Heart Failure May Not Be Causal Factor in Hepatorenal Syndrome With Acute Kidney Injury

By Patricia A. Sheiner MD

July 01, 2022

NEW YORK (Reuters Health) - A new study sheds light on the hemodynamic changes that accompany the progression of cirrhosis and suggests that cardiac dysfunction might not be causal factor in hepatorenal syndrome with acute kidney injury, researchers say.

Karen Vagner Danielsen and colleagues from the Copenhagen University Hospital Hvidovre in Denmark used magnetic resonance imaging (MRI) and hepatic portal pressure to characterize the spectrum of changes in peripheral, splanchnic, renal, and cardiac hemodynamics in 27 patients with compensated cirrhosis, 60 patients with decompensated cirrhosis, and 27 healthy controls.

The mean age of patients with cirrhosis was 60 years; 61% were male. The mean Model for End-Stage Liver Disease (MELD)-Na score was 15, and the mean hepatic venous pressure gradient (HVPG) was 15 mmHg. HVPG, MELD-Na score and mortality rate increased with the severity of liver disease.

Thirty-three patients had responsive ascites, 16 had refractory ascites, and 11 had hepatorenal syndrome-acute kidney injury (HRS-AKI).

Compared with healthy controls, patients with compensated cirrhosis showed higher azygos venous flow (p <0.05), higher azygos venous flow fraction of cardiac index (CI, p <0.05), and a higher hepatic artery flow fraction of CI (p <0.05). Although 74% of compensated cirrhotics had clinical portal hypertension, there were no differences from healthy controls in absolute flow in the superior mesenteric, splenic, femoral, renal, or hepatic arteries or the portal vein. Heart rate was higher in compensated cirrhotics than in healthy controls, but there were no differences in cardiac output (CO), mean arterial pressure (MAP), systemic vascular resistance (SVR) or global longitudinal strain(GLS).

Compared with compensated cirrhotics, decompensated cirrhotic patients with responsive ascites showed higher absolute flow in the superior mesenteric artery (p = 0.01), lower renal flow fraction of CI (p <0.01) and higher portal venous flow (p = 0.04). CI (p <0.01) and QTc interval (p<0.05) were significantly higher in the patients with responsive ascites. MAP (p <0.01) and SVR (p <0.001) were lower in the decompensated patients with responsive ascites. There was no difference in any parameters between patients with responsive and refractory ascites.

The 11 patients with HRS-AKI were found to have higher CO (9.9 L/min vs 7.3 L/min in patients with ascites vs 6.3 in compensated cirrhosis, p<0.01) and lower renal blood flow (233 ml/min vs 347 ml/min vs 393 ml/min, respectively, p≤0.02). Absolute renal flow was reduced by 43% compared with controls (410 ml/min vs. 233 mL/min, p <0.001).

Compared with all other groups, the patients with HRS-AKI had significantly higher CO and CI, and lower MAP and SVR.

Five of the 11 patients with HRS-AKI had acute on chronic liver failure (ACLF). The six with HRS-AKI without ACLF were hyperdynamic, with increased CO values similar to those found in decompensated patients and patients with refractory ascites, while patients with both HRS-AKI and ACLF had additional increases in CO and CI. The five patients with combined HRS-AKI and ACLF appeared to have higher peripheral flow (femoral) than the HRS-AKI group without ACLF. Both the absolute renal flow and the renal flow fraction of CI were lowest in patients with combined HRS-AKI and ACLF.

In the total patient population, the absolute renal flow, as well as the renal flow fraction of CI, correlated with serum creatinine and serum cystatin C (p <0.0001). Both creatinine, cystatin C, and renin correlated with MELD-Na score but not with HVPG.

Total mean follow-up time from inclusion until death or termination of the study was 726 days. Twenty-six patients died during follow-up. In a Cox proportional hazard model comprising CO, renal artery, superior mesenteric artery, and femoral artery blood flow, CO was the only parameter associated with death (hazard ratio 1.77, p <0.01). In a separate analysis, both MELD-Na score (hazard ratio 1.19, p <0.001) and HVPG (hazard ratio 1.10, p <0.05) were associated with death.

"HRS-AKI patients were characterized by severely compromised renal perfusion despite a high CO, which challenges the prevailing pathophysiological hypothesis of cardiac dysfunction being an important preceding factor for HRS-AKI," the authors noted.

In an editorial, Dr. Florence Wong from the University of Toronto in Ontario notes that the reported increase in CO, rather than decrease, in patients with advanced cirrhosis - and even in those with HRS-AKI- was an unexpected finding. However, she points out, the CO output was only one measurement in time, and cardiac dysfunction in cirrhosis comprises both systolic and diastolic dysfunction.

Also, Dr. Wong said, "Patients in the current study did not have type 1 HRS (HRS1) in the strictest sense, as their mean serum creatinine of 204 umol/L would not have fulfilled the diagnostic criteria of HRS1, and therefore had renal dysfunction that was less severe than in patients from previous studies."

She concludes, "While the findings of the current study are interesting, the results on 11 patients with HRS-AKI certainly cannot explain the complex role of cardiac dysfunction in the pathogenesis of HRS-AKI in cirrhosis, but it will certainly set the impetus to explore other aspects of cirrhotic cardiomyopathy in the development of HRS-AKI."

SOURCE: American Journal of Gastroenterology, online June 3, 2022.

processing....