According to a 2017 report from the Government Accountability Office, the FDA grants hundreds of requests similar to Anna’s every year, and the vast majority are single patients. Most are seeking access to anti-infective, antiviral, hematology, or oncology drugs, with relatively few coming in the areas of neurology, psychiatry, or anesthesia.11 The FDA has repeatedly taken steps to simplify the “complex and cumbersome” expanded access process, the GAO report says.

  Patients’ families sometimes mount social media campaigns to convince companies to grant them expanded access requests.12 With Roche management on board, Anna did not need to do so, but her request still ruffled some feathers inside the company. “There was no process for handling this type of request,” Baker wrote in his contemporaneous notes. “Some ‘old timers’ had never heard of Roche doing this sort of thing.”

  In August 2007, Anna wrote to Baker, saying that with flumazenil, “I felt awake for the first time in years. It felt incredible. I was animated, could focus, and simply felt alive,” and “We are desperate to get Romazicon [flumazenil’s brand name then] powder.” She outlined a scenario in which she would undergo surgery to implant a device to pump flumazenil into her body. She concluded: “In short, without Romazicon in some formulation, I will be permanently disabled.” In her letter, she also mentioned that the Emory team had begun to look into three other patients, including her youngest brother James, who might have similar conditions.

  “There was no way I could get enough pure flumazenil powder from Switzerland for three people, so I advised them to stay focused on Anna,” Baker said. He began to consider the “what ifs.” If success with one patient led to something bigger, who would own the rights? If something went wrong, who would be liable? Who would monitor for adverse events and report them to the FDA?

  According to Baker’s notes, “there was some debate about who was the Principal Investigator and whether she had the credentials to be such.” Despite not having a physician’s title, Parker ended up completing a master course in FDA and IRB paperwork. In a way, she was devising an experimental study for just one person, with no comparisons possible between Anna and any control group. For several years, Anna was the only person in the world taking flumazenil long-term for a sleep disorder.

  SLEEPING BEAUTY AT THE BACK

  In October, before the regulatory approvals had come through, Parker gave a lecture on Anna’s case at Emory. Anna sat in the back of the room listening but did not participate. It was the first time anyone from the Emory team had talked about Anna in public. Parker identified her as an attorney, using her first name only. “We have a sleeping beauty here,” Parker said. “She can’t stay awake for more than six hours. She can’t go out on a date. She had to take leave from her job.”

  Parker gave a name to Anna’s condition, endozepine-induced recurrent stupor, which embraced the endozepine terminology despite the controversy over cases in Italy. It also matched the language Anna used in her letter to Baker. Parker acknowledged Anna’s diagnosis of idiopathic hypersomnia but said it was only used “when we really don’t know what the issue is.”

  This blip of publicity generated concern when an article appeared in the university newspaper, leading to inquiries to Parker from People magazine. Nobody on the Emory team was ready for the media spotlight at that time. Rye and Jenkins didn’t want more attention before they could gather more information and define what was behind Anna’s condition.

  The flumazenil finally arrived in February 2008. Parker laughed while recalling how a tin with white powder in a plastic bag arrived at the Atlanta airport from Switzerland. She asked someone at the FDA, with whom she had been conferring for weeks, to send customs officials a message: it was not cocaine. To Parker, it seemed more like gold.

  The container made its way to Emory, where Parker had the pharmacy produce lozenges small enough to slip under the tongue. Anna came back to the hospital for another trial. This time, Anna had T-shirts emblazoned with the word “Annazam” (analogous to diazepam and midazolam) made for her and her family. If her doctors believed that her sleepiness was caused by a benzodiazepine-like substance, she would embrace it. But Kathy Parker was nervous. “The night before we started, my greatest fear hit me,” Parker said. “What if it doesn’t work? It was a sickening thought.”13

  At the hospital, Anna watched television and knitted, again with EEG electrodes on her head. The effect was subtler than the first time around, but she stayed awake for hours. Parker felt a sense of relief.

