Comparison of performance between Fast Track Diagnostics Respiratory Kit and the CDC global reference laboratory for influenza rRT‐PCR panel for detection of influenza A and influenza B

Abstract Background Reliable diagnostics are a key to identifying influenza infections. Objectives Our objectives were to describe the detection of influenza among severe acute respiratory infection (SARI) cases, to compare test results from the Fast Track Diagnostics (FTD) Kit for influenza detection to the Centers for Disease Control (CDC) human influenza virus detection and characterization panel, and to assess seasonality of influenza in Burkina Faso. Methods Nasopharyngeal and oropharyngeal specimens from SARI cases (hospitalized patients with fever, cough, and onset in the previous 10 days) were tested using the FTD‐33 Kit and the CDC rRT‐PCR influenza assays. We assessed sensitivity and specificity of the FTD‐33 Kit for detecting influenza A, influenza B, and the influenza A(H1N1)pdm09 strain using the CDC human influenza rRT‐PCR panel as the gold standard. Results From December 2016 to February 2019, 1706 SARI cases were identified, 1511 specimens were tested, and 211 were positive for influenza A (14.0%) and 100 for influenza B (6.6%) by either assay. Higher influenza circulation occurred between November and April with varying peaks of influenza A and influenza B. Sensitivity of the FTD‐33 assay was 91.9% for influenza A, 95.7% for influenza B, and 93.8% for A(H1N1)pdm09 subtype. Specificity was over 99% for all three tests. Conclusions Our study indicates that Burkina Faso has one peak of influenza each year which is similar to the Northern Hemisphere and differs from other countries in West Africa. We found high concordance of influenza results between the two assays indicating FTD‐33 can be used to reliably detect influenza among SARI cases.


Influenza viruses cause acute respiratory infections in all parts
of the world and can lead to hospitalization and death. An estimated 5%-10% of adults and 20%-30% of children worldwide are affected by seasonal influenza each year. 1 A recent model by Iuliano et al. estimated that 4-8.8 per 100000 influenza-associated deaths occur annually around the world, with the highest rates occurring in sub-Saharan Africa, the western Pacific, and South-East Asia. 2 Surveillance for influenza that includes accurate laboratory diagnostics plays a critical role for annual vaccine strain selection, identification of novel strains, and quick recognition of increased influenza circulation that could signal a pandemic. One report on the seasonality of influenza which included influenza-like illness (ILI) and/or severe acute respiratory infection (SARI) surveillance in eight West African countries described two peaks; however, data included in this report from Burkina Faso did not include SARI cases, and country-specific seasonality was limited. 3 Clinical presentation of influenza is similar to that of many other viral, bacterial, and fungal pathogens causing acute respiratory illnesses; thus, diagnostics play a key role in identifying which acute respiratory infections are caused by influenza. The Centers for Disease Control and Prevention (CDC) reverse transcription quantitative PCR (rRT-PCR) assay for detection and characterization of influenza 4 is utilized in influenza reference laboratories around the world and is considered a reliable standard for influenza testing by the World Health Organization. 5 Respiratory illnesses are caused by a variety of pathogens, and incorporating multi-pathogen diagnostic methods into respiratory disease surveillance platforms can help ministries of health determine potential causes of outbreaks or know when respiratory pathogens are circulating. Fast Track Diagnostics Respiratory Pathogens 33 Respiratory (FTD-33) Kit is a multi-pathogen platform using real-time reverse-transcriptase polymerase chain reaction (rRT-PCR) to detect 33 pathogens in humans with acute respiratory illness. 6 It is important that the individual targets in a multipathogen platform perform well compared with single-target assays.
Burkina Faso is a sub-Saharan country in West Africa with a population of approximately 20 million people. One of the leading causes of death is lower respiratory tract infections, which account for more than 14% of deaths annually. 7 The country implemented a sentinel surveillance system of ILI in 2010 and in late 2016, began conducting SARI sentinel surveillance to better understand respiratory pathogens circulating and their disease burden among the community. The country has a national influenza reference laboratory with established capacity to conduct influenza testing using the CDC panel and multipathogen testing using the FTD-33 Kit. 6 Our objectives were to describe the circulation of influenza among SARI cases in Burkina Faso and to determine sensitivity and specificity of the FTD-33 Kit for influenza detection compared with the CDC human influenza rRT-PCR detection and characterization panel.

