1360
Original article
Association of markers of bacterial translocation with immune activation in decompensated cirrhosis Christian Mortensena,b, Jørgen Skov Jensend, Lise Hobolthe, Sanne Dam-Larsenf, Bjørn S. Madseng, Ove Andersenc, Søren Møllerb and Flemming Bendtsena Background Bacterial translocation (BT) may cause infections, in particular, spontaneous bacterial peritonitis (SBP). In the absence of overt infection, BT may further stimulate the immune system and contribute to haemodynamic alterations and complications. Bacterial DNA (bDNA) is claimed to be a promising surrogate marker for BT, although its clinical relevance has been questioned. Materials and methods In 38 cirrhotic patients with and without SBP, bDNA in blood and ascites were assessed by 16S rDNA quantitative PCR. Levels of lipopolysaccharidebinding protein in plasma and highly sensitive C-reactive protein, tumour necrosis factor-α, soluble urokinase plasminogen activating receptor, interleukin-6, interleukin 8, interferon-γ inducible protein-10 and vascular endothelial growth factor in plasma and ascites were measured by multiplex cytokine and ELISA assays. Results In patients without signs of SBP or positive cultures, we found a high frequency of bDNA but low concordance of bDNA between blood and ascites. Markers of inflammation were not significantly different between blood bDNA-positive (22%), ascites bDNA-positive (52%), and bDNA-negative patients. The 16S rDNA PCR failed to
Introduction Impairment of the immune system in patients with cirrhosis leads to frequent lethal infections [1,2]. In patients with decompensated cirrhosis, development of spontaneous bacterial peritonitis (SBP) is associated with increased mortality [3,4]. Bacterial translocation (BT) refers to the passage of gut bacteria or bacterial products into extraintestinal lymph nodes [5], and it is thought to progress to SBP through viable gut bacteria colonizing the ascitic fluid. SBP is associated with high recurrence rates, and apart from high mortality, it is often accompanied by the development of hepatorenal syndrome [6]. Recently, it has been hypothesized that BT, even in the absence of clinical infection, may impact the course of the disease in patients with cirrhosis. Several studies, using different markers of BT have reported significant associations with the risk for infection and haemodynamic changes [7–11], which lends support to the hypothesis of an inappropriate activation of the immune 0954-691X © 2014 Wolters Kluwer Health | Lippincott Williams & Wilkins
show bDNA in two out of six samples with SBP. Sequencing of positive samples did not determine the source of bDNA. Conclusion bDNA as assessed by this PCR method was largely unrelated to markers of inflammation and does not seem to be of clinical value in the diagnosis of SBP. According to our results, bDNA is not a reliable marker of BT. Eur J Gastroenterol Hepatol 26:1360–1366 © 2014 Wolters Kluwer Health | Lippincott Williams & Wilkins. European Journal of Gastroenterology & Hepatology 2014, 26:1360–1366 Keywords: bacterial translocation, cirrhosis, inflammation, PCR Departments of aGastroenterology, bClinical Physiology and Nuclear Medicine, Centre of Functional Imaging and Research, cCentre for Clinical Research, Hvidovre University Hospital, Hvidovre, dMycoplasma Laboratory, Microbiology and Infection Control, Statens Serum Institut, eDepartment of Internal Medicine I, Bispebjerg Hospital, fDepartment of Internal Medicine, Glostrup Hospital, Glostrup and gDepartment of Medicine, Svendborg Sygehus, Svendborg, Denmark Correspondence to Christian Mortensen, MD, Department of Gastroenterology, Hvidovre University Hospital, Kettegaard Alle 30, DK-2650 Hvidovre, Denmark Tel: + 45 38 623 181; fax: + 45 3 8623 777; e-mail: [email protected] Received 26 June 2014 Accepted 29 August 2014
system on BT. This activation may further have adverse effects on circulation in cirrhosis by the release of systemic or local vasodilators, finally leading to secondary complications, even in the absence of clinically overt infection [12,13]. Such et al. [14] recently proposed the use of 16S rRNA PCR to detect bacterial DNA (bDNA) as a new marker of BT. They found bDNA in the serum and ascites of onethird of their patients. The bDNA isolated from ascites was consistently identical to bDNA isolated from serum. This finding was interpreted as a molecular evidence of BT in patients without signs of infection. In a subsequent follow-up multicenter study by the same group, bDNA was found to be strongly associated with survival and associated with complications of cirrhosis [15]. However, a later study did not confirm the prognostic value of bDNA, which was found to be unrelated to survival [16]. Several other markers of BT have been introduced including lipopolysaccharide (LPS) and DOI: 10.1097/MEG.0000000000000217
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Markers of bacterial translocation in cirrhosis Mortensen et al. 1361
lipopolysaccharide-binding protein (LBP). However, the ideal marker remains to be identified [17]. Experimental and clinical data on BT suggest that patients with cirrhosis and BT may benefit from pharmacological treatment, and markers of BT may contribute to our understanding of the haemodynamic alterations in cirrhosis [7,8,12,18]. The aims of this study were to assess bDNA in blood and ascites of patients with cirrhosis by quantitative 16S PCR, as well as its possible association with LBP, an alternative marker of BT, and markers of immune activation.
Materials and methods From 2010 to 2012, we prospectively enrolled 38 patients in three hospitals in the Copenhagen region. The study protocol adhered to the principles of the Helsinki Declaration, and the Scientific Ethics Committee of the Capital Region approved the study protocol (approval number: H-C-2009-020). All patients gave individual written consent for participation in the study. Eligible participants were patients with verified cirrhosis according to clinical, biochemical and ultrasonographic evaluation, who had undergone therapeutic or diagnostic paracentesis. Patients were either admitted to a hospital or seen at an outpatient clinic. Patients who were treated with antibiotics within 1 week before potential inclusion were excluded from participation. Accordingly, patients on SBP prophylaxis were not included in the present study. Paracentesis was performed by a sterile technique in all patients. Polymorphonuclear leucocyte (PMN) counts were obtained by automated cell counting, and SBP was diagnosed on the basis of a PMN count greater than 250/μl. Patients with positive cultures but a PMN count below 250/μl were considered to have bacterascites. The remaining patients, that is, those with normal PMN counts and negative cultures, were the main objective of the present study. Ascites were cultured in aerobic and anaerobic culture media (BACT-Alert; Biomeriux, Marcy l’Etoile, France). Ascites for DNA analysis were stored in sterile DNAse, in endotoxin-free tubes (Greiner Bio-One, Kremsmünster, Austria) at − 20°C for subsequent bDNA analysis. Aliquots of ascites were portioned after centrifugation at 3500g for 10 min and stored at − 20°C for later analysis of markers of inflammation. In a single vein puncture, blood for plasma separation, DNA analysis and culture was drawn by a sterile technique from a cubital vein. Blood was aspirated into sterile EDTA vacuum tubes and culture media. EDTA tubes for DNA analysis were stored unopened at − 20°C until analysis. Plasma was extracted by centrifugation at 3500g for 10 min, was aliquoted and was stored at − 20°C.
16S PCR analysis
The presence of bDNA was determined by quantitative PCR. The conserved regions in the gene encoding the 16S subunit of ribosomal RNA are contained in the genome of all known bacteria. In theory, using primers specific to a target sequence, including both conserved and variable regions, allows the amplification of minute quantities of DNA of any bacterium present in a sample and enables subsequent sequencing and subspecies characterization. Blood and ascites samples were thawed, and bDNA was extracted using Septifast M-grade lysis and prep kits (Roche Diagnostics, Basel, Switzerland), according to the manufacturer’s instructions. A commercially available mix, PerfeCTa qPCR FastMix II, UNG, Low ROX (Quanta BioSciences Inc., Gaithersburg, Maryland, USA), containing nucleotides, polymerase and magnesium chloride was mixed with 1% BSA (Sigma-Aldrich, Saint-Louis, Missouri, USA). The mix was pretreated with shrimp nuclease recombinant (ArcticZymes, Tromsø, Norway) to eliminate potential reagent contaminants. The nuclease was activated by heating the mixture at 30°C for 15 min and subsequently inactivated by heating it at 70°C for 20 min before the addition of samples, probe and primers. Primers with 16S rDNA specificity [forward primer, 16S-331F-mod: TCCTRCGGGAGGCWGCAGT; reverse primer, 16S-797R: GGACTACCAGGGTATCTAAT CCTGTT; and real-time probe, 16S-528 R: FAMCGTATTACCGCGGCTGCTGGCAC-BHQ1 (TAG Copenhagen A/S, Frederiksberg, Denmark), slightly modified from Nadkarni et al. [19]] were filtered through a Microcon 100 kDa molecular weight cutoff spincolumn, added to the master mix and distributed in wells of 96-well reaction plates. Patient samples, positive and negative controls, filtered primers and probes were added to the mixture, and PCR was run on an ABI 7500 real-time PCR (Applied Biosystems, Foster City, California, USA) under the following schedule: 95°C for 2 min, 50 cycles at 95°C for 15 s and 60°C for 1 min. Molecular-grade water and blood from healthy laboratory workers yielded a similar bDNA content, and water was used as a negative control in all PCR reactions. Reactions were run in duplicate and all reactions included positive and negative controls. Samples were quantified by comparing the fluorescent signal of the probes in patient samples with the linear regression line of a standard curve of three positive controls of known bacterial content. Positive controls were Legionella pneumophila. Samples were considered positive if the bDNA concentration estimate exceeded the mean concentration estimate of negative moleculargrade water controls by 2 SDs. Samples showing inhibition, as demonstrated by a PCR signal of low amplitude, were reanalyzed with the addition of DNA corresponding
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
1362 European Journal of Gastroenterology & Hepatology 2014, Vol 26 No 12
to the lowest positive control. If the rerun resulted in an exponential curve, raised to the level of the lowest positive control, the samples were considered negative.
pathogens in non-SBP ascites were Enterococcus faecium, Sarcina spp. and coagulase-negative cocci.
