Neonatal dengue: an under-diagnosed entity

Abstract

Neonatal dengue is an under-diagnosed disease likely due to low index of suspicion along with its resemblance to sepsis. We hereby report two cases of neonatal dengue, highlighting the need of high degree of suspicion in infants born to febrile mothers even with maternal serology being negative. Moreover, severity of neonatal illness positively correlates with the maternal disease.

Background

Dengue is an emerging, pandemic-prone mosquito-borne viral infection, caused by the single-stranded RNA arbovirus, the dengue virus (DENV). High-grade fever, along with haemoconcentration and thrombocytopenia, is the hallmark of severe dengue disease.

An epidemiological study to study the incidence of perinatal transmission of dengue is consistently lacking, however there are few case reports and case series from endemic regions like Sri Lanka, Malaysia and France, which have shown vertical transmission.^1–3 Arragain et al had done a prospective observational study in pregnant women with symptoms of dengue infection 7 days before and 2 days after delivery, during the 2012–2013 dengue epidemic in New Caledonia and revealed vertical transmission in 9 out of 10 enrolled mother–infant pairs.¹ Another prospective cohort study by Tan et al collected paired maternal-umbilical cord blood samples at delivery and revealed that only one of the 64 IgM-positive maternal dengue infection had IgM positivity in umbilical cord blood; however, they were not able to demonstrate the positive real-time PCR (RT-PCR) in any infant.² Similarly, Basurko et al carried out a retrospective study and revealed that maternal–fetal transmission occurred in 3 out of 20 parturient women.³ However single-centred small-population studies from Delhi and Vietnam failed to show any vertical transmission.^{4 5} A single systematic review by Pouliot et al that has included 30 published studies (19 case reports, 9 case series and 2 comparison studies), has described vertical transmission in 64% and 12.6% of women in case reports and case series, respectively.⁶ Thus, the overall incidence of vertical transmission is quite variable. Non-specific symptoms in neonates mimicking sepsis along with low index of suspicion may be the likely reasons for this wide discrepancy.^7–9 We hereby report two cases of neonatal dengue from a tertiary care institute in Western India during the last dengue-epidemic.

Case report

Patient 1: A 22-year-old mother presented to us at 39 weeks and underwent emergency low-segment caesarean section (LSCS) for fetal distress. Mother had fever spikes 3 days prior to delivery and found to have thrombocytopenia with normal leucocyte count at admission. Mother developed severe leucopenia (lowest counts being 2.84X109/L) on day 5 of illness (2 days after delivery) and thrombocytopenia (lowest platelet count being 12X10⁹/L) on day 7 of illness with haemorrhagic manifestations, required intensive care for severe dengue. Maternal work-up for dengue revealed positive non-structural protein (NS-1) by ELISA 2 days after delivery (day 5 of illness) and positive dengue serology IgM by immunochromatography (Nano-Check Ditech) on day 6 of illness (3 days after LSCS). RT-PCR was positive for both DENV-1 and DENV-2 in mother. RT-PCR was also done for Zika virus in mother’s sample which was negative. She received multiple platelet transfusions and intensive care unit care for 72 hours after the delivery.

A singleton, male infant with a birth weight of 3050 g was born with meconium-stained liquor (MSL). Infant developed respiratory distress since birth and received continuous positive airway pressure until 40 hours of life (HOL). Chest X-ray was suggestive of meconium aspiration syndrome. Though infant had no clinical symptoms suggestive of dengue infection with normal haemogram, however due to unexplained maternal fever with leucopenia and thrombocytopenia, NS-1 and IgM dengue was done in infant at 36 HOL which turned out to be negative. RT-PCR was not done (institute’s policy of performing RT-PCR only in NS1 positive cases). Infant developed new onset fever spikes at 72 HOL. Sepsis screen revealed C reactive protein (CRP) of 102.4 mg/L, total leucocyte count of 4.4X109/L and platelet count of 92X109/L. Injection Augmentin and amikacin was started suspecting neonatal sepsis; however due to the strong suspicion of neonatal dengue in the setting of maternal dengue, work-up for dengue was repeated at 72 HOL, which revealed positive NS1 antigen and RT-PCR was positive for DENV-2, however the dengue IgM was negative. Infant developed generalised erythematous rash and decompensated shock at 96 HOL with worsening thrombocytopenia (lowest platelet count being 11.0 X109/L on day 5 of life) (table 1). There was severe transaminitis with features suggestive of third space loss (haemoconcentration, excessive weight gain, pleural effusion and moderate ascites). Infant was managed conservatively with inotrope and given three platelet transfusions for upper gastrointestinal bleeds (figure 1). Antibiotics were stopped after 3 days once blood culture turned out to be sterile. Fluid therapy was titrated based on serial haematocrit. Infant continued to have thrombocytopenia; however, there were no major bleeds with normal neurological examination. Platelet count got normalised by day 18 of life. Serial blood counts are depicted in figure 1. NS-1 antigen was positive until day 11 of life, while dengue IgM and IgG were negative even on day 14, day 21 and day 28 of life. The mother–infant dyad was discharged on day 20 of life, remained haemodynamically stable during subsequent follow-up. The infant is thriving well with normal neurological examination and developmental screening until 18 months of age.

