Total parenteral nutrition with lipid

Total parenteral nutrition with lipid

Total Parenteral Nutrition with Lipid Carmen Paradis, MD, Montreal, Quebec, Canada Allen H. Spanier, MD, Montreal, Quebec, Canada Moira Calder, BSc, M...

1MB Sizes 0 Downloads 21 Views

Total Parenteral Nutrition with Lipid Carmen Paradis, MD, Montreal, Quebec, Canada Allen H. Spanier, MD, Montreal, Quebec, Canada Moira Calder, BSc, Montreal, Quebec, Canada Harry M. Shizgal, MD, FRCS(C), FACS, Montreal, Quebec, Canada

The patient subjected to catabolic stress and/or starvation must rely on endogenous fuels to meet his daily energy requirements. A prolonged catabolic state results in a significant loss of body protein which does not exist in storage form but is rather an integral component of some functioning unit. A loss of body protein is therefore a loss of body cell mass. Survival of the chronically stressed individual is often dependent on the provision of adequate nutritional support. The importance of total parenteral nutrition (TPN) in the management of this type of patient has been well established [I]. However, the administration of TPN continues to be associated with significant and occtisionally life-threatening complications. Sepsis, hyperglycemia, hypoglycemia, nonketotic hyperosmolar coma, and central venous thrombosis are all complications that have been related to the use of hypertonic dextrose solutions [Z]. The caloric requirements of the patient receiving TPN [3] necessitate the infusion of a solution with a high caloric density. To achieve this with dextrose requires the use of a hypertonic solution. A central venous catheter is therefore a necessity. In addition, because of the Maillard reaction, the amino acids cannot be added to the hypertonic dextrose solution prior to heat sterilization. Instead, the two must be mixed subsequent to sterilization under aseptic conditions, preferably by a pharmacist. Many of the problems associated with the infusions of a hypertonic dextrose solution are avoided with a soybean oil emulsion solution ([email protected], Pharmacia Canada Ltd, Montreal, Quebec, Canada). A 10 per cent Intralipid solution is isotonic, with a high caloric density of 1.1 kcal/ml. Thus, theoretically, a central

venous

catheter

is not required.

In

From the Department of Surgery, McGill University and Royal Victoria Hospital, Montreal, Quebec, Canada. This work was supported by a grant from the Medical Research Council of Canada. Reprint requests should be adbessed to Harry M. Shizgal, MD, Department of Surgery, Royal Victoria Hospital, 687 Pine Avenue West, Montreal, Quebec, Canada H3A 1Al. Presented at the Eighteenth Annual Meeting of the Society for Surgery of the Alimentary Tract, Toronto, Ontario, Canada, May 24-25, 1977.

164

addition, the Intralipid solution can be infused simultaneously via a Y connector (Figure 1) with a solution containing 7 per cent crystalline L-amino acids and 10 per cent fructose ([email protected], Pharmacia Canada Ltd), thus avoiding the need to have a pharmacist premix the solutions. The present study was performed to determine the efficacy of the 10 per cent Intralipid and Vamin combination for TPN. To quantitatively measure the efficacy of these solutions, body composition measurements were performed before and at regular time intervals after the onset of TPN. Material and Methods

Body composition was characterized with a multiple isotope dilution technic described previously [4], which involves the simultaneous injection of 8 &i of sodium 22 and 500 j&i of tritiated water. Blood samples were obtained at 4 and 24 hours. All of the urine excreted during the 24 hours after the isotope injection was collected. The concentration of the isotopes in both plasma and urine was determined by differential counting with a liquid scintillation detector. Total body water and total exchangeable sodium (Na,) were determined by standard isotope dilution [4]. Total exchangeable potassium (K,) was determined by an indirect technic developed in our laboratory [5]. As previously described, these parameters were employed to calculate the lean body mass, body fat; body cell mass, and extracellular mass [3]. Body composition studies were performed on forty-seven patients who were referred for TPN at the Royal Victoria Hospital. The majority of these patients were on the surgical services, and the indications for TPN are listed in Table I. The body composition measurements were performed at the beginning, at approximately two week intervals during and at the end of a course of TPN. A total of 108 studies were performed, permitting the evaluation of sixty-one intervals of TPN of 15.0 f 0.4 days in duration. All the patients received equal volumes of a 10 per cent Intralipid (10 per cent soybean oil emulsion, 1.2 per cent egg yolk phosphatide, and 2.5 per cent glycerol) and Vamin (7 per cent crystalline L-amino acids and 10 per cent fructose) solution. Since the Intralipid and Vamin solutions

