Distinguishing Secondary Dengue Virus Infection From Zika Virus Infection With Previous Dengue by a Combination of 3 Simple Serological Tests

Abstract

Background: The explosive spread of Zika virus (ZIKV) and associated microcephaly present an urgent need for sensitive and specific serodiagnostic tests, particularly for pregnant women in dengue virus (DENV)-endemic regions. Recent reports of enhanced ZIKV replication by dengue-immune sera have raised concerns about the role of previous DENV infection on the risk and severity of microcephaly and other ZIKV complications.

Methods: Enzyme-linked immunosorbent assays (ELISAs) based on ZIKV and DENV nonstructural protein 1 (NS1) were established to test acute, convalescent phase, and post-convalescent phase serum/plasma samples from reverse-transcription polymerase chain reaction-confirmed cases including 20 primary ZIKV, 25 ZIKV with previous DENV, 58 secondary DENV, and 16 primary DENV1 infections.

Results: ZIKV-NS1 immunoglobulin M (IgM) and immunoglobulin G (IgG) ELISAs combined can detect ZIKV infection with a sensitivity of 95% and specificity of 66.7%. The ZIKV-NS1 IgG cross-reactivity by samples from secondary DENV infection cases ranged from 66.7% to 28.1% (within 1 month to 1-2 years post-illness, respectively). Addition of DENV1-NS1 IgG ELISA can distinguish primary ZIKV infection; the ratio of absorbance of ZIKV-NS1 to DENV1-NS1 IgG ELISA can distinguish ZIKV with previous DENV and secondary DENV infections with a sensitivity of 87.5% and specificity of 81.3%. These findings were supported by analysis of sequential samples.

Conclusions: An algorithm for ZIKV serodiagnosis based on 3 simple ELISAs is proposed to distinguish primary ZIKV, ZIKV with previous DENV, and secondary DENV infections; this could be applied to serodiagnosis for ZIKV, serosurveillance, and monitoring ZIKV infection during pregnancy to understand the epidemiology, pathogenesis, and complications of ZIKV in dengue-endemic regions.

Keywords: Zika virus; cross-reactivity; dengue virus; non-structural protein 1; serological test.

Figure 1.

Results of nonstructural protein 1 (NS1) immunoglobulin G (IgG) enzyme-linked immunosorbent assays (ELISAs) for secondary dengue virus (sDENV) and Zika virus (ZIKV) with previous (wp) DENV infection panels. DENV1-NS1 (A) and ZIKV-NS1 (B) IgG ELISAs in convalescent-phase samples from sDENV, ZIKVwpDENV (Nicaragua [Nic]), probable ZIKVwpDENV (Brazil [Bra]) and negative control (NC) panels. C, Relative optical density (rOD) ratio of ZIKV-NS1 to DENV1-NS1. The sensitivity and specificity are shown based on a cutoff rOD ratio at 0.24. DENV1-NS1 (D) and ZIKV-NS1 (E) IgG ELISAs in post–convalescent phase samples (3 months to 2 years post–symptom onset) from sDENV panels. F, Positive rates of DENV1- and ZIKV-NS1 IgG ELISAs in sDENV panels over time. Dotted lines indicate cutoff rOD values for ELISAs. Data are mean of 2 experiments (each in duplicate). Two-tailed Mann-Whitney test was used to compare the 2 groups.

Figure 2.

Results of nonstructural protein 1 (NS1) immunoglobulin G (IgG) enzyme-linked immunosorbent assay (ELISAs) in sequential samples from individuals with secondary dengue virus (sDENV) and Zika virus (ZIKV) infection with previous (wp) DENV infections. A, Three cases with sDENV infection, 3–18 months post–symptom onset. B, Four blood donors with sDENV infection, 7 days to 6 months post–index day (PID). C, Five blood donors with ZIKVwpDENV infection, from index day to 3 months PID. Three cases (1, 4, and 7) seroconverted to ZIKV-NS1; 2 cases (15 and 21) had ZIKV-NS1 IgG starting from the index day. Dotted lines indicate cutoff relative optical density (rOD) values for ELISAs. Data are mean of 2 experiments (each in duplicates).

Figure 3.

Proposed algorithm of using 3 serological tests (without neutralization tests) to distinguish different Zika virus (ZIKV) and dengue virus (DENV) infections in dengue- and Zika-endemic regions in the framework of Centers for Disease Control and Prevention guidelines for laboratory diagnosis of ZIKV infection [6, 7]. Only samples tested positive or equivocal by ZIKV or DENV E protein–based immunoglobulin M (IgM) enzyme-linked immunosorbent assays (ELISAs) are included in the algorithm. The total numbers from each panel and the numbers of positive or negative based on the 3 Nonstructural Protein 1 (NS1) ELISAs are shown in parentheses. Abbreviations: IgG, immunoglobulin G; OD, optical density; pDENV, primary dengue virus infection; pZIKV, primary Zika virus infection; rOD, relative OD; sDENV, secondary dengue virus infection; ZIKVwpDENV, Zika virus infection with previous dengue virus infection.