Heart rhythm turbulence and NT-proBNP in decompensated liver cirrhosis – a pilot study

Background

Heart rhythm turbulence (HRT) is a novel tool for evaluation of cardiovascular mortality. HRT is especially used to evaluate risk of death in a patient with ischemic heart disease and congestive heart failure [1]. HRT describes a short-term heart rhythm (HR) fluctuation in sinus rhythm that follows the ventricular premature beat. Early acceleration followed by late deceleration of sinus rhythm is observed. This phenomenon usually includes 15–20 consecutive beats. It is associated with autonomic nervous system activity and baroreceptor reflex [2]. Two phases of HRT are estimated by the following parameters: turbulence onset (TO) and turbulence slope (TS) (Figure 1).

Normal values of these parameters are calculated as TO <0% and TS >2.5 ms/RR [3]. Recent findings suggest that TO and TS values are comparable or even better predictors of sudden cardiac death in myocardial infarction than are accelerated basic HR, ventricular arrhythmia, lowered left ventricular ejection fraction (LVEF<30%), and age over 65 years. The highest risk of death was demonstrated for patients after MI with impaired HRT [1,4,5].

Cirrhosis is associated with various hemodynamic and myocardial disturbances termed cirrhotic cardiomyopathy [6,7]. Impaired cardiac contractility, systolic and diastolic dysfunction, electromechanical abnormalities, chronotropic incompetence and various arrhythmias are observed in cirrhotic patients [8–12]. Increased or normal cardiac output, decreased blood pressure and total peripheral resistance are also noted. Increased cardiac load results in contractility disturbances and decreased cardiac output. Diastolic function is also affected, which is associated with the increase in left ventricular filling pressure. These pathological processes are associated with LV hypertrophy and myocardial fibrosis [13,14], and is also observed during surgery, pharmacotherapy, stress, transjugular intrahepatic portosystemic shunt placement (TIPS) or liver transplantation [15–17].

Recent findings suggest that systolic myocardial dysfunction is positively correlated with brain natriuretic peptide (BNP) in the stable stage of cirrhosis, so this parameter seems to be a marker of cirrhotic cardiomyopathy [18–21]. A recently published study by Pimeta et al. [22] proved that elevated BNP level is also associated with cardiac contractility impairment in decompensated cirrhosis. This biomarker may by an independent mortality predictor in medium-term survival in end-stage cirrhosis.

Until now, HRT and its correlation with NT-proBNP in patients with cirrhosis have not been evaluated. The aim of our study was to evaluate the HRT and its correlation with NT-proBNP, creatinine, ammonium, sodium, total bilirubin, albumin, INR and echocardiographic parameters in patients with decompensation of cirrhosis. We estimated a role of HRT and NT-proBNP evaluation in patients with cirrhosis decompensation.

Material and Methods

STATISTICAL ANALYSIS:

Statistical analysis was performed using STATISTICA 8 PL (StatSoft Inc.). Results are presented as median, average, minimum and maximum values. Statistical significance was indicated for P<0.05. Variable distribution was assessed by Shapiro-Wilk W test. As continuous variables were non-normally distributed, we used nonparametric tests: Mann-Whitney and Wilcoxon test. Correlations between variables were tested by using the Spearman R test. We attempted to determine if RV or/and LV parameters and biochemical parameters of liver function have an impact on HRT and NT-proBNP.

Results

For the first 6 months of the study, 27 patients with presumptive diagnosis of decompensated liver cirrhosis were admitted to our department. Among them 4 patients had previously diagnosed ischemic heart disease, 2 patients had heart failure caused by alcoholic cardiomyopathy, 1 patient had hemorrhage from esophageal varices in the last 2 months, 1 patient had lung tumor (carcinoma plano epitheliale), and 1 patient did not give consent to participate in the study. Finally, 18 patients aged 33–73 years (13 men and 5 women) were enrolled to the study.

Clinical, echocardiographic and laboratory parameters are presented in Table 1.

