Textbook of Traumatic Brain Injury - part 9 potx

612 TEXTBOOK OF TRAUMATIC BRAIN INJURY Baclofen Usually after sudden withdrawal Ketamine Also produces hallucinations, crying, changes in body image, and delirium Levodopa Often after dosage increase Pentazocine During treatment Propranolol See above a Digitalis See above a Paranoia Asparaginase May be common Bromocriptine Not dose related Corticosteroids, ACTH See above a Amphetamines Even at low doses Indomethacin Especially in elderly patients Propranolol At any dose Sulindac Reported in a few patients Aggression Bromocriptine Not dose related; may persist Tranquilizers and hypnotics A release phenomenon Levodopa See above a Phenelzine May be separate from mania Digitalis See above a Carbamazepine In children and adolescents Note. ACTH=adrenocorticotropic hormone. a Same comments apply as for previous reactions on this drug. Source. Reprinted from Dubovsky SL: “Psychopharmacological Treatment in Neuropsychiatry,” in The American Psychiatric Press Textbook of Neu- ropsychiatry, 2nd Edition. Washington, DC, American Psychiatric Press, 1991, pp 694–695. Used with permission. TABLE 34–2. General principles of pharmacotherapy for patients with traumatic brain injuries Start low, go slow Initiate treatment at doses lower than those used in patients without brain injuries, and raise doses more slowly than in patients without brain injuries. Adequate therapeutic trial Although patients with brain injuries may be more sensitive to the side effects of many medications, standard doses of such medication may be needed to treat adequately the neuropsychiatric problems of these patients. Continuous reassessment The need for continued treatment should be reassessed in an ongoing fashion, and dose reduction or medication discontinuation should be attempted after achieving remission of target symptoms. Spontaneous recovery occurs, and in such circumstances continued pharmacotherapy is unnecessary. Monitor drug–drug interactions Because patients with brain injuries are often sensitive to medication side effects and because they may require treatment with several medications, it is essential to be aware of and to monitor these patients for possible drug–drug interactions. Augmentation A patient experiencing a partial response to treatment with a single agent may benefit from augmentation of that treatment with a second agent that has a different mechanism of action. Augmentation of partial responses is preferable to switching to an agent with the same pharmacological profile as that producing the partial response. Symptom intensification If targeted psychiatric symptoms worsen soon after initiation of pharmacotherapy, lower the dose of the medication; if symptom intensification persists, discontinue the medication entirely. TABLE 34–1. Psychiatric side effects of neurological drugs (continued) Symptom Medications Comments Psychopharmacology 613 multiple anticonvulsants. Additionally, alterations of pharmacodynamics may develop during the administra- tion of medications with additive or synergistic clinical ef- fects (i.e., increased sedative effects when several sedating medications are administered simultaneously). If a patient does not respond favorably to the initial medication prescribed, several alternatives are available. If there has been no response, changing to a medication with a different mechanism of action is suggested, much as is done in the treatment of depressed patients without brain injury. If there has been a partial response to the initial medication, addition of another medication may be useful. The selection of a second supplementary or augmenting medication should be based on consideration of the possi- ble complementary or contrary mechanisms of action of such agents, the individual and combined side-effect pro- files of the initial and secondary agents, and their potential pharmacokinetic and pharmacodynamic interactions. Although individuals after TBI may experience multi- ple concurrent neuropsychiatric symptoms (i.e., depressed mood, irritability, poor attention, fatigue, and sleep distur- bances), suggesting a single psychiatric diagnosis such as major depression, we have found that some of these symp- toms often persist despite treatment of the apparent “diag- nosis.” In other words, diagnostic parsimony should be sought but may not always be the best or most accurate di- agnostic approach in this population. For this reason, the neuropsychiatric approach of evaluating and monitoring individual symptoms is necessary and differs from the usual syndromal approach of the present conventional psychiat- ric paradigm. Several medications may be required to alle- viate several distinct symptoms after TBI, although it is prudent to initiate such treatments one at a time to deter- mine the efficacy and side effects of each prescribed drug. Studies of the effects of psychotropic medications in patients with TBI are few, and rigorous double-blind placebo-controlled studies are rare (see Arciniegas et al. 2000b). The recommendations contained in this chapter represent a synthesis of the available treatment literature in TBI, extensions of the known uses of these medications in phenotypically similar non–brain-injured psychiatric populations of patients with other types of brain injuries (e.g., stroke and multiple sclerosis), and the opinion of the authors of this chapter. We recognize that the pathophys- iology of these symptoms may differ in patients with TBI, and, thus, generalization of response to treatment seen in the context of other forms of brain dysfunction (e.g., stroke and Alzheimer’s disease) to TBI may not always be valid. Where there are treatment studies in the TBI pop- ulation to offer guidance regarding medication treat- ments, these are noted and referenced for further consid- eration by interested readers. Neurotransmitter Changes After TBI Neuropsychiatric symptoms arising from penetrating or focal trauma, or both, are often understandable given the functions known to be subserved by the site of injury (e.g., behavioral disinhibition and aggression after bilateral orb- itofrontal contusion), but the etiology of cognitive impair- ments after nonpenetrating (or “nonfocal”) injuries is rela- tively less well understood. Cytotoxic processes such as calcium and magnesium dysregulation, free radical– induced injury, neurotransmitter (especially glutamate and cholinergic) excitotoxicity, and diffuse axonal injury because of straining and shearing biomechanical forces may be produced by nonpenetrating injuries (see Chapter 2, Neuropathology, and Chapter 39, Pharmacotherapy of Prevention, as well as McIntosh et al. 1999 and Halliday 1999 for review). These processes functionally and struc- turally disrupt the neural networks, subserving many crit- ical neuropsychiatric functions (i.e., cognition, emotion, and behavior). Although TBI-induced glutamatergic dis- turbances are almost certainly important in the genesis of injury to areas critical to neuropsychiatric function (see Obrenovitch and Urenjak 1997 for review), there are at present no therapies available to directly ameliorate neu- ropsychiatric problems predicated on disturbances in this system. Several studies of neurochemical changes subse- quent