The rebuttal contained data from Jenkins’s graduate student Olivia Moody and included results from a laboratory at the University of Queensland that had used a different patch clamp technique on the same CSF samples. As a bonus, the Emory group added testimonials from colleagues in Florida and Minnesota who had begun prescribing flumazenil to their patients starting in 2014.

  Around this time, Rye told me he felt like he was on a “clinical treadmill,” unable to devote as much dedicated time to research as he had earlier in his career. And while the number of patients who had tried flumazenil had climbed above three hundred by mid-2016, there wasn’t much new biochemistry to throw back at the French. The identity of the sleepy juice factor in patients’ CSF remained undetermined.

  ALTERNATIVE MECHANISMS

  Other mechanisms could explain the sleep-dispelling action of flumazenil, and they do not require invoking the existence of an endogenous benzodiazepine-like substance. Instead of displacing something from GABA-A receptors, flumazenil may be weakly signaling through them or reducing the receptors’ abundance on brain cells’ surfaces. These changes would take hours to occur, rather than minutes—an extension of changes that occur all the time. In animals, sleep deprivation causes an increased level of GABA-A receptor expression on hypocretin-producing and other midbrain neurons.9 A subset of GABA-A receptor levels also change after acute alcohol consumption.10

  When asked about this possibility, Rye referred to patients such as Anna and Sigurjon, who noticed the wakening influence of flumazenil quickly. While others also say they felt flumazenil’s effects immediately, not everyone has the same experience. A few reported feeling dizzy or sleepy right after taking it. Some people said they felt its effect the next morning, after taking their first dose in the evening. For them, flumazenil seems to have a stronger effect on sleep inertia, the transition out of sleep, than on overall daytime alertness.

  “I noticed that it basically cured my sleep inertia in the morning when I put it on before bed,” said Amy Desmarais, who was diagnosed with IH in 2015 after an initial diagnosis of “probable narcolepsy” during graduate school a few years before. “I was too afraid to stop taking stimulants during the day and didn’t really notice a difference during the day, so decided to only use it at night when I knew it definitely helped.” She said: “I still don’t like waking up in the morning, but then again I haven’t met many people who do. But the difference is that when I put flumazenil on, I can wake up in the morning. My alarm goes off and I still want to snooze, but the fact is, I hear it and I wake up. Whereas before flumazenil, I would sleep through anything and everything.”11 Other IHers have reported that for them, flumazenil took days or a week to kick in. They gradually felt less sleep inertia and were able to reduce the number of hours of sleep they needed per week, but its wakening effect wasn’t as dramatic as what others experienced.

  For support for this alternative mechanism, we can look to researchers studying how flumazenil could be used to treat addiction and withdrawal. In rats exposed to chronic methamphetamine, levels of a benzodiazepine-insensitive form of the GABA-A receptor (alpha 4) increase in the hippocampus.12 Those levels stay elevated when the rats are in withdrawal. Flumazenil can push the levels of the alpha 4 form back down, reducing withdrawal-related jumpiness in the rats. The presence of flumazenil appears to drive neurons to recycle GABA-A receptors more quickly.13 In addition, researchers in Italy have observed that neurons respond strongly to flumazenil after the cells are exposed to alcohol. That responsiveness goes along with elevated levels of the alpha 4 GABA-A receptor.14 Adjustments in GABA-A receptor expression are part of compensatory changes in neural sensitivity that occur during addiction and withdrawal for several drugs.

  Addiction treatment is one area of medicine where a limited number of physicians have gained experience with flumazenil comparable to Rye’s and Trotti’s. Flumazenil found its way into a flashy addiction treatment called the Prometa protocol, which was promoted in Hollywood around the time that Anna was going through her struggles.15 The Prometa protocol failed in a rigorous clinical trial,16 but clinicians at the University of Verona have reported success using continuous low-dose flumazenil in treating benzodiazepine withdrawal.17 The leader of the Verona group, Fabio Lugoboni, has published commentaries with titles such as “What Is Stopping Us from Using Flumazenil?”18 Perhaps the sleep research and addiction medicine communities could learn more from each other, since there is no licensed treatment for benzodiazepine dependency or addiction.

