Textbook of Traumatic Brain Injury - part 5 pdf

Mild Brain Injury and the Postconcussion Syndrome 297 patterns) was consistent with predictable areas of brain injury. Of note is that the patients were referred largely because of persistent complaints and thus may not be rep- resentative of all patients with a mild brain injury. This group has subsequently demonstrated correlations between certain electroencephalographic characteristics such as electroencephalographic coherence (a measure of homogeneity of electrical activity across different dis- tances) and electroencephalographic amplitude within different wave frequencies (i.e., alpha, beta, delta, and theta) and the brain water proton relaxation times (T2) obtained with conventional MRI (Thatcher et al. 1998a, 1998b). The average T2 relaxation time is in part a function of the distribution of the H1 imaging agent in intracellular water, extracellular water, and protein/lipid membrane, and this distribution can, and often does, change after a tissue injury. Thus, changes in the T2 relaxation time can reflect past injury. Thatcher et al. (2001) compared a variety of electroencephalographic measures between groups with mild, moderate, and severe TBI and proposed an “EEG Severity Index” that showed promise in distinguishing MTBI from more severe forms. These reports suggest that quantitative electroencephalographic techniques may prove to be more valuable in the assessment of mild brain injury than standard EEGs, although they remain experimental and as yet are not recom- mended as routine diagnostic procedures in the guide- lines put forth by the American Academy of Neurology and the American Clinical Neurophysiology Society (Gaetz and Bernstein 2001). A similar picture emerges with respect to the EP and ERP literature. In their study of brainstem auditory evoked responses in 165 patients with mild brain injury (GCS, 13–15; LOC less than 20 minutes), Schoenhuber and Gentilini (1986) showed that approximately 10% of patients had at least one prolonged interpeak latency. However, these abnormalities did not correlate with the presence or absence of relevant postconcussion symptoms. Abd Al-Hady et al. (1990) also found prolongation of certain interpeak latencies in brainstem auditory responses in their group of 30 patients with mild brain injury. It was not clear whether these findings correlated with any subjective complaints. Pratrap-Chand et al. (1988) found increased P300 latencies in a group of 20 patients with mild brain injury compared with healthy control subjects when tested within 4 days after injury. The latencies were normal on retesting 30–250 days subsequent to initial testing. Only two of these patients were complaining of any postconcussive symptoms. Several studies have explored different EP paradigms, including aspects of the evoked response that represent subcortical and thalamocortical processing (Drake et al. 1996; Soustiel et al. 1995) and visual evoked responses (Freed and Hellerstein 1997; Gaetz and Weinberg 2000; Gaetz et al. 2000; Papathanasopoulos et al. 1994). In general, a subsample of individuals with MTBI can be found with abnormal findings, the percentage of which varies with the range of “normal” that is used. This highlights the fact that there as yet are no established norms for many of these measures, or at least the ranges of norms are not universally agreed on. Thus, it is difficult to state with certainty what percentage of individuals with MTBI has abnormal findings. Arciniegas et al. (2000a) have studied attentional gat- ing mechanisms in individuals with persistent attentional complaints after TBI using a P50 auditory evoked response paradigm. In most healthy individuals, the evoked response to the second of a paired auditory stimulus is suppressed, implying the ability to screen out, or gate, auditory stimuli. A significantly higher percentage of persis- tently symptomatic individuals with TBI did not suppress the response to the second stimulus. These individuals were also found to have smaller hippocampal volumes (Arciniegas et al. 2001) and, in an open-label study, showed symptomatic improvement while taking donepezil, suggesting that cholinergic deficits may underlie some of the attentional complaints in this group (Arcinie- gas et al. 1999). Thus, from a neurodiagnostic standpoint, both functional imaging techniques and some of the newer EPs and ERPs show promise for helping to clarify aspects of brain function after MTBI, particularly in those with persistent symptoms. However, none of these techniques can be considered part of a routine clinical evaluation at this time. Treatment Issues Evaluation At the risk of stating the obvious, the foundation of the approach to patients with mild brain injury is a proper evaluation. Significant effort must be expended to clarify premorbid history. In particular, one must look for a prior history of brain injury, which can be seen in as many as 30% of patients (Rimel et al. 1981). The association of substance abuse with brain injury is well described (Spa- radeo et al. 1990) and may contribute to postinjury sequelae. Interviews with significant others can be invaluable in gaining a clearer picture of these issues. Signs and symptoms must be clearly defined, as well as any changes in symptom picture as a function of time from the injury. The profile of the injury itself must be outlined, including the type of injury, the presence or ab- 298 TEXTBOOK OF TRAUMATIC BRAIN INJURY sence of LOC and its duration, and the presence, absence, and duration of any retrograde and anterograde amnesia. Corroborative information, including accounts from ob- servers, emergency medical technicians, ambulance and emergency department personnel, and inpatient hospital records, can be invaluable. When evaluating these records, phrases such as “normal mental status” without sufficient documentation do not eliminate the possibility that there were cognitive changes. This is particularly true when the emergency team is distracted by other trauma such as injury to the spinal cord (Davidoff et al. 1985). The absence or presence and location of complications such as depressed skull fractures, cerebral contusions, and extradural hematomas should be noted because of the potential prognostic implications (Williams et al. 1990). The neurodiagnostic tests done and the timing in relation to the injury should be clarified and the reports or actual studies obtained. All of