Table 1.

Serial haematological and biochemical investigations in both infants

DOL	Hb (g/L)	PCV (%)	TLC (X 109/L)	Platelet* (X109/L)	Hs-CRP	Procalcitonin†	SGOT‡	SGPT‡	NS-1	IgM	IgG
Patient 1
1	156	46.5	19	260
2	164	47.3	7.56	241					Neg	NR	NR
3	174	51.1	4.4	92	102	1.7			Posit	NR	NR
5	175	50.7	6.35	11
6	172	52.2	9.36	11	29.0	0.54
7	152	46.1	13.8	28					Posit	NR	NR
8	133	39.8	22.820	37
9	124	37.8	30.800	106			3763	866
10	146	47	38.900	47
11	121	37	21.000	39			1575	595	Posit	NR	NR
12	115	35.9	15.450	33
14	105	33.6	11.250	30	17.1	0.1	335	580	Neg	NR	NR
17	106	33	11.02	64
21	110	35.2	13.4	217			78	38	Neg	NR	NR
28									Neg	NR	NR
Patient 2
1	188	55.4	12.980	215					Neg	NR	NR
5	195	59.3	6.35	143					Posit	NR	NR
6	213	63.9	5.47	79	0.59		112	47	Posit	NR	NR
7	201	56	10.06	29
8	198	57	10.900	58					Neg	React	NR
10	178	51.7	10.8	94			62	34
11	181	53.6	9.3	149

*Platelets are expressed as X109/L

†Procalcitonin is mentioned as ng/mL.

‡SGOT and SGPT are expressed as International units per liter

DOL, day of life; Hb, haemoglobin; Hs-CRP, high-sensitivity C reactive protein; IgG, immunoglobulin G; IgM, immunoglobulin M; Neg, negative; NR, non-reactive; NS-1, non-structural protein-1; PCV, packed cell volume; Posit, positive; SGOT, serum glutamic-oxaloacetic transaminases; SGPT, serum glutamic-pyruvic transaminases; TLC, total leucocyte count.

Figure 1.

Serial haemograms in confirmed neonatal dengue cases.

Patient 2: Second mother had fever 2 days prior to delivery; developed thrombocytopenia 48 hours after LSCS. Maternal NS-1 was negative at admission. Mother had recurrence of fever spikes on postoperative day 1 without any other localising symptom. The haemogram revealed worsening thrombocytopenia with lowest platelet count of 14X109/L on day 6 of illness. However, various investigations including urine microscopy and culture, widal, smear for malarial parasite and blood culture were unremarkable. NS-1, RT-PCR and IgM dengue ELISA, repeated on postoperative day 7 in view of persistent fever spikes and confirmation of neonatal dengue, were also negative. Mother received antimicrobial agents for unexplained fever and became afebrile on postoperative day 9. RT-PCR for Zika virus in mother’s sample was also negative.

The infant was a term male infant with birth weight of 3490 g, born by emergency LSCS in view of variable decelerations. The liquor was meconium stained. Infant’s NS-1 protein sent on day 1 of life (due to unexplained acute febrile illness in the mother before delivery) was negative. Though infant remained clinically asymptomatic throughout the hospital stay, however due to strong clinical suspicion because of unexplained maternal fever with thrombocytopenia serial haemograms were done which revealed thrombocytopenia and rising haematocrit on day 5 of life (table 1). NS-1 and IgM dengue done on day 3 of life were negative with NS-1 protein becoming positive on day 5 of life. RT-PCR was positive for DENV-2. Infant was managed conservatively with measured feeds and supportive care. There were no clinical bleeds and did not require any blood transfusions. Infant had seroconversion with reactive IgM ELISA on day 8 of life. Haematocrit and platelets gradually improved and normalised by day 11 of life (figure 1).

