Intralipid

Bahman Venus, M.D., F.C.C.P.; Robert A. Smith, M.S.; Chandrakant Patel, M.D.; and Edmundo Sancoval, M.D., F.C.C.P.

Hemodynamic and pulmonary gas exchange consequences of 20 percent intravenous fat emulsion infusion (3.0 [+ or -] .3 mg/kg/min) were evaluated in 19 patients who demonstrated ARDS. Lipid infusion precipitated a significant reduction in [PaO.sub.2/FIo.sub.2] from 241[+ or -]50 to 184[+ or -]41 (mean [+ or -] SD) and increased MPAP from 26.0[+ or -]5.1 to 31.8[+ or -]4.8 mm Hg, pulmonary vascular resistance from 149[+ or -]78 to 179[+ or -]61 dyne*s/[cm.sub.5] and pulmonary venous admixture (Qva/Qt) from 20.7[+ or -]15.2 to 30.6[+ or -]8.6 percent. Further analysis revealed that the magnitude of increased Qva/Qt was greater in patients who manifested septicemia (N = 10) compared to those who did not (N = 9): 12.3 vs 7.3 percent, respectively. We conclude that intravenous lipid administration was associated with increased MPAP and Qva/Qt in patients with ARDS, particularly when accompanied by septicemia. Although these alterations resolved after the lipid infusion was terminated, we suggest that prudent measures should be taken to gurantee adequate oxygenation during intravenous fat emulsion therapy in patients suffering from ARDS.

(Chest 1989; 95:1278-81)

Intravenous fat emulsion is frequently administered as metabolic fuel to individuals receiving TPN. We have previously reported that patients demonstrating mild respiratory insufficiency receiving SIMV to augment spontaneous breathing exhibited a significant increase in pulmonary venous admixture (Qva/Qt), reduction in arterial oxygen tension ([PaO.sub.2]) and increased MPAP during fat emulsion infusion.[1] These effects were amplified by the presence of sepsis. Although lipid therapy is commonly delivered to critically ill patients with impaired lung function, the hemodynamic and gas exchange consequences in individuals afflicted with ARDS have not been thoroughly explored. The current investigation was designed to assess the cardiopulmonary effects of intravenous lipid administration in patients suffering from ARDS and to evaluate the impact of concurrent septicemia.

MATERIALS AND METHODS

Nineteen patients aged 52[+ or -]8 years with ARDS were included in the investigation. Precipitating factors for admission to the critical care service included the following: gunshot wound to chest/abdomen (three), aspiration pneumonitis (two), peritonitis (eight), pancreatitis (three), blunt trauma to chest/abdomen (two), and GI bleeding (one). The ARDS was diagnosed when at least three of the following criteria were present: (1) history of an acute catastrophic illness; (2) refractory hypoxemia; (3) roentgenographic evidence of new diffuse infiltrates; (4) signs of reduced lung compliance. Septicemia was confirmed by the presence of positive blood culture with clinical and biochemical signs and/or hemodynamic hyperdynamia.

Spontaneous ventilation was supported with SIMV that was titrated to maintain an arterial [Pco.sub.2] conducive to preserving pH [is greater than or equal to]7.35<7.45. The [FIo.sub.2] supplementation and CPAP were regulated to achieve optimum pulmonary function and oxygen delivery as previously described.[2] No alterations in therapy that might have effected cardiorespiratory function were made during the study period. No patient was receiving pulmonary vasoactive or anti-inflammatory pharmaceuticals during the study. All patients received mini-dose heparin for prevention of deep vein thrombosis. For the assessment of cardiorespiratory function, each patient had a peripheral arterial and thermistor tipped PA catheter inserted. Data collection included the following: arterial and mixed venous pH/blood gas tensions, MPAP, PA diastolic pressure, WP, RAP, and CO. Vascular catheter lumens were hydraulically interfaced via continuous flushing devices to quartz transducers. Intravascular pressure catheters were flushed with heparinized normal saline solution (2 IU/Ml) at a rate of 3 ml/h. Thermodilution CO was calculated from the mean of three electronically integrated temperature decay curves generated following indicator injection. A 10-ml injectate of 5 percent dextrose in room temperature water was injected via the proximal lumen of the PA catheter at end-exhalation. Arterial and mixed venous blood pH/gas tensions were analyzed and temperature corrected. Hemoglobin concentration and oxyhemoglobin saturation were measured with a spectrophotometric oximeter. The Qva/Qt, SVR, and PVR were calculated via standard formulas. [TABULAR DATA OMITTED]

