Skip to main content
NCBI home page
Medicine logoLink to Medicine
. 2018 Oct 12;97(41):e12758. doi: 10.1097/MD.0000000000012758

A model predicting short-term mortality in patients with advanced liver cirrhosis and concomitant infection

Ying Li a,b, Roongruedee Chaiteerakij b,c, Jung Hyun Kwon d, Jeong Won Jang b,e, Hae Lim Lee e, Stephen Cha f, Xi Wei Ding b, Charat Thongprayoon b, Fu Shuang Ha a, Cai Yun Nie a, Qian Zhang a, Zhen Yang b, Nasra H Giama b, Lewis R Roberts b,∗, Tao Han a,∗
Editor: Ewa Janczewska
PMCID: PMC6203558  PMID: 30313084

Abstract

Infection is a common cause of death in patients with advanced cirrhosis. We aimed to develop a predictive model in Child–Turcotte–Pugh (CTP) class C cirrhotics hospitalized with infection for optimizing treatment and improving outcomes.

Clinical information was retrospectively abstracted from 244 patients at Tianjin Third Central Hospital, China (cohort 1). Factors associated with mortality were determined using logistic regression. The model for predicting 90-day mortality was then constructed by decision tree analysis. The model was further validated in 91 patients at Mayo Clinic, Rochester, MN (cohort 2) and 82 patients at Seoul St. Mary's Hospital, Korea (cohort 3). The predictive performance of the model was compared with that of the CTP, model for end-stage liver disease (MELD), MELD-Na, Chronic Liver Failure–Sequential Organ Failure Assessment, and the North American consortium for the Study of End-stage Liver Disease (NACSELD) models.

The 3-month mortality was 58%, 58%, and 54% in cohort 1, 2, and 3, respectively. In cohort 1, respiratory failure, renal failure, international normalized ratio, total bilirubin, and neutrophil percentage were determinants of 3-month mortality, with odds ratios of 16.6, 3.3, 2.0, 1.1, and 1.03, respectively (P < .05). These parameters were incorporated into the decision tree model, yielding area under receiver operating characteristic (AUROC) of 0.804. The model had excellent reproducibility in the U.S. (AUROC 0.808) and Korea cohort (AUROC 0.809). The proposed model has the highest AUROC and best Youden index of 0.488 and greatest overall correctness of 75%, compared with other models evaluated.

The proposed model reliably predicts survival of advanced cirrhotics with infection in both Asian and U.S. populations.

Keywords: 3-month mortality, cirrhosis, infection, MELD score

1. Introduction

Infection is a common cause of liver decompensation and death in patients with cirrhosis. In addition, infection is a precipitating factor for development of acute-on-chronic liver failure (ACLF) and extrahepatic organ failure in cirrhosis patients. Further, infection is one of the most common reasons for rehospitalization, increased health care costs, and impaired quality of life of patients with cirrhosis.[1,2]

Cirrhotic patients have a nearly 5-fold greater risk of developing infection than the general population due to the cirrhosis-related immunodeficiency state.[3–7] Approximately, 25% to 35% of cirrhotic patients have infections at admission or during hospitalization.[8,9] The mortality of cirrhosis patients with infection is around 4-fold higher than cirrhosis patients without infection.[1,10,11] Advanced cirrhosis patients with poor liver function have higher infection risk and mortality rates; this group of patients should therefore be targeted for prevention, early diagnosis, and treatment of infection. Delays in therapy and inappropriate therapy result in increased mortality.[6,12,13] A reliable predictor of survival would be helpful for managing this patient population. The current most widely used clinical scoring systems, including the Child–Turcotte–Pugh (CTP) score, Model of End-Stage Liver Disease (MELD), and MELD-Na do not specifically focus on advanced cirrhotic patients with infection. In this study, we propose a new predictive model from variables, which are easy to obtain and efficient for prediction of outcome for this particular group of patients in China. As a secondary aim, the proposed model was externally validated in U.S. and Korean cohorts. Lastly, the predictive performance of the proposed model was compared with that of commonly used scoring systems, that is, CTP score, MELD, MELD-Na, Chronic Liver Failure–Sequential Organ Failure Assessment (CLIF-SOFA), and the North American Consortium for the Study of End-stage Liver Disease (NACSELD) model. Our predictive model should help physicians better manage patients and make appropriate individualized treatment plans based on the condition of each patient.

