Does the addition of glutamine to enteral feeds ... - Semantic Scholar

Does the addition of glutamine to enteral feeds affect patient mortality?* Alison Saalwachter Schulman, MD; Kate F. Willcutts, MS, RD, CNSD; Jeffrey A. Claridge, MD; Heather L. Evans, MD, MS; Amy E. Radigan, RD, CNSD; Kelly B. O’Donnell, MS, RD, CNSD; Jeremy R. Camden, BA; Tae W. Chong, MD; Shannon T. McElearney, MD; Robert L. Smith, MD; Leo M. Gazoni, MD; Heidi-Marie A. Farinholt, MD; Cara C. Heuser, MD; Stuart M. Lowson, MD; Bruce D. Schirmer, MD, FACS; Jeffrey S. Young, MD, FACS; Robert G. Sawyer, MD, FACS

Objective: Studies have failed to consistently demonstrate improved survival in intensive care unit (ICU) patients receiving immune-modulating nutrient-enhanced enteral feeds when compared with standard enteral feeds. The objective was to study in a prospective fashion the effects of adding glutamine to standard or immune-modulated (supplemented with omega-3 fatty acids, ␤-carotene, and amino acids such as glutamine and arginine) tube feeds. Design: Prospective, unblinded study using sequential allocation. Setting: A university surgical trauma ICU. Patients: All surgical and trauma patients admitted to the surgical trauma ICU at a university hospital over a 3-yr period who were to receive enteral feeds (n ⴝ 185). Interventions: Sequential assignment to three isocaloric, isonitrogenous diets was performed as follows: standard 1-kcal/mL feeds with added protein (group 1), standard feeds with the addition of 20 – 40 g/day (0.6 g/kg/day) glutamine (group 2), or an

G

lutamine becomes an essential amino acid in the human body during times of stress including during critical illness and after major trauma or surgery. Glutamine contributes to many functions including immune function (1–9), gluconeogenesis/cellular energy supply (10, 11), maintenance of organ integrity/ function (3, 11), wound healing (6, 11),

*See also p. 2692. From the Department of Surgery (ASS, HLE, JRC, TWC, STM, RLS, LMG, HAF, CCH, BDS), Digestive Health Department (KFW, AER, KBD), Department of Anesthesiology (SML), and Departments of Surgery and Health Evaluation Sciences (JSY, RGS), University of Virginia Health System, Charlottesville, VA; and the Metrohealth System, Case Western Reserve, Cleveland, OH. The authors have no financial interests to disclose. Copyright © 2005 by the Society of Critical Care Medicine and Lippincott Williams & Wilkins DOI: 10.1097/01.CCM.0000185643.02676.D3

Crit Care Med 2005 Vol. 33, No. 11

immune-modulated formula with similar addition of glutamine (group 3). The goal for all patients was 25–30 kcal/kg/day and 2 g/kg/day protein. Measurements and Main Results: Patients were followed until discharge from the hospital. The primary end point was inhospital mortality, and multiple secondary end points were recorded. In-hospital mortality for group 1 was 6.3% (four of 64) vs. 16.9% (ten of 59, p ⴝ .09) for group 2 and 16.1% (ten of 62, p ⴝ .09) for group 3. After controlling for age and severity of illness, the difference in mortality between patients receiving standard tube feeds and all patients receiving glutamine was not significant (p < .11). There were no statistically significant differences between the groups for secondary end points. Conclusions: The addition of glutamine to standard enteral feeds or to an immunomodulatory formula did not improve outcomes. These findings suggest that enteral glutamine should not be routinely administered to patients with surgical critical illness. (Crit Care Med 2005; 33:2501–2506)

acid-base homeostasis (10, 11), and nitrogen transport (1, 10). Therefore, decreased glutamine availability is proposed to contribute to poor immune function, lack of proper organ integrity, and decreased energy availability, all of which further contribute to morbidity and mortality in seriously injured and critically ill patients. Because early nutrition support following major trauma, surgery, or critical illness is recommended (11, 12), many researchers have studied concomitant replacement of glutamine, either parenterally or enterally, assuming this could lead to improved patient outcomes. The results of these studies, however, remain inconclusive. Some studies have supported the use of enteral glutamine (2, 13, 14), primarily for reduction in infectious morbidity (15–18), whereas others have found no benefit or conflicting results (19 –22). Some authors warn of po-

tential harm (23), and animal studies have demonstrated an increased growth rate of bacteria when incubated in the presence of glutamine (24). The results of studies evaluating the role of parenteral glutamine show a trend toward improved outcomes (2– 4, 9, 14). This study was designed to further delineate the role of supplemental glutamine during enteral nutrition support, testing the hypothesis that glutamine supplementation of standard tube feeds would improve patient outcomes and the addition of supplemental glutamine to immunomodulatory tube feeds would provide additional benefit.

