0.3, and 0.4 gN* kg-' * [day]-') of nitrogen input on N balance, ... Supported by a grant from the National Health and Medical Research. Council and a project ...
What Rate of Infusion of Intravenous Nutrition Solution Is Required to Stimulate Uptake of Amino Acids by Peripheral Tissues in Depleted Patients?
PETER B. LODER, M.B. B.S.,* ROSS C. SMITH, M.D., F.R.A.C.S.,* ANTHONY J. KEE, B.Sc.(Hons),* STANLEY R. KOHLHARDT, M.B. B.S., B.Sc. (Med),* MALCOLM McD. FISHER, M.D., F.F.A.R.A.C.S.,t MICHAEL JONES, B.Sc.(Hons),* and THOMAS S. REEVE, F.R.A.C.S., F.A.C.S.*
We examined the effect of varying the quantities (0, 0.1, 0.2, 0.3, and 0.4 gN * kg-' * [day]-') of nitrogen input on N balance, 3-methylhistidine (3MH) excretion, plasma amino acid concentration, and the net flux of amino acids across the leg in depleted patients requiring parenteral nutrition. The calorie-to-nitrogen ratio was 140 to 1 (kcal:1 gN) and consequently the patients received varying amounts of calories (8, 14, 28, 42, and 56 kcal c kg-' [dayj-'). There was negative nitrogen balance and net loss of amino acids from the limb during fasting. An infusion of 0.2 gN * kg-' * [day]-' of IVN reversed the net catabolic process and resulted in equilibrium of peripheral total amino acid flux and of tyrosine flux without a decrease in 3MH excretion. Net uptake of total amino acids and tyrosine in peripheral tissues was achieved with 0.4 gN * kg-' * [day]-' and 56 kcal * kg-' * [dayj-'. This was associated with a fivefold increase in 3MH excretion (p < 0.01), indicating that net anabolism occurred with increased protein turnover. Fifty per cent of the amino acids taken up by peripheral tissues during infusions of 0.4 gN * kg-' * [day]-' was due to the uptake of glutamate (Glu) and 20% was due to the uptake of branched chain amino acids (BCAA). Plasma Glu concentration, [Glul, did not increase with increasing IVN infusion, but BCAA concentrations did. Although the mean plasma [Glul did not change with IVN infusion, there was an independent effect of plasma [Glul (p < 0.0001) and of N input (p < 0.0001) on Glu flux, indicating that even at high infusion rates the maximal capacity of peripheral tissues to take up Glu had not been reached.
From the Sydney University Department of Surgery, * the Intensive Therapy Unit,t and the Health Information Systems Departmentt at the Royal North Shore Hospital, Sydney, New South Wales, Australia
I T HAS BEEN ASSUMED that nutritionally depleted patients efficiently use intravenous nutrition (IVN) and obtain a state of anabolism.1"2 This is a desirable outcome of preoperative IVN and of IVN prescribed with the aim to replete patients with marasmus-like malnutrition. Muscle repletion is important for survival when the Supported by a grant from the National Health and Medical Research Council and a project grant from KabiVitrum International. Address reprint requests to Ross C. Smith, M.D., Department of Surgery, Royal North Shore Hospital, St. Leonards, 2065 Sydney, New South Wales, Australia. Accepted for publication February 27, 1989.