the above information can then be integrated with findings from the clinical interview to determine the consistency of the history and examination with the known sequelae of mild brain injury. This process should determine the presence or absence of one or more of the specific syndromes outlined above, including postconcussive symptoms, depression, mania, anxiety syndromes (including PTSD), and psychotic syndromes. Treatment should then follow rationally from this diagnostic scheme. Medication Approaches Several general principles should be borne in mind when prescribing psychotropic agents in the population with MTBI. These patients seem to be more sensitive to common psychotropic side effects such as sedation, psychomo- tor slowing, and cognitive impairment (such as impairments of recent memory and attention). Although there are few actual data, most clinicians working with patients with TBI note this tendency toward increased side effects and a resultant narrowing of the benefit to toxicity ratio. In general, it is prudent to use lower starting and (often) final doses and prolong the titration intervals (Arciniegas et al. 2000b; Cope 1987; Gualtieri and Evans 1988; McAllister 1992c; McAllister and Price 1990; Silver et al. 1992). Medication approaches to the sequelae of MTBI have generally taken three broad approaches: 1) amelioration of psychiatric complications, 2) amelioration of specific symptoms (e.g., headache, dizziness, and sleep distur- bances; see Chapters 20, Fatigue and Sleep Problems; 21, Headaches; and 22, Balance Problems and Dizziness), and 3) approaches to cognitive complaints. With respect to amelioration of psychiatric complications, the same general approaches taken in the noninjured population are typically used, although therapeutic efficacy studies are lacking in this group. An older study by Saran (1985) of 10 patients with mild brain injury and depression suggests that some of these patients may be less responsive to antidepressants than patients without a brain injury. On the other hand, Wroblewski et al. (1996) found a good response to desipramine in the treatment of their population of depressed individuals after TBI, and Fann et al. (2000) found a good antidepressant response to sertraline in 15 individuals with depression after an MTBI. In the Neuropsychiatry Clinic at Dartmouth Medical School, it is our experience is that there are no dramatic antidepressant efficacy differences in individuals with TBI relative to the noninjured population. Hoff et al. (1988) reported a higher relapse rate in patients with central nervous system secondary mania, although these were not patients with mild brain injury. The phenomenology of depressive and manic syndromes can also be altered by a brain injury (McAllister 1992b; McAllister and Price 1990; Saran 1985; Shukla et al. 1987; Silver et al. 1991), resulting in a mixed and atypical clinical presentation. Thus, psychotropic use is complicated by enhanced sensitivity to side effects, a mixed and atypical clinical picture (which can complicate assessment of target symptoms and drug response), and, perhaps, a reduced efficacy of certain standard agents, although the evidence for this is tentative. The treatment of postconcussive cognitive symptoms is even less clear-cut. Work since the 1980s has focused more on the role of catecholaminergic and cholinergic mechanisms as mediators of the attentional and memory domains vulnerable to injury in TBI (McAllister and Ar- ciniegas 2002). Catecholaminergic mechanisms, particularly through dopaminergic (DA) and α 2 -adrenergic (A2A) systems, appear to play important roles in memory function, particularly WM function (see Arnsten 1998) both in healthy individuals and individuals with TBI. Lu- ciana et al. (1992) and Luciana and Collins (1997) have found improvements in spatial WM tasks in healthy individuals treated with bromocriptine (a D2 agonist). Elliot et al. (1997) found improved performance on a spatial WM task after administration of methylphenidate. It is difficult to know whether the observed effect is strictly related to DA augmentation, because methylphenidate also results in release of norepinephrine (NE) and A2A stimu- lation is also known to improve WM performance in animals and healthy humans (Arnsten et al. 1998; Jakala et al. 1999a, 1999b). There is some evidence that baseline WM capacity plays a role in DA enhancement. Kimberg et al. (1997) gave 2.5 mg of bromocriptine to 31 healthy human subjects and then administered several neurocognitive tasks, including a spatial WM task similar to that used by Luci- Mild Brain Injury and the Postconcussion Syndrome 299 ana and Collins (1997). The subjects were divided into two groups (high- and low-WM capacity) on the basis of their performance on a reading span task. Administration of bromocriptine resulted in improvement in WM performance only in the low-capacity group. There are a few studies assessing the efficacy of DA agents on WM in individuals after TBI. McDowell et al. (1998) found significant improvement in tasks requiring “executive function” (e.g., dual-task paradigm) but not WM storage capacity or prefrontal tasks that did not require executive functions after a single dose of 2.5 mg bromocriptine to 24 subjects with TBI. Several DA agonists, including bromocriptine and stimulants, particularly those with DA agonist properties such as methylphenidate, amphetamine, and levodopa, have been used to treat various cognitive and behavioral sequelae of TBI and other acquired brain injuries. Clini- cal observations suggested improvement in many subjects in areas as diverse as impulse control, attention, insight, cooperation, and memory (Arciniegas et al. 2000b; Crismon et al. 1988; Dobkin and Hanlon 1993; Glenn 1998; Gualtieri et al. 1989; Lal et al. 1988; McAllister 1992a, 1992c; Powell et al. 1996). Whyte et al. (1997) reported the results of a double-blind, placebo-controlled, crossover study of the effects of 0.25 mg/kg methylphenidate on measures of attention in 19 TBI subjects of mixed injury severity. Components of attention assessed included sustained arousal, phasic arousal, distraction, choice reaction time, and behavioral inattention. Meth- ylphenidate was found to have a differential effect on different attentional performance variables. There is limited but equally compelling evidence suggesting that A2A mechanisms play a prominent role in the activation and modulation of WM. Localized and