to TBI suggest that alterations in neurotransmitter production or delivery, or both, occur within these net- works both acutely and chronically and may therefore play a role in the development of neuropsychiatric prob- lems after TBI. These studies have shown that neu- rotransmitter systems, including norepinephrine, seroto- nin, dopamine, and acetylcholine, are altered by TBI, although the timing of such effects after TBI is important to consider. Multiple pharmacotherapies are available to modify the function of these neurotransmitter systems and the neuropsychiatric problems arising from distur- bances within them. In this chapter, we focus on TBI-induced neurotrans- mitter disturbances that are both related to neuropsychi- atric functioning and amenable to modification using agents presently available. These two limits focus this portion of the discussion on disturbances in dopamine, norepinephrine, serotonin, and acetylcholine. Catecholamines Discrete lesions to ascending monoaminergic projections may interfere with the function of systems dependent on such afferent pathways (Morrison et al. 1979). Monoam- inergic afferents course from the brainstem anteriorly, 614 TEXTBOOK OF TRAUMATIC BRAIN INJURY curving around the hypothalamus, the basal ganglia, and the frontal cortex, placing them in anatomical areas that are especially vulnerable to the effects of TBI. Two studies found markedly elevated plasma norepi- nephrine levels after acute brain injury (Clifton et al. 1981; Hamill et al. 1987). However, most of the studies in this area suggest only that acute elevations of striatal do- pamine are predictive of poor recovery from TBI (Don- nemiller et al. 2000; Hamill et al. 1987; Woolf et al. 1987). Only the study of Tang et al. (1997) related alter- ations in dopamine function to cognitive performance, and their findings suggest that dopamine antagonism, but not agonism, may improve performance speed on the wa- ter maze task in experimentally injured mice. The ob- served pairing of striatal hyperdopaminergia with post- TBI memory deficits in mice is puzzling in light of the long-standing inference of reduced dopamine function after TBI in humans. It is noteworthy that this inference is drawn from the observation of cognitive benefits after augmentation of dopaminergic function in persons with TBI, an observation for which several hypotheses (e.g., correction of primary dopamine deficiency or correction of secondary dopamine dysfunction because of dysregula- tion in complementary neurotransmitter systems) may be generated. Few other experimental injury studies (Egh- wrudjakpor et al. 1991; Kmeciak-Kolada et al. 1987; Tang et al. 1997) offer support for the hypothesis that ce- rebral catecholamine levels are chronically altered by TBI. No human studies have demonstrated a clear rela- tionship between in vivo markers of dopaminergic func- tion and long-term cognitive deficits in traumatically brain-injured humans. Thus, the extent of dopaminergic and noradrenergic dysfunction in the late period after TBI remains uncertain, and the implications of such find- ings with respect to long-term neuropsychiatric distur- bances require further study. Nonetheless, the observa- tion of cognitive improvements (e.g., arousal, speed of processing, attention, and, perhaps, memory) among some persons with TBIs during treatment with agents that increase dopaminergic neurotransmission suggests that dopamine dysfunction (primary, secondary, or both) may play an important role in the genesis of cognitive im- pairment after TBI. Serotonin Serotonergic projections to the frontal cortical areas are susceptible to biomechanical injury, and both diffuse axonal injury and contusions may produce dysfunction in this neurotransmitter system. Secondary neurotoxicity that is caused by excitotoxins and lipid peroxidation may also damage the neuronal systems that mediate serotonin (Karakucuk et al. 1997) and perhaps also norepinephrine. Studies of serotonin activity after TBI are somewhat vari- able in their findings, although differences in the method- ology (especially location of cerebrospinal fluid [CSF] sampling) appear to account for many of the differences in study findings. Pappius (1989) demonstrated wide- spread increases in hemispheric serotonin levels after experimentally induced brain injury in rats and noted that increases in serotonin appeared to produce decreases in cerebral glucose utilization. Busto et al. (1997) found a prompt increase in the extracellular levels of serotonin in cortical regions adjacent to the impact site in an experi- mental injury study in rats. Tsuiki et al. (1995) demon- strated in an experimental injury paradigm that serotonin synthesis was significantly increased in cortical areas throughout the injured hemisphere, and particularly in the dorsal hippocampus and area CA3, the medial genic- ulate, and the dorsal raphe, concurrent to a depression in cortical glucose use. Eghwrudjakpor et al. (1991) demon- strated a rapid increase in hemispheric concentration of serotonin, dopamine, and norepinephrine shortly after experimentally induced TBI in rats, with continued increases to three to four times control levels by 24–48 hours postinjury. These authors also reported significant regional differences in serotonin levels after experimental TBI, with increases in the hemispheres but decreases in the spinal cord. This may offer some explanation for the discrepancy of findings related to CSF serotonin, norepinephrine, and dopamine metabolites after TBI in humans; namely, that the site from which samples are obtained may yield sub- stantially different findings. Consistent with this experi- mental observation, Vecht et al. (1975) and Bareggi et al. (1975) found that lumbar CSF 5-hydroxyindoleacetic acid (5-HIAA) was below normal in conscious patients and normal in patients who were unconscious. Decreased CSF levels of serotonin were reported by Karakucuk et al. (1997) in 45 adults undergoing minor surgery with spinal anesthesia within 24 hours of TBI. However, Porta et al. (1975) demonstrated elevated ventricular CSF 5-HIAA levels in patients within days of severe TBI. Additionally, focal and diffuse lesions may result in differences with re- spect to monoaminergic alterations after TBI. For exam- ple, Van Woerkom et al. (1977) investigated patients with frontotemporal contusions and those with diffuse contu- sions. They documented decreased levels of 5-HIAA in patients with frontotemporal contusions but increased 5- HIAA levels in those with more diffuse contusions. In summary, the animal and human studies suggest acute in- creases in hemispheric serotonin levels after TBI and sug- gest that such increases are associated with decreased glu- cose utilization. Whether or to what extent similar Psychopharmacology 615 changes persist into the late period after TBI remains un- certain, as does the role of such changes in the genesis of neuropsychiatric symptoms after TBI. Acetylcholine Findings from both basic and clinical neuroscience sug- gest both acute and long-term alterations in cortical cho- linergic function develop after TBI. Multiple animal studies (Ciallella et al. 1998; DeAngelis et al. 1994; Dixon et al. 1994a, 1994b, 1997a, 1997b; Saija et al. 1988) dem- onstrate