Direct identification of Streptococcus agalactiae and capsular type by real-time PCR in vaginal swabs from pregnant women

Abstract

Most group B streptococcus (GBS) infections in newborns are with capsular type Ia, Ib, or III. To prevent these infections more effectively, we developed a real-time PCR method to simultaneously detect GBS species and identify these 3 capsular types in vaginal swab samples from women at 36–39 weeks of gestation. DNA to be detected included those of the dltS gene (encoding a histidine kinase specific to GBS) and cps genes encoding capsular types. PCR sensitivity was 10 CFU/well for a 33−35 threshold cycle. Results were obtained within 2 h. Direct PCR results were compared with results obtained from cultures. Samples numbering 1226 underwent PCR between September 2008 and August 2012. GBS positivity rates by direct PCR and after routine culture were 15.7% (n = 192) and 12.6% (n = 154), respectively. Sensitivity and specificity of direct PCR relative to culture were 96.1% and 95.9%. Of GBS positive samples identified by PCR, capsular types determined directly by real-time PCR were Ia (n = 24), Ib (n = 32), and III (n = 26). Real-time PCR using our designed cycling probe is a practical, highly sensitive method for identification of GBS in pregnant carriers, allowing use of prophylactic intrapartum antibiotics in time to cover the possibility of unexpected premature birth.

Introduction

Streptococcus agalactiae, or Lancefield group B streptococcus (GBS), is a leading cause of neonatal infections with high mortality and morbidity rates [1], [2]. Colonization of GBS, reported to occur in 10%–30% of pregnant women [3], [4], is a risk factor for birth canal infection associated with severe infections in newborns [4], [5], [6].

GBS infections are classified as early-onset disease (EOD), occurring within the first 6 days after birth; and late-onset disease (LOD), occurring 7–89 days after birth [7]. In EOD, sepsis was the most frequent infection, followed by meningitis [8], [9], [10]; vertical transmission from mothers to neonates has been implicated. In contrast, meningitis predominates in LOD, which involves other routes of transmission from the mother or from environmental sources to neonates [11], [12], [13].

To prevent neonatal GBS infection, CDC guidelines issued in 2002 recommend universal screening of pregnant women for GBS colonization at 35–37 weeks of gestation, followed by intrapartum use of an antimicrobial agent for prophylaxis in any carriers identified [14]. These preventive measures were followed by a significant decline in incidence of EOD over the past decade in the US, while incidence of LOD has remained unchanged [15], [16]. Guidelines were revised again in 2010, advocating use of rapid testing methods for GBS identification and changing the antimicrobials suggested for intrapartum prophylaxis [17].

In Japan, similar guidelines including universal screening of pregnant women at 33–37 weeks of gestation and administration of intrapartum antibiotic prophylaxis in GBS-positive cases, was recommended by the Japan Society of Obstetrics and Gynecology in 2008 [18]. Although incidence of GBS infections per 1000 births is low in our country [19] compared to others [4], [17], [20], GBS remains a leading cause of newborn morbidity and mortality.

Recently, rapid screening of GBS by direct PCR and real-time PCR methods that show high sensitivity and specificity has attracted attention as an improvement over conventional culture-based methods [21], [22], [23], [24], [25], [26], [27], [28], [29].

Although capsular polysaccharide serotypes in GBS include 10 types [30], [31], only 3-- Ia, Ib, and III-- account for most neonatal infections [32], [33], [34]. Among these types, about 70–80% of LOD were caused by type III, which has particularly high virulence [32], [33], [34]. With the aim of preventing GBS infection in neonates, a GBS conjugate vaccine targeted to capsular types Ia, Ib, and III has been developed, and clinical trials in pregnant women are underway [35]. Therefore, rapid identification of serotypes Ia, Ib, and III together with species identification in samples is increasingly important in GBS testing.

In the present study, we constructed a system to directly identify serotypes Ia, Ib, and III, as well as the GBS species, with high sensitivity and specificity by real-time PCR methods using a cycling probe. We describe results of this real-time PCR assay in clinical specimens from pregnant women, as compared with findings based on conventional cultures.