A Feverish Debate

Are common bacterial infections making children mentally ill?

Illustration by Gracia Lam

One morning in July 2011, five-year-old Sammy Kennedy woke up before dawn. He was surly and withdrawn, and he did not smile once, which was unusual. He trudged through his day, out of sorts. In the evening, when his mother, Margo, asked him to tidy up his Lego pieces, he blew up. He screamed, head-butted her, and dug his nails into her arms. He growled and spat on the living room floor. When she went to the bathroom, he followed her and pounded and kicked at the door. His rage lasted almost four hours.

These behaviours persisted for three more days. Sammy, a strong boy even at five, charged at Margo and punched her. He yelled at her to shut up, something he had never done before. She found his actions incomprehensible. He had always been such a happy, giggly, even-tempered kid. “It was like someone took Sammy and replaced him with another boy,” she says.

Margo, a tall, heavy-set woman, is a single mom in her early fifties. She adopted Sammy, who is Inuit, at birth while she was living in Yellowknife, and the two had since moved to Edmonton to be closer to family. Her own mother counselled her to use a firmer hand. Her sister chalked up Sammy’s behaviour to not having a man in the house.

But Margo and Sammy’s home had always been orderly and cheerful. The TV was turned off when Margo said so, and toys got put away at night. No one hit, or fought, or used bad words. Both she and her dimple-cheeked son had always been patient with each other. She did not know what had turned things upside down, but she was confident it was not her parenting.

On the fifth day, Sammy woke with a high fever, around 39°C, and the outbursts stopped. To Margo, it now made sense: he had been sick. After a few days, when his fever would not stay down and he started complaining of a sore throat, she took him to a medical clinic, where he was tested for a strep infection. It was positive, and he was treated with a course of antibiotics. Soon he was himself again.

Just a few weeks later, however, the rages resumed. Margo feared that she might get seriously hurt. Sammy also developed unusual habits: he washed his hands incessantly and filled sinks with soapy water. When she emptied the water, he became furious. Then his fever returned, climbing to 40°C, and a doctor prescribed another round of antibiotics.

This time, although the fever went away, the bizarre behaviours did not. By early September, Sammy was routinely biting, kicking, and punching his mother. He screamed at her in the car, admonishing her not to drive too close to other vehicles or make a turn when he did not want her to. More than once, he unbuckled his seat belt and grabbed the gearshift. He began to clench and unclench his fists, rub his fingers together, and roll his eyes up and to the side. He continued to wash his hands obsessively, and Margo found a cache of soap, deodorant, toothpaste, and air freshener hidden in his desk drawer.

She knew little about his biological family’s history. She is a social worker, though, and used to troubled children. She began to fear her son might have a serious mental illness: “I really thought, ‘He’s going to be institutionalized. He can’t live like that in society.’”

In mid-September, she took him to see a pediatric specialist and laid out the entire ordeal: the rages, the anxiety, the tics, the handwashing, and the strep. She showed him video clips from her phone of Sammy attacking her, screaming in the car, and rubbing his fingers.

The doctor listened, and watched the videos through twice. Then he said he thought he might know what it was, a condition he had read about but had never seen before. It had been identified about a decade earlier by researchers in the United States who had observed that following a strep infection, some children developed obsessive-compulsive behaviours, tics, and emotional problems, seemingly out of the blue. It was still controversial, he admitted, but Sammy did seem to fit the pattern of the illness, which was known as pediatric autoimmune neuropsychiatric disorders associated with streptococcal infections, or PANDAS.

According to those who believe it exists, PANDAS can suddenly turn healthy children into basket cases. The hallmarks are rapid onset, and extreme emotional instability marked by screaming rages or uncontrollable sobs. “Some families have told us that their children seem possessed,” says Susan Swedo, an American physician who was part of the team that first described and named the condition. Children often exhibit classic signs of obsessive-compulsive disorder, such as repeatedly checking that they have locked a door. Or they display the tic-like symptoms of Tourette’s syndrome: blinking, shaking their heads, clearing their throats, grunting. In still other cases, they stop eating, because they find the texture of food repulsive or fear that it might be contaminated.

A pediatrician by training, Swedo has spent most of her career as a salaried researcher with the National Institute of Mental Health, outside Washington, DC. She joined NIMH in 1986 to study OCD alongside Judith Rapoport, a world leader in the field. Rapoport was by this time convinced that OCD was a brain disorder with a biological cause, but most experts still thought it was a psychological consequence of bad parenting (like overly strict toilet training), and that it was best treated with psychotherapy.

