Towards accurate exclusion of neonatal bacterial meningitis: a feasibility study of a novel 16S rDNA PCR assay

Abstract

Background: PCRctic is an innovative assay based on 16S rDNA PCR technology that has been designed to detect a single intact bacterium in a specimen of cerebro-spinal fluid (CSF). The assay's potential for accurate, fast and inexpensive discrimination of bacteria-free CSF makes it an ideal adjunct for confident exclusion of bacterial meningitis in newborn babies where the negative predictive value of bacterial culture is poor. This study aimed to stress-test and optimize PCRctic in the "field conditions" to attain a clinically useful level of specificity.

Methods: The specificity of PCRctic was evaluated in CSF obtained from newborn babies investigated for meningitis on a tertiary neonatal unit. Following an interim analysis, the method of skin antisepsis was changed to increase bactericidal effect, and snap-top tubes (Eppendorf™) replaced standard universal containers for collection of CSF to reduce environmental contamination.

Results: The assay's specificity was 90.5% in CSF collected into the snap-top tubes - up from 60% in CSF in the universal containers. The method of skin antisepsis had no effect on the specificity. All CSF cultures were negative and no clinical cases of neonatal bacterial meningitis occurred during the study.

Conclusions: A simple and inexpensive optimization of CSF collection resulted in a high specificity output. The low prevalence of neonatal bacterial meningitis means that a large multi-centre study will be required to validate the assay's sensitivity and its negative predictive value.

Keywords: Broad-range PCR; Ethidium monoazide; Neonatal bacterial meningitis.

Fig. 1

PCRctic: a CSF samples; b Intact bacteria are pelleted along with patient cells and debris; c The pellet is resuspended in the assay mixture (PCR reagents, ethidium azide, buffer); d light activates the ethidium azide, destroying exposed DNA (from dead bacteria or from contaminants in the reagents) but leaving intact bacteria unaffected; during PCR e, heat causes the bacteria to lyse, making their DNA available for amplification. f amplification products are then detected by a simple fluorometric assay (melting-curve analysis) using a widely-available real-time PCR machine. Total assay time is approximately 2 h

Fig. 2

Elimination of false positives due to contaminating bacterial DNA, by photoactivation of EtA. A band at 126 bp indicates a positive result. Without photoactivation (“no light”) all specimens are positive. Exposure to light for 60 s completely eliminated false positive signal. Electrophoresis of PCRctic products in 3% agarose gel

Fig. 3

Sensitivity of PCRctic with 200-μl CSF samples “spiked” with bacteria. Electrophoresis of PCRctic products in 3% agarose gel. aE. coli (DH5-alpha strain) were titrated to an average of 1.5 or 5 CFU/sample. Occasional negative results at 1.5 CFU/sample can be accounted for by the random nature of titration at this level: with an average of 1.5 CFU/sample, approximately 22% of samples should be negative. bStreptococcus pneumoniae, Staphylococcus aureus and Listeria monocytogenes (patient-derived isolates) were titrated by 10-fold dilutions to ≤5 CFU/sample

Fig. 4

Sensitivity of PCRctic with 100-μl samples “spiked” with 10 and 5 CFUs of bacteria per reaction. Melting curve analysis of PCRctic results

Fig. 5

Melting curve analysis of PCRctic results obtained from clinical specimens