Mortality after cardiac surgery in patients with liver cirrhosis classified by the Child-Pugh score

Transcription

1 Interactive CardioVascular and Thoracic Surgery 20 (2015) doi: /icvts/ivu438 Advance Access publication 22 January 2015 STATE OF THE ART ADULTCARDIAC Cite this article as: Jacob KA, Hjortnaes J, Kranenburg G, de Heer F, Kluin J. Mortality after cardiac in patients with liver cirrhosis classified by the Child- Pugh score. Interact CardioVasc Thorac Surg 2015;20: Mortality after cardiac in patients with liver cirrhosis classified by the Child-Pugh score Kirolos A. Jacob a, *, Jesper Hjortnaes a, Guido Kranenburg b, Frederiek de Heer a and Jolanda Kluin a a b Department of Cardiothoracic Surgery, University Medical Centre Utrecht, Utrecht, Netherlands Julius Center for Health Sciences and Primary Care, University Medical Centre Utrecht, Utrecht, Netherlands * Corresponding author. Department of Cardiothoracic Surgery, University Medical Centre Utrecht, Mail Stop E03.511, PO Box 85500, 3508 GA Utrecht, Netherlands. Tel: ; fax: ; k.a.jacob@umcutrecht.nl (K.A. Jacob). Received 10 September 2014; received in revised form 25 November 2014; accepted 3 December 2014 Abstract Liver cirrhosis is a known risk factor for postoperative mortality in patients undergoing cardiac. Clinical assessment of liver cirrhosis using the widely accepted Child-Pugh (CP) score is thus vital for evaluation of surgical options and perioperative care. However, detailed mortality rates as a consequence of liver cirrhosis are unclear. This review aimed to stratify the risk of short-term (<30 days) and overall (up to 10 years) mortality after cardiac in patients with liver cirrhosis, classified by the CP score. Thus, PubMed, Embase, CINAHL and the Cochrane Library were systematically reviewed by two independent investigators for studies published up to February 2014, in which mortality in cirrhotic patients, classified by the CP classification, undergoing cardiac was evaluated postoperatively. A total of 993 articles were identified. After critical appraisal of 21 articles, 19 were selected for final analysis. Weighted short-term mortality of cirrhotic patients undergoing cardiac was 19.3% [95% confidence interval (CI): %]. Across the different CP groups, shortterm mortality appeared to be 9.0% (95% CI: %), 37.7% (95% CI: %) and 52.0% (95% CI: %) in Groups A, B and C, respectively. Weighted overall mortality within 1 year was 42.0% (95% CI: %) in all cirrhotic patients. Subdivided in groups, overall mortality within that 1 year was 27.2% (95% CI: %), 66.2% (95% CI: %) and 78.9% (95% CI: %) in Groups A, B and C, respectively. In conclusion, short-term mortality is considerably increased in patients with liver cirrhosis CP class B and C. Overall mortality is significantly high in all classes of liver cirrhosis. Keywords: Cardiac Liver cirrhosis Child-Pugh score Mortality INTRODUCTION Cardiac is among the most frequently performed surgical procedures, with over 2 million operations performed worldwide each year [1]. Significant improvements in outcomes have been achieved over the last decades, yet particularly in high-risk patients cardiac still carries a risk of complications. Liver cirrhosis is among a number of concomitant diseases, which are associated with a high risk when undergoing cardiac [2]. The prevalence of liver cirrhosis in patients undergoing cardiac is reported to be %, accounting for over 6000 cases worldwide yearly [3]. Due to a compromised health status, at the time of the postoperative clinical results of cirrhotic patients are poor. Various factors attributed to cirrhosis rather than cardiac are responsible for the poor prognosis [4]. These factors, such as thrombocytopaenia, compromised immune system and other gastrointestinal disorders are correlated with the severity of cirrhosis [4]. Severity of cirrhosis is mainly classified by the Child-Turcotte-Pugh or Child- Pugh (CP) criteria, based upon the following five clinical factors: (i) presence of encephalopathy, (ii) severity of ascites, (iii) total bilirubin level, (iv) albumin level and (v) prothrombin time [5]. The cirrhotic status is graded into three groups according to the sum of the score (Class A, from 5 to 6; Class B, from 7 to 9; Class C, from 10 to 15). Since these variables are easily acquired with minimally invasive methods in common clinical practice, the CP score has become a widely accepted method for determining liver function and its effects on the postoperative risk status [6, 7]. As liver cirrhosis is not included in the European and North- American systems for cardiac operative risk evaluation score, its effects on postoperative outcomes are unclear [8, 9]. Consequently, a review by Modi et al. in 2010 showed that several small studies and case reports have studied the effects of liver cirrhosis on the postoperative prognosis. High 30-day mortality and morbidity rates were demonstrated in patients with liver cirrhosis classified by the CP score [2, 4]. Nevertheless, since publication of the aforementioned review, eight studies have been published and thus, an up-to-date systematic review and detailed analysis of the postoperative outcomes, short term (<30 days) and long term (up to 10 years), in these patients are lacking, and recommendations whether to perform cardiac remain unclear. Hence, the goal of this study was to systematically review observational reports on short-term and overall mortality in patients The Author Published by Oxford University Press on behalf of the European Association for Cardio-Thoracic Surgery. All rights reserved.

