Model for end-stage liver disease predicts right ventricular failure in patients with left ventricular assist devices

Transcription

1 J Artif Organs (2016) 19:21 28 DOI /s x ORIGINAL ARTICLE Artificial Heart (Clinical) Model for end-stage liver disease predicts right ventricular failure in patients with left ventricular assist devices Gardner L. Yost 1 Laura Coyle 1 Geetha Bhat 1 Antone J. Tatooles 2 Received: 6 February 2015 / Accepted: 6 July 2015 / Published online: 18 July 2015 The Japanese Society for Artificial Organs 2015 Abstract High rates of right ventricular failure continue to affect postoperative outcomes in patients implanted with left ventricular assist devices (LVADs). Development of right ventricular failure and implantation with right ventricular assist devices is known to be associated with significantly increased mortality. The model for end-stage liver disease (MELD) score is an effective means of evaluating liver dysfunction. We investigated the prognostic utility of postoperative MELD on post-lvad implantation outcomes. MELD scores, demographic data, and outcomes including length of stay, survival, and postoperative right ventricular failure were collected for 256 patients implanted with continuous flow LVADs. Regression and Kaplan Meier analyses were used to investigate the relationship between MELD and all outcomes. Increased MELD score was found to be an independent predictor of both right heart failure and necessity for RVAD implantation (OR 1.097, CI , p = 0.001; OR 1.121, CI 1.015, p = 0.024, respectively). Patients with RV failure and who underwent RVAD implantation had reduced postoperative survival compared to patients with RV dysfunction (no RV failure = ± days, RV failure = ± days, RVAD = & Gardner L. Yost Gyost2@uic.edu & Geetha Bhat Geetha.Bhat@advocatehealth.com 1 2 Center for Heart Transplant and Assist Devices, Outpatient Pavilion, H&V Administration, Advocate Christ Medical Center, 6th Floor, 4440 West 95th Street, Oak Lawn, IL 60453, USA Division of Cardiothoracic Surgery, Advocate Christ Medical Center, Oak Lawn, IL, USA 89.3 ± 72.8 days; p \ 0.001). In conclusion, MELD can be used to reliably predict postoperative right heart failure and the necessity for RVAD implantation. Those patients with RV failure and RVADs experience significantly increased postoperative mortality compared to those without RV dysfunction. Keywords Ventricle, right Statistics, risk analysis Circulatory assist devices, LVAD, RVAD Liver Outcomes Introduction Mechanical circulatory support with continuous flow left ventricular assist device (LVAD) placement remains an effective treatment for advanced heart failure with 89 % 1-year survival in patients implanted as bridge to transplant [1]. Improved patient selection, advances in pump design, and continued clinical exposure have led to improved survival and quality of life [2, 3]. Though postoperative complication rates have been considerably lower with continuous flow LVADs than with the previous pulsatile devices, INTERMACs reports an average of events/ 100 patient-months between 2008 and 2010 [1, 4, 5]. Among these postoperative adverse events, the incidence of right ventricular (RV) failure ranges from 9 to 44 % with INTERMACs reporting a multi-institution average of 2.55 events/100 patient-months between 2008 and 2010 [1, 6]. Postoperative RV failure is associated with increased morbidity and mortality, and the survival of patients requiring right ventricular assist device (RVAD) support is significantly poorer than those not requiring RVAD support [1, 7, 8]. Postoperative mortality ranges from 40 to 70 % for patients requiring an RVAD [9, 10]. Several groups

