Journal of Clinical Virology 16 (2000) 135 – 144 www.elsevier.com/locate/jcv
Comparison of PanBio Dengue Duo Igm and IgG Capture ELISA and Venture Technologies Dengue IgM and IgG Dot Blot A.J. Cuzzubbo a, D.W. Vaughn b,1, A. Nisalak b, T. Solomon c,2, S. Kalayanarooj d, J. Aaskov e, N.M. Dung f, P.L. Devine a,* a PanBio Pty Ltd, 116 Lutwyche Rd., Windsor 4030, Qld, Australia Armed Forces Research Institute of Medical Sciences (AFRIMS), Bangkok, Thailand c Wellcome Trust Clinical Research Unit, Centre for Tropical Diseases, Cho Quan Hospital, 190 Ben Ham Tu, Quan 5, Ho Chi Minh City, Vietnam d Queen Sirikit National Institute of Child Health (Bangkok Children’s Hospital), Raji6ithi Road, 10400 Bangkok, Thailand e School of Life Sciences, Queensland Uni6ersity of Technology, Brisbane, Australia f Pediatric Intensi6e Care Unit, Centre for Tropical Diseases, Cho Quan Hospital, 190 Ben Ham Tu, Quan 5, Ho Chi Minh City, Vietnam b
Received 20 May 1999; received in revised form 11 November 1999; accepted 22 November 1999
Abstract Background: A number of commercial ELISA for dengue diagnosis have recently become available, though direct comparison between these assays have not been published. Objecti6es: The Venture Technologies Dengue IgM and IgG Dot Blot assays and the PanBio Dengue Duo IgM and IgG Capture ELISA were compared. Study Design: Paired sera from patients with dengue (n= 20) and Japanese encephalitis (JE, n= 10), and single sera from patients with typhoid (n=10), leptospirosis (n= 10) and scrub typhus (n =10) were assayed according to the manufacturer’s instructions. Results: The Dot Blot IgM ELISA showed higher sensitivity than the PanBio IgM ELISA (100 vs. 95%), while the PanBio IgM ELISA showed higher specificity in JE (100 vs. 20%) and non-flavivirus infections (100 vs. 97%). Defining elevation of either IgM or IgG as a positive result, the Dot Blot and ELISA tests both showed 100% sensitivity in dengue infection, while the PanBio test showed superior specificity in JE (70 vs. 0%) and non-flavivirus Abbre6iations: ANOVA, Analysis of variance; CSF, Cerebrospinal fluid; DHF, Dengue haemorrhagic fever; ELISA, Enzymelinked immunosorbent assay; HIA, Haemagglutination inhibition assay; JE, Japanese encephalitis; RT, Room temperature; S1, Sera at hospital admission; S2, Sera at hospital discharge. * Corresponding author. Tel.: + 61-7-33571177; fax: + 61-7-33571222. E-mail addresses:
[email protected] (D.W. Vaughn),
[email protected] (T. Solomon), peter –
[email protected] (P.L. Devine) 1 Present address: Walter Reid Army Institute of Research, Washington DC 20307, USA. Tel.: + 1-202-7823655; fax: + 1-2027820442. 2 Present address: Department of Neurological Science, University of Liverpool, Walton Centre for Neurology and Neurosurgery, Fazakerley, Liverpool L9 7LJ, UK. Tel.: + 44-151-5253611, ext. 2147; fax: +44-151-5295465 1386-6532/00/$ - see front matter © 2000 Elsevier Science B.V. All rights reserved. PII: S 1 3 8 6 - 6 5 3 2 ( 9 9 ) 0 0 0 7 1 - 2
136
A.J. Cuzzubbo et al. / Journal of Clinical Virology 16 (2000) 135–144
infections (100 vs. 67%). Conclusions: Both assays are useful aids to the serological diagnosis of dengue infection. The clinical setting, user preference and local conditions will be important in determining which test is more appropriate. © 2000 Elsevier Science B.V. All rights reserved. Keywords: ELISA; Dot blot; Dengue infection