2 K.A. Jacob et al. Interactive CardioVascular and Thoracic Surgery 521 Table 1: Search string Database Field Search date: 28 February 2014 Pubmed Title/abstract (((Mortality[MeSH] OR Survival[Mesh] OR mortality[tiab] OR survival[tiab] OR survival rate [tiab] OR death[tiab] OR fatality[tiab])) AND ( Cardiac Surgical Procedures [Mesh] OR Cardiac [tiab] OR Heart [tiab] OR Coronary artery bypass grafting [tiab] OR CABG[tiab] OR Valve [tiab] OR Cardiovascular operation [tiab] OR Cardiopulmonary bypass [tiab] OR Heart bypass [tiab])) AND ( Liver cirrhosis [Mesh] OR Liver cirrhosis [tiab] OR Liver cirrhoses [tiab] OR Hepatic cirrhoses [tiab] OR Hepatic cirrhosis [tiab] OR Liver fibrosis [tiab] OR Liver fibroses [tiab] OR Hepatic fibrosis [tiab] OR Hepatic fibroses [tiab] OR Liver dysfunction [tiab] OR Liver failure [tiab] OR Liver disease [tiab] OR Hepatic failure [tiab] OR Hepatic disease [tiab] OR Hepatic dysfunction [tiab] OR Child-Pugh [tiab] OR Child Pugh [tiab] OR Child-Turcotte-Pugh [tiab] OR Child Turcotte Pugh [tiab]) Embase Title/abstract ( Mortality /exp OR Survival /exp OR mortality:ti,ab OR survival:ti,ab OR survival rate :ti,ab OR death:ti,ab OR fatality:ti, ab) AND ( heart /exp OR Cardiac :ti,ab OR Heart :ti,ab OR Coronary artery bypass grafting : ti,ab OR CABG:ti,ab OR Valve :ti,ab OR Cardiovascular operation :ti,ab OR Cardiopulmonary bypass :ti,ab OR Heart bypass :ti,ab) AND ( Liver cirrhosis /exp OR Liver cirrhosis :ti,ab OR Liver cirrhoses :ti,ab OR Hepatic cirrhosis :ti,ab OR Liver fibrosis :ti,ab OR Liver fibroses :ti,ab OR Hepatic fibrosis :ti,ab OR Hepatic fibroses :ti,ab OR Liver dysfunction :ti,ab OR Liver failure :ti,ab OR Liver disease :ti,ab OR Hepatic failure :ti,ab OR Hepatic disease :ti, ab OR Hepatic dysfunction :ti,ab OR Child-Pugh :ti,ab OR Child Pugh :ti,ab OR Child-Turcotte-Pugh :ti,ab OR Child Turcotte Pugh :ti,ab) Cochrane Library with liver cirrhosis classified by the CP score undergoing cardiac and to discuss patient-related factors and surgical/technical considerations to improve insight into potential determinants of success. MATERIALS AND METHODS Search strategy Title/abstract/ keywords A systematic literature search of Medline via PubMed, Embase, CINAHL and the Cochrane Library via Ovid was performed reviewing articles published up to February A search filter was designed using synonyms for domain ( patients with liver cirrhosis undergoing cardiac ), determinant (CP classification) and outcome (mortality) (Table 1). Firstly, all articles were screened on title and abstract by two independent researchers (Kirolos A. Jacob and Guido Kranenburg). Additionally, a manual search through the references of selected articles and the detected reviews and metaanalyses was performed to pinpoint additional relevant studies. Authors were contacted when a publication could not be obtained or if not all required information could be retrieved from the publication. Consensus was achieved by discussion. Mortality or survival or survival rate or death or fatality:ti,ab,kw and heart or Cardiac or Coronary artery bypass grafting or CABG or Valve or Cardiovascular operation or Cardiopulmonary bypass or Heart bypass :ti,ab,kw and Liver cirrhosis or Liver cirrhoses or Hepatic cirrhosis or Liver fibrosis or Liver fibroses or Hepatic fibrosis or Hepatic fibroses or Liver dysfunction or Liver failure or Liver disease or Hepatic failure or Hepatic disease or Hepatic dysfunction or Child-Pugh or Child Pugh or Child-Turcotte-Pugh or Child Turcotte Pugh :ti,ab,kw CINAHL Abstract AB (mortality or survival or survival rate or death or fatality) AND AB ( heart or Cardiac or Coronary artery bypass grafting or CABG or Valve or Cardiovascular operation or Cardiopulmonary bypass or Heart bypass ) AND AB (Liver cirrhosis or Liver cirrhoses or Hepatic cirrhosis or Liver fibrosis or Liver fibroses or Hepatic fibrosis or Hepatic fibroses or Liver dysfunction or Liver failure or Liver disease or Hepatic failure or Hepatic disease or Hepatic dysfunction or Child-Pugh or Child Pugh or Child-Turcotte-Pugh or Child Turcotte Pugh ) was diagnosed preoperatively by a standardized classification of CP score. Articles had to include the primary outcome, that is, absolute rates of mortality in the postoperative period. Short-term mortality was classified as death within 30 days after cardiac. Overall mortality was mortality occurring within 10 years postoperatively. The latter was divided into subgroups of 1, 5 and 10 years of the follow-up. Secondary outcome included occurrence of major morbidity events within 30 days after. Case series, case reports, expert opinions or guidelines were excluded. Data were extracted from the articles by at least two researchers (Kirolos A. Jacob and Guido Kranenburg), which were year of publication, study design, study population, sample size, type of (coronary artery bypass grafting, valve or combination ), short-term and overall mortality rates. If stated, major morbidity event rates and relevant data on independent predictors for poor outcome after cardiac were also included. Predictors that were sought for were the independent variables