Timothy was 10 years old when his personality changed overnight. A concussion during a family ski trip in December 2016 left him unsteady on his feet, but that was just the first sign something was wrong. The strawberry-blond boy who played on the chess team and looked forward to Mandarin lessons became withdrawn, obsessive, and suicidal. Back home in Marin County, California, he said “bad men” had surrounded his family’s house and were trying to get him.
Timothy’s parents, Rita and John, took him out of school while doctors tried to decipher what was going on inside his head. (The family members’ names have been changed to protect their privacy.) Rita suggested that her son take up knitting to fill the time. Once he started, he couldn’t stop. Compulsive thoughts haunted him, and he refused to wear many of his clothes, fearing they were contaminated.
The boy’s doctors were stumped. Concussions can cause mood changes, but not like this. They ran test after test, searching for a diagnosis. When Timothy’s parents wrestled him into the car to take him to various clinics—for brain scans, blood draws, immunological workups—he told them he wanted to jump out onto the highway. “You’re not my mom,” he yelled at Rita. In March, he started leaving the house and running barefoot through the surrounding fields. His parents placed a go bag near each door with bottles of water and a walkie-talkie. When Timothy ran, his father would slip on sneakers, grab a bag, and run alongside him until he tired. Eventually, the couple hired a military veteran to keep an eye on their son day and night.
The tests kept coming back normal. Neurologists referred him to psychiatrists. Psychiatrists referred him back to neurologists. Pediatricians recommended therapists. Therapists suggested psychologists. In late March, with Timothy in a deepening depression, his parents and uncle made a plan: They would rent a car with no back doors, sedate him with Benadryl, and drive him overnight to the child psychiatric unit at UCLA.
Timothy stayed there for more than three weeks. The doctors prescribed Lexapro, an antidepressant, and steadily upped the dose. But the boy only became more agitated. It was as if an alien had crept into his body and stolen the real Timothy, Rita recalls. His intrusive thoughts suggested a diagnosis of obsessive-compulsive disorder; his mood changes pointed to a depressive disorder. Rita says one psychiatrist told her, “To be honest, he doesn’t really fit any category we have.”
While Timothy was undergoing treatment at UCLA, Rita spoke with a mom in the San Francisco Bay Area who worked with a support group for athletes who suffer concussion and brain injury. She told Rita that when a child’s concussion symptoms don’t go away, it’s sometimes because there is an underlying infection that disrupts the brain. Rita searched online and found a diagnosis that seemed to describe the full range of her son’s symptoms: pediatric acute-onset neuropsychiatric syndrome, or PANS. One possible trigger of the disease, she read, is infection with Streptococcus, the bacteria that cause strep throat.
Rita thought back to the winter. She couldn’t remember Timothy coming down with a sore throat, but just before the ski trip she had noticed that the skin around his anus looked a little red. She had put it down to irritation. But strep, she read, can cause rashes like that. She asked a neurologist at UCLA whether PANS might be making her son sick. The reply shocked her. “That’s a made-up disease,” she recalls the doctor saying. According to Rita, the UCLA team wanted to keep Timothy in the hospital and continue giving him antidepressants. She and John had watched their son become less and less like the boy they knew. They made a plan to get him home.
A couple of days after returning to Marin, the family met with a chiropractor in San Francisco who specialized in the treatment of neurological disorders. Chiropractors are not medical doctors, but by this point Rita and John were ready to speak with any professional who might be able to help. Rita mentioned the rash, and the chiropractor seemed to confirm her research: Timothy, he said, had a subset of PANS called pediatric autoimmune neuropsychiatric disorder associated with streptococcal infections, or PANDAS. If the bacteria were still there, circulating in the boy’s bloodstream, the first step toward alleviating his symptoms was to knock them out.
The chiropractor arranged for a doctor he worked with to write a prescription for azithromycin, an antibiotic used to treat strep. Rita had her doubts. She’d told other doctors about the skin irritation; why hadn’t any of them diagnosed Timothy with PANDAS? But the risks to her son were low, and she figured they might as well try.
Two days later, the boy was starting to become himself again. The bad men had disappeared. He wanted to go out for pizza and read his favorite sci-fi books. For the first time in almost five months, Rita and John recognized their son. The relief was immense, but it was tinged with uncertainty: If this disease was “made-up,” why was Timothy getting better? Would the improvement in his condition last? And the biggest question, the one that would dog the family well into Timothy’s adolescence: When doctors disagree on the cause of an illness, where does that leave the patient?
