Rec¸u le : 1er juin 2012 Accepte´ le : 26 novembre 2013 Available online 30 December 2013
Available online at
ScienceDirect www.sciencedirect.com
Original article
Comparison of 16S rRNA gene PCR and blood culture for diagnosis of neonatal sepsis Comparaison entre l’ARNr 16S et l’he´moculture pour le diagnostic de septice´mie ne´onatale C.L. Liua, H.W. Aib, W.P. Wangc, L. Chena, H.B. Hub, T. Yec, X.H. Zhua, F. Wanga, Y.L. Liaoa, Y. Wanga, G. Oua, L. Xua, M. Suna, C. Jiana, Z.J. Chena, L. Lia, B. Zhanga, L. Tiana, B. Wanga, S. Yana, Z.Y. Suna,* a Department of clinical laboratory, Tongji hospital, Tongji medical college, Huazhong university of science and technology, 430030 Wuhan, PR China b Department of clinical laboratory, children’s Hospital, 430030 Wuhan, PR China c Department of clinical laboratory, women and children hospital, 430030 Wuhan, PR China
Summary
Re´sume´
Septicemia is a common cause of morbidity and mortality among newborns in the developing world. However, accurate clinical diagnosis of neonatal sepsis is often difficult because symptoms and signs are often nonspecific. Blood culture has been the gold standard for confirmation of the diagnosis. However, the sensitivity is low and results are usually not promptly obtained. Therefore, the diagnosis of sepsis is often based on clinical signs in association with laboratory tests such as platelets count, immature/total neutrophils ratio (I/T), and a rise in C-reactive protein (CRP). Polymerase chain reaction (PCR) methods for the detection of neonatal sepsis represent new diagnostic tools for the early identification of pathogens. Methods. During a 4-month prospective study, 16S rRNA PCR was compared with conventional blood culture for the diagnosis of neonatal bacterial sepsis. In addition, the relationship between known risk factors, clinical signs, laboratory parameters, and the diagnosis of sepsis was considered. Results. Sepsis was suspected in 706 infants from the intensive neonatal care unit. They all were included in the study. The number of positive cultures and positive PCR results were 95 (13.5%) and 123 (17.4%), respectively. Compared with blood culture, the diagnosis of bacterial sepsis by PCR revealed a 100.0% sensitivity, 95.4% specificity, 77.2% positive predictive value, and 100.0% negative predictive value. In this study, Apgar scores at 5 min, weight, icterus, irritability, feeding difficulties, gestational age (GA), premature rupture of membrane (PRM), platelets count, I/T, and a marked rise in CRP were important in establishing the diagnosis of sepsis in the newborn. In addition, weight, GA, PRM, irritability,
La septice´mie ne´onatale est la cause principale de la morbidite´ et de mortalite´ ne´onatale dans les pays en de´veloppement. Les symptoˆmes cliniques de cette maladie e´tant peu spe´cifiques, son diagnostic clinique pre´coce est assez difficile. A` l’heure actuelle, l’he´moculture est la re´fe´rence pour le diagnostic d’infection ne´onatale, mais son re´sultat est tre`s long a` obtenir et sa sensibilite´ est faible. Par conse´quent, le diagnostic de septice´mie est formule´ sur la base des symptoˆmes cliniques et des tests de laboratoire, tels que la nume´ration plaquettaire, le rapport neutrophiles immatures/total des neutrophiles (I/T), la de´tection de la prote´ine C-re´active (CRP), etc. Ces dernie`res anne´es, la me´thode de re´action par polyme´risation en chaıˆne (PCR) est devenue une me´thode de de´tection pre´coce des germes responsables d’infections bacte´riennes ne´onatales. Me´thodes. Au cours d’une e´tude prospective de quatre mois, nous avons compare´ la PCR de l’ARNr 16S a` l’he´moculture conventionnelle pour la de´tection des septice´mies ne´onatales, et recherche´ la relation entre les facteurs de risque classiques, les symptoˆmes cliniques, les indicateurs biologiques avec un diagnostic de sepsis confirme´. Re´sultats. Sept cent six nouveau-ne´s de l’unite´ ne´onatale de soins intensifs pre´sentant un risque de septice´mie ne´onatale ont e´te´ inclus. Les he´mocultures et la PCR de l’ARNr 16S e´taient positives dans respectivement 95 (13,5 %) et 123 (17,4 %) cas. Par rapport a` l’he´moculture, la sensibilite´ de la PCR e´tait de 100 %, sa spe´cificite´ de 95,4 %, la valeur positive pre´dictive de 77,2 % et la valeur ne´gative pre´dictive de 100 %. Cette e´tude a par ailleurs re´ve´le´ que le score d’Apgar a` 5 minutes, le poids, l’icte`re, l’irritabilite´, les
* Corresponding author. e-mail: [email protected] (Z.Y. Sun). 0929-693X/$ - see front matter ß 2013 Published by Elsevier Masson SAS. http://dx.doi.org/10.1016/j.arcped.2013.11.015 Archives de Pe´diatrie 2014;21:162-169
162
Comparison of 16S rRNA gene PCR and blood culture for diagnosis of neonatal sepsis
duration of antibiotic usage, mortality rate, and number of purulent meningitis cases were significantly different between early-onset sepsis and late-onset sepsis. Conclusion. 16S rRNA PCR increased the sensitivity in detecting bacterial DNA in newborns with signs of sepsis, allowed a rapid detection of the pathogens, and led to shorter antibiotic courses. However, uncertainty about the bacterial cause of sepsis was not reduced by this method. 16S rRNA PCR needs to be further developed and improved. Blood culture is currently irreplaceable, since pure isolates are essential for antimicrobial drug susceptibility testing. ß 2013 Published by Elsevier Masson SAS.
difficulte´s d’alimentation, l’aˆge gestationnel, la rupture pre´mature´e des membranes, la nume´ration plaquettaire, l’augmentation de l’I/T et de la CRP e´taient utiles pour le diagnostic de septice´mie ne´onatale. En outre, le poids, l’AG, la RPM, l’irritabilite´, le temps d’utilisation des antibiotiques, le taux de mortalite´ et le nombre de me´ningites purulentes e´taient significativement diffe´rents selon que la septice´mie e´tait d’apparition pre´coce ou tardive. Conclusion. La PCR de l’ARNr 16S pre´sente des avantages pour le diagnostic de septice´mie ne´onatale comme une haute sensibilite´, une de´tection rapide des agents pathoge`nes ou une re´duction du temps de l’antibiothe´rapie. Cependant, cette me´thode doit eˆtre de´veloppe´e et perfectionne´e. L’he´moculture reste irremplac¸able afin d’isoler les souches responsables pour tester leur sensibilite´ aux antibiotiques et orienter la prescription en clinique. ß 2009 Publie´ par Elsevier Masson SAS.
1. Introduction
2. Materials and methods
A large proportion of infants admitted to neonatal intensive care units (NICUs) present signs of sepsis; however, accurate clinical diagnosis of neonatal sepsis is difficult, because symptoms and signs are often nonspecific [1–3]. Especially in early-onset sepsis (EOS, within 72 h after birth), the incidence of culture-positive sepsis is low (1–8 cases/1000 livebirths), and the risk of mortality is high, ranging from 10 to 50% [4,5]. Alterations in laboratory test parameters provide evidence for the diagnosis of neonatal sepsis. Because the sensitivity of each laboratory test is quite low [6,7], the search for a reliable testing method has continued. Blood culture is the gold standard for confirmation of the diagnosis; however, blood culturing techniques may have unacceptably low sensitivities, and it often takes 48–72 h to obtain a preliminary positive result [8]. Infants with ‘‘risk factors’’ for neonatal sepsis are thus treated with broad-spectrum antibiotics and require prolonged hospitalization. Recently, C-reactive protein (CRP), interleukin (IL)-6, and procalcitonin (PCT) have been shown to be useful markers of sepsis in many studies [9,10]. 16S rRNA gene polymerase chain reaction (PCR) and sequencing has been identified as a method for defining an organism as a bacterium. This method may prove useful for detecting septicemia in infants whose mothers received intrapartum antibiotics and for diagnosing nonculturable pathogens. It has an excellent sensitivity and a shorter turnaround time than the conventional blood culture technique [11,12]. The aim of this study was to evaluate the molecular diagnosis of neonatal sepsis using PCR amplification of 16S rRNA in comparison with risk factors or clinical evidence of sepsis and to understand the pathogenic spectrum of neonatal septicemia in the region. In addition, we also analyzed the relationship between known risk factors, clinical signs, and laboratory parameters in suspected neonatal sepsis and in the diagnosis of sepsis.