2. Methods

2.1. Patient cohorts

A total of 1819 cirrhosis patients (519 from China, 1015 from the U.S., and 285 from Korea) were screened for this study. The diagnosis of cirrhosis was made by liver histopathology, or a combination of laboratory biochemistry, radiologic, and endoscopic findings, if a liver biopsy result was not available.[14] Patients with CTP C cirrhosis who were diagnosed with infection on admission (<48 hours) were included. Patients with malignancy or concomitant human immunodeficiency virus infection, or those who received liver transplantation before the study period were excluded. The research protocol was approved by the Ethics Committees of the Tianjin Third Central Hospital, the Mayo Clinic and Seoul St. Mary's Hospital.

After screening, 244 patients admitted to Tianjin Third Central Hospital, China, between January 2010 and April 2013 were included in the test cohort. Ninety-one patients admitted to Mayo Clinic, Rochester, MN, between January 2007 and December 2013, and 82 patients admitted to Seoul St. Mary's Hospital, Korea, between January 2011 and May 2014 served as validation cohorts.

2.2. Data collection

Baseline demographic characteristics, type of infection, and the presence of organ failure were retrospectively abstracted from the medical record. Infections were classified as follows[14–19]:

  • (1)

    Spontaneous bacterial peritonitis (SBP): an absolute number of polymorphonuclear cells of ≥250/mm3 or positive bacteriological culture of ascitic fluid;

  • (2)

    Pneumonia: at least 1 of the respiratory symptoms (i.e., cough, sputum production, dyspnea, pleuritic pain) in combination with one of the following: rales and/or crepitation on auscultation; at least 1 sign of infection (i.e., shivering, body temperature >38°C or <36°C, or leucocyte count >10,000/mm3 or <4000/mm3) in the absence of antibiotics; presence of pulmonary infiltrate on radiologic imaging by chest X-ray or computed tomography (CT) scan; or positive bacteriological culture in sputum;

  • (3)

    Enterocolitis infection: diarrhea with the presence of leukocytes in stool or positive stool culture for pathogens, including Salmonella, Shigella, Yersinia, Campylobacter, pathogenic Escherichia coli, or a positive Clostridium difficile stool assay;

  • (4)

    Urinary tract infection (UTI): >15 urine leukocytes/high power field with either positive gram stain or positive urine bacterial culture;

  • (5)

    Skin/soft-tissue infection: cellulitis;

  • (6)

    Empyema: an absolute number of polymorphonuclear cells ≥250/mm3 or positive bacterial culture in pleural fluid;

  • (7)

    Other infections: infections not included above.

Diagnostic criteria for organ failure were as follows[14,20–29]:

  • (1)

    Acute on chronic liver failure: defined by the Canonic Study;

  • (2)

    Renal failure: serum creatinine (sCr) ≥2 mg/dL or urine volume < 0.5 mL/kg/h for 6 hours;

  • (3)

    Respiratory failure: paO2 <60 mm Hg on room air O2 or paCO2 >50 mm Hg on room air O2 and/or requirement for mechanical ventilation;

  • (4)

    Shock: use of dopamine, dobutamine, terlipressin, or mean arterial pressure <60 mm Hg despite adequate fluid resuscitation and cardiac output.

All patients were followed for 90 days. The primary outcome was death within 3 months after hospitalization.

2.3. Statistical analysis

Continuous variables were summarized as mean ± standard deviation (SD) and compared using Student t test. Categorical variables were summarized as frequency (%) and compared using Pearson Chi-square or Fisher Exact test as appropriate. Predictors of 90-day mortality were determined using logistic regression. All variables with P < .05 in the univariate analysis were included in the multivariate logistic regression model, and the stepwise elimination method was used to identify independent predictors. These predictors, identified by multivariate logistic regression model, were further evaluated by decision tree method to determine the most reliable predictors of death and their optimal cutoff values.[30–33] The patients were divided into 2 groups according to these defined predictors. Each subgroup was repeatedly assessed and divided according to this 2-choice branching method. Finally, a 90-day mortality prediction model was proposed using predictors identified by the decision tree analysis. Receiver operating characteristic (ROC) curves were constructed to assess sensitivity, specificity, and respective areas under the curves (AUCs) with 95% confidence interval (95% CI). The cut-off points were determined when sensitivity and specificity reached the peak (The best Youden index). A log-rank test for Kaplan–Meier analysis and hazard ratio were used to statistically test short-term mortality. A P value <.05 was considered as statistically significant. All statistical analyses were performed using SPSS software version 13.0 (SPSS Inc., Chicago, IL), JMP 10 (SAS Institute Inc., Cary, NC), R software, and MedCalc (version 11.4.2.0).

