Intravenous lipid composition affects hypoxic pulmonary ...

Download PDF
1 downloads 0 Views 84KB Size Report
Comment
Keith J. Barrington, George Chan, and John E. Van Aerde. Abstract: To examine the effects of altering the fatty acid (FA) composition of intravenous (IV) lipid ...
594

Intravenous lipid composition affects hypoxic pulmonary vasoconstriction in the newborn piglet Keith J. Barrington, George Chan, and John E. Van Aerde

Abstract: To examine the effects of altering the fatty acid (FA) composition of intravenous (IV) lipid emulsions on pulmonary vascular resistance (PVR) and thromboxane production, we studied three groups of newborn piglets after three days of either sow’s milk (milk), or total parenteral nutrition (TPN) with either iv soy bean oil (SBO, 52% n-6 and 8% n-3 FA) or fish oil (FO, 5% n-6 and 51% n-3 FA) emulsions. At baseline, and during hypoxia at 20 min and 2 h, cardiac output (Q) was measured, PVR calculated and plasma levels of a prostacyclin metabolite (6-keto-PgF1α) and thromboxane B2 (TxB2) were measured. Fatty acid composition of the lung phospholipids was analyzed. There was an exaggerated increase in PVR and decrease in Q during prolonged hypoxia in the TPN-SBO group as compared with the other two groups. There was no difference in PVR and Q between the milk and TPN-FO groups. FA of lung phospholipids reflected the high dietary level of long chain n-3 FA in the TPN-FO group. However, no differences in plasma levels of 6-keto-PgF1α or TxB2 were found. Intravenous emulsions made from SBO reduced cardiac output and increased pulmonary vascular resistance in the hypoxic newborn piglet, whereas iv FO emulsions did not. When subjects with pulmonary hypertension are receiving TPN iv SBO may be detrimental; iv FO may be beneficial, giving similar responses as in a milk-fed subject. Key words: total parenteral nutrition, fish oil, pulmonary hypertension, lipid emulsion, fatty acids. Résumé : Les effets de la modification de la composition en acides gras (AG) d’émulsions lipidiques Barrington intraveineuses et al. (iv) sur la résistance vasculaire pulmonaire (RVP) et la production de thromboxane ont été examinés chez trois groupes de porcelets nouveau-nés ayant reçu pendant trois jours du lait maternel (lait) ou une nutrition parentérale totale (NPT) d’émulsions à base d’huile de soya (HS, AG 52 % n-6 et 8 % n-4) ou d’huile de poisson (HP, AG 5 % n-6 et 51 % n-3). À l’état basal et durant une hypoxie à 20 minutes et à 2 heures, le débit cardiaque (Q) et les taux plasmatiques d’un métabolite de la prostacycline (6-kéto-PgF1 et de thromboxane B2 (TxB2) ont été mesurés, et la RVP calculée. La composition en acides gras des phospholipides pulmonaires a été analysée. Une augmentation de RVP et une diminution de Q excessives ont été notées durant l’hypoxie prolongée chez le groupe NPT-HS comparativement aux deux autres groupes. Aucune différence dans la RVP et Q n’a été observée entre les groupes lait et NPT-HP. Les AG des phospholipides pulmonaires ont reflété la forte teneur alimentaire en AG n-3 à longue chaîne chez le groupe NPT-HP. Toutefois, aucune différence n’a été relevée dans les taux plasmatiques de 6-kéto-PgF1“ ou de TxB2. Les émulsions intraveineuses à base d’HS, contrairement aux émulsions IV d’HP, réduisent le débit cardiaque et augmentent la résistance vasculaire pulmonaire chez le porcelet nouveau-né hypoxique. L’HS IV pourrait être dommageable chez les sujets souffrant d’hypertension artérielle qui reçoivent une NPT, alors que l’HP IV pourrait être bénéfique, donnant des réponses similaires au sujet nourri avec du lait. Mots clés : nutrition parentérale totale, huile de poisson, hypertension pulmonaire, émulsion lipidique, acides gras. [Traduit par la Rédaction]

600

Introduction Total parenteral nutrition (TPN) is the primary source of energy for sick neonates, including those with pulmonary hypertension, which is commonly seen in infants with hyaline membrane disease, hypoxia, asphyxia, group B Received February 9, 2001. Published on the NRC Research Press Web site at http://cjpp.nrc.ca on Juen 28, 2001. K.J. Barrington, G. Chan, and J.E. Van Aerde.1 Department of Pediatrics, Division of Newborn Medicine, Stollery Children’s Health Center, University of Alberta, Edmonton, AB T6G 2B7, Canada. 1

Corresponding author (e-mail : [email protected]).

