Risk prediction scores for acute on chronic liver failure development and mortality

Applying the Milan models to setting analytical performance specifications - considering all the information

Analytical performance specifications (APS) are used for decisions about the required analytical quality of pathology tests to meet clinical needs. The Milan models, based on clinical outcome, biological variation, or state of the art, were developed to provide a framework for setting APS. An approach has been proposed to assign each measurand to one of the models based on a defined clinical use, physiological control, or an absence of quality information about these factors. In this paper we propose that in addition to such assignment, available information from all models should be considered using a risk-based approach that considers the purpose and role of the actual test in a clinical pathway and its impact on medical decisions and clinical outcomes in addition to biological variation and the state-of-the-art. Consideration of APS already in use and the use of results in calculations may also need to be considered to determine the most appropriate APS for use in a specific setting.

Prognostic models in end stage liver disease

Cirrhosis is a major cause of death worldwide, and is associated with significant health care costs. Even if milestones have been recently reached in understanding and managing end-stage liver disease (ESLD), the disease course remains somewhat difficult to prognosticate. These difficulties have already been acknowledged already in the past, when scores instead of single parameters have been proposed as valuable tools for short-term prognosis. These standard scores, like Child Turcotte Pugh (CTP) and model for end-stage liver disease (MELD) score, relying on biochemical and clinical parameters, are still widely used in clinical practice to predict short- and medium-term prognosis. The MELD score, which remains an accurate, easy-to-use, objective predictive score, has received significant modifications over time, in order to improve its performance especially in the liver transplant (LT) setting, where it is widely used as prioritization tool. Although many attempts to improve prognostic accuracy have failed because of lack of replicability or poor benefit with the comparator (often the MELD score or its variants), few scores have been recently proposed and validated especially for subgroups of patients with ESLD, as those with acute-on-chronic liver failure. Artificial intelligence will probably help hepatologists in the near future to fill the current gaps in predicting disease course and long-term prognosis of such patients.

EASL Clinical Practice Guidelines on acute-on-chronic liver failure

Acute-on-chronic liver failure (ACLF), which was described relatively recently (2013), is a severe form of acutely decompensated cirrhosis characterised by the existence of organ system failure(s) and a high risk of short-term mortality. ACLF is caused by an excessive systemic inflammatory response triggered by precipitants that are clinically apparent (e.g., proven microbial infection with sepsis, severe alcohol-related hepatitis) or not. Since the description of ACLF, some important studies have suggested that patients with ACLF may benefit from liver transplantation and because of this, should be urgently stabilised for transplantation by receiving appropriate treatment of identified precipitants, and full general management, including support of organ systems in the intensive care unit (ICU). The objective of the present Clinical Practice Guidelines is to provide recommendations to help clinicians recognise ACLF, make triage decisions (ICU vs. no ICU), identify and manage acute precipitants, identify organ systems that require support or replacement, define potential criteria for futility of intensive care, and identify potential indications for liver transplantation. Based on an in-depth review of the relevant literature, we provide recommendations to navigate clinical dilemmas followed by supporting text. The recommendations are graded according to the Oxford Centre for Evidence-Based Medicine system and categorised as 'weak' or 'strong'. We aim to provide the best available evidence to aid the clinical decision-making process in the management of patients with ACLF.

Association between statin exposure and short-term mortality in patients with high-grade acute-on-chronic liver failure

Background & Aims

Acute-on-chronic liver failure (ACLF) carries a high short-term mortality for patients with cirrhosis. Prior literature suggests that statin exposure may reduce the likelihood of ACLF events. However, it is unclear if statin exposure is associated with ACLF-related mortality. This study sought to determine the association between statin use and short-term mortality among patients hospitalised with ACLF.

Methods

This was a retrospective cohort study of Veterans Health Administration (VHA) patients diagnosed with cirrhosis between 2008 and 2021 and hospitalised with high-grade (2 or 3) ACLF. Patients were stratified into those with and without continuous statin exposure for at least 90 days prior to hospitalisation. Multivariable logistic regression models were created to determine the adjusted association between statin exposure and 28-day and 90-day mortality. Categorical statin dose exposure, converted to simvastatin equivalents, was also explored.

