Clearance and Nutrition in Neonatal Continuous Renal Replacement Therapy (CRRT) Using the CARPEDIEM™ System

doi:10.21203/rs.3.rs-3244085/v1

Download PDF

Research Article

Clearance and Nutrition in Neonatal Continuous Renal Replacement Therapy (CRRT) Using the CARPEDIEM™ System

https://doi.org/10.21203/rs.3.rs-3244085/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 17 Jan, 2024

Read the published version in Pediatric Nephrology →

You are reading this latest preprint version

Background: Infants with end stage kidney disease (ESKD) demonstrate poor growth partly due to obligate fluid and protein restrictions. Delivery of liberalized nutrition on continuous renal replacement therapy (CRRT) is impacted by clinical instability, technical dialysis challenges, solute clearance, and nitrogen balance. We analyzed delivered nutrition and growth in infants receiving CRRT with the Cardio-Renal, Pediatric Dialysis Emergency Machine (CARPEDIEM™).

Methods: Single center observational study of infants receiving CRRT with the CARPEDIEM™ between June 1 and December 31, 2021. We collected prospective circuit characteristics, delivered nutrition, anthropometric measurements, and illness severity Score for Neonatal Acute Physiology-II. As a surrogate to normalized protein catabolic rate in maintenance hemodialysis, we calculated normalized protein nitrogen appearance (nPNA) using the Randerson II continuous dialysis model. Descriptive statistics, Spearman correlation coefficient, Mann Whitney, Wilcoxon signed rank, Receiver-Operating Characteristic curves, and Kruskal-Wallis analysis were performed using SAS version 9.4.

Results: Eight infants received 31.9 [22.0, 49.7] days of CRRT using mostly (90%) regional citrate anticoagulation. Delivered nutritional volume, protein, calories, enteral calories, nPNA, and nitrogen balance increased on CRRT. Using parenteral nutrition, 90 ml/kg/day should meet caloric and protein needs. Following initial weight loss of likely fluid overload, weight gain occurred after 14 days of CRRT. Despite adequate nutritional delivery, goal weight (z-score=0) and growth velocity were not achieved. Most (62.5%) survived and transitioned to peritoneal dialysis (PD).

Conclusions: CARPEDIEM™ is a safe and efficacious bridge to PD in neonatal ESKD. Growth velocity of infants on CRRT appears delayed despite delivery of adequate calories and protein.

Neonatal nephrology

CRRT

nutrition

neonatal ESKD

The incidence of neonatal end stage kidney disease (ESKD) is estimated around 7.1 per million, and infants make up 11–13% of all children on chronic dialysis [1, 2]. Mortality rates among infants and children receiving continuous renal replacement therapy (CRRT) remain high, up to 75% in those less than 3 kg and 56% in those less than 5 kg, with higher rates among infants who start dialysis before 31 days of age [1, 3, 4]. Neonatal dialysis is particularly challenging due to lower patient blood volume compared to extracorporeal volume (ECV) of dialysis filters, small vessel diameter with limited access alternatives, evolving nutritional and metabolic demands, and high anticoagulation risk. Further, neonates are a vulnerable population to anticoagulation. In addition to several reports of accidental unfractionated heparin (UFH) overdoses, with potentially fatal consequences, there are high rates of intraventricular hemorrhage (IVH) among preterm infants (15% compared to full term infants 0.2%) and even higher rates among low birth weight infants at baseline even without anticoagulation [5, 6].

Due to a growing population of neonates with ESKD and an urgent need for neonatal dialysis options, the Cardio-Renal, Pediatric Dialysis Emergency Machine (CARPEDIEM™, Bellco-Medtronic, Mirandola, Italy) was designed to deliver CRRT to neonates. CARPEDIEM™ offers more accurate fluid balance control with high precision scale sensitivity of 1 g, lower blood flow rates starting at 2 ml/min, and compatibility with smaller catheters (4–7 Fr have been reported) [7]. While Italy has used CARPEDIEM™ since June 2013, the United States Food and Drug Administration approved the device in April 2020 for CRRT with continuous veno-venous hemodialysis (CVVHD) in infants weighing 2.5 to 10 kg. Using 0.15 m² and 0.25 m² filters that have an ECV of 32 and 41ml, respectively, blood prime may not be necessary. The CARPEDIEM™ system was studied with UFH anticoagulation, and no serious adverse events were reported [8].

For many critically ill infants with severe acute kidney injury (AKI) or ESKD, especially among oliguric and anuric infants, obligate fluid and protein restrictions limit the ability to provide adequate nutrition. Preterm infants require higher protein intake (as high as 4 g/kg/day) compared to their term counterparts for optimal growth [9]. Among premature and extremely low birthweight (ELBW, less than 1000 g) infants, there is an independent relationship between weight growth velocity (defined as g/kg/day) and z-score with neurodevelopmental and growth outcomes at 18 to 22 months’ corrected gestational age [10, 11]. Both the North American Pediatric Renal Trials and Collaborative Studies and the International Pediatric Peritoneal Dialysis Network identify suboptimal nutrition as an indication to begin dialysis in neonates to improve weight and height velocity. Current literature about the benefits of early dialysis in neonates emphasizes the liberalization of fluid and protein restrictions to increase prescribed nutrition in order to improve growth [12, 13].

Ensuring adequate nutrition and protein delivery in dialyzed neonates is complicated by macro- and micronutrient losses through dialysis. Accounting for dialysis-related protein losses as high as 20% of total intake, term infants on CRRT require at least 2.5-3 g/kg/day of delivered protein [13, 14]. Poor growth among dialyzed infants can be impacted by inadequate solute clearance as well as severity of critical illness and catabolic state. Among maintenance hemodialysis (HD) patients in steady state, urea kinetic modeling provides both a measure of dialysis adequacy (Kt/V) as well as daily protein intake or normalized protein catabolic rate (nPCR), derived from interdialytic rise in urea and, therefore, the urea generation rate. Because CRRT is a continuous treatment, continuous dialysis models are used to calculate normalized protein nitrogen appearance (nPNA) as a surrogate to nPCR used in maintenance HD patients [15]. Nitrogen balance has also been described in adult and pediatric literature as an index of nutritional status in patients with ESKD who typically have high rates of protein catabolism and negative nitrogen balance [13, 14, 16]. In addition to maintaining positive nitrogen balance in neonates receiving CRRT, better tools are needed to monitor nutritional adequacy given challenges with precise anthropometrics, especially in this population in whom adequate growth has been shown to correlate with morbidity [10, 11, 17].

