ORIGINAL RESEARCH
Seizures and Epileptiform Discharges in Patients With Acute Subdural Hematoma Sebastian Pollandt,* Bichun Ouyang,† Thomas P. Bleck,†‡§k and Katharina M. Busl* Section of Neurocritical Care, Department of Neurological Sciences, Rush University Medical Center, Chicago, Illinois, U.S.A.; Departments of †Neurological Sciences, ‡Neurosurgery, §Internal Medicine, and kAnesthesiology, Rush University Medical Center, Chicago, Illinois, U.S.A. *
Purpose: Subdural hematomas (SDH) are associated with seizures and epileptiform discharges, but little is known about the prevalence and impact of seizures, status epilepticus (SE), and epileptiform discharges on outcomes in patients with isolated acute SDH (aSDH). Methods: Continuous EEG reports from 76 adult patients admitted to Rush University Medical Center with aSDH between January 2009 and March 2012 were reviewed. Clinical and radiographic findings, comorbidities, treatment, and outcome parameters, such as mortality, discharge destination, need for tracheostomy/percutaneous endoscopic gastrostomy placement, and length of stay (LOS), were assessed. Univariate and multivariate analyses were performed to assess the impact of clinical seizures, SE, and epileptiform EEG on outcomes. Results: Of 76 patients with aSDH who underwent EEG monitoring, 74 (97.4%) received antiseizure prophylaxis. Thirtytwo (41.1%) patients had seizures, most of which were clinical seizures. Twenty-four (32%) patients had epileptiform EEG
ubdural hematoma (SDH) is one of the most common diagnoses in neurointensive care units (ICUs), with more than 90,000 hospitalizations for SDH occurring annually in the United States (Frontera et al., 2011). With the increasing proportion of elderly in the population of Western countries, the prevalence of SDH is expected to grow further (Frontera et al., 2011). The association of SDH and seizures is well known; however, most data originate from chronic SDH (Sabo et al., 1995; Wang et al., 2008) or aSDH in the setting of severe traumatic brain injury (TBI), resulting in a wide variation in reported incidence and impact of seizures (Annegers and Coan, 2000). Much less data exist on patients with isolated aSDH and seizures (Rabinstein et al., 2010), and little is known about the impact of seizures, status epilepticus (SE), and epileptiform findings on outcomes for patients with aSDH. The focus of this study was to determine the prevalence of seizures, SE, and epileptiform EEG findings in a large cohort of patients with isolated aSDH and examine their impact on outcomes.
S
findings. Clinical or nonconvulsive SE was diagnosed in 12 (16%) patients. Clinical seizures were not associated with outcome parameters. Epileptiform EEG findings were independently associated with longer hospital LOS (13 vs. 8 days, P ¼ 0.04) and intensive care unit LOS (10 vs. 4 days, P ¼ 0.002). The SE also predicted longer intensive care unit LOS (10 vs. 4 days, P ¼ 0.002). Neither epileptiform EEG nor SE was significantly related to mortality, discharge destination, or need for tracheostomy/ percutaneous endoscopic gastrostomy placement. Conclusions: Seizures and epileptiform EEG findings are very common in patients with aSDH despite antiseizure prophylaxis. While clinical seizures did not affect outcomes, the presence of epileptiform EEG findings and SE was independently associated with longer intensive care unit LOS and hospital LOS. Key Words: Subdural hematoma, EEG, Seizures, Status epilepticus, Adult, Epileptiform discharges. (J Clin Neurophysiol 2017;34: 55–60)
MATERIALS AND METHODS Participants We conducted a retrospective review of all adult patients admitted to Rush University Medical Center with an admission diagnosis of aSDH between January 2009 and March 2012. Patients with chronic SDH or concomitant intracranial injuries (including parenchymal contusions, global cerebral edema, and diffuse axonal injury, when noted on computed tomography of the head on admission) were excluded at data collection. Patients with insufficient EEG data were excluded. Acute or acute-onchronic SDH was considered aSDH and included. Patients with aSDH who underwent EEG monitoring were selected for further analysis. The study was approved by the local institutional review board, including a waiver for obtaining informed consent to collect and publish data.
Clinical Measures The abstract to this manuscript was presented in form of 2 posters at the Annual Neurocritical Care Society Meeting, Scottsdale, Arizona, October 7–10, 2015. The authors have no funding or conflicts of interest to disclose. Address correspondence and reprint requests to Sebastian Pollandt, MD, Department of Neurological Sciences, Section of Neurocritical Care, Rush University Medical Center, 1725 West Harrison St, POB Suite 1106, Chicago, IL 60612, U.S.A.; e-mail: [email protected]. Copyright Ó 2016 by the American Clinical Neurophysiology Society
ISSN: 0736-0258/16/3401-0055
DOI 10.1097/WNP.0000000000000311
clinicalneurophys.com
Baseline demographics, medical history, comorbidities, clinical and radiographic findings, prophylactic antiseizure drug administration, treatment of aSDH, discharge disposition, and hospital length of stay (LOS) were abstracted from chart review. Daily progress notes (including progress notes by treating or consulting physician teams, nursing staff, and therapy team notes) were reviewed for occurrence of clinical seizures, symptoms concerning for seizures, or changes in clinical
Journal of Clinical Neurophysiology Volume 34, Number 1, January 2017
55
S. Pollandt, et al.
Seizures and Epileptiform Discharges in Acute SDH
condition prompting the treating physician to obtain an EEG. An occurrence of clinical seizure was concluded for characteristic transient symptoms and exclusion of alternative etiologies or if the clinical event in question had an electrographic correlate on EEG. Nonconvulsive seizures were recorded if a seizure per EEG criteria did not have a clinical correlate on video review by the electroencephalographer. For patients who underwent surgical treatment for their aSDH, temporal relation of the clinical event to surgical treatment was noted. All patients received a Glasgow Coma Scale score on admission. Radiologic measures were made on the worst preoperative CT scan on axial images and included maximum SDH diameter and degree of midline shift measured in millimeters at the level of the pineal gland. Outcome parameters included mortality, discharge destination, defined as favorable (home or rehabilitation) versus unfavorable (nursing home, skilled nursing facility, or long-term acute care facility), need for tracheostomy and percutaneous endoscopic gastrostomy placement, and ICU and hospital LOS.
EEG Measures All EEG reports were generated by a board-certified epileptologist at this institution and reviewed by a boardcertified epileptologist (S.P.) for occurrence of clinical or electrographic seizures, SE, interictal epileptiform discharges, and focal slowing. EEG tracings were rereviewed for selected studies by one of the authors (S.P.). EEGs included in the analysis were routine EEGs, multiple routine EEGs, continuous EEGs, or a combination. Epileptiform changes were defined as interictal epileptiform discharges (transients distinguishable from background activity with a characteristic spiky morphology; Noachtar et al., 1999) or electrographic seizures.
Statistical Analyses x2
Univariate analyses were performed with or Fisher exact test for categorical variables and Wilcoxon rank-sum test for continuous variables as indicated. Multivariate logistic regression analyses with stepwise selection, including all factors with a P value of ,0.15 from the univariate analyses, were used to identify baseline factors that independently predicted outcome measurements and to further examine the relationship of epileptiform changes to outcome parameters. For LOS analyses, binary logistic regression models were used for ICU-LOS ( $ 6 vs. ,6 days) and for hospital LOS ( $ 10 vs. ,10 days) based on the median LOS, respectively. For the analysis of discharge destination, three patients were excluded based on their residence at an unfavorable discharge destination before admission. All statistical analyses were performed with SAS 9.2 (SAS Institute Inc, Cary, NC). A P value ,0.05 was considered significant.
