Next Section

Criterion 1

As mentioned, several authors have suggested that tics and OCD may not be the only clinical features of PANDAS, and thus the true boundaries of the PANDAS clinical spectrum remain to be scientifically validated or standardized. Even considering only tics and OCD, the level and characteristics of symptom severity required for diagnosis have not been defined yet by controlled studies.

Criterion 2

The proposed age at onset, ranging between 3 years and the beginning of puberty, may be too arbitrary and may simply be the result of referral patterns for the 50 initially described patients. For example, the validity of the proposed age-at-onset criterion is raised by the report of a 25-year-old with sudden-onset OCD after pharyngitis.¹⁵ Furthermore, because the vast majority of typical Tourette’s syndrome (TS) manifestations (ie, tics) begin during the prepubertal period (75% having onset before 11 years old¹⁶), the age-at-onset criterion is not sufficiently specific in distinguishing PANDAS from the more established diagnostic criteria for TS.¹⁷ Recognizing a documented higher likelihood of exposure of schoolchildren to GABHS, no pathogenetic or immunologic basis has been confirmed to explain why the onset might be expected to include only young children.

Criterion 3

A unique specificity of a clinical course consisting of abrupt onset or dramatic exacerbations has not been documented adequately. With careful observation, tics are either present or not; thus, the onset (or parental awareness) of tic disorders is not gradual. Often with a sudden recognition of tics by parents, in retrospect a prior history of milder tics can be elicited, making the precise timing of clinical onset confusing. Experienced neurologists and psychiatrists are aware that tic disorders and OCD characteristically wax and wane spontaneously, with exacerbations followed by remissions. The degree of clinical change required for recognizing exacerbations to distinguish a “PANDAS-like process” from the typical course of more established tic and OCDs has not been differentiated or quantitated adequately. For example, 1 recent study of 80 consecutively examined (unselected) patients in a tic-disorders clinic found that 53% reported sudden, explosive onset or worsening of their tics.¹⁸ These patients were not considered to have met criteria for a diagnosis of PANDAS. Preliminary evaluation of our own ongoing, case-control epidemiologic study of PANDAS indicated that 36% of the carefully selected control subjects (who had no recognized link between symptoms and GABHS infection) also reported an abrupt onset or dramatic exacerbations (R. Kurlan, MD, unpublished data, 2003). Thus, clinical course does not seem particularly useful in distinguishing patients suspected of PANDAS from children with more typical cases of TS or OCD.

Criterion 4

Onset or exacerbations of neuropsychiatric symptoms temporally related to GABHS infection has been proposed as the sine qua non of the diagnosis of PANDAS. Yet, despite limited studies to establish this, it has not been documented adequately in prospective studies and remains 1 of the most controversial aspects of the hypothesis. GABHS infections are very common, and the prevalence of this organism in the upper respiratory tract of asymptomatic school children (“carriers”) often is relatively high.¹⁹ This epidemiologic phenomenon of high prevalence of GABHS has resulted in a number of other disease entities (among the best examples are Henoch-Schönlein purpura and Kawasaki disease), which also have been proposed to be associated with group A streptococcal infections. However, when adequately controlled studies were conducted, streptococci were eliminated as the causal factor.²⁰ The confusion about the role of GABHS may be promoted further because, clinically, it is appreciated that tic disorders and OCD worsen during times of stress or illness of any kind. The worsening of symptoms might be expected with streptococcal upper respiratory tract infection and could even occur in the relatively large percentage of children who are asymptomatic streptococcal carriers if simultaneous stressful situations occur (from any other cause).

Additionally, review of published studies indicates that the temporal boundaries of a proposed “temporal” relationship with streptococcal infections have never been established. In Sydenham’s chorea, often cited as the model for PANDAS, the choreoathetoid movements typically appear 3 to 5 months after GABHS infection. Swedo et al¹ implied that an infection up to 9 months before symptom onset may be acceptable for the diagnosis of PANDAS but also have indicated that the presence of GABHS in the upper respiratory tract (not differentiating infection from the carrier state) may not be detectable until weeks after onset or exacerbation. Such a lack of precision does not assist in establishing a distinct syndrome.