We aimed to describe a single center experience of infants who received CRRT with the CARPEDIEM™ system to assess the delivered nutritional intake, solute clearance effect on nPNA, and overall growth among these high-risk neonates.

This is a single center (Texas Children’s Hospital, Houston, Texas, United States) observational study of all patients who received CRRT with the CARPEDIEM™ system between June 1, 2021, and December 31, 2021. IRB approval was obtained. Analysis was performed on the initial CARPEDIEM™ treatment run for each patient, which included treatments without a gap exceeding 7 days between them. Clinicians with guidance from institutional protocols determined CRRT initiation, modality, prescription, vascular access, anticoagulation including venous thromboembolism prophylaxis (VTE), and termination. All patients received CVVHD with a starting prescribed clearance of 2000 ml/1.73m²/h. Data abstracted included demographics, clinical and laboratory variables, and CRRT characteristics. A previously established institutional quality improvement (QI) dashboard on all patients on CRRT provided prospective circuit characteristics and delivered nutrition. Urea Generation rate \(\left(\frac{{EffluentUrea}_{mg/dL}*{EffluentVolume}_{mL}*\frac{1 dL}{100 mL}}{{FilterLife}_{min}}\right)\) was used in the Randerson II equation for nPNA \(\left(\frac{1.46+\left(10.76*{UreaGenerationRate}_{mg/min}\right)}{{Weight}_{Kg}}\right)\) which is a model estimating nitrogen losses in chronic ambulatory PD (CAPD) patients with a goal of at least 1.2 used for children on PD [18]. Delivered clearance \(\left(\frac{{EffluentVolume}_{mL}}{{Weight}_{Kg}*{FilterLife}_{hr}}\right)\)and urea clearance \(\left(\frac{{EffluentUreaNitrogen}_{mg/dL}*{EffluentVolume}_{mL}}{{BloodUreaNitrogen}_{mg/dL}*{FilterLife}_{min}}\right)\)were calculated.

Anthropometric measurements were collected through electronic health record growth chart and reported in both relative (z-score) and absolute units (kg, cm). Clinically appropriate intrauterine or postnatal reference growth charts adjusted for sex and gestational age were used to obtain z-scores. Per unit protocol, weights were measured daily during circuit changes and both length and occipital frontal circumference (OFC) were measured weekly. Weight was analyzed over 6 weeks and 6 months. Length and OFC were analyzed over 6 months. Follow-up data was extracted through October 31, 2022.

Illness severity is described by the Score for Neonatal Acute Physiology-II (SNAP-II) score which has been validated in critically ill neonates both at admission and sequentially to predict mortality risk due to physiologic instability [19]. SNAP-II score is a summative score out of 115 and based on mean arterial pressure, temperature, ratio of arterial oxygen partial pressure to fractional inspired oxygen, lowest pH, seizures, and urine output [19]. SNAP-II scores greater than 30 suggest a 3.5 times increased risk of mortality [19]. Unlike many neonatal illness severity scores, SNAP-II score does not require data from the first 24 hours of life, which is preferable given not all infants were born in-center. SNAP-II score was used in this study to assess illness severity at admission, at CRRT start, and at CRRT stop.

The percentage of fluid overload was determined by weight compared to estimated dry weight (EDW). Descriptive statistics were reported as means with standard deviations or medians with interquartile ranges (IQR) according to normality of data. Correlational analysis of clinically relevant variables was assessed to evaluate the relationships between nutritional intake, growth, and dialytic treatment via Spearman correlation coefficient. Nonparametric unpaired t-tests (Mann Whitney) evaluated the relationship between survivors and dialytic treatment variables. Nonparametric paired t-tests (Wilcoxon signed rank) evaluated the growth metrics before and after CARPEDIEM™ treatment. Receiver-Operating Characteristic (ROC) curves were used to identify optimal cut-off points for: days from CRRT start to achieve weight gain and nutritional fluid intake to achieve protein intake of 2.5 g/kg and 3 g/kg via Youden index (maximal effectiveness). Kruskal-Wallis analysis assessed the difference between survivorship and provision of nutritional fluids. Statistical significance was defined as p < 0.05. Statistical analysis was performed using SAS version 9.4.

Patient and Maternal Characteristics

Eight infants (62.5% male, 75% White, 50% Hispanic) received 278 CRRT sessions over a median of 31.9 [22.0, 49.7] days with the CARPDIEM™ system. Most were delivered in-center (75%) by Cesarean section (62.5%) and were admitted to the neonatal (75%) intensive care unit (ICU). Most were preterm (25% extremely low gestational age < 28 weeks, 37.5% preterm, 25% late preterm) and 25% were ELBW. Maternal history was significant for 25% with chronic hypertension and 25% with diabetes (12.5% gestational diabetes, 12.5% type II diabetes).

At ICU admission, median age was 0 [0, 37] days, weight was 2.61 [2.55, 3.42] kg, and SNAP-II overall was 18.5 [13.25, 53.75]. At CRRT start, infants had been hospitalized in the ICU for a median of 12.7 [9.1, 41.1] days with EDW 2.6 [2.5, 3.4] kg, actual weight 3.3 [3.1, 4.4] kg, and 23% [16%, 26%] fluid overload by weight. At CRRT start, 100% were mechanically ventilated, 37.5% required vasoactive medication, and SNAP-II was 28 [13.25, 39.75]. Most had neonatal ESKD (87.5%). The most common renal etiology was congenital anomaly of kidney and urinary tract (CAKUT, 37.5%).

