Xpert GBS Assay for Rapid Detection of Group B Streptococcus in Gastric Fluid Samples from Newborns

Abstract

The Xpert GBS real-time PCR assay was applied to gastric fluid samples from 143 newborns, and it detected group B streptococcus (GBS) within 1 h for 16 (11.2%) cases, while microscopic examination detected only 2 cases. The sensitivity and specificity of the Xpert GBS were 80% and 100%, respectively, with regard to 20 cases of GBS colonization or infection. Concordance of Xpert GBS results versus culture was 92.3%. This test detects in a timely manner newborns at risk for invasive GBS disease.

TEXT

Early-onset group B streptococcus (GBS) neonatal disease is a serious and frequent complication of birth in developed countries. GBS causes more than one-third of the materno-fetal infections (MFI) of term infants, followed by Escherichia coli and other Streptococcus spp., as well as species now classified in the Enterococcus genus (1). GBS colonizes the maternal genitourinary tract (at reported frequencies from 4 to 36%) during pregnancy in Europe and the United States (2) and may cause an ascending infection after the onset of labor or the preterm rupture of membranes. Neonates may become colonized (in 40 to 70% of the cases [3]) during passage through the birth canal at the time of delivery. Two strategies have significantly reduced the rate of early neonatal GBS colonization and infection (4): (i) late antenatal GBS screening of pregnant women and (ii) intrapartum chemoprophylaxis for colonized women to reduce vertical transmission of GBS. Despite international guidelines, the worldwide incidence rate of GBS neonatal diseases is still 0.23 to 1.22‰ of births (5, 6).

Because early detection of infected newborns is clinically difficult and the newborn's health can worsen rapidly, GBS detection for newborns is one of the emergencies in clinical microbiology. Gastric fluid is the usual relevant specimen for GBS detection (7). Its microscopic examination is so far the only bacteriological finding rapidly available for clinicians who suspect an early MFI. GBS microbiological diagnosis is delayed, since it is based on an 18-h culture. Since microscopic examination has a low sensitivity (27.5%) (8), a rapid method of GBS detection that is reasonably sensitive and specific would be valuable in stratifying the risk of MFI among neonates.

The Xpert GBS assay (Cepheid) is a marketed real-time PCR for screening maternal carriage by detecting GBS DNA in vaginal or rectal specimens. Recently, two studies have shown its utility when it is performed on vaginal swabs during intrapartum instead of antenatal screening, at 35 to 37 weeks of gestation (9), and on amniotic fluids collected from pregnant women with premature rupture of membranes (10). Our study aimed to evaluate the test's performance when applied to gastric fluid specimens collected from newborns. Our aim was to shorten the time to detect GBS-colonized infants and thus help in the diagnosis of invasive GBS materno-fetal infections.

We conducted a prospective study at Lariboisière University Hospital (Paris, France), where 2,500 deliveries are processed per year in an IIA maternity center. In our institution, gastric fluid, ear fluid, and placenta were collected at the time of the birth and submitted for bacteriological analysis only for newborns at risk of MFI (11). The study included gastric fluid specimens received for 143 newborns and sent to the laboratory during daytime hours from Tuesday to Friday over the course of 3 months. Newborns were of 31 to 41 weeks gestation, and their median birth weight was 3,288 ± 469 g (mean ± standard deviation). A swab was soaked briefly in the gastric fluid, avoiding the mucus, and introduced into the cartridge as for a vaginal swab specimen. The GeneXpert system (Cepheid) automatically processed DNA extraction, purification, and target amplification (cfb gene DNA of Streptococcus agalactiae) in a GBS-specific real-time PCR. The assay controls included an internal PCR control (IC) and a sample control (SPC). Results were obtained in approximately 30 min for positive results and 50 min for negative results. In parallel, gastric fluids were studied by classical bacteriological techniques (microscopic examination after Gram staining and culture). Blood agar plates were incubated at 37°C anaerobically, and chocolate Polyvitex agar plates (bioMérieux, Marcy l'Etoile, France) were incubated aerobically in 5% CO₂ for 24 to 48 h. Colonies suspected to be GBS were identified by matrix-assisted laser desorption ionization–time of flight mass spectrometry (Bruker Biotyper database) (12).

Of 143 gastric fluid samples, 16 (11.2%) were positive by the Xpert GBS assay (Table 1), and 122 (85.3%) were negative. Cycle thresholds (C_T values) for positive samples ranged between 25.3 and 39.3. The assay was invalid for 5 cases (3.5%) with PCR inhibition (negative IC). These specimens were retested after freeze-thawing and showed a negative result. Microscopic examination showed chained Gram-positive cocci in 7 cases (4.9%), but in only 2 cases did the culture lead to a positive GBS culture; the 5 remaining cases were positive for streptococci other than GBS. Cultures were positive for GBS in 17 (11.8%) gastric fluid samples, and the concordance between the Xpert GBS assay results and culture results was 92.3% (kappa = 0.76). For the three specimens negative by culture but positive by PCR (Table 1), the mother had received intrapartum antibiotic prophylaxis, which explain the number of false-negative culture results. In these 3 cases, newborn GBS colonization was confirmed by culture of another specimen (positivity of GBS culture in the ear fluid). Conversely, three specimens were positive by culture but negative in PCR. These false-negative results of the Xpert GBS assay can be explained by the low quantity of bacteria (5 × 10² to 2 × 10³ CFU/ml) present in the fluid. The presence of bacteria trapped in the mucus of gastric fluid, resulting in heterogenous distributions of GBS in the samples, could also explain the discrepancies between the two methods. For newborns positive for GBS in the gastric fluid by the Xpert GBS assay or culture (n = 20), we found that 90% (18/20) and 27% (7/20) of them carried GBS in ear fluid and placenta, respectively. For newborns negative for GBS in the gastric fluid (n = 123), we found GBS in culture in 8.1% (10/123) and 2.4% (3/123) of ear fluid and placenta specimens, respectively.

TABLE 1.

Comparison of the Xpert GBS assay, microscopic examination, and standard culture for GBS detection in 143 gastric fluid samples from newborns

Detection method	Assay result	No. of newborns with GBS colonization result (total no. newborn colonization casesa)
		Positive (20)	Negative (123)	Total (143)
Xpert GBS assay	Positive	16	0	16
	Negative	3	119	122
	Invalidb	1	4	5
Microscopic examination	Positive	2	5	7
	Negative	18	118	136
Standard culturec	Positive	17	0	17
	Negative	3	123	126

^aPositive GBS detection in gastric fluid was defined as a case for which either the Xpert GBS assay, standard culture method, or both methods yielded positive results.

^bAll invalid Xpert GBS results were retested and found to be negative.

^cColumbia agar with 10% sheep blood.

In total, 20 gastric fluid samples were positive for GBS by culture or the Xpert GBS assay. The sensitivities of the Xpert GBS assay, culture, and microscopic examination were 80%, 85%, and 10%, respectively (Table 1). The specificities of the Xpert GBS assay, culture, and microscopic examination were 100%, 100%, and 96%, respectively. We noticed that were two materno-fetal infections due to other Streptococcus spp. (S. constellatus and S. tigurius) for which the Xpert GBS assay gave true-negative results. Informative clinical and biological data for theses 20 cases are detailed in Table 2.

TABLE 2.

Clinical and biological data for mothers and newborns of the 20 GBS cases^a

Result group	Data for mothers (N [%])				Data for newborns (N [%])
	GBS vaginal carriageb	Antibiotic prophylaxisc	Fever during labor	Vaginal delivery	CRP of >10 μg/mld	Neonatology care unite	GBS materno-fetal infection
Negative	8 (40)	12 (60)	18 (90)	3 (15)	13 (65)	14 (70)	15 (75)
Positive	11 (55)	8 (40)	2 (10)	17 (85)	5 (25)	6 (30)	5 (25)
Not done	1 (5)	0	0	0	2 (10)	0	0

^aCases were positive by standard culture and/or Xpert GBS assay for GBS in gastric fluid.

^bBased on antenatal GBS screening of pregnant women at 35 to 37 weeks of gestation.

^cAn intrapartum amoxicillin dose of 2 g was delivered.

^dA CRP value less than 10 μg/ml was considered negative for infection. The five newborns with GBS infection had CRP values ranging from 18 to 90 μg/ml.

^eNewborns with severe clinical conditions were hospitalized in the neonatology care unit.

Among the 20 newborns positive for GBS, 5 cases of MFI were diagnosed according to clinical and biological criteria, such as fever, low Apgar score, respiratory distress, icteria, and increase in C-reactive protein ranging from 18 to 90 μg/ml. No bacteremia or meningitis was observed. The rupture of membranes had occurred between 2 and 9 h before delivery, and the duration of labor ranged from 2 to 11 h. In only one of these five cases had GBS been previously detected as a case of maternal vaginal carriage. All five cases were culture positive, with quantities ranging from 2 × 10³ to >10⁶ CFU/ml. Microscopic examination was positive for only two of them. The Xpert GBS assay detected four of these five cases of infection. The false-negative case was a newborn delivered by cesarean section, and the gastric fluid showed a low quantity of GBS by culture (10³ CFU/ml).

To our knowledge, this is the first evaluation of the Xpert GBS assay in gastric fluids. GBS detection is a microbiological emergency and is conducted 24 hours a day in several places that rely on microscopic examination. However, the performance characteristics of the microscopic detection method are very poor and cannot help in the diagnosis of MFI. On the contrary, the Xpert GBS assay showed performance characteristics similar to those for culture but gave a more rapid result, with results in less than 1 h versus requiring at least 1 day. The identification of GBS in gastric fluid by the Xpert GBS assay should rapidly point out the infants at risk of GBS early-onset infection. This strategy may enhance newborn care by reducing the time to adequate antibiotic therapy.