RESULTS Study Population Of the 300 patients admitted with aSDH during the study period, 77 patients underwent EEG monitoring. One patient was excluded from the analysis because of insufficient available EEG data. When compared with the entire cohort of patients with 56
aSDH, patients with history of hypertension, prior intracerebral hemorrhage, prior stroke, history of seizures or epilepsy, lower admission Glasgow Coma Scale, seizure at aSDH onset, and intubated patients were more likely to receive an EEG (Table 1). The most common indications for EEG monitoring were alterations in mental status, seizure-like activity (focal movements or generalized convulsions), or new focal findings, such as weakness or aphasia (Table 2). In the overall cohort, eight patients had clinical seizures but did not undergo EEG TABLE 1. Baseline Characteristics of Patients With aSDH (n ¼ 299) EEG Performed Yes (n ¼ 76) No (n ¼ 223) Age, median (IQR), years Sex male, n (%) Medical history, n (%) Diabetes mellitus Hypertension Coronary artery disease Congestive heart failure Renal disease Liver disease Alcohol abuse Smoking Dementia Prior intracerebral hemorrhage Prior stroke Prior seizures or epilepsy Clinical and radiographic characteristics on admission, n (%) Glasgow coma scale score 13–15 9–12 3–8 Median (IQR) Intubated SDH size,* median (IQR), mm SDH side Left Right Bilateral SDH location Frontoparietal Holohemispheric Temporal Tentorial/falcine Other lobar Midline shift,† median (IQR), mm Surgical SDH evacuation Seizures at onset of aSDH
P
70 (23) 49 (64.5)
71 (24) 135 (60.5)
0.81 0.54
19 65 22 12 11 3 14 14 12 10
61 161 59 22 32 12 37 42 30 7
0.69 0.02 0.67 0.16 0.98 0.77 0.74 0.94 0.61 0.003
(25) (85.5) (29.0) (15.8) (14.5) (4.0) (18.4) (18.4) (15.8) (13.2)
17 (22.4) 16 (21.1)
(27.4) (72.2) (26.5) (9.9) (14.4) (5.4) (16.7) (18.8) (13.5) (3.1)
29 (13) 23 (10.3)
0.05 0.02
,0.0001 46 19 11 13 20 12.3
(60.5) (25) (14.5) (5) (26.3) (9.5)
183 11 29 15 32 12.2
(82.1) (5) (13) (1) (14.4) (13.1)
,0.0001 0.02 0.75 0.58
27 (35.5) 27 (35.5) 22 (29.0)
93 (41.7) 76 (34.1) 54 (24.2)
34 17 7 4 14 3.4
90 48 23 27 35 2.2
0.64 (44.7) (22.4) (9.2) (5.3) (18.4) (6.4)
43 (56.6) 16 (21.1)
(40.4) (21.5) (10.3) (9.0) (15.7) (7.3)
101 (45.5) 6 (2.7)
0.1 ,0.0001
*Maximal SDH diameter measured on horizontal images of computed tomography of the head on admission. †Measured at the level of the pineal gland on horizontal images of computed tomography of the head on admission. aSDH, acute SDH; IQR, interquartile range; SDH, subdural hematoma. Bold/Italic values indicate statistical significance, P , 0.05.
Journal of Clinical Neurophysiology Volume 34, Number 1, January 2017
clinicalneurophys.com
S. Pollandt, et al.
Seizures and Epileptiform Discharges in Acute SDH
TABLE 2. EEG Indications and Findings Among Patients With Acute SDH (n ¼ 76) Indication for EEG (n ¼ 76) Altered mental status Seizure-like activity New onset focal deficits Other* Clinical and EEG findings (n ¼ 76) Patients with seizures Clinical seizures Exclusively nonconvulsive seizures Both clinical and nonconvulsive seizures Epileptiform EEG Epileptiform EEG without seizures Patients in status epilepticus Exclusively clinical SE Exclusively NCSE Clinical SE and NCSE Seizures in surgical cases (n ¼ 43) Patients with seizures Clinical seizures Exclusively nonconvulsive seizures Both clinical and nonconvulsive seizures Seizure timing Before surgery After surgery
n 30 26 11 9 n 32 30 2 4 24 9 12 7 3 2 n 18 17 1 2
(%) (40) (34) (15) (12) (%) (41.1) (93.8) (2.6) (12.5) (32) (11.8) (15.8) (58) (25) (16.7) (%) (41.9) (94.1) (2.3) (11.8)
6 (33.3) 12 (66.7)
Epileptiform EEG: epileptiform discharges or electrographic seizures. *Neurologic condition worse than expected for SDH; history of seizures or epilepsy. NCSE, nonconvulsive SE; SDH, subdural hematoma; SE, status epilepticus.
EEG Findings There were 42 (55%) patients who received a routine EEG (average duration 35 minutes, range 20–210 minutes), and 28 (37%) patients were monitored with continuous EEG (average duration 64 hours, range 6–259 hours), 14 of which also received a routine EEG. Six (8%) patients had multiple routine EEGs. EEG findings are shown in Table 2. Among 76 patients monitored with EEG, clinical seizures or an epileptiform EEG was noted in 41 (53.9%) patients. Clinical seizures were noted in 30 patients (39.5%). Four of these 30 patients had subsequent nonconvulsive seizures on EEG. Exclusively nonconvulsive seizures were seen in 2 (2.6%) patients. Twenty-four patients (32%) had epileptiform findings on EEG. Status epilepticus was diagnosed in 12 (16%) patients. Nine patients were in clinical SE (12%), 2 of which were in nonconvulsive SE (NCSE) on subsequent EEG monitoring. Three patients (4%) exclusively had NCSE. In surgically treated patients, the incidence of seizures was similar compared with the overall cohort (18/43 patients, 41.9%). Most seizures in surgically treated patients occurred in the postoperative period (12/18, 66.7%). Of 16 patients with a medical history of epilepsy, 8 (50%) developed clinical seizures, one of those patients was subsequently diagnosed with NCSE. While the incidence of patients with seizures is higher in this group compared with the overall cohort of monitored patients, the difference was not significant. Between patients with acute and acute-on-chronic SDH, there was no statistically significant difference in rate of clinical seizures (13/30 vs. 17/30, P ¼ 0.06) or epileptiform EEG (14/24 vs. 10/24, P ¼ 0.83).
Antiseizure Prophylaxis monitoring. Six of those patients had a seizure at onset of their aSDH before their hospitalization. All had a normal mental status after admission and no further clinical events. One patient had a clinical seizure after admission during a V-tach arrest but subsequently recovered and had no further clinical events. One patient was transitioned to comfort care shortly after admission, and EEG was not pursued. Of all patients, 293 (97%) received antiseizure prophylaxis. Of the 76 patients with a diagnosis of aSDH who received an EEG, median age was 70 years (interquartile range, 23), 49 (65%) were male, and 43 (57%) underwent surgical SDH evacuation. The most common comorbidities were arterial hypertension in 65 patients (85.5%), coronary artery disease in 22 (29.0%), and diabetes mellitus in 19 (25%). Sixteen (21.1%) patients had a history of seizures. The inciting event triggering EEG monitoring occurred between hospital days 0 and 28 (mean and median, hospital day 2); in 28 (37%) patients, the seizurelike event occurred on day 0.
Radiographic Characteristics The radiographic characteristics of the aSDH in the study population are shown in Table 1. The median SDH diameter was 12.3 mm in its widest diameter on horizontal images of the admission CT scan. Frontoparietal SDH location was more common. Median midline shift of the deep cerebral structures, measured at the level of the pineal gland on horizontal imaging, was 3.4 mm. clinicalneurophys.com
Seventy-four patients (97.4%) received antiseizure prophylaxis during their admission, 73 (96%) on hospital day 0 or 1. Thirteen patients in the population monitored with EEG had seizures at aSDH onset before the start of antiseizure prophylaxis. Of the 63 remaining patients, 61 (96.8%) were started or maintained on seizure prophylaxis. Nineteen of those patients (30.2%) developed clinical (17 patients) or nonconvulsive (2 patients) seizures despite antiseizure prophylaxis or while on home antiseizure drugs. All patients who developed seizures after admission were on antiseizure prophylaxis at the time of their event and underwent EEG monitoring. Eight of 12 patients diagnosed with SE were on antiseizure prophylaxis at the time the diagnosis was made. The majority (43 patients, 56.6%) received levetiracetam, 4 (5.3%) received phenytoin, 19 (25%) a combination of medications, and 4 (5.3%) a different single antiseizure medication in standard dosing.
Association of Baseline, Clinical, and Radiographic Characteristics With Outcomes On univariate analysis, the occurrence of clinical seizures was not associated with outcomes. Demographic factors and baseline characteristics (as displayed in Table 1) were not associated with the presence of epileptiform discharges or SE in univariate analysis. The presence of epileptiform EEG findings was associated with longer hospital LOS (13 vs. 8 days, P ¼ 0.04), longer ICU-LOS (10 vs. 4 days, P ¼ 0.002), and higher mortality (33.3% vs. 11.5%, P ¼ 0.03) in univariate analysis
Journal of Clinical Neurophysiology Volume 34, Number 1, January 2017
57
S. Pollandt, et al.
Seizures and Epileptiform Discharges in Acute SDH
(Table 3). On multivariate analysis, epileptiform EEG findings independently predicted longer ICU-LOS (odds ratio [OR], 13.99; 95% confidence interval [CI], 2.92–67.06; P ¼ 0.001) and hospital LOS (OR, 6.74; 95% CI, 1.63–27.87; P ¼ 0.01). Status epilepticus predicted longer ICU-LOS (P ¼ 0.002) and longer hospital LOS (P ¼ 0.01) in univariate analysis. In multivariate analysis, only ICU-LOS was significantly higher in the presence of SE (OR, 12.76; 95% CI, 1.32–123.75; P ¼ 0.03) (Table 4). Status epilepticus was not significantly related to mortality, discharge destination, and the need for tracheostomy/ percutaneous endoscopic gastrostomy placement. In multivariate analysis among survivors, epileptiform EEG also independently predicted longer ICU-LOS (OR, 27.33; 95% CI, 3.49–213.86; P ¼ 0.002) and hospital LOS (OR, 13.51; 95% CI, 2.27–80.46; P ¼ 0.004). For survivors only, SE was not significantly associated with either LOS or ICU-LOS.
