B Lymphocyte Antigen D8/17 and Repetitive Behaviors in Autism

Eighteen children with autism (14 boys, four girls) ranging in age from 4 to 13 years (mean=6.9, SD=3.1) completed the protocol. They met the DSM-IV criteria for autistic disorder and were free of major comorbid psychiatric disorders. The DSM-IV diagnosis of autism for each of the 18 patients was made by semistructured interview by a psychiatrist who specializes in autism research (E.H. or C.C.). In addition, for nine of the subjects the diagnosis was confirmed by the Autism Diagnostic Interview (16), with complete agreement between the DSM-IV and Autism Diagnostic Interview diagnoses. The subjects’ full-scale IQs ranged from 52 to 109 (mean=80, SD=20). Fourteen medically ill children (six boys, eight girls) ranging in age from 1 to 18 (mean=6.0, SD=4.6), who were comparable on socioeconomic, racial, and geographic factors, were used as comparison subjects. The illnesses in this group included surgery for small bowel resection, cardiac catheterization, asthma, sepsis, Kawasaki’s disease, and polycystic kidney disease complicated by hypertension.

The Yale-Brown Obsessive Compulsive Scale (17), a clinician-administered 10-item questionnaire that uses 5-point scales to rate time spent, interference, distress, resistance, and control, was administered to assess the severity of repetitive thoughts and behaviors. It is a reliable and valid scale for severity of obsessions and compulsions, and the compulsion subscale reliably assesses compulsive and repetitive behaviors in autistic subjects and is sensitive to change in treatment studies (18). All patients and their parents gave written informed assent and consent, respectively, for participation in this study.

Heparinized blood for immunofluorescent studies was collected from the autistic subjects and comparison subjects on the same day and stored at room temperature for no more than 24 hours before the assay. Monoclonal antibody D8/17 is a mouse monoclonal IgM antibody originally prepared from fusions of spleen cells from mice repeatedly immunized with isolated B cells obtained from patients with rheumatic fever or rheumatic heart disease. The monoclonal antibody D8/17 was produced in the Laboratory of Clinical Microbiology and Immunology at Rockefeller University, and the immuno­fluorescent assay was run as described elsewhere (11, 19). In brief, by means of an indirect immunofluorescence assay, B cell staining was determined on fresh mononuclear cell preparations. The peripheral blood B cells were then stained with monoclonal antibody D8/17 and fluoresceinated antimouse IgM and counted with a fluorescent microscope. B cells were identified by using concomitant staining with phycoerythrin-conjugated monoclonal anti-DR human leukocyte antigen class II reagents. The results were expressed as the percentage of positive cells among 500 counted. All of the readings were done by an examiner blinded to subject status, and they were checked by another examiner for accuracy. On the basis of previous studies, a D8/17-positive case was defined as a value of one standard deviation above historical comparison values, i.e., ≥11% antigenic expression. The normal comparison mean values range from 0% to 7% (10). To assess for recent streptococcal infection, blood samples were collected and assayed for streptococcal antibodies (antistreptolysin-O and antiDNase B).

The results of the immunofluorescence assays were classified as negative (0%–10% D8/17 expression) or positive (≥11%). All tests of significance used the 0.05 level of significance and were two-sided. For goal 1, the proportion of positive expression, out of positive plus negative, was the estimate of the occurrence, and Pearson’s chi-square test was applied. For goal 2, the autistic subjects’ D8/17 assay value was correlated with their Yale-Brown Obsessive Compulsive Scale compulsion score by means of Pearson correlation. For goal 3, we compared the compulsion scores of the D8/17-negative and -positive autistic subjects, conservatively using the less significant of the pooled and separate variance estimate t tests.

The autistic and medically ill groups did not significantly differ in age (t=0.62, df=22, p=0.54) or sex distribution (p=0.07, Fisher’s exact test), and age did not significantly influence D8/17 level (r=0.23, df=16, p=0.35). Fourteen (77.8%) of the 18 autism subjects were positive for the D8/17 marker (≥11%) and four (22.2%) were negative (<11%). In contrast, three (21.4%) of the 14 comparison subjects were positive and 11 (78.6%) were negative for the marker. Thus, the autistic subjects were significantly more likely than the comparison subjects to be classified as D8/17 positive (Pearson χ²=10.04, df=1, p=0.002).