MATERIALS AND METHODS All adult surgical and trauma patients admitted to a university hospital surgical trauma intensive care unit (STICU) between January 1999 and January 2002 (with the exception of a 3-month period from July to October 2000

2501

Table 1. Enteral formula composition per 1000 mL of formula

% protein % carbohydrate % fat Glutamine, g Arginine, g Omega-3 fatty acids, g Beta carotene

Standard 1-kcal/mL

ImmuneModulated

25 45 30 9 0 0 0

25 36 39 17.5 8 5.4 Included

when glutamine was unavailable from the manufacturer) who were to receive enteral feeds were included in the study. Patients who were expected to tolerate oral feedings, be transferred out of the STICU, or die within 72 hrs of admission were not started on tube feedings and, therefore, were excluded from the study. Institutional review board approval was obtained before the start of the study. Due to the observational nature of the study and the use of standard therapies that were routinely being used on selected STICU patients at the time of study initiation, the need for informed consent was waived. Patients were assigned to one of three diet groups: Group 1: Standard 1-kcal/mL feeds (Replete; Nestle, Deerfield, IL) with added protein (Promod powder; Ross Laboratories, Columbus, OH) Group 2: Standard feeds (Replete) with the addition of 20 – 40 g/day glutamine (Cambridge Neutraceuticals, Cambridge, MA) Group 3: Immune-modulated formula (Crucial; Nestle) with similar addition of glutamine

Sequential rotating assignment was used to group the patients in the following fashion: The first patient was assigned to group 1, the second to group 2, the third to group 3, the fourth to group 1, and continuing in this pattern as each new patient was included in the study. Ideally we would have blinded the research team to the group designations, but this was not feasible given clinical resource availability. These study diets were received by the patients while in the STICU. The compositions of Replete and Crucial are depicted in Table 1. Promod is a protein powder from whey protein concentrate and soy lecithin. It contains 5 g of protein per scoop. The glutamine was powdered glutamine packaged in 10-g packets that also contained 5 g of maltodextran to prevent clumping. 2502

Energy requirements were calculated by the Surgery Nutrition Support Team using the estimated preadmission dry weight. Diets were isocaloric and isonitrogenous with daily goals of 25–30 kcal/kg and 2.0 g/kg protein, including 0.6 g/kg glutamine (for patients in groups 2 and 3). This was adjusted for patients with body weights ⬎120% of ideal body weight by adding 25% of the difference between the actual and ideal body weight to the ideal body weight. Patients were initially fed into the small intestine (in accordance with STICU protocol) via naso- or orointestinal tubes (Entriflex; Kendall, Mansfield, MA), and some were eventually converted to percutaneous endoscopic gastrostomy tubes. Tube feeding was started at 25 mL/hr and advanced, as tolerated, by 25 mL/hr every 8 hrs to goal. Powdered glutamine was mixed with 50 mL of water and flushed via the feeding tube in 10-g doses. Patients received 20 – 40 g of supplemental glutamine per day to bring their total glutamine to 0.6 g of glutamine per kilogram of body weight. Patients were followed until hospital discharge and categorized on treatment assignment regardless of actual amount of tube feeds received (intention-to-treat analysis). In-hospital mortality was the primary end point. Secondary end points included intensive care unit (ICU) length of stay (LOS) for survivors, hospital LOS for survivors, and number of days requiring mechanical ventilation and before tracheostomy for survivors. Other recorded variables included need for surgery or tracheostomy, episodes of diarrhea, need for total parenteral nutrition to meet energy goals, febrile episodes, and episodes of infection. Infections were defined as per Centers for Disease Control and Prevention (CDC) criteria (25). The number of hospital days before starting tube feeds, number of days to reach goal amounts of tube feeds, number of days the subjects received tube feeds, and average protein and calories received over the first week were also measured. Statistical Methods. A power calculation assuming a crude mortality of 30% for the control group (based on prior experience from this particular STICU) and 15% for either experimental group suggested the need for 120 patients per group to achieve an 80% power for an alpha of .05. Secondary to changes in personnel that would have necessitated significant rearrangement of duties to continue the study, data were analyzed by the investigators after the first approximately 60 patients per group. Based on the results, presented in this

