Oral bacterial DNA findings in pericardial fluid.

BACKGROUND: We recently reported that large amounts of oral bacterial DNA can be found in thrombus aspirates of myocardial infarction patients. Some case reports describe bacterial findings in pericardial fluid, mostly done with conventional culturing and a few with PCR; in purulent pericarditis, nevertheless, bacterial PCR has not been used as a diagnostic method before.
OBJECTIVE: To find out whether bacterial DNA can be measured in the pericardial fluid and if it correlates with pathologic-anatomic findings linked to cardiovascular diseases.
METHODS: Twenty-two pericardial aspirates were collected aseptically prior to forensic autopsy at Tampere University Hospital during 2009-2010. Of the autopsies, 10 (45.5%) were free of coronary artery disease (CAD), 7 (31.8%) had mild and 5 (22.7%) had severe CAD. Bacterial DNA amounts were determined using real-time quantitative PCR with specific primers and probes for all bacterial strains associated with endodontic disease (Streptococcus mitis group, Streptococcus anginosus group, Staphylococcus aureus/Staphylococcus epidermidis, Prevotella intermedia, Parvimonas micra) and periodontal disease (Aggregatibacter actinomycetemcomitans, Porphyromonas gingivalis, Treponema denticola, Fusobacterium nucleatus, and Dialister pneumosintes).
RESULTS: Of 22 cases, 14 (63.6%) were positive for endodontic and 8 (36.4%) for periodontal-disease-associated bacteria. Only one case was positive for bacterial culturing. There was a statistically significant association between the relative amount of bacterial DNA in the pericardial fluid and the severity of CAD (p=0.035).
CONCLUSIONS: Oral bacterial DNA was detectable in pericardial fluid and an association between the severity of CAD and the total amount of bacterial DNA in pericardial fluid was found, suggesting that this kind of measurement might be useful for clinical purposes.

The pericardium has a complex role in the normal performance of the heart and pericardial disease is an important cause of morbidity in patients with cardiovascular disease (1). In clinical practice, both pericarditis and myocarditis may coexist because they share common etiologic agents, although in most cases they remain idiopathic. Myopericarditis caused by special inflammatory agents is less frequently reported when cardiotropic viruses are thought to be substantively prevalent, and bacteria or fungi uncommon (2). The diagnosis of pericardial inflammation frequently remains clinically challenging. The true incidence and prevalence of pericarditis are difficult to measure, and according to autopsy studies pericarditis may be frequently subclinical (3).

It is well known that there is an association between cardiovascular diseases and dental infections. Many studies have shown that oral bacterial DNA can be found in atherosclerotic plaques but the role of these bacteria is still under debate (4–6). Recently, we found bacterial DNA, mainly from oral streptococci, in thrombi aspirated from the coronary arteries of patients who had had an acute myocardial infarction and had undergone percutaneous coronary intervention (7).

The aim of this study was to investigate whether oral bacterial DNA can be detected in the pericardial fluid using real-time quantitative polymerase chain reaction (RT-qPCR) and if there is a relationship with certain pathologic–anatomical (PAD) parameters associated with cardiovascular diseases. To the best of our knowledge, there are no previously published studies regarding the amount of bacterial DNA in the pericardial fluid.

Material and methods

Study cases and sample collection

The study material was composed of 22 pericardial aspirations (1 ml) collected from the Department of Forensic Medicine at The University of Tampere during 2009–2010. The samples were collected aseptically as described earlier (8) and clinical characteristics, demonstrated in Table 1, were obtained from post-mortem and microscopic reports.

Table 1

Characteristics of the autopsy cases (n=22)

Age at death, y, mean±SD	52.9±15.1
BMI, kg/m2, mean±SD	27.9±6.41
Post-mortem time, h, mean±SD	88.4±30.88
Heart weight, g, mean±SD	499±141
Coronary artery disease (CAD)
No CAD, n (%)	10 (45.5%)
Mild CAD, n (%)	7 (31.8%)
Serious CAD, n (%)	5 (22.7%)
Cardiovascular death (CVD), n (%)	8 (36.4%)
Myofibrosis (MF)
No MF, n (%)	16 (72.7%)
Mild/disseminated MF, n (%)	4 (18.2%)
Extensive MF, n (%)	2 (9.1%)
Unstable plaque of the coronary arteries, n (%)	2 (9.1%)
Calcification of the coronary arteries, n (%)	9 (40.9%)

