Indian Journal of Medical Microbiology Medknow Publications on behalf of Indian Association of Medical Microbiology ISSN: 0255-0857 EISSN: 1998-3646 Vol. 25, Num. 2, 2007, pp. 108-114

Indian Journal of Medical Microbiology, Vol. 25, No. 2, April-June, 2007, pp. 108-114

Original Article

Bacteriological profile of community acquired acute bacterial meningitis: A ten-year retrospective study in a tertiary neurocare centre in South India

R Mani, S Pradhan, S Nagarathna, R Wasiulla, A Chandramuki

Department of Neuromicrobiology, NIMHANS, Bangalore - 560 029, Karnataka

Correspondence Address: Department of Neuromicrobiology, NIMHANS, Bangalore - 560 029, Karnataka, India (email: acmuki@gmail.com)

Date of Submission: 17-July-2006
Date of Acceptance: 16-Nov-2006

Code Number: mb07033

Abstract

Keywords: Community acquired bacterial meningitis, bacterial meningitis, pyogenic meningitis, Streptococcus pneumoniae

Acute bacterial meningitis remains a major cause of mortality and long-term neurological sequelae worldwide. Despite the availability of potent newer antibiotics, the mortality rate due to acute bacterial meningitis remains significantly high in India and other developing countries, ranging from 16-32%. ^{[1],[2],[3],[4]} There is a need for a periodic review of bacterial meningitis worldwide, since the pathogens responsible for the infection vary with time, geography and patient age. ^[2] Increased awareness, availability and usage of vaccines may also reflect as a change in the epidemiological pattern of these pathogens.

The etiological agents of community-acquired meningitis differ from those causing nosocomial meningitis. Few studies in India and other developing countries have documented the incidence of community acquired acute bacterial meningitis (CAABM) in children and adults. Delay in diagnosis and initiation of antimicrobial therapy can result in a poor outcome of the disease. ^[5] Since clinical signs of meningitis cannot always be relied upon, ^[4] laboratory support is imperative to achieve an early diagnosis. As a result of emergence of antimicrobial resistance being reported, recommendations for therapy are changing. Laboratory surveillance of isolates is crucial to identify targets for immunization, chart preventive strategies and to help formulate rational empirical treatment for potentially fatal bacterial meningitis.

We reviewed the microbiological records of patients with CAABM admitted at NIMHANS, to characterize and determine the frequency of various bacterial pathogens causing CAABM in different age groups, compare the diagnostic efficacy of tests being used in our laboratory and to evaluate the presence of a changing pattern of aetiological agents and their antimicrobial resistance.

Materials and Methods

A retrospective study of all cases of community acquired acute bacterial meningitis (CAABM) diagnosed and treated at National Institute of Mental Health and Neurosciences (NIMHANS) over a 10-year period from January 1996 to December 2005, was undertaken. NIMHANS is a 900 bedded teaching institute and a tertiary care center catering to neurological, neurosurgical and psychiatric cases.

Criteria used for inclusion of cases in the study were the presence of a clinical picture compatible with a diagnosis of bacterial meningitis with either a cerebrospinal fluid (CSF) neutrophilic pleocytosis of at least 100 neutrophils per cubic mm and /or a positive CSF culture for bacterial pathogens and/or a positive latex agglutination test for antigen detection.

Cases of post-traumatic meningitis and meningitis developing after cranial surgery were excluded. Only one representative CSF sample from every patient was included, consecutive CSF samples from the same patient were ignored for the purpose of the study. The bacteriological records of the CSF samples of 385 patients were reviewed.

Cell counts, smear, culture

All CSF samples received at the Neuromicrobiology department were processed immediately. The macroscopic appearance of the CSF was recorded. A routine CSF total and differential count was done using a haemocytometer by standard methods. The CSF samples were subjected to a cytospin by using Shandon cytospin MODEL 001/ 002. CSF (500 microlitres) was added to the spin cups, spun for 10 minutes at 800 rpm and the resultant smear was Gram stained and examined microscopically. All CSF samples were cultured on sheep blood agar, chocolate agar, MacConkey′s agar and thioglycollate broth as a backup. The isolates were identified by standard techniques and their antimicrobial susceptibility testing was done by Kirby Bauer disc diffusion technique.