global depletion of catecholamines (DA and NE) as well as aging impair performance on spatial WM tasks similarly to that seen with ablation of neural tissue in the prefrontal region (Arnsten 1998; Bartus et al. 1978; Brozoski et al. 1979; Cai et al. 1993; Luine et al. 1990). Infusion of A2A antagonists produces spatial WM impairment in both monkeys and rats (Steere and Arnsten 1997; Tanila et al. 1996). These performance deficits can be reversed by administration of A2A agonists (see Arnsten 1998). Of note is that adrenergic enhancement of WM appears to be relatively specific to ma- nipulation of the α 2 receptors in that α 1 - and β-adrenergic antagonists had no effect on WM performance (Li and Mei 1994). However, A1A agonists can impair WM function, suggesting 1) that different adrenergic receptors have opposing effects on cognitive function (Arnsten 1998) and 2) that it is important to clarify the different roles of these re- ceptor families rather than simply administering broad- spectrum adrenergic agents such as stimulants. Thus, broad-spectrum adrenergic agents, or agents that increase the endogenous release of NE such as methylphenidate, may have opposing effects on WM function. Jakala et al. (1999a, 1999b) gave healthy control subjects several different doses of clonidine or guanfacine (both A2A agonists). Guanfacine at the higher dose (29 µg/kg) was associated with significant improvement in several tasks, including a spatial WM task, paired associate learning, and Tower of London. They interpreted these results as consistent with guanfacine-enhanced frontal functioning in both spatial WM and planning. Another hypothesis relates cognitive impairment after TBI to acute and long-term alterations in cortical cholinergic function (Arciniegas 2003). Animal studies (DeAn- gelis et al. 1994; Dixon et al. 1994; Saija et al. 1988) demonstrate chronic alterations in hippocampal cholinergic function after experimentally induced TBI and the relationship of such alterations to persistent cognitive impairments. Human postmortem studies (Dewar and Graham 1996; Murdoch et al. 1998) also demonstrate that TBI produces cortical cholinergic dysfunction via loss of cortical cholinergic afferents; these studies also demonstrate that postsynaptic muscarinic and nicotinic receptors are not reduced by TBI. Multiple studies have demonstrated that cholinergic augmentation, generally using one of several cholinester- ase inhibitors (e.g., physostigmine or donepezil) can improve TBI-induced memory deficits even in the late postinjury period (longer than 1 year) in some TBI survivors (Aigner 1995; Bogdanovitch et al. 1975; Cardenas et al. 1994; Eames and Sutton 1995; Goldberg et al. 1982; Levin et al. 1986; Tayerni et al. 1998; Whelan et al. 2000). Arciniegas and colleagues have advanced the theory that cholinergic mechanisms play a critical role, particularly in certain attentional deficits after TBI (Arciniegas et al. 1999) and have reported successful use of donepezil in some individuals with TBI (Arciniegas et al. 2001). Thus, there appears to be increasing evidence, both theoretical and clinical, that suggests that the cautious, empiric use of cholinergic and catecholaminergic agents is warranted for the treatment of chronic memory and attentional deficits. It is possible that specific genetic profiles contribute to response to neurotrauma and cognitive outcomes. As described above, the neuropathology of TBI and the neu- rochemistry of memory and attention suggest that genes that modulate cholinergic and catecholaminergic function and systems important to neural repair and plasticity are attractive candidate genes (McAllister and Summerall 2003). My group has hypothesized that individuals with alleles that reduce central catecholaminergic/cholinergic tone and neuronal repair/plasticity may well show greater 300 TEXTBOOK OF TRAUMATIC BRAIN INJURY cognitive deficits shortly after injury and less improvement in cognitive function over time than those with al- ternative alleles. Preliminary data for this hypothesis are encouraging (McAllister et al. 2004). Furthermore, the effect of these alleles may be additive, such that individuals with more of the “adverse” alleles may have poorer cognitive outcomes. Psychoeducation Often, the most effective intervention in patients with active neurobehavioral sequelae is a careful explanation of the pathophysiology, typical sequelae, and time course of recovery associated with minor brain injury (Kelly 1975; Minder- houd et al. 1980, 1997; Mittenberg et al. 1996; Paniak et al. 1998a; Wade et al. 1997, 1998; Wrightson 1989). Problems with slowing, attention, and memory, especially in the first 3–6 months, should be described. The potential for longer- term difficulties should be mentioned. This should be done soon after the injury and is best done in the presence of family, friends, or significant others (see Wrightson 1989). The realistic setting of goals for return to major activities is a difficult process that must be individualized for each patient. Psychiatrists often are involved in the later stages of the process, by which time there is frequently an unpleasant dynamic operating in which various individuals (including family, friends, employers, insurance carriers, and health care workers) are questioning the validity of complaints on the basis of the seemingly “minor” nature of the injury and the patient’s healthy appearance. Validating the complaints of the patient without undue fostering of illness behavior can be a difficult and lengthy process. Medical-Legal Issues Psychiatrists increasingly are involved in the assessment of patients with mild brain injury, often at the request of attorneys or insurance carriers (see Chapter 33, Ethical and Clincal Legal Issues). Typically, an opinion is requested about whether the nature of the patient’s complaints, as well as their severity and duration, is consistent with what is known about the injury. The evaluation of such cases is time consuming and requires procurement and perusal of all pertinent records, including school and/or employment records, testing and evaluation, accident and emergency transport reports, and subsequent treatment records. When possible the cli- nician should interview the patient and others who knew the patient before the event. Results of neurodiagnostic tests must be evaluated. If they have not been performed, an MRI, careful neuropsychological evaluation, EEG, and EPs can