Multiple systems organ failure: VII. Reduction in plasma branched-chain amino acids--correlations with liver failure and amino acid infusion.

Lower fractional concentrations of branched-chain amino acids were found in trauma-septic patients who did not survive than in those who survived (p values less than or equal to 0.046 to 0.001). A liver dysfunction scale was constructed on the basis of the levels of plasma bilirubin, albumin, SGOT, prothrombin time, and neurologic encephalopathy. Increased liver dysfunction was associated with reduced plasma fractional branched-chain amino acids for all branched-chain amino acids in both the surviving and nonsurviving patients except for valine in the nonsurviving group. This decrease was statistically significant (p less than or equal to 0.041 to 0.001) for leucine and isoleucine in the nonsurvivors and for valine in the survivors. The infusion of amino acids was associated with a decrease in the fractional concentrations in the nonsurvivors for leucine while the fractional concentrations of isoleucine in the nonsurvivors and isoleucine and valine in the survivors rose. The statistically significant changes (p less than 0.018 to 0.001) were for leucine in the nonsurvivors and isoleucine and valine in the survivors. When the liver dysfunction and amino acid infusion related changes are taken in;to account there were no significant differences in the fractional branched-chain concentrations between survivors and nonsurvivors. Liver dysfunction and low fractional branched-chain amino acids were linked in magnitude in a way that is consistent with the low branched-chain amino acids producing the liver dysfunction.

Multiple systems organ failure: V. Alterations in the plasma protein profile in septic trauma - effects of intravenous amino acids.

The correlations between the acute phase and nutritional plasma proteins and intravenous amino acid dosage have been explored in a group of 25 trauma septic patients of whom 14 survived. The two groups of patients appeared to have equal cardiopulmonary function and exogenous nutritional support. The surviving group showed significant associated changes (p less than or equal to 0.05) between alpha1 acid glycoprotein, alpha2 HS glycoprotein, and ceruloplasmin (acute-phase proteins) and between prealbumin, retinol-binding protein, and transferrin (nutritional proteins). There were no correlations in concentration changes between these two groups of plasma proteins. The surviving group showed significant positive correlations between the nutritional plasma protein and intravenous amino acid dosage (prealbumin, p less than or equal to 0.001; transferrin, p less than or equal to 0.008; retinol-binding protein, p less than or equal to 0.001; and albumin, p less than or equal to 0.004) but no correlations with the acute-phase proteins. The nonsurviving patients showed significant intercorrelations between the acute-phase and nutritional proteins that were not seen in the surviving patients, and showed no relationship between intravenous amino acid dosage and the plasma levels of nutritional proteins. The data are consistent with increased obligatory catabolism of amino acids in the nonsurviving patient which based upon the amino acid behavior documented in the first paper in this series probably involves the branched-chain amino acids.

Multiple systems organ failure: IV. Imbalances in plasma amino acids associated with exogenous albumin in the trauma-septic patient.

In a survey study of septic trauma patients, the response of plasma amino acid concentration to albumin infusion was contrasted in survivors (14 patients) and nonsurvivors (11 patients). Plasma albumin levels were maintained at 3 gm/dl by albumin infusion (0-128 gm/day) because of central venous pressure/adequate circulation considerations. Survivors showed no significant increase in plasma essential amino acid concentration as a function of albumin infusion. In nonsurvivors threonine, valine, leucine, phenylalanine, lysine, and histidine all rose significantly (p less than or equal to 0.025) with albumin infusion. Isoleucine (8 residues/molecule albumin), in contrast to leucine (60 residues/molecule) did not increase. As a result, the ratio of isoleucine to leucine (Ile/Leu) decreased with albumin infusion from 0.47 (no albumin infused) to 0.27 (60 gm albumin/day). Survivors did not exhibit a similar response. The low Ile/Leu increased in most nonsurvivors with amino acid infusion from 0.27 (no amino acids) to 0.59 (150 gm amino acids/day). The data strongly suggest that nonsurvivors had an increased rate of albumin catabolism with subsequent amino acid release. Moreover, hypoalbuminemia treated with albumin infusion without amino acid infusion appears to produce a relative isoleucine deficiency which may detrimentally affect protein synthesis.

Multiple systems organ failure: III Contrasts in plasma amino acid profiles in septic trauma patients who subsequently survive and do not survive-effects of intravenous amino acids.

