The Centers for Disease Control and Prevention (CDC) receives influenza-positive specimens collected globally for virological surveillance. Viral genomes are analyzed using next-generation sequencing (NGS) to identify strains of epidemiologic, virologic, and clinical importance (9). To supplement US national antiviral surveillance, pyrosequencing is used by public health laboratories to screen additional viruses either in-house or by the National Influenza Reference Center (10).

During the 2019–20 influenza season, the pH1N1 subtype predominated in the United States. Later in the season, fewer influenza samples were identified, likely because of COVID-19 pandemic mitigation strategies. Of 951 pH1N1 isolates collected nationwide during October 2019–September 2020, 4 (0.4%) had the NA-H275Y marker. Supplemental surveillance, conducted on 282 viruses from 18 states collected November 2019–March 2020, detected another 6 (2.1%) NA-H275Y viruses, bringing the total detected nationwide to 10 (10/1,233; 0.8%). Of these, 4 (7.7%) were detected among 52 viruses from Texas.

An investigation into a potential epidemiologic link revealed that these 4 virus isolates were collected from the same location, a border detention center in Webb County, Texas, on the same day (January 24, 2020). In January 2020, an influenza outbreak took place there; 8 cases were reported during January 19–28, 2020. All patients showed similar symptoms, such as fever, cough, sore throat, and body aches. Oseltamivir was prescribed on the same day, following specimen collection. Only 4 nasopharyngeal specimens from this outbreak were available for analysis; these samples were collected from men 25–59 years of age. San Antonio Metropolitan Health District Laboratory (San Antonio, TX, USA) conducted the initial diagnostic testing by real-time reverse transcription PCR and determined the cycle threshold (C_t) values as 16.7–25.9, indicating relatively high viral loads. NGS analysis showed that the viruses had the oseltamivir resistance-conferring mutation, NA-H275Y. To expand testing, the San Antonio Laboratory submitted to CDC all remaining pH1N1 positive respiratory specimens (n = 36), collected from Webb County residents during November 2019–March 2020. These specimens were collected from 19 male and 17 female patients with a median age of 6 years (range 0–65 years); C_t values were 21.1–37.5. Pyrosequencing analysis concluded that there were no additional specimens with the NA-H275Y mutation.

A unique genomic signature can help in tracing the origin and spread of viruses in an outbreak. NGS analysis (11) revealed that the codon-complete genomes of the 4 cluster viruses were identical at a nucleotide level. Although the chain of transmission is unknown, considering the close-contact setting, this finding might suggest that an oseltamivir-resistant virus was transmitted from a single source. The cluster viruses shared 2 rare substitutions, PB1-Q687H and PB2-R251G, the combination of which was not found in other sequences from the National Center for Biotechnology Information and GISAID (https://www.gisaid.org; accessed October 22, 2020). Therefore, this virus has a unique genomic signature that has not been detected in viruses collected in Texas or elsewhere.

We isolated the 4 cluster viruses and propagated them in MDCK cells, followed by sequence confirmation. We tested the virus isolates for susceptibility to NA inhibitors using the NA inhibition assay (10) and they showed highly reduced inhibition by oseltamivir (≈1,300-fold) and peramivir (≈350-fold), and normal inhibition by zanamivir and laninamivir (Table 1). Markers associated with resistance to the polymerase inhibitor, baloxavir, were not detected. To confirm baloxavir susceptibility, we tested viruses by a high-content imaging-based neutralization test (HINT) (12). The concentrations of drug needed to inhibit infection by 50% fell in a low nanomolar range (mean 1.86 nM, SD 0.26), consistent with a susceptible phenotype.

Figure 1

Figure 1. Evolutionary relationships of the HA (A) and NA (B) genes of influenza A(H1N1)pdm09 viruses circulating in the United States during the 2019–20 influenza season compared with reference viruses. We generated...

HA phylogenetic analysis placed the cluster viruses into the 6B.1A5A-156K subclade, which shares additional amino acid substitutions K130N, L161I, V250A in HA1 and E179D in HA2 (Figure 1, panel A). HA substitutions at residue 156 have been sporadically detected and shown to affect antigenicity, but no widespread circulation of such viruses was observed before summer 2019. In the United States, viruses with HA-5A-156K were first detected in fall 2019 and prevailed among pH1N1 by February 2020. Conversely, the circulation of another recently emerged subclade, 6B.1A5A-187A, 189E, had decreased by winter 2020. A/Hawaii/70/2019 virus (HI/70) from subclade 5A-187A, 189E, was selected as a pH1N1 vaccine component for the 2020–21 Northern Hemisphere influenza season; A/Wisconsin/588/2019 (WI/588), representing 5A-156K, was selected for the 2021 Southern Hemisphere vaccine.

We assessed the antigenicity of pH1N1 viruses representing distinct HA genetic groups circulating in the United States for antigenic relatedness by HINT and hemagglutination inhibition (HI) assays, using postinfection ferret antiserum (13,14). We used viruses A/Idaho/07/2019 (ID/07), HI/70, and WI/588, representing recent vaccines, and their homologous ferret antiserum as references. In the HINT assay, we found that the antiserum raised to ID/07 showed poor reactivity (65–78-fold reduction) to viruses with HA-N156K, including the cluster. The HI/70 antiserum reacted even more poorly (315–429-fold) against this group but maintained good reactivity to other HA groups. Antiserum raised to WI/588 (5A-156K) had very high titers against viruses of the same group, including the cluster, and reacted poorly (40–612-fold) to viruses of other groups (Table 2). Results obtained by the conventional HI assay corroborated the HA antigenic drift detected by HINT (Table 2). While analyzing antigenicity of pH1N1, it is prudent to consider that ferret antiserum may preferentially detect changes at HA antigenic site Sa, where N156K resides, compared with site Sb, where D187A is located (15). Nevertheless, the findings of this study and other reports indicate that viruses carrying HA-N156K may escape humoral immunity elicited by previous infections and vaccinations.

Figure 2

Figure 2. In vitro replicative fitness of influenza A(H1N1)pdm09 cluster and control viruses. The growth kinetics of the 2 viruses, the cluster virus A/Texas/137/2020, and A/New York/19/2020, were assessed using MDCK (A)...

We assessed the in vitro replicative fitness of the cluster virus A/Texas/137/2020, and A/New York/19/2020, which has identical HA and NA amino acid sequences except for NA-H275Y. These 2 viruses had very similar growth kinetics in MDCK and humanized MDCK (hCK) cells. In MDCK cells, the growth curves were alike at all time points (Figure 2, panel A). In hCK cells, the NA-H275Y-containing virus had better growth at 8 hours, but its titers tapered off slightly at later times (Figure 2, panel B).

Phylogenetic analysis of NA (Figure 1, panel B) showed that the cluster viruses had NA similar to the majority of viruses in the 5A-156K group, including characteristic substitutions NA-Y66F and NA-N222K. However, their NA contained a rare substitution, NA-V80M. Studies to evaluate the effects of these changes on HA–NA functional balance are ongoing.

Although no evidence of oseltamivir-resistant virus transmission outside the detention center was found, the properties of the cluster viruses are concerning. They belong to an HA antigenically drifted group, and escape from preexisting immunity may contribute to the spread of oseltamivir-resistant viruses in coming seasons.