  A THIRD GABA OPTION

  In 2014, just after the Living with Hypersomnia meeting, Rye became aware of the ongoing development of another drug acting on GABA that deserved consideration, given the limitations of flumazenil and clarithromycin. He was speaking at a sleep research conference in Texas, where he met the Stanford neuroscientist Craig Heller.19

  Heller was part of a group investigating GABA receptor antagonists for another purpose: as cognitive enhancers for people with Down syndrome. In mouse models, the Stanford researchers had observed benefits from both flumazenil and a related drug, pentylenetetrazol, or PTZ. Those benefits were on memory, not sleep.

  Some of the Stanford group had formed a company, Balance Therapeutics, to commercialize PTZ. When Rye met Heller, Balance had already started a clinical trial in Australia in people with Down syndrome, and company executives were looking around for other possibilities. Heller introduced Rye to Balance executives, and the company was willing to expand into sleep disorders. Balance thus became the first pharmaceutical company to sponsor clinical trials designed for people with IH. It demonstrated Diana Kimmel’s and Jennifer Beard’s idea from a few years before: if people with IH organized themselves, industry partners will make themselves available.

  FIGURE 10.1. Chemical structures of flumazenil, clarithromycin, and pentylenetetrazol.

  While the development was encouraging, Balance was a small company, with just a handful of employees and limited resources. Balance licensed the Parker-Rye-Jenkins flumazenil patent from Emory University but did not organize clinical trials for other drugs besides PTZ, such as flumazenil.

  In addition, PTZ differs from flumazenil in a critical way. It doesn’t act on the benzodiazepine site—it directly blocks GABA-A receptor function, and its action does not depend on the presence of a somnogenic ligand, peptide or otherwise.20 The rationale for PTZ in hypersomnia was that it might push back against overactive sleep-related signals from GABA. PTZ’s GABA-based pharmacology was different enough from conventional stimulants for it to be a possible alternative, but its success would not have proven Rye’s “sleepy juice” theory.

  FROM THE DUSTBIN OF PHARMACEUTICAL HISTORY

  PTZ’s long history—part notorious, part benign—may not inspire confidence. Pentylenetetrazol’s first medical uses in the 1920s predate the synthesis of amphetamines.21 Known first as Cardiazol and called Metrazol in the United States, it was manufactured by the German pharmaceutical firm Knoll.22 In the 1930s, it was used in psychiatric hospitals in Europe and the United States to deliberately induce seizures in people with schizophrenia.23 Decades later, it was advertised as a mild stimulant for the elderly before being removed from the market as part of FDA reforms.24

  PTZ’s revival in the twenty-first century was part of a burst of drug discovery efforts aimed at Down syndrome, which researchers had assumed for years was too complex and permanent for pharmaceutical treatment.25 The origins of the wave came from research on Ts65DN mice, which have an extra copy of almost one hundred genes, modeling the situation in Down syndrome, in which an extra copy of chromosome 21 leads to intellectual disability. In brain regions important for learning and memory, some of the extra genes are thought to shift the balance between excitation and inhibition more toward the latter.

  The neuroscientist Craig Garner, then at Stanford, and colleagues wanted to push back against GABA’s inhibitory signals as a way to recalibrate the balance. Although PTZ’s reputation as a seizure-induction tool put off some scientists, it did not intimidate Garner. “Every drug has a toxic range,” he said. “Seizures are just toxicity. In the mice, we could see positive effects without getting anywhere near that point.”26

  Ts65DN mice that were given small doses of PTZ every day for a couple weeks performed better on memory tests.27 Specifically, they were able to recognize new objects, when usually Ts65DN mice could not. The mice didn’t perk up right after the first dose, but the memory effects lasted for up to two months after the drug was withdrawn. The observation suggested that low-dose PTZ was not acting as a conventional stimulant, since its effects lasted after the drug had left the body.28 “We think that it’s more about network adaptation,” Garner said. “The circuits in the brain readjust.”

