© 2015 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd Transplant Infectious Disease, ISSN 1398-2273
The relevance of intestinal dysbiosis in liver transplant candidates wko, K.M. Wronka, K. Graz t, Z. Lewandowski, M. Graz t, W. Hoło ska, M. Krasnodez bski, M. Wasilewicz, M. Gałez cka, I. Kosin P. Szachta, H. Zborowska, W. Patkowski, M. Krawczyk. The relevance of intestinal dysbiosis in liver transplant candidates. Transpl Infect Dis 2015: 17: 174–184. All rights reserved Abstract: Background. The gut microbial ecosystem plays an important role in the pathogenesis of liver diseases. However, the association of microbial community structure with the severity of liver dysfunction is not completely understood. Methods. Fecal microflora was assessed in 40 patients with liver cirrhosis listed for primary liver transplantation (LT). Independent associations between fecal microbial counts and serum bilirubin, serum creatinine, international normalized ratio (INR), and the Model for End-stage Liver Disease (MELD) score were established in multiple linear regression models. Results. Bifidobacterium (standardized regression coefficient [sb] = 0.549; P < 0.001), Enterococcus (sb = 0.369; P = 0.004), and yeast (sb = 0.315; P = 0.018) numbers were independently associated with serum bilirubin, while Escherichia coli counts (sb = 0.318; P = 0.046) correlated with INR, and Bifidobacterium counts (sb = 0.410; P = 0.009) with serum creatinine. Only Bifidobacterium (sb = 0.468; P = 0.003) and Enterococcus (sb = 0.331; P = 0.029) counts were independent predictors of the MELD score. Bifidobacterium/Enterococcus ratio, proposed as a measure of pre-LT gut dysbiosis, was significantly related to the MELD score following the adjustment for the absolute Bifidobacterium (sb = 0.333; P = 0.029) and Enterococcus (sb = 0.966; P = 0.003) numbers. This pre-transplant dysbiosis ratio (PTDR) was significantly correlated with Enterococcus (R = 0.897; P < 0.001) but not with Bifidobacterium (R = 0.098; P = 0.546) counts. Among the other components of gut microflora, only hydrogen peroxide (H2O2)-producing Lactobacillus strains significantly influenced Enterococcus counts (sb = 0.349; P = 0.028) and PTDR (sb = 0.318; P = 0.046). Conclusion. While the abundance of both Bifidobacterium and Enterococcus is related to liver dysfunction, the size of the Enterococcus population seems to be the most important determinant of pre-LT gut dysbiosis in cirrhotic patients. The H2O2-producing Lactobacillus strains potentially ameliorate this dysbiotic state.
Approximately 100 trillion microorganisms form a unique microbial ecosystem inside the human gastrointestinal tract (1). Because of the wide magnitude and complexity of microbial interactions with the host, gut microecology plays an important role in health and
174
M. Graz t1, W. Hołowko1, K.M. Wronka1, K. Graz t1, Z. Lewandowski2, I. Kosinska3, M. Krasnodez bski1, M. Wasilewicz1, M. Gałez cka4, P. Szachta4, H. Zborowska5, W. Patkowski1, M. Krawczyk1 1
Department of General, Transplant, and Liver Surgery, Medical University of Warsaw, Warsaw, Poland, 2 Department of Epidemiology, Medical University of Warsaw, Warsaw, Poland, 3Department of Preventive Medicine and Hygiene, Institute of Social Medicine, Medical University of Warsaw, Warsaw, Poland, 4Institute of Microecology, Poznan, Poland, 5Department of Laboratory Diagnostics, Medical University of Warsaw, Warsaw, Poland
Key words: liver transplantation; liver cirrhosis; gut microbiota; Model for End-stage Liver Disease; Bifidobacterium; Enterococcus Correspondence to: Michał Graz t, Department of General, Transplant, and Liver Surgery, Medical University of Warsaw, 1A Banacha Street, 02-097 Warsaw, Poland Tel: +48 22 599 25 45 Fax: +48 22 599 15 45 E-mail: [email protected]
Received 4 March 2014, revised 18 June 2014, accepted for publication 11 December 2014 DOI: 10.1111/tid.12352 Transpl Infect Dis 2015: 17: 174–184
disease (2, 3). Alterations in intestinal microflora have been reported in patients with various pathologic conditions, such as inflammatory bowel diseases, obesity, diabetes, and colorectal cancer (4, 5). Most importantly, gut microflora has both direct and indirect
Graz t et al: Intestinal dysbiosis in LT candidates
