Validation of a new prognostic model to predict short and medium-term survival in patients with liver cirrhosis

Machine Learning for Short-Term Mortality in Acute Decompensation of Liver Cirrhosis: Better than MELD Score

Prediction of short-term mortality in patients with acute decompensation of liver cirrhosis could be improved. We aimed to develop and validate two machine learning (ML) models for predicting 28-day and 90-day mortality in patients hospitalized with acute decompensated liver cirrhosis. We trained two artificial neural network (ANN)-based ML models using a training sample of 165 out of 290 (56.9%) patients, and then tested their predictive performance against Model of End-stage Liver Disease-Sodium (MELD-Na) and MELD 3.0 scores using a different validation sample of 125 out of 290 (43.1%) patients. The area under the ROC curve (AUC) for predicting 28-day mortality for the ML model was 0.811 (95%CI: 0.714- 0.907; p < 0.001), while the AUC for the MELD-Na score was 0.577 (95%CI: 0.435-0.720; p = 0.226) and for MELD 3.0 was 0.600 (95%CI: 0.462-0.739; p = 0.117). The area under the ROC curve (AUC) for predicting 90-day mortality for the ML model was 0.839 (95%CI: 0.776- 0.884; p < 0.001), while the AUC for the MELD-Na score was 0.682 (95%CI: 0.575-0.790; p = 0.002) and for MELD 3.0 was 0.703 (95%CI: 0.590-0.816; p < 0.001). Our study demonstrates that ML-based models for predicting short-term mortality in patients with acute decompensation of liver cirrhosis perform significantly better than MELD-Na and MELD 3.0 scores in a validation cohort.

Peri-interventional LiMAx test for liver ablation - A feasibility study

BACKGROUND

Local ablation is an important treatment for liver cancer in cases of cirrhosis. Liver failure may occur after ablation, especially in advanced liver diseases. Currently, there is no standardization for peri-interventional risk assessment. The liver maximum capacity test (LiMAx) is used to assess functional liver capacity, but there is a lack of exploration of its use in this context.

OBJECTIVE

The aim of this study was to retrospectively evaluate the usefulness of peri-interventional LiMAx measurements in patients with primary or secondary liver cancer who underwent ablation treatment.

METHODS

A LiMAx test was performed at 24 hours pre- and postablation in 49 patients. Blood parameters were collected to determine liver function using MELD and ALBI scores. The results of the LiMAx test were related with these scores and to critical postintervention LiMAx values.

RESULTS

LiMAx values correlated strongly with MELD and ALBI scores before the intervention and reflected the change in liver function, as shown by an increase in scores after the intervention. Notably, LiMAx values decreased during the intervention. AUC analysis for patients at risk of reaching a critical liver level after the intervention showed a cutoff value of 186μg/kg/h.

CONCLUSIONS

The LiMAx test may be a valuable tool in liver ablation for both peri-interventional monitoring of liver function and preintervention risk assessment.

Clinical implications of hepatic structure and function evaluation based on vibration-controlled transient elastography and liver maximum function capacity test in patients with nonalcoholic fatty liver disease

BACKGROUND AND AIMS

Transient elastography [vibration-controlled transient elastography (VCTE)] noninvasively guides risk stratification in patients with nonalcoholic fatty liver disease (NAFLD). Patients with nonalcoholic steatohepatitis (NASH) and fibrosis can be identified using the FAST-score. The liver maximum function test (LiMAx) could be helpful in more precise risk stratification. This pilot study evaluated VCTE, FAST-score, and LiMAx in NAFLD patients.

METHODS

NAFLD patients prospectively underwent VCTE and LiMAx. The cutoffs for high fibrosis risk were 9.3/9.6 kPa (M/XL-probe) and 331 dB/m for steatosis. A FAST-score greater than 0.67 was used to identify patients with NASH and LiMAx values below 315 μg/kg/h for impaired liver function.

RESULTS

In total, 57 NAFLD patients (BMI 32 ± 6 kg/m2; 60% diabetes) were included. High risk for fibrosis and steatosis was observed in 26/57 and 28/57 cases, respectively. Overall, 19/57 patients presented impaired liver function. However, 14/26 of patients with a high risk for fibrosis had impaired liver function compared to 5/31 of those without (P = 0.0026). Similarly, 12/18 patients at high risk for NASH had impaired liver function compared to 7/39 without (P < 0.001). The subgroup with diabetes had a liver stiffness a factor of 1.8 higher, FAST-score was 0.13 higher and LiMAx values were 66 μg/kg/h lower compared to nondiabetics.

CONCLUSION

There is a significant correlation between the functional liver capacity (LiMAx) and the structural liver assessment by VCTE. In cases with high liver stiffness or FAST-score, low LiMAx results may identify NAFLD patients at risk for disease progression and reduce the risk of false-positive categorization.

Validation of Dynamic Aspartate-to-Alanine Aminotransferase Ratio for Predicting Liver Disease Mortality

The dynamic aspartate-to-alanine aminotransferase ratio (dAAR) was developed recently to predict the risk of incident chronic liver disease among the Nordic adult population; however, the dAAR has not been externally validated in other ethnic cohorts. Therefore, we aimed to examine the predictive ability of dAAR for liver disease mortality in the South Korean adult population. As a population-based cohort study, we used the National Health Screening Cohort database, which included adult individuals who underwent standardized medical examinations between 2002 and 2003 in South Korea. The primary endpoint was liver disease mortality, defined as death due to liver disease. Liver disease mortality was evaluated between 2004 and 2015 (12 years). Analysis of data from 512,749 adults showed that 4,052 (0.8%) individuals died due to liver disease. On receiver operating characteristic (ROC) analyses, the area under curve for alanine aminotransferase (ALT), aspartate-to-ALT ratio (AAR), and dAAR for liver disease mortality were 0.74, 0.55, and 0.81, respectively. The cutoff point of dAAR was determined to be 0.72 on ROC analysis, using the Youden index method. On competing risk analysis using the Fine and Gray model, the dAAR > 0.72 group demonstrated a 4.43-fold higher rate of liver disease mortality (subdistribution hazard ratio: 4.43, 95% confidence interval: 4.11, 4.77; P < 0.001) after adjustment for covariates. Conclusion: The performance of dAAR in predicting liver disease mortality was better than that of AAR or ALT in South Korea. Our study suggests that dAAR scores can potentially be used for screening and predicting liver disease mortality among the general Korean population.