Encephalopathy, oxygen consumption, visceral amino acid clearance, and mortality in cirrhotic surgical patients

Encephalopathy, Oxygen Consumption, Visceral Amino Acid Clearance, and Mortality in Cirrhotic Surgical Patients

Massimo Loda, MD, Boston, Massachusetts George H. A. Clowes, Jr., MD, Boston, Massachusetts Angelo Nespoli, MD, Milan, Italy Luca Bigatello, MD, Milan, Italy Desmond H. Birkett, MD, Boston, Massachusetts James 0. Menzoian, MD, Boston, Massachusetts

Coma in cirrhotic surgical patients, usually associated with elevated amino acid concentrations in the blood is recognized as a significant indicator of high mortality [l-4]. However, in hepatic failure of this magnitude major metabolic defects exist in energy production and amino acid utilization which contribute to a number of deleterious physiologic alterations as well as to high blood levels of amino acids and coma [S-9]. Since all too frequently the causes of death are coagulopathy and bleeding, or overwhelming infection and multisystem failure [5,10], it is important to understand the relative significance of these metabolic abnormalities as well as coma itself in the establishment of this fatal outcome. High blood concentrations of aromatic amino acids and low levels of branched-chain amino acids result in the production of octopamine or other false neurotransmitters [11]. Mercaptans derived from methionine, short-chain fatty acids, and a variety of other agents associated with metabolic defects are found in the plasma or cerebrospinal fluid of comatose patients, and experimentally are capable of inducing encephalopathy [12]. On the other hand, in recent years the importance of accelerated visceral protein synthesis to survival from trauma and sepsis has become apparent [13]. Under these conditions, amino acids released at a high rate by muscle protein degradation are transferred to the liver and other From the Departments of Surgery, Harvard MedIcal School at the New England Deaconess Hospital, Boston University Medical School at Boston City Hospital. and University Hospital, Boston, Massachusetts; and the Ftrst Surgical Clime at Ospedale Policlmico, University of Milan, Milan, Italy. Requests for reprints should be addressed to George H A Clowes, Jr, M), Deac@ness/Hafvard Surgical Service. New England Deaconess Hospital, 194 Pilgrim Road, Boston. Massachusetts 02215 Presented at the 64th Annual Meeting of the New England Surgical Society, Sretton Woods, New Hampshire, September 30-October 2. 1983

542

visceral tissues not only for gluconeogenesis or oxidation, but far more importantly, for synthesis of proteins required for maintenance of immunocompetence, organ function, and survival [14-161. In addition to the supply of amino acids, the protein synthetic process depends on the presence of an appropriate stimulus, a source of energy, and functioning organelles in intact cells. Failure of this integrated response in patients with liver failure results in death just as it does in those with sepsis or trauma

B,91.

The present study of noninfected cirrhotic surgical patients who were classified, using the coma scale [17], as alert, encephalopathic, or comatose, was undertaken to assess the relationship of mortality and coma inducing factors to abnormalities of oxidative energy production and to defects of amino acid metabolism. Whole body oxygen consumption and the central plasma clearance rate (CPCR) of amino acids were measured. The CPCR of amino acids was developed as a means of measuring the rate at which amino acids are removed from the plasma by the liver, bone marrow, lymphoid tissue, and other central tissues [8,14]. Clinical determinations of CPCR of amino acids were previously found to correlate well with the in vitro rates of hepatic amino acid incorporation into protein [8,9]. The data obtained from this series of patients are presented to assess the concept that failure of energy production accompanied by inadequate utilization of amino acids are the important determinants of death in the cirrhotic patients, whereas encephalopathy and coma are manifestations by the central nervous system of metabolic defects in other organs. The clinical importance of understanding the abnormal metabolism in such patients lies in the pos-

The American Journal of Surgery

Cirrhotic Surgical Patients

sibility of employing metabolic measurements as accurate predictors of death and as guides to treatment for improving the outcome.

Material and Methods Fifty-nine cirrhotic patients were studied. Informed consent was obtained from each patient or from the person legally responsible for him or her before the study which was carried out with concurrence of the responsible physician. Of these people, 28 were admitted to the New England Deaconess Hospital, the Boston City Hospital, and University Hospital in Boston. Thirty-one entered the First Surgical Clinic of the University of Milan in Italy. Admission was for variceal bleeding in 49 patients, upper gastrointestional ulcer in 4 patients, and for other causes in 6 patients. Forty-one patients were male and 18 were female. The ranged in age from 23 to 77 years. Of these people, 71 percent were operated on. In all but eight, the diagnosis of cirrhosis was made histologically from operative or needle biopsies. Two patients were found to have biliary cirrhosis, three postnecrotic cirrhosis, one chronic, active hepatitis with cirrhosis, and the remainder Laennet’s cirrhosis. At the time of study, no patient was febrile or gave evidence that infection was present. Bleeding, if present, had been stopped. None of the patients were in shock, and blood volume had been replaced in each. Studies which included classification according to the degree of encephalopathy, hemodynamic evaluation, and metabolic measurements, were carried out preoperatively in 51 patients and postoperatively in 8. The composition and rate of infusion of amino acid solution, if any, was recorded. Staging of encephalopatby: Patients were assigned a level of coma (from 0 to 4) on the basis of psychologic, neurologic, and electroencephalographic findings as described by Ritt et al [17]. Patients were then subdivided into three groups: Group A, the alert group, had no encephalopathy (16 patients); Group E, the encephalopathic group, had coma levels 1 and 2 (19 patients); and Group C, the comatose group, had coma levels 3 and 4 (24 patients). Hemodynamic and oxygen consumption: Cardiac output and cardiac index were determined in all patients by thermodilution using a Swan-Ganz catheter. Complete hemodynamic evaluation, including central venous pressure and mean arterial pressure, was obtained in 31 patients in order to determine systemic vascular resistance. In the same patients, arterial and mixed venous blood gases were also measured by employing a 1302 Hemogas Analyzer@ (Instrument Laboratories, Watertown, MA) to obtain the arterial-mixed venous oxygen difference and whole body oxygen consumption. Amino acid measurements: Arterial blood was drawn into heparinized syringes from all patients. An amino acid analyzer (Beckman Instruments) was employed for determination of plasma amino acid concentrations [18]. In 28 patients, simultaneous femoral venous heparinized samples were also obtained for amino acid analysis in order to determine the arterial-femoral vein difference (AAFV). The CPCR of amino acids developed as an indicator of the rate at which plasma was cleared of amino acids by all central tissues, including the liver. Assuming a state of equilibrium at the time of measurement, the rate of amino

volume 147, April 1984

acid entry into the blood plasma (K1) is equal to the rate of extraction from it. In the absence of enteric feeding, the rate of amino acid entry into the plasma (Kl) equals total peripheral production (principally from muscle protein degradation) plus the rate of intravenous amino acid infusion. Since the muscle mass of the lower extremity is approximately a fifth of the skeletal muscle of the body, the value for K1 can be expressed by the following formula: K1 (~mol/min/m2) = 5 X leg production (rmol/min/m2) + infusion rate (~mol/min/m2), in which leg production = AA-FV (pmolfiiter) X plasma flow index (liter/min/mz). Leg blood flow, adjusted to leg plasma flow by each patient’s hematocrit value, was measured by indocyanine dye dilution [14] or estimated from cardiac index values. Previously in patients who were not in shock, measurements of leg blood flow by indocyanine dye dilution demonstrated leg flow to be 5.13 & 0.45 percent of the cardiac index [8,14]. The two values correlated significantly (1:= 0.72, p