INTERLABORATORY CONCORDANCE OF DNA-SEQUENCE ANALYSIS TO DETECT REVERSE-TRANSCRIPTASE MUTATIONS IN HIV-1 PROVIRAL DNA

Thirteen laboratories evaluated the reproducibility of sequencing meth ods to detect drug resistance mutations in HIV-1 reverse transcriptase (RT). Blinded, cultured peripheral blood mononuclear cell pellets wer e distributed to each laboratory. Each laboratory used its preferred m ethod for sequencing proviral DNA. Differences in protocols included: DNA purification: number of PCR amplifications; PCR product purificati on; sequence/location of PCR/sequencing primers; sequencing template; sequencing reaction label; sequencing polymerase; and use of manual ve rsus automated methods to resolve sequencing reaction products. Five u nknowns were evaluated. Thirteen laboratories submitted 39 043 nucleot ide assignments spanning codons 10-256 of HIV-I RT. A consensus nucleo tide assignment (defined as agreement among greater than or equal to 7 5% of laboratories) could be made in over 99% of nucleotide positions, and was more frequent in the three laboratory isolates. The overall r ate of discrepant nucleotide assignments was 0.29%. A consensus nucleo tide assignment could not be made at RT codon 41 in the clinical isola te tested. Clonal analysis revealed that this was due to the presence of a mixture of wild-type and mutant genotypes. These observations sug gest that sequencing methodologies currently in use in ACTG laboratori es to sequence HIV-I RT yield highly concordant results for laboratory strains; however, more discrepancies among laboratories may occur whe n clinical isolates are tested. (C) 1998 Elsevier Science B.V.