A comparative study of HLA-DRB1 typing by standard serology and hybridization of non-radioactive sequence-specific oligonucleotide probes to PCR-amplified DNA.

A double-blind study was carried out to evaluate the relative performance and reliability of the PCR/SSOP assay compared to conventional serological typing in identifying HLA-DR alleles. A total of 268 consecutive samples were entered into the study. In 14 (5.2%) of the cases, HLA-DR serology could not be performed due to poor cell viability, while in seven (2.6%) of the cases, PCR/SSOP typing could not be performed due to poor amplification or to contamination with exogenous DNA. Among samples that were successfully typed by both methods, serologic typing correctly identified 455/465 (97.9%) DR antigens, while PCR/SSOP correctly identified 464/465 (99.8%) DR alleles (p = 0.0117, McNemar's test). The majority of discrepancies in serologic typing resulted from a lack of discriminative alloantisera to identify DR6 or DR103. For the overall sample set (N = 268), serology provided accurate results in 244 (91.0%) cases, while PCR/SSOP provided accurate results in 260 (97.0%) cases (p = 0.0037). The results of this study demonstrate that PCR/SSOP typing for HLA-DRB1 alleles provides results that are equal to or surpass serological typing for HLA-DR antigens. In addition, the PCR/SSOP approach offers the advantages of better reagent availability, lower cost, more rapid turn-around time, and greater accuracy, all of which would warrant its use as an HLA typing method of choice.