2 22 J Artif Organs (2016) 19:21 28 have proposed methodologies for predicting RV failure post-lvad implantation; however, identification of at-risk patients remains difficult as patients with advanced univentricular and biventricular failure demonstrate heterogeneous etiologies and clinical presentations [8, 11 13]. Consequently, there is opportunity for accurate, rapid, and inexpensive prediction of risk for postoperative RV failure to improve preoperative selection and management. In this study, we investigated the relationship between postoperative RV failure and preoperative model for endstage liver disease (MELD) score. Originally developed for use in predicting outcomes following the transjugular intrahepatic portosystemic shunt (TIPS) procedure, the MELD is a composite score which allows for robust evaluation of severity of liver dysfunction. The MELD provides a quantitative measure of liver dysfunction based on a patient s creatinine, total bilirubin, and international normalized ratio (INR). Recently, it has been used to predict survival in patients with heart failure as well as those undergoing heart transplantation and LVAD implantation [14 21]. Liver dysfunction in patients with end-stage heart failure is often attributed to poor end-organ perfusion leading to parenchymal ischemia and hepatocellular necrosis [18, 19]. Further, RV dysfunction and elevated right atrial pressures result in the development of hepatic congestion [20]. Elevated MELD scores (MELD C 17) have been shown to be associated with reduced overall survival following LVAD implantation as well as lower post-transplant survival in patients bridged to cardiac transplant [18]. Typically, after a short postoperative period of worsening, hepatic congestion improves in patients with pre-lvad liver dysfunction. This is presumably due to unloading of the left ventricle resulting in reduction of right-sided congestion. The biomarkers used in the MELD score were hypothesized to reflect risk of RV failure. In this study, we investigated the use of the MELD as a predictor of RV failure, RVAD implantation, and postoperative adverse outcomes following LVAD implantation. Patients and methods Patient selection This retrospective, institutional review board-approved study enrolled 256 patients who were implanted with LVADs between 2005 and 2013 at our institution. All LVADs were continuous flow and were implanted for destination therapy (DT) or bridge to transplant (BTT). Patients undergoing device exchange or those who had been supported with either pulsatile or temporary devices prior to receiving a continuous flow LVAD were excluded. Data collection Prior to implant, laboratory, hemodynamic, echocardiographic, and demographic data were collected. Preoperative MELD ð¼ 9:57½lnðcreatinineÞ þ3:78½lnðbilirubinþ þ11:2½lnðinrþ þ6:43þ and MELD without INR (MELD- XI) ð¼ 5:11½lnðbilirubinÞ þ11:76½lnðcreatinineþþ9:44þ scores were collected for all patients. All anticoagulation with warfarin and clopidogrel was held 3 days prior to surgery. Accordingly, laboratory testing of INR and serum creatinine for calculation of the MELD score was performed immediately prior to surgery. Outcome measures including RV failure, RVAD implantation, postoperative readmissions, and blood pressure, LVAD flow, and echocardiography at discharge were gathered for all patients. RV failure was defined as the necessity for 2 weeks or more of inotropic support or RVAD implantation. The indications for RVAD have been described elsewhere [22]. All right ventricular support was provided by the CentriMag extracorporeal circulatory assist device (Thoratec, Pleasanton, CA). Patients were divided into three groups: those without RV failure, those with RV failure but no RVAD, and those with an RVAD. Continuous variables were expressed as means ± SD if normally distributed and as medians with interquartile ranges for skewed distributions. Categorical variables were expressed as frequencies and percentages. Statistical analysis Data were compared between groups using independent samples t tests or Mann Whitney U tests for continuous variables and Chi square tests for categorical variables. Two forward stepwise multivariate logistic regression models were used to analyze outcomes for the two binary outcome measures: RVAD implantation versus no RVAD implantation and RV failure versus no RV failure. Independent effect variables were identified using univariate analysis (threshold p \ 0.10) for inclusion into the regression model. A receiver-operating characteristics (ROC) curve was generated for RV failure, and area under the curve (AUC) was calculated for MELD and other commonly used predictors for RV failure. Statistical analyses were performed using SPSS version 20 for Windows (SPSS, Inc., Chicago, IL). Results The baseline demographics for the study population of 256 patients are listed in Table 1 and revealed a group with high rates of comorbidities including chronic kidney disease (60 %), atrial fibrillation (41 %), coronary artery disease

