In recent years, much attention has been focused on enhancing the nutritional support of premature infants to improve both survival and quality of life. Greater efforts have been made to improve neonatal nutrition in very low birth weight (VLBW) and extremely low birth weight (ELBW) infants. It has been suggested that the early weeks of life represent a window of opportunity to avoid cumulative nutritional deficits to improve early growth, reduce postnatal growth restriction, and promote long-term development. The nutritional requirements of ELBW infants (<1000 g) have recently been reviewed (1–4), and the association between early protein and energy intake and neurodevelopment and growth outcomes has been suggested (5–7).
Because of immaturity and clinical conditions, parenteral nutrition (PN) is frequently the primary or exclusive route for providing nutritional support during the early weeks of life in preterm infants. Parenteral solutions can be prescribed using 1 of 2 methods: individually tailored or standardized solutions. Tailored solutions are based on the principle that no single parenteral regimen can be ideal for all patients because of wide variations of pathological processes, varying ages of patients, and changing nutritional requirements of patients throughout their disease process. Because of the need for daily preparation, its relative unavailability during the first hours of life, nights, and weekends, and the risks of nutrient imbalances and compounding errors, the use of standardized solutions containing more fixed amounts of each component per unit of volume has been proposed (8–17). The use of standardized solutions has increased rapidly in the neonatal intensive care unit (NICU) because premature infants do seem able to tolerate mild or moderate variations in nutritional intake (10). Recent studies even suggest that the appropriate use of standardized solutions from the first day of life increases nutritional intake closer to requirements and reduces the incidence of postnatal growth restriction observed in VLBW infants (8,9,11,15,16,18).
In adults, ready-to-use (RTU) industrially manufactured multichamber bags (MCBs) containing 3 sterilized macronutrient solutions in separate chambers of a single, closed plastic system are widely available and have been used for more than 10 years (19,20). The guaranteed sterility and longer shelf life of these MCB systems are major technological advances. RTU MCBs minimize the risks of nutrient imbalances and inadvertent contamination during compounding and storage (20). Consequently, an international group of experts in pediatric nutrition suggested that a standardized total parenteral nutrition (TPN) solution in an MCB system could be satisfactory for the nutritional support of pediatric patients, with minimal individualized adjustments in this patient population, through 3 different steps: choice of the amount of energy and protein, in gram per kilogram per day, to achieve intakes closer to individual needs and in accordance with recommendations (1–4); optional activation of the lipid compartment in the MCB system; and the ability to make minimal additions of water, macronutrients, electrolytes, and/or minerals according to the clinical and biological assessment of patients. Such additions should be pretested by the manufacturer to provide maximum limits for each addition.
The primary objective of this study was to evaluate the safety of an industrially manufactured MCB system specially designed for preterm infants for practical therapeutic use in the NICU. For that purpose, the ability to provide daily nutritional requirements in the range of the recommended values, the clinical use of optional activation of the lipid compartment in the MCB system, and the flexibility for individual adaptation were determined. The secondary objectives were to provide additional information on weight gain, biological parameters, and adverse events during the study period.
This prospective, multicenter, noncomparative, open-label phase III clinical trial was conducted in accordance with good clinical practice guidelines, the Declaration of Helsinki (2000 version), and national regulations. The study used a 3-chamber bag system (Ped3CB) (Fig. 1) that was specifically designed for premature infants (Ped3CB-A; Baxter, Lessines, Belgium). The Ped3CB bag is a non-PVC bag. The study was designed in consultation with European regulatory authorities, which recommended that the study be conducted to provide daily information on the performance safety of the Ped3CB in practical therapeutic use in preterm infants. Because all of the components of Ped3CB were already approved and determined to be safe, the study was not designed to test efficacy or safety of the individual nutritional agents. Five NICUs, 3 in France and 2 in Belgium, participated in this clinical trial.
