A new method for the high-precision assessment of tumor changes in response to treatment.

Motivation: Imaging demonstrates that preclinical and human tumors are heterogeneous, i.e. a single tumor can exhibit multiple regions that behave differently during both development and also in response to treatment. The large variations observed in control group, tumors can obscure detection of significant therapeutic effects due to the ambiguity in attributing causes of change. This can hinder development of effective therapies due to limitations in experimental design rather than due to therapeutic failure. An improved method to model biological variation and heterogeneity in imaging signals is described. Specifically, linear Poisson modeling (LPM) evaluates changes in apparent diffusion co-efficient between baseline and 72 h after radiotherapy, in two xenograft models of colorectal cancer. The statistical significance of measured changes is compared to those attainable using a conventional t-test analysis on basic apparent diffusion co-efficient distribution parameters. Results: When LPMs were applied to treated tumors, the LPMs detected highly significant changes. The analyses were significant for all tumors, equating to a gain in power of 4-fold (i.e. equivalent to having a sample size 16 times larger), compared with the conventional approach. In contrast, highly significant changes are only detected at a cohort level using t-tests, restricting their potential use within personalized medicine and increasing the number of animals required during testing. Furthermore, LPM enabled the relative volumes of responding and non-responding tissue to be estimated for each xenograft model. Leave-one-out analysis of the treated xenografts provided quality control and identified potential outliers, raising confidence in LPM data at clinically relevant sample sizes. Availability and implementation: TINA Vision open source software is available from www.tina-vision.net. Supplementary information: Supplementary data are available at Bioinformatics online.

The effect of selective decontamination of the digestive tract on mortality in multiple trauma patients: a multicenter randomized controlled trial.

OBJECTIVE: Evaluation of selective decontamination of the digestive tract (SDD) on late mortality in ventilated trauma patients in an intensive care unit (ICU). METHODS: A multicenter, randomized controlled trial was undertaken in 401 trauma patients with Hospital Trauma Index-Injury Severity Score of 16 or higher. Patients were randomized to control (n=200) or SDD (n=201), using polymyxin E, tobramycin, and amphotericin B in throat and gut throughout ICU treatment combined with cefotaxime for 4 days. Primary endpoint was late mortality excluding early death from hemorrhage or craniocerebral injury. Secondary endpoints were infection and organ dysfunction. RESULTS: Mortality was 20.9% with SDD and 22.0% in controls. Overall late mortality was 15.3% (57/372) as 29 patients died from cerebral injury, 16 SDD and 13 control. The odds ratio (95% confidence intervals) of late mortality for SDD relative to control was 0.75 (0.40-1.37), corresponding to estimates of 13.4% SDD and 17.2% control. The overall infection rate was reduced in the test group (48.8% vs. 61.0%). SDD reduced lower airway infections (30.9% vs. 50.0%) and bloodstream infections due to aerobic Gram-negative bacilli (2.5% vs. 7.5%). No difference in organ dysfunction was found. CONCLUSION: This study demonstrates that SDD significantly reduces infection in multiple trauma, although this RCT in 401 patients was underpowered to detect a mortality benefit.

Estimation of errors in determining intrathoracic blood volume using thermal dilution in pigs with acute lung injury and haemorrhage.

BACKGROUND: Global end diastolic volume (GEDV) has a constant and predictable relationship to intrathoracic blood volume (ITBV). The present study assesses the difference between ITBV derived from GEDV and ITBV measured directly in pigs with acute lung injury (ALI) and mild haemorrhage. METHODS: We caused ALI in 12 anaesthetized pigs by i.v. injection of oleic acid and removed 10% of estimated blood volume. EVLW, GEDV, ITBV (COLD; Pulsion Medical Systems), Pa(o(2))/Fi(o(2)), lung compliance and haemodynamic variables were measured at baseline (time 0) and at 30 and 120 min. All animals were volume-resuscitated, followed by measurements at 180 min. A linear equation estimated from the 44 pairs of ITBV and GEDV values in 11 animals was applied iteratively to the four GEDV measurements in the 12th animal, enabling 48 comparisons between measured (ITBVm) and derived ITBV (ITBVd) to be made. RESULTS: Increase in extravascular lung water index (EVLWi) was associated with significant pulmonary hypertension, worsening of oxygenation and compliance (repeated measures ANOVA; P<0.05). There was good within-subject correlation and agreement between ITBV(m) and ITBV(d) (r=0.72, mean bias 0.8 ml; sd 32 ml). Mean error in deriving ITBV from GEDV was 4.5%. (sd 4.2%; range 0.05-19%). There were no significant differences in errors in the presence of small (up to 10%) deficits in blood volume (F=1.0; P=0.41). CONCLUSIONS: ITBV estimated by thermodilution alone is comparable to measurements made by the thermo-dye dilution technique in the presence of pulmonary hypertension and mild deficits in total blood volume.