  After Garner unsuccessfully tried to get established pharmaceutical firms interested in his Down syndrome work, he formed Balance Therapeutics in 2009. He said his lab also had obtained positive results with flumazenil,29 but industry advisors told him that PTZ was a better drug candidate, because flumazenil didn’t last long in the body and had an inconvenient mode of delivery.

  Balance sponsored a study of PTZ in Australia called Compose21, which enrolled eighty-eight young participants with Down syndrome, up until 2015.30 The Compose21 results remain unpublished; positive results would have been reported. Similar studies in other neurodevelopmental disorders, such as fragile X syndrome, have stumbled over differences between homogeneous populations of lab mice and humans of varying ages and abilities.

  Confounding matters, humans with Down syndrome tend to have more fragmented sleep than neurotypical people, and they often have obstructive sleep apnea because of their altered craniofacial structures.31 On top of that, Ts65DN mice are not sleepier than standard lab mice.32 However, since brain fog can be a prominent feature of IH, PTZ’s effects on memory are intriguing, even if they have been demonstrated only in mice.

  PTZ FOR HYPERSOMNIA

  At the 2017 World Sleep Congress in Prague, Rye reported his positive initial findings with PTZ—but only with five patients, in a non-placebo-controlled format.33 Balance Therapeutics has sponsored two clinical trials testing PTZ in people with narcolepsy type 2 and idiopathic hypersomnia.34 The company was able to recruit participants through the Hypersomnia Foundation’s conferences, for which it was listed as a sponsor. Balance also recruited through the Foundation’s IH registry, showing how compiling data on people with IH could facilitate clinical trials.

  Some members of the hypersomnia community, including Meghan Mallare and Diana Kimmel, have reported favorable experiences in Balance’s clinical trials. Some said it was the only experimental medication that seemed to work for them. A few have gone as far as to seek unconventional access to the drug; one person obtained it as a component of the cough syrup Cardiazol Paracodina, which is available in Italy.35 Balance was not able to respond to expanded access requests, executives told me.

  When Balance evaluated the results from its second clinical trial for sleep disorders, development of PTZ was stopped, and the board of directors decided to wind down the company. In 2020, Balance was put up for sale by Gerbsman Partners, which specializes in restructuring technology and life science firms, but there were no buyers.36 And as of 2022, the results of Balance’s clinical trials of PTZ in people with sleep disorders are unpublished. Despite anecdotal success, PTZ did not provide benefits to enough people in the second clinical trial such that the company’s advisors thought it would succeed.

  Although PTZ did not appear to have a path forward, some valuable precedents emerged from Balance’s efforts. It was the first company to obtain an orphan drug designation for IH and conduct clinical studies focused on IH. Also, for its second clinical trial, Balance developed its own outcome measure panel: a questionnaire on IH symptoms. According to Gerbsman, the FDA has agreed that the panel can be used as the basis for a marketing approval. We will discuss orphan drugs and the FDA approval process more extensively in chapters 14 and 15.

  CHAPTER 11

  THE HEART OF THE BRAIN

  The hypothalamus does mundane things, but it does them well, and how it solves difficult challenges can tell us much about what neurons and networks are capable of.

  —Gareth Leng, The Heart of the Brain

  In this chapter, we will venture beyond Rye’s “sleepy stuff” and look at other mechanisms that may explain how and why idiopathic hypersomnia occurs. In general, sleep researchers have moved beyond looking at sleepiness as being brought on by a single substance. The field has shifted to brain circuits and cellular clocks, and two proposals we will examine involve disturbances of those circuits and clocks. Experts don’t yet know enough to say definitively whether IH comes from sleep being too “thin”—missing some element of its mysterious, restorative oscillations—or the reverse, abnormally deep and prolonged because of a distortion of regulatory forces. In addition, uncertainty remains regarding to what degree IH is imposed from outside by injury or stress, versus an inherited disorder that unfolds from within. Still, enough is known to sketch out two possibilities.