1. Introduction Serology is commonly used in the diagnosis of dengue infections (Lam, 1993; Vorndam and Kuno, 1997). Haemagglutination inhibition assay (HIA) has been used traditionally, but enzymelinked immunosorbent assay (ELISA) has gained acceptance as a faster and more convenient alternative (Gubler, 1996). Primary infection is characterised by a rise in specific IgM 3–5 days after the onset of infection and this is generally detectable for 30 – 90 days. IgG levels rise after IgM and are detectable for life (Innis, 1997). In contrast, the IgM response in secondary dengue infection can be slower, weaker and shorter lived, and some patients do not show a detectable IgM (Ruechusatawat et al., 1994; Innis, 1997; Vaughn et al., 1997; Vorndam and Kuno, 1997). However, IgG levels rise rapidly to higher levels than observed in primary or past dengue infection and remain at these levels for 30 – 40 days (Gubler, 1996; Innis, 1997). Consequently, the combined use of IgM and IgG has been proposed as an effective strategy for the serological diagnosis of dengue virus infection (Innis et al., 1989; Kuno et al., 1991; Ruechusatawat et al., 1994; Devine et al., 1997; Vaughn et al., 1997; Sang et al., 1998b; Vaughn et al. 1998). Two commercial ELISA have been developed that utilize this strategy for IgM and IgG detection: the Venture Technologies Dengue IgM and IgG Dot Blot kits (Venture Technologies, Penang, Malaysia) (Lam, 1993; Chong et al., 1994; Cardosa et al., 1995; Lam et al., 1996; Kuno et al., 1998) and the PanBio Dengue Duo IgM Capture and IgG Capture ELISA (PanBio Pty Ltd, Brisbane, Australia) (Lam and Devine, 1998; Sang et al., 1998b; Vaughn et al., 1999). However, direct comparison between the two assays has not been reported. In this study, these tests were compared using sera taken from patients with primary and
secondary dengue infection, Japanese encephalitis (JE) infection, and non-flavivirus infections.
2. Materials and methods
2.1. Case definitions for dengue and Japanese encephalitis In children experiencing a febrile illness consistent with dengue fever or dengue hemorrhagic fever (DHF), dengue infections were defined as the isolation of a dengue virus, the detection of IgM to dengue (as opposed to IgM to JE), or a sustained elevation ( ] 1:2560) or fourfold rise in dengue virus HIA antibody titre (Innis et al., 1989). Dengue infection was categorised as primary or secondary according to the World Health Organisation criteria and the standard operating procedure for the reference EIA (Innis et al., 1989; World Health Organisation, 1997). JE was defined as a febrile illness associated with a decrease in consciousness and the presence of IgM to JE virus in the CSF.
2.2. Serum samples Serum from patients with a suspected dengue infection was collected at the time of admission to hospital and at discharge from either the Queen Sirikit National Institute of Child’s Health (Bangkok Children’s Hospital) or the Kamphaeng Phet Provincial Hospital, Thailand, and frozen at − 70°C prior to assay. Paired sera from 20 patients with dengue (10 primary and 10 secondary infections) were used in this study. The interval between the collection of the first and second sera in patients with dengue ranged from 1–6 days (mean 2.4 days). Paired sera from 10 patients with Japanese encephalitis (JE) were collected from the Centre for Tropical Diseases, Ho Chi Minh City,
A.J. Cuzzubbo et al. / Journal of Clinical Virology 16 (2000) 135–144
137
Vietnam. The interval between the collection of the first and second sera ranged from 3 – 11 days (mean 7 days). A panel of single sera from non flavivirus infections was also included. These represented widal felix test positive cases of typhoid from Malaysia (n= 10), indirect immunoperoxidase positive cases of rickettsial scrub typhus from Thailand (n=10), and microscopic agglutination test positive cases of leptospirosis from Australia (n=10).
negative sample with a ratio B 1.0. Dengue infection was characterised by the elevation of either IgM or IgG, with a negative sample defined as having both IgM and IgG ratios B 1.0. The cutoff sera in the IgG ELISA represented a HIA titre of 1:2560 — the titre used to distinguish secondary dengue virus infection from primary or past dengue virus infection (World Health Organisation, 1997).