Decision to administer intralipid was determined individually predicated upon our approach to provide approximately 40 percent of the estimated caloric requirement (35 kcal/kg/day) as fat to patients receiving ventilatory support for ARDS. Investigational data were measured before the eight hour infusion of 500 ml 20 percent intralipid was started (before). One hundred grams of intralipid over alternating eight-hour periods is a manufacturer-suggested infusion rate to facilitate desired caloric intake for moderately stressed patients. Data collection was repeated immediately prior to (during) and three to four hours following completion of (after) intralipid infusion.

Statistical analysis was accomplished with an analysis of variance for repeated measures software program. When an F value indicated a difference, Scheffe's a posteriori multiple comparison method was employed to distinguish between means. Statistical significance was accepted at the 95 percent confidence level, and values are indicated as mean [+ or -] SD.

This investigation was approved by the hospital Institutional Review Committee and all appropriate recommendations for the Declaration of Helsinki were honored.

RESULTS

In 19 patients with ARDS, 500 ml of 20 percent intralipid infused over eight hours (3.0[+ or -].3 mg/kg/min) produced significant reduction in [PaO.sub.2]/[FIo.sub.2] and significantly increased Qva/Qt, MPAP, and PVR (Table 1). Further analysis of data revealed that during lipid infusion, nonseptic ARDS patients (n = 9) had reduced [PaO.sub.2]/[FIo.sub.2] and increased Qva/Qt and MPAP, but manifested no significant alternations in PADP-WP or PVR. After cessation of intravenous fat therapy, SVR increased about preinfusion level (Table 2). The ARDS patients with associated sepsis (n = 10) incurred a significant reduction in [PaO.sub.2]/[FIo.sub.2] and elevation in Qva/Qt, MPAP, PVR, and PADP-WP during lipid infusion (Table 3). With statistically similar lipid infusion rates, Qva/Qt increased to a greater extent in septic vs nonseptic ARDS patients; by 12.3 vs 7.3 percent, respectively. The magnitude of increased MPAP was not influenced by the presence of absence of septicemia. Although hyperlipemia may induce spectral distortion predisposing to an underestimation of oxyhemoglobin saturation, the reduction in calculated Qva/Qt was corraborated by concurently reduced [PaO.sub.2]/[FIo.sub.2] ratio coincident with statistically similar CO and mixed venous [Po.sub.2]. [TABULAR DATA OMITTED]

DISCUSSION

There are numerous advantages to providing nonprotein calories as lipid rather than glucose. Fat emulsion preparations are isotonic, and thus are appropriate for peripheral infusion. Intravenous emulsions of long-chain fatty acid triglycerides may prevent the development of essential fatty acid deficiency and may facilitate utilization of fat-soluble vitamins. Fat combustion is the preferred energy source in injured and septic patients despite intravenous administration of concentrated glucose solution.[3] Glucose infusion may induce hyperinsulinemia, which may reduce nitrogen retention when energy expenditure exceeds caloric intake.[4] However, with an isocaloric infusion of lipid versus carbohydrate, nitrogen retention is analogous while sustaining an appropriate glucose/insulin ratio.[5] Since carbohydrate-induced insulin hypersecretion is abnormal in critically ill patients, fat emulsion therapy contributes a substrate more likely to expedite a favorable nitrogen balance. Also, carbohydrate loading may significantly increase [CO.sub.2] production, which could compromise ventilatory competence in patients experiencing marginal function.[6]

