|Home | About | Journals | Submit | Contact Us | Français|
We thank Dr Abeles and colleagues for their comments on our recent paper on the predictive value of E/A ratio in cirrhotic patients with a transjugular intrahepatic portosystemic shunt (TIPS), which gives us the opportunity to extend the discussion on some particular aspects.
We agree that echocardiographic measurement of the E/A ratio can be an imperfect tool for diagnosing diastolic dysfunction compared with new more sophisticated methods such as Doppler tissue imaging. However, when we started our study this new methodology was not available in our centre. Nevertheless, almost all previous studies on cirrhotic cardiomyopathy have used E/A ratio, deceleration time and/or isovolumic relaxation time to diagnose diastolic dysfunction.1,2,3 Moreover, Abeles observes that echocardiographic calculations were not verified by an independent observer to avoid operator bias. However, this risk was limited in our study because there was just one echocardiographic operator who was unaware of the clinical conditions of the patients.
The most important concern of Abeles and colleagues is that interpretation of the E/A ratio is affected by two important variables: loading conditions and age. For the first variable, one should take into account that patients with advanced cirrhosis have central hypovolaemia owing to a redistribution of the blood volume from the central to the splanchnic bed, and that TIPS insertion markedly increases the blood volume return to the heart and often corrects central hypovolaemia.4 This is confirmed by our data, which show a significant increase of the average value of the left ventricular end diastolic volume (LVEDV) after TIPS in both groups of patients. Therefore, measuring the E/A ratio in patients with TIPS may lead to a wrong diagnosis of normal diastolic function (pseudonormalisation effect) but is less likely to lead to a wrong diagnosis of diastolic dysfunction. Abeles and colleagues affirm that the E/A ratio increased after TIPS in all patients, but this is incorrect as the E/A ratio decreased or remained unchanged in 11 of 32 patients. Indeed, we have already reported that TIPS in cirrhotic patients may be able to distinguish between patients who have a true diastolic dysfunction due to ventricular stiffness (those whose E/A ratio did not rise or decrease after TIPS) and those who have an E/A ratio <1 because of circulatory underfilling and an excessively low preload (those whose E/A ratio increases after TIPS).
We also agree that age must be taken into account when interpreting E/A ratios, because they tend to reduce with aging. In our study, however, the average age was not significantly different between the two groups of patients with low or normal E/A ratio after TIPS insertion. In addition, although we observed a marginal inverse correlation between the E/A ratio and age (p=0.1), age was not associated with death at univariate analysis, making it unlikely that the predictive value of the E/A ratio was dependent on aging.
The hypothesis that a deterioration of systolic function rather than of diastolic dysfunction was responsible for the poor outcome in our cirrhotic patients is attractive but not sufficiently supported by our data. Indeed, the statistical analysis excluded the possibility that a drop in the ejection fraction (EF) or any other echocardiographic variable, except E/A ratio, was associated with death. Moreover, none of the patients had cardiac symptoms and no death was caused by evident cardiac failure. Nevertheless, we hope that this subject will be investigated further in a larger group of patients.
We agree with Abeles and colleagues that the most important message of our study is that cardiac function should be carefully assessed in critically ill patients with cirrhosis. This concept has recently been corroborated by other investigations, which showed that a relatively insufficient cardiac output was associated with the development of renal failure in cirrhotic patients with spontaneous bacterial peritonitis.5 The prevalence of diastolic dysfunction in cirrhosis should be re‐evaluated with more accurate methods such as colour M‐mode echocardiography and Doppler tissue imaging, which are independent of the load conditions.