INTRODUCTION — Kleine-Levin syndrome (KLS), also called recurrent hypersomnia, is a rare sleep disorder characterized by recurrent episodes of severe hypersomnia associated with cognitive and behavioral disturbances such as confusion, derealization, apathy, compulsive eating, and hypersexuality. Episodes last a few days to several weeks and are separated by weeks or months of normal sleep and behavior [1]. KLS is categorized as a central disorder of hypersomnolence; other disorders in this category include narcolepsy type 1 and type 2 and idiopathic hypersomnia. (See "Classification of sleep disorders".)
The epidemiology, pathogenesis, clinical features, diagnosis, and treatment of KLS will be reviewed here. A general approach to the patient with excessive daytime sleepiness and other central disorders of hypersomnolence are discussed elsewhere. (See "Approach to the patient with excessive daytime sleepiness" and "Clinical features and diagnosis of narcolepsy in adults" and "Treatment of narcolepsy in adults" and "Idiopathic hypersomnia".)
EPIDEMIOLOGY — Kleine-Levin syndrome (KLS) is a rare disorder with an estimated prevalence of one to five cases per million population [2-8]. Patients with primary KLS can be found worldwide, with a higher prevalence among Ashkenazi Jews. (See 'Etiology and pathogenesis' below.)
KLS is more common in males. In several large observational series, the proportion of males ranges from 60 to 87 percent [4-8].
ETIOLOGY AND PATHOGENESIS — The etiology of Kleine-Levin syndrome (KLS) is unknown. Although a psychological basis was considered most likely when the disorder was first described, contemporary hypotheses have focused primarily on genetic and immune-mediated causes.
●Genetic factors – A genetic basis for the disorder has been suggested by the preponderance of cases in the Ashkenazi Jewish population, suggesting a founder effect, as well as numerous reports of familial cases. To date, however, no specific genes have been identified.
Although the large majority of KLS cases are sporadic, familial cases account for approximately 8 percent of patients with primary KLS [3,5,8-19]. In one of the larger families described, a father, three sons, and two daughters were affected [12]. Two pairs of monozygotic twins who were concordant for the syndrome have also been reported [10,11].
The estimated risk of KLS in a first-degree relative of a patient with KLS is 1 percent. While this is a relatively low risk, it is 800 to 4000 times higher than the general population using an estimated KLS prevalence of 2 to 10 per million and five first-degree relatives per proband. The risk of psychiatric disorders (eg, bipolar disorder and depression) in first-degree relatives of patients with KLS is similar to that of controls [5,6].
●Precipitating factors – A variety of potential precipitating factors have been reported in patients with KLS. There appears to be seasonal variation in incidence, with first episodes occurring most commonly in autumn or winter. In one large series, 89 percent of patients remembered one or more events closely associated with disease onset, including infection (72 percent), alcohol ingestion (23 percent), sleep deprivation (22 percent), unusual stress (20 percent), and physical exertion (19 percent) [5].
Other series have also reported trivial infection, alcohol consumption, and head trauma as common precipitating factors [4,6,20-23]. In a cohort of 30 patients with KLS in Taiwan, both KLS onset and recurrent attacks correlated with epidemic upper respiratory tract infections [7]. No specific infectious agents have been consistently identified.
KLS onset has also been reported after vaccinations, including typhoid and tetanus [24], tuberculosis [5], human papilloma virus [8], and H1N1 influenza [8].
●Infectious and immune-mediated theories – An infectious or immune-mediated etiology has been hypothesized based on frequent reports of infectious-like symptoms at disease onset.
Focal viral encephalitis or postinfectious autoimmune encephalitis affecting specific areas of the brain (ie, thalamus-hypothalamus, frontal and temporal lobes) would be consistent with the nature of the disease. These structures have a pivotal role in the regulation of sleep, perception, mood, and impulse control, and KLS-like symptoms can occur in association with hypothalamic, frontal, and temporal lobe tumors [25,26]. However, evidence for such pathology in the brains of KLS patients remains elusive. (See 'Neuroimaging' below.)
In most cases of apparently infection-triggered KLS, symptoms of KLS occur between three and five days after the onset of fever. Agents responsible for the infection are rarely identified and highly variable, however, and the interval to episode onset is too short for most immune-mediated processes. Specific infections reported in association with KLS include Epstein-Barr virus, varicella zoster, herpes zoster, influenza A and B, adenovirus, enterovirus, H1N1, Salmonella typhi, and Streptococcus [4,5,7,8,24,27-31]. The heterogeneity of identified agents and lack of measurable viral titers in most patients suggest no direct effect [4].
In support of autoimmunity as a possible etiology, the human leukocyte antigen (HLA) subtype DQB1*02 has been associated with the disease in 30 European patients [3], a finding not replicated in larger populations when controlling for the exact haplotype [5,7,19].
Notably, serum cytokine levels in 17 patients with KLS were similar during and between episodes, suggesting that KLS episodes are not accompanied by an abnormal systemic immune reaction [32].
●Associated conditions – A wide variety of genetic disorders have been reported with increased frequency in patients with KLS, including Klinefelter syndrome, von Willebrand syndrome, polycystic kidneys, and autism (both high and low functioning) (table 1) [5,6,33,34]. An abnormal birth history and early developmental abnormalities are reported in 35 to 45 percent of patients [5,8].
