Specificity of Dengue NS1 Antigen in Differential Diagnosis of Dengue and Zika Virus Infection

To the Editor: Circulation of new arboviruses of the genus Flavivirus poses a major problem for differential diagnosis. Zika virus, a mosquitoborne virus of the family Flaviviridae, is closely related to other arboviruses circulating in the Americas, including dengue, yellow fever, Saint Louis encephalitis, and West Nile viruses (1,2). Serologic cross-reactivity between these arboviruses is common; thus, to ensure optimal patient care and accurate epidemiologic surveillance, an effective differential diagnosis is required in regions with active transmission of dengue virus and circulation of Zika virus (2–4).


Specificity of Dengue NS1 Antigen in Differential Diagnosis of Dengue and Zika Virus Infection
To the Editor: Circulation of new arboviruses of the genus Flavivirus poses a major problem for differential diagnosis. Zika virus, a mosquitoborne virus of the family Flaviviridae, is closely related to other arboviruses circulating in the Americas, including dengue, yellow fever, Saint Louis encephalitis, and West Nile viruses (1,2). Serologic cross-reactivity between these arboviruses is common; thus, to ensure optimal patient care and accurate epidemiologic surveillance, an effective differential diagnosis is required in regions with active transmission of dengue virus and circulation of Zika virus (2)(3)(4).
Cross-reactivity between flaviviruses has been reported in antibody assays and in tests for Dengue nonstructural 1 glycoprotein (NS1) antigen. Gyurech et al. (5) reported false-positive test results for dengue NS1 antigen in a patient with acute Zika virus infection. Of the 3 NS1 tests used in that study, only the SD Bioline Dengue Duo (Standard Diagnostics, Inc., Gyeonggi-do, South Korea) showed positive results for 3 of 4 sequential serum samples from the patient.
Cross-reactivity in NS1 dengue tests (ELISA and immunochromatographic) using serum samples from patients with acute Zika virus infection would have medically significant consequences. We therefore conducted a retrospective analysis of the differential diagnosis for dengue and Zika virus infections since the beginning of the Zika virus outbreak in French Guiana, a department of France on the northeast coast of South America.
French Guiana is subject to endemoepidemic circulation of dengue and experienced a large outbreak of chikungunya in 2014. We conducted our study from December 17, 2015 (the time of biologic confirmation of the first case of Zika virus disease in French Guiana), through March 2, 2016. During that time, the incidence of dengue virus infection in French Guiana was low, and only 1 sporadic case was confirmed. We studied clinical samples collected during this period from all patients with suspected arbovirus infection.
Samples were analyzed for the differential diagnosis of dengue, chikungunya, and Zika virus disease. Zika virus diagnosis was conducted by the National Reference Centre for Arboviruses (NRC) at the Institut Pasteur of French Guiana, in Cayenne, according to the real-time reverse transcription PCR (rRT-PCR) protocol described by Lanciotti et al. (3). Dengue diagnosis was routinely performed by all medical diagnostic laboratories using various rRT-PCR techniques or dengue NS1 test kits. We used the same SD Bioline Dengue Duo test used by Gyurech et al. (5); this test was performed at the laboratory of the Centre Hospitalier de l'Ouest-Guyanais in Saint-Laurent du Maroni, French Guiana. We also used the Platelia Dengue NS1 Ag kit (Bio-Rad, Marnes-la-Coquette, France); the assay was performed at NRC.
Since Zika virus first appeared in French Guiana, the NRC has investigated 270 samples collected 0-5 days after fever onset for molecular diagnosis of Zika virus and dengue NS1. Of the 270 suspected patients, 65 were confirmed positive for acute Zika virus infection by rRT-PCR of serum, urine, or both. The mean cycle thresholds (± SDs) were 33.0 (± 3.4) for serum samples and 33.2 (± 3.5) for urine samples (Table)  To the Editor: The etiologic agent of the sixth pandemic of cholera was classical biotype of Vibrio cholerae O1. The ongoing seventh pandemic is caused by El Tor biotype. The biotypes are differentiated by phenotypic and genotypic characteristics. However, this differentiation blurred when V. cholerae O1 strains were detected in Matlab, Bangladesh, in 2006, in which characteristics were mixed. Genetically, the differences occurred in tcpA, which encodes the major adherence antigen rstR that regulates site-specific recombination of CTXφ phage and ctxB that encodes the B subunit of cholera toxin. These genes had the characteristics of classical biotype in Matlab variants of El Tor strains. Later, various types of El Tor variants were reported in Southeast Asia, Africa, and Haiti. Differentiating features also occur in repeat toxin A gene (rtxA), chromosomal location of CTXφ, the number of heptad repeats in ToxR binding region, and the occurrence of vibrio seventh pandemic islands I and II (1,2).
