Sumeet K. Asrani M.D.

Levitsky, J., S. K. Asrani, M. Abecassis, R. Ruiz, L. W. Jennings and G. Klintmalm (2019). “External Validation of a Pre-Transplant Biomarker Model (REVERSE) Predictive of Renal Recovery after Liver Transplantation.” Hepatology Apr 19. [Epub ahead of print].

Full text of this article.

In patients with end stage liver disease, the ability to predict recovery of renal function following liver transplantation alone (LTA) remains elusive. However, several important clinical decisions depend on whether renal dysfunction is recoverable after LTA. We used a cohort of patients undergoing LT to independently validate a pre-LT model predictive of post-LTA renal recovery (REVERSE: high osteopontin (OPN) and tissue inhibitor of metalloproteinases-1 (TIMP-1) levels, age <57, no diabetes). Serum samples pre-LT and 4-12 weeks post-LTA (n=117) were analyzed for kidney injury proteins from 3 groups of recipients: (1) estimated GFR (eGFR)<30ml/min/1.73m(2) prior to LTA and <30 ml/min/1.73m(2) after LTA (irreversible acute kidney injury = iAKI), (2) eGFR<30ml/min/1.73m(2) prior to LTA and >50 ml/min/1.73m(2) after LTA (reversible AKI = rAKI) (3) eGFR>50 ml/min/1.73m(2) prior to LTA and >50 ml/min/1.73m(2) after LTA (no AKI = nAKI). In patients with elevated pre-LTA serum levels of OPN and TIMP-1, recovery of renal function correlated with decreases in the level of both proteins. At four weeks post-LT (n=77 subset), the largest decline in OPN and TIMP-1 was seen in the rAKI group. Validation of the REVERSE model in this independent dataset had high area under the curve (AUC) (0.78) in predicting full post-LT renal recovery (sensitivity 0.86, specificity 0.6, PPV 0.81, NPV 0.69). Our eGFR findings were confirmed using measured GFR (mGFR). CONCLUSION: The REVERSE model, derived from an initial training set combining novel plasma biomarkers and clinical characteristics, demonstrated excellent external validation performance characteristics in an independent patient cohort using serum samples. Among patients with kidney injury pre-LTA, the predictive ability of this simple tool may prove beneficial in clinical decision-making both prior to and following transplantation.

Devarbhavi, H., S. K. Asrani and P. S. Kamath (2019). “Reply to: “Alcohol-associated liver disease, not hepatitis B, is the major cause of cirrhosis in Asia”.” J Hepatol 70(5): 1033.

Full text of this article.

A recent multi-centre study from India with the specific aim of determining the aetiology of chronic liver diseases among new patients seen in an inpatient setting (n = 13,014) concluded that HBV was the most common cause of chronic liver disease (33.3%), followed by HCV (21.6%), alcohol (17.3%), and non-alcoholic fatty liver disease (12.8%). When stratified by presence of cirrhosis (n = 4,413), alcohol was the most common cause (34.3%); a significant limitation of this data was that more than 99% of patients with cirrhosis were decompensated and may not accurately reflect the burden of cirrhosis. Further there was significant regional heterogeneity within different regions of India; viral hepatitis B and C being more common in northern and eastern regions compared to southern regions where alcohol was more common. With increasing global coverage of universal vaccination against hepatitis B in India and the easy availability and affordability of direct-acting antivirals including provision for free treatments for hepatitis C in certain groups and regions, the burden of liver disease from B and C is expected to decrease in the future. Liver disease from alcohol is likely to increase in India in the absence of government policies aimed at reducing alcohol consumption. (Excerpt from text, responding to commentary on authors’ study, Burden of liver diseases in the world. J Hepatol 2019; 70(1): pp. 151-171.)

Agbim, U. and S. K. Asrani (2019). “Cirrhosis and the Acute Kidney Injury/Chronic Kidney Disease Continuum: The Path Chosen Matters.” Clin Gastroenterol Hepatol Apr 17. [Epub ahead of print].

Full text of this article.

The study [Cullaro G., Verna E.C., and Lai J.C.: Association between renal function pattern and mortality in patients with cirrhosis. Clin Gastroenterol Hepatol 2019] notably raised several important issues. First, renal dysfunction is exceedingly common and the temporal dimension of where the patient lies on the AKI/CKD continuum may matter. Although a mathematical model such as the MELD or MELD-Na may attempt to summarize AKI, CKD, or AKI/CKD by a singular serum creatine concentration or its surrogates, these may be 3 separate entities, each with their own trajectory. A young cirrhotic patient with a MELD score of 25 and a serum creatinine concentration of 2 mg/dL driven by AKI owing to volume depletion may have a different trajectory than an older cirrhotic patient with a similar MELD but volume depletion complicating underlying CKD. Overall, AKI exerts a greater influence in risk of mortality on CKD than it does on those with normal renal function. Intuitively thinking, this makes sense because any insult, whether small or large, on an otherwise diseased kidney can drive cirrhotics in an unfavorable direction. This is relevant given the increasing prevalence of CKD in this population, necessitating favorable strategies to avoid or mitigate further renal injury, thereby minimizing the potential for waitlist mortality. Hence, mechanisms and chronicity of renal dysfunction may be important even before an eventual transplant. Second, this study emphasized the need to measure renal function effectively. All serum creatinine–based equations overestimate GFR in the presence of renal dysfunction. Furthermore, several equations assume a stable GFR, which is not often the case in cirrhotic patients on the waiting list. Risk stratification remains paramount, requiring continual enhancement of tools. Several biomarkers, in addition to patient characteristics, currently are being evaluated to assess renal function. The operationalization of AKI, particularly in cirrhosis, has been problematic throughout the literature, and it will be necessary to formalize a consistent definition to measure the real effect of renal dysfunction. Furthermore, AKI represents a heterogeneous entity with a multitude of phenotypes (hypovolemic nephropathy, vasogenic nephropathy, acute tubular necrosis, hepatorenal syndrome), as does CKD, and it is unclear in this analysis how any particular AKI phenotypic insult influences risk. Finally, the study highlighted the importance of extrahepatic factors in determining mortality on the waitlist. Although in all comers, a mathematical model such as the MELD-Na score may be able to predict that a registrant with a higher MELD-Na score has an increased risk of mortality than someone with a lower MELD-Na score (e.g., score of 40 vs 6), the ability of any statistical model to parse out differences in patients with similar MELD scores is difficult. The presence of comorbid conditions, malnutrition and sarcopenia, infections, critical illness, and now pattern of renal dysfunction, all may play a role. (Excerpt from text of editorial, article in press, not paginated.)