be helpful in es- tablishing the presence of brain injury. All of these studies, as previously noted, are not always abnormal in the presence of obvious brain injury. Furthermore, even when abnormal, these studies may not reveal abnormalities that are pathognomonic for mild brain injury. Be- cause few patients have these tests performed both before and after their injury, it is difficult to be certain that such abnormalities were caused by the traumatic event in question. Thus, the foundation of such evaluation remains the careful assessment of premorbid function; delineation of the type, location, and severity of the trauma; documentation of the profile and time course of subsequent changes in cognitive, behavioral, and somatic areas; and integration of this information with the appropriate neurodiagnostic studies. Many of the latter may not have been done until weeks to months after the injury, making the yield from such studies lower than if performed within a week or so of the trauma. Thus, even in the absence of positive neurodiagnostic findings, the history of a documented injury, with subsequent onset of the symptoms described above, should enable a reasonable opinion to be given about the relationship between the injury and the current clinical picture. Summary Mild brain injury is a significant public health problem. It can result in an array of common neurobehavioral sequelae. Several points in this chapter are worth highlighting: • Well over a million people experience a mild TBI in the United States each year. • Limited human data and more extensive animal data suggest that minor brain injury produces neuropatho- logical changes to a lesser extent but of similar quality and location to those seen in more severe brain injury. • Mild brain injury is associated with impairments in speed of information processing, attention, and memory. These deficits are most pronounced in the initial days to weeks after the injury. Most patients show a rapid, progressive improvement over the subsequent 1–3 months. A small percentage of patients have de- monstrable long-term sequelae. • A variety of predictable cognitive, somatic, and behavioral complaints, known as postconcussive symptoms, are seen subsequent to brain injury of all levels of severity. After mild brain injury, most patients show progressive resolution of these symptoms over the subsequent 1–6 months. A small but significant percentage has persistent symptoms 12 months or longer. A history of prior 309 16 Seizures Gary J. Tucker, M.D. THE PRESENCE OF posttraumatic seizures is a major complication in the recovery of the brain injury patient. It not only adds further cognitive and behavioral changes (in addition to the brain injury itself), it also connotes a worse prognosis. The many cognitive problems faced by the patient with traumatic brain injury (TBI), such as the inability to sustain attention (Parasuraman et al. 1991) and impairments in social interaction (Marsh and Knight 1991; Sarna 1980), are further exacerbated by the presence of seizures. Seizures in themselves can cause marked effects on cognitive functions and social performance (Matthews 1992). In addition, anticonvulsant medications can also cause cognitive changes (Farwell et al. 1990; Gillham et al. 1988; Meador et al. 1990). Aside from the cognitive effects, seizures have an enormous psychological impact on the patient’s self-confidence in social interactions because of the stigma that has been associated with seizure disorders (Temkin 1971). Seizures, the medications used to treat them, and the psychological impact of seizures significantly complicate the rehabilitation of the brain-injured patient. Epidemiology Several studies have examined the occurrence of seizures after TBI. TBI associated with closed head injuries (i.e., when the dura has not been penetrated) has a 5% incidence of posttraumatic seizures that can occur any time after brain injury; however, with open head injury (when the dura has been penetrated), 30%–50% of the patients develop posttraumatic seizures (Jennett 1975; Lishman 1987). Jennett (1975) estimated that only 1% of patients will develop seizures if no seizure occurs during the first week after injury; however, if a seizure occurs during the first week, the lifetime incidence increases to 25%. Tech- nically, if seizures occur after the first week postinjury and are recurrent, the term posttraumatic epilepsy should be used, but the literature uses the terms posttraumatic seizures and posttraumatic epilepsy interchangeably, and most seem to favor the use of posttraumatic seizures. Whatever term is used, there is almost no information in the literature on how many seizures a particular patient will have post-TBI. In those patients who develop seizures post- TBI, the long-term prognosis is good. Fifty percent of patients with posttraumatic seizures will no longer have seizures 5–10 years postinjury, 25% will have good seizure control while taking medication, and only 25% will continue to have seizures. The occurrence of seizures depends on the severity and type of the brain injury. Annegers et al. (1980) provided the best available epide- miological data on posttraumatic seizures from a large community-based survey using the community database developed by the Mayo Clinic. They surveyed all medical records of patients with reported brain injury in Olmsted County, Minnesota, from 1935 to 1974. This included all patients with head trauma who were admitted to a hospital or emergency department, who were seen as outpa- tients, or for whom a home visit was made. In this manner, they collected a total sample of 3,587 patients with TBI, 840 of whom were excluded either because of death within the first month or a prior history of epilepsy or TBI, or because the seizure was the result of other condi- tions. The remaining 2,747 patients with brain injuries were followed longitudinally for the development of posttraumatic seizures. Thus, the authors avoided one of the major pitfalls in many of the studies of patients with brain injury—that is, the lack of data on those patients lost to follow-up. However, this study was not without methodological problems. First, the authors noted the extreme complexity in estimating the risk of seizures due to the absence, at that time, of standardized definitions of brain trauma or severity of injury. (This lack of definition is 310 TEXTBOOK OF TRAUMATIC BRAIN INJURY present in most of the literature