Balanced forced‐diuresis as a renal protective approach in cardiac surgery: Secondary outcomes of electrolyte changes

Abstract Objectives Forced‐diuresis during cardiopulmonary bypass (CPB) can be associated with significant electrolyte shifts. This study reports on the serum electrolyte changes during balanced forced‐diuresis with the RenalGuard® system (RG) during CPB. Methods Patients at risk of acute kidney injury (AKI)—(history of diabetes &/or anaemia, e‐GFR 20–60 ml/min/1.73 m2, anticipated CPB time >120 min, Log EuroScore >5)—were randomized to either RG (study group) or managed as per current practice (control group). Results The use of RG reduced AKI rate (10% for RG and 20.9% in control, p = .03). Mean urine output was significantly higher in the RG group during surgery (2366 ± 877 ml vs. 765 ± 549 ml, p < .001). The serum potassium levels were maintained between 3.96 and 4.97 mmol/L for the RG group and 4.02 and 5.23 mmol/L for the controls. Median potassium supplemental dose was 60 (0–220) mmol (RG group) as compared to 30 (0–190) mmol for control group over first 24 h (p < .001). On Day 1 post‐op, there were no significant differences in the serum sodium, potassium, calcium, magnesium, phosphate, and chloride levels between the two groups. Otherwise, postoperative clinical recovery was also similar. Conclusions Balanced forced‐diuresis with the RG reduced AKI rates after on‐pump cardiac surgery compared to controls. Although the RG group required higher doses of IV potassium replacement in the postoperative period, normal serum levels of potassium were maintained by appropriate intravenous potassium supplementation and the clinical outcomes between groups were similar.


| INTRODUCTION
The development of acute kidney injury (AKI) after cardiac surgery has been reported to be associated with significant postoperative morbidity and mortality both in the short-term and the long-term. 1,2 Many strategies have been described to reduce AKI 3 and the use of the RenalGuard® system (RG) (RenalGuard Solutions Inc.) has recently been reported to reduce AKI rates after cardiac surgery. 4 The principles and components of RG have been extensively described in previous reports. 4,5 In brief, it is a closed-loop fluid management system which allows forced-diuresis to be induced with low dose It is well recognised that diuresis, especially by loop diuretics, is associated with electrolyte losses. The loop diuretics block the NKCC2 (sodium potassium chloride co-transporter) channels present on the apical membrane of the thick ascending limb of the loop of Henle causing sodium and potassium excretion in the urine. 6 Other effects of loop diuretics include metabolic alkalosis, hypocalcaemia, hypomagnesemia and hypochloraemia. Moreover, the use of the cardiopulmonary bypass machine (CPB) during cardiac surgery also causes significant volume and electrolytes shifts. 7 These changes can lead to arrhythmias, cardiac dysfunction, brain oedema and in severe cases, neuronal demyelination. 8 These effects are modulated due to alterations in cellular metabolism, cell membrane potentials and energy transformations. It is therefore imperative that the electrolytes levels are monitored during cardiac surgery and replenished appropriately and in a timely fashion.
This randomised control trial assessed the electrolyte changes (secondary outcomes) between patients treated with balanced forceddiuresis (RG group) compared to controls, during cardiac surgery and within the first 24-h post-op, along with the clinical impact during the inhospital stay.

| Ethical approval
The KIDNEY study (Kidney protection using the RenalGuard® system in cardiac surgery) was reviewed and approved by the

| Aims and objectives
The primary aim of the study assessed the impact of the RG system on the reduction of AKI (RIFLE-Risk, Injury, Failure, Loss of kidney function, End-stage renal disease-criteria definition-50% increase in pre-op "baseline" serum creatinine within first 3 days of surgery) in patients undergoing cardiac surgery. Baseline creatinine was defined as latest creatinine level available before surgery. Secondary aims included the electrolyte changes during surgery and within 24-h postop as well as their impact on clinical outcomes during hospital stay.
The primary outcome has already been published 4 and the secondary objectives are being reported in this study.

| Inclusion criteria
The inclusion criteria for recruited patients were one or more of the following: (i) diabetics (insulin or noninsulin dependent diabetes mellitus), (ii) eGFR of 20-60 ml/min/1.73 m 2 , (iii) Logistic Euroscore of 5 or above, (iv) haemoglobin level of 12.5 g/dl or below and (v) cardiac procedures when CPB time was likely to exceed 120 min. sodium-147 mmol/l, potassium-84 mmol/l, calcium-2 mmol/l, magnesium-80mmol/l, procaine-5 mmol/l and chloride 400 mmol/ l. This was mixed with cold blood in a ratio 1:4 for myocardial protection during surgery.