incorporated in the CP score such as hyperbilirubinaemia and hypoalbuminaemia. When two studies by the same institution reported the same data, the study was considered a duplicate publication, and the publication with either the better quality or more informative data was included. Statistical analysis STATE OF THE ART Selection criteria and data extraction Articles were selected based on predetermined inclusion criteria, being: observational studies reporting on adult (>18 years) human subjects who underwent cardiac in which liver cirrhosis Short-term (<30 days) and overall mortality risks at 1, 5 and 10 years along with major morbidity events within 30 days [(number of events/number of patients) 100] were calculated for each individual study. Furthermore, pooled and weighted, per study sample size, absolute risks and 95% confidence intervals (CIs)

3 522 K.A. Jacob et al. Interactive CardioVascular and Thoracic Surgery were estimated for the included studies. To compare the increased risk of liver cirrhotic patients with the general cardiac surgical population, and to the general cirrhotic population, the mortality rates of the general cirrhotic population along with mortality rates of the Dutch surgical population at 30 days, 1 and 5 years were set up against the mortality rates of cirrhotic patients undergoing cardiac. Data for mortality rates among the general cirrhotic population and among the Dutch cardiac surgical population were extracted from earlier literature [10 12]. Also, a standardized mortality ratio (SMR) was calculated in studies, which stated the European System for Cardiac Operative Risk Evaluation (EuroSCORE) of the study population. The SMR was calculated for each study, by dividing the observed number of deaths within 30 days of in the total study sample of liver cirrhotic patients, by the average logistic EuroSCORE (= expected mortality) of the same sample. This analysis was performed only for shortterm mortality. The acquired SMR of each study, along with the averaged weighted SMR of all studies put together, were set up against the SMR of several European countries as the Dutch, English and Swedish cardiac surgical population [12 14]. RESULTS The study selection process is illustrated in Fig. 1. Overall, 991 articles were identified of which 19 retrospective studies met the inclusion criteria, and were selected for further analysis. An overview of study characteristics, along with, if stated in a study, the predictors of poor outcome obtained by the systematic review is illustrated in Table 2. These studies included 638 cirrhotic patients, of whom 123 died postoperatively within 30 days, giving an observed weighted short-term mortality after cardiac of 19.3% (95% CI: %). When comparing the different CP groups, an Figure 1: Study selection process. GUCH: grown-up congenital heart disease; LC: liver cirrhosis; MELD: model for end-stage liver disease.

4 Table 2: Study characteristics Author, Country, year (reference) Study design (level of evidence) Surgery type Sample size (%) Mean age (±SD) Follow-up Mortality absolute risk (%) Major morbidity/ complications absolute risk Comments An et al. China, (2007) [15] (2b) CABG, valve, pericardectomy, aortic Arif et al. Germany, (2012) [16] Bizouarn et al. France, (1999) [17] (2b) PC (2a) CABG, valve, combined, transplantation, assist device implantation CABG and valve Filsoufi et al. USA, (2007) [6] (2b) CABG, valve, pericardectomy, aortic Gundling et al. Germany, (2010) [18] Hayashida et al. Japan, (2004) [19] Kaplan et al. Turkey, (2002) [20] (2b) (2b) (4) CABG, valve, aortic and cardiac tumour resection CABG, valve, aortic CABG and valve 24 A: 17 (71) B: 6 (25) C: 1 (4) 109 A: 74 (68) B: 29 (27) C: 6 (5) 12 A: 10 (83) B: 2 (17) 27 A: 10 (37) B: 11 (41) C: 6 (22) 47 A: 33 (70) B: 14 (30) 18 A: 10 (56) B: 7 (39) C: 1 (5) 10 A: 4 (40) B: 6 (60) 53 (13) In-hospital and 30 days after discharge 64 (11) In-hospital and 30 days after discharge and 1 and 5 years 59 (14) In-hospital stay and up to 2 years 58 (10) In-hospital and 30 days after discharge and 1 year 65 (12) In-hospital and 30 days after discharge and 8 years 64 (12) In-hospital and 30 days after discharge 6 (25) A: 1/17 (6) B: 4/6 (67) 28 (26) A: 14/74 (19) B: 12/29 (41) C: 2/6 (33) 1 year: A: 22/74 (30) B: 19/29 (67) C: 4/6 (67) 5 years: A: 55/74 (74) B: 27/29 (95) C: 6/6 (100) 1 (8) A: 2/10 (20) B: 1/2 (50) 7 (26) A: 1/10 (10) B: 2 (18) C: 4/6 (67) 1 year: A: 2/10 (20) B: 10/11 (90) C: 5/6 (83) 9 (19) A: 2/33 (6) B: 7/14 (50) 8 years: A: 7/33 (21) B: 7/14 (50) 3 (17) A: 0/10 (0) B: 2/7 (29) 57 (7) In-hospital stay 3 (30) A: 0/4 (0) B: 3/6 (50) 16 (67) A: 9/17 (53) B: 6/6 (100) 7 (58) A: 5/10 (50) B: 2/2 (100) 14 (52) A: 2/10 (20) B: 6/11 (56) C: 6/6 (100) 14 (78) A: 6/10 (60) B: CPB: 7/7 (100) Morbidity: renal failure, reoperation, MOF, respiratory failure, hepatic failure Reopening rates: A: 18% B: 50% Predictors for poor outcome: Thrombocytopaenia, low preoperative serum cholinesterase Predictors for poor outcome: bilirubinaemia, hypoalbuminaemia, low total protein, prolonged prothrombin time and high EuroSCORE Higher risk (2 ) after multivariable analyses in more severe cirrhotic patients of: invasive ventilation, sepsis, dialysis, tracheotomy, GI tract complications, neurological complications, blood transfusion, prolonged intensive care unit stay Morbidity: reoperation, cardiac tamponade MOF, respiratory failure, hepatic failure, encephalopathy Predictors for poor outcome: thrombocytopaenia No mortality observed in off-pump cardiac Morbidity: renal failure, reoperation, ischaemic GI tract complications, respiratory failure, deep sternal wound infection Higher risk after multivariable analyses in more severe cirrhotic patients of: prolonged ventilation time, sepsis, haemodialysis, blood transfusion, prolonged intensive care unit stay Morbidity: infection, renal failure, respiratory failure, bleeding, GI tract complications Major morbidity in Class B was only 33% in the off-pump cardiac when compared with 100% on-pump cardiac No mortality observed in off-pump cardiac High morbidity and mortality if: postoperative haemorrhagia, dependency Continued K.A. Jacob et al. Interactive CardioVascular and Thoracic Surgery 523 STATE OF THE ART

5 524 Table 2: (Continued) Author, Country, year (reference) Study design (level of evidence) Surgery type Sample size (%) Mean age (±SD) Follow-up Mortality absolute risk (%) Major morbidity/ complications absolute risk Comments on mechanical ventilation, hepatic and renal failure, GI bleeding, sepsis Klemperer et al. USA, (1998) [21] Komoda et al. Germany, (2013) [22] (2b) (2b) CABG and valve Radical pericardectomy (median sternotomy or anterolateral thoracotomy) Lin et al. China, (2005) [23] (2b) CABG and valve Lopez et al. Spain, (2013) [24] CABG and valve Morimoto et al. Japan, (2013) [25] Morisaki et al. Japan, (2010) [26] CABG, valve, aortic CABG and valve 13 A: 8 (62) B: 5 (38) 64 A: 45 (70) B/C: 19 (30) 18 A: 13 (72) B: 4 (22) C: 1 (6) 58 A: 34 (59) B: 21 (36) C: 3 (5) 32 A: 14 (44) B: 14 (44) C: 4 (13) 42 A: 30 (71) B: 12 (29) 65 (8.3) In-hospital stay 4 (31) A: (0) B: 4/5 (80) 52 (15) 63 (10) In-hospital and 30-day after discharge and 5 years 56 (range 35 76) In-hospital and 30 days after discharge and 2 years 65 (12) In-hospital and 30-day after discharge 70 (9) In-hospital and 30 days after discharge and 5 years 8 (13) A: 1/45 (2) B/C: 7/19 (37) 5 years: A: 9/45 (19) B/C: 12/19 (62) 1 (6) A: 1/13 (8) B: 0/4 (0) C: 0/0 (0) 2 years: A: 2/13 (15) B: 1/4 (25) C: 0/0 (0) 7 (12) A: 0/34 (0) B: 5/21 (24) C: 2/3 (67) 5 (16) A: 2/14 (14) B: 3/14 (21) C: 0/4 (0) 5 years A: 4/14 (31) B: 5/14 (33) C: 1/4 (25) 69 (9) In-hospital stay 4 (10) A: 0/30 (0) B: 4/12 (33) 7 (54) A: 2/8 (25) B: 5/5 (100) 9 (50) A: 5/13 (39) B: 3/4 (75) 17 (53) A: 5/14 (36) B: 8/14 (64) C: 4/4 (100) Reopening rates: A: 12.5% B: 100% Morbidity: renal failure, reoperation, respiratory failure, severe neurologicalsymptoms, GI tract complications, >24 h intensive care unit stay Predictors for poor outcome: end-stage renal disease, moderate to severe, ascites, hypoalbuminaemia, mediastinal irradiation, age Morbidity: postoperative bleeding, hepatic decompensation, renal failure, GI tract complications, deep sternal wound infection, bacteraemia and lumbar discitis Morbidity: bleeding requiring reoperation, sepsis, stroke, MOF, prolonged ventilation for >72 h, myocardial infarction, arrhythmia and new-onset renal failure K.A. Jacob et al. Interactive CardioVascular and Thoracic Surgery Continued

6 Murashita et al. Japan, (2008) [27] Sugimura et al. Japan, (2012) [28] CABG and valve CABG, valve, aortic Suman et al. USA, (2004) [7] CABG, valve, aortic Thielmann et al. Germany, (2010) [29] Vanhuyse et al. France, (2012) [30] Yamane et al. Japan, (2008) [31] (2b) CABG, valve, aortic and cardiac tumour resection CABG, valve, aortic, atrial septal defect closure CABG, valve, aortic, pericardectomy 12 A: 6 (50) B: 6 (50) 13 A: 7 (54) B: 4 (31) C: 1 (8) 44 A: 31 (70) B: 12 (27) C: 1 (2) 57 A: 39 (68) B: 14 (25) C: 4 (7) 34 A: 22 (65) B: 10 (29) C: 2 (6) 21 A: 13 (62) B: 7 (33) C: 1 (5) 70 (9) In-hospital and 30 days after discharge 61 (11) In-hospital and 30 days after discharge 4 (33) A: 3/6 (50) B: 1/6 (17) 1 (8) A: 0/7 (0) B: 1/4 (25) C: 0/1 (0) 65 (13) In-hospital stay 7 (16) A: 1/31 (3) B: 5/12 (42) 62 (10) In-hospital and 30 days after discharge and 1, 5 and 10 years 65 (11) In-hospital and 30 days after discharge and 5 years 17 (30) A: 6/39 (15) B: 7/14 (50) C: 4/4 (100) 1 year: A: 12/39 (32) B: 11/14 (73) C: 4/4 (100) 5 years: A: 21/39 (54) B: 11/14 (73) C: 4/4 (100) 10 years: A: 28/39 (72) B: 11/14 (73) C: 4/4 (100) 10 (26) A: 4/22 (18) B: 4/10 (40) C: 2/2 (100) 1 year: A: 5/22 (23) B: 4/10 (40) C: 2/2 (100) 5 years A: 9/22 (40) B: 8/10 (80) C: 2/2 (100) 66 (10) In-hospital stay 1 (5) A: 0/13 (0) B: 1/7 (14) C:0/1 (1) 9 (75) A: 3/6 (50) B: 6/6 (100) 9 (69) A: 4/7 (57) B: 4/4 (100) 10 (29) A: 4/22 (18) B: 4/10 (40) C: 2/2 (100) 6 (29) A: 1/13 (8) B: 4/7 (57) C:1/1 (100) Patients classified by Child-Pugh scores A, B and C. Predictors for poor outcome stated by studies were reported after multivariable logistic regression analyses. CABG: coronary artery bypass grafting; GI: gastrointestinal; MOF: multiorgan failure; : not specified; : retrospective cohort. Major morbidity: bleeding, hepatic failure, infection and renal failure Morbidity: re-ventilation, infection, renal failure and cerebral infarct Higher risk after multivariable analyses in more severe cirrhotic patients of: stroke, renal failure, major bleeding, reoperation and perioperative myocardial infarction Morbidity: respiratory failure, ischaemic/ haemorrhagic complications, renal failure, deep sterna infection, wound infection and reoperation Morbidity: new dialysis, mechanical ventilation >48 h, mediastinitis, hepatic decompensation K.A. Jacob et al. Interactive CardioVascular and Thoracic Surgery 525 STATE OF THE ART