effects on liver function and physiology, and is involved in the pathogenesis and progression of liver diseases. Potentially pathogenic microbial species and their products may infiltrate into the portal circulation and induce inflammatory responses and, consequently, hepatic injury. On the other hand, pathologic conditions associated with liver cirrhosis, such as decreased secretion of bile acids, increased gut permeability, lower motility, and immune dysfunction, have a direct impact on intestinal microflora and contribute to dysbiosis in the digestive tract (6, 7). Liver transplantation (LT) continues to be the only life-saving procedure for patients with end-stage liver disease. Considering a growing disparity between the number of patients requiring LT and that of available donors, patients listed for this procedure currently face a waitlist mortality rate of approximately 8–25% (8–10). The risk of death in patients awaiting LT is closely related to the severity of liver disease ranked by the Model for End-stage Liver Disease (MELD) score and its dynamic changes over time (10); both factors are also associated with post-transplantation survival (10, 11). Given that the administration of probiotic agents has been reported to improve liver function in the course of cirrhosis (12), the modulation of gut microflora may be a potential approach to slow down disease progression and decrease pre-transplant mortality rates and, consequently, to improve the general results of LT programs. However, the association between the composition of gut microflora and the MELD score widely used for the prioritization of patients on liver transplant waiting lists has not been sufficiently investigated. Besides alterations in gut microecology, patients with liver cirrhosis often experience the migration of gut residential microorganisms through the intestinal barrier, a phenomenon called bacterial translocation, which elevates the risk of related infections (13). Overall, infection-related complications occur in approximately 55–77% of these patients, accounting for a major proportion of morbidity and mortality rates (14, 15). Although no studies have been reported on the effects of long-term pre-transplant modulation of gut microflora using probiotics, perioperative administration of synbiotics has been previously reported to markedly reduce infection rates (16). Accordingly, the modulation of gut microflora in liver transplant recipients may not only slow down the progression of cirrhosis, but also improve postoperative outcomes. Nevertheless, no consensus exists on a particular type of probiotic supplement that may bring the most benefits to patients listed for liver transplant procedures.
Previous studies have revealed that cirrhosis is associated with a potentially deleterious shift in the structure of intestinal microbial community: a decrease in the abundance of multiple beneficial bacterial groups, namely members of the autochthonous families (Lachnospiraceae, Ruminococcaceae, Blautia), lactic acid bacteria, and Bifidobacterium species, and an increase in that of potentially pathogenic organisms, particularly Enterobacteriaceae and Enterococcus species (17–19). The purpose of the present study was to determine the components of gut microflora independently associated with the MELD score in cirrhotic patients listed for LT, with special focus on potentially beneficial bacterial species.
Patients and methods This prospective study was performed in the Department of General, Transplant, and Liver Surgery at the Medical University of Warsaw, Poland, in the period between November 2012 and September 2013. The study group comprised 40 patients with liver cirrhosis listed for first LT. None of the patients received probiotics within 4 weeks prior to inclusion in the study. The study protocol has been approved by the local ethics committee at the Medical University of Warsaw. Informed consent was obtained from each of the study participants prior to enrollment. All patients provided stool samples of approximately 100 mL, which were sent to the laboratory within 24 h after collection. Samples were serially diluted in phosphate-buffered saline (pH 7.2) and seeded on enriching and selective agar media to determine the number of colony-forming units per gram (cfu/g) of wet feces for the following microbial groups: Enterococcus species, Escherichia coli, Proteus species, other members of Enterobacteriaceae taxon (Klebsiella, Enterobacter, Citrobacter, and Serratia species), Pseudomonas, Bifidobacterium, Bacteroides, Lactobacillus, and Clostridium species, and yeasts. Microbial counts were assessed following 48-h incubation of agar plates at 37°C in anaerobic conditions for Bacteroides (Schaedler agar), Bifidobacterium (DIC agar), and Clostridium species (SPM agar), and in increased carbon dioxide concentration for Lactobacillus species (Rogosa agar); 24-h incubation in aerobic conditions at 37°C was performed to detect the growth of Enterobacteriaceae and Enterococcus species on U3G agar. Hydrogen peroxide (H2O2)-producing and nonproducing Lactobacillus strains were distinguished based on the reaction with 3,30 ,5,50 -tetramethylbenzidine and peroxidase.
