Transjugular intrahepatic portosystemic shunt in the management of ascites and hepatorenal syndrome

Improving Clinical Decisions in IR: Interpretable Machine Learning Models for Predicting Ascites Improvement after Transjugular Intrahepatic Portosystemic Shunt Procedures

PURPOSE

To evaluate the potential of interpretable machine learning (ML) models to predict ascites improvement in patients undergoing transjugular intrahepatic portosystemic shunt (TIPS) placement for refractory ascites.

MATERIALS AND METHODS

In this retrospective study, 218 patients with refractory ascites who underwent TIPS placement were analyzed. Data on 29 demographic, clinical, and procedural features were collected. Ascites improvement was defined as reduction in the need of paracentesis by 50% or more at the 1-month follow-up. Univariate statistical analysis was performed. Data were split into train and test sets. Feature selection was performed using a wrapper-based sequential feature selection algorithm. Two ML models were built using support vector machine (SVM) and CatBoost algorithms. Shapley additive explanations values were calculated to assess interpretability of ML models. Performance metrics were calculated using the test set.

RESULTS

Refractory ascites improved in 168 (77%) patients. Higher sodium (Na; 136 mEq/L vs 134 mEq/L; P = .001) and albumin (2.91 g/dL vs 2.68 g/dL; P = .03) levels, lower creatinine levels (1.01 mg/dL vs 1.17 mg/dL; P = .04), and lower Model for End-stage Liver Disease (MELD) (13 vs 15; P = .01) and MELD-Na (15 vs 17.5, P = .002) scores were associated with significant improvement, whereas main portal vein puncture was associated with a lower improvement rate (P = .02). SVM and CatBoost models had accuracy ratios of 83% and 87%, with area under the curve values of 0.83 and 0.87, respectively. No statistically significant difference was found between performances of the models in DeLong test (P = .3).

CONCLUSIONS

ML models may have potential in patient selection for TIPS placement by predicting the improvement in refractory ascites.

[Acute Kidney Injury and Hepatorenal Syndrome]

Acute kidney injury (AKI) is common in patients with liver cirrhosis, occurring in 13-20% of patients hospitalized with decompensated cirrhosis, and is significantly associated with the prognosis. The development and progression of AKI is an independent predictive factor for mortality in these patients. If AKI develops, the renal function declines progressively even if AKI is improved later, the patients have a poorer prognosis compared to those who have not developed AKI. In addition, in patients without appropriate treatment or no improvement with the initial treatment, AKI often progress to hepatorenal syndrome (HRS), which is associated with significant morbidity and mortality. Therefore, early detection and appropriate management for the development of AKI is very important in these patients. Recently, there have been significant revisions in the diagnostic criteria and treatment of AKI and HRS; this manuscript reviews these changes.

Development of an automated phenotyping algorithm for hepatorenal syndrome

OBJECTIVE

Hepatorenal Syndrome (HRS) is a devastating form of acute kidney injury (AKI) in advanced liver disease patients with high morbidity and mortality, but phenotyping algorithms have not yet been developed using large electronic health record (EHR) databases. We evaluated and compared multiple phenotyping methods to achieve an accurate algorithm for HRS identification.

MATERIALS AND METHODS

A national retrospective cohort of patients with cirrhosis and AKI admitted to 124 Veterans Affairs hospitals was assembled from electronic health record data collected from 2005 to 2013. AKI was defined by the Kidney Disease: Improving Global Outcomes criteria. Five hundred and four hospitalizations were selected for manual chart review and served as the gold standard. Electronic Health Record based predictors were identified using structured and free text clinical data, subjected through NLP from the clinical Text Analysis Knowledge Extraction System. We explored several dimension reduction techniques for the NLP data, including newer high-throughput phenotyping and word embedding methods, and ascertained their effectiveness in identifying the phenotype without structured predictor variables. With the combined structured and NLP variables, we analyzed five phenotyping algorithms: penalized logistic regression, naïve Bayes, support vector machines, random forest, and gradient boosting. Calibration and discrimination metrics were calculated using 100 bootstrap iterations. In the final model, we report odds ratios and 95% confidence intervals.

RESULTS

The area under the receiver operating characteristic curve (AUC) for the different models ranged from 0.73 to 0.93; with penalized logistic regression having the best discriminatory performance. Calibration for logistic regression was modest, but gradient boosting and support vector machines were superior. NLP identified 6985 variables; a priori variable selection performed similarly to dimensionality reduction using high-throughput phenotyping and semantic similarity informed clustering (AUC of 0.81 - 0.82).

