Rye had little to offer people like James, Valerie, or others who were struggling with sleepiness and found available medications unsatisfactory. It gnawed at him. However, Anna’s arrangement with Roche and the FDA was considered exceptional, and the supply of flumazenil available outside of that channel was limited. It took several years before Rye began surmounting these obstacles.

  CHAPTER 4

  RYE VERSUS MSLT

  This is a somewhat roundabout way of saying that the sleep-onset REM period contains the essence of all that we have previously called narcolepsy. Conversely, we would suggest that those patients without cataplexy or sleep paralysis who also fail to show sleep-onset REM periods in laboratory tests probably do not have narcolepsy.

  —Dement, Rechtschaffen, and Gulevich, “The Nature of the Narcoleptic Sleep Attack,” 1966

  When David Rye began his scientific and medical career, he was not set on becoming a champion of under-recognized sleep disorders. His traits of being scholarly yet stubborn and independent of established authority emerged early on.

  Rye grew up in Detroit and attended an all-boys Jesuit-run high school. He still avidly follows the Detroit Red Wings hockey team and cites his Jesuit education as important for learning how to be a critical thinker. The son of an industrial engineer who worked in the automotive industry, Rye studied chemistry on a merit scholarship at Wayne State University in Detroit, and he lived at home until his senior year. To make ends meet, he worked sweaty twelve-hour shifts at a Ford assembly plant one summer. During his PhD studies in Chicago, he also worked as a bouncer at a bar in Chicago called the Hangge-Uppe. Given his height, we can imagine him peering down at bar patrons coming in from nearby Division Street.

  When Rye started in the University of Chicago’s MD-PhD program, his initial attraction was to neuroanatomy, rather than sleep per se. In the early 1980s, the field of neuroscience was expanding rapidly, and the now-elaborate picture of which brain cells use what neurotransmitter was starting to unfold. For his PhD research, Rye chose to train with the pathologist Bruce Wainer, who was developing tools that were making it possible to map circuits within the brain with greater precision than ever before. Wainer was a relatively junior faculty member who was able to give Rye attention. Working with Wainer, Rye’s fellow MD/PhD student Allan Levey had generated monoclonal antibodies, then a relatively new technology, allowing the labeling of neurons that synthesized the most well-known neurotransmitter: acetylcholine.1

  Rye joined the lab soon afterward. The young scientists had fun together, organizing golf outings in Chicago and on some Friday afternoons mixing drinks in the lab using the same blender they used for homogenizing brain tissues. “At first, it was just the two of us,” said Levey, who later became chair of the neurology department at Emory. “We spent an enormous amount of time simply looking at microscope slides. It was like a gold mine of information.”

  In rats, Levey was mapping the connections of the thalamus, a central structure located above the brainstem that acts as a gateway for sensory information. He was inspired by a 1984 paper by the scientific titan Francis Crick, who had turned from DNA to the problem of consciousness. How does a collection of cells conjure up a sense of self, with memory and selective attention? Crick proposed the reticular complex, a net of cells surrounding the thalamus, as an organizing hub, directing an “attentional searchlight.”2

  AWAY FROM THE PACK

  Wainer’s main interest was in studying Alzheimer’s disease. He was adept with antibodies and immunology but not primarily a neurology expert.3 Levey and Rye soaked up knowledge of the brain from others at the University of Chicago. They both cited the neuroanatomist Rainer Guillery, head of the neuroscience program and an investigator of the thalamus, as a teacher. “It was an environment where we could go where we needed to learn,” Levey said.

  At the time, a major theory driving research on Alzheimer’s was that problems with acetylcholine were central, and much attention was being put on the degeneration of acetylcholine-producing cells in the forebrain. The Alzheimer’s field wasn’t yet the funding behemoth that it would later become, but to avoid the crowd, Guillery suggested that Rye find a different region of the brain to specialize in. “He’s had to carve out his own niche,” said Levey, who became an expert on Alzheimer’s. “For me, it’s been easier.”