3. Results

3.1. Patient characteristics

Demographics and baseline clinical characteristics of the 244 patients in the test cohort are summarized in Table 1. The mean (± SD) age was 52.1 ± 11.6 years and 169 (69%) were male. The mean MELD score was 26 ± 8. The 90-day mortality rate was 58%.

Table 1.

Baseline characteristics of the test cohort.

3.1.

Open in a new tab

3.2. Infection characteristics

In the test cohort, 156 (64%) patients had a single infection site, 68 (28%) had 2 infection sites, and 20 (8%) had ≥3 infection sites on admission. The 3-month mortality rates were 56%, 56%, and 80% for those with 1, 2, and ≥3 infection sites, respectively. Pneumonia was the most frequent infection (n = 157, 64%), followed by SBP (n = 105, 43%) and infectious enterocolitis (n = 33, 14%), with corresponding 3-month mortalities of 63%, 62%, and 52%, respectively.

3.3. Organ failure characteristics

One hundred seventy-five (72%) patients had no extrahepatic organ failure, 54 (22%) patients had 1 extrahepatic organ failure, and 15 (6%) patients had ≥2 extrahepatic organ failures, with corresponding 3-month mortality rates of 48%, 80%, and 93%, respectively. The most frequent extrahepatic organ failure was renal failure (n = 60, 25%), followed by respiratory failure (n = 18, 7%), and shock (n = 9, 4%), resulting in 3-month mortalities of 83%, 94%, and 78%, respectively.

3.4. Univariate and multivariate logistic regression analysis of factors determining 3-month survival

Table 2 summarizes univariate and multivariate analysis of variables associated with 90-day mortality. By multivariate analysis, respiratory failure, renal failure, international normalized ratio (INR), total bilirubin (TBIL), and neutrophil percentage (NEU%) were significantly associated with death, with odds ratios (ORs) (95% CI) of 16.6 (1.9–143.1), 3.3 (1.4–7.8), 2.0 (1.3–3.1), 1.1 (1.0–1.1), and 1.03 (1.0–1.1), respectively.

Table 2.

Univariate and multivariate logistic regression analysis.

3.4.

Open in a new tab

3.5. Prognostic model for 3-month mortality by decision tree analysis

The results of the decision tree analysis are shown in Fig. 1. After decision tree analysis by R software, INR was selected as the first predictor, with a cutoff value of 3.0. The 3-month mortality rate was 90% for patients with INR ≥3.0 versus 50% for patients with INR <3.0. For patients with INR <3.0, the next most important (second) predictor of death was respiratory failure, with a 3-month mortality rate of 92% for those with respiratory failure versus 47% for those without respiratory failure. For patients without respiratory failure, the third predictor was NEU%. The 3-month mortality rate was 54% for patients with higher NEU% (≥71%) versus 17% in patients with lower NEU% (<71%). The fourth predictor for NEU% <71% patients was TBIL, with a cutoff value of 7.2 mg/dL. The 3-month mortality rates were 40% for patients with TBIL ≥7.2 mg/dL versus 0% in patients with TBIL <7.2. For NEU% ≥71% patients, TBIL also served as a fourth predictor, at a cutoff value of 9.3 mg/dL. The 3-month mortality rates were 67% in patients with TBIL ≥9.3 mg/dL versus 42% in patients with TBIL <9.3 mg/dL. The fifth predictor for patients with TBIL <9.3 mg/dL was renal failure. The 3-month mortality was 73% in those with renal failure versus 35% in patients without renal failure. This analysis identified 7 subgroups with 3-month mortality rates ranging from 0% to 92%.

Figure 1.

Figure 1

Open in a new tab

The decision tree model of 3-month mortality for advanced cirrhosis patients with concomitant infection. Boxes show that the factors and cutoff values differentiate groups of patients. Pie charts showed the 3-month mortality rates of each group after differentiation. The classified groups are numbered from 1 to 7.

3.6. Assess the accuracy of the 3-month mortality prognostic model

The discriminatory values and predictive accuracies of the proposed decision tree model and MELD, MELD-Na, CTP, and CLIF-SOFA score, and NACSELD study model[19] are listed in Table 3. According to the analysis of the AUROC curves, the decision tree model provided the best discriminatory power (AUROC 0.804) and was significantly more predictive than the MELD-Na score (AUROC 0.734). Among the 6 scoring systems evaluated, the decision tree model had the highest overall correctness of 75% and the best Youden index of 0.488.

Table 3.

Prediction of 3-month mortality for infected patients with advanced cirrhosis.

3.6.

