For over a century, clinicians have sought to use pharmacological agents to remediate aphasia and/or to help compensate for it, but without much success (Small, 1994). However, in several limited areas, the use of drug treatment as an adjunct to traditional (behavioral) speech therapy has shown some promise. Furthermore, the future for pharmacological and other biological treatments is bright (Small, 2000).

In this brief article, we restrict our attention to the subacute and chronic phases of aphasia, rather than the treatments for acute neurological injury. Much of this work has focused on a class of neurotransmitters, the catecholamines, which occur throughout the brain. Two important catecholamines are dopamine, produced by the substantia nigra, and norepinephrine, produced by the locus coeruleus. Since the catecholamines do not cross the blood-brain barrier, typical therapy involves agents that increase catecholamine concentrations. Dextro-amphetamine is the most popular agent of this sort, acting nonspecifically to increase the concentrations of all the catecholamines at synaptic junctions. In the early studies, a single dose of dextro-amphetamine led to accelerated recovery in a beam-walking task in rats with unilateral motor cortex ablation (Feeney, Gonzalez, and Law, 1982). By contrast, a single dose of haloperidol, a dopamine antagonist, blocked the amphetamine effect. When given alone, haloperidol delays spontaneous recovery, whereas phenoxybenzamine, an a₁-adrenergic antagonist, reproduces the deficits in animals that have recovered. Similar results have now been obtained in several species and in motor and visual systems (Feeney and Sutton, 1987; Feeney, 1997).

The role of antidepressant medications in stroke recovery, including selective serotonin reuptake inhibitors (SSRIs) and the less selective tricyclics, is not straightforward. Neither fluoxetine (an SSRI) nor direct administration of serotonin seems effective in improving motor function in a rat model (Boyeson, Harmon, and Jones, 1994), whereas the tricyclics have produced mixed effects (Boyeson and Harmon, 1993; Boyeson, Harmon, and Jones, 1994).

The role of the inhibitory transmitter g-aminobutyric acid (GABA) has been investigated in several studies. Intracortical infusion of GABA exacerbates the hemiparesis produced by a small motor cortex lesion in rats (Schallert et al., 1992). The short-term administration of diazepam, a benzodiazepine and indirect GABA agonist, can permanently impede sensory cortical recovery. Furthermore, phenobarbital, which may have some GABA agonist effects, also impedes recovery (Hernandez and Holling, 1994).

A number of early studies were conducted with limited success and are summarized in two recent reviews (Small, 1994; Small, 2001). Modern studies of pharmacological treatment of aphasia have focused on neurotransmitter systems, particularly catecholaminergic systems. A number of studies have been conducted, not all well designed. At present, no drug has been adequately shown to help aphasia recovery to the degree that would be necessary to recommend its general use. Several biological approaches that have been tested for aphasia recovery have been shown to be ineffective (e.g., meprobamate, hyperbaric oxygen) or very poorly supported by published results (e.g., amobarbital, selegiline).

Several studies have examined the role of dopamine. Albert et al. (1988) described a case suggesting that the dopamine agonist bromocriptine helped restore speech fluency in a patient with transcortical motor aphasia resulting from stroke. Another case report failed to find a similar benefit in a similar patient (MacLennan et al., 1991). Two additional patients improved in speech fluency but not in other aspects of language function (Gupta and Mlcoch, 1992). Another open-label study suggested some effect in moderate but not severe aphasia (Sabe, Leiguarda, and Starkstein, 1992).

Dextro-amphetamine is perhaps the most widely studied biological treatment for the chronic effects of stroke, including aphasia (Walker-Batson, 2000), yet both its clinical efficacy and mode of action remain unclear (Goldstein, 2000). Nonetheless, evidence from both animal model systems and humans make this a somewhat promising drug for the treatment of aphasia.

In a study of motor rehabilitation from stroke, more than half of a group of 88 elderly patients who had been classified as rehabilitation failures because of poor progress in physical therapy benefited from dextro-amphetamine as an adjunct to physical therapy (Clark and Mankikar, 1979). A double-blind placebo-controlled study replicated this finding (Crisostomo et al., 1988) in eight patients with ischemic stroke.

An early study of aphasia pharmacotherapy with methylphenidate (similar to amphetamine) and chlordiazepoxide (a benzodiazepine) revealed no effects (Darley, Keith, and Sasanuma, 1977). A recent prospective double-blind study of motor recovery with methylphenidate found a significant difference in motor and depression scores on some measures but not others (Grade et al., 1998). Methylphenidate may play a role in the treatment of post-stroke depression (Lazarus et al., 1992).

