Editorial: cardiac volume overload and pulmonary hypertension in long‐term follow‐up of TIPSS patients

Cirrhotic cardiomyopathy is defined as a disease ‘characterised by impaired contractile responsiveness to stress and/or altered diastolic relaxation with electrophysiological abnormalities in the absence of other known cardiac diseases’.1 Thereby, left ventricular systolic dysfunction is manifested by the inability to generate an adequate arterial blood pressure and cardiac output. Due to the decreased afterload, however, overt heart failure is rare and the ejection fraction (EF) remains normal as long as the patient is at rest.1, 2 Exercise or pharmacological challenge is needed to objectify systolic dysfunction by the inability to sufficiently increase cardiac output. Thus, systolic dysfunction (EF <50%) was not seen at rest, but 38.5% of cirrhotic patients showed an insufficient response to dopamine challenge defined as an increase in left ventricular EF by <10%.3 Left ventricular diastolic dysfunction generally precedes systolic dysfunction. The morphological finding responsible for diastolic dysfunction is an increased stiffness of the myocardial wall due to myocardial hypertrophy, fibrosis, and subendothelial oedema.2 Physiologically, it is characterised by an impaired relaxation leading to an increased left ventricular filling pressure. Using echocardiography, the elevated filling pressure corresponds to a changed transmitral blood flow, which is seen in 45–56% of cirrhotic patients.1, 2 It can be objectified by measurement of the peak early filling (E-wave) velocity, late diastolic atrial filling (A-wave) velocity and the E/A ratio. An E/A ration >1 is indicative of diastolic dysfunction. In addition to these parameters, isovolumic relaxation times, left atrial enlargement and increased left ventricular wall thickness are recognised as indicators for diastolic dysfunction and found in all cirrhotic patients.4 Pulmonary hypertension (PH) is present when the mean pulmonary artery pressures exceeds 24 mmHg diagnosed by right-heart catheterisation at rest. The most common cause is left heart failure leading to post-capillary PH with an elevated pulmonary capillary wedge pressure (PCWP).5 Pre-capillary PH can be associated with various diseases, such as chronic thromboembolism or portal-pulmonary hypertension. Both types eventually lead to remodelling of pulmonary arteries with consecutive alterations of the arterial wall. The retrospective study by Wannhoff et al. 6 investigated the effects of TIPSS on cardiac function and systolic pulmonary arterial pressures using echocardiography. The findings were compared to a matched cohort of 110 patients waiting for liver transplantation. Due to the retrospective nature of the study, measurements were pooled into time intervals post TIPSS, i.e. 1 day–6 weeks, 2–3 months, 4–10 months and 1–5 years, including 17, 18, 28 and 55 patients respectively. Compared to baseline, measurements performed 1–5 years after TIPSS implantation showed a slight but significant increase in the left atrial diameter (LAD) from 37 to 40 mm, in the left ventricular end-diastolic diameter (LV-EDD) from 45 to 48 mm, and pulmonary arterial systolic pressure (PASP) from 25 to 30 mmHg. An E/A ratio of <1 was seen in 24 of 59 patients before TIPSS (40.7%) and did not change in the few patients available for paired comparison. The left ventricular ejection fraction (LV-EF) was reduced in 4 of 34 patients before TIPSS who normalised during long-term follow-up. When TIPSS patients are compared to matched controls, again a significant difference was seen with respect to long-term changes of LAD, LV-EDD and PASP. In eight patients with an increased PASP at echocardiography, right-heart catheterisation was performed showing PH in six patients. Three patients each had elevated or normal PCWP. TIPSS, like transplantation, has been regarded as a cardiac stress by suddenly increasing the preload leading to increased cardiac diastolic volumes and diameters, and a transient PH for 3–6 months following TIPSS implantation.2, 7 Left ventricular diastolic dysfunction may improve or deteriorate after TIPSS 2, 8 and be associated with a slower response of ascites mobilisation and improvement of renal function.2 In addition, the persistence of diastolic dysfunction 1 month after TIPSS implantation correlates with mortality.9 In the study by Wannhoff et al.,6 short and median-term findings (1 day to 10 months after TIPSS) did not show significant changes in systolic and diastolic function when compared to baseline values, except PASP increasing from 25 mmHg to 33 mmHg at time-point 1 (1 day to 6 weeks) and LV-EDD increasing from 47 mm to 50 mm at time-point 2 (2–3 months). The lack of a continuous development of changes over time reduces the strength of the long-term findings. A biphasic development (early and late) of cardiac dysfunction as discussed by the authors seems to be rather unlikely because findings at 4–10 months after TIPSS did not show any tendency towards the findings at 1–5 years. With respect to systolic dysfunction, findings suggest even improvement by TIPSS. This is in line with the fact that TIPSS rapidly improves renal function together with a decrease in plasma renin activity, aldosterone and noradrenalin concentrations.10 Thus, the findings do not entirely confirm the hypothesis that TIPSS exerts cardiac stress aggravating diastolic and unveiling subclinical systolic dysfunction.1, 2 However, the findings were not analysed with regard to overall or transplant-free survival (55 of 158 patients died and 22 were transplanted). Authors interpreted the long-term increases of LAD, LV-EDD and PASP as being due to cardiac volume overload, a statement which is not based on respective measurements, i.e. cardiac output. As demonstrated, increased preload and cardiac output reverse towards pre-TIPSS values during follow-up reaching baseline values within 1 year.10 Thus, the changes may rather be due to deterioration of the diastolic dysfunction. This is supported by the E/A ratio which increased from 1.0 at baseline to 1.3 at 1–5 years. Unfortunately, only nine patients could be evaluated, a probable reason why the difference did not reach significancy. With respect to the control group, comparability may be questioned. It is most likely that the control group did not have serious portal hypertensive complications indicating TIPSS. The conclusion that TIPSS increases the incidence of PH by comparing the two groups is therefore dubious. In addition, the lack of PH in the control group is not in agreement with the literature, where 7% of similar patients exhibited PH.11 In summary, the study by Wannhoff et al. nicely demonstrates the need for regular cardiac surveillance. The diagnosis of PH in an early stage is of paramount importance earning increased recognition in TIPSS patients as well as in cirrhotic patients without a TIPSS. The study is an eye opener by pointing at this rare clinical complication and of great value by providing important hints at future investigations. Declaration of personal interests: Jonas C. Schupp has received travel support from Bayer. Thomas C. Köhler has received fees for serving on advisory boards from Actelion, Bayer, travel support from Actelion, Bayer, Mundipharma and lecture fees from Actelion, Astra Zeneca, Bayer, GlaxoSmithKiine, Novartis. M. Schultheiss has served as a speaker for Falk Foundation and for Gore Company. He has been advisory board member for Bayer. Declaration of funding interests: None.