Results

A total of 11,731 patients with cirrhosis hospitalised with Grade 2 or 3 ACLF were included in the analytic cohort, 26% of whom had statin exposure. In adjusted logistic regression models, statin use was associated with 18% lower odds of ACLF-related 28-day mortality (odds ratio [OR] 0.82, 95% CI 0.73-0.93, p = 0.001) and 24% lower odds of 90-day mortality (OR 0.76, 95% CI 0.68-0.86, p <0.001). Increasing statin dose exposure was also associated with further reductions in 90-day mortality (e.g. OR 0.81, 95% CI 0.70-0.93 for 10-40 mg vs. 0 mg and OR 0.72, 95% CI 0.60-0.87 for 80 mg vs. 0 mg, p <0.001).

Conclusions

In this large, retrospective cohort study, statin exposure before high-grade ACLF hospitalisation was associated with reduced odds of 28-day and 90-day mortality in patients with cirrhosis. A statin dose-dependent reduction in 90-day ACLF-related mortality was also observed.

Impact and Implications

Statins have been identified as a class of medications with potential beneficial effects for patients with cirrhosis. In this large, retrospective cohort study of patients with cirrhosis who seek care at the Veterans Health Administration, statin use was associated with a decrease in short term (28-day and 90-day) mortality as a result of acute-on-chronic liver failure. Future prospective studies are needed to further clarify the relative safety and efficacy of statin therapy in reducing morbidity and mortality associated with acute-on-chronic liver failure in patients with cirrhosis.

Acute on chronic liver failure: prognostic models and artificial intelligence applications

Critically ill patients presenting with acute on chronic liver failure (ACLF) represent a particularly vulnerable population due to various considerations surrounding the syndrome definition, lack of robust prospective evaluation of outcomes, and allocation of resources such as organs for transplantation. Ninety-day mortality related to ACLF is high and patients who do leave the hospital are frequently readmitted. Artificial intelligence (AI), which encompasses various classical and modern machine learning techniques, natural language processing, and other methods of predictive, prognostic, probabilistic, and simulation modeling, has emerged as an effective tool in various areas of healthcare. These methods are now being leveraged to potentially minimize physician and provider cognitive load and impact both short-term and long-term patient outcomes. However, the enthusiasm is tempered by ethical considerations and a current lack of proven benefits. In addition to prognostic applications, AI models can likely help improve the understanding of various mechanisms of morbidity and mortality in ACLF. Their overall impact on patient-centered outcomes and countless other aspects of patient care remains unclear. In this review, we discuss various AI approaches being utilized in healthcare and discuss the recent and expected future impact of AI on patients with ACLF through prognostic modeling and AI-based approaches.

Type of Infection Is Associated with Prognosis in Acute-on-Chronic Liver Failure: A National Veterans Health Administration Study

BACKGROUND

Acute-on-chronic liver failure (ACLF) is a syndrome in patients with cirrhosis with high short-term mortality. Infection is a frequent precipitant of ACLF; however, it is unclear if prognosis varies by difference infectious sources. To address this knowledge gap, we utilized a large national database of patients with cirrhosis.

METHODS

This was a retrospective cohort study of patients with cirrhosis in the Veterans Health Administration between 2008 and 2016. First ACLF hospitalizations were identified and infections were classified using validated algorithms, categorized as bacteremia, fungal, spontaneous bacterial peritonitis (SBP), pyelonephritis/urinary tract infection, or skin and soft tissue/musculoskeletal infection (SST/MSK). Inverse probability treatment weighing for infection-associated ACLF followed by multivariable logistic regression was used to evaluate the association between infection type and 90-day mortality.

