Journal of Clinical Virology 47 (2010) 49–53

Contents lists available at ScienceDirect

Journal of Clinical Virology

journal homepage: www.elsevier.com/locate/jcv

Early diagnosis of dengue in travelers: Comparison of a novel real-time RT-PCR,

NS1 antigen detection and serology
Eili Huhtamo a,∗ , Essi Hasu a , Nathalie Y. Uzcátegui a , Elina Erra b , Simo Nikkari c , Anu Kantele b ,
Olli Vapalahti a,d,e , Heli Piiparinen a
a
Department of Virology, Haartman Institute, Faculty of Medicine, University of Helsinki, P.O. Box 21 (Haartmaninkatu 3), FI-00014 University of Helsinki, Finland
b
Helsinki University Central Hospital, Department of Medicine, Division of Infectious Diseases, Helsinki, Finland
c
Centres for Military Medicine and for Biological Threat Preparedness, Tukholmankatu 8A, 00290 Helsinki, Finland
d
Department of Virology, HUSLAB, Helsinki University Central Hospital Laboratory, P.O. Box 400 (Haartmaninkatu 3), FI-00029 HUS, Finland
e
Division of Microbiology and Epidemiology, Department of Basic Veterinary Sciences, P.O. Box 66 (Agnes Sjöbergin katu 2), FI-00014 University of Helsinki, Finland

a r t i c l e i n f o a b s t r a c t

Article history: Background: The increased traveling to dengue endemic regions and the numerous epidemics have led
Received 18 August 2009 to a rise in imported dengue. The laboratory diagnosis of acute dengue requires several types of tests and
Accepted 4 November 2009 often paired samples are needed for obtaining reliable results. Although several diagnostic methods are
available, proper comparative data on their performance are lacking.
Keywords: Objectives: To compare the performance of novel methods including a novel pan-DENV real-time RT-
Dengue
PCR and a commercially available NS1 capture-EIA in regard to IgM detection for optimizing the early
Traveler
diagnosis of DENV in travelers.
Diagnosis
Real-time RT-PCR
Study design: A panel of 99 selected early phase serum samples of dengue patients was studied by real-
NS1 antigen time RT-PCR, NS1 antigen ELISA, IgM-EIA, IgG-IFA and cell culture virus isolation.
Serology Results: The novel real-time RT-PCR was shown specific and sensitive for detection of DENV-1-4 RNA and
Virus isolation suitable for diagnostic use. The diagnostic rate using combination of RNA and IgM detection was 99% and
using NS1 and IgM detection 95.9%. The results of RNA and NS1 antigen detection disagreed in 15.5% of
samples that had only RNA or NS1 antigen detected.
Conclusions: The diagnostic rates of early samples are higher when either RNA or NS1 antigen detection is
combined with IgM detection. Besides the differences in the RNA and NS1 detection assays, the observed
discrepancy of results could suggest individual variation or differences in timing of these markers in
patient serum.
© 2009 Elsevier B.V. All rights reserved.

1. Background
Dengue viruses (DENV-1-4) are mosquito-borne enveloped RNA
viruses belonging to the family Flaviviridae in the genus Flavivirus.
The dengue disease is categorized by severity as dengue fever,
dengue hemorrhagic fever or dengue shock syndrome.1,2 DENV
are endemic in tropics and subtropics of the world and cause a
major public health problem. Increased traveling to these areas
has resulted in an increase in imported cases to non-endemic
countries.3,4
The laboratory diagnostics of dengue relies on the use of sev-
eral methods detecting markers of DENV infection present in
patient serum at different time-points and often paired samples are
required for reliable results. The commonly used tests include sero-
∗ Corresponding author. Tel.: +358 9 19126706; fax: +358 9 19126491.
E-mail address: eili.huhtamo@helsinki.fi (E. Huhtamo).
logical methods detecting antibodies (IgM and IgG) against DENV1,5
and additionally various methods are used in detecting DENV
RNA6–9 or antigens: non-structural protein 1 (NS1) and envelope
protein (E).10–12 The serological methods are vulnerable to cross-
reactions caused by antibodies against related flaviviruses and are
therefore not DENV-specific tests like DENV NS1 antigen and RNA
detection methods. Travelers usually experience primary DENV
infections serologically characterized by high IgM and low IgG lev-
els, whereas the opposite is observed in secondary infections.1
2. Objectives
The aim of this study was to optimize the early diagnostics of
DENV in travelers, and to evaluate and compare two methods, a
novel DENV RNA detection method and a commercial NS1 antigen
detection method in regard to serology. The samples were addi-
tionally studied by cell culture virus isolation.

