Objectives: Congestive hepatopathy as a result of advanced heart failure correlates with poor outcomes. Thus, risk-scoring systems have been established to assess the risks for cardiac surgery and heart transplant, although these systems were originally designed to measure mortality risk in patients with end-stage liver disease. We compared the scores for the Child-Pugh criteria and standard and modified Models for End-Stage Liver Disease to evaluate the effect of preoperative liver dysfunction on postoperative outcomes in patients with heart failure who underwent heart transplant.
Materials and Methods: Data of 60 consecutive patients who underwent orthotopic heart transplant were analyzed from a historical cohort study from January 1, 2015, to December 31, 2018. We calculated the scores for Child-Pugh criteria and the standard and modified Models for End-Stage Liver Disease.
Results: Of the 60 total patients, 48 were male patients, with a median age of 43 years (range, 13-69 years). Twenty patients died before the end of the study. The causes of death were cardiac, liver, and renal diseases. The mortality risk increased 25% (interquartile range, 0.05-0.51) for the patients with 1 point higher score compared with the patients with 1 point lower score based on a modified Model for End-Stage Liver Disease (P = .01).
Conclusions: Preoperative liver dysfunction has a significant effect on patient survival. The modified Model for End-Stage Liver Disease scoring system could be an effective predictor of perioperative risk stratification for patients with congestive hepatopathy who are undergoing cardiac transplant.
Key words : Congestive hepatopathy, Heart failure, Liver dysfunction, Risk factor prediction
Heart failure (HF) is a complicated clinical syndrome resulting from impairment of ventricular filling or blood ejection. Despite dramatic innovations in medical and device treatments in recent decades, the incidence of HF is increasing.1
The gold standard treatment for patients with refractory end-stage HF is heart transplant (HTx), a procedure with substantial functional limitations and a high mortality rate.2
Despite stepwise improvements in early transplant survival, the long-term survival rate remains unchanged. The survival rate for an entire patient cohort of the worldwide registry showed that, after a harsh fall in survival during the first 6 months, survival decreases at a linear rate of 3.4% per year.3
Because of the current shortage of donors, HTx procedures are limited and the number of recipients on wait lists is expected to increase over time.4 Because of organ shortages and the clinical complications after HTx, appropriate donor and recipient selection is essential for a successful outcome.5 Comorbidity reduces survival and hinders surgical treatment. Therefore, clinical assessment of the transplant recipient should include indications and contraindications of HTx.2
Congestive hepatopathy is a well-known consequence of terminal HF that correlates with poor outcomes. The prevalence of hepatic dysfunction in HF is 15% to 65%.6 Liver injury due to prolonged recurrent congestion, also known as a backward failure and/or impaired arterial perfusion (forward failure), is common in patients evaluated for HTx.7,8
The hallmark of chronic congestive hepatopathy is presence of cholestatic changes, with high serum bilirubin and alkaline phosphatase concentrations. Early stages are reversible, but long-term congestive hepatopathy leads to irreversible damage to the liver tissue and cirrhosis with associated transaminases. Hepatic function has been shown to benefit from HTx.8
Abnormalities in liver function have been linked to increased short-term and long-term morbidity and mortality in patients undergoing surgeries. Some risk-scoring systems have been established to assess risks and performance measures for cardiac surgery and HTx, although these systems were originally designed to measure mortality risk in patients with end-stage liver disease.5,9
The Model for End-Stage Liver Disease (MELD) scoring system is a marker of multisystem dysfunction and coagulopathy and also is currently applied to determine the prognosis of patients with chronic HF who are referred for mechanical circulatory support or HTx. A higher MELD score is associated with poor survival and a higher risk of clinical complications.
