burns & propranolol

Propranolol Does Not Increase Inflammation, Sepsis,or Infectious Episodes in Severely Burned ChildrenMarc G. Jeschke, MD, PhD, William B. Norbury, MD, Celeste C. Finnerty, PhD, Ludwik K. Branski, MD,and David N. Herndon, MD

Background: Propranolol, a nonse-lective �1–2 antagonist, attenuates hy-permetabolism and catabolism in severelyburned patients. However, recent datasuggest that propranolol impairs immunefunction and enhances inflammation. Thepurpose of the present study was to deter-mine the effect of propranolol administra-tion on infection, sepsis, and inflammationin severely burned pediatric patients.

Patients: A prospective, intent-to-treat study was performed; patient de-mographics (age, gender, burn size, andmortality); infectious episodes (colonycount greater then 105); and sepsis(guidelines by the society of critical care

medicine) were determined. Hypermeta-bolic response was determined by restingenergy expenditure (REE), and the in-flammatory response was determined bymeasuring serum cytokine expression.

Results: Two hundred forty-five pa-tients (143 controls, 102 propranolol) wereincluded into the study. There were nodifferences between the control and pro-pranolol groups for age, gender distribu-tion, burn size, third degree burn, andlength of stay. Mortality was 6% in the con-trol group and 5% in the propranololgroup. Propranolol significantly decreasedREE and predicted REE during acute hos-pital stay. Forty-three patients developed

infections in the control group (30%),whereas 21 developed infections in thepropranolol group (21%). The incidenceof sepsis was 10% for controls and 7% forpropranolol. Analysis of the cytokine ex-pression profile in 20 patients in eachgroup revealed that propranolol signifi-cantly decreased serum tumor necrosisfactor and interleukin-1� compared withcontrols (p < 0.05).

Conclusion: Propranolol treatmentattenuates hypermetabolism and does notcause increased incidence of infection andsepsis.

Key Words: Burns, Sepsis, Beta-blockade, Pediatric, Inflammation.

J Trauma. 2007;62:676–681.

Endogenous catecholamines are primary mediators of thehypermetabolic response to trauma or burn.1 Levels ofthese “fight-or-flight” stress hormones are increased ap-

proximately 10-fold shortly after severe blunt trauma or aburn of over 30% to 40% total body surface area (TBSA).2–4

This highly conserved systemic response to injury is charac-terized by development of a hyperdynamic circulation,5 re-setting of the hypothalamic temperature regulation point,4,6

elevated basal energy expenditure,7 peripheral insulin resis-tance with hyperglycemia,8–10 increased peripheral lipoly-sis,9 altered immune function,11 and skeletal muscle proteincatabolism.12 We and others have attenuated supraphysi-ologic thermogenesis,12 tachycardia,13 cardiac work,14 andresting energy expenditure16 with administration of a nonse-lective �1/�2-blocking agent (propranolol) after severe burn.Decreased cardiac morbidity and diminished overall mortal-ity have been documented in non-trauma patients who were

given �-blockers for control of tachycardia after iatrogenictissue trauma inflicted by a major surgical procedure.16 In alater study, propranolol was administered at a dose that re-duced heart rate by approximately 15% to 20% from basallevels; propranolol decreased resting energy expenditure, im-proved skeletal muscle protein kinetics, and preserved fat-free and lean body mass.15 However, there is evidence thatpropranolol decreases immune function and is detrimentalduring infectious episodes or septicemia.17,18 The purpose ofthe present study was to determine the effect of propranololadministration on infection, sepsis, and inflammation in se-verely burned pediatric patients.

PATIENTS AND METHODSAn intent-to-treat study was performed (intent-to-treat

means patients were randomized to control or propranololand received it for more than 3 days). We did not dividepatients based on the length of propranolol administration orbetween responders and nonresponders. This study was ap-proved by the University of Texas Medical Branch Institu-tional Review Board. Informed written consent was obtainedfrom each patient’s guardian before enrollment in the study.Inclusion criteria were children younger than 18 years of ageand TBSA burns of greater than 40%. From 1996 to 2006,severely burned patients with burns over 40% TBSA wererandomized into two groups: controls (no anabolic agenttreatment) and propranolol (0.5–1.5 mg/kg by mouth every 6hours).