3 J Artif Organs (2016) 19: Table 1 Baseline demographics (n = 256) Mean ± SD or median (IQR) Count (%) BMI (kg/m 2 ) ± 6.27 Male 197 (77 %) Sodium (mmol/l) ± 3.70 White 133 (52 %) BUN (mg/dl) ± Device type Creatinine (mg/dl) 1.42 ± 0.47 HeartMate II 212 (82.8 %) BNP (pg/dl) ± HeartWare 44 (17.2 %) AST (unit/l) 27.0 (19) Ventilator 14 (5 %) ALT (unit/l) 37 (22) Ischemic 138 (54 %) T. Bili (mg/dl) 1.11 ± 0.86 CKD 155 (61 %) Albumin (gm/dl) 2.96 ± 0.46 COPD 43 (17 %) INR 1.25 ± 0.35 VT 71 (28 %) White blood cell count (1000/mcl) 7.62 ± 2.80 AFib 106 (41 %) Hemoglobin (gm/dl) 11.1 ± 1.86 DM 109 (43 %) Platelets (1000/mcl) ± HTN 166 (65 %) MELD ± 0.12 MELD-XI ± 5.36 LVEDD (mm) ± LVEF (%) ± 6.40 CVP (mmhg) ± 5.74 PCWP (mmhg) ± 8.02 Cardiac output (L/min) 4.57 ± 1.53 PVR (WU) 2.93 ± 1.65 SVR (dynes * s/cm 5 ) 1302 ± 460 MAP (mmhg) ± RVSWI (mmhg * ml/m 2 ) ± BMI body mass index, BUN blood urea nitrogen, BNP b-type natriuretic peptide, AST aspartate aminotransferase, ALT alanine aminotransferase, T. Bili total bilirubin, INR international normalized ratio, MELD model for end-stage liver disease, MELD-XI model for end-stage liver disease without INR, LVEDD left ventricular end diastolic diameter, LVEF left ventricular ejection fraction, CVP central venous pressure, PCWP pulmonary capillary wedge pressure, PVR pulmonary vascular resistance, SVR systemic vascular resistance, MAP mean arterial pressure, RVSWI right ventricular stroke work index, CKD chronic kidney disease, COPD chronic obstructive pulmonary disease, VT ventricular tachycardia, AFib atrial fibrillation, DM diabetes mellitus, HTN hypertension (58 %), diabetes mellitus (43 %), and hypertension (65 %). Elevated creatinine (1.4 ± 0.5 mg/dl), B-type natriuretic peptide (805.2 ± pg/dl), and bilirubin levels (1.1 ± 0.9 mg/dl), as well as reduced ejection fraction (18.2 ± 6.4 %) and elevated central venous pressure (11.3 ± 5.7 mmhg) demonstrated cardiac dysfunction and hemodynamic derangements indicative of advanced heart failure. A total of 56 patients (21.9 %) developed RV failure and 15 (5.9 %) required RVAD implantation. During follow-up (median 36 months), 52 patients (20.3 %) died. Those who received an RVAD or had RV failure had increased mortality compared to patients without RV failure (no RV failure = ± days, RV failure = ± days, RVAD = 89.3 ± 72.8 days; p \ 0.001). The cause of death for all groups is detailed in Table 2. Kaplan Meier survival analysis indicated significantly reduced survival in patients implanted with an RVAD (Fig. 1). Subsequently, the cohort was split into groups with and without RV failure. Median postoperative survival was significantly greater in the group without RV failure than in the group with RV failure [524 (676) vs (400.4) days; p \ 0.001]. Univariate analysis indicated that postoperative RV failure was associated with increased MELD and MELD-XI (MELD without INR) scores, blood urea nitrogen (BUN), creatinine, and white blood cell count as well as decreased albumin. Patients who had RV failure tended to have higher central venous pressures (CVP) and were more likely to have ischemic etiology of heart failure, and histories of chronic kidney disease (CKD) and ventricular tachycardia (VT) (Table 3). Logistic regression analysis was performed using RV failure as the dependent variable, with independent covariates with p \ 0.1. MELD was found to be an independent predictor for postoperative RV failure with OR 1.125, CI , p = Other variables found to be predictive of postoperative RV

