Cognitive outcome of status epilepticus in adults

Authors


Address correspondence and reprint requests to Professor Christoph Helmstaedter, Department of Epileptology, University of Bonn, Sigmund-Freud-Strasse 25, 53105 Bonn, Germany. E-mail: Christoph.Helmstaedter@ukb.uni-bonn.de

Summary

There is no doubt that structural morphological brain lesions and malformations in epilepsy represent major etiological factors for the cognitive impairments seen in this disease. The role of epileptic activity and seizures for cognition and cognitive development, however, is less easily determined. Epileptic dysfunction ranges from interictal and periictal activity over self-terminating seizures to non-convulsive and convulsive status epilepticus, which appear the most severe conditions along this continuum. The decisive question in this regard is as to whether cognitive impairments observed in the acute epileptic condition are reversible or not. Impairments from interictal or postictal epileptic dysfunction are reversible and may interfere at most with brain maturation and cognitive development in the young patient. Seizures and ictal dysfunction in contrast, even when reversible, can leave a permanent trace which extends the phase of postictal recovery. As for status epilepticus and subsequent cognitive decline it often remains open whether the epileptic condition itself or the underlying clinical condition is causative for the aftermath. While there is evidence for both possibilities, group data from neuropsychological cross sectional and longitudinal studies indicate that more severe mental impairments, which in turn indicate more severe clinical conditions, appear to be a risk factor for sustaining status epilepticus, rather than that status epilepticus causes the cognitive decline. Reviewing the literature the cognitive condition in patients with status epilepticus varies with the type of epilepsy, the etiology of epilepsy, severity of the status, and the age of the patient.

Whereas the question of whether self-limiting seizures damage the brain is still a matter of debate—according to a recent discussion by Sutula and Pitkänen (2002) this appears to be the case “in some individuals in some conditions”—there is ample evidence that a status epilepticus can be a life threatening condition, which can also lead to permanent mental impairment. Convulsive and nonconvulsive statuus epileptici (CSE, NCSE) mark the negative endpoint of a suggested continuum of epileptic dysfunction where the more positive endpoint would be interictal transient cognitive impairment (TCI).

Subtle and mostly reversible cognitive change due to interictal epileptic activity can be demonstrated by informal or formal testing in parallel to EEG monitoring (Binnie, 2003; Aldenkamp and Arends, 2004). A more direct relation between epileptic EEG activity and transient cognitive change, mental retardation or decline is most obvious in the idiopathic and the benign partial epilepsies. In adult focal symptomatic epilepsies the relation is less obvious. However in this type of epilepsy, secondary involvement of frontal lobe functions in temporal lobe epilepsy, short- and long-term recovery from cognitive impairments after successful seizure control, evidence of a relation between epileptic activity and language dominance, and the phenomenon of accelerated forgetting provide indirect evidence of a chronic interictal epileptic dysfunction, which is not necessarily directly related to epileptiform discharges (Hermann et al., 1988; Jokeit et al., 1997; Helmstaedter et al., 2003; Janszky et al., 2006; Mameniskiene et al., 2006). Overt seizures leave significant marks in the postictal phase. Even when patients appear fully recovered in terms of regaining consciousness, orientation and speech, cognitive testing shows continuing impairment in partial functions. Recovery from seizures is hierarchic, starting with distant and ending with the focal area, it is related to seizure severity, and it may take hours if not days in single cases (Helmstaedter et al., 1994). Postictal cognitive impairments are fully reversible in most cases. Regarding the impact of uncontrolled epilepsy on the course of cognition, recent longitudinal studies in homogeneous patient groups with chronic uncontrolled temporal lobe epilepsy demonstrate some decline in memory over time. A progressive dementing effect of chronic epilepsy can hardly be demonstrated (Helmstaedter et al., 2002; Rausch et al., 2003; Helmstaedter et al., 2003; Hermann et al., 2006).

