Predictors of Acute Kidney Injury Resolution and Associated Clinical Outcomes Among Hospitalized Patients with...

Background: Acute kidney injury (AKI) is one of the common complications of liver cirrhosis. It occurs in nearly 20% of patients with cirrhosis who are hospitalized. Prior literature demonstrated that the AKI occurrence in patients with cirrhosis is independently associated with higher mortality. However, there are data assessing predictors and outcomes of AKI resolution in hospitalized patients with cirrhosis. Therefore, the aim of the current study was to identify clinical predictors of AKI resolution among inpatients with cirrhosis that are easily obtained and to evaluate the clinical outcomes of those patients. Methods: The current study is a retrospective cohort of patients with cirrhosis who were hospitalized and had AKI between 2012 and 2020 at a tertiary referral center. Patients included in this study were identified using the International Classification of Diseases 9 codes and then they were manually verified by two independent chart reviewers. AKI was classified according to the AKI Network (AKIN) serum creatinine (Cr) criteria, with AKIN resolution defined as AKIN stage 1 or lower at the time of discharge, while unresolved AKIN was defined as AKIN stage 2 or 3 at the time of discharge. For univariate analysis, Fisher's exact and the two-sample T-test were utilized. For multivariable analysis, stepwise logistic regression was performed to evaluate variables associated with AKIN resolution. Survival curves were estimated and compared using the Kaplan-Meier method and Log-Rank Test. A p-value cutoff of 0.05 was used for statistical significance. Results: Between 2012 and 2020, there were 140 patients who were included (59% males). The majority of patients had viral hepatitis (54%) as the cirrhosis etiology with 80% of them having hepatitis C virus. Most patients had fluid-responsive AKI (49%), and stage 1 AKIN (69%). In terms of outcomes, the majority of patients (117 patients; 84%) had AKIN resolution at the time of discharge. In the multivariable analysis, after adjusting for clinical meaningful variables, our study shows that higher albumin value at the time of admission (adjusted Odds Ratio "aOR" = 3.28; p = 0.01) and non-metabolic dysfunction-associated steatotic liver disease (non-MASLD) cirrhosis (aOR = 9.43; p < 0.01) were variables associated with higher odds of AKIN resolution at the time of discharge. Conversely, we show that a higher Cr value at the time of admission was associated with lower odds of AKIN resolution at the time of discharge (aOR = 0.31; p < 0.01). When evaluating mortality, patients with unresolved AKIN at the time of discharge had higher rates of in-hospital mortality (p < 0.01) compared to those with resolved AKIN. Survival curve analyses using the Kaplan-Meier method indicated that patients with resolved AKIN experienced higher 90-day survival rates (p < 0.01). Additionally, those with resolved AKIN demonstrated greater transplant-free survival compared to patients with unresolved AKIN at both the 1-year (p = 0.04) and 3-year (p < 0.01) follow-ups. Conclusions: When evaluating clinical predictors of AKIN resolution in admitted patients with cirrhosis, our study showed that a higher admission albumin value and non-MASLD etiology of cirrhosis were associated with higher odds of AKIN resolution at the time of discharge. Conversely, a higher admission Cr value was associated with lower odds of AKIN resolution at the time of discharge. We also demonstrate that AKIN resolution during index admission was associated with improved short- and long-term transplant-free survival (up to 3 years). Our findings warrant external validation in larger cohorts to further evaluate the impact of inpatient AKI resolution on cirrhosis outcomes. Our findings can help clinicians predict AKIN outcomes and encourage more aggressive management of AKI, especially in high-risk patients, which can improve mortality.