|Year : 2019 | Volume
| Issue : 4 | Page : 508-513
Predictors of myocardial injury in patients with cirrhosis presenting with upper gastrointestinal bleeding
Rehab Shaheen1, Tarek Gouda1, Monir Bahgat2, Tarek Selim3, Hazem Hakim El-Beltagy El-Menshawy2, Elsayed Awad4
1 Department of Internal Medicine, Critical Care, Faculty of Medicine, Mansoura University, Mansoura, Egypt
2 Department of Internal Medicine, Hepatology and Gastroenterology, Faculty of Medicine, Mansoura University, Mansoura, Egypt
3 Department of Clinical Pathology, Faculty of Medicine, Mansoura University, Mansoura, Egypt
4 Department of Anaesthesia and ICU, Faculty of Medicine, Al-Azhar University, Cairo, Egypt
|Date of Submission||02-Sep-2018|
|Date of Acceptance||18-Oct-2019|
|Date of Web Publication||18-Aug-2020|
Hazem Hakim El-Beltagy El-Menshawy
Specialized Medical Hospital, Mansoura University, Mansoura
Source of Support: None, Conflict of Interest: None
Background Myocardial injury in conditions other than coronary artery disease (CAD), known as type 2 myocardial infarction, is mostly related to mismatch between myocardial oxygen supply and demand. Cirrhotic patients with acute upper gastrointestinal bleeding (UGIB) are usually hemodynamically unstable. Hypovolemia, hypotension, and decreased oxygen-carrying capacity as consequences of UGIB may precipitate subclinical heart failure and myocardial injury.
Aim of work Assessment of the prevalence and potential risk factors of myocardial injury in patients with liver cirrhosis with acute UGIB.
Patients and methods The study was conducted on 132 patients diagnosed with liver cirrhosis presenting by UGIT bleeding at Mansoura University Hospitals during one year. Patients were divided into 2 groups: group 1 (76 patients) with myocardial injury or ischemic heart disease and group 2 (60 patients) without.
Results The incidence of myocardial injury in this study (elevated troponin levels above cutoff value and/or ECG changes) was 55% of patients. Troponin I was positive in 25% of patients. ECG ischemic changes were found in 36.3% of patients in the form of ST-segment deviation or T-wave inversion. On univariate analysis, predictors of myocardial injury in patients with UGIB included MELD score and variceal source of GI bleeding. On multivariate analysis variceal source of GI bleeding is an independent predictor of myocardial injury. Variceal bleeding was found in 95 % of the ischemic group versus 63% in the other group.
Conclusion More than half of the study patients presented with UGIB have suffered from unnoticed subclinical myocardial injury. Variceal source of GI bleeding was found to be an independent predictor of myocardial injury.
Keywords: GI bleeding, liver cirrhosis, myocardial injury
|How to cite this article:|
Shaheen R, Gouda T, Bahgat M, Selim T, El-Menshawy HE, Awad E. Predictors of myocardial injury in patients with cirrhosis presenting with upper gastrointestinal bleeding. Egypt J Intern Med 2019;31:508-13
|How to cite this URL:|
Shaheen R, Gouda T, Bahgat M, Selim T, El-Menshawy HE, Awad E. Predictors of myocardial injury in patients with cirrhosis presenting with upper gastrointestinal bleeding. Egypt J Intern Med [serial online] 2019 [cited 2020 Sep 20];31:508-13. Available from: http://www.esim.eg.net/text.asp?2019/31/4/508/292249
| Introduction|| |
Acute upper gastrointestinal bleeding (UGIB) is a leading health problem. Approximately 150 for each 100 000 adults experience this problem every year . Approximately 60% of the attacks of bleeding in patients with cirrhosis are due to variceal causes . Type 2 myocardial infarction (MI) is defined as myocardial necrosis owing to inequity in coronary blood flow in the face of increased myocardial oxygen demand. Recognition of this type has increased with widespread use of increasingly sensitive troponin assays . Hypovolemia, decrease in blood pressure, and decreased oxygen-carrying capacity as consequences of acute gastrointestinal bleeding contribute to myocardial ischemia and necrosis , precipitating in mostly subclinical heart failure and myocardial injury . Many studies revealed that major upper or lower gastrointestinal hemorrhage is associated with MI in ∼30–49% of cases in ICUs causing a mortality rate of ∼5–10% .