The American Journal of Surgery

Total Parenteral Nutrition

TABLE I

lndlcations for TPN No. of Patients

Peritonitis

9

Anorexia Major abdominal operation Gastric outlet obstruction Gastrointestinal obstruction Pancreatitis Gastrointestinal fistula Inflammatory bowel disease Total

9 8 7 5 4 3 2 47

Results

One hundred

Figure 1. The 10 per cenf lipid emulsion solution and a solution containing 7 per cent L-amino acids and 10 per cent fructose were infused simutianeously at equal rates via a Y connector. The 10 per cent IntraNpld solution Is isotonic and therefore can be administered into a peripheral vein. However, in Me present study a central venous catheter was required in the majority of petients.

were administered simultaneously through a Y connector at equal rates, the patients received a solution containing 5 per cent Intralipid, 3.5 per cent L-amino acids, 5 per cent fructose, 0.6 per cent egg yolk phosphatide, and 1.25 per cent glycerol. In all cases the nursing staff was instructed to administer the combined TPN solution at a rate of 60 ml/kg/day. At this rate the amount of lipid infused would not exceed 3 gm/kglday, and the patients would be receiving 52 kcallkglday and 2.1 gm/kg/day of amino acids. All the solutions were administered by gravity drip, as pumps were unavailable. In addition, the nursing staff was instructed not to exceed the above rate and not to attempt to compensate with a higher infusion rate if the volume infused during the preceding time period was insufficient. As a result, there was a wide variation in the kcallkglday actually infused. Throughout this report mean values are presented with the standard error of the mean (SEM). The significance of the differences between the measurements taken before and after TPN was determined with a paired Student t test. When more than two groups were compared, an analysis of variance with a Scheffe test was employed [S]. The latter analysis was only performed on data which were first normalized for body size by expressing the parameter as a function of t&al body water, which is the best available independent parameter of body size [ 71.

Volume 135, February 1376

and eight body composition

studies

were performed before and after sixty-one periods of TPN of 15.0 f 0.4 days duration in forty-seven patients. The mean daily change in K, ( AK,/day), a measure of the mean daily change in the body cell mass, was correlated with the average kcal/kg/day infused. (Figure 2.) The resulting regression line was not statistically significant (p
165

Paradis et al

5% Intralipid, 5% Fructose, 3.5% L-Amino Acids y= -9.6

+0.16x

n=61 r=0.08 p
The mean AKJday was correlated with the kcalf kg/day in-

Figure 2. ,

.

l

:

.

l

.

.

(kcal/kg/day)

fused during sixty-one periods of TPN in forty-seven patients. The resultant regression line was not statistically significant. Patients with and without malnutrition are included.

.

-50

5% Intralipid, 5% Fructose, 3.5% L-Amino Acids 50

1

y=-30

8 + 0.69x

n=35 r=O 4

.

p
.

1

80 (kcal/kg/day) c

AK,/day=O kcallkg/day=44 95%

C

L =41.6

.

2 -46

. 9

-50

TABLE II

Figure 3. Correlation as in Figure 2, except only patients with malnutrition are included. Malnutrit~n was defined by a Na,/K, ratio of more than 1.22. The regression line, which is statistics//y significant, intercepts the abscissa at 44.2 kcalkgf day. The 95 per cent confidence limits of this intercept are 41.6 and 46.9 kcalfkgf day.

Effect of TPN on Body Composition in Patients without Malnutrition Normal Volunteers (n = 25)

Body weight (kg) Body fat (kg) TBW (I) Lean body mass (kg) Extracellular mass (kg) Body cell mass (kg) Na, (mEq) Nh/TBW (mEq/l) K, (mEq) K,/TBW (mEq/l) Na,/ K, Days on TPN kcallkglday AK,Jday