Nucleic acid extraction
Total nucleic acid (TNA) was extracted from 400 µL of Universal Transport Media containing the NP and OP specimens and eluted into 110 µL using the EZ1 Advanced XL Instrument with the EZ1 Mini Viral 2.0 Kit (Qiagen) following the manufacturer's instructions.
Seven specimens were extracted at a time with an extraction control consisting of nuclease-free water.

Respiratory viruses, bacteria, and fungi real-time RT-PCR detection
The TNA extracts from NP/OP specimens were screened individually for detection of respiratory pathogens using eight multiplex reverse-transcriptase real-time polymerase chain reactions (rRT-PCR) from the FTD-33 Test Kit (Fast Track Diagnostics). The kit is used for detection of the following respiratory viruses, bacteria, and fungi: The eight multiplex rRT-PCR reactions were set up following the manufacturer's instructions. 6 Each reaction mix consisted of 10 µL TNA, 1.5 µL of oligonucleotide mix, 12.5 µL 2 X AgPath-ID TM One- Step RT-PCR buffer and 1 µL 25 X AgPath-ID TM One-Step RT-PCR Enzyme Mix (Thermo Fisher Scientific). The manufacturer's instructions recommend introducing exogenous internal control material provided with the FTD-33 Kit into each clinical specimen prior to nucleic acid extraction and testing with corresponding rRT-PCR oligonucleotide mix to monitor for PCR inhibition. This step was excluded to minimize potential contamination of primary specimen material. Instead, each specimen was tested in parallel for the presence of human ribonuclease P (RNase P). Detection of RNase P, which is ubiquitous in human cells, serves as a control for specimen extraction and PCR inhibition without manipulation of primary specimens. Reaction mixtures containing a no-template control, extraction control, and positive control (Resp21 PC or Resp33 PC2) were included for each multiplex reaction mix. An extraction control and no-template control materials were included for testing of the eight multiplex reaction mixes and the RNase P reaction mix. The positive control (Resp21 PC or Resp33 PC2) consisting of pooled plasmids 6 from the FTD-33 Kit was also included for testing of the multiplex reaction mixes and human DNA was included as the positive control for the RNase P reaction mix. All assays were tested using the Applied Biosystems 7500 Real-Time PCR Instrument (Thermo Fisher Scientific) with the following cycling conditions: 42°C for 15 minutes, 94°C for 3 minutes, 40 cycles of 94°C for 8 seconds, and 60°C for 34 seconds. Any assay or specimen with a control result deviating from the expected result was retested.

Human influenza virus real-time RT-PCR detection
Influenza viruses were detected using singleplex reverse-tran-

| Statistical analysis
We described the basic demographics of SARI cases and the fre-    however, for influenza B, discordant results were more frequently positive with the FTD-33 Kits (8 of 12 discordant results).

| D ISCUSS I ON
Our study found that the results for influenza A, influenza B, and A(H1N1)pdm09 assays for FTD-33 and the CDC rRT-PCR panel were similar. In particular, concordance of testing results was very good when there was a higher viral load (lower ct value) in the specimen. We reported sensitivity of greater than 93% and specificity of greater than 99% for all three assays. Other studies have reported on sensitivity and specificity or concordance of the FTD-33 assay compared with either other multipathogen platforms or in-house singleplex tests. [8][9][10][11][12][13] Most of these studies reported comparable results for the influenza A, influenza B, and H1N1 assays; however, two reported significant discrepancies with the influenza B assay. 8,10 Most specimens with discordant results had higher ct values, closer to the limit of detection; thus, discordant results were not unexpected given the lower concentrations of virus in the specimens. We did not find that these results differed by age-group or site, and all discordant specimens arrived in the laboratory in good condition.
We also found that in Burkina Faso, a country with distinct dry and rainy seasons, influenza did not circulate year-round, but rather circulated primarily during the same months as influenza  have implications for the number of influenza cases identified during peak seasons. We were unable to include detection of additional respiratory pathogens in this report. A further investigation of additional pathogen detections would provide additional information about other respiratory pathogens circulating in this population.

CO N FLI C T O F I NTE R E S T
The authors have no potential conflicts to disclose.

D I SCL A I M ER
The

Data may be available pending approval from the Burkina Faso
Ministry of Health.