Self Management of Fatal Familial Insomnia. Part 1: What Is FFI?

Abstract

Context

Fatal familial insomnia (FFI) is a genetically transmitted neurodegenerative prion disease that incurs great suffering and has neither a treatment nor a cure. The clinical literature is devoid of management plans (other than palliative). Part 1 of this article reviews the sparse literature about FFI, including case descriptions. Part 2 of this paper describes the efforts of 1 patient (with the rapid-course Met-Met subtype) to contend with his devastating symptoms and improve the quality of his life.

Design

Interventions were based on the premise that some symptoms may be secondary to insomnia and not a direct result of the disease itself. Strategies (derived by trial and error) were devised to induce sleep and increase alertness. Interventions included vitamin supplementation, narcoleptics, anesthesia, stimulants, sensory deprivation, exercise, light entrainment, growth hormone, and electroconvulsive therapy.

Results

The patient exceeded the average survival time by nearly 1 year, and during this time (when most patients are totally incapacitated), he was able to write a book and to successfully drive hundreds of miles.

Conclusion

Methods to induce sleep may extend and enhance life during the disease, although they do not prevent death. It is hoped that some of his methods might inspire further clinical studies.

What Is Fatal Familial Insomnia?

Fatal familial insomnia (FFI) is an inherited (autosomal dominant) prion disease in which asparagine is substituted for aspartic acid at the 178 codon of the PrNP gene. Additionally, methionine (Met) occurs at codon 129 of the same mutated gene.[1,2] When methionine is also found at codon 129 of the nonmutated allele (Met-Met), the disease tends to run a shorter course than when the position is occupied by valine (Met-Val).[3]

The chief clinical features of FFI include a progressive and ferocious insomnia, waking “sleep,” hallucinations, autonomic disturbances suggestive of sympathetic overdrive (tachycardia, hypertension, hyperhidrosis, hyperthermia), a rise in circulating catecholamine levels, cognitive changes (such as attentional disturbance and short-term memory deficits without a loss in general intelligence), motor system deficits (ataxia), and endocrine manifestations.[1] Later cognitive changes involve a confusional state resembling dementia and, ultimately, death.

Mean age at onset is approximately 50 years, with most cases occurring between age20 and 61 years of age.[3–5] Age of onset cannot be predicted from polymorphism at the 129 codon.[1] Before clinical onset, the presence of FFI PrP mutant does not seem to affect function. Sensitive tests such as the PET scan and the premorbid EEG are normal.[6] Onset occurs when a critical amount of PrP is converted to PrPres (prion) protein.[7] The trigger for this protein conversion is unknown.

The typical duration of FFI is between 7 and 36 months, with a mean duration of 18 months.[8] Population studies show shorter mean survival time for Met-Met patients (12 ± 4 months) than for Met-Val patients, (21 ± 15 months).[3,9] These values are of particular relevance, because our case involves a Met-Met patient who surpassed 26 months of illness.

Clinically, 4 stages of FFI have been delineated, progressing from moderately disturbing to totally disabling.[10] Although clinical symptomatology is related to polymorphism at codon 129,[3] most end-stage patients are noninteractive and unable to care for themselves.

Neuropathologic Manifestations

The neuropathologic manifestations of FFI include thalamic degeneration, with selective involvement of the anterior ventral and mediodorsal thalamic nuclei, inferior olivary and cerebellar changes, and some spongiform change of the cerebral cortex.[3] Pathology of the neocortex varies with the disease duration and appears essentially spared of spongiosis in cases lasting less than 1 year (eg, Met-Met patients). The occipital lobe is relatively unaffected compared with the frontal, parietal, and temporal lobes. By contrast, the entorhinal cortex shows spongiosis and astrogliosis in virtually all patients.[7] Insofar as the magnitude of degeneration increases with the disease duration, those who die early show far less neurologic involvement than those who live longer.

