Manufacturing change in one of the two lipid emulsion products being studied.

This will be a prospective, randomized trial to determine if differences exist in the tolerance of lipid injectable emulsions in the neonatal intensive care unit (NICU). Lipid injectable emulsions are an essential nutrient for neonatal growth and development. Traditionally, lipid injectable emulsions have been commercially available in sterile glass bottles, but in April of 2004, a new container was introduced as a sterile plastic bag. In January, 2005, NICU personnel observed what appeared to be a higher than usual incidence of hypertriglyceridemia. Upon further laboratory investigation of the lipid injectable emulsions stored in glass bottles versus those in plastic, significant differences were noted in the population of large-diameter fat globules by globule size analysis, reflective of a less stable emulsion in plastic. The United States Pharmacopeia (USP) which sets the standards for drug purity and safety in the U.S., and whose drug monographs are enforceable by the FDA, has proposed to limit this large diameter fat globule population to a volume-weighted percent fat greater than five micrometers or PFAT5 to be less than 0.05% of the total lipid concentration. (At the present time, the USP monograph is not officially recognized, but is on track for adoption in 2006.) Our preliminary analyses of four lots of 20% lipid injectable emulsion packaged in glass to have a PFAT5 of 0.003±0.0008%, compared to an approximate 55-fold increase in the large-diameter fat globule population or 0.166±0.016% for an equal number of products packaged in plastic. We hypothesize this difference may explain the recent clinical observations. We will compare the incidence of hypertriglyceridemia in neonates between lipids packaged in glass versus those in plastic. The study will attempt to discern whether the differences in packaging influence the stability and subsequent tolerance of lipid injectable emulsions.

The study is a single-center (BIDMC), prospective, single-blinded, randomized controlled study involving neonatal ICU patients who will receive lipid injectable emulsion infusions as part of their normal clinical care. The dosing protocol for all neonates is weight- and age-based, as shown below:

Birth Weight (< 1000 g)

Age(Day of Life,DOL)- Lipid Dose g/kg/day:

DOL1- 0; DOL2- 1; DOL3- 1.5; DOL4- 2; DOL5- 2.5; DOL6+-3

Birth Weight (>1000 g)

Age-Lipid Dose g/kg/day:

DOL1- 0; DOL2- 1; DOL3- 2; DOL4- 3; DOL5- 3; DOL6+-3

Serum triglycerides is the primary outcome indicator of tolerance to the initiation and advancement of lipid injectable emulsion. Blood samples will be drawn before initiation of lipids at either DOL-0 or DOL-1, and then at DOL-4 and DOL-7.

The procedures for initiating lipid emulsions in neonates will follow the usual protocol as per NICU guidelines. On DOL-2, an order for 20% soybean oil lipid emulsion will be written by a physician or designated licensed assistant as per protocol. The Pharmacy Department will receive the order, and compound and dispense it in an appropriately sized syringe. Patients will receive lipids taken from either glass bottles or plastic bags according to a pre-assigned randomization scheme.

All lipid injectable emulsion lots in glass and plastic will be recorded as used in the study. In addition, all lots will be analyzed with respect to the globule size distributions in the Nutrition/Infection Laboratory.

Sample Size Justification:

Approximately 82 randomized patients are required for analysis of the differences between lipid injectable emulsion formulations.

The sample size was generated from the data obtained from the retrospective analysis described above in B2. A two-sided test based on a normal distribution, a nominal significance level of 0.05, and a statistical power of 0.80 were assumed. Based on these assumptions, the power calculation reveals an equal sample size for both groups is calculated to be 41. Hence, the total sample size of 82 patients is required to detect statistically significant differences in the incidence of hypertriglyceridemia between infusion groups.

Data Analysis:

We will be comparing lipid tolerance in critically ill neonates receiving pharmacy-prepared syringes of 20% soybean oil-based lipid injectable emulsion taken from commercially available, FDA-approved products that use either a glass or plastic packaging container. The primary outcome variable in assessing lipid tolerance will be serum triglycerides. A two-way analysis of variance (ANOVA) will be employed to assess the differences between lipid injectable emulsion products. The two independent variables are 1) Lipid Container (or Treatment) and, 2) Time (over 7 days); while the dependent variable will be serum triglycerides. If significant differences are detected by ANOVA, then individual differences will be further evaluated using Bonferroni tests for pair wise comparisons. All data will be expressed as the mean ± SD with statistical significance set at a p-value of <0.05. Analysis will be done at BIDMC under the supervision of the study PI. Although not expected to play a significant role, drop-out and missing data will be adjusted accordingly.

All patients who present to the neonatal intensive care unit at BIDMC who require intravenous nutritional support and provide informed consent will be included in the study. The decision to initiate intravenous nutritional support will be made by clinical personnel, and the dosing protocol for lipid injectable emulsion will follow institutional procedures as shown above in B3 (Description of Study Protocol) (section1-a). Patients will be randomized to receive one of two study parenteral lipid emulsions. As hydrocortisone therapy is rarely used in the treatment of refractory hypotension in neonates, and has been shown to induce hypertriglyceridemia in very low birth weight infants (younger than 29 weeks' gestational age at birth), such patients will be excluded from study. Due to randomization, patients enrolled in either group are expected to be similarly distributed with respect to gestational age, gender and race.

Potential study subjects will be identified by neonatal intensive care unit physicians at BIDMC who have determined intravenous nutrition support is clinically indicated. Parent(s) or guardian(s) of the NICU patient will review and sign a written informed consent form. The form will be obtained by NICU physicians or one of the co-investigators listed above. Prospective parents or guardians of the patient will be presented, in lay terms, the two options for providing intravenous lipid emulsions. The purpose of the study and current data regarding the standard of care will be presented to parents or guardians of the patient, and that assignment to either group (glass vs. plastic) will be random or determined by chance. Parents or guardians will be informed that participation is completely voluntary and that all patient data will be kept confidential. All aspects of this study will take place in the neonatal intensive care unit.

Inclusion Criteria:

• All patients who present to the neonatal intensive care unit at BIDMC who require intravenous nutritional support

Exclusion Criteria:

• Any patients receiving intravenous steroids