The study protocol was approved by ethics committees, and written informed consent from parent(s) or a legal representative was obtained. Eligible patients included hospitalized preterm newborn infants younger than 37 weeks’ gestation requiring PN that provided at least 80% of their total estimated nutritional needs at baseline. The duration of use of Ped3CB-A was a minimum of 5 consecutive days and could be extended to a maximum of 10 days. Patients with nutritional requirements that could not be met with the Ped3CB-A formula were excluded.
Ped3CB-A Solution and Prescription
The prescription of the PN regimens for the preterm infants was based on individual patient's estimated calorie and protein needs, and parenteral or enteral intakes following the protocol of the participating NICUs. The volume of Ped3CB plus any supplements was chosen to administer the required amounts of fluid, macronutrients, and micronutrients. When lipid-free PN was required or when the estimated lipid needs of patients were lower than the amount provided by the ternary admixture, the lipid chamber was not activated and a lower lipid intake could be provided separately using a Y-line, according to local practice. The composition of the Ped3CB-A was relatively concentrated, providing 3.1 g amino acids (AA), 13.3 g glucose, 2.5 g lipid, and 91 kcal/100 mL with the ternary admixture and 3.9 g AA, 16.7 g glucose, and 83 kcal/100 mL with the binary admixture. Designed to be used from the first days of life, sodium and potassium content was limited (Table 1). The volume of the binary or ternary admixture could be adjusted directly in the chamber bag or by using a Y-line with free water, electrolytes, or minerals according to daily individual requirements. The total daily fluid volume administered from the bag and from any addition was determined by the prescribing physician and was recorded in the report form. A list of pretested additions to the Ped3CB-A, with maximum limits for each addition, was provided by the manufacturer. For example, for activation of the 3-in-1 bag, additional supplementation of sodium, potassium, calcium, and organic phosphate can be added, increasing concentrations by 78%, 67%, 118%, and 78%, respectively. Vitamins and trace elements are not included in the Ped3CB-A bag and were added to the activated bag or provided separately, according to the general practice of the NICU.
The Ped3CB-A was dispensed by hospital pharmacists to clinical departments for use in the NICU. According to usual clinical and local practice, the RTU parenteral solution could be adapted to the estimated individual nutrition requirements in the hospital pharmacy, the NICU, or directly at the bedside with a Y-line. The Ped3CB-A formulation was administered through an umbilical catheter, central venous catheter, peripherally inserted central catheter (PICC), or peripheral line. Centers were permitted to follow their individual institution's routine and practice regarding line selection, photooxidation prevention, and filter usage. These aspects were not required as components of the study protocol. Osmolality of the Ped3CB-A solution could be reduced by adding free water for peripheral administration or nonoptimal situation of the tip of the catheter. Nutritional intakes and all additional supplements were recorded daily in the report form.
Practical Handling and Ease of Use
The clinical (nursing and pharmacy) staff members responsible for preparing and administering Ped3CB-A compared the bag's practical handling and ease of use with those of the standard clinical practices for providing PN at their site. These qualities were measured daily during the treatment period using a questionnaire and a visual analog scale (VAS) through which staff members indicated their assessments during the treatment period by ticking a 10-cm straight line. Tick marks placed toward the left side of the line indicated a lower performance rating of the Ped3CB system as compared with their standard practice; tick marks toward the right side of the line indicated a better performance rating of the Ped3CB system. The questionnaire and VAS were specifically created for the study.
Staff members used the VAS to compare the Ped3CB system with individual bottles suspended at the bedside, bags compounded in the ward, RTU compounded bags prepared by the pharmacy or a service provider (outside of the ward), and tailored premixes (ie, standard formulations prepared by the pharmacy for tailoring by the nursing staff according to the patient's needs). The tick marks were converted to centimeters by data entry operators; a measure of 10 cm represented the best performance rating of the Ped3CB system compared with standard practice.
Body weight was recorded daily, and weight gain was evaluated during the study period.