RESULTS

A total 22,589 ACLF hospitalizations were included, 3998 (17.7%) of which had ACLF grade 3. Infection was associated with 12,405 (54.9%) of ACLF hospitalizations. In regression models, SBP was associated with a 1.79-fold increased odds of 90-day mortality vs. no infection (95% confidence interval [CI] 1.58-2.02, p < 0.001), whereas SST/MSK infections had a lower relative odds of mortality (odds ratio 0.48, 95% CI 0.42-0.53, p < 0.001). There was a significant interaction between infection category and ACLF grade on the outcome of 90-day mortality (p < 0.001).

CONCLUSIONS

The impact of infection on short-term mortality in ACLF varies depending on the source of infection. This has relevance for ACLF prognostication and challenges previous notions that bacterial infection invariably worsens prognosis among all patients with ACLF.

Artificial liver support systems for hepatitis B virus-associated acute-on-chronic liver failure: A meta-analysis of the clinical literature

To evaluate the short-term and long-term survival efficacy of an artificial liver support system (ALSS) in patients with acute-on-chronic liver failure (ACLF). A systematic search was performed for relevant published data in PubMed, Web of Science and Cochrane Library databases. Studies that evaluated the efficacy of ALSS in patients with ACLF and provided the short-term or long-term survival rate were included. A total of 10 studies involving 3685 patients were included in this analysis. The pooled 28-day survival rate and 90-day survival rate were 68.7% (95% CI: 64.5%-72.9%) and 53.4% (95% CI: 45.5%-61.4%), respectively. The pooled estimates of the OR for the 28-day and 90-day survival rates between the ALSS group and the control group were 1.91 (95% CI: 1.21-3.04) and 1.41 (95% CI: 1.17-1.70), respectively. Subgroup analysis showed that patients treated with lower levels of TBIL and MELD scores had a higher 28-day survival rate (χ²  = 15.75, p < 0.01; χ²  = 13.80, p < 0.01). The present meta-analysis suggests that ALSS treatment could remarkably improve short-term survival rates in HBV-ACLF patients, which implies that treatment with ALSS may help to reduce high mortality. Further prospective randomized trials are needed to validate these findings.

Accurate long-term prediction of death for patients with cirrhosis

BACKGROUND AND AIMS

Cirrhosis is a major cause of death and is associated with extensive health care use. Patients with cirrhosis have complex treatment choices due to risks of morbidity and mortality. To optimally counsel and treat patients with cirrhosis requires tools to predict their longer-term liver-related survival. We sought to develop and validate a risk score to predict longer-term survival of patients with cirrhosis.

APPROACH AND RESULTS

We conducted a retrospective cohort study of adults with cirrhosis with no major life-limiting comorbidities. Adults with cirrhosis within the Veterans Health Administration were used for model training and internal validation, and external validation used the OneFlorida Clinical Research Consortium. We used four model-building approaches including variables predictive of cirrhosis-related mortality, focused on discrimination at key time points (1, 3, 5, and 10 years). Among 30,263 patients with cirrhosis ≤75 years old without major life-limiting comorbidities and complete laboratory data during the baseline period, the boosted survival tree models had the highest discrimination, with 1-year, 3-year, 5-year, and 10-year survival rates of 0.77, 0.81, 0.84, and 0.88, respectively. The 1-year, 3-year, and 5-year discrimination was nearly identical in external validation. Secondary analyses with imputation of missing data and subgroups by etiology of liver disease had similar results to the primary model.

CONCLUSIONS

We developed and validated (internally and externally) a risk score to predict longer-term survival of patients with cirrhosis. This score would transform management of patients with cirrhosis in terms of referral to specialty care and treatment decision-making for non-liver-related care.

External Validation of the FIPS Score for Post-TIPS Mortality in a National Veterans Affairs Cohort

BACKGROUND

The Freiburg index of post-TIPS survival (FIPS) score was recently demonstrated to improve prediction of post-TIPS mortality relative to existing standards. As this score was derived from a German cohort over an extended time period, it is unclear if performance will translate well to other settings. This study aimed to externally validate the FIPS score in a large Veterans Affairs (VA) cohort over two separate eras of TIPS-related care.

