Serum Angiopoietin-2 Predicts Mortality and Kidney Outcomes in Decompensated Cirrhosis

Abstract

Acute kidney injury in decompensated cirrhosis has limited therapeutic options, and novel mechanistic targets are urgently needed. Angiopoietin-2 is a context-specific antagonist of Tie2, a receptor that signals vascular quiescence. Considering the prominence of vascular destabilization in decompensated cirrhosis, we evaluated Angiopoietin-2 to predict clinical outcomes. Serum Angiopoietin-2 was measured serially in a prospective cohort of hospitalized patients with decompensated cirrhosis and acute kidney injury. Clinical characteristics and outcomes were examined over a 90-day period and analyzed according to Angiopoietin-2 levels. Primary outcome was 90-day mortality. Our study included 191 inpatients (median Angiopoietin-2 level 18.2 [interquartile range 11.8, 26.5] ng/mL). Median Model for End-Stage Liver Disease (MELD) score was 23 [17, 30] and 90-day mortality was 41%. Increased Angiopoietin-2 levels were associated with increased mortality (died 21.9 [13.9, 30.3] ng/mL vs. alive 15.2 [9.8, 23.0] ng/mL; P < 0.001), higher Acute Kidney Injury Network stage (stage I 13.4 [9.8, 20.1] ng/mL vs. stage II 20.0 [14.1, 26.2] ng/mL vs. stage III 21.9 [13.0, 29.5] ng/mL; P = 0.002), and need for renal replacement therapy (16.5 [11.3, 23.6] ng/mL vs. 25.1 [13.3, 30.3] ng/mL; P = 0.005). The association between Angiopoietin-2 and mortality was significant in unadjusted and adjusted Cox regression models (P ≤ 0.001 for all models), and improved discrimination for mortality when added to MELD score (integrated discrimination increment 0.067; P = 0.001). Conclusion: Angiopoietin-2 was associated with mortality and other clinically relevant outcomes in a cohort of patients with decompensated cirrhosis with acute kidney injury. Further experimental study of Angiopoietin/Tie2 signaling is warranted to explore its potential mechanistic and therapeutic role in this population.

Figure 1:

Initial serum Angiopoietin-2 levels as predictors of clinical outcomes Key: AKIN (Acute Kidney Injury Network), HRS (hepatorenal syndrome), ATN (acute tubular necrosis)

Figure 2:

Correlation between initial serum Angiopoietin-2 and laboratory values. All variables are log transformed. Key: INR (international normalized ratio), AST (aspartate aminotransferase), ALT (alanine aminotransferase), WBC (white blood cell count)

Figure 3:

90-day probability of survival by tertile of serum Angiopoietin-2 Patients were divided by serum Angiopoietin-2 tertile at enrollment: Low Tertile (Angiopoietin −2 < 13.5 ng/mL; n = 64), Mid Tertile (Angiopoietin −2 between 13.5 and 23.0 ng/mL; n = 64), and High Tertile (Angiopoietin −2 > 23.0 ng/mL; n = 63).

Figure 4:

Forest plot of Angiopoietin-2’s hazard ratio for 90-day mortality Key: MELD (Model for End-Stage Liver Disease), CLIF-C ACLF (Chronic Liver Failure Consortium Acute-on-Chronic Liver Failure Score), AKIN (Acute Kidney Injury Network), AKI (acute kidney injury) Model 1: adjusted for MELD score and age Model 2: adjusted for MELD score, age, and infection Model 3: excluding prior liver transplant recipients (n = 5) Model 4: excluding hepatocellular carcinoma (n = 20) Model 5: composite outcome of liver transplant or death

Figure 5:

Change in serum Angiopoietin-2 at 30 days among: those who received an interval liver transplant (left panel; n = 12) and those who did not (right panel; n = 28). * p = 0.002 ** p = 0.001 Those who received an interval liver transplant saw a decrease in Angiopoietin-2 from 28.2 [16.6, 33.1] ng/mL to 13.5 [6.1, 17.2] ng/mL. Those who did not receive an interval liver transplant saw an increase in Angiopoietin-2 from 13.5 [9.6, 19.7] ng/mL to 19.7 [11.1, 26.9] ng/mL.