administration; and (3) relapse, that is, the resumption of self-administration after a period of abstinence.
Phase 1: Initiation of Drug Use—Prevention Research
The initiation of drug use follows a typical pattern for all the addictive drugs. At the first exposure, usually in adolescence, some don't like the effects on cognitive function and mood, and they never become users. Some enjoy the drug and become (and remain) casual social users. Others like the drug at the outset, use it repeatedly, and eventually become compulsive, addicted, users.
These individual differences are present before first contact with the drug, therefore they could reflect innate genetic differences in predisposition. The best evidence for that comes from Cloninger's famous cross-adoption studies in the field of alcohol addiction. Also, abnormal responses to the drug, as demonstrated by Schuckit for sons of alcoholics, may provide a direct measure of vulnerability. In general, diagnosing special vulnerability at a young age could be a useful aid in targeting prevention efforts to those most at risk.
A fascinating recent epidemiologic study by Denise Kandel suggests, remarkably, that intrauterine exposure to nicotine during a woman's pregnancy predisposes her daughters to the use of tobacco in adolescence and thereafter. This result points to the need for more extensive animal research, in which prospective controlled experiments can be done (impossible on ethical grounds in humans) to examine the long-term effects (not merely toxicologic) of fetal exposure to each addictive drug.
For the study of genetic predisposition, reliable animal models have been developed—strains of mice and rats that prefer or avoid specific drugs in a free-choice self-administration paradigm. Thus, a basis for sorting out the relevant genes has been laid.
There is growing evidence that the "rewarding" properties of all the addictive drugs are mediated by dopamine release at the terminals of neurons originating in the ventral tegmental area and projecting to nucleus accumbens and basal forebrain structures. In one way or another—but how, exactly, is a current research topic—the various addictive drugs, acting through their own different receptors, impinge on this pathway. One would think it a simple matter to find or synthesize antagonists for the relevant receptors, and thus block the self-administration behavior. However, the "reward pathway" stimulated by addictive drugs plays an essential role in modulating normal, goal-directed behavior. Naltrexone, the specific opioid receptor antagonist, will indeed block the rewarding properties of heroin; but in practice, very few heroin addicts will use it. We do not really understand why; but naltrexone itself is mildly aversive to normal subjects, perhaps because functional opioid receptors are essential to normal mood and feelings of satisfaction.
A similar inference—that one must not perturb the dopaminergic pathway itself—can be drawn from the finding by Caron's group that knockout mice lacking the dopamine transporter (which is the cocaine receptor) are not affected by cocaine and presumably will not self-administer it; but they display grossly aberrant excitatory behavior, due to a persistent excess of synaptic dopamine. This elegant study ushers gene knockout technology into the drug abuse field. Of course, one first has to know which gene to knock out.
An interesting model analogizes an addictive drug to the vector of an infectious disease. Destroy the vector or inactivate it, and it becomes harmless. For scientists with a practical bent, here is opportunity to develop novel methods of prevention, treatment, and relapse prevention. One development, not yet commercialized,
is a catalytic antibody to destroy cocaine more rapidly than the plasma esterases can. Another is active immunization against cocaine, reported by the Koob group in animal experiments. Neutralization approaches, based on binding and inactivating an addictive drug in the blood, have yet to prove their worth; stoichiometry, here, is likely to be a big problem, except for drugs used at very low dosage because of their high potency. A successful precedent is the commercially available antibody developed some years ago by Haber to counteract digoxin toxicity.
Phase 2: The Active Disease and Its Treatment
How best to treat the active disease is highly controversial—meaning, of course, that research has been inadequate. One school of thought holds that becoming drug-free is the only acceptable aim. Another holds that maintenance treatment, employing a long-acting surrogate of the addictive drug, can stabilize the addict's situation and permit social rehabilitation. If research revealed a neurochemical deficit in the abstinent state, whether antecedent to first drug use or consequent to chronic drug exposure, and persisting during abstinence, the maintenance approach to therapy would gain a more solid basis in neuroscience. Here the powerful new PET and functional MRI techniques might be useful. Thus far, although considered for several drug addictions, only opiate addiction is treated by maintenance.