Samples that were positive by DNA quantity were sequenced as previously described [20].
16S rDNA PCR
Markers of inflammation
Levels of LBP (Abnova GmbH, Heidelberg, Germany) in plasma and levels of high-sensitivity C-reactive protein, tumour necrosis factor-α (TNF-α; RG Instruments GmbH, Marburg, Germany), and soluble urokinase plasminogen activating receptor (Virogates, Birkeroed, Denmark) in plasma and ascites were determined using commercially available enzyme-linked immunosorbent assay (ELISA) kits, according to the manufacturer’s instructions. Concentrations of interleukin-6 (IL-6), interleukin 8 (IL-8), interferon-γ inducible protein-10 (IP-10) and vascular endothelial growth factor were determined with Luminex xMAP technology on the Luminex 200 (Luminex Corp., Austin, Texas, USA) using Milliplex Map Human Cytokine/Chemokine Magnetic Bead Panel (Millipore Corp., Billerica, Massachusetts, USA) in both plasma and ascites. All samples were run in duplicates. Statistics
Data were analysed by nonparametric statistical analysis. Proportions were compared using the χ2-test. Results are presented as medians and interquartile ranges (IQRs). Statistical significance was defined as P-values below 0.05.
Results Demographic characteristics of the patients are presented in Table 1. The aetiology of cirrhosis was predominantly alcohol-related, and the patients had advanced cirrhosis as evidenced by Child–Pugh class and MELD-scores. Three out of 38 patients had SBP (8%). All SBP patients had negative blood cultures. A further thee patients had positive ascites cultures in the absence of elevated granulocyte count. The cultured
Table 1
Demographic and biochemical characteristics
Age (years) Outpatients/admissions Sex (M/F) MELD score Child–Pugh class B/C Aetiology (alcohol/combined/nonalcoholic) Serum creatinine (60–105) (μmol/l) Serum sodium (137–144) (mmol/l) Serum albumin (36–45) (g/l) INR (< 1.2) Serum bilirubin (5–25) (μmol/l) Blood platelets (145–390) (billions/l)
57 5/33 11/26 14.04 11/27 36/1/1 65.5 133.5 23 1.5 37.0 155
In 32 patients, ascites and blood cultures were negative, with PMNs below 250 million per litre. Among these patients, the ascitic fluid was DNA positive in 17 patients (52%), whereas blood samples were bDNA positive in seven patients (22%). In four patients (13%), both ascites and blood were bDNA positive (Fig. 1). The prevalence of bDNA-positive patients was not significantly different between Child classes B and C (blood: χ2 = 0.017, P = 0.9; ascites: χ2 = 3.3, P = 0.07), although a trend favouring Child class B bDNA-positive patients was observed. Biochemical variables in Table 1 were not related to presence of bDNA in blood, with the exception of serum creatinine, which was higher in blood bDNA-positive patients (median 81 vs. 59 μmol/l, P = 0.011). Patients with bDNA-positive ascites did not differ from those with bDNA-negative ascites with respect to ascites lymphocytes (P = 0.67) and PMN counts (P = 0.24). bDNA-positive patients were not more likely have mild hepatic encephalopthy or previous episodes of variceal bleeding (χ2, P = 0.28 and 0.4, respectively). In the majority of blood samples 19/32 (59%), bDNA could not be quantified. PCR signals in these samples were of low amplitude, suggesting the presence of inhibitors. As described in the Materials and methods section, we reanalyzed spiked samples and demonstrated the absence of bDNA in all inhibitory samples. bDNA could be quantified in all but one of the ascites samples. Biochemical parameters, including haemoglobin and bilirubin levels, were unrelated to the presence of inhibition in the samples (data not shown). In the 17 patients with ascites-positive PCR, but negative cultures and cell counts, the PCR-estimated bacterial load was generally low (median 9.0 bacteria, IQR 7.7–14.4). The patients with positive blood DNA presented similar low blood bDNA levels (median 9.55, IQR 7.9–14.5). The performance of the assay in patients with SBP was evaluated by the addition of ascites samples from patients with known SBP from another study to our own population of three patients with SBP. Three out of six samples from SBP patients were culture positive, all of them with Escherichia coli.
(49.5–63.3)
(12.4–19.7)
(54.8–93.3) (128–137) (19.8–26) (1.4–1.8) (25–69) (90–215)
INR, international normalized ratio; MELD, model for end-stage liver disease.
All PCR-positive or culture-positive samples were sequenced, but the results were inconclusive in all cases. Markers of inflammation
Plasma markers of inflammation were similar in Child class B and C patients, although a tendency towards higher plasma IL-8 (P = 0.053) and high-sensitivity C-reactive protein (hsCRP; P = 0.059) levels were observed in Child class C patients. Moreover, the
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Markers of bacterial translocation in cirrhosis Mortensen et al. 1363
Fig. 1
Patients included, n= 38
SPB samples from separate cohort, n= 3 (see text)
Ascites cell count and culture
Non-neutrocytic, culture-negative, N = 32
Bacterascites, n= 3
SPB, n= 6
16S PCR N = 32
Ascites bDNA positive n= 17
Ascites bDNA negative n = 15
Blood bDNA positive n=7
n =4
n=3
Blood DNA negative N = 25
n =13
n = 12
Flow-chart illustrating the number of patients with SBP and bacterascites, as well as DNA presence in the blood and ascites of culture-negative, nonneutrocytotic patients. bDNA, bacterial DNA; SBP, spontaneous bacterial peritonitis.
concentrations of VEGF in the ascitic fluid were higher in Child class C patients (P = 0.013). A comparison between bDNA-positive and bDNAnegative samples with respect to levels of inflammatory markers is shown in Table 2. In plasma, VEGF levels were lower in bDNA-positive patients (P = 0.017), whereas in ascites, the VEGF level was significantly higher in bDNA-positive patients Table 2 Markers of inflammation in blood and ascites in patients with normal PMN count and culture-negative ascites (n = 32) Plasma levels
Ascites levels
Marker
Median
IQR
Median
IQR
P-value
IL-6 (pg/ml) IL-8 (pg/ml) IP-10 (pg/ml) VEGF (pg/ml) TNF-α (pg/ml) hsCRP (mg/l) suPAR (ng/ml)
19.9 504.0 3.1 107.3 8.8 49.6 15.9
(9.8–30.4) (46.7–796.0) (2.8–214.4) (82.2–327.8) (5.8–13.2) (13.4–83.3) (9.4–19.8)
2191.1 105.3 3097.3 112.7 6.9 5.4 10.6
(978.6–3488.1) (42.6–3232.6) (138.3–6153.4) (59.3–185.5) (4.1–10.2) (2.0–10.5) (8.2–15.5)
< 0.001 0.866 < 0.001 0.159 0.092 < 0.001 0.03
Significant differences are marked in bold. hsCRP, highly sensitive C-reactive protein; IL-6, interleukin-6; IL-8, interleukin 8; IP-10, interferon-γ inducible protein-10; LBP, lipopolysaccharide-binding protein; PMN, polymorphonuclear leucocyte; suPAR, soluble urokinase plasminogen activating receptor; TNF-α, tumour necrosis factor-α; VEGF, vascular endothelial growth factor.