7 526 K.A. Jacob et al. Interactive CardioVascular and Thoracic Surgery observed weighted short-term mortality of 9.0% (95% CI: %) was found in Group A, 37.7% (95% CI: %) in Group B and 52.0% (95% CI: %) in Group C (Table 3). The CP groups in studies were comparable with respect to demographic and surgical characteristics across studies. No data were found in literature on crude mortality within 30 days in the general liver cirrhotic population standardized by the CP score, thus only the general Dutch cardiac surgical population and cirrhotic patients undergoing cardiac were set up against each other. When setting the cardiac surgical crude observed mortality, in cirrhotic patients, within 30 days of the follow-up, against crude observed mortality in Dutch patients without liver cirrhosis undergoing cardiac, it appears that cirrhotic patients have higher deaths rates when compared with the general cardiac surgical population (Fig. 2). Weighted overall crude mortality in studies with a follow-up within 1 year was high, being 42.0% (95% CI: %) in cirrhotic cardiac surgical patients in general. Overall crude mortality subdivided in groups was 27.2% (95% CI: %) in Group A, 66.2% (95% CI: %) in Group B and 78.9% (95% CI: %) in Group C. Also, at 5 and 10 years, the overall mortality increased progressively in all CP classes (Table 4). When comparing the crude death rates of cirrhotic patients in general and the Dutch cardiac surgical population with the acquired data on cirrhotic patients undergoing cardiac, at 1 and 5 years of the follow-up, cirrhotic patients had a higher observed crude mortality rates in the years following (Fig. 2). The high crude mortality within 5 years in cirrhotic patients undergoing was comparable with the general cirrhotic population, but still augmented when compared with the general cardiac surgical population (Fig. 2). The SMR in studies (n = 4), which stated the EuroSCORE, is depicted in Fig. 3. An average SMR of 1.39 (95% CI: ) for cirrhotic patients was found, compared with 0.40 (95% CI: ) in the general cardiac population in Netherlands (2013), 0.49 (95% CI: ) in Sweden (2013) and 0.47 (95% CI: ) in the UK (2013) (Fig. 3). Liver cirrhosis, classified by CP score A, B or C, did not increase the SMR significantly when compared with the general cardiac surgical population. Additionally, an analysis was performed in studies stating shortterm major morbidity in cirrhotic patients after cardiac, for example, reoperation, multiorgan failure, encephalopathy, gastrointestinal (GI) tract complications, >24 h intensive care unit stay and deep wound infections (Table 2). A weighted percentage of 52.7 ( %) was found in patients with liver cirrhosis developing major morbidity after. This was 35.4% ( %) in CP group A, 72.4% ( %) in CP group B and 100% in CP group C (Table 5). Noteworthy, reoperation rates were stated by two studies and were for CP class A % and for class B % [15, 21] (Table 5). DISCUSSION This systematic review illustrates that the observed weighted short-term mortality is high in patients with liver cirrhosis, especially in CP class B and C when compared with patients without liver cirrhosis undergoing cardiac. Additionally, overall weighted mortality within 1, 5 and 10 years is high in all classes of liver cirrhosis. After adjustment for EuroSCORE, the high mortality compared with the general cardiac surgical population was not statistically significantly increased. The mean 30-day mortality after cardiac in all patients undergoing surgical procedures in Netherlands is 3.0%, while overall mortality after 1 year is 6.0% [33]. Few and scattered evidence exists on whether to perform cardiac on cirrhotic patients or not. As is previously stated by this systematic review, Table 3: Short-term mortality (<30 days) analysis per study, given in absolute numbers Study Child-Pugh A Child-Pugh B Child-Pugh C All LC patients Deaths Total Deaths Total Deaths Total Deaths Total An et al Arif et al Bizouarn et al Filsoufi et al Gundling et al Hayashida et al Kaplan et al Klemperer et al Komoda et al Lin et al Lopez et al Morimoto et al Morisaki et al Murashita et al Surgimura et al Suman et al Thielman et al Vanhuyse et al Yamane et al Total Overall weighted mortality 9.0% ( %) 37.3% ( %) 52.0% ( %) 19.3% ( %) Overall mortality rates (95% confidence intervals) are given after a weighted calculation to study sample size. LC: liver cirrhosis.