TABLE 4. Association of SE and Outcomes in Patients With Acute SDH (n ¼ 76) SE Yes (n ¼ 12) No (n ¼ 64) LOS, median (IQR), days ICU-LOS, median, (IQR), days Tracheostomy/PEG, n (%) Mortality, n (%) Discharge destination, n (%)† Favorable Unfavorable Death or hospice
14 10 4 3
(9) (5.5) (33.3) (25.0)
6 (50.0) 3 (25.0) 3 (25.0)
9.5 4 13 11
(9.5) (7.5) (20.3) (17.2)
P 0.01 0.002* 0.45 0.68 0.91
34 (55.7) 16 (28.1) 11 (17.2)
Method: univariate analysis. *Remained significant in multivariate analysis. †Three patients who were nursing home residents before admission were excluded. ICU, intensive care unit; IQR, interquartile range; LOS, length of stay; PEG, percutaneous endoscopic gastrostomy; SDH, subdural hematoma; SE, status epilepticus. Bold/Italic values indicate statistical significance, P , 0.05.
DISCUSSION In this study of patients with aSDH, clinical events suspicious for seizures occurred in one of four patients, typically despite prophylaxis with antiseizure medications. Of patients with suspicion of seizures who underwent EEG monitoring, 40% had clinical seizures during their hospital stay, 32% had epileptiform findings on EEG, 54% had either, and 16% were diagnosed with SE. While occurrence of clinical seizures did not affect mortality, discharge destination, or need for tracheostomy/ percutaneous endoscopic gastrostomy, the presence of SE was independently associated with longer ICU-LOS, while epileptiform findings on EEG were associated with both longer ICULOS and hospital LOS. The association of seizures and epileptiform findings with SDH is generally perceived as well known, but it is not well described. Most published data have been collected from patients with chronic SDH, with a described incidence of seizures between 5% and 16%, with higher incidence in patients with mixed-density lesions and postoperatively (Battaglia et al., 2012; TABLE 3. Association of Epileptiform EEG and Outcomes in Patients With Acute SDH (n ¼ 76) Epileptiform EEG Yes (n ¼ 24) No (n ¼ 52) LOS, median (IQR), days ICU-LOS, median, (IQR), days Tracheostomy/PEG, n (%) Mortality, n (%) Discharge destination, n (%)† Favorable Unfavorable Death or hospice
13 10 7 8
(8.5) (7) (29.2) (33.3)
9 (37.5) 7 (29.2) 8 (33.3)
8 4 10 6
(10) (7) (19.2) (11.5)
P 0.04* 0.002* 0.33 0.03 0.05
31 (63.3) 12 (24.5) 6 (12.2)
Method: univariate analysis. *Remained significant in multivariate analysis. †Three patients who were nursing home residents before admission were excluded. ICU, intensive care unit; LOS, length of stay; PEG, percutaneous endoscopic gastrostomy; SDH, subdural hematoma. Bold/Italic values indicate statistical significance, P , 0.05.
58
Chen et al., 2004; Hamasaki et al., 2013). Data on aSDH and seizures are scarce. Periodic lateralized epileptiform discharges have been described in a few patients with aSDH who experienced seizures or transient neurologic deficits (Chu, 1979; Mehryar and McIntyre, 1975; Westmoreland, 2001). In a large cohort of patients with TBI, SDH was identified as a strong risk factor for late unprovoked seizures (Annegers et al., 1998). A recent series evaluating baseline EEG patterns on mixed adult ICU patients contained 14 patients with SDH within the study cohort, with a high reported seizure frequency of 43% (Swisher et al., 2015). There has been only one larger study systematically assessing the incidence of seizures or epileptiform findings on EEG; this was a 134-patient cohort of patients with aSDH (Rabinstein et al., 2010). In this study, all patients underwent surgical SDH treatment, and seizures were much more common after the surgical intervention despite good evacuation, no recurrence of the SDH, and the absence of ischemic complications on postoperative imaging (Rabinstein et al., 2010). The incidence of seizures or epileptiform findings in our cohort of monitored patients was 53.9% higher than the reported incidence of 25% reported by Rabinstein et al. (2010). This is likely because of the exclusion of patients who were not monitored with EEG in our cohort. The inclusion of only surgical patients in their study could potentially lead to an increased incidence of seizures, but in our study, there was no statistically significant difference in the incidence of seizures between conservatively and surgically treated patients in patients monitored with EEG. The incidence of NCSE has been reported in 10% to 67% of critically ill patients in various ICU populations (Sutter, 2016). These include select patient populations, such as subarachnoid hemorrhage, with a reported frequency of NCSE of 3% to 13% (Kondziella et al., 2015), or mixed ICU populations, such as of a general surgical ICU with a reported frequency of 5% to 16% (Kurtz et al., 2014); however, there are no studies on a homogenous group of ICU patients with aSDH only. The proportion of patients with NCSE in our series, 6.5% of the monitored cohort or 1.6% of the entire patient population, is lower than the
Journal of Clinical Neurophysiology Volume 34, Number 1, January 2017
clinicalneurophys.com
S. Pollandt, et al.
Seizures and Epileptiform Discharges in Acute SDH
previously reported range. However, patients with aSDH may differ from many other studied patient cohorts in that the actual disease process of aSDH does not involve direct injury to the brain parenchyma. Furthermore, as opposed to other or mixed ICU populations, patients with aSDH are often not systemically critically ill. Most of our patient population had a Glasgow Coma Score of 14 or 15 (69%). Our results may, nonetheless, underestimate the proportion of NCSE because of the number of patients with routine rather than continuous EEG monitoring because increased capture rates of NCSE have been reported with increased EEG duration (Jette et al., 2006). Another interesting finding of our study is that the frequency of clinical seizures was much higher than that of nonconvulsive seizures or NCSE. Comparison with available data is difficult, as there are very few data on similar patient populations, but the question is raised whether aSDH may preferentially present with clinical seizures, possibly because of the different pathophysiology of a disease that does not directly affect the brain parenchyma. Additionally, our data may overestimate the rate of clinical seizures because of referral bias, as largely patients with distinct clinical events were monitored. A point rarely addressed in the TBI and SDH literatures discussing seizures is the question of whether epilepsy or seizures were present before the injury occurred. Not surprisingly, patients with history of seizures or epilepsy were significantly more likely to undergo EEG monitoring after a clinical event suspicious for seizures. However, among patients with EEG monitoring, preexisting seizures or epilepsy was not associated with a significant increase in the incidence of seizures. This finding may underline the innate association of SDH with seizures. Another factor to be considered is the use of prophylactic antiseizure medications. In our series, seizures and epileptiform EEG findings were common despite the standard use of antiseizure prophylaxis. In routine practice, most patients with TBI are prescribed an antiseizure medication for 7 days after the injury according to the Brain Trauma Foundation guidelines (Bratton et al., 2007). These guidelines were established according to the findings of a study showing significantly lower incidence of early seizures with phenytoin prophylaxis than without prophylaxis (3.6 vs. 14.2%) (Temkin et al., 1990) in patients with acute TBI. Instead of phenytoin, levetiracetam is increasingly used in neuroscience ICUs (Szaflarski et al., 2007, 2010) because of better tolerance than phenytoin. However, levetiracetam prophylaxis was also found to be associated with a higher risk of electrographic seizures and abnormal EEG (Jones et al., 2008; Radic et al., 2014). In a prospective observational study of pediatric patients with TBI, posttraumatic seizures occurred frequently despite prophylaxis with levetiracetam (Chung and O’Brien, 2016). In our study, all patients who developed seizures after admission were monitored with EEG and included in our study cohort. Our data support the notion that the incidence of seizures and SE is high despite antiseizure prophylaxis, most commonly levetiracetam, and call into question the effectiveness of current common practice. With respect to prognosis and outcomes, well-described poor prognostic factors after aSDH include older age, preoperative coma, lower coma scores, and pupillary abnormalities clinicalneurophys.com
(Cagetti et al., 1992; Kotwica and Brzezinski, 1993; Phuenpathom et al., 1993; Servadei, 1997; Servadei et al., 2000; Wiedemayer et al., 2002; Wilberger et al., 1991). Other factors influencing prognosis are radiologic findings, time to surgery, and the presence of coagulopathy (Bershad et al., 2008; D’Amato et al., 2007; Servadei, 1997; Servadei et al., 2000; Sawauchi and Abe, 2008). Seizures or EEG findings are not included in most of these analyses. Additionally, the patient cohorts examined were often patients with more complex TBI rather than isolated aSDH. We found in our series that the occurrence of clinical seizures did not affect outcomes. However, ICU-LOS was longer for patients in SE or patients with epileptiform findings on EEG, which also predicted longer hospital LOS. Mortality and discharge destination were not affected. In comparison Rabinstein et al. (2010) found that while clinical seizures or epileptiform discharges were independently associated with poor functional outcome at discharge, there was no significant effect at 1 to 6 months follow-up. The slight difference might be explained by the inclusion of conservatively treated patients in our cohort, as these patients presumably had less severe SDH. As with any retrospective study, limitations are inherent. However, in the absence of data in this patient group, our data on this well-defined homogenous subgroup of critically ill aSDH patients serve to deliver novel data and could guide a prospective approach in the future. EEG monitoring is not uniformly performed in every patient but rather requested by the attending physician based on clinical indication, resulting in an inherent selection bias. In our unit, the threshold for obtaining an EEG in patients with unexplained poor neurologic examination or acute change in neurologic examination is low, but the true incidence of epileptiform abnormalities is likely underestimated. Because of the design as an exploratory study, we did not specify one outcome as primary outcome, risking some type 1 error, however, decreasing the chance of type II error. Lastly, Acute Physiology and Chronic Health Evaluation scores or similar measures of systemic illness could not be included in our analysis because of lack of availability. In summary, events suspicious for seizures in patients with aSDH are frequent. Clinical seizures, nonconvulsive seizures, or epileptiform discharges are seen in more than half of the patients undergoing EEG monitoring. Given this high incidence, the threshold for EEG monitoring should be low and perhaps all inclusive especially in the presence of abnormalities on neurologic examination, in order to elucidate the true incidence of seizures and epileptiform findings. In judging prognosis for patients with aSDH, it is important to consider that epileptic complications contribute to increased LOS but are not associated with poor outcome.