2.1. Patients The study was carried out from September 1, 2011, to December 31, 2011, at the NICU of the Women and Children Hospital, the Children’s Hospital, and Tongji Hospital in Hubei Province, which are all tertiary care centers in China. All infants who had blood samples drawn for concomitant culture, a complete blood count (CBC), and a CRP assay were eligible for inclusion in the study. At the same time, an additional 0.5–1 ml EDTAblood sample was collected from the infants for the purposes of 16S rRNA gene PCR amplification. Sample rejection or exclusion criteria for PCR included blood volumes of less than 500 ml, grossly hemolyzed or clotted blood specimens, and inadequate or missing clinical data.
2.2. Blood culture processing We used an automated continuous-monitoring blood culture system, BACTEC 9240 (Becton Dickinson, Sparks, MD, USA) or BacT/Alert 3D (Biome´rieux, Durham, NC, USA). The pediatricsample-sized, resin-containing blood culture bottles (Peds Plus/F; Becton Dickinson) were sent from the NICU prefilled with blood collected in a syringe from infants at their besides. A minimum of 1 ml of whole-blood was added per blood culture bottle. The bottles were incubated immediately upon receipt in the microbiology laboratory in accordance with the manufacturer’s recommendation.
2.3. DNA extraction The EDTA-blood samples for PCR were blinded and stored at – 70 8C before analysis. Samples (500 ml) were extracted with the blood gDNA mini-kit (Biomiga, USA), supplemented with lytic enzyme (Puregene DNA extraction kit for Gram-positive bacteria; Qiagen, Germany). All DNA isolation procedures were carried out in a Class-II Biological Safety Cabinet
163
Archives de Pe´diatrie 2014;21:162-169
C.L. Liu et al.
(MicroFlow, Weston-super-Mare, UK) in order to minimize contamination and the possibility of false positive results. For each batch of extractions, a negative extraction control containing all reagents minus blood was performed, as well as positive controls with Escherichia coli ATCC 25922 and Staphylococcus aureus ATCC 25923.
2.4. PCR amplification and product detection All DNA extracts were examined by amplification of the 16S rRNA gene using two pairs of universal primers: PSL(f): 50 -AGG ATT AGA TAC CCT GGT AGT CCA-3, UN1(r): 50 -TRR YKT GAC GKG CGG TGT GTA C-30 ; and P11P(f): 50 -GAG GAA GGT GGG GAT GAC GT-30 P13P(r): 50 -AGG CCC GGG AAC GTA TTC AC-30 [13]. The primers provide PCR products of 630 bp and 216 bp, respectively. Each PCR reaction (50 ml) consisted of 1 Taq buffer with (NH4)2SO4, 2.5 U Taq polymerase, 2.5 mM MgCl2, 0.5 mMdNTP, 2-ml template DNA, and ddH2O (Applied MBI Fermentas, Canada). Cycling conditions included a 5-min denaturing step at 94 8C followed by 30 cycles of 1 min at 94 8C, 1 min at 55 8C, and 1 min at 72 8C; a final extension cycle of 72 8C for 10 min was then carried out. Two pairs of primer had similar PCR system and reaction conditions. Sterile water, E. coli ATCC 25922, and S. aureus ATCC 25923 were used as the negative and positive controls, respectively. Separate dedicated pipetting devices with filter-sealed tips were used for each procedure, and UVlight irradiation was used before each procedure to prevent DNA cross-contamination. Positive PCR products were directly sequence analyzed (Invitrogen, USA). The results were analyzed by PCR using two pairs of primer P11/P13, PSL/UNI. PCR results were considered positive when visible PCR products of the correct size were found simultaneously in two reactions. PCR results were considered negative when no visible PCR products of correct size were found. Results were considered inconclusive when single positive runs could not be repeated. The results should then be combined with clinical signs and laboratory parameters: if the newborn has manifest clinical signs and laboratory parameters of sepsis it is determined to be neonatal sepsis, otherwise the results are considered negative.
Figure 1. Specificity of 16S rRNA PCR. 1: marker; 2: Escherichia coli ATCC 25922; 3: Staphylococcus aureus ATCC 25923; 4: negative qc 1; 5: negative specimen; 6: positive specimen.
2.5. PCR assay sensitivity The sensitivity of the 16S rRNA PCR assay was determined using whole-blood samples spiked with known amounts of viable organisms, not purified bacterial DNA extracts. To this end, known numbers of CFU of control strains, E. coli ATCC 25922 or S. aureus ATCC 25923, were added to 0.5-ml volumes of wholeblood containing EDTA. The spiked whole-blood samples were prepared and amplified under exactly the same conditions as those described for the clinical blood specimens. Using this approach, the sensitivity for the Gram-positive bacteria S. aureus and the Gram-negative organisms E. coli was 104 CFU/ml and 103 CFU/ml, respectively. The detection limits of two pairs of primer were similar (figs. 1–3).
164
Figure 2. Sensitivity of 16S rRNA PCR for Escherichia coli. 1: marker; 2: 7 6 5 4 3 10 CFU/ml; 3: 10 CFU/ml; 4: 10 CFU/ml; 5: 10 CFU/ml; 6: 10 CFU/ml; 7: 2 10 CFU/ml.
2.6. Data analysis Study patients were allocated into two groups: P (blood culture or PCR or both positive) and N (blood culture and PCR negative). Risk factors, clinical signs, and laboratory
Comparison of 16S rRNA gene PCR and blood culture for diagnosis of neonatal sepsis
Table II Microbiological results of 16S rRNA and conventional blood culture. Bacteria
Coagulase-negative Staphylococcus sp. Staphylococcus aureus Enterococcus sp. Group B Streptococcus Streptococcus viridans Streptococcus pneumoniae Other Gram-positive cocci Escherichia coli Klebsiella pneumoniae Enterobacter cloacae Pseudomonas sp. Other nonfermenting genus sp. Other Gram-negative bacilli Total Figure 3. Sensitivity of 16S rRNA PCR for Staphylococcus aureus. 1: marker; 2: 7 6 5 4 3 10 CFU/ml; 3: 10 CFU/ml; 4: 10 CFU/ml; 5: 10 CFU/ml; 6: 10 CFU/ml; 7: 102CFU/ml.
parameters were compared between the two groups. A P level of < 0.05 was considered statistically significant. The statistical program SPSS V13.0W was used for data analysis.
3. Results 3.1. Clinical data During the study period, 706 neonates were admitted (432 boys) to the NICU: 332 cases in the Children’s Hospital, 335 cases in the Women and Children Hospital, and 39 cases in Tongji Hospital. The mean birth-weight was 2,431 504 g. The average age was 7.8 days, ranging from 4 h to 28 days.
3.2. Comparison of 16S rRNA PCR and blood culture The positivity rates of culture and PCR were 13.5% (95/706) and 17.4% (123/706), respectively. Compared with blood culture, the diagnosis of bacterial sepsis by PCR revealed a 100.0% sensitivity, 95.4% specificity, 77.2% positive predictive value, and 100.0% negative predictive value (table I). Blood culture pathogens were all detected by the PCR method, Table I Comparison of 16S rRNA PCR and conventional blood culture. PCR test results
Blood culture results Positive
Negative
Positive Negative Total specimens
95 0 95
28 583 611
Total results 123 583 706
Blood culture result
PCR result
EOS
LOS
EOS
LOS
4 1 3 2 0 0 0 6 2 0 1 0 0 19
43 7 4 0 2 0 1 5 1 2 4 4 3 76
6 2 7 3 0 1 2 8 3 0 2 1 0 35
51 8 2 0 2 2 2 5 1 3 4 4 3 88
EOS: early-onset sepsis; LOS: late-onset sepsis.
whereas 28 specimens were PCR positive but culture negative. Clinical data showed that 19 patients out of 28 who had obvious clinical signs and laboratory parameters of sepsis had been diagnosed as having neonatal sepsis. In addition, the PCR method detected three strains of Streptococcus pneumoniae. The most frequently isolated pathogens in PCR positive neonates included coagulase-negative Staphylococcus species (46.3%) and E. coli (10.6%) (table II). The isolates were significantly different between EOS and late-onset sepsis (LOS, beyond 72 h after birth). In the EOS group (n = 35), the main isolates were E. coli (n = 8), Enterococcus sp. (n = 7), coagulasenegative Staphylococcus sp. (n = 6), group B Streptococcus (n = 3) and Klebsiella pneumoniae (n = 3). In the LOS group (n = 88), the most frequently isolated pathogens were coagulase-negative Staphylococcus sp. (n = 51), S. aureus (n = 8), and nonfermenting genus sp. (n = 8).