catabolic state has left the patient weak and immobile. Survival is particularly threatened when wasted intercostal muscles reduce the patient's ability to cough, which leads to sputum retention and pneumonia.3'4 The importance of this is emphasized by the finding that early death was most commonly due to respiratory complications in patients selected because of poor nutritional status and that these complications were only partly prevented by preoperative IVN, probably because the IVN did not completely reverse their malnourished state.5 Although intravenous nutrition can be shown to improve nitrogen balance,6'2 this has not always been associated with an improvement in muscle function7 or muscle chemistry.8 Deterioration of muscle chemistry has been shown to occur during preoperative intravenous nutrition delivering 43.5 kcal kg-' [day]-' and 0.15 gN * kg-' [day]-',9 but others'0 have demonstrated an increase in the size and number of type II muscle fibers and increased phosphofructokinase activity after 14 days of intravenous nutrition delivering 46 kcal * kg-' [day]-'. It may be necessary to deliver the greater amount of nutrition to obtain the desired effect of therapy. A net uptake of amino acids by peripheral tissues is essential if there is to be an increase in the muscle mass and supporting structures. There has been a number of studies on normal human subjects that have shown avid uptake of amino acids by peripheral tissues when supplied with a exogenous protein load.'1'-4 Three of these studies"" 2"'4 showed stimulation of total amino acid uptake after a short (2 to 4 hours) period of IVN infusion. -
Vol. 211 - No. 3
IVN INFJSION RATE FOR MUSCLE REPLETION
For many individual amino acids this uptake was related to their plasma levels. In a long-term study'3 of normal volunteers who underwent a 10-day fast followed by 10 days of parenteral nutrition, a net uptake of total amino acids was achieved with an infusion rate of 0.29 gN * kg-' * [day]-'. Again the forearm uptake was related to the input of individual amino acids. Few studies, however, have demonstrated uptake of amino acids during prolonged intravenous nutrition in patients undergoing intravenous nutrition for clinical indications. It is important to determine the effect of IVN infusion rate in the clinical setting in patients who require nutrition to replete peripheral tissues. This study examines the effect of different amounts of IVN on peripheral amino acid flux to determine the rate of IVN infusion necessary to produce uptake of amino acids into tissues. To reduce the effect of acute illness we have studied nonseptic nutritionally depleted preoperative patients. Methods Patients
Nine patients, who were admitted to a surgical ward for preoperative IVN because they were considered to have increased surgical risk due to malnutrition, gave informed consent to undergo repeated metabolic studies. Four patients had pancreatic pseudocysts, 3 had chronic pancreatitis, 1 had pseudo-obstruction ofthe small bowel, 1 had gastric outlet obstruction secondary to peptic ulceration, and 1 had gastric carcinoma that was resectable at a subsequent operation. No patients had a fever and no patients were considered, on clinical grounds, to be septic at the time of entry to the study. The patients were encouraged to walk around the ward, but no exercise program was undertaken. The Regional Ethics Committee approved the study. Investigative Protocol All measurements were performed on patients when they were in a clinically stable state. Patients entered the study after at least 48 hours of constant infusion of 5% dextrose/NaCl (about 8 kcal' kg-' [day]` and zero nitrogen intake). At this time flux of amino acids across the leg and urinary excretion of nitrogen (N) and 3-methylhistidine (3MH) were measured. A balanced glucose-based solution of amino acids (Vamin 9; KabiVitrum, Sweden, Table 1) was then infused at increasing rates corresponding to 0. 1, 0.2, 0.3, and 0.4 gN *kg`' [day]f and 14, 28, 42, and 56 kcal kg-' * [day]-'. After 48 hours of constant infusion at each level of nitrogen intake, a urine collection was commenced for measurement of 24-hour urinary N excretion and 3-methylhistidine excretion. The 24-hour -
TABLE 1. Amino Acid Composition of Vamin 9 (g- L')
Concentration Amino acid
Threonine (Thr) Serine (Ser) Glutamate (Glu) Glycine (Gly) Alanine (Ala) Valine (Val) /2 Cystine/Cysteine (Cys) Methionine (Met) Isoleucine (Ile) Leucine (Leu) Tyrosine (Tyr) Phenylalanine (Phe) Histidine (His) Lysine (Lys) Arginine (Arg)
3.0 7.5 9.0 2.1 3.0 4.3 1.4 1.9 3.9 5.3 0.5 5.5 2.4 3.9 3.3
Also contains 50 mmol sodium, 20 mmol potassium, 2.5 mmol calcium, 1.5 mmol magnesium, 50 mmol chloride, and 1.5 mmol sulfate.
urine samples were collected in acidified bottles. During the period that this urine collection was made, and before the patients got out of bed (7 to 8 A.M.), femoral blood flow measurements (mean of 10 determinations) were performed before collection ofarterial and femoral venous blood samples.
