Persistence of H7N9 virus antibody response 2 years after infection

Abstract We measured antibodies against H7N9 virus 2 years after infection in 14 patients who were infected during October 2016‐September 2017. Approximately 2 years after infection, antibody titers ≥10 were detectable in 13 (92.9%) patients. Three (21.4%) of 14 patients had hemagglutination inhibition titers ≥40, and their geometric mean titer (GMT) was 20 (95% CI 15.7‐28.1), whereas 10 (71.4%) and all 14 (100%) of the 14 patients had titers ≥40, and GMTs at 34.4 (95% CI 25.7‐51.2) and 73.45 (54.7‐106.7) for neuraminidase inhibition and microneutralization antibodies, respectively. Our findings suggest that H7N9 infection may induce long‐term antibody response at least 2 years after infection.

a mass poultry vaccination program was implemented. This campaign was highly effective in preventing H7N9 virus infection in both poultry and humans. 2 However, novel, highly pathogenic subtypes (H7N9 and H7N2) adapted to ducks now pose new challenges to public health. [2][3][4] Among patients with H5N1 virus infection, neutralizing antibodies are thought to persist for nearly 5 years, 5 although few patients have been studied. Antibodies induced by natural infection with the 2009 pandemic H1N1 virus persist for at least 15 months. 6 In our cohort of H7N9 patients, 7 antibodies against H7N9 virus were detected in the majority of patients about one year after symptom onset, although the antibodies decayed over time. However, the duration of antibody responses in patients beyond 2 years has not been previously studied. In this report, we examined antibody responses against H7N9 virus among 14 patients from our prior cohort, 2 years after their symptom onset. 7

| Study design and subjects
In our previous study, 7

| Serological testing
The hemagglutination inhibition (HI) assay, the enzyme-linked lectin assay to measure neuraminidase inhibition (NI) antibodies, and a microneutralization (MN) assay were used to measure antibodies as described in our previous study, 7 We defined the HI titer as the reciprocal of the highest serum dilution that completely inhibited hemagglutination, the NI titer as the reciprocal of the highest serum dilution that exhibited 50% inhibition concentration, and the MN titer as the reciprocal of the highest serum dilution that yielded >50% neutralization. For final titers <10 of HI, NI, and MN antibodies, we assigned a value of 5 as seronegative, and a titer ≥40 was reported as 50% protective threshold.

| Quality control
Although this study is the continuation of our previous work reported, and assays for antibodies detection were consistent in both study, 7 the time of detection was not synchronized, which may lead to variation in results. Thus, considering the variation and the specificity of the assays to measure antibodies to H7N9 virus, five serum samples from these 14 patients at each time point of acute phase, 100, 200, and 300 days after infection and five serum samples from control subjects in our previous study were used as positive and negative controls when testing the serum samples from these 14 patients.

| RE SULTS
Among 22 patients who participated in the last follow-up visit (about 1 year after infection), 7 14 consented to this new follow-up testing, with a median follow-up of 850 days (interquartile range 841-865) after symptom onset. Participants ranged from 41 to 77 years of age (median 60.5 years), and 6 (42.9%) were female (Table 1). Two  Figure 1A. Overall, approximately 2 years after symptom onset, HI and NI GMTs substantially declined and were lower than the titer of 40 and the GMTs in the acute phase. In contrast, although the MN GMTs declined 2 years after infection, yet they remained considerably above the GMTs in the acute phase and the titer of 40. We further examined the changes in antibody titers approximately 2 years after symptom onset in comparison with the titers at 300 days after symptom onset. Most patients maintained antibody titers compared with their titers 300 days after infection, but HI assay titers were lower ranging from 5 to 40, while NI and MN antibodies were relatively high, ranging from 10 to 80 and 40 to 160, respectively. In particular, four patients (patients 9, 10, 13, and 23) experienced a decrease in HI antibody titer, three (patients 3, 10, and 21) experienced a decrease in NI antibody titer, and five (patients 3, 9, 10, 13, and 18) experienced a decrease in MN antibody titer approximately 2 years after symptom onset. However, none of these patients had antibody titers that were considered as seronegative.

| D ISCUSS I ON
In this study, we examined the levels of virus-specific antibodies approximately 2 years after infection. Detectable levels of HI, NI, and MN antibodies were observed in all but one patient who has already had a HI antibody titer <10 at the last follow-up visit (about 300 days after infection). While over 70% of patients maintained NI and MN titer ≥40 2 years after infection, this was observed in only 21.4% of patients for HI antibody. We also found that most patients (10 for HI antibody, 11 for NI antibody, and 9 for MN antibody) maintained antibody titers compared with their titers 300 days after infection.
Although few patients had a decrease in one or more antibody titers, no patient had a decrease in antibody titer to become totally seronegative. These results suggest that natural infection of the H7N9 virus could induce the persistence of antibody response at least 2 years after infection.
An HI titer of 40 is associated with 50% protection of patients from seasonal influenza illness. [8][9][10] Evidence for the contribution of NI antibody to the protection against seasonal influenza, independent of the effect of HA antibody, has been suggested by patterns of infection during the 1968 pandemic and more recently using multivariable regression analyses. [11][12][13] Given the relative high NI  14,15 In our previous cohort study, 7 22 patients who were laboratory-confirmed H7N9 virus infection were prospectively evaluated the dynamic changes in their antibody response at the acute phase, 100, 200, and 300 days after symptom onset. In this study, 14 of these 22 patients were followed up. Although the sample size is small, this is the largest cohort of H7N9 patient to study long-term antibody response of H7N9 patients, and all 14 patients had a complete dataset. In conclusion, this study provides evidence that infection with the H7N9 virus could induce long-term antibody response at least 2 years after infection.

ACK N OWLED G EM ENTS
We thank all patients for their participation in the study and the staff of the local Centers for Disease Control and Prevention, whose collaboration made this study possible. This study was

CO N FLI C T O F I NTE R E S T
The authors declare no conflict of interest.