| Aim of work|| |
The aim is to assess the prevalence and potential risk factors of myocardial injury in patients with liver cirrhosis presenting with acute UGIB.
| Patients and methods|| |
This prospective study underwent from March 2017 to March 2018 at Hematemesis and Melena Units, Mansoura University Hospitals. It involved 132 patients (99 males and 33 females) diagnosed with liver cirrhosis presenting with upper gastrointestinal tract (GIT) bleeding. According to the presence or absence of myocardial injury, patients were divided into two groups:
- Group 1: with myocardial injury or ischemic heart disease (IHD) (76 patients).
- Group 2: without myocardial injury or IHD (60 patients).
This study was conducted on all patients with liver cirrhosis presenting with acute upper GIT bleeding (hematemesis and/or melena).
Patients with a history of coronary heart disease, chronic renal insufficiency, septic shock, cerebral stroke, and hematological malignancy were excluded.
Oral informed consents were obtained from the patients or their relatives participating in this study after informing them about the steps of the study. The study was approved by the Ethical Committee of Faculty of Medicine Mansoura University.
Methods of the study
Between March 2017 and March 2018, 132 patients diagnosed with liver cirrhosis presenting with upper GIT bleeding were considered eligible for analysis in this study. Patients were divided into two groups, as previously described, which were subjected to thorough medical history taking (stressing on age; sex; smoking; hypertension; diabetes mellitus; history of previous attacks; history of IHD; presentation with hematemesis, melena or both; and presence of anginal chest pain) and thorough examination (including vital signs and abdominal and cardiopulmonary examination). Both groups were subjected to standardized 12-lead ECG and cardiac troponin-I measurement on admission and 12 h later. Severity of liver disease and UGIB were determined according to Child–Turcot–Pugh (CTP) score , model for end‐stage liver disease (MELD) score , and Rockall score . All patients within 24 h from UGIB, except for those who refused or were contraindicated (irritable or shocked patients), underwent upper gastrointestinal endoscopy as an essential demonstrative and restorative device.
(1) CTP score .
(2) MELD score: MELD=3.8[log serum bilirubin (mg/dl)]+11.2[log INR]+9.6 [log serum creatinine (mg/dl)]+6.4 .
(3) Rockall score.
The total score is calculated by simple addition. A score less than 3 carries good prognosis, but a total score more than 8 carries high risk of mortality .
Data were fed to the computer and analyzed using IBM SPSS software package, version 21 (SPSS Inc., Chicago, IL, USA). Qualitative data were expressed as count and percent. Quantitative data were initially tested for normality using Kolmogorov–Smirnov and Shapiro–Wilk test, with data being normally distributed if P value more than 0.050. Quantitative data were expressed as mean±SD if normally distributed or median and interquartile range if not. Regarding data comparison, for qualitative data for two groups (2×2) table, χ2 test (or Fisher’s exact test) was used. Regarding qualitative data for more than two groups (e.g. 2×3 table), χ2 test (with Bonferroni’s method to adjust P values when comparing column proportions) was used. For quantitative data between two groups, independent samples t test was used if data were normally distributed in both groups, and the nonparametric alternative Mann–Whitney U test was used if not. Standard logistic regression was used to predict the likelihood of a diagnosis using only one predictor, and standard logistic regression analysis was used to calculate the OR with its 95% confidence interval. Multivariable logistic regression was used to create a prediction model of the likelihood of a diagnosis to detect the significant ‘independent’ predictors with their OR (95% confidence interval). Results were considered as statistically significant if P value less than or equal to 0.050.
