TY - JOUR
AU - Driscoll, David, F.
AB - Abstract Purpose. The compliance of injectable 20% lipid emulsions with the globule-size limits in chapter 729 of the U.S. Pharmacopeia (USP) was examined. Methods. As established in chapter 729, dynamic light scattering was applied to determine mean droplet diameter (MDD), with an upper limit of 500 nm. Light obscuration was used to determine the size of fat globules found in the large-diameter tail, expressed as the volume-weighted percent fat exceeding 5 μ m (PFAT5), with an upper limit of 0.05%. Compliance of seven different emulsions, six of which were stored in plastic bags, with USP limits was assessed. To avoid reaching coincidence limits during the application of method II from overly concentrated emulsion samples, a variable dilution scheme was used to optimize the globule-size measurements for each emulsion. One-way analysis of variance of globule-size distribution (GSD) data was conducted if any results of method I or II exceeded the respective upper limits. Results. Most injectable lipid emulsions complied with limits established by USP chapter 729, with the exception of those of one manufacturer, which failed limits as proposed for to meet the PFAT5 three of the emulsions tested. In contrast, all others studied (one packaged in glass and three packaged in plastic) met both criteria. Conclusion. Among seven injectable lipid emulsions tested for GSD, all met USP chapter 729 MDD requirements and three, all from the same manufacturer and packaged in plastic, did not meet PFAT5 requirements. Control, quality, Fat emulsions, Glass, Packaging, Particle size distribution, Plastics, Standards, United States Pharmacopeia The U.S. Pharmacopeia (USP) has established globule-size limits (effective December 1, 2007) for all commercial i.v. nutritional lipids in chapter 7291 and in an accompanying monograph.2 Specifically, these limits involve two key regions of the globule-size distribution (GSD): (1) the mean droplet diameter (MDD), expressed as the intensity-weighted MDD in nanometers in method I and (2) the large-diameter tail, expressed as the volume-weighted percent of fat greater than 5 μm (PFAT5) in method II of chapter 729. The upper limit in method I is an MDD of <500 must be nm; in method II, the PFAT5 less than 0.05%. In both cases, these upper limits must be met irrespective of the final concentration of the lipid injectable emulsion dosage form (i.e., 10%, 20%, or 30% w/v). On February 26, 2004, a notice was sent to directors of pharmacy announcing a change in the packaging container for Intralipid (Fresenius Kabi) from conventional glass bottles to plastic containers, with availability beginning in March or April 2004.3 In 2004, a preliminary analysis of this new dosage form was assessed in our laboratory, showing it to be coarse (containing an inordinate number of large-diameter fat globules) and therefore unable to meet the PFAT5 limit of <0.05%. Our findings were subsequently confirmed by three separate pharmaceutical laboratories (Mikrut B, Hospira Labs, personal communication, 2004 Jul 11; Nehne J, B. Braun, personal communication, 2005 Jan 20; Washington C, AstraZeneca, personal communication, 2006 Mar 22). Clearly, something changed in the integrity of the emulsion in plastic, as the version of Intralipid packaged in glass containers had previously met these limits.4 Consequently, a formal study of lipid injectable emulsions comparing four lots of Intralipid 20% in newly introduced plastic bags with four lots of Liposyn III 20% in conventional glass bottles further