4 24 J Artif Organs (2016) 19:21 28 Table 2 Cause of death by RV failure status Cause of death No RV failure (n = 172 with 30 deaths) (%) RV failure (n = 56 with 15 deaths) (%) RVAD (n = 15 with 7 deaths) (%) Hemorrhagic CVA Ischemic CVA Pump failure or driveline fracture Sepsis Multisystem organ failure Pump disconnect Gastrointestinal bleeding Other Each column shows the respective relative frequencies of death within each group for No RV failure, RV failure, and RVAD that MELD score was also an independent predictor for RVAD implantation (OR 1.12, 95 % CI , p = 0.024). CVP was also an independent predictor of RVAD implantation (Table 4). Subsequently, ROC analysis indicated that a MELD cutoff of 9 was the most sensitive value as a predictor of RV failure. Univariate analysis revealed that patients with MELD [ 9 had higher rates of RVAD implantation and RV failure, significantly longer postoperative length of stay, and more frequent readmissions for heart failure (Table 6). Discussion Fig. 1 Kaplan Meier survival curve showing survival in the no RV failure group, the RV failure group, and the RVAD group failure were ischemic etiology, history of ventricular tachycardia, and serum albumin (Table 4). The AUC for the MELD was ROC curves were also generated for other common predictors of RV failure including pulmonary capillary wedge pressure (PCWP), mean pulmonary artery pressure (MPAP), pulmonary vascular resistance (PVR), right ventricular stroke work index (RVSWI), and transpulmonary gradient (TPG) (Fig. 2, Table 5). Comparison indicates that MELD score is a more discriminative predictor for RV failure. Univariate analysis was used to compare patients who received an RVAD to those who did not. Patients who underwent RVAD implantation tended to be younger, and to have lower serum sodium and higher MELD, MELD-XI, and CVP. These patients were also more likely to have a history of ventricular tachycardia. A second multivariate model utilizing parameters from the univariate analysis of RVAD versus non-rvad patients, with p \ 0.1, indicated Patients supported by LVADs are living longer as technology and medical and surgical therapies improve. However, the development of early postoperative RV dysfunction is associated with poorer survival, greater length of stay, and increased postoperative bleeding and renal failure following LVAD implantation [1, 6]. Consequently, as we continue to treat end-stage heart failure, careful patient selection will remain critical in ensuring optimal postoperative outcomes. The future of device therapy is likely to depend on length of stay, cost-effectiveness, readmission rates, and quality of life. Consequently, the development of reliable and reproducible risk stratification tools is vital to the improvement of both patient selection and outcomes. With rates of RV failure as high as 44 % in patients receiving LVADs in some institutions, we sought to explore the relationship between liver function and postoperative RV failure in our population of 256 patients in an effort to develop improved preoperative screening measures [6, 9]. Current predictive risk factors for the development of RV failure are often costly and time consuming. Demographics, type or indication of LVAD support, preoperative circulatory failure, need for

5 J Artif Organs (2016) 19: Table 3 Demographics by RV failure status No RV failure (n = 172) RV failure (n = 56) P value No RVAD (241) RVAD (15) P value Age (years) 59.6 ± ± ± ± Weight (kg) 84.0 ± ± ± ± BMI (kg/m 2 ) 27.4 ± ± ± ± Sodium (mmol/l) ± ± ± ± BUN (mg/dl) 23.3 ± ± 17.5 \ ± ± Creatinine (mg/dl) 1.3 ± ± 0.6 \ ± ± BNP (pg/dl) (728) (696) (726) 799 (660) 0.22 AST (unit/l) 32.2 ± ± ± ± ALT (unit/l) 50.7 ± ± ± ± T. Bili (mg/dl) 0.95 ± ± ± ± INR 1.1 ± ± ± ± Albumin (gm/dl) 3.0 ± ± ± ± White blood cell count (1000/mcl) 7.2 ± ± ± ± Hemoglobin (gm/dl) 11.3 ± ± ± ± Platelets (1000/mcl) ± ± ± ± MELD 10.3 ± ± 5.2 \ ± ± MELD-XI 11.5 ± ± ± ± LVEDD (mm) 69.3 ± ± ± ± LVEF (%) 18.3 ± ± ± ± Central venous pressure (mmhg) 10.7 ± ± ± ± PCWP (mmhg) 22.8 ± ± ± ± CVP/PCWP 0.47 ± ± ± ± Mean pulmonary artery pressure 34.2 ± ± ± ± (mmhg) Cardiac output (L/min) 4.5 ± ± ± ± PVR (WU) 2.9 ± ± ± ± SVR (dynes * s/cm 5 ) ± ± ± ± MAP (mmhg) 80.3 ± ± ± ± RVSWI (mmhg * ml/m 2 ) ± ± ± ± Count (%) Male 133 (76.4 %) 64 (78.0 %) (78.0 %) 9 (60.0 %) 0.11 African American 66 (37.9 %) 34 (41.5 %) (39.0 %) 6 (40.0 %) 0.94 Caucasian 91 (52.3 %) 42 (51.2 %) (51.9 %) 8 (53.3 %) 0.91 Ischemic 85 (48.9 %) 53 (64.6 %) (53.5 %) 9 (60.0 %) 0.63 Ventilator dependent 8 (4.7 %) 4 (7.1 %) (5.0 %) 2 (13.3 %) Indication BTT 32 (18.6 %) 8 (14.2 %) 39 (16.2 %) 4 (26.7 %) DT 140 (81.6 %) 48 (85.7 %) (83.8 %) 11 (73.3 %) 0.29 Chronic kidney disease 96 (55.2 %) 59 (72.0 %) (60.6 %) 9 (60.0 %) 0.96 Chronic obstructive pulmonary disease 31 (17.8 %) 12 (14.6 %) (17.0 %) 2 (13.3 %) 0.71 Ventricular tachycardia 39 (22.4 %) 32 (39.0 %) (29.0 %) 1 (6.7 %) 0.06 Atrial fibrillation 72 (41.4 %) 34 (41.5 %) (42.7 %) 3 (20.0 %) 0.08 Diabetes mellitus 71 (40.8 %) 38 (46.3 %) (42.7 %) 6 (40.0 %) 0.84 Hypertension 112 (64.4 %) 54 (65.9 %) (65.1 %) 9 (60.0 %) 0.69 BUN blood urea nitrogen, BNP b-type natriuretic peptide, AST aspartate aminotransferase, ALT alanine aminotransferase, T. Bili total bilirubin, INR international normalized ratio, MELD model for end-stage liver disease, MELD-XI model for end-stage liver disease without INR, LVEDD left ventricular end diastolic diameter, LVEF left ventricular ejection fraction, PCWP pulmonary capillary wedge pressure, PVR pulmonary vascular resistance, SVR systemic vascular resistance, MAP mean arterial pressure, RVSWI right ventricular stroke work index, BTT bridge to transplantation, DT destination therapy

6 26 J Artif Organs (2016) 19:21 28 Table 4 Multivariate regression analysis Parameter Odds ratio 95 % confidence interval P value Table 5 Area under the ROC curve for MELD and other predictors of RV failure AUC (95 % CI) p value RV failure MELD score Ischemic etiology Ventricular tachycardia BUN (mg/dl) Albumin (gm/dl) MAP (mmhg) RVAD implantation MELD score CVP (mmhg) BUN blood urea nitrogen, MAP mean arterial pressure, CVP central venous pressure Fig. 2 ROC curve for MELD score and other predictors of right ventricular failure preoperative intra-aortic balloon pump, end-organ dysfunction, and markers of RV dysfunction including elevated CVP, CVP/PCWP, pulmonary vascular resistance, or abnormal RV strain have been previously shown to predict RV failure [10 13]. We found that the MELD, an inexpensive and easy-touse tool, may be used to assist preoperative risk assessment for prediction of RV failure and its associated poor outcomes. MELD evidenced better discriminative capability than other commonly used hemodynamic markers for RV MELD ( ) PCWP ( ) MPAP ( ) PVR ( ) RVSWI ( ) TPG ( ) PCWP pulmonary capillary wedge pressure, MPAP mean pulmonary artery pressure, PVR pulmonary vascular resistance, RVSWI right ventricular stroke work index, TPG transpulmonary gradient failure, potentially because the renal and hepatic laboratory markers used to calculate MELD reflect congestion and poor perfusion associated with RV dysfunction rather than changes in RV afterload or pre-load reflected by hemodynamic measurements. The MELD is a composite score based on creatinine, total bilirubin, and INR. Of these, bilirubin and creatinine were elevated in the RV failure and RVAD groups. However, INR showed no trend associated with the severity of RV dysfunction. This suggests that the differences in MELD score are primarily driven by creatinine and, to a lesser extent, total bilirubin. However, the multivariate regression model indicates that the utility of MELD as a composite score based on total bilirubin, creatinine, and INR is of greater value in the prediction of RV failure than any of the individual components alone. It has been shown that the MELD without INR (MELD-XI) is predictive of overall survival in patients implanted with LVADs, but because anticoagulation is held prior to surgery we suggest that INR remains a pertinent marker for liver and right heart dysfunction [18]. As expected based on our classification of RV failure, RVAD placement was considered to be the worst outcome. Patients with postoperative RV failure tended to have worse postoperative laboratory values and hemodynamics. Patients with RVADs had significantly reduced survival compared to those with RV failure alone or no RV failure (Fig. 1). Those patients with no RV failure tended to live longer than either of the other two groups. The systemic effects of heart failure include end-organ congestion and subsequent hepatic and renal dysfunction. The MELD utilizes three non-cardiac markers for hepatic and renal function and has been shown to be predictive of adverse outcomes in separate populations with heart failure and mechanical circulatory support. This study indicates that the MELD may also be used as a pre-implantation prognostic indicator for risk of development of RV failure

