Dengue is the most common mosquito-borne infection, with an estimated 100 million infections worldwide per year.1–3 The recent dramatic expansion of dengue fueled by urbanization, increased population density, air travel, and limited resources for dengue prevention has led to dengue becoming a major public health problem in the tropics.3,4 Of the 100 million annual infections, 250–500 thousand persons manifest severe disease, with the remainder being mild, nonspecific, or even asymptomatic.1–3
There are four serotypes of the dengue virus: infection by one serotype is thought to produce lifelong immunity to that serotype but only a few months of immunity to the others.3 In Malaysia where dengue is hyperendemic, all four dengue virus strains have been shown to be in local circulation.5 We hypothesize that the infection rate of both primary and secondary dengue would be relatively high in our population during the 3–6 months before delivery as reported cases of clinical dengue in Malaysia have increased almost 27-fold, from 1,487 to 39,654 cases between 1973 and 2005.6 In 2006, there were 38,556 clinical dengue cases reported to the Malaysian authorities: 19,605 of these cases were from the territories of Kuala Lumpur and Selangor (private correspondence with the Vector Borne Disease Control Section, Ministry of Health of Malaysia), which are encompassed in our hospital’s catchment area. A study of pregnant women from Thailand indicates that 94.7% had serological evidence of previous dengue infection.7
Vertical transmission of dengue has been reported in case reports or small series.8–14 Maternal and pregnancy effects have also been investigated in case reports or small series of hospitalized cases.8,15,16 These reports, by their selective nature, tended to overemphasize the severity of dengue infection in pregnancy. A recent review of dengue in pregnancy has found that little systematic research had been done, and based on the limited data available, pregnancy does not appear to increase the incidence or severity of dengue.17
Detection of dengue-specific immunoglobulin M (IgM) is a suitable method for ascertaining acute or recent infection because it is present early in infection and can persist for up to 6 months.2,18 The dengue IgM test cross-reacts across all four dengue serotypes.19 Immunoglobulin M is also produced in a secondary infection.20 Because maternal IgM does not cross the placenta, detection of specific IgM in the neonates is considered to be evidence of vertical transmission of infection.9 Reverse transcriptase polymerase chain reaction (PCR) for dengue has also been used to detect vertically transmitted neonatal infection.11,12
We sought to evaluate the prevalence of dengue infection in pregnancy, to estimate the vertical transmission rate, and to compare pregnancy outcome with the dengue IgM–negative parturient population as controls.
MATERIALS AND METHODS
We performed a prospective study on women delivering at our center based on dengue IgM detection in paired maternal and umbilical cord blood sera to evaluate dengue prevalence and the vertical transmission rate. Labor ward records of these women were checked for pregnancy outcome. Participants answered a three-part questionnaire about whether they had received a clinical diagnosis of dengue during pregnancy, a febrile illness during the pregnancy, or a febrile illness within the last 10 days of delivery.
We anticipated a maternal dengue IgM–positive prevalence rate of 2% in our parturients. We calculated that 50 dengue IgM–positive parturients would be required to demonstrate an increase in cesarean delivery rate from 30%, which is the background rate in our institution, to 50% (alpha 0.05, power 80%, control/case ratio of 49:1). We planned to obtain 2,500 paired maternal-cord blood samples for our study. Ethics approval for the study was obtained from the University of Malaya Medical Centre Medical Ethics Committee.
From May 22 to November 30, 2006, all the women who delivered at or delivered en route to our hospital were approached for their written consent to participate in the study. Women who delivered macerated stillbirths were excluded because we felt that umbilical cord blood was not likely to be obtainable. In 2006, a total of 5,435 women delivered at our university hospital, which served a metropolitan population. We had intended to recruit as many eligible women as possible during our recruitment period. Recruitment, questionnaire administration, and maternal and umbilical cord blood collection were performed by frontline providers and ward nurses after they were trained regarding the procedures by the investigators.
Venous blood was collected from recruited women in labor or within 24 hours of delivery and placed in plain blood bottles. The umbilical cord blood was taken directly from the cord immediately after delivery. No neonate was bled directly. Unpaired or missing samples were discarded and excluded from the study.
The blood samples obtained were kept in a refrigerator at 4°C before transportation to the laboratory for processing. Blood was transferred to the laboratory twice daily on working days and once a day during holidays. We sent our samples to a World Health Organization Collaborating Center for Arbovirus Reference and Research, based at our university, for testing.
