By Jonathan P. Fischer, MPH, and Philip R. Fischer, MD, DTM&H

Jonathan Fischer is a medical student at the Pritzker School of Medicine at the University of Chicago. Dr. Philip Fischer is Professor of Pediatrics, Department of Pediatric and Adolescent Medicine, Mayo Clinic, Rochester, MN.

Jonathan Fischer and Dr. Philip Fischer report no financial relationships relevant to this field of study.

SYNOPSIS: Pediatric Autoimmune Neuropsychiatric Disorders Associated with Streptococcal Infections (PANDAS) is a proposed disorder first named in 1998 that has been the subject of controversy in the literature. Although the debate has waned somewhat in the United States, it remains a topic of discussion in many European medical journals. A recent systematic review found no evidence for antimicrobial or immunomodulatory treatment for this condition.

SOURCE: Sigra S, Hesselmark E, Bejerot S. Treatment of PANDAS and PANS: A systematic review. Neurosci Biobehav Rev 2018;86:51-65.

The diagnosis of PANDAS is based on a set of criteria first proposed in 1998.1 These criteria include: obsessive compulsive disorder (OCD) and/or tic syndrome; prepubertal onset; abrupt onset with an episodic course; and association with group A streptococcus (GAS) infection (although this was not a uniformly required diagnostic criterion).

Discussion has been raised about these criteria, and other broader disease processes have been proposed that include non-OCD psychiatric symptoms, such as eating disorders, as well as other infectious etiologies beyond GAS. These alternative labels include Pediatric Infection-Triggered Autoimmune Neuropsychiatric Disorders (PITAND), Childhood Acute Neuropsychiatric Symptoms (CANS), and Pediatric Acute Onset Neuropsychiatric Syndrome (PANS). Each term was proposed with its own set of diagnostic criteria, but they share common features, such as recurrent, episodic, and acute exacerbations of a psychiatric illness with a presumed autoimmune etiology. There is not yet consensus on these definitions, and the variety of terms and criteria perhaps have contributed to the complexity surrounding this disorder.

The proposed pathogenesis of PANDAS and its related disorders is similar to that of Sydenham’s chorea. It is hypothesized that antibodies made in response to an infection cross-react in the basal ganglia and cortex to cause psychiatric disease. Because this proposed etiology differs significantly from typical psychiatric illnesses, a plethora of non-psychiatric treatments has been used to attempt to treat cases of suspected PANDAS. Last year, a consortium of clinicians and researchers released a series of practice guidelines, which recommend treating PANDAS and related disorders with standard psychiatric treatment, immunomodulatory therapy, and antibiotics.2,3,4 The systematic review by Sigra and colleagues attempted to summarize the current data on treatment modalities, including those that are part of the new guidelines.

The review included studies that applied diagnostic criteria for any of the four proposed disorders (PANDAS, PITAND, PANS, and CANS). The authors found 12 papers that included diagnostic criteria and treatment or outcome data. There were four randomized controlled trials, one cross-over trial, two open trials, four observational studies, and one survey study. Sixty-five case series or case reports also were included.

Sigra et al summarized the findings of each of these studies. Three treatment studies that evaluated antibiotics were included. Data showed that antibiotics were not effective as prophylaxis against future symptom flares, but that they did reduce psychiatric symptoms during an acute flair using one measure (Clinical Global Impressions Scale-Severity of Illness), but not another (Children’s Yale-Brown Obsessive Compulsive Scale). One treatment study evaluated plasma exchange therapy, which showed significant improvement over placebo at one year. Two treatment studies investigated intravenous immune globulin (IVIG) therapy. One study found that IVIG effectively reduced symptoms, while the other study found no effect over placebo.

There are ample studies about cognitive behavioral therapy (CBT) in typical OCD, but only two studies specifically about OCD in patients with a PANDAS, PITAND, PANS, or CANS diagnosis. These two studies found that CBT effectively reduced symptoms. There also are observational studies, survey studies, case reports, and case series that described patients showing improved symptoms after receiving all of the above treatments as well as nonsteroidal anti-inflammatory drugs, corticosteroids, selective serotonin reuptake inhibitors (SSRIs), and no treatment.

It is interesting to note that the one case report that published the result of providing no treatment found that the patient experienced spontaneous remission. An observational study of tonsillectomy found no difference in symptom resolution compared to no treatment. Finally, there are a number of other treatments that have been described that were given in conjunction with the above treatments, and so were not studied individually. These included attention deficit hyperactivity disorder medication, antipsychotics, anxiolytics, mood stabilizers, probiotics, omega-3 fatty acid, vitamin D, homeopathy, gluten-free diet, sinus surgery, and rituximab.

However, it is important to note that 11 of the 12 papers included in the systematic review and all of the case reports were found to have a moderate or high risk for bias. Each paper was ranked on several criteria. All but one of the included papers had at least one major methodological issue, such as no control group, inadequate description of randomization, small sample size, reliance on self-reported data, administration of multiple treatments at once, or variable doses of medications given in a treatment arm. The only study that the review authors ranked as not having moderate or high risk for bias found that IVIG is not beneficial compared to placebo at reducing symptoms.5

The authors of this new systematic review concluded that, based on the poor quality of the available data, there is insufficient evidence favoring medication and immunomodulatory treatment of PANDAS. They highlighted the need for future studies on this topic. They also underscored the importance of reducing bias in future studies if reliable results are to be attained.