70.4 20.2 36.8 50.3 25.6 24.7 2,869 77.5 2,955 80.0 0.98

f f f f f f f f f f f

2.5 1.4 1.4 1.9 0.9 1.1 95 0.9 130 1.0 0.02

Group IA (n = 12) (<44 kcal/kg/day) Before TPN After TPN 67.4 21.4 33.6 46.0 23.8 22.2 2,593 77.5 2,660 78.7 0.99

f f f f f f f f f f f

4.5 67.3 4.1 20.8 1.4 33.9 1.9 46.5 0.7 24.5 1.3 22.0 106 2,578 1.8 76.2 161 2,637 1.9 77.1 0.03 1.01 14.1 f 0.4 34.6 f 2.3 -1.3 f 6.1

f f f f f f f f f f f

4.6 4.4 1.6 2.1 1.0 1.6 137 2.2 189 2.6 0.06

Group IB (n = 14) (>44 kcal/kg/day) Before TPN After TPN 57.3 11.5 33.5 45.9 24.2 21.7 2,556 76.6 2,602 77.5 1.00

f f f f f f f f f f f

2.9 58.3 1.9 11.9 2.0 33.9 2.8 46.4 1.5 26.1 1.4 20.3 153 2,738 1.2 81.2 167 2,437 1.6 71.1 0.03 1.17 15.6 f 1.0 51.7 f 1.5 -9.7 f 5.4

Note: All values represent the mean f standard error of the mean. TBW = total body water: Na, = total exchangeable exchangeable potassium. + Significantly different (p <0.05) from measurements taken before TPN by paired Student t test. + Significantly different (p <0.05) from normal by an analysis of variance and Scheffe’s test.

166

f f f f f f f f f f f

3.0 1.7 2.0 2.7 1.5’ 1.5 163” 1.9’ 182 2.6’+ 0.07*+

sodium; K, = total

The American Journal of Surgery

Total Parenteral Nutrition

(Figure 3.) To demonstrate the latter, each group of patients was further subdivided according to whether the mean daily infusion of calories was more or less than the intercept of the regression line, that is, more or less than 44.2 kcal/kg/day. The body composition data obtained before and after TPN in the patients without malnutrition are listed in Table II. There were twelve periods of TPN during which the patients received less than 44 kcal/kg/day (group IA). These patients received a mean of 34.6 f 2.3 kcallkgfday for 14.1 f 0.4 days. The body composition of this group was not significantly different from that recorded in twenty-five normal volunteers both before and after TPN. (Table II.) None of the means calculated after TPN were significantly different from those calculated before TPN by the Student paired t test. Group IB comprised those patients who received more than 44 kcal/kg/day. They received a mean of 51.7 f 1.5 kcal/kg/day for 15.6 f 1.0 days. At the onset of TPN their body composition was normal. However, at the end of the study period there was a slight deterioration of their body composition. The body cell mass decreased accompanied by an expansion of the extracellular mass, resulting in an elevation of the Nae/Ke ratio. (Table II.) However, in all cases the changes were small, and not all of the differences were statistically significant. The NaJIL value after TPN in this group was 1.17, which is still within the normal range. In twenty-five normal volunteers, the upper 95 per cent confidence limit for this ratio was 1.22 [5]. Although statistically significant, the changes observed after TPN are probably not biologically significant since the differences were small and the body composition after TPN remained

TABLE III

normal. Thus, in the patients without malnutrition, regardless of the kcal/kg/day infused, TPN maintained normal body composition. Table III lists the data of the patients whose body composition was characteristic of malnutrition at the start of TPN. Group IIA consisted of the patients who received less than 44 kcal/kg/day. There were eighteen periods of TPN in this group, lasting for a mean of 14.6 f 0.5 days, during which a mean of 37.8 kcallkglday was infused. The body composition of these patients was significantly (p <0.05) different from that of the normal group by an analysis of variance and the Scheffe test [6]. The body composition of this group was characteristic of malnutrition both before and after TPN; that is, the body cell mass as measured by K, was decreased whereas the extracellular mass as measured by Na, was expanded, resulting in an elevated Na,/K, ratio. After a period of TPN, there was a further deterioration in body composition. The mean K, decreased from 1,844 f 93 to 1,795 f 97 mEq and the NaJK, ratio increased from 1.57 f 0.07 to 1.64 f 0.07. However, none of the differences between the measurements taken before and after TPN were statistically significant by a paired Student t test. The group IIB data (Table III) comprised the seventeen periods of TPN which lasted for 15.6 f 0.7 days. These patients received an average of 50.9 f 1.4 kcal/kg/day, and all of these patients received more than 44 kcalikglday, that is, more than required to maintain the body cell mass. The body composition of this group was also significantly (p <0.05) different from that of the normal group and were characteristic of malnutrition at the start of TPN. However, TPN resulted in a slight improvement in body composi-

