Originally developed by Bioprojet in France, pitolisant was approved in the European Union in 2016. Because of its different mode of action, pitolisant was thought to have less abuse potential, and it was the first wake-promoting medication to avoid Drug Enforcement Agency restrictions. For narcolepsy type 1 symptom management, another attractive property of pitolisant was its ability, demonstrated in some clinical trials, to reduce the frequency of cataplexy.59

  TABLE 14.1 Wake-promoting medications investigated for idiopathic hypersomnia

  Medication

  Neurotransmitter

  Mechanism

  Clinical studies for people with IH?

  Modafinil (Provigil)

  Dopamine

  Inhibits reuptake

  Germany 2009–2011, Japan/Alfresa 2014–2015, Emory head-to-head 2019–2023

  Armodafinil (Nuvigil)

  Dopamine

  Same as modafinil

  No

  Solriamfetol (Sunosi)

  Dopamine, NE

  Inhibits reuptake

  No

  Methylphenidate (many)

  Dopamine, NE

  Inhibits reuptake, promotes release

  Commonly prescribed

  Amphetamine (many)

  Dopamine, NE

  Inhibits reuptake, promotes release

  Commonly prescribedEmory head-to-head 2019–2023

  Mazindol (Quilience)

  Dopamine, NE

  Inhibits reuptake, promotes release

  No; expanded access program in Europe

  Pitolisant (Wakix)

  Histamine

  H3 receptor antagonist

  Harmony: 2022–

  GHB/Oxybate (Xyrem/Xywav)

  GABA

  GABA-B receptor agonist (see chapter 15)

  Jazz: 2018–2020

  Flumazenil (Anexate/Romazicon)

  GABA

  GABA-A receptor antagonist

  Emory: 2010–2012

  Clarithromycin (Biaxin)

  GABA

  GABA-A receptor inhibitor

  Emory: 2010–2012, 2019–ongoing

  Pentylenetetrazol (Cardiazol, Metrazol)

  GABA

  GABA-A receptor inhibitor

  Balance: 2015–2018, 2018–2020

  GR3027

  GABA

  GABA-A receptor inverse agonist (neurosteroid)

  Umecrine: 2017–2019

  TAK-925

  Hypocretin

  Hypocretin receptor agonist

  Takeda: 2020

  A study of pitolisant from Isabelle Arnulf’s sleep disorders clinic in Paris was relevant for IHers seeking additional options. Arnulf’s group had people with IH resistant to conventional stimulants try pitolisant in an open-ended fashion.60 Arnulf’s retrospective analysis was similar to how the Emory group assessed the efficacy of flumazenil. About a third of those whose records were reviewed reported an improvement on the Epworth Sleepiness Scale more or equal to 3, a significant alleviation of daytime sleepiness. A majority stopped taking the drug, mainly because of lack of efficacy and more rarely because of side effects, but the proportion who decided to keep taking pitolisant was about the same as with flumazenil.

  PLAYING POOL: OBJECTIVE VERSUS SUBJECTIVE

  Observers of the pharmaceutical industry have compared clinical trials to playing pool. Like committing to sinking a ball in the corner pocket, study organizers need to declare a primary endpoint ahead of time to get credit for it. When researchers are designing clinical trials for people with IH, what should be the measures of success? This decision cuts to the core of what IH is and how it is defined.

  Going back to Bedřich Roth, we have seen that the sleepiness of hypersomnia needs to be evaluated and measured differently, in comparison with narcolepsy. For IHers, sleepiness is multifaceted, including an inability to stay alert, the total hours of sleep someone needs or can’t push away, and sleep inertia. Measuring how quickly someone falls asleep only captures one part of IH, and sleep latency is often not the strongest aspect of sleepiness. Thus, to measure sleepiness in IH patients, the field may need to make a shift away from narcolepsy-oriented measurements.