When fat emulsion infusion is substituted for carbohydrate derived calories during TPN management, the complications of excess [CO.sub.2] production and hormone activation are blunted or obviated. However, studies in healthy adult volunteers and acutely ill patients have revealed lipid-induced changes in pulmonary function. Talbott and Frayser[7] demonstrated a reduction in [PaO.sub.2] and a 15 percent decline in pulmonary diffusing capacity for carbon monoxide (Dco) after a two-hour infusion of 15 percent Lipomul (37.5 g/h) in healthy subjects. Sundstrom et al[8] administered 20 percent intralipid for 20 minutes (approximately 56 g/h) to volunteers and documented a similar reduction in DCO that returned to control within 45 minutes after infusion was terminated. The investigators found no alteration in MPAP. Although these investigations employed lipid infusion rates substantially greater than those commonly used clinically. Greene et al,[9] also utilizing normal subjects, showed a reduction in DCO during rest and exercise during lipid infusion at 12.5 g/h that was correlated with a rise in serum triglyceride levels. In two individuals, heparin administration prevented both the rise in triglycerides and impairment of DCO. Kuo and Whereat[10] observed arterial desaturation following the infusion of 15 to 20 g of a 15 percent fat emulsion solution in patients afflicted with atherosclerotic heart disease. Recently, we evaluated the cardiopulmonary effects of fat emulsion infusion (12.5 g/h) in septic and nonseptic patients who were receiving some mechanical ventilatory support ostensibly to augment spontaneous minute ventilation rather than to support oxygenation.[1] Intralipid infusion precipitated a significant elevation of Qva/Qt and MPAP in both groups which returned to preinfusion levels coincident with the resolution of serum lipemia. Therefore, we were concerned that individuals with compromised lung function might experience further impairment during fat emulsion infusion. Additionally, we previously observed that during lipid infusion, septic versus nonseptic patients manifested a significantly greater increase in PVR and that serum lipemia persisted longer after discontinuation of infusion.[1] Thus, we subdivided the ARDS patients into septic and nonseptic groups to compare the impact that intravenous fat emulsion therapy might render upon hemodynamics and pulmonary gas exchange. We observed that septic patients demonstrated a significant increase in PVR and a larger alteration in Qva/Qt compared to those void of septic symptoms. The PVR in nonseptic patients was not changed, while SVR significantly increased after termination of lipid infusion. The explanation for observed postinfusion SVR elevation was not readily apparent.

As previously indicated, alterations in lung function have been generally attributed to triglyceride-associated reduction in DCO. However, experimental evidence derived from a long-term sheep[11] and short-term rabbit[12] model suggest that some of the diverse vasoactive intermediates and end-products of lipid metabolism are probably responsible for the observed alterations in pulmonary vascular tone and gas exchange. These alterations were not obviated by heparin pretreatment despite the accelerated clearance of serum lipemia.[11] Furthermore, both investigator groups were able to prevent the reduction in [PaO.sub.2] consequent to fat emulsion infusion with indomethacin contreatment. McKeen et al[11] also demonstrated that the 10 percent intralipid infusion (4.2 mg/kg/min) induced increase in MPAP, and protein-poor lung lymph flow was prevented by cyclo-oxygenase inhibition. The increased transvascular fluid filtration appeared to result exclusively from increased microvascular pressure in the exchanging vessels rather than via altered alveolar-capillary membrane permeability. The authors concluded that vasoconstricting arachidonic acid metabolites rather than hyperlipemia were the likely culprits responsible for the observed pulmonary hypertension, increased lung microvascular pressure, and arterial hypoxemia. In an apparent contradiction, Hageman et al[13] reported that 10 percent intralipid infusion for one hour (6.7 mg/kg/min) following experimental acute lung injury in chronically instrumented rabbits increased pulmonary production of vasodilating prostaglandins ([PGE.sub.2] and [PGI.sub.2]) and increased Qva/Qt. The authors speculated that the increased Qva/Qt resulted from a release of HPV due to a net vasolilatory effect prompted by the production of [PGE.sub.2] and [PGI.sub.2]. Since HVC is an autoregulatory mechanism that attempts to match regional ventilation to perfusion (VA/Q) by diverting blood flow from poorly to well ventilated lung areas, lipid infusion may negate physiologic compensation to improve arterial oxygenation in individuals with diseased lungs.

Skeie and colleagues[14] have suggested that the lipid infusion concentration, rate, duration and fatty acid chemical structure characteristics will determine the predominant metabolic pathway. Their hypothetical explanation for the conflicting data is that there may be a net increase in vasodilatory and anti-inflammatory prostaglandin metabolites during slow lipid infusion, whereas, consequent to bolus or rapid administration, fatty acid substrate may overwhelm the enzymatic pathways for [PGI.sub.2] and [PGE.sub.2] production, resulting in increased manufacturing of vasopressor and inflammatory prostaglandin metabolites (eg, thromboxane [A.sub.2]). With a similar 10 percent intralipid solution, McKeen et al[11] utilized an infusion rate 38 percent slower than Hageman et al[13] and observed increased rather than decreased pulmonary vascular tone. This conclusion disparity may be the result of species variability. Thus, further investigations would appear warranted.