  First, we must acknowledge that in clinical practice, the label of idiopathic hypersomnia is applied to individuals with a variety of underlying issues. More than one pathological mechanism may apply to people who fit under this umbrella. As currently defined, IH includes both people who display extra-long sleep periods—eleven hours of sleep or more on most nights (figure 11.1)—and those who do not.1 Some with IH diagnoses may have chronic pain or other factors disturbing their nighttime sleep, enough that they fall asleep quickly in the setting of an MSLT. It is also possible that some diagnosed with IH may have other conditions that are difficult to spot right away: metabolic diseases, vitamin deficiencies, or even covert hepatic encephalopathy. One of the seven flumazenil-responsive patients described in Rye and Jenkins’s 2012 paper was discovered later to have myotonic dystrophy, an inherited multisystem disorder that includes daytime sleepiness as a prominent symptom.2

  FIGURE 11.1. Sleep recordings from a patient with idiopathic hypersomnia with long sleep and sleep drunkenness.

  Source: U. Voderholzer et al., “A 19-Hour Spontaneous Sleep Period in Idiopathic Central Nervous System Hypersomnia,” Journal of Sleep Research 7, no. 2 (1998). Reprinted by permission from John Wiley & Sons ©1998.

  Putting those possibilities aside, one proposed mechanism for IH is an old idea, extending back before Bedřich Roth’s time. It is partly based on an analogy with narcolepsy. The basic idea is that an infection somehow damages part of the brain whose function is to regulate sleep and wake. More than one hundred years ago, this idea manifested in a way that shaped the history of neurology and produced a group of famously sleepy patients.

  SMOLDERING WILDFIRES

  In the middle of World War I, doctors at a hospital in Vienna began to see patients with signs of inflammation in the central nervous system: encephalitis. Amid wounded soldiers who had returned from Austria-Hungary’s battles, these civilians with encephalitis were distinctive. Affected people usually had fever and headaches. They developed double vision, and the muscles controlling their eyes became paralyzed. And they were profoundly sleepy. They could be wakened and answer questions but would fall into delirious quasi-sleep if left alone and would sometimes progress to stupor and coma. More than a quarter died. Despite differences between individual cases, they had enough in common that a visionary neurologist named Constantin von Economo (figure 11.2) grouped them together and named their disease “encephalitis lethargica.” “The most striking symptom of this disease is the sleepiness of the patients, which at times is associated with delirium and at times is not. This somnolence can vary from light sleep, entirely resembling physiological sleep, to the deeper stupor, independent of any fever,” von Economo wrote.3

  FIGURE 11.2. Constantin von Economo was one of Austria-Hungary’s first credentialed pilots. He was serving as a wartime pilot until 1916, when he returned to Vienna, where he would investigate encephalitis lethargica.

  Source: Photo by Max Schneider. Public domain.

  Independently from von Economo, doctors in the United Kingdom and France made similar observations. An epidemic of encephalitis appeared in Europe and the United States in 1918. A second wave in the 1920s was more varied, with a dominant “somnolent-ophthalmoplegic” form (sleepiness plus eye paralysis), but other patients experienced different symptoms, such as insomnia, agitation, and involuntary movements.4

  For those that recovered, the acute state of lethargy lifted after days or weeks. However, the epidemic left thousands of people with a variety of neurological and psychiatric disorders—usually with a puzzling delay of months or years before the appearance of chronic symptoms. Encephalitis lethargica was like a series of smoldering wildfires, burning across the brain along trails that differed from person to person.5

  Many survivors had symptoms similar to Parkinson’s disease: rigidity, immobility, and tremors. Those with symptoms like these lingered in New York City hospitals long enough for Oliver Sacks to learn their stories in the 1960s for his book Awakenings, describing people who had been immobilized for years. Other survivors developed sleep disorders, such as severe insomnia or sleep inversion: being drowsy during the day but having disturbed sleep at night. Some of those with sleep inversion found night jobs because of their need to sleep during the day.6 A handful of encephalitis survivors were reported to have narcolepsy, although terminology was then looser, and part of this group did not display clear cataplexy.7 A few individuals developed a sleep disorder that resembled idiopathic hypersomnia.