Narcolepsy & Hypersomnia
The main symptoms of narcolepsy type 1 (NT1) are excessive daytime sleepiness, cataplexy, sleep paralysis, hypnagogic hallucinations, and disturbed nocturnal sleep. It affects about 0.03% of the population (1 for 3,000 individuals) in most countries in the world. About half of all patients start narcolepsy before age 18, in rare cases as young as 3 years old, and in general the younger it starts, the more severe and abruptly the disease strikes; in young children one often observed a regression where tantrums and bad behavior restarts because the child is always exhausted. A rapid gain of weight often also occurs.
Since the 1960s it has been known that several of the disabling symptoms of narcolepsy, such as sleep paralysis, cataplexy and hypnagogic hallucinations, are pathological equivalents of REM sleep (a stage of sleep when we dream but are paralyzed to avoid moving in our dreams). Indeed, patients with narcolepsy enter REM sleep abnormally fast, minutes after falling asleep unlike normal people where REM sleep only appear after one hour of sleep (see History of Narcolepsy).
Cataplexy is unique to NT1. Cataplexies are sudden, brief episodes of muscle weakness triggered by emotions. Typically, the patient's knees buckle and may give way upon laughing, elation, surprise, or anger. In other typical cataplectic attacks, the head may drop or the jaw may become slack. In severe cases, the patient might fall and become completely paralyzed for a few seconds to several minutes. Reflexes are abolished during the attack. Although cataplexy is most often triggered by emotions such as a good joke or a funny cartoon, in children close to onset, it can be atypical and manifest by mouth opening, jaw dropping with tongue protrusions that are not always obviously triggered by emotions. Children can also feel generally weak, having trouble walking. Only after many months the cataplexy starts then to transform into the “adult” form only triggered by jokes or laughing.
Sleepiness is the most problematic symptom in narcolepsy. In general, patients feel exhausted all the time and would fall asleep as soon as they are not moving or being stimulated (for example as a passenger in a car). On top of this background of constant sleepiness, sudden sleep attacks can appear that are so strong that the patient cannot stay awake and struggles. If the patient can, he/she would take a nap and very often would feel better even after 15-30 min of sleep. Other times, patients would fight though sleep attacks and this can lead to “automatic behaviors”. During these events, patients continue the activity they initiated half asleep and have no recollection, for example continuing to write gibberish on a piece of paper.
In sleep paralysis, a frightening symptom considered to be an abnormal episode of REM sleep atonia, the patient suddenly finds himself unable to move for a few minutes, most often upon falling asleep or waking up. During hypnagogic hallucinations, patients experience dream-like auditory or visual hallucinations, while dozing (during sleep attacks) or falling asleep. These events, together with vivid dreams that often feel so real that it takes minutes after waking up to make sure it did not happen, perturb sleep in patients with narcolepsy. Sleep is often perceived as exhausting and not restful. As the disease progress, patients start to sleep less over the entire day and night, but this improvement during the day becomes associated with dreadful episodes of time when at night the patient is wide awake and unable to sleep. Nocturnal sleep becomes disturbed not only by vivid dreams and sleep paralysis, but by intractable bouts of insomnia inthe middle of the night.
Many patients with NT1 also develop obesity. The reason for this is partly behavioral, as patients are exhausted and stop moving or exercising, and likely partially metabolic, as the burning of calories becomes slower. Obesity also leads to sleep apnea, and some patients as a result are misdiagnosed. This is particularly problematic in children, when the onset of the disease is abrupt.
Thanks to research conducted by our lab in the 1990-2000, we now know the cause of NT1. Much of the research that led to these discoveries was made thanks to dogs, as some dogs have a genetic form of narcolepsy, unlike humans were the disease is more often sporadic and rarely familial (see History of Narcolepsy). All the symptoms of type 1 narcolepsy are due to the loss of about 20,000 neurons (brain cells) producing a peptide chemical called hypocretin or orexin. The cause of the loss of hypocretin neurons is autoimmune, which means that the immune system, normally involved in fighting infections, misdirects its efforts and attacks the hypocretin neurons thinking they are infected or foreign. In many ways, narcolepsy is very similar to type 1 diabetes, but instead of the immune system attacking pancreas islet cells secreting insulin (type 1 diabetes patients need insulin treatment all their live to regulate their glucose), in type 1 narcolepsy it is the hypocretin/orexin cells that are attacked.