●Neuropathologic findings – Neuropathologic studies of KLS are limited to a small number of case reports. In two cases of primary KLS, the following findings were described:
•Severe inflammation of the thalamus and mild inflammation of the hypothalamus were described in a 46-year-old man who died from aspiration pneumonia during an attack of megaphagia [35].
•A 17-year-old male with a classic form of KLS and autonomic dysfunction during attacks died from cardiopulmonary arrest during an attack [36]. The neuropathologic examination revealed mild depigmentation of the substantia nigra and locus coeruleus. The cortex, thalamus, and hypothalamus were normal.
In two other reported cases, symptoms of KLS appeared to be secondary to a paraneoplastic syndrome, with abnormalities in the hypothalamus, amygdala, and gray matter of the temporal lobes [37], or inflammatory encephalitis within the thalamus and hypothalamus [38].
CLINICAL FEATURES — Kleine-Levin syndrome (KLS) is characterized by recurrent episodes of hypersomnia, confusion, derealization, apathy, and decreased mood.
Clinical presentation — The median age of onset of KLS is 16 years [5]. Approximately 80 percent of patients present during the second decade of life [4-6], with a range of 4 to 82 years [39,40]. Patients with late onset (>35 years) usually have typical symptoms but tend to have longer episodes [24,35,40-44], while childhood onset (<12 years) is associated with more frequent episodes [4,5,45-48]. In one large series, girls were older at onset than boys (18.5 versus 15.0 years), despite an earlier onset of puberty (11.8 versus 13.2 years) [5].
Symptom severity during each episode tends to progress rapidly, within hours to days, and symptoms last for a median of 10 days [4,5,9,49]. There is a broad range, however, and episodes can last for a few days in mild cases to several weeks or even months in more severe cases. Milder cases of KLS have one to three short episodes per year, but moderate cases can have monthly episodes of 7 to 10 days each [4,5,27].
Prolonged episodes (eg, four weeks or longer) have been described in approximately one-third of patients and are associated with higher levels of anxiety, agitation, and amnesia compared with shorter episodes [8]. In severe cases, patients can experience 40 to 80 episodes in a rapid succession, with a possible permanent impairment of attention and mood [4].
Cardinal symptoms — Episodic hypersomnia is a defining feature of KLS. Additional symptoms present in most patients during episodes of hypersomnia include cognitive disturbance, an altered perceptive state, referred to as derealization, and severe apathy. Although hyperphagia and hypersexuality have traditionally been considered hallmark clinical features, these symptoms are more variable across patients and episodes than previously thought and are better viewed as expressions of general disinhibition.
Hypersomnia — Hypersomnia is a major clinical symptom of KLS and is mandatory for the diagnosis. Hypersomnia refers to an increase in actual sleep time and is distinct from symptoms of fatigue or low energy. Sleep is enormously increased during episodes, with a median total sleep time of 18 hours per day [4,5], sometimes with loss of the normal circadian rhythm of sleep [9].
Most patients are difficult to awaken and report intense dreaming and hypnagogic hallucinations; sleep paralysis is uncommon. The need for sleep is so intense that patients may be found sleeping in the hallway outside a classroom or on the sidewalk [50-52]. Patients remain arousable, waking up spontaneously to void and eat, but are usually irritable or aggressive when awakened or prevented from sleeping.
There is frequently a brief overshoot of insomnia at the end of an episode. This insomnia is sometimes associated with euphoria [5,41,53,54]. As an example, one patient described the feeling as follows: "All I want to do is stay awake and get things done that I was unable to while in the episode. Plus, it is as if I have been reenergized." The need for sleep and total sleep time tend to diminish during later episodes [8].
Cognitive disturbance — Cognitive disturbance is the second most common symptom after hypersomnia. The large majority of patients have difficulty reading and speaking during episodes. Abnormal, effortful speech has been described in more than two-thirds of patients [4,5]. Symptoms include mutism, absence of spontaneous speech, monosyllabic or short sentences, limited vocabulary, and slurred, muddled, incoherent, or childish stereotypical language. Patients may also have slowness of speech and comprehension, verbal perseveration, and echolalia.
Memory disturbances are common, and the majority of patients report incomplete recollection of episodes. One patient described this symptom as follows: "It seems like a movie that is cut up, so scenes are missing." Eye-hand coordination and simple gestures, such as finding one's key in a pocket, may be difficult and perceived abnormally (astereognosia). Clumsiness resulting in broken limbs has been described. Temporal disorientation is twice as frequent as spatial disorientation. Altered perception is always present and can affect all senses: "Things seemed hazy, foggy"; "I would break a cup to see if it would break to reassure me things were normal."
Formal cognitive and memory tests have been rarely performed during a KLS episode, and their validity is questionable as patients can be uncooperative, irritable, sleepy, and inattentive [55-57].
In between episodes, most patients are described as completely normal. With formal testing, however, patients with KLS score lower than age-matched controls on nonverbal intelligence quotient (IQ), processing speed, attention, and episodic memory retrieval [24,26,55,58-61]. In one study, episodic memory performance declined after a mean follow-up of 1.7 years, despite a decrease in yearly episode frequency [61].