Up until the 1980s, psychiatry in the United States was still a quasi-Freudian undertaking. If a child developed tics or obsessive-compulsive disorder, the thinking went, it must be because her parents were emotionally frigid or had punished her during toilet training. (Mothers were also blamed for a number of other conditions, including autism.) So when a pediatrician named Susan Swedo joined the National Institute of Mental Health in 1986, she was delighted to be part of a new vanguard. Her mentor there, Judith Rapoport, was challenging the prevailing theories and seeking a medical explanation for OCD.
A few old papers in the literature had piqued Rapoport’s interest. They concerned a childhood illness that causes tics in the face, hands, and feet. Patients jerk their limbs in a strange and uncontrollable dance; their tongues flicker; their fingers seem to hammer the keys of an invisible piano. Thomas Sydenham, the 17th-century English physician who first described the condition, called it Saint Vitus’ dance, after the dancing manias that emerged in continental Europe during the Black Death, when large groups of people, sometimes thousands at a time, would cavort in the streets until they collapsed from exhaustion. He attributed the cause to “some humor falling on the nerves.”
Not until the 1930s did scientists discover that children suffering Saint Vitus’ dance, now known as Sydenham’s chorea, had something else in common: Their blood contained antibodies for Streptococcus. Left untreated, the pathogen can cause acute rheumatic fever, a serious autoimmune disease of the heart, joints, and skin. Sydenham’s chorea, it turned out, is the neurological manifestation of acute rheumatic fever.
Scientists have yet to work out exactly how one progresses to the other, but the theory goes something like this: Pathogens and antibodies in the bloodstream generally have a tough time getting past the tight-knit barrier of cells and blood vessels that protect the brain. But some Streptococcus seem to carry a secret key. They are believed to secrete toxins that open up the blood-brain barrier, allowing antibodies in. The antibodies try to grab hold of the distinctive sugar-protein clumps on the exterior of the bacteria—but, in a stroke of evolutionary bad luck, some brain cells wear similar clumps. Unable to differentiate friend from foe, the antibodies attack both. The worst damage occurs in the basal ganglia, the part of the brain that controls habits and movement.
Rapoport had found that children with OCD showed increased activity in the basal ganglia. And when she looked at the case reports of patients with Sydenham’s chorea, she discovered that many had developed compulsive thoughts and obsessive behaviors weeks before their tics began. Was it possible that this autoimmune disease, an illness of the body, was triggering illness in the brain? If you could cure one, would the other disappear?
Over the next few years, Swedo and her colleagues treated a number of children with Sydenham’s chorea and OCD. Most had already tried the standard neuropsychiatric medications, but the drugs didn’t seem to work. The next step was to see whether the standard autoimmune treatments had any effect. Some of the kids received intravenous immunoglobulin, which can help reboot the immune system with a mixture of antibodies from healthy donors. Some underwent plasmapheresis, a process in which all of the patient’s blood plasma is run through a filter. They seemed to improve.
Around this time, the first case of what would come to be known as PANDAS was referred to Swedo’s lab. The patient, an 8-year-old boy, had begun flailing his arms seemingly at random, and he had trouble speaking. His doctor suspected Sydenham’s chorea, but Swedo and her colleagues ruled it out. To them, his symptoms looked like OCD. The flailing wasn’t a physical tic; it was a mental compulsion. The boy was trying to fling bad germs away. He hoarded pieces of paper in a brown bag and refused to swallow his saliva because he feared it was contaminated.
Swedo has credited the boy’s mother, a medical technologist, for drawing the crucial connection. She told Swedo that her other son—the patient’s elder brother—had Tourette’s syndrome. She’d noticed that his tics were worse when he had a sore throat, so she’d started taking swabs and culturing them in her lab. Sure enough, they sprouted colonies of Streptococcus. The same was true of the younger brother; his strep infections and OCD symptoms waxed and waned in tandem. This raised a new possibility: You might not need a full-blown infection to trigger mental illness. Something as minor as a sore throat could be enough.
By 1996, Swedo and her colleagues were feeling confident enough of the strep-tic link to give the condition a name: PANDAS. Then, in 1998, they published a paper in the American Journal of Psychiatry laying out the diagnostic criteria, with the goal of developing “treatment and prevention strategies.” They spent the following year working on those strategies, publishing case studies on various therapies—immunoglobulin, plasmapheresis, and prophylactic doses of antibiotics to reduce the severity of strep-triggered neuropsychiatric symptoms. (Swedo, who retired from the NIMH in 2019, is still active in PANDAS work. She did not respond to numerous interview requests.)
