Clinical characteristics and outcomes of continuous renal replacement therapy performed on younger children weighing up to 10 kg

Background/aim This study aimed to investigate the clinical features, modality, complications, and effecting factors on the survival of children weighing up to 10 kg who received continuous renal replacement therapy (CRRT). Materials and methods This study was a retrospective observational study conducted in five pediatric intensive care units in tertiary hospitals in Turkey between January 2015 and December 2019. Results One hundred and forty-one children who underwent CRRT were enrolled in the study. The median age was 6 (range, 2–12) months, and 74 (52.5%) were male. The median weight of the patients was 6 (range, 4–8.35) kg and 52 (36.9%) weighed less than 5 kg. The most common indication for CRRT was fluid overload in 75 (53.2%) patients, and sepsis together with multiorgan failure in 62 (44%). The overall mortality was 48.2%. Conclusion Despite its complexity, CRRT in children weighing less than 10 kg is a beneficial, lifesaving extracorporeal treatment modality.


Introduction
Continuous renal replacement therapy (CRRT) has been the preferred treatment method for critically ill children with acute kidney injury (AKI) and fluid overload (FO) [1]. The indications for CRRT in pediatric intensive care units (PICUs) are varied [2]. Diuretic-resistant fluid overload, metabolic acidosis, poisoning, electrolyte abnormalities, and attacks of inborn errors of metabolic disease are the most common reasons for CRRT use in PICU settings [2,3]. The first report on CRRT use in adults was reported by Kramer et al. [4] in 1977, and Leone et al. [5] documented a successful fluid-electrolyte and acid-base balance with CRRT in anuric children in 1986 [5].
Among the renal replacement therapy (RRT) modalities, peritoneal dialysis (PD) is the most used method in neonatal intensive care units (NICU) due to its applicability and availability [6]. However, CRRT is more effective in terms of the removal of free water and ammonia, and it is preferred over PD, which has limited benefits in younger children who have undergone abdominal surgery recently [6,7]. However, CRRT use in younger children brings some risks due to higher extracorporeal blood volumes in the circuit than the patient's blood volume and relatively higher blood flows [8]. The main concerns about CRRT in younger children are establishing vascular access, hemodynamic instability during CRRT, hypothermia, and hemodilution [9]. These problems are more frequently observed in younger children, and it causes challenges in all centers; therefore, CRRT use in younger children can only be performed in developed and dedicated PICUs [9]. The improved technology of pumpdriven volumetric-controlled CRRT devices with small extracorporeal volumes has increased the use of CRRT in younger children [10]. The purposes of this study were to determine the clinical characteristics of the patients and risk factors, which were mortality and CRRT-related complications, and the effecting factors on the survival of children weighing up to 10 kg who received CRRT.

Materials and methods
This was a retrospective, multicenter, descriptive study. We included children who weighed up to 10 kg of body weight and underwent CRRT between January 2015 and March 2021 in five PICUs. The study was approved by the Institutional Review Board of the Local University Faculty of Medicine (approval number: I10-667-21).
CRRT therapies were performed using PrismaflexTM HF20 Gambro (USA) and PrismaflexTM M60 Gambro (USA) sets. Two-lumen 6.5-Fr, 7-Fr, and 8-Fr dialysis catheters were used for vascular access in the patients. In patients undergoing extracorporeal membrane oxygenation (ECMO), the dialysis circuit was connected through the connectors on the venous lines of the ECMO circuit, not through a venous catheter. CRRT circuit type (HF20, M60), circuit priming solution (red packed cell-normal saline mixture, albumin or normal saline), type of anticoagulation (heparin, citrate or no anticoagulation), ultrafiltration, number of circuits used, complete blood count and biochemistry results pre-CRRT and on the first day, use of blood products during CRRT, complications, vasoactive inotropic score (VIS), duration of CRRT, and outcomes (survivors, nonsurvivors). After enrollment, patients were further analyzed in two groups according to their body weight: less than 5 kg, and between 5 kg and 10 kg.
Acute kidney injury was categorized according to the RIFLE (Risk, Injury, Failure, Loss, and End-stage Kidney) and pediatric RIFLE (pRIFLE) criteria according to serum creatinine, estimated creatinine clearance (eCCl), and urine output within 24 h before the initiation of CRRT [11]. Fluid overload was calculated using the following equation: Fluid overload (%) = (Fluid in -fluid out)/admission weight in PICU × 100%. Multi-organ dysfunction syndrome (MODS) was defined as the 'involvement of three or more organ systems' [12].
VIS was calculated using the following equation:  [13].
Statistical analyses were performed using the SPSS v26.0 software package (Statistical Package for the Social Sciences for MacOS, SPSS Inc., USA). The patients were divided into survivors, nonsurvivors, and '<5 kg and 5-10 kg patients' . Numbers (n) and proportions (%) were used for descriptives of categorical variables. Means and standard deviations were used for normally distributed variables. The Mann-Whitney U test was used for comparison purposes as the continuous variables do not have a normal distribution with categorical variables having two categories. The results were reported as median values and interquartile ranges (IQR 25-75). Normal distribution was tested using histograms, the Shapiro-Wilk and Kolmogorov-Smirnov tests, and variation coefficients. The Wilcoxon test was used for related variables of the groups. The Chi-square test and Fisher's exact test were used to compare nonnumerical parameters between categorical groups. The Kaplan-Meier method was used for survival analysis. P-values below 0.05 were accepted as statistically significant.
Upon comparison of the laboratory parameters of patients before CRRT and on the first day of CRRT, a statistically significant difference was observed in the values of BUN, creatinine, potassium, uric acid, phosphorus, WBC, platelet, pH, HCO3, and ammonia (Table 5).
According to the characteristics of CRRT treatment, survivors and nonsurvivors showed significant differences in days of CRRT [ Survival time analysis was calculated using the Kaplan-Meier method. The survival time showed significant differences in hepatic failure and circulation failure (p = 0.042 and p = 0.010, respectively); these data are presented in Figures 1 and 2. Survival times showed no significant differences in FO, body weight, sepsis, and ECMO support.