2.3. PanBio Dengue Duo ELISA
2.4. Venture Technologies Dengue IgM and IgG Dot Blot Tests
The PanBio Dengue Duo IgM and IgG Capture ELISA (Product Code DEC-100) was performed by AC at PanBio Pty Ltd according to the manufacturer’s instructions (tested once). Two microtitre plates were supplied, one containing stabilised dengue virus antigen types 1 – 4 (antigen plate) and the other containing either anti-human IgM or anti-human IgG bound to separate microwells (assay plate). Peroxidase labelled antidengue monoclonal antibody (125 ml/well) was added to the antigen plate to resuspend the antigens and form antibody – antigen complexes. Concurrently 100 ml/well of patient sera, diluted 1:100 in the diluent provided, was added to the assay plate containing either bound anti-human IgM or anti-human IgG, and human IgM or IgG in the patient’s sera was captured. Both plates were incubated for 1 h at room temperature (antigen plate) or 37°C (assay plate), after which time the assay plate was washed, and 100 ml/well antibody–antigen complexes are transferred from the antigen plate to the assay plate. These complexes were then captured by dengue specific IgM or IgG during an incubation of 1 h at 37°C. The plate was then washed and bound complexes were visualized through the addition of a 100 ml/well tetramethylbenzidine substrate. After 10 min, the reaction was stopped by the addition of 100 ml/well 1 M phosphoric acid, and the strips were read at 450 nm using a microtitre plate reader. Positivity was determined by comparison to IgM and IgG reference sera provided (cutoff calibrators). A positive sample was defined as having a sample:calibrator absorbance ratio ] 1.0 and a
The Dot Blot tests were performed by AC at PanBio Pty Ltd according to the manufacturer’s instructions. Each sera was tested once and the results were read without knowledge of the clinical diagnosis or results of other tests. IgM was determined using an antibody capture format assay, while IgG was determined using an indirect format assay. In the IgM Dot Blot, nitrocellulose membrane coated with a spot of anti-human IgM was incubated for 2 h at room temperature (RT, 23°C) in patient sera diluted 1:100 in the diluent provided. After washing for 3× 5 min in the buffer provided, the membrane was incubated with dengue antigen overnight at 4°C. The membrane was washed again and then incubated with anti-dengue monoclonal antibody for 1 h at RT before washing again and incubating with conjugate for 1 h at RT. After washing, reactivity was visualised by incubation of the membrane for 30 min at RT in the substrate solution provided. The reaction was stopped by washing with distilled water after aspirating the substrate. In the IgG test, sera was sera was diluted 1:1000 in the diluent provided and incubated with membrane-bound dengue antigen for 60 min at RT. After washing, conjugate was added for 60 min at room temperature to detect bound IgG and colour development was performed as described above. Dots produced by different sera were compared to reactive and nonreactive controls, with any sign of color interpreted as a positive result. The dots were scored arbitrarily as negative, weak positive and positive.
A.J. Cuzzubbo et al. / Journal of Clinical Virology 16 (2000) 135–144
138
2.5. Data analysis The proportion of patients with antibody levels above the designated cutoff for ELISA was determined. The Fisher’s Exact Test was performed to compare sensitivity, specificity and F-values. ANOVA was used to compare the mean assay values of different diagnostic groups. Statistics were performed using Instat® (Graphpad Software Inc, San Diego, CA, USA).
3. Results
3.1. Sensiti6ity and specificity of IgM assays The sensitivity obtained with the IgM assays using sera from dengue patients is summarized in Table 1. The Dot Blot IgM test detected all 20 cases of dengue when paired sera were used (either S1 or S2 positive), while the PanBio IgM ELISA detected 19/20 (95%) of cases. In the admission (S1) sera, the Dot Blot IgM test