Effect of TPN on Body Composition in Patients with Malnutrition Normal Volunteers

Body weight (kg) Body fat (kg) TBW (I) Lean body mass (kg) Extracellular mass (kg) Body cell mass (kg) Na, (mEq) Na,/TBW (mEq/l) K, (mEq) K,/TBW (mEq/l) Na,/K, Days on TPN kcallkgiday AK,lday

70.4 20.2 36.8 50.3 25.6 24.7 2,869 77.5 2,955 80.0 0.98

f f f f f f f f f f f

2.5 1.4 1.4 1.9 0.9 1.1 95 0.9 130 1.0 0.02

Group HA (n = 18) (<44 kcal/kg/day) Before TPN After TPN 62.4 20.0 30.9 42.4 27.0 15.4 2,855 92.1 1,844 59.7 1.57

f f f f f f f f f f f

2.7 61.4 2.0 18.2 1.4 31.8 1.9 43.3 1.4 28.3 0.8 15.0 159 2,892 2.1’ 91.7 93 1,795 1.8” 57.3 0.07’ 1.64 14.6 f 0.5 37.8 f 0.9 -4.0 f 2.8

f f f f f f f f f f f

2.8 2.5 1.5 2.1 1.6 0.8 152 2.3’ 97 2.2’ 0.07’

Group IIB (n = 17) (>44 kcal/kglday) After TPN Before TPN 53.0 11.1 30.6 41.9 26.6 15.2 2,849 93.4 1,826 59.7 1.58

f f f f f f f f f f f

2.6 53.9 2.2 11.6 1.2 30.9 1.6 42.3 1.0 26.5 0.7 15.8 111 2,816 1.1’ 91.0 83 1,894 1.1’ 61.9 0.04’ 1.54 15.6 f 0.7 50.9 f 1.4 3.4 f 4.3

f f f f f f f f f f f

2.3 2.1 1.0 1.4 1.2 0.7 123 2.2’ 84 2.6’ 0.15’

Note: All values represent the mean f standard error of the mean. TBW = total body water; Na, = total exchangeable sodium; K, = total exchangeable potassium. Significantly different (p <0.05) from normal by an analysis of variance and Scheffe’s test. l

Volume 135, February 1978

167

Paradis et al

y” 60

< 44 CONTROL

NORMAL

>44 NUTRITION

<44

244

MALNUTRITION

Figure 4. The mean ( f SEM) body composition of twentyfive normal vokmteers and of the patients with and withouf malnutrition both beiore and after TPN. Each pafient group was further subdivided according to whether they received more than or less than 44.2 kcaU kg/day.

tion. The mean AK,/day was 3.4 f 4.3 mEq/day, which was more than that recorded in any of the other subgroups. (Table III.) However, this difference was not statistically significant. The body cell mass increased slightly, and the Na,/K, ratio decreased from 1.58 f 0.04 to 1.54 f 0.15. However, the differences between these measurements, taken before and after TPN, were not statistically significant, and the body composition of this group after TPN remained abnormal and consistent with malnutrition. Comments and Conclusions

Total parenteral nutrition in the chronically stressed and/or starved patient is intended to maintain a body cell mass that is normal and rebuild one that has become depleted. The simultaneous infusion of a 10 per cent Intralipid emulsion with Vamin solution accomplished the former but not the latter objective. The body composition of the patients without malnutrition remained normal during approximately two weeks of TPN, regardless of the amount of calories infused. (Figure 4.) Paradoxically, group IB patients, who received a larger amount of calories (51.7 f 1.6 kcal/kg/day for 15.6 f 1.0 days), experienced a slight deterioration of their body composition during this period. Although the small changes observed were in some cases statistically significant, they were probably not biologically significant, since the body composition of this subgroup remained within the normal range. The results with the Intralipid and Vamin combination were, however, poor in the patients with preexisting malnutrition. (Figure 4.) In these patients, the body cell mass was depleted at the onset

