Field evaluation of performance of dried blood spots (DBS) from finger-prick for the determination of viral load in a resource-constrained setting in urban and rural Zimbabwe Sekesai Mtapuri-Zinyowera,1 Fabian Taziwa,2 Carol Metcalf,3 Elton Mbofana, 4 Silvia De Weerdt, 5 Laurence Flevaud, 5 Sandra Simons,2 Jean-François Saint-Sauveur,2 Helen Bygrave,3 Emmanuel Fajardo3 1National Microbiology Reference Laboratory, Harare, Zimbabwe, 2MSF, Harare, Zimbabwe, 3MSF, Cape Town, South Africa, 4MSF, Buhera District, Zimbabwe, 5MSF, Barcelona, Spain
In resource-limited countries, the use of dried blood spots (DBS) as an alternative to plasma for the determination of HIV-1 viral load is an important intervention for ART programmes intending to scale up access to viral load monitoring at national level. DBS have been shown to correlate well with plasma using RNAspecific viral load assays that use the nucleic acid sequenced-based amplification (NASBA) technology; however, higher viral load results have previously been reported with assays that use non RNA-specific amplification methods such as the Roche COBAS AmpliPrep/COBAS TaqMan or the Abbott m2000sp due to the contribution of pro-viral DNA present in white blood cells from whole blood.
118 finger-prick DBS, venous blood DBS and plasma specimens were analyzed in the National Microbiology Reference Laboratory (NMRL) in Harare, Zimbabwe. Specimens were collected from ART patients attending two rural clinics in Buhera and Tsholotsho districts and one urban clinic in Harare.
Sensitivities for finger prick-DBS were 100%, 97%, 85% and 89% and specificities were 13% 25%, 74% and 91% for thresholds of 400, 1,000, 5,000 and 10,000 copies/ml (Table 1).
Aim To assess the diagnostic accuracy of dried blood spot (DBS) samples from finger prick and EDTA-whole blood compared to plasma for the quantification of HIV-1 RNA viral load using the COBAS AmpliPrep/COBAS TaqMan HIV-1 v2.0 (CAP/CTM).
Sensitivity, specificity, correlation and agreement analysis were used to assess diagnostic performance using different thresholds for virologic failure. External Quality Control was undertaken with the CDC HIV-1 viral load proficiency testing panel and results obtained were satisfactory (>90% agreement). Positive [high and low] and negative controls were also included in each batch testing to ensure validity of results.
Fig. 1 (A) shows a moderate correlation between finger prick DBS and plasma (R2=0.49). For finger prick DBS the Bland-Altman analysis revealed an overall substantial difference to plasma mean (SD) difference of 0.85 0.93 log copies/ml and limits of agreement showed a wide variation (95% LOA -0.96 log10 – 2.6 log10 copies/ml) (Fig 1 B). 60% of the differences were higher than 0.5 log10 and 19% higher than 1.0 log10 compared to plasma. Similar results were obtained for venous blood DBS. ( Table 1). Sensitivity and Specificity using different viral load thresholds
±70 μL of capillary and EDTA blood (air dry min. 3h)
+1000 μL SPEX
560C, 10 min, 1000 rpm
detach and transfer 1 spot (Ø 12 mm)
Conclusion The COBAS® AmpliPrep/ COBAS TaqMan® is not suitable for measuring viral load on DBS samples using current extraction techniques. The low specificity of viral load measured on DBS compared to plasma would result in a higher rate of false-positive results, and lead to patients being switched to a second-line regimen inappropriately. The low specificity is most likely due to the contribution of proviral DNA from polymorphonuclear (PMN) cells present in whole blood. Falsely elevated viral load results in DBS samples due to the detection of proviral DNA could be avoided by using RNA-specific platforms such as NASBA (that do not amplify DNA given the isothermal amplification). As the COBAS® AmpliPrep/ COBAS TaqMan® is among the most widely used analysers for the determination of HIV-1 viral load, in resourcelimited settings, ways need to be found to overcome the low specificity with DBS specimens. One possible solution that warrants research is the use of DNAse to remove contaminating DNA.
Fully automated nucleic acid extraction and amplification/detection
Ethical approval for this study was obtained through the Research Council of Zimbabwe and the MSF External Review Board.
We thank all study participants and we are grateful to NMRL laboratory staff in Harare for carrying out laboratory testing and MSF and MoH clinical staff for patient recruitment. This work was funded by Médecins Sans Frontières Operational Centres of Brussels and Barcelona.
COBAS® AmpliPrep/COBAS® TaqMan® HIV-1 Test Dried Fluid Spot Procedure RUO .