Effects of sustained post-traumatic shock and initial fluid resuscitation on extravascular lung water content and pulmonary vascular pressures in a porcine model of shock.

BACKGROUND: The temporal evolution of lung injury following post-traumatic shock is poorly understood. In the present study we have tested the hypothesis that manifestations of pulmonary vascular dysfunction may be demonstrable within the first hour after the onset of shock. METHODS: Twenty-nine anaesthetized pigs (mean weight 27.4 kg; (SD) 3.2) were randomly allocated to three groups: control (C, n=9), shock resuscitated with either NaCl 0.9% (S, n=10), or 4% gelatine (G, n=10). Shock was maintained for 1 h followed by fluid resuscitation with either normal saline or 4% gelatine solution. Cardiac output (CO), mean arterial pressure (MAP), mixed venous saturation (Sv(O(2))), blood lactate concentration, mean pulmonary artery pressure (MPAP), MPAP/MAP, pulmonary vascular resistance (PVR), extravascular lung water index (EVLWi), Pa(O(2))/FI(O(2)), venous admixture (Q(.)(S)/Q(.)(T)), and dynamic lung compliance (C(dyn)) were measured at baseline, beginning of shock phase, end of shock phase, and post-resuscitation. RESULTS: At the end of volume resuscitation CO was restored to control values in both shock groups. MAP remained significantly below control values (95% CI: C=70-95, S=28-52, G=45-69 mm Hg) in both shock groups. MPAP/MAP was significantly greater in both shock groups at the end of the shock phase (95% CI; C=0.15-0.24, S=0.28-0.38, G=0.32-0.42) and at the post-resuscitation phase (95% CI: C=0.12-0.30, S=0.43-0.61, G=0.32-0.49) indicating the presence of relative pulmonary hypertension. This was associated with a significant increase in PVR in Group S (F=3.9; P<0.05). There were no significant changes in Pa(O(2))/FI(O(2)), Q(.)(S)/Q(.)(T), EVLWi, or C(dyn). In a small cohort of animals a measurable increase in EVLWi (>30%) and reduction in C(dyn) (>10%) were observed. CONCLUSIONS: Pulmonary vascular injury manifesting as relative pulmonary hypertension and increased PVR may occur within the first hour after the onset of shock. These changes may not be accompanied by overt changes in oxygenation, compliance, or EVLWi. Br J Anaesth 2003; 91: 224-32

Cardiovascular response to graded lower body negative pressure in young and elderly man.

Lower body negative pressure (LBNP) reduces central venous pressure (CVP) and cardiac output. The elderly are reported to have a limited capacity to increase cardiac output by increasing heart rate (HR), are especially dependent on end diastolic volume to maintain stroke volume and therefore should be especially vulnerable to LBNP. The present study compared the effects of LBNP in the young and old. Stroke volume was assessed non-invasively as stroke distance (SD) by aortovelography. Two groups of healthy male volunteers were studied: eight young (29.7 +/- 2.0 years, mean +/- S.E.M.) and nine old (70.1 +/- 0.9 years). LBNP was applied progressively at 17.5, 35 and 50 mmHg in 20 min steps, with measurements taken during each steady state. There were similar, significant, falls in CVP in both groups. SD fell significantly in both groups from respective control values of 24.8 +/- 1.6 and 16.6 +/- 0.9 cm to 12.5 +/- 1.3 and 8.9 +/- 0.4 cm at a LBNP of 50 mmHg. Although SD in the elderly was significantly lower than in the young, the LBNP-induced changes were not different between groups. Both groups produced similar significant increases in vascular resistance, HR, plasma vasopressin (AVP) and noradrenaline. Mean arterial blood pressure (MBP) and plasma adrenaline did not change significantly. Therefore healthy old men respond to LBNP in a similar manner to the young, although MBP and SD are regulated around different baselines in the two groups.

Tracheal gas insufflation apparatus. Performance of a novel phase-specific device.