  Characteristically, Rye wanted to make his own way. Here, we can glimpse the origins of his interest in “how the brain keeps the lights on.” “I wanted to stay away from the pack,” Rye said. “I had access to a useful set of tools. Where was there another area of the brain that predominantly used acetylcholine?”

  The answer: the pedunculopontine nucleus, or PPN, a bundle of neurons in the pons, part of the brainstem. The PPN was thought to be a component of the reticular activating system, the network of neural circuits in the midbrain and brainstem that keeps us awake. Rye worked with slices of rat brain, staining the PPN with antibodies and tracing acetylcholine-producing neurons’ paths on microscope slides.4 In his thesis, Rye suggested that the PPN corresponds to neurons that initiate REM sleep, known to be sensitive to acetylcholine.

  Even after Rye had left Wainer’s lab, he would sometimes return and hold court, smoking cigarettes and giving advice on what current students should be pursuing. Stick to neuroanatomy rather than follow the latest trends, he told them. A fellow Wainer trainee recalled: “I thought he would not mind working in an area where others thought that something else was more important.”5

  Rye said he learned a great deal during graduate school from Clifford Saper, then a young faculty member at Washington University–St. Louis, who moved to Chicago in 1985. He visited Saper’s lab to learn neuron-tracing techniques, and they kept in touch. Rye sometimes asked for Saper’s help in editing and trimming his papers on neuroanatomy, but on one of those papers, Rye was stubborn. “I took out my red pen and cut the length in half,” Saper recalled. “The next version was still too long. I eventually ran out of red pens and told him: ‘OK wise guy, send it in.’ ” The reviewers wrote back: “This manuscript would have been wonderful in the time of Charles Darwin.”

  During his neurology residency, Rye worked under the supervision of Jean-Paul Spire, director of the University of Chicago’s sleep disorders clinic. Spire had wide-ranging interests in neurology and electrophysiology; in newspaper articles, he was depicted probing the brain’s responses to sounds to detect brain tumors and monitoring a patient’s EEGs and other neurological signs during epilepsy surgery.6

  Rye’s first research paper on human sleep was a case report on a young woman who had a stroke affecting her left pons. The location of her stroke provided a chance to test theories about the regions of the brain controlling REM sleep, which at that point had mostly been evaluated in cats by the French sleep researcher Michel Jouvet.

  Over several months in 1989, the paralysis on the right side of the woman’s body improved. Rye and Clete Kushida, a medical student, studied the woman’s REM sleep, supervised by Spire.7 They noticed a disruption of REM-sleep EEG patterns coming from the left side of her brain, although she continued to have rapid movements of both eyes.

  The experience deepened Rye’s interest in sleep neurology, although he did not focus primarily on REM sleep in his later research. Instead, he saw the junction between the study of movement disorders and sleep as attractive. As a resident, Rye worked in several Chicago hospitals, where he encountered a variety of patients with Parkinson’s and related movement disorders. Through these experiences, he developed a hunch that “Parkinson’s was unlikely to shut off when someone is asleep.” Rye’s later work on restless leg syndrome and REM sleep behavior disorder, when people act out their dreams, both grew out of that hunch.

  PARKINSON’S AND ATLANTA

  Rye first met the neurologist Mahlon DeLong, who would recruit him to Emory, on a visit to Baltimore as a medical student, when he was scouting out places to apply for residency. At Johns Hopkins, DeLong was in the middle of transformative research.

  In the early 1980s, several young people in California were poisoned by a contaminant in synthetic heroin, reproducing the symptoms of Parkinson’s.8 DeLong and his colleagues were using the same chemical to create a model system for Parkinson’s in monkeys. In this model system, symptoms such as tremors and muscular rigidity could be alleviated by surgery. This led to the reemergence of surgery, and later electrical brain stimulation, as tactics for the treatment of Parkinson’s.