8 K.A. Jacob et al. Interactive CardioVascular and Thoracic Surgery 527 Figure 2: Mortality within 30 days (A), 1 year (B) and up to 5 years (C) compared between the general Dutch cardiac surgical population, liver cirrhotic patients in general and liver cirrhotic patients undergoing cardiac across the three classes of the Child-Pugh score. Mortality at 5 years in the general cardiac surgical population includes only the valve population [32]. CTS: cardiothoracic ; LC: liver cirrhosis. the observed crude mortality percentages within <30 days and for 1, 5 and 10 years tend to be higher in patients with liver cirrhosis compared with the general cardiac surgical population, making liver cirrhosis of great clinical relevance in assessing the perioperative care. When short-term mortality is adjusted for by the EuroSCORE, these increased death rates were still three times as high as the general cardiac surgical population. Yet the new calculated high SMR was insignificant, which implies that not all cirrhotic patients, especially CP class A, may have a higher risk of <30-day mortality. Hence, cirrhosis is to be first classified into CP scores to identify the higher risk patients. Patients with liver cirrhosis classified by the CP score have a 1-year survival of 100, 81and 45%, in Class A, B and C, respectively (Fig. 2) [10]. Accordingly, it is generally accepted that CP class A patients, unlike classes B and C, may undergo cardiac since the increase of short-term mortality compared with the general cardiac surgical population, is relatively low. Interestingly, this review demonstrates that overall survival in cirrhotic patients after cardiac is poor in all CP classes. Patients with liver cirrhosis class A have an almost seven times higher risk of dying within 12 years after cardiac. Nevertheless, in spite of the high early death rates among patients with liver cirrhosis CP class B and C, case reports have shown that thorough perioperative management of patients can result in good early clinical outcome [34, 35]. Hence, careful and meticulous consideration of perioperative management techniques and surgical strategy, that is, off-pump for isolated coronary artery bypass grafting, is vital and recommended for the safety of cirrhotic patients postoperatively to decrease not only early but also late mortality. Cirrhosis is classified mainly by the CP score, a subjective score, however readily performed in clinical practice. Another internationally accepted score for the classification of liver cirrhosis is the Model for End-stage Liver Disease (MELD) score. The MELD score has also been shown to predict mortality after cardiac significantly, as it also includes renal function, and has less subjective variables as the CP score, that is, severity of ascites and encephalopathy [25]. Nonetheless, recent studies have shown that CP scores are similar to or even better than MELD scores in predicting mortality after cardiac [6, 7, 36], though, on the other hand, other studies have shown MELD to be a more superior score to predict outcomes after cardiac [24]. Indeed, in some studies, mortality was better predicted by MELD score than CP score after cardiac [25, 30, 37]. Hepatic cirrhosis is well known to often be accompanied by anaemia, leucopaenia and thrombocytopaenia due to poor nutritional status, splenomegaly and bleeding from varices [4]. Additionally, impaired coagulation may occur in cirrhotic patients since many of the coagulation factors are synthesized by the liver. The most serious and frequently encountered postoperative complications in cirrhotic patients are infections and excessive mediastinal bleeding (Table 2). Liver dysfunction results in increased infection rates not only by altering the immune status, but also due to the fact that a higher incidence of reoperations takes place in cirrhotic patients as they have multiple bleeding episodes [38] (Table 2). As infections are the leading cause of deaths among hospitalized patients in the postoperative period, scrupulous management of perioperative immune status of patients is crucial in decreasing the rates of infections and diminishing death rates. Hence, meticulous measurements need to be taken to reduce those infections, for example, screening preoperatively for pathogens carriage, early diagnosis and antibiotics or immune globulin administration are vital. The second major complication following cardiac in cirrhotic patients is, as aforementioned, excessive bleeding, especially from the mediastinum. Thrombocytopaenia and the coexisting coagulopathy lead to increased risk of bleeding and reoperations after cardiac (Table 2). In addition, the cardiopulmonary bypass aggravates these haematological disorders in coagulopathy by hypothermia, haemodilution and hypoperfusion [39]. It has been illustrated that off-pump cardiac in cirrhotic patients results into fewer deaths, especially in patients with severe CP classes, that is, B and C [6, 19, 20]. Though off-pump cardiac is not always possible, its employment as an alternative treatment strategy should be assessed in every patient individually. Noteworthy, despite the good results of off-pump shown by studies analysed in this review, only a small number of patients were included and the overall survival after this technique is unknown [6, 19]. Other treatments that can be performed as preventive measures are vitamin K and tranexamic acid administration along with fresh frozen plasma and packed cell transfusions. STATE OF THE ART

9 528 K.A. Jacob et al. Interactive CardioVascular and Thoracic Surgery Table 4: Pooled analysis of overall mortality within 1, 5 and 10 years per study, given in absolute numbers Study Child-Pugh A Child-Pugh B Child-Pugh C All LC patients Deaths Total Deaths Total Deaths Total Deaths Total Mortality within 1 year of the follow-up Arif et al Filsoufi et al Lin et al Thielmann et al Vanhuyse et al Total Overall weighted mortality 27.2% ( %) 66.2% ( %) 78.9% ( %) 42.0% ( %) Mortality within 5 years of the follow-up Arif et al Komoda et al Morimoto et al Thielmann et al Vanhuyse et al Total Overall Weighted Mortality 50.5% ( %) 73.2% ( %) 81.3% ( %) 58.1% ( %) Mortality within 10 years of the follow-up Gundling et al Thielmann et al Total Overall Weighted Mortality 56.9% ( %) 64.3% ( %) 100% 60.6% ( %) Overall mortality rates (95% confidence intervals) are given after a weighted calculation to study sample size. LC: liver cirrhosis. Figure 3: Short-term standardized mortality ratio (SMR) in cardiac according to the four studies stating the EuroSCORE. Weighted average SMR of the five studies is given along with the national Dutch, English and Swedish SMR in cardiac surgical patients. CI: confidence interval; LC: liver cirrhosis; N: number; NL: Netherlands; SW: Sweden; UK: United Kingdom. Limitations Selection bias is likely since selection criteria were implemented and because unpublished data, abstracts and presentations were excluded. Also, selection bias within studies could have occurred, nevertheless the detailed observations by the two independent researchers of the demographic and clinical characteristics of included patients, such as age, perioperative procedures and postoperative management protocols within studies, were comparable across the three CP scores. Another drawback of this systematic review is that heterogeneous populations from various countries, with different kinds of surgical procedures, were found across studies that were used for this mortality analysis. Moreover, the number of studies included was retrospective and incorporated small sample sizes, studying a subjective score as the CP score. Nonetheless, the retrieved results were considered to be the best available evidence. Additionally, pooling data illustrated a strong relationship between liver cirrhosis and mortality after cardiac across all CP classes, which illustrate that differences between populations in this situation are not significantly present.

