Abstract

AIMS--To determine whether the polymerase chain reaction with sequence specific primers (PCR-SSP) can assign HLA-DR type more accurately than serology in a routine hospital laboratory. METHODS--The 93 patients currently awaiting kidney transplants have been DR typed by serology over the past 14 years, 82% within the past five years. They have now been retyped using the PCR-SSP method described by Bein et al. Where the two results differed, PCR-SSP was repeated, once by the same method and once using the primer set of Olerup and Zetterquist. RESULTS--There were 13 (14%) discrepancies between the results. Of these, two were PCR-SSP failures, later overcome: three were failure to detect DRB1*0103 by serology; five assignment of other alleles by PCR-SSP to serological "blanks"; and three alleles were differently assigned by serology and PCR. The serological typing of the final patient when repeated for this study was at variance with the original findings (14 years ago), but in agreement with PCR. In the remaining patients, serology had not determined the split of 36 DR3 alleles (all DR17 by PCR-SSP) or 13 DR6 alleles (12 DR13 and one DR14 by PCR-SSP). One patient in each case had their antigen splits of DR2 and DR5 assigned by PCR-SSP (DR15 and DR11, respectively) but not by serology. CONCLUSIONS--PCR-SSP provides more reliable and detailed information on HLA-DR polymorphism than serology, and does so within a routine tissue typing laboratory.