7 J Artif Organs (2016) 19: Table 6 Postoperative outcomes comparison based on MELD score cutoff of 9 Outcomes MELD \ 9(n = 90) MELD C 9(n = 156) P value RVAD 1 (0.9 %) 17 (9.4 %) RV failure 18 (16.8 %) 71 (39.2 %) \0.001 Postoperative length of stay ± ± Discharge MAP (mmhg) ± ± Day MAP (mmhg) ± ± Discharge LVEF (%) ± ± Discharge LVEDD (mm) ± ± Discharge LVAD Flow (L/min) 5.22 ± ± Readmission for heart failure 15 (14.3 %) 42 (25.5 %) Readmission for VT 11 (10.5 %) 25 (15.2 %) Readmission for stroke 24 (22.9 %) 40 (24.2 %) Readmission for GI bleed 28 (26.7 %) 43 (26.1 %) RVAD right ventricular assist device, RV right ventricular, MAP mean arterial pressure, LVEF left ventricular ejection fraction, LVEDD left ventricular end diastolic diameter, VT ventricular tachycardia, GI gastrointestinal and for postoperative need for RVAD. Our analysis indicates that a MELD score of [9 may be an effective indicator for increased risk of RVAD implantation, RV failure, and longer-term complications. Though a MELD score of 9 is relatively low, its discriminative capability, especially for RVAD implantation, makes it a valuable reference. The increased incidence of readmission for heart failure symptoms in patients with MELD over 9 suggests that this preoperative diagnostic may be used to anticipate longterm adverse events tied to poor right ventricular function. Conclusions The MELD score, calculated rapidly from three common laboratory values, is an excellent marker for liver function in patients with heart failure. When measured preoperatively, the MELD score may be used to estimate the risk of postoperative RV failure and the necessity for RVAD implantation in the LVAD population. Limitations This study is subject to the limitations and inherent biases of a single center, retrospective study. The laboratory values used for the calculation of MELD were drawn immediately preoperatively and may have been affected by the hospital admission. Compliance with ethical standards Conflicts of interest Drs. Bhat and Tatooles are occasional consultants to Thoratec and HeartWare. References 1. Kirklin JK, Naftel DC, Pagani FD, Kormos RL, Stevenson LW, Blume ED, Miller MA, Baldwin JT, Young JB. Sixth INTER- MACS annual report: a 10,000-patient database. J Heart Lung Transplant. 2014;33: Aaronson KD, Patel H, Pagani FD. Patient selection for left ventricular assist device therapy. Ann Thorac Surg. 2003;75:S Slaughter MS, Pagani FD, Rogers JG, Miller LW, Sun B, Russell SD, Starling RC, Chen L, Boyle AJ, Chillcott S, Adamson RM, Blood MS, Camacho MT, Idrissi KA, Petty M, Sobieski M, Wright S, Myers TJ, Farrar DJ; HeartMate II Clinical Investigators. Clinical management of continuous-flow left ventricular assist devices in advanced heart failure. J Heart Lung Transplant. 2010;29(4 Suppl):S Patel ND, Weiss ES, Schaffer J, et al. Right heart dysfunction after left ventricular assist device implantation: a comparison of the pulsatile HeartMate I and axial-flow HeartMate II devices. Ann Thorac Surg. 2008;86: Lazar JF, Swartz MF, Schiralli MP, Schneider M, Pisula B, Hallinan W, Hicks GL, Massey HT. Survival after left ventricular assist device with and without temporary right ventricular support. Ann Thorac Surg. 2013;96: Patlolla B, Beygui R, Haddad F. Right-ventricular failure following left ventricle assist device implantation. Curr Opin Cardiol. 2013;28: Baumwol J, Macdonald PS, Keogh AM, Kotlyar E, Spratt P, Jansz P, Hayward CS. Right heart failure and failure to thrive after left ventricular assist device: clinical predictors and outcomes. J Heart Lung Transplant. 2011;30: Kormos RL, Teuteberg JJ, Pagani FD, Russel SD, John R, Miller LW, Massey T, Milano CA, Moazami N, Sundareswaran KS, Farrar DJ. Right ventricular failure in patients with the HeartMate II continuous-flow left ventricular assist device: incidence, risk factors, and effect on outcomes. J Thorac Cardiovasc Surg. 2012;139: Dang NC, Topkara VK, Michelle M, et al. Right heart failure after left ventricular assist device implantation in patients with chronic congestive heart failure. J Heart Lung Transplant. 2006;25:1 6.

