bs_bs_banner
doi:10.1111/jgh.12667
H E PAT O L O G Y
Use of proton pump inhibitors decrease cellular oxidative burst in patients with decompensated cirrhosis Irma Garcia-Martinez,* Rubén Francés,*,† Pedro Zapater,*,‡,§ Paula Giménez,* Isabel Gómez-Hurtado,* Alba Moratalla,* Beatriz Lozano-Ruiz,* Pablo Bellot,¶ José M González-Navajas* and José Such*,† *CIBERehd Hospital General Universitario de Alicante, Alicante, †Department of Clinical Medicine, §Instituto de Bioengeniería, Miguel Hernández University, Elche, Alicante, and ‡Clinical Pharmacology Unit and ¶Liver Unit, Hospital General Universitario de Alicante, Alicante, Spain
Key words bacterial DNA, cirrhosis, norfloxacin, oxidative burst, proton pump inhibitors. Accepted for publication 21 May 2014. Correspondence Dr José Such, Liver Unit, Hospital General Universitario de Alicante, and Department of Clinical Medicine, Miguel Hernández University, Elche, Alicante 03202, Spain. Email: [email protected]
Abstract Background and Aims: Proton pump inhibitors (PPIs) are commonly used antisecretory drugs and have been linked to an increased risk of bacterial infections in cirrhosis. We investigated whether the treatment with PPIs in cirrhosis affects the oxidative burst activity of granulocytes and monocytes and its possible interference with serum norfloxacin (Nflx) levels in these patients. Methods: 70 patients with cirrhosis and ascitic fluid and 24 healthy controls were included in the study and distributed into groups according to the regular use of PPIs and/or norfloxacin. The blood granulocyte and monocyte’s phagocytic activity and oxidative burst were evaluated by flow cytometry. Blood levels of norfloxacin were measured by HPLC and bacterial translocation was evaluated by detection of bacterial DNA in blood. Results: Use of PPIs was associated with a decreased granulocyte and monocyte oxidative burst, but not of phagocytic activity, as compared with patients not receiving PPIs. PPIs use did not affect serum norfloxacin levels in patients. A not significant trend to an increased bacterial DNA translocation was observed in patients receiving PPIs, including patients simultaneously receiving norfloxacin. Conclusions: PPIs significantly decrease cellular oxidative burst in cirrhosis. This fact may provide a pathogenic explanation to the reported high rates of bacterial infections in this setting, and strongly suggests that PPIs should only be used in patients with cirrhosis when clinically indicated.
Introduction Bacterial infections are common complications arising in patients with decompensated cirrhosis.1 The most relevant of them, such as bacteraemia or bacterial peritonitis, are called “spontaneous” since they have their origin in the intestinal bacterial content that can cross an excessively permeable intestinal wall accessing mesenteric lymph nodes and other extraintestinal sites in a process termed bacterial translocation (BT).2 The likelihood of developing bacterial infections is therefore related to the viability of the accessing bacteria and the capacity of the host immune system to mount an adequate bactericidal response.3 Several immune deficiencies have been described in advanced cirrhosis, either local or systemic, that favor the development of bacterial infections.1 In fact, cirrhosis may be considered as an acquired immunodeficient disease.4 Therefore, circumstances leading to an increased risk of infections in these patients may be taken into serious consideration, not only because bacterial infections compromise the survival of patients per se, but also because
the inflammatory response associated with these infections is disproportionate when compared to similar infections in noncirrhotic patients.5 We have previously shown that bacterial DNA (bactDNA) translocation induces an inflammatory response that is similar to that observed in fully developed infections6 and may impair prognosis.7 Proton pump inhibitors (PPIs), such as omeprazole, constitute a family of frequently used antisecretory drugs, either with or without a clear clinical indication, according to the general belief that are “protective” drugs. An adequate indication for omeprazole use exists only in half of cases prescribed,8,9 and this same proportion is thought to occur in cirrhotic patients as well, converting PPIs in one of the most frequently prescribed drugs in this population.10 The use of PPIs has been associated with an increased risk of developing nosocomial pneumonia in general population,11 or spontaneous bacterial peritonitis (SBP) among patients with cirrhosis and ascites.12,13 PPIs can favor intestinal bacterial overgrowth14 and affect the immune cellular activity in healthy donors,15 thereby increasing the risk of developing bacterial infections.
Journal of Gastroenterology and Hepatology 30 (2015) 147–154 © 2014 Journal of Gastroenterology and Hepatology Foundation and Wiley Publishing Asia Pty Ltd
147
Proton pump inhibitors and cirrhosis
I Garcia-Martinez et al.
Norfloxacin (Nflx) is a widely used medication in patients with decompensated cirrhosis either as primary or secondary prophylaxis of bacterial infections.16–18 We have previously reported that Nflx modulates neutrophil function and apoptosis,19 a mechanism probably beneficial in this setting. However, no information is available regarding the possible interactions between PPIs and Nflx and their effect on neutrophil function. The aims of this work have been (i) to assess the effect of PPIs treatment on both the phagocytosis and oxidative burst of innate immune cells from patients with decompensated cirrhosis, (ii) to investigate the relationship between PPI use and the presence of bactDNA, and (iii) to evaluate the possible interactions of PPIs with the concomitant use of Nflx.
Patients and methods Patients and study design. Seventy consecutively admitted patients with cirrhosis and ascitic fluid (AF) and 24 healthy donors from Centro de Transfusión de Alicante, San Juan de Alicante, Spain, were included in the study and distributed into groups according to the regular use of PPIs, starting at least 2 weeks before inclusion. Patients were further stratified according to the concomitant use of Nflx as primary prophylaxis or secondary prophylaxis of SBP. Cirrhosis was diagnosed by histological, clinical, laboratory, and/or ultrasonographic findings. Exclusion criteria were the presence of a culture-positive blood or AF, intake of antibiotics other than Nflx in the previous 2 weeks, SBP, upper gastrointestinal bleeding, hepatocellular carcinoma that exceeds Milan criteria,20 and/or portal thrombosis. The Ethics Committee of the General Hospital of Alicante approved the study protocol, and all patients provided their informed consent before being included in the study. Blood was obtained for routine hematological, biochemical, and coagulation studies. Simultaneously, a large-volume paracentesis was performed in all patients at admission in aseptic conditions following the usual procedures,21 and samples for routine biochemical study and Polymorphonuclear leucocytes (PMN) count were obtained. Total protein, albumin, leukocyte count, and PMN count were performed in all AF specimens. Both blood and AF were inoculated at bedside in aerobic and anaerobic blood culture bottles (10 mL each). Finally, separate blood and AF samples were inoculated, under aseptic conditions, in rubber-sealed sterile Vacutainer SST II tubes (BD Diagnostics, Belgium) that were never exposed to free air. Phagocytic assay. The phagocytic activity was evaluated using a Phagotest kit (Orpegen Pharma, Heidelberg, Germany) following the manufacturer’s protocol. Two cellular activities are measured in this assay; the in vitro percentage of granulocytes and monocytes that are able to ingest fluorescein-isothiocyanate (FITC)-labeled opsonized Escherichia coli bacteria, and their phagocytic capacity (total number of bacteria that is ingested per cell). In brief, 100 μL of heparinized whole blood were incubated with 20 μL of FITC-labeled opsonized E. coli or without any stimulus (negative control). The subsequent ingestion of the bacteria by phagocytes generates a green fluorescence signal that is quantified by flow cytometry. Phagocytosis was expressed as a percentage of activated phagocytes and as the mean fluorescence intensity (MFI). All samples were analyzed in triplicate. 148
Oxidative burst assay. The oxidative burst was measured with a Phagoburst kit (Orpegen Pharma, Heidelberg, Germany) according to the manufacturer’s protocol. Similar to the phagocytic assay, this kit measures both the percentage of granulocytes and monocytes that generate reactive oxygen species (ROS), and their oxidative burst activity (number of ROS per cell). In brief, 100 μL of heparinized whole blood was incubated with either 20 μL of unlabelled opsonized E. coli, 20 μL of N-formylmethionyl-leucyl-phenylalanine (fMLP), 20 μL of phorbol-12-myristate-13-acetate (PMA) or left unstimulated (negative control). The formation of ROS was monitored by the oxidation of dihydrorhodamine 123 to rhodamine, which produces a green fluorescence signal. Oxidative burst was expressed as a percentage of activated phagocytes and as the MFI. All samples were analyzed in triplicate. Flow cytometric analysis. Cells are analyzed by flow cytometry using the blue-green excitation light (488 nm) in a FacsCanto cytometer (Becton Dickinson, Heidelberg, Germany), and the data were processed using the FacsDiva 12.0 software. Neutrophils were identified and gated by the software program in the scatter diagram (Forward light scatter (FSC) vs Side light scatter (SSC)) and their green fluorescence histogram (FL1) was analyzed. We collected 10.000–15.000 leukocytes per sample. A control sample was used for setting an FL1 marker so that less than 1–3% of the events are positive. The parameters in the test sample can then be determined by counting the number of events above this marker position. Bacterial DNA isolation, amplification, and sequencing. Bacterial translocation was defined as the presence on bactDNA in blood. DNA extraction was performed by handling Qiagen Blood Minikit (Hilden, Germany) according to manufacturer’s directions. A broad-range polymerase chain reaction (PCR) and partial nucleotide sequencing analysis of prokaryote 16S rRNA were run in all serum and AF samples for amplification an identification bacterial DNA fragments as previously described.22 The primers used were: 5_ TTCCGGTTGATC CTGCCGGA 3_ as forward, and 5_ GGTTACCTTGTTACGA CTT 3_ as reverse.23 The same forward primer used for PCR amplification was used for sequencing. Sequences obtained were compared with 16S rRNA sequences available both in the Ribosomal Database Project24 and the GenBank and European Molecular Biology Laboratory obtained from the National Center for Biotechnology Information Database by the advanced Basic Local Alignment Search Tool (BLAST) search tool.25 Measurement of Norfloxacin levels. Simultaneously to the other clinical and analytical measurements described above, we analyzed the concentration of Nflx in plasma samples from patients taking Nflx using a HPLC method as previously described.19 Briefly, chromatographic separation was performed using a reverse-phase Eclipse XDB-C18 (Agilent, Santa Clara, CA, USA.) column (150x4.6 mm, Φ = 5μm). The mobile phase, Acetonitrile-tetrabutyl ammonium hydroxiphosphate buffer, was eluted at a flow rate of 1 mL/min, and effluent was monitored at
Journal of Gastroenterology and Hepatology 30 (2015) 147–154 © 2014 Journal of Gastroenterology and Hepatology Foundation and Wiley Publishing Asia Pty Ltd
I Garcia-Martinez et al.
