Background and aims
Critical illness is characterized by several metabolic alterations, including an upregulation of catabolic pathways promoting endogenous energy substrate production. In contrast to starvation catabolism, this endogenous energy supply cannot be suppressed by feeding in the early acute phase of critical illness. This observation is one of the reasons that current guidelines recommend hypocaloric nutrition during the first week in ICU \[1\]. However, it is not known when this anabolic resistance subsides and a transition towards a normal response to feeding occurs.
The aims of this study are two-fold: 1) to investigate the temporal changes in non-suppressible endogenous energy production during critical illness, and 2) identify potential biomarkers indicating a normalized response to exogenous nutrients.
Protocol
For ICU patients, the protocol is first performed within 24-72 hours (early acute phase) from ICU admission. The protocol will be repeated if an enrolled patient is still in the ICU 120-168 hours (late acute phase) and 240-288 hours (late phase) after the first study session.
Infusions of intravenous (i.v.) glucose, enteral/parenteral nutrition (EN/PN) and insulin are stopped at 02:00 hours. Blood glucose concentrations will be monitored at least hourly during the fasting period. The protocol will be terminated in case of hypoglycaemia (\<4 mmol/L), and the patient will be excluded from the trial. The infusion rate of propofol will remain unchanged from 04:00 unless a change is clinically indicated, determined by the nurse or physician. Healthy controls will be asked to fast from midnight. In the control group, a peripheral i.v. line and a radial arterial line will be placed under local anaesthesia.
Baseline blood samples will be drawn from the arterial catheter, whereafter primed i.v. infusions of ring-2H5-phenylalanine, 2H2-glucose, and 2H5-glycerol will be started at 07:00. Measured resting energy expenditure (mREE) is then determined by 20-minute indirect calorimetry (IC) using the Q-NRG device (Cosmed, Rome, Italy) in ventilator or canopy mode. After 165 minutes, four blood samples will be drawn at 5-minute intervals from the arterial catheter to determine glucose, glycerol, and amino acid kinetics in a fasted state. Additional blood samples are drawn to analyze circulating hormones, cytokines, metabolites, and markers of autophagy.
PN (Olimel perifer N4E, Baxter) is then started at a rate that, including calories from propofol and other non-nutritional energy sources, corresponds to 100% of mREE. Repeated blood samples (same as above) are drawn at 45-60, 105-120 and 165-180 minutes after the start of PN. An additional IC is performed during the final 30 minutes of the PN. If hyperglycemia occurs during the PN administration, an insulin infusion will be started to keep blood glucose levels \<14 mmol/L.
After the final blood samples, all tracer infusions and PN are stopped. In ICU patients, nutritional therapy will be restarted as prescribed by the care team. In healthy controls, all catheters are removed.
Patients
Adult patients with an estimated ICU length of stay (LOS) \>72 hours will be recruited from the intensive care unit (ICU) at Karolinska University Hospital Huddinge, a tertiary university hospital with approximately 800 annual ICU admissions. Exclusion criteria include liver disease (liver transplant, acute or acute-on-chronic liver failure, cirrhosis), diabetes, pancreatic surgery or pancreatitis, pregnancy, intubation for airway protection only, mitochondrial disease, amino acid metabolism disorder, familial hypertriglyceridemia, severe hypertriglyceridaemia (≥ 10 mmol/L), hypoglycemia within the last 72 h, requiring ongoing large-volume resuscitation and/or blood transfusions, readmission within one week of ICU discharge, BMI ≥35, limitations to best supportive care or ongoing insulin/glucose treatment related to hyperkalemia.
Control group
Healthy, age- and sex-matched controls.
Statistical analysis and sample size considerations
Normal distribution will be assumed a priori. Primary and secondary outcomes will be analyzed using Student's T-test for paired or independent samples or one-way ANOVA as appropriate. The level of statistical significance is p ≤ 0.05. Correction for multiple comparisons will not be applied.
Based on previous data describing endogenous glucose production in a fasted and fed state in critically ill patients, 10 subjects in each group are required to detect a 20% difference in change in glucose rate of appearance with 80% power. No similar studies describing glycerol or amino acid kinetics are available for sample size calculations. Due to missing data, dropouts and inter-/intraparticipant variability, 10 patients with measurements of substrate kinetics in both the early acute and late acute phases are required to answer the primary and secondary outcome measures. To identify potential biomarkers of attenuated catabolism, 30 patients who complete the protocol in both the early acute and late acute phases will be included, i.e. two measurements per patient. Ten volunteers will be recruited to an age-matched control group.