REFERENCES Annegers JF, Coan SP. The risks of epilepsy after traumatic brain injury. Seizure 2000;9:453–457. Annegers JF, Hauser WA, Coan SP, Rocca WA. A population-based study of seizures after traumatic brain injuries. N Engl J Med 1998;338:20–24. Battaglia F, Lubrano V, Ribeiro-Filho T, et al. Incidence and clinical impact of seizures after surgery for chronic subdural haematoma [in French]. Neurochirurgie 2012;58:230–234.
Journal of Clinical Neurophysiology Volume 34, Number 1, January 2017
59
S. Pollandt, et al.
Bershad EM, Farhadi S, Suri MF, et al. Coagulopathy and inhospital deaths in patients with acute subdural hematoma. J Neurosurg 2008;109:664–669. Bratton SL, Chestnut RM, Ghajar J, et al. Guidelines for the management of severe traumatic brain injury. XIII. Antiseizure prophylaxis. J Neurotrauma 2007;24(suppl 1):S83–S86. Cagetti B, Cossu M, Pau A, et al. The outcome from acute subdural and epidural intracranial haematomas in very elderly patients. Br J Neurosurg 1992;6:227–231. Chen CW, Kuo JR, Lin HJ, et al. Early post-operative seizures after burr-hole drainage for chronic subdural hematoma: correlation with brain CT findings. J Clin Neurosci 2004;11:706–709. Chu NS. Acute subdural hematoma and the periodic lateralized epileptiform discharges. Clin Electroencephalogr 1979;10:145–150. Chung MG, O’Brien NF. Prevalence of early posttraumatic seizures in children with Moderate to severe traumatic brain injury despite levetiracetam prophylaxis. Pediatric critical care medicine. Pediatr Crit Care Med 2016;17:150–156. D’Amato L, Piazza O, Alliata L, et al. Prognosis of isolated acute posttraumatic subdural haematoma. J Neurosurg Sci 2007;51:107–111. Frontera JA, Egorova N, Moskowitz AJ. National trend in prevalence, cost, and discharge disposition after subdural hematoma from 1998-2007. Crit Care Med 2011;39:1619–1625. Hamasaki T, Yamada K, Kuratsu J. Seizures as a presenting symptom in neurosurgical patients: a retrospective single-institution analysis. Clin Neurol Neurosurg 2013;115:2336–2340. Jette N, Claassen J, Emerson RG, Hirsch LJ. Frequency and predictors of nonconvulsive seizures during continuous electroencephalographic monitoring in critically ill children. Arch Neurol 2006;63:1750–1755. Jones KE, Puccio AM, Harshman KJ, et al. Levetiracetam versus phenytoin for seizure prophylaxis in severe traumatic brain injury. Neurosurg Focus 2008;25:E3. Kondziella D, Friberg CK, Wellwood I, et al. Continuous EEG monitoring in aneurysmal subarachnoid hemorrhage: a systematic review. Neurocrit Care 2015;22:450–461. Kotwica Z, Brzezinski J. Acute subdural haematoma in adults: an analysis of outcome in comatose patients. Acta Neurochir (Wien) 1993;121:95–99. Kurtz P, Gaspard N, Wahl AS, et al. Continuous electroencephalography in a surgical intensive care unit. Intensive Care Med 2014;40:228–234. Mehryar GR, McIntyre HB. Periodic lateralized epileptiform discharges associated with subdural hematoma. Bull Los Angeles Neurol Soc 1975;40:8–12. Noachtar S, Binnie C, Ebersole J, et al. A glossary of terms most commonly used by clinical electroencephalographers and proposal for the report form for the EEG findings. The International Federation of Clinical
60
Seizures and Epileptiform Discharges in Acute SDH
Neurophysiology. Electroencephalogr Clin Neurophysiol Suppl 1999; 52:21–41. Phuenpathom N, Choomuang M, Ratanalert S. Outcome and outcome prediction in acute subdural hematoma. Surg Neurol 1993;40:22–25. Rabinstein AA, Chung SY, Rudzinski LA, Lanzino G. Seizures after evacuation of subdural hematomas: incidence, risk factors, and functional impact. J Neurosurg 2010;112:455–460. Radic JA, Chou SH, Du R, Lee JW. Levetiracetam versus phenytoin: a comparison of efficacy of seizure prophylaxis and adverse event risk following acute or subacute subdural hematoma diagnosis. Neurocrit Care 2014;21:228–237. Sabo RA, Hanigan WC, Aldag JC. Chronic subdural hematomas and seizures: the role of prophylactic anticonvulsive medication. Surg Neurol 1995;43:579–582. Sawauchi S, Abe T. The effect of haematoma, brain injury, and secondary insult on brain swelling in traumatic acute subdural haemorrhage. Acta Neurochir (Wien) 2008;150:531–536. Servadei F. Prognostic factors in severely head injured adult patients with acute subdural haematoma’s. Acta Neurochir (Wien) 1997;139:279–285. Servadei F, Nasi MT, Giuliani G, et al. CT prognostic factors in acute subdural haematomas: the value of the “worst” CT scan. Br J Neurosurg 2000;14:110–116. Sutter R. Are We Prepared to Detect Subtle and nonconvulsive status epilepticus in critically ill patients? J Clin Neurophysiol 2016;33:25–31. Swisher CB, Shah D, Sinha SR, Husain AM. Baseline EEG pattern on continuous ICU EEG monitoring and incidence of seizures. J Clin Neurophysiol 2015;32:147–151. Szaflarski JP, Meckler JM, Szaflarski M, et al. Levetiracetam use in critically ill patients. Neurocrit Care 2007;7:140–147. Szaflarski JP, Sangha KS, Lindsell CJ, Shutter LA. Prospective, randomized, single-blinded comparative trial of intravenous levetiracetam versus phenytoin for seizure prophylaxis. Neurocrit Care 2010;12:165–172. Temkin NR, Dikmen SS, Wilensky AJ, et al. A randomized, double-blind study of phenytoin for the prevention of post-traumatic seizures. N Engl J Med 1990;323:497–502. Wang HC, Chang WN, Chang HW, et al. Factors predictive of outcome in posttraumatic seizures. J Trauma 2008;64:883–888. Westmoreland BF. Periodic lateralized epileptiform discharges after evacuation of subdural hematomas. J Clin Neurophysiol 2001;18:20–24. Wiedemayer H, Triesch K, Schafer H, Stolke D. Early seizures following non-penetrating traumatic brain injury in adults: risk factors and clinical significance. Brain Inj 2002;16:323–330. Wilberger JE Jr, Harris M, Diamond DL. Acute subdural hematoma: morbidity, mortality, and operative timing. J Neurosurg 1991;74:212–218.
