Risk of further decompensation/mortality in patients with cirrhosis and ascites as the first single decompensation event

Abstract

Background & aims: Although ascites is the most frequent first decompensating event in cirrhosis, the clinical course after ascites as the single index decompensation is not well defined. The aim of this multicentre study was thus to systematically investigate the incidence and type of further decompensation after ascites as the first decompensating event and to assess risk factors for mortality.

Methods: A total of 622 patients with cirrhosis presenting with grade 2/3 ascites as the single index decompensating event at 2 university hospitals (Padova and Vienna) between 2003 and 2021 were included. Events of further decompensation, liver transplantation, and death were recorded.

Results: The mean age was 57 ± 11 years, and most patients were male (n = 423, 68%) with alcohol-related (n = 366, 59%) and viral (n = 200,32%) liver disease as the main aetiologies. In total, 323 (52%) patients presented with grade 2 and 299 (48%) with grade 3 ascites. The median Child-Pugh score at presentation was 8 (IQR 7-9), and the mean model for end-stage liver disease (MELD) was 15 ± 6. During a median follow-up period of 49 months, 350 (56%) patients experienced further decompensation: refractory ascites (n = 130, 21%), hepatic encephalopathy (n = 112, 18%), spontaneous bacterial peritonitis (n = 32, 5%), hepatorenal syndrome-acute kidney injury (n = 29, 5%). Variceal bleeding as an isolated further decompensation event was rare (n = 18, 3%), whereas non-bleeding further decompensation (n = 161, 26%) and ≥2 concomitant further decompensation events (n = 171, 27%) were frequent. Transjugular intrahepatic portosystemic shunt was used in only 81 (13%) patients. In patients presenting with grade 2 ascites, MELD ≥15 indicated a considerable risk for further decompensation (subdistribution hazard ratio [SHR] 2.18; p <0.001; 1-year incidences: <10: 10% vs. 10-14: 13% vs. ≥15: 28%) and of mortality (SHR 1.89; p = 0.004; 1-year incidences: <10: 3% vs. 10-14: 6% vs. ≥15: 14%). Importantly, mortality was similarly high throughout MELD strata in grade 3 ascites (p = n.s. for different MELD strata; 1-year incidences: <10: 14% vs. 10-14: 15% vs. ≥15: 20%).

Conclusions: Further decompensation is frequent in patients with ascites as a single index decompensation event and only rarely owing to bleeding. Although patients with grade 2 ascites and MELD <15 seem to have a favourable prognosis, those with grade 3 ascites are at a high risk of mortality across all MELD strata.

Lay summary: Decompensation (the development of symptoms as a result of worsening liver function) marks a turning point in the disease course for patients with cirrhosis. Ascites (i.e. , the accumulation of fluid in the abdomen) is the most common first decompensating event, yet little is known about the clinical course of patients who develop ascites as a single first decompensating event. Herein, we show that the severity of ascites is associated with mortality and that in patients with moderate ascites, the widely used prognostic MELD score can predict patient outcomes.

Keywords: ACLF, acute-on-chronic liver failure; Acute kidney injury; Baveno; CI, confidence interval; CPS, Child–Pugh score; HCC, hepatocellular carcinoma; HE, hepatic encephalopathy; HRS-AKI, hepatorenal syndrome–acute kidney injury; Hepatic decompensation; ICA, International Club of Ascites; INR, international normalised ratio; IQR, interquartile range; LT, liver transplantation; MELD, model for end-stage liver disease; NAFLD, non-alcoholic fatty liver disease; NASH, non-alcoholic steatohepatitis; PVT, portal vein thrombosis; Portal hypertension; RA, refractory ascites; SBP, spontaneous bacterial peritonitis; SHR, subdistribution hazard ratio; Spontaneous bacterial peritonitis; TFS, transplant-free survival; TIPS, transjugular intrahepatic portosystemic shunt; UNOS MELD (2016), United Network for Organ Sharing model for end-stage liver disease (2016); VB, variceal bleeding; aSHR, adjusted subdistribution hazard ratio.

Graphical abstract

Fig. 1

Clinical course after ascites as the index decompensation event. A Sankey plot of the first further decompensation event in patients with ascites as the first index decompensation.

Fig. 2

Temporal sequence of further decompensating events. A Sankey plot depicting the temporal sequence of further decompensating events according to ascites graduation and outcome at the last follow-up. HRS, hepatorenal syndrome (-AKI, acute kidney injury); LT, liver transplantation; multiple, more than 1 concomitant decompensating event; no f. dec., no further decompensation events; RA, refractory ascites; SBP, spontaneous bacterial peritonitis; VB, variceal bleeding.

Fig. 3

Further decompensation according to MELD and ascites severity at index presentation. Cumulative incidence plot of further decompensation stratified according to MELD in (A) the study cohort, (B) patients with grade 2 ascites, and (C) patients with grade 3 ascites. MELD, model for end-stage liver disease; SHR, sub-distribution hazard ratio.

Fig. 4

Transplant-free survival according to MELD and ascites severity at index presentation. Probability of transplant-free survival stratified according to MELD strata in (A) the study cohort, (B) patients with grade 2 ascites, and (C) patients with grade 3 ascites. MELD, model for end-stage liver disease; SHR, sub-distribution hazard ratio.