Rehabilitation issues in Landau–Kleffner Syndrome

Landau–Kleffner Syndrome (LKS), first described in 1957 by neurologist William Landau and speech pathologist Frank Kleffner, is a rare childhood disorder in which sudden onset of a severe language disturbance and pattern of very frequent electroencephalograph (EEG) abnormalities (that appear either exclusively or predominantly during sleep) are associated [1]. LKS is considered to be on a continuum of epileptic encephalopathies that generally involves sleep potentiated epileptiform abnormalities (focal or generalized) and associated change in cognitive function. Benign rolandic epilepsy is on the mild end, while the severe end is comprised of LKS and Continuous Spike and Waves Syndrome (CSWS), which is sometimes called electrical status epilepticus of sleep (ESES) [2].

The current understanding of LKS is that it represents a primary auditory or verbal agnosia (i.e. loss of ability to give semantic significance to environmental sounds or spoken language), with associated rapid reduction in speech. Although prevalence data are limited, LKS is recognized as a rare disease by NIH [3]. Since the original case series over 50 years ago, ∼350 cases have been described in published reports internationally [4]. Onset of LKS is typically between the ages of 3–8 years and males are affected more often than females, with an approximate ratio of 2:1 [5]. All children present with a global language disorder and EEG abnormalities. Clinical seizures are the first symptom of LKS in ∼60% of cases and a language disturbance is the first symptom in the remainder [6]. Clinical seizures are not necessarily present and never appear in ∼20–30% of cases [7]. Neuroimaging studies show no obvious anatomic abnormalities [7], [8].

Perhaps because of its relatively low prevalence, LKS has received relatively little attention from rehabilitation researchers and professionals. The aim of this editorial is to raise awareness of LKS by providing a brief overview of the associated developmental and behavioural characteristics as well as available intervention research. It is hoped that this overview will stimulate interest and research into the rehabilitation of children with this syndrome. To stimulate research, some areas in need of empirical investigation are suggested.

The course of LKS can be described in three stages [9]. However, it should be noted that the course of LKS can be discontinuous such that a waxing and waning of symptoms is often observed, with alternating periods of regression and stabilization. The first stage can be represented as a sudden and significant deterioration of receptive language in children with previously normal cognitive and developmental course. The initial deterioration is followed by a period of decreased speech, which can lead to mutism. Hearing is not compromised [10], although sounds may be experienced as unpleasant and aversive [11]. It has been suggested that aphasia may be secondary to defective analysis of the child's own speech [12].

Awake EEG shows focal or multi = focal epileptiform discharges mostly on the temporal lobes [5], [8], [13], [14]. Sleep EEG shows increased activity with bilateral and generalized diffusion of epileptiform discharges [5]. Abnormal epileptiform discharges are potentiated by sleep, such that even the transition to sleep can involve significant spikes and wakes in contrast to the awake state. In many cases, spike–wake activity occupies more than 85% of slow sleep [13], which is referred to as continuous spike-and-waves (CSWS). Non-verbal skills are preserved [7]. In many cases, children present with challenging behaviour often directly linked to deficits in communication [6], [15].

The second stage is a chronic phase, during which mild cases show continuing abnormalities in language and EEG, while more severe cases show further deterioration of language as well as manifestation of behavioural problems, which can be seen in 50–70% of cases [6], [9]. These can include attention deficits, hyperactivity, oppositionality, aggression, inappropriate social behaviour, mood problems and hyperkinesia [7], [10] and have been found to be relatively independent of the severity of the language disturbance [4]. Severe behavioural problems have been linked to frontal EEG abnormalities, hypothesized to have spread anteriorly from original foci [12], in a sub-set of children [9]. Severe cases also demonstrate general cognitive decline if the duration of disturbance is long [6]. In addition, social adjustment is affected for obvious reasons [8]. The second stage can last between 1–7 years, with a mean of 4 years [9].

The third stage is that of spontaneous normalization of EEG and typically begins by 15 years of age [9]. At this time, seizures, if present, remit (rare cases have been reported with continuing seizures [8]) and problematic behaviour may improve. Studies appear to agree that EEG normalization is necessary for good language outcome, but that language difficulties can persist despite disappearance of epileptiform discharges [5], [8]. Reports vary on long-term prognosis, ranging from age-appropriate vocational and social functioning [4] to extremely limited independence [10], [15]. An estimated 50% of children experience partial remission and 10–20% of children experience almost total remission [16]. However, persistent language deficits are seen in most cases and specific deficits in phonological short-term memory (but not necessarily lexico-semantic short-term memory [17]) are seen in all cases [8], [18]. In addition, one-ear dichotic extinction in the channel contralateral to the focal side is often seen at follow-up [18], [19]. In one study, all adults with LKS considered residual language issues to be the main difficulty in their lives [8]. Language outcome has been linked strongly to length of CSWS [15] (with normalization unlikely if CSWS lasts more than 3 years [8], [9]), earlier onset of EEG disturbance [8] and fluctuations in language skills [4].

Little is known about the aetiology of LKS. There is some preliminary evidence indicating familial predisposition in cognitive impairments associated with abnormal EEGs [19]; therefore, it is premature to rule out genetics entirely as a factor in LKS. However, out of more than 300 cases reported, only two have involved sets of siblings and one set of monozygotic twins did not show concordance [20]. Therefore, a currently unknown environmental trigger may be necessary for the condition to manifest [6]. Single cases have been reported of LKS following infection or other insult, although these are rare and most cases are idiopathic [5].

