Anthony P. Silvestri

Physical compatibility of neonatal total parenteral nutrition admixtures containing organic calcium and inorganic phosphate salts in a simulated infusion at 37°C*

Objective: Within an incubator environment at 37°C, we evaluated the physical compatibility of calcium (Ca) and phosphate (P) in simulated neonatal total parenteral nutrition (TPN) infusions. Design: Representative neonatal TPN formulas containing four concentrations of amino acids [AA] (1–4 g/dL) and two levels of dextrose [D] (5 and 10 g/dL) were monitored for calcium-phosphate precipitate formation over 24 hours using a laser-based particle counting technique. Setting: Research Laboratory. Subjects: None. Measurements and Main Results: Of the eight AA-D combinations studied, two formulas containing 1 g/dL AA with either 5 g/dL or 10 g/dL of D and high amounts of Ca and P resulted in significant increases in particle counts and visibly evident precipitation during the 24-hour infusion. Precipitation was noted only after the infusion passed through the incubator, but not in the TPN bags, demonstrating the heightened risk of calcium-phosphate incompatibility with increased temperatures. Conclusions: Low AA concentrations and high temperatures are significant risk factors for the precipitation of the insoluble dibasic calcium phosphate that may be fatal upon intravenous infusion. When thermoregulation is required, the AA concentration of the TPN formula should not be less than 1.5 g/dL (and ideally ≥3 g/dL) when high intakes of calcium (60 mg/dL) and phosphate (46.5 mg/dL) are prescribed.

Calcium and Phosphate Compatibility in Low-Osmolarity Parenteral Nutrition Admixtures Intended for Peripheral Vein Administration

BACKGROUND Precipitation of calcium (Ca) and phosphate (P) salts can lead to potentially lethal outcomes, especially in low-osmolarity parenteral nutrition (LO-PN) formulations. Three concentrations of amino acids (AA) and 2 concentrations of calcium gluconate and sodium phosphate injections on the compatibility of Ca and P in LO-PN admixtures were studied. METHODS Final AA concentrations of 1%, 2%, or 3% (n = 3) and 5% glucose (G) were prepared with either 2.5 or 5 mmol/L (5 or 10 mEq) of Ca (n = 2) and 15 or 30 mmol/L of P (n = 2) for a total of 12 base (3 x 2 x 2) formulations. Triplicate bags of each were analyzed for subvisible micro-precipitates using the light obscuration (or extinction) method for particle counts per milliliter (PC) in the size range of 1.8-50 mum at 7 time intervals over 48 hours stored at 30 degrees C +/- 0.2 degrees C. Visual evaluation was performed using a high-intensity lamp against a black background for detection of macro-precipitates. The pH of all 36 admixtures was measured at 0 and 48 hours. Any precipitated material was isolated and characterized by polarized light microscopy and infrared spectroscopy. RESULTS Of the 12 base LO-PN formulations tested, those containing 1% and 2% AA with 5 mmol/L of Ca and 30 mmol/L of P showed significant increases in PC, and some resulted in visible dibasic calcium phosphate precipitation. Analyses of variance based on concentrations of AA, Ca, P, and time were highly significant independent variables for increases in the PC of potentially embolic particles, that is, sizes >5 mum (P < .0001). The lowest concentrations of Ca and P, 2.5 and 15 mmol/L, respectively, had significantly lower PC (P < .05) for all sizes compared with the other Ca and P combinations. CONCLUSIONS LO-PN admixtures (AA </= 3%; G = 5%) should not contain more than 2.5 mmol/L of Ca from calcium gluconate injection and no more than 15 mmol/L of P from potassium or sodium phosphates injection.

Fine vs. coarse complete all-in-one admixture infusions over 96 hours in rats: Fat globule size and hepatic function

BACKGROUND & AIMS The United States Pharmacopeia (USP) has adopted Chapter <729> that set two globule size limits for all lipid emulsions with the mean droplet size at no >500 nm, while large-diameter fat globules as the percent fat>5 microm or PFAT(5) must be <0.05%. A quantitative risk assessment of toxicity from the intravenous infusion of all-in-one (AIO) admixtures made from a lipid emulsion that meets USP standards (fine) vs. one that does not (coarse), was conducted. METHODS Two separate 96-h infusion studies in rats receiving nutritionally complete AIOs made from a fine (F) vs. a coarse (C) 20% starting lipid emulsion (SLE) with either 18 or 36% as fat calories were performed. The animals were equally divided in each (18% fat, n=18; 36% fat, n=22) to receive AIOs made from F or C lipids. PFAT(5) levels were measured at the outset and every 24h at the change of infusions and blood levels of liver enzymes AST and GST, and serum triglycerides (TG) were measured at the end of study. RESULTS On average, the starting PFAT(5) values for infusions of F-AIOs were 0.018+/-0.007 (n=48) vs. C-AIOs at 0.183+/-0.026% (n=48), whereas the 24-h average was 0.234+/-0.211% (n=168) vs. 1.033+/-0.224% (n=180), respectively. No significant differences in the blood-based parameters were noted in rats between F-AIOs and C-AIOs in the studies comparing 18 or 36% of fat calories, respectively. When the data were combined into all F- vs. all C-AIOs, AST was significantly higher in C-AIOs (157+/-41) vs. F-AIOs (130+/-37), p=0.036. TG was lower in C (69+/-37) vs. F (106+/-70), nearly reaching statistical significance (p=0.056) with no differences in GST levels for C (21+/-9) vs. F (17+/-9), p=0.199. When stratified according to a PFAT(5) of 0.4%, C-AIOs were significantly higher than F-AIOs for AST (157+/-41 vs. 130+/-37, p=0.004), and TG was significantly lower in C- vs. F-AIOs (67+/-36 vs. 117+/-71, p=0.022), respectively. CONCLUSIONS Coarse lipid emulsions that fail pharmacopeial limits produce less stable AIOs and are associated with evidence of worsened hepatic injury.

Stability of MCT/LCT-based total nutrient admixtures for neonatal use over 30 hours at room temperature: applying pharmacopeial standards.

BACKGROUND United States Pharmacopeial Chapter <729> places a limit on the percentage of large fat globules >5 microm, expressed as a PFAT5 of <0.05% for all native lipid emulsions. Some adult total nutrient admixtures (TNAs) have also remained below this limit for up to 48 hours. In 2003, medium-chain/long-chain triglyceride (MCT/LCT)-based neonatal TNAs with between 2% and 3% amino acid (AA) concentrations were shown to be similarly stable by the PFAT5 parameter. Stability assessment of neonatal TNAs with AA <2% or > or =3% were tested. METHODS Eight neonatal TNAs with various combinations of AA (1%, 1.5%, 3%, and 4%), glucose (G; 5% and 10%), and MCT/LCT (ML; 2% and 4%) and standard concentrations of additives were tested in triplicate (n = 24) over 30 hours (immediately after mixing, then at 6, 24, and 30 hours) at 25 degrees C +/- 2 degrees C. PFAT5 determinations for all 24 formulations were made in duplicate, immediately after mixing, and then at 6, 24, and 30 hours later. Mean droplet size (MDS) and pH were assessed at the outset and end of the study. RESULTS The differences in the PFAT5 levels were significant (P < .001) by a 2-way analysis of variance based on formula and time as the independent variables. The TNAs with 1% and 1.5% AA with all Gs and MLs (group 1, n = 12) had PFAT5 levels >0.05% (up to 0.50%) in most samples (68 of 96 samples, or 71% of cases) in the study, whereas in the same TNAs, but made with 3% and 4% AA (group 2, n = 12), 100% of samples (all 96 cases) had PFAT5 levels <0.05% (up to 0.04%), and this difference was significant (P < .001). Pairwise comparisons between groups based on overall values of PFAT5, MDS, and pH showed significant differences between groups for all variables. CONCLUSIONS For neonatal TNAs, AA level is the most sensitive determinant of stability, and the PFAT5 parameter is the most sensitive indicator of stability.

Effects of glucose or fat calories in total parenteral nutrition on fat metabolism and systemic inflammation in rats.

This study compared the effects of total parenteral nutrition (TPN) by central vein with or without fat provided at maintenance energy requirement on fatty acid metabolism, de novo lipogenesis, and the risk of hepatic and systemic inflammation in rats. Study 1 was conducted in 2 groups: high glucose (HG), where fat-free TPN was given at maintenance levels of 180 kcal/(kg d), and low glucose (LG), where fat-free TPN containing 30% fewer calories at 126 kcal/(kg d) was provided by reducing 54 kcal/(kg d) from parenteral glucose. Study 2 contained 3 TPN groups: 1 LG group at 126 kcal/(kg d) and 2 groups at 180 kcal/(kg d) with 30% of total calories (54 kcal/[kg d]) either from soybean or fish oil emulsion. In both studies, animals fed a chow diet ad libitum were included. Plasma and hepatic triglyceride and phospholipid fatty acid profiles, enzymes indicating hepatic injury, and C-reactive protein levels (CRP) reflecting systemic injury were measured. In study 1, evidence of de novo lipogenesis was noted in LG and was more prominent in HG with elevation of CRP in HG. In study 2, de novo lipogenesis was reduced by adding either fat to LG to achieve maintenance energy levels. Moreover, adding fat as soybean oil but not fish oil significantly increased plasma and hepatic triglyceride and also elevated aspartate aminotransferase and CRP levels, reflecting inflammation. Thus, in rats, either hypocaloric feeding as glucose-based TPN or TPN provided at maintenance energy levels with the addition of fish oil limits hepatic lipid accumulation and prevents the evidence of hepatic and systemic injury found with maintenance level TPN as glucose only or glucose plus soybean oil.