Cerebrospinal Fluid Bacterial Antigen
No Benefit Beyond Gram Stain
Click CC box for captions; full transcript is below.
Published: May 2010Print Record of Viewing
Dr. Robin Patel discusses a recent study performed at Mayo Clinic on the use of bacterial antigen testing (BAT) on cerebrospinal fluid. The study, which showed no benefit for BAT over Gram stain, was published in a Letter to the Editor in the Journal of Clinical Microbiology (April 2010).
Presenter: Robin Patel, MD
- Director of the Initial Processing and Media Laboratory
- Director of the Bacteriology Laboratory in the Division of Clinical Microbiology at Mayo Clinic
Welcome to Mayo Medical Laboratories' Hot Topics. These presentations provide short discussions of current topics and may be helpful to you in your practice.
Our presenter for this program is Robin Patel, MD, Director of the Initial Processing and Media Laboratory and Director of the Bacteriology Laboratory in the Division of Clinical Microbiology at Mayo Clinic in Rochester, Minnesota. Dr. Patel discusses a recent study performed at Mayo Clinic on the use of bacterial antigen testing (BAT) on cerebrospinal fluid. The study, which showed no benefit for BAT over Gram stain, was published in a Letter to the Editor in the Journal of Clinical Microbiology (April 2010).
Bacterial Antigen Test (BAT)
Bacterial meningitis is a serious disease with potentially lethal consequences. Rapid and accurate diagnosis and treatment are, therefore, critical. The bacterial antigen test (or BAT) was introduced in the 1980s as a rapid test for common bacterial causes of meningitis. It screens cerebrospinal fluid for antigens of the classic bacterial meningitis pathogens, Streptococcus pneumoniae, Haemophilus influenzae type b, Group B streptococcus, Neisseria meningitidis, and Escherichia coli K1.
The bacterial antigen test uses antibodies to bacterial antigens linked to polystyrene latex particles, which agglutinate in the presence of corresponding antigen. The clinical usefulness of bacterial antigen testing has been questioned in several studies. Further, changes in the epidemiology of bacterial meningitis pathogens related to the pneumococcal, meningococcal, and H. influenzaetype b vaccines have likely affected the positive predictive value of bacterial antigen testing.
We compared cerebrospinal fluid Gram stain to bacterial antigen testing in a recent population. To our knowledge, this is the largest study comparing bacterial antigen testing to Gram stain in cerebrospinal fluid in adults and children.
We retrospectively analyzed 918 cerebrospinal fluid specimens that our institution (Mayo Clinic, Rochester MN) tested with the bacterial antigen test, Gram stain, and culture, between January 2000 and March 2009. A subset of these in which at least 1 of the following criteria was met, was further analyzed: A positive bacterial antigen test; A positive Gram stain consistent with 1 of the classical bacterial meningitis pathogens; or a classical bacterial meningitis pathogen identified by culture from cerebrospinal fluid. 42 cases were identified.
The bacterial antigen test was performed according to the manufacturer’s instructions. A positive result was indicated by the development of clearly visible agglutination after 3 minutes of mixing. Agglutination of a single test latex without agglutination in the other tests or negative control latexes indicated the presence and identity of the bacterial antigen. Cerebrospinal fluid Gram stains performed after January 2007 utilized a cytocentrifugation step. Prior to that time, cerebrospinal fluid Gram stains had not been subjected to cytocentrifugation.
Comparison of BAT and Gram Stain
Results of bacterial antigen testing and Gram stain were compared. Of the 42 identified cases, 26 were positive by both the bacterial antigen test and Gram stain, 3 were positive by the bacterial antigen test but negative by Gram stain, 11 were positive by Gram stain but negative by the bacterial antigen test , and 2 were negative by both methods. The results of the bacterial antigen test and Gram stain were not statistically significantly different from one another (p = 0.64, Fisher's exact test). We demonstrated that the bacterial antigen test did not offer statistically significant sensitivity improvement compared to Gram stain for detection of classic bacterial meningitis organisms. In the 3 cases in which bacterial antigen test was positive and Gram stain was negative, culture was also positive, rendering bacterial antigen testing redundant. None of the culture-negative cases were missed by Gram stain.
Cases With and Without Classic Bacterial Meningitis-Causing Organisms and Number Missed by BAT and Gram Stain
Classic meningitis-causing organisms were grown from cerebrospinal fluid culture in 29 of 42 cases. In 9 of the culture-negative cerebrospinal fluid cases, culture of blood or ear drainage yielded classic meningitis organisms. In the cases with positive cultures from cerebrospinal fluid, blood, or ear drainage, the results of bacterial antigen test and Gram stain were analyzed by organism type and are shown here. There were 22 Streptococcus pneumoniae cases, 6 of which were missed by bacterial antigen test, and 3 of which were missed by Gram stain. There were 7 group B streptococcus cases, 3 of which were missed by bacterial antigen test, and none of which were missed by Gram stain. There were 7 meningococcal cases, 3 of which were missed by bacterial antigen test, and 1 of which were missed by Gram stain. There were 2 Haemophilus influenzae type b cases, none of which were missed by bacterial antigen test, and 1 of which was missed by Gram stain.
In an effort to increase the sensitivity of cerebrospinal fluid Gram stain, a cytocentrifugation procedure was implemented in January of 2007. All 5 specimens with classic bacterial meningitis pathogens present in which the Gram stain was negative (including those in which the bacterial antigen test was positive) were performed prior to the implementation of the cytocentrifugation procedure. Had they been performed using the more sensitive cytocentrifugation protocol, the Gram stains in these cases may have been positive.
Studies by Other Investigators
Although previous studies have examined the bacterial antigen test, they differ from the current study in that the populations studied were primarily pediatric or fewer positive cerebrospinal fluid specimens were included in the analysis. Ours is, to our knowledge, the only study that includes large numbers of adult and pediatric patients.
Hayden et al and Perkins et al found results similar to ours in smaller groups of patients. Bhisitkul et al found the sensitivities of cerebrospinal fluid Gram stain and bacterial antigen test to be equivalent in pediatric patients. In a study by Hill et al, nearly 50% of the meningitis cases were due to organisms not detectable by the bacterial antigen test, suggesting that the changing epidemiology of bacterial meningitis pathogens reduces the usefulness of this test in current practice.
In conclusion, the findings of our large, retrospective study indicate that bacterial antigen testing provides no substantial benefit beyond Gram stain in screening for bacterial meningitis. Although results may have returned a faster result than culture in the Gram stain-negative cases, routine antibacterial treatment is recommended until all test results are back. This moderate benefit must be contrasted with the risk of a false sense of security with a false-negative bacterial antigen test, or the risk of a false-positive bacterial antigen test, which would complicate care decisions.
I would like to acknowledge Dr. Tess Karre who performed the study I have described during her fellowship in clinical microbiology at Mayo Clinic, Ms. Emily Vetter, who collected the data, Dr. Jay Mandrekar, a statistician who assisted with analysis of results, Dr. Tom Grys, who kindly reviewed our manuscript, and the outstanding staff of the Mayo Clinic Bacteriology Laboratory.