Neurocrit Care DOI 10.1007/s12028-014-0070-0

ORIGINAL ARTICLE

Non-convulsive Status Epilepticus and Non-convulsive Seizures in Neurological ICU Patients Ikuko Laccheo • Hasan Sonmezturk • Amar B. Bhatt • Luke Tomycz Yaping Shi • Marianna Ringel • Gina DiCarlo • DeAngelo Harris • John Barwise • Bassel Abou-Khalil • Kevin F. Haas



Ó Springer Science+Business Media New York 2014

Abstract Background Non-convulsive seizures (NCS) or nonconvulsive status epilepticus (NCSE) has been reported in 8–20 % of critically ill patient populations, and delayed diagnosis and treatment of NCSE may lead to increased mortality. This study seeks to better understand the risk factors, characteristics, and outcome of NCS/NCSE in the neurological ICU. Methods This is a prospective observational study, recruiting consecutive patients admitted to the adult neurological ICU with altered mental status. Patients with anoxic brain injury were excluded from the study. Data were collected and analyzed for prevalence of NCSE/NCS, EEG patterns, associated risk factors, treatment response, and final outcome. Results NCSE/NCS was detected in 21 % of 170 subjects. Clinical seizures preceded EEG diagnosis of NCSE/ NCS in 25 % of cases. Significant risk factors for NCSE/ NCS were a past medical history of intracranial tumor, epilepsy, or meningitis/encephalitis, or MRI evidence of encephalomalacia. Subtle clinical findings such as

I. Laccheo (&) Epilepsy Center of Excellence, Veterans Affairs Medical Center, Richmond, VA, USA e-mail: [email protected] I. Laccheo Virginia Commonwealth University, Richmond, VA, USA H. Sonmezturk  A. B. Bhatt  Y. Shi  M. Ringel  G. DiCarlo  D. Harris  J. Barwise  B. Abou-Khalil  K. F. Haas Vanderbilt University Medical Center, Nashville, TN, USA L. Tomycz Seattle Children’s Hospital, Seattle, WA, USA

twitching of oral or ocular muscles and eye deviations were found on exam in 50 % of the NCSE/NCS group. Mortality was increased in NCSE cases as 31 % of NCSE/NCS patients died compared to 14 % in non-NCSE/NCS group. Conclusions Specific clinical features along with history and imaging findings may be used to identify patients at high risk of NCSE/NCS in the neurological ICU. Keywords Neurological ICU  Critical care  Altered mental status  Non-convulsive seizures  Non-convulsive status epilepticus  EEG

Introduction Continuous EEG (cEEG) monitoring in the critical care setting is increasingly utilized to detect and treat seizures associated with altered mental status (AMS) or coma. It is important in neurological critical care to diagnose and treat seizures in a timely manner; however, diagnosis of seizures in obtunded or comatose patients can be challenging as clinical manifestations are often subtle or absent. Approximately, 90 % of recorded seizures are purely nonconvulsive in this patient population [3, 6]. Previous retrospective and prospective observational single-center studies focusing on ICU patient populations have shown a variable frequency of non-convulsive seizures (NCS) or non-convulsive status epilepticus (NCSE), ranging from 8 to 20 % [1–11]. In patients with coma after convulsive status epilepticus, up to 48 % were found to have NCS or NCSE [12, 13]. In the setting of limited resources, many clinicians and intensivists may have limited access to cEEG, yet do not want to risk delayed diagnosis and treatment of NCSE which has been associated with higher patient morbidity and mortality [5]. Previous studies

123

Neurocrit Care

reported mortality rates of 3, 19, and 32 % for seizures lasting 12 h in 5 subjects). Patients with coma due to operative anesthesia, anoxic brain injury, or hypothermia were excluded. Patients underwent brain imaging (CT and/or MRI as medically indicated by the physicians caring for the patient) and were placed on EEG monitoring which was continued as medically indicated. NCSE was defined as per published criteria [21]. Periodic discharges without a clear evolution in amplitude, frequency, or morphology were noted but were not classified as seizures. NCSE was defined as continuous or recurrent ictal discharges lasting for at least 30 min. Data were collected and analyzed for prevalence of NCSE and/or NCS, EEG patterns, associated risk factors, AED use, treatment response, and eventual outcome. For continuous variables, the mean and standard deviation (SD) and for categorical variables, the frequency and percentage were presented. Unadjusted demographic and clinical

123

One hundred and seventy patients were enrolled between January 2012 and October 2013. Thirty seven percent (62/ 170) of all subjects were intubated and 25 % (42/170) were on intravenous (IV) continuous sedation at the time of EEG. Eighty eight percent (154/170) of subjects had EEG because of change in mental status. The most common reason was fluctuation of mental status (45 %, 69/154), followed by sudden decline in neurological status (29 %, 44/154), unexplained stupor (12 %, 18/154), and agitation (4 %, 7/154). Median duration of cEEG was 24 h, and 76 % (129/ 170) of subjects had cEEG > 12 h. Ninety one percent (154/170) of EEG results were abnormal (Table 1). NSCE/ NCS was detected in 21 % (36/170) of patient cases. In these cases, 36 % (13/36) had NCS (Fig. 1a). Eight percent (3/36) had NCSE lasting greater than 30 min with no further seizures after initial treatment, 42 % (15/36) had a continuous NCSE pattern, and 14 %(5/36) had an intermittent/recurrent NCSE pattern. Eighty six percent of cases (31/36) had seizures with focal onset while 14 % (5/36) had generalized onset. In cases with focal seizures, seizure onset was most often in the temporal and frontal regions (42 % and 26 % of all the NCS/NCSE cases) (Fig. 1b). Background was diffusely slow in 61 % of all cases (104/170). Lateralized abnormalities were found in 44 % of all cases (76/170). Lateralized periodic discharges (LPDs, formerly called periodic lateralized epileptiform discharges or PLEDs) were

Neurocrit Care

a NCSE: Intermittent/ recurrent seizure pattern (5) 14%

NCS (13) 36%

NCSE: continuous seizure pattern (15) 42%

NCSE: seizures resolved with initial treatment (3) 8%

b Parietal (1) 3%

Occipital (5)14%

Generalized (5) 14%

Frontal (9)25% Temporal (16) 44%

Fig. 1 a NCSE/NCS pattern. b NCSE/NCS location

seen in 8 % (13/170) of cases (Case example 1, Fig. 2), and focal interictal epileptiform discharges were found in 15 % (26/170) of all cases (Case example 2, Fig. 3). Patient characteristics are listed in Table 2. There were no significant differences in age or gender in patients found to have NCSE/NCS. Past medical history diagnoses associated with an increased risk of NCSE/NCS were epilepsy (31 %, 11/36 of NCSE/NCS vs. 13 %, 17/134 of nonNCSE/NCS group; OR 3.01, p = 0.01), history of intracranial tumor (19 %, 7/36 of NCSE/NCS group vs. 5 %, 7/134 of non-NCSE/NCS group; OR 4.46, p = 0.006), and history of meningitis/encephalitis (11 %, 4/36 of NCSE/ NCS group vs. 1 %, 2/134 of non-NCSE/NCS group; OR 12.24, p = 0.005). The only primary diagnosis at admission that was associated with increased risk of NCSE/NCS was clinical seizures, which was present in 61 % of the NCSE/NCS group compared with 25 % of those without NCSE/NCS (OR 4.69, p < 0.001).

Subtle signs such as eye deviation or twitches of facial/ oral muscles were found in 50 % of the NCSE/NCS group (18/36) compared to 19 % (25/134) of the non-NCSE/NCS group (p < 0.001, Table 3). Thirty seven percent of patients (62/170)) were intubated and 25 % (42/170) were on intravenous (IV) continuous sedation, most commonly with propofol (36 out of 42 patients), at the time of EEG. Fifty two percent (87/170) and 64 % (109/170) of all subjects received IV antiepileptic drugs (AED) and intravenous benzodiazepines (BZD) prior to EEG due to clinical concern for seizure, respectively (Table 3). Of the 14 NCSE/NCS patients who underwent IV BZD trial, 13 had EEG improvement, although only 1 improved clinically. Twenty five percent (43/170) of patients were on home AEDs at baseline and on admission. Overall, the EEG changed AED management in 39 % of all (66/170) subjects. We had hypothesized that imaging findings could help identify patients at greater risk for NCS/NCSE. All patients had a brain MRI or CT prior to EEG. The majority of patients had abnormal image findings: 94 % (34/36) in the NCSE/NCS group and 87 % (117/134) in the non-NCSE/ NCS group (Table 4). Acute (new) imaging abnormalities were common, and severity of acute abnormalities such as imaging evidence of mass effects was found at similar rates between NCSE/NCS and non-NCSE/NCS groups (Table 4). On the other hand, chronic abnormalities were found in 50 % (18/36) of NCSE/NCS subjects compared to 31 % (41/134) of non-NCSE/NCS subjects (p < 0.03). Changes consistent with encephalomalacia were found in 17 % of NCSE/NCS subjects (6/36) compared to 5 % (7/ 134) of non-NCSE/NCS subjects (p < 0.022). Chronic cortical abnormalities showed a trend to more risk for NCSE/NCS than subcortical abnormalities, and there was increased risk with abnormalities in the parietal and occipital regions (Table 4). We also evaluated response to treatment and outcome in patients with NCS/NCSE and those without (Table 5). Sixty seven percent of patients with NCSE/NCS (24/36) had improved mental status after treatment with AEDs, but mortality was significantly higher in the NCSE/NCS group (31 %, 11/36) compared to non-NCSE/NCS group (14 %, 19/134, p = 0.003). Comparing subgroups of NCSE/NCS, patients with generalized ictal onset had significantly higher mortality rate compared with the ones with lateralized onset (44 %, 4/9 vs. 28 % 7/26, p = 0.003). Glasgow coma scale (GCS) at initiation of EEG and age did not differ significantly between patients with and without NCSE/NCS. All eight patients who underwent continuous anesthetic infusion for burst suppression for the treatment of NCSE died: three patients died following withdrawal of care after family discussion, two patients had complications (one cardiopulmonary, the other propofol