Can. J. Physiol. Pharmacol. 79: 594–600 (2001)

streptococcal sepsis, or persistent fetal circulation. There is accumulating evidence that intravenous (iv) soy bean oil (SBO) emulsions may have negative effects on pulmonary vascular resistance, hypoxic pulmonary vascular responses, and pulmonary outcomes in the newborn infant (Cooke 1991; Hammerman and Aramburo 1988; Hammerman et al. 1989; Prasertsom et al. 1996). In chronically instrumented lambs, commercial iv SBO emulsions cause a dose-dependent pulmonary vasoconstriction with a decrease in PaO2 and a significant increase in thromboxane A2 without affecting 6-keto-PgF1α (Teague et al. 1987). In the preterm infant we demonstrated a doseresponse relationship between the dose of lipid infusion and the degree of increase in pulmonary artery pressure as esti-

DOI: 10.1139/cjpp-79-7-594

© 2001 NRC Canada

Barrington et al.

595 Table 1. Fatty acid composition of triglycerides and phospholipids of lipid emulsions (percent, w/w). Triglycerides

Phospholipids

Fatty acid

Milk

Soy bean oil

Fish oil

C14:0 C16:0 C18:0

34 4

11.4 3.6

8.8 10.5 1.7

33.3 13.7

C16:1 C18:1

11 35

24.3

10.5 9.7

1.2 26.6

52.1

2.8 1.4

16.4 4.9 0.9

7.1

1.1 4.7 29 2.9 13.2

1.1

22.8 21.7 4.6 50.9

47.5 29.1 22.3 1.1

C18:2n-6 C20:4n-6 C22:5n-6

8.3 0.3

C18:3n-3 C18:4n-3 C20:5n-3 C22:5n-3 C22:6n-3

1

Total Total Total Total

saturates monoenes n-6 n-3

0.1 0.1 0.1 44 46 8.7 1.4

15.2 24.9 52.3 7.5

mated by echocardiography (Prasertsom et al. 1996). There is also evidence of an association between thromboxane levels and pulmonary hypertension during the use of iv SBO emulsions (Hammerman et al. 1989). In adults, thromboxane A2 production can be decreased by an oral fish diet, whereas the production of thromboxane A3, a much weaker vasoconstrictor, tends to be enhanced by such a diet (Brox et al. 1981; Knapp et al. 1986). Similarly, in chronically hypoxic rats, mean pulmonary arterial pressure is lower when fed fish oil by mouth as compared with corn oil (Archer et al. 1989). This fish oil diet increases the lung content of eicosapentaenoic acid 50-fold and decreases the arachidonic acid content by 60%, resulting in lower thromboxane B2 levels and decreased mortality during chronic hypoxia (Archer et al. 1989). Studies in swine have shown that pulmonary hypertension and hypoxia in response to endotoxemia are significantly reduced after consuming a diet rich in fish oil derived n-3 fatty acids as compared with a diet enriched with n-6 fatty acids (Murray et al. 1993). Thromboxane B2 and 6-keto-PgF1α levels were also reduced in the fish oil group. It has, therefore, been postulated that the likely adverse effects of early iv soy bean emulsions in preterm infants may be caused by increased pulmonary vascular resistance, leading to increased requirements for positive pressure ventilation and oxygen requirements which, as previously suggested, can cause long-term morbidity such as bronchopulmonary dysplasia (Cooke 1991; Hammerman and Aramburo 1988). We hypothesized that isocaloric replacement of SBO by FO in newborn piglets receiving TPN would attenuate hypoxic pulmonary vasoconstriction responses. We used a previously developed FO emulsion that is high in eico-

sapentaenoic (EPA) and docosahexaenoic acid (DHA) and low in linoleic acid.

Materials and methods Animals and TPN Three groups of six newborn piglets (Camborough/Canabreed), up to 24 h old and with birth weights between 1.5 and 2.7 kg entered the study. Group 1 (milk) was fed sow’s milk. The animals in groups 2 (SBO) and 3 (FO) were fed intravenously for ≥72 h but less than 96 h after entering the study. Total parenteral nutrition was given at 213 kcal·kg–1·day–1 and consisted of 31% of calories as lipid, 40% as carbohydrate, and 29% as protein. The only difference in the feeding regimen between groups 2 and 3 was the composition of the lipid emulsion; group 2 was fed a lipid emulsion prepared from 10% (wt/wt) SBO triglycerides (Sigma Chemical Co., St. Louis, Mo.), 1.2% egg yolk lecithin (Cal Biochem, San Diego, Calif.), 2.5% glycerol and water, as described previously (Duerksen et al. 1996). The final fatty acid composition of this emulsion (Table 1) was similar to the fatty acid composition of lipid emulsions used in clinical practice. Group 3 was fed a lipid emulsion prepared from 10% (w/w) fish oil triglycerides (Nisshin Flour Milling Co., Japan), 1.2% egg yolk lecithin (Cal Biochem), 2.5% glycerol, and water (Table 1) (Van Aerde et al. 1999). On the day of arrival, under general anesthesia, the IV fed animals had external jugular catheters inserted bilaterally, which were advanced into the right atrium and tunneled through an incision in the posterior neck. The animals were kept in a thermoneutral environment in Plexiglas cages with swivels allowing freedom of movement and access to water ad libitum. A 10% amino acid solution (Travasol® Baxter, Toronto, Canada) was mixed with dextrose, multivitamins (1mL/kg every 2nd day; MVI-Pediatric, RhonerPoulenc Rorer, Canada), essential trace elements (MTE, Lyphomed, Canada), calcium gluconate (Abbott, Canada), and electrolytes. This solution was infused via a metered infusion © 2001 NRC Canada