detected 9/10 primary infections and 10/10 secondary infections and the PanBio IgM ELISA detected 9/10 primary infections and 4/10 secondary infections. In the discharge (S2) sera, the Dot Blot IgM test detected 9/10 primary infections and 10/10 secondary infections, while the PanBio IgM ELISA detected 10/10 primary infections and 9/10 secondary infections. The specificity of the tests is shown in Table 2. When paired sera were used for diagnosis (either S1 or S2 positive), the Dot Blot IgM test produced a large number of false positive reactions particularly in JE (20% specificity), while the PanBio IgM ELISA showed significantly higher specificity (100%, P= 0.0007). The PanBio IgM ELISA was negative in all sera collected from patients with non-flavivirus infections (scrub typhus, leptospirosis and typhoid), and the Dot Blot IgM test was also negative with all sera except one patient with typhoid (Table 2). The overall performance of the IgM tests were assessed by the sum of sensitivity and specificity
Table 1 Sensitivity of dengue tests in primary and secondary dengue infectionsa Number elevated/total (% sensitivity) Primary dengue
Venture Tech. IgM PanBio Duo IgM Venture Tech. IgG PanBio Duo IgG Venture Tech. IgM or IgG PanBio Duo IgM or IgG a
Secondary dengue
All dengue
S1
S2
S1 or S2
S1
S2
S1 or S2
S1
S2
S1 or S2
9/10 (90) 9/10 (90) 1/10 (10) 0/10 (0) 9/10 (90) 9/10 (90)
9/10 (90) 10/10 (100) 5/10 (50) 0/10 (0) 9/10 (90) 10/10 (100)
10/10 (100) 10/10 (100) 5/10 (50) 0/10 (0) 10/10 (100) 10/10 (100)
10/10 (100) 4/10 (40) 10/10 (100) 6/10 (60) 10/10 (100) 7/10 (70)
10/10 (100) 9/10 (90) 10/10 (100) 10/10 (100) 10/10 (100) 10/10 (100)
10/10 (100) 9/10 (90) 10/10 (100) 10/10 (100) 10/10 (100) 10/10 (100)
19/20 (95) 13/20 (65) 11/20 (55) 6/20 (30) 19/20 (95) 18/20 (90)
19/20 (95) 19/20 (95) 15/20 (75) 10/20 (50) 19/20 (95) 20/20 (100)
20/20 (100) 19/20 (95) 15/20 (75) 10/20 (50) 20/20 (100) 20/20 (100)
Abbreviations used: S1, admission sera; S2, discharge sera, Tech., Technologies.
A.J. Cuzzubbo et al. / Journal of Clinical Virology 16 (2000) 135–144
139
Table 2 Specificity of dengue ELISAs in JE and non-flavivirus infectionsa Number negative/total (% specificity)
Venture Tech. IgM PanBio Duo IgM Venture Tech. IgG PanBio Duo IgG Venture Tech. IgM or IgG PanBio Duo IgM or IgG a
JE S1
JE S2
JE S1 or S21
Scrub Typhus
Lepto.
Typh. (%)
Non flavivirus Total
Non dengue Total
3/10 (30) 10/10 (100) 3/10 (30) 8/10 (80) 0/10 (0) 8/10 (80)
2/10 (20) 10/10 (100) 3/10 (30) 7/10 (70) 0/10 (0) 7/10 (70)
2/10 (20) 10/10 (100) 2/10 (20) 7/10 (70) 0/10 (0) 7/10 (70)
10/10 (100) 10/10 (100) 9/10 (90) 10/10 (100) 9/10 (90) 10/10 (100)
10/10 (100) 10/10 (100) 8/10 (80) 10/10 (100) 8/10 (80) 10/10 (100)
9/10 (90) 10/10 (100) 3/10 (30) 10/10 (100) 3/10 (30) 10/10 (100)
29/30 (97) 30/30 (100) 20/30 (67) 30/30 (100) 20/30 (67) 30/30 (100)
31/40 (78) 40/40 (100) 22/40 (55) 37/40 (93) 20/40 (50) 37/40 (93)
Abbreviations used: JE, Japanese encephalitis; Lepto., leptospirosis; Typh., typhoid; S1, admission sera; S2, discharge sera.
Table 3 ELISA performance Sum (sensitivity+specificity)
Venture Technologies IgM PanBio Duo IgM Venture Technologies IgG PanBio Duo IgG Venture Technologies IgM or IgG PanBio Duo IgM or IgG
Dengue and JE
Dengue and non-flavivirus
Dengue and non-dengue
120
197
178
195
195
195
95
142
130
120
150
143
100
167
150
170
200
193
(F-value) (Table 3). The PanBio IgM ELISA also showed superior performance when dengue and all other infections were analyzed (F = 195 vs. F= 178 for the Dot Blot; P =0.0010). The Dot Blot IgM and PanBio IgM tests performed similarly in distinguishing between dengue and nonflavivirus infections (F = 197 vs. F = 195), while the PanBio test was superior in distinguishing between dengue and JE infections (F = 195 vs. F= 120; P B 0.0001).