Some of Swedo’s fellow researchers were skeptical. Stanford Shulman, an expert on Streptococcus who edited the journal Pediatric Annals for 14 years, called the evidence for PANDAS “tenuous at best.” More than two decades later, he finds the data even less convincing. For one thing, he notes, strep infection is extremely common in children, accounting for as many as a third of all sore throats—but you don’t see droves of kids with abnormal behavior crowding emergency departments and psychiatric clinics in late winter and early spring. What’s more, strep antibody levels can remain high for months after the infection is gone. “That creates a huge background noise,” Shulman says. “If a child develops PANDAS symptoms and by reflex some doctor draws anti-strep antibodies, they’ll say, ‘Oh look, there’s elevated titers!’”
As Shulman and others see it, that doesn’t constitute enough evidence to prescribe the child antibiotics, much less intensive immunological treatments. “If a child has psychiatric symptoms, that child needs psychiatric care,” he says. Conventional psychiatric drugs and talk therapy are backed up by decades of robust scientific evidence. This is not true, he notes, of the typical PANDAS therapies. The long-term use of antibiotics especially worries him, because it could contribute to the problem of drug-resistant bacteria. This is still the mainstream position: The most recent edition of the Red Book, an exhaustive guide to childhood infectious diseases published every three years by the American Academy of Pediatrics, goes out of its way to recommend that children with PANS and PANDAS symptoms not be given an extended course of antibiotics.
In 2010, after more than a decade of controversy, Swedo convened a group of colleagues to revisit the PANDAS diagnostic criteria. Physicians, patients, and their families had been “left confused” by the scientific shouting match, they later wrote in the journal Pediatrics & Therapeutics. Sick kids weren’t getting treated; researchers were having a tough time designing and funding rigorous studies. The group’s solution was to scrap the taboo letters in the PANDAS acronym, the ones standing in for “autoimmune” and “associated with streptococcal infections.” Rather than naming the condition for its supposed cause, they’d name it for its presentation in patients.
The clearest and most common characteristic was rapid onset: A kid could be himself one day and a stranger the next. That became the centerpiece of the new name, PANS, or pediatric acute-onset neuropsychiatric syndrome. The diagnosis was meant to be broad, allowing for a range of possible triggers—infection with strep or another microbe, environmental factors, metabolic disorders. PANDAS, in other words, wasn’t going away; it was just becoming a subset of the larger syndrome.
Swedo and her colleagues included a handful of children’s drawings in their paper, made before, during, and after the kids got sick. One triptych is especially moving, a course of illness in miniature. The “before” image shows a dark-haired woman in a teal cocktail dress, her cat-eye makeup meticulously rendered. The “during” image, drawn in the midst of a flare-up, feels addled by comparison. There are no colors or recognizable figures, just squiggles and disembodied eyes. The “after” image shows a girl in a red-striped shirt and sunglasses. She’s standing beside the Eiffel Tower, smiling.
About a week after his visit to the chiropractor in San Francisco, Timothy was sitting in a clinic at Stanford’s Lucile Packard Children’s Hospital telling a trio of board-certified doctors about his hellish ordeal. His life had been turned upside down, he said, but a few days of antibiotics had made him feel himself again. By this point, he told me, he thought of physicians as “clueless.” He’d have won a gold medal in the “100-meter ditch-your-doctor dash,” he said. He had been poked and prodded, his brain scanned, his mind trawled. The medical establishment had belittled his parents, and he felt he had been misdiagnosed and mistreated. So he expected these doctors might dismiss him, too. Instead, the head of the clinic, a rheumatologist named Jennifer Frankovich, promised that her team would help. (Although Frankovich and I are both employed by Stanford, our work has never intersected.)
Alongside Frankovich were Margo Thienemann, a child and adolescent psychiatrist, and Theresa Willett, a pediatrician with a PhD in immunology. The three doctors weren’t shocked by Timothy’s spiral into despair, the sudden psychiatric symptoms and the personality change. They weren’t surprised that psychiatric medicines made him feel worse and that antibiotics made him feel better, or that a multitude of doctors had not been able to offer a single, conclusive diagnosis. This was classic PANS, they said.
The doctors started Timothy on a new course of antibiotics. (Frankovich says she’s always reluctant to prescribe them, though some kids end up needing antibiotics for years.) They also gave him anti-inflammatories and intravenous steroids. Rita felt hope for the first time that year.
