Abstract
A new algorithm based upon the differential antibody responses to two gag gene products (p19 and p24) of human T lymphotropic virus (HTLV) has been suggested for serologic discrimination of HTLV type I (HTLV‐I) and type II (HTLV‐II) [Lillihoj et al., 1990]. To evaluate the practical usefulness of this algorithm, serum specimens from HTLV‐seropositive individuals whose infection was confirmed by PCR analysis to be HTLV‐I (n = 60) or HTLV‐II (n = 61) were analyzed by western blot. The intensities of the antibody response to p24^gag^and p19^gag^ were scored by one individual without prior knowledge of PCR results. According to the algorithm, specimens with p19 > p24 were classified as HTLV‐I, whereas specimens with p19 < p24 were classified as HTLV‐II. Of 60 PCR confirmed HTLV‐I specimens, 56 had p19 > p24 (93%) while 4 had p19 < p24. Of 61 PCR confirmed HTLV‐II specimens, 56 had p19< p24 (92%) and 5 had p19 > p24. The overall accuracy of serologic differentiation when using this algorithm was 92%, as 4 of 60 HTLV‐I (7%) and 5 of 61 HTLV‐II (8%) could have been wrongly classified. Although the differential antibody response to p19^gag^ and p2 4^gag^ provides a simple means of serologically distinguishing between HTLV‐I and HTLV‐II infection in population‐based epidemiological studies, in a clinical context more accurate means of confirmation are required. The dominant p19^gag^ responses were mapped to the C‐terminus of p19 (p19^102–117^) Competitive inhibition of p19^gag^ response by peptide p19^102–107^, however, did not abrogate the binding of serum specimens from HTLV‐I‐infected individuals.