Derealization — During episodes, sensations of unreality are common and are probably the most specific symptom of the syndrome [5]. Surroundings are seemingly wrong, distorted, or unreal. Perception of sight, sound, smell, taste, temperature, and pain can all feel abnormal or unreal. Objects and voices are perceived to be distant, and one patient noted that his own voice appeared strange to himself [46,62]. The symptom has been described by patients with KLS in various ways [9,63,64]:
●"Unpleasant perception, bizarre and wrong"
●"A nightmarish sense of the surroundings"
●"The feeling of being almost in a dream"
●"Feelings of depersonalization, anguish, and a belief of a split between mind and body"
In one cross-sectional series, derealization was both sensitive (100 percent of patients reported it) and specific, as it was not reported in patients with other sleep or psychiatric diseases [5].
Apathy — Apathy is one of the major and most frequent symptoms during KLS episodes, reported by more than 90 percent of patients, often to the point of frank abulia [8]. Patients and families report the absence of self-motivation for almost all activities, including reading, watching television, using cell phones, and social communication of any type. Approximately 80 percent of patients neglect their hygiene during episodes.
Change in eating behaviors — Most patients report abnormal eating during at least one episode. Approximately two-thirds of patients have megaphagia (eating large amounts of food at each meal).
The disinhibited eating is often directed towards specific tastes (eg, sweet, salty, or sour), even if these are atypical food choices for the individual. Other patients would eat any and all food before them, behavior that is reminiscent of frontal lobe syndromes. There are reports of patients who stole food in shops or off the plates of other patients [52,65,66], who searched for food in waste baskets [52], or who stuffed food in their mouths with both hands [67].
Hyperphagia and inactivity often leads to striking weight gain: 4.6 kg/episode in one large study [5] and 3.2 to 13.6 kg/episode in another [4]. In one patient’s words, "I would not eat for 10 to 14 hours, then would eat four or five peanut butter and jelly sandwiches—without even chewing—very fast, then would start to fall asleep again with food still in my mouth."
Notably, however, up to one-third of patients have decreased appetite or eat less during some episodes [5,7,15,44,68,69]. These patients usually have a more severe hypersomnia, compared with those with hyperphagia.
The eating habits of patients with KLS are distinct from those of patients with bulimia in that there are no periods with self-induced vomiting or other attempts to control weight.
Hypersexuality — Approximately half of males and one-third of females have hypersexuality during at least one episode [5,7]. Hypersexuality occasionally occurs in prepubescent children [46,48,70].
Symptoms include increased masturbation and inappropriate sexual behaviors such as exposing or touching genitals, masturbating in the presence of parents and clinicians, use of obscene language, or touching other people inappropriately [42]. Although hypersexuality is often emphasized as a key symptom of KLS, it likely represents just one aspect of disinhibition that is characteristic of KLS episodes. A less specific behavioral disinhibition is more common [5,7].
Other symptoms during KLS episodes
●Mood disturbances – Approximately half of patients with KLS have depressed mood during episodes, a finding significantly more frequent in women [4,5]. Suicidal thoughts have been reported in up to 15 percent of patients, often expressed as a desire to bring an end to the episode. Rare suicide attempts have been reported [41,71].
Patients are often anxious, scared of being left alone or of answering phone calls, or afraid of novelty in their environment. They usually refuse to visit friends or families or to receive visitors, mostly saying that they do not want to be seen in this state.
Irritability is present in almost all patients, especially when sleep or food intake is prohibited. Aggressive behavior, such as biting, spitting, throwing stones, or tantrums at school, is rare [44,50,52,64,72]. Some patients exhibit regressive behaviors such as using childish words and voice or asking a parent to sleep nearby [5,73].
In most cases, mood disturbances resolve by the end of the episode, although in rare cases they persist longer. A period of hypomania is not uncommon at the conclusion of an episode, extending over a day or two [74,75].
●Visual and auditory hallucinations – Visual and auditory hallucinations and paranoid delusions are occasionally reported. Examples include seeing "scary snakes on TV," "distorted faces, including Jesus," or "the dead bodies of parents" [67]. One teenage boy feared that people were trying to kill him, heard threatening voices, and took his razor to protect himself [50]. Another patient believed that his wife was trying to poison him and the neighbors were trying to steal from him; he changed his lock five times and stopped eating for fear of poisoning [35]. Such symptoms can lead to a misdiagnosis of schizophrenia. (See 'Differential diagnosis' below.)
●Other compulsive behaviors – In addition to hypersexuality and hyperphagia, various other compulsions are reported in approximately one-third of patients. Examples include inappropriate singing, tapping, clapping, body rocking, chewing lips, compulsive writing on walls, stripping down wallpaper, continuously switching lights on and off, pacing, wringing hands, tearing out hair, and even arson [18,50,56,63,67,76-81]. Some patients repeatedly listen to music or watch the same videos over and over again [5,7,56,78].
●Autonomic symptoms – In a series of 108 patients questioned retrospectively about autonomic symptoms during an episode, 46 percent recalled sweating, 24 percent hot flashes, and 18 percent nausea [5]. In other studies, there have been rare reports of flushed face [53], thermoregulatory changes [24], hyperventilation [82], short episodes of flushes, profuse sweating, excessive salivation, hypertension and tachycardia [43], and hypotension and bradycardia [36,44,78,83].
●Meningeal symptoms – Fever, photophobia, and headache were endorsed by half of patients in one retrospective series, but their significance is unclear since no signs of meningeal irritation or changes in cerebrospinal fluid (CSF) cell counts or protein have been reported [5].
