Response to Retroviral Therapy in the Mulago Hospital AIDS Clinic/Infectious Disease Institute: A
Start Date: 01/01/2007
End Date: 01/01/2010
A prospective comparison of two manual bead-based assays (Dynabeads and Cytospheres) to immunologic response by flow cytometry and virologic response by RT-PCR at a public, urban clinic in Kampala, UgandaLaboratory-based monitoring of antiretroviral therapy (ART) is essential, but adds a significant cost to HIV care. The optimal method of monitoring ART remains unknown. We compared CD4 lymphocyte counts (CD4) obtained on replicate samples from 497 HIV-positive Ugandans before and during ART and followed for 18 months by two manual bead-based assays, Dynabeads (Dynal Biotech) and Cytospheres (Coulter), to those generated by flow cytometry (FC) in Kampala, Uganda. In addition, we evaluated how well CD4 measurement by manual methods and FC predicted virologic failure (Roche, Amplicor 1.5). We tested 1670 samples (123 prior to ART) with Dynabeads and 1442 samples (91 prior to ART) with Cytospheres. Mean CD4 was 230 cells/mm3 (SD,139) and 239 cells/mm3 (SD,140) by Dynabeads and FC, respectively. Mean CD4 was 185 cells/mm3 (SD,101) and 242 cells/mm3 (SD,136) by Cytospheres and FC, respectively. Linear regression slopes were 0.85 and 0.58 and intercepts were 28 and 46 for Dynabeads and Cytospheres, respectively with Pearson correlation coefficients of 0.85 and 0.78. The mean differences in CD4 were 8.7 (95%CI, 5.1-12.4) and 56.8 (95%CI, 52.4-61.2) for Dynabeads and Cytospheres, respectively with differences ranging from -409 to 501 and -399 to 583. The differences between the manual bead-based methods and FC were significantly different (p-value, 0.0001). Whereas, manual methods to predict ART failure are characterized by low sensitivity and PPV, and high specificity and NPV. Our results suggest that manual methods can be used to initiate ART in resource-limited settings without a high risk of false negative results. In contrast, follow-up monitoring to evaluate for ART failure with CD4 enumeration by manual methods and FC is best used to identify those with a favorable response to ART, but is unreliable when used to identify failure of ART.