3.2. Sensiti6ity and specificity of IgG assays The sensitivity obtained with the IgG assays in sera from dengue patients is summarized in Table 1. As expected, due to the high IgG cutoff values used, all assays showed lower sensitivity in primary dengue infections than in secondary dengue infections. The Dot Blot IgG test detected 50% of primary dengue infections while the PanBio IgG ELISA was negative in all primary infections. The
140
A.J. Cuzzubbo et al. / Journal of Clinical Virology 16 (2000) 135–144
Dot Blot test produced a positive result with all S1 and S2 sera from secondary dengue, while the PanBio ELISA produced positive results with 6/ 10 (60%) S1 sera and all ten S2 sera. The specificity of the IgG ELISAs is shown in Table 2. When paired sera were used for diagnosis (either S1 or S2 positive), the Dot Blot IgG test produced a number of false positive reactions with sera from JE patients (20% specificity); while the PanBio IgG ELISA showed higher specificity (70%; P =0.0698). The PanBio IgG ELISA was negative with all sera collected from patients with non-flavivirus infections, while the Dot Blot IgG test produced false positive reactions in patients with scrub typhus (90% specificity), leptospirosis (80% specificity) and typhoid (30% specificity) (Table 2). The overall performance of the IgG tests was assessed by the sum of the sensitivity and specificity (F-value) (Table 3). The F-values for the IgG tests were lower than the IgM tests (Table 3) due to the poor sensitivity of the IgG tests in primary dengue (Table 1).
3.3. The combined use of IgM and IgG assays The sensitivity obtained if detection of either anti-dengue IgM or IgG was regarded as a positive result is summarized in Table 1. Both PanBio and Venture Technologies assays showed high sensitivity in dengue infections. The Dot Blot and PanBio tests detected all twenty dengue infections using paired sera. The combined use of the IgM and IgG Dot Blot tests gave false positive results in sera from all JE patients (0% specificity), while the PanBio Duo ELISA showed significantly higher specificity (70%; P = 0.0031). All sera collected from patients with non-flavivirus infections were negative in the PanBio Duo ELISA (100% specificity), while the combined use of the Dot Blot IgG and IgM tests showed a significant rate of false positive reactions with sera from patients with scrub typhus (90% specificity), leptospirosis (80% specificity) and typhoid (30% specificity) (P= 0.0008) (Table 2). The overall specificity in the PanBio Duo ELISA was 93%, while in the Dot Blot IgM and IgG tests specificity was 50% (PB 0.0001).
The overall performance of the IgM/IgG combination was assessed by the sum of the sensitivity and specificity (F-value) (Table 3). The PanBio tests showed superior performance to the dot blot tests when dengue was compared to either JE or non-flavivirus infections (P B 0.0001 for both).
3.4. Correlation analysis The individual results of the Dot Blot and ELISA were well correlated using either the IgM or IgG tests (Fig. 1). There was a significant association between the mean ELISA ratio and the Dot Blot score (ANOVA: F=18.52, PB 0.0001 for IgM; F= 27.30, PB0.0001 for IgG). However, it was apparent that many ELISA negative samples were positive in the Dot Blot test, suggesting that the cutoffs used in the Dot Blot tests were set below that of the ELISAs.
4. Discussion IgM is the marker of choice in dengue diagnosis, but some patients with secondary dengue have been reported to produce low or undetectable levels of IgM, and the IgM response in some secondary cases may be slow and short lived (Ruechusatawat et al., 1994; Innis, 1997; Vaughn et al., 1997; Vorndam and Kuno, 1997). To overcome this, it has been suggested that IgG be used as an adjunct to IgM, with the cutoff level set to detect high levels of IgG characteristic of secondary dengue but not primary or past dengue virus infection (Innis et al., 1989; Ruechusatawat et al., 1994; Vaughn et al., 1998). In this study, two commercial tests for both IgM and IgG were compared (PanBio Dengue Duo IgM and IgG Capture ELISA and the Venture Technologies IgM and IgG Dot Blot). The PanBio Dengue Duo ELISA gave the best distinction between dengue and other infections, as judged by the sum of sensitivity and specificity. The PanBio and Venture Technologies’ assays detected all dengue virus infections, but the PanBio Duo ELISA showed superior specificity (93 vs. 50%). The Venture Technologies IgM and IgG Dot Blot tests showed higher sensitivity than the
A.J. Cuzzubbo et al. / Journal of Clinical Virology 16 (2000) 135–144
corresponding PanBio ELISAs (100 vs. 95% for IgM and 75 vs. 50% for IgG), while the Dot Blot tests showed low specificity compared with the
141
PanBio ELISAs (78 vs. 100% for IgM and 55 vs. 93% for IgG). The PanBio IgM ELISA showed low sensitivity in acute phase sera of patients with
Fig. 1. Comparsion of individual assay values using PanBio Dengue Duo ELISAs and Venture Technologies Dengue Dot Blot tests for (a) IgM and (b) IgG. The cutoff ratios (1.0) for the PanBio ELISAs are shown by broken lines, while the mean assay value in the PanBio ELISAs are shown by a horizontal bar. The Dot Blot test was scored as negative, weakly positive or distinctly positive.