Tracheal gas insufflation is a technique in which gas is injected intratracheally during positive pressure ventilation. The fresh gas rinses expired gas from the tracheal tube and anatomical dead space, aiding carbon dioxide elimination. This reduces ventilatory volume and pressure, helping to reduce ventilator-induced lung damage. Complications of tracheal gas insufflation include interference with ventilator function, tracheal damage and barotrauma. Expiratory washout is a variation of tracheal gas insufflation. We designed and constructed an original expiratory washout system and evaluated its safety and performance in lung and animal models. We found that expiratory limb and tracheal tube occlusion tests caused the device to disable itself at acceptable intratracheal pressures. We also demonstrated up to 31% reduction in tidal volume compared with conventional ventilation, supporting the possibility of using this device clinically to lessen volutrauma. We concluded that aspects of this design might alleviate many of the safety concerns of using tracheal gas insufflation.

Insulin resistance and substrate utilization in human endotoxemia.

Infection results in a state of insulin resistance, but the pathogenesis is poorly understood. Intravenous administration of bacterial lipopolysaccharide (LPS) has been used to mimic the febrile and systemic inflammatory responses to infection, but it is unclear whether LPS induces insulin resistance in man. To investigate the effects of LPS on insulin sensitivity and substrate utilization, we administered, in paired cross-over studies, either 20 U/kg Escherichia coli endotoxin or saline control to healthy volunteers (n = 6) 120 min after the start of a 10-h euglycemic hyperinsulinemic clamp (insulin infusion rate, 80 mU/m2 x min). LPS induced a fever, tachycardia, and mild arterial hypotension. Glucose utilization increased abruptly 120 min after LPS administration (+64.1+/-12.0%; P < 0.003), but then declined progressively, and insulin resistance was evident by 420 min (+1.9+/-3.5%; P < 0.05). The reduction in glucose utilization, like that observed in sepsis, was related to impaired nonoxidative glucose disposal and not abnormal glucose oxidation. The cortisol and GH responses to LPS were of sufficient duration and magnitude to explain the insulin resistance. LPS administration results in metabolic responses very similar to those observed in sepsis and could provide a useful model for the study of insulin resistance in human critical illness.

Monitoring the circulatory responses of shocked patients during fluid resuscitation in the emergency department.

OBJECTIVES: To assess the feasibility of constructing left ventricular response curves non-invasively during the fluid resuscitation of critically ill patients in the emergency department (ED) using a portable suprasternal Doppler ultrasound (PSSDU) device. DESIGN: Prospective case series. SETTING: Emergency department, Catholic University of Leuven, Belgium. PATIENTS: Shocked patients in the ED were diagnosed by predefined criteria. Only those thought to require standardised intravenous colloid challenges were observed i. e., sequential boluses of 3.5 ml/kg/10 min titrated against changes in stroke distance (Doppler surrogate for left ventricular stroke volume). RESULTS: A total of 50 shocked patients were studied. Stroke distance was measurable in 45 patients. 35 patients were fluid responders in terms of stroke distance. Group mean stroke distance increased during resuscitation (8.6 +/- 4.1 cm to 19.5 +/- 4.6 cm, P < 0.001) and then reached a plateau value (19.6 +/- 4.6 cm, P = 0.488). No response to fluid was seen in nine patients of which eight had severe sepsis. Alternative therapeutic approaches increased stroke distance for all of these patients. Evidence for right ventricular dysfunction was found as a cause for fluid non-response in the majority of patients with sepsis. CONCLUSIONS: Previous experimental work has shown that changes in central blood flow can be derived using the PSSDU device. This clinical feasibility study suggests that the PSSDU can help tailor haemodynamic therapy for an individual patient and give an early indication of treatment failure in the ED.

Blood volume determination by the carbon monoxide method using a new delivery system: accuracy in critically ill humans and precision in an animal model.

