Severe neonatal infection secondary to prenatal transmembranous ascending vaginal candidiasis

Introduction

The prevalence of vaginal candidiasis in pregnancy is high (20–30%) but in the majority of the cases asymptomatic [1]. In women at risk for preterm delivery vaginal and cervical swabs are performed to diagnose infections associated with preterm labor or cervical insufficiency. The general consensus is that treatment for candidiasis in pregnancy is primarily for relief of symptoms and there is no recommendation for treatment of asymptomatic candida colonization [2]. The majority of systemic neonatal fungal infections occur in extremely premature deliveries. Fungal infections acquired through vertical transmission are uncommon [3]. There have been cases of neonatal leukemoid reactions associated with Candida without signs of infection [4]. Neonatal systemic infection with Candida despite being rare, poses risk for severe neonatal pneumonia, sepsis and even death. Especially neonates <27 weeks gestational age or with a birth weight <1000 g are at risk for a poor outcome [5].

Case report

A 41-year-old patient was hospitalized for preterm labor and vaginal bleeding in the 29th week of pregnancy. A McDonald’s cervical cerclage had been performed at 20 weeks of gestation due to amniotic membrane prolapse and cervical dilatation. At that time she received prophylactic antibiotic therapy (clindamycin 600 mg three times daily for 5 days), as well as a fungicidal cream for asymptomatic unspecified cervical and vaginal yeast growth. In her obstetric history, she had two terminations of pregnancy, one early miscarriage and two cesarean sections. In her previous pregnancy an emergency cerclage was performed in 22 weeks.

During her current admission she received antenatal steroid therapy (two doses of 12 mg betamethasone i.m.) to improve fetal lung maturity and tocolytic therapy with fenoterol hydrobromide (partusisten 5 μg every 3 minutes intravenously) was started. Her vital signs were stable, an elevated C-reactive protein CRP of 35.3 mg/L on admission decreased over the next 2 days to 8.7 mg/dL. The leucocytes also decreased from 23.5/nL to 20.2/nL despite the betamethasone treatment. The patient received prophylactic antibiotic therapy with clindamycin 600 mg intravenously three times daily. A low hemoglobin level of 8.0 g/dL was treated with iron supplement (ferrosanol 100 mg daily and later ferrlicit 62.5 mg intravenously). The culture of her cervical swab showed moderate growth of Candida and Escherichia coli, despite being asymptomatic. In 29+6 weeks she complained of increasing contractions despite tocolytic therapy.

A cesarean section was performed. The amniotic fluid was green and there was a 5 cm retroplacental hematoma. The placenta appeared clinically to be infected and was sent for microbiological and pathological examination.

A female infant (1810 g) was delivered with Apgar scores of 7/7/8 at 1, 5 and 10 minutes. The umbilical artery pH was 7.26 with a base excess of –5. The infant was intubated and received endotracheal surfactant and was admitted to the neonatal care unit. Upon delivery the baby showed signs of infection with hepatomegaly and a general maculopapular skin rash (Figure 1). A chest X-ray confirmed mild respiratory distress syndrome. The initial rise in IL-6 from 205 ng/L to 240.9 ng/L and leukocytes from 51.6/nL to 75.6/nL decreased during the course of the stay. Antibiotic therapy was started with ampicillin and gentamicin for 5 days and antimycotic therapy (fluconazole 12 mg/kg/day for 4 days) was added on day 2 when the result of the placenta culture confirmed Candida albicans.

Figure 1:

One-day-old female newborn with skin rash.

The pathology examination of the placenta confirmed severe ascending amnionitis with extensive maternal (intensity grade 3) and fetal reaction (intensity grade 1–2) and severe purulent phlegmonous chorioamnionitis. In addition, there was evidence of a fresh hematoma on the edge of the placenta consistent with preterm placental abruption. The microbiology results of the swabs of the placenta were positive for a moderate growth of C. albicans. There was no evidence of Listeria monocytogenes, which had been the initial clinical diagnosis, especially considering the newborn rash. The blood culture and culture samples of pharyngeal and gastric secretions of the newborn obtained on day 3 revealed growth of C. albicans. The skin rash improved dramatically after antimycotic therapy. Oxygen supplementation was necessary for 5 days and the newborn was stable for the rest of the admission. The neonate was discharged on the 37th day of life.

Discussion

Whilst it is generally accepted that the majority of cases of vaginal candidiasis do not lead to chorioamnionitis or neonatal infection, it is nevertheless worthwhile in women with a high risk of preterm delivery to start treatment with fungicides to prevent ascending infections [2]. In cases where a cervical cerclage is indicated, it is crucial that infections are ruled out preoperatively. In our case, it is reasonable to consider in view of the fact that there was no preterm rupture of membranes, that the neonatal infection was secondary to a transmembranous migration of Candida. The ensuing chorioamnionitis presumably led to preterm placental abruption. This correlates well with work from Nath et al. where a strong association between chorioamnionitis and placental abruption could be shown [6].

A Cochrane analysis has shown that screening for lower genital tract infections and treatment significantly lowered the rate of preterm birth (3% vs. 5%) with a risk ratio (RR) of 0.55 [95% confidence interval (CI) 0.41–0.75] [7].

Based on systematic reviews, antimycotic prophylaxis (fluconazole at 3–6 mg/kg/dose twice weekly) decreased the incidence of invasive Candida infections in infants <1000 g by 91% (OR 0.09; 95% CI 0.04–0.24; p=0.0004) and in infants <1500 g by 85% (OR 0.15; 95% CI 0.08–0.26; p<0.0001). In our case, the baby had a birth weight of 1810 g with a leukemoid reaction (defined as leukocyte count >50/nL), therefore, antimycotic therapy was indicated instead of prophylaxis. Leukemoid reactions have been described in preterm infants with chromosomal abnormalities, antenatal exposure to steroids and severe infections such as chorioamnionitis. In some cases extreme hyperleucocytosis in premature infants occurred in the absence of leukemia and had a mild course. Leukocytosis normalized without any therapy [8].

Conclusion

Our case highlights the importance of screening and treatment of vaginal colonization with Candida in women with threatened preterm labor. Topical treatment is insufficient in cases of preterm labor and especially cerclage placement. Additionally, follow up cultures should be performed to document clearance of vaginosis. If colonization persists, typing and susceptibilities of the Candida isolates should be performed to determine optimal antifungal therapy.

Whilst the incidence of neonatal infection with Candida is rare the consequences of neonatal morbidity and mortality is significant and a high suspicion with timely diagnosis and definitive management is warranted. Current guidelines need to take this into account and be modified accordingly.