To evaluate the safety of the Ped3CB system, all of the adverse events were recorded from the start of the first Ped3CB-A infusion (baseline) until 2 days after the last infusion. Vital signs and adverse events were recorded daily throughout the study. Biological laboratory parameters, including plasma ionogram (sodium, potassium, calcium, and phosphorus), urea, triglycerides, glucose, and bicarbonates, were evaluated at baseline, on day 5, and on day 10 or at the end of treatment. In addition, any other abnormal laboratory values were reported as adverse events in the report form.
To collect data on a relevant number of days of PN use in an NICU, 100 preterm infants per protocol analyzed were estimated to be representative. The investigators planned to enroll 115 preterm infants to account for an estimated dropout rate of approximately 15%. Demographic and baseline characteristics were analyzed using descriptive statistics. The numbers of all-in-one admixtures provided as binary or ternary solutions during the treatment period were calculated. Mean PN intake, maximum PN intake, and mean oral intake were calculated and compared with actual nutritional recommendations for preterm infants. Changes in safety variables, including clinical laboratory test results and vital signs, were analyzed using analysis of variance. Weight gain during the study was evaluated based on mean body weight during the study period and analyzed using descriptive statistics and analysis of variance. The occurrence of clinical and biological adverse events was summarized globally.
Between February 2008 and November 2008, 113 preterm infants were enrolled in the study from 3 NICUs in France and 2 NICUs in Belgium. Of these infants, 16 were excluded from the analysis because their gestational age was older than 37 weeks (n = 5), their medication was incompatible with Ped3CB-A (n = 3), they had a clinical condition for which Ped3CB-A was contraindicated (n = 3); their oral nutrition rates increased rapidly (n = 2), they were transferred to another hospital during the study period (n = 2), or they lacked a Social Security identifier (n = 1). In all, 97 preterm infants were included in the per protocol analysis.
Of the 97 preterm infants whose results were included in the analysis, 29 were ELBW, 33 were VLBW, and 35 were preterm infants weighing more than 1500 g. Patients were grouped based on their postnatal age at inclusion: from birth day through day 3, from day 4 through day 7, and after day 7 (Table 2). Low birth weight, respiratory distress syndrome (RDS), and gastrointestinal (GI) problems were the main indications for PN in the study population (Fig. 2).
Ped3CB-A Compliance and Flexibility
Administration Route and Bag Activation
Study participants received a total of 854 days of PN through an umbilical catheter (n = 157 days; 18.4%), peripheral line (n = 88 days; 10.3%), or PICC (n = 610 days; 71.4%). Only 32 PN infusion days (3.7%) in 11 infants (11.3%) involved double-chamber bag activation as a result of delayed lipid infusion practice (n = 3), an increase in oral nutrition intake (n = 7), or persistent hypertriglyceridemia from study enrollment (n = 1).
Vitamins and Trace Elements
All of the infants received water-soluble and lipid-soluble vitamins during the study. Water-soluble vitamin supplementation occurred on 84.9% of infusion days, and lipid-soluble vitamins were infused on 91.5% of the 822 infusion days that included IV lipids. Overall, vitamin preparations were added to the activated Ped3CB-A in 65.9% of additions of water-soluble vitamins and 88.7% of additions of lipid-soluble vitamins. Trace elements were provided on 78.2% of infusion days in 93 of the 97 infants, but addition of zinc alone was provided during an additional 19.4% of infusion days.
Addition of Water for Injection
In 53 infants, water for injection was added to the bag on 221 (26.0%) infusion days by peripheral line (82 of 88), umbilical catheter (44 of 157), or PICC (95 of 610) to reduce osmolality or increase water supply.
Macronutrient (AA, Glucose, and Lipid) Supplementation
Table 3 depicts macronutrient supplementation. Ten infants received additional AA, accounting for 17 infusion days and providing 0.69 ± 0.28 g · kg−1 · day−1 AA supplementation. Glucose supplementation occurred in only 7 infants, accounting for 34 PN infusion days and providing 3.2 ± 2.0 g · kg−1 · day−1. In 4 infants, accounting for 9 infusion days, the lipid chamber was not activated and lipids (1.12 ± 0.43 g · kg−1 · day−1) were provided through a separate perfusion using a Y-line. In 1 infant, limited lipid quantities (0.23 and 0.43 g · kg−1 · day−1) were added to the triple-chamber solution during the first 2 days. In all, supplementation directly to the triple-chamber bag occurred on only 18 infusion days (2.1%) in 11 infants. In 1 infant, AA and glucose were added simultaneously.
Electrolyte and Mineral Supplementation
Table 3 depicts electrolyte and mineral supplementation. Minerals or electrolytes, mainly sodium, were added on 387 infusion days (45.3%) in 65 infants, primarily using a Y-line. In all, compared with the intakes provided by the triple-chamber bags, electrolyte and mineral supplementation was negligible except for sodium (1.63 mmol · kg−1 · day−1). Chloride supplementation was the result of NaCl or KCl supplementation (1.85 mmol · kg−1 · day−1) and was also in the range of the amounts provided by Ped3CB-A. Additions directly to Ped3CB-A occurred on 185 (21.7%) infusion days in 36 infants (37.1%).
Use of Ped3CB-A for Macronutrient, Electrolyte, and Mineral Supplementation
Macronutrient, electrolyte, and mineral supplements were administered primarily through a Y-line. A total of 199 additions were made to the Ped3CB-A bags on 197 infusion days (23.1%) in 43 infants (44.3%). Multiple additions to the bag accounted for only 1 perfusion day.
As shown in Table 3, mean intakes of administered parenteral nutrients from the triple-chamber bag solution and from supplements were AA 2.8 g · kg−1 · day−1 and 81 kcal · kg−1 · day−1. The mean maximum intakes from Ped3CB-A were AA 3.6 g · kg−1 · day−1 and 104 kcal · kg−1 · day−1, within the actual recommended ranges for this population (2,4).
Enteral nutrition increased progressively during the study. Combined with enteral nutrition, mean protein and energy intakes represented AA 3.6 g · kg−1 · day−1 and 115 kcal · kg−1 · day−1, respectively (Table 4). Enteral protein and energy intakes accounted for 17.8% to 25.4% of total protein equivalent and 24.9% to 34.4% of total energy, respectively.
Weight gain and nutrient intakes were evaluated based on postnatal age at study enrollment (Table 4). In preterm infants included during the first 3 days of life, weight gain represented 10.0 ± 9.5 g · kg−1 · day−1. In infants included between 4 and 7 days of life or thereafter, weight gain was higher and reached 21.5 and 22.6 g · kg−1 · day−1, respectively.
Practical Handling and Ease of Use
VAS scores for Ped3CB-A were significantly higher for ease of bag manipulation and prescription-to-infusion time than were the scores for bags compounded in the ward, RTU compounded bags, and tailored premixes. The highest scores were for manipulation when Ped3CB-A was compared with the use of ward-compounded bags (Fig. 3).
Mean values of the biological parameters evaluated during the study are shown in Table 5 and Figures 4 and 5. Plasma triglyceride concentrations remained mainly in the normal range (<250 mg/dL; Fig. 4). Hypertriglyceridemia was observed in 8 infants at inclusion, 7 infants at day 5, and 2 infants at the end of the study. Blood urea nitrogen (BUN) levels decreased during the study. Only 4 infants had a BUN level >25 mmol/L at the end of the study. Plasma glucose concentrations remained satisfactory during the study. Four infants had plasma glucose concentrations >10 mmol/L at day 5. Insulin was required in 13 infants on 34 infusion days. During the study, a moderate but significant increase in plasma bicarbonate was observed. Sodium, potassium, and calcium variability decreased during the study and remained in the normal range (Fig. 5). By contrast, relative hypophosphoremia was frequent during the study, occurring in 27.5% of infants at baseline and in 30.7% of infants at the end of the study. Phosphorus supplements were administered to 11 infants, providing a mean 0.79 mmol/kg and increasing mean serum inorganic phosphorus concentrations by 0.45 mmol/L.
In addition, a few abnormal biological parameters were reported as adverse events, including hyperglycemia (n = 13), hypernatremia (n = 2), hyponatremia (n = 9), hyperkalemia (n = 4), hyperphosphoremia (n = 4), hypophosphoremia (n = 4), hypercalcemia (n = 3), and hypertriglyceridemia (n = 5).
Four patients experienced 6 treatment-emergent serious adverse events consisting of herpes simplex infection, staphylococcal sepsis, severe emphysema, pneumothorax, septic shock, or catheter sepsis. One 9-day-old infant (27 weeks’ gestational age) died during the study after septic shock. These serious adverse events were not believed to be related to the infusion of PN. Additional adverse events included patent ductus arteriosus (7 events), sepsis (18 events), anemia (12 events), and constipation (8 events). The majority of adverse events were thought not to be related to the infusion of PN.
The present study shows that the Ped3CB-A system, the first standardized industrially prepared RTU triple-chamber PN system designed for commercial use in preterm infants, is easy to use, has the flexibility required to meet the varied and changing nutritional needs of preterm infants, provides levels of nutrients within the ranges of recent recommendations (2,4), produces appropriate weight gain rates, and is safe.
Prescriptions of RTU, all-in-one–tailored PN solutions specially designed to satisfy the daily nutritional requirements of individual preterm infants appear to be the criterion standard to provide PN in the NICU, but they require optimal hospital pharmaceutical structures with day and night availability to safely provide, with minimal delay, PN to meet medical prescriptions and reformulations. Therefore, compounded RTU parenteral solutions prepared inside or outside the hospital pharmacy are frequently used in NICUs as “starter bags” or as standardized formulations immediately available after birth. In some units, 3 to 7 different RTU compounded PN solutions are available according to postnatal age and requirements (8–11). To improve stability and reduce peroxidation, trace elements and vitamins need to be added extemporaneously on a daily basis in the pharmacy or in the neonatal unit. In addition, to fulfill the nutritional requirements of an increasing population of preterm infants, supplementation with electrolytes and minerals is frequently requested. Thus, those standardized compounded parenteral solutions are not truly all-in-one solutions. Nevertheless, standardized PN, in general, is associated with fewer risks than tailored PN (21). Unlike individualized PN, standardized PN reduces the chance for potential errors with ordering and compounding; the risk of PN-associated infections; or adverse outcomes from the infusion of incompatible mixtures (21). In infants, standardized PN improves nutritional support and weight gain, and reduces electrolyte disturbances compared with individualized PN (8,9,12). Recently, it has been suggested that RTU compounded parenteral solutions are safe and that optimal nutritional intakes and adequate early growth rate can be achieved in preterm infants fed a RTU compounded parenteral solution with high nutritional density (12,18).
The use of multichamber PN bags has been suggested in adults (19,20); they are associated with few risks for virtually any indication or application (19). The industrially prepared PN MCBs offer the advantage of being designed in separate chambers, which reduces the interactions among the various components and allows storage at room temperature. Additionally, they are submitted to rigorous stability studies and checked according to good manufacturing practices before batches are released. The quality of the checks performed by the pharmaceutical industry is largely superior to that of the checks performed by hospital pharmacies. In hospital pharmacies, although all of the components of compounded or manually prepared PN admixtures undergo routine sterility and stability testing during the manufacturing process, subsequent extensive routine testing of the final admixtures is not performed. By contrast, during the manufacturing process for Ped3CB-A, the individual components and activated (mixed) bags are thoroughly tested for sterility, stability, free fatty acid levels, endotoxin levels, fat globule size, and degradation products. The testing of both the individual components and activated (mixed) bags provides an additional layer of safety that is not available from admixtures prepared from individual components. After reconstitution, the pediatric MCBs can be infused as is or can be specifically supplemented with adequate small, manufacturer-pretested volumes of micronutrients to deliver ready-to-infuse admixtures, whose final stability would be further ensured by manufacturer testing.
In the present study, the AA and energy intakes during the first week of life were in the range of “aggressive” nutrition recommendations for VLBW infants. In addition, the maximum parenteral AA and energy intakes coming from the bag and the total combined parenteral and enteral intakes recorded during the study were in the range of the most recent recommendations (2,4), demonstrating that clinicians were able to use and easily adapt the standardized Ped3CB-A system to meet the nutritional needs of preterm infants. These results suggest that the use of the Ped3CB-A in NICUs could be an important factor in helping to reduce the cumulative protein and energy deficit during the first few weeks of life.
The daily weight gain of approximately 22 g · kg−1 · day−1 observed in infants included after the first week of life is at the high end of the expected weight gain rate of 10 to 20 g · kg−1 · day−1 expected with PN (2). Early weight gain (10.0 g · kg−1 · day−1) was observed in the early inclusion group, suggesting that the use of the Ped3CB-A could reduce postnatal growth deficits (6,13,18) and improve long-term neurological development and growth outcomes (6,7).
The majority of prescription days were successfully fulfilled using Ped3CB-A as is, with the addition of vitamins and trace elements as usually performed in adult patients. Few additional supplementations into the bag, mainly sodium, were performed, representing only 23.1% of the perfusion days in 44.3% of the infants. Sodium supplementation was requested as sodium requirements in preterm infants range between 3 and 5 or 7 mmol · kg−1 · day−1, whereas the sodium content of the Ped3CB-A was specifically limited to be provided to preterm infants even during the first few days of life. In the future, the availability of 2 different MCBs, 1 designed for the first few days of life with low sodium and potassium content, as is the Ped3CB-A, and another with higher sodium content for VLBW infants after the third to the fourth day of life, could reduce the need for extra micronutrient supplementation. An increase in the sodium content of the Ped3CB-A, allowing an increase in the phosphorus and the calcium content, could also reduce the need for extra micronutrient supplementation without limiting its use during the first days of life. Indeed, it has been suggested that the use of early PN with high AA content in preterm infants increases the need for electrolyte supplies during the first 3 days of life (22). The results of our recent study in VLBW infants (18) support this hypothesis. Nevertheless, the question is still open and requires further investigation.
Mean values of the biological parameters evaluated during the study remained mainly in the normal range for preterm infants during the first weeks of life. Relative hypophosphoremia was observed at baseline and during the study, but it appears that adult reference levels were frequently used instead of preterm reference values. The rapid increase in fat intakes during the study provided by the Ped3CB-A was well tolerated, and the nonactivation of the lipid chamber was mainly the result of an increasing oral nutrition intake.
Adverse events were limited during the study, and none were directly attributable to the design of the Ped3CB system; all of the adverse events were consistent with the known clinical status and the invasive treatments required by clinical condition. Most of the adverse events were mild, and patient vital signs remained relatively stable throughout the study.
The main limitation of this prospective multicenter study is its noncomparative design. Nevertheless, this study confirms the results of our recent study (18) demonstrating that, in preterm infants with a BW <1250 g, the early use of an appropriate RTU compounded solution can provide nutritional intakes in the range of the more recent recommendations (ESPGHAN/ESPEN 2005 and ESPGHAN 2010). It enables the avoidance of both postnatal cumulative protein and energy deficits as well as postnatal growth restriction in ELBW and VLBW infants.
In conclusion, Ped3CB-A is the first standardized, industrially manufactured, triple-chamber bag developed to provide well-balanced and safe nutritional support in premature infants. This study demonstrated the ease of use of the Ped3CB system. Clinicians and staff were able to configure the system, including selective activation (triple- or double-chamber bag), and to provide additions and supplements to meet the nutritional needs of preterm infants. Early nutritional support, maximal nutrient intakes, and body weight gain were favorable, meeting the requirements and expectations in the more recent recommendations for preterm infants. The new design of the Ped3CB system provides well-balanced PN with the benefits of multiple manufacturing process controls, improving the safety of PN in preterm infants.
We thank Debby Berlyne, PhD, of ACCESS Medical, LLC, for editorial assistance. ACCESS Medical obtained funding from Baxter Healthcare for editorial assistance with this manuscript.