Identification of monoclonal B-cell populations by rapid cycle polymerase chain reaction. A practical screening method for the detection of immunoglobulin gene rearrangements.

Alternatives to Southern blot hybridization for gene rearrangement analysis are being studied because of the time, labor, cost, and radioisotopes required for this technique. We have utilized a rapid, hot air, thermocycling polymerase chain reaction (PCR) system to examine various lymphoproliferative disorders for immunoglobulin heavy chain (IgH) gene rearrangements. This unique system amplifies DNA from 10 microliters samples placed in glass capillary tubes, over a total cycle time of about 30 minutes. Amplified bands are easily visualized on ethidium bromide-stained agarose gels. Forty-one monoclonal B-cell proliferations, 27 reactive lymphoid hyperplasias, 17 T-cell lymphomas and 3 cases of Hodgkin's disease were studied. All 88 cases were fully characterized by morphologic, immunophenotypic, and genotypic (Southern blot) analyses. Each case was separately evaluated by PCR with two primer pairs: 1) IgH variable region (VH) and IgH joining region (JH) and 2) bcl-2 and JH. Thirty-four of 41 monoclonal B-cell proliferations revealed a distinct band (within an expected base pair range) with 1 or both primer combinations supporting B-cell monoclonality; the other 7 cases were considered false negatives. The 47 entities without IgH gene rearrangements detectable by Southern analysis demonstrated no amplified product or a smear of amplified DNA with no distinct band. The overall specificity of PCR was 100%, and the sensitivity was 83% when directly compared with Southern blot analysis. Although its sensitivity is currently less than optimal, PCR is a rapid and practical screening method for the detection of IgH gene rearrangements. If a positive result is obtained no further analysis is required; however, if there is a negative result, standard Southern blot analysis should be performed to definitively exclude the presence of a monoclonal B-cell population in the sample.