article, the study was stopped as further resource utilization was difficult to justify given the unlikelihood of demonstrating a positive effect with additional patients. Categorical data were analyzed using the chi-square test. Any calculations with cell values less than five were analyzed using the Fisher’s exact test. Continuous data were summarized by mean and SEM and compared using the Student’s t-test assuming equal or unequal variance based on the F test for variance. These data were analyzed using SPSS, version 11.5 software (SPSS, Chicago, IL). Except as otherwise specified, values expressed are mean ⫾ SE or percentage of the group of origin. All p values are two-tailed. Values ⱕ.05 are considered statistically significant. Multivariate logistic regression analysis was performed to evaluate multiple predictor variable effects on mortality using S-Plus 2000, Release 3 software (MathSoft, Seattle, WA). Predictor variables were selected a priori based on expert opinion, prior data from this particular STICU (26), and the literature (27), all of which support age and severity of illness score as predictors of mortality. Two separate models were developed, the first using the Acute Physiology and Chronic Health Evaluation (APACHE) II score (28) to control for severity of illness and the second using the Injury Severity Score (ISS) (29) and thus restricting the analysis to trauma patients. This second model was used because of the high percentage of trauma patients included in the study (only nine patients were nontrauma admissions) and the previously established suboptimal predictive value of admission APACHE II score in this population. For the first model, APACHE II scores were missing for 11 subjects and were imputed using the overall mean APACHE II score recorded. Protein and calorie counts were included as cross-products with tube feed groups to assess their possible role as covariates in influencing mortality. Data were not available on protein and calories for ten patients, and another ten patients did not receive the intended tube feeds and thus values of zero were used for this analysis.

RESULTS A total of 185 patients, about 67% male and the majority trauma admissions, were included in the study. Patient characteristics were similar between group 1 (n ⫽ 64), group 2 (n ⫽ 59), and group 3 (n ⫽ 62) in regard to age, gender, admitting service, injuries (93% of trauma patients admitted to our institution have sustained blunt trauma), ISS, initial Glasgow Coma Scale score, percent with Glasgow Coma Scale score ⱕ7, and percent intubated on arrival at this medical center (Table 2). Group 3 had a higher APACHE II score compared with group 1. The median APACHE II score Crit Care Med 2005 Vol. 33, No. 11

Table 2. Group characteristics, recorded as n (%) where appropriate

n Male Trauma Mean age Mean ISS Mean APACHE II Mean GCS GCS ⱕ7 Intubated at arrival Severe head injury Major hepatic injury Required surgery Required abdominal surgery Required tracheostomy

Group 1

Group 2

Group 3

64 44 (68.8) 61 (95.3) 44.1 ⫾ 2.3 30.1 ⫾ 1.4 17.9 ⫾ 0.9a 10.2 ⫾ 0.9 15 (23.4) 14 (21.9) 18 (28.1) 7 (10.9) 55 (85.9) 19 (29.7) 35 (54.7)

59 41 (69.5) 55 (93.2) 48.7 ⫾ 2.6 27.3 ⫾ 1.2 18.6 ⫾ 0.9 9.9 ⫾ 0.9 13 (22.0) 13 (22.0) 16 (27.1) 2 (3.4) 43 (72.9) 13 (22.0) 27 (45.8)

62 40 (64.5) 59 (95.2) 48.4 ⫾ 2.5 28.5 ⫾ 1.4 20.6 ⫾ 0.9a 10.4 ⫾ 0.9 13 (21.0) 16 (25.8) 22 (35.5) 4 (6.5) 51 (82.3) 15 (24.2) 39 (62.9)

ISS, Injury Severity Score; APACHE, Acute Physiology and Chronic Health Evaluation; GCS, Glasgow Coma Scale. a p ⫽ .03. Table 3. Secondary endpoints, recorded as n (%) where appropriate

Mean no. of infections ⱖ1 infection ⱖ1 febrile episode(s) ⱖ1 episode(s) of diarrhea Required TPN

Group 1

Group 2

Group 3

1.8 ⫾ 0.3 38 (59.4) 54 (84.4) 26 (40.6) 10 (15.6)

1.5 ⫾ 0.3 38 (64.4) 44 (74.6) 27 (45.8) 8 (13.6)

1.7 ⫾ 0.3 43 (69.4) 48 (77.4) 30 (48.4) 12 (19.4)

TPN, total parenteral nutrition. p ⫽ NS for all values.