Open in a new tab

For more convenient use in clinical practice, the patients were classified into low, intermediate, and high-risk groups (Table 4 and Fig. 2). On the basis of AUROC curves and hazard ratio analysis, the proposed 3 groups were more predictive of mortality than the 3 groups of MELD score classified by tertiles, that is, MELD <23, ≥23 to <28, and ≥28 [AUROC: 0.771 vs 0.713 in China cohort (Fig. 3A–C)].

Table 4.

Constitution of low-risk group, intermediate-risk group, and high-risk groups.

3.6.

Open in a new tab

Figure 2.

Figure 2

Open in a new tab

The patients were classified into 3 groups as low, intermediate, and high-risk groups according to the decision tree model.

Figure 3.

Figure 3

Open in a new tab

Three-month mortality prediction outcomes of decision tree model and MELD score for advanced cirrhosis patients with infection. (A–C) ROC curve, Kaplan–Meier curve, and hazard ratio for decision tree model and MELD score in China cohort.

3.7. External validation of the proposed prognostic model

Demographics of the 91 patients in the first validation cohort and 82 patients in the second validation cohort are summarized in Table 5. For the first validation cohort, the overall 3-month mortality rate was 58%. SBP and UTI were the most frequent infections (n = 21 each, with mortality rates of 67% and 57%, respectively). The patients were subdivided according to the prognostic model and subclassified into the low, intermediate, and high-risk subgroups (Fig. 4A–C), which had 3-month mortalities of 17%, 64%, and 84%, respectively (P < .001), compared with the 3 subgroups classified by MELD score, which had 3-month mortalities of 23%, 56%, and 79%. The decision tree model was more predictive of death than the MELD score model, with an AUROC (95% CI) of 0.808 (0.706–0.910) versus 0.756 (0.643–0.868).

Table 5.

Baseline characteristics of validation cohort.

3.7.

Open in a new tab

Figure 4.

Figure 4

Open in a new tab

Three-month mortality prediction outcomes of decision tree model and MELD score for advanced cirrhosis patients with infection. (A–C) ROC curve, Kaplan–Meier curve, and hazard ratio for decision tree model and MELD score in U.S. validation cohort. (D–F) ROC curve, Kaplan–Meier curve, and hazard ratio for decision tree model and MELD score in Korea validation cohort.

For the second validation cohort, the overall 3-month mortality rate was 54%. SBP was the most frequent infection (n = 24, mortality rate 67%). According to the proposed prognostic model, 3-month mortalities were 21%, 52%, and 87% in the low, intermediate, and high-risk subgroups, respectively (P < .001). When using classified MELD score model, patients had 3-month mortalities of 23%, 50%, and 79%, respectively. Consistent with the finding in the first validation cohort, the decision tree model was more predictive of death than the MELD score model, with an AUROC (95% CI) of 0.809 (0.711–0.906) versus 0.771 (0.664–0.878) (Fig. 4D–F).

4. Discussion

Patients with advanced cirrhosis are at a high risk of infection, which is commonly complicated by liver and/or extrahepatic organ failure(s), resulting in short-term death.[1,2,34,35] In this multicenter study, the overall 3-month mortality rate of this group of advanced cirrhosis patients with infection was almost 60%, which was higher than that found in a meta-analysis reporting a 44% 3-month mortality rate of all cirrhosis patients enrolled between 1978 and 2009 who were admitted with infection.[11] Thus, this is a major life-threatening health problem for patients with advanced liver cirrhosis. Early diagnosis and appropriate treatment are critical for improving outcomes. This investigation identified respiratory failure, renal failure, INR, TBIL, and NEU% as independent predictors of 3-month mortality. A decision tree model built using these independent predictors was proved to have excellent performance for predicting 3-month mortality and was applicable for patients from different areas with different etiologies. This system subclassified patients into low, intermediate, and high-risk subgroups; the 3-month mortality of the low-risk subgroup was around 20%, while the 3-month mortality of the high-risk subgroup was close to 90%. The system had better discriminatory ability than the MELD score, MELD-Na score, CTP score, CLIF-SOFA score, and NACSELD study models and may promote the more rational use of medical resources.

This study found some characteristics of infection for advanced cirrhosis patients. Approximately 36% of the patients in the test cohort had multiple infection sites. For patients who had more than 2 infection sites, the 3-month mortality rates significantly increased by over 20% compared with those with 1 or 2 infection sites. SBP and pneumonia were the most frequent infections in this multicenter study. Our findings were similar to a report from the NACSELD study, which included 18 referral centers across the US and Canada.[14] In the present study, patients with pneumonia or SBP infections had higher 3-month mortality than those with other types of infections, and also had higher mortality rates of 62% to 75% than a previous meta-analysis, which showed mortality rates of 37% to 43% for all cirrhosis patients admitted with infections.[11] Neither the infection type nor the number of infection site was independent risk factor for 3-month mortality. Nineteen percent of patients had virus coinfection, which was respiratory system infection. Sixteen percent of patients had fungus coinfection. The virus coinfection was not associated with 3-month mortality.

