Sensitivity and specificity of human T-lymphotropic virus (HTLV) types I and II polymerase chain reaction and several serologic assays in screening a population with a high prevalence of HTLV-II.

Abstract

Since 1988, all blood donations in the United States have been screened for antibodies to human T-lymphotropic virus type I (HTLV-I). However, the sensitivity of current serologic tests for the detection of HTLV type II (HTLV-II) antibodies and the diagnostic utility of direct tests for HTLV-I and -II using polymerase chain reaction (PCR) are poorly defined. Five hundred sixty-nine HTLV-I- or -II-seropositive and 687 age- and sex-matched seronegative samples from a high-risk population at an inner-city emergency department were selected. All samples were tested with four HTLV enzyme immunoassays (EIAs), one Western blot assay and one type-specific Western blot assay, one HTLV type-specific EIA, and a research HTLV-I/II PCR kit. Sensitivity of the various EIAs ranged from 95.1 to 99.5 percent, and specificity ranged from 97.2 to 99.4 percent. PCR performed in duplicate without selective retesting had lower sensitivity (85.1 %) and specificity (88.0%). However, PCR detected 20 (3.2%) HTLV-I-positive and 47 (7.5%) HTLV-II-positive samples among the 627 samples that were negative in all EIAs. The type-specific EIA and PCR assay had the highest rate of concordance in classifying samples as either HTLV-I or II, with the type-specific EIA and type-specific Western blot having the next highest rates of concordance. In this sample set from a population at high risk for HTLV-II, screening with HTLV-I/II PCR had lower sensitivity and specificity than that with EIAs. However, 4.1 to 10.8 percent of samples were PCR positive but seronegative for HTLV-I or -II, and their true infection status remains undetermined.