Table 1
Proton pump inhibitors and cirrhosis
Clinical and analytical characteristics of controls and patients Control
Age (years) Gender (M/F) Etiology of cirrhosis, n Alcohol Hepatitis C Virus (HCV) Alcohol+HCV Other Drug history β-blockers agents Spironolactone Furosemide Lactulose Norfloxacin Refractory Ascites Previous hepatic encephalopathy Previous variceal bleeding T (°C) MAP (mm Hg) MELD score Child-pugh score Serum creatinine (mg/dL) Serum sodium (mEq/L) Serum albumin (mg/dL) Serum bilirubin (mg/dL) Leukocytes (× 103/μL) Neutrophils (× 103/μL) Quick (%) AF glucose (mg/dL) AF total protein (g/dL) AF RBCs (/μL) AF leukocytes (/μL)
Patient
PPI- (n = 15)
PPI+ (n = 9)
37 ± 10 5/10
50 ± 21 4/5
— — — —
— — — —
— — — — — — — — — — — — 0.8 ± 0.2 141 ± 2 — 0.6 ± 0.3 5.5 ± 1.6 3.2 ± 1.0 97 ± 4 — — — —
— — — — — — — — — — — — 0.8 ± 0.2 139 ± 2 — 0.4 ± 0.2 6.2 ± 1.8 3.8 ± 1.6 85 ± 10 — — — —
PPI- (n = 26) 61 ± 11 20/6 17 4 4 1 4 (15.4%) 11 (42.3%) 10 (38.5%) 2 (7.7%) 24 (92.3%) 8 (30.8%) 2 (7.7%) 20 (76.9%) 36.1 ± 0.4 104.9 ± 17.8 11 ± 4 9±1 0.9 ± 0.4 136 ± 5 2828 ± 665 2.3 ± 1.1 5.9 ± 3.3 3.9 ± 2.4 56 ± 14 137 ± 50 1.7 ± 0.9 1691 ± 3635 133 ± 109
PPI+ (n = 44) 62 ± 10 37/7 25 4 10 5 23 (52.3%)* 27 (61.4%) 35 (79.5%)* 11 (25%) 26 (59.1%)* 24 (54.5%) 12 (27.3%) 39 (88.6%) 36.4 ± 0.5 97.9 ± 15.1 13 ± 5 9±2 1.1 ± 0.5 133 ± 6 * 3029 ± 605 2.5 ± 1.7 5.1 ± 2.6 3.3 ± 1.8 59 ± 12 143 ± 39 1.5 ± 0.7 1951 ± 3337 182 ± 163
Values are expressed as mean ± standard deviation. *P < 0.05 compared with patients proton pump inhibitor (PPI) group. AF, ascitic fluid; MAP, mean arterial pressure; MELD, model for end-stage liver disease; RBC, red blood cells.
excitation and emission wavelengths of 278 and 456 nm, respectively. The limit of detection was 0.06 μg/mL. The calibration graph was found to be linear from 0.0625 to 4 μg/mL. Norfloxacin was obtained from Sigma Chemical Co (Madrid, Spain). Serum samples preparation for HPLC injection: A 100 μL plasma sample was diluted with 100 μL of acetonitrile, and the mixture shaken and centrifuged at 6000 rpm for 10 min. A 60 μL aliquot of supernatant was injected, and peak areas were recorded. Serum Nflx concentrations were expressed in micrograms/ milliliters. Measurement of TNF-α and nitric oxide (NO) products. Immunoenzymometric assay for quantitative measurement of human tumor necrosis factor-alpha (TNF)-α and nitric oxide (NO) levels in serum samples were performed using human TNF-α HS Quantikine and NOx ELISA kit, (R&D Systems) respectively, according to the manufacturer’s instructions. All samples were tested in duplicate and read at 450 nm (TNF-α) and 540 nm (NOx) in a ThermoMax microplate reader (Molecular
Devices, Sunnyvale, CA, USA). Standard curves were generated for each plate, and the average zero standard optical densities were subtracted from the remaining standards, controls, and samples to obtain a corrected concentration for TNF-a.
Statistical analysis. Continuous variables are reported as mean ± standard deviation and categorical variables as frequencies or percentages. Statistical differences between groups were analyzed using the Fisher test for categorical data and the Mann– Whitney U-test for quantitative data. Spearman’s linear correlation and Mann–Whitney U-test were initially performed to identify quantitative and qualitative variables that were significantly associated with phagocytosis and bust oxidative parameters. Analysis of covariance (ANCOVA) was used to further examine the association between variables selected by the previous analysis and the parameters of phagocytosis and oxidative burst. All reported P values are two-sided, and P values lower than 0.05 were considered to indicate significance. All calculations were performed using the SPSS 19.0 software. (SPSS, Inc, Chicago, IL, USA).
Journal of Gastroenterology and Hepatology 30 (2015) 147–154 © 2014 Journal of Gastroenterology and Hepatology Foundation and Wiley Publishing Asia Pty Ltd
149
Proton pump inhibitors and cirrhosis
I Garcia-Martinez et al.
Results
Table 2 Clinical and analytical characteristics of patients receiving PPIs, according to the use of Nflx
Characteristics of patients and laboratory data. The overall series of patients (n = 70) were divided into two groups according to the previous use of PPIs: PPI(+), n = 44, PPI(-), n = 26. Similarly, controls were divided into two groups according to the same criteria: control PPI(+), n = 9, control PPI(-), n = 15. Most patients undergoing PPI treatment were on omeprazole (91%), while esomeprazole and pantoprazole use accounted for 4.5% of cases each. Within the omeprazole group, most patients were receiving a dose of 20 mg QD (82.5%), and the remaining patients received 40 mg (10%) or 60 mg/d (7.5%). Only 22.7% of patients had been treated with PPIs during less than 1 year. Clinical and analytical characteristics of all groups of patients and controls according to the use of PPIs are detailed in Table 1. No statistically significant differences were observed between groups of controls. As expected, parameters reflecting liver function were different between controls and patients, irrespective of the use of PPIs. Statistically significant differences were observed between patients receiving or not PPIs in the number of coadministered medications and serum sodium levels (Table 1). The patients on PPIs were taking beta (β)-blockers, furosemide, and norfloxacin in a higher proportion than patients without PPIs. And, although the episodes of refractory ascites, hepatic encephalopathy, and variceal bleeds were higher in those patients (PPIs+), differences were not statistically significant. Patients receiving PPIs were further distributed according to the use of Nflx as secondary prophylaxis of SBP (PPI+Nflx group, n = 26 and PPI+Nflx+ group, n = 18). The clinical and analytical characteristics of both groups are detailed in Table 2. The proportion of patients with refractory ascites, previous episodes of hepatic encephalopathy, or previous episodes of variceal bleeding was significantly higher in patients receiving Nflx. Also, a higher proportion of PPIs+Nflx+ were receiving β-blocker treatment, and their mean arterial pressure was significantly lower compared with PPIs+Nflx- patients. BactDNA was present in 4 (15.4%) patients without PPIs (E. coli (n = 3) and K pneumoniae) and in 12 (27.3%) patients with PPIs, although values did not reach significance (P NS). Among patients receiving PPIs but not Nflx, bactDNA was detected in eight cases (30.7%), a percentage similar to previously described in patients with ascites (E. coli [n = 6], S aureus and C. Freundii). Of interest, four out of the 18 patients receiving Nflx, and PPIs (22.2%) showed the presence of bactDNA in blood, a rate higher than expected in patients receiving Nflx. Species identified included Shigella, E. coli (n = 2), and Klebsiella pneumoniae. No relationship was found among the type or dosage of PPI, the time of its use and the presence of bactDNA. Similarly, neither the length of therapy with Nflx nor the serum Nflx levels in blood were related with the presence of bactDNA (data not shown).