No significant differences were observed in age, BMI, women percentage, maximal HR, TAPSE, LVEF, LVEDd, LVESd, IVSDd, creatinine level between the study group and the control group. Mean HR, minimal HR, ammonium, NT-proBNP, total bilirubin, INR, LAd, RVdd, SPAP, DT, and TO were significantly higher in patients with cirrhosis compared to the control group. Hemoglobin level, PLT, sodium and albumin level, E/A ratio, and Vprop were significantly lower in patients with cirrhosis compared to the control group. No significant changes were noted in other studied parameters between group I and group II. Patients from group I had significantly better stage in Child-Pugh classification (p=0.04) than patients with abnormal TO or TS. Correlations of selected biochemical and echocardiographic parameters with TO, TS and NT-proBNP were calculated (Table 2).

Significant negative correlation was found between creatinine and TO. For other studied parameters, significant correlations were not found with TO value. Significant negative correlation was found between mean HR and TS, and between Vprop and NT-proBNP level. Significant positive correlation was found between LAd and TS. During the study period 7 patients died. We compared selected parameters between patients that died and patients that survived (Table 3.).

Except for parameters of diastolic function (E/A ratio, DT and Vprop), any significant differences were noted in studied parameters between patients that died and patients that survived during the observational period. Diastolic function was significantly disturbed in patients that died during the observation period. There were no significant differences in all parameters regarding etiology of liver cirrhosis.

Discussion

Cirrhotic patients rarely present symptoms of apparent heart failure; however, some patients develop cardiomyopathy related to cirrhosis. At the beginning of this state we observed hyperkinetic circulation and increased or normal cardiac output; simultaneously, lowered peripheral resistance results in decreased blood pressure. Progressive fibrosis and hypertrophy of the myocardium results in disturbances of LV diastolic function [9,10,12,13,15]. In this condition, increased load may result in acute circulatory failure. Such a state may occur in patients after liver transplantation.

Histopathological findings and hemodynamic disturbances on early stages of cirrhotic cardiomyopathy are related to those observed in hypertrophic cardiomyopathy (HCM) [25]. Our study revealed significantly higher mean HR in patients with cirrhosis and significantly higher minimal HR compared to the control group. LV EF and TAPSE in cirrhotic patients did not differ significantly from the control group. This confirms that our patients developed hyperkinetic circulation. Disturbances of diastolic LV function were demonstrated for all parameters in the study group. Similar results were presented by Bernal et al. [26] in patients undergoing liver transplantation. Moreover, fluid overload and LV hypertrophy were noted in these patients. Our patients also presented other laboratory abnormalities typical for cirrhosis: decreased levels of hemoglobin, PLT, sodium, albumins and elevated concentrations of bilirubin and ammonium. Our results are consistent with previous data presented by Pimenta et al. [22], who also observed significantly higher BNP levels in cirrhotic patients, mainly in patients with cardiac systolic dysfunction, defined by low cardiac output and associated with an increased occurrence of death during 6 months observation. In our studies we also noted significantly elevated NT-proBNP concentration in patients with cirrhosis compared to the control group, which may be associated with excessive circulatory system load in decompensation of cirrhosis. Increased serum level of natriuretic peptides seems to be rather an effect of damage and structural lesion of myocardium and is not associated with hyperkinetic circulation. In our study we did not observe significant correlations between NT-proBNP and biochemical parameters, mean HR, LAd, RVdd, TAPSE and SPAP, as also noted by Pimenta et al. In contrast to the findings of Pimenta et al, we did not observe significant correlation between NT-proBNP and LVEF. Our results revealed that systolic LV function was not affected in cirrhotic patients. It is possible that classical echocardiography examination is not sufficiently sensitive to find mild LV systolic function disturbances that cause increased NT-proBNP in cirrhotic patients. Moreover, patients included for our studies had no past history of cardiovascular diseases. Because study participants had no previous cardiovascular diseases, it seems that increase in NT-proBNP is a result of myocardial damage caused by cirrhosis. In a stable stage of cirrhosis, elevated levels of NT-proBNP were also observed, and this biomarker seems to be an indicator of cirrhotic cardiomyopathy [18,19]. Readle-Hurst et al. [27] demonstrated that elevated NT-proBNP level may be a useful marker of LV diastolic function disturbances in liver cirrhosis. These authors suggest that NT-proBNP levels exceeding 290 pg/ml should undergo further cardiac evaluation. We also demonstrated significantly higher NT-proBNP level and LV diastolic dysfunction in cirrhotic patients, which confirms previous Readle-Hurst et al observations. Significant correlation was found between Vprop and NT-proBNP level, indicating that elevation of NT-proBNP levels follows diastolic dysfunction progress. In our studies, NT-proBNP levels and LVEF did not differ significantly for both values between patients that died during the study period and patients that survived. It is possible that NT-proBNP levels and LVEF, contrary to chronic heart failure, is not a predictor for survival in end-stage cirrhosis.