(P = 0.030). No other marker, including LBP (P = 0.684), differed significantly according to bDNA status. With respect to markers of inflammation in plasma and ascites, hsCRP and soluble urokinase plasminogen activating receptor levels were found to be lower in ascites (P < 0.001 and P = 0.03, respectively) compared with plasma, whereas IL-6 and IP-10 levels were markedly elevated in ascites compared with plasma (P < 0.001), as shown in Table 3. In ascites, TNF-α (ρ = 0.453, P = 0.018) and IP-10 (ρ = 0.423, P = 0.044) levels were positively correlated with PMN count, and IL-6 level was positively correlated with the lymphocyte count (ρ = 0.383, P = 0.044). The performance of the assay in patients with SBP was evaluated by the addition of ascites samples from patients with known SBP from another study to our own population of three patients with SBP. Three out of six samples from SBP patients were culture positive, all of them with E. coli. Cell counts, culture results and bDNA levels in the six patients with SBP are shown in Table 4. All culture-positive ascites samples were PCR positive, although bacterial estimates were low in several samples.
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
1364
European Journal of Gastroenterology & Hepatology 2014, Vol 26 No 12
Table 3 Levels of cytokines in blood and ascites according to the presence of bacterial DNA in patients with normal PMN counts and culturenegative ascites (n = 32) Blood DNA-positive patients (n = 7)
Plasma Plasma Plasma Plasma Plasma Plasma Plasma Plasma
IL-6 (pg/ml) IL-8 (pg/ml) IP-10 (pg/ml) VEGF (pg/ml) LBP (mg/ml) TNF-α (pg/ml) hsCRP (mg/l) suPAR (ng/ml)
Ascites Ascites Ascites Ascites Ascites Ascites Ascites
IL-6 (pg/ml) IL-8 (pg/ml) IP-10 (pg/ml) VEGF (pg/ml) TNF-α (pg/ml) hsCRP (mg/l) suPAR (ng/ml)
Blood DNA-negative patients (n = 25)
Median
Interquartile range
Median
Interquartile range
P-value
20.4 940.0 3.1 70.0 56.7 11.3 63.7 13.1
(12.9–30.3) (428.7–1444.6) (2.9–48.8) (41.1–82.1) (53.9–62.5) (8.4–15.8) (43.4–85.7) (9.3–16.8)
20.8 257.5 3.2 166.5 49.5 8.6 44.7 15.9
(8.2–39.3) (45.7–703.2) (2.8–301.9) (83.9–450.4) (28.9–75.2) (5.5–11.1) (9.1–83.1) (9.3–20.5)
0.957 0.053 0.788 0.017 0.741 0.117 0.226 0.576
Ascites DNA-positive patients (n = 17) 2584.9 122.5 3022.4 139.1 7.5 5.4 10.6
(992.5–4782) (37.2–4307.5) (3.8–4333.0) (72.3–225.9) (6.4–9.3) (1.4–8.5) (8.2–16.6)
Ascites DNA-negative patients (n = 15) 2086.1 119.6 4571.5 76.2 4.2 5.1 9.0
(696.3–3111.9) (43.3–257.6) (541.1–8728.1) (44.1–118.9) (3.0–10.5) (2.2–11.9) (8.1–13.2)
0.402 0.961 0.402 0.030 0.125 0.616 0.390
Significant differences are marked in bold. hsCRP, highly sensitive C-reactive protein, IL-6, interleukin-6, IL-8, interleukin 8, IP-10, interferon-γ inducible protein-10, LBP, lipopolysaccharide-binding protein, PMN, polymorphonuclear leucocyte; suPAR, soluble urokinase plasminogen activating receptor, TNF-α, tumour necrosis factor-α; VEGF, vascular endothelial growth factor.
Table 4
Sample 1 2 3 4 5 6
Bacterial DNA, PMN count and culture in SBP samples Ascite cell count (PMNs; millions/l)
Culture
bDNA positive?
Ascites bDNA quantity
650 545 616 9900 462 468
E. coli E. coli Negative E. coli Negative Negative
Yes Yes Yes Yes No No
276 5.1 4.7 3.4 0 0
bDNA, bacterial DNA; E. coli, Escherichia coli; PMN, polymorphonuclear leucocyte.
In cases of culture-negative SBP, two out of three were negative. As in PCR-positive sterile ascites, sequencing did not characterize any PCR product. The only culturepositive blood sample, with E. coli, was PCR negative.
Discussion The major finding of this study was the frequent presence of bDNA in decompensated cirrhotic patients, which was, however, inconsistently present in blood and ascites. In general, there was no relation between markers of inflammation and LBP. Our results differ markedly from previous results from the group of Such and colleagues, who repeatedly and consistently found clonally identical bDNA in ascites and serum of 30–35% of patients with cirrhosis [14,15,21], and bDNA related to inflammation [22] and endothelial dysfunction [10]. Diversity in methods, for instance, differences in quantitative assessments versus conventional PCR, and dissimilarities between blood and serum may in part explain the discrepancies in our findings. However, several recent studies have also shown differing results with respect to the presence of bDNA,
immune activation and SBP. A large study investigating the presence of bDNA in blood and ascites using an automated, commercially available PCR-based probe system found bDNA to be more frequent in sterile ascites (12%) than in blood (1%) and concluded that this method was imprecise in the diagnosis of SBP [23]. In addition, neither biochemical parameters nor variables with known prognostic value were significantly associated with the presence of bDNA. Vlachogiannakos et al. [17] reported the complete absence of plasma bDNA in 29 patients when they reanalyzed data from a previous study using LPS as a marker of BT. Soriano et al. [24] investigated the presence of bDNA in ascites in patients with SBP episodes and in 20 cirrhotic patients with sterile ascites using a quantitative PCR assay similar to ours. In sterile ascites, bDNA was positive in 60% of cases and could be sequenced in half of those cases, whereas bDNA was positive in 92% of culture-positive and 53% of culture-negative SBP patients. In our study, markers of inflammation did not reflect the stage of disease, although the trends we observed were similar to the results obtained by our group and other groups using similar techniques with regard to hsCRP [25] and IL-8 [16,25] levels. The only marker in our study that differed between bDNA-positive and bDNA-negative patients was VEGF, which was higher in ascitic fluid in bDNA-positive patients, but lower in the plasma of blood bDNApositive patients. A relationship between VEGF level and BT has been previously demonstrated by Perez-Ruiz et al. [26], who showed that peritoneal macrophages produced VEGF in response to TNF-α and LPS stimulation, thus linking an alternate marker of BT to VEGF
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Markers of bacterial translocation in cirrhosis Mortensen et al. 1365
overexpression in ascites. However, no relationship between bDNA and TNF-α was evident in our study. In culture-negative ascites with low levels of immune cells, we expected to find levels of cytokines comparable to plasma levels. Surprisingly, we found markedly elevated levels of two markers of inflammation. With regard to IL-6 levels, these findings confirm earlier findings of IL-6 levels being 24-fold higher in ascites compared with the serum of non-SBP patients [27]. Interferon-γ inducible protein-10 levels, a cytokine mediating macrophage and T-cell chemotaxis, were elevated by ∼ 1000-fold in ascites. These findings may be a consequence of marked production or reduced elimination of important regulators of immune function and cell trafficking in the ascitic fluid and warrant further investigation. The assay applied for bDNA detection was developed for the specific purpose of 16S PCR in blood and ascites in a PCR laboratory accredited for diagnostic PCR and with long-term experience in developing PCR methods for pathogen detection [28–30]. Despite extensive validation on simulated samples, technical issues interfere with the interpretation of the results of our study, constituting a weakness of the present study. The quantification of bDNA in a number of blood samples and in a single ascites sample failed because of inhibition of the PCR reaction, despite the application of an established strategy, the addition of BSA [31], to reduce this problem. We were, however, able to determine the presence or absence of DNA by our applied methods. The failure of sequencing the DNA-positive samples may be due to several factors. The amount of bDNA in ascites was generally modest, and low levels of bDNA in the samples may have hindered the determination of the target sequence. Furthermore, the amplified DNA targets may be polymicrobial in nature, therefore making the sequences difficult to interpret. This was also suggested by Soriano et al. [24], who were able to successfully sequence only half of the PCR-positive sterile ascites samples.
with low levels of bacteria, PCR may fail simply by missing the bacteria. Conclusion
Our results reveal a low quantity of bDNA in blood and ascites of patients with cirrhosis. We were unable to sequence the PCR products, likely because of the very low amounts of DNA and possibly a polyclonal origin of the DNA. bDNA levels in blood and ascites from individual patients were often not concordant, and bDNA appeared unrelated to markers of inflammation. These findings suggest that the presence of bDNA is unrelated to the immunological and consequent haemodynamic deterioration in cirrhosis. Moreover, bDNA failed to correctly diagnose SBP, rendering the developed assay unsuitable for such diagnostic purposes.