was 18. There were no differences between groups in the percent of patients who required surgery, abdominal surgery, and tracheostomy (Table 2) as well as thoracic, orthopedic, neurologic, or minor (i.e., feeding tube placement, wound debridement) surgical procedures (data not shown). In-hospital mortality rate for the group that received control tube feeds, group 1, was 6.3% (four of 64) vs. 16.9% (ten of 59, p ⫽ .09) for group 2 and 16.1% (ten of 62, p ⫽ .09) for group 3. Mortality rates among the two treatment groups (groups 2 and 3) were similar (p ⫽ .90). To assess the effect of glutamine on the outcome, the two treatment groups were combined and compared with the control group in a post hoc analysis. When this treatment group was compared with the control group, mortality rate was numerically higher in the treatment group, 16.5%, but this did not reach statistical significance (20 of 121, p ⫽ .06 comparing group 1 vs. groups 2 and 3). There were no statistically significant differences between the groups in mean number of infections, percent with one or more infections, or percent with fever Crit Care Med 2005 Vol. 33, No. 11

(Table 3). Indirectly measuring tube feed tolerance, there were no differences between the groups in percent with diarrhea and percent that required total parenteral nutrition to meet energy needs during the study (Table 3). Similarly, there were no statistically significant differences between the survivors in the three groups for STICU LOS, hospital LOS, days of mechanical ventilation, and days to tracheostomy (Table 4). In comparing the three study groups, some variables were best described for survivors only as the nonsurvivor data points were more likely to be outliers (e.g., one would generally expect LOS for nonsurvivors to be either very short or very long). Due to inherent concerns in performing an analysis without including all members of the study group, further analyses were performed, including survivors and nonsurvivors, all of which failed to demonstrate differences between the groups (data not shown). As expected in a predominantly trauma and, to a lesser extent, surgical ICU population, the majority of deaths in all groups involved sepsis/multiple system organ failure and neurologic devastation. There were statistically more deaths related to sepsis/multiple system

organ failure in group 2 compared with group 1 (seven of ten, 70% vs. one of four, 25%, p ⫽ .02). There were no differences between groups 1 or 2 when compared with group 3 (four of ten, 40%). A notable percentage of deaths in each group was at least in part attributable to a severe neurologic injury: two of four (50%) in group 1, two of ten (20%) in group 2, and six of ten (60%) in group 3 (p ⫽ not significant). Because interventions such as nutrition support are unlikely to affect mortality rate in patients with devastating neurologic injuries, the mortality analysis was repeated excluding those patients whose deaths were due in entirety or in part to a severe neurologic injury. In this analysis, there was a statistically higher mortality rate for group 2 when compared with group 1 (14.0% [eight of 57] for group 2 vs. 3.2% [two of 62] for group 1, p ⫽ .04). No differences were found between either group 1 or 2 and group 3 (7.1% [four of 56] mortality rate). There were no statistically significant differences between the groups in the percentages of patients for whom support was withdrawn or hospice care was initiated: three of four (75%) in group 1, four of ten (40%) in group 2, and eight of ten (80%) in group 3. Regarding nutritional variables (Table 5), the three study groups were similar in the number of days to starting tube feeds; the number of days to reach 50%, 75%, and 100% of calculated needs; and the total number of days of enteral feeding. However, protein received in the first week differed significantly between groups 2 and 3, and the mean number of calories received in the first week was also slightly different between groups. The first logistic regression model included the following predictor variables: tube feed group assignment (intent-totreat, control vs. either glutamine group), age, and APACHE II score. The model was found to perform well by area under the receiver operating curve analysis (C index ⫽ 0.74). Age was the only independent predictor of mortality (p ⫽ .01). However, the presence of glutamine in tube feeds demonstrated a weak association with mortality rate (p ⫽ .11), and this was similarly predicative as APACHE II score (p ⫽ .13). The second logistic regression model used ISS (excluding nontrauma patients) in place of APACHE II score with other variables unchanged. Again, the model had a good fit (C index ⫽ 0.72). As with the first model, increasing age was found 2503

Table 4. Secondary endpoints for survivors

STICU LOS, days Hospital LOS, days Days on ventilator Days to tracheostomy

Group 1 Survivors

Group 2 Survivors

Group 3 Survivors

15.2 ⫾ 2.1 25.6 ⫾ 1.9 13.2 ⫾ 1.3 4.4 ⫾ 0.7

16.7 ⫾ 1.9 24.1 ⫾ 2.5 10.9 ⫾ 1.3 4.0 ⫾ 0.7

16.1 ⫾ 1.6 24.1 ⫾ 1.5 13.6 ⫾ 1.3 4.7 ⫾ 0.7

STICU, surgical trauma intensive care unit; LOS, length of stay. p ⫽ NS for all values. Table 5. Measures of nutrition support received