Journal of Clinical Neurophysiology Volume 34, Number 1, January 2017
clinicalneurophys.com
Seizures and Epileptiform Discharges in Patients With Acute Subdural Hematoma Sebastian Pollandt,* Bichun Ouyang,† Thomas P. Bleck,†‡§k and Katharina M. Busl* Section of Neurocritical Care, Department of Neurological Sciences, Rush University Medical Center, Chicago, Illinois, U.S.A.; Departments of †Neurological Sciences, ‡Neurosurgery, §Internal Medicine, and kAnesthesiology, Rush University Medical Center, Chicago, Illinois, U.S.A. *
Purpose: Subdural hematomas (SDH) are associated with seizures and epileptiform discharges, but little is known about the prevalence and impact of seizures, status epilepticus (SE), and epileptiform discharges on outcomes in patients with isolated acute SDH (aSDH). Methods: Continuous EEG reports from 76 adult patients admitted to Rush University Medical Center with aSDH between January 2009 and March 2012 were reviewed. Clinical and radiographic findings, comorbidities, treatment, and outcome parameters, such as mortality, discharge destination, need for tracheostomy/percutaneous endoscopic gastrostomy placement, and length of stay (LOS), were assessed. Univariate and multivariate analyses were performed to assess the impact of clinical seizures, SE, and epileptiform EEG on outcomes. Results: Of 76 patients with aSDH who underwent EEG monitoring, 74 (97.4%) received antiseizure prophylaxis. Thirtytwo (41.1%) patients had seizures, most of which were clinical seizures. Twenty-four (32%) patients had epileptiform EEG
ubdural hematoma (SDH) is one of the most common diagnoses in neurointensive care units (ICUs), with more than 90,000 hospitalizations for SDH occurring annually in the United States (Frontera et al., 2011). With the increasing proportion of elderly in the population of Western countries, the prevalence of SDH is expected to grow further (Frontera et al., 2011). The association of SDH and seizures is well known; however, most data originate from chronic SDH (Sabo et al., 1995; Wang et al., 2008) or aSDH in the setting of severe traumatic brain injury (TBI), resulting in a wide variation in reported incidence and impact of seizures (Annegers and Coan, 2000). Much less data exist on patients with isolated aSDH and seizures (Rabinstein et al., 2010), and little is known about the impact of seizures, status epilepticus (SE), and epileptiform findings on outcomes for patients with aSDH. The focus of this study was to determine the prevalence of seizures, SE, and epileptiform EEG findings in a large cohort of patients with isolated aSDH and examine their impact on outcomes.
S
findings. Clinical or nonconvulsive SE was diagnosed in 12 (16%) patients. Clinical seizures were not associated with outcome parameters. Epileptiform EEG findings were independently associated with longer hospital LOS (13 vs. 8 days, P ¼ 0.04) and intensive care unit LOS (10 vs. 4 days, P ¼ 0.002). The SE also predicted longer intensive care unit LOS (10 vs. 4 days, P ¼ 0.002). Neither epileptiform EEG nor SE was significantly related to mortality, discharge destination, or need for tracheostomy/ percutaneous endoscopic gastrostomy placement. Conclusions: Seizures and epileptiform EEG findings are very common in patients with aSDH despite antiseizure prophylaxis. While clinical seizures did not affect outcomes, the presence of epileptiform EEG findings and SE was independently associated with longer intensive care unit LOS and hospital LOS. Key Words: Subdural hematoma, EEG, Seizures, Status epilepticus, Adult, Epileptiform discharges. (J Clin Neurophysiol 2017;34: 55–60)
MATERIALS AND METHODS Participants We conducted a retrospective review of all adult patients admitted to Rush University Medical Center with an admission diagnosis of aSDH between January 2009 and March 2012. Patients with chronic SDH or concomitant intracranial injuries (including parenchymal contusions, global cerebral edema, and diffuse axonal injury, when noted on computed tomography of the head on admission) were excluded at data collection. Patients with insufficient EEG data were excluded. Acute or acute-onchronic SDH was considered aSDH and included. Patients with aSDH who underwent EEG monitoring were selected for further analysis. The study was approved by the local institutional review board, including a waiver for obtaining informed consent to collect and publish data.
Clinical Measures The abstract to this manuscript was presented in form of 2 posters at the Annual Neurocritical Care Society Meeting, Scottsdale, Arizona, October 7–10, 2015. The authors have no funding or conflicts of interest to disclose. Address correspondence and reprint requests to Sebastian Pollandt, MD, Department of Neurological Sciences, Section of Neurocritical Care, Rush University Medical Center, 1725 West Harrison St, POB Suite 1106, Chicago, IL 60612, U.S.A.; e-mail: [email protected]. Copyright Ó 2016 by the American Clinical Neurophysiology Society
ISSN: 0736-0258/16/3401-0055
DOI 10.1097/WNP.0000000000000311
clinicalneurophys.com
Baseline demographics, medical history, comorbidities, clinical and radiographic findings, prophylactic antiseizure drug administration, treatment of aSDH, discharge disposition, and hospital length of stay (LOS) were abstracted from chart review. Daily progress notes (including progress notes by treating or consulting physician teams, nursing staff, and therapy team notes) were reviewed for occurrence of clinical seizures, symptoms concerning for seizures, or changes in clinical
Journal of Clinical Neurophysiology Volume 34, Number 1, January 2017
55
S. Pollandt, et al.
Seizures and Epileptiform Discharges in Acute SDH
condition prompting the treating physician to obtain an EEG. An occurrence of clinical seizure was concluded for characteristic transient symptoms and exclusion of alternative etiologies or if the clinical event in question had an electrographic correlate on EEG. Nonconvulsive seizures were recorded if a seizure per EEG criteria did not have a clinical correlate on video review by the electroencephalographer. For patients who underwent surgical treatment for their aSDH, temporal relation of the clinical event to surgical treatment was noted. All patients received a Glasgow Coma Scale score on admission. Radiologic measures were made on the worst preoperative CT scan on axial images and included maximum SDH diameter and degree of midline shift measured in millimeters at the level of the pineal gland. Outcome parameters included mortality, discharge destination, defined as favorable (home or rehabilitation) versus unfavorable (nursing home, skilled nursing facility, or long-term acute care facility), need for tracheostomy and percutaneous endoscopic gastrostomy placement, and ICU and hospital LOS.
EEG Measures All EEG reports were generated by a board-certified epileptologist at this institution and reviewed by a boardcertified epileptologist (S.P.) for occurrence of clinical or electrographic seizures, SE, interictal epileptiform discharges, and focal slowing. EEG tracings were rereviewed for selected studies by one of the authors (S.P.). EEGs included in the analysis were routine EEGs, multiple routine EEGs, continuous EEGs, or a combination. Epileptiform changes were defined as interictal epileptiform discharges (transients distinguishable from background activity with a characteristic spiky morphology; Noachtar et al., 1999) or electrographic seizures.
Statistical Analyses x2
Univariate analyses were performed with or Fisher exact test for categorical variables and Wilcoxon rank-sum test for continuous variables as indicated. Multivariate logistic regression analyses with stepwise selection, including all factors with a P value of ,0.15 from the univariate analyses, were used to identify baseline factors that independently predicted outcome measurements and to further examine the relationship of epileptiform changes to outcome parameters. For LOS analyses, binary logistic regression models were used for ICU-LOS ( $ 6 vs. ,6 days) and for hospital LOS ( $ 10 vs. ,10 days) based on the median LOS, respectively. For the analysis of discharge destination, three patients were excluded based on their residence at an unfavorable discharge destination before admission. All statistical analyses were performed with SAS 9.2 (SAS Institute Inc, Cary, NC). A P value ,0.05 was considered significant.