both acute and chronic alterations in hippocam- pal cholinergic function after experimentally induced TBI as well as a robust relationship between such alter- ations in cholinergic function and persistent cognitive impairments, including memory dysfunction. One of the most compelling demonstrations of relatively selective cholinergic injury after TBI is the report of Schmidt and Grady (1995). They induced a fluid-percussion brain injury sufficient to cause a 13- to 14-minute loss of right- ing reflex in rats anesthetized with halothane. Rats with experimentally induced midline injury had significant bilateral reductions in cholinergic neurons, including reductions in area Ch1 (medial septal nucleus; 36%), Ch2 (nucleus of the diagonal band of Broca; 44%), and Ch4 (nucleus basalis of Meynert; 41%). In animals with later- alized injuries, similarly severe losses of cholinergic neu- rons were observed ipsilaterally and lesser (11%–28%) losses were observed contralateral to the injury site. The authors noted that these losses did not extend to brain- stem cholinergic nuclei (Ch5 and Ch6), and there were no observable effects on forebrain dopaminergic or norad- renergic innervation. These findings suggest that cholin- ergic losses may exceed those of other neurotransmitter afferents. TBI appears to produce an acute increase in cholin- ergic neurotransmission followed by chronic reductions in neurotransmitter function and cholinergic afferents. Consistent with observations in experimental injury stud- ies, Grossman et al. (1975) demonstrated that patients with TBI had elevated acetylcholine levels in fluid ob- tained from intraventricular catheters or lumbar puncture in the acute period after TBI. Dewar and Graham (1996) and Murdoch et al. (1998) demonstrated cortical cholin- ergic dysfunction (loss of cortical cholinergic afferents with concurrent preservation of postsynaptic muscarinic and nicotinic receptors) weeks after severe TBI. Arcinie- gas et al. (1999, 2000a, 2001), using the hippocampally mediated cholinergically dependent P50-evoked wave- form response to paired auditory stimuli, demonstrated electrophysiological abnormalities consistent with re- duced hippocampal cholinergic function in patients with chronic symptoms of impaired auditory gating, attention, and memory in the late (longer than 1 year) period after TBI (see Chapter 7, Electrophysiological Techniques). Pharmacological Treatment of Specific Neuropsychiatric Syndromes Neuropsychiatric symptoms resulting from the neu- rotransmitter disturbances produced by TBI are amenable to treatment with a variety of medications. Where possible, selection of these medications should be guided by an understanding of the relationship between the neurochem- istry most likely related to the symptom, the injury location in the patient with that symptom, or (preferably) both. In this section, we review the major neuropsychiatric symp- toms and syndromes after TBI that may respond to medi- cations. We also present recommendations for the use of psychotropic medications to treat these syndromes as well as review their significant side effects. Emotional Disturbances Emotional disturbances, including mood disorders and disorders of affect regulation, are common conse- quences of TBI and may be detrimental to a patient’s rehabilitation and socialization (for reviews on these issues, see Arciniegas and Topkoff 2000; Arciniegas et al. 2000b; Hurley and Taber 2002; Silver et al. 1990, 1991). The literature regarding treatment of these con- ditions after TBI is limited when compared with that for phenotypically similar primary psychiatric disorders but is actively developing. Depression Depression after TBI can be responsive to psychophar- macologic treatment. Because of the safety profile, selec- tive serotonin reuptake inhibitors (SSRIs) are the pre- ferred medications. Cassidy (1989) conducted an open trial using fluoxetine for eight patients with severe TBI and associated depression. He found that two had marked improvement and three had moderate improvement. One-half of the patients experienced sedative side effects, and three out of the eight patients reported an increase in anxiety. Bessette and Peterson (1992) reported the case of a 41-year-old woman who experienced an episode of major depression after a mild brain injury and responded favorably to treatment with fluoxetine, 20 mg/day. Wrob- lewski et al. (1992a) reported a case in which improve- ment in depression after treatment with fluoxetine, 20 mg/day, after treatment with desipramine alleviated 616 TEXTBOOK OF TRAUMATIC BRAIN INJURY depressive symptoms but also precipitated posttraumatic seizures; however, this patient developed seizures while on fluoxetine as well, prompting the addition of pheny- toin. It is difficult to reach conclusions regarding the safety (or efficacy) of a medication based on single case reports. Thus, we remain circumspect with regard to the potential for fluoxetine to lower significantly the seizure threshold among patients with posttraumatic epilepsy. Nonetheless, the published observation of precipitation of posttraumatic seizures with both of these generally well-tolerated agents suggests that the possibility of alter- ing seizure threshold by their administration should not be dismissed offhandedly. Additionally, the observation supports the suggestion that this possibility should be dis- cussed during the process of providing informed consent to treatment with these (or almost any) antidepressant agents in this population. Fann et al. (2000) described improvement in depres- sion secondary to mild TBI using sertraline (dose range, 25–200 mg by end of study) in an 8-week, nonrandom- ized, single-blind, placebo run-in trial conducted on 15 patients diagnosed with major depression between 3 and 24 months after a mild TBI. Thirteen (87%) had a de- crease in Hamilton Rating Scale for Depression score of 50% or more (“response”), and 10 (67%) achieved a score of 7 (“remission”) or less by treatment week 8. Significant improvements were also observed in ratings of psycho- logical distress, anger and aggression, functioning, and postconcussive symptoms during treatment, and only one patient discontinued treatment because of side effects. In a subsequent report, Fann et al. (2001) described im- provements in psychomotor speed, recent verbal mem- ory, recent visual memory, and general cognitive effi- ciency as well as improvements in patient perception of cognitive symptoms as an effect of treatment of post-TBI depression with sertraline. Turner-Stokes et al. (2002) performed an open-label trial of sertraline for depression after brain injuries, in- cluding TBI, in 21 adult patients. They reported clinical improvement as assessed by DSM-IV (American Psychi- atric Association 1994) criteria in all of these patients. Among the 17 patients able to complete the Beck De- pression Inventory before and after treatment, signifi- cant decreases in depressive symptoms were associated with treatment in this group. Of these, 11 had failed pre- vious treatment with a different selective serotonin reup- take inhibitor. However, Meythaler et al. (2001) performed a pla- cebo-controlled trial of sertraline for arousal and atten- tional impairments in 11 subjects