In cirrhotic patients, 1 HRT parameter differed significantly compared to the control group; TO value in cirrhotic patients amounted to −0.385% v/s −0.920% in the control group. TS value did not differ significantly between cirrhotic patients and the control group. In spite of these differences, TO and TS values were within normal limits in a majority of study participants. Only 3 patients had abnormal TO values, and 4 patients had abnormal TS values. Other studied parameters, including NT-proBNP, did not differ significantly between patients with normal TO and TS and patients with abnormal TO or TS. Correlations between NT-proBNP and TO and TS values were not found in any of the studied groups. Interestingly, significant negative correlation between creatinine level and TO occurred, which may indicate that development of renal failure correlates with improvement in TO value. There are no data that may explain this phenomenon. Significant negative correlation was found between mean HR and TS. There are not at present any available data on heart rate turbulence in cirrhosis in the literature. In studies performed on patients with congestive heart failure by Szymanowska et al. [28], contrary to our results, no correlations were found between mean HR and HRT parameters (TO, TS). Authors of the cited study observed significant negative correlation between TS and BNP, but they did not observe correlations between TO and biochemical parameters. They also demonstrated a correlation between progression of heart failure and worsening of HRT parameters. Kawasaki et al. [29] studied patients with HCM; revealing that TO and TS stayed within normal limits (TO −2.1±3.4 vs. −1.1±2.9 vs. −1.4±5.1%; p=0.22; TS 18.0±13.9 vs. 10.6±8.6 vs. 16.6±9.7 ms/beat; p=0.00084, respectively). As mentioned above, myocardial lesion in HCM is similar to damage observed in cirrhotic cardiomyopathy. These results are similar to our results obtained from patients with cirrhosis. Kawasaki et al did not find significant differences in HRT parameters between patients with HCM and the control group. A significant difference was observed in TS for patients after MI vs. the control group. In cited work, HRT parameters stayed within normal limits during a 27-month observation of patients with HCM and did not affect survival. These observations suggest that HCM is not associated with HRT disturbances. HRT disturbances are present when myocardium is damaged due to ischemic heart disease.

Our studies indicate that worsening HRT parameters are associated with deterioration of cirrhosis assessed by Child-Pugh classification. This observation agrees with a previous study by Szymanowska et al, which demonstrated that worsening HRT parameters were correlated with deterioration of congestive heart failure estimated by NYHA scale. Thus, abnormalities in HRT (mainly TO) parameters may be associated with worse prognosis in cirrhotic patients.

Conclusions

To conclude, our study demonstrates that patients with decompensation of cirrhosis have elevated levels of NT-proBNP and left ventricle diastolic dysfunction. In a majority of study participants, TO and TS, or at least 1 of these parameters, stayed within normal limits. However, TO values in cirrhotic patients differed significantly from the control group.

Increased NT-proBNP level is a result of myocardium lesion typical for cirrhotic cardiomyopathy (mainly diastolic dysfunction) and not a result of biochemical changes and hyperkinetic circulation. Elevated NT-proBNP level can indicate risk of symptomatic heart failure development. NT-proBNP may be marker of cirrhotic cardiomyopathy. No correlations between NT-proBNP and TO and TS were found. Patients with normal TO and TS had significantly lower Child-Pugh class than patients with abnormal TO or TS. HRT parameters do not appear to be appropriate death predicators in cirrhosis decompensation. These parameters can be estimated in complex evaluation of autonomic disturbances that occurred in cirrhosis. This problem requires further studies.