Acknowledgements The authors gratefully acknowledge the help received from Drs Camilla Nøjgaard, Peter Thielsen and Lasse Bremholm, as well as the expert laboratory assistance of Diana Klüver and Lise Lotte Spurr Madsen. C.M., S.M. and F.B. designed the study and wrote the protocol. O.A. advised on immunological aspects of the protocol. The methods for DNA analysis were developed by J.S.J. The markers of immunity were developed by O.A. O.A. analysed markers of immunity. J.S.J. performed DNA analyses. C.M., S.D.L., B.S.M. and L.H. included patients and collected samples. C.M. carried out the statistical analyses. C.M. drafted the manuscript. All authors contributed to the final manuscript. This work is supported by grants from L.F. Foght’s Foundation, Hvidovre Hospitals’ Research Foundation, Hvidovre Hospitals Foundation for Liver Diseases, the Novo Nordisk Foundation and the The Capital Region of Denmark, Foundation for Health Research. Conflicts of interest
There are no conflicts of interest.
References 1 2
The pathogenesis of culture-positive SBP and of bacteria cultured from lymph nodes in experimental cirrhosis has led to the assumption of BT being a monomicrobial event. However, the bDNA detected by PCR in culturenegative ascites may be derived from pathogens that are neither viable nor cultivable, and may indeed be fragments opsonized or phagocytosed by cells of the immune system during previous translocations or subclinical infections. In the culture-positive sample, however, sequencing failure was more surprising. The typical sample volume used for PCR is much smaller compared with the volume used for culture. Accordingly, in samples
3
4
5
6
Tandon P, Garcia-Tsao G. Bacterial infections, sepsis, and multiorgan failure in cirrhosis. Semin Liver Dis 2008; 28:26–42. Arvaniti V, D’Amico G, Fede G, Manousou P, Tsochatzis E, Pleguezuelo M, Burroughs AK. Infections in patients with cirrhosis increase mortality four-fold and should be used in determining prognosis. Gastroenterology 2010; 139:1246–1256. European Association for the Study of the Liver. EASL clinical practice guidelines on the management of ascites, spontaneous bacterial peritonitis, and hepatorenal syndrome in cirrhosis. J Hepatol 2010; 53:397–417. Fernandez J, Navasa M, Gómez J, Colmenero J, Vila J, Arroyo V, Rodés J. Bacterial infections in cirrhosis: epidemiological changes with invasive procedures and norfloxacin prophylaxis. Hepatology 2002; 35:140–148. Berg RD, Garlington AW. Translocation of certain indigenous bacteria from the gastrointestinal tract to the mesenteric lymph nodes and other organs in a gnotobiotic mouse model. Infect Immun 1979; 23:403–411. Follo A, Llovet JM, Navasa M, Planas R, Forns X, Francitorra A, et al. Renal impairment after spontaneous bacterial peritonitis in cirrhosis: incidence, clinical course, predictive factors and prognosis. Hepatology 1994; 20:1495–1501.
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
1366
7
8
9
10
11
12 13
14
15
16
17
18
European Journal of Gastroenterology & Hepatology 2014, Vol 26 No 12
Rasaratnam B, Kaye D, Jennings G, Dudley F, Chin-Dusting J. The effect of selective intestinal decontamination on the hyperdynamic circulatory state in cirrhosis. A randomized trial. Ann Intern Med 2003; 139:186–193. Vlachogiannakos J, Saveriadis AS, Viazis N, Theodoropoulos I, Foudoulis K, Manolakopoulos S, et al. Intestinal decontamination improves liver haemodynamics in patients with alcohol-related decompensated cirrhosis. Aliment Pharmacol Ther 2009; 29:992–999. Albillos A, de la Hera A, González M, Moya JL, Calleja JL, Monserrat J, et al. Increased lipopolysaccharide binding protein in cirrhotic patients with marked immune and hemodynamic derangement. Hepatology 2003; 37:208–217. Bellot P, García-Pagán JC, Francés R, Abraldes JG, Navasa M, PérezMateo M, et al. Bacterial DNA translocation is associated with systemic circulatory abnormalities and intrahepatic endothelial dysfunction in patients with cirrhosis. Hepatology 2010; 52:2044–2052. Albillos A, de-la-Hera A, Alvarez-Mon M. Serum lipopolysaccharide-binding protein prediction of severe bacterial infection in cirrhotic patients with ascites. Lancet 2004; 15:1608–1610. Vallance P, Moncada S. Hyperdynamic circulation in cirrhosis: a role for nitric oxide? Lancet 1991; 337:776–778. Garcia-Tsao G, Wiest R. Gut microflora in the pathogenesis of the complications of cirrhosis. Best Pract Res Clin Gastroenterol 2004; 18:353–372. Such J, Francés R, Muñoz C, Zapater P, Casellas JA, Cifuentes A, et al. Detection and identification of bacterial DNA in patients with cirrhosis and culture-negative, nonneutrocytic ascites. Hepatology 2002; 36:135–141. Zapater P, Francés R, González-Navajas JM, de la Hoz MA, Moreu R, Pascual S, et al. Serum and ascitic fluid bacterial DNA: a new independent prognostic factor in noninfected patients with cirrhosis. Hepatology 2008; 48:1924–1931. Tritto G, Bechlis Z, Stadlbauer V, Davies N, Francés R, Shah N, et al. Evidence of neutrophil functional defect despite inflammation in stable cirrhosis. J Hepatol 2011; 55:574–581. Vlachogiannakos J, Daikos G, Thalheimer U, Burroughs AK, Ladas SD. Is bacterial DNA a better marker than endotoxin of bacterial translocation in decompensated cirrhosis? Hepatology 2011; 53:2140–2141. Wiest R, Das S, Cadelina G, Garcia-Tsao G, Milstien S, Groszmann RJ. Bacterial translocation in cirrhotic rats stimulates eNOS-derived NO production and impairs mesenteric vascular contractility. J Clin Invest 1999; 104:1223–1233.
19
20
21
22
23
24
25
26
27
28
29 30
31
Nadkarni MA, Martin FE, Jacques NA, Hunter N. Determination of bacterial load by real-time PCR using a broad-range (universal) probe and primers set. Microbiology 2002; 148:257–266. Hjorth SV, Björnelius E, Lidbrink P, Falk L, Dohn B, Berthelsen L, et al. Sequence-based typing of Mycoplasma genitalium reveals sexual transmission. J Clin Microbiol 2006; 44:2078–2083. Such J, Muñoz C, Zapater P, Pérez-Mateo M. Bacterial DNA induces a proinflammatory immune response in patients with decompensated cirrhosis. Gut 2005; 54:1500. González-Navajas JM, Bellot P, Francés R, Zapater P, Muñoz C, GarcíaPagan JC, et al. Presence of bacterial-DNA in cirrhosis identifies a subgroup of patients with marked inflammatory response not related to endotoxin. J Hepatol 2008; 48:61–67. Appenrodt B, Lehmann LE, Thyssen L, Gentemann M, Rabe C, Molitor E, et al. Is detection of bacterial DNA in ascitic fluid of clinical relevance? Eur J Gastroenterol Hepatol 2010; 22:1487–1494. Soriano G, Esparcia O, Montemayor M, Guarner-Argente C, Pericas R, Torras X, et al. Bacterial DNA in the diagnosis of spontaneous bacterial peritonitis. Aliment Pharmacol Ther 2011; 33:275–284. Mortensen C, Andersen O, Krag A, Bendtsen F, Møller S. High-sensitivity C-reactive protein levels predict survival and are related to haemodynamics in alcoholic cirrhosis. Eur J Gastroenterol Hepatol 2012; 24:619–626. Perez-Ruiz M, Ros J, Morales-Ruiz M, Navasa M, Colmenero J, Ruiz-delArbol L, et al. Vascular endothelial growth factor production in peritoneal macrophages of cirrhotic patients: regulation by cytokines and bacterial lipopolysaccharide. Hepatology 1999; 29:1057–1063. Rodriguez-Ramos C, Galan F, Díaz F, Elvira J, Martín-Herrera L, GirónGonzález JA. Expression of proinflammatory cytokines and their inhibitors during the course of spontaneous bacterial peritonitis. Dig Dis Sci 2001; 46:1668–1676. Dragsted DM, Dohn B, Madsen J, Jensen JS. Comparison of culture and PCR for detection of Bordetella pertussis and Bordetella parapertussis under routine laboratory conditions. J Med Microbiol 2004; 53:749–754. Jensen JS, Borre MB, Dohn B. Detection of Mycoplasma genitalium by PCR amplification of the 16S rRNA gene. J Clin Microbiol 2003; 41:261–266. Villumsen S, Pedersen R, Krogfelt KA, Jensen JS. Expanding the diagnostic use of PCR in leptospirosis: improved method for DNA extraction from blood cultures. PLoS One 2010; 5:e12095. Al-Soud WA, Radstrom P. Purification and characterization of PCR-inhibitory components in blood cells. J Clin Microbiol 2001; 39:485–493.
