Objectives: Each year in the United States, approximately 40 000 patients with a liver disorder will progress to end-stage liver disease and about 30 000 of those patients will subsequently die from this condition. Liver transplant remains the definitive treatment option for end-stage liver disease, and understanding the causes of posttransplant mortality is an ongoing area of investigation.
Materials and Methods: In this retrospective cohort study, patients who underwent orthotopic liver transplant between January 2012 and January 2015 at the Johns Hopkins Hospital Liver Transplant Program were reviewed by a single reviewer for cardiac events in the 30 days after transplant or during the index admission.
Results: Of the 145 patients included, 30 (20.6%) were identified as having experienced a cardiac event during the defined postoperative period. Overall 1-year mortality for the cohort of 145 patients was 11.7%; however, 1-year mortality in those who had a cardiac event was 36.7% compared with 5.2% in the noncardiac event group (odds ratio = 18.17; P < .001). Although there was a statistically significant difference in age between the groups (58.6 vs 52.3 years old), once accounted for in multivariate analysis, a posttransplant cardiac event was still a statistically significant variable in 1-year mortality (odds ratio = 89.16; 95% confidence interval, 2.71-2933.95; P = .012). Similarly, hepatocellular carcinoma, sex, age, and presence of diabetes had little effect on 1-year mortality when we compared those patients who experienced a cardiac event in the first 30 days versus those who did not (odds ratio = 100.82; 95% confidence interval, 2.15-4726.12; P = .019).
Conclusions: Recipients who experience cardiac events within 30 days after transplant have increased 1-year posttransplant mortality. This highlights the importance of cardiac risk stratification before transplant.
Key words : Candidate, Coronary artery calcium score, Coronary artery disease, End-stage liver disease, Liver transplantation
Each year in the United States, approximately 40 000 patients with a liver disorder will progress to end-stage liver disease and about 30 000 of those patients will subsequently die from this condition.1 During the past 25 years, liver transplant has become the definitive treatment for most patients with cirrhosis2; in 2015 alone, 7127 adult liver transplants were performed in the United States.3 Significant research has been undertaken to evaluate pretransplant risk assessments that could lead to numerous preoperative metrics being validated, including those involving cardiac risk stratification.4 This has typically involved echocardiography and noninvasive stress testing with the assumption that physiologic stress during the perioperative period may potentially “unmask” the underlying cardiac disease that may adversely affect transplant outcomes and therefore increase recipient mortality.5
Despite interest in predicting perioperative risk, there have been limited publications regarding the effects of postoperative cardiac events on the mortality of patients who have undergone liver transplant. These few studies have shown that 25% to 70% of liver transplant recipients have a cardiac complication within 6 months of transplant and that major cardiac events in the perioperative period were associated with a lower 5-year survival rate. Given the striking nature of these findings, if the goal of preoperative risk stratification is to be more fully implemented, a better understanding of cardiac complications in the posttransplant group is necessary.6,7
Materials and Methods
In this retrospective cohort study, adult patients with liver cirrhosis who underwent deceased-donor orthotopic liver transplant (OLT) with a minimum of 1-year follow-up were included. After Institutional Review Board approval, the discharge summaries and progress notes of patients who underwent OLT between January 1, 2012 and January 1, 2015 at the Johns Hopkins Hospital Liver Transplant Program were reviewed by a single reviewer for cardiac events in the 30-day postoperative period.
The pretransplant cardiac work-up protocol requires candidates over 40 years of age to be screened for underlying coronary artery disease in addition to undergoing a resting transthoracic 2-dimensional echocardiogram. Usually, dobutamine stress echocardiography is used as the screening tool. Single-photon imaging computed tomography was also used at the discretion of the ordering hepatologist in some cases. Candidates with additional cardiac risk factors (ie, hypertension, diabetes, smoking, dyslipidemia, and diagnosis of nonalcoholic steatohepatitis) underwent further testing with cardiac computed tomography scans to calculate the coronary artery calcium (CAC) score. Patients with abnormal dobutamine stress echocardiography and/or CAC score > 100 Agatston units were evaluated for appropriateness for cardiac catheterization. A cutoff CAC score of 100 was established based on the American College of Cardiology (ACC) Clinical Expert Consensus Task Force guidelines.8
Cardiac events were categorized as new arrhythmia (as documented by 12-lead electrocardiogram [ECG]), acute coronary syndrome (as defined by clinical assessment in conjunction with evaluation of ECG changes), and cardiac biomarker elevations according to ACC guidelines. An ST-elevation myocardial infarction was dependent on documented ECG findings,9 type 2 myocardial infarction (defined by the ACC as when myocardial oxygen supply does not equal myocardial oxygen demand9), new heart failure (defined by echocardiographic evidence of reduced left ventricular ejection fraction and/or diastolic dysfunction and/or increase in serum B-type natriuretic peptide), new valvular disease (defined on echocardiography), cardiac arrest, or other cardiac disease. Data were also collected on demographics, calcium score, if available, the presence or absence of diabetes, and mortality.