The response of the plasma substrate and hormone profile of survivor and nonsurvivor septic trauma patients to varying rates of amino acid infusion (IVAA) were contrasted. When IVAA=0 levels of most plasma amino acids (except aspartate, tryptophan, cysteine, and proline) were lower in nonsurvivors. At IVAA=1 to 100, however, 11 of 20 plasma amino acids were significantly (p less than or equal to 0.05) higher in nonsurvivors: only glutamate was significantly lower (p less than or equal to 0.001) and valine, isoleucine, and arginine on average lower. At IVAA less than or equal to 101 to 200, only alanine, methionine, tyrosine, and phenylalanine were significantly (p less than or equal to 0.005) higher in nonsurvivors; isoleucine was significantly (p less than or equal to 0.02) lower. The sharp increase in methionine and decrease in tryptophan in nonsurvivors with IVAA was particularly marked. Polynomial regression analysis showed that urea increased significantly with IVAA in both patient groups, while free fatty acids and cortisol decreased only in nonsurvivors. Insulin increased with IVAA only in survivors, glucagon only in nonsurvivors. Triglycerides, glycerol, acetoacetate, beta OH butyrate, and glucose appeared to show no significant response to IVAA in either patient group. The data are consistent with increased peripheral protein catabolism and branched-chain amino acid oxidation in association with decreased tissue uptake of conventional energetic fuels. These results may be interpreted to be consistent with an impairment of mitochondrial translocase systems.

Multiple systems organ failure: II. The effect of infusion of amino acids and glucose.

Amino acids and dextrose infusion were given for short periods of time to a young man whose basal state is characterized in the previous paper in this series, and their effects were documented in terms of plasma concentrations and splanchnic extraction. The basal state measurements show in the acute trauma state and its subsequent starvation state a largely balanced splanchnic extraction of amino acids but at a decreasing rate. Amino acid (FreAmine) infusion at low rates on this background produced a large increase in extraction of a largely balanced mixture of amino acids but a minimal change in glucose release. The septic state is characterized in both the basal and amino acid infusion state by splanchnic extraction of an unbalanced mixture of amino acids which is deficient in branched-chain amino acids and in relative excess of glucogenic amino acids with increased glucose release and increased utilization of amino acids for gluconeogenesis. In early sepsis this state can largely be repaired by exogenous amino acid infusion but in late sepsis can only be partially repaired. The data suggest that the patient in late sepsis should have a branched-chain rich amino acid mixture and that the hepatic failure of sepsis is strongly associated with peripheral release of an unbalanced mixture of amino acids secondary to enhanced branched-chain catabolism. Infused glucose produces a large increase in the plasma glucose but also improves the balance of the splanchnic amino acids extracted. The statistical validity of the preceding statements are examined in detail in the manuscript.

Location of the medium chain fatty acid site on human serum albumin. Residues involved and relationship to the indole site.

Human serum albumin (HSA) is found to contain one primary binding site for medium chain fatty acids which is competitive with the binding of N-acetyl-L-tryptophan. The association constant is higher with nondefatted HSA than with HSA defatted by either heptane-acetic acid or charcoal extraction. Affinity labeling of HSA by N-bromoacetyl-n-decylamine (1:1 molar ratio) blocks the primary binding site of hexanoate and decanoate at an equivalency of the molar ratio of label incorporated. CNBr fragmentation reveals that the primary medium chain fatty acid binding site is positioned between Fragment C (residues 124 to 297) and A (residues 298 to 585). With nondefatted HSA, 60% of the label is associated with Fragment C and 35% with Fragment A, whereas upon defatting, this labeling pattern is reversed. The labeled residue in Fragment C is Lys 199, the same residue as acetylated by acetylsalicylate (Hawkins, D., Pinckard, R.N., Crawford, I.P., and Farr, R.S. (1969) J. Clin. Invest. 48, 536). Forty per cent of the label in Fragment A from nondefatted albumin is located in the A-Pro II subfragment (residues 447 to 548) and 60% in the A-Phe subfragment (residues 330 to 446). With defatted HSA, the A-Pro II subfragment contains 25% and the A-Phe subfragment contains 75% of the label. In all cases, labeling in the subfragments of Fragment A occurred at several different residues, indicating that this part of the binding site has considerable flexibility. The secondary binding sites for medium chain fatty acid are found to be the same general region as the primary binding site. N-Bromoacetyl-n-tetradecylamine (a long chain fatty acid-labeling agent) labeled only the A fragment of HSA.

Septic autocannibalism. A failure of exogenous nutritional support.