  The symptoms of Parkinson’s were viewed, up to that point, as resulting from tissue damage that couldn’t be repaired: a loss of function. One of DeLong’s insights was that degeneration in one area of the brain, the basal ganglia, resulted in a gain of function, or excessive activity, in other areas: the subthalamic nucleus and globus pallidus. By restraining either with a surgical lesion or with electricity, a balance of signals could be restored.

  The late 1980s and early 1990s were a time of rapid expansion at Emory. Even before Science’s publication of DeLong’s monkey surgery experiments in 1990, leaders at Emory sought him out as a chair of neurology. When DeLong arrived, he brought with him several colleagues from Johns Hopkins, bolstering Emory’s neurology department. With dedicated funding from the American Parkinson’s Disease Association, Emory became a hot spot for the study of Parkinson’s and other movement disorders.

  Before the 1990s, Emory had provided a base for sleep researchers such as the psychologist David Foulkes, who studied dreams in children. Gerald Vogel, the first to publish observations on REM sleep in narcolepsy while at the University of Chicago, was based for many years at an Emory-affiliated psychiatric hospital, the Georgia Mental Health Institute. While there, Vogel conducted landmark experiments on the relationship between REM sleep and depression. However, neither was involved in clinical care for sleep disorders, which is what DeLong wanted to build up.

  DeLong remembered Rye as “bright and creative” and recruited both Rye and Bliwise to Emory around the same time. Rye also had an offer from Harvard, where his friend Saper moved in 1992. However, Atlanta and its airport looked like a better option for Rye. Convenient airport access could allow his wife, Catherine, to continue managing her family’s residential construction business in Chicago.

  Rye’s early emphasis on Parkinson’s made practical sense. The PPN, his favorite region of the brain, was emerging as a vulnerable area in neurodegenerative diseases such as Parkinson’s, and it was close to and connected with the basal ganglia. In Rye’s first few years at Emory, he was supported by an American Parkinson’s Disease Association fellowship, along with NIH grants overseen by DeLong and Bliwise.

  Rye had ambitions for himself as thinking and publishing ahead of his peers. As a role model, he admired Percival Bailey, a neurosurgeon and psychiatrist who taught at the University of Chicago in the 1930s. Bailey built the foundations of medical knowledge of brain cancers with his careful studies of anatomy and pathology. According to a colleague, Bailey “was feared because he was so blunt and outspoken.”9

  SNOW WHITE AND THE DWARFS

  In his first decade at Emory, Rye recruited several graduate students to work in his laboratory. They took advantage of the same monkey model of Parkinson’s that DeLong and his colleagues had used. In the lab, Rye would demonstrate his skills with animal surgeries but was also willing to pitch in with mundane tasks. When students and technicians needed to keep monkeys awake at night for experiments, Rye would join in, periodically flicking lights, playing the radio, or crinkling paper to make sure animals did not doze off.

  Among students, his lab had an intimidating reputation because his students, MD/PhDs especially, took a long time to complete their degrees. Yet one of Rye’s former graduate students, Gillian Hue, said she did not regret her experience training with him. Originally from Jamaica, Hue had started working as a technician with one of Rye’s collaborators and was unsure about graduate school. He encouraged her to apply, squeezing in her application after the deadline. She joined the band of Rye’s acolytes. For an opulent party at the Society for Neuroscience meeting one year, she and other members of his lab dressed up as Snow White and the dwarfs—with Rye as Snow White.

  Hue encountered a series of frustrations with her research in rats and mice on dopamine and the spinal cord—much of it went unpublished outside her thesis. She was more interested in a teaching career, but that did not lead to conflict with Rye. After finishing her doctorate, Hue went on to teach psychology at a local college and later returned to Emory. “I learned from him how to be a scientist,” Hue said. “He asks bigger questions than are possible to address in incremental papers. Sometimes that conflicts with how science is currently practiced.”

  THE NARCOLEPSY PENUMBRA

  After his first decade at Emory, Rye did not stay focused on the PPN. He was drawn to restless leg syndrome, and his personal interest in the condition converged with his knowledge of dopamine neurochemistry. But at the same time, starting from when Rye arrived in the early 1990s, his clinical services were in demand. Emory’s status as a referral center in the Southeast meant patients with a variety of sleep disorders were coming through the door.