Analytical Procedures Nitrogen balance. Urinary nitrogen was measured by the macro-Kjeldahl technique. Administered nitrogen was calculated from the volume of IVN infused as determined by a volumetric infusion pump (Imed) and the nitrogen concentration of the IVN solution. Fecal losses were not measured, but in these patients bowel actions were infrequent. A correction factor of 3 gN per day has been added to the urinary nitrogen losses to adjust for fecal, skin, and other insensible losses.'5 Nitrogen balance was then calculated as N Balance = N Input - (Urinary N + 3), where N Balance, N Input, and Urinary N are in units of gN* [day]-'. Bloodflow measurements. Femoral arterial blood flow was determined by duplex ultrasound and pulsed Doppler in milliliters per minute.'6 In the femoral artery the withinpatient, within-day coefficient of variation was 10% (10 observations on each patient). For interpatient comparisons of metabolic flux, the flow was corrected for lean mass of the leg and expressed as mL* 100 g lean tissue- * minute-'. The estimate of lean leg volume is based on the assumption that the leg comprises two cylinders the dimensions of which are: thigh length times thigh muscle area and leg length times calf muscle area. Calf muscle area was calculated from calf circumference and calf skin fold thickness (CSF) by the formula Mid-calf muscle area = [(Calf circumference/r-CSF)/2]2-r, where calf circumference and calf skin fold are measured
LODER AND OTHERS
in centimeters. Mid-thigh muscle area was similarly calculated. To validate this method a comparison of blood flow measured by duplex ultrasound and by strain-gauge plethysmography was performed on six volunteers. There was a good correlation between the two methods. The relationship is described by the following regression equation: B.FLOW1 =-0.4 + 0.99 B.FLOW2, (r = 0.93, p < 0.001) where B.FLOW1 = Blood Flow by plethysmography in units of mL* 100 mL tissue-' * [minute]-' and B.FLOW2 = Blood Flow measured by duplex Deep Doppler and Ultrasound in units of mL* 100 g tissue-' * [minute]-'.
Amino acid analysis. Blood for amino acid analysis collected in ice-cold heparinized tubes. Plasma was separated by centrifugation and was deproteinized with an equal volume of ice-cold 6% sulfosalicylate. The acid soluble extract was filtered and stored at -80 C until analysis. A -mL aliquot of the 24-hour urine collection was taken for 3MH determination and was prepared for amino acid analysis in a fashion identical to that of the plasma samples. Amino acids were separated by means of high-pressure liquid chromatography using a cationic exchange resin and postcolumn ninhydrin derivatization. Norleucine was used as an internal standard in each sample. A known amount of norleucine was added to each sample (plasma and urine) before deproteinization and the unknown amino acids were corrected for the recovery of norleucine. The chromatographic peak heights of each individual amino acid were measured by an automated integrator. To test the analytical precision of amino acid analysis for determination of arteriovenous differences, paired aliquots of 10 plasma samples were processed and analyzed as indicated above. Each of these paired aliquots were treated as if they were paired arterial and venous samples. A summed within-pair coefficient of variation was calculated for each amino acid (Table 2). Flux calculation. Plasma amino acid flux was calculated as (arterial concentration-deep vein concentration) X blood flow x (1-hematocrit). A positive (+) flux represents net uptake across the limb whereas a negative (-) flux represents a net release of substrate from the limb. Reproducibility offlux method. In order to assess the precision of the hindlimb flux method multiple flux measurements were performed on a group of seven surgical patients. At the time ofthe measurements they were clinically stable and had been on a constant infusion of IVN for 48 h. The variability of the individual flux measurements for between and within patient studies are presented in Table 2. Nutritional assessment. Plasma albumin was measured was
Ann. Surg. * March 1990
TABLE 2. Precision (% CV) ofAmino Acid Analysis and the Hindlimb Flux Technique Precision for Individual Amino Acids Amino Acid Flux
Amino Acid Analysis
2.5 4.5 4.3 3.4 2.2 1.4 1.4 3.4 5.6 3.9 3.1 5.0 3.6 2.5 3.6 2.7 3.3 3.7 2.8
13.8 6.8 7.2 20.8 7.7 14.6 11.5 11.7 17.0 7.8 13.4 10.0 13.5 3.9 3.0 17.3 4.7 6.2 9.6
2.5 2.5 3.1 4.1 2.7 4.3 5.4 1.7 7.2 2.3 2.1 3.0 3.2 0.5 3.3 5.2 1.9 2.8 4.0
Tau Thr Ser Glu Gln
Gly Ala Val Cys Met Ile Leu Tyr Phe His Orn Lys Arg ABA
For Abbreviations of amino acids see Table 1; abbreviations not found in Table 1 include Tau, taurine; Gln, glutamine; Orn, ornithine; and ABA, alpha-aminobutyrate.