Cases were excluded if the centrifuged sample appeared lysed or the quantity of serum available was inadequate for testing. The spun samples were aliquoted and stored at –70°C. Maternal sera were tested in batches using an in-house IgM-capture enzyme-linked immunosorbent assay (ELISA) for dengue.21,22 Samples were considered positive if the ratio of optical density of the positive to negative control was at least 2. Maternal sera positive for dengue IgM was also tested for dengue viral RNA particles with a reverse transcriptase PCR test.5 The umbilical cord sera samples paired to IgM-positive maternal samples were also similarly tested for dengue IgM and dengue virus by reverse transcriptase PCR. Pregnancy outcome data were extracted from labor ward records of the participants.
Data were entered into SPSS 14 (SPSS Inc., Chicago, IL). Comparison of the means of continuous variables was by the Student t test and ordinal variables by the Mann-Whitney U test. Fisher exact test was used on 2×2 categorical data sets and the χ2 test on larger categorical data sets. Confidence intervals of proportions were calculated. All tests were two-sided, and P<.05 was considered significant.
Recruitment process for the study is depicted in the flow chart in Figure 1. During the recruitment period, 2,963 women delivered at or en route to our hospital. Because there was no initial arrangement for delivery of samples to the laboratory during the weekend, a total of 208 women who delivered within the first six weekends of study commencement were not approached for recruitment. Also, during the initial stage of recruitment, 59 women admitted directly to the antenatal ward for elective cesarean delivery were missed; we classified them within miscellaneous group in Figure 1. The other eight women in the miscellaneous group were not approached for reasons that were uncertain; this was identified on cross-checking the study database against the birth register entries over the study period. Only 44 of 2,642 (1.7%) women who were formally approached declined to participate. After recruitment, 67 of 2,598 (2.6%) women-infant pairs were excluded for technical reasons: inadequate, missing, or lysed blood samples.
Of the 2,531 maternal sera tested for dengue IgM, 63 (2.5%, 95% confidence interval [CI] 1.9–3.2%) were positive; 64 (including one pair of twins) of the linked cord blood sera were subsequently tested for dengue IgM, with one (1.6%, 95% CI 0.0–9.5%) positive result (Fig. 2). None of the reverse transcriptase PCR tests for dengue viral RNA particles on the aforementioned maternal or umbilical cord samples was positive. The sequence of laboratory testing is illustrated in Figure 2.
The laboratory test results correlated with self-reported febrile illness and diagnosis of clinical dengue were shown in Fig. 3. Of dengue IgM–positive women, 56 of 63 (88.9%) did not report a febrile illness in pregnancy. Of the women who reported a febrile illness during their pregnancy, seven of 132 (5.3%) were dengue IgM–positive compared with 56 of 2,399 (2.3%) IgM positives in women who did not report a febrile illness (P=.043; odds ratio 2.34, 95% CI 1.05–5.25).
The characteristics of the IgM-positive and IgM-negative subgroups were similar except for mean (±standard deviation) maternal age of 30.6±5.2 compared with 29.2±4.9 years (P=.025), respectively, but when categorized into age groups 35 years or older compared with younger than 35 years, no significant difference was demonstrated (Table 1).
The pregnancy outcomes of IgM-positive participants compared with IgM-negative controls are shown on Table 2. No difference was found in the requirement for labor induction, mode of delivery, preterm delivery rate, low birth weight rate, postpartum hemorrhage rate, umbilical cord blood pH value, neonatal admission rate, or neonatal mortality rate for dengue IgM–positive women compared with control IgM–negative women. Correction for maternal age did not change findings.
Dengue IgM was found in 2.5% of our parturients. In a recent study of 245 parturients from Thailand, there were no positive dengue IgM tests.7 In their study, 94.7% had dengue IgG antibodies. This would suggest that, if IgM were detected, it would likely be provoked by a secondary infection by a different serotype. Immunoglobulin M response is less pronounced in a secondary infection2,20 and might become undetectable at a faster rate. Our study design did not permit us to differentiate between primary or secondary infections in our parturients. A recent community-based cross-sectional dengue antibody prevalence study from the same catchment area as that of our parturients indicates a lower dengue antibody rate of 63% in those aged 21–40 years,23 compared with the 94.7% rate of the Thai study.7
Vertical transmission is a relatively low 1.6% (with only a single fetus affected) as defined by the presence of dengue IgM in umbilical cord serum. The affected neonate did not manifest any problems at birth. A study from Cuba of infants born to women with clinically apparent dengue during pregnancy showed four of 59 (6.8%) were positive for dengue IgM.24 Another study from French Guyana, also of women with clinical dengue infection in pregnancy, reported two of 19 (10.5%) neonates vertically infected by dengue.25 However, a study from northern India of clinical dengue in pregnancy has shown no vertical infection in eight pregnancies.26 In our study with a 1.6% vertical transmission rate, 56 of 63 (88.9%) women who were dengue IgM–positive did not report any febrile illness during pregnancy. It is possible that the vertical transmission rate might be dependent on the severity of maternal dengue.