Another particularity of NT1 is that the abnormal immune response seems to be often triggered by flu infections (sometimes without symptoms) and strep throats. In fact, in narcolepsy, the immune response to some flu strains becomes too strong and abnormally cross react with hypocretin confusing the hypocretin molecule with a piece of the flu. Once it is started, the immune system only stops attacking the hypocretin cells when they are 95% destroyed. At this point the patient has all the symptoms of narcolepsy. This process is however rare, and it is fair to say that developing narcolepsy is the result of a lot of bad luck, having a certain genetic predisposition that many others have, and getting certain types of infections and flu at exactly the wrong time in your life. Because of this, patients with narcolepsy have only a 1-2% chance to have a child with narcolepsy, and even identical twins raised in the same family with similar infections only both have narcolepsy 25% of the time.
Once the cells that secrete hypocretin/orexin are killed, recovery is impossible, and the disease is life-long. Effective treatments are however available and are life changing, so that many patients do well if treated aggressively and early with the right medication and if given the right advice.
Narcolepsy can be diagnosed using specific medical procedures: the diagnosis of narcolepsy is usually easy if all the symptoms of the illness are present. More often, however, the symptoms of dissociated REM sleep such as cataplexy are mild, and a nocturnal polysomnogram (PSG), followed by the multiple sleep latency test (MSLT), a test where patients are asked to nap 4-5 times during the day, is suggested. This test, performed at a sleep disorders clinic, will confirm the daytime sleepiness by showing that patients fall asleep quickly, typically in less than 8 minutes as a mean, and go abnormally fast into REM period in multiple naps (“SOREMPs”). Other causes of daytime sleepiness, such as sleep apnea or periodic leg movements, are also excluded by the nocturnal recordings. The MSLT is good but not perfect, it can be negative in about 8% of true patients and can be positive in about 4% of controls.
In some cases, and more frequently in Europe, to make sure that patients have type 1 narcolepsy, hypocretin is directly measured but this necessitates a lumbar puncture as hypocretin/orexin is primarily produced in the brain. Showing that hypocretin/orexin is very low or absent in the CSF is the best test and the gold standard to make sure a patient has narcolepsy type 1. It is generally done after a genetic test called HLA testing is done and has showed that the patient has a HLA marker called DQB1*06:02, the reason being that 98% of type 1 narcolepsy versus 25% of controls are positive for this immune gene variant. The HLA testing is only useful to exclude type 1 narcolepsy as about 25% of the general population have the gene.
Many patients do not have all the symptoms of narcolepsy, for example, they can be tired and need to take multiple naps without having cataplexy. Some need 12 hours of sleep every day and still feel exhausted. Other are plagued by vivid dreams and sleep paralysis but never experienced cataplexy. Typically, in these cases, an MSLT is done and if the pattern looks like NT1 (mean sleep latency ≤ 8 min; ≥ 2 SOREMPs in 5 naps), the patient is called type 2 narcolepsy (NT2, or narcolepsy without cataplexy). If the MSLT shows sleepiness with a short mean sleep latency (≤ 8 minutes) but does not show REM sleep in more than 1 nap, or reports sleeping every day more than 10 hours while still feeling tired, the patient is diagnosed as Idiopathic Hypersomnia.
For all intent and purposes, narcolepsy type 2 and idiopathic hypersomnia are not really different diseases, and in fact recent data suggest that when MSLTs are repeated, patients often move from one type to the other. The treatment is also similar, but unlike NT1, little is known regarding evolution of the disease, which can sometimes improve with time, so that it is important to consider stopping the treatment if people are doing much better as they get older. For these reasons, we are very careful when using addictive stimulants such as short acting amphetamines in NT2 and IH patients.
Narcolepsy and hypersomnia are disabling and underdiagnosed illnesses. The effect of type 1 narcolepsy on its victims for example is devastating. Studies have shown that even treated narcoleptic patients are often markedly psychosocially impaired in the area of work, leisure, interpersonal relations, and are more prone to accidents. Patients often face stigma and struggle to be as effective as other people. These effects are even more severe than the well-documented deleterious effects of epilepsy when similar criteria are used for comparison.
Treatment is a combination of behavioral changes and medications. Every patient is different, even when they have the same cause, for example a loss of hypocretin/orexin neurons like in NT1. We like to say that narcolepsy does not develop in a “vacuum” but in a real person with a particular personality, particular interests, dislikes and likes, and wanting to live their life in a particular way. The situation is slightly different in children where it is even more important to treat rapidly and aggressively to avoid schooling problems and loss of confidence associated with not being able to do well and suddenly becoming obese. Treatment involves a combination of one or several medications that help stay awake, sleep better and block having cataplexy or abnormal dreams. Tailoring the treatment to each patient, and giving behavioral advice is also essential.