Of interest, the severity of repetitive behaviors among the autistic children, as measured by the compulsion subscale of the Yale-Brown Obsessive Compulsive Scale, significantly positively correlated with D8/17 values (r=0.73, df=16, p=0.001), such that the autistic patients with greater compulsive/repetitive behavior had greater D8/17 antigen positivity (figure 1, left side). In contrast, there was no significant (or nearly significant) correlation between D8/17 value and score on the Autism Diagnostic Interview social (r=0.01) or communication (r=0.29) algorithm. When a dichotomous approach was used it was seen that the D8/17-positive autistic patients had a significantly higher mean Yale-Brown Obsessive Compulsive Scale compulsion score (mean=11.9, SD=3.8) than did the D8/17-negative patients (mean=3.7, SD=4.5) (t=3.29, df=4.29, p=0.03) (figure 1, right side). Thus, D8/17 positivity was associated with greater compulsive severity in children with autism, and the D8/17-positive patients had greater compulsive severity.

Only one (5.6%) of the 18 autistic subjects had a high antistreptolysin-O titer (≥200 IU/ml), and only four (22.2%) had high levels of antiDNase B (≥1:64 for preschool and ≥1:170 for school-age children; smaller ratio equals greater dilutional titer, which corresponds to higher antistreptococcal antibody level), suggesting that the majority of the autistic children had had no recent streptococcal exposure. The study subjects and their parents did not report any history of rheumatic fever.

The cause of autism is not known, although there is evidence of immunologic dysfunction in the disorder (1, 3–8). Monoclonal antibody D8/17 identifies a B cell antigen that denotes susceptibility to rheumatic fever (10). Recently, high levels of D8/17 have been documented in a subgroup of OCD patients (13, 14).

Our data demonstrate significantly greater expression of monoclonal antibody D8/17 in a subgroup of autistic children than in matched medically ill children. D8/17 has been proposed as a trait, rather than state, marker of rheumatic fever, and thus it has been suggested that D8/17 should be treated as a dichotomous variable (i.e., positive or negative) (20). However, the percentage of B cells positive for D8/17 may fluctuate with disease exacerbation (J.B. Zabriskie, personal communication), and severity of repetitive behavior in autism may fluctuate in response to antiobsessional treatment with serotonin reuptake inhibitors (21), suggesting that a continuous or dimensional approach to studying their relationship may be justified. Severity of repetitive behaviors as assessed by the compulsions score on the Yale-Brown Obsessive Compulsive Scale strongly positively correlated with D8/17 expression. Alternatively, in a dichotomous approach, the D8/17-positive patients had significantly higher scores for repetitive behavior than did the D8/17-negative patients.

D8/17 antigen expression may represent a genetic vulnerability to autism and an environmental susceptibility to autism. The nature of the association between D8/17 positivity in autism and abnormal immune response to group A β-hemolytic streptococcal is unclear at this time. A major manifestation of rheumatic fever is Sydenham’s chorea, in which antibodies to β-hemo­lytic streptococcal cross-react with neuronal cells and cause motoric and behavioral abnormalities, including obsessive-compulsive symptoms (13, 14), and autoantibodies directed against the caudate and subthalamic nuclei have been identified (22). D8/17 expression and red cell Na⁺/H⁺ antiporter activity were positively correlated in rheumatic fever patients in one study (23), and it was suggested that D8/17 may be associated with cation transport in cell membranes (23), not just of erythrocytes but perhaps of brain cells as well (24).

Our preliminary data suggest that the B cell alloantigen identified by the monoclonal antibody D8/17 may identify a subgroup of autistic subjects. Future research should examine whether antineuronal antibodies are involved in the pathogenesis of autism and should compare D8/17-positive and D8/17-negative autistic subjects for the presence of antineuronal antibodies, differences in familial transmission, and differential response to specific therapeutic strategies.

Presented in part at the 35th annual meeting of the American College of Neuropsychopharmacology, San Juan, Puerto Rico, Dec. 11–15, 1996; the 23rd Northeast Region Meeting of the American College of Rheumatology, Boston, April 18–20, 1997; and the 150th annual meeting of the American Psychiatric Association, San Diego, May 17–22, 1997; Received Jan. 9, 1998; revision received May 20, 1998; accepted June 8, 1998. From the Department of Psychiatry and the Seaver Autism Research Center, Mt. Sinai School of Medicine; and the Laboratory of Clinical Microbiology and Immunology, Rockefeller University, New York. Address reprint requests to Dr. Hollander, Department of Psychiatry, Box 1230, Mt. Sinai School of Medicine, One Gustave L. Levy Place, New York, NY 10029-6574; [email protected]. edu (e-mail). Supported in part by grants from the Seaver Foundation and from Cure Autism Now, by a Distinguished Investigator Award from the National Alliance for Research in Schizophrenia and Affective Disorders, and by grant RR-00071 for the Mount Sinai Clinical Research Center from the NIH Center for Research Resources.