Frankovich had started out as a PANS and PANDAS skeptic. As a resident in Stanford’s pediatrics training program in the early 2000s, she’d given a presentation on an article that questioned the link between OCD, tics, and strep. She fell in line with the mainstream thinking, which attributed the disorders to faulty wiring in the brain. Then, in 2010, she met a 13-year-old girl who suffered from the autoimmune disorder lupus. The girl had endured years of treatment with steroids and other harsh medications, including an immune-suppressing drug called CellCept. The side effects had been horrible: Her cheeks ballooned and her belly became distended. Finally, though, she had gone into remission.
But when Frankovich began to taper the dose of CellCept, the girl became depressed; she found it hard to read, remember, and think. Frankovich upped the CellCept and started intravenous steroids again. “Right in front of my eyes, all those mental illness symptoms melted away,” she says. The same thing happened with a 10-year-old boy who had an inflammatory disease of the spine. Overnight, he had developed OCD and tics. Frankovich called his pediatrician and mentioned Swedo’s research. The pair treated the boy with steroids. His symptoms melted away. To Frankovich, this suggested a sobering possibility: Thousands of sick kids around the country were being treated with psychiatric medications while the underlying cause of their illness—inflammation—went unnoticed.
Frankovich cobbled together an informal clinic with two psychiatrists. Starting in 2012, the trio worked overtime, tagging hours onto their already packed schedules. They borrowed clinic space and a clinical coordinator from the rheumatology department and called themselves the Neuro-Psychiatric-Immunology Clinic, a non-catchy mouthful of a name chosen deliberately to avoid attention. They didn’t want controversy, and they certainly didn’t want to put the names of contested illnesses in their clinic’s title. The hypotheses they were testing placed them at the margins of their disciplines and at odds with mainstream medicine.
Frankovich shared her research and patients’ stories at small medical meetings around the country, in hopes that more doctors might consider treating PANS and PANDAS. But as she spoke, some of her peers would stand up and walk out of the room. Others approached her afterward. “What are you doing, Jenny?” she recalls one saying. “Why are you pushing this nonsense?” asked another. Word travels fast through the world of pediatric rheumatologists. At the time, there were fewer than 300 such doctors in the United States. Frankovich felt like an outcast.
In 2014, a story about one of Frankovich’s patients made the pages of a local newspaper. Other doctors had diagnosed the little girl with bipolar disorder, but the Stanford team treated her for PANS, and she’d made a dramatic recovery. The article, Frankovich says, marked “a very low point in my career and life.” It brought on a renewed wave of criticism, which was bad enough. Even worse, Frankovich says, it gave hope to vastly more patients and families than she and her colleagues would ever be able to treat. “We got absolutely crushed with phone calls and emails and people just showing up,” she recalls. “It was a nightmare.” But the article was also a turning point: Frankovich soon got an offer of support from the hospital’s chief operational officer. She requested a clinic room and a half-time coordinator.
As the calls and emails kept coming, Frankovich’s team would sift through thousands of medical records, looking for patients with the clearest-cut cases of PANS. She estimates they were able to treat one in 10 patients who applied, if that. They met families who had sold their cars and refinanced their homes to pay for their children’s medical care. Many said, like Rita, that Frankovich’s clinic was the first place they felt hope.
Doctors have been proving other doctors wrong for millennia. Established credo has been overturned many times, only to be replaced with new information and new beliefs about science and medicine. In the 19th century, perhaps one in five British men who were admitted to a mental hospital suffered what was then called general paresis of the insane, a crippling condition that ended in delusions of grandeur, paralysis, and death. As the poet Kelley Swain writes in The Lancet, the Victorians considered it “a disease of dissolution and disrepute,” more moral than biological. We have a different name for the disease now, neurosyphilis, and a treatment, penicillin. But in the decades it took for medical science to cross that threshold, people were left to suffer in shame without proper treatment.
Many PANS patients and their families feel stuck on the wrong side of the threshold. “The system is not there for them in the same way it is for other illnesses,” Frankovich says. She points out that a child undergoing treatment for a brain tumor gets access to a specialized ward and a team of medical professionals and social workers. “But when a kid comes in with a mental health deterioration and their brain MRI is normal,” she says, the support network “walks away from them.” The families become so desperate for treatment, Frankovich adds, that “they can appear very dysfunctional and disorganized, and they can be very aggressive with trying to get their child help.” (Several PANDAS skeptics declined to be interviewed for this story, saying they feared online harassment.)