The relationship between the topography of PrPres (prion protein) and neural dysfunction is unclear. In FFI, PrPres appears to be more widespread than suggested by lesions or areas of hypometabolic function.[11] Furthermore, the magnitude of PrPres is not correlated with clinical severity,[7] nor does its distribution parallel that of cellular apoptosis.[12] Finally, structures that contain equal amounts of PrPres (such as the thalamus and brainstem) are differentially vulnerable, the thalamus being much more so.[7] All FFI patients show similar amounts of PrPres in the thalamus and brainstem. However, those whose disease runs a short duration show the least amount and most focal distribution of PrPres, with the main accumulation being primarily in limbic areas (eg, entorhinal cortex or the cingulate gyrus) and in subcortical structures (including the thalamus, hypothalamus, and brainstem).[7] With disease of longer duration, the abnormal protein becomes detectable in the cerebral cortex and eventually exceeds the amount observed in subcortical areas.[7]

Thus, 2 interpretations regarding the role of PrPres are possible. Either PrPres is not toxic and programmed cell death results from the loss of important signals from the normally protective PrNP, or PrPres is toxic but different types of neural tissue are differentially vulnerable to it, in ways related to both tissue type and polymorphic status of C 129.[12] Tobler and colleagues[13] studied PrP knockout mice and suggested that normal PrP is necessary for sleep. Compared with the sleep of wild-type mice, sleep of knockout mice was more fragmented, showed smaller amplitude, slow waves of non-REM, and a stronger response to sleep deprivation.

Apoptosis or programmed cell death is a dynamic process, the consequence of activating signals and specific protein synthesis in the dying cell. It differs from necrosis in that it is not accompanied by local inflammation. In FFI patients (but not controls), apoptotic tissue was observed in a distribution and abundance that closely correlated with neuronal loss.[12] The mechanism of apoptosis in FFI is unknown, although a possible cause is the differentiation of microglia into macrophages, which is known to occur following neural damage.[12] This transformation is accompanied by the release of cytotoxins including free oxygen radicals, cytokines, or nitric oxide (NO), which may contribute to both cell death and the gliosis observed in FFI. In Dorandeu's study,[12] immunologic evidence for the presence of NO was supportive but relatively weak. To the extent that free oxygen radicals underlie cellular apoptosis, antioxidant therapy may possibly offer some protection against these ravages.

Does Insomnia Cause Death?

The precise cause of death in FFI patients is unclear. Although neural degeneration predicts long-term death, as in all prion diseases,[14] it is also possible that some functions that are disrupted or unbalanced by the ongoing disease (such as sleep and autonomic function) may terminate life even before critical degeneration. This might particularly be suggested by the rapid disease course of the Met-Met patients who, upon autopsy, show the least pathology, as measured either by apoptosis, neural degeneration, or PrPres distribution. This important possibility suggests that aggressive symptomatic treatment of insomnia may extend life.

The benefits of sleep are well documented in both humans and animals. In animal studies, total sleep deprivation resulted in death (within 4-6 days for puppies[15] and 2-4 weeks for rats[16]). Death is preceded by weight loss despite increased food intake, debilitation, a decline in thyroid hormone, elevated sympathetic activation, and poor resistance to infection.[17] Compared with yoked controls, glucose utilization decreases in the hypothalamus, thalamus, and limbic system. Hypocretin levels increase in the lateral hypothalamus[18] leading to wakefulness, stimulation of hypothalamic-pituitary axis, and sympathetic activation. A drop in body temperature 2-3 standard deviations below baseline is an irreversible harbinger of death.[17]

Postmortem analysis of sleep-deprived puppies describes capillary hemorrhages in the cerebral grey matter and chromatolysis and vacuolation in the cytoplasm of cells in the cerebral cortex, particularly the frontal lobe.[15] Rats show significantly lower liver and spleen weights and higher adrenal weight,[16] suggesting a stress response. Everson found no structural damage in the brains of these rats.[17] However, others have reported degenerative changes in the supraoptic nucleus (SON) of the anterior hypothalamus.[19] The SON, an integration site for surrounding hypothalamic nuclei, orchestrates many homeostatic mechanisms such as blood pressure, vascular resistance, and body temperature. If SON function is impaired in sleep deprivation, then subsequent changes in internal state may be a secondary consequence of the insomnia rather than a primary one.[19]

Sleep and Memory

Sleep has been implicated in the memory process.[20,21] However, describing its precise role is complicated by the fact that both sleep and learning consist of qualitatively different stages, each involving unique biochemistries and discrete neural loci. For example, memory, has been categorized as declarative and nondeclarative (with several subtypes), all of which involve various stages (including stabilization, enhancement, etc). Finally, determining whether learning has occurred may depend on the index of measurement.[20] Thus, conclusions regarding the role of sleep in memory sometimes differ between authors.