The use of PPIs does not affect cellular phagocytic capacity in patients. The percentage of granulocytes and monocytes showing phagocytosis at baseline, and their intrinsic phagocytic capacity (number of bacteria ingested per cell) were considered as two separated parameters. 150
Age (years) Gender (M/F) Etiology of cirrhosis, n Alcohol HCV Alcohol+HCV Other Drug history β-blockers agents Spironolactone Furosemide Lactulose Refractory ascites Previous hepatic encephalopathy Previous variceal bleeding Presence bacterial DNA T (°C) MAP (mm Hg) MELD Score Child-pugh score Serum creatinine (mg/dL) Serum sodium (mEq/L) Serum albumin (mg/dL) Serum bilirubin (mg/dL) Leukocytes (× 103/μL) Neutrophils (× 103/μL) Quick (%) AF glucose (mg/dL) AF total protein (g/dL) AF RBCs (/μL) AF leukocytes (/μL)
Nflx- (n = 26)
Nflx+ (n = 18)
63 ± 10 21/5
60 ± 9 16/2
14 2 6 4
11 2 4 1
9 (34.6%) 14 (53.8%) 19 (73.1%) 6 (23.1%) 11 (42.3%) 4 (15.4%) 21 (80.8%) 8 (30.8%) 36.5 ± 0.3 107 ± 9 13 ± 5 8.5 ± 1.6 1.1 ± 0.5 134 ± 7 2971 ± 673 2.5 ± 1.9 5.8 ± 2.8 3.8 ± 2.1 61 ± 13 132 ± 36 1.6 ± 0.7 1307 ± 3059 170 ± 149
14 (77.8%)* 13 (72.2%) 16 (88.9%) 5 (27.8%) 13 (72.2%)* 8 (44%)* 18 (100%)* 4 (22.2%) 36.2 ± 0.5 87 ± 14* 15 ± 3 9.5 ± 1.6 1.1 ± 0.5 131 ± 5 3175 ± 402 2.5 ± 1.3 4.1 ± 1.8 2.6 ± 0.9 54 ± 9 159 ± 39* 1.5 ± 0.8 2838 ± 3594 199 ± 184
Values are expressed as mean ± standard deviation. *P < 0.05 compared with patients Norfloxacin group. AF, ascitic fluid; MAP, mean arterial pressure; MELD, model for endstage liver disease; RBC, red blood cells.
Previous use of PPIs was not associated with differences in the baseline percentage of granulocytes and monocytes showing phagocytosis neither in controls nor in patients (Supplementary Fig. S1). Baseline phagocytic capacity in the absence of PPIs was similar between patients and controls. However, the use of PPIs was associated with a significantly higher phagocytic capacity in patients than in controls (Fig. 1a). The concomitant use of Nflx did not affect the phagocytic capacity observed among patients receiving PPIs and no correlation was found between serum Nflx levels and the total amount of ingested bacteria (Fig. 1b). The presence of bacterial DNA did not affect the phagocytic capacity in our series of PPI+ versus PPI- patients (Fig. 1c). The use of PPIs decreases the cellular oxidative burst in patients with decompensated cirrhosis. The percentage of granulocytes and monocytes showing oxidative burst at baseline and their intrinsic oxidative burst capacity (number of ROS released per cell) were considered as two
Journal of Gastroenterology and Hepatology 30 (2015) 147–154 © 2014 Journal of Gastroenterology and Hepatology Foundation and Wiley Publishing Asia Pty Ltd
I Garcia-Martinez et al.
Proton pump inhibitors and cirrhosis
(a)
(b)
(c)
Figure 1 Phagocytic capacity of granulocytes represented by the mean fluorescence intensity (MFI) measured by flow cytometry. (a) Phagocytic capacity of granulocytes from controls and patients, according to the use of proton pump inhibitors (PPIs). (b) Phagocytic capacity of granulocytes from patients distributed by norfloxacin, according to the use of PPIs and correlation between serum levels of norfloxacin and phagocytic capacity. (c) Phagocytic capacity of granulocytes from patients distributed by the presence of blood bactDNA according to the use of PPIs. Nflx: norfloxacin; bactDNA: bacterial DNA; *P < 0.05. □, PPI−; ■, PPI+.
separated parameters and was evaluated at baseline and after in vitro stimulation with fMLP (weak stimulus), PMA (strong stimulus), and E. coli. Previous use of PPIs did not induce changes in the percentage of granulocytes and monocytes showing oxidative burst either in controls or in patients (Supplementary Fig. S2). However, the intake of PPIs was associated with a significant decrease in the total amount of ROS released by monocytes and granulocytes, both in controls and patients at baseline and all stimulated conditions (Fig. 2). No significant differences were observed in the reduction of burst activity between controls and patients (data not shown). Univariate analysis showed a significant association between oxidative burst activity and previous use of PPIs (P = 0.001), age (categorized as < 60 years and > 60 years; P = 0.006), leukocytes (P = 0.03), neutrophils (P = 0.013), and potassium (P = 0.005). The use of PPIs (P = 0.001), as well as age (P = 0.004), remained
as significant independent variables associated with burst activity in the ANCOVA multivariate analysis including all previous variables. The use of Nflx did not affect the oxidative burst activity observed among cirrhotic patients receiving PPIs, and no correlation was found between serum Nflx levels and the total amount of ROS released by either monocytes or granulocytes. The presence of bacterial DNA did not affect the oxidative burst activity in our series of PPI+ versus PPI- patients. *The results on phagocytosis and oxidative burst shown in the present study were similar for granulocyte and monocyte populations in all cases.
Serum soluble inflammatory response not affected by PPIs in patients. TNF-alpha and NOx levels in the serum of patients distributed according to the use of PPIs revealed no
Journal of Gastroenterology and Hepatology 30 (2015) 147–154 © 2014 Journal of Gastroenterology and Hepatology Foundation and Wiley Publishing Asia Pty Ltd
151
Proton pump inhibitors and cirrhosis
I Garcia-Martinez et al.
Figure 2 Oxidative burst activity (amount of reactive oxygen species, ROS, released) of granulocytes from controls and patients. Oxidative burst activity of granulocytes from controls and patients, according to the use of proton pump inhibitors (PPIs) at baseline and in response to several stimuli. fMLP: N- formylmethionyl-leucyl-phenylalanine; PMA: phorbol-12-myristate-13-acetate. *P < 0.05 compared with the PPI- group. □, PPI−; ■, PPI+.
statistically significant differences at baseline. Distribution of patients by intake of Nflx or the presence of bacterial DNA showed no differences on these inflammatory mediators (data not shown).
Discussion In this investigation, we show evidences that omeprazole, as representative drug of the family of PPIs, significantly reduces the oxidative burst activity of granulocytes and monocytes in patients with decompensated cirrhosis, which is a key mechanism of the immune system bactericidal activity. Notably, these effects are not modified by the simultaneous administration of Nflx. PPIs constitute the third most prescribed medications in the United States,26 and its use has been recently associated with an increased risk of developing pulmonary infections in admitted patients11,27 and SBP in patients with decompensated cirrhosis.12,13 Although the reasons behind this fact may be several, the main deleterious effect associated with their use was initially attributed to the relationship between hypochlorhydria and bacterial survival in the stomach.28 Additional investigations in animal models showed that inhibition of gastric acid secretion may be related to increased intestinal paracellular permeability,29 overgrowth of small bowel microflora,30 and an increased rate of BT.31 In addition, Omeprazole has been shown to inhibit both the neutrophil function in vitro32 and the bactericidal activity in general population.15 However, no data are available on the effect of PPIs on monocyte and neutrophil function in patients with cirrhosis. This is relevant, since omeprazole is largely used in these patients, who already have a compromised neutrophil function at baseline.33 In the present study, we have demonstrated that PPIs indeed decrease both granulocyte and monocyte’s oxidative burst activity in patients with decompensated cirrhosis, a surrogate marker of an impaired bactericidal activity. This fact that may provide a mechanistic explanation to the reported increased incidence of bacterial infections observed in other studies, both in controls and in patients with cirrhosis.11,13 Although this effect was not associated 152
with a significantly increased rate of bacterial DNA translocation in this investigation, an interesting finding is the fact that patients receiving PPIs plus Nflx showed a rate of bacterial DNA translocation higher than usual (Table 2). This suggests that PPIs use may decrease the Nflx efficacy to prevent bacterial translocation in cirrhosis. This hypothesis though should be confirmed in a specifically designed study. Despite not observing differences in bactDNA translocation rates, the oxidative burst activity downregulation associated with the intake of PPIs in patients with decompensated cirrhosis remains relevant from an immunological point of view, since it may point towards a compromised situation in an eventual bacterial infection challenge. An impaired neutrophil function, as observed in cirrhotic patients in this study, compromises a key part of the innate immune system, the first line of host resistance against bacterial infection. Adding this fact to the already deranged cellular situation of these patients provides a concern for a more careful decision on whether start PPIs in this setting. Patients on Nflx did not show a different capacity of oxidative burst irrespective of the concomitant use of PPIs. The same lack of association was observed between the presence of bacterial DNA in blood, serum inflammatory markers, and oxidative burst activity in our series. On the contrary, the ANCOVA analysis determined that age, and the use of PPIs were the only variables independently associated with a decreased oxidative burst activity. Elderly patients (> 60years group) shown lower levels of ROS released than younger patients. In fact, a decrease in phagocytosis of opsonized bacteria by neutrophils34 and in the production of superoxide by PMNs or antioxidants such as glutathione in the have been shown to occur in the elderly.35,36 The fact that PPIs are not associated with a lower phagocytic capacity in patients with cirrhosis compared with controls may be a consequence of the general immunological derangement associated with the course of disease. This pre-activated state may account for preventing a downregulation of the phagocytic activity in patients. Although the effect of PPIs on oxidative burst activity had been previously evaluated in healthy donors and in in vitro assays, we
Journal of Gastroenterology and Hepatology 30 (2015) 147–154 © 2014 Journal of Gastroenterology and Hepatology Foundation and Wiley Publishing Asia Pty Ltd
I Garcia-Martinez et al.
report first-time evidence of their deleterious effect on the oxidative burst activity of granulocytes and monocytes from patients with decompensated cirrhosis and provide rationale for limiting PPIs use in cirrhosis when clinically indicated, far from the supposed role of “protective” drug. In summary, PPIs use decrease the amount of ROS released during the oxidative burst by granulocytes and monocytes in patients with decompensated cirrhosis, thereby decreasing bactericidal activity and giving a pathogenic explanation to the reported increased rate of bacterial infections in patients with cirrhosis receiving PPIs.