This systematic review highlights an interesting topic. Although PANDAS is rare (and its existence is not even fully established), this new systematic review provides interesting lessons that are broadly applicable. Two key questions are raised that provide insight to clinical practice more generally.

The first question that is sparked by this review is what should physicians do when there are insufficient data to guide decisions. Sometimes it is necessary to base treatment options on educated guesses using our understanding of pathophysiology, even when data are not available. On the other hand, it is important that clinicians “do no harm” and avoid exposing patients to unnecessary risks. This tension is especially salient in the case of PANDAS where, to date, there is no causal evidence that this disease exists.6 As Sigra’s review highlighted, there also are no high-quality data guiding what medications are most effective for patients with the symptoms attributed by some to PANDAS. Furthermore, many of the above treatments have known risks and side effects. Should clinicians treat aggressively when supportive data are lacking, or should they focus on doing no harm?

Both possibilities have been argued in the literature. The fact that the review by Sigra et al included so many different experimental treatment methods is an indication that many clinicians have subscribed to the view of aggressive treatment. In fact, Sigra and colleagues ended their review by suggesting that in spite of the lack of available data, the guidelines to treat PANDAS with immunomodulatory therapy and antibiotics still should be followed.

Another commentary argued against this perspective.6 Instead, treatments could include more holistic approaches to patient symptoms that, as one author stated, understand the child’s symptoms from broader medical and psychosocial perspectives while considering tangible ways to overcome the problems.7 Although patients may lose the theoretical benefit of the experimental therapies mentioned earlier, they also avoid the known harms associated with them. And, CBT, even without medication, does have proven effectiveness for children with OCD. For those with tic disorders, habit reversal therapy, with or without medication, also can be effective.

The second question arising from Sigra’s review is how should studies with potential for bias be interpreted and applied. This review did an excellent job of scoring the included studies on the possible level of bias in their findings. Regarding PANDAS specifically, some have argued that the fact that it became a heated topic in the media and online forums perhaps has influenced researchers to conclude more readily that it is a distinct illness from OCD.8 (Even Swedo’s original 1998 paper noted the limitation of using many patients from a patient support/advocacy group.1) Thus, it is even more important to maintain a critical eye when appraising medical literature on this topic to assess possible ways in which bias could have been introduced. The fact that the only study included in this review that was found to have a low level of bias also found no benefit over placebo for the studied treatment is telling. Could it be that methodologic issues influenced the other studies to find benefit in treatments that actually are not effective? Until more rigorous studies are conducted, this question may remain unanswered.

In spite of these issues and the lack of consensus about the management of PANDAS, a full understanding of the current data can aid a clinician in providing relief to patients experiencing OCD and/or tics. A recent editorial aptly indicated that the important point is not the physician’s belief in the condition, but the physician’s understanding that the patient experiences distress from the severe acute-onset neuropsychiatric symptoms.7 Although some clinicians may disagree with the treatment recommendation of Sigra and colleagues, their systematic review provides a thorough understanding of the background of PANDAS and the current lack of evidence for medical treatment modalities. It also provides key lessons that can be applied to medical practice more broadly.


  1. Swedo SE, Leonard HL, Garvey M, et al. Pediatric autoimmune neuropsychiatric disorders associated with streptococcal infections: Clinical description of the first 50 cases. Am J Psychiatry 1998;155:264-271.
  2. Thienemann M, Murphy T, Leckman J, et al. Clinical management of pediatric acute-onset neuropsychiatric syndrome: Part I – Psychiatric and behavioral interventions. J Child Adolesc Psychopharmacol 2017;27:566-573.
  3. Frankovich J, Swedo S, Murphy T, et al. Clinical management of pediatric acute-onset neuropsychiatric syndrome: Part II — use of immunomodulatory therapies. J Child Adolesc Psychopharmacol 2017;27:574-593.
  4. Cooperstock MS, Swedo SE, Pasternack MS, Murphy TK. Clinical management of pediatric acute-onset neuropsychiatric syndrome: Part III—Treatment and prevention of infections. J Child Adolesc Psychopharmacol 2017;27:595-606.
  5. Williams KA, Swedo SE, Farmer CA, et al. Randomized, controlled trial of intravenous immunoglobulin for pediatric autoimmune neuropsychiatric disorders associated with streptococcal infections. J Am Acad Child Adolesc Psychiatry 2016;55:860e2-867e2.
  6. Shulman ST. Pediatric autoimmune neuropsychiatric disorders associated with streptococci (PANDAS): Update. Curr Opin Pediatr 2009;21:127-130.
  7. Hedderly T, Malik O. Childhood acute neuropsychiatric syndromes...furthering the discussion PANS and PANDAS. Eur J Paediatr Neurol 2018;22:223-224.
  8. Murphy TK, Kurlan R, Leckman J. The immunobiology of Tourette’s disorder, pediatric autoimmune neuropsychiatric disorders associated with Streptococcus, and related disorders: A way forward. J Child Adolesc Psychopharmacol 2010;20:317-331.