168

of TPN and remained unchanged after a course of TPN. This was expected for group IIA patients, because they received fewer calories than was required to maintain the body cell mass constant, that is, less than 44 kcal/kg/day. However, group IIB patients all received more than this amount of calories. Nevertheless, their body composition did not improve significantly with TPN. This poor result was predicted by the regression curve obtained by correlating AK,/day with the number of kcallkglday infused in the patients with malnutrition. (Figure 3.) Although the regression was statistically significant, the slope of the curve was not steep, indicating very little change in the body cell mass with the infusion of calories in excess of 44.2 kcalfkglday. The equation of the regression line predicts a AK,/day of 4.32 mEq/day with the infusion of 50.9 kcallkglday, which was the mean caloric infusion in group IIB patients. Over a period of 15.6 days, this would result in a mean increase in K, of 67.4 mEq, a positive nitrogen balance of 22.5 gm, an increase in body protein of 140 gm, and an increase in lean body mass of 0.56 kg. These predicted changes are similar to those experimentally determined. (Table III.) The solutions infused were, thus, unable to rebuild a depleted body cell mass. One possible explanation was the use of lipid as the major caloric source. With the solutions employed in the present study, lipid was providing 68 per cent of the nonprotein calories and 57 per cent of the total calories. Long et al [9] reported a decrease in nitrogen retention when lipid provided more than 20 per cent of the metabolic requirements. In contrast, Jeejeebhoy et al [IO] contended that lipid is an efficient source of calories for TPN. They reported positive nitrogen balance in patients receiving a TPN solution in which lipid provided 83 per cent of the nonprotein calories. However, in the same study, the patients receiving isocaloric and isonitrogenous infusions of hypertonic dextrose experienced a slightly greater nitrogen retention than did the patients receiving lipid. A second possible explanation for the poor results in the patients with malnutrition is simply that insufficient calories were infused. Group IIB patients received a mean of 50.9 f 1.4 kcal/kg/day, which is only slightly greater than the amount required for maintenance, that is, 44.2 kcallkglday. It was impossible to administer a larger amount of calories without infusing more than 3 gmlkglday of lipid. To exceed the latter limit would be unwise since Intralipid is presently approved for a maximum infusion rate of only 2 gmlkglday. Since the present study was designed to test the efficacy of the Vamin and Intralipid combination, the

The

American

Journal

of Surgery

Total Parenteral Nutrition

data do not indicate the cause of the poor results obtained in the patients with a depleted body cell mass. Additional data are required to determine whether the inability to rebuild a depleted body cell mass was related to the use of lipid as the principal caloric source. However, regardless of the explanation, it is apparent from the data that the Intralipid and Vamin combination is not an ideal solution for TPN. The amount of protein infused was not a limiting factor in preventing an increase in the size of the body cell mass. Anderson, Patel, and Jeejeebhoy [11] demonstrated nitrogen balance in patients receiving 55 k&kg/day when 0.4 gm/kg/day of L-amino acids were infused. With the Intralipid and Vamin combination, this level of protein intake is achieved with 10 kcal/kg/day. Thus, all of the patients in the present study received more than 0.4 gmlkglday of L-amino acids. Group IIB patients, with an infusion of 50.9 f 1.4 kcal/kg/day, received 2.0 gm/kg/day of L-amino acids. The inability of TPN to increase the body cell mass of these patients was consequently not due to insufficient protein intake. The patients were divided into two groups on the basis of the Na,/K, ratio at the onset of TPN. Malnutrition was defined by the presence of a Nae/K, ratio that exceeded 1.22. This approach is supported by the experimental data. Body composition was normal in both patient groups without malnutrition. (Table II.) In the patients with malnutrition, the body cell mass was decreased significantly, whereas the extracellular mass was significantly expanded (Table III), changes which are characteristic of malnutrition [4,7]. The data therefore support the use of the Na& ratio to characterize an individual’s nutritional status. The correlation between AK,/day and kcal/kg/day resulted in a statistically significant regression curve which intercepted the abscissa at 44.2 kcal/kg/day (95 per cent confidence limits’= 41.6 to 46.9). (Figure 3.) A similar result was obtained in a previous study in which AK,/day was correlated with kcal/kg/day infused in thirty-five seriously ill patients who received a variety of TPN solutions [3]. In the latter study, the regression line was also statistically significant and intercepted the abscissa at 45.5 kcall kg/day (95 per cent confidence limits = 41.2 to 49.8). Thus, the number of calories required to maintain the body cell mass was similar in the two studies. However, the slope of the curve with the Intralipid and Vamin combination was much smaller. Nitrogen balance under steady state conditions is a measure of the net effect of protein anabolism and catabolism and has therefore been widely used to