with severe TBI in the acute rehabilitation setting and failed to find a statistically significant treatment effect on these cognitive functions. Horsfield et al. (2002) performed an 8-month open- label study of the effects of fluoxetine, 20–60 mg/day in five patients with TBI and varying levels of depression to determine whether this medication conferred mood and/ or cognitive benefits. They observed improvements in mood as well as improvement on several measures of at- tention, processing speed, and working memory in this small group of patients. They suggested that fluoxetine’s ability to stimulate expression of brain-derived neu- rotrophic factor and its specific tyrosine kinase receptor, which has in rodents been demonstrated to produce neu- ritic elongation and increased dendritic branching density of some hippocampal neurons, may explain the apparent benefits of this agent on posttraumatic cognitive impair- ments. Although their suggestion is intriguing, support for it in experimental injury models is lacking. For the present, it is simpler to interpret their findings as reflect- ing the well-known activating effects of fluoxetine. Kant et al. (1998) reported that sertraline may also re- duce irritability and aggression (as assessed using the Overt Aggression Scale—Modified for outpatients) and depressive symptoms (as assessed using the Beck Depres- sion Inventory) after TBI at doses of 50 mg or greater. Notably, in this study, sertraline appeared to have a more robust effect on irritability and aggression than on de- pressive symptoms. Although Khouzam and Donnelly (1998) reported a reduction in TBI-induced compulsive behavior in re- sponse to treatment with venlafaxine, there are at the time of this writing no reports offering support for the use of newer antidepressants such as venlafaxine or mirtazapine in the treatment of depression after TBI. Common clini- cal experience suggests that many of these agents may be useful in the treatment of depression after TBI, but their use must be undertaken knowing that there has been no published information in this population to assist clini- cians in ascertaining the likelihood of benefit and the risk of adverse consequences. Because of the concern about hepatotoxicity with nefazodone, we would consider this medication only for individuals who have not been re- sponsive or tolerant to other antidepressants. When using the SSRIs, we would start at equivalent dosages of sertraline, 25 mg, or citalopram, 10 mg, and gradually increase the dose on a weekly basis (i.e., sertra- line, 50 mg for 1 week, then 100 mg, or increase citalo- pram to 20 mg after 1 week). Usual antidepressant dos- ages may be required. Tricyclic antidepressants (TCAs) may not be as effec- tive a treatment for depression after TBI as for primary major depressive episodes, and they are associated with increased risks of adverse events in patients with TBI. Sa- ran (1985) conducted a crossover study of phenelzine and Psychopharmacology 617 amitriptyline administered at therapeutic doses to 10 pa- tients with “minor brain injury” and 12 patients with ma- jor depression without TBI. All of the patients with major depression improved after 4 weeks of amitriptyline, but none of the TBI patients improved. Of note, however, the patients were reported to be the “melancholic” subtype, but they did not have significant weight loss or difficulty sleeping, which are typical symptoms of melancholic de- pression; therefore, the diagnostic categorization of these patients must be questioned. A subsequent study by Var- ney et al. (1987) found that 82% of 51 patients with major depressive disorder and TBI who received treatment with either TCAs or carbamazepine reported at least moderate relief of depressive symptoms. However, Dinan and Mobayed (1992) subsequently reported 85% of patients with major depressive disorder responded to amitrip- tyline, whereas only 31% of similarly depressed TBI pa- tients responded to this treatment. Nortriptyline and desipramine are used commonly in clinical practice, but there remains less evidence to guide their use and with which to assess the risks entailed by their use in persons with TBI than in other populations. Wrob- lewski et al. (1996) performed a modified, blinded, placebo lead-in treatment study of 10 patients with depression after severe TBI using desipramine and demonstrated improve- ment in six of seven patients (86%) able to complete the study. However, three patients (30%) discontinued the study, including one who developed seizures and one who developed mania during treatment. An additional patient experienced a seizure during treatment with desipramine but continued treatment with this medication nonetheless. In a study comparing nortriptyline versus fluoxetine in poststroke depression, nortriptyline was superior in effi- cacy to fluoxetine, and fluoxetine demonstrated no benefit above placebo (Robinson et al. 2000). Stroke is not patho- physiologically equivalent to TBI, and the studies compar- ing antidepressant efficacy may not be equally applicable to both populations. Both stroke and TBI may produce dis- crete white matter lesions that interrupt catecholaminergic or serotonergic pathways (source, projection, or target), and mood disorders after such injuries may result from dys- function in these neurotransmitter systems. Many persons with TBI may not have discrete lesions to these systems but may instead experience diffuse axonal injuries; such injuries may modestly affect ascending catecholaminergic or sero- tonergic pathways and also glutmatergically dependent systems, cholinergic projections, and a host of other cor- tico-cortico or cortico-subcortical pathways and cortical and/or subcortical structures. Additionally, TBI, but not stroke, produces bihemispheric injury in this manner. Therefore, the neuroanatomical and neurochemical conse- quences of TBI may not be the same as those resulting from stroke. That being so, there is reason to predict and also to explain observed differences in treatment effects and side effects in these two populations. The published treatment data for these two populations suggest the possi- bility that there are differences in TCA efficacy in these two populations (more effective in stroke than in TBI) and also that there may be a greater risk of adverse effect in TBI patients. If a heterocyclic antidepressant is chosen, we suggest nortriptyline (initial doses of 10 mg/day), or desipramine (initial doses of 25 mg/day), and a careful plasma moni- toring to achieve plasma levels in the therapeutic range for the parent compound and its major metabolites (e.g., nortriptyline levels 50–150 ng/mL; desipramine levels greater than 125 ng/mL). Should the patient become se- dated, confused, or severely hypotensive, the dosage of these drugs should be reduced. Depressed mood because of TBI may respond to treatment with methylphenidate. Gualtieri and Evans (1988) reported significant improvement on ratings of mood and cognitive performance among 15 patients with TBI after treatment with methylphenidate using a double- blind, placebo-controlled crossover design study. Al- though these results were modest and suggestive of a pos- sible role for methylphenidate in the treatment of the mood