| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Technical Briefs |
1 Department of Pediatrics and Obstetrics and Gynecology, Giannina Gaslini Children’s University Hospital, Genoa, Italy2 Department of Neonatology, Children’s Memorial Health Institute, Warsaw, Poland3 Institute of Anatomy and Cell Biology, Catholic University, Rome, Italy
aaddress correspondence to this author at: Institute of Anatomy and Cell Biology, Catholic University, Largo Francesco Vito, 1, I-00168 Rome, Italy; fax 39-06-30154813, e-mail fabrizio.michetti{at}rm.unicatt.it
Perinatal infections such as bacterial meningitis (BM) are one of the major factors associated with perinatal brain damage (1)(2)(3)(4). Despite accurate monitoring, the early stages of meningitis are crucial because brain damage may occur at a subclinical stage when ultrasound assessment is still silent (5)(6). Laboratory assessment is based on chemical analysis of cerebrospinal fluid (CSF) and the detection of bacteria, and the possibility of detecting cases at risk of brain damage is to date limited. The measurement of brain constituents able to diagnose subclinical lesions at this stage could therefore be useful.
S100B is a calcium-binding protein primarily present in nervous tissue (7)(8)(9). Increased S100B in biological fluids has been shown to be a marker of brain damage both in adults and during the antenatal and postnatal periods (10)(11)(12)(13)(14)(15)(16).
The present case–control study is aimed at investigating whether the measurement of S100B in CSF could also be useful in infants with BM for the early detection of cases at risk of encephalitis.
Samples of CSF were collected from infants consecutively admitted between April 1998 and June 2000 to our tertiary referral center for intensive care for infectious diseases. For the present study we identified from our database 44 patients with BM and matched them for gestational age at sampling with 44 patients without BM (1 BM case vs 1 control). We retrieved clinical, laboratory, and routine CSF test data and CSF S100B concentrations.
Eligibility criteria for infants with BM were as follows: clinical (respiratory distress, lethargy, presence/absence of minor/major neurologic symptoms, feeding and abdominal distension problems, temperature instability or increases, unexplained recurrent hypoglycemia, poor vascular perfusion) and laboratory signs of septicemia with altered CSF results (leukocyte count, protein, glucose, visible bacteria) (17). Causative bacteria were gram-positive cocci (Streptococcus agalactiae) in 19 cases, unidentified gram-positive cocci in 12 cases, and gram-negative rods (Haemophilus influenzae or Escherichia coli) in 13.
For ethical reasons, the healthy group consisted of infants in whom CSF samples had been collected to investigate confirmed or probable meningitis. Infants were included in the control group when CSF findings of meningeal inflammation such as CSF leukocyte count and protein and glucose concentrations were normal and CSF culture or bacterial antigen test results were negative (17). An additional criterion for admission to the control group was that the ultrasound patterns were negative for encephalitis and for central nervous system diseases. Exclusion criteria were fetal or neonatal central nervous system malformations, chromosomal abnormalities, perinatal asphyxia, and congenital heart disease.
In the control group septicemia was identified (15 of 44 with unidentified gram-positive cocci, 22 of 44 with Staphylococcus epidermidis, and 7 of 44 with Staphylococcus species).
All recruited infants were delivered at term without apparent perinatal complications and/or clinical history of neurologic abnormalities or comorbidities between birth and admission to our neonatal intensive care units.
Cerebral ultrasound (US) was assessed at the time of CSF sampling, at 72 h after admission, and on discharge from hospital. Cerebral computerized tomography (CT) was performed, in the infants with BM plus encephalitis (BME), for the detection of the presence and extension of encephalitis when US was already suggestive of brain lesion.
Local Ethics Committees approved the study protocol, and the parents of the infants gave signed and informed consent that an aliquot of the CSF obtained at the time of CSF collection for clinical indications could be used for research purposes.
At sampling, 100 µL of CSF for S100B measurement was obtained by lumbar puncture for study. Samples were immediately centrifuged at 900g for 10 min, and the supernatants were stored at –70 °C.
The S100B protein concentration was measured by use of a commercially available immunoluminometric assay (Lia-mat Sangtec 100; AB-Sangtec Medical) (10)(18)(19). Each measurement was performed in duplicate according to the manufacturer’s recommendations. The limit of detection of the assay was 0.02 µg/L, the intraassay imprecision (CV) was <5%, and the interassay CV was <10%. S100B measurements were performed by a single expert who did not know the infants’ clinical conditions.
S100B concentrations are reported as medians and interquartile ranges. Statistical analyses were performed with the Kruskal–Wallis one-way ANOVA for comparisons between groups and the Mann–Whitney U-test when data did not follow a gaussian distribution. The Fisher exact test was used to compare the incidences of abnormal cerebral US and CT results. The correlation between the S100B concentrations in CSF and routine CSF patterns was assessed by linear regression analysis. Multiple logistic regression analysis was performed with the occurrence of BME as the independent variable to analyze the influence of various laboratory variables (S100B and CSF variables). S100B in CSF as a diagnostic test for BME was assessed by ROC curves. A P value <0.05 was considered significant.