Menstrual recurrent hypersomnia — Menstrual recurrent hypersomnia (MRH) is a rare variant of recurrent hypersomnia [1], identified in less than 20 women so far [6]. A pathophysiologic association between MRH and KLS is supported by report of a family with a daughter with MRH and a son with KLS [13]. Moreover, some girls with typical KLS have most of their episodes (but not all) in association with menstruation, making the differentiation between typical KLS and MRH difficult [24,84]. As a consequence, the third edition of the International Classification of Sleep Disorders (ICSD-3) considers MRH as a variant of KLS [1].
Episodes of hypersomnia occur just before or during menstruation and are associated with compulsive eating (65 percent), depressed mood (35 percent), and sexual disinhibition (29 percent) [6]. Compared with typical KLS, episodes of menstrual recurrent hypersomnia are usually shorter (3 to 15 days), less frequently involve cognitive changes, and sometimes respond to treatment with estrogen-containing contraceptives [6,85,86].
Features in between episodes — Individuals with KLS typically return to a normal baseline between episodes.
In a retrospective case-control study of 108 KLS patients and 108 matched controls, sleep parameters, eating, and psychological symptoms were virtually indistinguishable between groups [5]. Habitual sleep and wake times were similar in cases and controls. Patients with KLS were slightly more anxious than controls. There was no difference in mean depression scores or a family history of depression between groups.
In a prospective case-control series, patients with KLS slept longer and at an earlier time than controls, suggesting a mild persistence of prolonged sleep in patients during asymptomatic periods [8].
Physical characteristics may differ slightly, however. In the same series, patients with KLS had a higher body mass index (BMI), increased leptin and C-reactive protein (CRP) levels, and an increased prevalence of witnessed obstructive sleep apnea, compared with controls [5]. Of interest, BMI did not differ between patients with and without megaphagia, suggesting that the increased BMI could be a result of decreased activity during episodes [4,5]. Other studies have found no difference in BMI between patients with KLS and controls [8].
Cerebrospinal fluid — During and between episodes, the CSF in patients with KLS is normal. This includes both basic CSF parameters (eg, cell counts, glucose, total protein) as well as immunoelectrophoresis for oligoclonal bands [48,50,87,88]. No consistent alterations in neurotransmitter levels have been identified [35,36,47,89,90].
CSF levels of hypocretin-1, a hypothalamic peptide that has been shown to be deficient in narcolepsy type 1 (narcolepsy with cataplexy), are typically within normal range (above 110 pmol/L). Some studies have found that CSF hypocretin-1 levels are lower in patients with KLS compared with controls or lower during episodes than in between episodes [9,91-94], suggesting a possible role for hypocretin-1 in the pathogenesis of KLS.
Electroencephalography — Electroencephalogram (EEG) findings in patients with KLS are nonspecific. There is no evidence of seizure activity during episodes, even when sphenoid electrodes are placed. In 70 percent of patients who undergo EEG, a nonspecific, diffuse slowing (7 to 8 Hz) of background EEG activity is observed [4]. Less often, low-frequency, high-amplitude waves (delta or theta) are observed, mainly in bilateral temporal or frontotemporal areas.
Polysomnography — Polysomnography findings during an episode of hypersomnia are variable and may be affected by the duration of monitoring (ie, 12 versus 24 hours) and the timing of the study in relation to the episode and disease onset [2-4,95]. Total sleep time is generally increased, ranging from approximately 9 to 12 hours, but sleep structure is essentially normal [2,3].
Less consistent abnormalities that have been described in one or more studies include a mild reduction in slow-wave sleep during the first half of an episode, mild reduction in sleep efficiency due to frequent awakenings during non-rapid eye movement (NREM) stages 1 to 2, and reduced NREM stage 3 and 4 duration.
In one study that included 19 patients with KLS, approximately 40 percent of patients demonstrated a narcolepsy-like pattern, with two or more sleep-onset REM periods during multiple sleep latency test (MSLT) [95]. Reduced sleep latency is not a universal finding, however. In other cases, the duration of sleep stages during daytime sleep contains an excess of NREM stages 1 to 2 [4]. Rare sleep abnormalities, such as an absence of REM sleep [96], an absence of slow-wave sleep [74], and recurrent interruption of stage 2 by fear and stereotypical movements [96], have also been described.
Of note, polysomnographic findings may vary according to the stage of the disease. In one study, sleep studies obtained later in the disease course demonstrated that patients spent three to nine hours per day in a state of withdrawal with their eyes closed and an EEG pattern that resembled their awake state [97].
In between episodes, polysomnography in patients with KLS is generally normal.
Neuroimaging — Structural brain computed tomography (CT) and magnetic resonance imaging (MRI) studies are normal in patient with KLS [2-4,98]. Although various minor abnormalities have been reported (eg, frontal cysts, local atrophy), they are likely incidental findings [60].
By contrast, functional imaging studies are frequently abnormal in patients with KLS both during and between episodes. Case reports and small case series have described a variety of abnormalities on single-photon emission computed tomography (SPECT), positron emission tomography (PET), and functional MRI [7,55,69,98-101]. The most consistent finding in studies using visual or semiquantitative analysis is reduced temporal/frontotemporal lobe perfusion, found in approximately half of patients; the thalamus, parietotemporal junction, and association cortex are also commonly involved.