Mortality data for patients with a cardiac event within the first 30 days after surgery were compared with mortality data for patients without a cardiac event in the perioperative period using multivariate analysis. Subgroup analysis was performed using a Fisher exact test between the exposed subgroup and the group of patients with no perioperative cardiac events.
Our study included 145 transplant patients who had at least 1 year of follow-up. In the cohort, there were 86 male patients (59.3%). Mean age of those who experienced a cardiac event within the first 30 days posttransplant was 58.6 years, whereas the group without cardiac event had a mean age of 52.3 years. Furthermore, mean biological Model for End-Stage Liver Disease at time of transplant was 21.4 in the cohort without cardiac events compared with 22 in the cohort with cardiac events (Table 1). The cause of cirrhosis was hepatitis C virus (HCV) infection in 68 patients, alcohol-related disease in 28 patients, nonalcoholic steatohepatitis in 10 patients, autoimmune hepatitis in 5 patients, primary sclerosing cholangitis in 5 patients, primary biliary cirrhosis in 3 patients, and other causes of cirrhosis in 26 patients (Figure 1).
Of the 145 total patients, 30 (20.6%) were identified as having experienced a cardiac event during the defined postoperative period. Among these 30 patients, there were 33 documented events. Seven experienced a new arrhythmia, of which, supraventricular tachycardia was the most common arrhythmia occurring in 3 patients. This was followed by atrial fibrillation in 2 patients and stable ventricular tachycardia in 2 patients. Five experienced an acute coronary syndrome, and 1 of these was an ST-elevation myocardial infarction with the rest being non-ST- elevation myocardial infarctions. In addition, 16 patients had demand ischemia and 1 experienced myopericarditis (Figure 2). There were no cases of new heart failure or new valvular disease. Four patients had cardiac arrest with pulseless electrical activity. Of these, return of spontaneous circulation was achieved in 2 patients, and 1 was alive at 1 year.
Overall 1-year mortality for the cohort of 145 patients was 11.7%. However, 1-year mortality in the segment who experienced a cardiac event was 36.7% compared with 5.2% in the noncardiac event group (Figure 3) (odds ratio [OR] = 18.17; 95% confidence interval [CI], 7.96-41.44; P < .001). Because the average age of patients who experienced a cardiac event posttransplant was significantly higher than those who did not (Table 1), we performed multivariate analyses to account for this. When we accounted for age, there was still a significant difference in 1-year mortality between those who experienced a cardiac event and those who did not (OR = 89.16; 95% CI, 2.71-2933.95; P = .012). This is visually represented by age-matched brackets in Table 2. Similarly, presence of hepatocellular carcinoma, sex, age, and presence of diabetes had little effect on 1-year mortality when we compared those patients who experienced a cardiac event in the first 30 days versus those who did not (OR = 100.82; 95% CI, 2.15-4726.12; P = .019).
Subgroup analysis demonstrated a 1-year mortality of 42.9% among patients who experienced new arrhythmias (OR = 13.6; 95% CI, 2.470-75.164; P = .0027 vs patients with no cardiac events in the first 30 days) and 1-year mortality of 28.5% among patients who experienced demand ischemia (OR = 7.3; 95% CI, 1.754-30.102; P = .0063 vs patients with no cardiac events in the first 30 days).
In terms of preoperative assessment, among those who had a cardiac event in the first 30 days after transplant, 19 of 30 patients (63%) had undergone coronary computed tomography for calcium score calculation, and the average calcium score was 326.33 ± 394.53. Among those who did not have a cardiac event within the first 30 days after transplant, 42 of 115 patients (36.5%) had a calcium score calculated, with an average calcium score of 770.31 ± 2449.73 (Table 1).
In this retrospective single-center cohort study, among patients who have undergone OLT, cardiac events within the 30-day postoperative period were associated with decreased 1-year survival. Although patients with and without cardiac events had statistically significant differences in age (P < 0.005), this had no influence on mortality outside of a cardiac event. Similarly, differences in 1-year survival could not be explained by other risk factors such as diabetes, sex, or presence of hepatocellular carcinoma. Patients who experienced new arrhythmias appeared to be at particularly increased risk of death within the first year after transplant. As such, there is obvious interest in identifying patients at risk of developing such postoperative complications. Although the current screening modalities do attempt to identify patients at elevated risk for cardiac events, this study serves to illustrate the on-going need for new methods to detect occult obstructive coronary artery disease. Novel tools such as CAC scoring may represent a promising new avenue for risk stratification but will require further validation before being added to the existing screening process.
