Introduction

Dengue is considered the most important of the arboviruses affecting humans. It is transmitted mainly by the mosquito Aedes aegypti and has four known serotypes: DENV-1, DENV-2, DENV-3 and DENV-4. Approximately 50 million people in over 100 nations on all continents (except Europe) are infected annually [1].

The first dengue epidemic that was documented in both the clinic and laboratory in Brazil occurred in 1981 in Boa Vista, State of Roraima, and DENV-1 and DENV-4 were identified [2]. In 1990, DENV-2 was also detected in Rio de Janeiro and was responsible for an extensive outbreak in the country and for the emergence of severe cases [3], [4]. However, the greatest epidemic of dengue in Brazil occurred in 2002, when almost 800,000 cases of the disease were reported, most of them DENV-3 [5]. In 2010, DENV-4 was detected in Boa Vista [6].

Since 2009, the NS1 antigen-capture ELISA test has been implemented by the Health Ministry in the Public Health laboratories network. Patients arriving for care within five days of the onset of disease symptoms are screened for viral isolation in health care sentinel units and, thus, provide early detection of circulating serotypes in a given area [7].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 1. Reverse transcriptase-semi-nested-polymerase chain reaction (RT-PCR) product in 2% agarose gel stained with syber safe.

Lines 1–4: positive samples; 5–8: negative samples; 9–12: positive control for DENV-1, DENV-2, DENV-3 and DENV-4; NC: negative control; MW: molecular weight marker.

https://doi.org/10.1371/journal.pone.0064056.g001

In Sergipe state, routine NS1 Ag test detection protocol is applied to patients in the sentinel units according to Health Ministry determinations. Blood is collected and send to the Central Laboratory (Laboratório Central - LACEN/SE) for the Platelia™ test. Positive biological samples are then forwarded to the reference laboratory in Rio de Janeiro for viral isolation. Negative samples are stored in a −80°C and for this study were provided to our group in order to evaluate the presence of other arbovirus.

Methods

Results and Discussion

When processing the samples, 58 (48.7%) samples were unexpectedly found to be positive for DENV-4 (Fig). No other serotype was detected. Of the 58 positive samples, 33 (56.9%) were from the capital city, Aracaju, while two (3.4%) were from the municipality of Itabaiana, and 23(39.7%) samples were from the municipality of Simão Dias; these two municipalities are 51 and 101 km from the capital, respectively. Positive samples were subjected to viral isolation in Aedes albopictus cells, clone C6/36s, to confirm the diagnosis, and 31 (53.5%) were positive for DENV-4.

The DENV-4 serotype was detected for the first time in Brazil in 1982 during a dengue outbreak in the north of the country, and since then, no new cases of DENV-4 were detected until 2010 [6]. In Sergipe, DENV-4 was not reported before March 2012, but, when we analyzed blood samples that were negative for dengue by the Platelia™ NS1 Ag test collected between September 2011 and February 2012, we found that 48.7% were positive for DENV-4 by semi-nested RT-PCR. These results suggest that the Platelia™ NS1 Ag test has a low sensitivity for the DENV-4 serotype. These findings corroborate those of Hang et al. (2009) [10], who demonstrated that NS1 Ag had variable sensibility to the different Dengue serotypes, although not evaluating Dengue 4 cases. The simultaneous presence of IgG, which suggests a secondary response and low viremia in the analyzed samples, should be considered as the factor that contributes to the low sensitivity of the NS1 Ag test for the examined samples [10]. The Platelia™ NS1 Ag test was also assessed in a multicenter study and showed overall sensitivities ranging from 36% (Central America) to 88% (Asian countries) depending on the serotype and the number of days since the onset of symptoms. In a Cuban study, when evaluations were conducted based on serotype, DENV-4 showed a sensitivity of 79% (95%CI = 67–91); however, few samples were analyzed [11]. In Brazil, the sensitivity was 83.6% for confirmed positive samples for DENV-1, DENV-2, and DENV-3, and better results were obtained using samples obtained up to the fifth day in the viremic phase [12]. However, in the above mentioned study, no correlation between sensitivity and specificity could be determined for DENV-4, as this serotype was not yet circulating in the country.

The low sensitivity of the NS1 Ag test should be considered in relation to secondary dengue infections in the studied population [12], [13], as the state of Sergipe can be considered endemic for dengue. In Sergipe, the median incidence has been 195 cases per 100,000 inhabitants (22.9–1620.9) over the past 16 years, and successive outbreaks involving the circulation of the other three serotypes have occurred since 1996 [14].

Although the implementation of the NS1 Ag test for viremic sample screening has efficiently increased the viral isolation percentage and permitted the identification of the circulating serotypes in Brazil [7], the findings of the present study strongly suggest that DENV-4 was previously circulating with a lower prevalence and was not being detected by the NS1 Ag test. This hypothesis may explain the speed with which DENV-4 has spread across the country since its detection in northern Brazil in 2010, in the south and southeast in early 2011, and even in small municipalities of the northeastern region in mid-2011, as observed in the present study.

Most likely due to the low sensitivity of the Platelia™ NS1 Ag test for DENV-4, there was a delay in detection and an underreporting of DENV-4 dengue cases in Brazil. As this serotype has been increasingly detected in many regions [7], the low sensitivity may have resulted in the absence of detected DENV-4 dengue cases and, obviously, in underreporting. Therefore, we conclude that the sensitivity, specificity and accuracy of the Platelia™ NS1 Ag test must be reevaluated as a screening method for monitoring circulating dengue serotypes and that regional endo-epidemic profiles and the high incidence of secondary responses should be taken into account.

Acknowledgments

The authors would like to thank Dr. Pedro Vasconcelos for the critical evaluation of the manuscript, Danuza Duarte Costa and Sandra Maria Araujo Menezes Cavalcante for support with storage and provision of samples.