METHODS

This was a retrospective cohort study of patients with cirrhosis who underwent TIPS placement in the VA from 2008 to 2020. Cox regression models for post-TIPS survival were constructed using FIPS, MELD, MELD-Na, or CTP scores as predictors. Discrimination (Harrell's C) and calibration (joint tests of calibration curve slope and intercept) were evaluated for each score. A stratified analysis was performed for time periods between 2008-2013 and 2014-2020.

RESULTS

The cohort of 1,274 patients was 97.3% male with mean age 60.9 years and mean MELD-Na 14. The FIPS score demonstrated the highest overall discrimination versus MELD, MELD-Na, and CTP (0.634 vs. 0.585, 0.626, 0.612, respectively). However, in the modern treatment era (2014-2020), the FIPS score performed similarly to MELD-Na. Additionally, the FIPS score demonstrated poor calibration at one-month and six-month post-TIPS timepoints (joint p = 0.04 and 0.004, respectively). MELD, MELD-Na, and CTP were well-calibrated at each timepoint (each joint p > 0.05).

CONCLUSION

The FIPS score performed similarly to MELD-Na in the modern TIPS treatment era and demonstrated regions of poor calibration. Future models derived with contemporary data may improve prediction of post-TIPS mortality.

Major Shifts in Outpatient Cirrhosis Care Delivery Attributable to the COVID-19 Pandemic: A National Cohort Study

The coronavirus disease 2019 (COVID-19) pandemic has disrupted health care delivery in the United States, with increased reliance on telemedicine visits as opposed to in-person outpatient appointments. We used national data to evaluate shifts in modes of hepatology outpatient care for patients with cirrhosis during the pandemic. This was a retrospective cohort study among U.S. veterans with cirrhosis. We used linear regression to evaluate absolute and percentage changes from baseline in hepatology in-person visits and telemedicine visits from January 1, 2020, to August 11, 2020. The proportion of in-person and telemedicine visits were plotted geographically to demonstrate state-level shifts in care delivery over time. Patient-level characteristics in the pre-COVID and during-COVID periods were also compared. We identified 5,618 in-person and 6,210 telemedicine hepatology visits among patients with cirrhosis. In-person visits significantly declined (-16.0% per week; 95% confidence interval [CI] -20.7, -11.2; P < 0.001), while telemedicine visits significantly increased (61.3% per week; 95% CI 45.1, 77.5; P < 0.001) in the early during-COVID period. At the U.S. state level, we found that nearly all states experienced a significant shift toward telemedicine over the course of several weeks. Patients over the age of 70 years and Black patients were less likely to receive telemedicine visits in the pre-COVID period (each P < 0.05), although these differences were eliminated in the during-COVID periods. Conclusion: Among patients with cirrhosis, hepatology outpatient care delivery has shifted heavily toward telemedicine due to COVID-19. This occurred across the United States, and changes have been sustained through August 2020. Expanded telemedicine visits among older patients and Black patients may reflect dedicated efforts to increased access to care among these groups.

Liver transplantation in patients with liver failure: Twenty years of experience from China

Liver transplantation (LT) is the only effective method of treating end-stage liver disease, such as various types of liver failure. China has the largest number of patients with hepatitis B virus-related disease, which is also the main cause of liver failure. From the first LT performed in 1977, and especially over the past two decades, LT has experienced rapid development as a result of continuous research and innovation in China. China performs the second-highest number of LTs every year worldwide, and the quality of LT continues to improve. Starting January 1, 2015, all donor's livers have been from deceased donors and familial donors. Thus, China entered into a new era of LT. However, LT is still a challenging procedure in China. In this review, we introduced the brief history of LT in China, the epidemiology, aetiology and clinical outcomes of LT for liver failure in China and summarized the experience of LT from Chinese LT surgeons and scholars. The future perspectives of LT were also discussed, and it is expected that China's LT research could be further integrated elsewhere in the world.

Statin exposure is associated with reduced development of acute-on-chronic liver failure in a Veterans Affairs cohort

BACKGROUNDS & AIMS

There is a need to identify therapies that prevent the development of acute-on-chronic liver failure (ACLF) in patients with cirrhosis. This study sought to evaluate the association between statin exposure and the risk of developing ACLF in a large national cohort of patients with cirrhosis.