Forty-six patients with surgical sepsis were studied prospectively until death or survival to evaluate the effect of exogenous metabolic support on the observed plasma substrate levels and on the differential endogenous utilization of branch chain amino acids. There were no effects of administered glucose or colloid load. The administered amino acid load had little effect on substrate levels in patients who died; but significantly effected the observed levels of glycine, isoleucine, and methionine in patients who survived. Evidence is presented which suggests that fatal sepsis is associated with an increased release of endogenous valine and isoleucine into plasma, as well as increased plasma levels of tyrosine, proline, and methionine. These abnormalities are highly correlated with the increased levels of plasma alanine and occur at a time when the nonsurviving septic patient manifests a tendency toward reduced oxygen consumption and abnormal vascular tone relations--the septic B state. These data are consistent with the hypothesis that increased muscle protein catabolism is occurring with a differential utilization of branch chain amino acids and increased use of leucine and isoleucine and reduced use of valine. This autocannibalism of muscle mass appears to be the source of the increased plasma alanine and is little influenced by administered amino acid support in the absence of control of the septic process.

Proline metabolism in sepsis, cirrhosis and general surgery. The peripheral energy deficit.

Proline metabolism was prospectively evaluated in patients with surgical sepsis, cirrhosis, and elective surgical procedures. Significant correlations were found in the septic patients. Proline levels were an excellent indicator of mortality and correlated positively with lactate levels. Lactate and proline were inversely related to total peripheral resistance and oxygen consumption. In septic patients who expired: the metabolites involved in the hepatic pathways of proline degradation were elevated in proportion to proline; lactate, glutamate and proline were directly related to pyruvate; lactate/pyruvate ratios were constant; proline, glutamate, ammonia, ornithine, lactate and pyruvate levels were inversely proportional to oxygen consumption and total peripheral resistance. The primary defects in sepsis seem to be metabolic; there are very strong correlations in time between physiology and metabolism; the metabolic abnormality seems to be a progressive energy-fuel deficit, possibly from a progressive inhibition of substrate entry into the Krebs cycle.

Correlations between metabolic and cardiopulmonary measurements in patients after trauma, general surgery, and sepsis.

Simultaneous longitudinal hormonometabolic-physiologic (cardiopulmonary) profiles were measured in 14 nonseptic trauma/general surgery (T/GS) patients and in ten patients with Gram-negative abdominal surgical sepsis. The physiologic state classification system was used as the frame of reference. There were no response differences between the T and GS groups: they had A State responses. The sepsis (S) patients initially had exaggerated A State responses with significant changes in glucose, fat, amino acid, and glucagon plasma levels relative to T/GS. The S patients who survived (four) demonstrated profiles as in T/GS. The S patients who expired (six) progressively evolved an unbalanced, hyperdynamic B State response with progressive elevations of glucose, lactate, aromatic and branched-chain amino acids and glucagon, and low ketone bodies. There is definite correlation over time between metabolic and physiologic responses; the physiologic responses reflect the metabolic responses; the metabolic responses are consistent with a peripheral energy-fuel deficit.

Physiological and metabolic correlations in human sepsis. Invited commentary.

The septic response in man appears to be a disease in which the infecting agent induces a state of disordered metabolic control in the host. The abnormal regulation of metabolic pathways causes a diversion of substrate utilization toward gluconeogenesis and ketone body formation and a reduction in oxidative energy-producing metabolism. This state of metabolic insufficiency is reflected in the pattern of cardiorespiratory, vascular, and physiological compensation. The precise magnitude of the resultant physiological compensation and its rate and direction of change can be quantified by the use of physiological state trajectories which also reflect the magnitude of underlying metabolic derrangements. The prognostic and therapeutic implications of these changes are discussed.

Monoglyceryl acetoacetate: a ketone body-carbohydrate substrate for parenteral feeding of the rat.

The monoglyceride of acetoacetate was prepared from diketene and glycerol. The resulting mixture was composed of nearly equal amounts of 1- and 2-monoacetoacetin. This mixture was tested as a parenteral energy substrate by continuous intravenous infusion into the rat. This glyceride provided 71% of the daily energy for 7 consecutive days. Other groups were either fed ad libitum or fed ad libitum and supplemented with intravenous glucose isoenergetic to monoacetoacetin. All three groups had similar daily non-protein energy intake, and the two supplemented groups ate less protein than normal rats. All three groups gained weight similarly although the glucose group tended to gain fastest and the monoacetoacetin group tended to gain slowest. The rates were not significantly different. At the end of 7 days, the glucose group was hyperglycemic and the monoacetoacetin group was hyperketonemic compared to normal fed rats. The only significant differences among the livers was the small size found for the glucose group. Hepatic compositions were similar. It was concluded from these data that intravenous monoacetoacetin can support weight gain in rats and is a potential alternative to glucose as an energy source in parenteral nutrition.