before the development of the standardized rating scales for TBI.) Second, there is the possibility that this was an atypical sample because it was obtained from a major neurosurgical center. Third, the authors noted the poor follow-up for most patients with brain trauma. Last, it was often unclear whether the patient had a history of seizures before the injury. In spite of these methodological concerns, this community-based study is still valuable in presenting a most complete picture of the longitudinal course of patients with brain trauma. The patients were grouped into the following three categories: • Mild brain trauma (1,640 patients)—defined as those without skull fractures and without loss of consciousness, or with a period of posttraumatic amnesia of less than 30 minutes. • Moderate brain trauma (912 patients)—those patients who had more than a 30-minute period of unconsciousness or posttraumatic amnesia or had a skull fracture, or both. • Severe brain trauma (195 patients)—evidence of brain contusion, hematoma, or more than 24 hours of unconsciousness. With this classification, Annegers et al. (1980) followed the patients over the 40-year period from 1935 to 1974. Seizures developed in 51 patients during the first 4 years after injury. The risk for patients with severe injury (7.1% in the first year and 11.5% within the next 5 years) was much greater than for those with moderate (0.7% in the first year and 1.6% within 5 years) or mild injury (0.1% in the first year and 1.6% within 5 years). In chil- dren (younger than 14 years) with severe injury, the incidence of posttraumatic seizures was 30% compared with only 10% in adults with severe brain injury. Thus, the age of the patient and the severity of the injury are crucial de- terminants of the subsequent development of posttraumatic seizures. In 1998, Annegers et al. reported the results of an ex- tension of this study involving those who experienced TBI up to 1984, with a follow-up of these additional cases through 1994; in this manner, the sample was increased to 4,541 patients. In the total sample, 97 patients had unpro- voked seizures post-TBI; 22 of these had single seizures, and 75 had multiple seizures. The 30-year cumulative incidence for seizures post mild TBI was 2.1% (3.1% for the first year and 2.1% for the next 4 years), 4.2% for moderate TBI, and 16.7% for severe TBI. Brain contusion, subdural hematoma, and age older than 65 years were the major risk factors for seizures, whereas skull fracture and prolonged unconsciousness were slightly less so. Apparently, the early treatment of TBI can affect the occurrence of seizures as well. Temkin et al. (1990) treated patients with severe brain trauma with either phenytoin or a placebo immediately after the injury. Between drug load- ing and the seventh day after the trauma, 3.6% of the phenytoin group and 14.2% of the control group developed seizures. In the group in whom phenytoin was continued after day 8 through the end of the first year, 21.5% of the phenytoin group but only 15.7% of the placebo group had seizures. At the end of the second year, the seizure rates were 27.5% for the phenytoin group and 21.1% for the control group (these differences were statistically significant). The authors hypothesized that phenytoin exerts a prophylactic effect on reducing seizures during the first week post severe brain injury but may increase seizure frequency with prolonged treatment. In addition, patients who continued taking phenytoin longer than 1 week posttrauma had more cognitive deficits than those whose phenytoin was discon- tinued after the first week. The authors concluded that the drug has an early suppressive effect but not a true prophylactic one. In 1999, Temkin et al. repeated this study. Within 24 hours postinjury, 132 patients received 1-week treatment with phenytoin, 120 patients received 1-month treatment with valproate, and 127 received a 6-month course of treatment with valproate. The rate of early seizures was low and similar to that in the study by Annegers et al. (1998). The rates of late seizures (after 1 week) did not differ in the treatment groups (15% of the group taking phenytoin, 16% of the group taking valproate for 1 month, and 24% of the group taking valproate for 6 months). Al- though there was no difference in the treatment groups in the occurrence of side effects (e.g., coagulation problems or liver impairments), there was a trend toward a higher mortality rate in the valproate groups (7.2% vs. 13.4%). A study by Dikmen et al. (2000) also showed few cognitive effects of valproate but found a trend toward increased mortality with the use of valproate. A subsequent meta-analysis of controlled trials of post-TBI seizure prevention in late- occurring seizures (Temkin 2001) showed effectiveness for phenytoin and carbamazepine but not for valproate. There have been no studies to date evaluating the use of the more recently developed anticonvulsants such as gabapentin, lamotrigine, or topiramate for the treatment of posttraumatic seizures (Bazil 2001; Martin et al. 1999). In light of the findings with valproate, the newer drugs probably should be used with caution until detailed studies in patients with TBI are available. In view of the cognitive changes associated with phenytoin and other anticonvulsants, their continued use after the first week following brain injury may be contraindicated. The American Academy of Physical Medicine and Rehabil- itation (Brain Injury Special Interest Group of the American Seizures 311 Academy of Physical Medicine and Rehabilitation 1998) and the American Association of Neurological Surgeons (Brain Trauma Foundation 2000) recommend that only phenytoin, phenobarbital, or carbamazepine be used to prevent early (1 week post-TBI) seizures in patients without penetrating injuries of the dura and that no antiepileptic drug be used pro- phylactically in anticipation of late seizures. Diagnosis A major diagnostic indicator of a seizure disorder is an abnormal electroencephalogram (EEG), generally involving paroxysms or spikes, either focal or generalized (Tucker 2002). The presence of an epileptiform EEG pattern occurs more frequently with penetrating brain injury. It is important to emphasize, however, that even several EEGs will reveal seizure activity in only 41% of patients with symptomatic seizures (Desai et al. 1988). Consequently, this relatively low sensitivity of the EEG suggests that the presence or