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Original article
Association of markers of bacterial translocation with immune activation in decompensated cirrhosis Christian Mortensena,b, Jørgen Skov Jensend, Lise Hobolthe, Sanne Dam-Larsenf, Bjørn S. Madseng, Ove Andersenc, Søren Møllerb and Flemming Bendtsena Background Bacterial translocation (BT) may cause infections, in particular, spontaneous bacterial peritonitis (SBP). In the absence of overt infection, BT may further stimulate the immune system and contribute to haemodynamic alterations and complications. Bacterial DNA (bDNA) is claimed to be a promising surrogate marker for BT, although its clinical relevance has been questioned. Materials and methods In 38 cirrhotic patients with and without SBP, bDNA in blood and ascites were assessed by 16S rDNA quantitative PCR. Levels of lipopolysaccharidebinding protein in plasma and highly sensitive C-reactive protein, tumour necrosis factor-α, soluble urokinase plasminogen activating receptor, interleukin-6, interleukin 8, interferon-γ inducible protein-10 and vascular endothelial growth factor in plasma and ascites were measured by multiplex cytokine and ELISA assays. Results In patients without signs of SBP or positive cultures, we found a high frequency of bDNA but low concordance of bDNA between blood and ascites. Markers of inflammation were not significantly different between blood bDNA-positive (22%), ascites bDNA-positive (52%), and bDNA-negative patients. The 16S rDNA PCR failed to
Introduction Impairment of the immune system in patients with cirrhosis leads to frequent lethal infections [1,2]. In patients with decompensated cirrhosis, development of spontaneous bacterial peritonitis (SBP) is associated with increased mortality [3,4]. Bacterial translocation (BT) refers to the passage of gut bacteria or bacterial products into extraintestinal lymph nodes [5], and it is thought to progress to SBP through viable gut bacteria colonizing the ascitic fluid. SBP is associated with high recurrence rates, and apart from high mortality, it is often accompanied by the development of hepatorenal syndrome [6]. Recently, it has been hypothesized that BT, even in the absence of clinical infection, may impact the course of the disease in patients with cirrhosis. Several studies, using different markers of BT have reported significant associations with the risk for infection and haemodynamic changes [7–11], which lends support to the hypothesis of an inappropriate activation of the immune 0954-691X © 2014 Wolters Kluwer Health | Lippincott Williams & Wilkins
show bDNA in two out of six samples with SBP. Sequencing of positive samples did not determine the source of bDNA. Conclusion bDNA as assessed by this PCR method was largely unrelated to markers of inflammation and does not seem to be of clinical value in the diagnosis of SBP. According to our results, bDNA is not a reliable marker of BT. Eur J Gastroenterol Hepatol 26:1360–1366 © 2014 Wolters Kluwer Health | Lippincott Williams & Wilkins. European Journal of Gastroenterology & Hepatology 2014, 26:1360–1366 Keywords: bacterial translocation, cirrhosis, inflammation, PCR Departments of aGastroenterology, bClinical Physiology and Nuclear Medicine, Centre of Functional Imaging and Research, cCentre for Clinical Research, Hvidovre University Hospital, Hvidovre, dMycoplasma Laboratory, Microbiology and Infection Control, Statens Serum Institut, eDepartment of Internal Medicine I, Bispebjerg Hospital, fDepartment of Internal Medicine, Glostrup Hospital, Glostrup and gDepartment of Medicine, Svendborg Sygehus, Svendborg, Denmark Correspondence to Christian Mortensen, MD, Department of Gastroenterology, Hvidovre University Hospital, Kettegaard Alle 30, DK-2650 Hvidovre, Denmark Tel: + 45 38 623 181; fax: + 45 3 8623 777; e-mail: [email protected] Received 26 June 2014 Accepted 29 August 2014
system on BT. This activation may further have adverse effects on circulation in cirrhosis by the release of systemic or local vasodilators, finally leading to secondary complications, even in the absence of clinically overt infection [12,13]. Such et al. [14] recently proposed the use of 16S rRNA PCR to detect bacterial DNA (bDNA) as a new marker of BT. They found bDNA in the serum and ascites of onethird of their patients. The bDNA isolated from ascites was consistently identical to bDNA isolated from serum. This finding was interpreted as a molecular evidence of BT in patients without signs of infection. In a subsequent follow-up multicenter study by the same group, bDNA was found to be strongly associated with survival and associated with complications of cirrhosis [15]. However, a later study did not confirm the prognostic value of bDNA, which was found to be unrelated to survival [16]. Several other markers of BT have been introduced including lipopolysaccharide (LPS) and DOI: 10.1097/MEG.0000000000000217
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Markers of bacterial translocation in cirrhosis Mortensen et al. 1361
lipopolysaccharide-binding protein (LBP). However, the ideal marker remains to be identified [17]. Experimental and clinical data on BT suggest that patients with cirrhosis and BT may benefit from pharmacological treatment, and markers of BT may contribute to our understanding of the haemodynamic alterations in cirrhosis [7,8,12,18]. The aims of this study were to assess bDNA in blood and ascites of patients with cirrhosis by quantitative 16S PCR, as well as its possible association with LBP, an alternative marker of BT, and markers of immune activation.
Materials and methods From 2010 to 2012, we prospectively enrolled 38 patients in three hospitals in the Copenhagen region. The study protocol adhered to the principles of the Helsinki Declaration, and the Scientific Ethics Committee of the Capital Region approved the study protocol (approval number: H-C-2009-020). All patients gave individual written consent for participation in the study. Eligible participants were patients with verified cirrhosis according to clinical, biochemical and ultrasonographic evaluation, who had undergone therapeutic or diagnostic paracentesis. Patients were either admitted to a hospital or seen at an outpatient clinic. Patients who were treated with antibiotics within 1 week before potential inclusion were excluded from participation. Accordingly, patients on SBP prophylaxis were not included in the present study. Paracentesis was performed by a sterile technique in all patients. Polymorphonuclear leucocyte (PMN) counts were obtained by automated cell counting, and SBP was diagnosed on the basis of a PMN count greater than 250/μl. Patients with positive cultures but a PMN count below 250/μl were considered to have bacterascites. The remaining patients, that is, those with normal PMN counts and negative cultures, were the main objective of the present study. Ascites were cultured in aerobic and anaerobic culture media (BACT-Alert; Biomeriux, Marcy l’Etoile, France). Ascites for DNA analysis were stored in sterile DNAse, in endotoxin-free tubes (Greiner Bio-One, Kremsmünster, Austria) at − 20°C for subsequent bDNA analysis. Aliquots of ascites were portioned after centrifugation at 3500g for 10 min and stored at − 20°C for later analysis of markers of inflammation. In a single vein puncture, blood for plasma separation, DNA analysis and culture was drawn by a sterile technique from a cubital vein. Blood was aspirated into sterile EDTA vacuum tubes and culture media. EDTA tubes for DNA analysis were stored unopened at − 20°C until analysis. Plasma was extracted by centrifugation at 3500g for 10 min, was aliquoted and was stored at − 20°C.
16S PCR analysis
The presence of bDNA was determined by quantitative PCR. The conserved regions in the gene encoding the 16S subunit of ribosomal RNA are contained in the genome of all known bacteria. In theory, using primers specific to a target sequence, including both conserved and variable regions, allows the amplification of minute quantities of DNA of any bacterium present in a sample and enables subsequent sequencing and subspecies characterization. Blood and ascites samples were thawed, and bDNA was extracted using Septifast M-grade lysis and prep kits (Roche Diagnostics, Basel, Switzerland), according to the manufacturer’s instructions. A commercially available mix, PerfeCTa qPCR FastMix II, UNG, Low ROX (Quanta BioSciences Inc., Gaithersburg, Maryland, USA), containing nucleotides, polymerase and magnesium chloride was mixed with 1% BSA (Sigma-Aldrich, Saint-Louis, Missouri, USA). The mix was pretreated with shrimp nuclease recombinant (ArcticZymes, Tromsø, Norway) to eliminate potential reagent contaminants. The nuclease was activated by heating the mixture at 30°C for 15 min and subsequently inactivated by heating it at 70°C for 20 min before the addition of samples, probe and primers. Primers with 16S rDNA specificity [forward primer, 16S-331F-mod: TCCTRCGGGAGGCWGCAGT; reverse primer, 16S-797R: GGACTACCAGGGTATCTAAT CCTGTT; and real-time probe, 16S-528 R: FAMCGTATTACCGCGGCTGCTGGCAC-BHQ1 (TAG Copenhagen A/S, Frederiksberg, Denmark), slightly modified from Nadkarni et al. [19]] were filtered through a Microcon 100 kDa molecular weight cutoff spincolumn, added to the master mix and distributed in wells of 96-well reaction plates. Patient samples, positive and negative controls, filtered primers and probes were added to the mixture, and PCR was run on an ABI 7500 real-time PCR (Applied Biosystems, Foster City, California, USA) under the following schedule: 95°C for 2 min, 50 cycles at 95°C for 15 s and 60°C for 1 min. Molecular-grade water and blood from healthy laboratory workers yielded a similar bDNA content, and water was used as a negative control in all PCR reactions. Reactions were run in duplicate and all reactions included positive and negative controls. Samples were quantified by comparing the fluorescent signal of the probes in patient samples with the linear regression line of a standard curve of three positive controls of known bacterial content. Positive controls were Legionella pneumophila. Samples were considered positive if the bDNA concentration estimate exceeded the mean concentration estimate of negative moleculargrade water controls by 2 SDs. Samples showing inhibition, as demonstrated by a PCR signal of low amplitude, were reanalyzed with the addition of DNA corresponding
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
1362 European Journal of Gastroenterology & Hepatology 2014, Vol 26 No 12
to the lowest positive control. If the rerun resulted in an exponential curve, raised to the level of the lowest positive control, the samples were considered negative.
pathogens in non-SBP ascites were Enterococcus faecium, Sarcina spp. and coagulase-negative cocci.