Table 1

Patient and maternal characteristics.
Patient and Maternal Characteristics	N (%), Median [IQR]
Male	5 (62.5%)
Race
White, Hispanic	4 (50%)
White, Non-Hispanic	2 (25%)
Black	2 (25%)
Born in-center	6 (75%)
Cesarean section	5 (62.5%)
Maternal history
Chronic hypertension	2 (25%)
Diabetes (Gestational and Type II)	2 (25%)
Gestational age (GA)
Extremely low GA (< 28 weeks)	2 (25%)
Preterm (32–35 weeks)	3 (37.5%)
Late preterm (35–37 weeks)	2 (25%)
Term (37–40 weeks)	1 (12.5%)
Renal etiology
Neonatal ESKD	7 (87.5%)
CAKUT	3 (37.5%)
Congenital nephrotic syndrome	2 (25%)
Autosomal recessive polycystic kidney disease	1 (12.5%)
Congenital cyanotic heart disease, recurrent prolonged AKI	1 (12.5%)
Neonatal AKI
CAKUT and hydrops fetalis	1 (12.5%)
ICU admission
Neonatal ICU	6 (75%)
Age in days	0 [0, 37]
SNAP-II overall	18.5 [13.25, 53.75]
SNAP-II among survivors	16 [5, 16]
SNAP-II among non-survivors	56 [38.5, 59.5]
CARPEDIEM™ start
Age in days	18.5 [8.5, 73.5]
Intubated and mechanically ventilated	8 (100%)
Vasoactive medication	3 (37.5%)
SNAP-II overall	28 [13.25, 39.75]
SNAP-II among survivors	32 [24, 39]
SNAP-II among non-survivors	16 [10.5, 32.5]
CARPEDIEM™ stop
Age in days	81 [39, 117]
SNAP-II overall	29.5 [22.5, 55]
SNAP-II among survivors	24 [18, 27]
SNAP-II among non-survivors	61 [57, 69]
Outcome
Survival and transition to peritoneal dialysis	5 (62.5%)
Non-survivors and transition to comfort care	3 (37.5%)

Abbreviations: ESKD, end stage kidney disease; AKI, acute kidney injury; CAKUT, congenital anomaly of the kidney and urinary tract; ICU, intensive care unit; SNAP-II, Score for Neonatal Acute Physiology-II; CARPEDIEM^TM, Cardio-Renal, Pediatric Dialysis Emergency Machine; CRRT, continuous renal replacement therapy; N, number; %, percentage; IQR, interquartile range.

CRRT Characteristics

At CRRT initiation, most were dialyzed with the 015 m² (84%) hemofilter, and 40.6% of circuits were primed with normal saline. Circuit-to-circuit exchange was used for infants requiring blood prime to decrease blood exposure. All patients received CVVHD with a prescribed clearance of 2000 ml/1.73m²/h. Median prescribed blood flow was 7.1 [5.0, 8.7] ml/kg/min, actual effluent dose was 34.4 [23.9, 42.2] ml/kg/h, delivered clearance was 86.4 [75.8, 97.1] ml/kg/h, and urea clearance was 2.9 [2.6, 3.4] ml/min. Overall, median filter life was 21.8 [17.0, 23.4] hours with premature filter clotting occurring 25.0% [14.3%, 35.0%] of the time. Patients received 97.7% [86.7%, 103.2%] of their prescribed treatment.

Table 2. Patient continuous renal replacement therapy (CRRT) characteristics. Delivered clearance was calculated using effluent volume, patient estimated dry weight (EDW), and filter life in hours. Urea clearance was calculated using effluent and blood urea nitrogen, effluent volume, and filter life in minutes.

CRRT Characteristics and Outcomes	N (%), Median [IQR]
CRRT Sessions	278
015 m²filter	233 (84%)
Normal saline prime	113 (40.6%)
CVVHD with 2000ml/1.73m²/h clearance	8 (100%)
Regional citrate anticoagulation	251 (90.3%)
Treatment duration (days)	31.9 [22.0, 49.7]
Blood flow (ml/kg/min)	7.1 [5.0, 8.7]
Effluent dose (ml/kg/h)	34.4 [23.9, 42.2]
Filter life (h)	21.8 [17.0, 23.4]
EDW at CRRT start (kg)	2.6 [2.5, 3.4]
Last weight at CRRT start (kg)	3.3 [3.1, 4.4]
Fluid overload at CRRT start by weight (%)	23% [16%, 26%]
Delivered clearance (ml/kg/h)	86.4 [75.8, 97.1]
Urea clearance (ml/min)	2.9 [2.6, 3.4]
Hemodialysis catheters per patient	2 [1, 3]
Survival to discharge and transition to PD	5 (62.5%)

Abbreviations: CRRT, continuous renal replacement therapy; CVVHD, continuous veno-venous hemodialysis; EDW, estimated dry weight; PD, peritoneal dialysis; N, number; %, percentage; IQR, interquartile range.

Access and Anticoagulation

Initial dialysis catheter access was 7 Fr (12.5%), 7.5 Fr (50%), and 8 Fr (37.5%), mostly placed in the internal jugular vein (87.5%). Patients required a median of 2 [1, 3] dialysis catheters. Most infants (87.5%) were started on regional citrate anticoagulation (RCA) with 37.5% also receiving concurrent UFH for catheter-associated VTE prophylaxis (12.5% at 5 units/kg/h, 25% at 10 units/kg/h). One infant (12.5%) required UFH due to thrombus prior to CRRT start. RCA was used in most (90.3%) treatments.