absence of an epileptiform spike should not be the sole factor in determining disability benefits for individuals with epilepsy, and one should not use such abnormalities as an entry criterion for research (Desai et al. 1988). Jabbari et al. (1986) performed EEG evaluations on 515 Vietnam War veterans 12–16 years after penetrating brain injury. They found that 42% of the subjects had abnormal EEGs, but only 9% demonstrated epileptiform findings. There was a significant correlation between EEG findings and the extent of brain volume loss visualized by computed tomography. All patients with anterior temporal or central spike foci experienced posttraumatic seizures. Focal slowing, as would be expected, correlated significantly with localized neurological deficits such as hemiplegia (Jabbari et al. 1986). Salazar et al. (1985) studied 421 Vietnam veterans with penetrating brain injuries. Posttraumatic seizures developed in 53% of these patients. However, only 12% of patients with seizures had EEG results diagnostic of a seizure disorder. The authors concluded that the EEG might not always be diagnostically helpful. The severity of the injury increases the probability of EEG abnormality. Koufen and Hagel (1987) evaluated 100 patients with posttraumatic late seizures who also had at least 1 week of amnesia after brain injury and found that 95% had focal EEG abnormalities, 70% of which were bilateral. Many of these patients had focal neurological symptoms and skull fractures as well. The EEG nor- malized in 48% of patients after 2 years, but foci persisted in 22% of the patients, and 30% remained diffusely abnormal. The most common abnormalities were delta rhythms (85%) and focal dysrhythmias with temporal lo- calization (58%–82%, depending on criteria). Although many clinicians have the impression that most posttraumatic seizures are generalized, all types of partial seizures can also occur (Salazar et al. 1985) and, in fact, are equal in presentation to the generalized seizures. The diagnosis of seizure disorders is a clinical diagnosis because the best diagnostic test is to observe someone having a seizure. All evaluations of suspected seizure disorders should include regular EEGs, especially a sleep EEG, which is four times more likely to show an abnormality than a waking EEG (Ba- zil et al. 2000; Crespel et al. 2000; Foldvary et al. 2000; Gibbs and Gibbs 1952; Malow et al. 2000). Although some researchers advocate the use of na- sopharyngeal leads, these actually increase the rate of abnormal findings by only 10% (Bickford 1979). Although a recent study by Pacia et al. (1998) reports an increased diagnostic yield for the diagnosis of temporal lobe seizures with sphenoidal leads, a previous study (Sadler and Goodwin 1989) shows that submandibular notch place- ment on the buccal skin surface is as effective as either na- sopharyngeal or sphenoidal leads. Prolactin levels have been shown to rise in patients with seizures and may be of some use in diagnosis (Dana- Haer and Trimble 1984). Recent studies using single- photon emission tomography show approximately a 30%–40% chance of demonstrating a seizure focus inter- ictally and a 70%–80% chance if the study is done ictally (Lassen and Holm 1992; Lee et al. 1988). This may prove to be a useful technique for the confirmation of seizure foci in patients with TBI. Pathogenesis Although the etiology of posttraumatic seizures is not certain, the most frequently associated factor is the actual disruption of brain tissue. Almost any injury that penetrates the dura and the cortex results in a higher incidence of posttraumatic seizures. The incidence of posttraumatic seizures in penetrating injuries reported in the literature varies from 28% to 50% (Salazar et al. 1985). Some seizure disorders can be treated successfully by the surgical removal of cortical scar tissue (Spencer and Katz 1990). We can infer that cortical disruption, scarring, or irritability and the release of various endogenous neurotoxins (e.g., glutamate) can lead to the onset of posttraumatic seizures. Vespa et al. (1998), using implanted extracellular microdialysis probes, studied 17 patients with severe TBI. They found that extracellular glutamate was increased in these patients, particularly in relation to seizure activity. Heikkmen et al. (1990) noted that although the severity of injury was most predictive of the development of early seizures (within the first 7 days postinjury), other specific 312 TEXTBOOK OF TRAUMATIC BRAIN INJURY factors were also associated with the onset of seizures, including periods of unconsciousness over 24 hours, skull fracture with dural tears, contusions, hematomas, and/or hemorrhage. The presence of subcortical atrophy or impaired local cerebral blood flow was most predictive of late- onset seizures occurring in the 3- to 12-month period after injury (Table 16–1). There is some recent evidence that mesial temporal sclerosis may be important in the development of post-TBI seizures (Marks et al. 1998). Diaz-Arrastia et al. (2000) studied 23 patients with intractable epilepsy after TBI and found that 35% had hippocampal sclerosis, and 2 of the patients had temporal lobectomies with relief of seizures. In a prospective, observational study of 647 individuals admitted to trauma centers after TBI who had abnormal CT findings or a Glasgow Coma Scale score of 10 or lower during the first 24 hours, 66 patients developed a late seizure during a 24-month follow-up period. Patients with biparietal contusions (66%), dural penetration with bone and metal fragments (62.5%), multiple intracranial operations (36.5%), multiple subcortical contusions (33.4%), subdural hematoma with evacuation (27.8%), midline shift greater than 5 mm (25.8%), or multiple or bilateral cortical contusions (25%) (Englander et al. 2003) had the highest cumulative probability for the development of seizures. Mazzini et al. (2003) found that the degrees of hydro- cephalus and temporal lobe hypoperfusion (found on single-photon emission tomography) were risk factors for the development of late posttraumatic seizures. After severe brain injury, hyperexcitable neurons may produce an epileptic focus between the time of the trauma and the seizure occurrence (Kuhl et al. 1990). There is biochemical evidence from animal studies (Mori et al. 1990) that the occurrence of posttraumatic seizures may be related to a breakdown of red blood cells and hemoglo- bin in the cerebral cortex, leading