However, if the conjecture regarding preferred pathways for fatty acid catabolism is accurate, then lipid-induced pathophysiology may be caused by either a rapid or slow infusion rate depending upon the underlying condition of the lungs. Predominant [PGE.sub.2] and [PGI.sub.2] production may significantly increase Qva/Qt and reduce arterial oxygenation when the compensatory HVC mechanism is active, but not under normal circumstances. An increased thromboxane level presumably could cause or exacerbate pulmonary inflammation in normal or injured lungs respectively.

In summary, we have shown that an eight-hour, 500 ml 20 percent intralipid infusion in patients with ARDS caused statistically significant elevation in MPAP and Qva/Qt, which is particularly evident with concurrent sepsis. Clinically significant alterations in gas exchange could be expected to manifest in patients with marginal arterial oxygenation, especially when adequate VA/Q is HVC dependent. These abberations are generally transient and return to control levels within three to four hours following suspension of lipid infusion. Despite mechanism, observed cardiopulmonary events during intralipid infusion may precipitate deleterious effects in patients with ARDS. Our results do not suggest abandonment of fat emulsion therapy in susceptible patients, but rather cognizance of the temporary alterations in gas exchange and hemodynamics that may ensue during a 3.0[+ or -].3 mg/kg/min lipid infusion. Whether these alterations are infusion-rate dependent, remains unclear. Adequate oxygenation should be verified for patients with compromised pulmonary function receiving intravenous fat, particularly when associated with septicemia.

REFERENCES

1 Venus B, Prager R, Patel CB, Sandoval E, Sloan P, Smith RA. Cardiopulmonary effects of intralipid infusion in critically ill patients. Crit Care Med 1988; 16:587-90

2 Venus B. Jacobs HK, Lim L. Treatment of the adult respiratory distress syndrome with continuous positive airway pressure. Chest 1979; 76:257-61

3 Askanazi J, Carpentier YA, Elwin DH. Influence of total parenteral nutrition on fuel utilization in injury and sepsis. Ann Surg 1980; 191:40-46

4 Nordenstrom J, Carpentier YA, Askanazi J, Robin A, Elwyn DH, Hensle TW, et al. Metabolic utilization of intravenous fat emulsion during total parenteral nutrition. Ann Surg 1982; 196:221-31

5 Jeejeebhoy KN, Anderson GH, Nakhooda AF, Greenberg GR, Sanderson I, Marliss EB. Metabolic studies in total parenteral nutrition with lipid in man. J Clin Invest 1976; 57:125-36

6 Askanazi J, Nordenstrom J, Rosenbaum SH, Elwyn DH, Hyman AI, Carpentier YA, et al. Nutrition for the patient with respiratory failure: Glucose vs fat. Anesthesiology 1981; 54:373-77

7 Talbott GD, Frayser R. Hyperlipidaemia: a cause of decreased oxygen saturation. Nature 1963; 200:684-86

8 Sundstrom G, Zauner CW, Arborelius M. Decrease in pulmonary diffusing capacity during lipid infusion in healthy men. J Appl Physiol 1973; 34:816-20

9 Greene HL, Hazlett D, Demaree R. Relationship between intralipid-induced hyperlipemia and pulmonary function. Am J Clin Nutr 1976; 29:127-35

10 Kuo PT, Whereat AF. Lipemia as a cause of arterial oxygen unsaturation, and the effect of its control in patients with atherosclerosis. Circulation 1957; 16:493-97

11 McKeen CR, Brigham KL, Bowers RE, Harris TR. Pulmonry vascular effects of fat emulsion infusion in unanesthetized sheep. Am Soc Clin Invest 1978; 57:1291-97

12 Inwood RJ, Gora P, Hunt CE. Indomethacin inhibition of intralipid-induced lung dysfunction. Prostaglandins Med 1981; 6:503-14

13 Hageman JR, McCulloch K, Gora P, Olsen EK, Pachman L, Hunt CE. Intralipid alterations in pulmonary prostaglandin metabolism and gas exchange. Crit Care Med 1983; 11:794-98

14 Skeie B, Askanazi J, Rothkope MM, Rosenbaum SH, Kvetan V, Thomashow B. Intravenous fat emulsions and lung function: a review. Crit Care Med 1988; 16:183-94

COPYRIGHT 1989 American College of Chest Physicians
COPYRIGHT 2004 Gale Group