Recent evidence suggests that obsessive-compulsive disorder (OCD) and Tourette’s syndrome (TS), both of which frequently begin in childhood, are in some cases caused by autoimmune-mediated cerebral inflammation that is triggered by a streptococcal infection. This conceptualization, which has been termed PANDAS (Pediatric Autoimmune Neuropsychiatric Disorders Associated with Streptococcal infections), suggests that immunotherapy may be a useful treatment for these disorders, and indeed this seems to be the case, said Rosario R. Trifiletti, MD, PhD, Assistant Professor of Neurology and Neuroscience at Weill Medical College of Cornell University, in New York City.

Speaking at the annual meeting of the American Academy of Child and Adolescent Psychiatry, Dr. Trifiletti noted that although the idea of using immunosuppressive drugs is somewhat unusual in psychiatry, neurologists routinely use these medications to treat disorders such as multiple sclerosis and myasthenia gravis. Thus far, only a single randomized trial has shown immunotherapy to be effective in the treatment of OCD and TS, but those findings, coupled with published case reports and clinical observations, suggest that immunological approaches may be a worthwhile option in severe cases of these disorders.

EVIDENCE FROM SYDENHAM’S CHOREA

It has long been known that there is a relationship between movement disorders and autoimmune disease. This relationship is perhaps illustrated most clearly by Sydenham’s chorea, a component of rheumatic fever. The manifestations of this dramatic disorder include motor abnormalities—ranging from a darting tongue to motor impersistence (the inability to sustain a position or posture)—as well as restlessness, emotional lability, and OCD-like features. A link between Sydenham’s chorea and streptococcal infection was established definitively a half century ago, when researchers correlated the onset of chorea with evidence of streptococcus in throat cultures. The finding that therapies targeting immune function, such as corticosteroids or antibiotics, greatly improved the symptoms of patients with Sydenham’s chorea provided further evidence that immune processes are important in the etiology of this disorder. Most patients recover completely without treatment, although the process often takes months or years; recovery is much faster if patients receive corticosteroids.

Sydenham’s chorea is probably caused by the formation of antistreptococcal antibodies that cross-react with one or more proteins in the brain, a phenomenon known as molecular mimicry. In the 1970s, in a series of groundbreaking studies, John Zabriskie, MD, and colleagues analyzed serum samples obtained from patients with rheumatic fever to look for antibodies directed against various brain regions. In samples from patients with Sydenham’s chorea, the investigators found IgG antibodies that selectively targeted neurons of the subthalamic and caudate nuclei. These antibodies also recognized streptococcal cellular components, a finding consistent with the hypothesis that Sydenham’s chorea is caused by molecular mimicry. “Importantly, antibody [titre] correlates with the severity and the duration of the chorea,” Dr. Trifiletti noted. Furthermore, these antibodies cross-react with group A ß-hemolytic streptococcus, the etiologic agent for rheumatic fever, but not with other streptococcus strains.

GENETIC VULNERABILITY

Although researchers have known for years that the likelihood of developing rheumatic fever varies with the strain of streptococcus involved, more recent research suggests that features of the host are also important, Dr. Trifiletti said. For example, family studies have shown that there is an inherited susceptibility to rheumatic fever following streptococcus infection. In order to evaluate differences in immune function that could contribute to this vulnerability (and thus to chorea), Dr. Zabriskie and colleagues developed a panel of monoclonal antibodies directed against a variety of B lymphocyte antigens; they then compared lymphocyte characteristics between normal individuals and those with Sydenham’s chorea. One antibody that differentiated the two groups recognized a B lymphocyte cell-surface antigen known as D8/17; this important immunologic marker is associated with a susceptibility to rheumatic fever. Patients who develop rheumatic fever following a streptococcal infection express D8/17 on the surface of about 30% to 40% of circulating B lymphocytes, compared with less than 10% for individuals with uncomplicated infection.

THE OCD CONNECTION

Other research, meanwhile, has revealed intriguing parallels between Sydenham’s chorea and symptoms of OCD, Dr. Trifiletti said. In the 1980s, investigators from the National Institute of Mental Health (NIMH) observed that a high percentage of patients with rheumatic fever also met diagnostic criteria for OCD; conversely, the NIMH group has reported that as many as a third of patients with OCD show signs of chorea. Moreover, the changes in mental status associated with acute Sydenham’s chorea are often similar to those of OCD (though the neuropsychiatric symptoms of Sydenham’s patients may resolve rapidly with improvement of chorea). These findings suggested the possibility that, as in Sydenham’s chorea, OCD might be triggered, at least in some cases, by a streptococcal infection.

Thus, Susan E. Swedo, MD, and colleagues at the NIMH formally proposed the PANDAS concept, focusing on the relationship between OCD, TS (which shares many biologic and etiologic features with OCD), and streptococcal infection. The investigators developed five diagnostic criteria for PANDAS: presence of OCD or a tic disorder (not necessarily TS); onset between age 3 and puberty; abrupt onset or dramatic exacerbations of symptoms; repeated episodes in which symptom severity suddenly increases and then decreases; and symptom exacerbation that correlates temporally with signs of group A ß-hemolytic streptococcal infection, such as a positive throat culture. Perhaps as many as 30% of these patients will exhibit soft neurologic signs, such as choreiform activity, which can help to support the diagnosis, Dr. Trifiletti said.