Acknowledgments This work has been supported in part with grants from Instituto de Salud Carlos III (PI08/1075), Excma. Diputación Provincial de Alicante, Spain and Fundación FCVI-HGUA, Alicante, Spain.
References 1 Such J, Runyon BA. Spontaneous bacterial peritonitis. Clin. Infect. Dis. 1998; 27: 669–74; quiz 75–6. 2 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: 403–11. 3 Rimola A, Soto R, Bory F, Arroyo V, Piera C, Rodes J. Reticuloendothelial system phagocytic activity in cirrhosis and its relation to bacterial infections and prognosis. Hepatology 1984; 4: 53–8. 4 Guarner C, Runyon BA. Spontaneous bacterial peritonitis: pathogenesis, diagnosis, and management. Gastroenterologist 1995; 3: 311–28. 5 Byl B, Roucloux I, Crusiaux A, Dupont E, Deviere J. Tumor necrosis factor alpha and interleukin 6 plasma levels in infected cirrhotic patients. Gastroenterology 1993; 104: 1492–7. 6 Frances R, Zapater P, Gonzalez-Navajas JM et al. Bacterial DNA in patients with cirrhosis and noninfected ascites mimics the soluble immune response established in patients with spontaneous bacterial peritonitis. Hepatology 2008; 47: 978–85. 7 Zapater P, Frances R, Gonzalez-Navajas JM et al. Serum and ascitic fluid bacterial DNA: a new independent prognostic factor in noninfected patients with cirrhosis. Hepatology 2008; 48: 1924–31. 8 Naunton M, Peterson GM, Bleasel MD. Overuse of proton pump inhibitors. J. Clin. Pharm. Ther. 2000; 25: 333–40. 9 Nardino RJ, Vender RJ, Herbert PN. Overuse of acid-suppressive therapy in hospitalized patients. Am. J. Gastroenterol. 2000; 95: 3118–22. 10 Chavez-Tapia NC, Tellez-Avila FI, Garcia-Leiva J, Valdovinos MA. Use and overuse of proton pump inhibitors in cirrhotic patients. Med. Sci. Monit. 2008; 14: CR468–72. 11 Herzig SJ, Howell MD, Ngo LH, Marcantonio ER. Acid-suppressive medication use and the risk for hospital-acquired pneumonia. JAMA 2009; 301: 2120–8. 12 Goel GA, Deshpande A, Lopez R, Hall GS, van Duin D, Carey WD. Increased rate of spontaneous bacterial peritonitis among cirrhotic patients receiving pharmacologic Acid suppression. Clin. Gastroenterol. Hepatol. 2012; 10: 422–7. 13 Bajaj JS, Zadvornova Y, Heuman DM et al. Association of proton pump inhibitor therapy with spontaneous bacterial peritonitis in cirrhotic patients with ascites. Am. J. Gastroenterol. 2009; 104: 1130–4.
Proton pump inhibitors and cirrhosis
14 Fried M, Siegrist H, Frei R et al. Duodenal bacterial overgrowth during treatment in outpatients with omeprazole. Gut [Clinical Trial Controlled Clinical Trial Research Support, Non-U.S. Gov’t]. 1994;35: 23–6. 15 Zedtwitz-Liebenstein K, Wenisch C, Patruta S, Parschalk B, Daxbock F, Graninger W. Omeprazole treatment diminishes intraand extracellular neutrophil reactive oxygen production and bactericidal activity. Crit. Care Med. 2002; 30: 1118–22. 16 Soriano G, Guarner C, Tomas A et al. Norfloxacin prevents bacterial infection in cirrhotics with gastrointestinal hemorrhage. Gastroenterology 1992; 103: 1267–72. 17 Fernandez J, Ruiz del Arbol L, Gomez C et al. Norfloxacin vs ceftriaxone in the prophylaxis of infections in patients with advanced cirrhosis and hemorrhage. Gastroenterology 2006; 131: 1049–56; quiz 285. 18 Fernandez J, Navasa M, Planas R et al. Primary prophylaxis of spontaneous bacterial peritonitis delays hepatorenal syndrome and improves survival in cirrhosis. Gastroenterology 2007; 133: 818–24. 19 Zapater P, Cano R, Llanos L et al. Norfloxacin modulates the inflammatory response and directly affects neutrophils in patients with decompensated cirrhosis. Gastroenterology 2009; 137: 1669–79. 20 Mazzaferro V, Regalia E, Doci R et al. Liver transplantation for the treatment of small hepatocellular carcinomas in patients with cirrhosis. N. Engl. J. Med. 1996; 334: 693–9. 21 Runyon BA. Paracentesis of ascitic fluid. A safe procedure. Arch. Intern. Med. 1986; 146: 2259–61. 22 Such J, Frances R, Munoz C et al. Detection and identification of bacterial DNA in patients with cirrhosis and culture-negative, nonneutrocytic ascites. Hepatology 2002; 36: 135–41. 23 Lane DJ. 16S/23S rRNA sequencing. In: Stackebrandt E, Goodfellow M, eds. Nucleic Acid Techniques in Bacterial Systematics. New York: John Wiley and Sons, 1991; 115–75. 24 Maidak BL, Cole JR, Lilburn TG et al. The RDP-II (Ribosomal Database Project). Nucleic Acids Res. 2001; 29: 173–4. 25 Altschul SF, Madden TL, Schaffer AA et al. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 1997; 25: 3389–402. 26 Bavishi C, Dupont HL. Systematic review: the use of proton pump inhibitors and increased susceptibility to enteric infection. Aliment. Pharmacol. Ther. 2011; 34: 1269–81. 27 Laheij RJ, Sturkenboom MC, Hassing RJ, Dieleman J, Stricker BH, Jansen JB. Risk of community-acquired pneumonia and use of gastric acid-suppressive drugs. JAMA 2004; 292: 1955–60. 28 Tennant SM, Hartland EL, Phumoonna T et al. Influence of gastric acid on susceptibility to infection with ingested bacterial pathogens. Infect. Immun. 2008; 76: 639–45. 29 Hopkins AM, McDonnell C, Breslin NP, O’Morain CA, Baird AW. Omeprazole increases permeability across isolated rat gastric mucosa pre-treated with an acid secretagogue. J. Pharm. Pharmacol. 2002; 54: 341–7. 30 Lichtman SM. Bacterial [correction of bacterial] translocation in humans. J. Pediatr. Gastroenterol. Nutr. 2001; 33: 1–10. 31 Basaran UN, Celayir S, Eray N, Ozturk R, Senyuz OF. The effect of an H2-receptor antagonist on small-bowel colonization and bacterial translocation in newborn rats. Pediatr. Surg. Int. 1998; 13: 118–20. 32 Wandall JH. Effects of omeprazole on neutrophil chemotaxis, super oxide production, degranulation, and translocation of cytochrome b-245. Gut 1992; 33: 617–21. 33 Fiuza C, Salcedo M, Clemente G, Tellado JM. In vivo neutrophil dysfunction in cirrhotic patients with advanced liver disease. J. Infect. Dis. 2000; 182: 526–33. 34 Wenisch C, Patruta S, Daxbock F, Krause R, Horl W. Effect of age on human neutrophil function. J. Leukoc. Biol. 2000; 67: 40–5.
Journal of Gastroenterology and Hepatology 30 (2015) 147–154 © 2014 Journal of Gastroenterology and Hepatology Foundation and Wiley Publishing Asia Pty Ltd
153
Proton pump inhibitors and cirrhosis
I Garcia-Martinez et al.
35 Braga PC, Sala MT, Dal Sasso M, Mancini L, Sandrini MC, Annoni G. Influence of age on oxidative bursts (chemiluminescence) of polymorphonuclear neutrophil leukocytes. Gerontology 1998; 44: 192–7. 36 Corberand J, Ngyen F, Laharrague P et al. Polymorphonuclear functions and aging in humans. J. Am. Geriatr. Soc. 1981; 29: 391–7.
of granulocytes from controls and patients, according to the use of PPIs. (b) Percentage of granulocytes from patients distributed by norfloxacin, according to the use of PPIs and correlation between serum levels of norfloxacin and percentage of phagocytosis. (c) Percentage of granulocytes from patients distributed by the presence of blood bactDNA according to the use of PPIs. Nflx, norfloxacin; bactDNA, bacterial DNA; *P < 0.05.
Supporting information
Figure S2 Percentage of granulocytes from controls and patients that generate reactive oxygen species (ROS). Percentage of granulocytes generating ROS from controls and patients, according to the use of PPIs at baseline and in response to several stimuli. fMLP, N-formylmethionyl-leucyl-phenylalanine; PMA, phorbol-12-myristate-13-acetate.