  At the same time, the EMA’s safety review of modafinil does indicate regulators’ preference for objective measures of patient benefit, given risks of adverse side effects. We can also see this preference in the FDA’s clinical review history for pitolisant: “DNP [Division of Neurology Products] advised that efficacy for EDS [excessive daytime sleepiness] should be supported by positive findings in two adequate and well-controlled studies on both an objective and subjective measure.… The Agency does have reservations about the use of the ESS, as it requires patients to answer hypothetical questions and is also subject to recall bias.”61

  In the clarithromycin study organized at Emory, participants reported a strong effect on the subjective ESS but did not experience improvements in their PVT reaction time, the declared primary outcome. That result was not a problem, given how the antibiotic was being used clinically. Trotti and Rye regarded it as a temporary stopgap, and clarithromycin was not destined for commercial approval as a sleepiness remedy anyway.

  Regarding outcome measures, the two trials for PTZ sponsored by Balance Therapeutics have taken a mixed approach. The first study, conducted from 2015 to 2018, included people with both narcolepsy type 2 and IH diagnoses, and any use of conventional stimulants had to be halted for the trial.62 The primary outcome was the subjective Epworth Sleepiness Scale, with the MWT a secondary outcome. The second study lasted from 2018 to 2020 and allowed participants to be taking other medications, including stimulants, at the same time.63 For this study, which included only people with IH diagnoses, Balance engaged the Michigan-based sleep researcher Tom Roth, who played a key part in establishing the MWT and MSLT.

  For the second study, Roth and colleagues interviewed IHers and developed a symptom diary as the primary outcome measure. It includes questions on brain fog, fatigue, and memory impairment. Before any results were announced, Balance presented analyses of patient data at the 2019 World Sleep Congress in Vancouver. The study team observed that the IH patients they recruited had high subjective sleepiness on the Epworth Sleepiness Scale but could stay awake during a MWT for a normal length of time. Without breaking the blind of the placebo-controlled study, the analysis revealed that some participants’ subjective levels of sleepiness and brain fog improved but their objective sleepiness did not. The investigators’ conclusion was that patients’ reports of sleepiness in IH related more to mental fatigue or daytime impairment rather than physiological sleepiness, as measured by tests like the MWT or PVT. This does not mean that study participants’ sleepiness was less real, only that the physiological tests currently used to measure it were inadequate.

  The question of subjective versus objective endpoints has also come up for other neurological disorders in which excessive daytime sleepiness and fatigue are prominent symptoms, such as multiple sclerosis and Parkinson’s disease. These are conditions for which modafinil is widely accepted but not FDA approved. Studies in Parkinson’s have shown stronger effects from modafinil on subjective versus objective sleepiness, and in both Parkinson’s and multiple sclerosis, modafinil has stronger reported effects on sleepiness than fatigue.64 Looking ahead, it highlights the need to develop standardized, well-accepted measurements for other aspects of hypersomnia, such as weekly total sleep time and symptoms such as brain fog and sleep inertia.

  CHAPTER 15

  KNOCK YOURSELF OUT

  Disease states will also define the essential scientific and medical audiences that must be persuaded if the drug is to gain legitimacy, whether through FDA approval or through other means.… Whenever a drug is developed for a certain disease, the various communities organized about this category quickly become the therapy’s most important constituencies.

  —Daniel Carpenter, Reputation and Power, 2010

  In 2018, Jazz Pharmaceuticals began organizing a clinical trial of a drug called JZP-258 for idiopathic hypersomnia.1 Jazz’s interest was a sign that part of the pharmaceutical industry was taking IH more seriously than before. Although small compared with pharmaceutical giants such as Roche, Jazz was a bigger player than Balance Therapeutics.

  Jazz’s drug was a low-sodium formulation of the sedative gamma-hydroxybutyrate (GHB), eventually named Xywav. It was a new version of a drug, GHB, which had been used to treat narcolepsy since the 1970s. Jazz had already begun clinical trials of JZP-258/Xywav for narcolepsy before initiating studies on IH.2

  In the narcolepsy community, Xywav’s predecessor Xyrem was well known. Xyrem had the distinction of being the first drug approved by the FDA specifically for the treatment of cataplexy in 2002. In the scientific literature, it was known as sodium oxybate or by its active ingredient: gamma-hydroxybutyrate.