Narcolepsy and hypersomnia are not hopeless conditions, and even if sometimes a clear cause is not found, there are active therapies. The first difficulty when facing a patient with excessive daytime sleepiness is to assess what is the cause of the problem. Is it a loss of hypocretin/orexin cells? What is the participation of sleep apnea, even if mild, or of poor sleep at night, abnormal circadian rhythms, or other complex neuropsychiatric issues that remain yet to be understood at the neurobiological level?
To establish that the cause of the problem is hypocretin/orexin deficiency as in NT1 is useful as it allows the patient to stop searching for the cause of the major portions of its symptom. We also have a lot of experience on how NT1 patients react to specific medications and what would help them. For example, we were among the first to realize that sodium oxybate is more active than any other therapies in NT1, notably in children, although use of this drug needs a specialize expertise as it does not come without serious side effect. Finally, it will likely be more and more important to know if patients are hypocretin/orexin deficient or not, as new drugs that replace orexin/hypocretin are now in clinical trials at Stanford. These are likely to be more effective in people who have this defect, not unlike insulin is uniquely effective in insulin dependent diabetes.
Unfortunately, however, the only way to be sure that a patient has hypocretin/orexin deficiency to date is to measure hypocretin/orexin in the cerebrospinal fluid, a test invented at Stanford. This test is only conducted in complex cases when there is a doubt, after verifying that the patient has the HLA-DQB1*06:02 marker mentioned above. Research we conduct at Stanford aims at finding a more convenient diagnostic blood test (probably immune or proteomics), or at finding ways to record and monitor sleep at home for better diagnostic and monitoring.
For other patients that do not have narcolepsy due to an orexin/hypocretin defect (NT1), whether they have narcolepsy type 2 (NT2) or idiopathic hypersomnia (IH) is largely academic, although sadly, it can make a big difference in the way patients get their medications covered by insurance. We and other have shown that to differentiate NT2 and IH based the MSLT as a diagnostic test results is meaningless. Often one patient who will have multiple SOREMPs on one MSLT test (consistent with NT2) will not have any SOREMP if retested a week later (consistent with IG). In brief, the MSLT only works repetitively to diagnose narcolepsy when there is hypocretin deficiency, as only in these cases it is then repeatable (although of course even in NT1, it is not perfect, and has about 5% false negative).
In NT2 and IH, what is more important therapeutically is to find the most likely contributing factors to daytime sleepiness for each patient. A better understanding of the cause(s) leads to personalized treatment, although sometimes, a physician comes to the realization there is no clear answer and trial and error with known medications is needed. In this context, treating mild sleep apnea with surgery or CPAP, behavioral therapies, removing existing sedative medications, using various types of stimulants, psychiatric medications such as antidepressants or lithium, or treating with sodium oxybate to ensure deeper nighttime sleep can all have life transforming effects.
It is clear the MSLT has outlived its purpose as a diagnostic test. First as mentioned above, the MSLT is only reliable to diagnose NT1, i.e., cases with orexin/hypocretin deficiency and typically cataplexy. Yet these cases are the ones where an MSLT is not useful, as a simple blood test (HLA typing, see above) plus clinical acumen is generally sufficient to get a reliable diagnosis. In doubt, CSF hypocretin/orexin can always be measured, as increasingly done in countries outside of the United States. Second, the MSLT not measuring the true problem of a patient with daytime sleepiness and is outdated technologically. Indeed, the MSLT can be confounded by shiftwork, sleep deprivation, and this creates false positive. It is also a test that measures the ability to fall asleep during the daytime, not the inability to stay awake which is the true complain of these patients. It is also an artificial test, not a real-life situation.
Criticism is easy but art is difficult. What do we do if the MSLT is not adequate anymore? Before the MSLT was invented at Stanford, assessment of narcolepsy or hypersomnia involved 24-48 hours of continuous EEG recordings in a sleep laboratory. The MSLT replaced these older tests because it was faster, cheaper and because the primary goal of diagnosis at the time was to identify NT1 patients. One advantage of the older 24-48 h continuous recording tests was that it was possible to objectively evaluate sleep and wake during the day and the night. It is still used in Europe, with the caveat that an exact protocol is not fully agreed upon, with some laboratories mandating subjects to stay in bed trying to sleep as much as possible, while others are asking patients to walk around and come to bed only when they need it.