| Results|| |
A total of 132 patients (99 males and 33 females) were involved in this trial, with an age range of 43–79 years (mean±SD, 59.29±9 years). They were divided into two groups according to myocardial injury presence. Both groups were comparable in their baseline characteristics as shown in [Table 1], whereas the main outcomes are shown in [Figure 1], [Table 2] and [Table 3], and predictors for myocardial injury in the context of UGIB are shown in Tables 4 and 5. INR, MELD score, and variceal bleeding were statistically significantly higher in IHD groups than non-IHD group. Presence of variceal bleeding has an 11-fold increased odds of occurrence of ischemia. The logistic regression model was statistically significant, χ2 (2)=9.235, P value=0.010. The model explained 28.1% (Nagelkerke R2) of the variance in coronary ischemia and correctly classified 69.2% of cases. Sensitivity was 75%, specificity was 63.2%, positive predictive value was 68.2%, and negative predictive value was 70.6%. Of the two predictor variables, variceal bleeding as a source of bleeding was statistically significant. Patients variceal bleeding had 10.5 times higher odds to exhibit coronary ischemia.
| Discussion|| |
Patients with cirrhosis with UGIB are unstable hemodynamically most of the time, which may precipitate to subclinical heart failure and myocardial injury . The current study revealed that the incidence of myocardial injury (elevated troponin levels above cutoff value and/or ECG changes) was seen in 55% (72) of patients, in contrast to Emenike et al. , who found that the frequency of myocardial injury was seen in 13%. This result can be explained by that in the Emenike study myocardial injury was diagnosed by enzymatic elevation plus ECG changes, whereas in our study, the presence of any of them is enough for the diagnosis. Troponin-I, which is considered as a superior biologic marker to identify myocardial ischemia in gastrointestinal bleeders, was positive in 25% (33) of patients compared with 12% in the study by Bellotto et al. , 8.4% in Wu et al. , and 19% in Iser et al. .
ECG ischemic changes were found in 36.3% (48) of patients in the form of ST-segment deviation or T-wave inversion compared with 24% in the study by Emenike et al. . In the study by Wu et al. , the percentage of myocardial injury was 40.46% using high cutoff troponin value (500 pg/ml), and in the study by Hamid et al. , 37% of the patients were reported to have myocardial injury. In the mentioned study, the number of patients included was 105 versus 132 in our study. Regarding age and sex, patients diagnosed with having myocardial injury were older, and mostly males were more affected than those without myocardial injury. However, age and sex were not significant predictors of myocardial injury in cirrhotic cases presented with GIT bleeding, which was in parallel to the reports by Bellotto et al.  and Hamid et al. .
Hypertension, diabetes mellitus, and smoking had no significant statistical difference between the ischemic and the nonischemic groups, which is in agreement with Bellotto et al. . This result is in contrast to the studies by Bhatti et al.  and Prendergast et al.  who proposed that these factors are the only variables capable of predicting myocardial injury. This discrepancy could be explained by the small proportion of the patients who were smokers, diabetic, and hypertensive in our study. The type of MI in our study is type 2 MI, which results from an inequity between the actual blood supply and the need of oxygen without CAD.
Regarding clinical data, Prendergast et al.  found that patients presented with hematemesis were less likely to develop complicated course that could result in ICU admission, cardiac complications, and even death than those presented with melena or hematochezia. However, in our study, we found no significant difference between the two groups regarding the presenting sign (hematemesis and/or melena) or the first or recurrent attacks. Severity of blood loss could be estimated by the presence of hypotension and reflex tachycardia. In this study, admission heart rate in patients with myocardial injury was 100.8±16 versus 96.2±9.5 in the other group. Systolic blood pressure/diastolic blood pressure in the ischemic group was 100 (90–110)/64.58±14.7, and in the nonischemic group was 110 (90–127)/68.50±15.6. However, there is no significant difference between the two groups regarding heart rate and blood pressure. This result is in same line with those of Bellotto et al. , Prendergast et al. , and Hamid et al. .