Thus, the absence of regular, continuous, prospective throat cultures along with streptococcal antibody determinations has made it impossible to establish conclusively a temporal relationship between streptococcal infection and onset or exacerbation of neuropsychiatric symptoms. Sufficiently long, longitudinal, prospective studies conclusively demonstrating streptococcal infection and correlating this with clinical findings are required.

Criterion 5

The final diagnostic criterion, that there must be abnormalities on neurologic examination, presents a difficult and practical clinical problem. For example, the 1998 report that “choreiform” movements are often present in patients in whom a diagnosis of PANDAS is considered indicates that the cohort on which the proposed diagnostic criteria were based may well have included subjects with mild chorea that might also be compatible with more conventional Sydenham’s chorea.¹

Laboratory Testing

An initial concern of pediatricians and primary care physicians about clinical management of patients with this symptom complex is which, if any, diagnostic laboratory studies should be done. Given the current status of as-yet incompletely defined support for the existence of a specific entity called PANDAS, the physician may be cautious in deciding whether to obtain throat cultures or determine antistreptococcal antibodies in all children presenting with neurobehavioral symptoms such as tics, obsessions, or compulsions. Because of potential difficulties in interpretation of single tests, even with “positive” culture results, the high prevalence of so many GABHS carriers in normal populations potentially confuses the issue.

Patients with signs and symptoms compatible with streptococcal upper respiratory tract infections should be studied for the presence of GABHS just as with any other patient. In the absence of symptoms of streptococcal infection, the relevance of a positive culture is difficult to interpret. The many uncertainties encountered in interpreting streptococcal antibody titers in otherwise normal children are recognized.²¹ Therefore, to use streptococcal antibody tests or throat cultures in an asymptomatic child either as a screening test or to confirm a diagnosis of PANDAS is problematic.

Two research laboratory tests, measurement of circulating antineuronal antibodies and an assay for the presence of an alleged rheumatic fever-associated T lymphocyte alloantigen called D8/17,²² have not been able to correlate reliably and consistently with suspected PANDAS cases.¹⁴

Although radiologically suspected enlargement of caudate nuclei has been described in some suspected patients, routine magnetic resonance imaging cannot be recommended currently, because the validity has not been substantiated.²³

Therapy

What treatment should be considered for patients who present with the manifestations associated with the PANDAS constellation of symptoms?

Neuropsychiatric Drugs

Children with disabling tics or OCD should receive appropriate treatment for their symptoms, including medications (eg, tic suppressants such as α-agonists, classical or atypical antipsychotics, or antiobsessional drugs such as selective serotonin-reuptake inhibitors) or nonpharmacologic approaches (eg, cognitive behavioral therapy for OCD).²⁴

Antibiotics to Prevent Strep Infections

The only published clinical trial of prophylactic penicillin in children with this syndrome revealed no conclusive evidence that the antibiotic reduced clinical exacerbations.²⁵ Admittedly, the duration of that study was too short to allow definite conclusions. A later report suggesting improvement of new-onset or acute exacerbations of symptoms in such children with antibiotics provides inadequate support for such an approach, because treatment was not placebo-controlled and was unblinded.²⁶ It is well known in treatment studies of TS that there is a substantial placebo effect; the natural course of TS and OCD is such that exacerbations are followed by remissions. This latter phenomenon of “reversion to the mean” implies that virtually any intervention at the time of peak symptoms may seem successful. Only a double-blind, placebo-controlled study can identify a true therapeutic effect.

Another reason to feel comfortable with avoiding antibiotic treatment for these patients is that, to date, no cases have been reported to develop any rheumatic carditis as occurs in patients with Sydenham’s chorea. This is a very important point to remember, because some have attempted to relate this syndrome to rheumatic fever. It is recognized that as many as one third of patients with Sydenham’s chorea ultimately will have evidence of rheumatic valvular heart disease.²⁷ This issue requires additional investigation, because the lack of heart disease strongly argues against a relationship between PANDAS and Sydenham’s chorea or other forms of rheumatic fever.