Discussion
We report our experience in 141 small children weighing less than or equal to 10 kg who were treated with CRRT. One of the most critical points required for CRRT in this extreme group is providing good vascular access for an efficient process. In a study on CRRT, especially in younger children, it was stated that the region used primarily for dialysis catheters was the left IJV under the guidance of ultrasonography [6,9]. The neck veins are preferred sites for patients weighing less than 5 kg [14]. In the studies of Kaempfen et al. [13] and Symons et al. [10], the most frequently used catheter sites were the femoral vein in CRRT studies performed in children weighing less than or equal to 10 kg. Unlike their results, in this multicenter study, we observed that the RJVI was the most used site for vascular access. A comparison of catheter sites in patients weighing less than 5 kg and 5-10 kg showed that the most used catheter site was RJVI in both groups (59.6% and 62.9%, respectively), and there was no significant difference between the two groups.
The CRRT circuit onto the ECMO circuit becomes an advantage in children with low body weight, especially in patients without venous access [15,16]. However, it has an increased risk of systemic inflammation and increased   hemolysis due to flow turbulences [16]. We performed this technique on 16 patients who were on ECMO run without dialysis catheters. Another critical element is the establishment of the CRRT circuit. High-volume sets are avoided because the blood volume is small in patients with low body weight [17,18]. The blood volumes in devices such as CARPEDIEM (27 mL) and NIDUS (10 mL) used primarily for newborns are relatively low compared with other sets, but these CRRT devices are not available in most centers [18]. Conventional HD and CRRT devices are not approved for children weighing under 20 kg [19]. The NIDUS device is licensed for children weighing 0.8-8 kg [17,18]. CARPEDIEM can be used in children who weigh more than 2.5 kg and blood priming may not be required in children weighing up to 2.5 kg [18]. There are no CARPEDIEM and NIDUS devices in our centers. The most frequently used set (%) in our study was the Gambro Prismaflex HF20 set. The volume of this set is 60 mL, and it is recommended to prime with blood for babies weighing less than 5 kg [18]. In some patients, without the HF20 circuit, we had to use the M60 circuit. However, this rate was relatively low.
Circuit anticoagulation is an essential technical issue in CRRT, and its primary purpose is to prevent thrombotic processes when blood encounters the extracorporeal circuit [20]. Heparin is the most widely used anticoagulant [20]. Regional citrate anticoagulation (RCA) with citrate is another alternative anticoagulation method [3,20,21]. Compared with anticoagulation with heparin, RCA has been shown to reduce the risk of bleeding and improve circuit life [21][22][23]. Younger children who have undergone recent surgery often require RCA because of the postoperative condition or coagulopathy [22]. Despite increasing knowledge and experience in performing CRRT with RCA in the pediatric population, RCA in younger children remains a challenge because it increases the risk of citrate accumulation (mainly because of an imbalance between blood flow and body weight) [22]. To our knowledge, only a few studies have reported detailed data on RCA performance during CRRT in younger children, and a recent review suggested more reports were required on the use of citrate [22]. CARPEDIEM and NIDUS devices, which are specifically designed for younger children, do not yet provide automated RCA [18,24]. The standard anticoagulation method in our study was heparin. Although we prefer citrate in older children, we found that this rate is low for this age group.