Presumably, the higher mortality rates in the more advanced cirrhosis patients in this study reflect their more severe level of liver dysfunction, which is more often complicated by other extrahepatic organ failure(s).[1,2] In this study, renal failure and respiratory failure were independent risk factors for 3-month mortality. Many studies have shown that bacterial infections induce a profound pro-inflammatory response of the host immune system that can precipitate organ failure.[36] The most frequent organ failure was renal failure. Bacterial infections are well-known triggers of kidney failure in cirrhosis, primarily due to the induction of the host inflammatory response and hemodynamic alterations.[1,37–39] Many studies have shown that renal failure is a major determinant of mortality in cirrhotic patients.[11,40,41] Consistent with the other studies, renal failure was an independent risk factor for 3-month mortality in this study. Of the patients who developed renal failure, the 3-month mortality was 81% to 84%. Respiratory failure was another independent predictor of mortality; indeed, it was the most lethal organ failure complication, being associated with 3-month mortality rates of 94% to 100%. The NACSELD study also showed that patients with cirrhosis complicated with respiratory failure had the highest 30-day mortality of any organ failure of around 61%.[14] However, unlike the situation with renal failure, for which a number of studies have explored the underlying mechanisms of interaction with advanced cirrhosis, there are relatively few studies focusing on the mechanisms underlying respiratory failure and its progression in cirrhotic patients. This is a research area of unmet need.

TBIL and INR are both measures of liver function and components of the MELD score. As expected, multivariate logistic analysis showed that TBIL and INR were also determinants of 3-month mortality. We found that NEU% was also a determinant of 3-month mortality in this special population, with higher levels of NEU% associated with higher mortality rates. Similar to the NACSELD study,[14] after multivariate analysis, NEU% was independent predictor for 3-month mortality. In the inflammatory process, neutrophils are rapidly recruited to sites of acute inflammation; neutrophils and the innate immune system, in particular, play a central role in the inflammatory pathology and multiple organ dysfunction.[42–44]

We also tested the component parameters of the NACSELD study model for their utility in this model[19]; heart rate and second infection were statistically significant in the univariate logistic analysis, but not independent predictors for 3-month mortality in the multivariate analysis. This difference is presumably due to the difference in the study populations. In this study, the patients all had advanced cirrhosis, CTP C grade, and MELD scores around 30, compared with the broader distribution of the NACSELD study patients, with CTP grades A–C and a mean MELD of 20.

The final decision tree model included INR, respiratory failure, NEU%, TBIL, and renal failure as independent predictors. The model efficiently classified patients into low, intermediate, and high-risk subgroups. The 3-month mortality in the high-risk subgroup was almost 90%, compared with a 3-month mortality of only around 20% in the low-risk subgroup. The predictive accuracy of the model was superior to MELD, CTP, MELD-Na, CLIF-SOFA, and the NACSELD study model for advanced cirrhotic patients with infection and had the best Youden index and overall accuracy of prediction. Moreover, in the present study, we also validated the model in both an Asian and a western cohort. Overall, the results showed excellent reproducibility, suggesting that the model is generalizable to patients with similar conditions from different geographic regions. The model is readily obtained from routine physical examination and laboratory testing, and is thus available even in lower resource settings.

Although the study results were encouraging, this study still had several potential limitations. First, this was a retrospective study and the conclusions will need to be further verified by prospective studies. Second, because of the relatively small number of CTP grade C patients presenting with infection at each medical center, the predictive accuracy of the decision tree model may be limited.

In conclusion, patients with advanced cirrhosis and infection were more vulnerable to liver and/or extrahepatic organ failure with resultant high short-term mortality. SBP and pneumonia were the most frequent infections and were associated with higher short-term mortality than infections at other sites. The INR, respiratory failure, NEU%, TBIL, and renal failure all contributed independently to a model that achieved excellent discrimination of patients into low, intermediate, and high-risk subgroup. The accuracy and reproducibility of the model were validated at 2 different medical centers and this model may prove useful for physicians taking care of these patients to identify those patients at highest risk and help reduce mortality rates.

Author contributions

Ying Li, Roongruedee Chaiteerakij, Jung Hyun Kwon, and Jeong Won Jang were of the same contribution to this article (cofirst author). The work included study concept and design, acquisition of data, analysis and interpretation of data, drafting of the manuscript, and critical revision of the manuscript for important intellectual content and statistical analysis.