1386-6532/$ – see front matter © 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.jcv.2009.11.001
50 E. Huhtamo et al. / Journal of Clinical Virology 47 (2010) 49–53

Table 1 and the probe were designed using DENV sequences in Gen-
Travel history of the patients.
Bank database targeted to a conserved region in 3 UTR using
Country or region n of patients Primer Express Software version 3.0 (Applied Biosystems) (Table 2).
Thailand 32 The assay was optimized to 50 ␮l reaction volume using 5 ␮l of
India 10 template RNA and 900 nM of primers and 250 nM of probe on
Sri Lanka 7 Quantitect One Step Probe RT-PCR Kit (Qiagen). The test was
Indonesia 5 run on 96-well plates on duplicates using a program of 50 ◦ C
Asia, other countries* 11
for 30 min, 95 ◦ C for 15 min followed by 45× cycle of 95 ◦ C for
Mexico 5
Americas, other countries* 8 15 s and 60 ◦ C for 1 min on ABI Prism 7700 Sequence Detection
Africa, all countries* 10 System.
Unknown travel history 10 The test was evaluated using external quality control samples
*
Countries with 5 or more patients. Countries with fewer patients are com- from European Network for Diagnostics of Imported Viral Diseases
bined regionally, Africa: Kenya (n = 3), Eritrea (n = 1), Ethiopia (n = 1), Congo (n = 1), (ENIVD)13 The control panel included seven sera containing vari-
Ghana (n = 1), Somalia (n = 2); Seychelles (n = 1), the Americas: Brazil (n = 2), Costa able amounts of DENV-1-4 RNA and two non-DENV templates,
Rica (n = 1), Cuba (n = 1), Venezuela (n = 1), Martinique (n = 1), Peru (n = 1) Dominican
TBEV and YFV RNA, and one negative sample. Additionally RNA
Republic (n = 1); Asia: Vietnam (n = 2), Cambodia (n = 1), Bangladesh (n = 2), Philip-
pines (n = 3), Malaysia (n = 1), China (n = 2). Other travel destinations: Papua New from cell cultured flaviviruses including YFV (17D vaccine strain),
Guinea (n = 1). JEV (Nakayama), TBEV (Kumlinge A52), USUV (Vienna 2001), WNV
(Egypt 101) and RNA from sera of patients suspected for non-
3. Study design related viral infections (n = 32) were tested. For sensitivity evalua-
tion, viral RNA was extracted from DENV-1 (Hawaii) DENV-2 (New
3.1. Patient serum samples Guinea C) DENV-3 (H87), DENV-4 (H241) culture supernatants in
the absence of carrier RNA. The concentrations were measured
Sera from DENV patients were obtained from the diagnostic unit spectrophotometrically. The RNA samples were tested on 10-fold
of Department of Virology, HUSLAB, where all DEN diagnostics are dilution series in real-time RT-PCR and the results, RNA concentra-
performed in Finland. The panel consisted of first available, sin- tions and DENV genome size were used in calculation of the assay
gle serum from 99 serologically confirmed patients collected in sensitivity.
1999-2009. Information on travel history and timing of sampling
was available in most cases. However, in 21/99 (21.2%) of samples
the information was not available or was incomplete. For choos- 3.4. NS1 antigen detection by EIA
ing early samples a criterion of IgG titer ≤320 was used. The travel
history of the patients varied (Table 1), however all patients expe- The NS1 antigen tests were done according to the manufac-
rienced dengue fever without complications. The aliquots of sera turer’s instructions using BIORAD Platelia Dengue NS1 AG EIA
for virus isolation and RNA extraction were stored at −70 ◦ C and test. The cross-reactivity of the assay was tested using antigen-
for the NS1 antigen assay at −20 ◦ C prior to use. containing supernatants from Vero E6 cells infected with TBEV, JEV,
WNV and YFV 17 D vaccine. DENV-1-4 supernatants were included
3.2. RNA extraction as positive controls.

