"Chemical restraint" and "symptom management lacking curative actions" are longstanding and frequently voiced criticisms of pharmacotherapeutic interventions in violent individuals. Although these sweeping criticisms are mostly of historical relevance, they do remind us of the trial-and-error approach of the early clinical psychopharmacologists. In recent years, more rational pharmacotherapeutic developments have been based on a more adequate understanding of the drugs' sites and mechanisms of action and the more selective neurochemical systems that are targeted. The essence of these increasingly selective pharmacologic tools for violence research remains the assessment of their behavioral specificity: How specifically and permanently are the excessive aggressive and violent acts controlled without compromising physiologic functions or the remaining behavioral repertoire? The following examination of the evidence for the management of violent individuals with major classes of therapeutic drugs demonstrates that (1) there is no universally effective antiviolence drug, but rather specific agents exert efficacious antiaggressive effects in specific types of violent individuals; (2) acute and long-term management are achieved with different types of agents; and (3) the multiply interacting neurochemical systems mediating aggressive and violent acts offer several targets for pharmacologic interventions.

The management of violent psychiatric patients, whether requiring institutionalization or not, relies primarily on antipsychotic drugs (e.g., Itil and Wadud, 1975; Tupin, 1985; Eichelman, 1986, 1987; Yudofsky et al., 1987). The clinical effectiveness of most typical antipsychotic or neuroleptic drugs is believed to be due primarily to their action on the D2 subtype of brain dopamine receptors (see above; e.g., Carlsson, 1987; Seeman et al., 1976). Under appropriate treatment conditions, these substances effectively control the florid, positive symptoms of schizophrenia (Kellam et al., 1967; Ebert et al., 1977). Social dysfunctions as part of the negative symptomatology may be based on separate neurochemical mechanisms and have not been treated as successfully. Neuroleptic drugs are effective in reducing aggressive and violent behavior not only in patients with an effective pathology, but also in individuals without diagnosed disorder.

During the past 30 years, experimental evidence from studies

on animal aggression has repeatedly demonstrated that phenothiazines, butyrophenones, and thioxanthines as well as the more recent atypical neuroleptics effectively decrease aggressive behavior by isolated mice, by mice or rats exposed to various noxious and provocative stimuli, by rats rendered irritably due to brain injury or neurotoxicity, by brain-stimulated cats, by mice or rats confronting an intruder, and by fish, mice, rats, cats, pigs, and monkeys establishing and maintaining their dominance (see Table 7A; e.g., Yen et al., 1959; Tedeschi et al., 1959a; Horovitz et al., 1963; Lister et al., 1971; Lal et al., 1975; Kido et al., 1967; Olivier and van Dalen, 1982). This substantial literature defines several features of the antiaggressive effects of neuroleptics. One of the most important characteristics of neuroleptics is their potent sedative and tranquilizing effect. The antiaggressive effects of these drugs appear to be part of the tranquil state. Depending on the demand that is placed on the organism, assessments of the motor capacities show that neuroleptics suppress a range of active forms of behavior, but preserve reactive response capacities. A characteristic effect of moderate doses of these drugs is a lengthening of the time to initiate behavioral action, including aggressive acts. Detailed ethological analyses of behavioral sequences reveal a shift from rapid short-duration acts to prolonged postures (e.g., Silverman, 1965a,b, 1966; Schmidt and Apfelbach, 1977). A further increase in neuroleptic dose renders animals incapable of engaging in coordinated motor behavior.

Defensive reactions to brain stimulation or painful environmental events have proven relatively immune to the suppressive effects of neuroleptic drugs (see Table 7A; e.g., Dubinsky and Goldberg, 1971; Andy and Velamati, 1978; Tedeschi et al., 1969). Studies in mice and rats point to potent modulatory effects on escape and defensive responses by treatment with D2 receptor antagonists (Silverman, 1965a,b) and agonists (Puglisi-Allegra and Cabib, 1988; Cabib and Puglisi-Allegra, 1989).

The clinical experiences with neuroleptics confirm their immediate effectiveness in controlling aggressive and violent behavior, which makes these drugs suitable for emergency treatment (e.g., Ananth et al., 1972; Poldinger, 1981; Tardiff, 1982, 1984; Sheard, 1983; Conn and Lion, 1984; Tupin, 1985; Clinton et al., 1987; Itil and Reisberg, 1978). As summarized in Table 7B, neuroleptics potently decrease a wide spectrum of aggressive and violent acts ranging from inpatients and outpatients with amphetamine or endogenous psychosis (e.g., Itil and Wadud, 1975; Yesavage, 1982; Sheard, 1983; Klar and Siever, 1984), to hostile depressive patients

(Overall et al., 1964), mentally retarded patients (e.g., Hacke, 1980; Read and Batchelor, 1986), and opiate addicts or alcoholics during withdrawal (e.g., Itil and Seaman, 1978). The generality of the antiaggressive effects of neuroleptics does not extend to patients with intermittent explosive and epileptic disorders. Neuroleptics are not recommended for use in controlling aggressive behavior in epileptics (Itil, 1981; Sheard, 1983). Overall, antipsychotic drugs, along with benzodiazepines (see below), may be considered relatively safe emergency medications in managing violent outbursts. However, the neuroleptics' impressive efficacy and wide generality, with relatively few nonresponders or contraindications, should not be confused with a specific action on neurobiologic mechanisms mediating aggressive and violent behavior. As a matter of fact, neuroleptics engender serious compromising autonomic and neurologic conditions in the long-term treatment of aggressive, violent, hostile, and explosive individuals that argue against their use as maintenance therapies of choice (e.g., Lion, 1975; Leventhal and Brodie, 1981; Tupin, 1985; Yudofsky et al., 1987; Itil and Reisberg, 1978).

Aggressive and destructive behavior in children, usually diagnosed with conduct disorder, minimal brain damage, or mental retardation, has been successfully decreased with thioridazine and similar piperidylalkylphenothiazines (see Table 7B; e.g., Alderton and Hoddinott, 1964; Alexandris and Lundell, 1968). In addition to the reduction in aggressive behavior, these neuroleptics also had pronounced effects on the level of motor activity, not only decreasing hyperactivity, but often lengthening reaction times and sometimes leading to apathy and drowsiness (e.g., Shaw et al., 1963; Alexandris and Lundell, 1968; Le Vann, 1971; Campbell et al., 1972; Campbell, 1987). These observations parallel those in animal preparation and again question the specificity of these drugs for the treatment of aggressive behavior. Winsberg et al. (1976) recommend psychostimulants and tricyclic antidepressants in preference to antipsychotics as pharmacotherapy for hyperactive and aggressive children.

Very violent psychiatric inmates, 44 of whom had committed violent crimes, were considered to have benefited in terms of their functioning and social adjustment when maintained on fluphenazine deconate, thioridazine, or other long-lasting antipsychotics (Scarnati, 1986). It appears that this most violent psychiatric inmate population requires higher than normal doses of antipsychotic medication, despite the risk of tardive dyskinesias.

Several decades of clinical experience and research history have