Transplant Infectious Disease 2015: 17: 174–184
175
Graz t et al: Intestinal dysbiosis in LT candidates
Data on patients’ age and gender, etiology of liver disease, serum bilirubin concentration, international normalized ratio (INR), serum creatinine concentration, and the MELD score were also collected. Laboratory measurements used to calculate MELD score were performed in a week preceding collection of stool sample. Microbial counts were presented as natural logarithms of cfu/g. Quantitative variables were compared using the Mann–Whitney U-test. Spearman’s correlation coefficient was used for unadjusted analyses of the associations between microbial counts and the MELD components (serum bilirubin concentration, INR, and serum creatinine concentration). Linear regression models were used to assess microbial counts independently associated with the 3 MELD components. Following the results of these analyses, microbial counts independently associated with serum bilirubin concentration, serum creatinine concentration, or INR were tested for an independent impact on the MELD score in multivariate analysis. Finally, the remaining microbiota was tested for the impact on the abundance of bacterial groups independently associated with the MELD score. Multivariate models were created based on forward stepwise method, using a P-value of 0.15 for both inclusion and removal of variables from the model, as previously recommended (20). Crude (cb) and standardized (sb) regression coefficients were presented with standard errors (SE); regression plots were presented with 95% confidence intervals (95% CI). For the plotting of regression lines in bivariate analyses, a median of the second variable was included in the equation. P-values
The relevance of intestinal dysbiosis in liver transplant candidates wko, K.M. Wronka, K. Graz t, Z. Lewandowski, M. Graz t, W. Hoło ska, M. Krasnodez bski, M. Wasilewicz, M. Gałez cka, I. Kosin P. Szachta, H. Zborowska, W. Patkowski, M. Krawczyk. The relevance of intestinal dysbiosis in liver transplant candidates. Transpl Infect Dis 2015: 17: 174–184. All rights reserved Abstract: Background. The gut microbial ecosystem plays an important role in the pathogenesis of liver diseases. However, the association of microbial community structure with the severity of liver dysfunction is not completely understood. Methods. Fecal microflora was assessed in 40 patients with liver cirrhosis listed for primary liver transplantation (LT). Independent associations between fecal microbial counts and serum bilirubin, serum creatinine, international normalized ratio (INR), and the Model for End-stage Liver Disease (MELD) score were established in multiple linear regression models. Results. Bifidobacterium (standardized regression coefficient [sb] = 0.549; P < 0.001), Enterococcus (sb = 0.369; P = 0.004), and yeast (sb = 0.315; P = 0.018) numbers were independently associated with serum bilirubin, while Escherichia coli counts (sb = 0.318; P = 0.046) correlated with INR, and Bifidobacterium counts (sb = 0.410; P = 0.009) with serum creatinine. Only Bifidobacterium (sb = 0.468; P = 0.003) and Enterococcus (sb = 0.331; P = 0.029) counts were independent predictors of the MELD score. Bifidobacterium/Enterococcus ratio, proposed as a measure of pre-LT gut dysbiosis, was significantly related to the MELD score following the adjustment for the absolute Bifidobacterium (sb = 0.333; P = 0.029) and Enterococcus (sb = 0.966; P = 0.003) numbers. This pre-transplant dysbiosis ratio (PTDR) was significantly correlated with Enterococcus (R = 0.897; P < 0.001) but not with Bifidobacterium (R = 0.098; P = 0.546) counts. Among the other components of gut microflora, only hydrogen peroxide (H2O2)-producing Lactobacillus strains significantly influenced Enterococcus counts (sb = 0.349; P = 0.028) and PTDR (sb = 0.318; P = 0.046). Conclusion. While the abundance of both Bifidobacterium and Enterococcus is related to liver dysfunction, the size of the Enterococcus population seems to be the most important determinant of pre-LT gut dysbiosis in cirrhotic patients. The H2O2-producing Lactobacillus strains potentially ameliorate this dysbiotic state.