Bacterial analyses

Bacterial culturing from samples was performed in an accredited laboratory using standard protocols at the Department of Microbiology, Fimlab Laboratories Ltd, Pirkanmaa Hospital District, Tampere, Finland. Briefly, 10 µl aliquots of pericardial fluid were streaked on chocolate, 5% horse blood, MacConkey and fastidious anaerobe (FAA) agar plates, and into fastidious anaerobe broth (FAB) tubes. Plates were incubated at 35°C for 24–48 hours in aerobic (chocolate, blood, and MacConkey) and 48 hours in anaerobic (FAA) conditions. FAB tubes were incubated at 35°C for 10 days. Species identification was obtained by conventional identification methods (including gram staining, specific tests and API tests, and susceptibility testing) according to standard procedures. The amount of total bacterial DNA and the presence of candidate bacterial DNA for endodontic-disease-associated bacteria (Streptococcus sp. mainly S. mitis group, S. mitis & S. oralis, S. sanguinis & S. gordonii, S. anginosus group, Staphylococcus aureus & Staphylococcus epidermidis, Parvimonas micra and Prevotella intermedia) and periodontal-disease-associated bacteria (Porphyromonas gingivalis, Aggregatibacter (née Actinobacillus) actinomycetemcomitans, Fusobacterium nucleatum, Dialister pneumosintes, and Treponema denticola) were determined using real-time quantitative PCR as earlier described (7, 8) (Table 2 and Supplementary file). Human DNA determined by the TaqMan RNase P Detection Reagents–kit (Applied Biosystems Ltd, Foster City, CA) were used as references and venous blood from a healthy person was used as a negative control. Total amounts of human and bacterial DNA were analyzed using a standard curve, for which a dilution series was made from both human and Escherichia coli DNA. One case found to be both culture and qPCR positive for streptococci was excluded from association analyses.

Table 2

Primers and probes used

Primer and probe	Sequence (5′→3′)	Reference
Streptococcus sp. mainly S. mitis groupa		(8)
Forward	CCAGCAGCCGCGGTAATA
Reverse	CCTGCGCTCGCTTTACG
Probe	ACGCTCGGGACCTACG
S. mitis and S. oralis		(9)
Forward	GCCATTGAAGCGGTTACTTTG
Reverse	CATCCGACATTAACGCAAGTTC
Probe	ATGATTGAGCGTGGAACGGTGGGT
Streptococcal virulence factor gftG/gtfPb		(7)
Forward	TTTACCATGGATGAGCTCAAGCAA
Reverse	TTGGAGAGCATGAGAGCATG
Probe	ACGCAGTTCAATATCC
S. anginosus groupc		(7)
Forward	CCGTAGTGTTTGTGCTAGGTGAAA
Reverse	CCAGAGACGTAGCTGTTTCGT
Probe	CCGTAACGATTTCTCG
Staphylococcus aureus and S. epidermidis		(8)
Forward	GCGTTTTTCACGTGGAATATC
Reverse	AATCCAAAACACAAACAAAGACAAGGT
Probe	ACGTGCCATATTAATTTAC
Chlamydophila (née Chlamydia) pneumoniae		(10)
Forward	GATCCGCTGCTGCAAACTATACT
Reverse	GTGAACCACTCTGCATCGTGTAA
Probe	TAGGCCGGGTTAGGTCTATCTACGGCAGT
Treponema denticola		(11)
Forward	CCGAATGTGCTCATTTACATAAAGGT
Reverse	GATACCCATCGTTGCCTTGGT
Probe	ATGGGCCCGCGTCCCATTAGC
Prevotella intermedia		(11)
Forward	TCCACCGATGAATCTTTGGTC
Reverse	ATCCAACCTTCCCTCCACTC
Probe	CGTCAGATGCCATATGTGGACAACATCG
Fusobacterium nucleatum		(7)
Forward	AGGGTGAACGGCCACAAG
Reverse	TCTCGGTCCATTGTCCAATATTCC
Probe	ACACGGCCCTTACTCC
Aggregatibacter (née Actinobacillus) actinomycetemcomitans		(12)
Forward	CAAGTGTGATTAGGTAGTTGGTGG G
Reverse	CCTTCCTCATCACCGAAAGAA
Probe	ATCGCTAGCTGGTCTGAGAGGATGGCC
Dialister pneumosintes		(12)
Forward	GAGGGGTTTGCGACTGATTA
Reverse	CCGTCAGAC TTT CGTCCATT
Probe	CACCAAGCCGACGATCAGTAGCCG
Porphyromonas gingivalis		(12)
Forward	TGCAACTTGCCTTACAGAGGG
Reverse	ACTCGTATCGCCCGTTATTC
Probe	CACGAGCTGACGACA[A/G]CCATGCA
Parvimonas micra		(12)
Forward	AAACGACGATTAATACCACATGAGAC
Reverse	ACTGCTGCCTCCCGTAGGA
Probe	TCAAAGATTTATCGGTGTAAGAAGGGCTCGC
Universald		(12)
Forward	TGGAGCATGTGGTTTAATTCGA
Reverse	TGCGGGACTTAACCCAACA
Probe	CACGAGCTGACGACA[A/G]CCATGCA