CONCLUSION

This study demonstrated improved phenotyping of a challenging AKI etiology, HRS, over ICD-9 coding. We also compared performance among multiple approaches to EHR-derived phenotyping, and found similar results between methods. Lastly, we showed that automated NLP dimension reduction is viable for acute illness.

Prior to the oral therapy, what do we know about HCV-4 in Egypt: a randomized survey of prevalence and risks using data mining computed analysis

Hepatitis C virus (HCV) affects over 180 million people worldwide and it's the leading cause of chronic liver diseases and hepatocellular carcinoma. HCV is classified into seven major genotypes and a series of subtypes. In general, HCV genotype 4 (HCV-4) is common in the Middle East and Africa, where it is responsible for more than 80% of HCV infections. Although HCV-4 is the cause of approximately 20% of the 180 million cases of chronic hepatitis C worldwide, it has not been a major subject of research yet. The aim of the current study is to survey the morbidities and disease complications among Egyptian population infected with HCV-4 using data mining advanced computing methods mainly and other complementary statistical analysis. Six thousand six hundred sixty subjects, aged between 17 and 58 years old, from different Egyptian Governorates were screened for HCV infection by ELISA and qualitative PCR. HCV-positive patients were further investigated for the incidence of liver cirrhosis and esophageal varices. Obtained data were analyzed by data mining approach. Among 6660 subjects enrolled in this survey, 1018 patients (15.28%) were HCV-positive. Proportion of infected-males was significantly higher than females; 61.6% versus 38.4% (P=0.0052). Around two-third of infected-patients (635/1018; 62.4%) were presented with liver cirrhosis. Additionally, approximately half of the cirrhotic patients (301/635; 47.4%) showed degrees of large esophageal varices (LEVs), with higher variceal grade observed in males. Age for esophageal variceal development was 47±1. Data mining analysis yielded esophageal wall thickness (>6.5 mm), determined by conventional U/S, as the only independent predictor for esophageal varices. This study emphasizes the high prevalence of HCV infection among Egyptian population, in particular among males. Egyptians with HCV-4 infection are at a higher risk to develop cirrhotic liver and esophageal varices. Data mining, a new analytic technique in medical field, shed light in this study on the clinical importance of esophageal wall thickness as a useful predictor for risky esophageal varices using decision tree algorithm.

Predictive factors for improvement of ascites after transjugular intrahepatic portosystemic shunt in patients with refractory ascites

AIM

The aim of this study was to investigate the predictive factors for the response of ascites to a transjugular intrahepatic portosystemic shunt (TIPS) and the impact of improvement of ascites on the overall prognosis of patients with cirrhosis and refractory ascites.

METHODS

Forty-seven consecutive patients with liver cirrhosis who underwent TIPS for refractory ascites were studied retrospectively. The mean follow-up period was 615 ± 566 days.

RESULTS

Thirty-six of the patients (77%) were responders at 4 weeks after TIPS (early responders) and 37 (79%) were responders at 8 weeks after TIPS. Of the 11 non-responders at 4 weeks, four showed an improvement of ascites at 8 weeks. Multivariate analysis showed that only the serum creatinine level before TIPS was an independent predictor of an early response. The cumulative survival rate of early responders was significantly higher than that of non-responders. The survival of patients grouped according to creatinine level was better in patients with serum creatinine of 1.9 mg/dL or less than in those with serum creatinine of more than 1.9 mg/dL.

CONCLUSION

A low serum creatinine level in patients with refractory ascites is associated with an early response to TIPS. An early response of ascites to TIPS provides better survival. A serum creatinine level below 1.9 mg/dL is required for a good response to TIPS.

Transjugular intrahepatic portosystemic shunt versus paracentesis plus albumin in patients with refractory ascites who have good hepatic and renal function: a prospective randomized trial

BACKGROUND

Transjugular intrahepatic portosystemic shunt (TIPS) has recently been reported to be effective in the treatment of cirrhotic patients with refractory ascites. However, the clinical utility of TIPS in the subset of refractory ascitic patients with good hepatic and renal function is uncertain. The aim of this study was to compare the efficacy of TIPS to that of large-volume paracentesis in cirrhotic patients with refractory ascites who have good hepatic and renal function.

METHODS

Sixty cirrhotic patients with refractory ascites who presented with a Child-Pugh score of <11, serum bilirubin of <3 mg/dl and creatinine of <1.9 mg/dl were assigned randomly to TIPS (n = 30) or large-volume paracentesis plus albumin (n = 30). The primary endpoint was survival. The secondary endpoints were response to treatment and development of hepatic encephalopathy.