In rat studies, selective deprivation of either REM or slow wave sleep (SWS) interferes with long-term memory encoding.[22] A current hypothesis is that SWS is involved in clearing unusable memories and that REM sleep is necessary for the retention of newly acquired learning. Moreover, a temporal relationship has been established between the high frequency oscillations (ripple activity) of the rat hippocampus and the neocortical K complex spindles of SWS.[23] This activity is believed to underlie the learning process by strengthening connections between memory-representing cells in diverse parts of the cortex.

Some controversy exists regarding the importance of sleep in human learning.[21,24] In particular, the role of REM sleep in memory consolidation is challenged by the case of an Israeli man with a REM-abolishing shrapnel injury to the brainstem. In the 35 years since his injury, the man has prospered professionally as a lawyer, painter, and puzzle columnist.[21] He is described by others as entirely normal. Unfortunately, empirical studies using material known to depend on REM sleep have not been performed on either this (or similar types) of patients.[24]

Procedural (motor) learning improves in both speed and accuracy after sleep, although not to the extent that it replaces the need for practice.[24] At present, it is accurate to say that not all stages of sleep are required for any single form of memory consolidation. Nor is any one stage required for all forms of consolidation. Finally, some forms of memory may not require facilitation by sleep at all. Nevertheless, compelling evidence does exist for an important role of sleep in human learning.[20,21]

Human Studies

Human sleep deprivation studies differ from animal research in that human studies are of shorter duration than animal studies. Second, humans know why they're being kept awake and may not respond with the same degree of stress as animals.[25] Sleep deprivation may result from endogenous factors (such as aging or chronic insomnia) or exogenous prevention of sleep state (such as deliberate waking).

Total sleep deprivation stimulates the HPA axis and suppresses the growth hormone (GH) axis.[26] After a 90-hour vigil (3.7 days), human subjects show the same drop in body temperature as animals (although the circadian rhythm remains) and the same loss of sensory acuity, motor speed, and short-term memory.[15] Visual hallucinations and psychotic-like behavior have been reported.[15] As opposed to rats, who are not clinically diabetic after long-term deprivation, sleep deprivation in humans is associated with a prediabetic state.[27]

The longest period of voluntary sleep deprivation, 264 hours, was associated with irritability, incoordination, slurred speech, blurred vision, hypnagogic reveries, lapses in attention, and disturbances of short-term memory.[28] Although the vigil appeared to have no lasting deleterious effect, photographs taken during this episode show a subject with a haggard and apathetic facade.

Relative sleep deprivation in humans often occurs as a natural consequence of aging. By the fourth decade of life, both SWS and GH are reduced by 75%.[29] GH deficiency in the elderly is associated with weight gain, loss of muscle mass, and reduced exercise capacity. Increasing deep sleep triggers a proportional increase in GH secretion.[29] By the sixth decade, total sleep time is reduced by approximately 27 minutes per decade, dramatically affecting REM sleep as well. The loss of REM sleep appears to be associated with elevated evening levels of the stress-related hormone cortisol. Cortisol levels normally peakin the morning and decline during the day to very low levels in the evening, giving the body time to recover. Subjects with decreased REM sleep, however, were less able to achieve evening quiescence. Lack of hormonal “down time,” a recovery period for the stress-response system, has been linked to memory deficits and insulin resistance, a risk factor for diabetes. Elevated evening cortisol levels could also cause additional sleep loss. Although the elderly person does not die of his relative insomnia, supplements that reverse sleep decay have been promoted as rejuvenating.

Chronic Insomnia

Chronic insomnia (CI) refers to the subjective experience of poor sleep. People differ vastly in their nightly sleep requirements from as little as 1 hour[30] to more than 10 hours. Meddis[30] described a cheerful, active, 70-year-old retired nurse who claimed to have always slept little (even in childhood) and who, upon 5 nights of testing in a sleep lab, was shown to average roughly 67 minutes of sleep per night. Approximately 3% of the population experience insomnia, particularly older people and women.[31] CI is associated with an overall rise in ACTH levels and cortisol secretion, which retains a normal circadian pattern. It is not clear whether the increased cortisol level is a cause or consequence of the insomnia.[32] Physiologic hyperarousal (eg, elevated metabolic rate; sympathetic nervous system activity) is considered to be characteristic of primary insomnia[33] and also occurs in FFI.[34]

Of interest, people who are sleep-deprived (including FFI patients) tend to be sleepy during the day. Insomniacs are not. Nevertheless, insomniacs are more likely to complain of daytime problems of mood and concentration than normal sleepers. Chronic primary insomnia is also associated with an elevated risk for major depressive and anxiety disorders (symptoms also noted in FFI).