Kuwait is free of endemic cholera, but imported cases occur there (3). Cholera is endemic to neighboring Iraq. An outbreak caused by V. cholerae O1 Inaba serotype started in Iraq in September 2015 (4). However, a full characterization of the strain is lacking. A thorough characterization of the strain assumes urgency in light of the spread of variants. We characterized isolates from 2 recent cholera cases imported to Kuwait from Iraq.
The first case was in a 19-year-old Kuwaiti man who visited Najaf and Karbala in Iraq in September 2015; the second case was in a 52-year-old Kuwaiti woman who visited the same 2 locations in October 2015. Both had watery diarrhea 3-4 times daily and vomiting; they returned to Kuwait and were admitted to Al Amiri Hospital (Sharq, Kuwait). They gave histories of drinking local water in Iraq, had moderate dehydration, and were treated with intravenous rehydration solution and a single doxycycline dose (500 mg). Diarrhea resolved after 2-3 days.
Fecal specimens collected at admission from both patients grew yellow colonies on thiosulfate bile salt sucrose agar (Eiken, Tokyo, Japan); these colonies were confirmed as V. cholerae O1 Inaba serotype by biochemical reactions and agglutination with specific antiserum (Denka Seiken, Tokyo, Japan). The woman's isolate was designated as Kuwait 36 and the man's as Kuwait 37. The isolates were positive for chicken cell agglutination and Voges-Proskauer tests and were polymyxin B resistant, characteristics of El Tor biotype. The isolates were resistant to nalidixic acid but susceptible to ciprofloxacin, norfloxacin, ofloxacin, tetracycline, meropenem, ampicillin, ceftriaxone, trimethoprim/sulfamethoxazole, chloramphenicol, erythromycin, azithromycin, streptomycin, neomycin, and gentamicin by disk diffusion test. Tetracycline susceptibility confirmed favorable response to doxycycline.
We studied the genotype of ctxB using a double-mismatch amplification mutation PCR (i.e., mismatches in both primers). PCR with classical ctxB-specific primers ctxBF4/ctxBRvCla yielded an amplicon of 191 bp, but not with Haitian ctxB specific primers ctxBF3/ctxBRvCla, indicating that the isolates had a ctxB of classical biotype (genotype 1) (5,6). Mismatch amplification mutation assay PCR (MAMA-PCR, i.e., mismatch in only 1 primer) with Haitian-specific tcpA primers tcpAF2/tcpARev produced an amplicon of 167 bp but not with El Tor tcpA-specific primers tcpAF1/tcpAElRev, suggesting these isolates had the Haitian variant tcpA (2). MAMA-PCR for rtxA with El Tor-specific primer pair rtxAF/rtxAR1 yielded a 187bp amplicon but no amplicon for Haitian variant primer pair rtxAF/rtxAR2, suggesting the occurrence of rtxA of El Tor variety (2). The isolates possessed El Tor type rstR because they produced a 500-bp amplicon with primer pair rstR2/rstA3R (7). The isolates were positive for rstC, a repeat sequence activator found in El Tor biotype, because they yielded an amplicon of 238 bp with primer pair rstC1/ rstC2 (8). rstB is required for CTXφ phage integration. The Haitian strain has a GTA deletion at positions 77-79. MAMA-PCR with primer pair rstB F1/rstB R1 produced a 160-bp amplicon, suggesting the absence of deletion in El Tor type rstB (2). The isolates had CTXφ integrated in the large chromosome with RS element downstream because they produced a 766-bp amplicon with CII F/CII R primers