Background: The purpose of this study was to assess whether laboratories which do not use CD45 for gating lymphocytes with three-(or four-) color flow cytometry (non-CD45 laboratories) for CD3 ุ 4 ุ % and CD3 ุ 8 ุ % do worse on quality assessment (QA) studies than laboratories which do use CD45 (CD45 laboratories). Methods: Data came from blood specimens donated by 62 donors (50 HIV-positive) assayed over 2 years (November, 1996-October, 1998) by 35 laboratories in the NIAID DAIDS Flow Cytometry QA Program.
Results: Non-CD45 laboratories were significantly more likely to be classified as having unacceptable inter-laboratory results (far from the group median) than CD45 laboratories (5.6% vs 1.5%, P โซุโฌ 0.005 for CD3 ุ 4 ุ %; 10.4% vs 5.0%, P โซุโฌ 0.007 for CD3 ุ 8 ุ %). The intra-laboratory range of results on blinded replicates was significantly more likely to be deemed unacceptable (range >4%) in non-CD45 laboratories than in CD45 laboratories for CD3 ุ 8 ุ % (14.5% vs 3.5%, P โซุโฌ 0.002) but not for CD3 ุ 4 ุ % (2.6% vs 1.5%, P โซุโฌ 0.62). These differences in favor of CD45 gating were observed even though the non-CD45 laboratories had been doing three-color flow cytometry in the QA program significantly longer (P โซุโฌ 0.05) than the CD45 laboratories, and so would be expected to have fewer problems with the assay.
Conclusions: Laboratories which choose to use a single CD3/CD4/CD8 tube for immunophenotyping may be sacrificing both accuracy and reproducibility.