RESULTS

The baseline characteristics were similar in the two groups. Seventeen patients treated with TIPS and 21 treated with paracentesis died during the study period. The cumulative probabilities of survival at 1 and 2 years were 80 and 64% in the TIPS group and 49 and 35% in the paracentesis group (p < 0.005). TIPS was significantly superior to paracentesis in the control of ascites (p < 0.005). Treatment failure was more frequent in the paracentesis group, whereas the frequency of hepatic encephalopathy was greater in the TIPS group.

CONCLUSIONS

In cirrhotic patients with refractory ascites who have good hepatic and renal function, TIPS improves survival and provides better control of ascites than large-volume paracentesis.

Hepatorenal syndrome: a comprehensive overview for the critical care nurse

Over the past 50 years, the pathophysiology and features of the hepatorenal syndrome have been illuminated. The syndrome can be divided into 2 distinct clinical patterns: a rapidly progressive renal failure with an extremely poor prognosis (type 1) and a slow progressive renal failure that correlates with the degree of cirrhosis (type 2). Although our understanding of hepatorenal syndrome continues to grow, our current methods of treating this condition remain limited in their effectiveness. The only definitive therapy is liver transplantation. This is a review of the definition, pathophysiology, and current recommendations for management of hepatorenal syndrome with the critical care nurse in mind.

Fifteen years' experience with transjugular intrahepatic portosystemic shunt (TIPS) using bare stents: retrospective review of clinical and technical aspects

PURPOSE

The authors present a retrospective analysis of a large series of patients who underwent transjugular intrahepatic portosystemic shunt (TIPS) placement.

MATERIALS AND METHODS

Between March 1992 and December 2006, 658 patients were referred to our centre for TIPS placement. Indications for the procedure were digestive tract bleeding (52.8%), refractory ascites (35.3%), preservation of portal vein patency prior to liver transplantation (3.0%) and thrombosis of the suprahepatic veins (2.3%). Other indications (6.6%) included pleural ascites, portal thrombosis and hepatorenal and hepatopulmonary syndromes. All patients were evaluated with colour Doppler ultrasonography and in a few cases with computed tomography. The portal system was punctured under sonographic guidance. Wallstent, Palmaz and Nitinol thermosensitive stents were used. Embolisation of persistent varices was performed in 6.8% of cases.

RESULTS

Technical success was 98.9%. During a 1,500-day follow-up, the cumulative incidence of stent revision was 25.7% (Nitinol), 32.9% (Wallstent) and 1.8% (Palmaz). Mortality rates were 31.1%, 38.5% and 56.4%, respectively. The technical complications included six cases of heart failure, six of haematobilia, three of stent migration, two of intrahepatic haematoma and one of haemoperitoneum. Eight patients with severe portosystemic encephalopathy (PSE) were treated with a reduction stent.

CONCLUSIONS

TIPS placement is safe and effective and may act as a bridge to liver transplantation. Ultrasonography plays a fundamental role in the preliminary assessment, in portal vein puncture and during the follow-up. Stent patency is satisfactory.

Renal dysfunction in the perioperative liver transplant period

PURPOSE OF REVIEW

Acute kidney failure in the perioperative liver transplant recipients results in an increased hospital length of stay, acute rejection, infection rate and overall mortality. Thus, it is of great importance to be able to recognize, prevent and treat kidney injury.

RECENT FINDINGS

Immediate post liver transplant kidney dysfunction is increased in those with pretransplant kidney failure, hepato-renal syndrome, intraoperative hypotension, intraoperative hypovolemia, aprotinin use and those requiring transfusion of more units of blood products or needing to return to the operating room.

SUMMARY

To date, avoiding risky clinical situations, maintaining homeostasis and a multidisciplinary approach to care have been reasonable approaches to decrease the incidence of postoperative acute kidney injury.

The transjugular route: the key hole to the liver world

Portal hypertensive complications are major causes of morbidity and mortality in patients with liver cirrhosis. The advent of the transjugular route with its minimal access allows non-surgical management of portal hypertension, therapy of venous complications of liver transplantation, monitoring of therapy for portal hypertension, hepatic venous pressure gradient and is also the major route to treat hepatic venous obstruction syndromes. In addition, the transjugular route is a safe route to perform a liver biopsy (transjugular liver biopsy) and allows retrograde evaluation of the portal vein. All these procedures can be combined in the same session. These hepatic interventional radiological skills should be incorporated into the expertise of the liver team in specialised hepatological centres, particularly in liver transplant centres as they are especially useful in improving outcomes of cirrhotic patients on the liver transplantation waiting list. A limitation in achieving this goal, could be the number of experienced radiologists, but hepatologists can be trained, at least for the most simple procedures (transjugular liver biopsy and hepatic venous pressure gradient). This would allow wider applicability and use of these diagnostic and therapeutic techniques, all through a 2 mm hole in the neck--the key hole to the liver world.