  Over the next couple days, Anna discovered that having the level of flumazenil in her system swing too drastically made her nauseated or groggy. Parker asked the hospital pharmacy to mix up a cream containing the drug, to be applied to Anna’s forearms. The cream was originally blue, making Anna remark that she felt like a Smurf.

  This combined regimen seemed to do the trick. It evened out the fluctuations and helped her wake up in the morning. During the day, Anna kept a supply of lozenges in a bracelet around her wrist, so they could be taken as needed. Having spent so much time in bed over the past year, she was physically weak and initially had trouble walking across a parking lot without feeling tired.

  In April, she went back to work part-time and told others she was gaining strength and able to get by with ten or eleven hours of sleep. She could watch entire movies or join friends for dinner for the first time in years. On top of that, she could drive, something that hadn’t felt safe when sleep was such an overpowering presence in her life.

  BREAKING THE MOLD

  In May 2008, Kathy Parker accompanied Anna to Roche’s headquarters in New Jersey to thank employees there. An internal newsletter lists employees who had a hand in overcoming logistical and regulatory obstacles: forty people in ten departments. Talking with them was one of Anna’s first chances to tell her own story. “The people at Roche helped save me. You have given me back my friends, my family, my work, my hobbies and my ability to drive, which, to a Southerner, is no small thing,” Anna told the group. Parker was effusive, saying: “Watching Anna get her life back has been the most fulfilling thing I have ever experienced. I have had grants, awards, papers published, and more, but this was the peak experience of my entire career.”

  Despite the heartwarming tone of the occasion, Baker was not eager to spread the news beyond his company. When he was dealing with the Emory researchers, Baker asked his Roche colleagues whether they were interested in repurposing flumazenil for sleep disorders. Despite its life-changing effects with Anna, they declined. “After the luncheon in Nutley, someone asked me if we were going to put out a news release or article in the Newark newspaper,” Baker told me. “While it would show a generous side of pharma that few people are aware of, I explained that we could not. First, Roche did this for Anna because it was the right thing to do—not for public relations. Second, if we did, the FDA would view the article as an off-label promotion of the drug. And third, it would drive more people to send letters to the CEO to ask for it.”

  Back at Emory, David Rye had a set of concerns that overlapped Baker’s. He grumbled about what he perceived as Parker’s victory lap. She asked him: “Why aren’t you more excited about this?” “We need to look at the larger picture,” he recalled telling her. “It’s very likely to be true for other people besides Anna. There’s a lot more to this.”

  Rye wanted to figure out how Anna’s case fit in with established categories—or perhaps broke the mold. He had high hopes for flumazenil and the test for GABA-enhancing activity. He thought those two tools, despite their limitations, could redraw the map of sleep disorders. A new category might include people who conventionally belonged in several baskets: idiopathic hypersomnia, narcolepsy type 2, or others. The patch clamp assay—or some refined version of it—could indicate whether someone belonged in the new category and would be likely to respond to flumazenil.

  The Roche newsletter contained clues about what was going on at Emory. One was a shift in language, suggesting a search for an appropriate name for the new category; Parker said she preferred to call Anna’s condition “endogenous hypersomnia” rather than “recurring stupor,” the term she had used in her lecture. The term “endogenous hypersomnia” was not found in the International Classification of Sleep Disorders and resembles “essential hypersomnia,” a term used by Japanese researchers to describe cases of excessive sleepiness distinct from narcolepsy.14 The newsletter mentioned thirteen probable endogenous hypersomnia patients besides Anna. That number had increased from three since a year before, when Anna wrote to Roche asking for help.

  At the Emory sleep center, then based at a faded former retirement home, Rye had been sorting through patient charts and digging through freezers to retrieve samples from patients he had already examined. He was conducting clinical research on a shoestring, without nursing staff or dedicated space. Recruited patients came in on Saturdays in order to avoid his regular clinical schedule. Meanwhile, in the laboratory, Jenkins and sometimes Garcia were testing patient samples via patch clamp. They wanted to know: how many sleepy people had GABA-enhancing substances in their spinal fluid, like Anna did?