METHODS

We performed a retrospective cohort study of patients diagnosed with cirrhosis within the Veterans Health Administration from 2008 and 2018. Patients were stratified into 3 groups based on statin exposure (statin naïve, existing statin user, and new statin initiator). Cox proportional hazards regression models with inverse probability treatment weighting and marginal structural models were utilized to comprehensively address potential confounding in estimating the association between time-updated statin exposure and first occurrence of high-grade ACLF.

RESULTS

The cohort included 84,963 patients, of whom 26.9% were on a statin at baseline. A total of 8,558 (10.1%) patients with cirrhosis were hospitalized with high-grade ACLF over a median follow-up time of 51.6 months (IQR 27.5-81.4). Time-updated statin use was associated with a significant reduction in the hazard of developing ACLF (hazard ratio [HR] 0.62, 95% CI 0.59-0.65, p <0.001). Increasing doses of statin were associated with progressively reduced hazard of developing ACLF (HR 0.75, 95% CI 0.66-0.86, p <0.001 for <20 mg vs. 0 mg of time-updated statin exposure, in simvastatin equivalents; HR 0.61, 95%, CI 0.58-0.64, p <0.001 for >20 mg vs. 0 mg statin exposure). Furthermore, every additional 5 months of statin exposure was associated with a 9% reduced hazard of high-grade ACLF (HR 0.91, 95% CI 0.90-0.92, p <0.001).

CONCLUSIONS

In this large, retrospective, cohort study in patients with cirrhosis, statin use was significantly associated with reduced development of high-grade ACLF.

LAY SUMMARY

Statin therapy has been shown to have numerous beneficial effects in patients with chronic liver disease. This study demonstrated a strong association between statin therapy and a reduced risk of acute-on-chronic liver failure development in patients with cirrhosis. The results of this study support the promising role that statins may play in future prevention of acute-on-chronic liver failure in patients with cirrhosis.

Recompensation factors for patients with decompensated cirrhosis: a multicentre retrospective case-control study

OBJECTIVES

We aimed to evaluate recompensation factors among patients with decompensated cirrhosis.

DESIGN

A multicentre retrospective case-control study was conducted. Data were collected from and compared between groups of patients with recompensated and acute decompensated cirrhosis. Univariable and multivariable logistic regressions were used to select indicators associated with recompensation among patients with decompensated cirrhosis with different complications. A decision tree with 10-fold cross-validation was used to develop the model to identify patients with recompensation. We followed the transparent reporting of a multivariable prediction model for individual prognosis or diagnosis (TRIPOD) guideline for development and reporting of the new model.

SETTING

The study was conducted in six tertiary public hospitals in Chongqing, China.

PARTICIPANTS

This study included 3953 patients with decompensated cirrhosis.

RESULTS

In the total sample of included patients, there were 553 patients with recompensation and 3400 patients with acute decompensation, including 1158 patients with gastrointestinal bleeding, 1715 patients with a bacterial infection, 104 patients with hepatic encephalopathy and 423 patients with ascites. The most relevant indicator of recompensation selected by the decision tree model was albumin, with a threshold of 40 g/L. Total protein, haemoglobin, basophil percentage, alanine aminotransferase, neutrophil-to-lymphocyte ratio and diabetes were also selected to subsequently distinguish patients. The terminal nodes with a probability of recompensation was 0.89. The overall accuracy rate of the model was 0.92 (0.91-0.93), and it exhibited high specificity (86.9%) and sensitivity (92.6%).

CONCLUSIONS

The occurrence of recompensated cirrhosis could be identified by albumin, total protein, haemoglobin, basophil percentage, alanine aminotransferase, neutrophil-to-lymphocyte ratio and diabetes. These simple variables may help clinicians develop a treatment plan to encourage patients with decompensated cirrhosis to recompensate.