by an autoanalytical technique in our Hospital's biochemistry laboratory. Plasma transferrin and prealbumin were measured by radial-immunodiffusion using commercially available kits (Behring Instit. Sydney, Australia). Weight loss was calculated from difference between patient's admission weight and their recalled weight prior to the onset of the illness and expressed as a percentage. Mid arm triceps skinfold (TSF) was measured using a Harpenden caliper and mid arm circumference was measured using a plasticized tape measure. These measures were used to calculate arm muscle circumference (AMC). A commercially available device for the measurement of cell mediated immunity (CMI) was used to measure delayed skin hypersensitivity response to a batch of 7 antigens. The prognostic nutritional index (PNI) of Mullen's'7 group was calculated for the individual patients using the CMI score to categorize patients as being allergic, partially allergic or anergic.
Statistical Methods Repeated Measures Analysis of Variance (ANOV), using program 5V of the BMDP Statistical Software,18 was used to examine differences in individual amino acid concentrations and fluxes and nitrogen balance with IVN infusion rate. If this demonstrated significant variation with infusion rate, the baseline values were compared with
Vol. 211 No. 3 -
IVN INFUSION RATE FOR MUSCLE REPLETION
TABLE 3. Nutritional Parameters ofthe Patients Studied
Age (yrs.) Weight (kg) Weight Loss (%) TSF (mm) AMC (cm) Albumin (g/L) Transferrin (mg/dL) Prealbumin (mg/dL) CMI (mm) TLC (cc-') PNI (%)
60 52.9 18.4 14.7 20.0 32.3 190.0 18.6 8.4 1,400 48.2
9.4 15.5 15.3 8.3 4.2 4.7 76.0 9.8 2.3 179 20.6
TSF, triceps skin fold; AMC, arm muscle circumference; CMI, cellmediated immunity; TLC, total lymphocyte count; and PNI, the prognostic nutritional index of Mullen's group.'7
the four positive infusion rates using two-tailed Dunnett comparisons.1920 Dunnett comparisons were similarly used to compare the results of nitrogen balance, loglo(3MH) excretion, and total amino acid flux with their initial values. This procedure ensures that the experimental error rate does not exceed the nominated level. One sample Student's t test was also used to determine at which infusion rates nitrogen balance, total amino acid flux, and the individual amino acid fluxes differed from zero. Linear regression analysis, which included a patient effect due to repeated measurements, was used to determine the relationship between total amino acid flux, log10 (3MH) excretion, and nitrogen infusion, and to determine the relationship between individual amino acid concentrations and their fluxes. Multiple linear regression analysis was used to examine the independent effects of infusion rate and plasma glutamate concentration on glutamate flux. Results The nine patients' nutritional states are summarized in Table 3, demonstrating that they had lost weight but their plasma albumin, transferrin, prealbumin, and immunologic values were not significantly suppressed below normal. There were no significant complications of the parenteral nutrition therapy; in particular there were no septic or febrile episodes related to the central line.