596 Fig. 1. Pulmonary vascular resistance index (PVRI) during the experiment (means ± SD). PVRI was significantly higher in the group given iv soy bean oil (SBO) after prolonged hypoxia (* P < 0.05) compared with the other two groups ( †P < 0.05 compared with normoxia). There was no difference between the milk-fed group (milk) and the iv fish oil group (FO). There were no differences between groups during normoxia or short term hypoxia. Values during hypoxia were significantly higher in each group than during normoxia.

pump. This resulted in an intake of electrolytes and trace elements similar to those previously reported for TPN-fed animals (Duerksen et al. 1996; Van Aerde et al. 1999). The lipid emulsion was infused at 60 mL·kg–1·day–1; the amino acid–glucose solution and the lipid emulsion were infused through separate catheters. Approval of the Health Sciences Animal Welfare Committee of the University of Alberta was obtained for all procedures according to the guidelines of the Canadian Council of Animal Care.

Cardiovascular measurements On the day of the physiological measurements, the TPN was discontinued about 2 hours prior to transport into the cardiovascular physiology lab area and 0.2 mg·kg–1 iv acepromazine was given just prior to transport. To minimize circadian variability, all measurements were started close to 0900 h. The investigator performing the instrumentation and hypoxic pulmonary vasoconstriction responses (K.J.B.) was blinded as to which feeding group the animals belonged. Halothane was given until laryngeal reflexes were abolished and animals were intubated then placed on a pressurelimited time cycled ventilator at a peak inspiratory pressure of 16 cm H2O and a peak end-expiratory pressure of 3 cm H2O. The initial ventilator rate was 20 per min and the rate was adjusted to keep the arterial Pa CO 2 between 36 and 44 torr. This range of Pa CO 2 was maintained throughout the experiment. After intubation, halothane was discontinued and an infusion of 20 µg·kg–1·h–1 of fentanyl was commenced. At this time one of the carotid arteries was catheterized and connected to a pressure transducer; a 6 French sheath was inserted into the ipsilateral external jugular vein. Under fluoroscopy, a #5 French flow directed balloon-tipped thermodilution catheter was advanced into the main pulmonary artery. Position was checked by fluoroscopy and pressure wave, and was confirmed by autopsy at the end of the experiment. Central venous

Can. J. Physiol. Pharmacol. Vol. 79, 2001 and arterial pressures were continuously monitored using a computerized data acquisition system, digitizing the data at a rate of 24 samples per second. Pulmonary pressures were continuously monitored from the distal tip of the balloon-tipped catheter. Thermodilution cardiac output was measured using 3 mL of iced water that was manually injected as quickly as possible during expiration, into the contralateral venous catheter, which terminated in the right atrium. Thermodilution estimates were always performed in duplicate and if more than a 10% difference between measurements was noted, triplicate and quadruplicate measurements were performed. Inflation of the catheter balloon to record “wedge” pressure leads to acute bradycardia in animals of this size, therefore no estimate of left atrial pressure was available. Pulmonary vascular resistance was calculated as the quotient of mean pulmonary artery pressure and cardiac index. Hemodynamic variables were recorded after 20 min of stabilization and a 2 mL blood sample was taken in an EDTA tube containing 2 µg·mL–1 of indomethacin for later analysis of TxB2 and 6-keto-PgF1α. The animal was then made hypoxic by reducing the inspired oxygen concentration to approximately 12%. FiO2 was adjusted in order to achieve an arterial oxygen saturation of 55–65%, which was continuously monitored by pulse oximetry and confirmed by in vitro cooximetry. After 20 min of hypoxia, hemodynamic measurements were repeated and the animals were made normoxic again by ventilating with an FiO2 of 0.45 to ensure 100% saturation. After 15 min of normoxia, the saturation was confirmed by co-oximetry, and hemodynamic baseline measurements were repeated before the animals were made hypoxic again and maintained hypoxic for 2 h, using the same technique. At the end of the two hours, the hemodynamic measurements and saturations were again recorded; a 2-mL blood sample was taken from the arterial catheter in an EDTA tube containing 2 µg·ml–1 of indomethacin for later analysis of TxB2 and 6-keto-PgF1α. After euthanization of the animal with an overdose of IV pentobarbital, and after opening the animal’s chest, the lungs were removed, rinsed in ice-cold saline, and cut in small segments before being frozen in liquid nitrogen and stored at –80oC. Finally, the correct position of the catheters was confirmed.