Antigen detection

Detection of soluble antigens of H. influenzae (type b), S. pneumoniae and N. meningitidis in CSF was performed by latex agglutination tests (LAT), using commercial kits (Wellcogen, UK). Antigen detection was performed subject to availability of diagnostic kits, when Gram stain of CSF was inconclusive despite a neutrophilic pleocytosis and clinical findings suggestive of meningitis. LAT for antigen detection could not be done when the quantity of CSF received was insufficient, if the CSF was blood stained or when Gram stain was morphologically suggestive of pathogens other than the three aforementioned pathogens.

Polymerase chain reaction (PCR)

Since S. pneumoniae is one of the most important causes of bacterial meningitis, a nested PCR (nPCR) assay was developed and evaluated for the detection of pneumococcal DNA in CSF samples. The gene coding for pneumolysin, a 52kDa pneumococcal toxin protein, was chosen as a target for amplification. This toxin is produced by all pneumococcal strains isolated from clinical samples and its virulence properties are well known.

Eighteen random CSF samples from patients included in the above mentioned study group with a clinical diagnosis of pyogenic meningitis were selected. An additional 32 CSF samples were used as the control group, sufficient water blanks and culture grown pneumococci serving as negative and positive controls respectively.

The control group consisted of tuberculous meningitis (TBM) (n=10), noninfectious illnesses of the CNS (n=12), Japanese encephalitis (n=5) and herpes simplex encephalitis (n=5). The tuberculous meningitis cases were confirmed at autopsy by gross and histopathological features characteristic of tuberculous pathology. Japanese encephalitis was confirmed by detection of antibodies in CSF by IgM capture enzyme-linked immunosorbent assay and herpes simplex encephalitis was confirmed by CSF PCR for glycoprotein D gene. Confirmed by standard clinical and laboratory parameters the noninfectious illnesses included patients with Guillain-Barrι syndrome (n=2), lower motor neuron palsy (n=3), motor neuron disease (n=4) and myelopathy (n=3).

DNA extraction

100 μL of CSF was added to 100 μL of lysis buffer (Tris EDTA Buffer, pH-7.4) and 50 μL of lysozyme (2 mg/mL) and incubated at 37°C for 60 minutes. In the next step, 10 μL of Proteinase K (20 mg/mL) was added and incubated at 56°C for 60 minutes in a water bath. Subsequently, extraction was carried out by adding a 200 μL solution containing 5 M-guanidium thiocyanate and 0.1M EDTA (pH-7.0), 150 μL of 7.5 M-ammonium acetate and 450 μL of chloroform-isoamyl alcohol mixture (24:1) and centrifuged at 14000 rpm for 15 minutes. DNA was precipitated using 0.54 volume of isopropanol. The precipitated DNA was then washed with 80% alcohol, dried and reconstituted in 50 μL of TE buffer.

Primers

The two primer pairs used were based on the published nPCR assay. ^[6] Amplification was done with a programmable thermal cycler (Omn-E, Thermo Hybaid, Middlesex, UK). A 50 μL reaction mixture contained 10x assay buffer (Bangalore Genei, Bangalore),10 mM dNTPs (MBI Fermentas), 10pmol of each primer, 2.5 units of Taq DNA polymerase (Bangalore Genei, Bangalore) and 10 μL of DNA extracted from the CSF specimens. Amplification involved 30 cycles of denaturation at 94°C for one minute, annealing at 55°C for one minute and extension at 72°C for one minute. The amplified product was analysed by electrophoresis on a 2% agarose gel and visualized by ethidium bromide staining.