The solution lies in the explosion of new hardware and software technologies. As an example of software development, our team has created a deep learning program which analyzes nocturnal sleep PSG results and diagnoses NT1 patients as well as the 2-day long MSLT procedure. The program works because it can detect atypical half-dreaming half-awake “states” in patients with hypocretin/orexin deficiency, as predicted from the description of their symptoms. We are also starting to develop artificial intelligent programs that can automatically detect a whole host of sleep problems and predict development of various cardiovascular or neurodegenerative diseases (see Mignotlab.com).
More excitingly however, it is now increasingly possible to monitor sleep using various consumer devices, although those that are currently available do not typically include EEG, the signal needed to identify sleep and study brain activity. We believe that the solution, attainable today, is to build a home monitoring device that can monitor wake and sleep EEG during the day, and breathing, EEG and leg movements during the night. This device would be used for 48 hr., for example during a weekend, and the signal sent to us by internet for automatic analysis. Analysis of sleep and wake at home in real life circumstances would allow sleep doctors to objectively evaluate what is wrong with each patient. Differential patterns of “wakefulness” may also start to be identified, reflecting the fact people cannot concentrate, are sleep deprived, in brain fog, etc. We would also be able to compare daytime alertness to sleep quality the night before, therefore properly classifying patients into subjects who are tired because they don’t have good sleep at night versus patients who seem to need to sleep all the time. This would for the rationale for a truly useful new classification of patients with narcolepsy type 2 or Idiopathic Hypersomnia. The test could also be repeated after treatment, ensuring response to any intervention and better titration of medications.
In parallel with this, new biological tests in sleep medicine are needed. Indeed, one of the other revolutions in medicine today is large scale analytics of metabolites and proteins. As described in my research lab page, we are now developing new diagnostic tests based on the multivariate integration of multiple analytes, notably proteins. These novel biomarkers will be able to differentiate if a specific patient is tired because she or he doesn’t get enough sleep at night and is somehow sleep deprived, or because their endogenous circadian clock is abnormal and they are in permanent jet lag, or because they are hypoxic at night because of sleep apnea. These tests may also be usable to monitor these variables in response to treatment.
We see a future where a combination of biological and home monitoring tests will, with proper analytics, revolutionize sleep medicine and the therapy of patients.
Our current results indicate that NT1 is associated with specialized autoreactive CD4+ T cells recognizing fragments of hypocretin presented by DQ0602, the HLA allele strongly associated with the disease. As the cause of the symptoms is the loss of hypocretin, we believe this population of autoreactive CD4+ T cells is likely within the causal pathway for narcolepsy. Influenza A, notably 2009 pH1N1 is a likely environmental trigger of the autoreactive CD4 + responses. This would explain why cases of narcolepsy in young children (where onset is abrupt), often start in the spring or summer, a few months after a presumed flu infection (that may sometimes be asymptomatic). It would also explain why cases of narcolepsy have been triggered by a specific swine flu vaccine called Pandemrix in 2009-2010.
Finding the causative narcolepsy autoimmune cells is high priority research in our laboratory. Indeed, once identified, we may be able to use their presence in blood as a diagnostic marker for narcolepsy. This would replace measuring CSF hypocretin-1 which requires a lumbar puncture. Second, understanding this process could lead us to modify flu vaccines to prevent not only the flu but the development of narcolepsy. Third, whereas a few years ago researchers believed that the brain and neurons were somewhat protected from autoimmune attacks, this knowledge is now outdated as new immune diseases affecting the brain are now being identified at a rapid pace (see mignotlab.com). Even neurodegenerative diseases like Parkinson’s and Alzheimer’s diseases are also believed to have important immune component. Understanding narcolepsy may thus serve as a model to understand T cell autoimmunity.
Finally, suffice to say that neurological complications following Covid -19 are now well recognized, not unlike what happened after the 1918 flu and encephalitis lethargica or after the 2009 H1N1 swine flu and narcolepsy. In this direction, understanding the interplay of normal and pathological immunity is going to be more and more important to understand and treat cancer, neurodegeneration, and a host of new diseases. Further, vaccines like mRNA vaccines, will become more and more potent and important in the fight against diseases, and with these active therapies will come cross-reactivity and side effects.