This prospective randomised feasibility study was conducted at a large London teaching hospital in the adult ICU from May 2012 to December 2012.
Inclusion criteria were: age > 18 years; admission for trauma (head or multiple trauma); expected to be mechanically ventilated for > 48 h; within 72 h of ICU admission. Exclusion criteria were: non-trauma admissions; pregnant or lactating; contraindication to enteral feeding or a pre-existing condition where muscle wasting was exhibited, i.e. patients with congestive cardiac failure, neuromuscular disorders, end-stage renal failure, chronic obstructive pulmonary disease and liver cirrhosis. Trauma patients were selected as these patients are generally younger, fitter and usually present with fewer co-morbidities than other ICU cohorts, thereby representing a more homogenous cohort. The onset of injury can also usually be identified enabling baseline representation of muscle mass and the anticipated length of stay for head or multiple trauma is often significant.
An independent research nurse performed the randomisation of patients via computer-generated schedule to either standard care (control group) or the L-EAA supplement. A branched-chain amino acid ratio for valine : leucine : isoleucine of 1 : 5 : 1 was used based on a previous study where additional leucine was shown to promote muscle protein synthesis [
8]. In addition to standard enteral feeding, patients in the treatment group received a 5-g L-EAA powder dissolved in 100 mL of sterile water five times per day via the enteral feeding tube. This was continued for up to 14 days. Control patients received standard enteral feeding via ICU protocol. No placebo was used since at the time of study design the manufacturer could not supply an appropriate product. The study was not blinded as the researcher was also measuring the outcome.
Study outcome measures
The following measures of feasibility were collected: the number of patients recruited compared with those screened; compliance with administration of the L-EAA dose; and the number and reason for missed outcome measures (inflammatory markers, cytokines, nitrogen balance studies, muscle depth change, protein turnover and functional outcome measures).
Candidate outcome measures to determine rates of muscle wasting, protein turnover, strength and physical function were explored to examine their feasibility in a large RCT. These were muscle depth change on ultrasound, whole-body protein turnover using an intravenous (i.v.) infusion of 1–13C leucine, Katz and Barthel indices to assess functional change and the Medical Research Council (MRC) sum score to assess ICU-acquired weakness (ICU-AW).
C-reactive protein (CRP) (mg/L), albumin (g/L), interleukin (IL)-6 (pg/mL) and IL-10 (pg/mL) levels were measured serially on days 1, 3, 7 and 14 of the study. Cytokines were measured from plasma samples and analysed via enzyme-linked immunosorbent assay (ELISA) using ELISA grade streptavidin HRP and human IL-10 and IL-6 duosets obtained from R&D systems, Oxford, UK.
Muscle depth change (cm) was measured on alternate days throughout the ICU admission. The protocol has been described by Reid et al. [
2], but, in brief, includes taking muscle measurements from the bicep, forearm and thigh and adding the measurements together to obtain a total muscle depth (cm). Ultrasound measurements were undertaken using a Sonosite M Turbo™ ultrasound machine with a 5-MHz linear array transducer (Sonosite Ltd., Hitchin, Hertfordshire, UK).
Urinary urea (mmol/24 h) and urinary 3-methylhistidine (μmol/24 h) samples were collected on study days 1, 3, 7 and 14. 3-Methylhistidine (3-MH) i was used as a surrogate marker for skeletal muscle breakdown. An amino acid analyser with cation exchange chromatography (JEOL UK Ltd., Hertfordshire, UK) was used to analyse samples with ninhydrin detection and one inferred standard. Urine urea was measured by a kinetic urease using an Abbott Architect assay with Abbott reagents (Abbott, Maidenhead, UK).
Nitrogen balance (= nitrogen in – nitrogen out) (g/day) was calculated using both the British Dietetic Association’s Parenteral and Enteral Nutrition Group (PENG) recommended equation and the Deacon equation [
9] (see Additional file
1).
Katz and Barthel indices were used to assess activities of daily living data. The MRC sum score was used upon awakening and at discharge from ICU as a measure of strength and to determine presence of ICU-AW as recommended by International Guidelines [
10].
To measure protein turnover a primed intravenous infusion of 1–13C Leucine 1 mg/kg; 1 mg/kg/h (MassTrace, Somerville, MA, US) was administered for 3 h as previously described [
11] on days 1, 3, 7 and 14. Regular blood and breath samples were taken for the measurement of the enrichment of alpha-ketoisocaproic acid (α-KIC), the concentration of leucine and the enrichment of expired CO
2 at – 10 min, 0 min, 150 min (steady state), 160 min, 170 min and 180 min. Blood samples were taken from indwelling arterial catheters and breath samples were collected from the expiratory port of the ventilator (Servo 300, Siemens, Berlin, Germany or Servo I, Maquet, Sweden) via a Douglas bag with flow meter attached. Enteral feed was stopped 5 h before and during the 3 h of the protein turnover study (see Additional file
1 for protocol and laboratory analysis).
Continuous variables are shown as median (95% confidence interval) due to small sample size; categorical variables are shown as frequencies (%). Descriptive statistics were performed using SPSS V.20 (IBM Corp. 2011).