and cognitive disturbances after TBI, they have of- ten been interpreted as strong evidence of a role for this medication in the treatment of neuropsychiatric sequelae of TBI. Although other studies offer support for the role of methylphenidate in the treatment of cognitive impair- ment after TBI (discussed in the section Cognitive Im- pairment), it is not clear if or for how long such benefits on either mood or cognition might be sustained by this treatment. Common clinical experience suggests that dextroamphetamine may be similar in its effects on mood and cognition after TBI, but no reports document a clear role for this medication in the treatment of depression af- ter TBI. Monoamine oxidase inhibitors (MAOIs) are not often used in persons with depression after TBI. This may re- flect the high likelihood of difficulties with compliance to the complex dietary restrictions required during use of these medications given the cognitive impairments expe- rienced by many TBI patients. Additionally, the literature offers little support for the effectiveness of these medica- tions in the TBI population. In the studies by Saran (1985) and Dinan and Mobayed (1992) noted above, phenelzine was tried unsuccessfully in patients who had depression after TBI, even among those failing to re- spond to amitriptyline. Moclobemide, a selective MAO- A inhibitor, afforded improvement in 23 of 26 patients (88%) with depression after TBI (Newburn et al. 1999). 618 TEXTBOOK OF TRAUMATIC BRAIN INJURY Because moclobemide does not affect the isoenzyme MAO-B, its use does not entail the dietary restrictions as- sociated with other MAOIs. However, moclobemide is not available in the United States. Electroconvulsive therapy (ECT) remains a highly ef- fective and underused modality for the treatment of de- pression in general, and it appears to be an effective treat- ment of depression after acute TBI (Crow et al. 1996; Ruedrich et al. 1983; Zwil et al. 1992). Kant et al. (1999) re- ported on the safety and efficacy of ECT in patients with brain injury in a retrospective review of 11 patients hospi- talized as a result of neuropsychiatric problems after TBI. Of these subjects, 9 experienced a major depression or other mood disorder because of TBI. All of the patients with neuropsychiatric problems because of TBI responded favorably to ECT, as assessed by the Montgomery-Åsberg Rating Scale for Depression and Global Assessment Scale, and did so without significant adverse cognitive or physical sequelae. Functional improvement occurred irrespective of baseline cognitive functioning or severity of injury. These studies suggest that ECT may be a safe treatment for chronic and severe neuropsychiatric disorders because of TBI. When ECT is used, we recommend treatment with the lowest possible energy levels that will generate a seizure of adequate duration (longer than 20 seconds), using pulsa- tile currents, increased spacing of treatments (2–5 days be- tween treatments), and fewer treatments in an entire course (four to six). If the patient also has significant cogni- tive (especially memory) impairments because of TBI, nondominant unilateral ECT may be the preferable tech- nique if this treatment is used in this population. Adverse effects of antidepressants. The most common and disabling side effects of antidepressants in patients with neurological disorders are those associated with the anticholinergic properties of these medications, which can impair attention, concentration, and memory. For example, patients with Parkinson’s disease have shown increased confusion when treated with anticholinergic medications (De Smet et al. 1982; Dubois et al. 1990). Experimental evidence in traumatically brain-injured rats supports this observation (Dixon et al. 1994b, 1995), as does common clinical experience in the treatment of patients with TBI. Such observations are consistent with the observed effects of both experimental and human TBI on cortical cholinergic function noted in the section Ace- tylcholine. The antidepressants amitriptyline, trimip- ramine, doxepin, and protriptyline have high affinities for the muscarinic receptors; given their strong anticho- linergic properties, these medications should be pre- scribed only after careful consideration of alternative medications. The choice of SSRI may require similar consider- ation; Schmitt et al. (2001) demonstrated that healthy middle-aged adults experienced significantly greater im- pairments of delayed recall in a word learning test during treatment with paroxetine, 20–40 mg/day, than during treatment with placebo, an effect attributed to paroxe- tine’s nontrivial antimuscarinic properties. This study also demonstrated significant improvements in verbal fluency among healthy middle-aged adults treated with sertraline, 50–100 mg, when compared with treatment with placebo, an effect attributed to sertraline’s dopamine reuptake in- hibition. Whether similar differences in cognitive profiles distinguish between these and other SSRIs in the TBI population is not yet clear. Nonetheless, observations of distinct cognitive profiles among these agents may merit consideration when selecting an agent in this population. Additionally, many antidepressants (e.g., doxepin, am- itriptyline, trimipramine, imipramine, maprotiline, and trazodone) are highly sedating, resulting in significant problems of arousal in the TBI patient. Again, these med- ications should be prescribed only after careful consider- ation of other therapies. TCAs may be associated with nontrivial rates of ad- verse events, particularly seizures. Wroblewski et al. (1990) reviewed the records of 68 patients with TBI who received antidepressant and, predominantly, TCA treat- ment for at least 3 months. The frequency of seizures was compared for the 3 months before treatment, during treatment, and after treatment. Seizures occurred among 6 patients during the baseline period, 16 during antide- pressant treatment, and 4 after treatment was discontin- ued. Fourteen patients (20%) had seizures shortly after the initiation of treatment. For 12 of these patients, no seizures occurred after treatment with the antidepressant was discontinued. Importantly, 7 of these patients were receiving anticonvulsant medication before and during antidepressant treatment. Also, the occurrence of seizures was related to greater severity of brain injury. Wroblewski et al. (1992a) also observed seizures in a patient receiving fluoxetine for depression after TBI, suggesting that this medication, and perhaps other SSRIs, may be associated with an increased risk of seizures during antidepressant therapy after TBI. In addition to the TCAs, maprotiline and bupropion are often suggested to be associated with a higher incidence of seizures in otherwise healthy psychi- atric patients (Davidson 1989; Pinder et al. 1977). Such suggestions prompt caution before prescribing these agents in patients with depression after TBI. However, Johnston et al. (1991), in a 102-site study of 1,986 patients treated with bupropion for depression, reported seizure rates of 0.24%–0.40%, and, among those receiving 300– 450 mg/day, the cumulative rate of seizure was 0.36%. Psychopharmacology 619 This large data set suggests that bupropion may not be more