Nitrogen Balance and Infusion Rate (Fig. 1) The fasting negative nitrogen balance (-6.0 ± 0.43 gN, mean ± sem, p < 0.00001) was almost completely reversed (- 1.0 ± 0.65 gN, NS) by the infusion of only 0.1 gN* kg- [day]-' and 14 kcal kg-' [day]-' (Dunnett comparison, p < 0.05). At each stepwise increase in infusion of IVN there was an increase in net nitrogen retention with a significantly positive nitrogen balance of 3.2 ± 0.65 gN * [day]-' (p < 0.02) from the IVN infusion,
363 giving 0.3 gN and 42 kcal. kg-' * [day]-'. Increasing the IVN input to 0.4 gN and 56 kcal kg-' * [day]f resulted in a further increase in nitrogen balance to 5.4 ± 1.9 gN- [day]-', (p < 0.05). -
Plasma Amino Acid Concentration and Infusion Rate (Table 4)
Plasma concentrations of the branched chain amino acids (isoleucine and valine) rose with increasing infusion rates, starting with levels in the low normal range and increasing to postprandial levels. The phenylalanine concentration also rose sharply with increasing infusion rates, although there was no significant change in tyrosine concentration. Threonine, serine, and methionine values rose above baseline levels with infusions of 0.3 and 0.4 gN * kg-' * [day]-'. The concentrations of the important gluconeogenic amino acid, glutamine, did not change significantly when tested by Repeated Measure ANOV. In contrast the levels of the other important gluconeogenic amino acid, alanine, rose with infusions of 0.3 and 0.4 gN. kg- * [day]-'; these values were significantly different from the fasting values (p < 0.01) when tested by Dunnett's multiple comparisons procedure. No amino acid concentration fell during IVN infusion. Several amino acids remained at basal concentrations despite being present in the IVN solution. Even at the high infusion rate of 0.4 gN * kg-' * [day]-', the concentrations of glutamate, glycine, '/2cystine/cysteine, leucine, tyrosine, histidine, lysine, and arginine were not significantly elevated. N Balance ( gN/day ) 8
6 4 2
0.3 0.4 N Input ( gN/kg/day )
FIG. 1. N balance at different IVN infusion rates (mean ± SEM). Calorie inputs associated with levels of N intake are 8, 14, 28, 42, and 56 kcal * kg- * [day]-', respectively. Mean values during IVN were significantly different from basal values by Repeated Measures Analysis of Variance (p < 0.00001) with means at IVN input giving 0.0, 0.3 and 0.4 gN kg- I [day]-' being significantly different from zero N balance (p < 0.00001, p < 0.02, and p < 0.05, respectively). -
LODER AND OTHERS
TABLE 4. Arterial Amino Acid Concentration During Different IVN Infusion Rates (Mean ± SD,
Nitrogen Input in gN * kg1AA
11± 104 ± 121 ± 114± 460 ± 291 ± 226 ± 184 ± 33± 22 ± 60 ± 108 ± 54± 64 ± 64 ± 87± 93 ± 74 ±
Ser* Glu Gln Gly Ala* Val* Cys Met Ile* Leu Tyr Phe* His Orn Lys Arg
0.1 9 57 51
36 159 124 110 53 28 12 23 34 18 32 32 88 59 52
9± 139 ± 150 ± 153± 470± 300± 272 ± 232 ± 26± 24 ± 81 ± 120 ± 49± 92 ± 82 ± 107± 83 ± 69 ±
4 56 54 105 106 109 87 58 17 5 24 36 14 45 43 106 49 30
14±15 184 ± 77 183 ± 74 148±29 455 ± 86 278 ± 72 296 ± 70 266 ± 59t 25± 16 24 ± 8 86 ± 18 135 ± 37 54± 18 127 ± 35* 77 ± 32 84±71 77 ±43 75 ± 31
0.3 8± 244 ± 208 ± 150± 528 ± 289 ± 357 ± 342 ± 39± 36 ± 110 ± 147 ± 60± 151 ± 69 ± 110± 85 ± 79 ±
116t 77t 76 134 85
130t 79t 25
13* 28t 34 15
32t 33 90 37 38
11± 257± 245± 150± 506± 275± 406 ± 402 ± 25 ± 39 ± 134 ± 152 ± 65 ± 207 ± 83 ± 118± 82 ± 70 ±
87t 76t 76 95 77
144* 104t 17
12t 47t 79 18
54t 20 114 35 26
The effect of differences in concentration were tested by "within-patient analysis of variance" with * indicating p < 0.001 and the absence of"*" indicating p > 0.05. Results of Dunnett's multiple comparisons procedure was used to
compare baseline amino acid concentrations with their concentrations after IVN are depicted by t if p < 0.05, t if p < 0.01. See Tables 1 and 2 for abbreviation definifions.
