RESUMO
BACKGROUND AND AIMS: It has been described that recompensation can improve prognosis in patients with cirrhosis. However, recompensation after transjugular intrahepatic portosystemic shunt (TIPS) has not been studied. We evaluated the impact of recompensation after TIPS on the risk of hepatocellular carcinoma (HCC) and death, and we compared it with compensated cirrhosis patients. METHODS: An observational study of consecutive patients with cirrhosis undergoing TIPS between 2008 and 2022 was performed. Baveno VII definition of recompensation was used including patients with or without diuretics/Hepatic encephalopathy prophylaxis. A prospective cohort of consecutive compensated cirrhosis patients was used for comparison. RESULTS: Overall, 208 patients with cirrhosis were included, 92 compensated and 116 decompensated who underwent TIPS. After 1 year, 24% achieved recompensation. Liver function (MELD 12 ± 5 vs. 15 ± 6; p = .049), LDL-cholesterol (97 mg/dL vs. 76 mg/dL, p = .018), white cell count (7.96 × 109/dL vs. 6.24 × 109/dL, p = .039) and platelets (129 × 109/dL vs. 101 × 109/dL, p = .039) were associated with recompensation. Recompensation was associated with a reduction in the risk of HCC (p = .020). Multivariable analysis showed that this risk was significantly higher in non-recompensated patients (p = .003) but no differences were observed in recompensated compared with compensated patients (p = .816). Similarly, decompensated patients presented lower survival rates (p = .011), while no differences were observed between recompensated and compensated patients (p = .677). CONCLUSIONS: Recompensation after TIPS has a clear impact on the incidence of HCC and death, with a similar prognosis than patients with compensated cirrhosis. Liver function is associated with recompensation, suggesting the importance of considering early TIPS in patients with indication.
RESUMO
INTRODUCTION: Caustic ingestion management could be improved with a diagnostic approach based on risk factors. This study aimed to develop an algorithm derived from predictive factors of a poor clinical course, to evaluate its diagnostic accuracy and resource consumption, and to compare it with 2 other approaches, a radiological one based on computed tomography and a classical one based on symptoms and endoscopy. METHODS: All patients older than 15 years presenting with caustic ingestion in our tertiary care hospital between 1995 and 2021 were prospectively included. Adverse outcome was defined as intensive care unit admission, emergency surgery, or death. Ingestion characteristics, symptoms, and laboratory and endoscopic findings were analyzed to determine the most relevant risk factors. Diagnostic accuracy and the number of examinations required were estimated and compared with the other 2 algorithms applied to our series. RESULTS: The sample included 532 cases of caustic ingestion, 13.2% (95% confidence interval [CI]: 10.3-16.0) of which had adverse outcomes. Volume and type of caustic substance; presence of symptoms and pharyngolaryngeal involvement; and neutrophilia, acidosis, and endoscopic injury were combined to develop an algorithm that would provide the highest diagnostic odds ratio (167.2; 95% CI: 71.9-388.7). Following this approach, half of the patients (50.6%; 95% CI: 46.2-55.1) would not require any examination and, overall, the need for endoscopy (20.0%; 95% CI: 16.4-23.5) and computed tomography (16.3%; 95% CI: 13.0-19.5) would be lower than that for the other 2 algorithms. DISCUSSION: A risk-based algorithm could improve caustic ingestion management by maintaining high diagnostic accuracy while reducing diagnostic test requirements.