Ultrafiltration and cardiopulmonary bypass associated acute kidney injury: A systematic review and meta‐analysis

Abstract Background Cardiopulmonary bypass is known to raise the risk of acute kidney injury (AKI). Previous studies have identified numerous risk factors of cardiopulmonary bypass including the possible impact of perioperative ultrafiltration. However, the association between ultrafiltration (UF) and AKI remains conflicting. Thus, we conducted a meta‐analysis to further examine the relationship between UF and AKI. Hypothesis Ultrafiltration during cardiac surgery increases the risk of developping Acute kidney Injury. Methods We searched PubMed, Web of Science, EBSCO, and SCOPUS through July 2021. The RevMan (version 5.4) software was used to calculate the pooled risk ratios (RRs) and mean differences along with their associated confidence intervals (95% CI). Results We identified 12 studies with a total of 8005 patients. There was no statistically significant difference in the incidence of AKI between the group who underwent UF and the control group who did not (RR = 0.90, 95% CI = 0.64−1). Subgroup analysis on patients with previous renal insufficiency also yielded nonsignificant difference (RR = 0.84, 95% CI = 0.53 −1.33, p = .47). Subgroup analysis based on volume of ultrafiltrate removed (> or <2900 ml) was not significant and did not increase the AKI risk as predicted (RR = 0.82, 95% CI = 0.63 −1.07, p  = .15). We also did subgroup analysis according to the type of UF and again no significant difference in AKI incidence between UF groups and controls was observed in either the conventional ultrafiltration (CUF), modified ultrafiltration (MUF), zero‐balanced ultrafiltration (ZBUF), or combined MUF and CUF subgroups. Conclusion UF in cardiac surgery is not associated with increased AKI incidence and may be safely used even in baseline chronic injury patients.


| INTRODUCTION
Cardiopulmonary bypass (CPB)-associated AKI occurs in 18.2% of adult patients who undergo CPB and is associated with a twofold increase in early mortality. 1 Risks for CPB-associated AKI have been classified as patient-related and procedure-related factors. 2 Procedure-related factors include systemic inflammatory response, anemia, oxygen delivery, coagulopathy subsequent to foreign surface exposure, and hemodilution associated with the use of a priming solution leading to renal vasoconstriction, and impaired oxygenation including the typically nonpulsatile flow used in CPB.
Ultrafiltration (UF) is a technique commonly used during CPB for volume management and/or filtration of blood to reduce deleterious components. 3 An ultrafilter can be incorporated into a CPB circuit and plasma water, and its soluble components are removed as blood passes through the ultrafilter fibers. Common UF techniques include modified ultrafiltration (MUF), conventional ultrafiltration (CUF), and zero-balanced ultrafiltration (ZBUF). All of these techniques share a common goal: blood concentration, filtration, and the balancing of shifts in the electrolyte plasma concentration as potassium overload, thus protecting the kidney and avoiding homologous blood transfusions. 4,5 Some studies suggest that the use of UF during CPB to remove excessive fluid is not renal protective and may even lead to kidney damage if the fluid removed is more than what is needed. 3 Furthermore, recent reviews warn that we should limit UF in patients with reduced kidney function to prevent AKI. 4 Some studies have set a limit to the volume of ultrafiltrate removed above which AKI can occur to 2900 ml (knowing that 2200 ml is equivalent = 32 ml/kg in an average 70 kg adult). 5 The aim of our present study is to investigate if ultrafiltration is associated with an increased risk of AKI and the safety of its use in patients with previous kidney problems. The present study will also examine if the removal of an ultrafiltrate volume above the 2900 ml suggested limit is associated with an increased risk of AKI.

| Selection process and inclusion criteria
Yielded results from databases were imported into Covidence. 5 From the searches, we reviewed the title and abstract of each paper and retrieved potentially relevant references. Following this initial screening, we obtained the full text of potentially relevant studies and did the full-text screening for the papers using predetermined inclusion criteria, which are any trial or observational study on patients who underwent any type of filtration procedure during cardiac surgery. We excluded case reports and non-English articles.

| Data extraction
Details about the occurrence of AKI in the study groups along with the volume of filtrate removed and type of ultrafiltration used were extracted for subsequent analysis. In addition to study design, participant characteristics and study setting were also extracted to be presented in tables.

| Statistical analysis
A meta-analysis was carried out comparing the occurrence of AKI in patients who underwent cardiac surgery and received ultrafiltration and controls who did not receive ultrafiltration.
RevMan 5.4 was used to calculate the pooled risk ratios (RRs) along with their confidence intervals (CI). We used random-effects model when we observed significant heterogeneity, and when heterogeneity was not solved by random effects, we did the leave one out test. Our analysis was reviewed following the PRISMA Statement checklist to ensure its high quality.  Figure 3A. Quantity of volume removed was divided into two subgroups (less than 2900 ml and more than 2900 ml). The pooled analyses between ultrafiltration group and control group in volume <2900 ml and volume >2900 ml subgroups were (RR = 1.12, 95% CI = 0.78-1.61, p = .54) and (RR = 0.82, 95% CI = 0.63-1.07, p = .15), respectively, as shown in

| DISCUSSION
Our analysis found no significant difference in AKI incidence between patients undergoing cardiac surgery having undergone UF and those who have not (RR = 0.90, 95% CI = 0.64-1.27, p = .55). We also found no increase in AKI incidence in studies that removed an ultrafiltrate volume above what was set to be a critical value of 2900 ml (RR = 0.82, 95% CI = 0.63-1.07, p = .15).
We did also subgroup analysis according to the type of ultrafiltration procedure performed and again no significant difference in AKI incidence between ultrafiltration groups and controls was observed in either the MUF, CUF, ZBUF, and combined CUF and MUF.
Our results (RR = 0.84, 95% CI = 0.53-1.33, p = .47) contradict some studies that claim the UF increase AKI risk in patients with previous kidney disease like Paugh 15 in which the rate of AKI was higher in the filtration group and another study Musleh 13 where there is observed a higher number of AKI in the filter group. It is important to mention that even in these two studies although there is an observed increase in AKI incidence, patients have similar rates of death, postoperative length of stay, and readmissions compared to the group who did not undergo ultrafiltration, so this claimed dysfunction does not really affect the clinical outcome and patients did not require either dialysis or support to treat their AKI.

| Limitations of our study
Several studies we included in our analysis had not mentioned the weight indexed volume of filtrate removed so we had to run the analysis based on the total volume of filtrate removed instead.

| CONCLUSION
UF in cardiac surgery is safe and does not increase the risk of AKI, even in patients with previous kidney problems. Also, the removal of a volume of filtrate above 2900 ml during the procedure was not shown to negatively affect outcomes.