In addition to working on the OCD study, Swedo began looking into a rare movement disorder called Sydenham’s chorea, an autoimmune reaction that affects the brain and is brought on by a prior streptococcal infection such as strep throat; it causes jerky, uncontrollable writhing movements, mostly in children. Many of the chorea patients reported OCD symptoms, which often arose days to weeks before the movement problems began. She began to wonder if, like Sydenham’s chorea, OCD was triggered by strep.

Several cases suggested that she might be on to something. There was the fourteen-year-old boy who had displayed mild OCD for over a year, but was referred to NIMH when he suddenly became incapacitated by it following a confirmed bout of strep throat. Swedo prescribed six rounds of plasmapheresis, a procedure that involves removing the patient’s blood plasma, treating it, and transfusing it back in. She had used this therapy successfully with Sydenham’s chorea, and in the boy’s case the OCD pretty much disappeared within two weeks. An MRI scan of his brain done before treatment revealed inflammation in part of the basal ganglia (which are also involved in Sydeham’s chorea). After the fourth round of plasmapheresis, the area had shrunk by 25 percent, back to its normal size.

Then there was the thirteen-year-old with a long history of Tourette’s, OCD, and hyperactivity. Again, a strep infection dramatically worsened his symptoms. When Swedo treated him with intravenous immunoglobulin, or IVIg, infusing him with antibodies pooled from healthy blood donors, his tics diminished significantly.

Another case from the early ’90s stands out for Swedo. A ten-year-old boy suddenly exhibited flailing arm movements, hoarding behaviours, and fears about being contaminated. At the same time, his older brother developed a strep infection. The brother had lived with Tourette’s for years, and their mother, a medical technologist, had observed that whenever it became more pronounced he inevitably came down with strep throat.

Between 1987 and 1997, Swedo documented 144 instances in which an infection led to sudden emotional and behavioural changes in a child, accompanied by an onset or worsening of OCD or movement disorders. In forty-five cases, strep was confirmed by a throat swab; sixty more involved a sore throat or upper respiratory illness, but throat cultures had not been taken; and in a further six, children had been exposed to others with confirmed strep. In 1998, Swedo published her landmark paper in the American Journal of Psychiatry, proposing that this condition, named PANDAS, resulted from an immune response to strep gone rogue.

To lay people, it may sound outrageous: a common childhood infection can lead to an immune reaction that attacks the brain. But Swedo’s theory was not new. Her hypothesis was similar to the explanation for rheumatic fever. She thought then, and still thinks now, that PANDAS might be a variant of that disease.

Rheumatic fever is now best known for causing inflammation that damages the mitral valve of the heart, but until the late 1700s it was a fairly benign illness with fever, chills, aches, and transient joint problems. Symptoms were not so well documented in the eighteenth century, acknowledges Peter English, a retired pediatrician and medical historian at Duke University in North Carolina who has studied the history of the disease. However, doctors’ records indicate that it was considered a common, non-fatal condition.

That changed during the nineteenth century, he says, when deadly cases began to appear in the medical literature. Autopsies suggested that some patients died of pericarditis, an inflammation of the membranous sac surrounding the heart. By the mid-1800s, post-mortems revealed inflamed mitral valves, and by the end of the century it was understood that the heart muscle itself was the target of attack. What’s more, rheumatic fever revealed other characteristics: in the 1830s, patients began to demonstrate signs of Sydenham’s chorea, which had been identified in the late 1600s but had not been connected to rheumatic fever; by the 1880s, the movement disorder had become an accepted part of the symptom constellation for rheumatic fever.

English theorizes that rheumatic fever has changed over time. It was not that heart problems went undetected until stethoscopes were introduced (they did not come into widespread use until the 1840s); nor does he think that chorea was simply overlooked. “It is unimaginable,” he says, “that an affliction that required a straitjacket, prevented ambulation, eating, and drinking, and distorted facial and bodily features could be easily missed when treating a patient for rheumatism.” Rather, rheumatic fever itself has been a moving target. He speculates that streptococcal bacteria have evolved, inciting attacks on different bodily tissues.

By the 1920s, it is estimated that tens of thousands of people in the Western world were dying each year of the disease, mostly from damage to heart tissue. Mortality was especially high among children; only tuberculosis and perhaps pneumonia took more lives. The disease had a long and disparate checklist of symptoms, but the cause remained unknown. Even after doctors noted that tonsillitis, scarlet fever, and sore throat often preceded it, decades passed before anyone put forward a convincing disease theory.

In 1931, Alvin Coburn, a physician at Presbyterian Hospital in New York, proposed that rheumatic fever has three phases: The first is infection with group A strep. The second is a silent period of about two weeks, during which the immune system produces antibodies. The third occurs when symptoms appear and the antibodies meant to attack the strep set upon the joints, the heart, and the brain.

When he suggested this idea, the evidence was circumstantial, but other researchers would go on to bolster his case. It is now widely believed that some individuals are genetically vulnerable, and that only particular strains of strep will result in the illness. To this day, though, rheumatic fever is diagnosed primarily on the basis of symptoms rather than by lab tests. Still, few dispute that it is a bona fide disease. That battle has been decided.

We tend to think of new diseases being “discovered.” But they are not discrete entities, like ancient pots buried in the sand, waiting to be unearthed, says Jacalyn Duffin, a physician and medical historian at Queen’s University in Kingston, Ontario. Even with the right combination of clinical observation and science, there may be no ultimate truth to be determined. She argues in her book, Lovers and Livers: Disease Concepts in History, that diseases are just ideas formulated to make sense of the evidence: the symptoms patients report, the details doctors observe or find through testing, the data scientific research generates.

Upstart theories are seldom a shoo-in. Only when enough experts are persuaded about the meaning of symptoms, the cause and effect, and so on, can a disease enter the medical canon (for a time, at least), and that is where PANDAS stands now: in the thick of the debate.

Some theories never prevail; for others, the struggle for acceptance lasts decades. The belief held today, that rheumatic fever is a post-streptococcal autoimmune disease, took decades to formulate and test, and to win over the medical establishment. Sometimes, a new idea triumphs quickly. I am old enough to remember when peptic ulcers were blamed on nervous dispositions. Since 1994, however, the Helicobacter pylori bacterium has been the agreed-upon culprit. Barry Marshall, the Australian physician who helped to identify the bacterium and prove that it led to ulcers, was ridiculed for his theory at first, and went so far as to drink a petri dish of the cultured bacteria to substantiate his argument.

When Swedo formally proposed PANDAS in 1998, it was received with cautious enthusiasm. At the time, no credible competing theories existed about what caused symptoms such as Sammy’s, so other researchers began to investigate.

According to Swedo’s original paper, PANDAS has five key features: it affects kids; they display obsessions, compulsions, or tics; they present other neuropsychiatric oddities; their symptoms either start abruptly or have sharp worsenings; and a clear association has been established between those worsenings and infection with strep. Though passionate debates have been held about all of these criteria, which is commonplace in medicine, the real controversy concerns the final point: to what extent does the streptococcus bacterium play a role in triggering the illness?

This is critical because if Swedo is right, doctors will be compelled to change how they respond to kids who turn up with sudden-onset OCD, tics, and rages; it would be negligent not to swab them. Doctors would also have to consider prescribing long-term preventive antibiotics, according to Edward Kaplan, an infectious disease specialist at the University of Minnesota Medical School in Minneapolis. “Were this association definitely proven,” he wrote in Contemporary Pediatrics in 2000, “the management of patients with tics and obsessive-compulsive disorder (OCD) would dictate long-term antibiotic prophylaxis to prevent recurring group A streptococcal infections just as in patients who have had classic rheumatic fever.”

But a temporal link has yet to be definitively shown. In the absence of clinching evidence, doctors disagree about how to proceed. Some are not convinced that PANDAS exists. Those who are find themselves at a loss about how to treat it. Should they test for evidence of current or past infection and address that, or should they simply concentrate on the symptoms?

Sammy’s pediatrician opted for the latter approach. When he met with Margo and Sammy in September 2011 and suggested that the boy might have PANDAS, he explained that Sammy probably had inflammation in his brain. The doctor admitted that he was not an expert on the condition, and he referred the boy to a behavioural specialist, who put him on a six-month waiting list.

Not long after Swedo’s seminal paper was published in 1998, Marie Lynd Murphy and Michael Pichichero, pediatricians in Rochester, New York, decided to formally track the condition. Between 1998 and 2000, they saw twelve children with abrupt, explosive behavioural changes. All had obsessive thought patterns, mostly about germs and illness, and they were emotionally “labile,” or changeable. Two boys also exhibited recurrent tics.

All but one of the children were diagnosed with an active strep infection or had had one in the previous four weeks (the exception had just undergone a tonsillectomy following earlier confirmed strep), and all were treated with antibiotics for ten days. In every case, the symptoms improved dramatically within days and disappeared within three weeks.

In half, however, the behaviours returned, in one child six times, and the recurrence was always associated with a new strep infection. (These patients were not carriers, who always test positive but may exhibit no symptoms.) All were treated, usually with a class of antibiotic called a cephalosporin, and all showed quick improvement. No new instances of OCD or associated behaviours took place without strep infection, and no confirmed strep occurred in these children without concurrent behavioural symptoms.

The illnesses that developed alongside the psychiatric symptoms tended to be particularly mild. None of the children presented the classic signs of strep; few had open sores in their throats, swollen lymph nodes, or even fevers, and their throats were moderately red. A few times, when parents noticed a recurrence of behavioural problems, no strep was detected. Murphy and Pichichero told them to return in seventy-two hours for further tests, all of which came back positive.

The results were striking. Similar patterns were described in subsequent case studies and personal accounts elsewhere. However, at any given time, about one in five school-age children will have a strep infection, and many more will have evidence in their blood of past strep infection. (Testing for strep requires multiple samples, to determine patterns.) Skeptics argued that these findings could be coincidental: the children just happened to have strep when the behavioural symptoms appeared.

Murphy and Pichichero investigated that question in 2001. Along with their colleagues, they swabbed 814 children; 411 of them with sore throats had strep. But an infection, the results showed, did not make them more likely to develop PANDAS—at least, not if they were promptly treated with antibiotics, as were all of the children with active infections. As the researchers point out, and even critics acknowledge, the treatment might have prevented PANDAS from developing. (When the connection between strep and rheumatic fever was first being studied, in the early 1900s, antibiotics had not yet been invented, so infections went untreated; today failing to treat strep would be considered unethical.)

Another team of researchers, led by Loren Mell at Group Health Cooperative’s Center for Health Studies in Seattle, examined medical records collected between 1992 and 1999. They wanted to see if children who developed tics, like Sammy’s eye rolling and finger rubbing, and OCD, like his handwashing and hoarding, were more likely than tightly matched controls to have suffered from a strep infection beforehand.

They found that kids with these disorders had been diagnosed more than twice as often with a group A strep infection in the three months before onset, and that those with Tourette’s were thirteen times more likely to have been diagnosed with multiple cases of strep in the year before their symptoms first appeared. The authors stated that the study supported the PANDAS hypothesis, but skeptics remained unpersuaded; they wanted a direct line between infection and symptom exacerbation.

Researchers from six separate medical centres got together in 2002 to address the issue. For two years, they followed thirty-one children who met the criteria for PANDAS and fifty-three who did not. About every three months, they swabbed the subjects’ throats and tested their blood. In the end, there was no evidence that strep uniquely intensified the symptoms in PANDAS kids. Only six strep infections led to worse symptoms within two months, and all of them occurred in the control group, three in one child. The study, published in 2011, has been hailed as strong evidence against strep as the prime instigator.

This study has since come under fire, including criticism from its lead author, James Leckman, a psychiatrist at Yale University in New Haven, Connecticut. He now thinks researchers made mistakes in allocating subjects. The control group may have accidentally included PANDAS patients, and not all of the patients in the PANDAS group were necessarily real cases.

As it turns out, even before he saw the raw data from the study he had already undergone a change of heart, after talking extensively with a mother whose son had been diagnosed with PANDAS. Leckman had also treated a child with sudden-onset OCD. Like many of Murphy and Pichichero’s patients, the girl had not had a sore throat but tested positive for strep. Plasmapheresis improved her condition, and for a time she was prescribed prophylactic antibiotics. However, within three weeks of Leckman taking her off the antibiotics she contracted strep again and was back in hospital with severe OCD.

Leckman believes a genuine and distinct condition exists that involves sudden-onset psychiatric problems caused by infection. Whether strep is the only trigger, the inciting one, or one of many, he is still uncertain, but he is persuaded by the larger theory. At a pediatrics conference in Boston last year, he told an audience, “I’m what you call a PANDAS convert.”

Many physicians remain skeptical. Another problem is that PANDAS lacks a clear biomarker or a set of them. Type 1 diabetes, for instance, can be definitively diagnosed using a blood sugar test. Plenty of recognized conditions lack definitive markers—schizophrenia and Parkinson’s, to name just two—but having one can help an upstart disease theory win acceptance.

Madeleine Cunningham, an immunologist at the University of Oklahoma Health Sciences Center, has spent much of her career researching rheumatic heart disease. Her work has been key in demonstrating how surface molecules on strep bacteria mimic those on the surface of heart tissue, essentially providing a molecular explanation for how rheumatic fever leads to heart damage. She has also identified a target in the brain that is involved in Sydenham’s chorea. Since 1999, she has studied PANDAS, and this past spring she began marketing a panel of blood tests she claims can identify the condition.

She also worked with a group of researchers in Tel Aviv that exposed rats to strep, which led to behavioural changes that resembled those seen in Sydenham’s chorea and PANDAS. They succeeded in demonstrating that strep antibodies attached themselves to brain tissue. Another research team in New York induced PANDAS-like symptoms in mice by infecting them with the equivalent of a strep vaccine. They then produced the same symptoms in other mice by injecting them with antibodies from the first mice. When they removed the blood serum component that contained the antibodies, the OCD-like symptoms no longer arose. However, other researchers have not generated convincing symptoms in animal models. Moreover, critics of the theory are not convinced that what the Tel Aviv and New York teams observed in the rats and mice was indeed PANDAS.

By 2010, after more than a decade of disagreements, the PANDAS disease theory had reached an impasse. Opponents charged that primary care physicians were jumping on the bandwagon merely because it was an easy, appealing diagnosis, while proponents argued that doctors were ignoring compelling evidence.

A major sticking point seemed to be the criteria. What if a child had sudden-onset OCD and evidence of strep but none of the other detectable neurological oddities that Swedo’s criteria required? What if there were no abrupt symptom exacerbations and the condition just exploded and then stayed that way? What if the child showed all of the symptoms of PANDAS but no evidence of strep?

This last point in particular continued to provoke controversy. Some medical professionals complained that the criteria for strep kept changing. In Swedo’s earliest publications, she called the condition pediatric infection-triggered autoimmune neuropsychiatric disorders, or PITANDs, and cited such triggers as chicken pox and influenza; later she referred to it as PANDAS and suggested the trigger was strep only. Over the years, various proponents, including Swedo, have suggested that strep probably sets off PANDAS initially, but other types of infections might reignite it, by opening up the brain to residual strep antibodies.

In intellectual battles over disease theories, the recasting of boundaries is common. Indeed, if the history of rheumatic fever teaches us anything, it is that illnesses can be enigmatic, changeable, controversial, and exasperating. For decades, according to Peter English, doctors struggled with the disease’s contradictory signals and progressions. Some patients had rashes and some had fevers, but in no particular sequence. Some were uncomfortable and others in great pain, while others became mentally ill. Some responded to therapy, others got better on their own, and others died. Even after its trigger was known, the Jones criteria, a set of symptoms used to this day in diagnosing rheumatic fever, have changed four times since 1944.

In the summer of 2010, NIMH convened a meeting of forty-one experts, in an attempt to unify the warring sides. At the suggestion of the chair, Noel Rose, a specialist in autoimmunity at Johns Hopkins, it was agreed in advance that etiology of any kind would be dropped from the criteria. In fact, nothing definitive would be said about the cause of the illness. Instead, the experts would try to describe what the syndrome would look like to a physician confronted with it, and the meeting achieved some consensus that sudden onset set the condition apart.

A new definition was proposed that expanded on PANDAS: PANS, for “pediatric acute-onset neuropsychiatric syndrome.” Swedo hoped it would encourage physicians to investigate infectious origins for sudden-onset symptoms. “All you have to do is recognize the acute onset,” she says. “At that point, a definite diagnostic workup should be done.” For his part, Leckman lobbied for greater focus on the behavioural aspects. He also persuaded the group to drop tics as a major manifestation, and to add the strange eating restrictions many clinicians observed.

By the end of the two-day session, however, the discussion remained as polarized as ever. Harvey Singer, a professor of pediatric neurology at Johns Hopkins, considers PANDAS an interesting hypothesis but wants more data—in particular, definitive evidence of a causal link between strep and the condition.