Acute-on-chronic liver failure: update on pathogenesis, therapeutic targets, predictive models, and liver transplantation

PURPOSE OF REVIEW

Acute-on-chronic liver failure (ACLF) is a clinical syndrome in patients with chronic liver disease that is associated with multiple organ failures and a high short-term mortality. Systemic inflammation is suggested to play a key role in its pathogenesis, although the precise causative mechanism is unknown. The purpose of this review is to present and discuss new findings related to: mechanisms underlying ACLF, therapeutic targets, risk prediction models for developing ACLF, and liver transplantation for ACLF.

RECENT FINDINGS

Recent studies of ACLF pathophysiology classified the immunosuppressive phenotype in monocytes. Investigation of therapeutic strategies identified inhibition of toll-like receptor-4 (TLR-4) and glutamine synthetase (GLUL) as potential targets. Recent studies identified novel risk prediction models for developing ACLF and enhanced our understanding of liver transplantation for ACLF to guide clinicians in determining that patients will benefit from transplantation.

SUMMARY

Improved knowledge on the pathogenesis of ACLF and identification of TLR-4 and GLUL may lead to clinical trials to study the efficacy of these novel therapeutic targets for patients with ACLF. Liver transplantation is the only current treatment for ACLF. Given the limited availability of donor organs, recent studies have identified ACLF patients who may merit the highest waitlist priority.

The Predictive Role of Model for End-Stage Liver Disease-Lactate and Lactate Clearance for In-Hospital Mortality Among a National Cirrhosis Cohort

The burden of cirrhosis hospitalizations is increasing. The admission Model for End-Stage Liver Disease-lactate (MELD-lactate) was recently demonstrated to be a superior predictor of in-hospital mortality compared with MELD in limited cohorts. We identified specific classes of hospitalizations where MELD-lactate may be especially useful and evaluated the predictive role of lactate clearance. This was a retrospective cohort study of 1036 cirrhosis hospitalizations for gastrointestinal bleeding, infection, or other portal hypertension-related indications in the Veterans Health Administration where MELD-lactate was measured on admission. Performance characteristics for in-hospital mortality were compared between MELD-lactate and MELD/MELD-sodium (MELD-Na), with stratified analyses of MELD categories (≤15, >15 to <25, ≥25) and reason for admission. We also incorporated day 3 lactate levels into modeling and tested for an interaction between day 1 MELD-lactate and day 3 lactate clearance. MELD-lactate had superior discrimination for in-hospital mortality compared with MELD or MELD-Na (area under the curve [AUC] 0.789 versus 0.776 versus 0.760, respectively; P < 0.001) and superior calibration. MELD-lactate had higher discrimination among hospitalizations with MELD ≤15 (AUC 0.763 versus 0.608 for MELD, global P = 0.01) and hospitalizations for infection (AUC 0.791 versus 0.674 for MELD, global P < 0.001). We found a significant interaction between day 1 MELD-lactate and day 3 lactate clearance; heat maps were created as clinical tools to risk-stratify patients based on these clinical data. MELD-lactate had significantly superior performance in predicting in-hospital mortality among patients hospitalized for infection and/or with MELD ≤15 when compared with MELD or MELD-Na. Incorporating day 3 lactate clearance may further improve prognostication.

Patient Frailty Is Independently Associated With the Risk of Hospitalization for Acute-on-Chronic Liver Failure

There is significant interest in identifying risk factors associated with acute-on-chronic liver failure (ACLF). In transplant candidates, frailty predicts wait-list mortality and posttransplant outcomes. However, the impact of frailty on ACLF development and mortality is unknown. This was a retrospective study of US veterans with cirrhosis identified between 2008 and 2016. First hospitalizations were characterized as ACLF or non-ACLF admissions. Prehospitalization patient frailty was ascertained using a validated score based on administrative coding data. We used logistic regression to investigate the impact of an increasing frailty score on the odds of ACLF hospitalization and short-term ACLF mortality. Cox regression was used to analyze the association between frailty and longterm survival from hospitalization. We identified 16,561 cirrhosis hospitalizations over a median follow-up of 4.19 years (interquartile range, 2.47-6.34 years). In adjusted models, increasing frailty score was associated with significantly increased odds of ACLF hospitalization versus non-ACLF hospitalization (odds ratio, 1.03 per point; 95% CI 1.02-1.03; P < 0.001). By contrast, frailty score was not associated with ACLF 28- or 90-day mortality (P = 0.13 and P = 0.33, respectively). In an adjusted Cox analysis of all hospitalizations, increasing frailty scores were associated with poorer longterm survival from the time of hospitalization (hazard ratio, 1.02 per 5 points; 95% confidence interval, 1.01-1.03; P = 0.004). Frailty increases the likelihood of ACLF hospitalization among patients with cirrhosis, but it does not impact short-term ACLF mortality. These findings have implications for clinicians caring for frail outpatients with cirrhosis, including tailored follow-up, risk mitigation strategies, and possible expedited transplant evaluation.

