Affordable pediatric CD4 counting by flow cytometry in Malawi

Abstract

Background: Rapid expansion of antiretroviral therapy in Malawi has occurred in the relative absence of suitable pediatric CD4 counting facilities. We have recently validated in adults a simplified affordable flow cytometric CD4 counting method, the Blantyre count. There is a need for this technology to transfer to government laboratories run by local staff, and to be validated in children, where %CD4/lymphocyte values are required.

Methods: We assessed agreement of %CD4/lymphocyte values determined by the Blantyre count and Panleucogate methods on an EPICS XL-MCL flow cytometer on 113 venous blood samples from HIV-seropositive children in Blantyre, Malawi. All assays were performed by two Malawian laboratory technicians.

Results: Overall bias between the two methods was -0.13% (95% CI -0.37 to 0.11) and limits of agreement were -2.69 to 2.43% (95% CI -3.11 to -2.27 and 2.01 to 2.85). Limits of agreement were within -3.00 and 3.00 for each laboratory technician. Coefficient of variation for the Blantyre count assay was 2.0% and samples showed good stability over 5 days.

Conclusions: The Blantyre count method can accurately determine %CD4/lymphocyte values in blood of HIV-seropositive children on an EPIC XL-MCL flow cytometer at a reagent cost of US $0.21 per test or less. The assay can be competently carried out by local laboratory technicians.

Fig. 1

Panleucogate and Blantyre count gating strategies for %CD4/lymphocyte determination using System II Software. (a and b) Panleucogate strategy on venous blood sample with (a) CD45(log)/sidescatter(log) dot plot showing total leucocytes gate (the Panleucogate) (A) and total lymphocytes (B), (b) CD4(log)/sidescatter(log) dot plot of total lymphocytes (A) showing CD4-positive lymphocytes (D). (c and d) Blantyre count strategy on venous blood sample with (c) CD45(log)/sidescatter(linear) dot plot showing total lymphocytes (A), (d) CD4(log)/sidescatter(linear) dot plot of total lymphocytes (A) showing CD4-positive lymphocytes (B). Same venous blood sample as used in (a) and (b). (e and f) Blantyre count strategy on finger-prick blood sample with (e) CD45(log)/sidescatter(linear) dot plot showing total lymphocyte gate (A), (f) CD4(log)/sidescatter(linear) dot plot of total lymphocytes showing CD4-positive lymphocytes (B). %CD4/lymphocyte values are calculated directly from the number of cells in the CD4-positive lymphocyte region and number of cells in the total lymphocyte region. SS, side-scattered light.

Fig. 2

Comparison of CD4 counts determined as a percentage of total lymphocyte count (%CD4/lymphocyte) using the Blantyre count and Panleucogate methods with 113 venous blood samples from HIV-seropositive Malawian children. Mean %CD4/lymphocyte values are the average of the %CD4/lymphocyte values generated by the Blantyre count and Panleucogate methods for each sample. Black lines depict bias and upper and lower limits of agreement. Gray lines denote 95% confidence intervals for these values.