Regarding laboratory data, the patient group with myocardial injury had hemoglobin level of 7.724±1.9 in contrast to 8.658±2 in the other group, but with no significant difference, which was similar to what was reported in Prendergast et al.  and Hamid et al.  study. In contrast to our study, the study by Bellotto et al.  found that minimum hemoglobin during admission was positively correlated with myocardial ischemia, but in our study, we used hemoglobin on admission only. INR was significantly higher in the ischemic group than the nonischemic one to be 1.456±0.2 versus 1.188±0.13, respectively. This result can be explained by that increased INR is a self-determining prognostic factor of mortality in patients without anticoagulant drugs, as in Kırıs et al. . Activated coagulation, inflammation, neurohumoral stimulation, and liver failure are related to elevated INR .Regarding scoring systems, ischemic group had a Child score of 9.5 (7–10.75) in contrast to 8 (7–9.75) in the other group. Rockall score has been done also to be 5.33±1.958 in the ischemic group and 5.16±1.47 in the nonischemic group. However, there is no significant difference between the two groups regarding CTP and Rockall score. MELD score was significantly higher in the ischemic than the nonischemic group. The MELD had been initially established to evaluate the short‐term mortality in cirrhotic cases scheduled to undergo transjugular intrahepatic porto-systemic shunts . The importance of MELD is increased to involve ranking of liver transplantation .
More lately, data have been found proposing that the MELD score could be used as a new biomarker of medical and cardiovascular hazards, surprisingly in cases who do not have known liver disease ,. MELD score has been lately recognized to be related to predominant CVD in patients with nonalcoholic fatty liver disease . MELD‐Na score is known to be linked to an elevated possibility of major adverse CV events including CV death, nonfatal MI, angina in need for hospital admission, PCI, CABG, stroke, or transient ischemic attack . MELD score has been found to have a significant elevation in nonsurvived cases of ACS who underwent PCI than those who survived .
Regarding endoscopic findings, the source of bleeding was found to be a significant predictor of myocardial injury. Variceal bleeding was found in 95% of the ischemic group versus 63% in the other group. This result is in agreement with Wu et al. . However, Bellotto et al.  and Prendergast et al.  showed insignificant difference between the group diagnosed with IHD and the other group; this may be explained by small number of patients diagnosed as variceal bleeders, that is, 26 of 227 in the Belloto study, and only four of 68 patients in the Prendergast study. Other endoscopic findings were that active spurter in 30% of ischemic patients and 5.3% in nonischemic group. Portal hypertensive gastropathy has been found in 25% of patients with myocardial injury versus 42% of those without myocardial injury. None of these variables were statistically different between the two study groups. On multivariable logistic regression analysis involving MELD score and variceal bleeding as compared with nonvariceal bleeding on the likelihood of having coronary ischemia, variceal bleeding was an independent predictor of myocardial injury.
| Conclusion|| |
In conclusion, our trial revealed that more than half of the study patients presented with UGIB have had unnoticed subclinical myocardial injury. Predictors of myocardial injury in patients with UGIB must concern MELD score and variceal source of GI bleeding as our study revealed that they are independent risk factors. So, frequent monitoring, careful fluid resuscitation, and blood transfusion in addition to ECG should be taken into account for every patient presenting with UGIB.
The members involved in the study are the financiers.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
| References|| |
Hopper AD, Sanders DS. Upper GI bleeding requires prompt investigation. Practitioner 2011; 255:15–19.
Garcia‐Tsao G, Sanyal AJ, Grace ND, Carey W. Prevention and management of gastroesophageal varices and variceal hemorrhage in cirrhosis. Hepatology 2007; 46:922–938.
Sandoval Y, Smith SW, Schulz KM, Murakami MM, Love SA, Nicholson J et al.
Diagnosis of type 1 and type 2 myocardial infarction using a high-sensitivity cardiac troponin I assay with sex-specific 99th percentiles based on the third universal definition of myocardial infarction classification system. Clin Chem 2015; 61:657–663.
Bhatti N, Amoateng-Adjepong Y, Qamar A, Manthous CA. Myocardial infarction in critically III patients presenting with gastrointestinal hemorrhage: retrospective analysis of risks and outcomes. Chest 1998; 114:1137–1142.
Lee MY, Chu CS, Lee KT, Lee HC, Su HM, Cheng KH et al.
Terlipressin-related acute myocardial infarction: a case report and literature review. Kaohsiung J Med Sci 2004; 20:604–608.
Emenike E, Srivastava S, Amoateng-Adjepong Y, Al-Kharrat T, Zarich S, Manthous CA. Myocardial infarction complicating gastrointestinal hemorrhage. Mayo Clin Proc 1999; 74:235–241.