D to start TF D to 50% D to 75% D to 100% Total days TF Mean protein, g/kga Mean calories, kcal/kga

Group 1

Group 2

Group 3

3.3 ⫾ 0.4 3.0 ⫾ 0.6 6.5 ⫾ 0.9 10.5 ⫾ 1.4 14.9 ⫾ 1.2 1.3 ⫾ 0.1 17.0 ⫾ 0.8c

2.5 ⫾ 0.2 2.8 ⫾ 0.5 5.7 ⫾ 0.9 8.8 ⫾ 1.3 13.2 ⫾ 1.2 1.2 ⫾ 0.1b 14.5 ⫾ 0.8c

3.2 ⫾ 0.3 2.9 ⫾ 0.5 6.4 ⫾ 0.9 8.8 ⫾ 1.0 14.8 ⫾ 1.1 1.4 ⫾ 0.1b 19.4 ⫾ 0.8c

D, days; TF, tube feeds. a Mean protein and calorie amounts represent the first week only; b p ⫽ .006 between groups 2 and c 3; p ⬍ .05 for each group compared with the other two groups.

to be an independent predictor of mortality (p ⫽ .04), and glutamine supplementation of tube feeds was almost significant (p ⫽ .08). ISS, like APACHE II, was not an independent predictor of mortality in this population (p ⫽ .13). Given the statistically significant differences between the three groups in protein and calories received in the univariate analysis, both of the previously detailed models were assessed for interactions between tube feed assignment and protein and calories delivered. These models were significantly overfit due to the large number of variables relative to the number of outcomes (deaths). Interaction between the tube feed group and protein and calorie counts was not identified (p ⫽ .9), and the presence of glutamine in tube feeds was still not a significant predictor of mortality (p ⫽ .07 using APACHE II scores, p ⫽ .33 using ISS).

DISCUSSION In this study, 185 trauma and general surgery ICU patients were sequentially assigned to one of three diets: standard tube feeds, standard tube feeds with supplemental glutamine, or immunomodulatory tube feeds with supplemental glutamine. Contrary to our initial hypothesis, we were unable to demonstrate improved outcomes with enteral glutamine supplementation. Our analysis 2504

instead showed no beneficial effect with a trend toward increased mortality. When the control group was compared with the groups receiving standard and immunomodulatory tube feedings supplemented with glutamine, the increased mortality in the groups receiving glutamine approached statistical significance (p ⫽ .09, group 1 compared with both group 2 and group 3). In multivariate analysis, age was the only independent predictor of mortality, although the presence of glutamine in tube feeds approached significance in the ISS model. There was a suggestion that the presence of glutamine in tube feeds might be more predicative of death than severity of injury score in both models, likely due to the relatively narrow range of ISS and APACHE II scores in the study population. Enteral glutamine supplementation had no significant effects on secondary outcomes. The depletion of glutamine stores in patients with critical illness, severe trauma, burns, or postoperative status is well recognized (6, 10, 20, 30). ICU patient mortality has been shown to be significantly higher for patients with low plasma glutamine levels (31). It therefore might seem logical that glutamine replacement could potentially help to maintain glutamine-dependent processes such as immune function (1, 2, 5–9), maintenance of organ integrity/function (3, 11), wound healing (6, 11), acid-base