Sustained periods of insufficient sleep (eg, 4 hours of sleep for 6 consecutive days) result in impaired glucose tolerance[35] (possibly contributing to a prediabetic state),[29] a decrease in natural killer immune cells,[25] an increase in thyrotropin and evening cortisol concentrations, and elevated sympathetic nervous system activity.[35] This degree of sleep loss also results in poorer performance on neuropsychological tests of temporal order[36] and complicated motor skills.[37]

Selective sleep deprivation restricted to specific stages (eg, REM or SWS only) results in a rebound of the deficient activity during the recovery period. During SWS deprivation, characteristic delta waves also increase during the REM portion,[38] suggesting a low-frequency homeostatic mechanism common to non-REM and REM sleep.

Sleep Deprivation in FFI

Despite its suggestive name, the insomnia of FFI may not be an early or essential symptom of the disorder. Among a series of German patients, sleep disturbances were mild and often recognized only in retrospect after detailed questioning of the family or reinvestigation of the hospital records.[5] Similar observations have been reported in other international populations.[14,39,40] We propose that the term insomnia must be used in a technical sense, meaning not just the complaint of subjective insomnia, but a disorganization of processes intrinsic to sleep. This requires lab confirmation, which is not often available. For example, Julien and colleagues[41] reported a patient who “sometimes lapsed into a short-lasting behavioral state of sleep with closed eyes, but EEG performed during these periods did not disclose any sleep pattern.” Charts and retrospective reports do not clarify this issue. Moreover, FFI patients may be admitted late in the course of the disease, when insomnia is already lost as a symptom, or may receive a single polysomnography which is not sufficient to demonstrate abnormalities.

The EEG of normal sleep-deprived subjects shows a predominance of delta activity[42] and episodes of “micro-sleep,” brief bursts of high-amplitude slow waves, K-complexes, and sleep spindles.[43] The absence of this response is associated with disturbed perception, lapses of consciousness, and erratic behavior.[44] In patients with FFI, sleep EEGs contain a mixture of rhythms that are neither typical of wake nor of light sleep, and may be called “subwakefulness.”[3] Similarly, EEG testing of an FFI patient “without insomnia” demonstrated an absence of sleep spindles and K-complexes.[45] Total sleep time was reduced and showed only slow activity or desynchronization without rapid eye movements.

The impact of sleep deprivation upon the course of FFI is not trivial. Many of the other symptoms of FFI that contribute to dysfunction and death may be secondary to the insomnia. Chronic sleep deprivation has been associated with hypometabolism in the thalamo-limbic circuits (in a way similar to the PET markers of FFI).[17] epileptic seizures,[46] aggravation of the autonomic functions,[3,47] increased cortisol, and reductions in both thyroid-stimulating hormone and melatonin.[48] Observed deficits in memory encoding may result from degeneration in the dorsomedial thalamus[49] or the disruption of consolidation that normally occurs during sleep.[23] In addition, the secretion of GH that normally occurs during deep sleep is reduced in FFI.[47] Because GH plays a role in body growth, fat mobilization, and inhibition of glucose utilization, its disruption might underlie the rapid aging and weight loss noted in the FFI patient. Neuroimaging studies of cerebral energy metabolism during the sleep-wake cycles show a global decrease in energy consumption during SWS as opposed to REM and waking states, suggesting that, for the normal, it is a time of relative rest of brain cells.[50]

Sleep Alteration in FFI

As described by Montagna,[34] sleep in FFI is characterized by an early and progressive reduction in sleep spindles and K-complexes, a reduction in total sleep time, and disruption of the cyclical organization of sleep; SWS is lost first, then REM disengages from its circadian cycle and intrudes into the waking state. Lugaresi and colleagues[47] explain that circadian rhythms involving melatonin gradually decrease and shift in phase, and finally disappear. The rhythmicity of somatotropin (GH) shows a similar reduction or total loss in tandem with the loss of deep sleep. Only prolactin rhythmicity remains unaltered.[3]