The physiologic recovery trajectory as the organizing principle for the quantification of hormonometabolic adaptation to surgical stress and severe sepsis.

Preoperative and serial postoperative clinical, cardiovascular, physiologic, and metabolic studies were carried out on 86 patients undergoing coronary artery bypass surgery (CABG); and 48 patients undergoing abdominal general surgical procedures (GSEL). Multivariable statistical analysis of these data showed the patients to be in different physiologic states and to manifest several types of recovery trajectories that could not be discerned on clinical grounds alone. The CABG patients followed one of three types of cardiogenic recovery trajectories. In contrast, GSEL patients show a normal recovery trajectory different from all CABG types. When sepsis develops, and exaggerated stress response (A state) occurs, with increased oxygen consumption and a pattern of amino acids, fat, and glucose breakdown products, which is heightened but similar to the response of nonseptic GSEL patients. With progression of sepsis severity, an unbalanced hyperdynamic recovery trajectory (B state) develops in which a decrease in oxygen consumption is associated with increases in the aromatic amino acids tyrosine, tryptophane, and phenylalanine; and decreases in the branched-chain amino acids, leucine and isoleucine. Triglycerides rise as keto acids fall, but both lactate and pyruvate rise. Glucagon is persistently high, regardless of insulin levels. The quantifiably different physiologic recovery trajectories reflect altered hormone and metabolic states and imply different responses to therapy.

Plasma concentrations and tissue uptake of free amino acids in dogs in sepsis and starvation: effects of glucose infusion--some effects of low alimentation.

The plasma concentrations of substrates, together with transhepatic and transgut balances, have been studied in six control and eight septic awake fasted dogs. Four severely ill septic dogs (typically fluid in chest and/or abdomen, extensive peritonitis, respiratory difficulties) had high concentrations of threonine, glycine, tyrosine, lysine, histidine, tryptophan, and triglycerides (p less than or equal to 0.05). The other septic dogs (less severely ill) showed fewer and less pronounced alterations in the plasma substrates (aspartate and tryptophan were elevated, p less than or equal to 0.05). The infusion of glucose increased the concentration of glucose, lactate, and pyruvate and depressed the concentrations of most amino acids in both normal and septic dogs. Threonine, asparagine, glutamine, leucine, isoleucine, alpha-aminobutyrate, and tyrosine were significantly depressed in the severely ill septic dogs (p less than or equal to 0.05). In the normal dogs most amino acids were removed by the liver, with alanine accounting for approximately 40% of the total. Glutamine removal was negligible. In the septic dogs hepatic removal of amino acids was variable; livers of two severely ill septic dogs did not remove amino acids. In the control dogs glucose infusion (0.015--0.017 g/kg/min) tended to lower hepatic removal of amino acids. Hepatic dye removal in the septic dogs was always very poor. In the gut glutamine was removed and alanine, glutamate, glycine, and ammonia produced, but the overall sum of amino acid uptake was negligible in both the control and septic dogs. The ratio of tryptophan to the sum of valine, isoleucine, leucine, tyrosine, and phenylalanine concentrations was greatly elevated in all septic dogs in which it was measured. The free concentrations of amino acids in the liver, heart, and muscle tissues were grossly elevated in the low intravenous alimented septic state relative to the fasted normal state, whereas the tissue concentrative ability as measured by nonmetabolizable amino acids, alpha-aminoisobutyrate and cycloleucine, was not similarly increased. Sepsis clearly alters plasma and tissue concentrations, and in some instances hepatic uptake of amino acids.

Intravenous feeding of the rat with short chain fatty acid esters. I. Glycerol monobutyrate.

Intravenous nutrition was investigated using butyric acid because it is oxidized independent of carnitine transport into the cell mitochondria. This carnitine independent fatty acid, in the form of water soluble monobutyrin, was continuously infused into rats for seven days at 27 g monobutyrin per kilogram of body weight per day and provided half the daily energy requirement. All experimental animals survived the alimentation in good health and were free of detectable physiological and behavioral abnormalities. The intravenous infusion depressed the test animals spontaneous food intake to half their preinfusion level. These experimental rats still demonstrated continuous weight pain in contrast to weight loss by pair-fed controls. At decapitation, the monobutyrin infused rats had hepatic glycogen levels three times that of the controls, along with lower soluble hepatic protein, and normal lipid and water content. The plasma acetoacetate was also elevated in experimental rats. It was inferred from these results that monobutyrin was hydrolyzed, and the metabolites were oxidized by the rat. It is concluded from these observations that monobutyrin produces no obvious toxic affects during short infusion periods and provides calories for the rat when given intravenously.