Oxidation of parenteral lipid emulsion by ambient and phototherapy lights: Potential toxicity of routine parenteral feeding

Oxidation of parenteral lipid emulsion by ambient and phototherapy lights: Potential toxicity of routine parenteral feeding Jii'i Neu2il, PhD, Brian A. Darlow, MD, Terrie E. Inder, MBChB, Karl B. Sluis, BSc, Christine C. W i n t e r b o u r n , PhD, a n d Roland Stocker, PhD From the Biochemistry Group, The Heart Research Institute, Camperdown, Sydney, New South Wales, Austral[a, and the Departments of Paediatrics and Pathology, Christchurch School of Medic[ne, Christchurch Hospital, Christchurch, New Zealand Vitamin E can be a p r o o x i d a n t in isolated lipoprotein suspensions. Because lipid emulsions used in parenteral nutrition are lipoprotein-like suspensions rich in polyunsaturated fatty acids and vitamin E, we hypothesized that vitamin E may act as a prooxidant in lipid emulsions, as it is in lipoprotein suspensions. We therefore e x p o s e d an intravenously administered lipid emulsion (Intralipid) to a single spotlight c o m m o n l y used in the treatment of neonatal jaundice, and measured the formation of triglyceride hydroperoxides by using high-perform a n c e liquid c h r o m a t o g r a p h y with postcolumn c h e m i l u m i n e s c e n c e detection. Concentrations of these hydroperoxides in different batches of fresh Intralipid were usually a p p r o x i m a t e l y 10 ~mol/L but increased up to 60 times after exposure to p h o t o t h e r a p y light for a period of 24 hours, even though significant amounts of vitamin E were present at the end of the exposure. Triglyceride hyd r o p e r o x i d e s were formed during p h o t o t h e r a p y light exposure whether the Intralipid was in plastic tubing used routinely tor infusion or in glass containers. Ambient light also caused significant peroxidation of the formula lipids, although to a much lesser extent than observed with p h o t o t h e r a p y light. For infants in the neonatal intensive care unit who were receiving Intralipid but not phototherapy, solutions being infused at the end of 24 hours c o n t a i n e d a mean of 40 ~mol/L hydroperoxides. For infants receiving phototherapy, the mean was 97 /~mol/L. Phototherapy l i g h t - i n d u c e d formation of triglyceride hydroperoxides was prevented by covering the Intralipid with aluminum foil or supplementation with sodium ascorbate before light exposure. We c o n c l u d e that Intralipid is highly susceptible to oxidation and that e l e v a t e d levels of oxidized lipids can be formed during its clinical use, especiaily when Intralipid infusion is c o m b i n e d with phototherapy. Because lipid hydroperoxides are cytotoxic and can cause adverse effects, inadvertent infusion of rancid Intralipid may a d d to the numerous problems e n c o u n t e r e d by premature neonates. (J PEDIATR1995;126:785-90)

Supported by the National Health and Medical Research Council of Australia (grants No. 910284 and No. 940915 to R. S.) and the Health Research Council of New Zealand. Presented at the Annual Meeting of the American Pediatric Society and the Society for Pediatric Research, San Diego, Calif., May 1995.

Submitted for pnblication Aug. 23, 1994; accepted Dec. 9, 1994. Reprint requests: Roland Stocker, PhD, Biochemistry Group, The Heart Research Institute, 145 Missenden Road, Camperdown, Sydney, 2050 New South Wales, Australia. Copyright © 1995 by Mosby-Year Book, Inc. 0022-3476/95/$3.00 + 0 9/23/62652

785

786

Neu~il et al.

The Journal of Pediatrics May 1995

See commentary, p. 747.

NICU

Neonatal intensive care unit

] I

Parenteral lipid emulsions derived from either soybean or safflower oil provide essential fatty acids and a r e a concentrated source of calories for preterm infants. However, lipid infusions in preterm neonates are associated with problems (reviewed by Stahl et a1.1), including inefficient plasma lipid clearance, displacement of bilirubin from albumin, and impaired pulmonary function. Earlier reports suggested that lipid emulsions used in parenteral feeding of infants are susceptible to peroxidation, as judged indirectly by the formation of pentane and malondialdehyde. 2 More recently Helbock et al. 3 demonstrated the presence of high concentrations of oxidized lipids in stored lipid emulsions before their administration. Lipid hydroperoxides in either free form or within lipid emulsions are cytotoxic and cause marked damage to endothelial cells of the thoracic aorta when administered intravenously. 4 Lipid hydroperoxides also modulate lipoxygenases and cyclooxygenase6; the latter enzyme is responsib!e for the biosynthesis of important vasoregulatory prostaglandins. Moreover, perfusion of umbilical veins with lipid emulsions causes vasoconstriction, which has been linked to an inhibition in the production of the vasodilator prostaglandin I2 and an increase in the vasoconstrictor prostaglandin F2,»7 These findings, together with the observation that the concentration of lipid hydroperoxides in plasma of healthy human beings is very low (i.e., measurable in nanomolar amounts), s suggest that intravenous infusion of oxidized Intralipid presents a potentially serious problem. Phototherapy could potentially enhance lipid oxidation. We have exposed an intravenously administered lipid emulsion (Intralipid; Kabi Pharmacia AB, Stockholm, Sweden) to ambient room light and single spotlight under simulated and real phototherapy conditions in a neonatal intensive care unit to determine whether this produces high levels of lipid hydroperoxides.

METHODS Exposure of Intralipid to various conditions was performed in the Christchureh Women's Hospital NICU, where the ambient temperature is 25 ° C. A single spotlight unit (Healthdyne Phototherapy Light, model PT 1400-2; Air-Shields, Hatboro, Pa.) was used for phototherapy. The lamps emit light of 400 to 520 nm wavelength. The light is routinely directed at the neonate's chest and abdomen, with the drip site and ext•nsion tubing eontaining ftuids fre-

quently exposed, whereas the syringe and infusion pump are located on one side. Solutions of 20% Intralipid are imported by the local distributor (Baxter Healthcare Ltd., Auckland, New Zealand) in 250 ml clear glass bottles with a 2-year expiration date. The distributor prepares 50 ml aliquots in plastie syringes that have a 30-day expiration date if stored at 4 ° C, and these are dispensed by the hospital pharmacy. Solutions of Intralipid were exposed either to ambient room light or a single phototherapy spotlight at a distance of 32 to 46 cm. Light fiuxes were quantified at baseline and at 24 hours with a mode1451 Minolta Air-Shields Fluoro-Lite Meter, which measures irradiances in # W / c m 2 per nanometer within a wavelength range of 400 to 520 nm. 3 The irradiances used were as recommended for effective phototberapy. 9 To simulate the clinical situation, the syringes were placed on the bench in the N I C U in ambient lighting and connected to standard minibore plastic extension tubing of 1.8 ml capacity (catalog number BC 567, Codan Medlon, Inc., Burbank, Calif.). In the experimental system, the tubing but not the syringe was exposed to phototherapy light. In control samples, syringes and tubing remained in ambient lighting throughout. Samples of lipid (1.5 tal) were taken from the syringes and distal end of the tubing at baseline and after 24 hours. To investigate the time course of oxidation of Intralipid in different conditions, syringes containing 20% Intralipid were exposed to phototherapy light and samples were taken at intervals up to 24 hours. The syringes were glass, plastic, plastic covered in aluminum foil, or glass to which 2 cm of plastic infusion tubing was added. In further experiments, sodium ascorbate was added to Intralipid in plastic syringes at a final concentration of 1 mmol/L and at pH 7.4. Control conditions were provided by Intralipid with ascorbate added and exposed to ambient light, and Intralipid with an equivalent volume of distilled water added in place of ascorbate and exposed to either phototherapy or ambient light for 24 hours. Samples of direct clinical relevance were obtained from six neonates receiving Intralipid infusion. Intralipid was removed at baseline and at 24 hours from both the syringe and distal end of the infusion tubing from three infants exposed to phototherapy light and three neonates in ambient lighting. The Intralipid samples were placed immediätely on dry ice and stored at - 8 0 ° C until they were transported to Sydney on dry ice for analyses; they remained frozen until analyzed. For each shipment, baseline samples that have not been exposed to ambient or phototherapy light served as control samples. Results were included only if these control samples contained < 15.9 ~tmol/L triglyceride hydroperoxides (i.e., two SD values greater than the mean value [see below]).

The Journal o f Pediatrics Volume 126, Number 5, Part 1

Neu~il et al.

787

Table. Levels of triglyceride hydroperoxides in Intralipid taken from the reservoir and distal end of the infusion tubing before and after a 24-hour exposure to ambient room or phototherapy light Triglyceride hydroperoxides (~mol/L) Intralipid reservoir Condition

Simulated Ambient light Phototherapy light Clinical Ambient light Phototherapy light

Infusion tubing

Baseline

After 24 hr

Baseline

After 24 hr

9.9 ___1.8 9.2 _+ 0.5

27.4 _+ 7.3* 32.4 _+ 6.2*

7.5 _+ 1.2 9.6 + 2.9

59.2 ___9.7* 633.9 + 82.5*

5.4 6.8

26.9 _+ 18.0 77,8 _+ 17.0"

5.5 7.0

39.9 _+ 9.0* 92.7 _+ 4.7*

The resultsof experimentsthat simulatedthe clinicalsituation(Simulated)and Intralipidsamplesactuallyusedfor parenteralnutritionof neonates(Clinical) are shown.The resultsrepresentmeanvalues +_SD ofthree separateexperiments,exceptfor the baselineelinicalvalues,whichrepresentthe meanof two separate experiments.All baselinevalueswere within 2 SD of the meanvalueof fresh Intralipidsamples. *Values different(« < 0.0005) from those in fresh Intralipid;WilcoxonRank Sum Test (two-sample).

Intralipid samples were extracted and analyzed for unoxidized lipids and triglyceride hydroperoxides by highperformance liquid chromatography with ultraviolet and postcolumn chemiluminescence detectiõn. ~°, 11 To verify chemiluminescence-positive peaks as hydroperoxides, an aliquot of the Intralipid sample was treated with sodium borohydride. 12 The amounts of triglyceride hydroperoxides in Intralipid extracts were determined by area comparison of the eluting peaks with those of a standard of cholesteryl linoleate hydroperoxide, assuming an equal response of both hydroperoxides.13 Vitamin E isomers in the hexane layer in the Intralipid biphasic extract were analyzed by high-performance liquid chromatography with electrochemical detection. 14 For sodium ascorbate stability studies, fresh Intralipid was supplemented with sodium ascorbate (Aldrich Chemical, St. Louis, Mo.) at 100 #mol/L (final concentration) and incubated at 25 ° C. At various times an aliquot was removed, extracted, l° and the aqueous methanol layer was analyzed for ascorbate. 15 RESULTS We initially examined extracts of Intralipid as obtained from the local pharmacy before exposure to light (i.e., ùbaseline" Intralipid) for their contents of lipid hydroperoxides by means of high-performance liquid chromatography with specific postcolumn chemiluminescence detection for hydroperoxides. Analyses of 59 such baseline Intralipid samples revealed concentrations of triglyceride hydroperoxides of 7.8 + 4.0 #mol/L (mean value + SD). One of the seven shipments contained baseline Intralipid samples with an initial content of triglyceride hydroperoxides of "--100 #mol/L; the samples of that shipment were excluded. Treatment of these samples with sodium borohydride, which reduces hydroperoxides to the corresponding alcohols that themselves are not chemiluminescenceactive,13 largely eliminated the chemiluminescence-positivepeaks detected

in untreated samples and typical for triglyceride hydroperoxides. This demonstrated that the chemiluminescencepositive peaks were indeed hydroperoxides, t°, 12, t3 Analysis of a baseline Intralipid sample for its endogenous vitamin E showed ~ 6 0 ~mol/L 3,-tocopherol and 20/~mol/L each of the 6- and c~-isomers;«-tocopherol is biologically the most active form of vitamin E. t6 To simulate the clinical situation in the NICU, we exposed Intralipid dispensed in standard infusion plastic tubing to either ambient or phototherapy light at irradiances reeommended for effective therapy. The plastic tubing was connected to a syringe (used as the Intralipid reservoir). The syringes always remained in ambient light. As shown in the Table, levels of triglyceride hydroperoxides in all the syringes increased threefold during the 24-hour period. There was an additional doubling of the peroxide concentration in the tubing exposed to ambient light, whereas the concentration in the tubing exposed to phototherapy light increased to 65-fold that of the original Intralipid solution. Such high levels of triglyceride hydroperoxides occurred despite the presence of vitamin E, because even at the 24-hour exposure to phototherapy light, 72.5% + 5.5% and 88% _+ 2% (means _+ SD, n = 3) of the original c~- and ~/-tocopherol, respectively, remained in the most heavily oxidized Intralipid sample. For Intralipid samples used clinically for infusion into neonates (Table), the increases in triglyceride hydroperoxides under ambient light were similar to those seen in the simulated conditions. For infants undergoing phototherapy, hydroperoxide accumulation both in the syringes and in the tubing after 24 hours was higher than with ambient lighting. The values for the accumulated hydroperoxides in the tubing were lower than in the simulated experiments because the Intralipid solution was passing through the light beam rather than remaining static under the phototherapy light. Furthermore, in the clinical situation, with alterations

788

Neu~il et al.

i

"".-

The Journal of Pediatrics May 1995

i

I

i

I

i

300 3

~-~

200

DISCUSSION

2

1

0

e[-ù

ited peroxide formation in ambient light (data not shown). When added to Intralipid at 25 ° C and under ambient light, ascorbate at 100 #mol/L autooxidized linearly at a rate of 13.5 #mol/L per hour, so that significant and increasing amounts of its oxidation product(s) will be present in the Intralipid.

100 5 I

I

I

I

I

I

0

5

10

15

20

25

Time ( h ) Figure. Kinetics of accumulation of triglyceride hydroperoxides in Intralipid exposed to phototherapy light under conditions that mimic the situation in an NICU. A solution of 20% Intralipid was dispensed into a glass reservoir containing plastic tubing (line 3), plastic (line 1) õr glass reservoir (line 2), plastic reservoir covered with aluminum foil (line 4), or a plastic reservoir after supplementation with 1 mmol/L sodium ascorbate (line 5) and exposed to phototherapy light for up to 24 hours. For all samples the light fluxes measured were 11.7 to 25.9 ~W/cm 2 per nanometer for 24 hours. At the time points indicated, aliquots were withdrawn, extracted, and their hexane extracts analyzed for the presence of trigylceride hydroperoxides and c~-tocopherol as described in the Methods section. in the neonate's position and the placement of the light, a variable length of the tubing was exposed to phototherapy for a variable proportion of the 24 hours. To determine whether photosensitization by the plastic tubing was an important factor, we performed additional experiments, again under conditions that simulated the situation in the NICU. The time-dependent increase in triglyceride hydroperoxides on exposure to phototherapy light was similar for Intralipid in plastic or glass syringes (Figure, lines 1 and 2). At least 8% more triglyceride hydroperoxides accumulated in Intralipid dispensed in a glass syringe to which plastic infusion tubing was added (line 3), indicating some photosensitizing effect of the plastic tubing. However, these differences were small compared with those related to phototherapy versus ambient light exposure (Table), suggesting that the irradiance applied was largely responsible for hydroperoxide formation. Oxidation of Intralipid was inhibited almost completely by covering the syringe with aluminum foil (Figure). Supplementation of Intralipid with 1 mmol/L sodium ascorbate before exposure to phototherapy light also inhibited hydroperoxide formation by ~95%. Addition of water (as a control substance instead of ascorbate) did not affect the rate of triglyceride hydroperoxide formation, and ascorbate also inhib-

These data show that Intralipid is highly oxidizable under routine clinical conditions in an NICU, and this may result in high concentrations of lipid hydroperoxides being administered intravenously to large numbers of patients receiving this form of parenteral nutrition. This problem is particularly marked for neonates undergoing phototherapy, because this form of lighting is associated with greater prooxidant effect. Lipid oxidation occurs despite high concentrations of vitamin E, but appears to be preventable by the relatively simple maneuvers of wrapping the syringes and tubing in aluminum foil, or adding ascorbate to the infusate. The high susceptibility of Intralipid to oxidation may appear surprising because it contains large amounts of endogenous vitamin E. The latter is generally regarded as the most efficient chain-breaking, lipid-soluble antioxidant. ]6 However, recent reports have revealed that for vitamin E to be an efficient antioxidant in isolated lipoprotein emulsions, it requires a suitable "coantioxidant," such as vitamin C (ascorbate),17, 18 ubiquinol,19, 2o or albumin-bound bilirubin. 21 These coantioxidants eliminate the vitamin E radical that is formed from the interaction of the vitamin with initiating radical oxidants, which can have an adverse effect by promoting lipid peroxidation. ]8,19 Intralipid is also an emulsion of large particles, made of a triacylglycerol core surrounded by a monolayer of phospholipids.22 Hence vitamin E is likely to act similarly in Intralipid as in lipoproteins, and in the absence of suitable coantioxidants may cause formation of triglyceride hydroperoxides during storage. ~7-19Therefore increasing the concentration of this antioxidant alone is not likely to increase the oxidation resistance of Intralipid. Two recent reports have examined the formation of lipid hydroperoxides in infant feeding formulas. Helbock et al., 3 who used a method similar to ours, reported values of lipid hydroperoxides as high as 650 izmol/L in 20% Intralipid taken from the samples prepared by the pharmacy of the University of California, both before and after a 20-hour exposure to ambient light. The extent to which the differences in initial lipid peroxide concentrations are related to differences in the manufactured batches, handling, the processing of Intralipid in the local pharmacy or hospital before its use, or a combination of these factors, is not clear. Another report 23 showed that the level of lipid peroxides in human milk and infant formula was very low (<0.5 /zmol/L for the formula, ~1.5 #mol/L for

The Journal o f Pediatrics Volurne 126, Number 5, Part 1

the milk). Unlike Intralipid, the formula contained vitamin C, and human milk possesses natural antioxidant defenses. Exposure of these preparations to phototherapy light did not increase the concentration of lipid hydroperoxides. 23 Intravenous infusion with oxidized Intralipid in preterm neonates may be a clinically significant hazard. The highest level of lipid hydroperoxides detected in infused Intralipid ( ~ 1 0 0 #mol/L, Table) is some five orders of magnitude higher than that present in plasma of healthy adult human subjects. 8 Thus, even when the slow infusion rate, hemodilution, and rapid in vivo clearance of Intralipid (i.e., in the minute time scale) are considered, the overall plasma concentration of lipid hydroperoxides derived from the infused oxidized Intralipid may weil be within nr eren above the concentration range of endogenous hydroperoxides; local hydroperoxide concentrations would be expected to be eren higher. In addition to a direct cytotoxic effect, 4 lipid bydroperoxides in oxidized Intralipid could conceivably cause (pulmonary) vasoconstriction through interference with the endogenous synthesis of vasoregulatory prostaglandins,5' 6 inactivation of endothelium-derived relaxing factor, nr both, as has been shown for oxidized low-density lipoprotein and the lipids extracted from it. 24 In vitro oxidation of low-density lipoprotein results in non-cyclooxygenasederived esterified prostanoids (F2-isoprostanes),25 some of which are potent vasoconstrictors in their free form. Therefore esterified isoprostanes may also be formed during phototherapy light-induced oxidation of Intralipid and become biologically active during infusion and hydrolysis. Several strategies might minimize this potential!y serious problem. One simple solution is to protect the Intralipid from light exposure by wrapping the tubing and syringe with aluminum foil; in the future an aluminum foil sleeve nr dark tubing may be used to solve the problem identified here. Addition of ascorbate to Intralipid, although very effective and inexpensive, cannot currently be recommended, because side effects of relatively high concentrations of ascorbate, its oxidation products, nr both cannot be excluded. However, combining commercial water-soluble vitamin preparations used in parenteral nutrition with Intralipid in the syringe may be beneficial because some of these preparations contain ascorbate. We found a greater accumulation of hydroperoxides in the syringes for neonates receiving phototherapy even though they were theoretically not exposed to the spotlight. However, syringes may have been exposed while lights were redirected during periods when access to the infant was required. Thus phototherapy lights should be turned oft rather than redirected during infant care procedures. In this study, the analyses of lipids actually given to neonates were similar to the simulated conditions of light ex-

Neu~il et al.

789

posure described, and this will provide a useful model for further testing. Although we exclusively tested only one commonly used brand of lipid supplement, one form of phototherapy, and in only orte NICU, we believe that the results are widely applicable because these variables are similar in many hospital settings. We conclude that parenteral nutrition with Intralipid administered under standard practices in an N I C U results in the formation and intravenous infusion of relatively high levels of oxidized lipids. This may present an additiònal source of morbidity in premature neonates, but appears preventable by minimizing the exposure of the lipid to ambient and phototherapy lights. We thank Drs. W. Jessup, L. Krithrides, and D. Celermajer for critical reading of the manuscript. REFERENCES

1. Stahl GE, Spear ML, Mamosh M. Intravenous administration of lipid emulsions to premature infants. Clin Perinatol 1986; 13:133-62. 2. Pitkänen O, Hallman M, Andersson S. Generation of free radicals in lipid emulsionused in parenteral nutrition. Pediatr Res 1991;29:56-9. 3. Helbock HJ, Motchnik PA, Ames BN. Toxic hydroperoxides in intravenous lipid emulsionsused in preterm infants. Pediatrics 1993;91:83-7. 4. Yagi K, Ohkawa H, Ohnishi N, Yamashita M, Nakashima T. Lesions of aortic intima caused by intravenous administration of linoleicacid hydroperoxides. J Appl Biochem 1981;3:5865. 5. Hatzelmann A, Schatz M, Ullrich V. Involvement of glutathione peroxidase activity in the stimulation of 5-1ipoxygenase activity by glutathione-depleting agents in human polymorphonuclearleukocytes. Eur J Biochem 1989;180:52733. 6. Markey CM, Alward A, Weller PE, Marnett LJ. Quantitative studies of hydroperoxide reduction by prostaglandin H synthase. J Biol Chem 1987;262:6266-79. 7. LavoieJC, Chessex P. The increase in vasomotortorte induced by parenteral lipid emulsionis linked to an inhibition of prostacyclin production. Free Radic Biol Med 1994;16:795-9. 8. Bowry VB, Stanley KK, Stocker R. High-density lipoprotein is the major carrier of lipid hydroperoxides in human blond plasma from fasting donors. Proc Natl Acad Sci U S A 1992;89:10316-20. 9. Modi N, Keay AJ. Phototherapy for neonatal hyperbilirubinaemia: the importance of dose. Arch Dis Child 1983; 58:406-9. 10. Sattler W, Mohr D, Stocker R. Rapid isolation of lipoproteins and assessment of their peroxidation by HPLC postcolumn chemiluminescence.Methods Enzymol 1994;233:469-89. 11. Mohr D, Stocker R. Radical-mediated oxidation of isolated human very low density lipoprotein. Arterioscler Thromb 1994;14:1186-92. 12. Frei B, Yamamoto Y, Niclas D, Ames BN. Evaluation of an isnluminol chemiluminescence assay for the detection of hydroperoxides in human blond plasma. Anal Biochem 1988; 175:120-30. 13. Yamamoto Y, Brodsky MH, Baker JC, Ames BN. Detection and characterization of lipid hydroperoxides at picomoleler-

790

14.

15.

16.

17.

18.

19.

Neu}il et al.

The Journal of Pediatrics May 1995

els by high-performance liquid chromatography. Anal Biochem 1987;60:7-13. Pascoe GA, Duda CT, Reed DJ. Determination of a-tocopherol and ~-tocopherylquinonein small biological samples by high-performance liquid chromatography with electrochemical detection. J Chromatogr 1987;414:440-8. Behrens WA, Madere R. A highly sensitive high-performance liquid chromatography method for the estimation of ascorbic and dehydroascorbic acid in tissues, biological fluids, and foods. Anal Biochem 1987;165:102-7. Burton GW, Ingold KU. Vitamin E: application of the principles of physical organic chemistry to the exploration of its structure and function. Acc Chem Res 1986;19:194-201. Bowry VW, Ingold KU, Stocker R. Vitamin E in human lowdensity lipoprotein. When and how this antioxidant becomes a pro-oxidant. Biochem J 1992;288:341-4. Bowry VW, Stocker R. Tocopherol-mediated peroxidation. The pro-oxidant effect of vitamin E on the radical-initiated oxidation of human low-density lipoprotein. J Am Chem Soc 1993;115:6029-40. Ingold, KU, Bowry VW, Stocker R, Walling C. Autooxidation of lipids and antioxidation by a-tocopherol and ubiquinol in homogeneous solution and in aqueous dispersions of lipids. The unrecognized consequences of lipid particle size as exemplified

20.

21.

22.

23.

24.

25.

by the oxidation of human low-density lipoprotein. Proc Natl Acad Sei U S A 1993;90:45-9. Stocker R, Bowry VW, Frei B. Ubiquinol-10 protects human low-density lipoprotein more efficiently against lipid peroxidation than does c~-tocopherol. Proc Natl Acad Sei U S A 1991; 88:1646-50. Neu~il J, Stocker R. Free and albumin-boundbilirubin are efficient co-antioxidants for «-tocopherol, inhibiting plasma and low density lipoprotein lipid peroxidation. J Biol Chem 1994; 269:16712-9. Rotenberg M, Rubin M, Bot A, Meyuhas D, Talmon Y, Lichtenberg D. Physico-chemical characterization of [ntralipid emMsions. Biochim Biophys Acta 1991;1086:265-72. van Zoeren-Grobben D, M0ison RMW, Ester WM, Berger HM. Lipid peroxidation in human milk and infant formula: effect of storage, tube feeding and exposure to phototherapy. Acta Paediatr 1993;82:645-9. Chin JH, Azhar S, Hoffman BB. Inactivation of endothelialderived relaxing factor by oxidized lipoproteins. J Clin Invest 1992;89:10-8. Lynch SM, Morrow JD, Roberts LJ II, Frei B. Formation of noncyclooxygenase-derived prostanoids (F2-isoprostanes) in plasma and low-density lipoprotein exposed to oxidative stress in vitro. J Clin Invest 1994;93:998-1005.

FELLOWSHIPS Fellowships available in pediatrie subspecialties and those for general academic pediatric training are listed once a year, in January, in THE JOURNAL OF PEDIATRICS. Eaeh June, forms for listing fellowships available for the academic year beginning 18 months after publication are seht to the Chairman of the Department of Pediatries at major hospitals in the United Stares and Canada. In addition, a eopy of the application form appears in the July, August, and September issues of THE JOURNAL (please use the current form). Should you desire to list fellowships, a separate application must be made each year for each position. All applications must be returned to Mosby-Year Book, Inc., by October 15 preeeding the listing year to ensure publieation. Additional forms will be supplied on request from the Journal Editing Department, Mosby-Year Book, Inc., 11830 Westline Industrial Drive, St. Louis, M O 63146-3318/800-325-4177, ext. 4317, or 314453-4317.