Met-Met patients demonstrate a different sleep pattern than Met-Val patients, including severe fragmentation, brief but repeated episodes of sudden-onset REM (with oneiric enactment), and an earlier loss of total sleep. Sleep loss in Met-Val patients progresses more slowly, although they, too, ultimately lose deep-sleep stages, slow-wave EEG activity, and circadian motor rhythms. Nevertheless, body temperature, heart rate, and blood pressure remain elevated (Reder, personal communication, 2005).

Circadian Rhythms in FFI

Circadian factors regulating sleep are primarily localized in the hypothalamic suprachiasmatic nuclei (SCN) and promote sleep in concert with other biological events, including lowering core temperature and cortisol levels.[51] An obvious mystery of FFI is the disappearance of circadian rhythms despite ostensible integrity of the hypothalamus.[7,39,52] Mignot and associates[18] attribute disruption of circadian rhythms to an interdependence between hypothalamic and thalamic nuclei. Specifically, the dorsomedial hypothalamus innervates not only the mediodorsal (MD) and paraventricular (PVT) nuclei of the thalamus, but also other hypothalamic areas involved in non-REM sleep and thermoregulation (ie, medial preoptic area), circadian rhythms (eg, SCN), and adrenocortical axis (ie, PVN), as well as other neuroendocrine secretions. In addition, hypothalamic function in FFI may still be altered despite the lack of histologic evidence. Indeed, an abnormal accumulation of PrPres is noted in the FFI hypothalamus as early as 7-8 months into the disease.[7]

Rodent studies suggest that circadian rhythms can be promoted even in the absence of neural connections between SCN and other brain structures. Transplants of fetal SCN tissue into the brains of SCN-lesioned hamsters restore circadian rhythms even when the transplanted tissue is encapsulated and has no neural connections to the recipient animal.[53] The SCN of pregnant mice synchronizes fetal SCN activity,[54] and its removal during the gestational period results in desynchronization of the unborn. The means of entrainment is probably hormonal. These findings raise the question of whether administration of missing hormones (eg, GH or melatonin) may help restore circadian rhythms or compensate in some way for their absence. Administering melatonin induces sleep sooner at night and has been used to treat jet lag.[54] Several studies have shown that circadian rhythms are influenced by external stimuli, such as light. Direct retinohypothalamic pathways exist in rodents.[55] Human subjects deprived of light stimulation follow rhythms slightly out of sync with the 24-hour diurnal solar pattern, indicating that light information finely tunes these rhythms.

Differences Between FFI and Sleep Deprivation

Although patients with FFI share many features with those who are sleep deprived, certain differences are apparent:

1. Experimental subjects experience constant sleep pressure and immediately lapse into sleep if permitted. The FFI patient cannot fall asleep.

2. Selective deprivation of REM or non-REM sleep is followed by selective rebound, indicating the necessity of each. As compared with the pre-deprivation period, human adults recovering from SWS deprivation show rapidly appearing and long-lasting delta. This increase in delta waves (most notably in the frontal area), occurs in both REM and non-REM sleep.[38] In FFI, sleep loss is initially SWS, which does not show any form of rebound. Loss of nocturnal REM is replaced by brief, periodic episodes of “parasomnia,” a state resembling REM but without atonia. This parasomniac REM is not experienced as refreshing. Except when the patient is distracted by others, he or she remains almost continually in this stuporous state.[34]

3. Normal sleep-deprived subjects are most impaired during the nighttime hours because circadian rhythms depress overall functioning level. In FFI, circadian rhythms break down and the distinction between day and night is blurred.

4. Temperature decline associated with chronic sleep deprivation does not occur in patients with FFI. On the contrary, temperature is frequently elevated.