He emphasized, however, that a diagnosis of PANDAS cannot be made on the basis of a single evaluation: It is necessary to establish a correlation between multiple symptom exacerbations and streptococcus infections; it may be necessary to document symptoms for six months to several years before PANDAS can be diagnosed with certainty. It remains unclear what proportion of patients with OCD or tics meets the criteria for PANDAS.

Recent studies have demonstrated that the D8/17 B lymphocyte marker associated with Sydenham’s chorea is also found in patients with OCD and tic disorders. When D8/17 positivity is defined as having the antigen on at least 20% of one’s B lymphocytes, no more than 20% of normal individuals are D8/17 positive, compared with about 90% of patients with Sydenham’s chorea; a similar proportion of patients with OCD or TS is D8/17 positive as well, supporting the view that there is an immunologic similarity between the two disorders. In one study of children with PANDAS, 85% were D8/17 positive, compared with only 17% of healthy controls.

THE SEARCH FOR ANTIGENS

Dr. Trifiletti and colleagues have recently sought to determine whether patients with OCD or Tourette’s syndrome also produce antibrain antibodies and, if so, to identify the proteins recognized by these antibodies. Using serum samples from patients with TS and/or OCD, the researchers detected an antibody that targets a brain protein with an apparent molecular weight of 83 kilodaltons (kD). (A second group of investigators has since replicated the finding.) Although this discovery suggested a potential role for antibrain antibodies in the etiology of OCD and TS, it was not known whether the antibody is capable of crossing the blood-brain barrier. Therefore, samples of cerebrospinal fluid (CSF) were assayed from adult patients with OCD. Once again, the researchers found antibodies to the 83-kD protein; in contrast, no antibody was detected in CSF samples from normal controls or patients with multiple sclerosis. Furthermore, immunologic treatment significantly reduced the production of these antibrain antibodies in patients with severe OCD or tic disorders.

In a subsequent study, Dr. Trifiletti and colleagues found antibodies against the 83-kD protein in 25 of 29 adult patients with TS or OCD, compared to only two of 20 individuals who did not have OCD. In children, they detected the antibody in all of 11 patients with OCD but just two of 13 children without the disorder. Additional work by Dr. Trifiletti and by researchers at Johns Hopkins University has revealed antibodies to a second brain protein with an apparent molecular weight of 60 kD. Nearly all patients with TS or OCD have antibodies to one of these two proteins, Dr. Trifiletti said, although very few have both. (Interestingly, patients with Sydenham’s chorea have neither antibody but produce an entirely different one.)

Although the brain proteins targeted by these antibodies have not been definitively identified, Dr. Trifiletti noted that the 83-kD protein is found exclusively in the gray matter and may be a brain-specific form of the enzyme calpain; the 60-kD protein is thought to be a calpain-regulating enzyme known as calpastatin. Together, Dr. Trifiletti said, the available data suggest that PANDAS occurs when streptococcus infection or another triggering agent produces alterations in the calpain/calpastatin system in genetically susceptible individuals, ultimately resulting in the neurochemical alterations that produce the symptoms of OCD and TS.

TREATING WITH IMMUNOTHERAPY

Because many cases of OCD and TS appear to develop as a result of an immunologic process, immunotherapy has the potential to reduce or prevent the changes in neurochemistry that cause them. However, the adverse effects of immunotherapy are potentially more severe than those associated with other treatments for OCD or TS. “These therapies, at this point in time, are really reserved for patients whose symptoms are extremely severe, who don’t respond to therapy,” Dr. Trifiletti said. One possible immunologic approach, he noted, is the use of intravenous glucocorticoids, such as a high-dose methylprednisolone regimen similar to that used for patients with multiple sclerosis. Oral corticosteroids are easier to tolerate but may not be as effective. Another alternative is intravenous immunoglobulin (IVIG), which, although effective, is very expensive. Plasmapheresis (plasma exchange) removes circulating antibodies but requires specialized equipment and expertise to administer. In some patients, antibiotics may also be worth a try, Dr. Trifiletti said.

The use of immunotherapy in patients with OCD or tic disorders is supported by several published reports. For example, Dr. Trifiletti described a 1996 case study of a boy whose OCD symptoms and associated neuroimaging abnormalities worsened following group A ß-hemolytic streptococcus infection. Treatment with plasmapheresis produced marked improvement in clinical symptoms and in magnetic resonance imaging changes. The results of a randomized trial by the NIMH group, published in 1999, demonstrated that a single course of either plasmapheresis or IVIG significantly improved OCD and tic disorders for up to one year; at one month after treatment, for example, scores on the Yale-Brown Obsessive Compulsive Scale had improved by 45% in IVIG recipients but only 3% in a saline placebo group. In contrast, a recent report by the same researchers found no improvement in OCD symptoms when five patients without a history of streptococcus-associated exacerbations received IVIG. Nonetheless, a growing body of evidence and clinical experience suggests that for “the most severely affected patients with Tourette’s syndrome or OCD, immunotherapy might be a viable addition to our current management,” Dr. Trifiletti concluded.

—Mark Bowes, PhD

Suggested Reading
1. Perlmutter SJ, Leitman S, Garvey MA, et al. Therapeutic plasma exchange and intravenous immunoglobulin for obsessive-compulsive disorder and tic disorders in childhood. Lancet. 1999;354:1153-1158.
2. 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.

Return to table of contents