Total Amino Acid Flux (Fig. 2) The total amino acid flux value was calculated by summing the net flux results of the 20 individual amino acids. The total amino acid flux increased with increasing infusion rate. During the basal period there was net effilux of total amino acids from the leg (different from zero at p < 0.05). This was reversed to a net influx of total amino
acids at the 0.4 gN * kg-' from zero at p < 0.05).
Total Amino Acid Flux ( umol/DOOg/min ) 10 8 6 4
-2 -4 -6
0.4 0.3 N Input ( gN/kg/day )
FIG. 2. Total amino acid flux at different IVN infusion rates (mean + SEM). Calorie inputs associated with levels of N intake are 8, 14, 28, 42, and 56 kcal * kg- * [day]-', respecively. The regression equation describing the relationship is: T Flux = -2.65 + 21.6(N Input), (s = 4.65, RI = 0.46, p < 0.0003) where, T Flux = total of net amino acid fluxes in nmol 100 g tissue' * [minute]-' and N Input is in gN * kg-' [day]-'. -
[day]-' infusion rate (different
IVN Infusion and Flux of Individual Amino Acids During the basal state, the loss of total amino acids from the leg, as shown in Figure 2, is mainly due to the loss of threonine, alanine, leucine, glutamine, tyrosine, histidine, and arginine (Table 5). In contrast there was a net influx of glutamate into peripheral tissues, but this did not reach statistical significance. Infusion of 0.1 gN* kg-' * [day]-' of IVN abolished the net efflux of leucine, arginine, and histidine and produced a significant net influx of glutamate into the hindlimb. Increasing the infusion rate to 0.2 gN * kg-' * [day]' increased the uptake of glutamate and leucine by the hindlimb. During infusion of 0.3 gN* kg-' * [day]-' the only amino acid that showed significant net influx was glutamate. In contrast there was a significant net efflux of glutamine and alanine. At an IVN infusion rate of 0.4 gN * kg-' * [day]-' there was a significant net influx of isoleucine, tyrosine, and glutamate. In addition there were small gains of other amino acids, but these were not statistically significant. Fifty per cent of the net amino acid influx at 0.4 gN. kg`' [day]-' was accounted for by the flux of glutamate and 20% was due to the influx of branched chain amino acids (leucine, isoleucine, and valine) when the IVN solution was composed of 16% and 25% of these amino acids, respectively. Net flux of tyrosine, an indicator of anabolism or catabolism, was related to infusion rate (p < 0.001; Fig. 3).
VOl. 211.o NO. 3
IVN INFUSION RATE FOR MUSCLE REPLETION
TABLE 5. Amino Acid Hindlimb Fluxes (nmol/100 glminute) at Different IVN Infusion Rates, Mean ± SD
Nitrogen Input in gN * kg-
-19 ± 11 ± 0± 71 ± -112 ± -51 ± -133 ± -25± 4± -5 ± -4 ± 12 ± -8 ± -14 ± 18 ± 3± -4 ± -23 ±
0 ± 24 -31 ±3611 2±29 39 ± 59 -63 ± 7011 -41 ±45 -103 ± 73¶ -11±54 -13±26 -1 ± 11 -4 ± 10 -22 ±1911 -16 ± 131 -23 ± 55 -16 ± 1311 -3 ± 90 -25 ± 52 -27 ± 14#
Glu§,ll II Gln Gly
Val Cys Met* Ile§,§§ Leut Tyr§,|| || Phe His* Om Lys
50 68 32
10811 53 12411 41 16
611 13 72
101l 21 43 13 52 46
Repeated Measures ANOV results are indicated by: * if p < 0.05, t if p