homeostasis (10, 11), nitrogen transport (1, 10), cellular energy supply/gluconeogenesis (10, 11), and perhaps bowel integrity (32–35). The role of glutamine in immune function has been explored, and studies have concluded that glutamine supplementation can decrease incidence of infection (4, 15– 18), primarily pneumonia, bacteremia, and sepsis (15, 18). In a larger (n ⫽ 84), multiple-center trial, Conejero et al. (16) demonstrated a decrease in nosocomial pneumonia in patients randomized to receive enteral glutamine; however, they were unable to demonstrate an association between glutamine use and decreased incidence of other infections or mortality. Similarly, Jones et al. (13) were unable to show decreased mortality in critically ill patients receiving enteral glutamine. Published trials evaluating supplemental glutamine in enteral nutrition have enrolled between 35 and 84 patients. The current study includes 185 patients and should thus be better able to detect a true difference. Another large study examining a mixed group of 363 ICU (56% surgical, 23% trauma) patients randomized to enteral feedings with or without glutamine similarly failed to demonstrate a beneficial role of supplemental glutamine (19). Novak et al. (14) performed a meta-analysis of 14 randomized trials (751 patients) and demonstrated that the presence or lack of glutamine supplementation did not affect mortality in surgical patients (any mortality benefit noted in the study was observed in the critically ill population) and that, overall, although enteral glutamine supplementation had no effect on mortality rate, patients supplemented parenterally had improved mortality rate. This and other studies have supported the use of parenteral glutamine supplementation for decreasing mortality (2, 3, 14). In our post hoc subgroup mortality analysis that excluded all patients whose deaths were due at least in part to a devastating neurologic injury, we found a statistically higher mortality in group 2 compared with group 1, which could suggest a harmful effect of glutamine. This analysis failed to demonstrate, however, a difference between groups 1 and 3, which suggests that there was no difference between the standard and immune-modulated enteral feeds. In fact, the largest study to date (n ⫽ 597) on immunemodulated feeds compared with standard feeds also showed no effect on mortality rate (36). It is, however, impossible to differentiate between the effects of supplemental glutamine and the possible inCrit Care Med 2005 Vol. 33, No. 11

E

nteral glutamine supplementation should not be

used routinely in the trauma and surgical critically ill population (outside of a study setting), as we have shown no clear benefit.

teractions between additional glutamine and the other immunonutrients in the formula. There is no way to determine which, if any, of these nutrients are affecting patient outcomes. At this point we can only speculate on why enteral glutamine supplementation may not have a positive effect on STICU patient mortality and could possibly have a negative effect. Glutamine is crucial for many important functions, including homeostasis and energy supply, and the proposed benefit of glutamine for human cells may also extend to bacteria. Thus, when glutamine is administered enterally, multiple pathogens residing in the gastrointestinal tract have access to it and its benefits. The growth rate of mixed luminal bacteria, for example, was found to increase when exposed to high levels of glutamine (24). This theory may also explain why parenteral glutamine supplementation has been shown more definitively to have a positive effect on mortality rate as the glutamine is not being introduced directly to the bacteria in the gut flora. Additional theories might include alteration of bowel flora causing imbalances in absorption and function or, in the case of this study, glutamine-related exacerbation of hepatic or renal disease as no patients were excluded based on underlying disease. This is unlikely because the majority of the enrolled patients were admitted for trauma, and there is a relatively low occurrence of these comorbidities in this population. Although these results demonstrate an important trend away from the current belief that supplemental enteral glutamine is beneficial for ICU patients, some important weaknesses of this study must be noted. First, the study was not Crit Care Med 2005 Vol. 33, No. 11

blinded due to the necessity of the nursing staff to mix and administer the supplemental glutamine and the need for the nutrition support team to monitor protein and calorie intake daily for each patient. Second, the study was not formally randomized. The likelihood of patient selection bias is low, however, as the only statistically significant difference found to exist between the group compositions and characteristics was an increased APACHE II score in group 3 compared with group 1 and the majority of the enrolled patients arrived in the STICU after a traumatic event, a presumed random occurrence. Third, because the exact compositions of immune-modulated formula differ between manufacturers, these results are specific to the formula used, and the activity of each individual ingredient cannot be distinguished. Finally, glutamine absorption in critically ill patients may not be predictable or complete. The original power calculations were based on a projected 30% mortality rate for patients in the control group, which does not correlate with the observed rate of 6.1%. The 30% mortality statistic for this particular STICU is based on the usual surgery/trauma admission ratio of 3:2, which is considerably larger than that observed in this study, likely because surgical patients more frequently receive total parenteral nutrition than enteral nutrition. In our experience, trauma patients have improved survival over surgery patients. Additionally, patients who died within 72 hrs of admission (who are included in the 30% mortality estimate) did not have enteral feeds started and were therefore not included in the study. In hindsight, the 30% statistic was erroneously high. Additionally, the study was underpowered to show that glutamine was harmful, an unexpected trend. However, since the initial hypothesis was that glutamine would be helpful, we felt that we had enough data to support the null hypothesis. Despite efforts to maintain our subjects on 25–30 kcal/kg and 2.0 g/kg protein per day, significant differences in intake were seen. This does not appear to correlate with intolerance of feeds, and multivariate analysis did not demonstrate a relationship between protein or calories received and mortality. Because this was an intent-to-treat analysis, all patients were included regardless of calories and protein received and whether goal rates were reached. Meeting energy and pro-