All tests with reverse transcriptase PCR were negative in the maternal IgM–positive samples as well as in the paired cord samples. This is in keeping with a previous finding that reverse transcriptase PCR to detect viral RNA particles was invariably negative a few days after defervescence.27 There were also no positive reverse transcriptase PCR results among parturients with recent fever or in their infants’ cord samples, indicating that we did not detect any active dengue infection around delivery in our study population. It is possible that a few cases of dengue viremia in asymptomatic women with dengue virus in their circulation might have been missed because we did not test all maternal sera with reverse transcriptase PCR.
We did not detect any significant differences in pregnancy outcome comparing dengue IgM–positive with IgM–negative women. Small case series have reported that maternal clinical dengue may be associated with pregnancy complications, including maternal mortality,16 preterm delivery,10,16,25,28 fetal death,10,16,25 low birth weigh,15,28 neonatal admissions,16 fetal anomalies,28,29 and miscarriage.15,16 Other small case series have shown a more benign outcome26,30 similar to our finding.
There are some limitations in our study design. Because dengue IgM typically persists for only 3–6 months2,18 after infection, our methodology was capable of identifying infection only within the last 6 months of the pregnancy. In addition, specific IgM might not be detectable in a small proportion of secondary infections.2,20 This would decrease the sensitivity of our test for detecting dengue infection throughout the whole of pregnancy. Our IgM-capture ELISA assay for dengue cross-reacts with IgM against other flaviviruses, notably Japanese encephalitis and yellow fever viruses, but IgM-capture ELISA for dengue has demonstrated only a low cross-reactivity with Japanese encephalitis virus.31 None of the 18 cases of Japanese encephalitis cases reported to the Malaysian authorities in 2006 came from our catchment area, and no yellow fever case was reported in 2006 (private communication, the Vector Borne Disease Control Division of the Ministry of Health of Malaysia). We used both IgM and reverse transcriptase PCR on cord sera to establish fetal infection and confirm vertical transmission. The fetus is able to mount an immune response to dengue infection with resultant IgM production,9,24 but the dengue IgM test’s sensitivity for detecting fetal dengue infection is not established. We have one dengue IgM–positive result from cord serum. It remains unclear how effectively the fetus can mount an IgM response or how for long IgM would continue to be detectable in the fetal circulation in response to in utero dengue exposure. Reverse transcriptase PCR, although highly sensitive and specific,32 will only be positive during the acute phase of infection and is typically negative shortly after defervescence, so the window of opportunity for detection of dengue virus RNA to confirm infection is relatively narrow.27 Also, our study design would have underestimated the vertical transmission rate due to clearance of maternal dengue IgM while the fetus was still dengue IgM–positive.
In summary, maternal dengue infection affects at least 2.5% of pregnancies in our hyperendemic setting. Vertical transmission, as defined by detection of dengue IgM in cord serum, is only 1.6%. Pregnancy outcome does not appear to be adversely affected in women recently infected with dengue. Further research should focus on the effect that dengue infection in the first trimester may have on outcome.
1. Deen JL, Harris E, Wills B, Balmaseda A, Hammond SN, Rocha C, et al. The WHO dengue classification and case definitions: time for a reassessment. Lancet 2006;368:170–3.
2. Wilder-Smith A, Schwartz E. Dengue in travelers. N Engl J Med 2005;353:924–32.
3. Gibbons RV, Vaughn DW. Dengue: an escalating problem. BMJ 2002;324:1563–6.
5. Kong YY, Thay CH, Tin TC, Devi S. Rapid detection, serotyping and quantitation of dengue viruses by TaqMan real-time one-step RT-PCR. J Virol Methods 2006;138:123–30.
7. Perret C, Chanthavanich P, Pengsaa K, Limkittikul K, Hutajaroen P, Bunn JE, et al. Dengue infection during pregnancy and transplacental antibody transfer in Thai mothers. J Infect 2005;51:287–93.
8. Thaithumyanon P, Thisyakorn U, Deerojnawong J, Innis BL. Dengue infection complicated by severe hemorrhage and vertical transmission in a parturient woman. Clin Infect Dis 1994;18:248–9.
9. Chye JK, Lim CT, Ng KB, Lim JM, George R, Lam SK. Vertical transmission of dengue. Clin Infect Dis 1997;25:1374–7.
10. Carles G, Peiffer H, Talarmin A. Effects of dengue fever during pregnancy in French Guiana. Clin Infect Dis 1999;28:637–40.