Samples that were positive by DNA quantity were sequenced as previously described [20].
16S rDNA PCR
Markers of inflammation
Levels of LBP (Abnova GmbH, Heidelberg, Germany) in plasma and levels of high-sensitivity C-reactive protein, tumour necrosis factor-α (TNF-α; RG Instruments GmbH, Marburg, Germany), and soluble urokinase plasminogen activating receptor (Virogates, Birkeroed, Denmark) in plasma and ascites were determined using commercially available enzyme-linked immunosorbent assay (ELISA) kits, according to the manufacturer’s instructions. Concentrations of interleukin-6 (IL-6), interleukin 8 (IL-8), interferon-γ inducible protein-10 (IP-10) and vascular endothelial growth factor were determined with Luminex xMAP technology on the Luminex 200 (Luminex Corp., Austin, Texas, USA) using Milliplex Map Human Cytokine/Chemokine Magnetic Bead Panel (Millipore Corp., Billerica, Massachusetts, USA) in both plasma and ascites. All samples were run in duplicates. Statistics
Data were analysed by nonparametric statistical analysis. Proportions were compared using the χ2-test. Results are presented as medians and interquartile ranges (IQRs). Statistical significance was defined as P-values below 0.05.
Results Demographic characteristics of the patients are presented in Table 1. The aetiology of cirrhosis was predominantly alcohol-related, and the patients had advanced cirrhosis as evidenced by Child–Pugh class and MELD-scores. Three out of 38 patients had SBP (8%). All SBP patients had negative blood cultures. A further thee patients had positive ascites cultures in the absence of elevated granulocyte count. The cultured
Table 1
Demographic and biochemical characteristics
Age (years) Outpatients/admissions Sex (M/F) MELD score Child–Pugh class B/C Aetiology (alcohol/combined/nonalcoholic) Serum creatinine (60–105) (μmol/l) Serum sodium (137–144) (mmol/l) Serum albumin (36–45) (g/l) INR (< 1.2) Serum bilirubin (5–25) (μmol/l) Blood platelets (145–390) (billions/l)
57 5/33 11/26 14.04 11/27 36/1/1 65.5 133.5 23 1.5 37.0 155
In 32 patients, ascites and blood cultures were negative, with PMNs below 250 million per litre. Among these patients, the ascitic fluid was DNA positive in 17 patients (52%), whereas blood samples were bDNA positive in seven patients (22%). In four patients (13%), both ascites and blood were bDNA positive (Fig. 1). The prevalence of bDNA-positive patients was not significantly different between Child classes B and C (blood: χ2 = 0.017, P = 0.9; ascites: χ2 = 3.3, P = 0.07), although a trend favouring Child class B bDNA-positive patients was observed. Biochemical variables in Table 1 were not related to presence of bDNA in blood, with the exception of serum creatinine, which was higher in blood bDNA-positive patients (median 81 vs. 59 μmol/l, P = 0.011). Patients with bDNA-positive ascites did not differ from those with bDNA-negative ascites with respect to ascites lymphocytes (P = 0.67) and PMN counts (P = 0.24). bDNA-positive patients were not more likely have mild hepatic encephalopthy or previous episodes of variceal bleeding (χ2, P = 0.28 and 0.4, respectively). In the majority of blood samples 19/32 (59%), bDNA could not be quantified. PCR signals in these samples were of low amplitude, suggesting the presence of inhibitors. As described in the Materials and methods section, we reanalyzed spiked samples and demonstrated the absence of bDNA in all inhibitory samples. bDNA could be quantified in all but one of the ascites samples. Biochemical parameters, including haemoglobin and bilirubin levels, were unrelated to the presence of inhibition in the samples (data not shown). In the 17 patients with ascites-positive PCR, but negative cultures and cell counts, the PCR-estimated bacterial load was generally low (median 9.0 bacteria, IQR 7.7–14.4). The patients with positive blood DNA presented similar low blood bDNA levels (median 9.55, IQR 7.9–14.5). The performance of the assay in patients with SBP was evaluated by the addition of ascites samples from patients with known SBP from another study to our own population of three patients with SBP. Three out of six samples from SBP patients were culture positive, all of them with Escherichia coli.
(49.5–63.3)
(12.4–19.7)
(54.8–93.3) (128–137) (19.8–26) (1.4–1.8) (25–69) (90–215)
INR, international normalized ratio; MELD, model for end-stage liver disease.
All PCR-positive or culture-positive samples were sequenced, but the results were inconclusive in all cases. Markers of inflammation
Plasma markers of inflammation were similar in Child class B and C patients, although a tendency towards higher plasma IL-8 (P = 0.053) and high-sensitivity C-reactive protein (hsCRP; P = 0.059) levels were observed in Child class C patients. Moreover, the
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Markers of bacterial translocation in cirrhosis Mortensen et al. 1363
Fig. 1
Patients included, n= 38
SPB samples from separate cohort, n= 3 (see text)
Ascites cell count and culture
Non-neutrocytic, culture-negative, N = 32
Bacterascites, n= 3
SPB, n= 6
16S PCR N = 32
Ascites bDNA positive n= 17
Ascites bDNA negative n = 15
Blood bDNA positive n=7
n =4
n=3
Blood DNA negative N = 25
n =13
n = 12
Flow-chart illustrating the number of patients with SBP and bacterascites, as well as DNA presence in the blood and ascites of culture-negative, nonneutrocytotic patients. bDNA, bacterial DNA; SBP, spontaneous bacterial peritonitis.
concentrations of VEGF in the ascitic fluid were higher in Child class C patients (P = 0.013). A comparison between bDNA-positive and bDNAnegative samples with respect to levels of inflammatory markers is shown in Table 2. In plasma, VEGF levels were lower in bDNA-positive patients (P = 0.017), whereas in ascites, the VEGF level was significantly higher in bDNA-positive patients Table 2 Markers of inflammation in blood and ascites in patients with normal PMN count and culture-negative ascites (n = 32) Plasma levels
Ascites levels
Marker
Median
IQR
Median
IQR
P-value
IL-6 (pg/ml) IL-8 (pg/ml) IP-10 (pg/ml) VEGF (pg/ml) TNF-α (pg/ml) hsCRP (mg/l) suPAR (ng/ml)
19.9 504.0 3.1 107.3 8.8 49.6 15.9
(9.8–30.4) (46.7–796.0) (2.8–214.4) (82.2–327.8) (5.8–13.2) (13.4–83.3) (9.4–19.8)
2191.1 105.3 3097.3 112.7 6.9 5.4 10.6
(978.6–3488.1) (42.6–3232.6) (138.3–6153.4) (59.3–185.5) (4.1–10.2) (2.0–10.5) (8.2–15.5)
< 0.001 0.866 < 0.001 0.159 0.092 < 0.001 0.03
Significant differences are marked in bold. hsCRP, highly sensitive C-reactive protein; IL-6, interleukin-6; IL-8, interleukin 8; IP-10, interferon-γ inducible protein-10; LBP, lipopolysaccharide-binding protein; PMN, polymorphonuclear leucocyte; suPAR, soluble urokinase plasminogen activating receptor; TNF-α, tumour necrosis factor-α; VEGF, vascular endothelial growth factor.