likely to reduce seizure threshold than other antide- pressants. Whether the same is true of bupropion’s effects on seizure threshold after TBI is not clear at present, nor are there any data with which to assess the likelihood of similar problems during treatment with maprotiline in this population. Among patients with established epilepsy, Ojemann et al. (1987) found that seizure control does not appear to worsen if psychotropic medication is introduced cautiously and if the patient is on an effective anticonvulsant regimen. There are, at present, no indications that treatment of de- pression in patients with posttraumatic epilepsy differs from that in patients with epilepsy of other etiologies. Al- though we conclude that antidepressants can be used safely and effectively in patients with TBI, including patients with posttraumatic epilepsy, we recommend that these agents be prescribed with caution and that treatment with them should include assiduous monitoring for adverse ef- fects, including change in seizure frequency. There are several important drug interactions that may occur among antidepressants and other drugs com- monly prescribed for neurological conditions (Dubovsky 1992). Many antiparkinsonian drugs and neuroleptics have anticholinergic effects that are additive to those of the antidepressants. Antidepressant levels are likely to be decreased—often below therapeutic range—by the anti- convulsants phenytoin, carbamazepine, and phenobar- bital. Similarly, antidepressants such as fluoxetine may raise the plasma levels of the anticonvulsants phenytoin (Jalil 1992), valproate (Sovner and Davis 1991), and car- bamazepine (Grimsley et al. 1991). Carbamazepine in- duces the metabolism of sertraline. Therefore, patients receiving treatment with medications that require ther- apeutic blood level monitoring should have more fre- quent monitoring when antidepressants are adminis- tered. Although they may be highly efficacious drugs in patients with primary major depression, MAOIs should be less frequently prescribed for the treatment of de- pression in patients with TBI and particularly among those who are also taking other drugs that affect the cen- tral nervous system (CNS). For example, interactions with stimulants such as dextroamphetamine and with le- vodopa may result in lethal hypertensive reactions. (For a review of the safe use of MAOIs, see Marangell et al. 2003.) Mania Mania and bipolar disorder are less common conse- quences of TBI, although we believe they have been underdiagnosed in these individuals (see Chapter 10, Mood Disorders, and Hurley and Taber 2002 for review). Several small case series suggest that lithium carbonate may be useful for the treatment of mania after TBI, although partial response, relapse of symptoms, or need for a second mood stabilizer is often observed (Bamrah and Johnson 1991; Parmalee and O’Shanick 1988; Stark- stein et al. 1988, 1990; Stewart and Hemsath 1988; Zwil et al. 1993). Lithium has been reported to aggravate con- fusion in patients with brain damage (Schiff et al. 1982) and may relatively easily produce nausea, tremor, ataxia, and lethargy in persons with neurological disorders. In addition, lithium may lower seizure threshold (Massey and Folger 1984). Hornstein and Seliger (1989) reported a patient with preexisting bipolar disorder who experi- enced a recurrence of mania after closed head injury. This patient’s mania, before injury, was controlled with lithium carbonate without side effects. However, subsequent to brain injury, dysfunctions of attention and concentration emerged that reversed when the lithium dosage was low- ered. Because lithium carbonate may exacerbate cognitive impairments or cause confusion, especially in combina- tion antidepressants, anticonvulsants, and antipsychotic medications, we suggest limiting the use of lithium in patients with TBI to those with mania or recurrent depressive illness that preceded their brain damage and who previously responded well to this treatment. Fur- thermore, and to minimize lithium-related side effects, we begin with low doses (300 mg/day). Patients with mania after TBI may respond to treatment with lithium despite relatively low blood levels (e.g., 0.2–0.5 mEq/L), highlighting the need for a “start low, go slow” approach to the care of these patients. Manic episodes occurring after TBI may also respond to carbamazepine (Nizamie et al. 1988; Stewart and Hem- sath 1988), although often only after addition of lithium (Stewart and Hemsath 1988) or antipsychotics (Sayal et al. 2000; Starkstein et al. 1988). For patients with mania sub- sequent to TBI, carbamazepine should be initiated at a dosage of 200 mg bid and adjusted to obtain plasma levels of 8–12 µg/mL. Because carbamazepine may produce or exacerbate cognitive impairments (Massagli 1991), moni- toring for this effect when using this agent in patients with TBI is suggested. Brain damage appears to increase suscep- tibility to neurotoxicity induced by combination therapy with carbamazepine and lithium (Parmelee and O’Shanick 1988). As is true for patients without histories of TBI, cli- nicians should be aware of the potential risks associated with carbamazepine treatment, particularly bone marrow suppression (including aplastic anemia) and hepatotoxicity. Complete blood cell counts and liver function tests should be regularly monitored (Marangell et al. 1999). The most common signs of carbamazepine-induced neurotoxicity in- clude lethargy, confusion, drowsiness, weakness, ataxia, 620 TEXTBOOK OF TRAUMATIC BRAIN INJURY nystagmus, and increased seizures. Pleak et al. (1988) de- scribed the development of mania, irritability, and aggres- sion with carbamazepine treatment; however, in our expe- rience, this reaction is unusual. Pope et al. (1988) suggested that sodium valproate may be a useful mood stabilizer for patients with symp- toms of bipolar disorder after TBI, and Monji et al. (1999) suggested that this benefit may extend to patients with rapid cycling mood disorders after TBI. In Monji et al.’s retrospective report, patients with such symptoms after TBI appeared to respond more robustly than those with similar symptoms in the absence of TBI (88% vs. 46%). The small sample sizes in this study do not permit extrap- olation of this observation to TBI patients more gener- ally, but are nonetheless encouraging of the use of this medication in the TBI population. As with carbamaz- epine, valproate may exacerbate cognitive impairments (Massagli 1991), and its use should include ongoing as- sessment of cognition in persons with TBI. Valproate is begun at a dosage of 250 mg bid and gradually increased to obtain plasma levels of 50–100 µg/mL. Tremor and weight gain are common side effects. Hepatotoxicity is rare and usually occurs in children who are treated with multiple anticonvulsants (Dreifuss et al. 1987). For mania or manic-like syndromes after TBI that do not respond to conventional mood-stabilizing therapies, relatively more novel approaches may be useful to con- sider. Bakchine et al. (1989) described a manic-like state in a 44-year-old right-handed woman with bilateral or- bitofrontal and right temporoparietal traumatic contu- sions that responded to clonidine after her behavior failed to respond to