11. Janjindamai W, Pruekprasert P. Perinatal dengue infection: a case report and review of literature. Southeast Asian J Trop Med Public Health 2003;34:793–6.
12. Witayathawornwong P. Parturient and perinatal dengue hemorrhagic fever. Southeast Asian J Trop Med Public Health 2003;34:797–9.
13. Sirinavin S, Nuntnarumit P, Supapannachart S, Boonkasidecha S, Techasaensiri C, Yoksarn S. Vertical dengue infection: case reports and review. Pediatr Infect Dis J 2004;23:1042–7.
14. Poli L, Chungue E, Soulignac O, Gestas P, Kuo P, Papouin-Rauzy M. Materno-fetal dengue. Apropos of 5 cases observed during the epidemic in Tahiti (1989) [in French]. Bull Soc Pathol Exot 1991;84:513–21.
15. Waduge R, Malavige GN, Pradeepan M, Wijeyaratne CN, Fernando S, Seneviratne SL. Dengue infections during pregnancy: a case series from Sri Lanka and review of the literature. J Clin Virol 2006;37:27–33.
16. Ismail NA, Kampan N, Mahdy ZA, Jamil MA, Razi ZR. Dengue in pregnancy. Southeast Asian J Trop Med Public Health 2006;37:681–3.
17. Carroll ID, Toovey S, Van Gompel A. Dengue fever and pregnancy - a review and comment. Travel Med Infect Dis 2007;5:183–8.
18. Nogueira RM, Miagostovich MP, Cavalcanti SM, Marzochi KB, Schatzmayr HG. Levels of IgM antibodies against dengue virus in Rio de Janeiro, Brazil. Res Virol 1992;143:423–7.
19. De Paula SO, Fonseca BA. Dengue: a review of the laboratory tests a clinician must know to achieve a correct diagnosis. Braz J Infect Dis 2004;8:390–8.
20. Sa-Ngasang A, Anantapreecha S, A, AN, Chanama S, Wibulwattanakij S, Pattanakul K, et al. Specific IgM and IgG responses in primary and secondary dengue virus infections determined by enzyme-linked immunosorbent assay. Epidemiol Infect 2006;134:820–5.
21. Osman O, Fong MY, Devi S. A preliminary study of dengue infection in Brunei. Jpn J Infect Dis 2007;60:205–8.
22. Lam SK, Devi S, Pang T. Detection of specific IgM in dengue infection. Southeast Asian J Trop Med Public Health 1987;18:532–8.
23. Chen WS, Wong CH, Cillekens L. Dengue antibodies in a suburban community in Malaysia. Med J Malaysia 2003;58:142–3.
24. Fernandez R, Rodriguez T, Borbonet F, Vazquez S, Guzman MG, Kouri G. Study of the relationship dengue-pregnancy in a group of Cuban-mothers [in Spanish]. Rev Cubana Med Trop 1994;46:76–8.
25. Carles G, Talarmin A, Peneau C, Bertsch M. Dengue fever and pregnancy. A study of 38 cases in french Guiana [in French]. J Gynecol Obstet Biol Reprod (Paris) 2000;29:758–62.
26. Malhotra N, Chanana C, Kumar S. Dengue infection in pregnancy. Int J Gynaecol Obstet 2006;94:131–2.
27. Sa-ngasang A, Wibulwattanakij S, Chanama S, O-rapinpatipat A, A-nuegoonpipat A, Anantapreecha S, et al. Evaluation of RT-PCR as a tool for diagnosis of secondary dengue virus infection. Jpn J Infect Dis 2003;56:205–9.
28. Restrepo BN, Isaza DM, Salazar CL, Ramirez JL, Upegui GE, Ospina M, et al. Prenatal and postnatal effects of dengue infection during pregnancy [in Spanish]. Biomedica 2003;23:416–23.
29. Sharma JB, Gulati N. Potential relationship between dengue fever and neural tube defects in a northern district of India. Int J Gynaecol Obstet 1992;39:291–5.
30. Figueiredo LT, Carlucci RH, Duarte G. Prospective study with infants whose mothers had dengue during pregnancy [in Portuguese]. Rev Inst Med Trop Sao Paulo 1994;36:417–21.
31. Yabe S, Nakayama M, Yamada K, Kitano T, Arai Y, Horimoto T, et al. Laboratory virological diagnosis of imported dengue cases [in Japanese]. Kansenshogaku Zasshi 1996;70:1160–9.
32. Chan SY, Kautner I, Lam SK. Detection and serotyping of dengue viruses by PCR: a simple, rapid method for the isolation of viral RNA from infected mosquito larvae. Southeast Asian J Trop Med Public Health 1994;25:258–61.