Outcomes and Adverse Events

Overall, 62.5% survived their CRRT course with CARPEDIEM™, and all survivors were successfully transitioned to PD. No neonates had renal recovery. Survival among infants with EDW < 3 kg at CRRT start (3/5, 60%) was similar to infants with EDW > 3 kg at CRRT start (2/3, 66%; p = 0.85). The median duration of hospitalization was 228 [177, 249] days among survivors compared to 31 [21.5, 33.5] days for non-survivors. Despite RCA with heparin prophylaxis, 12.5% had dialysis catheter-associated thrombus and 12.5% had umbilical line associated thrombus. Overall, 37.5% (3/8) of infants had intracranial hemorrhage but the majority (67%, 2/3) occurred prior to CRRT start.

Delivered Nutrition and Nutritional Assessment

Over 6 weeks of CRRT with the CARPEDIEM™ system, median total delivered nutritional volume increased from 74.9 [65.2, 88.8] to 89.8 [78.2, 100.6] ml/kg/day (Figure 1a), total protein increased from 2.5 [2.2, 2.7] to 3.4 [2.6, 4.0] g/kg/day (Figure 1c), delivered calories increased from 77.5% [61.8%, 94.3%] to 91.5% [79.8%, 103.5%] of goal (Figure 1e), and delivered calories from enteral nutrition increased from 0% [0%, 31.0%] to 79.5% [64.5, 98.8%] of goal total calories (Figure 1g).

Kruskal-Wallis analysis revealed a difference between survivorship and the ability to deliver nutritional (includes enteral and parenteral) fluid (ml/kg/day) (survivor: median 94.41 [80.16, 102.55] (95%CI: 89.39, 96.43), non-survivors: median 70.3 [53.58, 88.51] (95%CI: 56.84, 84.32), p 0.0001). Moderate to strong positive correlations were seen with survivorship and delivered enteral feeding volume (r=0.73687, p=0.037), enteral protein intake (r=0.73687; p=0.037), and enteral caloric intake (r=0.85023; p=0.0075).

Over 6 weeks of CRRT with the CARPEDIEM™ system with unchanged prescribed dialysis dose, median nPNA initially increased from 1.54 [1.29, 2.49] during week 1 to 2.26 [1.81, 2.42] during week 4 before decreasing to 1.90 [1.31, 2.29] by week 6 (Figure 2a). Median net nitrogen balance increased from +0.73 [+0.05, +1.25] to +1.05 [+0.98, +1.15] g/kg/day over 6 weeks (Figure 2c). Between the first and last week of CARPEDIEM™, median nPNA increased from 1.54 [1.29, 2.49] to 1.82 [1.74, 2.20] with a narrower IQR and net nitrogen balance increased from +0.73 [+0.05, +1.25] to +1.19 [+0.15, +1.21] g/kg/day (Figure 2d). There was no significant difference in Randerson nPNA by survivorship (all p values greater than 0.4875).

Absolute and Relative Growth Velocity

Change in absolute weight, length, and OFC were graphed over time relative to CARPEDIEM™ start in Figures 3-5, showing sustained, slow growth that continued even after discontinuation of CARPEDIEM™. Median weight z-score gradually declined from 0.90 [-0.18-1.95] the week prior to CARPEDIEM™ to -1.21 [-1.42, -0.95] by week 6 (Figure 3a). When studied over a 6-month period (Figure 3b), the z-score decreased to -1.58 [-1.97, -0.58] by month 3 before increasing to -0.32 [-2.13, 0] by month 6.

A similar pattern is seen with median weight gain velocity. The week prior to CARPEDIEM™ start, weight gain velocity was +9.42 [-5.29, 46.30] g/kg/day with an initial decline in weight gain velocity by week 3 of -16.77 [-24.89, 2.94] g/kg/day followed by improvement to +6.37 [-2.09, 20.30] g/kg/day by week 6 (Figure 3c). Sensitivity analysis via ROC, used to analyze the relationship between days on CRRT and the probability of weight gain, indicated that our cohort took approximately 14 days of CRRT with CARPEDIEM™ to begin gaining weight (J=0.2745; sensitivity 0.7602; specificity: 0.51428; probability 0.563).

Length initially decreased by z-score from -0.20 [-9.54, 1.34] the month prior to CARPEDIEM™ to -2.65 [-3.49, -0.35] at month 4 before increasing to -1.26 [-2.01, -0.92] by month 6 (Figure 4a). Linear growth velocity suggests the slowest period of linear growth of 0 [-0.11, 0.96] cm/week at month 1 before gradually increasing to 0.48 [0, 1] cm/week by month 6 (Figure 4b).

OFC growth trends differ between z-score (Figure 5a) and growth velocity (Figure 5b). Median OFC z-scores initially decreased from -0.48 [-4.03, 0.08] the month prior to CARPEDIEM™ to -4.24 [-5.07, -1.86] by month 2 before increasing to -3.10 [-3.95, -2.00] by month 6. Median OFC growth velocity remained positive with improvement from 0.0 [-0.5, 1.1] cm/week the month prior to CARPEDIEM™ to 0.2 [-0.3, 0.4] cm/week by month 6.

When analyzing by survivorship, both pre- and post-weight z-scores were significantly higher in non-survivors compared to survivors (p= 0.0369, p=0.0369; respectively). Survivorship did correlate with linear growth z-score (r=-0.92217, p=0.0011) and approached significance with OFC z-score (r=0.948, p = 0.0513). Interestingly, there was no statistically significant clear relationship between percent fluid overload and survivorship (r=-0.39, p=0.33).

Total Fluid Goals

Analysis via ROC curves to estimate the optimal volume of nutritional fluid (total fluids, TF) cut-points to meet specific nutritional provision goals indicated a TF intake of 79.49 ml/kg/day was maximally effective in achieving 2.5 g/kg of protein intake (AUC 0.884, J cutoff = 0.6843, sensitivity 0.88944, specificity 0.79487, probability 0.6993) and a TF intake of 84.989 ml/kg/day was maximally effective in achieving 3 g/kg of protein intake (AUC 0.7951, J cutoff = 0.5315, sensitivity 0.8648, specificity 0.6666, probability 0.5003). Achievement of 100% of caloric needs required TF 89.64 ml/kg (AUC 0.8435, J cutoff = 0.53138, sensitivity 0.7569, specificity 0.7744; probability 0.55521); with a high correlation of caloric intake associated with the use of parenteral nutrition (Spearman correlation, R=0.76190; p=0.028).