Biochemical analyses Plasma was separated from cells by centrifugation at 2000 g for 10 min and stored frozen at –80oC until analysis. Plasma 6keto-F1α was purified via Amprec C2 columns (Amersham, Buckinghamshire, U.K.) and assayed with a RIA kit (Amersham TRK 790). Plasma TxB2 was measured using a RIA kit (NEN NEK-007, Dupont, Mississauga, Ont.) and stripped porcine plasma was included as the analyte in constructing the standard graph. Weighed aliquots of lung were homogenized in 0.1 M phosphate buffer (pH 7.4), acidified to pH 3 with 1 M HCl, and passed through Amprec C2 columns. Purified 6-keto-F1α and TxB2 were analyzed with the respective RIA kits. Other portions of lung tissue were homogenized and lipids extracted using the modified Folch procedure. Total phospholipids were separated on a silica G plate, using a petroleum ether : diethylether : glacial acetic acid 80 : 20 : 1 solvent system. The phospholipid fraction was methylated with boron trifluoridemethanol, and was quantitated by gas liquid chromatography using diheptadecanoyl phosphatidylcholine as an internal standard.

Statistical analysis All values are expressed as means ± standard deviation (SD). Using ANOVA, and in order to detect a minimum mean difference of 50% between groups with an expected standard deviation in PVRI measurements of 20% (as we have previously shown, Cheung et al. 1998), and, assuming an alpha value of 0.05, a minimum sample size of 6 in each group was required to reach a power of 0.8. Multivariate analysis and repeated measures analysis of © 2001 NRC Canada

Barrington et al.

597 Table 2. Cardiovascular parameters of the milk-fed, iv SBO, and iv FO groups. Parameters Baseline Q PAP SAP SVR Early hypoxia Q PAP SAP SVR 2nd baseline Q PAP SAP SVR Late hypoxia Q PAP SAP SVR

Milk

SBO

FO

262 25 77 0. 30

± ± ± ±

51 5 6 0.07

213 20 74 0.36

± ± ± ±

50† 5 7 0.10*

263 24 80 0.32

± ± ± ±

51 3 2 0.06

232 34 88 0.29

± ± ± ±

87# 7# 14 0.08

240 26 75 0.32

± ± ± ±

50# 6# 18 0.07

288 32 73 0.28

± ± ± ±

68# 6# 12 0.06

246 ± 64 18.5 ± 6.8 77 ± 5 0.32 ± 0.07

204 19.1 77 0.40

± ± ± ±

47 4.9 9 0.11

250 22.1 76 0.32

± ± ± ±

52 3.1 9 0.11

278 42 90 0.33

193 34 74 0.38

± ± ± ±

33† 5# 22 0.08

233 36 62 0.28

± ± ± ±

36 7# 13 0.06

± ± ± ±

75 7# 20 0.05

Note: Values are means ± SD. Milk, sow milk fed; SBO, soy bean oil; FO, fish oil; Q, cardiac index (mL·kg–1·min–1); PAP, pulmonary arterial pressure (torr); PVR, pulmonary vascular resistance (torr·mL–1·kg–1·min); SAP, systemic arterial pressure (torr); SVR, systemic vascular resistance (torr·mL–1·kg–1·min).*P < 0.05 as compared with milk—fed group, †P < 0.05 compared with both milk—fed and FO groups. #P < 0.05 compared with the baseline results in the same group.

variance were applied where appropriate. When a significant difference was noted, a two-tailed t test with Bonnferoni correction to compensate for multiple comparisons was performed. When the normality or equal variance tests failed, a Kruskal-Wallis one-way analysis of variance on ranks was used with Dunn’s method for pairwise multiple comparisons. Linear regression of TxB2 production after 2 h of hypoxia versus pulmonary vascular resistance was used. A P value of < 0.05 was considered statistically significant.

Results The mean cardiac index (Q) of the control group at baseline was not significantly different from Q in the FO group, but the SBO group had a lower Q (P < 0.05) (Table 2). Mean arterial blood pressure was not different between the groups (Table 2). Calculated systemic vascular resistance (SVR) was significantly greater in the SBO group at baseline (P < 0.05). Pulmonary artery pressure (PAP) and pulmonary vascular resistance (PVR) at baseline did not differ significantly between groups (Fig. 1). The Pa CO2 did not differ significantly throughout the experiment; the mean Pa CO2 was between 39 and 41 throughout, with no difference between groups. The pH remained unchanged during the first 3 phases of the experiment; pH was 7.44 (SD 0.07) during the baseline period, 7.39 (SD 0.09) during the first hypoxic period, and 7.40 (SD 0.06) during the second normoxic period. There was a significant fall in pH due to metabolic acidosis because of prolonged hypoxia by the end of the second hypoxic phase (pH 7.31 SD 0.15, P < 0.05). There was no difference between groups.

