TO THE EDITOR: My colleagues and I appreciate the comments of Dr. Hollander and his colleagues and welcome the opportunity to provide clarification of the results described in our recent article.

As we reported, D8/17 was first identified as a trait marker of rheumatic fever susceptibility and has been widely tested in a variety of patient groups and in various epidemiologic samples throughout the world. There is no evidence to suggest that attack rates of acute rheumatic fever differ between genders (although Sydenham's chorea is slightly more common among female adolescents with rheumatic fever than among male adolescents). Similarly, there has been no evidence for male-female differences in the rates of D8/17 positivity in previous investigations (1–3), nor did we find differences in relative rates of D8/17 positivity among male and female patients in our study. Thus, at present, there is no evidence to suggest that gender is related to D8/17 status.

In interpreting the results of recent neuropsychiatric investigations (4), it is important to remember that D8/17 was developed as a trait marker of rheumatic fever susceptibility. Numerous rheumatic fever investigations and our increasing experience with longitudinal D8/17 assessments in patients with OCD and tic disorders (including Tourette's disorder) clearly demonstrate that D8/17 is not a state marker of streptococcal reactivity. Subjects who are initially identified as being D8/17 positive remain in that category even when their antistreptococcal titers fall to normal levels; conversely, numerous subjects have been found to be D8/17 negative despite markedly elevated antistreptococcal antibody titers, as seen in the patients with well-documented acute poststreptococcal glomerulonephritis, in which all patients had decreased complement, high anti-streptolysin O or anti-DNase B titers, and urinary signs of disease, yet had low D8/17 values (1). Because the relative percentage of D8/17+ cells remains constant among individuals across time, it is highly unlikely that the percentage of D8/17+ cells will be found to correlate with symptom severity. In fact, since D8/17 status is reported as a dichotomous variable (positive or negative), it is difficult to envision how it might be used as a "dimensional" variable.

We agree with Dr. Hollander and colleagues that D8/17 is an interesting biologic marker worthy of further investigation. Studies that examine rates of D8/17 positivity in various neuropsychiatric disorders will help determine whether the marker is related only to poststreptococcal immune dysfunction (as postulated) or if it may also serve as a marker of neuropsychiatric vulnerability. The recent report by Dr. Murphy and colleagues suggests that at the least, D8/17 is able to identify an unselected group of patients with childhood-onset obsessive-compulsive disorder (4).