Serological evidence of possible high levels of undetected transmission of Zika virus among Papua New Guinea military personnel, 2019

Highlights • In total, 208 Papua New Guinea military personnel (PNGMP) participated in this survey.• The overall seroprevalence of Zika virus IgG using indirect enzyme-linked immunosorbent assay was 67%.• The overall seroprevalence of Zika virus by neutralizing assay was 65%.• Five of 19 anti-ZIKV-IgM+ samples met the criteria of the World Health Organization for confirmed ZIKV infection.• An undetected Zika virus outbreak may have occurred in PNGMP in 2019.


Introduction
Zika virus (ZIKV) infection is associated with congenital neurological abnormalities, such as fetal microcephaly and Guillain-Barré syndrome ( Baker et al., 2022 ). ZIKV infection can cause an acute febrile illness with symptoms of fever, rash, joint pain and conjunctivitis, which may be misdiagnosed as malaria, dengue or chikungunya in co-circulation regions ( Haby et al., 2018 ). Retrospective testing of samples collected from febrile patients during a dengue and malaria outbreak in Papua New Guinea (PNG) in 2014-2016 identified six patients infected with ZIKV with no history of travel outside of PNG ( World Health Organization, 2016 ).
Due to a lack of testing capability, the actual incidence of ZIKV infection in PNG remains largely unknown. Currently, laboratory diagnosis of ZIKV infection depends on the detection of anti-ZIKV-IgM antibody, which normally appears in serum 5-7 days after disease onset. The au-  samples were also tested for anti-dengue Nabs with strains of dengue-1 Hawaii , dengue-2 NGC , dengue-3 H-87 and denuge-4 H-241 . Detailed methods are described in Appendix 1 (see online supplementary material).

Discussion and conclusions
Anti-ZIKV-IgM typically develops by the end of the first week after symptom onset, and remains detectable for approximately 3 months ( Griffin et al., 2019 ). The present data, showing that five PNGMP met the WHO criteria for recent ZIKV infection, and a prevalence of anti-ZIKV-Nab of 65% in PNGMP, suggest that there may have been significant, undetected transmission of ZIKV in PNGMP recently.
Seroprevalence rates of approximately 71% for anti-ZIKV ELISA IgG/M antibodies and approximately 65% for anti-ZIKV-Nabs among PNGMP are comparable with rates in French Polynesia (66%, ELISA IgG) ( Aubry et al., 2017 ) and Yap Island, Micronesia (73%, ELISA IgG/IgM) ( Duffy et al., 2009 ) where ZIKV is endemic. It remains unclear why ZIKV-related microcephaly has not been reported in PNG despite the high seroprevalence of antibodies in PNGMP. The shortage of reliable diagnostic, reporting and monitoring systems that track virus transmission may be a partial explanation. High endemicity of dengue in PNG may also be a contributor, as a high level of multi-type dengue virus antibody and anti-dengue CD8 + T cells may be protective against congenital ZIKV syndrome ( Wen et al., 2017 ;Pedroso et al., 2019 ). Another hypothesis is that phenotypic changes in Asian lineage ZIKV strains may have led to differing disease outcomes ( Weaver et al., 2016 ). Differences in ZIKV exposures between PNGMP and the general population may also be a factor.
The higher proportion of samples showing anti-ZIKV-IgG positivity compared with anti-ZIKV-Nab positivity could be due to the endemicity of other flavivirus that are antigenically closely related to ZIKV, such as dengue virus and JEV. The finding of two samples that were anti-ZIKV-IgM + but anti-ZIKA-Nab − indicates that the current commercially available ELISA detection kits for ZIKV may not be suitable for diagnostic or seroprevalence survey purposes in dengue-and JEV-endemic areas, such as PNG, due to possible serological cross-reactivity among flaviviruses ( Maeki et al., 2019 ;Montecillo-Aguado et al., 2019 ) ( Table 2 ). All samples that tested anti-ZIKV-IgG/IgM + on ELISA should be confirmed by neutralizing assay for diagnostic/surveillance purposes.
This preliminary finding requires further support from additional investigations, as the present study had a small sample size and only PNGMP were included. The authors intend to expand their arbovirus surveillance programme in PNG to include investigation of circulating ZIKV strains, mosquito behaviours and antibody prevalence amongst the entire population of PNG in order to better understand the potential risk of ZIKV transmission in PNG.