volume 135, February 1979

determine the efficacy of TPN. However, the experimental error of nitrogen balance measurements is large, because nitrogen balance is usually a small difference between a large intake, which tends to be overestimated, and a large output, which tends to be underestimated [12]. Another difficulty with nitrogen balance was demonstrated by the present study. Ideally, TPN will increase the size of a depleted body cell mass, that is, produce positive nitrogen balance. However, in the normal individual without preexisting malnutrition, the body cell mass will not increase-nitrogen balance will remain zero, regardless of the amount of TPN infused. The excessive administration of protein and calories to the normal individual will simply increase body fat without any change in the body cell mass. The interpretation of nitrogen balance data, therefore, requires knowledge of the preexisting state of nutrition. The importance of the latter is illustrated by the change that occurred in the correlation between AK,/day and kcal/kg/day when the patients without malnutrition were excluded. (Figures 2 and 3.) One of the major advantages of Intralipid is that it is isotonic. Administration by a peripheral vein is therefore possible. However, in our experience, the Vamin and Intralipid combination is irritating and causes thrombophlebitis unless the infusion site is changed frequently. Similar findings have been reported by others [10,13]. In addition, the majority of patients requiring TPN are seriously ill and have been hospitalized for a prolonged period. As a result, many of them no longer have peripheral veins suitable for TPN administration. Therefore, in the present study, a central venous catheter was required in almost all of the patients. Summary

The efficacy of TPN, with lipid as the major source of calories, was assessed with body composition measurements. A multiple isotope dilution technic was employed to measure body composition before, at two week intervals during, and at the completion of a course of TPN. All the patients received equal volumes of a 10 per cent Intralipid solution and a solution containing 7 per cent L-amino acids with 10 per cent fructose (Vamin). The mean daily change in exchangeable potassium ( AK,/day), a measure of the mean daily change in the body cell mass, was correlated with the number of kcal/kg/day infused to determine the efficacy of the solution. The correlation was not statistically significant. However, it became statistically significant (p <0.05) when the patients without malnutrition were excluded, as their

169

Paradiset al

body cell mass was not expected to increase regardless of the amount of calories and protein infused. The resultant regression line indicated that the body cell mass was maintained when 44.2 kcal/kg/day was infused. However, the slope of the curve indicated that the body cell mass did not increase significantly when calories in excess of this amount were infused. The Intralipid and Vamin combination successfully maintained a normal body cell mass but did not rebuild one that was depleted. In addition, the data emphasize the importance of knowledge of a patient’s preexisting nutritional status in evaluating the response to TPN. References 1. Dudrick SJ, Ruberg RL: Principles and practice of parenteral nutrition. Gasfroenferology 6 1: 901, 197 1. 2. Ryan JA: Complications of total parenteral nutrition, p 55. Total Parenteral Nutrition (Fischer JE, ed). Boston, Little Brown, 1976. 3. Spanier AH, Shizgal HM: Caloric requirements of the critically ill patient receiving intravenous hyperalimentation. Am J Surg 133: 99, 1977. 4. Shizgal HM, Spanier AH, Kurtz RS: Effect of parenteral nutrition on body composition in the critically ill patient. Am J Surg 131: 156, 1976. 5. Shizgal HM, Spanier AH, Humes J, Wood CD: The indirect measurement of total exchangeable potassium. Am J phvsiol 233: F253.1977. 6. Armitage P: Statistical Methods in Medical Research. New York, John Wiley and Sons, 1971. 7. Moore FD, Olesen KH, f&Murray, Parker HV. Ball MR. Boyden CM: The Body Cell Mass and its Supporting Environment. Body Composition in Health and Disease. Philadelphia, WB Saunders, 1963. 8. Spanier AH, Kurtz RS, Shibata HR. MacLean LD, Shizgal HM: Alterations in body composition following intestinal bypass for morbid obesity. Surgery80: 171, 1976. 9. Long JM, Wilmore DW, Masson AD Jr, Pruitt BA Jr: Fat carbohydrate interaction: nitrogen-sparing effect of varying caloric sources for total intravenous feeding. Surg forum 25: 61, 1974. 10. Jeejeebhoy KN, Anderson GH, Nakhooda AF, Greenberg GR, Sanderson I, Marliss EB: Metabolic studies in total parenteral nutrition with lipid in man. J C/in /west 57: 125, 1976. 11. Anderson GH, Pate1 DG, Jeejeebhoy KN: Design and evaluation by nitrogen balance and blood aminograms of an amino acid mixture for total parenteral nutrition of adults with gastrointestinal disease. J C/in /west 53: 904, 1974. 12. Vinnars E: Effect of intravenous amino acid administration on nitrogen retention. Stand J C/in Lab Invest 27:Suppl 117: 55, 1971. 13. Grotte G, Jacobson S, Wretlind A: Lipid emulsions and technique of peripheral administration in parenteral nutrition, p 346. Total Parenteral Nutrition (Fischer JE. ed). Boston, Little Brown, 1976.