Sundaram, V., R. Jalan, T. Wu, M. L. Volk, S. K. Asrani, A. S. Klein and R. J. Wong (2019). “Factors Associated with Survival of Patients With Severe Acute-On-Chronic Liver Failure Before and After Liver Transplantation.” Gastroenterology 156(5): 1381-1391.e1383.

Full text of this article.

BACKGROUND & AIMS: Liver transplantation for patients with acute-on-chronic liver failure (ACLF) with 3 or more failing organs (ACLF-3) is controversial. We compared liver waitlist mortality or removal according to model for end-stage liver disease (MELD) score vs ACLF category. We also studied factors associated with reduced odds of survival for 1 year after liver transplantation in patients with ACLF-3. METHODS: We analyzed data from the United Network for Organ Sharing (UNOS) from 2005 through 2016. We identified patients who were on the waitlist (100,594) and those who received liver transplants (50,552). Patients with ACLF were identified based on the European Association for the Study of the Liver-chronic liver failure criteria. Outcomes were evaluated with competing risks regression, Kaplan-Meier analysis, and Cox proportional hazards regression. RESULTS: Patients with ACLF-3 were more likely to die or be removed from the waitlist, regardless of MELD-sodium (MELD-Na) score, compared with the other ACLF groups; the proportion was greatest for patients with an ACLF-3 score and MELD-Na score below 25 (43.8% at 28 days). Mechanical ventilation at liver transplantation (hazard ratio [HR] 1.49; 95% confidence interval [CI] 1.22-1.84), donor risk index above 1.7 (HR 1.22; 95% CI 1.09-1.35), and liver transplantation within 30 days of listing (HR 0.89; 95% CI 0.81-0.98) were independently associated with survival for 1 year after liver transplantation CONCLUSIONS: In an analysis of data from the UNOS registry, we found high mortality among patients with ACLF-3 on the liver transplant waitlist, even among those with lower MELD-Na scores. So, certain patients with ACLF-3 have poor outcomes regardless of MELD-Na score. Liver transplantation increases odds of survival for these patients, particularly if performed within 30 days of placement on the waitlist. Mechanical ventilation at liver transplantation and use of marginal organs were associated with increased risk of death.

Kanwal, F., E. B. Tapper, C. Ho, S. K. Asrani, N. Ovchinsky, J. Poterucha, A. Flores, V. Ankoma-Sey, B. Luxon and M. Volk (2019). “Development of Quality Measures in Cirrhosis by the Practice Metrics Committee of the American Association for the Study of Liver Diseases.” Hepatology 69(4): 1787-1797.

Full text of this article.

Health care delivery is increasingly evaluated according to quality measures, yet such measures are underdeveloped for cirrhosis. The Practice Metrics Committee of the American Association for the Study of Liver Diseases was charged with developing explicit process-based and outcome-based measures for adults with cirrhosis. We identified candidate measures from comprehensive reviews of the literature and input from expert clinicians and patient focus groups. We conducted an 11-member expert clinician panel and used a modified Delphi method to systematically identify a set of quality measures in cirrhosis. Among 119 candidate measures, 46 were identified as important measures to define the quality of cirrhosis care, including 26 process measures, 7 clinical outcome measures, and 13 patient-reported outcome measures. The final process measures captured care processes for ascites (n = 5), varices/bleeding (n = 7), hepatic encephalopathy (n = 4), hepatocellular cancer (HCC) screening (n = 1), liver transplantation evaluation (n = 2), and other care (n = 7). Clinical outcome measures included survival, variceal bleeding and rebleeding, early-stage HCC, liver-related hospitalization, and rehospitalization within 7 and 30 days. Patient-reported outcome measures covered physical symptoms, physical function, mental health, general function, cognition, social life, and satisfaction with care. The final list of patient-reported outcomes was validated in 79 patients with cirrhosis from nine institutions in the United States. Conclusion: We developed an explicit set of evidence-based quality measures for adult patients with cirrhosis. These measures are a tool for providers and institutions to evaluate their care quality, drive quality improvement, and deliver high-value cirrhosis care. The quality measures are intended to be applicable in any clinical care setting in which care for patients with cirrhosis is provided.