Discussion

There is a dogma among epidemiologists that vertical transmission rate of dengue is low and severe dengue is rarely seen among neonates hence, neonate-specific dengue guidelines are not available.^{10 11} Although the wide range of clinical signs that the infected neonate can present, delays the diagnosis and predisposes to concealment of true burden of this condition, some severe cases urges the clinicians to be increasingly watchful for this emergent arbovirus.8 9 12–15 Most recommendations are based on experiences with older children and adults. However, the ‘infection-enhancing antibodies’ acquired by the mother from previous Flavivirus infections are passively transmitted to the newborn and this may result in serious manifestations in the newborn.¹⁶

Our first case had severe dengue requiring inotropic support, while second case had only milder features of haemoconcentration and thrombocytopenia and never required intravenous fluids or platelet transfusion. This very well correlates with the severity of maternal illness.⁶ In our reported first case, pleural effusion and ascites appeared on day 6 of life, which gradually improved by day 15 of life, similar to available literature.¹⁷

In our reported cases, leucopenia and haemoconcentration were the first to appear, followed by worsening thrombocytopenia (figure 1). It is not uncommon to have raised CRP values as the disease advances as was seen in our case and reported in literature.¹⁸

Serological diagnoses include anti-dengue IgM and 90% of infected newborns are reported to have IgM positivity by the sixth day.¹⁹ Our first case had no seroconversion until day 28 while second case had seroconversion on day 8 of life.

There was history of MSL in both the affected infants which is in line with the previous case reports which have also reported vertical transmission in newborns born out of MSL. This may be explained by the fact that the dengue infection could have resulted in fetal distress leading to passage of meconium in-utero.3 6 20 21 However, well-designed studies are required to prove this association.

Glycoprotein NS-1 can be detected early in illness and may persist until first 9 days of life.²² In our first patient, NS-1 was negative at 36 HOL, appeared at 72 HOL and remained positive until day 11 of life. NS-1 can be positive in Zika virus infection also,²³ hence Zika viral PCR was also done in both the cases, which was negative.

Though the tell-tale sign of vertical transmission of dengue in our cases as we could not demonstrate the NS-1 positivity in cord blood and placenta, there are indirect evidences favouring the vertical transmission in our cases. WHO and Centers for Disease Control and Prevention (CDC)have defined an incubation period of 5–7 days (range, 3–10 days) and thrombocytopenia usually develops 3–5 days after onset of fever.^{17 24} We detected thrombocytopenia on day 5 of life in our second patient. Assuming a minimum incubation period of 3 days as per available literature along with a minimum period of another 3 days to develop thrombocytopenia from the day of onset of illness, we assume that thrombocytopenia should have appeared beyond 6 days of life but not on day 5 of life as observed in our case. Thus, there is a likely possibility of vertical transmission though horizontal transmission could not be reliably ruled out.

Viral PCR has the highest sensitivity (80%–100%) in detecting virus during the acute phase of the disease and is considered confirmatory evidence of dengue infection as seen in both the cases. Viral RNA may be affected by transport and storage conditions and may be the reason of negative result in the mother of the second case.¹⁷

To conclude, one should have high index of suspicion of dengue illness among infants born to mothers with febrile illness in perinatal period during dengue epidemics even if the maternal investigations are negative. Moreover, the presence of MSL in mothers with suspected dengue warrants an early suspicion and close monitoring for severe neonatal dengue which positively correlates with the severity of maternal illness in the peripartum period.

Learning points.

• Neonatal dengue is an under-diagnosed entity; a high index of suspicion and prompt investigation is required for timely diagnosis of this sepsis mimicker.

• During dengue epidemic, infants born to mothers with unexplained peripartum fever should be actively evaluated for dengue infection despite negative maternal reports.

• Infants with culture-negative sepsis should be evaluated for seasonal or endemic tropical infections like dengue and accordingly worked-up to minimise unnecessary antibiotic exposure.

• There may be a possibility that the severity of neonatal dengue may positively correlate with severity of maternal illness in the peripartum period.

• Proper neonate-specific dengue guidelines are required as neonates are not miniature adults.

Ethics statements

Patient consent for publication

Parental/guardian consent obtained.