5. Hypocretin levels are elevated in sleep-deprived animals but not in FFI patients.[18]

6. Rats who die of sleep deprivation do not show cortical apoptosis,[17,56] as seen in patients with FFI.

Model of Death in FFI as a Result of Sleep Deprivation

FFI patients appear to have some “sleep” relief in a quasi-REM state. However, the REM state is associated with a significant rise in sympathetic activity (heart rate and blood pressure) compared with the waking state.[57] This suggests that the waking REM of FFI offers no protection from excessive sympathetic overdrive. A plausible hypothesis regarding early death in the Met-Met patient may be “burnout” of the sympathetic nervous system. Indeed, energy expenditure of FFI patients is very much higher than that of normal controls, which suggests severe metabolic exhaustion as a consequence.[34]

Symptomatic Treatment

Although the ultimate therapy for FFI must address the destructive prion, present treatment of FFI is mainly palliative. FFI patients respond poorly to conventional drugs such as sedatives[41] and benzodiazepines.[52] Cibula and colleagues[58] mention giving a cocktail of medications to successfully induce SWS and sleep spindles for EEG testing of an insomniac, non-FFI patient without thalamic degeneration. These investigators did not discuss their potential therapeutic use for patients with FFI. Gamma-hydroxybutyrate (GHB), an investigational drug that reliably stimulates SWS in normal and insomniac patients, resulted in 3 hours of SWS sleep in a patient with FFI of 14 months' duration (3-mg dose).[48]

Whereas before receiving this medication, the patient was unresponsive to his environment, when he awoke from his drug-induced sleep, he was alert, attentive, and responsive to questions. The continued nightly use of this medication yielded similar positive results for an additional 2 weeks. During the third week, the patient developed fevers and restlessness, and subsequently died. Obviously, whether the medication hastened or delayed his death cannot be determined. GHB is a normally endogenous metabolite and precursor of GABA; it is found in all mammalian cells, with the greatest concentrations being found in the substantia nigra, thalamus, and hypothalamus. Known effects include increases in acetylcholine and serotonin and time-dependent changes in dopamine (DA) levels. DA suppression is followed by increased secretion of both GH and prolactin.

At appropriate doses, GHB induces sleep, lowers the body temperature, and slows the heart rate without loss of blood pressure.[59] Indeed, many of the problems inherent to FFI are influenced by this drug. Although recreational use of GHB has been associated with adverse reactions and a few fatalities,[59] low doses have been regarded as safe. GHB has been used for over 40 years in surgical procedures.

According to Reder (personal communication; 2005), the patient's death followed shortly after the termination of treatment trials, suggesting either that GHB administration must not be stopped abruptly or, perhaps, that the patient had reached a fatal level of degeneration and would not have survived under any circumstance. Autopsy of this patient showed that only 5% of dorsomedial and anteroventral thalamic cells were preserved. It is significant that SWS can still be induced with this very low number of cells and that cognitive function should be so markedly enhanced as a result. These findings suggest that the functional suppression is reversible and not totally a consequence of degeneration.

Efforts to treat the confusional state of FFI are not systematically made. In one patient, flumazenil (a benzodiazepine antagonist) produced a dramatic alerting reaction on 2 occasions, but the drug was not administered therapeutically.[1,45] Treatment of insomnia, even if only moderate, may positively affect the patient's symptoms and enhance the quality of the patient's remaining days. The accompanying case describes the ongoing self-initiated therapeutic efforts of a Met-Met patient, DF, whose symptoms began around March 2001 and who managed, through both conventional and nontraditional approaches, to ameliorate some of the devastating symptomatology of FFI, to achieve sporadic but substantial periods of sleep (even in the 26th month of his illness), and to enhance the quality of his life despite the ongoing ravages of his disease. The purpose of this paper is not to suggest that FFI can be cured by this approach or even that life can be extended by any appreciable amount (although DF did outlive the typical Met-Met patient). Primarily, it is to offer some suggestions for disease management and research directions, as well as to offer neuropsychological insights into the disorder.

All of DF's approaches were based on trial and error. Many worked for a short time, and some were accompanied by serious medical side effects. Some of these approaches stirred ethical concerns among the participating clinicians. To entreat compliance, DF stressed his fatal diagnosis and the interest of science. Although his good days were often punctuated with symptoms of his illness (such as fever, high blood pressure, ataxia, dysarthria, tremor, memory loss), his various strategies to promote sleep gave him months of mental clarity and productivity; during the 25th month of his illness, he was able to drive 950 miles in a single 15-hour stretch.