  By the time Rye encountered Anna Sumner, he had been treating patients with narcolepsy for more than a decade. Some of them stuck in his mind, such as an airline pilot who had successfully managed without mishaps for twenty-five years. His condition only surfaced when the pilot wanted to move from shorter to longer flights, and when it did, it threatened to derail his career. When the pilot experienced cataplexy during a test in a flight simulator, his examiners thought he was having a heart attack. An observant cardiologist referred the pilot to Rye.

  In 1997, Rye, Bliwise, and others had compiled a survey of forty-one narcolepsy cases, including the airline pilot. About half had been diagnosed after the age of forty, and most with later-life onset did not have cataplexy. In some individuals, sleepiness appeared decades before they began to experience cataplexy.10 At the time, the Emory paper went against conventional thinking about narcolepsy, usually observed to have its onset during childhood or adolescence. “This has always been a theme in our clinical research,” Bliwise said. “We saw that the spectrum of patients with narcolepsy was large. We thought that the penumbra was where the action was.”

  While exploring this territory, Rye gradually became dissatisfied with the Multiple Sleep Latency Test, the daytime nap procedure used to diagnose narcolepsy and idiopathic hypersomnia, and, more generally, with the categories it generated. He later adopted a rebellious stance with respect to the International Classification of Sleep Disorders—the consensus guidebook for the sleep medicine field.

  To understand where Rye’s criticisms came from, we have to back up and away from Atlanta. The concept of IH has evolved at the edges of narcolepsy, and running through narcolepsy’s history has been a tug of war over how to define it.

  IS CATAPLEXY REQUIRED?

  In the first half of the twentieth century, the term narcolepsy was sometimes used loosely to refer to the symptom of overwhelming sleep, or “sleep attacks.” German-, French-, and English-speaking neurologists disagreed on whether narcolepsy should be considered a disease of its own or simply a sign of some other condition that made someone sleepy. Some mistakenly thought narcolepsy was related to epilepsy or originated in psychological conflicts, with dubious psychoanalytic interpretations.

  Among neurologists, encephalitis and traumatic brain injury were recognized as possible causes of narcolepsy. The distinctive symptom of cataplexy, which did not appear in everyone with narcolepsy, did suggest a neurological origin. But an unresolved question was whether cataplexy was necessary for a patient to receive the narcolepsy label. The Australian-born neurologist William Adie wrote one of the clearest arguments for one side of this debate: “To my mind cataplexy in narcolepsy is as characteristic of the disease as the sleep attacks themselves. Given a case with sleep attacks alone the diagnosis is difficult, for sleep attacks indistinguishable from those of true narcolepsy occur as a symptom in many dissimilar diseases; but if definite cataplectic attacks are present as well the diagnosis is made certain, for the combination is seen in no other condition whatsoever.”11

  Others publishing around the same time thought narcolepsy should be defined more broadly, based on short attacks of sleepiness plus other symptoms. The British neurologist S. A. Kinnier Wilson dismissed the idea that narcolepsy should be thought of as one disease and referred to “the narcolepsies.”

  In the mid-1950s, neurologists at the Mayo Clinic solidified the clinical picture, describing the “narcoleptic tetrad”: excessive daytime sleepiness, cataplexy, sleep paralysis, and hypnogogic hallucinations. However, all four symptoms were not needed for a narcolepsy diagnosis. In the Mayo compilation of 241 people with narcolepsy, only a few displayed all four, and more than 30 percent didn’t have cataplexy.12

  The discovery of REM sleep transformed the debate. Pioneer sleep researchers such as Vogel, along with William Dement, described as the “father of sleep medicine,” recognized that when people with narcolepsy went to bed at night, they tended to enter REM sleep immediately or within a few minutes.13 This didn’t occur every time people with narcolepsy were studied in the laboratory, but rapid REM onset rarely occurred in healthy people, who usually took about ninety minutes to enter REM sleep at night.14