Low-chloride versus high-chloride hypertonic solution for the treatment of subarachnoid hemorrhage-related complications (The ACETatE trial): study protocol for a pilot randomized controlled trial

Aneurysmal subarachnoid hemorrhage (aSAH) is a life-threatening condition that results from a rupture in a cerebral artery [1]. Risk factors include female gender, hypertension, smoking, and excessive alcohol consumption [2]. The overall incidence is estimated at 5–10 per 100,000 per year [3]. Although treatment has improved in recent years, aSAH still represents a potentially devastating clinical condition, with a reported in-hospital mortality of approximately 20% [4]. The clinical course is often complicated by aneurysm re-bleeding, cerebral vasospasm with delayed cerebral ischemia, cerebral edema, and systemic complications, such as cardiopulmonary failure, electrolyte dysregulation, infections, and acute kidney injury (AKI). Systemic complications account for up to 40% of all deteriorations during inpatient management [5]. AKI in aSAH patients is associated with a greater than four-fold increase in hospital mortality, along with a less favorable neurological recovery in survivors [6, 7].

Hyperosmolar therapy is a cornerstone in the treatment of aSAH-related cerebral edema, which occurs in approximately 20 to 30% of this patient population [8]. Osmotherapy is based on elevation of serum osmolality via hypernatremia (e.g., hypertonic NaCl) or the use of an osmotically active agent (e.g., mannitol) [9]. The goal of osmotherapy is to reduce intracranial pressure and improve cerebral blood flow by raising serum osmolarity and subsequently inducing excess fluid shifts from brain tissue into the blood stream. Hypertonic sodium chloride (NaCl) solutions, 3% to 23.4%, are commonly used to induce hypernatremia, but carry the risk of significant hyperchloremia.

Intravenous chloride load has been reported to cause negative effects on blood pressure, kidney blood flow, and fluid retention in animals [10] and, subsequently, in healthy human subjects [11, 12]. In a variety of critically ill patient populations, hyperchloremia has also been observed to directly correlate with AKI and mortality [13, 14]. Furthermore, comparative studies have noted an association between isotonic NaCl infusions and an increased risk of AKI and poor outcome in critically ill patients [15, 16]. Summarizing 21 prospective studies, a meta-analysis by Krajewski et al. [17] demonstrated that high Cl⁻ containing solutions increase the risk of AKI, the number of blood transfusions, and the need for mechanical ventilation in perioperative or critically ill patients. Surprisingly, a multicenter trial failed to show a benefit of balanced intravenous (IV) solutions, i.e., low Cl⁻ solution, over high Cl⁻ solutions (NaCl 0.9%) in a mixed population of critically ill patients [18]. However, even more recent publications did demonstrate favorable renal outcomes in non-critically ill [19] and critically ill [20] patients receiving balanced crystalloid solution when compared with NaCl 0.9%. Since hypertonic NaCl solutions are not balanced solutions, it is not surprising to encounter controversy regarding their role in the development of AKI [21, 22].

In a recently published retrospective analysis of SAH patients, we observed correlations between hyperchloremia and AKI, and between AKI and increased morbidity and mortality. We identified treatment with hypertonic NaCl as a risk factor for the development of hyperchloremia [7]. Nonetheless, we were not able to show a dose-response relationship between chloride load and subsequent development of hyperchloremia [7]; instead we found great variability in chloride exposure. The large variability appeared to be related to a high incidence of cerebral salt wasting syndrome in our study population, resulting in the need for extensive sodium and volume replacement with IV fluids, especially NaCl solutions at various concentrations [23].

Therefore, the question arises whether avoiding further worsening of hyperchloremia by differential use of hypertonic solutions might be a better strategy to prevent AKI in critically ill patients (with SAH) than simply replacing high Cl⁻ solutions with low Cl⁻ solutions in all cases.

Additional analyses of our recent study further support this approach [7]. The frequency of maximum chloride levels in patients without AKI increased up to a chloride concentration of 109 mmol/L and decreased thereafter (Fig. 1). On the contrary, the frequency of maximum Cl⁻ levels increased and remained elevated for serum Cl⁻ levels greater than 109 mmol/L in patients with AKI (Fig. 1). Moreover, only 8.5% of all AKI patients had maximum serum Cl⁻ concentrations < 109 mmol/L. By setting a threshold of 109 mmol/L, we would therefore exclude only a small proportion of AKI patients. Moreover, patients without hyperchloremia appear unlikely to benefit from preventive measures with low Cl⁻ solutions. Focusing on at-risk patients should allow us to reduce variability and consequently provide more stringent answers in this highly controversial area of AKI research.