Approximately 100 trillion microorganisms form a unique microbial ecosystem inside the human gastrointestinal tract (1). Because of the wide magnitude and complexity of microbial interactions with the host, gut microecology plays an important role in health and
174
M. Graz t1, W. Hołowko1, K.M. Wronka1, K. Graz t1, Z. Lewandowski2, I. Kosinska3, M. Krasnodez bski1, M. Wasilewicz1, M. Gałez cka4, P. Szachta4, H. Zborowska5, W. Patkowski1, M. Krawczyk1 1
Department of General, Transplant, and Liver Surgery, Medical University of Warsaw, Warsaw, Poland, 2 Department of Epidemiology, Medical University of Warsaw, Warsaw, Poland, 3Department of Preventive Medicine and Hygiene, Institute of Social Medicine, Medical University of Warsaw, Warsaw, Poland, 4Institute of Microecology, Poznan, Poland, 5Department of Laboratory Diagnostics, Medical University of Warsaw, Warsaw, Poland
Key words: liver transplantation; liver cirrhosis; gut microbiota; Model for End-stage Liver Disease; Bifidobacterium; Enterococcus Correspondence to: Michał Graz t, Department of General, Transplant, and Liver Surgery, Medical University of Warsaw, 1A Banacha Street, 02-097 Warsaw, Poland Tel: +48 22 599 25 45 Fax: +48 22 599 15 45 E-mail: [email protected]
Received 4 March 2014, revised 18 June 2014, accepted for publication 11 December 2014 DOI: 10.1111/tid.12352 Transpl Infect Dis 2015: 17: 174–184
disease (2, 3). Alterations in intestinal microflora have been reported in patients with various pathologic conditions, such as inflammatory bowel diseases, obesity, diabetes, and colorectal cancer (4, 5). Most importantly, gut microflora has both direct and indirect
Graz t et al: Intestinal dysbiosis in LT candidates
effects on liver function and physiology, and is involved in the pathogenesis and progression of liver diseases. Potentially pathogenic microbial species and their products may infiltrate into the portal circulation and induce inflammatory responses and, consequently, hepatic injury. On the other hand, pathologic conditions associated with liver cirrhosis, such as decreased secretion of bile acids, increased gut permeability, lower motility, and immune dysfunction, have a direct impact on intestinal microflora and contribute to dysbiosis in the digestive tract (6, 7). Liver transplantation (LT) continues to be the only life-saving procedure for patients with end-stage liver disease. Considering a growing disparity between the number of patients requiring LT and that of available donors, patients listed for this procedure currently face a waitlist mortality rate of approximately 8–25% (8–10). The risk of death in patients awaiting LT is closely related to the severity of liver disease ranked by the Model for End-stage Liver Disease (MELD) score and its dynamic changes over time (10); both factors are also associated with post-transplantation survival (10, 11). Given that the administration of probiotic agents has been reported to improve liver function in the course of cirrhosis (12), the modulation of gut microflora may be a potential approach to slow down disease progression and decrease pre-transplant mortality rates and, consequently, to improve the general results of LT programs. However, the association between the composition of gut microflora and the MELD score widely used for the prioritization of patients on liver transplant waiting lists has not been sufficiently investigated. Besides alterations in gut microecology, patients with liver cirrhosis often experience the migration of gut residential microorganisms through the intestinal barrier, a phenomenon called bacterial translocation, which elevates the risk of related infections (13). Overall, infection-related complications occur in approximately 55–77% of these patients, accounting for a major proportion of morbidity and mortality rates (14, 15). Although no studies have been reported on the effects of long-term pre-transplant modulation of gut microflora using probiotics, perioperative administration of synbiotics has been previously reported to markedly reduce infection rates (16). Accordingly, the modulation of gut microflora in liver transplant recipients may not only slow down the progression of cirrhosis, but also improve postoperative outcomes. Nevertheless, no consensus exists on a particular type of probiotic supplement that may bring the most benefits to patients listed for liver transplant procedures.