Walker-Batson et al. (1991) have reported a study of six aphasic patients with ischemic cerebral infarction. Each patient took dextro-amphetamine every 4 days, about an hour prior to a session of speech and language therapy, for a total of ten sessions. When evaluated after this period, the patients performed at significantly above expected levels.

Of potential significance, the studies showing beneficial effects of dextro-amphetamine, that is, the study by Walker-Batson et al. (1991), a motor study by the same group (Walker-Batson et al., 1995), and the other study of motor rehabilitation (Crisostomo et al., 1988), share the common feature of evaluating the drug as an enhancement to behavioral or physical therapy rather than as a monotherapeutic panacea.

Piracetam is a GABA derivative that acts as a nootropic agent on the central nervous system (CNS) and facilitates cholinergic and excitatory amine neurotransmission (Giurgea, Greindl, and Preat, 1983; Vernon and Sorkin, 1991). A large multicenter trial (De Deyn et al., 1997) showed no effect on the primary outcome measure of neurological status at 4 weeks. Another study showed improvement at 12 weeks that was no longer present at 24 weeks (Enderby et al., 1994). A later study (Huber et al., 1997) showed that improvement occurred on only one subtest (written language) of a large battery.

A crucial issue that must be addressed as part of aphasia rehabilitation is depression, since it can adversely affect language recovery. Following stroke, patients with depression have more cognitive impairment than patients with comparable lesions but no depression (Downhill and Robinson, 1994). Furthermore, in stroke patients matched for severity and lesion localization, patients with depression experience a poorer recovery than their nondepressed counterparts in functional status and cognitive performance (Morris, Raphael, and Robinson, 1992).

Growth factors have been advocated for a variety of purposes in the treatment of stroke, particularly in the acute phase of ischemic brain injury (Zhang et al., 1999), but also as neuroprotective agents useful in the chronic phase of recovery from brain injury (Olson et al., 1994). Gene transfer into the CNS might ultimately play a role in delivering trophins or other agents to damaged brain areas and thus to help stimulate recovery or increased synaptic connectivity.

Neural stem cells are multipotential precursors to neurons and glia. Attempts have been made to induce differentiation into neurons and glial cells, and further into specific types of such cells. Specifically with regard to stroke and the treatment of cortical lesions, fetal neocortical cells have been successfully transplanted into the site of cortical lesions (Johansson, 2000), and have even been shown to migrate selectively into areas of experimental cell death (Macklis, 1993; Snyder et al., 1997).

One important consequence of this research into the pharmacology of aphasia is the realization that drugs are not only potential therapeutic adjuncts but can also serve as inhibitors of successful recovery. The first study of this type, by Porch and colleagues (1985), showed that patients taking certain medicines performed more poorly on an aphasia battery than those who were not taking medicines.

In a formal retrospective (chart review) study, patients with motor deficits after stroke were divided into one group taking a number of specific drugs at the time of stroke (clonidine, prazosin, any dopamine receptor antagonist [e.g., neuroleptics], benzodiazepines, phenytoin, or phenobarbital) and another group that was not (Goldstein, 1995). Statistical analysis revealed that whereas patient demographics and stroke severity were similar between groups, motor recovery time was significantly shorter in the patients who were not taking one of these drugs.

This work has profound relevance to aphasia rehabilitation. To maximize functional recovery, it is important not only to ensure adequate behavioral treatment, but also to ensure the appropriate neurobiological substrate for this treatment (or, more concretely, to ensure that this substrate is not pharmacologically inhibited from responding to the therapy). It is thus advisable for patients in aphasia therapy to avoid drugs that might interfere with catecholaminergic or GABAergic function or that are thought to delay recovery by empirical study.

Current knowledge suggests a potential beneficial effect of increased CNS catecholamines on human motor recovery and aphasia rehabilitation. Although pharmacotherapy cannot be used as a replacement for speech and language therapy, it might play a role as an adjunct, and other biological therapies, such as cell transplantation, might play a role in concert with carefully designed, adaptive learning approaches. In the published cases where pharmacotherapy improved language functioning in people with aphasia, it was used adjunctively, not alone. It is very likely that pharmacotherapy has a valuable role to play as an adjunct to behavioral rehabilitation to decrease performance variability and to improve mean performance in patients with mild to moderate language dysfunction from cerebral infarctions.