Positive fluid balance was associated with mortality in patients with acute-on-chronic liver failure: A cohort study

Highlights

• International retrospective cohort of 333 patients with cirrhosis and organ failures.

• Fluid balance at 7 days post admission was associated with higher hospital mortality.

• Fluid balance may be a therapeutic target to improve these patients' outcomes.

Abstract

Purpose

We aimed to study the effect of FB in the outcomes of critically-ill patients with cirrhosis.

Materials

Retrospective analysis of all adult consecutive admissions of patients with cirrhosis and organ failures to the Intensive Care Unit (ICU) at Curry Cabral Hospital (Lisbon, Portugal) and University of Alberta Hospital (Edmonton, Canada) on 08/2013–08/2017. Primary exposure was FB at 3 and 7 days post ICU admission. Primary endpoint was hospital mortality.

Results

Amongst 333 patients, median age was 56 years and 67.6% were men. Median MELD, APACHEII, CLIF-SOFA, and CLIF-C-ACLF scores on ICU admission were 27, 28, 14, and 54, respectively. ICU and hospital mortality rates were 33.0% and 49.2%, respectively. While median FB at 3 days post ICU admission (+5.46 l vs. +6.62 l; P = 0.74) was not associated with hospital mortality, higher median FB at 7 days post ICU admission (+13.50 l vs. +6.90 l; P = 0.036) was associated with higher hospital mortality. This association remained significant (OR 95%CI = 1.04 [1.01;1.07] per each l) after adjustment for confounders (age, ascites, infection, lactate, and number of organ failures).

Conclusions

FB may be a therapeutic target that helps to improve the outcomes of patients with acute-on-chronic liver failure. This data may inform future clinical trials.

Introduction

Patients with cirrhosis and organ failures (acute-on-chronic liver failure (ACLF)) admitted to the intensive care unit (ICU) have high short-term mortality rates, especially if they do not receive a liver transplant (LT) [[1], [2], [3]].

Virtually all critically-ill patients receive variable amounts of fluid therapy during their ICU stay. Fluid resuscitation of critically-ill patients encompasses 4 phases: rescue, optimization, stabilization, and de-escalation [4]. The duration and sequence of these phases varies widely amongst different patients.

During resuscitation, intravenous fluids are generally administered as early as possible and targeted to appropriate physiologic endpoints, commonly a mean arterial pressure of ≥65-70 mmHg and a serum lactate of ≤2 mmol/L. However, following achievement of hemodynamic stability, patients often continue to receive variable amounts of fluid therapy which ultimately may lead to fluid overload (FO). This additional fluid has the potential to accumulate in the tissues, especially in the context of increased capillary permeability such as in sepsis, and lead to end-organ edema and dysfunction [5]. In fact, several studies have shown that a persistent cumulative positive fluid balance (FB) negatively impacts the outcomes of critically-ill patients [[6], [7], [8]]. For example, in patients with sepsis, a positive FB was associated with worse 28-day survival [6]. Furthermore, in patients with acute lung injury, a positive FB was associated with further impaired oxygenation and prolonged need for mechanical ventilation [7]. Finally, in patients with acute renal failure, a positive FB was associated with worse 60-day survival [8].

In patients with cirrhosis and portal hypertension, the hemodynamic assessment has specific features. The splanchnic vasodilatation often results in effective hypovolemia which leads to the activation of the renin-angiotensin-aldosterone system causing further sodium and water retention [9]. The ensuing renal vasoconstriction, coupled with the potentially increased intra-abdominal pressure due to the ascites, may lead to acute renal failure and further contribute to the fluid accumulation. In the context of critical illness, circulatory failure in these patients often results in greater reduction in the mean arterial pressure and tissue impaired perfusion [10]. Therefore, the resuscitation of patients with cirrhosis and portal hypertension is often challenging, especially when it comes to monitor early signs of FO.

There is lack of data about the impact of FB on the outcomes of patients with ACLF. We hypothesized that FO may contribute to higher short-term mortality in patients with ACLF. Accordingly, the objectives of this study were: (1) characterize the FB of patients with cirrhosis and organ failures admitted to the ICU; and (2) study the association of FB with these patient's outcomes.