Type 2 diabetes mellitus is also a risk factor for cardiac events. In a recent study, Li and associates revealed that new-onset diabetes mellitus at 1 year after liver transplant was significantly higher in HCV-positive liver transplant recipients than in HCV-negative transplant recipients, which may contribute to cardiac events after liver transplant.10 Although specific mention of cardiac events was not made, overall 5-year mortality was improved in posttransplant patients with new-onset diabetes mellitus. Our retrospective study did not identify any specific influence of diabetes on postoperative mortality. Diabetes was present at similar rates in groups with and without cardiac event in the first 30 days posttransplant, but its presence did not influence 1-year mortality posttransplant.
In the present study, we observed that posttransplant cardiac events most commonly occurred among patients with HCV infection. In a previous study that included 32 810 patients, 11% had major cardiac events 90 days after liver transplant, with most occurring in patients with nonalcoholic steatohepatitis and alcohol cirrhosis.11 Multiple studies have suggested a relationship between HCV infection and cardiac disease. Hepatitis C virus infection was found to be related with reversible myocardial dysfunction, which improved after sustained virologic response after HCV treatment.12-14 Although cardiac events were similar in patients with alcoholic and nonalcoholic cirrhosis, concomitant HCV positivity and male sex corresponded to an increased risk of cardiac events (9- and 3-fold, respectively).15 This is similar to our findings in which cardiac events and 1-year mortality rates were also higher in this group. Cirrhosis secondary to HCV was shown in 42.9% of the 7 patients who experienced arrhythmia and 75% of the 4 patients who experienced acute coronary syndrome. Although our sample sizes were small, these findings corroborate previously published results; however, the mechanism behind HCV infection and cardiac events has yet to be elucidated.
Although there is no consensus on preoperative cardiac evaluation of liver transplant candidates, recent studies have suggested cardiac computed tomography and calcium scoring can be used for further risk stratification.16-18 There are current screening modalities that attempt to identify patients at elevated risk for cardiac events; however, our study serves to illustrate the on-going need for preoperative methods to detect occult obstructive coronary artery disease. Interestingly, in our cohort, there was not a single patient with a positive dobutamine stress or nuclear stress imaging who subsequently underwent transplant. In total, 60 patients underwent CAC scoring as part of preoperative risk assessment; however, given the wide range of scores, there was no statistical difference between the groups. We note the retrospective nature of our study and its inherent limitations; as such, potential candidates deemed too high risk for transplant were not included.
In conclusion, cardiac events occurring within 30 days after transplant were shown to be correlated with increased 1-year posttransplant mortality. Pretransplant cardiac evaluation, risk stratification, and intervention can potentially reduce the 1-year mortality after transplant. In particular, noninvasive imaging modalities, such as CAC scoring, may allow easier, higher-yield identification of potential at-risk patient populations compared with the current standard of functional stress imaging. Multicenter, prospective studies are warranted to determine the optimal preoperative evaluation and risk-stratification algorithms for liver transplant candidates.
DOI : 10.6002/ect.2017.0276
From the 1Osler Internal Medicine Residency Program, the 2Division of
Cardiology, the 3Division of Liver Transplantation, Department of Anesthesiology
and Critical Care, the 4Transplant Hepatology, Division of Gastroenterology and
Hepatology, and the 5Division of Liver Transplantation, Department of Surgery,
Johns Hopkins University School of Medicine, Baltimore, Maryland, USA
Acknowledgements: The authors have no conflicts of interest to disclose. Statistical analysis was funded by the National Center for Research Resources and the National Center for Advancing Translational Sciences through National Institutes of Health Grant 1UL1TR001079. We acknowledge Carol Thompson and Sarah Salter who performed statistical analysis and consult for the project. *Theodore W. James and Jessica E. S. Shay contributed equally to this work.
Corresponding author: Ahmet Gurakar, Johns Hopkins University School of Medicine, Transplant Hepatology, Division of Gastroenterology and Hepatology, 720 Rutland Avenue, Ross Research Building #918, Baltimore, Maryland, USA
Phone: +1 410 614 3369
Table 1. Demographics and Causes of Disease in Patient Cohorts
Table 2. One-Year Mortality as Influenced by Cardiac Events Posttransplant and Age
Figure 1. Transplant Characteristics
Figure 2. Cardiac Event Type
Figure 3. One-Year Survival Posttransplant