In this observational study of 8 neonates who received CRRT with the CARPEDIEM™ system, we show that critically ill infants can successfully receive CRRT with a smaller ECV system, relatively few adverse events, and a 62.5% survival while achieving delivered nutrition goals and sustained growth among survivors. CARPEDIEM™ served as a successful bridge therapy in this population while they underwent placement and healing of PD catheter.

Neonatal ESKD

Our cohort was primarily neonatal ESKD (87.5%) due to lower urinary tract obstruction (25%), renal cystic dysplasia (12.5%), autosomal recessive polycystic kidney disease (12.5%), congenital nephrotic syndrome (25%), and repeated AKI from congenital cyanotic heart disease (12.5%). The median duration of CRRT with CARPEDIEM™ was 31.9 [22.0, 49.7] days with a little over half (60%) requiring blood prime. The ability to bridge neonates with ESKD to alternative renal replacement therapy (RRT) such as PD is critical as this is the recommended mode of RRT for infants [20]. Neonates who receive PD rather than HD have lower rates of hypertension and anemia, highlighting the importance of optimizing long term PD [20]. Given improved outcomes of long-term PD as well as the eventual need for renal transplantation, safely providing prolonged CRRT with CARPEDIEM™ while optimizing PD site healing and limiting blood product exposure is critical for this high-risk group of neonates with ESKD.

Growth Patterns

Identifying malnutrition in children receiving CRRT is critical as it can lead to long term growth delays, neurocognitive delays, and increased mortality [2]. However, characterization of growth is not standardized and has been described in the literature with both absolute (e.g. g/kg/day, g/day, cm/week) and relative (e.g. change in z-score) values, leading to significant variation in reporting and interpretation of growth patterns [21, 22]. In our study, the analysis of absolute and relative growth patterns differed. Absolute changes in weight suggested an initial decline in weight gain velocity, likely due to initial fluid positive status. Weight gain velocity nadired at week 3 followed by a positive weight gain through week 6. Sensitivity analysis found sustained weight gain was a delayed response occurring 14 days after start of CARPEDIEM^TM which supports the growth trends reported by Sutherland et al of 7 ESKD infants (86% CAKUT) who either primarily started on RRT with continuous ultrafiltration therapy or first failed a trial of PD [23]. Given difficulties performing accurate length measurements in this population, failure to gain weight can be more sensitive for poor overall growth, especially when occurring in the first few months of life [24]. Infants are most sensitive to weight failure in their first 3 months of life: if they do not experience weight gain for 2 weeks they can lose 1 weight centile, and if this persists for 4 weeks they can lose 2 weight centiles [25]. Our cohort’s delay in achieving sustained growth is likely due to multiple factors including an initial period of CRRT stabilization to reach euvolemia, delays or challenges in titrating nutritional fluids to deliver adequate nutrition, and the potential unclear impact of metabolic alterations on CRRT that may limit nutrient utilization for growth.

Previous work has shown impaired linear growth among neonates receiving RRT within 1 month of life, irrespective of modality [20]. Our cohort did not achieve positive linear velocity until month 2, but linear growth remained below goal of 0.8-1cm/week [21]. Factors affecting linear growth were likely similar to those contributing to delays in sustained weight gain. Additionally, neonates with CKD may experience unique factors such as growth hormone insensitivity and a delayed infantile phase, leading to a notable reduction in their height potential [24, 26–30].

Linear growth can also be impacted by inaccurate length measurements, so weight and OFC have been found to be more sensitive markers of poor growth among infants with CKD than length [25]. Similar to linear growth, relative OFC growth in our cohort was also worse by month 2 but began improving by month 6. However, these infants had positive OFC growth velocity starting in month 1 (+0.68 [0.00-1.79] cm/week, slightly above goal for age of 0.4-0.5 cm/week) and remained appropriate through month 6 (+0.24 [-0.26,0.43] cm/week, slightly above goal for age of 0.12-0.2 cm/week) [21].

Analyzing growth and survivorship in our cohort, survivorship correlated with linear z-score at the end of treatment and approached significance with OFC z-score, possibly suggesting that growth is a delayed outcome impacted by survivorship bias and not solely based on adequate nutritional provision earlier in the clinical course [31]. We found both pre- and post-weight z-scores were significantly higher in non-survivors compared to survivors (r=-0.84515, p=0.0082), possibly suggesting worse fluid overload at time of CARPEDIEM^TM start and stop. However, we did not find a statistically significant relationship between survivorship and percent fluid overload by weight at CRRT start (r=-0.39, p=0.33). Defining an infant’s euvolemic weight and degree of fluid overload is challenging, especially in the context of postnatal physiologic fluid shifts with loss of extracellular total body water, which can be as high as 90% in preterm infants [2, 32, 33]. The impact of oliguria and anuria on physiologic initial weight loss is unclear and has not been described, adding complexity to interpretation of our findings.