tein goals with enteral nutrition is notoriously difficult and is not unique to this study. ICU studies evaluating nutrient delivery in patients receiving enteral nutrition have demonstrated that only between 52% and 64% of goal calories and protein are actually administered in the clinical setting (37–39), and other studies on enteral immune-enhanced nutrition have had delayed start times (22). Additionally, to attempt to match protein intake between the patients, varying quantities of protein powder were used under the assumption that it would not have a therapeutic effect. Similarly, we assumed no therapeutic effect from the maltodextran, which was included in the packets of powdered glutamine. Although these data do not demonstrate improved outcomes in patients receiving supplemental enteral glutamine, further investigation should be carried out to explore why this might be the case, perhaps better elucidating the role of glutamine on bacterial proliferation. As parenteral glutamine supplementation has had more conclusive results and because the preferred route of nutrient administration in trauma patients is enteral (12, 22, 40), future studies could evaluate the role of intravenous glutamine as a neutraceutical while the patient receives enteral feedings.

CONCLUSIONS Although numerous studies have evaluated the role of parenteral and enteral glutamine supplementation in critically ill, trauma, and surgical patients, the results have been inconclusive in regard to mortality rate in the trauma and surgical population. This study, one of the largest to date, failed to demonstrate improved outcomes with enteral glutamine supplementation. Enteral glutamine supplementation should not be used routinely in the trauma and surgical critically ill population (outside of a study setting), as we have shown no clear benefit.

REFERENCES 1. Andrews FJ, Griffiths RD: Glutamine: Essential for immune nutrition in the critically ill. Br J Nutr 2002; 87(Suppl 1):S3–S8 2. Goeters C, Wenn A, Mertes N, et al: Parenteral L-alanyl-L-glutamine improves 6-month outcome in critically ill patients. Crit Care Med 2002; 30:2032–2037 3. Griffiths RD, Jones C, Palmer TEA: Sixmonth outcome of critically ill patients given glutamine-supplemented parenteral nutrition. Nutrition 1997; 13:295–302

2505

4. Griffiths RD, Allen KD, Andrews FJ, et al: Infection, multiple organ failure, and survival in the intensive care unit: Influence of glutamine-supplemented parenteral nutrition on acquired infection. Nutrition 2002; 18:546 –552 5. Grimm H, Kraus A: Immunonutrition— Supplementary amino acids and fatty acids ameliorate immune deficiency in critically ill patients. Langenbecks Arch Surg 2001; 386: 369 –376 6. Houdijk APJ, van Leeuwen PAM: Glutamineenriched enteral nutrition in multiple trauma patients. Nutrition 2000; 16:70 –71 7. Karinch AM, Pan M, Lin CM, et al: Glutamine metabolism in sepsis and infection. J Nutr 2001; 131:2535S–2538S 8. Neu J, DeMarco V, Li N: Glutamine: Clinical applications and mechanisms of action. Curr Opin Clin Nutr Metab Care 2002; 5:69 –75 9. Wischmeyer PE: Clinical applications of Lglutamine: past, present, and future. Nutr Clin Pract 2003; 18:377–385 10. Lebow BI, Souba WW: Glutamine. World J Surg 2000; 24:1503–1513 11. Wilmore DW: The effect of glutamine supplementation in patients following elective surgery and accidental injury. J Nutr 2001; 131: 2543S–2549S 12. Moore FA, Moore EE, Jones TN, et al: TEN versus TPN following major abdominal trauma—Reduced septic morbidity. J Trauma 1989; 29:916 –922 13. Jones C, Palmer TEA, Griffiths RD: Randomized clinical outcome study of critically ill patients given glutamine-supplemented enteral nutrition. Nutrition 1999; 15:108 –115 14. Novak F, Heyland DK, Avenell A, et al: Glutamine supplementation in serious illness: A systematic review of the evidence. Crit Care Med 2002; 30:2022–2029 15. Boelens PG, Houdijk APJ, Fonk JCM, et al: Glutamine-enriched enteral nutrition increases HLA-DR expression on monocytes of trauma patients. J Nutr 2002; 132: 2580 –2586 16. Conejero R, Bonet A, Grau T, et al: Effect of a glutamine-enriched enteral diet on intestinal permeability and infectious morbidity at 28 days in critically ill patients with systemic inflammatory response syndrome: A ran-

2506

17.

18.

19.

20.

21.

22.

23. 24.

25.

26.

27.

