OBJECTIVE. Chorea is characterized by involuntary, fleeting, irregular, nonrhythmic movements that flow from 1 body region to another. There are many causes of childhood chorea, including cerebrovascular accidents, collagen vascular diseases, drug intoxication, hyperthyroidism, Wilson's disease, Huntington's disease, and infectious agents. Although Sydenham's chorea (SC), a nonsuppurative sequela of group A streptococcal infection, is known to be a common cause of chorea, multiple laboratory and radiographic studies are often obtained to determine the cause of pediatric chorea. We conducted a retrospective chart review to determine the causes of childhood chorea seen in a large children's hospital in an area endemic for acute rheumatic fever (ARF). The utility of neuroimaging in establishing a final diagnosis of SC is discussed.
METHODS. Patients who received a diagnosis of chorea between 1980 and 2004 at the Children's Hospital of Pittsburgh were identified from databases that are maintained by the divisions of Infectious Diseases and Cardiology and from the hospital's medical records department. Charts were abstracted retrospectively. All patients who had new-onset chorea and did not have any underlying neurologic disorders were included in this study. Patient demographic, clinical, laboratory, and imaging information was analyzed. Follow-up information was not found consistently and therefore was not included. Charts of patients with questionable diagnoses were reviewed with a neurologist.
RESULTS. A total of 144 patients met the search criterion. Eleven patients had incomplete charts, and 6 charts could not be located. Thirty patients were excluded because they had preexisting neurologic diagnoses, eg, cerebral palsy. Fifteen patients were excluded because they were miscoded as having chorea. Eighty-two patients had new-onset chorea. The cause was SC (n = 79), postoperative cerebral ischemia (n = 1), and basal ganglion infarct (n = 2). Seventy-six (71%) children with SC were female. The mean age of presentation was 9.8 years (range: 5–14.5 years). Chorea was unilateral in 23 (30%) patients. Family history of ARF existed in 30% of patients. Neurologic symptoms of SC included behavior change (46%), dysarthria (67%), gait change (51%), deterioration of handwriting (29%), and headache (11%). Nonneurologic manifestations of ARF were carditis (44%), arthritis (11%), erythema marginatum (3%), and subcutaneous nodules (0%). Antecedent group A streptococcal infection was documented in 99% of patients who were tested by an elevated antistreptolysin O titer (n = 53), an elevated anti-deoxyribonuclease B titer (n = 7), a positive streptozyme (n = 53), or acute throat infection with Streptococcus pyogenes (n = 19). A total of 52 neuroimaging tests were obtained from 46 patients with SC. In patients with SC, brain MRI was abnormal in 8 of 32 patients, and brain computed tomography was abnormal in 1 of 20 patients. Abnormalities did not aid in diagnosis and included nonspecific increased signal (n = 2), nonspecific punctate lesions (n = 2), asymmetry of the hippocampal fissures, unrelated petrous bone anomaly, Arnold Chiari malformation, and medulloblastoma in a macrocephalic patient. Three patients with chorea that was not attributed to ARF had atypical presentations: 1 developed chorea after a perioperative hypoxic/ischemic central nervous system insult; 1 had an episode of disorientation, aphasia, and transient facial droop (angiography showed basal ganglia infarct); and 1 with hemichorea had basal ganglion infarct seen on MRI.
CONCLUSIONS. Ninety-six percent of children who had acute chorea and presented to a large children's hospital in an area that is endemic for ARF had SC. These patients had characteristic demographic and clinical features of SC. The most common concurrent major Jones criterion was carditis. Arthritis, erythema marginatum, and subcutaneous nodules were uncommon in this population. Neuroimaging was obtained in 58% of patients with SC and did not aid in any of their diagnoses. The 3 patients with chorea that was not caused by SC had histories that were atypical for SC and warranted neuroimaging. SC can be readily diagnosed on the basis of history, physical examination, and laboratory evaluation; neuroimaging is not necessary and should be reserved for patients who have an atypical presentation, including hemichorea.