The RNA was extracted from 100 ␮l of serum specimens or viral

culture supernatants using QIAamp Viral RNA Mini Kit (Qiagen)
according to manufacturer’s instructions.
3.3. Dengue virus RNA detection by Taqman real-time one-step
RT-PCR
A one-step real-time TaqMan RT-PCR was developed for diag-
nostic use detecting simultaneously DENV-1-4 RNA. The primers
3.5. Antibody tests (IgM and IgG)
The sera were tested for anti-DENV IgM by enzyme immunoas-
say (Focus Technologies) according to manufacturer’s instructions.
The IgG tests were carried out as dilution series of 1:10–1:640
by an in-house immunofluorescence assay (IFA) as previ-
ously described14 using acetone-fixed DENV-3 infected Vero
E6 cells.

Table 2
Multiple sequence alignment of DENV 3 UTR region showing primers and probe used in the real-time one-step RT-PCR assay. FAM: carboxyfluorescein, MGB: minor groove
binder, BHQ1: Black hole quencher 1. Alignment made by using ClustalW2 available at http://www.ebi.ac.uk/Tools/clustalw2/index.html.

Forward primer: 5 -GGACTAGAGGTTAGAGGAGACCCC.

Probe 6-FAM-AGCATATTGACGCTGGGA-MGB-BHQ1.
Reverse primer: 5 -GAGACAGCAGGATCTCTGGTC.
 E. Huhtamo et al. / Journal of Clinical Virology 47 (2010) 49–53 51

3.6. Virus isolation 4. Results

For 40/99 sera, the virus isolation trial was done separately prior
this study (Huhtamo et al., 2008). In this study, 59 samples were
studied similarly using a 50 ␮l aliquot of serum in Aedes albopic-
tus C6/36 cells and screened for the presence of flavivirus antigens
in IFA as described previously.15 From the IFA positive isolation
culture supernatants RNA was extracted and studied using DENV-
typing RT-PCR.6
3.7. Confirmation of positive real-time RT-PCR results
The real-time RT-PCR positive samples that remained
virus isolation negative were studied additionally using two
separate conventional RT-PCR methods6,16 for confirmation
of the result and for obtaining information of the infect-
ing DENV serotype. The products were sequenced and the
sequences compared to sequences in the GenBank using BLAST N
algorithm.
4.1. Specificity and sensitivity of the novel one-step real-time
RT-PCR
The one-step real-time RT-PCR assay was shown specific to
DENV, as no amplification was observed from other flavivirus-
or non-dengue patient sample templates. Results of the external
controls of ENIVD were 100% correct, as only samples containing
DENV-1-4 RNA were positive. The sensitivity calculations based
on the results of the end-point titration of spectrophotometrically
quantified viral RNA in one-step real-time RT-PCR resulted in assay
detection sensitivities of 200-900 DENV genomes/reaction depend-
ing on the DENV serotype; 600 DENV-1, 200 DENV-2, 500 DENV-3
and 900 of DENV-4.
4.2. Specificity of NS1 antigen ELISA
The results of cell culture derived viral antigens demonstrated
Platelia Dengue NS1 antigen ELISA specific to DENV NS1. No cross-