The MELD score may not be a valid prognostic index in patients who are undergoing anticoagulant therapy, because these drugs affect the international normalized ratio (INR). There are some alternative MELD scoring systems that do not include INR as a factor. In 2007, Heuman and colleagues created a modified MELD scoring system that replaced INR with a factor for albumin levels.10 This modified MELD system (which excluded INR, hence, MELD-XI) better accounts for lack of interaction with oral anticoagulants than does the standard MELD score. Elevation in MELD-XI score before HTx has been associated with higher morbidity and mortality after the operation.5,11
The Child-Turcotte-Pugh system (now known simply as the Child-Pugh system) is based on criteria originally conceived by Child and Turcotte in 1964 to predict mortality in patients with cirrhosis; this system was later modified by Pugh and colleagues, who substituted prothrombin time for clinical nutrition status. The Child-Pugh score can facilitate prediction of all-cause mortality risk and the risk of other complications of liver dysfunction.12-14
We compared the scores for Child-Pugh, standard MELD, and MELD-XI to evaluate the effect of preoperative liver function on early and 1-year postoperative outcomes in patients with HF who underwent HTx.
Materials and Methods
This historical cohort study included patients with end-stage HF who underwent HTx from January 1, 2015, to December 31, 2018, with baseline laboratory data available from a tertiary referral center. Clinical and laboratory data were obtained from a computerized database and patient files. Patients with elevations in hepatobiliary markers secondary to known viral hepatitis and drug-related hepatitis were excluded. All patients were followed up for 3 years. This study was approved by the Iran National Committee for Ethics in Biomedical Research and followed the ethical guidelines outlined in the 1975 Helsinki Declaration. Informed consent was obtained for all study participants before inclusion in the study.
The inclusion criteria for this study were (1) patients with end-stage HF who underwent HTx and (2) patients with congestive hepatopathy as a consequence of terminal HF. Exclusion criteria were (1) patients with elevations in hepatobiliary markers secondary to known viral hepatitis and drug-related hepatitis and (2) patients with posttransplant drug protocol deviation and poor compliance.
Preoperative data were collected for the most recent laboratory analyses before HTx. Postoperative data were collected from electronic medical records at 1 week and at 1, 3, 6, and 12 months after the original transplant date. A data collection form was designed, and data were recorded, including age, sex, etiology of HF, start date of disease, pretransplant medications, comorbidities, and history of liver dysfunction. Laboratory test results were also recorded, including albumin, protein, aspartate aminotransferase (AST), alanine aminotransferase (ALT), alkaline phosphatase, total bilirubin and direct bilirubin, creatinine, urea, prothrombin time, and INR, as well as signs including encephalopathy and ascites.
Laboratory parameters and hepatopathy
Cutoff values for laboratory tests were used according to international guidelines with consideration of the age and sex of each patient. Pathological or abnormal values were defined as those that fell outside the normal ranges. The upper limits normally used for AST, ALT, and total bilirubin were 35, 45, and 1.3 mg/dL, respectively. The lower limits normally used for total protein and albumin were 6.7 and 4.1 g/dL, respectively. The upper limit normally used for alkaline phosphatase was 130 mg/dL.15,16
Definition of scoring systems for Child-Pugh Criteria and standard and modified Models for End-Stage Liver Disease
The standard MELD score was determined according to the United Nation for Organ Sharing modification, which uses the logarithmic formula, 3.78 × ln (total bilirubin) + 11.2 × ln (INR) + 9.57 × ln (creatinine) + 6.43, where total bilirubin and creatinine are equal to 1 if the laboratory value is <1.17 Calculation of the MELD-XI score followed the same formula, but with albumin in place of INR.5 The calculation of the Child-Pugh score was based on 5 factors: encephalopathy, ascites, bilirubin, albumin, and prothrombin time or INR. Child-Pugh severity of cirrhosis is categorized into 3 classes: A, 5 to 6 points; B, 7 to 9 points; and C, 10 to 15 points. The Child-Pugh score has been validated as a predictor of postoperative mortality after portocaval shunt surgery and other major operations,18 and patients designated as Child-Pugh class C have a high mortality rate.12 The scores for standard MELD, MELD-XI, and Child-Pugh were all calculated from laboratory measurements.
Continuous variables are presented as mean values ± SD. Categorical variables are reported as number of patients (%). Analysis of variance was used to compare the continuous variables between groups. The survival rates were obtained by the Kaplan-Meier method. The predictability of the criteria on the mortality risk was assessed with the Cox proportional hazard model. Receiver operating characteristic (ROC) analysis was used to detect an acceptable cutoff point for the 3 scoring systems by evaluation of specificity and sensitivity. Data were analyzed with the R software package (version 3.6.0). The significance level was set to P < .05.