Cardiac output increased with short-term hypoxia. There was a similar increase in PVR in each group after a 20-min period of hypoxia (Fig. 1). However, cardiac output was lower at each stage in the SBO group, and therefore the PAP during short term hypoxia was lower in the SBO group. Following the prolonged 2-h period of hypoxia, there was a progressive increase in pulmonary vascular resistance in all groups, but this was significantly greater in the SBO than in the milk or FO group (Fig. 1). After prolonged hypoxia, the cardiac index was also significantly lower in the SBO as compared with the other two groups, leading to a significant greater rise in PVR in the SBO group. PVR and Q for the FO and control group were not different from each other (Table 2; Fig.1). The fatty acid composition of lung phospholipids reflected to a certain extent the fatty acid composition of the respective diets (Table 3). The long chain n-3 fatty acids were predominant in the FO group with both eicosapentaenoic and docosahexaenoic acid being significantly higher in the FO group. Similarly, the total n-6 fatty acid content was significantly lower in the FO group. However, despite the linoleic acid content of the lung phospholipids being highest in the SBO group, this did not result in an increased arachidonic acid content. The ratio of eicosapentaenoic to arachidonic acid was 20 times higher in the fish oil than in the milk-fed and SBO groups (Table 3). No differences were found in plasma PgF1α (baseline: milk 520 ± 68 ng·mL–1, SBO 533 ± 149 ng·mL–1, FO 519 ± 188 ng·mL–1; 2-h hypoxia: milk 524 ± 72 ng·mL–1, SBO 642 ± 195 ng·mL–1, FO 636 ± 222 ng·mL–1; not significant) and TxB2 content (baseline: © 2001 NRC Canada

598

Can. J. Physiol. Pharmacol. Vol. 79, 2001

Table 3. Phospholipid fatty acid composition of whole lung tissue (percent, w/w). Study Group Milk SBO FO

C18:2n-6 a

3.87 ± 0.46 6.87 ± 0.51b 2.13 ± 0.24c

C20:4n-6

C20:5n-3

C22:6n-3

EPA/AA

Total n-6

Total 3-n

8.33 ± 1.15* 4.29 ± 0.95 3.97 ± 0.50

0.22 ± 0.01 0.11 ± 0.04 2.34 ± 0.29*

1.13 ± 0.16 0.83 ± 0.39 1.77 ± 0.21*

0.026 ± 0.01 0.027 ± 0.01 0.591 ± 0.06*

14.0 ± 1.7 12.3 ± 1.6 6.9 ± 0.8*

2.30 ± 0.50 1.61 ± 0.43 5.57 ± 0.64*

Note: Values are means ±SD. Milk, sow milk fed; SBO, soy bean oil; FO, fish oil. Significantly different from the two other groups within the same column (*P < 0.05). Groups with different superscripts within the same column are significantly different from each other (P < 0.05).

milk 1717 ± 997 ng·mL–1, SBO 1153 ± 610 ng·mL–1, FO 1143 ± 827 ng·mL–1; 2-h hypoxia: milk 1550 ± 912 ng·mL–1, SBO 1086 ± 241 ng·mL–1, FO 836 ± 693 ng·mL–1; not significant) either for different diets or between baseline and after prolonged hypoxia. No correlation was found between plasma TxB2 level and Q or PVR.

Discussion Postnatal pulmonary development follows a similar pattern in swine and man, although development in the piglet is completed faster (Coe et al. 1987; Rendas et al. 1978). The piglet pulmonary vasculature is very responsive to hypoxia (Kato et al. 1980). The piglet has also been validated as a newborn animal model for TPN in the neonate (Duerksen et al. 1996). We have demonstrated that hypoxic pulmonary vasoconstriction in response to prolonged hypoxia is significantly worse when piglets have received three days of TPN with iv soy bean oil as the lipid source as compared to animals fed with iv fish oil or milk-fed animals. Baseline PVR was not affected by the diet. As in the human neonate, administering iv lipid emulsions might not cause significant pulmonary vasoconstriction unless there is another cause for elevated pulmonary vascular resistance, such as when the infant is experiencing respiratory distress (Prasertsom et al. 1996). It should also be noted that PVR was measured about 2 h after stopping TPN and it is possible that baseline PVR changed during this period. This would be consistent with our previous results in the human neonate in whom stopping iv SBO did lead to a rapid decrease in right ventricular pre-ejection time (Prasertsom et al. 1996). However, during hypoxia and 4 h after discontinuing the TPN, the SBO group demonstrated increased PVR, consistent with results in other models. It was demonstrated that there are several phases to the hypoxic PVR response; an initial vasoconstriction is followed by partial relaxation, which in turn is followed by a progressive slow rise in pulmonary vascular resistance (Bennie et al. 1991; Leach et al. 1994). The mechanisms for these responses are still somewhat uncertain, but it appears that the initial phase is modified by the release of nitric oxide and eicosanoids (Gordon et al. 1997). This is followed by a prolonged and entirely endothelium-dependent phase (Leach et al. 1994). The biphasic nature of the hypoxic PVR and the likelihood of two different mechanisms underlying the two phases may explain the varying results reported with regard to the role of cyclo-oxygenase and lipoxygenase products in hypoxic PVR in the neonate (Cassin et al. 1989; Fike and Kaplowitz 1994; Gordon et al. 1997; Taylor et al. 1992). In neonatal studies,