Discussion Josef E. Fischer (Boston, MA): This study is another bit of evidence added onto a lot of evidence already reported which fails to demonstrate that intravenously administered fat, particularly in sick people, contributes to the economy of the lean body mass.

170

I still have some questions as to whether the technic (although it is very sophisticated) is really capable of differentiating some of the undoubtedly small changes that one is trying to measure. It would have been helpful if perhaps concomitant nitrogen balance studies had been carried out to give a little better idea of whether this definition of malnutrition changes as these people accumulate nitrogen. Perhaps you have some data that was not included. If you have some data on what happens to these people when a more standard type of repletion (probably amino acids in hypertonic dextrose) is utilized, it would probably be critical to know whether or not there were any differences in data on malnourished patients. In future studies I would urge nitrogen balance studies be done despite the fact that they do not differentiate between patients who are malnourished and those who are not. However, nitrogen balance studies verify the technic and its efficacy.

Joel B. Freeman (Iowa City, IA): Doctor Shizgal has been one of the few people in this country to use a parameter other than nitrogen balance as an index of nutritional

repletion. Those of us who use nitrogen balance in metabolic studies are aware of its many vagaries and appreciate that we are looking at only nitrogen in and nitrogen out rather than actual utilization. The current paper is an attempt to measure the ratio of total body sodium to total exchangeable potassium in a group of patients receiving TPN with the Intralipid system. During hypertonic TPN with 25 per cent dextrose and 5 per cent protein, a’nonprotein calorie to nitrogen ratio of 150~1 is used. This ratio was worked out some forty to fifty years ago. In this classic experiment, progressive increments of nonprotein (dextrose) calories are given to a patient who is receiving a constant amount of nitrogen. There is a progressive increase in the amount of nitrogen retained as the amount of dextrose is increased. On the other hand, in the Intralipid system, one normally mixes equal volumes of Intralipid and 5 per cent Amigen@ in 5 per cent glucose (although the present authors used fructose instead of dextrose). In an 8 hour period, a typical patient would receive 650 cal but only 3 gm of nitrogen. In 24 hours, the patient would receive 9 gm of nitrogen, which is considerably less than the amount. normally administered during TPN. The nonprotein calorie to nitrogen ratio (216:l) is adequate although higher than necessary. Why did the sodium to potassium ratios not reflect nutritional improvement during nutrition with the Intralipid system? I believe this can be answered by the fact that the patients were receiving inadequate amounts of nitrogen. I would very much like to see nitrogen balance data on these patients, since a study comparing this parameter to the sodium to potassium ratio has never been reported. Even if the authors do not have capabilities for determining Kjeldahl microdigestion, a rough approximation of nitrogen balance could be made by measuring daily 24 hour urinary urea concentration.