Our study included 60 patients with end-stage HF who underwent HTx, with 48 (80%) male and median age of 43 years (range, 13-69 years). Eight (13%) had type 2 diabetes, and 15 (25%) had hypertension. Two patients received pretransplant support: 1 underwent intra-aortic balloon pump therapy and 1 received life support by an extracorporeal membrane oxygenation machine. Nineteen patients had early mortality (1 month after transplant), and 1 died after 3 months. The causes of death were as follows: 14 cardiac (70%), 5 liver (25%), and 1 renal disease (5%), primary graft failure (graft rejection and right-side HF), and bleeding. The median follow-up time was 844 days (interquartile range, 742-1021), 9 days (interquartile range, 4-17), and 3 days (interquartile range, 1-52) for survival, death due to liver dysfunction, and death due to other causes, respectively.
Laboratory values were followed from pretransplant to 3 years after surgery. Laboratory values generally improved after transplant. Transaminases, albumin, total protein, and INR improved on average (Figure 1). Liver enzyme assessment showed that, in 30% of patients with pretransplant elevated AST (more than 38 mg/dL), there was a 41% improvement in 1 week, 11% in 1 month, 8% in 3 months, 3% in 6 months, and 3% in 12 months after transplant. Changes in the patterns of other liver parameters are shown in Figure 1 (P < .05). There was a peak increase in most of the values around the first week followed by gradual improvement. Cholestatic values were worse compared with hepatocellular values, and the increases in alkaline phosphatase and bilirubin were more apparent (Figure 1).
Table 1 shows that the mean scores for standard MELD, MELD-XI, and Child-Pugh were not significantly different between outcomes of survival and death due to liver dysfunction versus other causes at baseline (pretransplant) (P > .05). The ROC curve analysis revealed that the optimal cutoff scores for standard MELD, MELD-XI, and Child-Pugh were 11.3, 22.22, and 5.5, respectively, based on 3-year survival rate (sensitivity of 85%, 85%, and 75%; specificity of 72%, 85%, and 60%, respectively) for all-cause mortality (Table 2). Kaplan-Meier estimation showed the 3-year survival rate for the patients was 66% (Figure 2). The MELD-XI score was a significant predictor for patient mortality; patients with 1 point higher MELD-XI score had a higher mortality risk of 1.25 (95% CI, 1.05-1.51) compared with the patients with 1 point lower (P = .01; Table 3).
The present study compared 3 risk-prediction systems (Child-Pugh, standard MELD, and MELD-XI) for cardiac hepatopathy assessment in patients with HF who underwent HTx. During periods of organ shortages, it is essential to define objective methods that will best select patients at high risk for HTx. Although the scoring systems were originally used to measure mortality risk in patients with end-stage liver disease, the use of these systems has been expanded to predict mortality risk in cirrhotic patients who undergo orthopedic, digestive, and cardiovascular surgery. Recently, these scoring systems have been applied to patients with cardiac hepatopathy who are candidates for HTx.19 The study demonstrated a significant improvement in elevated hepatobiliary serum markers within 12 months after HTx, which decreased gradually after surgery. Our results also indicated that the MELD-XI score was a significant predictor for patient mortality, and the standard MELD score and Child-Pugh score were not significantly associated with the prediction of mortality.
According to data from the International Society for Heart and Lung Transplantation, the 30-day survival rate after HTx has continually improved from 84% during the 1979-1985 era to 91% for the 1996-2001 era.20 The most recent data from the registry of the International Society for Heart and Lung Transplantation shows the current 1-year and 5-year survival rates are 84.5% and 72.5%, respectively.21 This improvement has likely been achieved by better donor-recipient selection, meticulous surgical technique, improved postoperative care, and appropriate immunosuppressive drug regimens.20
One multinational prospective study on patients with HF across 11 Asian regions has shown that patients from Southeast Asia (particularly countries with low incomes and with the youngest patients) had the worst outcomes regardless of the left ventricular ejection fraction. Region?specific risk factors, gaps in guideline?directed therapy, potential regional differences in health care systems, limited access to device therapy, and unfavorable genetic and environmental factors should be addressed to better understand these poor outcomes.22
In a study conducted in Iran, the 1-month and 1-year survival rates after HTx were 82.6% and 70%, respectively.23 In our study, we observed a lower 30-day survival rate because of shortages of above-mentioned items and, moreover, because of donor shortage and a long wait list. Most of our recipients were classified as Interagency Registry for Mechanically Assisted Circulatory Support (INTERMACS) profile 3, and 3 of our recipients were INTERMACS profile 1, so some degree of multiorgan failure in these patients was common, with consequential effects on postoperative outcomes.