| RESULTS
A total of 220 patients were randomised to the study (110 patients per group). Pre and intraoperative patients' data were not significantly different between the two groups ( Table 1). The primary outcome of postoperative AKI was significantly lower in the RG group as compared to controls (10% vs. 20.9%, p = .025). The changes in serum electrolytes (sodium, potassium), lactate and pH levels are depicted in Figure 3. On the first postoperative day, there were no significant differences in the sodium, potassium, calcium, magnesium, phosphate and chloride serum levels, between the two groups (Table 1).
There was no significant difference in incidence of atrial fibrillation rate, infections (chest, surgical site infection, sepsis), postoperative cerebro-vascular events rate and median durations of CICU stays between the two groups. The median in-hospital stay was 6 days for both groups.
One patient in RG group and two patients in control group died before hospital discharge. Causes of death were not related to the use of device and included cardiogenic shock, cardiac failure and sepsis (pneumonia) respectively.

| DISCUSSION
The primary aim of this study was confirmed 4 as AKI rate was reduced when the RG system was used perioperatively in cardiac surgery. The study also found that there were statistically significant electrolyte shifts during the use of the CPB and when it is combined with forced-diuresis provided by the RG system. However, these changes did not impact on clinical outcomes.

Polderman reported electrolytes depletion in patients under-
going surgery with the CPB machine. The author suggested that these could be due to the intracellular shifts of electrolytes which are induced by hypothermia as well as urinary excretion of these electrolytes which was independent of the use of perioperative diuretics. 9 The current study confirmed similar electrolytes shifts.
However, clinically there was no significant difference in the patients' recovery as the levels of these electrolytes were closely monitored and supplementation of any electrolyte depletion was carried out in a timely manner.
Another important aspect which could influence electrolytes shifts is the acid-base status of the patient in particular, alkalosis. 7 The latter is also assessed regularly during cardiac surgery and in the LUCKRAZ ET AL.
| 4127 postoperative phase. There was no significant difference between the two groups in terms of bicarbonate and pH levels. Correction of acid-base disturbances can be through respiratory, renal or pharmacological routes.
The use of CPB circuit is recognized to lead to metabolic acidosis possible as a consequence of bicarbonate dilution. 10 As per the Stewart calculations, 11 metabolic acid-base status is a function of two independent variables interacting in intravascular and interstitial compartments: the strong ion difference and the concentration of nonvolatile weak acid. The former is the net charge of all fully dissociated ions (sum of Na + , K + , Ca2 + , Mg2 + levels minus the sum of Cl − and lactate levels). As for the weak-acid, it consists of albumin and phosphate in the extracellular fluid, whereas within erythrocytes, it is primarily haemoglobin. An isolated increase in weak-acid or a decrease in strong ion difference creates a metabolic acidosis. Changes in opposite directions, respectively, cause a metabolic alkalosis. In normal individuals not undergoing surgery a strong ion difference of around 40-45 mEq/L is considered normal 12 whereas during a CPB run, a crystalloid strong ion difference of 24 mEq/L produces a balanced CPB with normal acid-base status. 13 In this study the mean strong ion difference between the RG group and control group was similar at most time-points except during the period of 6-12 h post ICU admission ( Figure 2). Nevertheless, it was maintained between 30 and 45 mEq/L throughout the perioperative period.
The CPB circuit also causes significant haemodilution, hypomagnesaemia, and hyponatraemia. Haemodilution, as also discussed, causes acid-base changes. The degree of haemodilution can be minimized by using the retrograde autologous priming technique (RAP) where the priming volume is kept to around 1 L. 14 This approach was used during this study and despite large volume diuresis and volume replacement in the RG group, the transfusion rate was similar between both groups as were the haemoglobin levels around the perioperative period. Precipitous hyponatraemia on CPB initiation can lead to osmotic demyelination syndrome causing cerebral injury and paralysis. 8

| Limitations
This was a single centre study. Moreover, as the study was not blinded, the management of the control group could have been influenced by some aspects of the Hawthorne effect.

| CONCLUSION
In patients at-risk for AKI, undergoing cardiac surgery with CPB, balanced forced-diuresis as provided by the RG system significantly reduced the incidence of AKI. Compared to the control group, serum potassium levels were lower in the RG group. However, normal levels could be maintained by administration of IV potassium replacement in the postoperative period, thus maintaining similar clinical outcomes, with no adverse safety concerns.