Alimentary Pharmacology & Therapeutics ( IF 6.6 ) Pub Date : 2024-11-05 , DOI: 10.1111/apt.18371 Ting Wang, Deli Zou, Xingshun Qi
At present, non-selective beta-blockers (NSBBs) are recommended for the management of clinically significant portal hypertension and primary or secondary prophylaxis of variceal bleeding in cirrhosis [1, 2]. However, as mentioned in the editorial by Alferink and de Knegt [3], the right time to discontinue NSBBs remains inconclusive. The window hypothesis, which has been for the first time proposed by Krag et al. suggests that the beneficial effects of NSBBs may be lost in end-stage cirrhosis [4]. Since then, numerous studies have confirmed their deleterious effects in some specified populations (Table 1), which may negatively correlate with the Model of End-Stage Liver Disease (MELD) score [5]. In a previous meta-analysis by our group, the use of NSBBs significantly increased the risk of renal dysfunction in patients with a MELD score of > 15 [6]. Calès et al. also found that the use of NSBBs worsened the survival in patients with a MELD score of > 12 [7]. Similarly, our current study further suggested that NSBBs should increase the risk of further decompensation in decompensated patients with a MELD score of > 9 [8].
| First author (year) | No. Pt. | Study design | Study population | Follow-up duration | NSBBs group | Control group | Outcomes | Major findings | |
|---|---|---|---|---|---|---|---|---|---|
| Type | Dose (mg/d) | ||||||||
| Sersté (2010) [11] | 151 | Retrospective | Cirrhosis and refractory ascites | 8 months | Propranolol | 113.00 ± 46.00 | No NSBBs | Mortality | The use of NSBBs was associated with poor survival in patients with refractory ascites |
| Mandorfer (2014) [12] | 607 | Retrospective | Cirrhosis and SBP | > 90 days | Propranolol | 20.00–120.00 | No NSBBs | Mortality | NSBBs increased mortality, time of hospitalisation, and risks for HRS and AKI in cirrhotic patients with SBP |
| Time of hospitalisation | |||||||||
| Carvedilol | 6.25–25.00 | ||||||||
| HRS and AKI | |||||||||
| Sersté (2015) [13] | 139 | Retrospective | Severe alcoholic hepatitis | 168 days | Propranolol | Various | No NSBBs | AKI | The use of NSBBs was associated with a higher cumulative incidence of AKI in patients with severe alcoholic hepatitis |
| Madsen (2016) [14] | 81 | Retrospective | Cirrhosis and SBP | 120 days | NA | NA | No NSBBs | Mortality | The use of NSBBs was associated with increased mortality in cirrhotic patients with SBP |
| Kim (2017) [15] | 510 | Retrospective | Cirrhosis and ascites | 18.2 months | Propranolol | Various | No NSBBs | AKI | NSBBs increased the risk of AKI by more than 3-fold among the patients with cirrhosis and ascites |
| Nadolol | |||||||||
| Yoo (2020) [16] | 271 | Retrospective | Ascites received primary prevention of EV | 42.1 months | NSBBs+EVL | NA | EVL | Mortality | All-cause mortality was significantly higher in the EVL+ NSBBs group than in the EVL |
| Cales (2021) [7] | 258 | Retrospective | Alcoholic cirrhosis | 5.3 ± 2.6 years | Propranolol | 20.00–240.00 | No NSBBs | Mortality | The use of NSBBs was significantly associated with increased overall and liver-related mortality among the patients with a MELD score of ≥ 12 |
| Nadolol | 80.00 | ||||||||
| Carvedilol | 12.50 | ||||||||
| Sotalol | 80.00 | ||||||||
| Singh (2022) [17] | 160 | Prospective | Cirrhosis and ≥ grade 2 ascites | 18 months | Propranolol | 20.00–80.00 | EVL | Mortality | The use of NSBBs was associated with poor survival, increased recurrence of ascites and risk of AKI in cirrhotic patients with grade ≥ 2 ascites |
| Relapse of ascites | |||||||||
| AKI | |||||||||
| Fallahzadeh (2022) [18] | 393 | Retrospective | Cirrhosis | 3.8 years | Nadolol | 20.00–80.00 | No NSBBs | HE-related readmissions | The use of NSBBs was independently associated with increased HE-related readmissions in cirrhosis |
| Propranolol | 10.00–80.00 | ||||||||
| Carvedilol | 6.25–50.00 | ||||||||
| Metoprolol | 12.50–200.00 | ||||||||
| Atenolol | 25.00–100.00 | ||||||||
| Tonon (2024) [5] | 617 | Prospective | Cirrhotic outpatients | 60 months | NA | NA | No NSBBs | NAD | The use of NSBBs was associated with an increased risk of AD and NAD in cirrhosis |
| AD | |||||||||
| Wang (2024) [8] | 332 | Retrospective | Decompensated cirrhosis | 3.93 years | Propranolol | 5.00–90.00 | No NSBBs | Further decompensation | NSBBs may be beneficial for the prevention of further decompensation in cirrhotic patients with a MELD score of ≤ 9, but deleterious in those with a MELD score of > 9 |
| Mortality | |||||||||
- Abbreviations: AD, acute decompensation; AKI, acute kidney injury; EV, oesophageal varices; EVL, oesophageal variceal ligation; HE, hepatic encephalopathy; HRS, hepatorenal syndrome; MELD, model for end-stage liver disease; NAD, non-acute decompensation; NSBBs, non-selective beta-blockers; SBP, spontaneous bacterial peritonitis.
As pointed out in the editorial by Alferink and de Knegt mentioned above [3], there was a significant interaction between the MELD score and the use of NSBBs in our current study. Accordingly, we compared the incidence of further decompensation stratified by MELD score deciles between the NSBBs group and the non-NSBBs group. We found that a MELD score of 9 should be a specific turning point to identify a bidirectional impact of NSBBs on the risk of further decompensation in cirrhosis. Specifically, NSBBs may be harmful in patients whose MELD score is above nine, and vice versa.
As outlined in the editorial [3], there were some limitations in our current study. First, the retrospective study design may cause bias in patient selection and recall. Thus, propensity score matching analyses were performed to minimise known and unknown confounders. Second, only Chinese patients were included, thus, our findings may not be generalisable. Third, our patients needed a relatively low dosage of propranolol, probably due to their differences in body constitution and ethnics from other regions, which is consistent with the dosage of propranolol used in many previous studies from China [9, 10]. Lastly, all patients included in our study had been treated with propranolol. Therefore, it was impossible to identify whether carvedilol could also cause such a bidirectional effect on the patients' outcomes.
Regardless, our findings further support the window hypothesis of NSBBs in cirrhosis. However, we require more solid evidence to determine which is the right time to discontinue NSBBs.
京公网安备 11010802027423号