Until now, a large number of studies were limited to the discriminative abilities of these scoring systems, and the results of these studies have remained controversial. Some studies favored a particular system, but others did not report distinctive differences among the systems.24 A recent integrative review confirmed the predictive power of the MELD score and derivatives and demonstrated its relevance to cardiology. The use of these predictive scoring systems may contribute to the early evaluation of the severity and priority of the patients with HF who are candidates for HTx, in a manner similar to the traditional use of these systems in the context of liver transplant.9 Our present study found the MELD-XI score was a significant predictor of survival after transplant (ie, a higher MELD-XI score predicted a higher risk of mortality), whereas the scores for standard MELD and Child-Pugh were not significant predictors of risk. However, mean scores of MELD, MELD-XI, and Child-Pugh were not significantly different for outcomes of survival and death due to liver dysfunction versus baseline.
Various predictive models have been used to assess prognosis in patients with cardiac hepatopathy, but few studies have assessed the ability of all 3 scores (Child-Pugh, standard MELD, and MELD-XI) to predict survival. In our study, we recruited 60 patients with end-stage HF who underwent HTx, and we calculated their scores for Child-Pugh, standard MELD, and MELD-XI. As such, all 3 prognostic scores were of no value for assessment of hepatopathy in patients with HF.
In a study from Korea that was in agreement with our study, Kim and colleagues25 assessed 49 consecutive patients with liver cirrhosis among 170 patients who underwent HTx and compared the risk prediction of the standard MELD score, the MELD-XI score, and the Child-Pugh score. The optimal cutoff scores for standard MELD, MELD-XI, and Child-Pugh in ROC curve analysis were 12.2, 12.0, and 7.5, respectively (sensitivity of 69.2%, 61.5%, and 69.2%; specificity of 68.6%, 60.0%, and 62.9%, respectively) for all-cause mortality. The standard MELD score and the MELD-XI score were significantly different for survival versus death outcomes. Cox regression analysis showed that high scores for standard MELD, MELD-XI, and Child-Pugh were associated with higher risk of all-cause mortality.25 In contrast, we found that the MELD-XI score was a significant predictor for patient mortality; patients with 1 point higher in MELD-XI score had a higher mortality risk of 0.25 (95% CI, 0.05-0.51), and scores for standard MELD and Child-Pugh were not significantly associated with the prediction of mortality.
In a 2012 study by Chokshi and colleagues, the 1-year survival rate was 85.5% in 617 adult patients with a standard MELD score >20 and 75% in patients with a MELD-XI score >20. According to that study, both the standard MELD score and the MELD-XI score provide valuable and efficient prognostic tools for evaluation of candidates for HTx.5
Suman and colleagues in 2004 showed a signi?cant association of MELD score with hepatic decompensation in HTx patients and mortality. The best cutoff value of this score for prediction of mortality and hepatic decompensation was >13 for the standard MELD score, and it had high speci?city (89%) and low sensitivity (71%).26 In our present study, although the standard MELD score correlated with the MELD-XI score, the standard MELD score had a lower sensitivity (85%) and specificity (72%) and poor power to predict mortality after HTx; these limitations may have been the result of the low rate of hepatic failure among patients in our study versus the rates of hepatic failure that have been reported in other studies.4 Our study demonstrated that the MELD-XI was the preferable scoring system to predict mortality after HT in patients with severe cardiac hepatopathy who are under treatment with anticoagulant drugs.