to release of free hy- droxyl radicals into the central nervous system, subsequently affecting the neuronal membranes and leading to seizures. Although a recent review (Maas 2001) found that no study had demonstrated any positive effect with any neuroprotective antioxidants, it was also noted that the heterogeneity of the brain trauma group may prevent the demonstration of effectiveness. Weiss et al. (1982) noted a higher incidence of cerebral vascular accidents in patients with posttraumatic epilepsy. Proctor et al. (1988) used an experimental model for seizure development in closed head injury. Their research involved cats subjected to significant atmospheric fluid percussion impact (3.5 at- mospheres administered to the cerebral cortex). They found that there were significant differences in seizure development related to measures of oxygenation and cy- tochrome A and adenosine triphosphate. It remains unclear why one person develops seizures and another, with the same degree of brain trauma, does not. Weiss et al. (1982) and Salazar et al. (1985) reported no genetic predisposition or a family history of seizures in those who developed seizures. Inheritance of the APOE ε4 allele was found to be associated with increased risk of late posttraumatic seizures (Diaz-Arrastia et al. 2003). Two recent animal studies (Koh et al. 1999; Schmid et al. 1999) demonstrated that neonatal seizures, even though they did not cause cellular injury, predisposed the animals to brain-damaging effects of seizures in later life. Cer- tainly age, as noted in the section Epidemiology, seems to be a factor, with both younger patients (younger than 14 years) and older patients (older than 65 years) being more prone to posttraumatic seizures (Annegers et al. 1998). It is also unclear why the prolonged prophylactic use of anticonvulsants leads to a greater incidence of seizures (Temkin et al. 1999). Prognosis What are the implications of seizures for the person with TBI? In most cases, seizures indicate that the person has had a more severe brain injury. This factor constantly leaves one with the question of whether the seizures further complicate the clinical course of a patient with severe brain injury or simply reflect the more extensive injury. In favor of the latter, Dikmen and Reitan (1978) reported TABLE 16–1. Factors associated with early and late seizures after traumatic brain injury Early seizures (within the first week) Late seizures (after the first week) Younger age (especially<5 years) Age>65 years Posttraumatic amnesia>24 hours Posttraumatic amnesia>24 hours Skull fracture (especially depressed) Depressed skull fracture Intracranial hemorrhage Hematoma Seizures during first week posttrauma Early seizures Penetrating injury Penetrating injury High Glasgow Coma Scale score Source. Data from Heikkmen ER, Routy HS, Tolonen H, et al: “Devel- opment of Posttraumatic Epilepsy.” Stereotactic and Functional Neurosur- gery 54/55:25–33, 1990. Seizures 313 that a group of posttraumatic epilepsy patients with cortical defects on neuropsychological testing had a worse prognosis than those with posttraumatic epilepsy who showed no cortical deficits. The patients with cortical deficits and seizures would be expected to do poorly because they are usually the most severely injured. Corkin et al. (1984) showed that patients with posttraumatic epilepsy had shorter life expectancies than brain-injured patients without seizures. Walker and Blumer (1989) followed, over a 40-year period, 244 World War II veterans who had penetrating brain injuries and seizure disorders and found that 101 had died (a figure much higher than expected in a general population). Thus, patients with posttraumatic epilepsy have an increased mortality. Weiss et al. (1982) confirmed this increased mortality in patients with post-TBI seizures and also demonstrated that 25% of all brain injury survivors showed deterioration in cognitive functions and earlier signs of aging. The prognosis for posttraumatic seizures is good. Walker and Blumer (1989) studied a group of World War II veterans with TBI and noted that in those with seizures, 75% had no seizures after 10 years. They also pointed out that the type of injury that occurs in the military differs from civilian brain injuries. Civilian brain injuries are usually in the frontal-temporal region, whereas those associated with military injuries are usually penetrating and in rolandic (motor) and parietal regions and involve several lobes. Thus, the mortality and neurological deficit studies may not be generalizable to civilian populations. Weiss et al. (1986), in a 15-year follow-up study of 520 veterans, noted that 95% of the patients were seizure free 3 years after the trauma. The presence of substance or al- cohol abuse was not a factor in the cessation of seizure activity. However, Salazar et al. (1985) noted that seizures could occur up to 15 years posttrauma in a group of Viet- nam veterans. Although the majority of veterans (57%) developed seizures within the first year of injury, 15% did not develop seizures until 2 years after brain injury, and 18% developed seizures within 5 years (Weiss et al. 1986). Armstrong et al. (1990) surveyed 300 consecutive brain trauma admissions to a rehabilitation hospital and, after ex- cluding those with penetrating brain injuries or prior his- tories of epilepsy, found 87 patients with posttraumatic epilepsy (37%) and 151 patients (63%) with brain trauma and no posttraumatic epilepsy. In comparing these patients, they noted that the posttraumatic epilepsy group had a greater incidence of males than females. There were no differences between the two groups in frequency of skull fractures, hematomas, or hemorrhages, or in Halstead- Reitan Neuropsychological Test Battery results; however, there were marked differences in outcome in the patients who had posttraumatic epilepsy. Patients with posttraumatic epilepsy had a longer stay in the hospital, more diffi- culty with receptive language and intelligibility, decreased ability to perform activities of daily living, decreased motor function, and more mood and affective changes, as well as more problems with orientation. Although all of the patients made gains from admission to discharge, the posttraumatic epilepsy group started lower and ended lower, a further indication that posttraumatic seizures may simply be a marker of TBI severity. Table 16–2 summarizes factors associated with the presence of