Results

Three hundred and eighty-five patients with community acquired acute bacterial meningitis were diagnosed and treated at NIMHANS. Two hundred and ninety-three (76.1%) were male and 92 (23.9%) were female. Fifty-one (13.2%) patients belonged to the pediatric age group (less than 12 years) while 334 (86.8%) were adults.

The bacterial pathogen could be demonstrated by the cytospin Gram stain in the CSF samples of 253 (65.7%) patients, while 157 (40.8%) samples yielded growth on culture [Table - 1].

Streptococcus pneumoniae was the most common pathogen isolated on culture from 116 cases. An additional 122 cases of culture negative pneumoccal meningitis could be identified by LAT (62 cases) and typical morphology on Gram stain (60 cases). Of the 62 samples positive by LAT, 53 samples were also positive by Gram stain, while nine samples were smear negative. In the 60 samples where pneumococci could be identified only by typical Gram stain morphology, LAT was negative in 23 cases, while it could not be done in the remaining 37 samples. Thus a total of 238 (61.8%) cases of pneumococcal meningitis were identified in all age groups. [Table - 2]

H. influenzae was the causative agent in 7 (1.8%) cases. Culture yielded the pathogen in four cases. The organism could be demonstrated on Gram stain in all these four cases. LAT was positive in one sample; it was not performed for the rest of the three culture positive samples. Antigen detection could identify three more cases, which were negative by Gram stain and culture.

Only four (1%) cases of meningococcal meningitis were identified. Gram stain demonstrated the gram negative diplococci in all four cases. Two samples yielded growth on culture, LAT was positive in one of the samples on which it was done. The two culture negative samples were positive by LAT. [Table - 3] and [Table - 4] show the correlation of LAT for antigen detection with smear and culture results. Gram-negative bacilli (other than H. influenzae ) accounted for 19 (4.9%) cases, Staphylococcus aureus for 7(1.8%) cases, while Streptococcus spp. accounted for 9 (2.3%) cases.

The cell counts of the CSF samples ranged from no cells to sheets of cells, which could not be counted on the hemocytometer. A predominance of polymorphonuclear cells was the common feature in all cases with high cell counts. [Table - 5] includes the correlation between cell count range and the bacterial pathogens identified.

All the 116 isolates of S. pneumoniae were sensitive to penicillin and vancomycin. The resistance pattern to other drugs tested was ofloxacin (23.2%), cotrimoxazole (22.5%), cefotaxime (7.8%), ceftriaxone (6.8%) and tetracycline (2.5%). Two of the four isolates of H. influenzae that could be tested were sensitive to ampicillin, cefuroxime and chloramphenicol. The two isolates of N. meningitidis were sensitive to penicillin. Other gram negative bacilli showed a high percentage of resistance to several common drugs tested; ampicillin (79%), amikacin (26.3%), gentamicin (21%), ciprofloxacin (47.4%), oflaxacin (58%), cefotaxime (52.6%), ceftazidime (63.2%) and ceftriaxone (52.6%).

PCR

PCR for detection of Streptococcus pneumoniae was positive in all 11 cases that could be diagnosed by culture, smear or LAT, thus demonstrating 100% sensitivity. An additional four cases, which were negative both by culture and smear, were positive by PCR. LAT was negative in two of these samples and was not done on two samples [Table - 6]. Out of the 32 control samples that were employed in the assay, the test showed a positive result in only two cases of autopsy proven TBM. All the remaining controls were negative by the test.

Discussion

Acute bacterial meningitis is a medical emergency, which warrants early diagnosis and aggressive therapy. Most often therapy for bacterial meningitis has to be initiated before the etiology is known. The choice of initial antimicrobial therapy in CAABM is based on the most common pathogen prevalent in a particular geographical area and age group and its antibiotic sensitivity pattern.

Though the common pathogens associated with CAABM are S. pneumoniae, H. influenzae and N. meningitidis , the aetiological agents and their relative frequency may vary in different geographical areas. Some changing trends in the epidemiology of CAABM have also been reported worldwide over the past few decades.

Bacterial meningitis is being reported predominantly in adults in USA because of the immunization practices adopted and also due to a relative increase in frequency of nosocomial meningitis. In North America and Europe, because of the vaccine related decline in H. influenzae disease, S. pneumoniae and N. meningitidis remain important pathogens in children and young adults. Group B streptococcus is the most common pathogen associated with meningitis in newborns. Listeria monocytogenes is also recognized as a significant cause of meningitis in newborns and the elderly in the United States. ^[7]

As compared to Western studies, the relative incidence of meningitis caused by H. influenzae, N. meningitidis and Listeria is less in South-East Asia. On the contrary, gram negative bacilli such as Klebsiella pneumoniae and Pseudomonas aeruginosa are increasingly being recognised as important pathogens of community-acquired as well as nosocomial meningitis especially among the elderly and in patients with chronic debilitating diseases like cirrhosis, diabetes and malignancies. ^[2] As life expectancy is increasing, it may not be uncommon to see an increased incidence of CAABM in the elderly in the coming years. ^[2]

Streptococcus pneumoniae was the most common etiological agent of community acquired meningitis in all age groups accounting for 238 (61.8%) cases in our study, reflecting a similar trend reported in an earlier study from our institute (1978-1988). ^[8] Most Indian studies have also reported a high incidence of pneumococcal meningitis. ^[1],[4]

In our present study all 7(1.8%) cases of H. influenzae meningitis occurred in the pediatric age group (0-12 years). Some Indian authors have reported a high incidence of H. influenzae meningitis in the pediatric age group, ^[4] while others have experienced a low incidence. ^[1] In our set-up, the incidence has remained low for the past several decades. However, the commonly held view of H. influenzae disease being rare in Asia has been challenged by a research study, which recommends large-scale vaccination for H. influenzae type b (Hib) in Asian countries. ^[9]

Only four (1%) cases of meningococcal meningitis were detected in the last ten years, all in adult patients. Our study reiterates the finding of a low prevalence of meningococcal meningitis ^[4] except during epidemics, in various Indian studies.

Though this study and various other studies ^[1],[4] have implicated various gram negative bacilli (other than H. influenzae ) as etiological agents of CAABM both in children and adults, they account for a minor percentage of the cases. More data from systematic studies all over India, in the coming years needs to be analysed to comment whether meningitis caused by gram negative bacilli is on the increase, reflecting a trend similar to some other countries.

Though streptococci were isolated from nine cases of meningitis, we did not encounter any case of Group B streptococcal meningitis, a common cause of neonatal meningitis. It was possibly because patients in the zero to two years age group formed only 5.9% (23 cases) of the study group. It can be attributed to the fact that a public tertiary care children′s hospital lies in close proximity to our institute, where most pediatric cases are referred for further management. In recent years Group B streptococci, known to cause meningitis in neonates, is being increasingly recognized as a cause of meningitis in adults in Southeast Asia and is associated with a high case fatality. ^[10]

No case of Listeria monocytogenes meningitis was seen in this study. Though the incidence is low in most Indian studies, it should be considered especially in the elderly and immunocompromised patients, since it is known to be resistant to third generation cephalosporins used in the empirical treatment of bacterial meningitis.

Salmonella species, Flavobacterium meningosepticum and Bacillus anthracis, reported as causative agents of acute meningitis in earlier studies from our institute, were not encountered in the present study. ^{[11],[12],[13]}

A simple Gram stained smear can offer immediate clues to aid a diagnosis of pyogenic meningitis. Some studies have reported a CSF Gram stain sensitivity of 60-90% and a high specificity of> 97%, stressing its importance in the rapid and accurate diagnosis of the causative bacteria. ^[4],[14] In a study, patients with pneumococcal meningitis were found to be at risk for an unfavorable outcome, even after correction for other clinical predictors. The authors stress the importance of the need for clinicians to know the causative organism in predicting the outcome and advocate Gram staining of CSF as a routine procedure for prompt identification of the pathogen. ^[14] The yield of bacteria on a Gram stain depends on several factors like the number of organisms present, prior use of antibiotics, technique used for smear preparation (centrifuged deposit, cytospin, direct smear etc.), staining techniques and the observer′s skill and experience. Despite low Gram stain smear positivity from CSF samples and the fact that a negative Gram stain does not rule out infection, the importance of a positive smear cannot be over-emphasized, especially in developing countries where financial constraints limit the use of other rapid diagnostic tests to diagnose this potentially fatal infection.

In our study, Gram stain provided an evidence of the causative bacteria in 253 (65.7%) patients. Of these 138 CSF samples yielded growth on culture, while 115 were culture negative. Fifty-five of the culture negative cases were positive by antigen detection tests. Thus in a total of 60 culture negative cases (23-antigen negative and 37-antigen detection not done) the etiological pathogen was recognized only by Gram stain [Table - 3]. Most Indian studies report only culture findings or a low smear positivity. ^[15] Our relatively high yield of pathogens on Gram stain can be attributed to the routine use of cytospin to concentrate the smear. A cytospin also provides several other diagnostic benefits like good preservation of morphology of cells and bacteria and an increased rate of detection of bacterial pathogens especially in partially treated pyogenic meningitis, which can mimic tuberculous meningitis posing a diagnostic dilemma for clinicians. ^[8]

CSF samples from 157 (40.8%) patients were positive on culture in our study. Several studies report culture negative cases of meningitis or a low CSF culture positivity, ranging from 6 to 50%. ^[1],[4],[15] Various reasons cited in the literature for a low yield of bacteria on culture are prior antibiotic therapy, delay in transport of specimens to the laboratory, nonavailability of special media for specific pathogens, presence of autolysis enzymes in CSF and lack of a 24 hour facility for processing CSF samples. The need for development of Neuromicrobiology, especially in centres catering to neurological/neurosurgical patients, ^[8] encouraging a careful Gram staining and prompt bed-side inoculation of CSF samples on culture media by resident doctors, ^[1] has been stressed by some authors.

Latex agglutination tests (LAT) for detection of capsular antigen of Streptococcus pneumoniae , N. meningitidis and H. influenzae (type b) were performed on a total of 185 CSF samples. The test was positive in 101(54.6%) cases, accounting for 26.2% of the total 385 cases. In the 228 culture negative cases in our study, LAT was performed on 133 CSF samples, of which 66(49.6%) were positive. 55 of these 66 cases were also positive by Gram′s stain, but 12(18%) samples which did not show any evidence of the pathogen on either Gram′s stain or culture, were positive by LAT, thus helping clinch the diagnosis. [Table - 3] LAT was performed in 52 of the 122 culture positive cases [Table - 4]. It was positive in 34 (65.4%) cases, but negative in 18 (34.6%) cases. These 18 cases represent false negative LAT. All these 18 cases yielded Streptococcus pneumoniae on culture. A US-based study to determine the efficacy of antigen detection tests did not identify any false negative LAT, suggesting a high sensitivity of these tests. ^[16] However a similar Indian study has reported a sensitivity of 83%. ^[15]

The false-negative LAT in our study could be possibly because of low antigen titres in the CSF. ^[1] Though capsular serotyping of our isolates was not performed, it is possible that the antiserum in diagnostic LAT kits does not detect all the capsular serotypes prevalent in our geographical area or probably as yet unrecognized serotypes are the causative agents in such cases. Some workers have questioned the clinical usefulness of antigen detection tests, ^[16] explaining that a negative test does not rule out infection and false positive results may lead to unnecessary prolonged course of antibiotics, lengthened hospital stay and in some cases important clinical complications. Recent immunization with Hib conjugate vaccine and infection with cross-reacting organisms were quoted as common reasons for false positive LAT. However, several studies ^[4],[15] advocate the usefulness of LAT, especially in pretreated cases and to differentiate partially treated pyogenic meningitis from tuberculous meningitis, which is rampant in India. Despite its drawbacks, we found LAT to be a simple, rapid procedure suitable to be used as an adjunct laboratory test, but needs to be interpreted cautiously, taking the patient′s clinical condition and several other factors into consideration. In developing countries like India where many laboratories lack facilities for culture and other elaborate investigations, LAT can help establish the crucial diagnosis. However, the high costs of LAT kits remain a prohibitive factor for its routine use in most laboratories. Since prompt diagnosis and aggressive treatment for pyogenic meningitis determines a favorable outcome, it is the need of the hour to develop affordable indigenous kits for this purpose, designed to detect the serotypes prevalent in our geographical area and also to help detect other prevalent pathogens.