Yalcin and colleagues27 conducted a study in 2020 about the effects of preoperative liver dysfunction on outcomes in patients with left ventricular assist devices and reported that the 1-year mortality rate optimal cutoff value for the standard MELD score was 12.6.
Some recent studies are in agreement with results from our study that showed hepatobiliary serum markers improve and decrease gradually. Dichtl and colleagues7 showed that HTx significantly restores liver function within 3 months in patients with end stage HF who had preoperative chronic cardiac hepatopathy. They also confirmed recently published data from Loforte and colleagues that suggested elevated cholestatic parameters (?-glutamyl transferase, alkaline phosphatase, and bilirubin) rather than enzymes indicative of hepatocellular injury (AST and ALT) are the hallmark of chronic cardiac hepatopathy.4 In a study by Yang and colleagues, there was evidence that cholestasis was exacerbated during early ventricular assist device support.8 Also, a review study by Ford and colleagues reported that elevated levels of alkaline phosphatase and ?-glutamyl transferase reflect a progressive increase in HF class and are significant predictors of all-cause mortality in patients with HF.28
In some studies, there was a peak in the results of liver function tests that decreased gradually over time, which is consistent with our results. For example, the study by Ford and colleagues showed that transaminase levels usually peak at 1 to 3 days after the hemodynamic insult, decrease by 50% within 72 hours, and often normalize within 7 to 10 days.28 Chou and colleagues clearly indicated that postoperative hypoxic hepatitis can be an outcome of open heart surgery for patients with liver cirrhosis, especially in patients with high MELD scores.29
The present study was limited by its small number of cases as a single-center, retrospective study. Larger, multicenter prospective studies are needed to further confirm the reliability and clinical utility of the scoring systems discussed here. Also, the diagnosis of cardiac hepatopathy was not confirmed by liver biopsy. A liver biopsy may be performed in HTx candidates with HF in an attempt to assess the degree of liver injury, and the findings of severe fibrosis or frank cirrhosis on liver biopsy may be used to exclude the candidate from undergoing HTx.30 However, because of its association with a high rate of complications, liver biopsy may be indicated only in patients with severely compromised hepatic function.7
In this single-center historical cohort study, we demonstrated the dynamics of liver dysfunction after HTx, and we also showed that an elevated MELD-XI score that indicates impaired liver function is associated with poor clinical outcomes after HTx. Preoperative liver dysfunction has a significant effect on patient survival after HTx, and these scoring systems can be a tool for risk stratification for patients with congestive hepatopathy who are undergoing HTx. Therefore, early recognition of hepatic impairment may warrant more intensive treatment of HF to define the parameters of liver dysfunction and potentially correct the hepatic dysfunction before transplant. Also, we advise HTx centers to consider the MELD-XI score, because this particular score can be an effective predictor of mortality in HTx patients.
Volume : 19
Issue : 9
Pages : 963 - 969
DOI : 10.6002/ect.2020.0559
From the 1Lung Transplantation Research Center and the 2Tobacco Prevention and Control Research Center, National Research Institute of Tuberculosis and Lung Diseases, Shahid Beheshti University of Medical Sciences; and the 3Department of Biostatistics, School of Public Health, Iran University of Medical Sciences, Tehran, Iran
Acknowledgements: The authors have not received any funding or grants in support of the presented research or for the preparation of this work and have no declarations of potential conflicts of interest.
Corresponding author: Shadi Shafaghi, Lung Transplantation Research Center, National Research Institute of Tuberculosis and Lung Diseases, Shahid Beheshti University of Medical Sciences, Masih Daneshvari Hospital, Darabad Avenue, Shahid Bahonar roundabout, Tehran, Iran
Phone: +98 91 2590 1135
Table 1. Baseline (Pretransplant) Scores for Child-Pugh Criteria and Standard and Modified Models for End-Stage Liver Disease
Table 2. Cutoff Criteria for Survival Outcomes (Sensitivity 0.58-0.9 and Specificity 0.6-0.86)
Figure 1. Trend of Liver Parameters During 12 Months After Transplant
Figure 2. Kaplan-Meier Estimation of 3-Year Survival Rate
Table 3. Predictability of Each Criterion for Mortality With the Cox Proportional Hazard Model)