142
A.J. Cuzzubbo et al. / Journal of Clinical Virology 16 (2000) 135–144
secondary dengue infections (40%). However, significant levels of anti-dengue IgG antibody was detected in sera from many of these patients with the PanBio IgG ELISA (overall sensitivity of IgM and IgG combination in S1 sera was 70%). Similar results using the PanBio Dengue Duo ELISA have been reported previously (Lam and Devine, 1998; Sang et al., 1998b; Vaughn et al., 1999). The low specificity of the Venture Technologies IgM Dot Blot was due mainly to false positive reactions with sera from JE patients. High levels of cross-reactivity (80%) have been reported previously using the Dot Blot test when CSF and sera from flavivirus infections other than dengue (JE and yellow fever) were used (Kuno et al., 1998). In addition, another study with the Dot Blot IgM test showed high levels of cross reactivity (23%) in sera collected from patients with malaria (Lam et al., 1996). Apart from the differences in cutoff values, the differences observed between the two assays with sera from JE infections may have been due to the differences in incubation times and procedures used in these two tests. In the Dot Blot IgM test antigen was incubated overnight with captured IgM, while the PanBio IgM ELISA antigen was incubated for only 1 h with captured antibody. It is possible the longer incubation time used in the Dot Blot IgM test led to capture of dengue antigen by lower affinity IgM produced during JE infection. Alternatively, differences in the timing of conjugate addition in the two assays may affect the binding of dengue antigen to antiJE IgM. That is, in the Dot Blot IgM test conjugated monoclonal antibody is added after the binding of antigen by captured IgM, while in the PanBio test dengue antigen and conjugate are complexed before incubation in the capture plate. The latter method has been reported previously to halve the number of incubation steps required in the dengue IgM capture ELISA without affecting the correlation of the assay with HAI (Chong et al., 1994). The Venture Technologies IgG Dot Blot test produced false positive reactions with sera collected from patients with JE (80%), scrub typhus (10%), leptospirosis (20%) and typhoid (70%). High levels of cross reactivity (29%) have also been reported previously in patients with rubella
infections (Lam, 1993). These studies highlight the importance of selecting appropriate serum panels when evaluating dengue diagnostic assays. Of particular importance are other flavivirus infections (e.g. JE, yellow fever) and non-flavivirus infections with similar clinical presentation (e.g. scrub typhus, typhoid, leptospirosis and perhaps rubella). Comparison of the individual assay values revealed that the cutoffs used in the IgM and IgG Dot Blot tests were set below those used in the corresponding PanBio ELISAs. This resulted in better sensitivity but much poorer specificity in the Dot Blot tests. Similarly, another study concluded that the cutoff of the Dot Blot IgM test was too low and it was shown that a higher cutoff value would have improved performance (Kuno et al., 1998). The inverse relationship between sensitivity and specificity has been well documented (Metz, 1978; Greiner et al., 1995; Xu et al., 1997). Consequently, manufacturer’s of commercial assays need to set the cut-off to maximize sensitivity and/or specificity, depending on the target disease. One method that has been suggested regarding cut-off determination is to maximize the combined sensitivity and specificity (F-value) (Greiner et al., 1995; Xu et al., 1997). In this study the PanBio Dengue Duo ELISA showed superior performance (highest F-value) in distinguishing between dengue and other infections and this was significantly higher than the F-value obtained with the Dot Blot tests. Measurement of the relative levels of antidengue IgM and IgG has been suggested previously to be useful in distinguishing between primary and secondary dengue virus infections (Innis et al., 1989; Ruechusatawat et al., 1994; Vaughn et al., 1998). By employing this approach, the PanBio Dengue Duo ELISA has a high predictive value in distinguishing between primary and secondary dengue virus infections (Devine et al., 1997; Lam and Devine, 1998; Sang et al., 1998b; Vaughn et al., 1999). In this study, sera from all primary dengue infections assayed with the PanBio Duo ELISA contained anti-dengue IgM but not anti-dengue IgG, while all secondary dengue virus infections detected in the PanBio test were IgG positive. In contrast, the Dot Blot test
A.J. Cuzzubbo et al. / Journal of Clinical Virology 16 (2000) 135–144
could not be used for this purpose as 50% of primary dengue infections and all secondary dengue infections showed elevation of both IgM and IgG. Similarly, 29% of primary dengue cases have been detected when the IgG Dot Blot test was used in another study (Lam, 1993) and this rose to 61% when paired sera were used for diagnosis of primary dengue virus infections (Cardosa and Tio, 1991). The pattern of IgM and IgG reactivity detected with the PanBio ELISA could also be used to distinguish between dengue and other infections since 19/20 (95%) of dengue patients showed elevation of IgM, while patients with non-dengue infections (including JE) did not show elevation of IgM. Consequently, a positive IgM test in the PanBio Dengue Duo ELISA had a positive predictive value of 100% for dengue virus infection and cross-reactivity was only of concern in the minority of patients with dengue virus infection who were IgM negative, IgG positive. In contrast, the Dot Blot test could not be used in this way since cross-reactivity was observed in both IgM and IgG tests. The ELISA and Dot Blot tests described here have utility as aids in the diagnosis of dengue virus infections. The ELISA is particularly suited to the diagnostic pathology laboratory where automated equipment is available. In addition, the ELISA is rapid and simple to perform, large numbers of samples can be accommodated easily and interpretation of results is more objective than the Dot Blot test. Similar problems with the Dot Blot test have been reported previously (Lam et al., 1996). However, the Dot Blot test is more appropriate in a field setting as it does not require sophisticated equipment. Another commercially available test for dengue has been described that is also well suited to this application (Devine et al., 1997; Lam and Devine, 1998; Sang et al., 1998a; Vaughn et al., 1998). The dot blot test may also have utility in areas of low dengue prevalence (e.g. returning travellers from non-endemic areas) since higher sensitivity is needed. However, trials with asymptomatic donors and appropriate disease controls from these regions would be needed to confirm this.
143
Acknowledgements This work was supported by the US Army Medical Research and Materiel Command and PanBio Pty, Ltd (Brisbane, Australia) through a co-operative research and development agreement. The opinions or assertions contained herein are the private ones of the authors and are not to be construed as official or as reflecting the views of the US Government. The Dengue Duo ELISA was developed through an Australian Government sponsored Cooperative Research Centre for Diagnostic Technologies. We thank Ms Panor Srisongkram for performing the AFRIMS reference ELISA; Mr Ming Choohong for performing the hemagglutination inhibition assay; Dr Rachel Kneen for specimen and data management; Ms Tipawan Kungvanrattana for data entry; and the Directors and staff of the Queen Sirikit National Institute of Child’s Health and the Centre for Tropical Diseases, and Professor Nicholas White for their support. We also thank Dr George Watt, Department of Medicine, AFRIMS, Bangkok, Thailand for the provision of sera from patients with rickettsial scrub typhus; Mr William Winslow, Institute for Medical and Veterinary Science, Adelaide, Australia for the provision of sera and MAT results from patients with leptospirosis; and Dr Rodney Jones, Gribbles Pathology, Kuala Lumpur, Malaysia, for the provision of Widal felix positive sera from patients with typhoid.
References Cardosa MJ, Tio PH. Dot enzyme immunoassay: as alternative diagnostic aid for dengue fever and dengue haemorrhagic fever. Bull WHO 1991;69:741 – 5. Cardosa MJ, Baharudin F, Hamid S, Hooi TP, Nimmanitya S. A nitrocellulose membrane based IgM capture enzyme immunoassay for etiological diagnosis of dengue virus infections. Clin Diagn Virol 1995;3:343 – 50. Chong CF, Ngoh BL, Tan HC, Yap EH, Singh M, Chan L, Chan YC. A shortened dengue IgM capture ELISA using simultaneous incubation of antigen and peroxidase-labeled monoclonal antibody. Clin Diagn Virol 1994;1:335 – 41. Devine P, Cuzzubbo A, Marlborough D. Dengue fever testing. Today’s Life Sci 1997;9:26 – 30.
144
A.J. Cuzzubbo et al. / Journal of Clinical Virology 16 (2000) 135–144
Greiner M, Sohr D, Gobel P. A modified ROC analysis for the selection of cut-off values and the definition of intermediate results of serodiagnostic tests. J Immunol Methods 1995;185:123 – 32. Gubler DJ. Serological diagnosis of dengue haemorrhagic fever. Dengue Bull 1996;20:20–3. Innis BL, Nisalak A, Nammanitya S, Kusalerdchariya S, Chongswasdi V, Suntayakorn S, Puttisri P, Hoke CH. An enzyme-linked immunosorbant assay to characterise dengue infections where dengue and Japanese encephalitis cocirculate. Am J Trop Med Hyg 1989;40:418–27. Innis B. Antibody responses to dengue virus infection. In: Gubler DJ, Kuno G, editors. Dengue and Dengue Haemorrhagic Fever. New York: CAB International, 1997:221– 43. Kuno G, Gomez I, Gubler DJ. An ELISA procedure for the diagnosis of dengue infections. J Virol Methods 1991;33:101 – 13. Kuno G, Cropp CB, Wong-Lee J, Gubler DJ. Evaluation of an IgM Immunoblot kit for dengue diagnosis. Am J Trop Med Hyg 1998;59:757–62. Lam SK. Rapid dengue diagnosis and interpretation. Malays J Pathol 1993;15:9 – 12. Lam SK, Fong MY, Chungue E, Doraisingham S, Igarashi A, Khin MA, Kyaw ZT, Nisalak A, Roche C, Vaughn DW, Vorndam V. Multicentre evaluation of dengue IgM dot immunoassay. Clin Diagn Virol 1996;7:93–8. Lam SK, Devine PL. Evaluation of capture ELISA and rapid immunochromatographic test for the determination of IgM and IgG antibodies produced during dengue infection. Clin Diagn Virol 1998;10:75–81. Metz CE. Basic principles of ROC analysis. Semin Nucl Med 1978;8:283 – 98. Ruechusatawat K, Morita K, Tanaka M, Vongcheree S, Rojanasuphot S, Warachit P, Kanai K, Thongtradol P, Nimnakorn P, Kanungkid S, Igarashi A. Daily observation of antibody levels among dengue patients detected by en-
.
zyme-linked immunosorbent assay (ELISA). Jpn J Trop Med Hyg 1994;22:9 – 12. Sang CT, Lim SH, Cuzzubbo A, Devine PL. Clinical evaluation of rapid immunochromatographic test for the diagnosis of dengue infection. Clin Diagn Lab Immunol 1998a;5:407 – 9. Sang CT, Cuzzubbo A, Devine PL. Evaluation of commercial capture enzyme-linked immunosorbent assay for the detection of Immunoglobulin M (IgM) and IgG antibodies produced during dengue infection. Clin Diagn Lab Immunol 1998b;5:710. Vaughn DW, Green S, Kalayanarooj S, Innis BL, Nirnmannitya S, Suntayakorn S, Rothman AL, Ennis FA, Nisalak A. Dengue in early febrile phase: viremia and antibody responses. J Infect Dis 1997;176:322 – 30. Vaughn DW, Nisalak A, Kalayanarooj S, Solomon T, Dung NM, Cuzzubbo A, Devine PL. Evaluation of rapid immunochromatographic test for diagnosis of dengue virus infection. J Clin Microbiol 1998;36:234 – 8. Vaughn DW, Nisalak A, Kalayanarooj S, Solomon T, Kalayanarooj S, Dung NM, Kneen R, Cuzzubbo A, Devine PL. Rapid serological diagnosis of dengue virus infection using a commercial capture enzyme-linked immunosorbent assay that distinguishes primary and secondary infections. Am J Trop Med Hyg 1999;60:693 – 98. Vorndam V, Kuno G. Laboratory diagnosis of dengue virus infections. In: Gubler DJ, Kuno G, editors. Dengue and Dengue Haemorrhagic Fever. New York: CAB International, 1997:313 – 33. World Health Organisation. Dengue haemorrhagic fever diagnosis, treatment, prevention and control. Geneva: WHO, 1997. Xu H, Lohr J, Greiner M. The selection of ELISA cut-off points for testing antibody to Newcastle disease by twograph receiver-operating characteristic (TG-ROC) analysis. J Immunol Methods 1997;13:61 – 4.