inhibition of cyclo-oxygenase with indomethacin reduces hypoxic PVR in some studies, and leukotrienes may also be involved, but probably only in the late phase of the response (Schreiber et al. 1985). In isolated lungs of rabbits, a 3-h infusion of an iv FO emulsion resulted in a shift from leukotriene LtC4 to LtC5, resulting in an attenuated pulmonary arterial pressure reaction and decreased formation of lung edema (Koch et al. 1993, 1995; Breil et al. 1996). LtC4 and LtB4 have strong vasoconstrictor actions on pulmonary arteries (Coe et al. 1988). Dietary fish oil is also known to alter the LtB4 : LtB5 ratio in alveolar macrophages (Kobayashi et al. 1995) and leukocytes (Grimminger et al. 1996; Morlion et al. 1996). Unfortunately, we were unable to measure leukotriene metabolism in our animals, thus the contribution of shifts in leukotriene metabolites to our results is unknown. The fatty acid composition of lung phospholipids did reflect the dietary fatty acid composition. Dietary fatty acids have been shown to be incorporated into lung phospholipids shortly after ingestion (Koch et al. 1993). As in our study, other investigators have reported that oral fish oil for several weeks results in an increased EPA and DHA content in lung tissue with a reduction in linoleic and arachidonic acid (Archer et al. 1989; Baybutt et al. 1993; Rayon et al. 1997). In juvenile pigs receiving an oral n-3 enriched diet for 9 days, the administration of endotoxin did result in a lower pulmonary vascular resistance as compared with the control and n-6 enriched diet groups (Murray et al. 1993). This was accompanied by changes in plasma and platelet phospholipid fatty acid composition that reflected the dietary fatty acid composition; in the n-3 group, both 6-keto-PgF1α and TxB2 were decreased. We did not find the latter in our study, possibly because of the shortness of the experiment, the size of the animal groups, or the variability of the metabolites. However, as we were unable to measure TxB3 and PgI3 metabolites, we do not have data on the ratio of the EPA and arachidonate derivatives, yet it is the ratio and not the absolute concentration that is of major concern (Morlion et al. 1996). We presume that the finding of significant alterations in haemodynamics and hypoxia responses, despite there being no measured difference in plasma prostanoid concentrations, may be because there was an unmeasured increase in the n-3 series prostanoids. If the diets had been continued for longer we postulate that a decrease in plasma n-6 series prostanoids would have taken place. We also found, unexpectedly, that there was an significant difference between baseline cardiac output in the SBO group as compared with the other groups. As this was accompanied by similar systemic blood pressures in all groups, the calculated SVR was higher in the SBO group. The reason behind the lower baseline cardiac output and higher SVR © 2001 NRC Canada

Barrington et al.

may lie in the neonatal animal model we used. It is well recognized that the neonatal myocardium is very intolerant of increased afterload (Belik and Light 1989; Van Hare et al. 1990). Altering systemic and pulmonary vasomotor tone by administering SBO and the generation of vasoconstrictor mediators for >72 h may have effects on cardiac function that persist beyond cessation of the lipid infusion. We postulate that the animals receiving SBO had impaired ventricular function, which was due to 3 days of SBO-induced vasoconstriction, that led to decreased cardiac output and increased SVR. We are unaware of any other controlled comparisons of the administration of SBO emulsions to the newborn that investigated SVR or cardiac function. There are very few randomized controlled clinical trials of iv lipid usage in the human newborn. None of these studies investigated cardiac output or SVR. It is possible that impaired cardiac function due to iv SBO oil emulsions accounted in part for the increased mortality seen in the extremely low birth weight infant group that received iv lipid very early in the randomized study by Sosenko et al. (1993). If this explanation for the decreased baseline cardiac output is confirmed by future studies, then it strongly suggests that alternatives to SBO emulsions should be sought for nutritional support of the newborn. Our results give particular cause for concern regarding the clinical use of iv SBO emulsions in hypoxia and in clinical situations with increased pulmonary vascular pressure such as meconium aspiration syndrome, asphyxia, and persistent fetal circulation. Although these results can not be directly related to the human neonate, the combined weight of the evidence to date would suggest that PVR is increased during infusion of a SBO emulsion and that the hypoxic PVR response remains elevated even several hours after the iv SBO has been discontinued. This is in contrast with iv FO, which results in increased levels of long chain n-3 fatty acids in lung tissue and does not adversely affect pulmonary vascular responses. The exact mechanism of these effects remains unclear, but may be due to changes in leukotriene or isoprostane metabolism. We suggest caution in using iv SBO during acute illness in the newborn infant and propose further research into emulsions containing long chain n-3 fatty acids.