RESULTS Study Population Of the 300 patients admitted with aSDH during the study period, 77 patients underwent EEG monitoring. One patient was excluded from the analysis because of insufficient available EEG data. When compared with the entire cohort of patients with 56
aSDH, patients with history of hypertension, prior intracerebral hemorrhage, prior stroke, history of seizures or epilepsy, lower admission Glasgow Coma Scale, seizure at aSDH onset, and intubated patients were more likely to receive an EEG (Table 1). The most common indications for EEG monitoring were alterations in mental status, seizure-like activity (focal movements or generalized convulsions), or new focal findings, such as weakness or aphasia (Table 2). In the overall cohort, eight patients had clinical seizures but did not undergo EEG TABLE 1. Baseline Characteristics of Patients With aSDH (n ¼ 299) EEG Performed Yes (n ¼ 76) No (n ¼ 223) Age, median (IQR), years Sex male, n (%) Medical history, n (%) Diabetes mellitus Hypertension Coronary artery disease Congestive heart failure Renal disease Liver disease Alcohol abuse Smoking Dementia Prior intracerebral hemorrhage Prior stroke Prior seizures or epilepsy Clinical and radiographic characteristics on admission, n (%) Glasgow coma scale score 13–15 9–12 3–8 Median (IQR) Intubated SDH size,* median (IQR), mm SDH side Left Right Bilateral SDH location Frontoparietal Holohemispheric Temporal Tentorial/falcine Other lobar Midline shift,† median (IQR), mm Surgical SDH evacuation Seizures at onset of aSDH
P
70 (23) 49 (64.5)
71 (24) 135 (60.5)
0.81 0.54
19 65 22 12 11 3 14 14 12 10
61 161 59 22 32 12 37 42 30 7
0.69 0.02 0.67 0.16 0.98 0.77 0.74 0.94 0.61 0.003
(25) (85.5) (29.0) (15.8) (14.5) (4.0) (18.4) (18.4) (15.8) (13.2)
17 (22.4) 16 (21.1)
(27.4) (72.2) (26.5) (9.9) (14.4) (5.4) (16.7) (18.8) (13.5) (3.1)
29 (13) 23 (10.3)
0.05 0.02
,0.0001 46 19 11 13 20 12.3
(60.5) (25) (14.5) (5) (26.3) (9.5)
183 11 29 15 32 12.2
(82.1) (5) (13) (1) (14.4) (13.1)
,0.0001 0.02 0.75 0.58
27 (35.5) 27 (35.5) 22 (29.0)
93 (41.7) 76 (34.1) 54 (24.2)
34 17 7 4 14 3.4
90 48 23 27 35 2.2
0.64 (44.7) (22.4) (9.2) (5.3) (18.4) (6.4)
43 (56.6) 16 (21.1)
(40.4) (21.5) (10.3) (9.0) (15.7) (7.3)
101 (45.5) 6 (2.7)
0.1 ,0.0001
*Maximal SDH diameter measured on horizontal images of computed tomography of the head on admission. †Measured at the level of the pineal gland on horizontal images of computed tomography of the head on admission. aSDH, acute SDH; IQR, interquartile range; SDH, subdural hematoma. Bold/Italic values indicate statistical significance, P , 0.05.
Journal of Clinical Neurophysiology Volume 34, Number 1, January 2017
clinicalneurophys.com
S. Pollandt, et al.
Seizures and Epileptiform Discharges in Acute SDH
TABLE 2. EEG Indications and Findings Among Patients With Acute SDH (n ¼ 76) Indication for EEG (n ¼ 76) Altered mental status Seizure-like activity New onset focal deficits Other* Clinical and EEG findings (n ¼ 76) Patients with seizures Clinical seizures Exclusively nonconvulsive seizures Both clinical and nonconvulsive seizures Epileptiform EEG Epileptiform EEG without seizures Patients in status epilepticus Exclusively clinical SE Exclusively NCSE Clinical SE and NCSE Seizures in surgical cases (n ¼ 43) Patients with seizures Clinical seizures Exclusively nonconvulsive seizures Both clinical and nonconvulsive seizures Seizure timing Before surgery After surgery
n 30 26 11 9 n 32 30 2 4 24 9 12 7 3 2 n 18 17 1 2
(%) (40) (34) (15) (12) (%) (41.1) (93.8) (2.6) (12.5) (32) (11.8) (15.8) (58) (25) (16.7) (%) (41.9) (94.1) (2.3) (11.8)
6 (33.3) 12 (66.7)
Epileptiform EEG: epileptiform discharges or electrographic seizures. *Neurologic condition worse than expected for SDH; history of seizures or epilepsy. NCSE, nonconvulsive SE; SDH, subdural hematoma; SE, status epilepticus.
EEG Findings There were 42 (55%) patients who received a routine EEG (average duration 35 minutes, range 20–210 minutes), and 28 (37%) patients were monitored with continuous EEG (average duration 64 hours, range 6–259 hours), 14 of which also received a routine EEG. Six (8%) patients had multiple routine EEGs. EEG findings are shown in Table 2. Among 76 patients monitored with EEG, clinical seizures or an epileptiform EEG was noted in 41 (53.9%) patients. Clinical seizures were noted in 30 patients (39.5%). Four of these 30 patients had subsequent nonconvulsive seizures on EEG. Exclusively nonconvulsive seizures were seen in 2 (2.6%) patients. Twenty-four patients (32%) had epileptiform findings on EEG. Status epilepticus was diagnosed in 12 (16%) patients. Nine patients were in clinical SE (12%), 2 of which were in nonconvulsive SE (NCSE) on subsequent EEG monitoring. Three patients (4%) exclusively had NCSE. In surgically treated patients, the incidence of seizures was similar compared with the overall cohort (18/43 patients, 41.9%). Most seizures in surgically treated patients occurred in the postoperative period (12/18, 66.7%). Of 16 patients with a medical history of epilepsy, 8 (50%) developed clinical seizures, one of those patients was subsequently diagnosed with NCSE. While the incidence of patients with seizures is higher in this group compared with the overall cohort of monitored patients, the difference was not significant. Between patients with acute and acute-on-chronic SDH, there was no statistically significant difference in rate of clinical seizures (13/30 vs. 17/30, P ¼ 0.06) or epileptiform EEG (14/24 vs. 10/24, P ¼ 0.83).
Antiseizure Prophylaxis monitoring. Six of those patients had a seizure at onset of their aSDH before their hospitalization. All had a normal mental status after admission and no further clinical events. One patient had a clinical seizure after admission during a V-tach arrest but subsequently recovered and had no further clinical events. One patient was transitioned to comfort care shortly after admission, and EEG was not pursued. Of all patients, 293 (97%) received antiseizure prophylaxis. Of the 76 patients with a diagnosis of aSDH who received an EEG, median age was 70 years (interquartile range, 23), 49 (65%) were male, and 43 (57%) underwent surgical SDH evacuation. The most common comorbidities were arterial hypertension in 65 patients (85.5%), coronary artery disease in 22 (29.0%), and diabetes mellitus in 19 (25%). Sixteen (21.1%) patients had a history of seizures. The inciting event triggering EEG monitoring occurred between hospital days 0 and 28 (mean and median, hospital day 2); in 28 (37%) patients, the seizurelike event occurred on day 0.
Radiographic Characteristics The radiographic characteristics of the aSDH in the study population are shown in Table 1. The median SDH diameter was 12.3 mm in its widest diameter on horizontal images of the admission CT scan. Frontoparietal SDH location was more common. Median midline shift of the deep cerebral structures, measured at the level of the pineal gland on horizontal imaging, was 3.4 mm. clinicalneurophys.com
Seventy-four patients (97.4%) received antiseizure prophylaxis during their admission, 73 (96%) on hospital day 0 or 1. Thirteen patients in the population monitored with EEG had seizures at aSDH onset before the start of antiseizure prophylaxis. Of the 63 remaining patients, 61 (96.8%) were started or maintained on seizure prophylaxis. Nineteen of those patients (30.2%) developed clinical (17 patients) or nonconvulsive (2 patients) seizures despite antiseizure prophylaxis or while on home antiseizure drugs. All patients who developed seizures after admission were on antiseizure prophylaxis at the time of their event and underwent EEG monitoring. Eight of 12 patients diagnosed with SE were on antiseizure prophylaxis at the time the diagnosis was made. The majority (43 patients, 56.6%) received levetiracetam, 4 (5.3%) received phenytoin, 19 (25%) a combination of medications, and 4 (5.3%) a different single antiseizure medication in standard dosing.
Association of Baseline, Clinical, and Radiographic Characteristics With Outcomes On univariate analysis, the occurrence of clinical seizures was not associated with outcomes. Demographic factors and baseline characteristics (as displayed in Table 1) were not associated with the presence of epileptiform discharges or SE in univariate analysis. The presence of epileptiform EEG findings was associated with longer hospital LOS (13 vs. 8 days, P ¼ 0.04), longer ICU-LOS (10 vs. 4 days, P ¼ 0.002), and higher mortality (33.3% vs. 11.5%, P ¼ 0.03) in univariate analysis
Journal of Clinical Neurophysiology Volume 34, Number 1, January 2017
57
S. Pollandt, et al.
Seizures and Epileptiform Discharges in Acute SDH
(Table 3). On multivariate analysis, epileptiform EEG findings independently predicted longer ICU-LOS (odds ratio [OR], 13.99; 95% confidence interval [CI], 2.92–67.06; P ¼ 0.001) and hospital LOS (OR, 6.74; 95% CI, 1.63–27.87; P ¼ 0.01). Status epilepticus predicted longer ICU-LOS (P ¼ 0.002) and longer hospital LOS (P ¼ 0.01) in univariate analysis. In multivariate analysis, only ICU-LOS was significantly higher in the presence of SE (OR, 12.76; 95% CI, 1.32–123.75; P ¼ 0.03) (Table 4). Status epilepticus was not significantly related to mortality, discharge destination, and the need for tracheostomy/ percutaneous endoscopic gastrostomy placement. In multivariate analysis among survivors, epileptiform EEG also independently predicted longer ICU-LOS (OR, 27.33; 95% CI, 3.49–213.86; P ¼ 0.002) and hospital LOS (OR, 13.51; 95% CI, 2.27–80.46; P ¼ 0.004). For survivors only, SE was not significantly associated with either LOS or ICU-LOS.
