EVALUATION OF DIFFERENT CRITERIA OF INTERPRETATION OF THE WESTERN-BLOT FOR THE DIAGNOSIS OF INFECTION BY THE HUMAN-IMMUNODEFICIENCY-VIRUS

BACKGROUND: The Western blot (WB) is the most commonly used test to co nfirm the presence of antibodies against the human immunodeficiency vi rus type 1 (HIV-1). Different criteria of interpretation of the band p rofile have been proposed with there being no unanimity as to its reli ability. The sensitivity and specificity of several criteria proposed for the interpretation of WB were evaluated and the individual signifi cance of the reactivity of each band of the WB was analyzed. METHODS: The presence of antibodies against HIV-1 was prospectively studied in 8,073 samples of subjects with risk of infection. A total of 1,993 (25 %) were reactive by ELISA and 1,261 were analyzed by WB, with a semiq uantitative reading of the bands with a point scale from 0 to 2 being performed. The final interpretation of the WB (negative, daubtful, or positive) was carried out following 5 recommendations of usage. A test designed with synthetic peptides (Pepti-lav) was used as a reference and in discordant cases, other more specific serologic tests and/or ge netic analysis by polymerase chain reaction (PCR) were performed. RESU LTS: In order of frequency, the greater sensitivity was found to be fo r the CRSS (Consortium for Retrovirus Serology Standarization) criteri a (97.9 %), OMS (96.6 %), CDC (Center for Disease Control) (95.9 %), A RC (American Red Cross) (95.6 %) and FDA (99.8 %). The greatest specif icity was for the criteria of the OMS, and FDA (99.8 %). In order of f requency, the most frequent bands in HIV-1 + individuals were gp160 (9 9 %), gp120, p24, p31, p55, p68, gp41, and p17 (68 %). In non infected individuals, the recognized bands were, in decreasing order, p24, p17 , p55, p68, p31, and glucoproteins. CONCLUSIONS: Different criteria of interpretation of the Western blot provide different degrees of sensi tivity and specificity. The Western blot is a non standardized, expens ive, laborious technique of subjective interpretation which provides a n appreciable number of undetermined results.