(P = 0.030). No other marker, including LBP (P = 0.684), differed significantly according to bDNA status. With respect to markers of inflammation in plasma and ascites, hsCRP and soluble urokinase plasminogen activating receptor levels were found to be lower in ascites (P < 0.001 and P = 0.03, respectively) compared with plasma, whereas IL-6 and IP-10 levels were markedly elevated in ascites compared with plasma (P < 0.001), as shown in Table 3. In ascites, TNF-α (ρ = 0.453, P = 0.018) and IP-10 (ρ = 0.423, P = 0.044) levels were positively correlated with PMN count, and IL-6 level was positively correlated with the lymphocyte count (ρ = 0.383, P = 0.044). The performance of the assay in patients with SBP was evaluated by the addition of ascites samples from patients with known SBP from another study to our own population of three patients with SBP. Three out of six samples from SBP patients were culture positive, all of them with E. coli. Cell counts, culture results and bDNA levels in the six patients with SBP are shown in Table 4. All culture-positive ascites samples were PCR positive, although bacterial estimates were low in several samples.
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
1364
European Journal of Gastroenterology & Hepatology 2014, Vol 26 No 12
Table 3 Levels of cytokines in blood and ascites according to the presence of bacterial DNA in patients with normal PMN counts and culturenegative ascites (n = 32) Blood DNA-positive patients (n = 7)
Plasma Plasma Plasma Plasma Plasma Plasma Plasma Plasma
IL-6 (pg/ml) IL-8 (pg/ml) IP-10 (pg/ml) VEGF (pg/ml) LBP (mg/ml) TNF-α (pg/ml) hsCRP (mg/l) suPAR (ng/ml)
Ascites Ascites Ascites Ascites Ascites Ascites Ascites
IL-6 (pg/ml) IL-8 (pg/ml) IP-10 (pg/ml) VEGF (pg/ml) TNF-α (pg/ml) hsCRP (mg/l) suPAR (ng/ml)
Blood DNA-negative patients (n = 25)
Median
Interquartile range
Median
Interquartile range
P-value
20.4 940.0 3.1 70.0 56.7 11.3 63.7 13.1
(12.9–30.3) (428.7–1444.6) (2.9–48.8) (41.1–82.1) (53.9–62.5) (8.4–15.8) (43.4–85.7) (9.3–16.8)
20.8 257.5 3.2 166.5 49.5 8.6 44.7 15.9
(8.2–39.3) (45.7–703.2) (2.8–301.9) (83.9–450.4) (28.9–75.2) (5.5–11.1) (9.1–83.1) (9.3–20.5)
0.957 0.053 0.788 0.017 0.741 0.117 0.226 0.576
Ascites DNA-positive patients (n = 17) 2584.9 122.5 3022.4 139.1 7.5 5.4 10.6
(992.5–4782) (37.2–4307.5) (3.8–4333.0) (72.3–225.9) (6.4–9.3) (1.4–8.5) (8.2–16.6)
Ascites DNA-negative patients (n = 15) 2086.1 119.6 4571.5 76.2 4.2 5.1 9.0
(696.3–3111.9) (43.3–257.6) (541.1–8728.1) (44.1–118.9) (3.0–10.5) (2.2–11.9) (8.1–13.2)
0.402 0.961 0.402 0.030 0.125 0.616 0.390
Significant differences are marked in bold. hsCRP, highly sensitive C-reactive protein, IL-6, interleukin-6, IL-8, interleukin 8, IP-10, interferon-γ inducible protein-10, LBP, lipopolysaccharide-binding protein, PMN, polymorphonuclear leucocyte; suPAR, soluble urokinase plasminogen activating receptor, TNF-α, tumour necrosis factor-α; VEGF, vascular endothelial growth factor.
Table 4
Sample 1 2 3 4 5 6
Bacterial DNA, PMN count and culture in SBP samples Ascite cell count (PMNs; millions/l)
Culture
bDNA positive?
Ascites bDNA quantity
650 545 616 9900 462 468
E. coli E. coli Negative E. coli Negative Negative
Yes Yes Yes Yes No No
276 5.1 4.7 3.4 0 0
bDNA, bacterial DNA; E. coli, Escherichia coli; PMN, polymorphonuclear leucocyte.
In cases of culture-negative SBP, two out of three were negative. As in PCR-positive sterile ascites, sequencing did not characterize any PCR product. The only culturepositive blood sample, with E. coli, was PCR negative.
Discussion The major finding of this study was the frequent presence of bDNA in decompensated cirrhotic patients, which was, however, inconsistently present in blood and ascites. In general, there was no relation between markers of inflammation and LBP. Our results differ markedly from previous results from the group of Such and colleagues, who repeatedly and consistently found clonally identical bDNA in ascites and serum of 30–35% of patients with cirrhosis [14,15,21], and bDNA related to inflammation [22] and endothelial dysfunction [10]. Diversity in methods, for instance, differences in quantitative assessments versus conventional PCR, and dissimilarities between blood and serum may in part explain the discrepancies in our findings. However, several recent studies have also shown differing results with respect to the presence of bDNA,
immune activation and SBP. A large study investigating the presence of bDNA in blood and ascites using an automated, commercially available PCR-based probe system found bDNA to be more frequent in sterile ascites (12%) than in blood (1%) and concluded that this method was imprecise in the diagnosis of SBP [23]. In addition, neither biochemical parameters nor variables with known prognostic value were significantly associated with the presence of bDNA. Vlachogiannakos et al. [17] reported the complete absence of plasma bDNA in 29 patients when they reanalyzed data from a previous study using LPS as a marker of BT. Soriano et al. [24] investigated the presence of bDNA in ascites in patients with SBP episodes and in 20 cirrhotic patients with sterile ascites using a quantitative PCR assay similar to ours. In sterile ascites, bDNA was positive in 60% of cases and could be sequenced in half of those cases, whereas bDNA was positive in 92% of culture-positive and 53% of culture-negative SBP patients. In our study, markers of inflammation did not reflect the stage of disease, although the trends we observed were similar to the results obtained by our group and other groups using similar techniques with regard to hsCRP [25] and IL-8 [16,25] levels. The only marker in our study that differed between bDNA-positive and bDNA-negative patients was VEGF, which was higher in ascitic fluid in bDNA-positive patients, but lower in the plasma of blood bDNApositive patients. A relationship between VEGF level and BT has been previously demonstrated by Perez-Ruiz et al. [26], who showed that peritoneal macrophages produced VEGF in response to TNF-α and LPS stimulation, thus linking an alternate marker of BT to VEGF
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Markers of bacterial translocation in cirrhosis Mortensen et al. 1365
overexpression in ascites. However, no relationship between bDNA and TNF-α was evident in our study. In culture-negative ascites with low levels of immune cells, we expected to find levels of cytokines comparable to plasma levels. Surprisingly, we found markedly elevated levels of two markers of inflammation. With regard to IL-6 levels, these findings confirm earlier findings of IL-6 levels being 24-fold higher in ascites compared with the serum of non-SBP patients [27]. Interferon-γ inducible protein-10 levels, a cytokine mediating macrophage and T-cell chemotaxis, were elevated by ∼ 1000-fold in ascites. These findings may be a consequence of marked production or reduced elimination of important regulators of immune function and cell trafficking in the ascitic fluid and warrant further investigation. The assay applied for bDNA detection was developed for the specific purpose of 16S PCR in blood and ascites in a PCR laboratory accredited for diagnostic PCR and with long-term experience in developing PCR methods for pathogen detection [28–30]. Despite extensive validation on simulated samples, technical issues interfere with the interpretation of the results of our study, constituting a weakness of the present study. The quantification of bDNA in a number of blood samples and in a single ascites sample failed because of inhibition of the PCR reaction, despite the application of an established strategy, the addition of BSA [31], to reduce this problem. We were, however, able to determine the presence or absence of DNA by our applied methods. The failure of sequencing the DNA-positive samples may be due to several factors. The amount of bDNA in ascites was generally modest, and low levels of bDNA in the samples may have hindered the determination of the target sequence. Furthermore, the amplified DNA targets may be polymicrobial in nature, therefore making the sequences difficult to interpret. This was also suggested by Soriano et al. [24], who were able to successfully sequence only half of the PCR-positive sterile ascites samples.
with low levels of bacteria, PCR may fail simply by missing the bacteria. Conclusion
Our results reveal a low quantity of bDNA in blood and ascites of patients with cirrhosis. We were unable to sequence the PCR products, likely because of the very low amounts of DNA and possibly a polyclonal origin of the DNA. bDNA levels in blood and ascites from individual patients were often not concordant, and bDNA appeared unrelated to markers of inflammation. These findings suggest that the presence of bDNA is unrelated to the immunological and consequent haemodynamic deterioration in cirrhosis. Moreover, bDNA failed to correctly diagnose SBP, rendering the developed assay unsuitable for such diagnostic purposes.