Phase 3: Relapse and Its Prevention
What drives relapse after even long periods of abstinence? One view gives prominence to the abstinence syndrome, which is the immediate consequence of drug withdrawal. In rats, withdrawal is associated with decreased brain reward, that is, elevated thresholds for intracranial electrical self-stimulation. In humans, the feeling of "being sick," of "needing the drug," can drive drug-seeking behavior. But there is a problem. How do we account for relapse that occurs long after the abstinence syndrome has subsided?
Animal models date back to the pioneering work of Wikler, who first showed that conditioned cues play a key role. Current experiments by O'Brien's group demonstrate that drug-related stimuli can evoke craving and associated physiologic changes in abstinent heroin or cocaine addicts. A fascinating question is how a classic psychologic phenomenon—the conditioned cue—can activate the drug-seeking behavior. Whether this is mediated in the dopaminergic pathway itself or elsewhere in the brain is unknown.
Is relapse a consequence of irreversible neurochemical changes caused by prolonged chronic exposure to an addictive drug? Animal experiments show
that after a period of drug self-administration followed by a period of abstinence, resumption of self-administration occurs much more readily than in animals never exposed to the drug. This "priming effect" of the first small dose after prolonged abstinence is thought to play an important part in provoking full-blown relapse. Many research projects are suggested by the speculation that the conditioned cue causes "priming" by release of an endogenous activator of the reward pathway. But which endogenous compound? Acting where?
A recent study in rats by the Nestler group showed that an agonist selective for D1 (but not D2) dopamine receptors can block both the self-administration of cocaine and its priming effect after abstinence. An unexpected finding with possible therapeutic potential, this work exemplifies the value of basic research that is consciously focused on a drug abuse behavior.
It is noteworthy that millions of patients who receive an opiate chronically in hospital are withdrawn and sent on their way without any desire to find and use an opiate again. A difference between them and opiate addicts is, of course, that they do not self-administer. But Lee Robins' follow-up of Americans who became addicted to heroin in Vietnam did deal with classic self-administration behavior; and yet very few of these veterans ever sought out heroin again after their return home. This happy finding of a circumstance in which addicts did not relapse adds an interesting fact to the relapse puzzle.
Quite interesting is the extensive evidence developed in China by Han concerning the neurochemical concomitants of electroacupuncture analgesia. In animal experiments, he observed a frequency-dependent release of enkephalins (at 2 Hz) and dynorphins (at 100 Hz) in spinal cord and other regions of the central nervous system. Together with Terenius in Sweden, he found corresponding changes in human cerebrospinal fluid after transdermal electrical stimulation. Then it was claimed that the same procedure, in heroin addicts, relieved the abstinence syndrome and inhibited craving. These provocative and potentially useful findings have thus far been ignored; clearly, these data need to be replicated and if confirmed, followed up vigorously.
Questions are raised by the surprising recent finding that naltrexone, a specific antagonist at mu opioid receptors, is beneficial in preventing relapse to alcohol addiction. Does this imply that activation of an endogenous opioid system is what drives relapse to alcohol addiction? And would naltrexone be effective in preventing relapse in other addictions, too? Relapse, occurring even after prolonged abstinence, is one of the key mysteries—perhaps the most important one—in drug addiction research. We need to learn the circuitry and the neurochemical steps between the conditioned cue and the relapse behavior to be able to intervene therapeutically. The truth is, whatever the addictive drug, clinicians are able to bring addicts through withdrawal safely and comfortably. Thus, if we could prevent relapse, we could make significant progress toward actually eliminating drug addiction once and for all.
If we understood the neurobiology of this strange behavior—drug abuse and drug addiction—if we understood the mechanisms whereby a drug takes control of a person's thoughts and actions, we might some day be able to intervene more effectively at one or another of the three phases. Especially useful would be techniques for preventing relapse.
There is a special challenge here for the young scientist who is not afraid to pioneer in a confusing area that needs innovative, creative ideas. There is a special appeal for those who would like to see their basic findings translated into therapeutic interventions with major societal impact on the most prevalent disease of our times. Finally, the field of drug abuse is especially ripe for interdisciplinary research, for the establishment of more centers of excellence in which molecular geneticists, neurobiologists, animal behavior experts, clinicians, and social scientists will collaborate closely (and close collaboration is the key to keeping a conscious focus on drug abuse) to increase our understanding of the peculiar compulsive and self-destructive behavior of the drug addict.