OBJECTIVE: To evaluate accuracy and repeatability of blood volume determinations made by the carbon monoxide method, using a ventilator-driven administration system. DESIGN: Prospective within-patient comparison, using simultaneous measurements by two methods to determine accuracy. Prospective laboratory investigation in animals to estimate repeatability. SUBJECTS: For accuracy: Nineteen ventilated critically ill patients in a university hospital intensive care unit. For repeatability: Six anesthetized, mechanically ventilated normovolemic pigs because this is impossible to perform in humans. INTERVENTIONS: In the accuracy study, a small mass of carbon monoxide was administered via a closed breathing system and arterial blood samples were taken from existing cannulas. In the repeatability study, an intramuscular sedative was given, followed by an inhalational anesthetic induction and mechanical ventilation via a tracheal tube. Left axillary artery and external jugular vein cannulas were sited. Anesthesia was maintained using an intravenous infusion. Five sequential circulating hemoglobin and blood volume estimations were made using the carbon monoxide method. MEASUREMENTS AND MAIN RESULTS: The small carboxyhemoglobin increase produced by uptake of a small, known mass of carbon monoxide was used to estimate the circulating blood volume. Simultaneous measurement, using 51Cr-labeled red blood cells, was performed. Twenty measurements were made in 19 patients. The bias (mean difference between blood volume measurements by the two methods) was 397 mL (5.53 mL x kg(-1)) +/-415 mL (+/-5.95 mL x kg(-1)); the limits of agreement (mean difference +/-2 SD) were -433 mL and 1227 mL (-6.36 mL x kg(-1) and 17.42 mL x kg(-1)). Therefore, 95% of expected differences will lie between these limits. The mean blood volume was 75.8 mL x kg(-1) in the animals. The coefficient of variation of repeated estimates was 9.49%. Mean circulating hemoglobin mass was 7.31 mmol with a coefficient of variation of 10.18%. The mean hemoglobin concentration, by co-oximetry, was 5.014 mmol x L(-1), coefficient of variation, 2.99%. CONCLUSION: This arrangement is a potential bedside method of estimating blood volume and circulating hemoglobin mass. We have rendered the technique more acceptable clinically by creating a ventilator-driven administration system.

Increased quantal release of acetylcholine at the neuromuscular junction following scald injury in the rat.

Following severe burns, patients frequently develop a profound resistance to nondepolarizing neuromuscular blockers. Several mechanisms have been proposed to account for this, including upregulation of nicotinic acetylcholine receptors. We investigated the effects of a 30% body surface area (BSA) scald on neuromuscular transmission in slow-twitch soleus (SOL) and fast-twitch extensor digitorum longus (EDL) of rats. Rats were sacrificed 72 h after the injury, a time at which sepsis is unlikely and body weight gain and core temperature have returned to normal. Further groups of rats were sham operated and either pair fed to the scalded rats or freely fed to assess the influence of food restriction. When compared with muscle from pair-fed control rats, scald resulted in an almost 50% increase in miniature endplate potential (mEPP) frequency in both SOL and EDL. However, scald did not increase mean mEPP amplitude in SOL, although it did cause a 10% increase in EDL. Scald injury did produce a significant increase in the size of the evoked endplate potential in SOL (33%) and EDL (37%). These data indicate that a significant increase in the quantal content of evoked transmitter released in SOL (38%) and EDL (30%) occurred by 72 h after scald. Such an increase may contribute to the resistance to nondepolarizing neuromuscular blockers documented in patients following thermal injury.

Modification of the cardiovascular response to hemorrhage by somatic afferent nerve stimulation with special reference to gut and skeletal muscle blood flow.

BACKGROUND: Tissue injury modifies heart rate and blood pressure responses to hemorrhage. The effect of concomitant injury on the hemorrhage-induced redistribution of cardiac output is much less clear. However, if injury elicits the visceral alerting response of the defense reaction, then a change in this redistribution of peripheral blood flow might be expected. If such a change compromised the gut circulation, then it might explain the deleterious effects of injury on the ability to withstand hemorrhage. METHODS: Immature pigs anesthetized with Saffan were bled 30% of blood volume with or without concomitant somatic afferent (brachial) nerve stimulation (to mimic injury). In addition to global cardiovascular and oxygen transport variables, blood flow was measured in the cranial mesenteric (gut) and right femoral (skeletal muscle) arteries after a 60-minute stabilization period after surgery, at the end of the 30-minute hemorrhage, and after a 30-minute shock period. RESULTS: Hemorrhage induced the expected cardiovascular and oxygen transport changes accompanied by a reduction in skeletal muscle blood flow and a 55% increase in skeletal muscle vascular resistance, but gut blood flow and vascular resistance were unchanged. However, in the presence of brachial nerve stimulation, the pattern of response to hemorrhage was modified, such that gut blood flow was now reduced and gut vascular resistance increased. CONCLUSION: The sparing of the gut circulation after hemorrhage was abolished in the presence of "injury." This finding is consistent with injury eliciting the defense reaction and may help explain the deleterious effects of injury on resistance to hypovolemia.

Selective impairment of glucose storage in human sepsis.