Fig. 1. Results of Real-time RT-PCR, NS1 antigen and IgM detection. NoI, no information available, Neg, negative, Pos, positive. (a) All samples (n = 99); (b) samples taken on
days 1–3 (n = 15); (c) days 4–5 (n = 27); (d) days 6–7 (n = 19); (e) day 8 and later (n = 17) after onset of symptoms.
52 E. Huhtamo et al. / Journal of Clinical Virology 47 (2010) 49–53
reactivity was observed from virus culture supernatants of other
flaviviruses tested (YFV, JEV, WNV, TBEV).
4.3. Real-time RT-PCR, virus isolation and confirmatory tests on
patient samples
A total of 59/98 (60.2%) the tested patient samples were found
positive in real-time RT-PCR (Fig. 1a). The proportion of posi-
tive samples in real-time RT-PCR was highest on days 1–3, 86.7%
(13/15), followed by 76.9% (20/26) on days 4–5 and 63.2% (12/19)
on days 6–7 after onset (Fig. 1b–d). In samples taken on day 8 after
onset or later, only 58.8% (10/17) were positive (Fig. 1e). Using a
threshold of 0.15, the threshold cycle (Ct) values of positive patient
samples varied from 16.6 to 39.9. The positive result was confirmed
by virus isolation in 43 samples, and by conventional RT-PCR6,16
and sequencing in 12 samples. In three real-time RT-PCR positive
samples the positive result was not confirmable using other meth-
ods, leaving the serotype undetermined. All virus isolation positive
sera were positive in real-time RT-PCR on low Ct values. The results
of virus isolation, DENV type specific RT-PCR and sequences of con-
ventional RT-PCR products (∼200 bp) demonstrated that all four
DENV serotypes were reliably detected by real-time RT-PCR from
patient serum. A total of 27 patient samples contained RNA of
DENV-1, 9 of DENV-2, 17 of DENV-3 and 3 of DENV-4.
4.4. NS1 antigen detection
The NS1 antigen ELISA was positive in 66/98 (67.3%) of tested
samples (Fig. 1a). In one patient sample borderline result was
obtained repeatedly and regarded as a negative result. The highest
percentages of NS1 antigen positive samples 84.2% (16/19) were
taken on days 6–7 after onset of symptoms (Fig. 1d). The propor-
tion of NS1 positive samples was lower on samples taken on days
1–3 78.6% (11/14) and on days 4–5 74.1% (20/27) and in samples
taken on day 8 or later 70.6% (12/17) (Fig. 1b, c and e).
4.5. Serological tests
A total of 78/98 (79.6%) of the tested samples were IgM positive
(Fig. 1a). Detectable levels of IgM were present only in 42.9% (6/14)
of the samples on days 1–3 rising on samples taken on days 4–5 to
81.5% (22/27). From samples taken on days 6–7, IgM was detectable
in 94.7% (18/19) of the samples, and day 8 and later 82.4% (14/17).
Detectable levels of IgG were present in 69.4% of samples. On days
1–3 35.7% (5/14) of samples, on days 4–5 55.6% (15/27) and on days
6–7 78.9% (15/19), on day 8 and later 82.4% (14/17) of samples had
measurable levels of IgG (supplementary table).
Table 3
Comparison of results obtained with the different diagnostic methods: (a) real-time
RT-PCR and NS1 antigen ELISA n = 97; (b) NS1 antigen ELISA and IgM-ELISA, n = 98
(neither NS1 antigen or IgM has been performed for one sample); (c) real-time RT-
PCR and IgM-ELISA, n = 97 (real-time RT-PCR has not been performed for one sample
and IgM for one sample).
NS1+
(a)
Real-time RT-PCR + 54
Real-time RT-PCR - 11
(b)
IgM+ 50
IgM- 16
IgM+
(c)
Real-time RT-PCR + 39
Real-time RT-PCR - 38
4.6. Comparison and combination of test results
When comparing the results of RNA, NS1 antigen and IgM detec-
tion without considering the timing of the sampling, IgM-EIA was
the most useful marker of DENV infection (Fig. 1a). Using a diag-