confirmed the GSD abnormalities of lipids in plastic containers.5 To determine whether the coarsening observed with the newly introduced Intralipid was due to the plastic container, we conducted an additional preliminary study of other lipid injectable emulsions packaged in plastic and available in Europe.6 We found the coarsening of the GSD (as measured by PFAT5) to be unique to one product, which suggested the problem was not related to plastic but rather the manufacturing process. The purpose of this study was to further analyze GSD in injectable 20% lipid emulsions and compliance with the limits specified in USP chapter 729. Methods Selection of lipid injectable emulsions and sampling Seven different injectable lipid emulsions were studied (Table 11). Two emulsions available in the United States (one marketed in glass and one in plastic) were included along with five products from Europe that are packaged in plastic. All products studied were injectable 20% lipid emulsions. Of the European products studied, one was a mixture of 80% olive oil and 20% soybean oil stored in a single-chamber plastic bag. The other four products were mixtures of medium-chain triglycerides (MCT) and long-chain triglycerides (LCT) in three-chamber plastic bags that also contained amino acids and dextrose (and electrolytes) in separate compartments. For the purposes of this study, samples of the 20% lipid injectable emulsion were removed from the dedicated compartment (to ensure that only the lipid fraction was analyzed) and then compared with the other products. The months remaining on the manufacturer-assigned expiration date at the time of the GSD analysis for each product was recorded, and at least three separate samples from each batch were tested (Table 1 1). Table 1. Characteristics of Injectable 20% Lipid Emulsions Studied Lipid Emulsion Oil(s) Container Manufacturer Lot Months to Expirationa No. Dosage Units Tested No. Replicates aAt time of globule-size analyses. bMCT = medium-chain triglycerides. U.S. Lipids     Liposyn III 100% soybean Glass Hospira 31-349-DE 1.5 3 9     Intralipid 100% soybean Plastic Fresenius Kabi 1031220 5 3 9 Non-U.S. Lipids     StructoKabiven Peri 64% soybean, 36% MCTb Plastic Fresenius Kabi 1032674 16 6 12     StructoKabiven 1100 64% soybean, 36% MCT Plastic Fresenius Kabi 1033521 18 6 12     ClinOleic 80% olive, 20% soybean Plastic Baxter 06F23A90 9 3 9     Nutriflex Peri 50% soybean, 50% MCT Plastic B. Braun 5422A159 18 6 12     Nutriflex Special 50% soybean, 50% MCT Plastic B. Braun 5404A159 17 6 12 Lipid Emulsion Oil(s) Container Manufacturer Lot Months to Expirationa No. Dosage Units Tested No. Replicates aAt time of globule-size analyses. bMCT = medium-chain triglycerides. U.S. Lipids     Liposyn III 100% soybean Glass Hospira 31-349-DE 1.5 3 9     Intralipid 100% soybean Plastic Fresenius Kabi 1031220 5 3 9 Non-U.S. Lipids     StructoKabiven Peri 64% soybean, 36% MCTb Plastic Fresenius Kabi 1032674 16 6 12     StructoKabiven 1100 64% soybean, 36% MCT Plastic Fresenius Kabi 1033521 18 6 12     ClinOleic 80% olive, 20% soybean Plastic Baxter 06F23A90 9 3 9     Nutriflex Peri 50% soybean, 50% MCT Plastic B. Braun 5422A159 18 6 12     Nutriflex Special 50% soybean, 50% MCT Plastic B. Braun 5404A159 17 6 12 Open in new tab Table 1. Characteristics of Injectable 20% Lipid Emulsions Studied Lipid Emulsion Oil(s) Container Manufacturer Lot Months to Expirationa No. Dosage Units Tested No. Replicates aAt time of globule-size analyses. bMCT = medium-chain triglycerides. U.S. Lipids     Liposyn III 100% soybean Glass