Additional Supporting Information may be found in the online version of this article at the publisher’s web-site: Figure S1 Percentage of granulocytes from controls and patients that ingest bacteria, measured by flow cytometry. (a). Percentage
154
Journal of Gastroenterology and Hepatology 30 (2015) 147–154 © 2014 Journal of Gastroenterology and Hepatology Foundation and Wiley Publishing Asia Pty Ltd
doi:10.1111/jgh.12667
H E PAT O L O G Y
Use of proton pump inhibitors decrease cellular oxidative burst in patients with decompensated cirrhosis Irma Garcia-Martinez,* Rubén Francés,*,† Pedro Zapater,*,‡,§ Paula Giménez,* Isabel Gómez-Hurtado,* Alba Moratalla,* Beatriz Lozano-Ruiz,* Pablo Bellot,¶ José M González-Navajas* and José Such*,† *CIBERehd Hospital General Universitario de Alicante, Alicante, †Department of Clinical Medicine, §Instituto de Bioengeniería, Miguel Hernández University, Elche, Alicante, and ‡Clinical Pharmacology Unit and ¶Liver Unit, Hospital General Universitario de Alicante, Alicante, Spain
Key words bacterial DNA, cirrhosis, norfloxacin, oxidative burst, proton pump inhibitors. Accepted for publication 21 May 2014. Correspondence Dr José Such, Liver Unit, Hospital General Universitario de Alicante, and Department of Clinical Medicine, Miguel Hernández University, Elche, Alicante 03202, Spain. Email: [email protected]
Abstract Background and Aims: Proton pump inhibitors (PPIs) are commonly used antisecretory drugs and have been linked to an increased risk of bacterial infections in cirrhosis. We investigated whether the treatment with PPIs in cirrhosis affects the oxidative burst activity of granulocytes and monocytes and its possible interference with serum norfloxacin (Nflx) levels in these patients. Methods: 70 patients with cirrhosis and ascitic fluid and 24 healthy controls were included in the study and distributed into groups according to the regular use of PPIs and/or norfloxacin. The blood granulocyte and monocyte’s phagocytic activity and oxidative burst were evaluated by flow cytometry. Blood levels of norfloxacin were measured by HPLC and bacterial translocation was evaluated by detection of bacterial DNA in blood. Results: Use of PPIs was associated with a decreased granulocyte and monocyte oxidative burst, but not of phagocytic activity, as compared with patients not receiving PPIs. PPIs use did not affect serum norfloxacin levels in patients. A not significant trend to an increased bacterial DNA translocation was observed in patients receiving PPIs, including patients simultaneously receiving norfloxacin. Conclusions: PPIs significantly decrease cellular oxidative burst in cirrhosis. This fact may provide a pathogenic explanation to the reported high rates of bacterial infections in this setting, and strongly suggests that PPIs should only be used in patients with cirrhosis when clinically indicated.
Introduction Bacterial infections are common complications arising in patients with decompensated cirrhosis.1 The most relevant of them, such as bacteraemia or bacterial peritonitis, are called “spontaneous” since they have their origin in the intestinal bacterial content that can cross an excessively permeable intestinal wall accessing mesenteric lymph nodes and other extraintestinal sites in a process termed bacterial translocation (BT).2 The likelihood of developing bacterial infections is therefore related to the viability of the accessing bacteria and the capacity of the host immune system to mount an adequate bactericidal response.3 Several immune deficiencies have been described in advanced cirrhosis, either local or systemic, that favor the development of bacterial infections.1 In fact, cirrhosis may be considered as an acquired immunodeficient disease.4 Therefore, circumstances leading to an increased risk of infections in these patients may be taken into serious consideration, not only because bacterial infections compromise the survival of patients per se, but also because
the inflammatory response associated with these infections is disproportionate when compared to similar infections in noncirrhotic patients.5 We have previously shown that bacterial DNA (bactDNA) translocation induces an inflammatory response that is similar to that observed in fully developed infections6 and may impair prognosis.7 Proton pump inhibitors (PPIs), such as omeprazole, constitute a family of frequently used antisecretory drugs, either with or without a clear clinical indication, according to the general belief that are “protective” drugs. An adequate indication for omeprazole use exists only in half of cases prescribed,8,9 and this same proportion is thought to occur in cirrhotic patients as well, converting PPIs in one of the most frequently prescribed drugs in this population.10 The use of PPIs has been associated with an increased risk of developing nosocomial pneumonia in general population,11 or spontaneous bacterial peritonitis (SBP) among patients with cirrhosis and ascites.12,13 PPIs can favor intestinal bacterial overgrowth14 and affect the immune cellular activity in healthy donors,15 thereby increasing the risk of developing bacterial infections.
Journal of Gastroenterology and Hepatology 30 (2015) 147–154 © 2014 Journal of Gastroenterology and Hepatology Foundation and Wiley Publishing Asia Pty Ltd
147
Proton pump inhibitors and cirrhosis
I Garcia-Martinez et al.
Norfloxacin (Nflx) is a widely used medication in patients with decompensated cirrhosis either as primary or secondary prophylaxis of bacterial infections.16–18 We have previously reported that Nflx modulates neutrophil function and apoptosis,19 a mechanism probably beneficial in this setting. However, no information is available regarding the possible interactions between PPIs and Nflx and their effect on neutrophil function. The aims of this work have been (i) to assess the effect of PPIs treatment on both the phagocytosis and oxidative burst of innate immune cells from patients with decompensated cirrhosis, (ii) to investigate the relationship between PPI use and the presence of bactDNA, and (iii) to evaluate the possible interactions of PPIs with the concomitant use of Nflx.
Patients and methods Patients and study design. Seventy consecutively admitted patients with cirrhosis and ascitic fluid (AF) and 24 healthy donors from Centro de Transfusión de Alicante, San Juan de Alicante, Spain, were included in the study and distributed into groups according to the regular use of PPIs, starting at least 2 weeks before inclusion. Patients were further stratified according to the concomitant use of Nflx as primary prophylaxis or secondary prophylaxis of SBP. Cirrhosis was diagnosed by histological, clinical, laboratory, and/or ultrasonographic findings. Exclusion criteria were the presence of a culture-positive blood or AF, intake of antibiotics other than Nflx in the previous 2 weeks, SBP, upper gastrointestinal bleeding, hepatocellular carcinoma that exceeds Milan criteria,20 and/or portal thrombosis. The Ethics Committee of the General Hospital of Alicante approved the study protocol, and all patients provided their informed consent before being included in the study. Blood was obtained for routine hematological, biochemical, and coagulation studies. Simultaneously, a large-volume paracentesis was performed in all patients at admission in aseptic conditions following the usual procedures,21 and samples for routine biochemical study and Polymorphonuclear leucocytes (PMN) count were obtained. Total protein, albumin, leukocyte count, and PMN count were performed in all AF specimens. Both blood and AF were inoculated at bedside in aerobic and anaerobic blood culture bottles (10 mL each). Finally, separate blood and AF samples were inoculated, under aseptic conditions, in rubber-sealed sterile Vacutainer SST II tubes (BD Diagnostics, Belgium) that were never exposed to free air. Phagocytic assay. The phagocytic activity was evaluated using a Phagotest kit (Orpegen Pharma, Heidelberg, Germany) following the manufacturer’s protocol. Two cellular activities are measured in this assay; the in vitro percentage of granulocytes and monocytes that are able to ingest fluorescein-isothiocyanate (FITC)-labeled opsonized Escherichia coli bacteria, and their phagocytic capacity (total number of bacteria that is ingested per cell). In brief, 100 μL of heparinized whole blood were incubated with 20 μL of FITC-labeled opsonized E. coli or without any stimulus (negative control). The subsequent ingestion of the bacteria by phagocytes generates a green fluorescence signal that is quantified by flow cytometry. Phagocytosis was expressed as a percentage of activated phagocytes and as the mean fluorescence intensity (MFI). All samples were analyzed in triplicate. 148
Oxidative burst assay. The oxidative burst was measured with a Phagoburst kit (Orpegen Pharma, Heidelberg, Germany) according to the manufacturer’s protocol. Similar to the phagocytic assay, this kit measures both the percentage of granulocytes and monocytes that generate reactive oxygen species (ROS), and their oxidative burst activity (number of ROS per cell). In brief, 100 μL of heparinized whole blood was incubated with either 20 μL of unlabelled opsonized E. coli, 20 μL of N-formylmethionyl-leucyl-phenylalanine (fMLP), 20 μL of phorbol-12-myristate-13-acetate (PMA) or left unstimulated (negative control). The formation of ROS was monitored by the oxidation of dihydrorhodamine 123 to rhodamine, which produces a green fluorescence signal. Oxidative burst was expressed as a percentage of activated phagocytes and as the MFI. All samples were analyzed in triplicate. Flow cytometric analysis. Cells are analyzed by flow cytometry using the blue-green excitation light (488 nm) in a FacsCanto cytometer (Becton Dickinson, Heidelberg, Germany), and the data were processed using the FacsDiva 12.0 software. Neutrophils were identified and gated by the software program in the scatter diagram (Forward light scatter (FSC) vs Side light scatter (SSC)) and their green fluorescence histogram (FL1) was analyzed. We collected 10.000–15.000 leukocytes per sample. A control sample was used for setting an FL1 marker so that less than 1–3% of the events are positive. The parameters in the test sample can then be determined by counting the number of events above this marker position. Bacterial DNA isolation, amplification, and sequencing. Bacterial translocation was defined as the presence on bactDNA in blood. DNA extraction was performed by handling Qiagen Blood Minikit (Hilden, Germany) according to manufacturer’s directions. A broad-range polymerase chain reaction (PCR) and partial nucleotide sequencing analysis of prokaryote 16S rRNA were run in all serum and AF samples for amplification an identification bacterial DNA fragments as previously described.22 The primers used were: 5_ TTCCGGTTGATC CTGCCGGA 3_ as forward, and 5_ GGTTACCTTGTTACGA CTT 3_ as reverse.23 The same forward primer used for PCR amplification was used for sequencing. Sequences obtained were compared with 16S rRNA sequences available both in the Ribosomal Database Project24 and the GenBank and European Molecular Biology Laboratory obtained from the National Center for Biotechnology Information Database by the advanced Basic Local Alignment Search Tool (BLAST) search tool.25 Measurement of Norfloxacin levels. Simultaneously to the other clinical and analytical measurements described above, we analyzed the concentration of Nflx in plasma samples from patients taking Nflx using a HPLC method as previously described.19 Briefly, chromatographic separation was performed using a reverse-phase Eclipse XDB-C18 (Agilent, Santa Clara, CA, USA.) column (150x4.6 mm, Φ = 5μm). The mobile phase, Acetonitrile-tetrabutyl ammonium hydroxiphosphate buffer, was eluted at a flow rate of 1 mL/min, and effluent was monitored at
Journal of Gastroenterology and Hepatology 30 (2015) 147–154 © 2014 Journal of Gastroenterology and Hepatology Foundation and Wiley Publishing Asia Pty Ltd
I Garcia-Martinez et al.