  According to Jazz, around 15,000 patients were regularly taking Xyrem in the United States by 2020.3 Some, like Ann in chapter 12, swore by it. Others with narcolepsy were intimidated by the drug and vowed not to take it. One woman in Atlanta told me she cried herself to sleep her first night because she was so afraid of its side effects, yet she acknowledged its positive effects on her narcolepsy symptoms.

  The drug, mixed in water and consumed as a clear or opalescent liquid, was supposed to enforce the continuous, unfragmented sleep that people with narcolepsy tend to have trouble achieving. When taken in the evening, Xyrem quickly plunged recipients into deep, nonphysiological sleep,4 a state that could prevent them from hearing a fire alarm or getting up to go to the bathroom. Its effects were not sustained for more than a few hours, so they were supposed to take another dose in the middle of the night. The drug was only available from one central pharmacy in the United States, which shipped it directly to patients because of federal regulations. It came with rules that can constrain someone’s schedule: no food two hours before bed, and no combining it with alcohol.5

  For people with narcolepsy or IH, regularly quaffing a strong sedative may seem counterintuitive. Yet if someone lacks the feeling of restorative sleep, then it may make sense to push them into a state where some components of sleep occur, even if that state is artificial. We might compare the drug to a cast allowing a broken bone to heal.6 “I really thought for a long time that idiopathic hypersomnia patients should never try Xyrem,” Lynn Marie Trotti said in 2018.7 “People with idiopathic hypersomnia already sleep really well—like arguably, much much too well, and certainly much too long.”

  What changed Trotti’s mind—and was generally credited with opening her colleagues’ eyes to the possibility—was a 2016 publication from Isabelle Arnulf’s group in Paris. Arnulf found that her patients with IH responded to Xyrem as well as patients with narcolepsy type 1, often managing with a smaller dose.8 More than a third experienced side effects such as nausea and dizziness, and more than half stopped taking it after fifteen months. Still, for more than 70 percent of the patients with IH who tried it, Xyrem improved their severe sleep inertia, a debilitating symptom of IH that had been poorly studied at that point.

  A full dose of Xyrem (9 grams, spread out into two quaffs) contains more than 1,600 milligrams of sodium, more than the American Heart Association’s ideal limit for daily food intake. Some sleep medicine specialists had been concerned about cardiovascular health in people with narcolepsy, since they often rely on blood pressure–increasing stimulants during the day.

  Among those who had tried it, the drug was known for its soapy, salty taste. When the FDA gathered information on narcolepsy for its 2013 Patient-Focused Drug Development program, Xyrem was mentioned hundreds of times—often positively, because it controlled commenters’ symptoms. Others reported that the saltiness led to nausea and loss of appetite. One person wrote that the drug had “the taste of the bottom of the inside of a shoe after it has been worn all day.” A high school student wrote an entire essay on how “it tastes so bad any attempt to describe it would not do it justice.”

  Thus, a low-sodium version of Xyrem could be a more palatable and cardiovascular-healthy option to offer patients. But Xywav’s predecessor, like pentylenetetrazol, was another drug whose past cast a shadow over its present.

  THE LONG-SOUGHT SLEEP HUMOR

  The first to study GHB’s clinical utility was Henri Laborit, a polymathic French physician. One of his primary interests was anesthesia. In the 1950s, Laborit shared a Lasker Award for his part in developing chlorpromazine, the first antipsychotic drug. Laborit was interested in GHB because he thought it might function as a GABA precursor and facilitate GABA’s synthesis in the brain.9 GHB has a close chemical relationship to GABA (figure 15.1). Laborit favored GHB as a sedative because it induced hypothermia and because he thought it did not suppress respiration, unlike other sedatives or anesthetics. It also had a “spectacular action on the dilation of the cervix.” In the 1960s, GHB was used as a sedative during childbirth in France and Italy.10 GHB is not an analgesic—it has to be combined with something else to block pain signals.

  FIGURE 15.1. The chemical structures of GHB (gamma-hydroxybutyrate) and GABA (gamma-aminobutyric acid) are similar.