  FUTURE DIRECTIONS

  Kathy Parker’s time at Emory was waning. Leaders at the University of Rochester had started recruiting her to become dean of nursing, a position she began in August 2008. Once Anna had recovered, she felt like her role in the hypersomnia story was complete. “I’m not a basic scientist,” she said. “I was less sure about what needed to happen next.”

  Rye had more of an inkling about future directions. Although he gave no public lectures on Anna’s case, like Parker did, through documents generated around this time, it is possible to discern his plans. Rye had two main goals: test flumazenil in other people with sleep disorders resembling Anna’s and identify the somnogen, the GABA-enhancing substance in patients’ cerebrospinal fluid. The first project got off the ground; the other ran into difficulties, which are explored in chapter 6.

  In March 2008, after Anna started her lozenge and skin cream regimen of flumazenil, a provisional application for a patent was filed on behalf of Parker, Rye, and Jenkins. The full application came a year later.15 While it is full of technical language and legalese, the patent also reveals something about the inventors’ scientific thinking. It contains the basics of Anna’s story, recipes for flavored flumazenil lozenges, and a plan for a clinical trial. Casting a wide net, the patent covers the use of flumazenil and several other GABA-receptor antagonists to treat the broad category of “excessive sleepiness and sleep disorders associated with excessive sleepiness.”

  The “endogenous hypersomnia” patients Parker mentioned when visiting Roche show up in the patent, too. Two besides Anna had idiopathic hypersomnia, and two had narcolepsy without cataplexy. Five others had “medically refractory sleepiness,” a term suggesting either (1) diagnosis with sleep apnea or restless leg syndrome, but having persistent symptoms; or (2) subjective sleepiness unmatched by falling asleep quickly enough in a Multiple Sleep Latency Test. One woman had hypersomnia that fluctuated along with her menstrual cycle, and another man had the rare episodic Kleine-Levin syndrome, with a disoriented mental state lasting days or weeks at a time. Even though other characteristics separated them, Rye reasoned that these patients’ conditions all reflected an increased need or drive for sleep. Grouping them together showed Rye’s willingness to color outside the boundaries set by the rest of the sleep medicine field.

  All of these people had levels of GABA-enhancing activity in their CSF that were higher than controls, according to Jenkins’s patch clamp assay. For five of them, the patent and a 2009 National Institute of Neurological Disorders and Stroke grant application16 include reaction time measurements, before and after intravenous flumazenil.

  Remember Anna’s complaints about the “world’s most boring video game”? Out of the five, her reaction time actually improved the least with flumazenil. One of her comrades had an average reaction time of almost two seconds—about a quarter of a second is typical. We can interpret this as intermittently nodding off or being almost asleep while taking the test. The higher dose of flumazenil chopped one and a half seconds off this person’s average reaction time, even though it didn’t change her subjective sleepiness much. Her response on the 1–7 Stanford Sleepiness Scale went from 6 (“Sleepy, woozy, fighting sleep; prefer to lie down”) to 4 (“Somewhat foggy, let down”), while Anna’s changed from 6 to 1 (“Feeling active, vital, alert, or wide awake”).

  Besides Anna, who were these sleepy people? For the sake of their privacy, we don’t need to know everything about them, but since Rye was taking a conceptual leap in putting these people together in the same group, knowing something about them might help us understand the category he was outlining.

  PRACTICE WHAT YOU PREACH

  As with restless leg syndrome, Rye felt a personal connection to what he was seeing in Anna and others. Around this time, he was diagnosed with hypothyroidism, a common condition that can drain someone’s energy and produce excessive daytime sleepiness. And as an empathetic physician, he was touched by his patients’ struggles.