Table 1
Proton pump inhibitors and cirrhosis
Clinical and analytical characteristics of controls and patients Control
Age (years) Gender (M/F) Etiology of cirrhosis, n Alcohol Hepatitis C Virus (HCV) Alcohol+HCV Other Drug history β-blockers agents Spironolactone Furosemide Lactulose Norfloxacin Refractory Ascites Previous hepatic encephalopathy Previous variceal bleeding T (°C) MAP (mm Hg) MELD score Child-pugh score Serum creatinine (mg/dL) Serum sodium (mEq/L) Serum albumin (mg/dL) Serum bilirubin (mg/dL) Leukocytes (× 103/μL) Neutrophils (× 103/μL) Quick (%) AF glucose (mg/dL) AF total protein (g/dL) AF RBCs (/μL) AF leukocytes (/μL)
Patient
PPI- (n = 15)
PPI+ (n = 9)
37 ± 10 5/10
50 ± 21 4/5
— — — —
— — — —
— — — — — — — — — — — — 0.8 ± 0.2 141 ± 2 — 0.6 ± 0.3 5.5 ± 1.6 3.2 ± 1.0 97 ± 4 — — — —
— — — — — — — — — — — — 0.8 ± 0.2 139 ± 2 — 0.4 ± 0.2 6.2 ± 1.8 3.8 ± 1.6 85 ± 10 — — — —
PPI- (n = 26) 61 ± 11 20/6 17 4 4 1 4 (15.4%) 11 (42.3%) 10 (38.5%) 2 (7.7%) 24 (92.3%) 8 (30.8%) 2 (7.7%) 20 (76.9%) 36.1 ± 0.4 104.9 ± 17.8 11 ± 4 9±1 0.9 ± 0.4 136 ± 5 2828 ± 665 2.3 ± 1.1 5.9 ± 3.3 3.9 ± 2.4 56 ± 14 137 ± 50 1.7 ± 0.9 1691 ± 3635 133 ± 109
PPI+ (n = 44) 62 ± 10 37/7 25 4 10 5 23 (52.3%)* 27 (61.4%) 35 (79.5%)* 11 (25%) 26 (59.1%)* 24 (54.5%) 12 (27.3%) 39 (88.6%) 36.4 ± 0.5 97.9 ± 15.1 13 ± 5 9±2 1.1 ± 0.5 133 ± 6 * 3029 ± 605 2.5 ± 1.7 5.1 ± 2.6 3.3 ± 1.8 59 ± 12 143 ± 39 1.5 ± 0.7 1951 ± 3337 182 ± 163
Values are expressed as mean ± standard deviation. *P < 0.05 compared with patients proton pump inhibitor (PPI) group. AF, ascitic fluid; MAP, mean arterial pressure; MELD, model for end-stage liver disease; RBC, red blood cells.
excitation and emission wavelengths of 278 and 456 nm, respectively. The limit of detection was 0.06 μg/mL. The calibration graph was found to be linear from 0.0625 to 4 μg/mL. Norfloxacin was obtained from Sigma Chemical Co (Madrid, Spain). Serum samples preparation for HPLC injection: A 100 μL plasma sample was diluted with 100 μL of acetonitrile, and the mixture shaken and centrifuged at 6000 rpm for 10 min. A 60 μL aliquot of supernatant was injected, and peak areas were recorded. Serum Nflx concentrations were expressed in micrograms/ milliliters. Measurement of TNF-α and nitric oxide (NO) products. Immunoenzymometric assay for quantitative measurement of human tumor necrosis factor-alpha (TNF)-α and nitric oxide (NO) levels in serum samples were performed using human TNF-α HS Quantikine and NOx ELISA kit, (R&D Systems) respectively, according to the manufacturer’s instructions. All samples were tested in duplicate and read at 450 nm (TNF-α) and 540 nm (NOx) in a ThermoMax microplate reader (Molecular
Devices, Sunnyvale, CA, USA). Standard curves were generated for each plate, and the average zero standard optical densities were subtracted from the remaining standards, controls, and samples to obtain a corrected concentration for TNF-a.
Statistical analysis. Continuous variables are reported as mean ± standard deviation and categorical variables as frequencies or percentages. Statistical differences between groups were analyzed using the Fisher test for categorical data and the Mann– Whitney U-test for quantitative data. Spearman’s linear correlation and Mann–Whitney U-test were initially performed to identify quantitative and qualitative variables that were significantly associated with phagocytosis and bust oxidative parameters. Analysis of covariance (ANCOVA) was used to further examine the association between variables selected by the previous analysis and the parameters of phagocytosis and oxidative burst. All reported P values are two-sided, and P values lower than 0.05 were considered to indicate significance. All calculations were performed using the SPSS 19.0 software. (SPSS, Inc, Chicago, IL, USA).
Journal of Gastroenterology and Hepatology 30 (2015) 147–154 © 2014 Journal of Gastroenterology and Hepatology Foundation and Wiley Publishing Asia Pty Ltd
149
Proton pump inhibitors and cirrhosis
I Garcia-Martinez et al.
Results
Table 2 Clinical and analytical characteristics of patients receiving PPIs, according to the use of Nflx
Characteristics of patients and laboratory data. The overall series of patients (n = 70) were divided into two groups according to the previous use of PPIs: PPI(+), n = 44, PPI(-), n = 26. Similarly, controls were divided into two groups according to the same criteria: control PPI(+), n = 9, control PPI(-), n = 15. Most patients undergoing PPI treatment were on omeprazole (91%), while esomeprazole and pantoprazole use accounted for 4.5% of cases each. Within the omeprazole group, most patients were receiving a dose of 20 mg QD (82.5%), and the remaining patients received 40 mg (10%) or 60 mg/d (7.5%). Only 22.7% of patients had been treated with PPIs during less than 1 year. Clinical and analytical characteristics of all groups of patients and controls according to the use of PPIs are detailed in Table 1. No statistically significant differences were observed between groups of controls. As expected, parameters reflecting liver function were different between controls and patients, irrespective of the use of PPIs. Statistically significant differences were observed between patients receiving or not PPIs in the number of coadministered medications and serum sodium levels (Table 1). The patients on PPIs were taking beta (β)-blockers, furosemide, and norfloxacin in a higher proportion than patients without PPIs. And, although the episodes of refractory ascites, hepatic encephalopathy, and variceal bleeds were higher in those patients (PPIs+), differences were not statistically significant. Patients receiving PPIs were further distributed according to the use of Nflx as secondary prophylaxis of SBP (PPI+Nflx group, n = 26 and PPI+Nflx+ group, n = 18). The clinical and analytical characteristics of both groups are detailed in Table 2. The proportion of patients with refractory ascites, previous episodes of hepatic encephalopathy, or previous episodes of variceal bleeding was significantly higher in patients receiving Nflx. Also, a higher proportion of PPIs+Nflx+ were receiving β-blocker treatment, and their mean arterial pressure was significantly lower compared with PPIs+Nflx- patients. BactDNA was present in 4 (15.4%) patients without PPIs (E. coli (n = 3) and K pneumoniae) and in 12 (27.3%) patients with PPIs, although values did not reach significance (P NS). Among patients receiving PPIs but not Nflx, bactDNA was detected in eight cases (30.7%), a percentage similar to previously described in patients with ascites (E. coli [n = 6], S aureus and C. Freundii). Of interest, four out of the 18 patients receiving Nflx, and PPIs (22.2%) showed the presence of bactDNA in blood, a rate higher than expected in patients receiving Nflx. Species identified included Shigella, E. coli (n = 2), and Klebsiella pneumoniae. No relationship was found among the type or dosage of PPI, the time of its use and the presence of bactDNA. Similarly, neither the length of therapy with Nflx nor the serum Nflx levels in blood were related with the presence of bactDNA (data not shown).
