Sampling variability between two mid-turbinate swabs of the same patient has implications for influenza viral load monitoring

Abstract

Background: With the clinical development of several antiviral intervention strategies for influenza, it becomes crucial to explore viral load shedding in the nasal cavity as a biomarker for treatment success, but also to explore sampling strategies for sensible and reliable virus collection.

Findings: In this study, 244 patients suffering from Influenza like Illness and/or acute respiratory tract infection were enrolled. Sampling was done using mid-turbinate flocked swabs and two swabs per patient were collected (one swab per nostril). The influenza A viral loads of two mid-turbinate flocked swabs (one for each nostril) per patient were compared and we have also assessed whether normalization for human cellular DNA in the swabs could be useful. The Influenza mid-turbinate nasal swab testing resulted in considerable sampling variability that could not be normalized against co-isolated human cellular DNA.

Conclusions: Influenza viral load monitoring in nasal swabs could be very valuable as virological endpoints in clinical trials to monitor treatment efficacy, in analogy to HIV, HBV & HCV viral load monitoring. However, the differences between left and right nostrils, as observed in this study, highlight the importance of proper sampling and the need for standardized sampling procedures.

Figure 1

Variability plot of influenza A viral load or RnaseP DNA as observed for two mid-turbinate swabs taken from the same patient (n = 244). A: Influenza A viral load (r2=0.733): One hundred samples have an InfA VL > LLOQ for both swabs (r2 = 0.183). Dotted lines represent upper limit of quantification (ULOQ) = 10.3 log10 copies/ml & lower limit of quantification (LLOQ) = 4.3 log10 copies/ml. Viral load values below LLOQ are plotted in the respective axes or on the origin. B: RnaseP DNA (r2 = 0.242).

Figure 2

Plot of the RNaseP Ct values versus InfA log ₁₀ copies/ml (n = 488; two swabs/patient) (r ²  = 0,009 and slope = -0,06). Dotted lines represent LLOQ & ULOQ of the InfA viral load assay.

Figure 3

Correction of the influenza A viral load for amount of human cells in the swabs. A: Process flow. B: Comparison of the influenza A viral load of the swabs from both nostrils before and after correction for human cell content (n=100). Values < LLOQ are omitted. The Swab A and B InfA viral load (corrected for human cellular DNA) were classified into 4 groups delimited by the first, second and third quartile and the relative frequency for each group were tabulated. (VL=viral load).