Our study also was impacted by using growth curves based on gestational age and sex without accounting for growth impairment that can be inherent to the ESKD disease process. Linear growth may be more severely impaired in CAKUT and ESKD due to a combination of inadequate protein and caloric intake, losses of water and electrolytes, metabolic acidosis, and poorly controlled mineral bone disease, despite the use of growth hormone [28, 32]. Adequate nutrition is necessary but may not be sufficient for normal or catch-up growth in dialyzed neonates [20, 34]. Slagle et al also found non-ESKD infants with CAKUT had worse severe growth impairment compared to their counterparts on dialysis [35]. Those on dialysis experienced the most prominent growth failure in the first 2 months of life but had better catch-up growth by 1 year of life than their non-dialysis peers [35]. However, given infants with CAKUT demonstrate variable degrees of renal dysfunction, it would be difficult and likely inaccurate to develop CAKUT-specific growth charts. The Pediatric Renal Nutrition Taskforce currently recommends the use of worldwide standard growth charts at this time as the goal is to obtain growth that is as close to normal as possible [24]. Interpretation of relative growth is challenging when disease-specific growth charts do not exist as growth velocity slower than the growth chart population will progressively lead to worsening discrepancy and declining z-score.

nPNA and Nitrogen Balance

Achieving a positive nitrogen balance supports nutritional growth by ensuring patients receive adequate protein to promote an anabolic state. This can be difficult in patients on CRRT due to increased uremic burden and hypercatabolic state [15]. Dietary protein intake is an independent predictor of survival among patients requiring CRRT, and nPNA can predict hospitalization and mortality even among patients with an adequate dialysis dose (Kt/V) [18, 36, 37]. Adults on CRRT demonstrate hypercatabolism with negative protein balance and median nPNA around 1.57 g/kg/day [15]. While there was no significant difference in Randerson nPNA by survivorship, this was likely impacted by low sample size. Our patients achieved their protein delivery goals while on CRRT allowing a sustained positive nitrogen balance while on CARPEDIEM^TM.

Nutritional Fluids

A particular challenge in caring for neonates with oliguric renal failure and/or fluid overload is the ability to deliver optimal nutrition in the minimum amount of fluid. Generally, infants can require around 150 ml/kg/day of nutritional provision to achieve growth goals [24]. Especially among neonates, even subtle differences between prescribed and delivered nutrition can substantially impact infant nutrition and growth outcomes [38, 39]. Yet, provision of nutritional fluids at 150 ml/kg/day in this subset of neonates may lead to substantial difficulties with volume management. Therefore, with the careful use of concentrated nutritional fluids, our analysis shows that only 85 ml/kg/day and 90 ml/kg/day are required to meet protein and caloric goals, respectively. Since inadequate caloric intake leads to increased reliance on protein catabolism for energy, a minimum delivered nutritional fluid of 90 ml/kg/day is necessary to promote growth goals [16]. This will require parenteral nutrition and/or fortification of feeds. As infants become older and more clinically stable, fortification of feeds can optimize time off RRT to ensure appropriate development and rehabilitative therapies [24].

Survivorship in our cohort correlated with the ability to deliver nutritional fluids of at least 90 ml/kg/day (p=0.0001). Non-survivors may have been unable to liberalize nutritional fluids due to CRRT or access challenges, frequent treatment interruptions, and/or significant clinical instability (e.g., hypotension). Survivorship correlated with enteral feed volume (p=0.037), enteral protein (p=0.037), and enteral calories (p=0.0075), likely also a reflection of patient clinical stability. Survivorship was likely impacted by the overall clinical status and baseline morbidities rather than the ability to meet early nutritional goals. Achieving nutritional goals early in CRRT may lead to longer-term benefits with regards to growth rather than to initial survivorship on CRRT.

Access and Anticoagulation

The CARPEDIEM™ system has been previously reported using systemic heparin anticoagulation during most sessions (71%) with no serious adverse events [8]. Most neonates in our cohort were dialyzed using a 7.5 Fr catheter and started on RCA with some receiving VTE prophylaxis. We were able to successfully anticoagulate our patients using RCA shown by median filter life 21.8 [17.0, 23.4] hours and patients receiving 97.7% [86.7%, 103.2%] of their prescribed treatment. Pediatric CRRT studies using RCA show longer hemofilter survival compared to systemic heparin anticoagulation and a better safety profile including lower rates of bleeding complications than UFH, but this has yet to be reported in neonatal dialysis [40]. Potential challenges using RCA with the CARPEDIEM™ system include calcium-containing dialysate and delayed citrate metabolism in infants, especially premature neonates or those with liver dysfunction [40]. There is limited data about VTE prophylaxis, but currently UFH 10 units/kg/h is used in clinical practice [5, 41]. A Cochrane review found UFH for VTE prophylaxis did not have a statistically significant difference in new or extended IVH [41]. Newer anticoagulation agents for this population have been reported including prostacyclin and bivalirudin [42].

Strengths and Limitations

The strengths of our study include prospective quantification of delivered nutrition and collection of detailed CRRT circuit characteristics using existing institutional databases. Growth was analyzed using both actual and relative values to provide a comprehensive overview of our patients. While our study limitations include a small number of patients (n=8), this is comparable to previously published studies on neonatal CRRT. Due to the observational nature of this study, some infants did have missing anthropometric measurements (4.6% missing), possibly due to clinical instability, or missing lab values needed to calculate nPNA (10.8% missing) which led to incomplete data. As there is no validated nPNA equation for pediatric patients receiving CRRT, our study used the Randerson II equation modelled off CAPD data, which reflects a form of continuous dialysis albeit a different modality but should be similar in nature to CRRT. Effluent urea used in this equation may have been impacted by our institutional practice of measurement with morning labs, which typically correlated with the end of a planned circuit change for CARPEDIEM™. It is possible that protein caking on the filter towards the end of a 24-hour treatment could lead to a decreased sieving coefficient and subsequent decreased measured effluent urea, urea generation rate, and nPNA. Lastly, most of our patients were neonates with ESKD which may limit external applicability to neonates with AKI requiring CRRT using the CARPEDIEM™ system.

CARPEDIEM™ is a safe and efficacious bridge to PD in neonatal ESKD. To our knowledge, this is the first report to define exact nutritional provision delivered to infants who received CRRT with the CARPEDIEM™ system. Our study found survivorship correlated with the ability to deliver nutritional fluids of at least 90 ml/kg/day which should be sufficient to provide full caloric and protein needs using parenteral nutrition. Analysis with both absolute and relative growth parameters suggest that growth velocity of infants on CRRT appears delayed despite the delivery of adequate calories and protein. Advanced measures to assess fluid status and nutritional intake, such as nPNA, are needed. More investigations are necessary to determine other CRRT variables such as micronutrient losses or metabolic alterations limiting growth as well as the impact of oliguria and anuria on physiologic loss of total body water in the postnatal period.