Hae Lim Lee, Xi-Wei Ding, Charat Thongprayoon, Fu-Shuang Ha, Cai-Yun Nie, Qian Zhang, Zhen Yang, and Nasra H. Giama worked on acquisition of data.

Stephen Cha worked on statistical analysis.

Lewis R. Roberts and Tao Han worked on obtaining funding, administration, material support, study supervision, study concept and design, and critical revision of the manuscript for important intellectual content.

Conceptualization: ying Li, Roongruedee Chaiteerakij, Jung Hyun Kwon, Jeong Won Jang, Lewis R. Roberts, Tao Han.

Data curation: ying Li, Roongruedee Chaiteerakij, Jung Hyun Kwon, Jeong Won Jang, Hae Lim Lee, Xi Wei Ding, Charat Thongprayoon, Fu Shuang Ha, Cai Yun Nie, Qian Zhang, Zhen Yang, Nasra H. Giama, Lewis R. Roberts, Tao Han.

Formal analysis: ying Li, Roongruedee Chaiteerakij, Jung Hyun Kwon, Jeong Won Jang, Stephen Cha, Lewis R. Roberts, Tao Han.

Investigation: ying Li, Lewis R. Roberts, Tao Han.

Writing – original draft: ying Li, Roongruedee Chaiteerakij, Jung Hyun Kwon, Jeong Won Jang.

Funding acquisition: Lewis R. Roberts, Tao Han.

Methodology: Lewis R. Roberts, Tao Han.

Project administration: Lewis R. Roberts, Tao Han.

Supervision: Lewis R. Roberts, Tao Han.

Validation: Lewis R. Roberts, Tao Han.

Writing – review & editing: Lewis R. Roberts, Tao Han.

Footnotes

Abbreviations: ACLF = acute-on-chronic liver failure, AFP = alpha fetal protein, ALT = alanine aminotransferase, AST = aspartate aminotransferase, BUN = blood urea nitrogen, CHE = cholinesterase, CLIF-SOFA = Chronic Liver Failure–Sequential Organ Failure Assessment, CTP = Child–Turcotte–Pugh, INR = international normalized ratio, MELD = model for end-stage liver disease, NACSELD = the North American consortium for the Study of End-stage Liver Disease, NEU% = neutrophil percentage, PALB = prealbumin, PBC = primary biliary cirrhosis, ROC = receiver operating characteristic, SBP = spontaneous bacterial peritonitis, sCr = serum creatinine, TBIL = total bilirubin, UTI = urinary tract infection, WBC = white blood cell.

Funding/support: This work was supported by the National 12th 5-year Plan for Hepatitis Research (No.2012ZX10002004–011); Tianjin Science and Technology Fund, China (No.12JCYBJC17300); and Tianjin Science and Technology Fund, China (No.13RCGFSY19200). National Institutes of Health Grants CA165076 (to LRR); and the Mayo Clinic Center for Translational Science Activities (NIH/NCRR CTSA Grant Number UL1 TR000135). Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the NIH.

There is no conflict of interest to declare.