In our study, we could identify some proof of the infecting agent (culture and/or smear and/or LAT) in 284 (73.7%) of the total 385 cases. The aetiology could not be established in the rest of the 101 (26.2%) cases. Though all these CSF samples showed a neutrophilic pleocytosis and were clinically compatible with a diagnosis of pyogenic meningitis, in such cases the final diagnosis can be established only after analyzing other biochemical parameters and tests to rule out the initial stages of tuberculous or viral meningitis.

Polymorphonuclear leucocytosis was the common CSF finding in our study; however a predominance of lymphocytes is sometimes known to occur in acute bacterial meningitis. 14 cases in our study had a CSF cell count of < 100 cells/cumm, 2 of which had a cell count of 10 cells/cmm and one had no cells (data not shown). All these cases yielded S. pneumoniae on culture. Normal or marginally elevated CSF white cell counts are known to occur in 5-10% patients and are associated with an adverse outcome. ^[14]

All 116 isolates of S. pneumoniae in our study were sensitive to penicillin. Some Indian studies have reported the emergence of penicillin resistance in their strains. ^[17] There is a need for continued monitoring for penicillin resistance in pnemococcal isolates, though at present it is not a significant problem in India. ^[18]

The PCR assay showed an overall sensitivity of 100% by positively identifying all the culture, smear and antigen positive isolates. The assay could pick four additional cases, which could not be diagnosed by the conventional techniques. Out of the 32 controls employed for the assay, the assay showed a positive result in only two cases of autopsy positive TBM cases, suggesting the possibility of a mixed underlying infection in those particular patients. The remaining controls were all negative, which showed a high specificity of the assay. Most often, because of the inadequacy of the CSF samples, delay in transport or initiation of antibiotics, conventional methods may not yield the pathogen. This re-emphasizes the need for a molecular technique, not requiring the organism to be viable. ^[19] Nested PCR was found to be a sensitive, specific and rapid tool for diagnosis of pneumococcal meningitis. Several new molecular techniques for detecting bacteria in CSF by PCR with a high sensitivity and specificity have emerged as powerful tools in diagnosis of patients with culture-negative meningitis. However, further refinements and more studies to evaluate the practical value of such techniques in a clinical setting may be needed before they can be recommended for routine use.

In summary, the spectrum of bacteria causing CAABM has remained more or less the same in our set-up. It needs to be reiterated that simple, rapid, inexpensive tests like the Gram stain remain significant means of diagnosis of CAABM in developing countries. To increase the cost-effectiveness in a resource limited setting, LAT for pneumococcal antigen should be performed first, since it is the most common pathogen causing CAABM in all age groups. Smear negative CSF samples with neutrophilic pleocytosis in patients with a clinical suspicion of bacterial meningitis warrant antigen testing with LAT.

Streptococcus pneumoniae remains the major aetiological agent of CAABM both in adults and children not only in India, but worldwide. High fatality rates have been reported in CAABM by Streptococcus pnemoniae with long-term neurological sequelae in survivors. Meningitis caused by H. influenzae has almost been eliminated from the Western world following routine vaccination with Hib conjugate vaccine. ^[7] Introduction of conjugate vaccines against Streptococcus pneumoniae can reduce the burden of childhood meningitis and may produce herd immunity among adults. ^[20]

References

Copyright 2007 - Indian Journal of Medical Microbiology

The following images related to this document are available:

Photo images