BACKGROUND: Glucose utilization in sepsis is impaired but the mechanisms are unclear. This study examined the effect of sepsis on total glucose utilization, oxidation and storage, and the energetic costs of these metabolic processes. METHODS: Glucose infusion rate (GIR), glucose oxidation rate (GOR), non-oxidative disposal rate and the energetic cost of glucose storage were studied in 24 patients with abdominal sepsis and in 26 healthy controls, using indirect calorimetry and the euglycaemic hyperinsulinaemic clamp with insulin infusion rates of 40 and 240 mU m-2 min-1. RESULTS: Basal GOR was significantly lower in septic patients than in controls (1.5 versus 2.3 mg per kg fat-free mass (FFM) per min, P < 0.001). Septic patients had a significantly lower GIR at 40 mU m-2 min-1 (4.2 versus 9.1 mg per kg FFM per min) and at 240 mU m-2 min-1 (7.5 versus 11.8 mg per kg FFM per min), relative to controls (P < 0.001). GOR was similar in septic and control subjects at both rates of insulin infusion whereas non-oxidative disposal was significantly lower in septic patients (P < 0.001) and accounted entirely for the reduction in GIR. The energetic cost of glucose disposal was unaffected by sepsis. CONCLUSION: Sepsis is associated with selective impairment of glucose storage but the energetic cost of non-oxidative disposal is unaffected.

Serum leptin concentrations and their relation to metabolic abnormalities in human sepsis.

Circulating leptin concentrations are raised in animal models of inflammation and sepsis. The purpose of this study was to determine the effect of sepsis on serum leptin concentration in humans and to examine the relationship between leptin and the metabolic consequences of sepsis. Resting energy expenditure, insulin sensitivity, and fasting serum leptin, plasma insulin, and cortisol concentrations were measured in 20 subjects with intra-abdominal sepsis and 20 healthy control subjects, before and during a 2-h period of euglycemic hyperinsulinemia. Fasting serum leptin concentrations were similar in septic and control subjects. In simple regression analysis, serum leptin concentrations correlated significantly with percent body fat in both septic patients (r = 0. 64, P < 0.005) and healthy subjects (r = 0.75, P < 0.0001). Multiple regression analyses additionally indicated that percent body fat, fasting plasma insulin, and plasma cortisol, but not sepsis, were significant and independent determinants of serum leptin concentration. No relationship between leptin and resting energy expenditure or insulin sensitivity was identifiable. A major metabolic role for leptin in human sepsis therefore appears unlikely.

Manipulating the metabolic response to injury.

In this short review we will concentrate on just one of the features of the metabolic response to injury (classified as accidental trauma, injury or sepsis) which are collectively known as the 'flow' phase. These include an increase in energy expenditure (hypermetabolism), changes in substrate utilisation (insulin resistance) and the focus of this chapter muscle wasting or catabolism. It is recognised that the three features are interrelated, for example insulin is believed to be an important factor in controlling amino acid flux in skeletal muscle and increasing environmental temperature which may reduce flow phase hypermetabolism has been shown to reduce postoperative nitrogen excretion (a marker of protein catabolism). However, we will concentrate on muscle wasting and refer the reader to other reviews on insulin resistance and metabolic rate.

Bradycardia and hypotension associated with severe hemorrhage are reversed by morphine given centrally or peripherally in anesthetized rats.

BACKGROUND: Severe simple hemorrhage (blood loss in the absence of tissue damage and nociception) leads to a reflex bradycardia and hypotension. Earlier studies showed that this reflex can be attenuated by prior administration of morphine. However, some patients may receive morphine, e.g., for analgesia after they have suffered severe hemorrhage. The aim of this study was to determine whether an established bradycardia and hypotension could be reversed by morphine. METHODS: Four groups of male Wistar rats (236-258 g) were anesthetized with alphadolone/alphaxalone (16-19 mg x hg x h(-1) intravenously). All groups received a hemorrhage of 40% total blood volume (BV) at 2% BV x min(-1). After the loss of 27% BV, bradycardia and hypotension were established equally in groups I and II and III and IV. Groups I (n=8) and III (n=10) received 0.9% saline (20 microL intracerebroventricularly or 1 mL x kg(-1) intravenously, respectively), whereas groups II (n=10) and IV (n=10) received morphine (10 microg intracerebroventricularly or 0.5 mg x kg(-1) intravenously, respectively). RESULTS: In groups I and III, heart rate and mean arterial blood pressure continued to fall, whereas the bradycardia was completely reversed and the hypotension partly reversed in groups II and IV after treatment with morphine. CONCLUSION: Morphine, administered centrally or peripherally, can reverse the bradycardia and markedly can attenuate the hypotension associated with severe hemorrhage. However, any benefit may be more apparent than real because other studies suggest that mortality may be increased.