References

  • [1].Jalan R, Fernandez J, Wiest R, et al. Bacterial infections in cirrhosis: a position statement based on the EASL Special Conference 2013. J Hepatol 2014;60:1310–24. [DOI] [PubMed] [Google Scholar]
  • [2].Bajaj JS, O’Leary JG, Wong F, et al. Bacterial infections in end-stage liver disease: current challenges and future directions. Gut 2012;61:1219–25. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [3].Tritto G, Bechlis Z, Stadlbauer V, et al. Evidence of neutrophil functional defect despite inflammation in stable cirrhosis. J Hepatol 2011;55:574–81. [DOI] [PubMed] [Google Scholar]
  • [4].Wasmuth HE, Kunz D, Yagmur E, et al. Patients with acute on chronic liver failure display “sepsis-like” immune paralysis. J Hepatol 2005;42:195–201. [DOI] [PubMed] [Google Scholar]
  • [5].Malik R, Mookerjee RP, Jalan R. Infection and inflammation in liver failure: two sides of the same coin. J Hepatol 2009;51:426–9. [DOI] [PubMed] [Google Scholar]
  • [6].Fernandez J, Gustot T. Management of bacterial infections in cirrhosis. J Hepatol 2012;56:S1–2. [DOI] [PubMed] [Google Scholar]
  • [7].Gustot T, Durand F, Lebrec D, et al. Severe sepsis in cirrhosis. Hepatology 2009;50:2022–33. [DOI] [PubMed] [Google Scholar]
  • [8].Fernandez J, Navasa M, Gomez J, et al. Bacterial infections in cirrhosis: epidemiological changes with invasive procedures and norfloxacin prophylaxis. Hepatology 2002;35:140–8. [DOI] [PubMed] [Google Scholar]
  • [9].Fernández J, Acevedo J, Castro M, et al. Prevalence and risk factors of infections by multiresistant bacteria in cirrhosis: a prospective study. Hepatology 2012;55:1551–61. [DOI] [PubMed] [Google Scholar]
  • [10].Foreman MG, Mannino DM, Moss M. Cirrhosis as a risk factor for sepsis and death: analysis of the National Hospital Discharge Survey. Chest 2003;124:1016–20. [DOI] [PubMed] [Google Scholar]
  • [11].Arvaniti V, D’ Amico G, Fede G, et al. Infections in patients with cirrhosis increase mortality four-fold and should be used in determining prognosis. Gastroenterology 2010;139:1246–56. [DOI] [PubMed] [Google Scholar]
  • [12].Arabi YM, Dara SI, Memish Z, et al. Antimicrobial therapeutic determinants of outcomes from septic shock among patients with cirrhosis. Hepatology 2012;56:2305–15. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [13].Ginès P, Angeli P, Lenz K, et al. EASL clinical practice guidelines on the management of ascites, spontaneous bacterial peritonitis, and hepatorenal syndrome in cirrhosis. J Hepatol 2010;53:397–417. [DOI] [PubMed] [Google Scholar]
  • [14].Bajaj JS, O’Leary JG, Reddy KR, et al. Survival in infection-related acute-on-chronic liver failure is defined by extrahepatic organ failures. Hepatology 2014;60:250–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [15].Cazzaniga M, Dionigi E, Gobbo G, et al. The systemic inflammatory response syndrome in cirrhotic patients: relationship with their in-hospital outcome. J Hepatol 2009;51:475–82. [DOI] [PubMed] [Google Scholar]
  • [16].Rimola A, Garcia-Tsao G, Navasa M, et al. Diagnosis, treatment and prophylaxis of spontaneous bacterial peritonitis: a consensus document. J Hepatol 2000;32:142–53. [DOI] [PubMed] [Google Scholar]
  • [17].Sobel JD, Kaye D. Mandell GL, Bennet JE, Dolin R. Urinary tract infections. Principles and Practice of Infectious Diseases 4th ed.New York: Churchill Livingstone; 1995. 662–90. [Google Scholar]
  • [18].Fasolato S, Angeli P, Dallagnese L, et al. Renal failure and bacterial infections in patients with cirrhosis: epidemiology and clinical features. Hepatology 2007;45:223–9. [DOI] [PubMed] [Google Scholar]
  • [19].Bajaj JS, O’Leary JG, Reddy KR, et al. Second infections independently increase mortality in hospitalized patients with cirrhosis: the North American consortium for the study of end-stage liver disease (NACSELD) experience. Hepatology 2012;56:2328–35. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [20].Moreau R, Jalan R, Gines P, et al. Acute-on-chronic liver failure is a distinct syndrome that develops in patients with acute decompensation of cirrhosis. Gastroenterology 2013;144:1426–37. [DOI] [PubMed] [Google Scholar]
  • [21].Cordoba J, Ventura-Cots M, Simón-Talero M, et al. Characteristics, risk factors, and mortality of cirrhotic patients hospitalized for hepatic encephalopathy with and without acute-on-chronic liver failure (ACLF). J Hepatol 2014;60:275–81. [DOI] [PubMed] [Google Scholar]