Acknowledgements This study was supported by grants from the Medical Research Council of Canada. The advice and support of Dr. M.T. Clandinin with the fatty acid analysis is greatly appreciated. The authors wish to thank E. Paslawski, and A. Wierzbicki for technical assistance, and Kathleen-Ruth Eck and Judy Minckler for secretarial support. MTE-conc trace elements were kindly donated by Lyphomed, Canada, calcium gluconate by Abbott, Canada and MVI-Pediatric by Rhone-Poulenc Rorer, Canada.

References Archer, S., Nelson, D., Gebhard, R., Levine, A.S., Prigge, W., and Weir, E.K. 1987. Effects of dietary fish oil on lung phospholipid fatty acid composition and intrinsic pulmonary vascular reactivity. Cardiovasc. Res. 21: 928–932.

599 Archer, S., Johnson, G., and Gebhard, R. 1989. Effect of dietary fish oil on lung lipid profile and hypoxic pulmonary hypertension. J. Appl. Physiol. 66: 1662–1673. Baybutt, R.C., Smith, J.E., and Yeh, Y.Y. 1993. The effects of dietary fish oil on alveolar type II cell fatty acids and lung surfactant phospholipids. Lipids, 28: 167–172. Belik, J., and Light, R.B. 1989. Effect of increased afterload on right ventricular function in newborn pigs. J. Appl. Physiol. 66: 863–869. Bennie, R.E., Packer, C.S., Powell, D.R., Jin, N., and Rhoades, R.A. 1991. Biphasic contractile response of the pulmonary artery to hypoxia. Am. J. Physiol. 261: L156–L163. Breil, I., Koch, T., Heller, A., Schlotzer, E., Grunert, A., van Ackern, K., and Neuhof, H. 1996. Alteration of n-3 fatty acid composition in lung tissue after short-term infusion of fish oil emulsion attenuates inflammatory vascular reaction. Crit. Care Med. 24: 1893–1902. Brox, J., Killie, J., and Gunnes, S. 1981. The effects of cod liver oil and corn oil on platelets and vessel wall in man. Thromb. Haemostasis, 46: 604–611. Cassin, S., Gause, G., Ter Riet, M., and Baker, R. 1989. Do inhibitors of lipoxygenase and cyclo-oxygenase block neonatal hypoxic pulmonary vasoconstriction? J. Appl. Physiol. 66: 1779–1784. Cheung, P.-Y., Barrington, K., and Bigham, D.L. 1998. Temporal effects of prolonged hypoxaemia and reoxygenation on systemic, pulmonary and mesenteric perfusions in newborn piglets. Cardiovasc. Res. 39: 451–458. Coe, Y., Olley, P., and Hamilton, F.A. 1987. Method to study pulmonary vasculary response in the conscious newborn piglet. Can. J. Physiol. Pharmacol. 65: 785–790. Coe, Y., Olley, P., and Coceani, F. 1988. The effect of leukotriene4 on pulmonary and systemic circulation in conscious newborn piglets. Prostaglandins, 36: 31–47. Cooke, R.W.I. 1991. Factors associated with chronic lung disease in preterm infants. Arch. Dis. Child. 66: 776–779. Duerksen, D.R., Van Aerde, J.E., Gramlich, L., Meddings, J.B., Chan, G., Thomson, A.B.R., and Clandinin, M.T. 1996. Intravenous ursodeoxycholic acid reduces cholestasis in parenterally fed newborn piglets. Gastroenterology, 111: 1111–1117. Fike, C.D., and Kaplowitz, M.R. 1994. Developmental differences in vascular responses to hypoxia in lungs of rabbits. J. Appl. Physiol. 77: 507–516. Gordon, J., Moreira, G., O’Donnell, D., Aldinger, A.M., and Tod, M.L. 1997. Relative effects of cyclooxygenase and nitric oxide synthase inhibition on vascular resistances in neonatal lamb lungs. Pediatr. Res. 42: 738–743. Grimminger, F., Grimm, H., Fuhrer, D., Lindemann, G., Blecher, C., Mayer, K., Tabesch, F., Kramer, H.J., Stevens, J., and Seeger, W. 1996. ω-3 Lipid infusion in a heart allotransplant model. Shift in fatty acid and lipid mediator profiles and prolongation of transplant survival. Circulation, 93: 365–371. Hammerman, C., and Aramburo, M.J. 1988. Decreased lipid intake reduces morbidity in sick premature neonates. J. Pediatr. 113: 1083–1088. Hammerman, C., Aramburo, M.J., and Hill, V. 1989. Intravenous lipids in newborn lungs: Thromboxane-mediated effects. Crit. Care Med. 17: 430–436. Kato, H., Kidd, L., and Olley, P. 1980. Effect of hypoxia on pulmonary vascular response in conscious newborn piglets. J. Appl. Physiol. 48: 188–196. Knapp, H., Reilly, P., and Allessandrini, P. 1986. In vivo indices of © 2001 NRC Canada