carbamazepine and worsened with levo- dopa. Dubovsky et al. (1987), Levy and Janicak (2000), and others have suggested that verapamil may be a useful agent for the treatment of mania alone or in combination with other mood stabilizers. To date, there are no studies of verapamil for the treatment of mania after TBI, but this agent might be worth considering when other con- ventional treatments fail or produce intolerable side ef- fects. Clark and Davison (1987) also reported that ECT effected improvement in manic symptoms after nonpen- etrating trauma, and the authors suggested that this ther- apy may be valuable to consider in such cases. Lamotrig- ine, oxcarbazepine, and gabapentin are other options, although evidence as to efficacy in individuals with TBI is not presently available. Affective Dysregulation (Affective Lability and Pathological Crying/Laughing) In contrast to mood disorders, conditions in which the base- line emotional state is pervasively disturbed over a relatively long period (i.e., weeks), disorders of affect denote condi- tions in which the more moment-to-moment variation and regulation of emotion is disturbed. The classic disorder of affective dysregulation is pathological laughing and/or cry- ing (PLC), also sometimes referred to as emotional inconti- nence or pseudobulbar affect. Patients with this condition expe- rience episodes of involuntary crying and/or laughing that may occur many times per day, often provoked by trivial (i.e., not sentimental) stimuli, are quite stereotyped in their pre- sentation, are uncontrollable, do not evoke a concordant subjective affective experience, and do not produce a persis- tent change in the prevailing mood (Poeck 1985). In this classic presentation, PLC appears to be a relatively infre- quent (5.3%) consequence of TBI (Zeilig et al. 1996). Affec- tive lability differs from PLC in that both affective expres- sion and experience are episodically dysregulated, the inciting stimulus may be relatively minor but is often some- what sentimental, and the episodes are somewhat more ame- nable to voluntary control and are less stereotyped. How- ever, these episodes do not produce a persistent change in mood and are often sources of significant distress and embarrassment to patients who otherwise (quite correctly) report their mood as “fine” (euthymic). The prevalence of affective lability after TBI is not clear, although Jorge and Robinson (2003) suggested a 1-year prevalence of approxi- mately 12% among persons with TBI. Although the neurobiology of mood and affect regu- lation overlap, the treatment of affective dysregulation in patients with brain injury overlaps but is not identical with the treatment of “uncomplicated” depression after TBI (Lauterbach and Schweri 1991; Panzer and Mellow 1992; Schiffer et al. 1985; Seliger et al. 1992; Sloan et al. 1992). The treatment literature overwhelmingly supports the use and effectiveness of relatively low doses (below typical antidepressant doses) of serotonergically and nor- adrenergically active antidepressants (Andersen et al. 1993; Lawson et al. 1969; Robinson et al. 1993; Schiffer et al. 1985) and to a lesser extent dopaminergic (Udaka et al. 1984) and noradrenergic (Evans et al. 1987; Sandyk and Gillman 1985) agents for the treatment of PLC and affective lability. Whether the lack of distinct therapies for these two disorders of affect reflects inseparable com- monalities in their neurobiology or is instead an artifact of the diagnostic heterogeneity of patients included in the available treatment reports is unclear (Arciniegas and Topkoff 2000). It is noteworthy that the majority of treat- ment studies of these problems derives from the stroke, and not TBI, literature. Nonetheless, similar findings in multiple case series support the benefit of these agents for affective lability and PLC after TBI. There are multiple reports of the beneficial effects of fluoxetine for “emotional incontinence” secondary to neu- rological disorders (Panzer and Mellow 1992; Seliger et al. Psychopharmacology 621 1992), including TBI (Nahas et al. 1998; Sloan et al. 1992). Brown et al. (1998) treated 20 patients with poststroke “emotionalism” (either PLC or affective lability) with flu- oxetine in a double-blind placebo-controlled study. Those individuals receiving fluoxetine exhibited statistically and clinically significant improvement. In general, these inves- tigators began treatment with 20 mg/day of fluoxetine, and patients often exhibited response within 5 days. We have had similar success with fluoxetine raised to higher doses (40–80 mg/day) and with sertraline, often starting and re- maining at 25 mg/day and occasionally increasing gradu- ally to 100 mg/day. A single-case report (Breen and Gold- man 1997) and a small open-label trial (Muller et al. 1999) demonstrated reductions in affective lability during treat- ment with paroxetine; the latter of these two reports also compared the effectiveness of paroxetine and citalopram for the treatment of affective lability after brain injury and found both medications effective and citalopram somewhat better tolerated. Although only 2 of 26 patients included in the series described by Muller et al. (1999) were patients with TBI (the remainder being patients with strokes), both remained successfully treated for 1 year with paroxetine and relapsed after drug discontinuation. Andersen et al. (1999) also describe improvement in episodic crying after TBI in a 6-year-old child with citalopram, 2.5 mg daily. As is often seen in the treatment of affective lability, treatment response occurred within 2 days of beginning treatment, a response more rapid than that usually encountered in the treatment of depressed mood or major depressive episode. TCAs may also be effective for affective lability and PLC. Allman (1992) described a marked decrease in patho- logical laughter in a patient treated with imipramine, 150 mg/day, with improvement occurring by the second week of treatment. Common clinical practice using TCA for PLC and affective lability after stroke (Robinson et al. 1993) suggests that nortriptyline may be of considerable benefit to patients with these conditions, and often at doses lower than those generally used to treat major depressive episodes. However, we emphasize that for many patients it may be necessary to administer these medications at stan- dard antidepressant dosages to obtain full therapeutic ef- fects, even when patients begin responding within days of initiating treatment at relatively low doses. Although psychostimulants and dopaminergic agents are used most often for the treatment of cognitive impair- ments or diminished motivation, or both, after TBI, they may also offer some relief from affective lability during treatment of these other problems as well. Evans et al. (1987) reported reduced affective lability as well as cogni- tive improvements in a young man treated with methyl- phenidate or dextroamphetamine during a single-case, double-blind, placebo-controlled, dose-response study. Gualtieri et al. (1989) described a sustained reduction of agitation and aggression, decreased distractibility, and improvement in affective stability among 19 of 30 TBI patients taking amantadine, 50 to 400 mg/day (average dose of 