The pathophysiology of LKS remains poorly understood, although many hypotheses have been advanced. Some autoantibodies to brain endothelial cells have been found to be present at higher levels in some children with LKS than in healthy controls [21] and the sera of children with LKS tested for anti-brain antibodies have shown greater reactivity against the cerebellum, auditory cortex and brainstem as compared to those of healthy controls [22], suggesting immune involvement. As previous research has shown that language function can be preserved with only one hemisphere, it seems clear that LKS causes bilateral involvement, either independently or induced by the primary hemisphere [12], [23]. Bilateral epileptic discharges are generated in the perisylvian cortex, which is related to auditory and language processing [24]. Spike-waves have been demonstrated to transiently induce impairment of auditory processing in children with LKS [25]. In addition, abnormal epileptiform discharges have been shown to be associated with transient cognitive impairment [2]. Cortical reorganization in response to continuous epileptiform activity has been suggested, which would account for residual language problems [4], [23], [26], [27]. One theory postulates that near-continuous epileptiform activity hampers synaptic pruning [28], [29], although volumetric studies do not necessarily support this conclusion [30].

LKS has been posited to be a disorder of brain maturation, given its onset in childhood and typical remission of seizures in adolescence [19]. Positron emission tomography (PET) scans have shown hypermetabolism of glucose in the superior temporal gyrus during the active phase and hypometabolism during the recovery phase [6], [18], [19], with this decrease correlated with the severity of residual phonological short-term memory impairment [18], [27]. An overarching theory of pathophysiology suggests that hyperexcitability of neurons in the perisylvian cortices related to spike-wave generation leads to the continuous spike-waves seen in sleep following thalamo-cortical uncoupling and subsequent changes in the blood–brain barrier lead to suppressed seizures and onset of auditory agnosia [18].

Interventions for LKS can be broadly divided into pharmacologic, surgical and behavioural. Unfortunately, there is a paucity of research involving participants with LKS and there is no well-developed protocol for intervention [5], [28]. There is consensus in the field that multi-modal treatment is essential and that educational and behavioural therapies focusing on communication skills should be ongoing while pharmacologic and/or surgical therapies are trialled [5], [14].

There have been no controlled studies conducted of pharmacalogic therapies for LKS. Anti-epileptic drugs (AEDs) are the first line pharmacologic treatment for LKS and have been found in many cases to reduce epileptiform discharges with inconsistent effects on language [15], [31]. Some studies indicate that high dose corticosteroids for a short period (3–6 months) result in improvement in language and EEG disturbances [32–34]; others have shown mixed results for language [35]. Side-effects of corticosteroids can be problematic [35]. Intravenous immunoglobulin has been used both with good effect [36] and no effect [37]. LKS refractory to pharmacologic treatment may be considered for multiple subpial transaction (MST). This surgical technique selectively disrupts horizontal intra-cortical fibres, believed to spread epileptiform discharges, while preserving vertical fibres, which maintain nerve pathways, typically in the superior and middle temporal gyri and substantial suprasylvian area [38], [39]. Although immediate and dramatic improvements have been reported for seizure activity and behaviour following MST, recovery of language function has been variable [39], [40].

Rehabilitative behavioural intervention approaches are extremely under-studied in LKS. Additionally, unlike most conditions that present to rehabilitation centres, LKS is not the result of a static injury. As the syndrome is a dynamic process with a discontinuous course, studying treatment effectiveness can be difficult. Clinical practice guidelines encourage intensive speech therapy with a focus on maintaining residual language [41], although no empirical literature is currently available on the efficacy of speech therapy. Case studies inconsistently demonstrate that children with LKS may rely more on prosodic cues than healthy controls to support comprehension and this may be an appropriate target of speech therapy [42]. As non-verbal abilities are preserved, visual methods of communication are often encouraged [15], [41]. Case reports suggest that sign language allows communication and does not impede reacquisition of spoken language [11], [43]. Patients and families have reported that sign language has been pivotal in allowing fuller participation of the child in family and social settings, with a concomitant increase in quality-of-life [11]. Finally, cued speech, which involves handshapes made near the mouth to supplement lip-reading, has been suggested as promising based on anecdotal data [15], [44].

Classroom rehabilitative approaches have not been studied empirically. A qualitative case series describing four students with LKS reports that substantially separate classrooms, individualized attention, pre-teaching and a functional approach to communication are highly recommended [41]. Case reports document success with using highly visual strategies in the classroom, as processing of non-verbal information is unimpaired in LKS [15]. For example, picture schedules, colour coding to support letter and sound identification while reading, storyboards for instructions and graphic conversation have been reported to be helpful to students [15], [41], [44]. Computer programs are suggested to support reading comprehension [41].

Behavioural intervention has been encouraged as a treatment modality in this population [5], [45], [46], but research remains extremely scarce. In a report concerning a 9-year old boy in a psychiatric inpatient facility, functional analysis was used successfully to assess verbal and physical aggression and encopresis [47]. Response cost, over-correction and differential reinforcement of other behaviours were implemented for this child. Although quantitative data were not collected, the researchers reported that behaviours decreased to below baseline frequencies and the child was discharged from the facility, with no readmissions at 1- and 3-year follow-up checks.

Finally, a single case report of an intense 12-week hemispheric integration therapy without concurrent AED therapy had anecdotal positive results in the domains of communication, auditory processing, adaptive skills and EEG normalization. The protocol involved pre-training multi-modal sensory inputs (e.g. optokinetic and vestibular) and metronome therapy focusing on improving timing, rhythmicity and sustained attention [48].

In conclusion, LKS is a disease whose development and characteristics are becoming increasingly clear. The evidence on intervention strategies, however, is still basic or preliminary. Research is therefore warranted on medical (i.e. pharmacological, surgical and dietary) strategies as well as on educational and rehabilitation strategies, so as to produce evidence-based guidelines that could help families and school contexts deal with this problem satisfactorily.