الفهرس 
Bacterial MeningitisOnly 14 pages are availabe for public view
 |  | from | 208 |  |  |
| | | from | 208 | | |
Abstract
A
ppropriate diagnosis and early antibiotic treatment of bacterial meningitis reduce the risk of dying from meningitis to below 15%, although the risk remains higher among young infants and the elderly. It is also very important to differentiate between acute bacterial meningitis (ABM) and acute viral meningitis (AVM) which is the most common type of meningitis that does not require antibiotics and is usually of better prognosis (Guiddir et al., 2014).
Although considered the gold standard, culture requires at least 48 hours before results can be available. Conventional culture and Gram stain are associated with high false negative rates particularly when patients had already started empirical antibiotic therapy (Brouwer et al., 2012). The value of latex agglutination test in the diagnosis of bacterial meningitis has been questionable (Fouad et al., 2014). Polymerase chain reaction, even multiplex-PCR, although proven to be helpful in etiological diagnoses during the past decade, it is capable only of detecting the pathogens that are already suspected and included in the primer mix (Li et al., 2017).
Therefore, the use of diagnostic tools characterized by easy technique, reduced turnaround time, and high diagnostic accuracy is advisable (Lippi et al., 2014).
The aim of the current study was to determine the value of LCN 2 as a new diagnostic marker for acute bacterial meningitis by comparing its diagnostic performance with that of conventional CSF culture, latex agglutination test and multiplex PCR.
The study included eighty CSF specimens collected from patients with signs and symptoms of acute meningitis that were admitted at Al-Abbassia Fever hospital during the period between September 2015 and August 2017. They were 34 females (42.5%) and 46 males (57.5%) with their age ranging from 15 years to 60 years (median 33 years). Specimens were subjected to routine microbiological investigations for the isolation and identification of rapidly growing aerobic and anaerobic bacteria; latex agglutination test for the detection of Streptococcus group B, S. pneumoniae, H. influenza type b, Neisseria and E. coli type K antigens; multiplex-PCR for the detection of S. pneumoniae, H. influenzae type b, and N. meningitides; ELISA for the estimation of LCN2 level, and chemical analysis for the estimation of CSF glucose and total protein levels.
In the current study, out of the 80 examined CSF specimens, 43 samples (53.8%) were found to be positive by either one or more of the used bacterial detection method (conventional culture, multiplex PCR, and latex agglutination test) denoting bacterial meningitis. The remaining 37 specimens (46.3%) were negative by all of the 3 used methods suggesting possible aseptic meningitis.
Chemical examination of the CSF specimens, in the present study, revealed that the median CSF glucose level was 40mg/dl (IQR= 24 - 66.5 mg/dl) with 47.5% of the patients having CSF glucose levels <40 mg/dl. On the other hand, the median CSF protein level was 159.5mg/dl (IQR= 85.5 – 278 mg/dl) with 100% of the patients having CSF protein levels > 50mg/dl.
There was a significant difference between the patients with evidence of bacterial meningitis and those with possible aseptic meningitis regarding the CSF glucose levels (P<0.001). Among patients with bacterial meningitis the median CSF glucose level was 25mg/dl (IQR=19-38) whereas the patients with aseptic meningitis had median CSF glucose level of 67mg/dl (IQR=50-73mg/dl). Also, a significant difference was observed between the patients with evidence of bacterial meningitis and those with possible aseptic meningitis regarding the CSF protein levels (P<0.001). Patients with bacterial meningitis had a median CSF protein level of 270mg/dl (IQR= 194-376mg/dl) whereas those with suspected aseptic meningitis had a median CSF protein level of 85mg/dl (IQR=69-100mg/dl).
In the current work, the median cell count was 1252.5 cells/cmm (IQR= 650 - 2665). With regard to the cell type, 50 out of the 80 studied CSF samples (62.5%) showed predominance of polymorphnuclear leucocytes, while the remaining thirty samples (30/80; 37.5%) showed predominance of lymphocytes.
In the present study, 21.3% of the studied CSF samples (17/80) gave positive findings on microscopic examination using the Gram stain. Gram positive cocci were detected in 58.8% (10/17) of the positive CSF gram-stained films whereas gram negative bacilli constituted the remaining 41.2% (7/17).
Using conventional culture, 21 out of the 80 (26.2%) studied CSF samples gave positive results. All of the positive cultures (21/21, 100%) recovered a single isolate. Among positive cultures, Streptococcus pneumoniae was the most commonly identified pathogen (12/21; 57.1%). E.coli and Pseudomonous spp. were isolated from 3 out of the 21 positive cultures (14.3%), respectively. Also, growth of each of Haemophilus spp., Klebsiella spp. and Enterococcus spp. was detected in 1/21 (4.8%) of the positive cultures, respectively.
A statistically excellent agreement was observed, in the current study, between the results of conventional culture and those of Gram stain (Kappa = 0.862, P<0.001). Compared to conventional culture, regarding the diagnosis of bacterial meningitis, the sensitivity and specificity of Gram stain were 100% and 81%, respectively. The overall diagnostic accuracy of gram was 95% with PPV and NPV of 93.7% and 100%, respectively.