Pugh R, Murray‐Lyon IM, Dawson JL, Pietroni MC, Williams R. Transection of the oesophagus for bleeding oesophageal varices. Br J Surg 1973; 60:646–649.
Freeman RB, Wiesner RH, Harper A, McDiarmid SV, Lake J, Edwards E et al.
The new liver allocation system: moving toward evidence‐based transplantation policy. Liver Transpl 2002; 8:851–858.
Rockall TA, Devlin HB, Logan RF, Northfield TC. Selection of patients for early discharge or outpatient care after acute upper gastrointestinal haemorrhage. Lancet 1996; 347:1138–1140.
Kırıs T, Avcı E, Çelik A. Combined value of left ventricular ejection fraction and the model for end-stage liver disease (MELD) score for predicting mortality in patients with acute coronary syndrome who were undergoing percutaneous coronary intervention. BMC Cardiovasc Disord 2018; 18:44.
Sorbi D, Gostout CJ, Peura D, Johnson D, Lanza F, Foutch PG et al.
An assessment of the management of acute bleeding varices: a multicenter prospective member-based study. Am J Gastroenterol 2003; 98:2424–2434.
Bellotto F, Fagiuoli S, Pavei A, Gregory SA, Cati A, Silverj E et al.
Anemia and ischemia: myocardial injury in patients with gastrointestinal bleeding. Am J Med 2005; 118:548–551.
Wu I-C., Yu F-J., Chou J-J., Lin T-J., Chen H-W., Lee C-S. et al.
Predictive risk factors for upper gastrointestinal bleeding with simultaneous myocardial injury. Kaohsiung J Med Sci 2007; 23:8–16.
Iser DM, Thompson AJ, Sia KK, Yeomans ND, Chen RY. Prospective study of cardiac troponin I release in patients with upper gastrointestinal bleeding. J Gastroenterol Hepatol 2008; 23:938–942.
Hamid EF, Emara MH, El-Hawari SA, Ateya AA, Sherbeni AS. Upper gastrointestinal bleeding associated with myocardial injury in Egyptian patients: frequency and risk factors. J Gastroenterol Hepatol Res 2013; 2:419–422.
Prendergast HM, Sloan EP, Cumpston K, Schlichting AB. Myocardial infarction and cardiac complications in emergency department patients admitted to the intensive care unit with gastrointestinal hemorrhage. J Emerg Med 2005; 28:19–25.
Wong CC, Ng AC, Lau JK, Chow V, Chen V, Ng AC et al.
High mortality in patients presenting with acute pulmonary embolism and elevated INR not on anticoagulant therapy. Thromb Haemost 2016; 115:1191–1199.
Malinchoc M, Kamath PS, Gordon FD, Peine CJ, Rank J, Ter Borg PC. A model to predict poor survival in patients undergoing transjugular intrahepatic portosystemic shunts. Hepatology 2000; 31:864–871.
Wiesner R, Edwards E, Freeman R, Harper A, Kim R, Kamath P et al.
Model for end-stage liver disease (MELD) and allocation of donor livers. Gastroenterology 2003; 124:91–96.
Kim MS, Kato TS, Farr M, Wu C, Givens RC, Collado E et al.
Hepatic dysfunction in ambulatory patients with heart failure: application of the MELD scoring system for outcome prediction. J Am Coll Cardiol 2013; 61:2253–2261.
Wernly B, Lichtenauer M, Franz M, Kabisch B, Muessig J, Masyuk M et al.
Model for End-stage Liver Disease excluding INR (MELD-XI) score in critically ill patients: Easily available and of prognostic relevance. PloS One 2017; 12:e0170987.
Corey KE, Kartoun U, Zheng H, Chung RT, Shaw SY. Using an electronic medical records database to identify non-traditional cardiovascular risk factors in nonalcoholic fatty liver disease. Am J Gastroenterol 2016; 111:671.
Simon TG, Kartoun U, Zheng H, Chan AT, Chung RT, Shaw S, Corey KE. Model for end‐stage liver disease Na score predicts incident major cardiovascular events in patients with nonalcoholic fatty liver disease. Hepatology Commun 2017; 1:429–438.
[Table 1], [Table 2], [Table 3]