As for status epilepticus, the convulsive status (CSE) represents a clear-cut condition, after which its short and long-term cognitive consequences can be assessed. A nonconvulsive status epilepticus (NCSE) in contrast is sometimes difficult to diagnose. NCSE may become manifest only in subtle behavioral changes or cognitive impairments. Symptoms range from mild dysphoric alterations to confusional states, nonresponsiveness, aphasia, apraxia, amnesia, sometimes going along with bizarre behavioral changes (Wieser et al., 1985; Bauer et al., 1997; Helmstaedter, 2001). In an own series of six patients tested during NCSE, the most prominent behavioral feature was that patients showed responsive and reflexive behaviors but almost no self initiated directed behavior (see Table 1). In NCSE, negative symptoms clearly dominate positive symptoms. Testing during NCSE indicates an inhibition of the affected hemisphere (Helmstaedter, 2001; Profitlich et al., 2007). However, disinhibition or release of the nonaffected hemisphere has also been described (Regard et al., 1985). NCSE responds well to diazepam, particularly when the EEG shows generalized EEG patterns (Granner and Lee 1994). In doubtful cases slow titration of diazepam going along with repeated neurocognitive testing can unmask a NCSE.

Table 1.  Ictal symptoms in nonconvulsive status epilepticus (n = 6)
PerformanceImpairment
  1. Main feature: “pathological inhibition” and “negative symptoms” rater than “exitatory” and “positive” symptomatology [frontal dysexecutive?].

Consciousness:Partly reduced, great fluctuations
[impaired in 4 of 6]Never completely lost
Executive functions:No self-initiated directed behaviors
[impaired in 6 of 6]Preserved responsiveness, but slowed, reflexive, and restricted to single modalities preservative behaviors
Inadequate intrusive behaviors
Problems with concept formation & response inhibition (color/form…)
Very limited working memory
Higher order functions:Apractical signs in object use and imitation
[impaired in 6 of 6]Receptive/expressive dysphasia
Transcortical aphasia
Dyscalculia
Dyslexia
Amnesia
No global amnesia!
EmotionEmotional instability [dysphoric, irritated, angry]
[changed in 4 of 6] 

Status epilepticus represents the most extreme condition in epilepsy since it can have permanent mentally disabling or retarding consequences for the patient. This however varies with the type of epilepsy (idiopathic vs. symptomatic), its etiology (acute or progredient neurological condition versus evolving within the course of epilepsy), seizure severity (generalized vs. focal), duration, and chronological age of the patient. Whether convulsions are a necessary prerequisite for cerebral damage and thus for mental impairment due to a status is a matter of debate (Young and Jordan, 1998; Aminoff, 1998).

According to the literature, the direct cognitive consequences of a status epilepticus appear less severe in idiopathic epilepsies as compared to symptomatic epilepsies, and in children as compared to adults (Lothman and Bertram, 1993; Shorvon and Walker, 2005). However in children, the less severe direct sequels of a status may by part be outweighed by its negative impact on mental development in the long run as it is the case in the ESES or Landau-Kleffner syndrome.

In symptomatic epilepsies, reported cognitive impairments after a grand mal status range from 8% to 26% (Rowan and Scott, 1970; Oxbury and Whitty, 1971; Aminoff and Simon, 1980). Impairments following a focal status appear related to the localization of the status. At least for seizures with a limbic origin, animal research and single case reports indicate specific mnestic impairments as a possible consequence of a status (Ben-Ari, 1985; Engel et al., 1978; DeGiorgio et al., 1999). Observation of two cases of permanent global amnesia after a status in temporal lobe epilepsy supports this suggestion (Dietl et al., 2004). With both patients however it is important to note that they had temporal lobe surgery before they experienced the status. Performance change in one of these patients is demonstrated in Table 2. In this patient, who had had a right standard 2/3 temporal lobectomy, figural memory was throughout very poor. After initial postsurgical improvements, the patient was about 1 year seizure free, verbal memory severely dropped after a status that occurred in the time after the 1 year postoperative follow-up evaluation. IQ did not experience a significant change over time.

Table 2.  Global amnesia after SE experienced one year after right temporal lobe surgery (Female patient with right TLE, 60 years at first visit)
 Preop 02.19903 m. post 06.901year post 01.91After status 05.91After status 1994
  1. sv, standard value; rs, raw score.

Vocabulary IQ124
MWT-B 
Full scale IQ107105
WAIS 
Attention (sv)9597100
Letter cancellation 
d2 
Perceptual speed2325232725
Symbol counting 
c.I.-Test (rs) 
Interference2925253028
A/B test 
c.I.-Test (rs) 
Digits f/b (rs)6/58/66/66/46/5
Corsi f/b (rs)5/55/45/66/56/4
Verbal learning3454422424
VLMT (rs) 
Verbal retention58410
VLMT (rs) 
Verbal recognition (rs) VLMT131414−2−5
(Correct–errors) 
Figural learning11111
DCS-R (rs) 
Visual retention5/85/96/74/93/10
Benton Test 
Correct/errors (rs) 

In 1990 Dodrill and Wilensky reviewed the literature on cognitive outcome of status epilepticus and the authors stated that most studies attributed at least some negative effects to SE, and that studies in children showed negative effects at least in the very young. However, the data were medical records and interviews rather than formal psychometric tests, most studies were retrospective, and the older studies reported greater deterioration than the more recent. The latter observation was interpreted as being due to unknown etiologies. Of greater interest are Dodrill's and Wilensky's own longitudinal data from 143 adults where those nine patients, who suffered from a status within the 5-year follow-up interval, turned out to be duller (IQ difference of 10–18 pts.) than those without a SE already at baseline. Analyses indicated some intellectual decline over time in the status group but looking closer at the data one may as well discuss that the status group did not show amelioration over time, which was observed in the group without SE. With caution it was concluded that “in some individuals the effects were definitely greater, but that in many persons no effects were discernible ….” A similar and more recent study by Adachi et al., (2005) confirms the impression from this study that there are no major cognitive changes after SE. Fifteen patients with SE, in comparison to those without SE, failed to show any significant post-SE intellectual decline. Furthermore, neither the SE-related variables nor the clinical characteristics were correlated with intellectual outcome.

These two studies gave reason to retrospectively analyze the cognitive profiles of own patients with focal epilepsy as dependent on whether they had a status in their history or not. The results indicate significantly poorer education, IQ, motor functions and attention in patients with a CSE or NCSE. Interestingly, memory functions, which are particularly vulnerable with respect to hypoxia and brain damage in general, were not differentially affected (see Table 3). Furthermore, the cognitive profiles of patients with NCSE versus CSE were not different. Clinically, an earlier onset of epilepsy in the status group and an involvement of frontal lobe structures in NCSE were of particular interest. In summary the profiles indicate retardation rather than decline or a loss of partial functions in the patients with a status in their history. Together with the fact that the status groups did not differ from each other this may well be taken as evidence that brain damage and impairments preceding the status prepare the ground for its occurrence, rather than that the status itself adds to the impairments (Hilkens and de Weerd, 1995). Mental retardation or decline in epilepsy may thus be seen as a marker of a pathological condition, which can lead to a status. The other way around fitter and more differentiated brains would then appear less likely to experience a status. After all, etiology of the epilepsy appears to be the key variable for occurrence and the prognosis of SE. Therefore, change in neuropsychological measures in the time after CSE or NCSE needs to be differentiated from possible change due to progression of the pathology underpinning the epilepsy (see also Krumholz et al. 1995).

Table 3.  Patients with CSE or NCSE versus patients without SE in their history
 NCSE n = 30CSE n = 36No SE n = 88Sign. p
  1. # onset is earlier in patients with status versus without status (m 17/SD14 vs. m21/SD14, p < 0.1).

  2. ## in patients with NCSE frontal lobe involvement is more likely (73%, vs. 25% and 23%) (chi2 = 15.26, p = 0.0005).

Sex (n = 154)
 Male 19/63%17/47%47/53%n.s.
 Female 11/37%19/53%41/47% 
Age (n = 154)Years m/SD35.1/22.535.9/14.835.9/14.0n.s.
Handedness (n = 150)
 Right 21/78%34/94%78/90%n.s.
Age at epilepsy onset (n = 146)Years m/SD.17/1616/1321/15n.s.#
Duration of epilepsy (n = 146)Years m/SD17/1719/1214/12 
Localization (n = 89)
[MRI & or EEG]
 Frontal 8/42%5/22%8/17%0.027
 Temporo-mesial 4/21%9/40%20/43%##
 Temporal 1/ 5%5/22%12/25% 
 Parietal/occipital 0/ 0%1/ 4%4/ 9% 
 Hemispheric 3/16%1/ 4%0/ 0% 
 Multifocal 3/16%2/ 8%3/ 6% 
Lateralization (n = 89)
 Right 8/44%11/48%19/40%n.s.
 Left 5/28%10/44%22/45% 
 Bilateral 5/28%2/ 8%7/15% 
Education (n = 130)
 No formal education 3/13%7/23%1/ 1%0.001
 <10 years 17/74%15/50%48/62% 
 >10 years 3/13%8/27%28/37% 
IQ (n = 88) 
 <85 3/21%9/38%5/10%0.02
Attention (n = 135) <m –1SD
 18/78%24/75%36/45%0.001
Motor fct. (n = 122) <m –1SD 16/84%22/79%37/49%0.001
Language (n = 116) <m –1SD 14/73%18/67%44/63% 
Visuoconstruction (n = 102) 10/67%14/64%30/46% 
<m –1SD 
Verbal memory (n = 127) 12/57%21/70%52/68% 
<m –1SD 
Figural memory (n = 126) 13/62%21/70%46/61% 
<m –1SD 

This may be demonstrated with a very recent case of a female patient, who after normal development, had her first seizure with 11 years. In the time after she showed a slow but progressive worsening of her general condition and of her intellectual capabilities as well (Table 4). First neuropsychological testing was done when she was 17 years old, ½ year before the first status. Testing indicated a learning disability with poor performances in almost all domains except for motor and verbal memory functions. Seizures were complex partial and secondarily generalized and at the time of admittance, seizure semiology and EEG indicated right frontal epilepsy. One week after experiencing a first CSE she showed declined figural memory, a month after a second status, which required intensive care, she had declined in all domains but motor functions and verbal memory. The last follow-up was done two months after the second and three weeks after a third status which required intensive care in another clinic. Testing indicated a worsened condition with extreme general slowing and a drop now also in verbal memory and motor functions. The condition did not any longer allow formal testing. MRI indicated progressive changes over time with no abnormalities at the first visit, increasing thalamic and right parietal signal intensities after the first two SE and with additional contralateral left frontal and parietal signal changes after the third SE. Accumulation of MRI changes over time, which did not yet show any recovery, indicated that in this case the status appears a sequel rather than the cause of the mental decline. After having excluded mitochondrial diseases, encephalitis etc., a diagnosis is awaited now from biopsy.

Table 4.  Patient with progressive mental decline
IntervalsBaseline6 months 1 CSE1 month 2 CSE2 months 3 CSE
IQ<1SD (WAIS-R IQ 76)
Execut fct.<2SD↓↓↓
Motor fct.o.K.↓↓
Verb memory<1SD↓↓
Fig. memory<1SD↓↓
Language<1SD↓↓
Visuo-constr.<2SD↓↓
Agnosia/Anopsia/Neglectno
Disorientationno

When severe seizures and an SE can lead to irreversible damage the question is whether this might be prevented. In this respect recent reports of neuroprotective effects of the SE treatment in animal models are of major interest. In particular topiramate (TPM) and maybe also levetiracetam (LEV) appear to have positive effects on mitochondrial dysfunction, hippocampal cell survival, and behavior (Cha et al., 2002; Kudin et al., 2004; Gibbs et al., 2006; Frisch et al., 2006) (see Fig. 1). Despite some unsolved inconsistencies, in our own study the positive effects of TPM on a cellular hippocampal level were not correlated with the positive effects on water maze performance, and only a medium but not a high dose TPM had a positive effect on behavior, such findings are promising and await confirmation in human epilepsy.

Figure 1.


(Independent) positive effects of TPM on poststatus hippocampal cell counts and water maze performance.

Ancillary