Jonathan Mink, a pediatric neurologist at the University of Rochester Medical Center, attributes the heightened emotions to a mismatch between what families want—an answer, a treatment—and what medical science is equipped to provide: “Some people come up to me and say, ‘I know you’re not a believer in PANDAS,’ and I say, ‘It’s not about believing in PANDAS. I believe in the data, and right now the data on PANS and PANDAS is inconclusive.’” He adds, “The underlying hypothesis is reasonable, but the data is very mixed. So how do we approach things when we physicians are uncertain?”
Stanford Shulman, the early PANDAS critic, also stressed the need for better data. “Should all older adults take an aspirin once a day? Because that was dogma for a long, long time,” he says. “But then studies came along in the New England Journal of Medicine, very large studies demonstrating no benefit and potential side effects, so we do have to change our mind.” He adds, “If we’re proven wrong, and really proven wrong, then we have to change our opinions, and that’s true for all medicine.”
For the past several years, Frankovich has been trying to raise money and recruit patients for a comprehensive, long-term study of PANS, which would follow 600 children for as long as 12 years. “We need proper funds to provide the kind of robust evidence that could end the controversy,” she says. “My colleagues have applied for NIH grants to study PANS and PANDAS, and despite their proven accomplishments they failed to get government funding. So how do we provide the evidence that this is real?”
There are PANS and PANDAS programs at a number of respected academic institutions, including Dartmouth, Massachusetts General Hospital, and—as of recently—UCLA. Most lack enough resources to study the thousands of children who walk through their doors. Late last year, though, a wealthy couple donated $2.4 million to Frankovich’s clinic to fund the completion of a “biorepository.” For now, Frankovich is focused on stocking the biorepository with blood and tissue specimens; she is also collecting MRI and EEG scans and sleep study data that might reveal the disease’s pathways through the immune system.
One concern Frankovich and Shulman share is that some practitioners try to capitalize on the medical and scientific uncertainty. They open up cash-only clinics and dangle the promise of a cure. “They promote this idea that this is an easily fixed problem,” he says. “They find a sizable market, and they give therapies that have never been demonstrated to be beneficial. That’s dangerous.” At Stanford, Frankovich says, “we’ve been studying the disease for eight years, and we never promise a cure.”
Timothy likens the fighting doctors to gods in a Greek myth, doing battle on Olympus while the mortals down on the ground try to survive. He believes that the PANS treatments have worked for him and that they could do the same for other patients. “I love debate—I’m in the debate club—and I love science,” he told me. “So give people the help they need and keep debating.”
Timothy is 14 now. He plays drums in a rock band and acts in school plays, but he still struggles to make sense of his ordeal in the early months of 2017. “I’m seeking absolution,” he says, a process that requires ongoing talks with a therapist. He has been on antibiotics since his diagnosis at age 10, although Frankovich has significant concerns about long-term antibiotic use and is collaborating with a pediatric infectious disease specialist who is managing Timothy’s treatment. In the winter of 2019, Timothy suffered a significant recurrence of symptoms, including obsessive thoughts about contamination, which his doctors said were possibly triggered by another strep infection. At that time, he says, flare-ups of differing severity occurred as often as once a month. Today, he and his parents consider him in remission.
Likening doctors to gods, whether those gods are beneficent or truculent, seems fitting for a profession that rules over the bodies of the dying, the voiceless, and the vulnerable. Margo Thienemann, the child and adolescent psychiatrist at the Stanford clinic, teaches her patients’ parents how to engage with the people who choose this profession: Don’t put them on the defensive, don’t appear aggressive or hostile, don’t lead with your own diagnosis, don’t say the words PANS or PANDAS when you first see a doctor, she tells them. “Doctors are used to knowing the answer,” she says. “Doctors are people who got all the good grades in school; they want to be the ones to figure everything out. So if you go in saying, ‘I’m telling you my child has this,’ then the doctor will feel like, who’s the doctor here?” She advises families to “stick to the symptoms, stick to the presentation. If the doctor can come up with the answer, then that’s better for everyone.”
But while we wait for proof to possibly sway opinion, children are in limbo, stuck between highly qualified doctors who say these diseases are real and equally qualified doctors who say they are made up. Timothy gushes over Frankovich and the help she offered when he was scared, wary, and sick. She’s a rare kind of god, he says. “That god goes down and tries to help the people. That god is extraordinary.”
Let us know what you think about this article. Submit a letter to the editor at mail@WIRED.com.