Analytical reactivity of 13 commercially available rapid influenza diagnostic tests with H3N2v and recently circulating influenza viruses

Abstract

Objectives: Rapid influenza diagnostic tests (RIDTs) used widely in clinical practice are simple to use and provide results within 15 minutes; however, reported performance is variable, which causes concern when novel or variant viruses emerge. This study's goal was to assess the analytical reactivity of 13 RIDTs with recently circulating seasonal and H3N2v influenza viruses, using three different viral measures.

Design: Virus stocks were characterized by infectious dose (ID50 ) and nucleoprotein (NP) concentration, diluted at half-log dilutions, and tested with each RIDT and real-time RT-PCR.

Results: Strong correlation was observed between NP concentration and RIDT reactivity; however, only weak correlation was seen with ID50 or Ct values. Only four RIDTs detected viral NP at the lowest dilution for all influenza A viruses (IAV). Influenza A viruses not detected by more than one RIDT had lower NP levels. Of the 13 RIDTs, 9 had no significant differences in reactivity across IAV when compared to NP levels.

Conclusions: Previous reports of RIDT performance typically compare reactivity based on ID50 titers, which in this study correlated only weakly with proportional amounts of viral NP in prepared virus samples. In the context of the strong correlation of RIDT reactivity with NP concentration, H3N2v was found to be as reactive as seasonal circulating IAV. While these findings may not reflect clinical performance of these RIDTs, measuring NP concentration can be useful in the future to assess comparable reactivity of available RIDTs, or to assess reactivity with newly evolving or emerging viruses.

Keywords: Diagnostic; FDA; H3N2v; influenza; rapid.

Figure 1

Viruses used in this study, with measurements by TCID₅₀/ml or EID₅₀/ml, NP in μg/ml as determined by mass spectrometry, and the C_t value of the 10⁻¹ dilution. Viruses 13, 15, and 18 were quantified by TCID₅₀/ml (red numbering). All others were quantified by EID₅₀/ml. Also shown is the number of reactive rapid influenza diagnostic tests results (at least 2/3 positive) for each influenza virus dilution. A maximum of 13 test kits could be positive for each dilution.

Figure 2

Reactivity of each rapid influenza diagnostic tests across influenza virus groups. Six viruses in each group were tested at three replicates per dilution for a maximum of 18 positive results per dilution. CLIA-waived tests are marked with an *.

Figure 3

Scatter plots showing the mean log highest dilution reactive (HDR) across all rapid influenza diagnostic tests for each virus tested against (A) the log stock NP concentration, (B) the log stock ID₅₀, and (C) the 10⁻¹ dilution C_t value. The black line shows the linear regression trend line and the black dotted lines show the 95% confidence interval, along with equations and R² values for each trend line. Only viruses quantified by EID₅₀/ml were used for the trendline in B.

Figure 4

Graph showing the mean NP concentration at the highest dilution reactive for each of the virus groups (pH1N1, H3N2, H3N2v, and influenza B) for each of the rapid influenza diagnostic tests (RIDTs). * indicates that an influenza A viruses (IAV) group is significantly different from other IAV groups for that RIDT based on Tukey's HSD test. The QuickVue test was significantly less reactive with H3N2v than with pH1N1 and H3N2. The Status test was significantly more reactive with H3N2 than with pH1N1 or H3N2v. The Flu Alert test was significantly more reactive with H3N2 than with pH1N1 (H3N2v could not be statistically evaluated). The TRU FLU test was significantly more reactive with H3N2v than with pH1N1. A † represents situations in which an RIDT was reactive with 3 or less viruses in a group. CLIA indicates that an RIDT is CLIA-waived.