3.3. Comparison of risk factors, clinical signs and laboratory parameters Patients were divided into two groups: group P (blood culture or PCR positive) and group N (blood culture and PCR negative). Known risk factors [14] such as invasive examinations or surgery were not significantly higher in group P; neither was prolonged labor, placental abnormalities, or amniotic fluid abnormalities during birth (table III). Birth-weight < 1500 g, age 3 days, Apgar score < 8 at 5 min, gestational age (GA) < 37 weeks, premature rupture of membranes, and multiple pregnancies were significantly more frequent in the P group. Clinical signs of sepsis, such as fever in the newborn (defined as a rectal temperature > 38 8C), impetigo, ompha-
165
Archives de Pe´diatrie 2014;21:162-169
C.L. Liu et al.
Table III Comparison of the P group and N group. Basic data
P group (n = 123)
N group (n = 583)
P
Weight < 1500 g Age 3 days Risk factors Gestational age (GA) < 37 weeks Premature rupture of membrane Prolonged labor Placental abnormalities Multiple pregnancy Amniotic fluid abnormalities Invasive examination or surgery Maternal antibiotics Apgar score at 5 min < 8 Clinical signs Fever > 38 8C Icterus Irritability Feeding difficulties Impetigo Omphalitis Lung infection Laboratory parameters Heart rate > 160 beats minute Respiratory rate > 60 minute Base excess < 5 SaO2 < 88% Leukocyte count < 5 109/L Platelets < 100 109/l I/T 2 12-h CRP (mg/l)
49/123 35/123
61/583 258/583
< 0.0001 0.001
65/123 63/123 52/123 69/123 21/123 59/123 4/123 35/123 39/123
123/583 173/583 194/583 276/583 33/583 278/583 11/583 171/583 77/583
< 0.0001 < 0.0001 0.0570 0.0775 < 0.0001 0.9545 0.3400 0.8461 < 0.0001
95/123 73/123 61/123 69/123 5/123 13/123 67/123
416//583 231/583 178/583 169/583 9/583 42/583 300/583
0.1850 < 0.0001 < 0.0001 < 0.0001 0.0684 0.2058 0.5432
69/123 66/123 14/123 19/123 40/123 65/123 38/123 45.5 28.9
325/583 278/583 86/583 64/583 172/583 194/583 71/583 4.3 1.7
0.9431 0.2284 0.3301 0.1620 0.5070 < 0.0001 < 0.0001 0.00
P group: blood culture or PCR positive; N group: blood culture and PCR negative; I/T: immature/total neutrophils ratio; CRP: C-reactive protein.
litis, and lung infection did not correlate significantly in either group. Furthermore, icterus, irritability, and feeding difficulties seemed to be the only clinical signs to be trusted in our study when diagnosing sepsis in the newborn. Finally, infants in group P had a mean 12-h CRP concentration of 45.5 mg/l, far higher than in the N group (4.3 mg/l). Platelet counts (< 100 109/L) and the immature/total neutrophils ratio (I/ T; 2) differed significantly between the two groups.
3.4. Comparison of risk factors, clinical signs, laboratory parameters, and prognosis in EOS and LOS The 123 cases of proved neonatal sepsis were allocated into two groups: EOS group (n = 35) and LOS group (n = 88). Known impact factors (statistically significant factors in table III, antibiotics use and prognosis) such as weight, GA, premature rupture of membranes, irritability, length of antibiotics usage, mortality rate, and cases of purulent meningitis were significantly different between the two groups (table IV). Weight of < 1500 g and GA < 37 weeks had an obvious link with LOS; premature rupture of membranes and irritability were more
166
common in the EOS group. The average length of antibiotics usage was longer in the LOS group. The mortality rate in the EOS group (6/35) was higher than in the LOS group (2/88), and purulent meningitis was more common in the LOS group.
4. Discussion This study was designed to compare the utility of a 16S rRNA PCR assay with conventional blood culture for detecting bacteria in blood obtained from neonates suspected of having bacterial sepsis. The comparison revealed a high level of agreement between the two methodologies, with sensitivity, specificity, positive and negative predictive values of 100.0%, 95.4%, 77.2%, and 100.0%, respectively, for PCR. The high negative predictive value of the PCR assay compared with culture is indicative of the assay’s usefulness in accurately ruling out the diagnosis of bacterial sepsis in noninfected term neonates admitted to the NICU for such an evaluation [15]. In addition, blood culture pathogens were all detected by the PCR method, whereas 28 of 706 specimens were PCR positive but culture negative. Clinical data showed that 19
Comparison of 16S rRNA gene PCR and blood culture for diagnosis of neonatal sepsis
Table IV Comparison according to the early- or late-onset of sepsis. Factors
EOS group (n = 35)
LOS group (n = 88)
P
Weight < 1500 g Gestational age (GA) < 37 weeks Premature rupture of membrane Multiple pregnancy Apgar score at 5 min < 8 Platelets < 100 109/l I/T (immature/total neutrophils) 2 12-h CRP (mg/l) Icterus Irritability Feeding difficulties Length of antibiotics < 14 days Prognosis Recovery Death Purulent meningitis Pneumonia Osteomyelitis Cellulitis
12/35 16/35 24/35 9/35 21/35 29/35 25/35 42.35 18.24 17/35 26/35 13/35 24/35
67/88 63/88 25/88 15/88 37/88 73/88 66/88 47.24 22.48 35/88 41/88 27/88 41/88
< 0.0001 0.0069 < 0.0001 0.2737 0.0719 0.9897 0.6837 0.00 0.3728 0.0054 0.3442 0.0276
24/35 6/35 1/35 2/35 1/35 0/35
65/88 2/88 11/88 7/88 3/88 1/88
0.5538 < 0.0001 0.0132 0.6668 0.8763 –
EOS: early-onset sepsis; LOS: late-onset sepsis; I/T: immature/total neutrophils ratio; CRP: C-reactive protein.
patients out of 28 with obvious clinical signs and laboratory parameters of sepsis had been diagnosed as having neonatal sepsis. Moreover, the PCR method detected three strains of S. pneumoniae and may prove to be more useful for diagnosing nonculturable pathogens than blood culture is. Jordan et al. [16] also showed a higher level of agreement between the two methodologies when preincubation was performed before PCR testing. They used 200–500 ml EDTAfull blood preincubated at 37 8C for 5 h before PCR testing, and found 96% sensitivity, 99.4% specificity, and 88.9% positive and 99.8% negative predictive values. However, a drawback with this procedure is that only living bacteria able to grow in blood culture bottles will be detected. Compared with blood culture, 16S rRNA PCR may prove useful for detecting septicemia in infants whose mothers received intrapartum antibiotics and for diagnosing nonculturable or severe oxygen pathogens. It has an excellent sensitivity and a shorter turnaround time. In this study, two pairs of primers detected the same specimen. The sensitivity for the Gram-positive bacteria S. aureus and the Gram-negative organisms E. coli was 104 CFU/ml and 103 CFU/ml, respectively [3,13]. The detection limits of two pairs of primers were similar. PCR results were considered negative when no visible PCR products of correct size were found in both reactions. Results were considered inconclusive when single positive runs could not be repeated. The results should then be combined with clinical signs and laboratory parameters: If the newborn has manifest clinical signs and laboratory parameters of sepsis it is determined to be neonatal sepsis, otherwise the results are considered negative. Furthermore, UV-light irradiation was used before each pro-
cedure in order to prevent DNA contamination [17]. LOS is mainly a nosocomial infection, and thus the most common organisms differ significantly between EOS and LOS [18]. In our study, the major isolates of EOS were E. coli, followed by Enterococcus sp., coagulase-negative Staphylococcus sp., group B Streptococcus and K. pneumoniae. The most frequently found isolates in the LOS group were coagulasenegative Staphylococcus sp., S. aureus, and nonfermenting genus sp. Similar findings have been previously noted [19,20]. The World Health Organization (WHO) has published detailed criteria for the diagnosis of sepsis [21]. These criteria were easily adapted to most developing countries. In our study, the number of infants with risk factors such as weight under 1500 g, Apgar scores less than 8 at 5 min, premature rupture of membranes, GAs less than 37 weeks, and multiple pregnancies was significantly higher in the P group (blood culture or PCR positive). Icterus, irritability, and feeding difficulties were the most frequently observed clinical signs in infants with the diagnosis of sepsis. In addition, platelets count and I/T differed significantly according to the presence or not of sepsis. The absence of the other clinical criteria suggested by the WHO could be explained by the small number of patients with serious bacterial illness in our study [22]. This study also found weight under 1500 g, GA less than 37 weeks, Apgar score at 5 min less than 8, platelet counts of less than 100 109/l, an I/ T ratio of 2, and length of antibiotics use to have a significant effect on the prognosis of neonatal sepsis. CRP is the most commonly used acute-phase reactant. Its sensitivity substantially increases with serial determinations 24–48 h after the onset of symptoms. The reported sensitivity of CRP for detection of bacterial infection varies widely from
167
Archives de Pe´diatrie 2014;21:162-169
C.L. Liu et al.
47% to 100% [23,24], and depends on the chosen cutoff-point, on the definition of infection, and on the sampling time. Erdeve et al. [25] found that CRP and serum procalcitonin are probably the best single diagnostic test for the various indicators of sepsis in neonates. In this study, we considered a 12-h CRP value of more than 10 mg/l as positive. Infants with sepsis had a mean 12-h CRP value of 45.5 mg/l, far higher than infants without sepsis (4.3 mg/l). Infants with complications had a mean 12-h CRP value of 69.14 mg/l, significantly higher than the recovery group. It has been suggested that elevated CRP levels are significant in diagnosing sepsis and assessing prognosis. In this study, the relationship between known impact factors and the time of sepsis onset was analyzed. Known risk factors such as very low birth-weight (VLBW; birth-weight < 1500 g), GA, premature rupture of membranes, and multiple pregnancies were closely related with neonatal sepsis [18,19]. Our results showed that weight under 1500 g and GA less than 37 weeks had a more obvious link with LOS. Prematurity and VLBW were more common in LOS than in EOS, which was an interesting finding. Premature rupture of membranes was more common in EOS, as noted in previous studies [2,18]. Furthermore, irritability, length of antibiotic usage, mortality rate, and cases of purulent meningitis were significantly different between both groups and irritability was a more obvious sign of EOS. The mode of acquisition of late-onset neonatal infections is similar to that of nosocomial infections. Nosocomial infections in older children and adults involve the individual’s own skin, respiratory, and gastrointestinal flora, when a neonate is sterile at birth and acquires surface flora in the nursery. Normal flora confers resistance to colonization with new organisms and is an important constituent of host resistance to infection. This explains why the LOS group had longer antibiotic usage and hospitalization than the EOS group did. Although the incidence of culture-positive sepsis is low in EOS, the risk of mortality is quite high, ranging from 10 to 50% [4,5]. Our study showed that the mortality rate in the EOS group (17.1%) was significantly higher than in the LOS group (2.3%); however, purulent meningitis was more common in the LOS group.
potentially novel bacterial species. However, this technology needs to be further developed and improved. Blood culture is irreplaceable at present, since pure isolates are essential for antimicrobial drug susceptibility testing.
Disclosure of interest The authors declare that they have no conflicts of interest concerning this article. Acknowledgements This study was supported by a grant from the National 11th FiveYear Plan major scientific and technological issues (No. 2009ZX10004207).
References [1] [2] [3]
[4] [5]
[6]
[7]
[8]
[9] [10]
5. Conclusion The broad-range 16S rRNA PCR method in this study increased the sensitivity and negative predictive value of detecting bacterial DNA in newborns with signs of sepsis. The PCR method could quickly detect pathogens within 5 h and lead to shorter antibiotic courses and NICU stays. The use of a broadrange PCR followed, when necessary, by sequencing provides an assay complementary to culture for identifying microorganisms. This tool could be used in patients highly suspected of infection for whom blood culture is negative and it should allow the detection of previously unknown pathogens and
168
[11]
[12]
[13]
[14]
Mukhopadhyay S, Puopolo KM. Risk assessment in neonatal early-onset sepsis. Semin Perinatol 2012;36:408–15. Stefanovic IM. Neonatal sepsis. Biochem Med (Zagreb) 2011;21:276–81. Reier-Nilsen T, Farstad T, Nakstad B, et al. Comparison of broadrange 16S rDNA PCR and conventional blood culture for diagnosis of sepsis in the newborn: a case control study. BMC Pediatrics 2009;9:5. Klein JO. Current antibacterial therapy for neonatal sepsis and meningitis. Pediatr Infec Dis J 1990;9:783–4. Al-Taiar A, Hammoud MS, Cuiqing L, et al. Neonatal infections in China, Malaysia, Hong Kong and Thailand. Arch Dis Child Fetal Neonatal Ed 2013;98:F249–55. Wu TW, Tabangin M, Kusano R, et al. The utility of serum hepcidin as a biomarker for late-onset neonatal sepsis. J Pediatr 2013;162:67–71. Schrag SJ, Cutland CL, Zell ER, et al. Risk factors for neonatal sepsis and perinatal death among infants enrolled in the prevention of perinatal sepsis trial, Soweto, South Africa. Pediatr Infect Dis J 2012;31:821–6. Yadav AK, Wilson CG, Prasad PL, et al. Polymerase chain reaction in rapid diagnosis of neonatal sepsis. Indian Pediatr 2005;42:681–5. Arnon S, Litmanovitz I. Diagnostic tests in neonatal sepsis. Curr Opin Infect Dis 2008;21:223–7. Edmond K, Zaidi A. New approaches to preventing, diagnosing, and treating neonatal sepsis. PLoS Med 2010;7:e1000213. Ohlin A, Ba ¨ckman A, Ewald U, et al. Diagnosis of neonatal sepsis by broad-range 16S real-time polymerase chain reaction. Neonatology 2012;101:241–6. Pammi M, Flores A, Leeflang M, et al. Molecular assays in the diagnosis of neonatal sepsis: a systematic review and metaanalysis. Pediatrics 2011;128:e973–85. Xu J, Millar BC, Moore JE, et al. Employment of broad-range 16S rRNA PCR to detect aetiological agents of infection from clinical specimens in patients with acute meningitis: rapid separation of 16S rRNA PCR amplicons without the need for cloning. J Appl Microbiol 2003;94:197–206. Reinhart K, Brunkhorst FM, Bone HG, et al. Prevention, diagnosis, therapy and follow-up care of sepsis: 1st revision of S-2k guidelines of the German Sepsis Society (Deutsche SepsisGesellschaft e.V. (DSG)) and the German Interdisciplinary Asso-
Comparison of 16S rRNA gene PCR and blood culture for diagnosis of neonatal sepsis
[15]
[16]
[17] [18] [19]
ciation of Intensive Care and Emergency Medicine (Deutsche Interdisziplina ¨re Vereinigung fu ¨r Intensiv- und Notfallmedizin (DIVI)). Ger Med Sci 2010;8:Doc14. Jordan JA, Durso MB. Comparison of 16S rRNA gene PCR and BACTEC 9240 for detection of neonatal bacteremia. J Clin Microbiol 2000;8:2574–8. Jordan JA, Durso MB, Butchko AR, et al. Evaluating the nearterm infant for early b onset sepsis: progress and challenges to consider with 16S rDNA polymerase chain reaction testing. J Mol Diagn 2006;8:357–63. Kitchin PA, Szotyori Z, Fromholc C, et al. Avoidance of false positives [corrected]. Nature 1990;344:201. Mukhopadhyay S, Puopolo MK. Risk assessment in neonatal early-onset sepsis. Semin Perinatol 2012;36:408–15. Ha ¨rtel C, Faust K, Avenarius S, et al. Epidemic microclusters of blood culture proven sepsis in very low birth-weight infants: experience of the German Neonatal Network. PLoS One 2012;7:e38304.
[20]
[21]
[22]
[23]
[24]
[25]
Didier C, Streicher MP, Chognot D, et al. Late-onset neonatal infections: incidences and pathogens in the era of antenatal antibiotics. Eur J Pediatr 2012;171:681–7. The WHO Young Infants Study Group. Clinical prediction of serious bacterial infections in young infants in developing countries. Pediatr Infect Dis J 1999;18:S23–31. Vergnano S, Sharland M, Kazembe P, et al. Neonatal sepsis: an international perspective. Arch Dis Child Fetal Neonatal Ed 2005;90:F220–4. Mussap M, Noto A, Cibecchini F, et al. The importance of biomarkers in neonatology. Semin Fetal Neonatal Med 2013;18:56–64. Meem M, Modak JK, Mortuza R, et al. Biomarkers for diagnosis of neonatal infections: a systematic analysis of their potential as a point-of-care diagnostics. J Glob Health 2011;1:201–9. Erdeve O, Celik IH, Uras N, et al. CRP as a predictive of neonatal sepsis and its role in differentiating the aetiologies. Acta Paediatr 2011;100:160–1.
169
Available online at
ScienceDirect www.sciencedirect.com
Original article
Comparison of 16S rRNA gene PCR and blood culture for diagnosis of neonatal sepsis Comparaison entre l’ARNr 16S et l’he´moculture pour le diagnostic de septice´mie ne´onatale C.L. Liua, H.W. Aib, W.P. Wangc, L. Chena, H.B. Hub, T. Yec, X.H. Zhua, F. Wanga, Y.L. Liaoa, Y. Wanga, G. Oua, L. Xua, M. Suna, C. Jiana, Z.J. Chena, L. Lia, B. Zhanga, L. Tiana, B. Wanga, S. Yana, Z.Y. Suna,* a Department of clinical laboratory, Tongji hospital, Tongji medical college, Huazhong university of science and technology, 430030 Wuhan, PR China b Department of clinical laboratory, children’s Hospital, 430030 Wuhan, PR China c Department of clinical laboratory, women and children hospital, 430030 Wuhan, PR China
Summary
Re´sume´
Septicemia is a common cause of morbidity and mortality among newborns in the developing world. However, accurate clinical diagnosis of neonatal sepsis is often difficult because symptoms and signs are often nonspecific. Blood culture has been the gold standard for confirmation of the diagnosis. However, the sensitivity is low and results are usually not promptly obtained. Therefore, the diagnosis of sepsis is often based on clinical signs in association with laboratory tests such as platelets count, immature/total neutrophils ratio (I/T), and a rise in C-reactive protein (CRP). Polymerase chain reaction (PCR) methods for the detection of neonatal sepsis represent new diagnostic tools for the early identification of pathogens. Methods. During a 4-month prospective study, 16S rRNA PCR was compared with conventional blood culture for the diagnosis of neonatal bacterial sepsis. In addition, the relationship between known risk factors, clinical signs, laboratory parameters, and the diagnosis of sepsis was considered. Results. Sepsis was suspected in 706 infants from the intensive neonatal care unit. They all were included in the study. The number of positive cultures and positive PCR results were 95 (13.5%) and 123 (17.4%), respectively. Compared with blood culture, the diagnosis of bacterial sepsis by PCR revealed a 100.0% sensitivity, 95.4% specificity, 77.2% positive predictive value, and 100.0% negative predictive value. In this study, Apgar scores at 5 min, weight, icterus, irritability, feeding difficulties, gestational age (GA), premature rupture of membrane (PRM), platelets count, I/T, and a marked rise in CRP were important in establishing the diagnosis of sepsis in the newborn. In addition, weight, GA, PRM, irritability,
La septice´mie ne´onatale est la cause principale de la morbidite´ et de mortalite´ ne´onatale dans les pays en de´veloppement. Les symptoˆmes cliniques de cette maladie e´tant peu spe´cifiques, son diagnostic clinique pre´coce est assez difficile. A` l’heure actuelle, l’he´moculture est la re´fe´rence pour le diagnostic d’infection ne´onatale, mais son re´sultat est tre`s long a` obtenir et sa sensibilite´ est faible. Par conse´quent, le diagnostic de septice´mie est formule´ sur la base des symptoˆmes cliniques et des tests de laboratoire, tels que la nume´ration plaquettaire, le rapport neutrophiles immatures/total des neutrophiles (I/T), la de´tection de la prote´ine C-re´active (CRP), etc. Ces dernie`res anne´es, la me´thode de re´action par polyme´risation en chaıˆne (PCR) est devenue une me´thode de de´tection pre´coce des germes responsables d’infections bacte´riennes ne´onatales. Me´thodes. Au cours d’une e´tude prospective de quatre mois, nous avons compare´ la PCR de l’ARNr 16S a` l’he´moculture conventionnelle pour la de´tection des septice´mies ne´onatales, et recherche´ la relation entre les facteurs de risque classiques, les symptoˆmes cliniques, les indicateurs biologiques avec un diagnostic de sepsis confirme´. Re´sultats. Sept cent six nouveau-ne´s de l’unite´ ne´onatale de soins intensifs pre´sentant un risque de septice´mie ne´onatale ont e´te´ inclus. Les he´mocultures et la PCR de l’ARNr 16S e´taient positives dans respectivement 95 (13,5 %) et 123 (17,4 %) cas. Par rapport a` l’he´moculture, la sensibilite´ de la PCR e´tait de 100 %, sa spe´cificite´ de 95,4 %, la valeur positive pre´dictive de 77,2 % et la valeur ne´gative pre´dictive de 100 %. Cette e´tude a par ailleurs re´ve´le´ que le score d’Apgar a` 5 minutes, le poids, l’icte`re, l’irritabilite´, les
* Corresponding author. e-mail: [email protected] (Z.Y. Sun). 0929-693X/$ - see front matter ß 2013 Published by Elsevier Masson SAS. http://dx.doi.org/10.1016/j.arcped.2013.11.015 Archives de Pe´diatrie 2014;21:162-169
162
Comparison of 16S rRNA gene PCR and blood culture for diagnosis of neonatal sepsis
duration of antibiotic usage, mortality rate, and number of purulent meningitis cases were significantly different between early-onset sepsis and late-onset sepsis. Conclusion. 16S rRNA PCR increased the sensitivity in detecting bacterial DNA in newborns with signs of sepsis, allowed a rapid detection of the pathogens, and led to shorter antibiotic courses. However, uncertainty about the bacterial cause of sepsis was not reduced by this method. 16S rRNA PCR needs to be further developed and improved. Blood culture is currently irreplaceable, since pure isolates are essential for antimicrobial drug susceptibility testing. ß 2013 Published by Elsevier Masson SAS.
difficulte´s d’alimentation, l’aˆge gestationnel, la rupture pre´mature´e des membranes, la nume´ration plaquettaire, l’augmentation de l’I/T et de la CRP e´taient utiles pour le diagnostic de septice´mie ne´onatale. En outre, le poids, l’AG, la RPM, l’irritabilite´, le temps d’utilisation des antibiotiques, le taux de mortalite´ et le nombre de me´ningites purulentes e´taient significativement diffe´rents selon que la septice´mie e´tait d’apparition pre´coce ou tardive. Conclusion. La PCR de l’ARNr 16S pre´sente des avantages pour le diagnostic de septice´mie ne´onatale comme une haute sensibilite´, une de´tection rapide des agents pathoge`nes ou une re´duction du temps de l’antibiothe´rapie. Cependant, cette me´thode doit eˆtre de´veloppe´e et perfectionne´e. L’he´moculture reste irremplac¸able afin d’isoler les souches responsables pour tester leur sensibilite´ aux antibiotiques et orienter la prescription en clinique. ß 2009 Publie´ par Elsevier Masson SAS.
1. Introduction
2. Materials and methods
A large proportion of infants admitted to neonatal intensive care units (NICUs) present signs of sepsis; however, accurate clinical diagnosis of neonatal sepsis is difficult, because symptoms and signs are often nonspecific [1–3]. Especially in early-onset sepsis (EOS, within 72 h after birth), the incidence of culture-positive sepsis is low (1–8 cases/1000 livebirths), and the risk of mortality is high, ranging from 10 to 50% [4,5]. Alterations in laboratory test parameters provide evidence for the diagnosis of neonatal sepsis. Because the sensitivity of each laboratory test is quite low [6,7], the search for a reliable testing method has continued. Blood culture is the gold standard for confirmation of the diagnosis; however, blood culturing techniques may have unacceptably low sensitivities, and it often takes 48–72 h to obtain a preliminary positive result [8]. Infants with ‘‘risk factors’’ for neonatal sepsis are thus treated with broad-spectrum antibiotics and require prolonged hospitalization. Recently, C-reactive protein (CRP), interleukin (IL)-6, and procalcitonin (PCT) have been shown to be useful markers of sepsis in many studies [9,10]. 16S rRNA gene polymerase chain reaction (PCR) and sequencing has been identified as a method for defining an organism as a bacterium. This method may prove useful for detecting septicemia in infants whose mothers received intrapartum antibiotics and for diagnosing nonculturable pathogens. It has an excellent sensitivity and a shorter turnaround time than the conventional blood culture technique [11,12]. The aim of this study was to evaluate the molecular diagnosis of neonatal sepsis using PCR amplification of 16S rRNA in comparison with risk factors or clinical evidence of sepsis and to understand the pathogenic spectrum of neonatal septicemia in the region. In addition, we also analyzed the relationship between known risk factors, clinical signs, and laboratory parameters in suspected neonatal sepsis and in the diagnosis of sepsis.
2.1. Patients The study was carried out from September 1, 2011, to December 31, 2011, at the NICU of the Women and Children Hospital, the Children’s Hospital, and Tongji Hospital in Hubei Province, which are all tertiary care centers in China. All infants who had blood samples drawn for concomitant culture, a complete blood count (CBC), and a CRP assay were eligible for inclusion in the study. At the same time, an additional 0.5–1 ml EDTAblood sample was collected from the infants for the purposes of 16S rRNA gene PCR amplification. Sample rejection or exclusion criteria for PCR included blood volumes of less than 500 ml, grossly hemolyzed or clotted blood specimens, and inadequate or missing clinical data.
2.2. Blood culture processing We used an automated continuous-monitoring blood culture system, BACTEC 9240 (Becton Dickinson, Sparks, MD, USA) or BacT/Alert 3D (Biome´rieux, Durham, NC, USA). The pediatricsample-sized, resin-containing blood culture bottles (Peds Plus/F; Becton Dickinson) were sent from the NICU prefilled with blood collected in a syringe from infants at their besides. A minimum of 1 ml of whole-blood was added per blood culture bottle. The bottles were incubated immediately upon receipt in the microbiology laboratory in accordance with the manufacturer’s recommendation.
2.3. DNA extraction The EDTA-blood samples for PCR were blinded and stored at – 70 8C before analysis. Samples (500 ml) were extracted with the blood gDNA mini-kit (Biomiga, USA), supplemented with lytic enzyme (Puregene DNA extraction kit for Gram-positive bacteria; Qiagen, Germany). All DNA isolation procedures were carried out in a Class-II Biological Safety Cabinet
163
Archives de Pe´diatrie 2014;21:162-169
C.L. Liu et al.
(MicroFlow, Weston-super-Mare, UK) in order to minimize contamination and the possibility of false positive results. For each batch of extractions, a negative extraction control containing all reagents minus blood was performed, as well as positive controls with Escherichia coli ATCC 25922 and Staphylococcus aureus ATCC 25923.
2.4. PCR amplification and product detection All DNA extracts were examined by amplification of the 16S rRNA gene using two pairs of universal primers: PSL(f): 50 -AGG ATT AGA TAC CCT GGT AGT CCA-3, UN1(r): 50 -TRR YKT GAC GKG CGG TGT GTA C-30 ; and P11P(f): 50 -GAG GAA GGT GGG GAT GAC GT-30 P13P(r): 50 -AGG CCC GGG AAC GTA TTC AC-30 [13]. The primers provide PCR products of 630 bp and 216 bp, respectively. Each PCR reaction (50 ml) consisted of 1 Taq buffer with (NH4)2SO4, 2.5 U Taq polymerase, 2.5 mM MgCl2, 0.5 mMdNTP, 2-ml template DNA, and ddH2O (Applied MBI Fermentas, Canada). Cycling conditions included a 5-min denaturing step at 94 8C followed by 30 cycles of 1 min at 94 8C, 1 min at 55 8C, and 1 min at 72 8C; a final extension cycle of 72 8C for 10 min was then carried out. Two pairs of primer had similar PCR system and reaction conditions. Sterile water, E. coli ATCC 25922, and S. aureus ATCC 25923 were used as the negative and positive controls, respectively. Separate dedicated pipetting devices with filter-sealed tips were used for each procedure, and UVlight irradiation was used before each procedure to prevent DNA cross-contamination. Positive PCR products were directly sequence analyzed (Invitrogen, USA). The results were analyzed by PCR using two pairs of primer P11/P13, PSL/UNI. PCR results were considered positive when visible PCR products of the correct size were found simultaneously in two reactions. PCR results were considered negative when no visible PCR products of correct size were found. Results were considered inconclusive when single positive runs could not be repeated. The results should then be combined with clinical signs and laboratory parameters: if the newborn has manifest clinical signs and laboratory parameters of sepsis it is determined to be neonatal sepsis, otherwise the results are considered negative.
Figure 1. Specificity of 16S rRNA PCR. 1: marker; 2: Escherichia coli ATCC 25922; 3: Staphylococcus aureus ATCC 25923; 4: negative qc 1; 5: negative specimen; 6: positive specimen.
2.5. PCR assay sensitivity The sensitivity of the 16S rRNA PCR assay was determined using whole-blood samples spiked with known amounts of viable organisms, not purified bacterial DNA extracts. To this end, known numbers of CFU of control strains, E. coli ATCC 25922 or S. aureus ATCC 25923, were added to 0.5-ml volumes of wholeblood containing EDTA. The spiked whole-blood samples were prepared and amplified under exactly the same conditions as those described for the clinical blood specimens. Using this approach, the sensitivity for the Gram-positive bacteria S. aureus and the Gram-negative organisms E. coli was 104 CFU/ml and 103 CFU/ml, respectively. The detection limits of two pairs of primer were similar (figs. 1–3).
164
Figure 2. Sensitivity of 16S rRNA PCR for Escherichia coli. 1: marker; 2: 7 6 5 4 3 10 CFU/ml; 3: 10 CFU/ml; 4: 10 CFU/ml; 5: 10 CFU/ml; 6: 10 CFU/ml; 7: 2 10 CFU/ml.
2.6. Data analysis Study patients were allocated into two groups: P (blood culture or PCR or both positive) and N (blood culture and PCR negative). Risk factors, clinical signs, and laboratory
Comparison of 16S rRNA gene PCR and blood culture for diagnosis of neonatal sepsis
Table II Microbiological results of 16S rRNA and conventional blood culture. Bacteria
Coagulase-negative Staphylococcus sp. Staphylococcus aureus Enterococcus sp. Group B Streptococcus Streptococcus viridans Streptococcus pneumoniae Other Gram-positive cocci Escherichia coli Klebsiella pneumoniae Enterobacter cloacae Pseudomonas sp. Other nonfermenting genus sp. Other Gram-negative bacilli Total Figure 3. Sensitivity of 16S rRNA PCR for Staphylococcus aureus. 1: marker; 2: 7 6 5 4 3 10 CFU/ml; 3: 10 CFU/ml; 4: 10 CFU/ml; 5: 10 CFU/ml; 6: 10 CFU/ml; 7: 102CFU/ml.
parameters were compared between the two groups. A P level of < 0.05 was considered statistically significant. The statistical program SPSS V13.0W was used for data analysis.
3. Results 3.1. Clinical data During the study period, 706 neonates were admitted (432 boys) to the NICU: 332 cases in the Children’s Hospital, 335 cases in the Women and Children Hospital, and 39 cases in Tongji Hospital. The mean birth-weight was 2,431 504 g. The average age was 7.8 days, ranging from 4 h to 28 days.
3.2. Comparison of 16S rRNA PCR and blood culture The positivity rates of culture and PCR were 13.5% (95/706) and 17.4% (123/706), respectively. Compared with blood culture, the diagnosis of bacterial sepsis by PCR revealed a 100.0% sensitivity, 95.4% specificity, 77.2% positive predictive value, and 100.0% negative predictive value (table I). Blood culture pathogens were all detected by the PCR method, Table I Comparison of 16S rRNA PCR and conventional blood culture. PCR test results
Blood culture results Positive
Negative
Positive Negative Total specimens
95 0 95
28 583 611
Total results 123 583 706
Blood culture result
PCR result
EOS
LOS
EOS
LOS
4 1 3 2 0 0 0 6 2 0 1 0 0 19
43 7 4 0 2 0 1 5 1 2 4 4 3 76
6 2 7 3 0 1 2 8 3 0 2 1 0 35
51 8 2 0 2 2 2 5 1 3 4 4 3 88
EOS: early-onset sepsis; LOS: late-onset sepsis.
whereas 28 specimens were PCR positive but culture negative. Clinical data showed that 19 patients out of 28 who had obvious clinical signs and laboratory parameters of sepsis had been diagnosed as having neonatal sepsis. In addition, the PCR method detected three strains of Streptococcus pneumoniae. The most frequently isolated pathogens in PCR positive neonates included coagulase-negative Staphylococcus species (46.3%) and E. coli (10.6%) (table II). The isolates were significantly different between EOS and late-onset sepsis (LOS, beyond 72 h after birth). In the EOS group (n = 35), the main isolates were E. coli (n = 8), Enterococcus sp. (n = 7), coagulasenegative Staphylococcus sp. (n = 6), group B Streptococcus (n = 3) and Klebsiella pneumoniae (n = 3). In the LOS group (n = 88), the most frequently isolated pathogens were coagulase-negative Staphylococcus sp. (n = 51), S. aureus (n = 8), and nonfermenting genus sp. (n = 8).
3.3. Comparison of risk factors, clinical signs and laboratory parameters Patients were divided into two groups: group P (blood culture or PCR positive) and group N (blood culture and PCR negative). Known risk factors [14] such as invasive examinations or surgery were not significantly higher in group P; neither was prolonged labor, placental abnormalities, or amniotic fluid abnormalities during birth (table III). Birth-weight < 1500 g, age 3 days, Apgar score < 8 at 5 min, gestational age (GA) < 37 weeks, premature rupture of membranes, and multiple pregnancies were significantly more frequent in the P group. Clinical signs of sepsis, such as fever in the newborn (defined as a rectal temperature > 38 8C), impetigo, ompha-
165
Archives de Pe´diatrie 2014;21:162-169
C.L. Liu et al.
Table III Comparison of the P group and N group. Basic data
P group (n = 123)
N group (n = 583)
P
Weight < 1500 g Age 3 days Risk factors Gestational age (GA) < 37 weeks Premature rupture of membrane Prolonged labor Placental abnormalities Multiple pregnancy Amniotic fluid abnormalities Invasive examination or surgery Maternal antibiotics Apgar score at 5 min < 8 Clinical signs Fever > 38 8C Icterus Irritability Feeding difficulties Impetigo Omphalitis Lung infection Laboratory parameters Heart rate > 160 beats minute Respiratory rate > 60 minute Base excess < 5 SaO2 < 88% Leukocyte count < 5 109/L Platelets < 100 109/l I/T 2 12-h CRP (mg/l)
49/123 35/123
61/583 258/583
< 0.0001 0.001
65/123 63/123 52/123 69/123 21/123 59/123 4/123 35/123 39/123
123/583 173/583 194/583 276/583 33/583 278/583 11/583 171/583 77/583
< 0.0001 < 0.0001 0.0570 0.0775 < 0.0001 0.9545 0.3400 0.8461 < 0.0001
95/123 73/123 61/123 69/123 5/123 13/123 67/123
416//583 231/583 178/583 169/583 9/583 42/583 300/583
0.1850 < 0.0001 < 0.0001 < 0.0001 0.0684 0.2058 0.5432
69/123 66/123 14/123 19/123 40/123 65/123 38/123 45.5 28.9
325/583 278/583 86/583 64/583 172/583 194/583 71/583 4.3 1.7
0.9431 0.2284 0.3301 0.1620 0.5070 < 0.0001 < 0.0001 0.00
P group: blood culture or PCR positive; N group: blood culture and PCR negative; I/T: immature/total neutrophils ratio; CRP: C-reactive protein.
litis, and lung infection did not correlate significantly in either group. Furthermore, icterus, irritability, and feeding difficulties seemed to be the only clinical signs to be trusted in our study when diagnosing sepsis in the newborn. Finally, infants in group P had a mean 12-h CRP concentration of 45.5 mg/l, far higher than in the N group (4.3 mg/l). Platelet counts (< 100 109/L) and the immature/total neutrophils ratio (I/ T; 2) differed significantly between the two groups.
3.4. Comparison of risk factors, clinical signs, laboratory parameters, and prognosis in EOS and LOS The 123 cases of proved neonatal sepsis were allocated into two groups: EOS group (n = 35) and LOS group (n = 88). Known impact factors (statistically significant factors in table III, antibiotics use and prognosis) such as weight, GA, premature rupture of membranes, irritability, length of antibiotics usage, mortality rate, and cases of purulent meningitis were significantly different between the two groups (table IV). Weight of < 1500 g and GA < 37 weeks had an obvious link with LOS; premature rupture of membranes and irritability were more
166
common in the EOS group. The average length of antibiotics usage was longer in the LOS group. The mortality rate in the EOS group (6/35) was higher than in the LOS group (2/88), and purulent meningitis was more common in the LOS group.
4. Discussion This study was designed to compare the utility of a 16S rRNA PCR assay with conventional blood culture for detecting bacteria in blood obtained from neonates suspected of having bacterial sepsis. The comparison revealed a high level of agreement between the two methodologies, with sensitivity, specificity, positive and negative predictive values of 100.0%, 95.4%, 77.2%, and 100.0%, respectively, for PCR. The high negative predictive value of the PCR assay compared with culture is indicative of the assay’s usefulness in accurately ruling out the diagnosis of bacterial sepsis in noninfected term neonates admitted to the NICU for such an evaluation [15]. In addition, blood culture pathogens were all detected by the PCR method, whereas 28 of 706 specimens were PCR positive but culture negative. Clinical data showed that 19
Comparison of 16S rRNA gene PCR and blood culture for diagnosis of neonatal sepsis
Table IV Comparison according to the early- or late-onset of sepsis. Factors
EOS group (n = 35)
LOS group (n = 88)
P
Weight < 1500 g Gestational age (GA) < 37 weeks Premature rupture of membrane Multiple pregnancy Apgar score at 5 min < 8 Platelets < 100 109/l I/T (immature/total neutrophils) 2 12-h CRP (mg/l) Icterus Irritability Feeding difficulties Length of antibiotics < 14 days Prognosis Recovery Death Purulent meningitis Pneumonia Osteomyelitis Cellulitis
12/35 16/35 24/35 9/35 21/35 29/35 25/35 42.35 18.24 17/35 26/35 13/35 24/35
67/88 63/88 25/88 15/88 37/88 73/88 66/88 47.24 22.48 35/88 41/88 27/88 41/88
< 0.0001 0.0069 < 0.0001 0.2737 0.0719 0.9897 0.6837 0.00 0.3728 0.0054 0.3442 0.0276
24/35 6/35 1/35 2/35 1/35 0/35
65/88 2/88 11/88 7/88 3/88 1/88
0.5538 < 0.0001 0.0132 0.6668 0.8763 –
EOS: early-onset sepsis; LOS: late-onset sepsis; I/T: immature/total neutrophils ratio; CRP: C-reactive protein.
patients out of 28 with obvious clinical signs and laboratory parameters of sepsis had been diagnosed as having neonatal sepsis. Moreover, the PCR method detected three strains of S. pneumoniae and may prove to be more useful for diagnosing nonculturable pathogens than blood culture is. Jordan et al. [16] also showed a higher level of agreement between the two methodologies when preincubation was performed before PCR testing. They used 200–500 ml EDTAfull blood preincubated at 37 8C for 5 h before PCR testing, and found 96% sensitivity, 99.4% specificity, and 88.9% positive and 99.8% negative predictive values. However, a drawback with this procedure is that only living bacteria able to grow in blood culture bottles will be detected. Compared with blood culture, 16S rRNA PCR may prove useful for detecting septicemia in infants whose mothers received intrapartum antibiotics and for diagnosing nonculturable or severe oxygen pathogens. It has an excellent sensitivity and a shorter turnaround time. In this study, two pairs of primers detected the same specimen. The sensitivity for the Gram-positive bacteria S. aureus and the Gram-negative organisms E. coli was 104 CFU/ml and 103 CFU/ml, respectively [3,13]. The detection limits of two pairs of primers were similar. PCR results were considered negative when no visible PCR products of correct size were found in both reactions. Results were considered inconclusive when single positive runs could not be repeated. The results should then be combined with clinical signs and laboratory parameters: If the newborn has manifest clinical signs and laboratory parameters of sepsis it is determined to be neonatal sepsis, otherwise the results are considered negative. Furthermore, UV-light irradiation was used before each pro-
cedure in order to prevent DNA contamination [17]. LOS is mainly a nosocomial infection, and thus the most common organisms differ significantly between EOS and LOS [18]. In our study, the major isolates of EOS were E. coli, followed by Enterococcus sp., coagulase-negative Staphylococcus sp., group B Streptococcus and K. pneumoniae. The most frequently found isolates in the LOS group were coagulasenegative Staphylococcus sp., S. aureus, and nonfermenting genus sp. Similar findings have been previously noted [19,20]. The World Health Organization (WHO) has published detailed criteria for the diagnosis of sepsis [21]. These criteria were easily adapted to most developing countries. In our study, the number of infants with risk factors such as weight under 1500 g, Apgar scores less than 8 at 5 min, premature rupture of membranes, GAs less than 37 weeks, and multiple pregnancies was significantly higher in the P group (blood culture or PCR positive). Icterus, irritability, and feeding difficulties were the most frequently observed clinical signs in infants with the diagnosis of sepsis. In addition, platelets count and I/T differed significantly according to the presence or not of sepsis. The absence of the other clinical criteria suggested by the WHO could be explained by the small number of patients with serious bacterial illness in our study [22]. This study also found weight under 1500 g, GA less than 37 weeks, Apgar score at 5 min less than 8, platelet counts of less than 100 109/l, an I/ T ratio of 2, and length of antibiotics use to have a significant effect on the prognosis of neonatal sepsis. CRP is the most commonly used acute-phase reactant. Its sensitivity substantially increases with serial determinations 24–48 h after the onset of symptoms. The reported sensitivity of CRP for detection of bacterial infection varies widely from
167
Archives de Pe´diatrie 2014;21:162-169
C.L. Liu et al.
47% to 100% [23,24], and depends on the chosen cutoff-point, on the definition of infection, and on the sampling time. Erdeve et al. [25] found that CRP and serum procalcitonin are probably the best single diagnostic test for the various indicators of sepsis in neonates. In this study, we considered a 12-h CRP value of more than 10 mg/l as positive. Infants with sepsis had a mean 12-h CRP value of 45.5 mg/l, far higher than infants without sepsis (4.3 mg/l). Infants with complications had a mean 12-h CRP value of 69.14 mg/l, significantly higher than the recovery group. It has been suggested that elevated CRP levels are significant in diagnosing sepsis and assessing prognosis. In this study, the relationship between known impact factors and the time of sepsis onset was analyzed. Known risk factors such as very low birth-weight (VLBW; birth-weight < 1500 g), GA, premature rupture of membranes, and multiple pregnancies were closely related with neonatal sepsis [18,19]. Our results showed that weight under 1500 g and GA less than 37 weeks had a more obvious link with LOS. Prematurity and VLBW were more common in LOS than in EOS, which was an interesting finding. Premature rupture of membranes was more common in EOS, as noted in previous studies [2,18]. Furthermore, irritability, length of antibiotic usage, mortality rate, and cases of purulent meningitis were significantly different between both groups and irritability was a more obvious sign of EOS. The mode of acquisition of late-onset neonatal infections is similar to that of nosocomial infections. Nosocomial infections in older children and adults involve the individual’s own skin, respiratory, and gastrointestinal flora, when a neonate is sterile at birth and acquires surface flora in the nursery. Normal flora confers resistance to colonization with new organisms and is an important constituent of host resistance to infection. This explains why the LOS group had longer antibiotic usage and hospitalization than the EOS group did. Although the incidence of culture-positive sepsis is low in EOS, the risk of mortality is quite high, ranging from 10 to 50% [4,5]. Our study showed that the mortality rate in the EOS group (17.1%) was significantly higher than in the LOS group (2.3%); however, purulent meningitis was more common in the LOS group.
potentially novel bacterial species. However, this technology needs to be further developed and improved. Blood culture is irreplaceable at present, since pure isolates are essential for antimicrobial drug susceptibility testing.
Disclosure of interest The authors declare that they have no conflicts of interest concerning this article. Acknowledgements This study was supported by a grant from the National 11th FiveYear Plan major scientific and technological issues (No. 2009ZX10004207).
References [1] [2] [3]
[4] [5]
[6]
[7]
[8]
[9] [10]
5. Conclusion The broad-range 16S rRNA PCR method in this study increased the sensitivity and negative predictive value of detecting bacterial DNA in newborns with signs of sepsis. The PCR method could quickly detect pathogens within 5 h and lead to shorter antibiotic courses and NICU stays. The use of a broadrange PCR followed, when necessary, by sequencing provides an assay complementary to culture for identifying microorganisms. This tool could be used in patients highly suspected of infection for whom blood culture is negative and it should allow the detection of previously unknown pathogens and
168
[11]
[12]
[13]
[14]
Mukhopadhyay S, Puopolo KM. Risk assessment in neonatal early-onset sepsis. Semin Perinatol 2012;36:408–15. Stefanovic IM. Neonatal sepsis. Biochem Med (Zagreb) 2011;21:276–81. Reier-Nilsen T, Farstad T, Nakstad B, et al. Comparison of broadrange 16S rDNA PCR and conventional blood culture for diagnosis of sepsis in the newborn: a case control study. BMC Pediatrics 2009;9:5. Klein JO. Current antibacterial therapy for neonatal sepsis and meningitis. Pediatr Infec Dis J 1990;9:783–4. Al-Taiar A, Hammoud MS, Cuiqing L, et al. Neonatal infections in China, Malaysia, Hong Kong and Thailand. Arch Dis Child Fetal Neonatal Ed 2013;98:F249–55. Wu TW, Tabangin M, Kusano R, et al. The utility of serum hepcidin as a biomarker for late-onset neonatal sepsis. J Pediatr 2013;162:67–71. Schrag SJ, Cutland CL, Zell ER, et al. Risk factors for neonatal sepsis and perinatal death among infants enrolled in the prevention of perinatal sepsis trial, Soweto, South Africa. Pediatr Infect Dis J 2012;31:821–6. Yadav AK, Wilson CG, Prasad PL, et al. Polymerase chain reaction in rapid diagnosis of neonatal sepsis. Indian Pediatr 2005;42:681–5. Arnon S, Litmanovitz I. Diagnostic tests in neonatal sepsis. Curr Opin Infect Dis 2008;21:223–7. Edmond K, Zaidi A. New approaches to preventing, diagnosing, and treating neonatal sepsis. PLoS Med 2010;7:e1000213. Ohlin A, Ba ¨ckman A, Ewald U, et al. Diagnosis of neonatal sepsis by broad-range 16S real-time polymerase chain reaction. Neonatology 2012;101:241–6. Pammi M, Flores A, Leeflang M, et al. Molecular assays in the diagnosis of neonatal sepsis: a systematic review and metaanalysis. Pediatrics 2011;128:e973–85. Xu J, Millar BC, Moore JE, et al. Employment of broad-range 16S rRNA PCR to detect aetiological agents of infection from clinical specimens in patients with acute meningitis: rapid separation of 16S rRNA PCR amplicons without the need for cloning. J Appl Microbiol 2003;94:197–206. Reinhart K, Brunkhorst FM, Bone HG, et al. Prevention, diagnosis, therapy and follow-up care of sepsis: 1st revision of S-2k guidelines of the German Sepsis Society (Deutsche SepsisGesellschaft e.V. (DSG)) and the German Interdisciplinary Asso-
Comparison of 16S rRNA gene PCR and blood culture for diagnosis of neonatal sepsis
[15]
[16]
[17] [18] [19]
ciation of Intensive Care and Emergency Medicine (Deutsche Interdisziplina ¨re Vereinigung fu ¨r Intensiv- und Notfallmedizin (DIVI)). Ger Med Sci 2010;8:Doc14. Jordan JA, Durso MB. Comparison of 16S rRNA gene PCR and BACTEC 9240 for detection of neonatal bacteremia. J Clin Microbiol 2000;8:2574–8. Jordan JA, Durso MB, Butchko AR, et al. Evaluating the nearterm infant for early b onset sepsis: progress and challenges to consider with 16S rDNA polymerase chain reaction testing. J Mol Diagn 2006;8:357–63. Kitchin PA, Szotyori Z, Fromholc C, et al. Avoidance of false positives [corrected]. Nature 1990;344:201. Mukhopadhyay S, Puopolo MK. Risk assessment in neonatal early-onset sepsis. Semin Perinatol 2012;36:408–15. Ha ¨rtel C, Faust K, Avenarius S, et al. Epidemic microclusters of blood culture proven sepsis in very low birth-weight infants: experience of the German Neonatal Network. PLoS One 2012;7:e38304.
[20]
[21]
[22]
[23]
[24]
[25]
Didier C, Streicher MP, Chognot D, et al. Late-onset neonatal infections: incidences and pathogens in the era of antenatal antibiotics. Eur J Pediatr 2012;171:681–7. The WHO Young Infants Study Group. Clinical prediction of serious bacterial infections in young infants in developing countries. Pediatr Infect Dis J 1999;18:S23–31. Vergnano S, Sharland M, Kazembe P, et al. Neonatal sepsis: an international perspective. Arch Dis Child Fetal Neonatal Ed 2005;90:F220–4. Mussap M, Noto A, Cibecchini F, et al. The importance of biomarkers in neonatology. Semin Fetal Neonatal Med 2013;18:56–64. Meem M, Modak JK, Mortuza R, et al. Biomarkers for diagnosis of neonatal infections: a systematic analysis of their potential as a point-of-care diagnostics. J Glob Health 2011;1:201–9. Erdeve O, Celik IH, Uras N, et al. CRP as a predictive of neonatal sepsis and its role in differentiating the aetiologies. Acta Paediatr 2011;100:160–1.
169