10 K.A. Jacob et al. Interactive CardioVascular and Thoracic Surgery 529 Table 5: Major short-term (<30 days) morbidity analysis per study, given in absolute numbers Study Child-Pugh A Child-Pugh B Child-Pugh C All LC patients Morbidity Total Morbidity Total Morbidity Total Morbidity Total An et al Bizouarn et al Filsoufi et al Hayashida et al Klemperer et al Lin et al Morimoto et al Murashita et al Surgimura et al Vanhuyse et al Yamane et al Total Overall weighted morbidity 35.4% ( %) 72.4% ( %) 100% 52.7% ( %) Overall morbidity rates (95% confidence intervals) are given after a weighted calculation to study sample size. LC: liver cirrhosis. CONCLUSION Strong evidence on the exact risks of liver cirrhosis after cardiac is lacking. The current evidence from cohort studies with small sample sizes shows that short-term mortality in patients undergoing cardiac is high in patients with liver cirrhosis, especially in Grades B and C classified by the CP score. Overall mortality after up to 10 years of the follow-up is high in all classes of liver cirrhosis. The authors recommend high caution in patients with liver cirrhosis CP classes B and C undergoing cardiac. Cardiothoracic surgeons operating patients with cirrhosis CP class A are advised to carefully and accurately consider and surgical strategy, that is, off-pump if possible. Conflict of interest: none declared. REFERENCES [1] Roger VL, Go AS, Lloyd-Jones DM, Adams RJ, Berry JD, Brown TM et al. Heart disease and stroke statistics 2011 update: a report from the American Heart Association. Circulation 2011;123:e18 e209. [2] Modi A, Vohra HA, Barlow CW. Do patients with liver cirrhosis undergoing cardiac have acceptable outcomes? Interact CardioVasc Thorac Surg 2010;11: [3] Shaheen AA, Kaplan GG, Hubbard JN, Myers RP. Morbidity and mortality following coronary artery bypass graft in patients with cirrhosis: a population-based study. Liver Int 2009;29: [4] Hayashida N, Aoyagi S. Cardiac operations in cirrhotic patients. Ann Thorac Cardiovasc Surg 2004;10: [5] Pugh RN, Murray-Lyon IM, Dawson JL, Pietroni MC, Williams R. Transection of the oesophagus for bleeding oesophageal varices. Br J Surg 1973;60: [6] Filsoufi F, Salzberg SP, Rahmanian PB, Schiano TD, Elsiesy H, Squire A et al. Early and late outcome of cardiac in patients with liver cirrhosis. Liver Transpl 2007;13: [7] Suman A, Barnes DS, Zein NN, Levinthal GN, Connor JT, Carey WD. Predicting outcome after cardiac in patients with cirrhosis: a comparison of Child-Pugh and MELD scores. Clin Gastroenterol Hepatol 2004; 2: [8] Geissler HJ, Holzl P, Marohl S, Kuhn-Regnier F, Mehlhorn U, Sudkamp M et al. Risk stratification in heart : comparison of six score systems. Eur J Cardiothorac Surg 2000;17: [9] Nashef SA, Roques F, Michel P, Gauducheau E, Lemeshow S, Salamon R. European system for cardiac operative risk evaluation (EuroSCORE). Eur J Cardiothorac Surg 1999;16:9 13. [10] Cholongitas E, Papatheodoridis GV, Vangeli M, Terreni N, Patch D, Burroughs AK. Systematic review: the model for end-stage liver disease should it replace Child-Pugh s classification for assessing prognosis in cirrhosis? Aliment Pharmacol Ther 2005;22: [11] Infante-Rivard C, Esnaola S, Villeneuve JP. Clinical and statistical validity of conventional prognostic factors in predicting short-term survival among cirrhotics. Hepatology 1987;7: [12] Siregar S, Groenwold RH, de Mol BA, Speekenbrink RG, Versteegh MI, Brandon Bravo Bruinsma GJ et al. Evaluation of cardiac mortality rates: 30-day mortality or longer follow-up? Eur J Cardiothorac Surg 2013;44: [13] Swedish Heart Surgery Registry Annual Report Scand Cardiovasc J 2013;47(Suppl 62): [14] Hickey LG, Grant SW, Freemantle N, Cunningham D, Munsch CM, Livesey SA et al. Surgeon length of service and risk-adjusted outcomes: linked observational analysis of the UK National Adult Cardiac Surgery Audit Registry and General Medical Council Register. J R Soc Med. 2014;107: [15] An Y, Xiao YB, Zhong QJ. Open-heart in patients with liver cirrhosis. Eur J Cardiothorac Surg 2007;31: [16] Arif R, Seppelt P, Schwill S, Kojic D, Ghodsizad A, Ruhparwar A et al. Predictive risk factors for patients with cirrhosis undergoing heart. Ann Thorac