1. Introduction

Bacterial pneumonia claims the lives of 10.8 million children annually [1], yet the diagnosis of both bacteraemic and non-bacteraemic pneumonia remains a challenge. Definitive diagnosis of bacteraemic pneumonia is difficult due to the reliance on culture-based methods [2]. Currently, blood culture detects less than 10% of pneumococcal pneumonia cases in children [3].

PCR methods have greatly improved the sensitivity of pneumococcus detection from clinical specimens. The advantages of quantitative realtime PCR (qPCR) processes are its speed, limit of detection and ability to process large numbers of specimens. These features make qPCR an appealing tool for the diagnosis of disease.

Several studies have specifically used qPCR to detect Streptococcus pneumoniae and Haemophilus influenzae. qPCR of the lytA gene has demonstrated a sensitivity of between 47% to 100%, and a specificity of 80% to 100% across studies investigating serum, whole blood, middle ear fluid and cerebrospinal fluid [3–6]. One study by Rello et al. (2009) was able to correlate S. pneumoniae bacterial loads in whole blood ≥ 103 copies/ml with poorer clinical outcomes [7]. However, there are concerns over false positive results with the enhanced sensitivity of molecular methods, stemming from issues with the pneumolysin-based PCR for detection of S. pneumoniae detecting other streptococci [8]. For our population, false positive results are of great concern due to the dense nasopharyngea colonisation in Indigenous children which has seen assays such as the rapid antigen-based assay BinaxNOW (Inverness Medical Innovations Inc, USA) not considered for use due to the false-positives in children who were colonised with S. pneumoniae [9].

The first aim of this study was to adapt and optimise published S. pneumoniae and H. influenzae DNA extraction and qPCR methods [10] for detection of these pathogens in serum samples. The qPCR sensitivity and specificity was determined using spiked serum samples. The second aim of the study was to determine if the sensitivity of a qPCR assay would detect false positives in serum from Indigenous children with varying density of nasopharyngeal colonisation. This was determined through testing of matched serum and nasopharyngeal swabs (NPS) from Indigenous children less than 5 years of age, without pneumonia, but with varying density of nasopharyngeal colonisation.

2. Methods and results

2.1. Method

To determine the most sensitive DNA extraction method, serum samples from healthy volunteers were spiked with reference strains of S. pneumoniae (ATCC 49619) and H. influenzae (ATCC 49274) at approximately 107 colony forming units/ml or to 0.6 optical density. DNA was extracted from the spiked samples using three commercially available kits; the UltraClean DNA BloodSpin Kit (MO BIO Laboratories, Inc., USA), the Wizard SV Genomic DNA Purification System (Promega Corporation, USA) and the QIAamp DNA Blood Mini Kit (QIAGEN, Germany). The blood protocol was followed for each kit. The QIAamp DNA Blood Mini Kit consistently yielded the greatest quantity of bacterial DNA when an optimised serum starting volume and DNA elution volume of 100 µl were used. Testing of pneumococcal serotypes with varying degrees of capsule expression (serotype 3, 19F and 14) did not affect the results.

1, 2 and 4 µl of each extraction was subsequently used as template in S. pneumoniae and H. influenzae qPCR assays [10], targeting lytA [5] and glpQ [11] genes, respectively. Each qPCR included 5 quantitative DNA standards with an unspiked serum extraction as a negative control. S. pneumoniae standards ranged from 8.92 × 104 to 8.92 genome copies and H. influenzae standards ranged from 9.91 × 104 to 9.91 genome copies. 2 µl of template enhanced sensitivity but inhibition was evident with 4 µl of template. In summary, using the QIAamp DNA Blood Mini Kit of elution volume and 2 µl of extraction as template for the qPCR enabled detection of 446 and 496 cellular equivalents of S. pneumoniae and H. influenzae, respectively. No qPCR inhibition was evident with these conditions.

2.2. Clinical evaluation

For clinical validation of this method it was imperative to rule out the detection of serum positives during nasopharyngeal carriage. Thirtynine serum samples (stored at −80°C) and 30 corresponding NPS (stored in 1ml in skim-milk tryptone glucose glycerol broth [12] at −80°C) from Indigenous children under the age of 5 years were selected based on microbiologically documented positive carriage for S. pneumoniae or H. influenzae or carriage of both bacteria. Briefly, 100 µl of serum was extracted and quantified as above and 100 µl of corresponding NPS was extracted and quantified as reported previously [10]. All sera were negative for both S. pneumoniae and H. influenzae as determined by qPCR while over >105 bacterial copies of both bacteria were found in at least 22 culture positive NPS (Table 1).

Table 1 qPCR detection of S. pneumoniae and H. influenzae in clinical samples
Full size table

To further ensure the absence of detectable bacteria in the blood of high carriage sera, ethanol precipitation was performed on all DNA extracts to increase template concentration five-fold. No additional positives were exposed.

2.3 Ethical approval

Ethical approval was granted from the Human Research Ethics Committee of Northern Territory Department of Health and Menzies School of Health Research with reference number 08/84.

3. Discussion

Detection and diagnosis of bacterial pathogens in paediatric populations is limited by small blood volume and a slow, insensitive gold standard (blood culture). While serum qPCR of adult patients with community-acquired pneumonia has been shown to detect as low as 8 cellular equivalents per ml, larger blood volumes are used [7]. In this study, using paediatric volumes of serum, we were able to detect to 446 and 496 cellular equivalents of S. pneumoniae and H. influenzae per 100 µl. This equates to a blood infection of approximately 104 cells/ml. These qPCR assays can be run concurrently and take 5 hours from receipt of sample to final result.

4. Conclusion

qPCR of S. pneumoniae or H. influenzae DNA in serum is rapid, reproducible and appropriate for children in populations experiencing dense nasopharyngeal colonisation of the target pathogens.