nostic combination of RNA and IgM detection positive diagnosis
was obtained in 99% (96/97) and by using the combination of NS1
antigen and IgM detection 95.9% (94/98) (Table 3). When compar-
ing NS1 antigen and RNA detection, the results were in agreement
in 82/97 (84.5%) and contradictory in 15.5% of the samples (15
individual samples, indicated in supplementary table).
5. Discussion
For early diagnosis of DENV in travelers, a simple one-step
real-time RT-PCR method was developed. The method was shown
sensitive and specific for DENV RNA detecting reliably DENV-1-
4 RNA of different strains and genotypes from sera of patients
infected in different continents.
When comparing the results of RNA and NS1 antigen detection
and serology, during the first days after onset viral RNA and NS1
antigen were expectedly more suitable markers of DENV infection
than IgM. However, already on days 4–5 IgM detection was the
most efficient diagnostic method. On days 6–7 NS1 antigen and
IgM detection performed superior to RNA detection, and this was
also the case for the samples taken on later time-points.
It is assumed that NS1 is present in the serum during viremia.17
However, our results suggest that NS1 antigen is detectable at later
time points than the viral RNA. In approximately half of the NS1
antigen positive, RNA negative samples the late phase of the dis-
ease possibly favored NS1 detection, however also RNA degradation
in samples could have caused lowered RNA concentration and neg-
ative result.
The negative NS1 antigen results in RNA positive samples raise
questions as NS1 antigen is probably stable and measured directly
from serum. These samples were taken on days 2–7. If they are
true negative samples, they could suggest individual variation in
NS1 secretion or different timing of NS1-antigenemia and viremia
in these patients. Notably, 2 of these cases could be considered
as secondary infections as IgG but no IgM was detected at onset1
that may explain the negative NS1 test. In comparison to pri-
mary infections, lower NS1 antigen detection rates have been
reported in secondary infections explained by NS1 binding to
immunocomplexes.18 However, also opposite results12 and equal
NS1 detection rates in primary and secondary cases have been
reported.10 Further research is required on the sensitivity of NS1
detection in secondary infections as in future these may become
more common in travelers due to repeated traveling to DENV
endemic areas.
The date after onset of the disease needs to be taken into account
when choosing which diagnostic test to use. In diagnosis of dengue,
combining DENV RNA or NS1 detection methods to IgM detection
enhances the diagnostic rates in early samples.
NS1− Conflict of interest
4 None declared.
28
28 Acknowledgements
4
We thank the clinicians for their input in patient informa-
IgM− tion collection, ENIVD for the quality control samples, Sirkka Vene
and Norbert Nowotny for providing virus strains, Auli Saarinen,
19 Kirsti Räihä, Minna Ulmanen, and Raija Leveelahti for assistance,
1
Orion Farmos Research Foundation, Biomedicum Helsinki Founda-
 E. Huhtamo et al. / Journal of Clinical Virology 47 (2010) 49–53
tion, Centre for Biothreat Preparedness, Finnish Funding Agency
for Technology and Innovation (TEKES) and Hospital District of
Helsinki and Uusimaa (EVO/TYH6215 and TYH 2008309) for finan-
cial support.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.jcv.2009.11.001.
References
1. World Health Organization. Dengue haemorrhagic fever: diagnosis, treatment,