Larger studies with a control group have begun to explore correlations between perfusion changes and disease severity or symptom type. In a study that used whole brain voxel-based group analysis, SPECT was compared during (n = 11 patients) and in between (n = 41 patients) episodes in 41 KLS patients and 15 healthy control subjects. [102]. Key findings in patients with KLS included hypothalamic, thalamic, caudate nucleus, and associative area (anterior cingulate, orbitofrontal and right superior temporal cortex) hypoperfusion during asymptomatic periods. Additional hypoperfusion in the right dorsomedial prefrontal cortex and right parietotemporal junction (extending to the angular gyrus) occurred during symptomatic periods. The intensity of derealization strongly correlated with hypoperfusion of the right and left parietotemporal junctions.
Persistent hypoperfusion in the diencephalic and association cortical area during asymptomatic periods may be a marker of complex thalamocortical circuit dysfunction, compensated by supplemental activation of other areas [103]. The altered brain activation pattern in KLS patients has been accompanied by lower performance levels and longer reaction times [57,103-105]. In a study of working memory, KLS patients performed worse than controls, but the patients with thalamic hyperactivity had better memory performance, suggesting a compensatory mechanism in those subjects with high working memory capacity [105]. In another study, magnetic resonance spectroscopy (MRS) demonstrated low concentrations of N-acetylaspartate (NAA) in the left thalamus in patients with thalamic hyperactivation during a working memory task, also supporting the hypothesis of compensatory activation [57].
EVALUATION — Kleine-Levin syndrome (KLS) is a rare disorder that should be suspected in adolescents, particularly males, who present with episodic hypersomnia, cognitive disturbance, and behavioral changes with a return to baseline between attacks (see 'Clinical features' above). Because of the rarity of the syndrome and the nonspecific nature of some of the symptoms, all patients with suspected KLS should undergo intensive investigation for other potential etiologies.
The evaluation should include a comprehensive medical history from both the patient and family members for verification, since patients are commonly amnesic for episodes. Particular attention should be given to cognitive and behavioral changes associated with the episodes, provocative factors, prescribed medications, herbals, supplements, and drugs of abuse.
The physical examination should include detailed neurologic and psychiatric examinations. The physical examination is normal in patients with KLS, and there are no specific neurologic findings. Although some patients report fever, headache, and photophobia, nuchal rigidity or focal neurologic signs have not been observed during or in between episodes [4], and the presence of such abnormalities should prompt consideration of an alternative diagnosis such as meningitis or encephalitis.
Brain imaging with and without contrast should be obtained as part of the initial evaluation to exclude alternative structural etiologies. Magnetic resonance imaging (MRI) is preferable to computed tomography (CT), based on the higher resolution and sensitivity of magnetic resonance imaging (MRI) for a variety of structural brain pathologies. An electroencephalogram (EEG) during an episode may be helpful if aspects of the history are suggestive of seizure, and the EEG may need to be repeated multiple times if concern persists and the initial recordings are normal. Polysomnography performed during an episode may show increased total sleep time but is otherwise nonspecific; the purpose is primarily to exclude alternative sleep disorders that may explain or contribute to excessive daytime sleepiness.
Additional laboratory testing should be tailored to the presenting symptoms and the differential diagnosis; there are no characteristic cerebrospinal fluid (CSF), blood, or urine abnormalities in patients with KLS. Tests to consider include lumbar puncture for CSF analysis; blood for complete blood count, blood urea nitrogen, creatinine, liver function tests, ammonia, lactic and pyruvic acids, pH, amino acids, acylcarnitine, toxicology screen, lead level, vitamin B12, folate, thyroid-stimulating hormone, and antithyroglobulin; and urine for toxicology screen, organic acids, and porphobilinogen. Serologic markers for specific pathogens (eg, Lyme) should be considered in selected cases. (See 'Differential diagnosis' below.)
DIAGNOSIS — The diagnosis of Kleine-Levin syndrome (KLS) is made based on the clinical features of episodic hypersomnia with typical cognitive and behavioral disturbances and exclusion of alternative psychiatric, neurologic, and toxic or metabolic etiologies.
According to the third edition of the International Classification of Sleep Disorders (ICSD-3), diagnostic criteria for KLS include all of the following (table 2) [1]:
●At least two episodes of excessive sleepiness and sleep duration, each lasting for two days to five weeks.
●Episodes recur at least once every 18 months and usually more than once a year.
●Normal alertness, cognitive function, behavior, and mood between episodes.
●At least one of the following during episodes: cognitive dysfunction, altered perception, eating disorder (anorexia or hyperphagia), or disinhibited behavior (such as hypersexuality).
●Hypersomnolence and related symptoms are not better explained by another sleep disorder; other medical, neurologic, or psychiatric disorder (especially bipolar disorder); or use of drugs or medications.
Menstrual recurrent hypersomnia (MRH) is classified as a clinical subtype of KLS in which episodes are exclusively associated with menstruation [1]. (See 'Menstrual recurrent hypersomnia' above.)
DIFFERENTIAL DIAGNOSIS — The differential diagnosis of Kleine-Levin syndrome (KLS) includes a variety of psychiatric, neurologic, and sleep disorders, most of which are much more common than KLS. Most can be distinguished by clinical features, neuroimaging, polysomnography, and lumbar puncture in selected cases.
●Psychiatric disorders — Psychiatric disorders are the major differential diagnosis for KLS. Specific disorders that should be considered include depression, bipolar disorder, psychogenic/somatic hypersomnia, and acute psychosis. Features that may help distinguish KLS from psychiatric disorders include the following:
•KLS patients generally do not report psychiatric symptoms between episodes and do not have significant family history for these disorders [5]. This contrasts with bipolar disorder, depression, or schizophrenia, in which symptoms are persistent and a family history is often present.