8 28 J Artif Organs (2016) 19: Rich JD. Right ventricular failure in patients with left ventricular assist devices. Cardiol Clin. 2012;30: Ochiai Y, McCarthy PM, Smedira NG, et al. Predictors of severe right ventricular failure after implantable left ventricular assist device insertion: analysis of 245 patients. Circulation. 2002;106:I Matthews JC, Koelling TM, Pagani FD, Aaronson KD. The right ventricular failure risk score a preoperative tool for assessing the risk of right ventricular failure in left ventricular assist device candidates. J Am Coll Cardiol. 2008;51: Grant AD, Smedira NG, Starling RC, Marwick TH. Independent and incremental role of quantitative right ventricular evaluation for the prediction of right ventricular failure after left ventricular assist device implantation. J Am Coll Cardiol. 2012;60: Kamath PS, Wiesner RH, Malinchoc M, et al. A model to predict survival in patients with end-stage liver disease. Hepatology. 2001;33: Kim MS, Kato TS, Farr M, Wu C, Given RC, Collado E, Mancini DM, Schulze PC. Hepatic dysfunction in ambulatory patients with heart failure. J Am Coll Cardiol. 2013;61: Deo SV, Daly RC, Altarabsheh SE, Hasin T, Zhoa Y, Shah IK, Stulak JM, Boilson BA, Schirger JA, Joyce LD, Park SJ. Predictive value of the model for end-stage liver disease score in patients undergoing left ventricular assist device implantation. ASAIO J. 2013;59: Kato TS, Stevens GR, Jiang J, Schulze PC, Gukasyan N, Lippel M, Levin A, Homma S, Mancini DM, Farr M. Risk stratification of ambulatory patients with advanced heart failure undergoing evaluation for heart transplantation. J Heart Lung Transplant. 2013;32: Yang JA, Kato TS, Shulman BP, Takayama H, Farr M, Jorde UP, Mancini DM, Naka Y, Schulze PC. Liver dysfunction as a predictor of outcomes in patients with advanced heart failure requiring ventricular assist device support: use of the model of end-stage liver disease (MELD) and MELD excluding INR (MELD-XI) scoring system. J Heart Lung Transplant. 2012;31: Giallourakis CC, Rosenberg PM, Friedman LS. The liver in heart failure. Clin Liver Dis. 2002;6: (viii ix). 20. Myers RP, Cerini R, Sayegh R, et al. Cardiac hepatopathy: clinical, hemodynamic, and histologic characteristics and correlations. Hepatology. 2003;37: Malinchoc M, Kamath PS, Gordon FD, Peine CJ, Rank K, ter Borg PC. A model to predict poor survival in patients undergoing transjugular intrahepatic portosystemic shunts. Hepatology. 2000;31: Morgan JA, John R, Lee BJ, et al. Is severe right ventricular failure in left ventricular assist device recipients a risk factor for unsuccessful bridging to transplant and post-transplant mortality. Ann Thorac Surg. 2004;77: