Options to Expand HIV Viral Load Testing in South Africa: Evaluation of the GeneXpert® HIV-1 Viral Load Assay
Options to Expand HIV Viral Load Testing in South Africa: Evaluation of the GeneXpert® HIV-1 Viral Load Assay
复制标题
DOI：
10.1371/journal.pone.0168244
复制
发表时间：
2016-12-16
期刊：
影响因子：
3.7
通讯作者：
Stevens, Wendy
中科院分区：
文献类型：
--
作者：
Gous, Natasha;Scott, Lesley;Stevens, Wendy
背景需要扩大HIV病毒载量(VL)检测服务，以满足监测患者接受抗逆转录病毒治疗的更多目标。南非目前每年在16个高度集中的自动化高通量实验室测试400万英镑的VL。Xpert HIV-1 VL检测(Cepheid)与国内的罗氏Cobas Taqmanv2和雅培HIV-1RT进行了评估，以调查利用GeneXpert的随机访问、多价能力和Xpert MTB/RIF检测在南非(207个地点)已经建立的足迹来扩大VL检测的选择。方法将Xpert HIV-1VL与Taqmanv2(n=155)和Abbott HIV-1 RT(n=145)的精密度、准确性(一致性)和临床误判(1000cp/ml)进行比较。在DBS(n=145)和全血(n=147)上进一步测试Xpert HIV-1VL的错误分类。结果Xpert HIV-1VL在标准化冰冻血浆板(n=42)上与预测平台的符合率为100%，与Taqmanv2和Abbott HIV-1 RT相比，总体相似性CV分别为1.5%和0.9%。在配对血浆临床标本上，Xpert HIV-1VL与Taqmanv2(新鲜)和Abbott HIV-1 RT(冷冻)相比，在1000cp/ml阈值下具有较低的偏倚(SD为0.32-0.37logcp/ml)和3%的误判。全血和DBS上的HIV-1VL错误分类增加了高达14%，无效率增加。所有标本检测均易于进行，并与Xpert结核分枝杆菌/RIF结核病同步检测在同一仪器上兼容。结论Xpert血浆HIV-1VL可以与南非现有的预测平台互换使用。全血和DBS检测需要进一步研究，但GeneXpert的多价性为扩展VL检测服务提供了解决方案。
BackgroundExpansion of HIV viral load (VL) testing services are required to meet increased targets for monitoring patients on antiretroviral treatment. South Africa currently tests > 4million VLs per annum in 16 highly centralised, automated high-throughput laboratories. The Xpert HIV-1 VL assay (Cepheid) was evaluated against in-country predicates, the Roche Cobas Taqmanv2 and Abbott HIV-1RT, to investigate options for expanding VL testing using GeneXpert's random access, polyvalent capabilities and already established footprint in South Africa with the Xpert MTB/RIF assay (207 sites). Additionally, the performance of Xpert HIV-1VL on alternative, off-label specimen types, Dried Blood Spots (DBS) and whole blood, was investigated.MethodPrecision, accuracy (agreement) and clinical misclassification (1000cp/ml) of Xpert HIV-1VL plasma was compared to Taqmanv2 (n = 155) and Abbott HIV-1 RT (n = 145). Misclassification of Xpert HIV-1VL was further tested on DBS (n = 145) and whole blood (n = 147).ResultsXpert HIV-1VL demonstrated 100% concordance with predicate platforms on a standardised frozen, plasma panel (n = 42) and low overall percentage similarity CV of 1.5% and 0.9% compared to Taqmanv2 and Abbott HIV-1 RT, respectively. On paired plasma clinical specimens, Xpert HIV-1VL had low bias (SD 0.32-0.37logcp/ml) and 3% misclassification at the 1000cp/ml threshold compared to Taqmanv2 (fresh) and Abbott HIV-1 RT (frozen), respectively. Xpert HIV-1VL on whole blood and DBS increased misclassification (upward) by up to 14% with increased invalid rate. All specimen testing was easy to perform and compatible with concurrent Xpert MTB/RIF Tuberculosis testing on the same instrument.ConclusionThe Xpert HIV-1VL on plasma can be used interchangeably with existing predicate platforms in South Africa. Whole blood and DBS testing requires further investigation, but polyvalency of the GeneXpert offers a solution to extending VL testing services.