To this end, our proposed trial described below intends to make use of the different Cl⁻ content in NaCl 23.4% or NaCl/Na-Acetate 16.4% (Table 1) and thereby to examine the causative role of hypertonic IV solutions in the development of hyperchloremia (and subsequently AKI) in patients with SAH. Our study design represents a double-blinded, double-dummy, pragmatic approach.

We propose a double-blind, double-dummy, prospective clinical trial to compare the effect of two hypertonic saline solutions with different Cl⁻ content on serum Cl⁻ levels in patients with cerebral edema due to aSAH. Our immediate goal is to collect high-quality prospective data regarding the adverse effects, effectiveness, and feasibility of our current hypertonic solution dosing strategies. In the future, we plan to design a larger randomized trial.

This is the first trial to prospectively assess the relationship between the Cl⁻ content of two hypertonic solutions with different Cl⁻ content and their effect on serum electrolytes as well as renal function.

The question whether balanced IV fluid solutions, i.e., low Cl⁻ solutions, improve patient outcome when compared with isotonic NaCl has been controversial for several years [30, 31]. Although some studies have demonstrated associations between higher IV chloride content and AKI, recent prospective studies have produced conflicting results [18‐20, 32]. Failure to focus on specific patient populations, i.e., patients at risk for AKI, could be one reason for these inconsistent results.

Hypertonic solutions are, by definition, non-balanced and represent an extreme case of IV Cl⁻ exposure. In a recent retrospective study, we have demonstrated a relationship between hyperchloremia, the use of hypertonic solutions, and the development of AKI in patients with SAH. Unfortunately, our data did not reveal a dose-response relationship between IV Cl⁻ exposure and AKI [7]. However, our findings have allowed us to identify a group of SAH patients who are at risk for AKI. As only 8.5% of our patients with maximal serum Cl⁻ concentrations < 109 mmol/L developed AKI, we consider serum Cl⁻ concentrations ≥ 109 mmol/L as a risk factor for AKI in patients with SAH, and propose a corresponding study design. So far, no clinical trial has based commencement of randomized treatment on the evolution or presence of a specific risk factor, such as hyperchloremia, when studying the effects of IV solutions with different Cl⁻ content.

The ACETatE trial will allow for a direct assessment of the dose-response relationship between the IV solution and serum concentrations, for both sodium and chloride. We are following a pragmatic trial design, as there is no clear, widely accepted protocol for use of hypertonic solutions in patients with SAH. Dose and frequency of osmotherapy will be at the discretion of the attending physician. The treatment for study patients will be nearly identical to that of any other patient with SAH in our ICU. Treatment will only differ by the Cl⁻ content of the hypertonic solution to which the patients will be randomized.

An important, potential limitation of our study is the comparison of two solutions with different volumes and Na⁺ concentrations. Although the total Na⁺ load will be lower in the more concentrated solution, the precise effects of 30 ml of 23.4% Na⁺ solution versus those of 50 ml of 16.4% Na⁺ solutions on serum osmolarity, serum sodium, and ICP remain unknown.

Another potential pitfall is insufficient enrollment due to the limitations of the informed consent process in critically ill patients [33]. First, patients are often unable to consent themselves due to the cerebral injury while a next of kin cannot be contacted in time. Second, patients or their families have great difficulty consenting for a trial given the seriousness of the illness and deciding on it within a relatively short period of time.

Even if we achieve our goal of patients, our power to detect the needed difference is only 63%. These calculations have dictated the primary goal of preventing extreme hyperchloremia, and not AKI, which is too rare to change in a small, single-center trial.

However, the ACETatE trial, with all the mentioned limitations and possible confounding circumstances, will follow the changes in serum electrolyte concentration as a function of IV exposure. This high-quality information will help us to design a large-scale clinical trial which is desperately needed in the debate about the optimal Cl⁻ content in the IV solution [30, 31].

This trial has been approved by the Emory University IRB (protocol version #4, dated 6/8/2017, clinicaltrials.gov #NCT03204955). Recruitment is ongoing.