123

Neurocrit Care

Fig. 2 a–d Case example 1. A 64-year-old male presented with intermittent aphasia and disorientation with a GCS of 9. a–c Initial EEG showed lateralized periodic discharges with a modifier fast

activity (formerly called PLEDS-plus), which transitioned into ictal discharges arising from the left posterior region. d Brain MRI showed a hemorrhagic mass, known metastatic squamous cell lung carcinoma

infusion syndrome), two had drug resistant recurrent seizures, and one had brain herniation from stroke. Four patients were on pentobarbital, three were on propofol, and one on midazolam continuous infusion. In patients who

survived 26 % (9/36) of NCSE/NCS subjects and 35 % (46/134) of non-NCSE/NCS had returned to baseline neurological status at discharge. Forty three percent (15/36) of NCSE/NCS subjects and 51 % (68/134) of non-NCSE/

123

Neurocrit Care

Fig. 2 continued

NCS subjects were discharged with new neurological deficits. Patients with NCSE/NCS had both prolonged ICU and hospital stay compared to those without NCSE/NCS, but this did not reach statistical significance. Median total duration of ICU stay was 10 days for NCSE/NCS group and 7 days for the non-NCSE/NCS group. Median total duration of hospital stay was 12 days for the NCSE/NCS group and 9 days for the non-NCSE/NCS group (Table 6).

Discussion This study detected electrographic seizures in 21 % (36 out of 170) of patients in neurological ICU with AMS (13 had NCS, 23 had NCSE). This prevalence is similar to that reported in Claassen et al., which found NCS in 9 % and NCSE in 19 % of comatose patients [3]. Other similar studies in neurological ICUs have also found NCS in

123

Neurocrit Care

Fig. 3 a–c Case example 2. A 69-year-old male who was found down with a GCS of three and intubated. a The patient had a focal ictal discharge arising from left frontal region 17 h into EEG, despite

an ongoing propofol drip. b, c Brain MRI brain showed a butterfly glioblastoma multiforme (GBM)

13–34 % [6, 10, 22]. As with previous studies, our study showed that NCSE/NCS was often found in patients with persistent AMS following clinical seizures. Half of the patients who found to have electrographic seizures did have subtle twitching of facial and/or perioral muscles or eye deviation. This highlights the importance of subtle clinical findings in raising the suspicion for NCSE/NCS. A previous study also indicated that certain clinical features such as abnormal ocular movements are more likely to be

present in patients in NCSE compared with other types of AMS [20], and were useful in deciding which patients should have an urgent EEG. Importantly, we identified several past medical history diagnoses that are associated with increased risk of NCSE/ NCS. Similar to a previous retrospective study, we found that a past history of epilepsy is highly associated with an increased risk of NCSE/NCS [3]. A novel finding in our study is that a past history of meningitis/encephalitis was

123

Neurocrit Care Table 2 Patient characteristics

Table 4 Image finding

Total n = 170

NCSE/NCS n = 36 (21 %)

Others n = 134

Age

56 ± 16

59 ± 16

p value OR

0.18

0.99

Gender Male Female

42 % (15) 58 % (21)

57 % (77) 43 % (57)

0.091

0.55

Others n = 134

p value

Abnormal image

94 % (34)

87 % (117)

0.23

Acute/new

75 % (27)

69 % (94)

0.51

Chronic/old

50 % (18)

31 % (41)

0.03

Encephalomalacia

17 % (6)

5 % (7)

0.022

62 % (21)

63 % (72)

0.93

22 % (8)

19 % (25)

0.15

Hydrocephalus Herniation

8 % (3) 3 % (1)

10 % (13) 8 % (11)

0.8 0.26

Others

6 % (2)

1 % (2)

0.15

8 % (3)

5 % (7)

0.48

Frontal

22 % (8)

10 % (14)

0.062

Parietal

22 % (8)

7 % (9)

0.006

8 % (3)

1 % (2)

0.031

14 % (5)

15 % (20)

0.88

Midline Shift

Epilepsy

31 % (11)

13 % (17)

0.01

3.01

CNS tumor/cancer

19 % (7)

5 % (7)

0.006

4.46

Meningitis/ encephalitis

11 % (4)

1 % (2)

0.005 12.24

ICH

6 % (2)

1 % (2)

0.15

6.32

Ischemic stroke

8 % (3)

13 % (18)

0.41

0.58

Other tumor/cancer

6 % (2)

7 % (9)

0.8

0.85

Traumatic brain injury 3 % (1)

1 % (1)

0.32

3.06

VPS

6 % (2)

2 % (3)

0.3

3.13

61 % (22)

25 % (34)

Non-convulsive status epilepticus and non-convulsive seizures in neurological ICU patients.

Non-convulsive seizures (NCS) or non-convulsive status epilepticus (NCSE) has been reported in 8-20 % of critically ill patient populations, and delay...
2MB Sizes 0 Downloads 6 Views