Effects of colloidal resuscitation fluids on reticuloendothelial function and resistance to infection after hemorrhage.

The effects of three resuscitation fluids, hydroxyethyl starch (HES), Haemaccel, and fresh autologous blood, on reticuloendothelial system phagocytic and catabolic functions and resistance to infection after 40% hemorrhages in BALB/c mice were studied. The mice, anesthetized with isoflurane, were bled over a 10-min period, left hypovolemic for 30 min, and then resuscitated with their shed blood or the same volume of asanguineous fluid. Normothermia was maintained throughout the experiments. The uptake and catabolism of intravenously injected double-labelled sheep erythrocytes (51Cr-125I-SRBC) in liver and spleen were determined at 1 and 48 h after hemorrhage. No significant changes in the uptake or catabolism of SRBC in liver or spleen were found at 1 h after hemorrhage and resuscitation with any of the fluids. However, at 48 h a significant increase in liver uptake of SRBC was seen in animals resuscitated with either Haemaccel or HES compared to that in animals resuscitated with shed blood or in animals subjected to a sham operation. The increase in liver uptake was accompanied by a small decrease in spleen uptake in animals resuscitated with Haemaccel but not with HES. No great changes in catabolic activity were seen at 48 h, although activity levels tended to be higher in animals resuscitated with Haemaccel. Separate groups of animals were challenged by an intraperitoneal injection with live Escherichia coli at 1 or 48 h after hemorrhage and resuscitation. Sixty-four percent of the animals resuscitated with shed blood survived the challenge with E. coli at 1 h after hemorrhage, whereas only 10 and 0% survival was seen for animals resuscitated with Haemaccel and HES, respectively. At 48 h survival was 80% for shed-blood-resuscitated animals and 60 and 70% for Haemaccel- and HES-resuscitated animals, respectively.

Prediction of hemorrhagic blood loss with a genetic algorithm neural network.

There is no established method for accurately predicting how much blood loss has occurred during hemorrhage. In the present study, we examine whether a genetic algorithm neural network (GANN) can predict volume of hemorrhage in an experimental model in rats and we compare its accuracy to stepwise linear regression (SLR). Serial measurements of heart period; diastolic, systolic, and mean blood pressures; hemoglobin; pH; arterial PO2; arterial PCO2; bicarbonate; base deficit; and blood loss as percent of total estimated blood volume were made in 33 male Wistar rats during a stepwise hemorrhage. The GANN and SLR used a randomly assigned training set to predict actual volume of hemorrhage in a test set. Diastolic blood pressure, arterial PO2, and base deficit were selected by the GANN as the optimal predictors set. Root mean square error in prediction of estimated blood volume by GANN was significantly lower than by SLR (2.63%, SD 1.44, and 4.22%, SD 3.48, respectively; P < 0.001). A GANN can predict highly accurately and significantly better than SLR volume of hemorrhage without knowledge of prehemorrhage status, rate of blood loss, or trend in physiological variables.

Thermogenic, hormonal and metabolic effects of intravenous glucose infusion in human sepsis.

BACKGROUND: Sepsis is associated with alterations in glucose metabolism and the effect of intravenous feeding on energy expenditure is unclear. Many studies of glucose metabolism in humans with sepsis have employed techniques that are not relevant to the practice of intravenous feeding. METHODS: The thermogenic, hormonal and metabolic effects of glucose were evaluated in a prospective experimental study of septic (n = 6) and non-septic (n = 6) subjects, by administering glucose intravenously under conditions and at a rate similar to those used in total parenteral nutrition. RESULTS: Patients with sepsis had a higher fasting metabolic rate than control subjects (P < 0.001) and a lower fasting respiratory quotient (P < 0.03). The thermic effect of glucose in both groups was small and not statistically significant (median 3.2 and 0.4 per cent in septic and non-septic subjects respectively, P > 0.1). Patients with sepsis had an attenuated plasma insulin response to glucose administration compared with control subjects (P < 0.001) and less marked suppression of plasma fatty acid and glycerol concentrations (P < 0.001). Glucose administration was not associated with significant changes in plasma catecholamine concentrations in either group of patients. CONCLUSION: Intravenous infusion of glucose at clinically relevant rates is associated with a negligible thermogenic response, with no activation of the sympathetic nervous system.