CRRT has become a preferred treatment modality for AKI and fluid overload in critically ill pediatric patients in the last two decades [25]. CRRT has advantages over other renal replacement therapies (PD and IHD) [25]. Compared with PD, CRRT has a better capability of solute filtration and liquid removal efficiency, and PD cannot provide adequate clearance of toxic metabolites, which can be achieved with CRRT, especially in congenital metabolic diseases [25]. CRRT is particularly effective in maple syrup disease, organic acidemia with hyperammonemia, and inborn metabolism in urea cycle disorder [25][26][27][28]. Higher peak ammonia levels during hyperammonemia episodes are associated with worse survival rates, and the duration of coma before dialysis is negatively associated with cognitive outcomes [26,29]. The rate of ammonia clearance with CRRT has been associated with improved outcomes [25][26][27][28][29][30]. Therefore, timely and aggressive treatment, including the use of dialysis, should be applied to rapidly lower ammonia levels [26][27][28][29][30]. In our study, CRRT was performed in 36 patients (25.5%) for metabolic reasons.
The frequency of CRRT performed mortality in children under 10 kg was 62% (total of 85 cases) in the study of Symons et al. [10], 38% (n = 16) in the study of Pedersen et al. [31], and 42.3% (71) in the study of Kaempfen et al. [13]. The mortality rate in our study was 48% (n = 141). Our study's mildly higher mortality rate may be due to the high MODS rate (44%). The mortality rate of the patients with MODS in our study was 66.2%. MODS itself occurs in 30% to 50% of children in the PICU and is responsible for a disproportionately higher percentage of total deaths in PICUs, reaching over 90% in some studies [32]. In our study, patients who were treated with CRRT and died had higher PRISM scores, consistent with previous data [33,34].
One of the important indications of CRRT is fluid overload [1,31,35]. As an independent indicator of mortality, early initiation of CRRT has been recommended ECMO 12 (23)   in patients with a >10% FO [36,37]. It has been reported that mortality increases as the percentage of FO increases when it passes 10% [31][32][33][34][35][36][37][38][39]. In our study, CVVH, CVVHD, and CVVHDF modes were used in CRRT, and these modes showed no significant differences between survivors and nonsurvivors. These data were similar to the findings of Symons et al. [10]. In the studies of Symons et al. [10] and Kaempfen et al. [13], CVVH mode was used in all patients. In our study, CVVHD mode was used more frequently in patients with a body weight of 5-10 kg.
The major limitation of this study was its retrospective design. We were not able to determine the exact reason for starting CRRT because we based it on what was documented in the patient records. In addition, the reason and timing of starting CRRT treatment may vary, especially according to the preference of the 'pediatric intensivist' in different centers. Also, we could not determine mechanical ventilation settings and ECMO flow rates. We could not detect circuit changes and circuit times from retrospective data; therefore, we could not compare circuit times between subgroups. No data on the exact cause of death are presented because some of the patients had multiple organ failures. The absence of a control group is another limitation.

Conclusion
To our knowledge, this multicenter study comprised the largest study group to date of children weighing less than or equal to 10 kg who underwent CRRT. Despite the technical difficulties of using modified equipment in critically ill children with a body weight of less than or equal to 10 kg, CRRT is a lifesaving extracorporeal treatment modality in dedicated and experienced PICUs. Although new CRRT devices are designed for younger children, access to these devices is restricted in most parts of the world. Especially in newborns, CRRT is an indispensable treatment method because it removes toxic substances faster than peritoneal dialysis in metabolic diseases.

Conflicts of interest/cCompeting interests
The authors have no financial disclosures that would be a potential conflict of interest with the current manuscript.

Source of fFunding
The original article received no external funding.

Consent to participate
Approval was obtained from the family of the participants.

Consent for publication
Approval was obtained from the family of the participants.

Ethics approval
The study was approved by the Institutional Review Board of the Ankara University Faculty of Medicine (aApproval nNumber: I10-667-21).