Evaluation of a flow cytometry method for CD4 T cell enumeration based on volumetric primary CD4 gating using thermoresistant reagents.

Laboratory follow-up of HIV patients in resource-limited settings requires appropriate instruments for CD4 T cell enumeration. In this study, we evaluated the application of a simplified, mobile and robust flow cytometry system, the Apogee Auto 40 analyzer (Auto40) using thermoresistant reagents, for CD4 T cell enumeration. We measured the absolute CD4 counts in fresh whole blood samples from 170 Senegalese subjects, including 129 HIV-positive (HIV+) patients and 41 HIV-negative (HIV-) controls. Based on volumetric primary CD4 gating, cells were stained with commercially available reagents (Easy MoAb CD4;Bio-D, Valenzano, Italy) and analyzed on the Auto40. The results were compared with those from the FACSCount system (Becton Dickinson, San Jose, USA). Repeatability analysis was performed on duplicate testing of 49 samples on both FACSCount and Auto40. The intra-run precision was measured by 10 replicates using 3 clinical blood samples with low, intermediate and high CD4 concentrations. The results from the two instruments were in good agreement. The percent similarity between the results of both instruments was 99%±relative standard deviation of 12.7%. The concordance correlation coefficient was 0.99. The absolute bias and limits of agreement (LOA) between the two instruments, calculated by Bland-Altman analysis, were clinically acceptable (bias: +4 cells/μl; LOA: -111 to +120 cells/μl). The clinical agreement between the two instruments at a cutoff of 200 CD4 cells/μl was 94%. The repeatability of measurements on the Auto40 was also similar to that observed with FACSCount system (bias +0.1 cells/μl, coefficient of variation 2.5% vs bias -1.1cells/μl, coefficient of variation 2.9% respectively). In conclusion, our results indicate that the Auto 40 system, using thermoresistant reagents, is suitable for CD4 T cell enumeration and will be a helpful tool to improve HIV laboratory monitoring in resource-limited settings.