Acknowledgements The authors gratefully acknowledge the help received from Drs Camilla Nøjgaard, Peter Thielsen and Lasse Bremholm, as well as the expert laboratory assistance of Diana Klüver and Lise Lotte Spurr Madsen. C.M., S.M. and F.B. designed the study and wrote the protocol. O.A. advised on immunological aspects of the protocol. The methods for DNA analysis were developed by J.S.J. The markers of immunity were developed by O.A. O.A. analysed markers of immunity. J.S.J. performed DNA analyses. C.M., S.D.L., B.S.M. and L.H. included patients and collected samples. C.M. carried out the statistical analyses. C.M. drafted the manuscript. All authors contributed to the final manuscript. This work is supported by grants from L.F. Foght’s Foundation, Hvidovre Hospitals’ Research Foundation, Hvidovre Hospitals Foundation for Liver Diseases, the Novo Nordisk Foundation and the The Capital Region of Denmark, Foundation for Health Research. Conflicts of interest
There are no conflicts of interest.
References 1 2
The pathogenesis of culture-positive SBP and of bacteria cultured from lymph nodes in experimental cirrhosis has led to the assumption of BT being a monomicrobial event. However, the bDNA detected by PCR in culturenegative ascites may be derived from pathogens that are neither viable nor cultivable, and may indeed be fragments opsonized or phagocytosed by cells of the immune system during previous translocations or subclinical infections. In the culture-positive sample, however, sequencing failure was more surprising. The typical sample volume used for PCR is much smaller compared with the volume used for culture. Accordingly, in samples
3
4
5
6
Tandon P, Garcia-Tsao G. Bacterial infections, sepsis, and multiorgan failure in cirrhosis. Semin Liver Dis 2008; 28:26–42. Arvaniti V, D’Amico G, Fede G, Manousou P, Tsochatzis E, Pleguezuelo M, Burroughs AK. Infections in patients with cirrhosis increase mortality four-fold and should be used in determining prognosis. Gastroenterology 2010; 139:1246–1256. European Association for the Study of the Liver. EASL clinical practice guidelines on the management of ascites, spontaneous bacterial peritonitis, and hepatorenal syndrome in cirrhosis. J Hepatol 2010; 53:397–417. Fernandez J, Navasa M, Gómez J, Colmenero J, Vila J, Arroyo V, Rodés J. Bacterial infections in cirrhosis: epidemiological changes with invasive procedures and norfloxacin prophylaxis. Hepatology 2002; 35:140–148. Berg RD, Garlington AW. Translocation of certain indigenous bacteria from the gastrointestinal tract to the mesenteric lymph nodes and other organs in a gnotobiotic mouse model. Infect Immun 1979; 23:403–411. Follo A, Llovet JM, Navasa M, Planas R, Forns X, Francitorra A, et al. Renal impairment after spontaneous bacterial peritonitis in cirrhosis: incidence, clinical course, predictive factors and prognosis. Hepatology 1994; 20:1495–1501.
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
1366
7
8
9
10
11
12 13
14
15
16
17
18
European Journal of Gastroenterology & Hepatology 2014, Vol 26 No 12
Rasaratnam B, Kaye D, Jennings G, Dudley F, Chin-Dusting J. The effect of selective intestinal decontamination on the hyperdynamic circulatory state in cirrhosis. A randomized trial. Ann Intern Med 2003; 139:186–193. Vlachogiannakos J, Saveriadis AS, Viazis N, Theodoropoulos I, Foudoulis K, Manolakopoulos S, et al. Intestinal decontamination improves liver haemodynamics in patients with alcohol-related decompensated cirrhosis. Aliment Pharmacol Ther 2009; 29:992–999. Albillos A, de la Hera A, González M, Moya JL, Calleja JL, Monserrat J, et al. Increased lipopolysaccharide binding protein in cirrhotic patients with marked immune and hemodynamic derangement. Hepatology 2003; 37:208–217. Bellot P, García-Pagán JC, Francés R, Abraldes JG, Navasa M, PérezMateo M, et al. Bacterial DNA translocation is associated with systemic circulatory abnormalities and intrahepatic endothelial dysfunction in patients with cirrhosis. Hepatology 2010; 52:2044–2052. Albillos A, de-la-Hera A, Alvarez-Mon M. Serum lipopolysaccharide-binding protein prediction of severe bacterial infection in cirrhotic patients with ascites. Lancet 2004; 15:1608–1610. Vallance P, Moncada S. Hyperdynamic circulation in cirrhosis: a role for nitric oxide? Lancet 1991; 337:776–778. Garcia-Tsao G, Wiest R. Gut microflora in the pathogenesis of the complications of cirrhosis. Best Pract Res Clin Gastroenterol 2004; 18:353–372. Such J, Francés R, Muñoz C, Zapater P, Casellas JA, Cifuentes A, et al. Detection and identification of bacterial DNA in patients with cirrhosis and culture-negative, nonneutrocytic ascites. Hepatology 2002; 36:135–141. Zapater P, Francés R, González-Navajas JM, de la Hoz MA, Moreu R, Pascual S, et al. Serum and ascitic fluid bacterial DNA: a new independent prognostic factor in noninfected patients with cirrhosis. Hepatology 2008; 48:1924–1931. Tritto G, Bechlis Z, Stadlbauer V, Davies N, Francés R, Shah N, et al. Evidence of neutrophil functional defect despite inflammation in stable cirrhosis. J Hepatol 2011; 55:574–581. Vlachogiannakos J, Daikos G, Thalheimer U, Burroughs AK, Ladas SD. Is bacterial DNA a better marker than endotoxin of bacterial translocation in decompensated cirrhosis? Hepatology 2011; 53:2140–2141. Wiest R, Das S, Cadelina G, Garcia-Tsao G, Milstien S, Groszmann RJ. Bacterial translocation in cirrhotic rats stimulates eNOS-derived NO production and impairs mesenteric vascular contractility. J Clin Invest 1999; 104:1223–1233.
19
20
21
22
23
24
25
26
27
28
29 30
31
Nadkarni MA, Martin FE, Jacques NA, Hunter N. Determination of bacterial load by real-time PCR using a broad-range (universal) probe and primers set. Microbiology 2002; 148:257–266. Hjorth SV, Björnelius E, Lidbrink P, Falk L, Dohn B, Berthelsen L, et al. Sequence-based typing of Mycoplasma genitalium reveals sexual transmission. J Clin Microbiol 2006; 44:2078–2083. Such J, Muñoz C, Zapater P, Pérez-Mateo M. Bacterial DNA induces a proinflammatory immune response in patients with decompensated cirrhosis. Gut 2005; 54:1500. González-Navajas JM, Bellot P, Francés R, Zapater P, Muñoz C, GarcíaPagan JC, et al. Presence of bacterial-DNA in cirrhosis identifies a subgroup of patients with marked inflammatory response not related to endotoxin. J Hepatol 2008; 48:61–67. Appenrodt B, Lehmann LE, Thyssen L, Gentemann M, Rabe C, Molitor E, et al. Is detection of bacterial DNA in ascitic fluid of clinical relevance? Eur J Gastroenterol Hepatol 2010; 22:1487–1494. Soriano G, Esparcia O, Montemayor M, Guarner-Argente C, Pericas R, Torras X, et al. Bacterial DNA in the diagnosis of spontaneous bacterial peritonitis. Aliment Pharmacol Ther 2011; 33:275–284. Mortensen C, Andersen O, Krag A, Bendtsen F, Møller S. High-sensitivity C-reactive protein levels predict survival and are related to haemodynamics in alcoholic cirrhosis. Eur J Gastroenterol Hepatol 2012; 24:619–626. Perez-Ruiz M, Ros J, Morales-Ruiz M, Navasa M, Colmenero J, Ruiz-delArbol L, et al. Vascular endothelial growth factor production in peritoneal macrophages of cirrhotic patients: regulation by cytokines and bacterial lipopolysaccharide. Hepatology 1999; 29:1057–1063. Rodriguez-Ramos C, Galan F, Díaz F, Elvira J, Martín-Herrera L, GirónGonzález JA. Expression of proinflammatory cytokines and their inhibitors during the course of spontaneous bacterial peritonitis. Dig Dis Sci 2001; 46:1668–1676. Dragsted DM, Dohn B, Madsen J, Jensen JS. Comparison of culture and PCR for detection of Bordetella pertussis and Bordetella parapertussis under routine laboratory conditions. J Med Microbiol 2004; 53:749–754. Jensen JS, Borre MB, Dohn B. Detection of Mycoplasma genitalium by PCR amplification of the 16S rRNA gene. J Clin Microbiol 2003; 41:261–266. Villumsen S, Pedersen R, Krogfelt KA, Jensen JS. Expanding the diagnostic use of PCR in leptospirosis: improved method for DNA extraction from blood cultures. PLoS One 2010; 5:e12095. Al-Soud WA, Radstrom P. Purification and characterization of PCR-inhibitory components in blood cells. J Clin Microbiol 2001; 39:485–493.
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