  • [22].Kellum JA, Lameire N. for the KDIGO AKI Guideline Work Group. Diagnosis, evaluation, and management of acute kidney injury: a KDIGO summary (Part 1). Crit Care 2013;17:204. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [23].Blei AT, Córdoba J. Practice Parameters Committee of the American College of Gastroenterology. Hepatic encephalopathy. Am J Gastroenterol 2001;96:1968–76. [DOI] [PubMed] [Google Scholar]
  • [24].Sun QF, Ding JG, Xu DZ, et al. Prediction of the prognosis of patients with acute-on-chronic hepatitis B liver failure using the model for end-stage liver disease scoring system and a novel logistic regression model. J Viral Hepat 2009;16:464–70. [DOI] [PubMed] [Google Scholar]
  • [25].Wehler M, Kokoska J, Reulbach U, et al. Short-term prognosis in critically ill patients with cirrhosis assessed by prognostic scoring systems. Hepatology 2001;34:255–61. [DOI] [PubMed] [Google Scholar]
  • [26].Moreau R, Lebrec D. Acute renal failure in patients with cirrhosis: perspectives in the age of MELD. Hepatology 2003;37:233–43. [DOI] [PubMed] [Google Scholar]
  • [27].Ginès P, Schrier RW. Renal failure in cirrhosis. N Engl J Med 2009;361:1279–90. [DOI] [PubMed] [Google Scholar]
  • [28].Wiesner R, Edwards E, Freeman R, et al. Model for end-stage liver disease (MELD) and allocation of donor livers. Gastroenterology 2003;124:91–6. [DOI] [PubMed] [Google Scholar]
  • [29].Rice TW, Wheeler AP, Bernard GR, et al. Comparison of the SpO2/FIO2 ratio and the PaO2/FIO2 ratio in patients with acute lung injury or ARDS. Chest 2007;132:410–7. [DOI] [PubMed] [Google Scholar]
  • [30].Kurosaki M, Hiramatsu N, Sakamoto M, et al. Data mining model using simple and readily available factors could identify patients at high risk for hepatocellular carcinoma in chronic hepatitis C. J Hepatol 2012;56:602–8. [DOI] [PubMed] [Google Scholar]
  • [31].Kurosaki M, Sakamoto N, Iwasaki M, et al. Pretreatment prediction of response to peginterferon plus ribavirin therapy in genotype 1 chronic hepatitis C using data mining analysis. J Gastroenterol 2011;46:401–9. [DOI] [PubMed] [Google Scholar]
  • [32].Kurosaki M, Tanaka Y, Nishida N, et al. Pre-treatment prediction of response to pegylated-interferon plus ribavirin for chronic hepatitis C using genetic polymorphism in IL28B and viral factors. J Hepatol 2011;54:439–48. [DOI] [PubMed] [Google Scholar]
  • [33].Hiramatsu N, Kurosaki M, Sakamoto N, et al. Pretreatment prediction of anemia progression by pegylated interferon alpha-2b plus ribavirin combination therapy in chronic hepatitis C infection: decision-tree analysis. J Gastroenterol 2011;46:1111–9. [DOI] [PubMed] [Google Scholar]
  • [34].Jalan R, Gines P, Olson J, et al. Acute on-chronic liver failure. J Hepatol 2012;57:1336–48. [DOI] [PubMed] [Google Scholar]
  • [35].Plessier A, Denninger MH, Consigny Y, et al. Coagulation disorders in patients with cirrhosis and severe sepsis. Liver Int 2003;23:440–8. [DOI] [PubMed] [Google Scholar]
  • [36].Medzhitov R, Schneider DS, Soares MP. Disease tolerance as a defense strategy. Science 2012;335:936–41. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [37].Navasa M, Follo A, Filella X, et al. Tumor necrosis factor and interleukin-6 in spontaneous bacterial peritonitis in cirrhosis: relationship with the development of renal impairment and mortality. Hepatology 1998;27:1227–32. [DOI] [PubMed] [Google Scholar]
  • [38].Terra C, Guevara M, Torre A, et al. Renal failure in patients with cirrhosis and sepsis unrelated to spontaneous bacterial peritonitis: value of MELD score. Gastroenterology 2005;129:1944–53. [DOI] [PubMed] [Google Scholar]
  • [39].Sort P, Navasa M, Arroyo V, et al. Effect of intravenous albumin on renal impairment and mortality in patients with cirrhosis and spontaneous bacterial peritonitis. N Engl J Med 1999;341:403–9. [DOI] [PubMed] [Google Scholar]
  • [40].Garcia-Tsao G, Parikh CR, Viola A. Acute kidney injury in cirrhosis. Hepatology 2008;48:2064–77. [DOI] [PubMed] [Google Scholar]
  • [41].Belcher JM, Garcia-Tsao G, Sanyal AJ, et al. Association of AKI with mortality and complications in hospitalized patients with cirrhosis. Hepatology 2013;57:753–62. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [42].Brown KA, Brain SD, Pearson JD, et al. Neutrophils in development of multiple organ failure in sepsis. Lancet 2006;368:157–69. [DOI] [PubMed] [Google Scholar]
  • [43].McDonald B, Kubes P. Neutrophils and intravascular immunity in the liver during infection and sterile inflammation. Toxicol Pathol 2012;40:157–65. [DOI] [PubMed] [Google Scholar]
  • [44].Nathan C. Neutrophils and immunity: challenges and opportunities. Nat Rev Immunol 2006;6:173–82. [DOI] [PubMed] [Google Scholar]

Articles from Medicine are provided here courtesy of Wolters Kluwer Health

RESOURCES