The clinical findings recorded at sampling are shown in Table 1
. As expected, CSF patterns were higher in infants with BM than in controls (P <0.05). US patterns recorded on admission to neonatal intensive care units were negative in all infants examined (P >0.05). At 72 h after admission, the US patterns showed the presence of encephalitis in 18 of 44 infants with BM. The abnormalities were ventriculitis (n = 9) and abnormal parenchymal echogenicities (periventricular, n = 5; focal, n = 4). Identical patterns were obtained in the CT analyses. According to the presence/absence of encephalitis, the BM group was subdivided into the BMO group (BM only; n = 26) and the BME group (with encephalitis; n = 18). Routine CSF patterns did not differ significantly between the two groups (P >0.05).
|
The concentrations of S100B in CSF were significantly higher in the total BM group (median, 1.34 µg/L; 25th percentile, 0.84 µg/L; 75th percentile, 1.78 µg/L) than in the controls (median, 0.16 µg/L; 25th percentile, 0.11 µg/L; 75th percentile, 0.33 µg/L; P <0.001). CSF S100B concentrations were significantly higher in the infants of the BME group than in the BMO group and the controls (P <0.001; Fig. 1
). The difference between the BMO group and controls was also significant (P <0.01).
|
CSF S100B correlated with CSF variables such as leukocyte count (r = 0.44; P <0.001), glucose (r = 0.25; P <0.01), and protein (r = 0.29; P <0.01).
Multiple logistic regression analysis showed a positive correlation only between CSF S100B and the occurrence of BME (P <0.001; odds ratio, 7.0). ROC curve analysis showed that higher S100B values were diagnostic for early BME detection, with a S100B cutoff of 1.0 µg/L, a sensitivity of 91% (95% confidence interval, 71–98.6%), a specificity of 82% (95% confidence interval, 70–91.4%), and an area under the curve of 0.918.
The present findings suggest the potential usefulness of S100B measurements in CSF as a possible tool for the early detection of infants at risk of brain lesions from BM. BM represents a major cause of perinatal mortality and morbidity, and approximately one-third of affected infants develop neurologic sequelae (20).
Increased S100B has been reported in CSF, cord blood, amniotic fluid, and urine in preterm/term infants with cerebral bleeding (10)(14)(15)(16), in high-risk fetuses (13)(21), in asphyxiated newborns (16)(22), and in children subjected to extracorporeal membrane oxygenation (23) and cardiopulmonary by-pass (12). In these conditions, S100B concentrations have been shown to detect individuals at high risk when routine monitoring procedures are silent. S100B cutoff values in urine and blood differ from that in CSF. The discrepancy can reasonably be attributed to the difference in biological fluids and differences in the perinatal disease and study population investigated.
Increased S100B concentrations in CSF are a reasonably direct manifestation of damage to the central nervous system, which is known to contain protein at this stage.
Higher S100B concentrations may reflect a bacterial infection because such infections are known to induce a cascade of events, including activation of proinflammatory cytokines, which in the brain trigger exaggerated activation of glial cells, which produce S100B (24). High extracellular concentrations of S100B have been shown to be neurotoxic, leading to apoptosis and neuronal death via a nitric oxide-mediated pathway (10)(25). We think it is possible that at least part of the S100B measured in the CSF of infants with encephalitis derives from this process and participates in the pathologic events accompanying parenchymal damage.
Standard monitoring procedures performed when infants were admitted to our neonatal intensive care units were unable to indicate which infants would develop encephalitis. At this stage, S100B was significantly higher in the infants who later developed encephalitis than in those who did not develop encephalitis or in controls. S100B was also higher in BM cases without encephalitis than in controls. One explanation involves the possibility that some of the S100B in the CSF of infants without apparent brain damage may derive from changes in brain–blood barrier permeability (a common finding in meningitis). Another possible explanation is subclinical brain damage not detectable by standard diagnostic procedures.
The present findings support the notion that assessment of S100B in CSF may provide additional information to physicians. Measurement of a brain constituent could be particularly useful for monitoring the effectiveness of therapeutic strategies. It is even possible that S100B assessment in other biological fluids may help in the early diagnosis of BM and encephalitis in infants with infectious diseases. The availability of longitudinal S100B measurements [half-life of 
1 h (26)] in other biological fluids (e.g., peripheral blood and urine) has been already reported (14)(15)(16).
The correlation between S100B and CSF patterns is intriguing. The correlation between S100B and polymorphonuclear leukocytes, a manifestation of the severity of an infection, offers additional support to the notion that a more severe of infection is associated with a higher risk of neurologic sequelae (20). This concept is supported by the correlation between S100B and other CSF variables.
In conclusion, the present investigation provides a new perspective for the clinical study of S100B, with special reference to neurologic sequelae after bacterial infections. Although CSF remains the biological fluid of choice, further investigations are needed in other biological fluids to improve the care of newborns.
Acknowledgments
This work was supported in part by grants to Fabrizio Michetti from Università Cattolica del S. Cuore, from the Ministero dell’Università e Ricerca Scientifica e Tecnologica, and from the Ministero della Salute and to Diego Gazzolo from the "Let’s Improve Prenatal Life" Foundation. We also thank Sangtec Medical (Bromma, Sweden) and Byk Goulden Italia for supplying analysis reagents. We offer a special thanks to the Nursery Team of the Department of Neonatology of the Giannina Gaslini Children’s University Hospital (Genoa, Italy) and the Department of Neonatology, Warsaw Medical University Hospital (Warsaw, Poland) for their constant enthusiasm and participation during this research.
References
2000 grams at birth: a population-based study. Pediatrics 2000;106:256-263.
expression produces increased blood-brain barrier permeability following temporary focal cerebral ischemia in rat. Brain Res Mol Brain Res 1999;69:135-143.[Medline]
[Order article via Infotrieve]
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