prevention and control. 2nd ed.; 1997.
2. Rigau-Perez JG, Clark GG, Gubler DJ, Reiter P, Sanders EJ, Vorndam AV. Dengue
and dengue haemorrhagic fever. Lancet 1998;352(September (9132)):971–7.
3. Jelinek T, Muhlberger N, Harms G, Corachan M, Grobusch MP, Knobloch J, et al.
Epidemiology and clinical features of imported dengue fever in Europe: sen-
tinel surveillance data from TropNetEurop. Clin Infect Dis 2002;35(November
(9)):1047–52.
4. Jelinek T. Trends in the epidemiology of dengue fever and their relevance for
importation to Europe. Euro Surveill 2009;14(June (25)):19250.
5. Shu PY, Huang JH. Current advances in dengue diagnosis. Clin Diagn Lab
Immunol 2004;11(July (4)):642–50.
6. Lanciotti RS, Calisher CH, Gubler DJ, Chang GJ, Vorndam AV. Rapid detection and
typing of dengue viruses from clinical samples by using reverse transcriptase-
polymerase chain reaction. J Clin Microbiol 1992;30(March (3)):545–51.
7. Johnson BW, Russell BJ, Lanciotti RS. Serotype-specific detection of dengue
viruses in a fourplex real-time reverse transcriptase PCR assay. J Clin Microbiol
2005;43(October (10)):4977–83.
8. Dumoulin A, Marti H, Panning M, Hatz C, Hirsch HH. Pan-dengue virus
detection by PCR for travelers returning from the tropics. J Clin Microbiol
2008;46(September (9)):3104–6.
53
9. Callahan JD, Wu SJ, Dion-Schultz A, Mangold BE, Peruski LF, Watts DM, et
al. Development and evaluation of serotype- and group-specific fluorogenic
reverse transcriptase PCR (TaqMan) assays for dengue virus. J Clin Microbiol
2001;39(November (11)):4119–24.
10. Alcon S, Talarmin A, Debruyne M, Falconar A, Deubel V, Flamand M. Enzyme-
linked immunosorbent assay specific to Dengue virus type 1 nonstructural
protein NS1 reveals circulation of the antigen in the blood during the acute
phase of disease in patients experiencing primary or secondary infections. J
Clin Microbiol 2002;40(February (2)):376–81.
11. Dussart P, Labeau B, Lagathu G, Louis P, Nunes MR, Rodrigues SG, et al. Evalua-
tion of an enzyme immunoassay for detection of dengue virus NS1 antigen in
human serum. Clin Vac Immunol 2006;13(November (11)):1185–9.
12. Young PR, Hilditch PA, Bletchly C, Halloran W. An antigen capture enzyme-
linked immunosorbent assay reveals high levels of the dengue virus protein
NS1 in the sera of infected patients. J Clin Microbiol 2000;38(March (3)):
1053–7.
13. Lemmer K, Donoso Mantke O, Bae HG, Groen J, Drosten C, Niedrig M. External
quality control assessment in PCR diagnostics of dengue virus infections. J Clin
Virol 2004;30(August (4)):291–6.
14. Vene S, Mangiafico J, Niklasson B. Indirect immunofluorescence for serologi-
cal diagnosis of dengue virus infections in Swedish patients. Clin Diagn Virol
1995;4(July (1)):43–50.
15. Huhtamo E, Uzcategui NY, Siikamaki H, Saarinen A, Piiparinen H, Vaheri A, et al.
Molecular epidemiology of dengue virus strains from Finnish travelers. Emerg
Infect Dis 2008;14(January (1)):80–3.
16. Scaramozzino N, Crance JM, Jouan A, DeBriel DA, Stoll F, Garin D. Comparison
of flavivirus universal primer pairs and development of a rapid, highly sensi-
tive heminested reverse transcription-PCR assay for detection of flaviviruses
targeted to a conserved region of the NS5 gene sequences. J Clin Microbiol
2001;39(May (5)):1922–7.
17. Halstead SB, Dengue. Lancet 2007;370(November (9599)):1644–52.
18. Kumarasamy V, Chua SK, Hassan Z, Wahab AH, Chem YK, Mohamad M, et al.
Evaluating the sensitivity of a commercial dengue NS1 antigen-capture ELISA
for early diagnosis of acute dengue virus infection. Singapore Med J 2007;48(July
(7)):669–73.