Development, Validation, and Evaluation of a Simple Machine Learning Model to Predict Cirrhosis Mortality

Importance

Machine-learning algorithms offer better predictive accuracy than traditional prognostic models but are too complex and opaque for clinical use.

Objective

To compare different machine learning methods in predicting overall mortality in cirrhosis and to use machine learning to select easily scored clinical variables for a novel cirrhosis prognostic model.

Design, Setting, and Participants

This prognostic study used a retrospective cohort of adult patients with cirrhosis or its complications seen in 130 hospitals and affiliated ambulatory clinics in the integrated, national Veterans Affairs health care system from October 1, 2011, to September 30, 2015. Patients were followed up through December 31, 2018. Data were analyzed from October 1, 2017, to May 31, 2020.

Exposures

Potential predictors included demographic characteristics; liver disease etiology, severity, and complications; use of health care resources; comorbid conditions; and comprehensive laboratory and medication data. Patients were randomly selected for model development (66.7%) and validation (33.3%). Three different statistical and machine learning methods were evaluated: gradient descent boosting, logistic regression with least absolute shrinkage and selection operator (LASSO) regularization, and logistic regression with LASSO constrained to select no more than 10 predictors (partial pathway model). Predictor inclusion and model performance were evaluated in a 5-fold cross-validation. Last, the predictors identified in the most parsimonious (the partial path) model were refit using maximum-likelihood estimation (Cirrhosis Mortality Model [CiMM]), and its predictive performance was compared with that of the widely used Model for End Stage Liver Disease with sodium (MELD-Na) score.

Main Outcomes and Measures

All-cause mortality.

Results

Of the 107 939 patients with cirrhosis (mean [SD] age, 62.7 [9.6] years; 96.6% male; 66.3% white, 18.4% African American), the annual mortality rate ranged from 8.8% to 15.3%. In total, 32.7% of patients died within 3 years, and 46.2% died within 5 years after the index date. Models predicting 1-year mortality had good discrimination for the gradient descent boosting (area under the receiver operating characteristics curve [AUC], 0.81; 95% CI, 0.80-0.82), logistic regression with LASSO regularization (AUC, 0.78; 95% CI, 0.77-0.79), and the partial path logistic model (AUC, 0.78; 95% CI, 0.76-0.78). All models showed good calibration. The final CiMM model with machine learning-derived clinical variables offered significantly better discrimination than the MELD-Na score, with AUCs of 0.78 (95% CI, 0.77-0.79) vs 0.67 (95% CI, 0.66-0.68) for 1-year mortality, respectively (DeLong z = 17.00; P < .001).

Conclusions and Relevance

In this study, simple machine learning techniques performed as well as the more advanced ensemble gradient boosting. Using the clinical variables identified from simple machine learning in a cirrhosis mortality model produced a new score more transparent than machine learning and more predictive than the MELD-Na score.

Early Transplantation in Acute on Chronic Liver Failure: Who and When

Purpose of Review

Acute on chronic liver failure (ACLF) is a unique syndrome that afflicts patients with chronic liver disease and results in high short-term mortality, > 50% at 28 days in patients with severe ACLF (grade 3, ACLF-3). Given this prognosis, there is an urgent need to understand risk factors for this condition, as well as selection criteria for patients who may benefit from liver transplantation (LT).