290 mg/day). Udaka et al. (1984) also reported re- ductions of PLC in response to amantadine or levodopa in approximately 50% of stroke or TBI patients. When patients present with affective lability or PLC in addition to cognitive and/or motivational impairments, methyl- phenidate, dextroamphetamine, amantadine, or levodopa may offer some relief from both sets of problems. In the event that the first-line therapies (i.e., seroton- ergically and/or dopaminergically active agents) do not provide adequate relief from affective lability after TBI, particularly if affective lability is comorbid with posttrau- matic aggression, treatment with mood-stabilizing agents may be necessary and of some benefit. Glenn et al. (1989) described an open-label trial of lithium carbonate for the treatment of affective instability and aggressive behavior in 10 patients (8 TBI and 2 stroke). The patients’ symp- toms included episodic aggressive or self-destructive be- havior, “mood swings,” tearfulness, and euphoria. Six of these patients demonstrated marked or moderate im- provement in these target symptoms, one improved tran- siently, one failed to respond, and two patients worsened with this treatment. Three patients were on concomitant neuroleptic therapy and experienced neurotoxic side ef- fects that prompted discontinuation of the lithium. Addi- tionally, one patient experienced decreased attentiveness, and one patient experienced a seizure during this treat- ment. Lithium levels associated with clinical improve- ment ranged between 0.5 and 1.4 mEq/L. Lewin and Sumners (1992) described a single case re- port of carbamazepine treatment of posttraumatic “epi- sodic dyscontrol,” a term used in their report to denote uncontrolled disproportionate episodic violence, depres- sion, tearfulness, and irritability toward and intolerance of others. Treatment with carbamazepine, 200 mg/day, produced a good response, with no violent outbursts over the 12-month period of observation. Both of these reports suggest possible benefit of mood-stabilizing agents for the treatment of some forms of affective lability after TBI, especially when mixed with irritability, aggression, or both. However, and as noted before, a cautious approach to dosing and continuous re- assessment of benefit and adverse effects is needed in this population when using such agents. Cognitive Impairment Medication treatments for cognitive impairments after TBI follow one or both of two major neuropharmacolog- [...]... (see Diller and Gordon 198 1a, 198 1b, for a discussion of this literature) In addition, several review papers have been published on this topic (Ben-Yishay and Diller 199 3; Gordon 199 0; Gordon and Hibbard 199 1, 199 2; Gordon et al 198 9; Mateer and Raskin 199 9; Prigatano 199 9) The rapid development of brain injury rehabilitation programs mirrored the development of this new form of rehabilitation therapy... and after the brain injury (Cicerone 198 9; Ellis 198 9; Prigatano 198 9) Events surrounding the injury can have far-reaching experiential and symbolic significance for the injured person, and the disinhibition that frequently follows as a consequence of brain injury can result in the reemergence of previously resolved psychological issues dating back to childhood (Bennett 198 9; Silver et al 199 2) These factors... the early 199 0s, 95 % of brain injury rehabilitation programs were providing some form of cognitive rehabilitation or remediation (Mazmanian et al 199 3) Approaches have been developed to remediate the most commonly recognized cognitive difficulties experienced by individuals with brain injury: in attention and concentration, memory, executive functions, visual per- TEXTBOOK OF TRAUMATIC BRAIN INJURY ception,... patients with brain injuries to breach the walls of their isolation and to begin to relate to other people effectively again, therapists and their patients must find areas of shared meaning (Stuewe-Portnoff 198 8) The therapist and the patient must come to share an understanding of the nature of the problem as it is experienced by the patient (Cicerone 198 9; Pollack 198 9; Prigatano 198 9) Prigatano ( 198 9) expressed... patients with brain injury are more sensitive to the development of extrapyramidal side effects during treatment with typical antipsychotic medications (Rosebush and Stewart 198 9; Vincent et al 198 6; Wolf et al 198 9; Yassa et al 198 4a, 198 4b) Given this literature and the availability of several atypical antipsychotic medications, we strongly discourage the use of typical and, particularly, the low-potency... treatment outcome Mild Brain Injury A number of animal and human studies have demonstrated that there is a continuum of neurological damage and functional impairment from mild to severe brain injury (Eisenberg and Levin 198 9; Genarelli 198 1; Rutherford 198 9) The cognitive impairments that result from mild brain injuries are essentially the same as many of those that are seen after major brain trauma, although... potential for lowering the seizure threshold (Marangell et al 199 9; Oliver et al 198 2) Clozapine treatment is associated with a significant dose-related incidence of seizures (ranging from 1% to 2% of patients who receive doses below 300 mg/day, and 5% of patients who receive 600 90 0 mg/day) (Lieberman et al 198 9) The observations of Michals et al ( 199 3) suggest that this risk may be increased in patients... discontinuation of thioridazine is attributable to the brain- injured patients’ reduced tolerance to the anticholinergic properties of this agent Similarly, Sandel et al ( 199 3) observed new-onset delusions in a TBI patient receiving chlorpromazine for the treatment of agitation after TEXTBOOK OF TRAUMATIC BRAIN INJURY TBI, an effect that may also be attributable to the significant anticholinergic properties of this... aware of the pervasive impact of brain injury on everyday function Does Severity of Injury Play a Role in the Efficacy of Cognitive Remediation? The nature of the interaction between severity of brain injury and the ability to profit from cognitive remediation, although not specifically studied, may be inferred from research and clinical experience: • Ben-Yishay et al ( 197 0) found that the number of cues... at large (Brooks 199 1; Lezak 198 6; Thomsen 198 4) 652 TABLE 35–2 TEXTBOOK OF TRAUMATIC BRAIN INJURY Special therapeutic problems and management issues Condition Description Therapist response Transference and countertransference reactions Be aware of the probability of some Transference: Loss or threat to sense of self, limited disruptive transference and adaptability, intolerance of anxiety; all promote . animal studies (Ciallella et al. 199 8; DeAngelis et al. 199 4; Dixon et al. 199 4a, 199 4b, 199 7a, 199 7b; Saija et al. 198 8) dem- onstrate both acute and chronic alterations in hippocam- pal cholinergic function. al. 198 3; Zwil et al. 199 2). Kant et al. ( 199 9) re- ported on the safety and efficacy of ECT in patients with brain injury in a retrospective review of 11 patients hospi- talized as a result of. 198 2; Dubois et al. 199 0). Experimental evidence in traumatically brain- injured rats supports this observation (Dixon et al. 199 4b, 199 5), as does common clinical experience in the treatment of patients