TABLE 4. Association of SE and Outcomes in Patients With Acute SDH (n ¼ 76) SE Yes (n ¼ 12) No (n ¼ 64) LOS, median (IQR), days ICU-LOS, median, (IQR), days Tracheostomy/PEG, n (%) Mortality, n (%) Discharge destination, n (%)† Favorable Unfavorable Death or hospice
14 10 4 3
(9) (5.5) (33.3) (25.0)
6 (50.0) 3 (25.0) 3 (25.0)
9.5 4 13 11
(9.5) (7.5) (20.3) (17.2)
P 0.01 0.002* 0.45 0.68 0.91
34 (55.7) 16 (28.1) 11 (17.2)
Method: univariate analysis. *Remained significant in multivariate analysis. †Three patients who were nursing home residents before admission were excluded. ICU, intensive care unit; IQR, interquartile range; LOS, length of stay; PEG, percutaneous endoscopic gastrostomy; SDH, subdural hematoma; SE, status epilepticus. Bold/Italic values indicate statistical significance, P , 0.05.
DISCUSSION In this study of patients with aSDH, clinical events suspicious for seizures occurred in one of four patients, typically despite prophylaxis with antiseizure medications. Of patients with suspicion of seizures who underwent EEG monitoring, 40% had clinical seizures during their hospital stay, 32% had epileptiform findings on EEG, 54% had either, and 16% were diagnosed with SE. While occurrence of clinical seizures did not affect mortality, discharge destination, or need for tracheostomy/ percutaneous endoscopic gastrostomy, the presence of SE was independently associated with longer ICU-LOS, while epileptiform findings on EEG were associated with both longer ICULOS and hospital LOS. The association of seizures and epileptiform findings with SDH is generally perceived as well known, but it is not well described. Most published data have been collected from patients with chronic SDH, with a described incidence of seizures between 5% and 16%, with higher incidence in patients with mixed-density lesions and postoperatively (Battaglia et al., 2012; TABLE 3. Association of Epileptiform EEG and Outcomes in Patients With Acute SDH (n ¼ 76) Epileptiform EEG Yes (n ¼ 24) No (n ¼ 52) LOS, median (IQR), days ICU-LOS, median, (IQR), days Tracheostomy/PEG, n (%) Mortality, n (%) Discharge destination, n (%)† Favorable Unfavorable Death or hospice
13 10 7 8
(8.5) (7) (29.2) (33.3)
9 (37.5) 7 (29.2) 8 (33.3)
8 4 10 6
(10) (7) (19.2) (11.5)
P 0.04* 0.002* 0.33 0.03 0.05
31 (63.3) 12 (24.5) 6 (12.2)
Method: univariate analysis. *Remained significant in multivariate analysis. †Three patients who were nursing home residents before admission were excluded. ICU, intensive care unit; LOS, length of stay; PEG, percutaneous endoscopic gastrostomy; SDH, subdural hematoma. Bold/Italic values indicate statistical significance, P , 0.05.
58
Chen et al., 2004; Hamasaki et al., 2013). Data on aSDH and seizures are scarce. Periodic lateralized epileptiform discharges have been described in a few patients with aSDH who experienced seizures or transient neurologic deficits (Chu, 1979; Mehryar and McIntyre, 1975; Westmoreland, 2001). In a large cohort of patients with TBI, SDH was identified as a strong risk factor for late unprovoked seizures (Annegers et al., 1998). A recent series evaluating baseline EEG patterns on mixed adult ICU patients contained 14 patients with SDH within the study cohort, with a high reported seizure frequency of 43% (Swisher et al., 2015). There has been only one larger study systematically assessing the incidence of seizures or epileptiform findings on EEG; this was a 134-patient cohort of patients with aSDH (Rabinstein et al., 2010). In this study, all patients underwent surgical SDH treatment, and seizures were much more common after the surgical intervention despite good evacuation, no recurrence of the SDH, and the absence of ischemic complications on postoperative imaging (Rabinstein et al., 2010). The incidence of seizures or epileptiform findings in our cohort of monitored patients was 53.9% higher than the reported incidence of 25% reported by Rabinstein et al. (2010). This is likely because of the exclusion of patients who were not monitored with EEG in our cohort. The inclusion of only surgical patients in their study could potentially lead to an increased incidence of seizures, but in our study, there was no statistically significant difference in the incidence of seizures between conservatively and surgically treated patients in patients monitored with EEG. The incidence of NCSE has been reported in 10% to 67% of critically ill patients in various ICU populations (Sutter, 2016). These include select patient populations, such as subarachnoid hemorrhage, with a reported frequency of NCSE of 3% to 13% (Kondziella et al., 2015), or mixed ICU populations, such as of a general surgical ICU with a reported frequency of 5% to 16% (Kurtz et al., 2014); however, there are no studies on a homogenous group of ICU patients with aSDH only. The proportion of patients with NCSE in our series, 6.5% of the monitored cohort or 1.6% of the entire patient population, is lower than the
Journal of Clinical Neurophysiology Volume 34, Number 1, January 2017
clinicalneurophys.com
S. Pollandt, et al.
Seizures and Epileptiform Discharges in Acute SDH
previously reported range. However, patients with aSDH may differ from many other studied patient cohorts in that the actual disease process of aSDH does not involve direct injury to the brain parenchyma. Furthermore, as opposed to other or mixed ICU populations, patients with aSDH are often not systemically critically ill. Most of our patient population had a Glasgow Coma Score of 14 or 15 (69%). Our results may, nonetheless, underestimate the proportion of NCSE because of the number of patients with routine rather than continuous EEG monitoring because increased capture rates of NCSE have been reported with increased EEG duration (Jette et al., 2006). Another interesting finding of our study is that the frequency of clinical seizures was much higher than that of nonconvulsive seizures or NCSE. Comparison with available data is difficult, as there are very few data on similar patient populations, but the question is raised whether aSDH may preferentially present with clinical seizures, possibly because of the different pathophysiology of a disease that does not directly affect the brain parenchyma. Additionally, our data may overestimate the rate of clinical seizures because of referral bias, as largely patients with distinct clinical events were monitored. A point rarely addressed in the TBI and SDH literatures discussing seizures is the question of whether epilepsy or seizures were present before the injury occurred. Not surprisingly, patients with history of seizures or epilepsy were significantly more likely to undergo EEG monitoring after a clinical event suspicious for seizures. However, among patients with EEG monitoring, preexisting seizures or epilepsy was not associated with a significant increase in the incidence of seizures. This finding may underline the innate association of SDH with seizures. Another factor to be considered is the use of prophylactic antiseizure medications. In our series, seizures and epileptiform EEG findings were common despite the standard use of antiseizure prophylaxis. In routine practice, most patients with TBI are prescribed an antiseizure medication for 7 days after the injury according to the Brain Trauma Foundation guidelines (Bratton et al., 2007). These guidelines were established according to the findings of a study showing significantly lower incidence of early seizures with phenytoin prophylaxis than without prophylaxis (3.6 vs. 14.2%) (Temkin et al., 1990) in patients with acute TBI. Instead of phenytoin, levetiracetam is increasingly used in neuroscience ICUs (Szaflarski et al., 2007, 2010) because of better tolerance than phenytoin. However, levetiracetam prophylaxis was also found to be associated with a higher risk of electrographic seizures and abnormal EEG (Jones et al., 2008; Radic et al., 2014). In a prospective observational study of pediatric patients with TBI, posttraumatic seizures occurred frequently despite prophylaxis with levetiracetam (Chung and O’Brien, 2016). In our study, all patients who developed seizures after admission were monitored with EEG and included in our study cohort. Our data support the notion that the incidence of seizures and SE is high despite antiseizure prophylaxis, most commonly levetiracetam, and call into question the effectiveness of current common practice. With respect to prognosis and outcomes, well-described poor prognostic factors after aSDH include older age, preoperative coma, lower coma scores, and pupillary abnormalities clinicalneurophys.com
(Cagetti et al., 1992; Kotwica and Brzezinski, 1993; Phuenpathom et al., 1993; Servadei, 1997; Servadei et al., 2000; Wiedemayer et al., 2002; Wilberger et al., 1991). Other factors influencing prognosis are radiologic findings, time to surgery, and the presence of coagulopathy (Bershad et al., 2008; D’Amato et al., 2007; Servadei, 1997; Servadei et al., 2000; Sawauchi and Abe, 2008). Seizures or EEG findings are not included in most of these analyses. Additionally, the patient cohorts examined were often patients with more complex TBI rather than isolated aSDH. We found in our series that the occurrence of clinical seizures did not affect outcomes. However, ICU-LOS was longer for patients in SE or patients with epileptiform findings on EEG, which also predicted longer hospital LOS. Mortality and discharge destination were not affected. In comparison Rabinstein et al. (2010) found that while clinical seizures or epileptiform discharges were independently associated with poor functional outcome at discharge, there was no significant effect at 1 to 6 months follow-up. The slight difference might be explained by the inclusion of conservatively treated patients in our cohort, as these patients presumably had less severe SDH. As with any retrospective study, limitations are inherent. However, in the absence of data in this patient group, our data on this well-defined homogenous subgroup of critically ill aSDH patients serve to deliver novel data and could guide a prospective approach in the future. EEG monitoring is not uniformly performed in every patient but rather requested by the attending physician based on clinical indication, resulting in an inherent selection bias. In our unit, the threshold for obtaining an EEG in patients with unexplained poor neurologic examination or acute change in neurologic examination is low, but the true incidence of epileptiform abnormalities is likely underestimated. Because of the design as an exploratory study, we did not specify one outcome as primary outcome, risking some type 1 error, however, decreasing the chance of type II error. Lastly, Acute Physiology and Chronic Health Evaluation scores or similar measures of systemic illness could not be included in our analysis because of lack of availability. In summary, events suspicious for seizures in patients with aSDH are frequent. Clinical seizures, nonconvulsive seizures, or epileptiform discharges are seen in more than half of the patients undergoing EEG monitoring. Given this high incidence, the threshold for EEG monitoring should be low and perhaps all inclusive especially in the presence of abnormalities on neurologic examination, in order to elucidate the true incidence of seizures and epileptiform findings. In judging prognosis for patients with aSDH, it is important to consider that epileptic complications contribute to increased LOS but are not associated with poor outcome.
