Introduction TIPSS is widely used to treat refractory ascites and variceal haemorrhage. Cardiac dysfunction is frequently observed in patients with cirrhosis. However, there remains a paucity of data from routine clinical practice regarding the use of echocardiography in the pre-assessment of TIPSS.
Aim To investigate if echocardiography predicts outcomes post-TIPSS in patients with cirrhosis.
Method Patients who underwent echocardiography and TIPSS at the liver transplant centre (Birmingham, UK) between 1999–2016 were included. All echocardiography measures [left ventricle (LV) ejection fraction (EF); LV diastolic diameter (LVDD); LV systolic diameter (LVSD); early maximal ventricular filling velocity/late filling velocity (E/A) ratio, deceleration time (DT), diastolic dysfunction as per ASE/BSE guidelines; regional wall abnormality) were independently reviewed/graded by a senior cardiologist (RS). Clinical predictors of 30 day, 90 day and overall transplant free-survival were assessed.
Results 117 patients with cirrhosis (median age 56 years; 54% alcohol; CPB/C 71/14.5%) underwent TIPSS for ascites (n=78) and variceal haemorrhage (n=39). Median MELD and UKELD was 12 (IQR 9–17) and 53 (IQR 50–57), respectively. Median post-TIPSS portal pressure gradient was 8 (IQR 6–10) mmHg. Post-TIPSS complications (<90 days) include arterial puncture (n-1), stent thrombosis (n=4), de novo encephalopathy (n=16), cardiac failure (n=1) and fulminant liver failure (n=2).
30 day, 90 day and overall transplant-free survival was 90% (n=105), 80% (n=93), and 31% (n=36) over a median 663 (IQR 385–2368) days follow-up. MELD (p<0.001), UKELD (p<0.01) and CP Score (p<0.01) significantly predicted 30 day and overall transplant-free survival. A MELD >/=15 had an AUROC 0.855 (95% CI 0.74–0.98) for 30 day transplant free-survival.
6% (n=7) of patients pre-TIPSS had a history of IHD and 34% (n=40) had 1 or more CVD risk factors. 50% (n=59) had a normal echocardiography, 33% (n=39) had grade 1–3 LV diastolic dysfunction and 6% (n=7) had LVEF<55%. On univariate analysis none of the pre=TIPSS echocardiography measures, including LVEF% (p=0.41), LVDD (p=0.07), LVSD (p=0.98), pulmonary artery pressure (p=0.87), E/A (p=0.27), DT (p=0.57), regional wall abnormality (p=0.97) and grade 1–3 LV diastolic dysfunction (p=0.47), were related to survival post-TIPSS.
Conclusion Echocardiography in patients with cirrhosis does not predict post-TIPSS survival. MELD score remains the best predictor of early and late mortality post-TIPSS.
Disclosure of Interest None Declared