The use of PPIs does not affect cellular phagocytic capacity in patients. The percentage of granulocytes and monocytes showing phagocytosis at baseline, and their intrinsic phagocytic capacity (number of bacteria ingested per cell) were considered as two separated parameters. 150
Age (years) Gender (M/F) Etiology of cirrhosis, n Alcohol HCV Alcohol+HCV Other Drug history β-blockers agents Spironolactone Furosemide Lactulose Refractory ascites Previous hepatic encephalopathy Previous variceal bleeding Presence bacterial DNA T (°C) MAP (mm Hg) MELD Score Child-pugh score Serum creatinine (mg/dL) Serum sodium (mEq/L) Serum albumin (mg/dL) Serum bilirubin (mg/dL) Leukocytes (× 103/μL) Neutrophils (× 103/μL) Quick (%) AF glucose (mg/dL) AF total protein (g/dL) AF RBCs (/μL) AF leukocytes (/μL)
Nflx- (n = 26)
Nflx+ (n = 18)
63 ± 10 21/5
60 ± 9 16/2
14 2 6 4
11 2 4 1
9 (34.6%) 14 (53.8%) 19 (73.1%) 6 (23.1%) 11 (42.3%) 4 (15.4%) 21 (80.8%) 8 (30.8%) 36.5 ± 0.3 107 ± 9 13 ± 5 8.5 ± 1.6 1.1 ± 0.5 134 ± 7 2971 ± 673 2.5 ± 1.9 5.8 ± 2.8 3.8 ± 2.1 61 ± 13 132 ± 36 1.6 ± 0.7 1307 ± 3059 170 ± 149
14 (77.8%)* 13 (72.2%) 16 (88.9%) 5 (27.8%) 13 (72.2%)* 8 (44%)* 18 (100%)* 4 (22.2%) 36.2 ± 0.5 87 ± 14* 15 ± 3 9.5 ± 1.6 1.1 ± 0.5 131 ± 5 3175 ± 402 2.5 ± 1.3 4.1 ± 1.8 2.6 ± 0.9 54 ± 9 159 ± 39* 1.5 ± 0.8 2838 ± 3594 199 ± 184
Values are expressed as mean ± standard deviation. *P < 0.05 compared with patients Norfloxacin group. AF, ascitic fluid; MAP, mean arterial pressure; MELD, model for endstage liver disease; RBC, red blood cells.
Previous use of PPIs was not associated with differences in the baseline percentage of granulocytes and monocytes showing phagocytosis neither in controls nor in patients (Supplementary Fig. S1). Baseline phagocytic capacity in the absence of PPIs was similar between patients and controls. However, the use of PPIs was associated with a significantly higher phagocytic capacity in patients than in controls (Fig. 1a). The concomitant use of Nflx did not affect the phagocytic capacity observed among patients receiving PPIs and no correlation was found between serum Nflx levels and the total amount of ingested bacteria (Fig. 1b). The presence of bacterial DNA did not affect the phagocytic capacity in our series of PPI+ versus PPI- patients (Fig. 1c). The use of PPIs decreases the cellular oxidative burst in patients with decompensated cirrhosis. The percentage of granulocytes and monocytes showing oxidative burst at baseline and their intrinsic oxidative burst capacity (number of ROS released per cell) were considered as two
Journal of Gastroenterology and Hepatology 30 (2015) 147–154 © 2014 Journal of Gastroenterology and Hepatology Foundation and Wiley Publishing Asia Pty Ltd
I Garcia-Martinez et al.
Proton pump inhibitors and cirrhosis
(a)
(b)
(c)
Figure 1 Phagocytic capacity of granulocytes represented by the mean fluorescence intensity (MFI) measured by flow cytometry. (a) Phagocytic capacity of granulocytes from controls and patients, according to the use of proton pump inhibitors (PPIs). (b) Phagocytic capacity of granulocytes from patients distributed by norfloxacin, according to the use of PPIs and correlation between serum levels of norfloxacin and phagocytic capacity. (c) Phagocytic capacity of granulocytes from patients distributed by the presence of blood bactDNA according to the use of PPIs. Nflx: norfloxacin; bactDNA: bacterial DNA; *P < 0.05. □, PPI−; ■, PPI+.
separated parameters and was evaluated at baseline and after in vitro stimulation with fMLP (weak stimulus), PMA (strong stimulus), and E. coli. Previous use of PPIs did not induce changes in the percentage of granulocytes and monocytes showing oxidative burst either in controls or in patients (Supplementary Fig. S2). However, the intake of PPIs was associated with a significant decrease in the total amount of ROS released by monocytes and granulocytes, both in controls and patients at baseline and all stimulated conditions (Fig. 2). No significant differences were observed in the reduction of burst activity between controls and patients (data not shown). Univariate analysis showed a significant association between oxidative burst activity and previous use of PPIs (P = 0.001), age (categorized as < 60 years and > 60 years; P = 0.006), leukocytes (P = 0.03), neutrophils (P = 0.013), and potassium (P = 0.005). The use of PPIs (P = 0.001), as well as age (P = 0.004), remained
as significant independent variables associated with burst activity in the ANCOVA multivariate analysis including all previous variables. The use of Nflx did not affect the oxidative burst activity observed among cirrhotic patients receiving PPIs, and no correlation was found between serum Nflx levels and the total amount of ROS released by either monocytes or granulocytes. The presence of bacterial DNA did not affect the oxidative burst activity in our series of PPI+ versus PPI- patients. *The results on phagocytosis and oxidative burst shown in the present study were similar for granulocyte and monocyte populations in all cases.
Serum soluble inflammatory response not affected by PPIs in patients. TNF-alpha and NOx levels in the serum of patients distributed according to the use of PPIs revealed no
Journal of Gastroenterology and Hepatology 30 (2015) 147–154 © 2014 Journal of Gastroenterology and Hepatology Foundation and Wiley Publishing Asia Pty Ltd
151
Proton pump inhibitors and cirrhosis
I Garcia-Martinez et al.
Figure 2 Oxidative burst activity (amount of reactive oxygen species, ROS, released) of granulocytes from controls and patients. Oxidative burst activity of granulocytes from controls and patients, according to the use of proton pump inhibitors (PPIs) at baseline and in response to several stimuli. fMLP: N- formylmethionyl-leucyl-phenylalanine; PMA: phorbol-12-myristate-13-acetate. *P < 0.05 compared with the PPI- group. □, PPI−; ■, PPI+.
statistically significant differences at baseline. Distribution of patients by intake of Nflx or the presence of bacterial DNA showed no differences on these inflammatory mediators (data not shown).
Discussion In this investigation, we show evidences that omeprazole, as representative drug of the family of PPIs, significantly reduces the oxidative burst activity of granulocytes and monocytes in patients with decompensated cirrhosis, which is a key mechanism of the immune system bactericidal activity. Notably, these effects are not modified by the simultaneous administration of Nflx. PPIs constitute the third most prescribed medications in the United States,26 and its use has been recently associated with an increased risk of developing pulmonary infections in admitted patients11,27 and SBP in patients with decompensated cirrhosis.12,13 Although the reasons behind this fact may be several, the main deleterious effect associated with their use was initially attributed to the relationship between hypochlorhydria and bacterial survival in the stomach.28 Additional investigations in animal models showed that inhibition of gastric acid secretion may be related to increased intestinal paracellular permeability,29 overgrowth of small bowel microflora,30 and an increased rate of BT.31 In addition, Omeprazole has been shown to inhibit both the neutrophil function in vitro32 and the bactericidal activity in general population.15 However, no data are available on the effect of PPIs on monocyte and neutrophil function in patients with cirrhosis. This is relevant, since omeprazole is largely used in these patients, who already have a compromised neutrophil function at baseline.33 In the present study, we have demonstrated that PPIs indeed decrease both granulocyte and monocyte’s oxidative burst activity in patients with decompensated cirrhosis, a surrogate marker of an impaired bactericidal activity. This fact that may provide a mechanistic explanation to the reported increased incidence of bacterial infections observed in other studies, both in controls and in patients with cirrhosis.11,13 Although this effect was not associated 152
with a significantly increased rate of bacterial DNA translocation in this investigation, an interesting finding is the fact that patients receiving PPIs plus Nflx showed a rate of bacterial DNA translocation higher than usual (Table 2). This suggests that PPIs use may decrease the Nflx efficacy to prevent bacterial translocation in cirrhosis. This hypothesis though should be confirmed in a specifically designed study. Despite not observing differences in bactDNA translocation rates, the oxidative burst activity downregulation associated with the intake of PPIs in patients with decompensated cirrhosis remains relevant from an immunological point of view, since it may point towards a compromised situation in an eventual bacterial infection challenge. An impaired neutrophil function, as observed in cirrhotic patients in this study, compromises a key part of the innate immune system, the first line of host resistance against bacterial infection. Adding this fact to the already deranged cellular situation of these patients provides a concern for a more careful decision on whether start PPIs in this setting. Patients on Nflx did not show a different capacity of oxidative burst irrespective of the concomitant use of PPIs. The same lack of association was observed between the presence of bacterial DNA in blood, serum inflammatory markers, and oxidative burst activity in our series. On the contrary, the ANCOVA analysis determined that age, and the use of PPIs were the only variables independently associated with a decreased oxidative burst activity. Elderly patients (> 60years group) shown lower levels of ROS released than younger patients. In fact, a decrease in phagocytosis of opsonized bacteria by neutrophils34 and in the production of superoxide by PMNs or antioxidants such as glutathione in the have been shown to occur in the elderly.35,36 The fact that PPIs are not associated with a lower phagocytic capacity in patients with cirrhosis compared with controls may be a consequence of the general immunological derangement associated with the course of disease. This pre-activated state may account for preventing a downregulation of the phagocytic activity in patients. Although the effect of PPIs on oxidative burst activity had been previously evaluated in healthy donors and in in vitro assays, we
Journal of Gastroenterology and Hepatology 30 (2015) 147–154 © 2014 Journal of Gastroenterology and Hepatology Foundation and Wiley Publishing Asia Pty Ltd
I Garcia-Martinez et al.
report first-time evidence of their deleterious effect on the oxidative burst activity of granulocytes and monocytes from patients with decompensated cirrhosis and provide rationale for limiting PPIs use in cirrhosis when clinically indicated, far from the supposed role of “protective” drug. In summary, PPIs use decrease the amount of ROS released during the oxidative burst by granulocytes and monocytes in patients with decompensated cirrhosis, thereby decreasing bactericidal activity and giving a pathogenic explanation to the reported increased rate of bacterial infections in patients with cirrhosis receiving PPIs.