600 platelet and vascular function during fish oil administration in patients. N. Engl. J. Med. 314: 937–942. Kobayashi, J., Yokoyama, S., and Kitamura, S. 1995. Eicosapentaenoic acid modulates arachidonic acid metabolism in rat alveolar macrophages. Prostaglandins, Leukotrines and Essent. Fatty Acids, 52: 259–262. Koch, T., Duncker, H.P., Klein, A., Schlotzer, E., and Peskar, B.M. 1993. Modulation of pulmonary vascular resistance and edema formation by short-term infusion of a 10% fish oil emulsion. Infusiontherapie und Transfusionsmedizin, 20: 291–300. Koch, T., Heller, A., Breil, I., Van Ackern, K., and Neuhof, H. 1995. Alterations of pulmonary capillary leukotriene synthesis due to infusion of a lipid emulsion enriched with omega-3 fatty acids. Clin. Intensive Care, 6: 112–120. Leach, R.M., Robertson, T.P., Twort, C.H.C., and Ward, J.P.T. 1994. Hypoxic vasoconstriction in rat pulmonary and mesenteric arteries. Am. J. Physiol. 266: L223–L231. Morlion, B., Torwesten, E., Lessire, H., Sturm, G., Peskar, B., Furst, P., and Puchstein, C. 1996. The effect of parenteral fish oil on leukocyte membrane fatty acid composition and leukotriene-synthesizing capacity in patients with postoperative trauma. Metabolism, 45: 1208–1213. Murray, M., Svingen, B.A., Yaksh, T.L., and Holman, R.T. 1993. Effects of endotoxin on pigs prefed omega-3 vs. omega-6 fatty acid-enriched diets. Am. J. Physiol. 265: E920–E927. Prasertsom, W., Phillipos, E.Z., Van Aerde, J.E., and Robertson, M. 1996. Pulmonary vascular resistance during lipid infusions in neonates. Arch. Dis. Child. 74: F95–F98. Rayon, J.I., Carver, J.D., Wyble, L.E., Wiener, E., Dickey, S.S.,

Can. J. Physiol. Pharmacol. Vol. 79, 2001 Benford, V.J., Chen, L.T., and Lim, D.V. 1997. The fatty acid composition of maternal diet affects lung prostaglandin E2 levels and survival from group B streptococcal sepsis in neonatal rat pups. J. Nutr. 127: 1989–1992. Rendas A., Branthwaite, M., and Reid, L. 1978. Growth of pulmonary circulation in normal pig-structural analysis and cardiopulmonary function. J. Appl. Physiol. 45: 806–817. Schreiber, M.D., Heymann, M.A., and Soifer, S.J. 1985. Leukotriene inhibition prevents and reverses hypoxic pulmonary vasoconstriction in newborn lambs. Pediatr. Res. 19: 437–441. Sosenko, I.R., Rodriguez-Pierce, M., and Bancalari, E. 1993. Effect of early initiation of intravenous lipid administration on the incidence and severity of chronic lung disease in premature infants. J. Pediatr. 123: 975–982. Taylor, B.J., Sziszak, T.L., Sziszak, T.J., and Dang, A.Q. 1992. Effects of alveolar hypoxia on the pulmonary circulation and lung mechanics after cromolyn sodium and U-60, 257 in lambs. J. Dev. Physiol. 18: 73–80. Teague, W., Raj, V., and Braun, D. 1987. Lung vascular effects of lipid infusion in awake lambs. Pediatr. Res. 22: 714–719. Van Aerde, J.E., Duerksen, D.R., Gramlich, L., Meddings, J.B., Chan, G., Thomson, A.B.R., and Clandinin, M.T. 1999. Intravenous fish oil emulsion attenuates TPN-induced cholestasis in newborn piglets. Pediatr. Res. 45: 202–208. Van Hare, G.F., Hawkins, J.A., Schmidt, K.G., and Rudolph, M.A. 1990. The effects of increasing mean arterial pressure on left ventricular output in newborn lambs. Circ. Res. 67: 78–83.

© 2001 NRC Canada