Recognition of S. mitis-group (S. mitts, S. oralis, S. gordonii, S. sanguinis, S. pneumonia), S. salivarius, S. thermophilus, uncultured streptococci, Lactobacillus lactis.

Recognition of virulence factor for S. sanguinis (gftP) and S. gordonii (gftG).

Recognition of S. anginosus-group (S. anginosus, S. milleri, S. constellatus, S. intermedius).

Recognition of total amount of bacterial DNA.

Ethics

The study was approved by the Ethics Committee of the Pirkanmaa Hospital District, Finland, and the National Supervisory Authority for Welfare and Health (VALVIRA), Finland.

Statistical analyses

Frequencies of the variables were calculated and statistical analyses performed with PASW Statistical Software version 18 (SPSS Ltd, Quarry Bay, Hong Kong). Due to the skewed distribution of the total amount of bacterial and human DNA, they were converted into logarithmic values. One-way ANOVA was used to calculate the associations between bacterial findings and clinical parameters.

Results and discussion

Figure 1 shows the relative amounts of total bacterial DNA measured in the pericardium compared to control peripheral blood, divided into three subgroups of coronary artery disease (CAD). There was a statistically significant association (p=0.035, ANOVA) between the amount of bacterial DNA in the pericardial fluid and the stage of CAD. The standard curve shows that the median value for the total amount of bacteriall DNA was highest in patients with serious disease (3.03 pg/ml, 25th–75th percentile, 1.83–5.11 pg/ml). There was no statistically significant relationship between other PAD parameters and the amount of bacterial DNA in pericardial samples. The samples with one or more measurements of bacterial DNA findings were determined as positive. Of the 22 cases, 14 (63.6%) were positive for endodontic and 8 (36.4%) for periodontal-disease-associated bacteria (Fig. 2). The most frequent bacterial DNA finding was from the S. mitis group and Dialister pneumosintes.

Fig. 1

Association between the stage of coronary artery disease and relative amount of bacterial DNA in pericardial fluid (post hoc ANOVA). The boxes indicate lower and upper quartiles and each central line is median. The upper end of each bar is maximum value and the lower end is minimum value.

Fig. 2

Frequencies of bacterial-DNA-positive findings in pericardial fluid using specific primers and probes in RT-qPCR. Venous blood from a healthy subject was used as reference.

Bacteriological diagnosis is mainly based on conventional culturing. However, it is well known that some bacteria are poorly cultivable or non-cultivable, and inhibitory effects, like concurrent antibiotic therapy or the formation of a bacterial biofilm, may interfere with results (13). Many studies have already shown molecular methods to be dependable and practical in research use (7, 8, 13–15). Therefore, they have been proposed as an additional diagnostic tool, particularly when conventional bacterial culturing remains negative. By using different PCR methods, DNA from oral bacteria has been detected in atherosclerotic lesions (5, 16) and thrombus aspirates (7, 17). It is also known that oral bacteria such as streptococci have the capability to invade human heart endothelial cells in vitro (18). Similar concentrations of macromolecules have been measured in plasma as well as in pericardial fluid, suggesting active transport through the vascular wall (19). Since we have now found oral bacterial DNA in pericardial fluid, we suggest that in a similar manner, oral bacteria might be transported through the myocardium into the pericardial fluid.

This study material is limited and based on post-mortem samples. However, in our previous study we showed that the pericardial fluid is one of the most reliable post-mortem fluids for microbiological sampling (8). In terms of clinical application, more studies with antemortem material must be performed to evaluate the possible utility of pericardial fluid samples in patient care and diagnostics. Furthermore, the clinical use and relevance need to be debated after larger studies have been undertaken. Pericardial fluid aspirates are generally taken in cases where there is pericardial fluid effusion with an uncertain etiology. This is usually accompanied with endomyocardial biopsy (EMB), which highly increases the risk of complications (20). Nevertheless, analysis is usually restricted to histology, cytology, and conventional culturing (21). We believe that RT-qPCR for pericardial aspirates could be a sensible diagnostic tool and offer substantive additional value.

Click here for additional data file.