Hospira 31-349-DE 1.5 3 9     Intralipid 100% soybean Plastic Fresenius Kabi 1031220 5 3 9 Non-U.S. Lipids     StructoKabiven Peri 64% soybean, 36% MCTb Plastic Fresenius Kabi 1032674 16 6 12     StructoKabiven 1100 64% soybean, 36% MCT Plastic Fresenius Kabi 1033521 18 6 12     ClinOleic 80% olive, 20% soybean Plastic Baxter 06F23A90 9 3 9     Nutriflex Peri 50% soybean, 50% MCT Plastic B. Braun 5422A159 18 6 12     Nutriflex Special 50% soybean, 50% MCT Plastic B. Braun 5404A159 17 6 12 Lipid Emulsion Oil(s) Container Manufacturer Lot Months to Expirationa No. Dosage Units Tested No. Replicates aAt time of globule-size analyses. bMCT = medium-chain triglycerides. U.S. Lipids     Liposyn III 100% soybean Glass Hospira 31-349-DE 1.5 3 9     Intralipid 100% soybean Plastic Fresenius Kabi 1031220 5 3 9 Non-U.S. Lipids     StructoKabiven Peri 64% soybean, 36% MCTb Plastic Fresenius Kabi 1032674 16 6 12     StructoKabiven 1100 64% soybean, 36% MCT Plastic Fresenius Kabi 1033521 18 6 12     ClinOleic 80% olive, 20% soybean Plastic Baxter 06F23A90 9 3 9     Nutriflex Peri 50% soybean, 50% MCT Plastic B. Braun 5422A159 18 6 12     Nutriflex Special 50% soybean, 50% MCT Plastic B. Braun 5404A159 17 6 12 Open in new tab Analytical procedures GSD was assessed according to methods I and II of USP chapter 729. For MDD assessments of method I, a Nicomp 370 submicron analyzera was used for dynamic light scattering with an automated dilution procedure that applies classical rules of light scattering and the Stokes–Einstein relation to derive the intensity-weighted MDD. For large-diameter tail measurements in method II, the volume- value was obtained weighted PFAT5 via light obscuration (or extinction) using an automatic particle sizer.b For this measurement, a standard 1-mL sample of each product was analyzed using a dilution scheme that kept the per-milliliter droplet counts well below the coincidence limit (CL) for the sensor used (i.e., ≈10,000 droplets/mL), a potential source of error with this method. To avoid reaching CLs during the application of method II from overly concentrated emulsion samples, a variable dilution scheme was used to optimize the globule-size measurements for each emulsion. The wide range of dilutions for each emulsion (Table 22) most likely demonstrated the variability in manufacturing processes among companies. The details of these measurements have been reviewed7 and recently described when applied to total nutrient admixtures (TNAs),8 which operate on the same principle for undiluted, native lipid emulsions and only differ in terms of dilution requirements (i.e., higher dilution requirements for undiluted lipids [10–30% w/v] compared with TNAs [2–5% w/v]) to avoid the CL issues described above. Table 2. Lipid Emulsion Dilution Scheme for Optimizing Measurements of Large-Diameter Taila Lipid Emulsion Dilution Range (Water:Emulsion)b Globule Count Rangec aOptimized count from 1-mL sample of emulsion yielding >500 globules/mL but <4,000 globules/mL per sample run. bDouble-filtered water (0.2 μm) used for sample dilution. cCounted per sample run. Liposyn III 570:1–1,140:1 76,577–150,943 Intralipid 17,100:1–34,200:1 214,576–399,350 StructoKabiven Peri 17,100:1–34,200:1 216,449–426,002 StructoKabiven 1100 17,100:1–34,200:1 195,165–380,679 ClinOleic 570:1–1,140:1 204,119–370,225 Nutriflex Peri 855:1–1,710:1 71,044–197,081 Nutriflex Special 855:1–1,710:1 78,455–242,281 Lipid Emulsion Dilution Range (Water:Emulsion)b Globule Count Rangec aOptimized count from 1-mL sample of emulsion yielding >500 globules/mL but <4,000 