seizures in patients with brain injury. The onset of seizures after TBI is a poor prognostic sign for general recovery, although, as noted, the seizures themselves often remit during the recovery years. The presence of focal neurological and cognitive deficits markedly wors- ens the prognosis. However, it is difficult to determine the exact contribution of the seizures to this poor prognosis because, as noted, these patients usually have had more severe initial brain injuries. Psychopathology Seizure disorders are associated with increases in psychopathology (McKenna et al. 1985; Trimble 1991; Tucker 2002) as is TBI (van Reekum et al. 2000). It is not clear if the presence of seizures in patients with TBI increases the risk for the development of psychopathology. The psychopathology associated with seizure disorders can range from personality changes to frank episodic or chronic psychosis. Patients with seizure disorders, when assessed in large studies, often show statistically significant increased incidence of such personality traits as impul- siveness and irritability, emotional lability, hyposexuality, hypergraphia, viscosity, paranoia, nightmares, fluidity of thinking, chronic pain, aggression, and philosophical or religious preoccupation. Those individuals who developed posttraumatic seizures had a significantly higher incidence of personality disorders, including uninhibited TABLE 16–2. Factors associated with the presence of seizures in brain-injured patients Increased levels of Decreased levels of Rehabilitation hospital stays Communicative ability Mood and affective disorders Motor function Cerebrovascular accidents Activities of daily living Orientation Life expectancy [...]... set of higher-order capabilities that are considered the domain of the frontal TEXTBOOK OF TRAUMATIC BRAIN INJURY TABLE 17–3 Aspects of executive functions potentially impaired after traumatic brain injury Goal establishment, planning, and anticipation of consequences Initiation, sequencing, and inhibition of behavioral responses Generation of multiple response alternatives (in contrast to preserverative... cognitive 3 15 Seizures TABLE 16–4 Anticonvulsant Daily doses, effective blood levels, and serum half-lives of anticonvulsants Serum half-life (hours) Usual daily dose (mg) Effective blood level (µg/mL) 200–2,000 6–12 Clonazepam 1–10 0.01–0.07 Ethosuximide 1 ,50 0–2,000 40–100 Gabapentin 1,800–3,600 4–16 5 7 Lamotrigine 100 50 0 2–16 12–60 60–200 10–40 96 Phenytoin 100–600 10–20 24 Primidone 250 –1 ,50 0 5 15 12... function that often follow severe TBI (Levine et al 2000; Pachalska et al 2002; Sbordone 2001) These vital aspects of behavior are linked to the integrity of ventral frontal regions, which often bear the brunt of TBI-related damage and yet fall outside the domain of routine cognitive testing Given the prominence of the orbitofrontal cortex in emotional processing and mediation of stimulus-reward associations... feature of the model presented here is the core circuit (Marin 1996b), a postulated subsystem of the forebrain, composed of the anterior cingulum, nucleus accumbens (NA), ventral pallidum (VP), and ventral tegmental area (VTA) (Figure 18–1) One hypothesis, based on a growing body of research 340 FIGURE 18–1 TEXTBOOK OF TRAUMATIC BRAIN INJURY Motivational circuitry The core circuit (shaded) consists of anterior... working memory, and––in particular––frontal control functions are germane Stuss and Levine (2002) summarized that left prefrontal injury is associated with simplified, repetitive, and impoverished discourse In contrast, right prefrontal lesions may produce amplification of detail, insertion of irrelevant elements, and a tendency toward TEXTBOOK OF TRAUMATIC BRAIN INJURY TABLE 17 5 Medications reported... true for some aspects of memory, but the results to date TEXTBOOK OF TRAUMATIC BRAIN INJURY are more mixed (Whyte et al 2002) In general, benefits of stimulants appear quite modest when compared to the robust effects observed in primary ADHD In an effort to circumvent the shortcomings of earlier work, Whyte’s group systematically explored the domain of attention in two studies of patients with residual... decrements in volunteers given either tryptophan, a 5- hydroxytryptamine (5HT) precursor, or fenfluramine, a 5- HT agonist (Luciana et al 2001) The role of the 5- HT system in opposing certain dopamine-mediated cognitive functions, such as working memory, was cited to explain these findings (Luciana et al 2001) Two reports have indicated positive effects of lamotrigine on cognition after TBI This agent alters... premorbid levels of cognitive performance (McAllister et al 1999) In contrast to the spontaneous recovery seen in mild TBI, a recent longitudinal study of moderate to severe TBI confirmed the presence of substantial memory impairments in 50 % of subjects at 5 years postinjury (Millis et al 2001) Impairments of Frontal Executive Functions The term executive functions refers to a set of higher-order capabilities... in hand with the treatment of con- TEXTBOOK OF TRAUMATIC BRAIN INJURY ditions—for example, stupor, delirium, depression, dementia—that lead to these disorders Such treatments may include a variety of behavioral techniques (Campbell and Duffy 1997; Giles and Clark-Wilson 1988, 1993) or specialized cognitive rehabilitative approaches to accomplish, for example, enhancement of attention or performance... reflects a number of factors, the most important being 1) the severity of diffuse axonal injury, as indicated by the length of posttraumatic amnesia (PTA), the extent of generalized atrophy; and 2) the location, depth, and volume of focal cerebral lesions (Katz and Alexander 1994; Wilson et al 19 95) Other critical factors include the patient’s age, preexisting morbidities, and the occurrence of significant . estimating the risk of seizures due to the absence, at that time, of standardized definitions of brain trauma or severity of injury. (This lack of definition is 310 TEXTBOOK OF TRAUMATIC BRAIN INJURY present. the effects of traumatic brain injury. 324 TEXTBOOK OF TRAUMATIC BRAIN INJURY may be obtained by using a measure of “everyday memory” (Wills et al. 2000), which includes analogues of daily tasks. (1997) reported the results of a double-blind, placebo-controlled, crossover study of the effects of 0. 25 mg/kg methylphenidate on measures of attention in 19 TBI subjects of mixed injury severity.

Định dạng
Số trang	55
Dung lượng	886,36 KB