No conflict of interests exist. No funding was received for this work.

Misurac J (2017) Chronic kidney disease in the neonate: etiologies, management, and outcomes. Semin Fetal Neonatal Med 22:98–103. https://doi.org/10.1016/j.siny.2016.09.003
Sanderson KR, Warady BA (2020) End-stage kidney disease in infancy: an educational review. Pediatr Nephrol 35:229–240. https://doi.org/10.1007/s00467-018-4151-8
Ranchin B, Plaisant F, Demède D, et al (2021) Review: Neonatal dialysis is technically feasible but ethical and global issues need to be addressed. Acta Paediatr 110:781–788. https://doi.org/10.1111/apa.15539
Symons JM, Brophy PD, Gregory MJ, et al (2003) Continuous renal replacement therapy in children up to 10 kg. Am J of Kidney Dis 41:984–989. https://doi.org/10.1016/S0272-6386(03)00195-1
Hepponstall M, Chan A, Monagle P (2017) Anticoagulation therapy in neonates, children and adolescents. Blood Cells, Mol, and Dis 67:41–47. https://doi.org/10.1016/j.bcmd.2017.05.008
Inder TE, Perlman JM, Volpe JJ (2018) Intracranial Hemorrhage. In: Volpe’s Neurology of the Newborn. Elsevier, pp 593-622.e7
Garzotto F, Zanella M, Ronco C (2014) The Evolution of Pediatric Continuous Renal Replacement Therapy. Nephron Clin Pract 127:172–175. https://doi.org/10.1159/000363204
Garzotto F, Vidal E, Ricci Z, et al (2020) Continuous kidney replacement therapy in critically ill neonates and infants: a retrospective analysis of clinical results with a dedicated device. Pediatr Nephrol 35:1699–1705. https://doi.org/10.1007/s00467-020-04562-y
Hay WW, Thureen P (2010) Protein for Preterm Infants: How Much is Needed? How Much is Enough? How Much is Too Much? Pediatr & Neonatol 51:198–207. https://doi.org/10.1016/S1875-9572(10)60039-3
Frondas-Chauty A, Simon L, Branger B, et al (2014) Early growth and neurodevelopmental outcome in very preterm infants: impact of gender. Arch Dis Child Fetal Neonatal Ed 99:F366–F372. https://doi.org/10.1136/archdischild-2013-305464
Ehrenkranz RA, Dusick AM, Vohr BR, et al (2006) Growth in the Neonatal Intensive Care Unit Influences Neurodevelopmental and Growth Outcomes of Extremely Low Birth Weight Infants. Pediatrics 117:1253–1261. https://doi.org/10.1542/peds.2005-1368
Murphy HJ, Cahill JB, Twombley KE, Kiger JR (2018) Early Continuous Renal Replacement Therapy Improves Nutrition Delivery in Neonates During Extracorporeal Life Support. J Ren Nutr 28:64–70. https://doi.org/10.1053/j.jrn.2017.06.008
Zappitelli M, Goldstein SL, Symons JM, et al (2008) Protein and calorie prescription for children and young adults receiving continuous renal replacement therapy: A report from the Prospective Pediatric Continuous Renal Replacement Therapy Registry Group: Crit Care Med 36:3239–3245. https://doi.org/10.1097/CCM.0b013e31818f3f40
Wong Vega M, Juarez Calderon M, Tufan Pekkucuksen N, et al (2019) Feeding modality is a barrier to adequate protein provision in children receiving continuous renal replacement therapy (CRRT). Pediatr Nephrol 34:1147–1150. https://doi.org/10.1007/s00467-019-04211-z
Ganesan MV, Annigeri RA, Shankar B, et al (2009) The Protein Equivalent of Nitrogen Appearance in Critically Ill Acute Renal Failure Patients Undergoing Continuous Renal Replacement Therapy. J Ren Nutr 19:161–166. https://doi.org/10.1053/j.jrn.2008.11.009
Macias WL, Alaka KJ, Murphy MH, et al (1996) Impact of the Nutritional Regimen on Protein Catabolism and Nitrogen Balance in Patients With Acute Renal Failure. JPEN J Parenter Enteral Nutr 20:56–62. https://doi.org/10.1177/014860719602000156
Johnson RJ, Warady BA (2013) Long-term neurocognitive outcomes of patients with end-stage renal disease during infancy. Pediatr Nephrol 28:1283–1291. https://doi.org/10.1007/s00467-013-2458-z
Juarez-Congelosi M, Orellana P, Goldstein SL (2007) Normalized Protein Catabolic Rate Versus Serum Albumin as a Nutrition Status Marker in Pediatric Patients Receiving Hemodialysis. J Ren Nutr 17:269–274. https://doi.org/10.1053/j.jrn.2007.04.002
Morse S, Groer M, Shelton MM, et al (2015) A Systematic Review: The Utility of the Revised Version of the Score for Neonatal Acute Physiology Among Critically Ill Neonates. J Perinat Neonatal Nurs 29:315–344. https://doi.org/10.1097/JPN.0000000000000135
van Stralen KJ, Borzych-Dużalka D, Hataya H, et al (2014) Survival and clinical outcomes of children starting renal replacement therapy in the neonatal period. Kidney Int 86:168–174. https://doi.org/10.1038/ki.2013.561
Fenton TR, Chan HT, Madhu A, et al (2017) Preterm Infant Growth Velocity Calculations: A Systematic Review. Pediatrics 139:e20162045. https://doi.org/10.1542/peds.2016-2045
Simon L, Hanf M, Frondas-Chauty A, et al (2019) Neonatal growth velocity of preterm infants: The weight Z-score change versus Patel exponential model. PLoS ONE 14:e0218746. https://doi.org/10.1371/journal.pone.0218746
Sutherland SM, Davis AS, Powell D, et al (2022) Kidney Replacement Therapy in Low Birth Weight Preterm Newborns. Pediatrics 150:e2022056570. https://doi.org/10.1542/peds.2022-056570
Shaw V, Anderson C, Desloovere A, et al (2023) Nutritional management of the infant with chronic kidney disease stages 2–5 and on dialysis. Pediatr Nephrol 38:87–103. https://doi.org/10.1007/s00467-022-05529-x
Shaw V, Polderman N, Renken-Terhaerdt J, et al (2020) Energy and protein requirements for children with CKD stages 2-5 and on dialysis–clinical practice recommendations from the Pediatric Renal Nutrition Taskforce. Pediatr Nephrol 35:519–531. https://doi.org/10.1007/s00467-019-04426-0
Haffner D (2020) Strategies for Optimizing Growth in Children With Chronic Kidney Disease. Front Pediatr 8:399. https://doi.org/10.3389/fped.2020.00399
Mehls O, Wühl E, Tönshoff B, et al (2008) Growth hormone treatment in short children with chronic kidney disease. Acta Paediatrica 97:1159–1164. https://doi.org/10.1111/j.1651-2227.2008.00845.x
Drube J, Wan M, Bonthuis M, et al (2019) Clinical practice recommendations for growth hormone treatment in children with chronic kidney disease. Nat Rev Nephrol 15:577–589. https://doi.org/10.1038/s41581-019-0161-4
Mekahli D, Shaw V, Ledermann SE, Rees L (2010) Long-Term Outcome of Infants with Severe Chronic Kidney Disease. Clin J Am Soc Nephrol 5:10–17. https://doi.org/10.2215/CJN.05600809
Greenbaum LA, Muñoz A, Schneider MF, et al (2011) The Association between Abnormal Birth History and Growth in Children with CKD. Clinical Journal of the American Society of Nephrology 6:14–21. https://doi.org/10.2215/CJN.08481109
Goldstein SL, Akcan-Arikan A, Alobaidi R, et al (2022) Consensus-Based Recommendations on Priority Activities to Address Acute Kidney Injury in Children: A Modified Delphi Consensus Statement. JAMA Netw Open 5:e2229442. https://doi.org/10.1001/jamanetworkopen.2022.29442
Nelms CL, Shaw V, Greenbaum LA, et al (2021) Assessment of nutritional status in children with kidney diseases—clinical practice recommendations from the Pediatric Renal Nutrition Taskforce. Pediatr Nephrol 36:995–1010. https://doi.org/10.1007/s00467-020-04852-5
Young A, Brown LK, Ennis S, et al (2021) Total body water in full-term and preterm newborns: systematic review and meta-analysis. Arch Dis Child Fetal Neonatal Ed 106:542–548. https://doi.org/10.1136/archdischild-2020-321112
Laakkonen H, Happonen J-M, Marttinen E, et al (2010) Normal growth and intravascular volume status with good metabolic control during peritoneal dialysis in infancy. Pediatr Nephrol 25:1529–1538. https://doi.org/10.1007/s00467-010-1535-9
Slagle CL, Riddle SL, McNelis K, Claes D (2022) Single-Center Experience on Growth in Infants Born With End-Stage Kidney Disease. J Ren Nutr S1051227622001686. https://doi.org/10.1053/j.jrn.2022.09.007
Kritmetapak K, Peerapornratana S, Srisawat N, et al (2016) The Impact of Macro-and Micronutrients on Predicting Outcomes of Critically Ill Patients Requiring Continuous Renal Replacement Therapy. PLoS One 11:e0156634. https://doi.org/10.1371/journal.pone.0156634
Kalantar-Zadeh K, Supasyndh O, Lehn RS, et al (2003) Normalized protein nitrogen appearance is correlated with hospitalization and mortality in hemodialysis patients with Kt/V greater than 1.20. J Ren Nutr 13:15–25. https://doi.org/10.1053/jren.2003.50005
Iglesias SBDO, Leite HP, Meneses JFSE, De Carvalho WB (2007) Enteral Nutrition in Critically Ill Children: Are Prescription and Delivery According to Their Energy Requirements? Nutr Clin Pract 22:233–239. https://doi.org/10.1177/0115426507022002233
Ainsworth S, Furness J, Fenton A (2007) Randomized comparative trial between percutaneous longlines and peripheral cannulae in the delivery of neonatal parenteral nutrition. Acta Paediatrica 90:1016–1020. https://doi.org/10.1111/j.1651-2227.2001.tb01357.x
Rico MP, Fernández Sarmiento J, Rojas Velasquez AM, et al (2017) Regional citrate anticoagulation for continuous renal replacement therapy in children. Pediatr Nephrol 32:703–711. https://doi.org/10.1007/s00467-016-3544-9
Shah PS, Shah VS (2008) Continuous heparin infusion to prevent thrombosis and catheter occlusion in neonates with peripherally placed percutaneous central venous catheters. Cochrane Database of Systematic Reviews. https://doi.org/10.1002/14651858.CD002772.pub3
Deep A (2022) Anticoagulation strategies in continuous kidney replacement therapy — does one size fit all? Pediatr Nephrol 37:2525–2529. https://doi.org/10.1007/s00467-022-05567-5

Download PDF

Journal Publication

published 17 Jan, 2024

Read the published version in Pediatric Nephrology →

Editorial decision: Major Revisions Needed
31 Aug, 2023
Reviewers agreed at journal
08 Aug, 2023
Reviewers invited by journal
08 Aug, 2023
Editor assigned by journal
08 Aug, 2023
First submitted to journal
07 Aug, 2023

You are reading this latest preprint version

Clearance and Nutrition in Neonatal Continuous Renal Replacement Therapy (CRRT) Using the CARPEDIEM™ System

Status:

Journal Publication

Version 1

Abstract

Figures

Introduction

Methods

Results

Patient and Maternal Characteristics

Discussion

Conclusions

Declarations

References

Status:

Journal Publication

Version 1