Not long after the 2010 conference, a group that included Singer, Roger Kurlan from the Atlantic Neuroscience Institute in New Jersey, and Donald Gilbert from the Cincinnati Children’s Hospital Medical Center wrote a paper renaming the disorder CANS, for “childhood acute neuropsychiatric symptoms.” The report, which appeared in The Journal of Pediatrics, argued that CANS had numerous causes, and that they needed to be investigated. It advised that doctors faced with children showing sudden-onset psychiatric conditions should look beyond just evidence of strep. Further, the authors suggested that the PANDAS entity be “eliminated.”

When margo heard that it would take six months for her son to see a behavioural therapist, she was frustrated: it was too little, too late. She also thought therapy would be insufficient help for Sammy. “I remember thinking, It’s a medical problem, and I want medical treatment. We’re not waiting six months.” So she did what many parents do these days when they receive an unfamiliar diagnosis: she got on the Internet.

She believed that PANDAS fit Sammy to a T, and she discovered that Swedo was running a new clinical trial on IVIg for PANDAS patients. Swedo had already completed a trial on IVIg and plasmapheresis in 1999, which found that immunotherapy had significantly improved the symptoms of fourteen of seventeen kids treated with it. To address criticisms about possible bias, because she had both treated and assessed the patients, Swedo had launched a new five-year study in January 2011, just six months before Sammy got sick.

The trial would accept thirty-two children with PANDAS. Some would receive dummy IVIg, because researchers had to demonstrate that any improvement was due to more than just the placebo effect, but all of the children who continued to have symptoms would be offered a second round of the real product. While Swedo was overseeing the trial and treating the children, Leckman, at Yale—with no knowledge of who was getting IVIg or the placebo—would assess the degree to which her patients improved.

The day after they consulted Sammy’s pediatrician, Margo emailed Swedo to find out if her son qualified. After screening his case, the researchers accepted him into the study. Within a month, he was on his way to Washington, DC. By this time, his rages were occurring daily, and his tics and OCD had not abated. Margo was on leave from work, and she felt hopeful but anxious about the clinical trial: “I was thinking, What if his brain stays like this? What if the IVIg doesn’t work? Is this his lot in life?”

During the week he was at NIMH, Sammy underwent MRI scans, a spinal tap, and multiple blood draws. On the second day of treatment, Margo thought she noticed a change in his countenance. When later that day and all of the next he vomited and complained of a headache, typical side effects of IVIg, she became more convinced that he might have received the genuine treatment.

They flew home to Edmonton shortly afterward, and in a diary of his behaviour that first week Margo noted sporadic improvements: he stuck his tongue out less, shouted and screamed less, punched and hit less. She was seeing glimpses of the old Sammy—but he still had rages, tics, OCD, and severe anxiety.

Over the next six weeks, his progress became more concrete. His obsessions with handwashing and hygiene products disappeared. The meanness subsided. His anxiety remained high, though, and he could not attend school. At this point, the end of November 2011, they headed back to NIMH for the second treatment. After that round, the rages and meltdowns became even less frequent. To help with the lingering anxiety, he began taking citalopram, an antidepressant, in early December. On January 6, 2012, he had one last rage. Within another week, he was back to his happy, giggly self, and he returned to school, slipping back into grade one with surprising ease.

Margo felt a huge sense of relief. Then, just a month later, Sammy had another fever. Two days after that, she noticed that he was flicking at his neck and rubbing his finger under his nose. Over the next six weeks, a wide array of tics began to appear: he rolled his head around in circles, blinked, and pulled out his hair in individual strands. He walked almost sideways and seemed strangely uncoordinated. The sink filling and handwashing habits returned. Despite this, he seemed happy, and the rages did not resurface.

Nonetheless, Margo was alarmed. A throat swab in early March found no evidence of strep, but an ear nose and throat specialist suggested that Sammy’s tonsils and adenoids might be harbouring infection, and he suggested to Margo that they be removed. In April, Sammy had a tonsillectomy. Afterwards, the doctor told Margo that the tonsils had, as he feared, been infected and deeply embedded.

The boy’s recovery was horrible. He bled from the incision and twice had to be taken to the hospital for emergency surgery. Yet even in the middle of the crisis, Margo noticed that his tics and OCD were diminishing, and by the time he came home from the hospital they were mostly gone. Within ten days of that, they had disappeared, and they have not returned since.

Depending on where you stand on PANDAS, Sammy’s story can have one of two arcs. In the first, a little boy gets sick, a doctor recalls an obscure disorder, a mother presses for help, treatment is given, and the boy recovers. In the other, a little boy gets sick, no single treatment works consistently and quickly, and the symptoms wax and wane but resolve over time—by coincidence, right after a tonsillectomy. Luckily for Sammy, either way he is back to his old self.