  The Toronto psychiatrist Mortimer Mamelak brought GHB to the attention of the sleep research community in the 1970s. In people with histories of insomnia and depression, the drug promoted both deep slow-wave sleep and REM sleep. In the Lancet, Mamelak speculated about GHB as “an attractive candidate for the long-sought sleep humour.” That is, the compound might be one of the elusive endogenous sleep substances researchers had been hunting for years.11

  Mamelak and his Canadian colleague Roger Broughton gave a report on GHB at the landmark 1975 narcolepsy meeting in France.12 Following Laborit’s reasoning, they proposed that GHB could stoke GABA production and alleviate fragmentation of nighttime sleep. They had tested GHB in four women, all with narcolepsy and cataplexy, adding banana flavoring to mask the unpleasant taste.

  With pie charts, Mamelak and Broughton showed how before treatment, the women’s sleep was spread around a twenty-four-hour clock, but by the fifth day, it was confined to the hours between 11 pm and 7 am. They reported dryly: “Daily irresistible sleep attacks and cataplexy disappeared, the patients were better able to cope with daily chores, and their mood improved.” At the time, Mamelak and Broughton were uncertain about how GHB worked. It might have to do with increasing GABA levels in the brain, but it might also stimulate dopamine synthesis or act through acetylcholine. They wrote: “The mechanisms of the apparent therapeutic effect of GHB in human narcolepsy-cataplexy remain obscure.”

  The 1975 conference proceedings contain an exchange between Broughton and Stanford’s Christian Guilleminault, who expressed skepticism about a GABA-based mechanism.13 More than forty years later, the field has not resolved the question of how GHB works.

  EXERCISING LESS RESTRAINT

  Mamelak and Broughton published several papers on their clinical experience with GHB over the next decade. A few other sleep specialists in Canada and France tried the drug with their narcolepsy patients, but uptake elsewhere was slow.14 A few people in the United States traveled to Toronto to obtain access to it.15

  In 1983, a sleep specialist in Cincinnati, Martin Scharf, obtained permission from the FDA to prescribe GHB for narcolepsy patients. He began doing so under a treatment IND (Investigational New Drug) program, which the FDA describes as “a mechanism for providing eligible subjects with investigational drugs for the treatment of serious and life-threatening illnesses for which there are no satisfactory alternative treatments.”

  What Scharf observed with the first thirty participants was not dramatic, when it came to measures of sleepiness. They were allowed to continue taking stimulants—and they still needed to, although not as much. However, GHB had a stronger effect on the other symptoms of narcolepsy: cataplexy, sleep paralysis, and hypnogogic hallucinations.16 It was tricky, since most had been taking antidepressants in an effort to control those symptoms, so withdrawal tended to make them rebound.

  When Scharf began experimental treatment of patients with GHB, he admitted them to the hospital for safety reasons. He recalled that some were emotional when he came to visit them in the morning.17 They surprised both Scharf and one another by not experiencing cataplexy. Some began testing their limits, seeking out situations that normally induced cataplexy by “exercising less emotional restraint.” For one person, playing the video game Pac-Man had been problematic, Scharf said. “In every single case, we’ve seen improvement,” he told Cincinnati magazine. “People who haven’t been able to work, who’ve been on disability for years, are going back to their jobs.”18

  Without a corporate sponsor or a government grant, Scharf ran an open-label study of GHB for the treatment of narcolepsy, which eventually included about 140 people and lasted for more than a decade.

  AN ORPHAN DRUG WITH SEVERAL PARENTS

  When Congress was holding hearings for the Orphan Drug Act in the early 1980s, the American Narcolepsy Association had mentioned GHB as a promising drug that was not being tested.19 A few years later, the FDA’s Office of Orphan Products—created by the Orphan Drug Act—provided important support for GHB’s development as a narcolepsy treatment.

  The Office of Orphan Products provided a grant to researchers at the University of Arkansas, who performed a small placebo-controlled study of GHB.20 Also, FDA officials approached several pharmaceutical companies about developing GHB. One abandoned the project after being acquired by a larger firm. The Minnesota-based firm Orphan Medical began developing the drug in the late 1990s. “We’re the guys who give a damn when no one else does,” Patti Engel, Orphan’s director of sales and marketing, said in a Minnesota newspaper’s profile of the company.21