•Compared with psychiatric disorders, the onset and offset of symptoms in KLS are much more abrupt (few hours to one to two days), and the length of episodes is shorter and more stereotyped.
•Although hypersomnia may be reported by patients with depression, the symptom is more often increased time in bed (clinophilia) with depressed mentation rather than true hypersomnia (increased sleep). Cognitive disturbances when awake, as described in KLS, are also a distinguishing feature from depression. (See 'Cognitive disturbance' above.)
●Neurologic disorders — In the midst of an episode, particularly if it is the first, numerous neurologic disorders may be considered on the differential diagnosis.
One or more of the cardinal symptoms of KLS could be caused by structural pathology in the hypothalamus, frontal and/or temporal lobes, such as tumors (eg, craniopharyngioma, glioma, pinealoma, meningioma, colloid cyst), stroke, or inflammation/demyelination (eg, infectious, autoimmune, or paraneoplastic encephalitis, acute disseminated encephalomyelitis, multiple sclerosis). In contrast with KLS, however, neurologic symptoms caused by all of these processes are not typically episodic, and patients often have additional findings on neurologic exam. Cystic tumors or colloid cysts within the third ventricle, which may cause episodic symptoms due to a ball-valve mechanisms of obstruction, can be distinguished from KLS due to additional symptoms of increased intracranial pressure (headache, nausea, vomiting).
The episodic nature of the symptoms in KLS may raise the possibility of epilepsy, particularly temporal lobe epilepsy or nonconvulsive status epilepticus. However, cardinal features of KLS, including prolonged periods (days) of derealization, hypersomnia, hyperphagia, and hypersexuality, are not typical for epilepsy. An electroencephalogram (EEG) performed during an episode can often distinguish between KLS and epilepsy if the two cannot be distinguished clinically.
Patients with Lewy Body dementia may have fluctuations of alertness associated with behavioral disturbances, but they are older than KLS patients, have permanent cognitive deficits, and often have associated Parkinsonism. (See "Clinical features and diagnosis of dementia with Lewy bodies".)
●Other sleep disorders — Excessive daytime sleepiness is also a major symptom of obstructive sleep apnea, delayed sleep phase syndrome, narcolepsy, and idiopathic hypersomnia. However, the clinical picture in all of these entities is neither episodic nor characterized by such severe cognitive or behavioral problems. Sleep time may mildly fluctuate in patients with these syndromes, however; hence, polysomnography is generally indicated in those with residual sleepiness during asymptomatic periods to exclude an alternative sleep disorder. (See "Approach to the patient with excessive daytime sleepiness", section on 'Initial evaluation'.)
●Metabolic encephalopathy or intoxication — Less common causes of recurrent hypersomnia are benzodiazepine intoxication, drug and alcohol overdoses or abuse, and recurrent metabolic encephalopathies due to hyperammonemia, aminoaciduria, or organic acidemia. A metabolic encephalopathy should be suspected in young children with developmental regression, failure to thrive, recurrent vomiting, and a positive family history of developmental delay or parental consanguinity. Other signs and symptoms of increased intracranial pressure and acidosis (eg, papilledema, pyramidal signs, Kussmaul breathing) are also expected in metabolic encephalopathies and should not be present in patients with KLS.
NATURAL HISTORY — Episodes of Kleine-Levin syndrome (KLS) tend to decrease in frequency and intensity over time, especially prior to disease termination.
Disease duration from the first to the last episode is extremely variable, with a spectrum ranging from 6 months to 41 years [4]. In a series of 110 patients with KLS, the median disease duration was eight years [4]. In another survey of 108 new KLS cases, the mean disease duration was 14 years, and approximately 15 percent of patients had a disease course longer than 25 years [5]. Risk factors for a longer disease course were male gender, hypersexuality, and onset after age 20 years.
In between episodes and at the end of the disease course, patients with KLS typically return to a normal baseline. However, functional imaging studies and long-term neuropsychological follow-up in a growing number of patients suggest that there may be mild, long-lasting memory and attention defects [4,58-60,103-108]. (See 'Cognitive disturbance' above and 'Neuroimaging' above.)
Patients are at increased risk for the development of psychiatric disorders, most commonly mood disorders (other than bipolar disorder) and anxiety. In a study of 115 patients with KLS, psychiatric disorders were diagnosed in 20 percent of patients over one to six years of follow-up [109]. Women and patients with repeated episodes are at highest risk. It is not known whether emergent psychiatric disorders are the consequence of psychiatric symptoms during episodes or a result of difficulty adjusting to KLS.
TREATMENT — The etiology of Kleine-Levin syndrome (KLS) is unknown. Treatment is therefore primarily supportive and educational. Lithium prophylaxis reduces the frequency and duration of episodes in patients with frequent episodes, and intravenous (IV) steroids may be therapeutic in patients with long (>30 day) episodes. A variety of other pharmacologic agents have been used in an attempt to treat symptoms during episodes, but success is generally limited. Modafinil and other stimulants may occasionally be helpful in severe cases.
Nonpharmacologic management — Disease management in patients with KLS is primarily supportive and educational. Most patients benefit from reassurance, maintenance of a simple hygiene routine, and management at home [110].