Previous studies have revealed that cirrhosis is associated with a potentially deleterious shift in the structure of intestinal microbial community: a decrease in the abundance of multiple beneficial bacterial groups, namely members of the autochthonous families (Lachnospiraceae, Ruminococcaceae, Blautia), lactic acid bacteria, and Bifidobacterium species, and an increase in that of potentially pathogenic organisms, particularly Enterobacteriaceae and Enterococcus species (17–19). The purpose of the present study was to determine the components of gut microflora independently associated with the MELD score in cirrhotic patients listed for LT, with special focus on potentially beneficial bacterial species.
Patients and methods This prospective study was performed in the Department of General, Transplant, and Liver Surgery at the Medical University of Warsaw, Poland, in the period between November 2012 and September 2013. The study group comprised 40 patients with liver cirrhosis listed for first LT. None of the patients received probiotics within 4 weeks prior to inclusion in the study. The study protocol has been approved by the local ethics committee at the Medical University of Warsaw. Informed consent was obtained from each of the study participants prior to enrollment. All patients provided stool samples of approximately 100 mL, which were sent to the laboratory within 24 h after collection. Samples were serially diluted in phosphate-buffered saline (pH 7.2) and seeded on enriching and selective agar media to determine the number of colony-forming units per gram (cfu/g) of wet feces for the following microbial groups: Enterococcus species, Escherichia coli, Proteus species, other members of Enterobacteriaceae taxon (Klebsiella, Enterobacter, Citrobacter, and Serratia species), Pseudomonas, Bifidobacterium, Bacteroides, Lactobacillus, and Clostridium species, and yeasts. Microbial counts were assessed following 48-h incubation of agar plates at 37°C in anaerobic conditions for Bacteroides (Schaedler agar), Bifidobacterium (DIC agar), and Clostridium species (SPM agar), and in increased carbon dioxide concentration for Lactobacillus species (Rogosa agar); 24-h incubation in aerobic conditions at 37°C was performed to detect the growth of Enterobacteriaceae and Enterococcus species on U3G agar. Hydrogen peroxide (H2O2)-producing and nonproducing Lactobacillus strains were distinguished based on the reaction with 3,30 ,5,50 -tetramethylbenzidine and peroxidase.
Transplant Infectious Disease 2015: 17: 174–184
175
Graz t et al: Intestinal dysbiosis in LT candidates
Data on patients’ age and gender, etiology of liver disease, serum bilirubin concentration, international normalized ratio (INR), serum creatinine concentration, and the MELD score were also collected. Laboratory measurements used to calculate MELD score were performed in a week preceding collection of stool sample. Microbial counts were presented as natural logarithms of cfu/g. Quantitative variables were compared using the Mann–Whitney U-test. Spearman’s correlation coefficient was used for unadjusted analyses of the associations between microbial counts and the MELD components (serum bilirubin concentration, INR, and serum creatinine concentration). Linear regression models were used to assess microbial counts independently associated with the 3 MELD components. Following the results of these analyses, microbial counts independently associated with serum bilirubin concentration, serum creatinine concentration, or INR were tested for an independent impact on the MELD score in multivariate analysis. Finally, the remaining microbiota was tested for the impact on the abundance of bacterial groups independently associated with the MELD score. Multivariate models were created based on forward stepwise method, using a P-value of 0.15 for both inclusion and removal of variables from the model, as previously recommended (20). Crude (cb) and standardized (sb) regression coefficients were presented with standard errors (SE); regression plots were presented with 95% confidence intervals (95% CI). For the plotting of regression lines in bivariate analyses, a median of the second variable was included in the equation. P-values