The American Journal of Surgery

Total Parenteral Nutrition

In the text it is stated that a normal individual will have zero nitrogen balance regardless of the amount of TPN infused. Many surgeons, including myself, would disagree with this statement. Although comparative studies were not done, the authors appropriately reviewed the current status of lipid versus carbohydrate calories. There are several good papers in this regard. In brief, the results are somewhat conflicting. Jeejeebhoy et al [IO] showed good results when more than 80 per cent of nonprotein calories were supplied by Intralipid. The current study is not strictly comparable, since Intralipid constituted just less than 60 per cent of total calories, and the patients were receiving fructose instead of dextrose. However, the data do suggest that this system is probably not as efficient as conventional central venous nutrition. As well as inadequate nitrogen intake with the peripheral Intralipid system, there is also the problem of finding veins in a sick, malnourished patient so as to infuse adequate nitrogen and calories. As pointed out in the text, there was considerable variation in the amounts of nutrients infused, and I suspect that this was due to difficulty in maintaining patent peripheral veins. Assurance of an eucaloric state is another advantage of the central venous route. We all appreciate that TPN exposes the patient to technical and infectious risks which are largely bypassed with the peripheral Intralipid system. If all of these patients received peripheral nutrition I suspect that there were many variations in the infusion rates which could explain the data variations alluded to in the text. Knowing the exact amount of daily nitrogen input would help in interpretation of the material. I should like to conclude by attempting to put the current role of Intralipid in perspective. The peripheral Intralipid system is clearly advantageous when the central venous route is contraindicated due to sepsis or when it is technically impossible to cannulate the subclavian vein. Intralipid in smaller amounts (2 to 4 units per week) is indicated in any patient being maintained on fat-free hyperalimentation solutions for more than three weeks to prevent essential fatty acid deficiency. Thirdly, the role of Intralipid mixed with dextrose and protein has potential, but this remains unproven. Would it not be superior to give the patient more physiologic diets consisting of 60 per cent dextrose and 40 per cent fat rather than using a high fat diet, as was used here and in Jeejeebhoy’s study [IO], or a high (90 per cent) carbohydrate diet, as used in conventional hyperalimentation? Gerald 0. Strauch (Stamford, CN): I support the use of intravenous fat, based on data developed at the Rhode Island Hospital in Providence and recently presented at the meeting of the Southern Surgical Association by Doctor William R. Thompson et al. These data indicate that positive nitrogen balance in critically ill surgical patients does indeed require caloric intake of more than 40 kcal/ kg/day and of more than 18 gm of amino nitrogen daily, and that the bulk of the caloric requirement can be met as well with intravenous fat as with carbohydrate. These data are

Volume 135, February IS75

supported by metabolic balance studies done for more than 350 patient days, and they look very strong. Our own clinical experience in Stamford seems to bear out these data. We are at present convinced that parenteral alimentation can be safely and successfully carried out utilizing intravenous fat as the chief caloric source with infusions through peripheral veins and without many of the mechanical nuisances and the metabolic and septic threats of central venous aliment&ion in many if not most patients whose gastrointestinal function precludes enteral nutrition for sustained periods of time. I suspect one of the problems we are dealing with in the present study is that the levels of nitrogen and caloric intake were not sufficient. Katherine Bury (Toronto, Ontario, Canada): I would like you to clarify the reason for administering TPN in that group of patients who had a normal body composition to start with. Was there some clinical indication for it or were they clinically malnourished, but had a normal body composition on your isotopic studies? Robert Zeppa (Miami, FL): Did you separate the patients who were malnourished and who were not the victims of some malady that was causing an increase in oxygen consumption from those who were just malnourished and recovering without this stimulus, such as burn patients and septic patients? Otherwise, the data become a little bit difficult to interpret. Carmen Paradis (closing): Doctors Fischer and Freeman, regarding the nitrogen balance studies, we used body composition studies because we found previously that our error in the nitrogen balance studies was much greater than with body composition studies. I do agree it might be worthwhile in the future to do the two studies concomitantly. Doctor Fischer, we do now have data comparing patients receiving these solutions to similar patients receiving high concentrations of dextrose, and we find those results considerably different. Malnourished patients receiving hypertonic dextrose did considerably better than the patients receiving Vamin and Intralipid, although the amount of nitrogen they received was considerably less. Doctor Strauch, one of the big problems we have had is maintaining peripheral sites. Our patients develop thrombophlebitis quite quickly, and since many of the patients have been very ill, they have very few sites to begin with. Doctor Bury, the patients who were put on hyperalimentation and who had a normal body cell mass were severely stressed patients with fistulas or were trauma patients who needed some sort of nutritional support but in whom the gastrointestinal tract could not be used. Doctor Zeppa, we did not separate our patients into the two groups. Sometimes it is impossible to tell when we are starting what sort of patient we are going to be dealing with.

171