Our approach involves educating family members, caregivers, and occasionally teachers about the core symptoms of KLS and providing conditions so that patients will not injure themselves during attacks. The best course for most patients is to allow the patient to sleep in a safe and familiar environment, avoid driving, and be observed for medical and psychiatric complications. Alternative therapies, such as light therapy, melatonin, and vitamin supplements, have been tried with no success [4,5].
Patients are generally exhausted during attacks, and sleep is often the least harmful way to wait for the episode to end. In an effort to prevent new episodes, patients are advised to avoid alcohol intake, to use good hand hygiene and avoid direct contact with individuals who are infectious if possible, and to sleep with regular schedules.
Symptom-based pharmacotherapy — Pharmacologic therapies that may benefit some patients in the midst of a KLS episode are stimulants and nonamphetamine wakefulness promoting agents. Antipsychotic drugs have also been explored for cognitive and behavioral disturbances. Antidepressants seem to have no effect on symptom severity or episode frequency.
In the absence of controlled trials and prospective studies, supportive evidence for various treatments is based on case reports and retrospective, uncontrolled series [4-6,27,111-122]. For most patients, we do NOT suggest pharmacotherapy during episodes of hypersomnia.
A summary of subjective patient responses to a variety of symptomatic drug trials compiled from two observational series is provided in the table (table 3). Due to the unpredictable course of the disease and the paucity of consistent positive results, reports should be interpreted cautiously. Selection of drug therapy in patients with KLS should be individualized, taking into account the most prominent and disabling symptom, risk of side effects and cost, and patient preferences.
●In patients with prolonged episodes (>30 days), we suggest a trial of high-dose glucocorticoids. (See 'Patients with prolonged episodes (>30 days)' below.)
●Modafinil, armodafinil, or methylphenidate can be tried symptomatically during an episode, keeping in mind that KLS symptoms extend beyond sleepiness, and that stimulants have no effect on derealization and apathy, which are much more disabling. In one large series, only 20 percent of patients reported some benefit from a stimulant used during an episode [4].
In our experience, many patients feel better by sleeping rather than by being artificially awakened by the stimulant, only to experience the other symptoms. If a stimulant is used, patients should be monitored for potential toxicities and unintended side effects such as exacerbation of hypersexuality, aggression, or hallucinations [123].
●For patients with psychotic symptoms that are sustained and prominent during episodes, an atypical antipsychotic such as risperidone may provide symptomatic benefit. Anxiety can be intense during some episodes; hydroxyzine or a benzodiazepine can be used as needed in these cases.
●Amantadine, a drug with dopaminergic, stimulant, and antiviral properties, was reported to help terminate episodes or improve cognitive symptoms in 10 of 24 patients (41 percent) when administered on day 1 of an episode in one series [4]. In our experience with amantadine in more than 100 patients with KLS, however, this effect could not be replicated [8].
Other therapies that do not appear to provide benefit in patients with KLS include antidepressants, electroconvulsive therapy, and insulin coma therapy [41,42,67,71,77,124]. In two patients, flumazenil, a benzodiazepine receptor antagonist, failed to elicit wakefulness [4]. In a case report, sodium oxybate, usually used for excessive daytime sleepiness in narcolepsy, was associated with clinical improvement in KLS but without a challenge of cessation and readministration [116].
Patients with prolonged episodes (>30 days) — Up to one-third of patients have prolonged episodes (>30 days) that are disabling and tend to be associated with higher degrees of anxiety and agitation than shorter episodes [8]. Based on experience of the authors, one option to consider in these patients is a trial of high-dose glucocorticoids at the time of the next episode. We use IV methylprednisolone 1 g/day for three consecutive days, beginning within 10 days of the onset of symptoms. If glucocorticoid treatment is successful in shortening the duration or severity of the episode compared with prior episodes, we repeat with each future episode.
The potential benefit of glucocorticoids is supported by a retrospective study in which 26 patients with long episodes (>30 days) received IV methylprednisolone 1 g/day for three days during one to six episodes [125]. When each patient's preceding untreated episode duration was used as a comparison, 42 percent of patients experienced an improvement in episode duration (by seven days or more) with steroid treatment. Receipt of steroids within the first 10 days of the onset of symptoms was associated with a greater likelihood of response (60 percent). Among a cohort of 48 contemporaneous patients with KLS who did not receive glucocorticoids and who were matched with treated patients by age, gender, disease duration, and number of episodes, the rate of spontaneous improvement of seven days or more from one episode to the next was only 10 percent. Steroids were well tolerated aside from transient insomnia and did not induce mania.
Future studies should help to address some of the limitations to the existing evidence, including the retrospective nature of the study, lack of blinding, and lack of an objective biomarker to measure disease activity. It is not known whether oral steroids have a similar effect.
Although short-term high-dose IV methylprednisolone is relatively safe and associated with few side effects in most patients, potential complications include hyperglycemia, increased susceptibility to infection, bone loss and fractures, and neuropsychiatric side effects. (See "Major side effects of systemic glucocorticoids".)
Prevention of episodes — There are no established therapies for episode prophylaxis in patients with KLS, but a variety of drugs with potential mood stabilizing effects (eg, lithium, antiepileptic drugs, other miscellaneous agents) have been explored. In our experience, lithium is the most likely to be beneficial in patients with frequent episodes (eg, 4 to 12 episodes per year). Use of lithium for KLS is endorsed by the American Academy of Sleep Medicine (AASM) guidelines based on very low quality evidence [126].