Recent Findings

Several studies have identified risk factors for developing ACLF, including higher model for end-stage liver disease score, anemia, and morbid obesity, as well ACLF mortality, such as infection, increasing organ failures, and higher white blood cell count. Prognostic tools are now available as online calculators. Regarding LT in ACLF, data suggest that even patients with ACLF-3 may do well after LT, with 1-year survival > 80% in several studies. Improvement in organ failures prior to LT, higher donor quality, and lack of mechanical ventilation further improve outcomes. Importantly, ACLF-3 patients may have higher short-term wait list mortality than patients listed status-1a, suggesting that increased LT prioritization may be warranted.

Summary

ACLF is a high-mortality condition that frequently responds well to LT. Ongoing efforts to understand the natural history of ACLF and predictors of improved post-LT survival will facilitate LT criteria for this condition, which may ultimately include increased LT prioritization for selected patients.

Models for acute on chronic liver failure development and mortality in a veterans affairs cohort

BACKGROUND AND PURPOSE

The diagnosis of acute on chronic liver failure (ACLF) carries a high short-term mortality, making early identification of at-risk patients crucial. To date, there are no models that predict which patients with compensated cirrhosis will develop ACLF, and limited models exist to predict ACLF mortality. We sought to create novel risk prediction models using a large North American cohort.

METHODS

We performed a retrospective study of 75,922 patients with compensated cirrhosis from the Veterans Outcomes and Costs Associated with Liver Disease (VOCAL) dataset. Using 70% derivation/30% validation sets, we identified ACLF patients using the Asian Pacific Association of Liver (APASL) definition. Multivariable logistic regression was used to derive prediction models (called VOCAL-Penn) for developing ACLF at 3, 6, and 12 months. We then created prediction models for ACLF mortality at 28 and 90 days.

RESULTS

The VOCAL-Penn models for ACLF development had very good discrimination [concordance (C) statistics of 0.93, 0.92, and 0.89 at 3, 6, and 12 months, respectively] and calibration. The mortality models also had good discrimination at 28 and 90 days (C statistics 0.89 and 0.88, respectively), outperforming the Model for End-stage Liver Disease (MELD), MELD-sodium, and the APASL ACLF Research Consortium ACLF scores.

CONCLUSION

We have developed novel tools for predicting development of ACLF in compensated cirrhosis patients, as well as for ACLF mortality. These tools may be used to proactively guide patient follow-up, prognostication, escalation of care, and transplant evaluation. Receiver operating characteristic (ROC) curves for predicting development of APASL ACLF at 3 months (a), 6 months (b), and 1 year (c).

Combining the serum lactic acid level and the lactate clearance rate into the CLIF-SOFA score for evaluating the short-term prognosis of HBV-related ACLF patients

BACKGROUND

Acute-on-chronic liver failure due to hepatitis B virus infection is a subtype of liver failure. The aim of the current study was to investigate the clinical significance of serum lactic acid in combination with the Chronic Liver Failure-Sequential Organ Failure Assessment score (CLIF-SOFA) for evaluating the short-term prognosis of HBV-related ACLF patients.

RESEARCH DESIGN AND METHODS

The serum lactic acid level, the score model for end-stage liver disease (MELD), as well as the CLIF-SOFA of inpatients with HBV-related ACLF who were admitted to the Liver Disease Center of the First Affiliated Hospital of Fujian Medical University between 2009 and 2017 were analyzed.

RESULTS

Three hundred and ninety-one HBV-related ACLF inpatients were measured. In the survival group, the lactate clearance rate measured over the course of 1 week post-admission was significantly higher than that measured for the death group. The area under the curve (AUC) for predicting the prognosis of HBV-related ACLF patients in short-term (within 3-month) with a baseline lactic acid level was 0.776.

CONCLUSIONS

Prediction performance of the short-term prognosis of HBV-related ACLF patients by combining the lactic acid level into the CLIF-SOFA score was significantly improved in comparison to using the CLIF-SOFA score alone.