Acknowledgments This work has been supported in part with grants from Instituto de Salud Carlos III (PI08/1075), Excma. Diputación Provincial de Alicante, Spain and Fundación FCVI-HGUA, Alicante, Spain.
References 1 Such J, Runyon BA. Spontaneous bacterial peritonitis. Clin. Infect. Dis. 1998; 27: 669–74; quiz 75–6. 2 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: 403–11. 3 Rimola A, Soto R, Bory F, Arroyo V, Piera C, Rodes J. Reticuloendothelial system phagocytic activity in cirrhosis and its relation to bacterial infections and prognosis. Hepatology 1984; 4: 53–8. 4 Guarner C, Runyon BA. Spontaneous bacterial peritonitis: pathogenesis, diagnosis, and management. Gastroenterologist 1995; 3: 311–28. 5 Byl B, Roucloux I, Crusiaux A, Dupont E, Deviere J. Tumor necrosis factor alpha and interleukin 6 plasma levels in infected cirrhotic patients. Gastroenterology 1993; 104: 1492–7. 6 Frances R, Zapater P, Gonzalez-Navajas JM et al. Bacterial DNA in patients with cirrhosis and noninfected ascites mimics the soluble immune response established in patients with spontaneous bacterial peritonitis. Hepatology 2008; 47: 978–85. 7 Zapater P, Frances R, Gonzalez-Navajas JM et al. Serum and ascitic fluid bacterial DNA: a new independent prognostic factor in noninfected patients with cirrhosis. Hepatology 2008; 48: 1924–31. 8 Naunton M, Peterson GM, Bleasel MD. Overuse of proton pump inhibitors. J. Clin. Pharm. Ther. 2000; 25: 333–40. 9 Nardino RJ, Vender RJ, Herbert PN. Overuse of acid-suppressive therapy in hospitalized patients. Am. J. Gastroenterol. 2000; 95: 3118–22. 10 Chavez-Tapia NC, Tellez-Avila FI, Garcia-Leiva J, Valdovinos MA. Use and overuse of proton pump inhibitors in cirrhotic patients. Med. Sci. Monit. 2008; 14: CR468–72. 11 Herzig SJ, Howell MD, Ngo LH, Marcantonio ER. Acid-suppressive medication use and the risk for hospital-acquired pneumonia. JAMA 2009; 301: 2120–8. 12 Goel GA, Deshpande A, Lopez R, Hall GS, van Duin D, Carey WD. Increased rate of spontaneous bacterial peritonitis among cirrhotic patients receiving pharmacologic Acid suppression. Clin. Gastroenterol. Hepatol. 2012; 10: 422–7. 13 Bajaj JS, Zadvornova Y, Heuman DM et al. Association of proton pump inhibitor therapy with spontaneous bacterial peritonitis in cirrhotic patients with ascites. Am. J. Gastroenterol. 2009; 104: 1130–4.
Proton pump inhibitors and cirrhosis
14 Fried M, Siegrist H, Frei R et al. Duodenal bacterial overgrowth during treatment in outpatients with omeprazole. Gut [Clinical Trial Controlled Clinical Trial Research Support, Non-U.S. Gov’t]. 1994;35: 23–6. 15 Zedtwitz-Liebenstein K, Wenisch C, Patruta S, Parschalk B, Daxbock F, Graninger W. Omeprazole treatment diminishes intraand extracellular neutrophil reactive oxygen production and bactericidal activity. Crit. Care Med. 2002; 30: 1118–22. 16 Soriano G, Guarner C, Tomas A et al. Norfloxacin prevents bacterial infection in cirrhotics with gastrointestinal hemorrhage. Gastroenterology 1992; 103: 1267–72. 17 Fernandez J, Ruiz del Arbol L, Gomez C et al. Norfloxacin vs ceftriaxone in the prophylaxis of infections in patients with advanced cirrhosis and hemorrhage. Gastroenterology 2006; 131: 1049–56; quiz 285. 18 Fernandez J, Navasa M, Planas R et al. Primary prophylaxis of spontaneous bacterial peritonitis delays hepatorenal syndrome and improves survival in cirrhosis. Gastroenterology 2007; 133: 818–24. 19 Zapater P, Cano R, Llanos L et al. Norfloxacin modulates the inflammatory response and directly affects neutrophils in patients with decompensated cirrhosis. Gastroenterology 2009; 137: 1669–79. 20 Mazzaferro V, Regalia E, Doci R et al. Liver transplantation for the treatment of small hepatocellular carcinomas in patients with cirrhosis. N. Engl. J. Med. 1996; 334: 693–9. 21 Runyon BA. Paracentesis of ascitic fluid. A safe procedure. Arch. Intern. Med. 1986; 146: 2259–61. 22 Such J, Frances R, Munoz C et al. Detection and identification of bacterial DNA in patients with cirrhosis and culture-negative, nonneutrocytic ascites. Hepatology 2002; 36: 135–41. 23 Lane DJ. 16S/23S rRNA sequencing. In: Stackebrandt E, Goodfellow M, eds. Nucleic Acid Techniques in Bacterial Systematics. New York: John Wiley and Sons, 1991; 115–75. 24 Maidak BL, Cole JR, Lilburn TG et al. The RDP-II (Ribosomal Database Project). Nucleic Acids Res. 2001; 29: 173–4. 25 Altschul SF, Madden TL, Schaffer AA et al. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 1997; 25: 3389–402. 26 Bavishi C, Dupont HL. Systematic review: the use of proton pump inhibitors and increased susceptibility to enteric infection. Aliment. Pharmacol. Ther. 2011; 34: 1269–81. 27 Laheij RJ, Sturkenboom MC, Hassing RJ, Dieleman J, Stricker BH, Jansen JB. Risk of community-acquired pneumonia and use of gastric acid-suppressive drugs. JAMA 2004; 292: 1955–60. 28 Tennant SM, Hartland EL, Phumoonna T et al. Influence of gastric acid on susceptibility to infection with ingested bacterial pathogens. Infect. Immun. 2008; 76: 639–45. 29 Hopkins AM, McDonnell C, Breslin NP, O’Morain CA, Baird AW. Omeprazole increases permeability across isolated rat gastric mucosa pre-treated with an acid secretagogue. J. Pharm. Pharmacol. 2002; 54: 341–7. 30 Lichtman SM. Bacterial [correction of bacterial] translocation in humans. J. Pediatr. Gastroenterol. Nutr. 2001; 33: 1–10. 31 Basaran UN, Celayir S, Eray N, Ozturk R, Senyuz OF. The effect of an H2-receptor antagonist on small-bowel colonization and bacterial translocation in newborn rats. Pediatr. Surg. Int. 1998; 13: 118–20. 32 Wandall JH. Effects of omeprazole on neutrophil chemotaxis, super oxide production, degranulation, and translocation of cytochrome b-245. Gut 1992; 33: 617–21. 33 Fiuza C, Salcedo M, Clemente G, Tellado JM. In vivo neutrophil dysfunction in cirrhotic patients with advanced liver disease. J. Infect. Dis. 2000; 182: 526–33. 34 Wenisch C, Patruta S, Daxbock F, Krause R, Horl W. Effect of age on human neutrophil function. J. Leukoc. Biol. 2000; 67: 40–5.
Journal of Gastroenterology and Hepatology 30 (2015) 147–154 © 2014 Journal of Gastroenterology and Hepatology Foundation and Wiley Publishing Asia Pty Ltd
153
Proton pump inhibitors and cirrhosis
I Garcia-Martinez et al.
35 Braga PC, Sala MT, Dal Sasso M, Mancini L, Sandrini MC, Annoni G. Influence of age on oxidative bursts (chemiluminescence) of polymorphonuclear neutrophil leukocytes. Gerontology 1998; 44: 192–7. 36 Corberand J, Ngyen F, Laharrague P et al. Polymorphonuclear functions and aging in humans. J. Am. Geriatr. Soc. 1981; 29: 391–7.
of granulocytes from controls and patients, according to the use of PPIs. (b) Percentage of granulocytes from patients distributed by norfloxacin, according to the use of PPIs and correlation between serum levels of norfloxacin and percentage of phagocytosis. (c) Percentage of granulocytes from patients distributed by the presence of blood bactDNA according to the use of PPIs. Nflx, norfloxacin; bactDNA, bacterial DNA; *P < 0.05.
Supporting information
Figure S2 Percentage of granulocytes from controls and patients that generate reactive oxygen species (ROS). Percentage of granulocytes generating ROS from controls and patients, according to the use of PPIs at baseline and in response to several stimuli. fMLP, N-formylmethionyl-leucyl-phenylalanine; PMA, phorbol-12-myristate-13-acetate.
Additional Supporting Information may be found in the online version of this article at the publisher’s web-site: Figure S1 Percentage of granulocytes from controls and patients that ingest bacteria, measured by flow cytometry. (a). Percentage
154
Journal of Gastroenterology and Hepatology 30 (2015) 147–154 © 2014 Journal of Gastroenterology and Hepatology Foundation and Wiley Publishing Asia Pty Ltd