Surg 2012;94: [17] Bizouarn P, Ausseur A, Desseigne P, Le Teurnier Y, Nougarede B, Train M et al. Early and late outcome after elective cardiac in patients with cirrhosis. Ann Thorac Surg 1999;67: [18] Gundling F, Seidl H, Gansera L, Schuster T, Hoffmann E, Kemkes BM et al. Early and late outcomes of cardiac operations in patients with cirrhosis: a retrospective survival-rate analysis of 47 patients over 8 years. Eur J Gastroenterol Hepatol 2010;22: [19] Hayashida N, Shoujima T, Teshima H, Yokokura Y, Takagi K, Tomoeda H et al. Clinical outcome after cardiac operations in patients with cirrhosis. Ann Thorac Surg 2004;77: [20] Kaplan M, Cimen S, Kut MS, Demirtas MM. Cardiac operations for patients with chronic liver disease. Heart Surg Forum 2002;5:60 5. [21] Klemperer JD, Ko W, Krieger KH, Connolly M, Rosengart TK, Altorki NK et al. Cardiac operations in patients with cirrhosis. Ann Thorac Surg 1998;65:85 7. [22] Komoda T, Frumkin A, Knosalla C, Hetzer R. Child-Pugh score predicts survival after radical pericardiectomy for constrictive pericarditis. Ann Thorac Surg 2013;96: [23] Lin CH, Lin FY, Wang SS, Yu HY, Hsu RB. Cardiac in patients with liver cirrhosis. Ann Thorac Surg 2005;79: STATE OF THE ART

11 530 K.A. Jacob et al. Interactive CardioVascular and Thoracic Surgery [24] Lopez-Delgado JC, Esteve F, Javierre C, Perez X, Torrado H, Carrio ML et al. Short-term independent mortality risk factors in patients with cirrhosis undergoing cardiac. Interact CardioVasc Thorac Surg 2013;16: [25] Morimoto N, Okada K, Okita Y. The model for end-stage liver disease (MELD) predicts early and late outcomes of cardiovascular operations in patients with liver cirrhosis. Ann Thorac Surg 2013;96: [26] Morisaki A, Hosono M, Sasaki Y, Kubo S, Hirai H, Suehiro S et al. Risk factor analysis in patients with liver cirrhosis undergoing cardiovascular operations. Ann Thorac Surg 2010;89: [27] Murashita T, Komiya T, Tamura N, Sakaguchi G, Kobayashi T, Furukawa T et al. Preoperative evaluation of patients with liver cirrhosis undergoing open heart. Gen Thorac Cardiovasc Surg 2009;57: [28] Sugimura Y, Toyama M, Katoh M, Kato Y, Hisamoto K. Analysis of open heart in patients with liver cirrhosis. Asian Cardiovasc Thorac Ann 2012;20: [29] Thielmann M, Mechmet A, Neuhauser M, Wendt D, Tossios P, Canbay A et al. Risk prediction and outcomes in patients with liver cirrhosis undergoing open-heart. Eur J Cardiothorac Surg 2010;38: [30] Vanhuyse F, Maureira P, Portocarrero E, Laurent N, Lekehal M, Carteaux JP et al. Cardiac in cirrhotic patients: results and evaluation of risk factors. Eur J Cardiothorac Surg 2012;42: [31] Yamane K, Izumi K, Yamachika S, Hashizume K, Tanigawa K, Miura T et al. Operative outcome of cardiac in patients with liver cirrhosis. Acta Med Nagasaki 2008;53: [32] Siregar S, de Heer F, Groenwold RH, Versteegh MI, Bekkers JA, Brinkman ES et al. Trends and outcomes of valve : 16-year results of Netherlands Cardiac Surgery National Database. Eur J Cardiothorac Surg 2014;46:386 97; discussion 97. [33] Siregar S, Groenwold RH, Versteegh MI, Takkenberg JJ, Bots ML, van der Graaf Y et al. Data resource profile: adult cardiac database of the Netherlands Association for Cardio-Thoracic Surgery. Int J Epidemiol 2013;42: [34] Ben Ari A, Elinav E, Elami A, Matot I. Off-pump coronary artery bypass grafting in a patient with child class C liver cirrhosis awaiting liver transplantation. Br J Anaesth 2006;97: [35] Kubota Y, Sakaguchi T, Miyagawa S, Nishi H, Yoshikawa Y, Fukushima S et al. Successful management of complex open heart in a patient with Child-Pugh class C liver cirrhosis: report of a case. Surg Today 2013;43: [36] Ailawadi G, Lapar DJ, Swenson BR, Siefert SA, Lau C, Kern JA et al. Model for end-stage liver disease predicts mortality for tricuspid valve. Ann Thorac Surg 2009;87:1460 7; discussion [37] Morimoto N, Okada K, Okita Y. Results of cardiac in advanced liver cirrhosis. Gen Thorac Cardiovasc Surg 2013;61: [38] Bonnel AR, Bunchorntavakul C, Reddy KR. Immune dysfunction and infections in patients with cirrhosis. Clin Gastroenterol Hepatol 2011;9: [39] Mazer CD, Hornstein A, Freedman J. Platelet activation in warm and cold heart. Ann Thorac Surg 1995;59: