PLoS ONEplosplosonePLOS ONE1932-6203Public Library of ScienceSan Francisco, CA USA10.1371/journal.pone.0152253PONE-D-16-00861Research ArticleBiology and life sciencesAnatomyBody fluidsBloodMedicine and health sciencesAnatomyBody fluidsBloodBiology and life sciencesPhysiologyBody fluidsBloodMedicine and health sciencesPhysiologyBody fluidsBloodMedicine and health sciencesHematologyBloodBiology and life sciencesMolecular biologyMolecular biology techniquesArtificial gene amplification and extensionPolymerase chain reactionReverse transcriptase-polymerase chain reactionResearch and analysis methodsMolecular biology techniquesArtificial gene amplification and extensionPolymerase chain reactionReverse transcriptase-polymerase chain reactionBiology and life sciencesMolecular biologyMolecular biology techniquesArtificial gene amplification and extensionPolymerase chain reactionResearch and analysis methodsMolecular biology techniquesArtificial gene amplification and extensionPolymerase chain reactionBiology and life sciencesMicrobiologyMedical microbiologyMicrobial pathogensBacterial pathogensMedicine and health sciencesPathology and laboratory medicinePathogensMicrobial pathogensBacterial pathogensEngineering and technologyEquipmentLaboratory equipmentFilter paperResearch and analysis methodsExtraction techniquesDNA extractionBiology and life sciencesOrganismsBacteriaStreptococcusPneumococcusBiology and life sciencesMicrobiologyMedical microbiologyMicrobial pathogensBacterial pathogensStreptococcusPneumococcusMedicine and health sciencesPathology and laboratory medicinePathogensMicrobial pathogensBacterial pathogensStreptococcusPneumococcusPeople and placesPopulation groupingsAge groupsChildrenPeople and placesPopulation groupingsFamiliesChildrenMolecular Detection of Streptococcus pneumoniae on Dried Blood Spots from Febrile Nigerian Children Compared to CultureS. pneumonia Dried Blood Spot PCRIroh TamPui-Ying12*Hernandez-AlvaradoNelmary2SchleissMark R.12Hassan-HangaFatimah3OnuchukwuChuma4UmoruDominic5ObaroStephen K.6Division of Pediatric Infectious Diseases and Immunology, University of Minnesota Masonic Children’s Hospital, Minneapolis, Minnesota, United States of AmericaCenter for Infectious Diseases, Microbiology and Translational Research, Minneapolis, Minnesota, United States of AmericaDepartment of Paediatrics, Aminu Kano Teaching Hospital, Kano, NigeriaDepartment of Paediatrics, Federal Medical Center, Keffi, Nasarawa, NigeriaDepartment of Paediatrics, Nyanya General Hospital, Abuja, NigeriaDepartment of Pediatrics, University of Nebraska Medical Center, Omaha, Nebraska, United States of AmericaMiyajiEliane NamieEditorInstituto Butantan, BRAZIL
The authors have declared that no competing interests exist.
Conceived and designed the experiments: PIT NHA. Performed the experiments: PIT NHA. Analyzed the data: PIT NHA MRS SKO. Contributed reagents/materials/analysis tools: MRS SKO FHH CO DU. Wrote the paper: PIT NHA MRS SKO.
* E-mail: irohtam@umn.edu23320162016113e015225381201611320162016Iroh Tam et alThis is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.Background
Nigeria has one of the highest burdens of pneumococcal disease in the world, but accurate surveillance is lacking. Molecular detection of infectious pathogens in dried blood spots (DBS) is an ideal method for surveillance of infections in resource-limited settings because of its low cost, minimal blood volumes involved, and ease of storage at ambient temperature. Our study aim was to evaluate a Streptococcus pneumoniae real-time polymerase chain reaction (rt-PCR) assay on DBS from febrile Nigerian children on Whatman 903 and FTA filter papers, compared to the gold standard of culture.
Methods
Between September 2011 to May 2015, blood was collected from children 5 years of age or under who presented to six hospital study sites throughout northern and central Nigeria with febrile illness, and inoculated into blood culture bottles or spotted onto Whatman 903 or FTA filter paper. Culture and rt-PCR were performed on all samples.
Results
A total of 537 DBS specimens from 535 children were included in the study, of which 15 were culture-positive for S. pneumoniae. The rt-PCR assay detected S. pneumoniae in 12 DBS specimens (2.2%). One positive rt-PCR result was identified in a culture-negative specimen from a high-risk subject, and two positive rt-PCR results were negative on repeat testing. Six culture-confirmed cases of S. pneumoniae bacteremia were missed. Compared to culture, the overall sensitivities of Whatman 903 and FTA DBS for detection of S. pneumoniae were 57.1% (95% CI 18.4–90.1%) and 62.5% (95% CI 24.5–91.5%), respectively. Nonspecific amplification was noted in an additional 22 DBS (4.1%). Among these, six were positive for a non-S. pneumoniae pathogen on culture.
Conclusions
Rt-PCR was able to detect S. pneumoniae from clinical DBS specimens, including from a culture-negative specimen. Our findings show promise of this approach as a surveillance diagnostic, but also raise important cautionary questions. Several DBS specimens were detected as S. pneumoniae by rt-PCR despite growth of a non-S. pneumoniae pathogen on culture. A precise definition of what constitutes a positive result is required to avoid falsely over-identifying specimens.
This work was supported by Thrasher Research Fund (grant 12162 to PI), and the University of Minnesota Foundation. The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.Data AvailabilityAll relevant data are within the paper and its Supporting Information files.Introduction
Streptococcus pneumoniae, a commonly isolated bacterial pathogen in otitis media, can also cause invasive disease such as bacteremia, bacteremic pneumonia and meningitis, and is associated with high case fatality rates. Nigeria, the most populous nation in Africa, is one of the countries with the highest pneumococcal mortality rates in the world; in 2000, there were an estimated 86,000 deaths in children under 5 years of age, the second highest of any country worldwide [1]. However, invasive pneumococcal disease (IPD) is likely being underreported, based on the high prevalence of pneumococcal carriage in Nigerian infants [2] and reports of substantial pneumococcal disease burden in other parts of sub-Saharan Africa [3–5].
A major challenge in accurately determining the prevalence of IPD is the lack of resources necessary for obtaining and processing diagnostic specimens. In addition, the widespread use of non-prescription antibiotics prior to evaluation by a physician makes recovery of viable bacteria difficult in children who present for care [6, 7]. Dried blood spot (DBS) testing is an ideal method for diagnosing infections in resource-limited settings because of its low cost, minimal blood volumes involved, and capacity for storage and transport at ambient temperature. Furthermore, molecular methods such as polymerase chain reaction (PCR) have been able to identify pneumococcus in culture-negative specimens obtained from subjects that have been pre-treated with antibiotic therapy [7, 8].
Whatman 903 filter paper has been the standard for DBS used in newborn screening for decades [9]. More recently, newer materials have been introduced in which the cellulose matrix has been impregnated with surfactant, a chelating agent, buffer, and a free radical trap [10]. These Flinders Technology Associates (FTA) filter papers have the ability to lyse cells on contact, denature proteins, and protect DNA from degradation. Whatman FTA has been shown to lead to lower cycle thresholds (i.e. a lower limit of detection (LLD) on real-time (rt)-PCR) in cerebrospinal fluid (CSF) studies compared to Whatman 903 [11]. However, studies comparing the performance of these filter papers for subsequent detection of S. pneumoniae genes are lacking.
The aim of our study was to evaluate a S. pneumoniae rt-PCR assay on DBS using different types of filter paper among a cohort of febrile Nigerian children. We hypothesized that DBS can be used to detect S. pneumoniae with high sensitivity and specificity, compared to the gold standard of culture. Our aim was to evaluate DBS on Whatman 903 and FTA from a clinical cohort of febrile Nigerian children using rt-PCR and determine whether this method can more accurately detect S. pneumoniae compared to the gold standard of culture.
Materials and MethodsEthics statement
This study was approved by the Ethics Committees of the Federal Capital Territory, Federal Medical Center Keffi, and Aminu Kano Teaching Hospital, and by the University of Nebraska Medical Center and University of Minnesota Institutional Review Boards. Written signed informed consent was obtained from the parent or guardian.
Bacterial strains
For the laboratory validation, clinical isolates of S. pneumoniae used were collected as part of the Active Bacterial Core surveillance of the Centers for Disease Control and Prevention’s (CDC) Emerging Infections Program [12]. These reference strains were collected from culture-confirmed cases of meningitis, pneumonia, or bacteremia, and cover 12 serotypes (1, 3, 4, 5, 6A, 7F, 9V, 14, 18C, 19A, 19F, 23F) present in the 13-valent pneumococcal conjugate vaccine. Nine bacterial species, including Neisseria meningitidis, Haemophilus influenzae type e, H. influenzae indeterminate type, Streptococcus mitis, methicillin-resistant Staphylococcus aureus, and Group A, B and G Streptococcus, were used as negative controls. All isolates were provided in saline suspensions of 0.5 McFarland [13].
Samples and serial dilutions
The two papers selected for performance evaluation included the Whatman 903 and FTA classic cards (GE Healthcare Bio-Sciences, Pittsburgh, PA). S. pneumoniae isolates were suspended in phosphate buffered saline (PBS) to 0.5 McFarland, equivalent to 108 CFU/mL [13]. These isolates were spiked into whole blood collected in sodium citrate tubes, and ten-fold serial dilutions of whole blood were prepared ranging from 0.5-5x104 CFU/μL of blood. In addition, dilutions at 2 and 20 CFU/μL were prepared, based on previous studies identifying a detection limit in bacterial PCR within this range [14–16]. Hence, there were a total of 8 dilutions tested per serotype. Next, 100 μL from each dilution was spotted onto Whatman 903 and FTA filter paper until soaked through. Blood spots were allowed to dry overnight on the benchtop and then stored in sealed plastic bags at room temperature. We used as negative controls 9 strains of other bacterial species that are associated with invasive disease, as well as non-spiked samples of whole blood.
DNA extraction
For whole blood, 200 μL of S. pneumoniae isolate was added to 100 μL of 0.04 g/mL lysozyme and 75 U/mL mutanolysin (Sigma-Aldrich Corp, St. Louis, MO) in Tris EDTA (TE, pH 8.5) for 1 hour at 37°C. Next, 20 μL of proteinase K (Qiagen, Valencia, CA) was added and the specimen was incubated at 56°C for 30 minutes. Following this step, 200 μL of QiaAmp Lysis (AL) lysis buffer (Qiagen) was added and the specimen was incubated for 10 minutes at room temperature. DNA was extracted using the QiaAmp DNA Mini Kit (Qiagen) according to the manufacturer’s instructions [17].
For each filter paper, three punches of DBS were collected, with each set of 3 mm punches on DBS followed by 15 punches on clean filter paper to prevent contamination between samples. For Whatman 903 DBS, three punches were placed into a 1.5 mL microcentrifuge tube and 200 μL of TE was added. The punches were washed by shaking at 40 rpm at 56°C over 16–20 hours. After washing, DBS were processed as previously described for whole blood. For DNA extraction from Whatman FTA DBS, three different methods including a TE buffer extraction method [18], a methanol extraction method [19], and a FTA purification reagent extraction method were assessed as described in S1 Appendix.
Real-time PCR assay
Primers and probes previously developed by the CDC for detection of the lytA gene were used (S1 Table [17, 20]). Rt-PCR was performed using TaqMan Universal PCR master mix (Qiagen) and run on the LightCycler 480 (Roche Diagnostics, Indianapolis, IN). Each rt-PCR reaction consisted of a total volume of 23 μL, with 12.5 μL of the reconstituted master mix, and contained either 0.2 μM of primers and probe for lytA, or 0.32 μM primers and 0.08 μM probe for RNAse P. For rt-PCR assays of whole blood, 2 μL of the extracted DNA was used as template based on a standard protocol developed by the Minnesota Department of Health [21]. A template volume of 6.5 μL of extracted DNA was used for rt-PCR assays of DBS, in comparison to a template volume of 2 μL of extracted DNA from whole blood. This was done to correct for a smaller initial volume of blood used in DBS DNA extraction compared to whole blood. Rt-PCR reactions were cycled at 50°C for 2 minutes; 95°C for 10 minutes; and then 50 cycles at 95°C for 15 seconds and 60°C for 1 minute.
The LLD, defined as the lowest limit of detection in CFU/μL of blood, for each isolate was determined by the lowest dilution that was detectable on rt-PCR assay. DNA concentrations that yielded a cycle threshold (Ct) value of <40 on rt-PCR assay were defined as positive. Values ≥40 were considered equivocal and were repeated, and were considered positive if the repeat value was <40. The cutoff Ct value of 40 was chosen based on in vitro rt-PCR results on the spiked samples. For the laboratory validation, DBS samples were run in duplicate, and were counted as positive when both replicates tested positive. To demonstrate the presence and successful extraction of human cellular DNA and as a control for PCR inhibitors [22], rt-PCR assays were tested for the presence of the housekeeping gene RNase P on Whatman 903 and FTA DBS for a limited number of serotypes.
Clinical study sites and participants
Children 5 years of age or under who presented to six hospital study sites throughout northern and central Nigeria were prospectively screened from September 2011 through May 2015, and enrolled if they had a fever >38.5°C associated with difficulty breathing or altered consciousness. In addition, a subgroup of healthy neonatal controls were also enrolled. A detailed questionnaire was administered by a physician to obtain information on clinical history and physical examination findings. After informed consent was obtained, the skin was cleaned with alcohol swabs and 1–3 mL of blood was directly collected and inoculated into an aerobic Bactec (Becton Dickinson, Temse, Belgium) culture bottle and delivered to the regional microbiology laboratory within 4 hours of collection for processing. In addition, the finger was cleaned with an alcohol swab before being pricked, and blood was spotted onto either five spots (diameter 1.3 cm) on Whatman 903, or one spot (diameter 2.5 cm) on Whatman FTA. Whatman 903 was available for the majority of the study period at the study sites, however in the later period these were replaced by Whatman FTA. DBS were stored in a -80°C freezer until they were transported to the University of Minnesota for further processing.
Clinical laboratory methods
Aerobic blood culture bottles were held in the Bactec 9050 incubator (Becton Dickinson) for a maximum of 5 days. Positive cultures were subcultured using sheep blood, chocolate or McConkey agar. Subcultured plates were incubated under aerobic and 5% CO2 conditions at 35°C for 18–24 hours. S. pneumoniae isolates were identified by susceptibility to optochin and bile solubility [23].
For S. pneumoniae DNA amplification, each clinical specimen was run in triplicate for detection of lytA gene. Based on our validation studies, we defined a positive result for S. pneumoniae as a specimen that yielded a Ct value of <40 in 2/3 replicates by rt-PCR. Values ≥40 were considered equivocal and were repeated, and were considered positive if the repeat value was <40 in 2/3 replicates. Each specimen was tested for the presence of the housekeeping gene RNase P to demonstrate the presence and successful extraction of human cellular DNA [22].
Data analysis
All data management and analysis was performed using Microsoft Excel 2011 (v.14.4.5, Redmond, WA). Exact binomial distribution was used to calculate 95% confidence intervals. Chi-square test was performed, with statistical significance set at 5%.
ResultsHigh detection rate of pneumococcal DNA in spiked samples on both filter papers by rt-PCR assay
Amplification curves on rt-PCR for a representative dilution series demonstrated the integrity of the assay (S1 Fig). We detected pneumococcal DNA in 91/96 spiked samples (94.8% [95% CI 88.4–97.8%]). Five dilutions were not detected in whole blood by rt-PCR. Compared with spiked whole blood, pneumococcal DNA was detected in 83.5% (95% CI 74.6–89.8%) of Whatman FTA DBS, compared to detection in 81.3% (95% CI 72.1–88%) of 903 DBS. All of the non-pneumococcal samples were negative for S. pneumoniae DNA.
Comparable range of LLD for Whatman 903 DBS compared with FTA DBS
The LLD on Whatman 903 DBS ranged from 0.5 CFU/μL of blood for 1/12 serotypes to 20 CFU/μL for 3/12 serotypes (Table 1). On Whatman FTA DBS the LLD ranged from 0.5 CFU/μL for 2/12 serotypes to 20 CFU/μL for 1/12 serotypes. On spiked whole blood, the LLD ranged from 0.5 CFU/μL for 8/12 serotypes to 5 CFU/μL for 1/12 serotypes.
10.1371/journal.pone.0152253.t001
Comparison of LLD on whole blood versus Whatman 903 and FTA DBS.
A total of 537 DBS specimens from 535 children were included in the study. Complete data was available for 261 participants (73.1%; Table 2). Of these, the median age was 14 months with the largest cohort being those from 12–59 months (49.0%), and 53.3% were males. The most common clinical diagnoses on admission were malaria (31.8%) followed by sepsis (29.9%) and respiratory tract illness/pneumonia (24.1%).
Molecular detection of S. pneumoniae in clinical DBS compared to culture
Among the 537 clinical DBS specimens, including 481 Whatman 903 and 56 FTA, rt-PCR assay detected S. pneumoniae in 12 DBS specimens (2.2%). Five DBS were on Whatman 903 and seven were FTA. Nonspecific amplification was noted in an additional 22 DBS (4.1%).
Culture data were available for 436 isolates (81.2%; S2 Table). The majority of blood cultures had no growth (61.6%), and 15 were culture-positive for S. pneumoniae (3.4%). Among these culture-positive samples, S. pneumoniae was detected by rt-PCR in 9 DBS (60%), and nonspecific amplification was noted in one additional DBS that did not meet criteria for a positive result. Only two patients had blood collected on both Whatman 903 and FTA, and in each case both were culture-positive for S. pneumoniae and both types of DBS were positive by rt-PCR.
S. pneumoniae was detected by rt-PCR in an additional three specimens. One was a Whatman FTA DBS from a high-risk subject whose blood culture had no growth. On repeat rt-PCR testing the specimen again had Ct values <40 in 3/3 replicates. However, the other two positives were from a healthy neonatal control, and from a specimen that was culture-positive for H. influenzae type b. Both of these latter two specimens on repeat rt-PCR testing were negative.
For the other 22 DBS (64.7%) where S. pneumoniae was detected by rt-PCR but did not meet criteria for a positive result, two DBS detected S. pneumoniae in 2/3 replicates, but both of those Ct values were not <40. Furthermore, both of these clinical DBS were positive by culture for other bacteria (S. aureus and coagulase-negative Staphylococcus). S. pneumoniae was detected in 1/3 replicates of 20 DBS by rt-PCR, but only 11 had Ct values <40. Of these 11, one had no culture data, 8 had no growth on culture, and two were positive for Salmonella Typhi on culture.
The overall sensitivity of DBS was 60% (95% CI 32.3–83.7%) and specificity was 99.4% (95% CI 98.3–99.9%; Table 3). Both Whatman 903 and FTA showed similar performance for detection of S. pneumoniae from clinical DBS, with a sensitivity of 57.1% (95% CI 18.4–90.1) and specificity of 99.8% (95% CI 98.8–100) for Whatman 903, and sensitivity of 62.5% (95% CI 24.5–91.5) and specificity of 95.8% (95% CI 85.7–99.5%) for Whatman FTA.
10.1371/journal.pone.0152253.t003
Sensitivity and specificity for blood culture and real-time PCR assays with Ct values.
aEither Whatman 903 or FTA was available at the field site; subjects did not have DBS collected on both.
bS. pneumoniae PCR-positive detection was defined as lytA gene detected by rt-PCR with Ct value <40 in 2/3 replicates.
cClinical DBS specimen was positive for H. influenzae type b on culture. The specimen was negative on repeat rt-PCR testing.
dOne subject was a high-risk febrile patient, the second subject was a healthy neonatal control. On repeat rt-PCR testing, the first specimen was positive (Ct <40 in 3/3 replicates), the second specimen was negative.
Discussion
This study provides the first description of molecular methods on DBS in a clinical cohort of febrile Nigerian children to detect bacterial pathogens. As all specimens were paired and tested by culture and rt-PCR, we were able to compare their performance. We were able to detect S. pneumoniae in DBS by rt-PCR, with an overall sensitivity of 60% (95% CI 32.3–83.7%) and specificity of 99.4% (95% CI 98.3–99.9%) compared to culture. The disappointing sensitivity of both filter papers (57.1% for Whatman 903 and 62.5% for FTA) for the detection of pneumococcal DNA was most likely due to the small amounts of whole blood captured in filter paper. Nonetheless, rt-PCR led to the identification of an additional S. pneumoniae specimen from a Whatman FTA DBS in a high-risk subject.
Our results, in particular the identification of S. pneumoniae in a culture-negative specimen, support the promise of molecular diagnostics. Molecular methods, including nucleic acid amplification tests, have the ability to detect minute amounts of nucleic acid of a pathogen, and studies using PCR to detect pneumococcal meningitis have shown sensitivity and specificity rates of 92–100% and 100% compared to the gold standard of culture, respectively [24, 25]. The method also does not require a viable microbe, and hence is less likely to be affected by prior antimicrobial therapy than culture-based methods. This was illustrated in Brazil, where rt-PCR was incorporated into routine public health surveillance of bacterial meningitis. Researchers found that using this method increased the detection of bacteria by up to 85% [8]. In bacteremic pneumonia, a study evaluating DBS PCR identified a three-fold increase in pneumococcal-positive specimens [26].
Nonspecific amplification was noted in almost 5% of clinical DBS, including DBS that were culture-positive for other bacteria, such as S. aureus, coagulase-negative Staphylococcus, alpha-hemolytic Streptococcus and Salmonella Typhi. When we compared Ct values of these DBS that were S. pneumoniae PCR-positive but culture-positive for non-S. pneumoniae specimens (average Ct 40.47, range 39.32–42.33), there was no difference from the Ct values of S. pneumoniae PCR-positive but culture-negative DBS specimens (average Ct 40.36, range 38.85–45). This contrasts with S. pneumoniae PCR- and culture-positive DBS specimens where the average Ct was lower at 35.77 (range 33.16–39.31). For two DBS that did meet criteria for a positive result, where S. pneumoniae was detected in 2/3 replicates with a Ct value of <40, one was from a normal healthy control and one was from a subject who was culture-positive for H. influenzae type b. Both were negative on repeat testing.
These issues illustrate the difficulties with intra- and inter-user reliability as well as with interpretation of a positive test. Our definition of a positive requiring 2/3 replicates to have Ct counts <40 was based on our validation studies and was potentially too strict, compared to previous studies utilizing molecular methods [26–29], and thereby may have limited the number of specimens deemed positive by molecular detection. Our documented LLD for both types of filter paper corresponded with Ct values <40. As these LLDs fell within the range of bacterial counts in bacteremia [30], we hypothesized that rt-PCR assays described in this study should be able to detect the target pathogen from bacteremic specimens in DBS with typical bacterial loads. However, high Ct values can be more problematic in that they can be the result of nonspecific amplification, can be associated with higher rates of false positives due to poor replicate reproducibility, and can result from the presence of PCR inhibitors or target DNA degradation, or low DNA quantity due to suboptimal transport and storage conditions or antimicrobial treatment prior to specimen collection [31]. In addition, high Ct values have been associated with misclassification of pneumococcal specimens in a pneumococcal surveillance program in South Africa, particularly when the Ct values were ≥35 [32]. As mean Ct results in our study among our S. pneumoniae PCR- and culture-positive specimens extended up to 39.31, we may have missed more culture-positive specimens if we had lowered our Ct cutoff to below 40. Conversely, if we had relaxed our positive criteria to include detection in 1/3 replicates, sensitivity would have only increased minimally from 60% to 66.7%.
While initial studies described the substantially increased sensitivity of molecular diagnostics in detecting bacterial pathogens [29], there is also growing evidence that molecular diagnostics is not without its own pitfalls. A systematic review and meta-analysis of the use of blood PCR in the diagnosis of IPD found that the sensitivity and specificity of molecular testing increased as the reference standard relied more on using any culture, serology or clinical impression of the subject for diagnosis of IPD [33]. Thus, as the reference standard became more uncertain, the better PCR performed [33]. This raises the question of whether molecular diagnostics is sufficiently sensitive or specific to replace the reference standard in clinical practice.
The limitations of this study included the small number of S. pneumoniae-culture positive specimens in our clinical DBS collection, as well as the availability of either Whatman 903 or FTA at the field site, but rarely both, which restricted our ability to directly compare the detection sensitivity of the two filter papers. The two patients who had blood collected on both Whatman 903 and FTA had pneumococcal DNA successfully detected from both types of filter paper. The DNA extraction methods for Whatman 903 and FTA were disparate, and hence these varied methodologies may explain the differences in sensitivity observed. In addition, the prolonged storage of some of the DBS specimens may have affected the DNA content [34]. While we used a limited number of Gram positive and Gram negative pathogens as negative controls, we did not survey a diverse array of bacteria in our validation of spiked samples and hence cross-reactivity of the lytA gene with other bacteria is a possibility.
In conclusion, compared to the gold standard of culture, rt-PCR did identify one culture-negative sample on Whatman FTA DBS from a high-risk subject. Hence, there is promise in this surveillance methodology, which presents a potentially useful epidemiological tool in settings where culture is not available. However, nonspecific amplification for S. pneumoniae was documented in several clinical samples, including specimens positive for non-S. pneumoniae on culture. With better delineation of results deemed positive by molecular detection, we could improve our acquisition of disease epidemiology in future studies for resource-limited regions such as Nigeria to monitor pneumococcal disease and the impact of pneumococcal vaccination.
Supporting Information
DNA extraction from Whatman FTA paper.
(DOCX)
Amplification curves for real-time PCR, at 10-fold dilutions from 5x104 CFU/μL to 0.5 CFU/μL for serotype 5 in whole blood.
(DOCX)
Primer and probe information for real-time PCR detection.
(DOCX)
Summary of results of S. pneumoniae detection in 537 clinical dried blood spot specimens.
(DOCX)
We are indebted to David Boxrud, Ginette Dobbins and the Minnesota Department of Health for providing pneumococcal isolates and technical guidance. We thank Shane McAllister for reviewing manuscript drafts, and Kyle Rudser for statistical assistance. We thank GE Healthcare Life Sciences for donating Whatman FTA cards and reagents for this study. This work was presented at the Pediatric Academic Societies meeting in San Diego in April 2015, and ID Week in San Diego in October 2015.
AbbreviationsCSF
cerebrospinal fluid
DBS
dried blood spot
FTA
Flinders Technology Associates
LLD
lower limit of detection
IPD
invasive pneumococcal disease
rt-PCR
real-time polymerase chain reaction
ReferencesO'BrienKL, WolfsonLJ, WattJP, HenkleE, Deloria-KnollM, McCallN, et al. Burden of disease caused by Streptococcus pneumoniae in children younger than 5 years: global estimates. . 2009;374(9693):893–902. Epub 2009/09/15. doi: 10.1016/S0140-6736(09)61204-619748398.AdetifaIM, AntonioM, OkoromahCA, EbrukeC, InemV, NsekpongD, et al. Pre-vaccination nasopharyngeal pneumococcal carriage in a Nigerian population: epidemiology and population biology. . 2012;7(1):e30548. Epub 2012/02/01. doi: 10.1371/journal.pone.003054822291984; PubMed Central PMCID: PMC3265474.BerkleyJA, LoweBS, MwangiI, WilliamsT, BauniE, MwarumbaS, et al. Bacteremia among children admitted to a rural hospital in Kenya. . 2005;352(1):39–47. Epub 2005/01/07. doi: 10.1056/NEJMoa04027515635111.EnwereG, BineyE, CheungYB, ZamanSM, OkokoB, OluwalanaC, et al. Epidemiologic and clinical characteristics of community-acquired invasive bacterial infections in children aged 2–29 months in The Gambia. . 2006;25(8):700–5. Epub 2006/07/29. doi: 10.1097/01.inf.0000226839.30925.a516874169.SigauqueB, RocaA, MandomandoI, MoraisL, QuintoL, SacarlalJ, et al. Community-acquired bacteremia among children admitted to a rural hospital in Mozambique. . 2009;28(2):108–13. Epub 2009/01/10. doi: 10.1097/INF.0b013e318187a87d19131902.ObaroS, LawsonL, EssenU, IbrahimK, BrooksK, OtuneyeA, et al. Community acquired bacteremia in young children from central Nigeria—a pilot study. . 2011;11:137. Epub 2011/05/21. doi: 10.1186/1471-2334-11-13721595963; PubMed Central PMCID: PMC3111365.TarragoD, FenollA, Sanchez-TatayD, ArroyoLA, Munoz-AlmagroC, EstevaC, et al. Identification of pneumococcal serotypes from culture-negative clinical specimens by novel real-time PCR. . 2008;14(9):828–34. Epub 2008/10/11. doi: 10.1111/j.1469-0691.2008.02028.x18844683.SacchiCT, FukasawaLO, GoncalvesMG, SalgadoMM, ShuttKA, CarvalhanasTR, et al. Incorporation of real-time PCR into routine public health surveillance of culture negative bacterial meningitis in Sao Paulo, Brazil. . 2011;6(6):e20675. Epub 2011/07/07. doi: 10.1371/journal.pone.002067521731621; PubMed Central PMCID: PMC3120771.GuthrieR, SusiA. A Simple Phenylalanine Method for Detecting Phenylketonuria in Large Populations of Newborn Infants. . 1963;32:338–43. Epub 1963/09/01. 14063511.Flinders Technology Associates, inventor. Whatman FTA (Flinders Techology Associates) Cards. GE Health Care. US Patent No. 5985327.ElliottI, DittrichS, ParisD, SengduanphachanhA, PhouminP, NewtonPN. The use of dried cerebrospinal fluid filter paper spots as a substrate for PCR diagnosis of the aetiology of bacterial meningitis in the Lao PDR. . 2013;19(10):E466–72. Epub 2013/06/07. doi: 10.1111/1469-0691.1226023738720.Centers for Disease Control and Prevention. Active Bacterial Core Surveillance Report, Emerging Infections Program Network, Streptococcus pneumoniae, 2014.SuttonS. Measurement of microbial cells by optical density. . 2011;17:46–9.RothmanR, RamachandranP, YangS, HardickA, WonH, KecojevicA, et al. Use of quantitative broad-based polymerase chain reaction for detection and identification of common bacterial pathogens in cerebrospinal fluid. . 2010;17(7):741–7. Epub 2010/07/27. doi: 10.1111/j.1553-2712.2010.00790.x20653589; PubMed Central PMCID: PMC3689214.WroblewskiD, HalseTA, HayesJ, KohlerschmidtD, MusserKA. Utilization of a real-time PCR approach for Haemophilus influenzae serotype determination as an alternative to the slide agglutination test. . 2013;27(2):86–9. Epub 2012/12/01. doi: 10.1016/j.mcp.2012.11.00323195602.ZhuH, WangQ, WenL, XuJ, ShaoZ, ChenM, et al. Development of a multiplex PCR assay for detection and genogrouping of Neisseria meningitidis. . 2012;50(1):46–51. Epub 2011/11/18. doi: 10.1128/JCM.00918-1122090406; PubMed Central PMCID: PMC3256684.CarvalhoMG, TondellaML, McCaustlandK, WeidlichL, McGeeL, MayerLW, et al. Evaluation and improvement of real-time PCR assays targeting lytA, ply, and psaA genes for detection of pneumococcal DNA. . 2007;45(8):2460–6. Epub 2007/06/01. doi: 10.1128/JCM.02498-0617537936; PubMed Central PMCID: PMC1951257.ElbadryMA, ExisteA, VictorYS, MemnonG, FukudaM, DameJB, et al. Survey of Plasmodium falciparum multidrug resistance-1 and chloroquine resistance transporter alleles in Haiti. . 2013;12:426. Epub 2013/11/21. doi: 10.1186/1475-2875-12-42624252305; PubMed Central PMCID: PMC3879102.BereczkyS, MartenssonA, GilJP, FarnertA. Short report: Rapid DNA extraction from archive blood spots on filter paper for genotyping of Plasmodium falciparum. . 2005;72(3):249–51. Epub 2005/03/18. 15772315.Centers for Disease Control and Prevention. Real-time PCR targets for Streptococcus species detection and antibiotic susceptibility in S. pneumoniae [cited 8 February 2016]. Available from: http://www.cdc.gov/streplab/downloads/pcr-realtime-target.pdf.Minnesota Department of Health. TaqMan PCR assay for detection of Neisseria meningitidis, Streptococcus pneumoniae, and Haemophilus influenzae. ME 337 Bacterial meningitis panel, 2013.StoddardRA, GeeJE, WilkinsPP, McCaustlandK, HoffmasterAR. Detection of pathogenic Leptospira spp. through TaqMan polymerase chain reaction targeting the LipL32 gene. . 2009;64(3):247–55. Epub 2009/04/28. doi: 10.1016/j.diagmicrobio.2009.03.01419395218.WernoAM, MurdochDR. Medical microbiology: laboratory diagnosis of invasive pneumococcal disease. . 2008;46(6):926–32. Epub 2008/02/12. doi: 10.1086/52879818260752.CorlessCE, GuiverM, BorrowR, Edwards-JonesV, FoxAJ, KaczmarskiEB. Simultaneous detection of Neisseria meningitidis, Haemophilus influenzae, and Streptococcus pneumoniae in suspected cases of meningitis and septicemia using real-time PCR. . 2001;39(4):1553–8. Epub 2001/04/03. doi: 10.1128/JCM.39.4.1553-1558.200111283086; PubMed Central PMCID: PMC87969.TzanakakiG, TsopanomichalouM, KesanopoulosK, MatzouraniR, SioumalaM, TabakiA, et al. Simultaneous single-tube PCR assay for the detection of Neisseria meningitidis, Haemophilus influenzae type b and Streptococcus pneumoniae. . 2005;11(5):386–90. Epub 2005/04/12. doi: 10.1111/j.1469-0691.2005.01109.x15819865.SelvaL, BenmessaoudR, LanaspaM, JroundiI, MoraledaC, AcacioS, et al. Detection of Streptococcus pneumoniae and Haemophilus influenzae type B by real-time PCR from dried blood spot samples among children with pneumonia: a useful approach for developing countries. . 2013;8(10):e76970. Epub 2013/10/12. doi: 10.1371/journal.pone.007697024116190; PubMed Central PMCID: PMC3792875.LundgrenIS, HeltsheSL, SmithAL, ChibwanaJ, FriedMW, DuffyPE. Bacteremia and malaria in Tanzanian children hospitalized for acute febrile illness. . 2015;61(2):81–5. Epub 2014/12/17. doi: 10.1093/tropej/fmu06925505140; PubMed Central PMCID: PMC4402358.NhantumboAA, CantarelliVV, CaireaoJ, MunguambeAM, ComeCE, Pinto GdoC, et al. Frequency of Pathogenic Paediatric Bacterial Meningitis in Mozambique: The Critical Role of Multiplex Real-Time Polymerase Chain Reaction to Estimate the Burden of Disease. . 2015;10(9):e0138249. Epub 2015/09/24. doi: 10.1371/journal.pone.013824926393933.SelvaL, KrauelX, PallaresR, Munoz-AlmagroC. Easy diagnosis of invasive pneumococcal disease. . 2011;17(6):1125–7. Epub 2011/07/14. doi: 10.3201/eid/1706.10099721749788; PubMed Central PMCID: PMC3358190.SantoshamM, MoxonER. Detection and quantitation of bacteremia in childhood. . 1977;91(5):719–21. Epub 1977/11/01. 333075.WangX, TheodoreMJ, MairR, Trujillo-LopezE, du PlessisM, WolterN, et al. Clinical validation of multiplex real-time PCR assays for detection of bacterial meningitis pathogens. . 2012;50(3):702–8. Epub 2011/12/16. doi: 10.1128/JCM.06087-1122170919; PubMed Central PMCID: PMC3295090.TempiaS, WolterN, CohenC, WalazaS, von MollendorfC, CohenAL, et al. Assessing the impact of pneumococcal conjugate vaccines on invasive pneumococcal disease using polymerase chain reaction-based surveillance: an experience from South Africa. . 2015;15:450. Epub 2015/10/27. doi: 10.1186/s12879-015-1198-z26496761; PubMed Central PMCID: PMC4620746.AvniT, MansurN, LeiboviciL, PaulM. PCR using blood for diagnosis of invasive pneumococcal disease: systematic review and meta-analysis. . 2010;48(2):489–96. Epub 2009/12/17. doi: 10.1128/JCM.01636-0920007385; PubMed Central PMCID: PMC2815606.SigurdsonAJ, HaM, CosentinoM, FranklinT, HaqueKA, QiY, et al. Long-term storage and recovery of buccal cell DNA from treated cards. . 2006;15(2):385–8. Epub 2006/02/24. doi: 10.1158/1055-9965.EPI-05-066216492933.