Many Canadian kids who turn up at their doctors’ offices with sudden rages or tics or OCD are not so lucky. They take drug after drug, with little improvement. Some spend time in psychiatric wards. Some are treated with antidepressants, antipsychotics, sedatives, and cognitive behavioural therapies—every remedy doctors have offered—without being released from their symptoms.

To make matters worse, many are given a revolving door of diagnoses: OCD, Tourette’s (or “Tourette’s plus”), ADHD, oppositional defiant disorder, psychosis. It is easy to lose faith in the experts when there is so little consensus about these symptoms, their causes, and the appropriate treatment. (In the interests of full disclosure, a few years ago my own daughter suddenly developed obsessive thoughts and emotional rages immediately following a serious ear infection. Doctors variously explained it as anxiety, PANDAS, and lupus psychosis.)

Canadian doctors are still wary of PANDAS. Sammy’s pediatrician did not want his name used in this article. The names of physicians who will even consider the diagnosis are traded among families, in a manner reminiscent of how women tracked down abortionists in the ’60s. The Canadian Pediatric Society has no formal position on the condition. “More research is needed,” says Danielle Grenier, the society’s medical affairs director and a practising pediatrician. Yet even gathering evidence is discouraged: in 2009, Asif Doja, a neurologist at the Children’s Hospital of Eastern Ontario in Ottawa, proposed a surveillance study, to see how many kids actually had the disorder, but to date he has failed to secure funding. Oddly, despite this caution about the condition, IVIg is an approved treatment for PANDAS in Canada. Still, most parents who believe their kids have PANDAS eventually go to the US for help.

Doctors must walk a fine line. We expect them to be our gatekeepers, to spare us from quack diagnoses and useless treatments. We also expect them to weigh the harm of doing something versus not doing it—swabbing the throat, for instance, of a child who presents with sudden-onset psychiatric problems. Good doctoring means navigating a world of uncertain and evolving evidence. PANDAS has not won the disease theory war, but neither has it lost.

To be sure, many unanswered questions remain. If PANDAS exists, it is not a straightforward disease. Cunningham hopes her panel of markers will at least manage to identify those who suffer from it. Swedo hopes that her clinical trial will demonstrate that even if we do not fully understand the mechanisms behind the illness, IVIg may provide relief.

The Sammy I met in the summer of 2012 was symptom-free: no tics, no OCD, no rages. He introduced me to his Lego collection, and annoyed me with a loud toy gun. He could not fathom that I had never tried a 7-Eleven Slurpee, and instructed me earnestly on how to make one.

When I visited, his supply of prophylactic antibiotics from NIMH had run out. Proponents recommend that after IVIg, kids take regular preventive doses until age eighteen or nineteen, to prevent reinfection with strep and any relapses. For many reasons, including a legitimate concern about antibiotic resistance and the lack of evidence that such prophylaxis will prevent PANDAS recurrences, the idea is controversial, and Margo was having trouble convincing Sammy’s Edmonton pediatrician to prescribe them.

At night, after Sammy was asleep, she confided that she worries constantly about strep. “Not a day goes by that it doesn’t go through my mind that I hope it’s not this illness again,” she said. Then she laughed, and told me about an incident from just after he had his tonsils out, once all of the tics and the OCD and the raging had gone away. Sammy had whacked her. She immediately asked, “How’s your throat? ”

“It’s not PANDAS, Mommy,” Sammy replied, “it’s me.”

“Then you need to go to your bedroom for hitting Mommy,” she said sternly.

And he did.

The Walrus thanks the Writers’ Trust of Canada for its financial support of this story.

Since the June issue of The Walrus went to press, Sammy, now seven, suddenly became ill again. In April, he developed a fever and a sore throat. Around the same time, he also began to display an array of odd behaviours, including spitting, swearing, hand flapping, high-pitched wailing, separation anxiety, and rages. A throat swab for strep infection was negative.

This appeared in the June 2013 issue.

Alison Motluk
Alison Motluk is a former staff writer for New Scientist and a regular contributor to CBC Radio One’s Quirks and Quarks. She won a National Magazine Award for her 2011 Walrus feature, “A Political Meltdown.”
Gracia Lam
Gracia Lam counts The New Yorker and Maisonneuve among her clients.