Assessing the fitness of a dual-antiviral drug resistant human influenza virus in the ferret model

Abstract

Influenza antivirals are important tools in our fight against annual influenza epidemics and future influenza pandemics. Combinations of antivirals may reduce the likelihood of drug resistance and improve clinical outcomes. Previously, two hospitalised immunocompromised influenza patients, who received a combination of a neuraminidase inhibitor and baloxavir marboxil, shed influenza viruses resistant to both drugs. Here-in, the replicative fitness of one of these A(H1N1)pdm09 virus isolates with dual resistance mutations (NA-H275Y and PA-I38T) was similar to wild type virus (WT) in vitro, but reduced in the upper respiratory tracts of challenged ferrets. The dual-mutant virus transmitted well between ferrets in an airborne transmission model, but was outcompeted by the WT when the two viruses were co-administered. These results indicate the dual-mutant virus had a moderate loss of viral fitness compared to the WT virus, suggesting that while person-to-person transmission of the dual-resistant virus may be possible, widespread community transmission is unlikely.

Trial registration: ClinicalTrials.gov NCT03684044.

Conflict of interest statement

These authors declare the following competing interests: K.K, S.W., and A.C.H are employees of F. Hoffmann-La Roche. All other authors declare no competing interests.

© 2022. The Author(s).

Figures

Fig. 1. Virus growth in single and multi-stage replication cycles.

MDCK cells were infected with virus isolates containing one or both NA-H275Y and PA-I38T mutations. a Single-stage replication assay (MOI 5) with sampling every 2 h for a total of 12 h. b Multi-stage replication assay (MOI 0.01) and sampled at 6, 24, 30, 48, 54, 72, 78, 96, 102 h. Infectious virus in the supernatant was quantified by TCID50. Each symbol represents the mean amount of virus present at each time point and the error bars depict one standard deviation (SD). Data in each graph is the combination of three independently conducted experiments performed in triplicate. Limit of detection (LOD) is shown as a dotted horizontal line, all values below LOD equal two log10TCID50/mL. No statistically significant differences (adjusted p < 0.05) were observed (Two-way ANOVA with Sidak’s multiple comparisons).

Fig. 2. Viral shedding and lung lobe viral load of WT and NA-H275Y + PA-I38T virus isolate infected ferrets.

Ferrets (n = 8) infected by the intranasal route with 4 log10TCID50/mL (in 500 µL of PBS) with WT (d1 isolate, black) or NA-H275Y + PA-I38T (d10 isolate, red) purified virus isolates were nasally washed daily for 9 days. Half of the ferrets were culled at day five post-inoculation to harvest the five major lung lobes. a The viral shedding from the ferret nasal washes was measured by TCID50 assay. b Viral titre of each homogenised lung lobe aliquot wase enumerated using a TCID50 assay, the fraction representing the number of major lung lobes with detectable viral titres (above LOD) per ferret. Each dot represents the titre in a single lung lobe and the line shows the mean. Error bars represent one SD. LOD is shown as a dotted horizontal line, all values beyond LOD equal zero viral titre. Adjusted p-values and asterisks show statistically significant (p < 0.05) differences (Two-way ANOVA with Sidak’s multiple comparisons).

Fig. 3. Infectious viral shedding in nasal washes from ferrets infected with WT or NA-H275Y + PA-I38T clinical isolates and in ferrets infected by airborne virus exposure.

Donor ferrets (n = 4) were infected by the intranasal route with 5 log10TCID50/mL of pure WT or NA-H275Y + PA-I38T clinically isolated virus and co-housed with naïve recipient ferret 1 (RF1) separated by an airborne virus-permeable barrier at day 1 (D1) post inoculation. Infected RF1s were then co-housed with RF2 ferrets, enabling a second airborne transmission event to occur. Viral titre mean from nasal washes from donor (a), RF1 (b) and RF2 ferrets (c), by TCID50 assay. Fractions above each symbol represent the number of ferrets shedding virus above LOD (dotted horizontal line), all values below LOD equal zero. No statistically significant (adjusted p < 0.05) differences was observed (Sidak’s Two-way ANOVA multiple comparisons). d Area under the curve analysis was performed for each individual ferret (n = 4) and overall viral shedding was compared between WT (black) and NA-H275Y + PA-I38T (D.Mt, red). e Duration of infectious viral shedding above LOD was tallied for each ferret. f Estimation plot of the combined duration of viral shedding for all donor, RF1 and RF2 ferrets was analysed by Unpaired t-test to display a 95% confidence interval of the difference between the mean duration of shedding for WT and NA-H275Y + PA-I38T infected ferrets. Error bars indicate one SD. Statistically significant differences (p < 0.05) were compared by Tukey’s Ordinary One-way ANOVA.

Fig. 4. Infectious virus titres and pyrosequencing analysis of ferret nasal washes from 50:50 WT:NA-H275Y + PA-I38T clinical isolate inoculated donor and airborne transmission infected recipient ferrets.

Donor ferrets (n = 4) were infected via the intranasal route with 5 log10TCID50/mL with 50:50 WT: NA-H275Y + PA-I38T clinically isolated virus and co-housed with naïve recipient ferrets (RF1) virus separated by an airborne virus permeable barrier, infected RF1s were then co-housed with RF2 ferrets. Nasal wash samples were titrated by TCID50 assay for each Group A-D Donor-RF1-RF2 transmission pairs, LOD is shown as a dotted horizontal line, and all values beyond LOD equal zero viral titre (a). The relative proportion of virus encoding WT PA (white bars) and PA-I38T (dark blue bars) and WT NA (white bars) and NA-H275Y (light blue bars) were determined by pyrosequencing analysis and are shown side-by-side in bar charts for each donor (b), RF1 (c) and RF2 (d) ferret.

Fig. 5. Summary of whole-genome sequencing on ferret inoculum and nasal wash samples from pure WT, pure NA-H275Y + PA-I38T and 50% WT: 50%NA-H275Y + PA-I38T isolate infected ferrets.

Ferret inoculum and nasal wash samples post inoculation/exposure (PI/E) from pure WT, pure NA-H275Y + PA-I38T or 50% WT: 50%NA-H275Y + PA-I38T virus infected ferrets were sequenced using WGS. Five amino acid (AA) changes (across the HA, NA, PA and PB1 genes) distinguished ferret inoculum WT and NA-H275Y + PA-I38T virus genotypes. The bar charts display the proportion of AA codon changes detected by NGS for each sample associated with the WT (white) and NA-H275Y + PA-I38T (black) virus inoculum for each gene loci (along X-axis; HA-294, NA-275, PA-38, PA-325, PB1-129). Percentage frequency of AA codons is determined from variant bases called with a read depth over 1000 and frequency of greater than 5%. Created with BioRender.com.