  One was a nurse working at Emory, whom he first encountered in 2001. In this book, she is called Valerie.17 Rye has said that she had helped reorient his thinking. At first, he looked at Valerie’s sleepiness through the prism of restless leg syndrome. But after talking with her several times, he thought that some other explanation was necessary. “I was always telling other doctors to listen to their patients,” he said. “In this case, I realized, maybe I should practice what I preach.”

  Sleepiness seemed to run in Valerie’s family. “That was my normal,” she said. She had often taken long naps as a child, but her sleepiness intensified after she reached adulthood and gave birth to her daughter. When her daughter reached high school, her daughter also often fell asleep in class and needed an Adderall prescription to manage in college.

  Valerie initially met Rye through her mother, whom he diagnosed first with RLS and then later with narcolepsy type 2. Her mother had leg pains that led her to walk on a treadmill at night. In addition, she drank coffee throughout the day yet frequently nodded off, exclaiming: “It’s hard enough to wake up once a day. I don’t want to do that twice.”

  While living in Florida in the 1990s, Valerie regularly had to drive across Tampa Bay and would often have to stop at a familiar convenience store to take a fifteen-minute nap and load up on caffeine and sugar so that she could continue somewhat safely. Once while driving on Interstate 75, she fell asleep for several seconds and woke up just in time to swerve when traffic stopped in front of her.

  Valerie’s RLS symptoms were sporadic, becoming more intense every few months. Even after treatment with pramipexole under Rye’s supervision, sleepiness continued to be a problem for her. She found she could manage weekend shifts as a staff nurse, when she’d stay moving and on her feet, but a regular five-day work week, along with taking care of her daughter, would leave her exhausted. To cope, she tried modafinil, but it gave her heart palpitations. “Even coffee made me uncomfortable,” she said.

  When Valerie participated in a test of intravenous flumazenil with Rye, she felt more awake and her reaction time measurements improved as well. “It was like realizing that day-to-day I am really living in somewhat of a fog, and the flumazenil made everything seem sharper and clearer,” she said.

  FAMILY CONNECTIONS

  In addition to Valerie, another connection suggesting a genetic basis for the new sleep disorder category was Anna’s brother James, who seemed to have a condition like hers. James has also participated in Emory studies; his and Anna’s parents, along with a third sibling, have not reported similar issues.

  For both James and Anna, their excessive sleepiness arrived around late adolescence. James noticed his around the time when his sister’s difficulties were intensifying. He recalled occasional instances of what he called “sleeping heavy” throughout college: having intense dreams that seemed to last years, then waking up feeling tired. For a short time, he stayed with Anna in Atlanta when she was struggling with her job as an attorney, and he witnessed her crashes. Hearing Anna discuss her own epic dreams and perpetual drowsiness made James wonder if he had something similar.

  James, a Teen Jeopardy! contestant, has a record of academic achievement similar to his sister’s. In 2003, he graduated from Yale, where he studied linguistics and learned to read several languages, including Greek and Turkish. His path after university led in a different direction. He taught himself how to use animation software and began making elaborate, phantasmagorical videos. Collaborating with a friend from school, he created videos to accompany an album by an experimental rock group. He then traveled around the country to present the video at music and art festivals.

  After moving to the Northwest, James sprained his ankle. He found that he was “sleeping heavy” more often, which he initially attributed to being less mobile. By 2008, in Portland, his sleepiness had intensified; he began dipping his head into a bucket of cold water several times a day. It got the blood flowing for a short time, at least.

  In Atlanta, James had taken (and sometimes lost) various jobs as a pest exterminator, rideshare service driver, or bookstore clerk. He managed with conventional stimulants, which he had begun taking in college for attention deficit disorder. But he found that those medications didn’t help him regain the ability to concentrate. The change caught him off guard, because he thought he generally needed less sleep than his peers. He said in an email: “Additionally, I’d always been very self-driven with art, and I found from 2008 on I was incredibly dull-witted and sluggish and could produce very little.”