Fifty CSF specimens, suspected to be acute bacterial meningitis (WBCs count 100 cells/mm3 with predominance of PMNLs), were examined by the latex agglutination test for the detection of antigens of Streptococcus group B, H. influenza type b, S. pneumoniae, N. meningitidis groups A, B, C, Y or W135 and E. coli K1. Positive results were obtained in 23/50 (46%) samples with S. pneumoniae being the only organism detected.
A statistically poor agreement was found between the results of the conventional culture and the results of latex agglutination test with regard to the diagnosis of bacterial meningitis (Kappa = 0.109, P=0.441). Comparing the latex agglutination test to conventional culture, regarding the recovery of S. pneumoniae, H. influenza and E.coli, the sensitivity and specificity of the latex agglutination test were found to be 43.5% and 77.8%, respectively with an overall diagnostic accuracy of 62.0%, PPV of 62.5% and a NPV of 61.8%.
The same 50 specimens, suspected to be bacterial meningitis, were examined by the multiplex-PCR for the detection of Streptococcus pneumoniae, Haemophilus influenzae type b, Neisseria meningitides DNA. Results showed that 40 samples (40/50; 80%) had positive results with S.pneumoniae being the most commonly identified pathogen since it was detected in 39/40 (97.5%) of the PCR positive samples. Only one sample gave positive result with H. influenza (1/40; 2.5%). N. meningitides was not detected in any of the studied CSF samples (0/40; 0%).
A statistically poor agreement was found between the results of conventional culture and those of PCR regarding the detection of S. pneumoniae, H. influenza and N. meningitides (Kappa = 0.161, P = 0.036). Comparing the results of conventional culture with those of multiplex PCR, regarding the detection of S. pneumoniae, H. influenza and N. meningitides, the sensitivity and specificity of culture were 32.5% and 100.0%, respectively. The overall diagnostic accuracy of culture was 46.0% with PPV and NPV of 100.0% and 27.0%, respectively.
Considering the culture as the gold standard method for the recovery of the causative organisms of bacterial meningitis, the multiplex PCR was found to have 100% sensitivity, 32.5% specificity, 27% PPV and 100% NPV.
In the present study, fair agreement was observed between the results of the latex agglutination test and those of the multiplex PCR with regard to the diagnosis of bacterial meningitis (Kappa = 0.351, P=0.001). The sensitivity and specificity of the latex agglutination test were found to be 57.5% and 100.0%, respectively with an overall diagnostic accuracy of 66.0%, a PPV of 100% and a NPV of 37.0%.
As we studied the effect of prior antibiotic administration on the results of the different studied tests, we observed a statistically significant difference between patients on antibiotics and those without prior antibiotic treatment regarding the results of gram stain (P<0.001), and conventional culture (P<0.001). However, no statistically significant difference was found between the patients on antibiotics and those without prior antibiotic treatment regarding the results of the latex agglutination test (P >0.05) as well as the results of the multiplex PCR (P > 0.05).
Estimation of (LCN 2) levels by ELISA was performed on the eighty CSF samples. The mean value was 53.68 ± 49.27 pg/ml with median (IQR) of 35.5 (12.5 - 98.5) pg/ml. There was a highly significant difference between the patients with bacterial meningitis and those with suspected aseptic meningitis regarding the median CSF lipocalin 2 levels (P<0.001). Among patients with bacterial meningitis, the median LCN2 level in the CSF was 60pg/ml (IQR= 36-128pg/ml). On the other hand, in patients with suspected aseptic meningitis, the median LCN2 level in the CSF was 13pg/ml (IQR= 5-30pg/ml).
A statistically significant positive correlation was found between the levels of CSF LCN 2 and CSF total protein (rs = 0.595; P 0.001). On the other hand, there was a statistically significant negative correlation between the levels of CSF LCN 2 and the CSF glucose levels (rs = -0.558; P 0.001). No statistically significant correlation was found between the CSF LCN2 level and the cell count (rs=0.038, P> 0.05).
The receiver operating characteristic (ROC) curve analysis was used to calculate the cutoff point of LCN 2 that can discriminate between bacterial and viral meningitis. An excellent diagnostic accuracy was found for LCN 2 by ROC curve analysis, with an area under the curve (AUC) of 0.86 (95% CI, 0.765 to 0.928; p value=0.0001) for identifying patients with positive bacterial meningitis. At a cut off of
18 pg/mL, the sensitivity and specificity of LCN 2 for identifying positive CSFs were 78.4% and 97.7%, respectively. The overall diagnostic accuracy of LCN2 in CSF was 88.8% with PPV and NPV of 84% and 96.7%, respectively. This is clinically meaningful, because a LCN 2 value in CSF above this threshold would provide strong support for public health decision making and a reliable basis to continue empiric antibiotic therapy until results of fluid culture and/or molecular biology become available.