28.

domized, single-blind, prospective, multicenter study. Nutrition 2002; 18:716 –721 Garrel D, Patenaude J, Nedelec B, et al: Decreased mortality and infectious morbidity in adult burn patients given enteral glutamine supplements: A prospective, controlled, randomized clinical trial. Crit Care Med 2003; 31:2444 –2449 Houdijk APJ, Rijnsburger ER, Jansen J, et al: Randomised trial of glutamine-enriches enteral nutrition on infectious morbidity in patients with multiple trauma. Lancet 1998; 352:772–776 Hall JC, Dobb G, Hall J, et al: A prospective randomized trial of enteral glutamine in critical illness. Intensive Care Med 2003; 29: 1710 –1716 Long CL, Borghesi L, Stahl R, et al: Impact of enteral feeding of a glutamine-supplemented formula on the hypoaminoacidemia response in trauma patients. J Trauma 1996; 40: 97–102 McCowen KC, Bistrian BR: Immunonutrition: Problematic or problem solving? Am J Clin Nutr 2003; 77:764 –770 Jacobs DG, Jacobs DO, Kudsk KA, et al: Practice management guidelines for nutritional support of the trauma patient. J Trauma 2004; 57:660 – 679 Buchman A: The role of glutamine: Counterpoint. Nutr Clin Pract 2003; 18:391–396 Kajikawa H, Mitsumori M, Ohmomo S: Stimulatory and inhibitory effects of protein amino acids on growth rate and efficiency of mixed ruminal bacteria. J Dairy Sci 2002; 85:2015–2022 Garner JS, Jarvis WR, Emori TG, et al: CDC definitions for nosocomial infections. Am J Infect Control 1988; 16:128 –140 Evans HL, Raymond DP, Pelletier SJ, et al: Tertiary peritonitis (recurrent diffuse or localized disease) is not an independent predictor of mortality in surgical patients with intraabdominal infection. Surg Infect (Larchmt) 2001; 2:255–263 Offner PJ, Moore EE, Biffl WL: Male gender is a risk factor for major infections after surgery. Arch Surg 1999; 134:935–938 Knaus WA, Draper EA, Wagner DP, et al: APACHE II: A severity of disease classification. Crit Care Med 1985; 13:818 – 829

29. Baker SP, O’Neill B, Haddon W Jr, et al: The injury severity score: A method for describing patients with multiple injuries and evaluating emergency care. J Trauma 1974; 14: 187–196 30. Parry-Billings M, Evans J, Calder PC, et al: Does glutamine contribute to immunosuppression after major burns? Lancet 1990; 336:523–525 31. Oudemans-van Straaten HM, Bosman RJ, Treskes M, et al: Plasma glutamine depletion and patient outcome in acute ICU admissions. Intensive Care Med 2001; 27:84 –90 32. Alverdy JC: Effects of glutamine-supplemented diets on immunology of the gut. JPEN J Parenter Enteral Nutr 1990; 14: 109S–113S 33. Foitzik T, Kruschewski M, Kroesen AJ, et al: Does glutamine reduce bacterial translocation? Int J Colorectal Dis 1999; 14:143–149 34. Panigrahi P, Gewolb IH, Bamford P, et al: Role of glutamine in bacterial transcytosis and epithelial cell injury. JPEN J Parenter Enteral Nutr 1997; 21:75– 80 35. De-Souza DA, Greene LJ: Intestinal permeability and systemic infections in critically ill patients: Effect of glutamine. Crit Care Med 2005; 33:1175–1178 36. Kieft H, Roos AN, van Drunen JDE, et al: Clinical outcomes of immunonutrition in a heterogeneous intensive care population. Intensive Care Med 2005; 31:524 –532 37. Krishnan JA, Parce PB, Martinez A, et al: Caloric intake in medical ICU patients: Consistency of care with guidelines and relationship to clinical outcomes. Chest 2003; 124: 297–305 38. Heyland DK, Schroter-Noppe D, Drover JW, et al: Nutrition support in the critical care setting: current practice in Canadian ICUs— Opportunities for improvement? JPEN J Parenter Enteral Nutr 2003; 27:74 – 83 39. Elpern EH, Stutz L, Peterson S, et al: Outcomes associated with enteral tube feeds in a medical intensive care unit. Am J Crit Care 2004; 13:221–227 40. Kudsk KA, Croce MA, Fabian TC, et al: Enteral versus parenteral feeding: effects on septic morbidity after blunt and penetrating abdominal trauma. Ann Surg 1992; 215: 503–511

Crit Care Med 2005 Vol. 33, No. 11