globules/mL per sample run. bDouble-filtered water (0.2 μm) used for sample dilution. cCounted per sample run. Liposyn III 570:1–1,140:1 76,577–150,943 Intralipid 17,100:1–34,200:1 214,576–399,350 StructoKabiven Peri 17,100:1–34,200:1 216,449–426,002 StructoKabiven 1100 17,100:1–34,200:1 195,165–380,679 ClinOleic 570:1–1,140:1 204,119–370,225 Nutriflex Peri 855:1–1,710:1 71,044–197,081 Nutriflex Special 855:1–1,710:1 78,455–242,281 Open in new tab Table 2. Lipid Emulsion Dilution Scheme for Optimizing Measurements of Large-Diameter Taila Lipid Emulsion Dilution Range (Water:Emulsion)b Globule Count Rangec aOptimized count from 1-mL sample of emulsion yielding >500 globules/mL but <4,000 globules/mL per sample run. bDouble-filtered water (0.2 μm) used for sample dilution. cCounted per sample run. Liposyn III 570:1–1,140:1 76,577–150,943 Intralipid 17,100:1–34,200:1 214,576–399,350 StructoKabiven Peri 17,100:1–34,200:1 216,449–426,002 StructoKabiven 1100 17,100:1–34,200:1 195,165–380,679 ClinOleic 570:1–1,140:1 204,119–370,225 Nutriflex Peri 855:1–1,710:1 71,044–197,081 Nutriflex Special 855:1–1,710:1 78,455–242,281 Lipid Emulsion Dilution Range (Water:Emulsion)b Globule Count Rangec aOptimized count from 1-mL sample of emulsion yielding >500 globules/mL but <4,000 globules/mL per sample run. bDouble-filtered water (0.2 μm) used for sample dilution. cCounted per sample run. Liposyn III 570:1–1,140:1 76,577–150,943 Intralipid 17,100:1–34,200:1 214,576–399,350 StructoKabiven Peri 17,100:1–34,200:1 216,449–426,002 StructoKabiven 1100 17,100:1–34,200:1 195,165–380,679 ClinOleic 570:1–1,140:1 204,119–370,225 Nutriflex Peri 855:1–1,710:1 71,044–197,081 Nutriflex Special 855:1–1,710:1 78,455–242,281 Open in new tab One-way analysis of variance (ANOVA) of GSD data was conducted if any results of method I or II exceeded the respective upper limits, using the lipid emulsion as the independent variable. A comparison of out-of-specification GSD measurements with those in compliance was the dependent variable among products. The analysis was conducted using Systat (Systat Software, Point Richmond, CA). Significant differences detected by ANOVA (p < 0.05) were further evaluated by Bonferroni pairwise comparisons. All GSD data were expressed as the mean ± S.D. Results With respect to method I of chapter 729, all injectable lipid emulsions studied met the requirement of an MDD of <500 nm (Table 33). All lipids studied complied with the of method II requirement of a PFAT5 <0.05%, except those from Fresenius Kabi (Intralipid, StructoKabiven Peri, and StructoKabiven 1100). In fact, these failures occurred for both soybean oil-only and mixed LCT–MCT emulsions (Table 3 3). ANOVA revealed highly significant differences in the large-diameter tail for all globule sizes, whether the variable was expressed as the number of globules of a specific size (globule number [GN]) per milliliter or as PFAT (p < 0.001). Bonferroni pairwise comparisons confirmed that the significant differences were confined to the three emulsions manufactured by Fresenius Kabi. The coarseness of these emulsions is clearly evident from the data presented in Table 33. Table 3. Normalized Results of Globule-Size Analysis Applying Methods I and II of USP Chapter 729a Mean ± S.D. No. Globules/mL Lipid Emulsion Globule Diameter >1.8 μm Globule Diameter >5 μm Globule Diameter >10 μm PFAT1 PFAT5b PFAT1 MDDc aAfter accounting for sample dilution, the counts per milliliter per size range were normalized to the undiluted sample. PFAT1.8, PFAT5, and PFAT10 = volume-weighted percentages of fat globules with a diameter greater than 1.8, 5, and 10 μm, respectively. MDD = mean droplet diameter. bUpper limit of method II: PFAT5 < 0.05%. cUpper limit of method I: MDD < 500 nm. Liposyn III 861,570 ± 40,742 55,985 ± 6,138 1,810 ± 316 0.011 ± 0.001 0.005 ± 0.001 0.001 ± 0.000 290 ± 1.2 Intralipid 63,959,781 ± 3,688,525 4,059,240 ± 255,549 21,788 ± 6,601 0.696 ± 0.041 0.226 ± 0.015 0.012 ± 0.005 326 ± 3.5 StructoKabiven Peri 76,597,163 ± 16,462,250 3,649,600 ± 750,197 51,876 ± 11,598 0.728 ± 0.151 0.226 ± 0.046 0.024 ± 0.005 313 ± 1.4 StructoKabiven 1100 61,759,182 ± 2,869,770 2,869,770 ± 208,275 48,506 ± 5,057 0.583 ± 0.046 0.180 ± 0.014 0.022 ± 0.003 318 ± 3.7 ClinOleic 2,344,481 ± 84,184 92,915 ± 14,463 1,536 ± 322 0.022 ± 0.002 0.006 ± 0.001 0.001 ± 0.000 284 ± 1.1 Nutriflex Peri 1,401,723 ± 163,205 133,998 ± 25,345 2,888 ± 1,092 0.019 ± 0.003 0.009 ± 0.002 0.002 ± 0.001 310 ± 8.9 Nutriflex Special 1,593,037 ± 251,554 112,927 ± 16,083 2,272 ± 816 0.018 ± 0.002 0.008 ± 0.001 0.001 ± 0.000 308 ± 2.4 Mean ± S.D. No. Globules/mL Lipid Emulsion Globule Diameter >1.8 μm Globule Diameter >5 μm Globule Diameter >10 μm PFAT1 PFAT5b PFAT1 MDDc aAfter accounting for sample dilution, the counts per milliliter per size range were normalized to the undiluted sample. PFAT1.8, PFAT5, and PFAT10 = volume-weighted percentages of fat globules with a diameter greater than 1.8, 5, and 10 μm, respectively. MDD = mean droplet diameter. bUpper limit of method II: PFAT5 < 0.05%. cUpper limit of method I: MDD < 500 nm. Liposyn III 861,570 ± 40,742 55,985 ± 6,138 1,810 ± 316 0.011 ± 0.001 0.005 ± 0.001 0.001 ± 0.000 290 ± 1.2 Intralipid 63,959,781 ± 3,688,525 4,059,240 ± 255,549 21,788 ± 6,601 0.696 ± 0.041 0.226 ± 0.015 0.012 ± 0.005 326 ± 3.5 StructoKabiven Peri 76,597,163 ± 16,462,250 3,649,600 ± 750,197 51,876 ± 11,598 0.728 ± 0.151 0.226 ± 0.046 0.024 ± 0.005 313 ± 1.4 StructoKabiven 1100 61,759,182 ± 2,869,770 2,869,770 ± 208,275 48,506 ± 5,057 0.583 ± 0.046 0.180 ± 0.014 0.022 ± 0.003 318 ± 3.7 ClinOleic 2,344,481 ± 84,184 92,915 ± 14,463 1,536 ± 322 0.022 ± 0.002 0.006 ± 0.001 0.001 ± 0.000 284 ± 1.1 Nutriflex Peri 1,401,723 ± 163,205 133,998 ± 25,345 2,888 ± 1,092 0.019 ± 0.003 0.009 ± 0.002 0.002 ± 0.001 310 ± 8.9 Nutriflex Special 1,593,037 ± 251,554 112,927 ± 16,083 2,272 ± 816 0.018 ± 0.002 0.008 ± 0.001 0.001 ± 0.000 308 ± 2.4 Open in new tab Table 3. Normalized Results of Globule-Size Analysis Applying Methods I and II of USP Chapter 729a Mean ± S.D. No. Globules/mL Lipid Emulsion Globule Diameter >1.8 μm Globule Diameter >5 μm Globule Diameter >10 μm PFAT1 PFAT5b PFAT1 MDDc aAfter accounting for sample dilution, the counts per milliliter per size range were normalized to the undiluted sample. PFAT1.8, PFAT5, and PFAT10 = volume-weighted percentages of fat globules with a diameter greater than 1.8, 5, and 10 μm, respectively. MDD = mean droplet diameter. bUpper limit of method II: PFAT5 < 0.05%. cUpper limit of method I: MDD < 500 nm. Liposyn III 861,570 ± 40,742 55,985 ± 6,138 1,810 ± 316 0.011 ± 0.001 0.005 ± 0.001 0.001 ± 0.000 290 ± 1.2 Intralipid 63,959,781 ± 3,688,525 4,059,240 ± 255,549 21,788 ± 6,601 0.696 ± 0.041 0.226 ± 0.015 0.012 ± 0.005 326 ± 3.5 StructoKabiven Peri 76,597,163 ± 16,462,250 3,649,600 ± 750,197 51,876 ± 11,598 0.728 ± 0.151 0.226 ± 0.046 0.024 ± 0.005 313 ± 1.4 StructoKabiven 1100 61,759,182 ± 2,869,770 2,869,770 ± 208,275 48,506 ± 5,057 0.583 ± 0.046 0.180 ± 0.014 0.022 ± 0.003 318 ± 3.7 ClinOleic 2,344,481 ± 84,184 92,915 ± 14,463 1,536 ± 322 0.022 ± 0.002 0.006 ± 0.001 0.001 ± 0.000 284 ± 1.1 Nutriflex Peri 1,401,723 ± 163,205 133,998 ± 25,345 2,888 ± 1,092 0.019 ± 0.003 0.009 ± 0.002 0.002 ± 0.001 310 ± 8.9 Nutriflex Special 1,593,037 ± 251,554 112,927 ± 16,083 2,272 ± 816 0.018 ± 0.002 0.008 ± 0.001 0.001 ± 0.000 308 ± 2.4 Mean ± S.D. No. Globules/mL Lipid Emulsion Globule Diameter >1.8 μm Globule Diameter >5 μm Globule Diameter >10 μm PFAT1 PFAT5b PFAT1 MDDc aAfter accounting for sample dilution, the counts per milliliter per size range were normalized to the undiluted sample. PFAT1.8, PFAT5, and PFAT10 = volume-weighted percentages of fat globules with a diameter greater than 1.8, 5, and 10 μm, respectively. MDD = mean droplet diameter. bUpper limit of method II: PFAT5 < 0.05%. cUpper limit of method I: MDD < 500 nm. Liposyn III 861,570 ± 40,742 55,985 ± 6,138 1,810 ± 316 0.011 ± 0.001 0.005 ± 0.001 0.001 ± 0.000 290 ± 1.2 Intralipid 63,959,781 ± 3,688,525 4,059,240 ± 255,549 21,788 ± 6,601 0.696 ± 0.041 0.226 ± 0.015 0.012 ± 0.005 326 ± 3.5 StructoKabiven Peri 76,597,163 ± 16,462,250 3,649,600 ± 750,197 51,876 ± 11,598 0.728 ± 0.151 0.226 ± 0.046 0.024 ± 0.005 313 ± 1.4 StructoKabiven 1100 61,759,182 ± 2,869,770 2,869,770 ± 208,275 48,506 ± 5,057 0.583 ± 0.046 0.180 ± 0.014 0.022 ± 0.003 318 ± 3.7 ClinOleic 2,344,481 ± 84,184 92,915 ± 14,463 1,536 ± 322 0.022 ± 0.002 0.006 ± 0.001 0.001 ± 0.000 284 ± 1.1 Nutriflex Peri 1,401,723 ± 163,205 133,998 ± 25,345 2,888 ± 1,092 0.019 ± 0.003 0.009 ± 0.002 0.002 ± 0.001 310 ± 8.9 Nutriflex Special 1,593,037 ± 251,554 112,927 ± 16,083 2,272 ± 816 0.018 ± 0.002 0.008 ± 0.001 0.001 ± 0.000 308 ± 2.4 Open in new tab The coarsened GSD profile observed among emulsions is most discernible when assessing large-diameter tail results expressed as PFAT5; the Fresenius Kabi products had a PFAT5 value well above the threshold of 0.05%. A modest degree of emulsion coarsening was also evident, even among formulations that meet method II size limits (e.g., at the 5-μm threshold, Liposyn III has approximately 56,000 globules/mL, whereas the ClinOleic and Nutriflex formulations had nearly 100,000 or more globules per milliliter), demonstrating that qualitative differences can be seen even among formulations that fully comply with the GSD limits of chapter 729. The GN per milliliter values for globules larger than 5 μm for lipids packaged in plastic and meeting the PFAT5 limit of chapter 729 (i.e., ClinOleic, Nutriflex Peri, and Nutriflex Special) are approximately an order of magnitude lower than those for the emulsions whose PFAT5 values exceeded the threshold. Discussion The production of a fine dispersion is an essential pharmaceutical characteristic that should be met by every commercial batch of injectable lipid emulsion. Until recently, there were no official pharmacopeial standards for these complex i.v. formulations, despite the wide clinical use of these emulsions over the past 40 years. Prior attempts to set large-diameter globule-size limits were attempted but subsequently dropped, as the limits could not be reasonably confirmed.9 With the application of light obscuration as a counting method, reasonable quantifiable limits were proposed in 2001 based on a consensus developed through the author’s collaboration with three major pharmaceutical manufacturers from an analysis of 16 different commercial lipid injectable emulsions: For commercial IVLE [intravenous lipid emulsions] from the manufacturer, we would suggest a mean droplet size (MDS) that does not exceed 450 nm, and an upper limit for PFAT > 5 μm that does not exceed 0.05%. Based on our experience and the data shown . . . these limits are easily met. Therefore, such a proposed range for both MDS and PFAT (> 5 μm) is pharmaceutically reasonable, in that the ranges are not only sufficiently broad, but are also likely safe, given the current use conditions and available data.4 Subsequently, USP first proposed such limits in 2004 that seemed achievable for injectable lipid emulsions available in the United States given the above recommendations. Then, a change in the packaging of one product from conventional glass bottles to newly introduced plastic bags occurred,3 and the product failed to meet the proposed PFAT5 limits of method II of USP chapter 729.5 This product was a unique and significant departure from its glass counterpart, as well as all others packaged in glass bottles. The essential question that arose from these findings was whether the coarsening of the emulsion now seen in plastic was a result of the container or the manufacturing process or both. A preliminary follow-up assessment of other lipids in plastic suggested that the problem was not the plastic packaging, as lipids packaged in plastic containers by other manufacturers were able to meet the proposed PFAT5 limits.6 The present data confirmed that the abnormality in the GSD of injectable lipid emulsions packaged in plastic was indeed confined to Fresenius Kabi and occurred in both its soybean and mixed-oil emulsions. In contrast, two other plastic-packaged injectable lipid emulsions of other manufacturers easily met the GSD limits for large-diameter fat globules, irrespective of the oil composition. It is also noteworthy that the MDD data from method I of USP chapter 729 cannot distinguish between fine and coarse emulsions, pointing out the absolute need for method II to discern such differences. This finding is consistent with earlier findings10 and may also have implications for the safety of i.v. infusions. For example, we recently showed that a coarse emulsion in plastic that failed to meet the PFAT5 limits resulted in less-stable TNAs8; the problem also exists when the emulsion is repackaged in syringes for neonatal administration.11 Because of the demonstrated characteristics of plastic-packaged Intralipid, the historical stability profile for Intralipid in glass containers may need to be reestablished with new limits to ensure its future safety in the clinical setting. Moreover, the plastic-packaged Intralipid has recently been associated with significant hypertriglyceridemia in critically ill premature infants when compared with its glass-packaged counterpart that met PFAT5 limits.12 This latter finding is most concerning clinically, especially in light of the black-box warning about the reduced abilities of critically ill neonates to clear lipids from the circulation. The clinical concern is heightened by the 1995 study that showed that hyper-triglyceridemia in this population is associated with liver dysfunction and growth retardation.13 USP chapter 729 is intended to set manufacturing standards for lipid injectable emulsions and is long overdue. To date, studies suggest that all but one manufacturer is capable of meeting the proposed globule-size limits. This finding suggests that the problem is less likely related to the plastic container but rather to a defect in the manufacturing of these complex pharmaceutical dosage forms, perhaps related to a change in the process coincident with the change in container. USP chapter 729 will establish pharmaceutical equivalence and possibly improve the stability and safety profile of injectable lipid emulsions. Conclusion Among seven injectable lipid emulsions tested for GSD, all met USP chapter 729 MDD requirements and three, all from the same manufacturer and packaged in plastic, did not meet PFAT5 requirements. a Nicomp 370 submicron particle sizer, Particle Sizing Systems, Santa Barbara, CA. b Accusizer 780/APS, automatic particle sizer, Particle Sizing Systems. References 1 Chapter <729>: globule size distribution in lipid injectable emulsions. In: The United States pharmacopeia, 30th rev., and The national formulary, 25th ed., suppl. 2. Rockville, MD: United States Pharmacopeial Convention; 2007 :3968–70. 2 Lipid injectable emulsion monograph. Pharm Forum . 2006 ; 32 : 350 –3. 3 Baxter Healthcare Corporation. Announcement to directors of pharmacy: Intralipid i.v. fat emulsion now available in plastic containers. Deerfield, IL: 2004 Feb 26. 4 Driscoll DF, Etzler F, Barber TA et al. Physicochemical assessments of parenteral lipid emulsions: light obscuration versus laser diffraction. Int J Pharm . 2001 ; 219 : 21 –37. Crossref Search ADS PubMed 5 Driscoll DF, Bistrian BR. The effects of packaging containers on the large-diameter tail of the globule size distribution (GSD) of lipid emulsions. Clin Nutr. 2005 ; 24 : 699 . Abstract. 6 Driscoll DF, Thoma A, Klütsch K et al. The effects of packaging containers and manufacturer on the large-diameter tail of the globule size distribution of lipid injectable emulsions—part 2. Clin Nutr. 2005 ; 24 : 1126 –7. Abstract. 7 Driscoll DF. Examination of selection of light-scattering and light-obscuration acceptance criteria for lipid injectable emulsions. Pharm Forum . 2004 ; 30 : 2244 –53. 8 Driscoll DF, Silvestri AP, Bistrian BR et al. Stability of total nutrient admixtures with lipid injectable emulsions in glass vs. plastic packaging. Am J Health-Syst . 2007 ; 64 : 396 –403. 9 Driscoll DF. Lipid injectable emulsions: pharmacopeial and safety issues. Pharm Res . 2006 ; 23 : 1959 –69. Crossref Search ADS PubMed 10 Driscoll DF, Bhargava HN, Li L et al. The physicochemical stability of complex intravenous lipid dispersions as total nutrient admixtures. Am J Health-Syst Pharm . 1995 ; 52 : 623 –34. Crossref Search ADS PubMed 11 Driscoll DF, Ling PR, Bistrian BR. Physical stability of 20% lipid injectable emulsions via simulated syringe infusion: effects of glass vs. plastic product packaging. J Parenter Enteral Nutr. 2007 ; 31 : 148 –53. [Erratum, J Parenter Enteral Nutr. 2007; 31:261.] Crossref Search ADS 12 Martin CR, Dumas GJ, Shoaie C et al. Incidence of hypertriglyceridemia in critically ill neonates receiving lipid injectable emulsions in glass vs. plastic containers: a retrospective analysis. J Pediatr. In press. 13 Toce SS, Keenan WJ. Lipid intolerance in newborns is associated with hepatic dysfunction but not infection. Arch Ped Adolesc Med . 1995 ; 149 : 1249 –53. Crossref Search ADS Author notes Presented in part at the European Society of Parenteral and Enteral Nutrition Meeting, Istanbul, Turkey, October 21, 2006. Dr. Driscoll is a researcher, consultant, or both for AstraZeneca, B. Braun, Biolink, GlaxoSmithKline, Hospira, and The Medicine Company. Copyright © 2007, American Society of Health-System Pharmacists, Inc. All rights reserved.
TI - Globule-size distribution in injectable 20% lipid emulsions: Compliance with USP requirements
JF - American Journal of Health-System Pharmacy
DO - 10.2146/ajhp070097
DA - 2007-10-01
UR - https://www.deepdyve.com/lp/oxford-university-press/globule-size-distribution-in-injectable-20-lipid-emulsions-compliance-IemNex3MFL
SP - 2032
VL - 64
IS - 19
DP - DeepDyve
ER -