REFERENCES Annegers JF, Coan SP. The risks of epilepsy after traumatic brain injury. Seizure 2000;9:453–457. Annegers JF, Hauser WA, Coan SP, Rocca WA. A population-based study of seizures after traumatic brain injuries. N Engl J Med 1998;338:20–24. Battaglia F, Lubrano V, Ribeiro-Filho T, et al. Incidence and clinical impact of seizures after surgery for chronic subdural haematoma [in French]. Neurochirurgie 2012;58:230–234.
Journal of Clinical Neurophysiology Volume 34, Number 1, January 2017
59
S. Pollandt, et al.
Bershad EM, Farhadi S, Suri MF, et al. Coagulopathy and inhospital deaths in patients with acute subdural hematoma. J Neurosurg 2008;109:664–669. Bratton SL, Chestnut RM, Ghajar J, et al. Guidelines for the management of severe traumatic brain injury. XIII. Antiseizure prophylaxis. J Neurotrauma 2007;24(suppl 1):S83–S86. Cagetti B, Cossu M, Pau A, et al. The outcome from acute subdural and epidural intracranial haematomas in very elderly patients. Br J Neurosurg 1992;6:227–231. Chen CW, Kuo JR, Lin HJ, et al. Early post-operative seizures after burr-hole drainage for chronic subdural hematoma: correlation with brain CT findings. J Clin Neurosci 2004;11:706–709. Chu NS. Acute subdural hematoma and the periodic lateralized epileptiform discharges. Clin Electroencephalogr 1979;10:145–150. Chung MG, O’Brien NF. Prevalence of early posttraumatic seizures in children with Moderate to severe traumatic brain injury despite levetiracetam prophylaxis. Pediatric critical care medicine. Pediatr Crit Care Med 2016;17:150–156. D’Amato L, Piazza O, Alliata L, et al. Prognosis of isolated acute posttraumatic subdural haematoma. J Neurosurg Sci 2007;51:107–111. Frontera JA, Egorova N, Moskowitz AJ. National trend in prevalence, cost, and discharge disposition after subdural hematoma from 1998-2007. Crit Care Med 2011;39:1619–1625. Hamasaki T, Yamada K, Kuratsu J. Seizures as a presenting symptom in neurosurgical patients: a retrospective single-institution analysis. Clin Neurol Neurosurg 2013;115:2336–2340. Jette N, Claassen J, Emerson RG, Hirsch LJ. Frequency and predictors of nonconvulsive seizures during continuous electroencephalographic monitoring in critically ill children. Arch Neurol 2006;63:1750–1755. Jones KE, Puccio AM, Harshman KJ, et al. Levetiracetam versus phenytoin for seizure prophylaxis in severe traumatic brain injury. Neurosurg Focus 2008;25:E3. Kondziella D, Friberg CK, Wellwood I, et al. Continuous EEG monitoring in aneurysmal subarachnoid hemorrhage: a systematic review. Neurocrit Care 2015;22:450–461. Kotwica Z, Brzezinski J. Acute subdural haematoma in adults: an analysis of outcome in comatose patients. Acta Neurochir (Wien) 1993;121:95–99. Kurtz P, Gaspard N, Wahl AS, et al. Continuous electroencephalography in a surgical intensive care unit. Intensive Care Med 2014;40:228–234. Mehryar GR, McIntyre HB. Periodic lateralized epileptiform discharges associated with subdural hematoma. Bull Los Angeles Neurol Soc 1975;40:8–12. Noachtar S, Binnie C, Ebersole J, et al. A glossary of terms most commonly used by clinical electroencephalographers and proposal for the report form for the EEG findings. The International Federation of Clinical
60
Seizures and Epileptiform Discharges in Acute SDH
Neurophysiology. Electroencephalogr Clin Neurophysiol Suppl 1999; 52:21–41. Phuenpathom N, Choomuang M, Ratanalert S. Outcome and outcome prediction in acute subdural hematoma. Surg Neurol 1993;40:22–25. Rabinstein AA, Chung SY, Rudzinski LA, Lanzino G. Seizures after evacuation of subdural hematomas: incidence, risk factors, and functional impact. J Neurosurg 2010;112:455–460. Radic JA, Chou SH, Du R, Lee JW. Levetiracetam versus phenytoin: a comparison of efficacy of seizure prophylaxis and adverse event risk following acute or subacute subdural hematoma diagnosis. Neurocrit Care 2014;21:228–237. Sabo RA, Hanigan WC, Aldag JC. Chronic subdural hematomas and seizures: the role of prophylactic anticonvulsive medication. Surg Neurol 1995;43:579–582. Sawauchi S, Abe T. The effect of haematoma, brain injury, and secondary insult on brain swelling in traumatic acute subdural haemorrhage. Acta Neurochir (Wien) 2008;150:531–536. Servadei F. Prognostic factors in severely head injured adult patients with acute subdural haematoma’s. Acta Neurochir (Wien) 1997;139:279–285. Servadei F, Nasi MT, Giuliani G, et al. CT prognostic factors in acute subdural haematomas: the value of the “worst” CT scan. Br J Neurosurg 2000;14:110–116. Sutter R. Are We Prepared to Detect Subtle and nonconvulsive status epilepticus in critically ill patients? J Clin Neurophysiol 2016;33:25–31. Swisher CB, Shah D, Sinha SR, Husain AM. Baseline EEG pattern on continuous ICU EEG monitoring and incidence of seizures. J Clin Neurophysiol 2015;32:147–151. Szaflarski JP, Meckler JM, Szaflarski M, et al. Levetiracetam use in critically ill patients. Neurocrit Care 2007;7:140–147. Szaflarski JP, Sangha KS, Lindsell CJ, Shutter LA. Prospective, randomized, single-blinded comparative trial of intravenous levetiracetam versus phenytoin for seizure prophylaxis. Neurocrit Care 2010;12:165–172. Temkin NR, Dikmen SS, Wilensky AJ, et al. A randomized, double-blind study of phenytoin for the prevention of post-traumatic seizures. N Engl J Med 1990;323:497–502. Wang HC, Chang WN, Chang HW, et al. Factors predictive of outcome in posttraumatic seizures. J Trauma 2008;64:883–888. Westmoreland BF. Periodic lateralized epileptiform discharges after evacuation of subdural hematomas. J Clin Neurophysiol 2001;18:20–24. Wiedemayer H, Triesch K, Schafer H, Stolke D. Early seizures following non-penetrating traumatic brain injury in adults: risk factors and clinical significance. Brain Inj 2002;16:323–330. Wilberger JE Jr, Harris M, Diamond DL. Acute subdural hematoma: morbidity, mortality, and operative timing. J Neurosurg 1991;74:212–218.
Journal of Clinical Neurophysiology Volume 34, Number 1, January 2017
clinicalneurophys.com