In two large case series that together included 59 patients treated with lithium, 22 percent reported at least mild improvement in episode frequency, and 24 percent reported clear benefit [4,5]. In a larger single-center observational study of 130 patients with KLS, 37 percent of patients treated with lithium (26 out of 71) experienced complete resolution of episodes over a mean follow-up of 22 months, compared with 3 percent of contemporaneous patients who were not treated with lithium [127]. The time spent incapacitated was one month lower per year in treated versus untreated patients. One or more side effects of lithium were reported by 50 percent of patients but were generally mild and led to treatment discontinuation in only 7 percent. The most common side effects were tremor, increased drinking, diarrhea, and subclinical hypothyroidism (11 percent; only two patients required levothyroxine).
In various series, the effective dose range of lithium is 600 to 2000 mg at bed time (to achieve serum levels of 0.8 to 1.2 mmol/L). The required concentrations are in the upper recommended range for bipolar disorder. Higher doses of lithium can have serious side effects (eg, thyroid and renal toxicities), and the risk to benefit ratio should be reassessed every six months. In some reports, episodes have recurred after two days of missed doses of lithium, highlighting the potential importance of adherence to therapy [110,127]. The pharmacology, administration, and side effects of lithium are reviewed in more detail elsewhere. (See "Bipolar disorder in adults and lithium: Pharmacology, administration, and management of adverse effects".)
More limited and less consistent results have been described with valproate and carbamazepine [4,5].
SUMMARY AND RECOMMENDATIONS
●Kleine-Levin syndrome (KLS) is a rare disease, characterized by relapsing-remitting episodes of hypersomnia and cognitive disturbances, including apathy and derealization. KLS is categorized as a central disorder of hypersomnolence; other disorders in this category include narcolepsy type 1 and type 2 and idiopathic hypersomnia. (See "Classification of sleep disorders".)
●The etiology of KLS is unknown. Contemporary hypotheses have focused primarily on genetic, inflammatory, and immune-mediated causes. (See 'Etiology and pathogenesis' above.)
●The median age of onset of KLS is 16 years, and approximately 80 percent of patients present within the second decade of life. KLS primarily affects adolescent males and has an unpredictable course of recurrence and remission lasting for years. Symptom severity during each episode tends to progress rapidly, within hours to days, and symptoms last for a median of 10 days. (See 'Clinical presentation' above and 'Natural history' above.)
●Episodic hypersomnia is a defining feature of KLS. Additional symptoms present in most patients during episodes of hypersomnia include cognitive disturbances, an altered perceptive state (derealization), and prominent apathy. Hyperphagia and hypersexuality are variably present. (See 'Cardinal symptoms' above.)
●Additional symptoms that may be reported during episodes include mood disturbances, visual and auditory hallucinations, and compulsive behaviors aside from hyperphagia and hypersexuality. (See 'Other symptoms during KLS episodes' above.)
●In between episodes, most patients with KLS return to a normal baseline and are generally indistinguishable from controls on measures of sleep, cognition, and mood. Over time, patients are at increased risk for mild neurocognitive impairment and emergent psychiatric disorders. (See 'Features in between episodes' above and 'Natural history' above.).
●KLS is a rare disorder that should be suspected in adolescents, particularly males, who present with episodic hypersomnia, cognitive disturbance, and behavioral changes with a return to baseline between attacks. Because of the rarity of the syndrome and the nonspecific nature of some of the symptoms, all patients with suspected KLS should undergo intensive investigation for other potential etiologies. (See 'Evaluation' above and 'Diagnosis' above.)
●The differential diagnosis of KLS includes a variety of psychiatric, neurologic, and sleep disorders, most of which are much more common than KLS. Most can be distinguished by clinical features, neuroimaging, electroencephalogram (EEG), polysomnography, and lumbar puncture in selected cases. (See 'Differential diagnosis' above.)
●Nonpharmacologic management, which includes establishing a safe sleeping environment, educating caregivers and teachers, and monitoring for medical and psychiatric complications, is the mainstay of therapy for all patients. (See 'Nonpharmacologic management' above.)
●Symptom-based pharmacotherapy may be warranted in severely affected patients, but results are often disappointing. For most patients with KLS, we suggest NOT treating with pharmacotherapy during episodes of hypersomnia (Grade 2C). The main pharmacologic therapies that have been explored in patients with KLS are stimulants and nonamphetamine wakefulness promoting agents. A trial of high-dose steroids is an option to consider in patients with prolonged episodes (>30 days). (See 'Symptom-based pharmacotherapy' above and 'Patients with prolonged episodes (>30 days)' above.)
●There are no established therapies for episode prophylaxis in patients with KLS, but a variety of drugs with potentially mood stabilizing effects (eg, lithium, antiepileptic drugs, other miscellaneous agents) have been explored. In our experience, lithium is the most likely to be beneficial and is suggested in patients with frequent episodes (eg, 4 to 12 per year) (Grade 2C). (See 'Prevention of episodes' above.)
●Episodes of KLS tend to decrease in frequency and intensity over time, especially prior to disease termination. Disease duration from the first to the last episode is extremely variable, with a spectrum ranging from 6 months to 41 years and a median of approximately 10 years. (See 'Natural history' above.)
ACKNOWLEDGMENT — The UpToDate editorial staff acknowledges Ruth S Shalev, MD, who contributed to an earlier version of this topic review.