Submitted for publication October 2, 2006.Accepted for publication December 14, 2006.Copyright © 2007 by Lippincott Williams & Wilkins, Inc.From the Department of Surgery, Shriners Hospitals for Children,

University Texas Medical Branch, Galveston, Texas.Supported by the Shriners Hospital for Children grants 8660, 8760, and

9145, NIH R01-GM56687, T32 GM008256, P50 GM60338, and NIDRRH133A020102.

Address for reprints: Marc G. Jeschke, MD, PhD, Shriners Hospitalsfor Children, Galveston Burns Unit, 815 Market Street, Galveston, TX77550; email: [email protected].

DOI: 10.1097/TA.0b013e318031afd3

The Journal of TRAUMA� Injury, Infection, and Critical Care

676 March 2007

We have a randomization list at our institution and, afterreceiving consent, patients were randomized according to thatlist. In this study, all patients received the same standardacute burn care. Within 48 hours of admission, each patientunderwent total burn wound excision and grafting with au-tograft skin and allograft. Patients returned to the operatingroom when autograft donor sites healed and became availablefor reharvest (usually 6–8 days from the last operation).Sequential, staged surgical procedures for repeat excision andgrafting were performed until the wounds were healed. Eachpatient received enteral nutrition via a nasoduodenal tubewith Vivonex TEN (Sandoz Nutritional Corp., Minneapolis,MN). The composition of Vivonex is 82% carbohydrate, 15%protein, and 3% fat. Daily caloric intake was given at a ratecalculated to deliver 1,500 kcal/m2 TBSA burned �1,500kcal/m2 TBSA. This feeding regimen was started at admis-sion and continued at a constant rate until the wounds werehealed. Caloric intake was based on resting energy expendi-ture (REE). REE was determined weekly and patients re-ceived the caloric amount (1.4 times REE).

Burned patients were connected to an Emtek vitals signstracking system (Eclipsys, Rockville, MD) by standard echo-cardiogram leads. Heart rate was measured hourly and veri-fied by each patient’s nurse. The average heart rate for eachentire 24-hour period was determined throughout the hospitalstay. Clinical data were prospectively collected.

After a control period of 7 days, patients were random-ized to either receive normal saline or propranolol at a dose of0.5 to 1.5 mg/kg by mouth every 6 hours. Patients wereclosely monitored for heart rate and blood pressure. Patientsdid not receive any other anabolic or anticatabolic agent. Allpatients received insulin if necessary (blood glucose �210mg/dL) to decrease blood glucose below 210 mg/dL, with atarget blood glucose of 140 to 160 mg/dL.

Indirect CalorimetryAll patients underwent REE measurements within 1

week after hospital admission and weekly thereafter duringtheir acute hospitalization. For the present study, we chosethe first metabolic study and compared it to the metabolicstudy at discharge. The studies were performed betweenmidnight and 5 AM while the patients were asleep and receiv-ing continuous feeding. REE was measured using a Sensor-Medics Vmax 29 metabolic cart (Yorba Linda, CA). Subjectswere tested in a supine position while under a large, clear,ventilated hood. The REE was calculated from the oxygenconsumption and carbon dioxide production by equations. AllREE measurements were made at ambient temperatures of30° C, which is the standard environmental setting for allpatient rooms in our acute burn intensive care unit. The REEmeasurements were used to guide nutritional managementand to assess the level of metabolism. The discharge REEmeasurement was used to determine the level of hyperme-tabolism when the burn wounds were 95% healed, and wereincluded as part of this study. Measured values were com-

pared with predicted norms based on the Harris-Benedictequation.

Infection Episodes and SepsisDiagnosis of infectious episodes was determined by col-

ony counts of colony forming units (CFUs) greater than 105

CFUs, with the identification of the bacteria. Diagnosis ofsepsis was made through hospitalization, and final diagnosiswas confirmed by autopsy-demonstrated sepsis (presence andidentification of organisms and inflammatory response inmultiple organs). Because the hypermetabolic state inducedby a severe burn can mimic conditions typically used todiagnose systemic inflammatory response syndrome or sepsisas defined by the guidelines of the Society of Critical CareMedicine, a modified list of criteria was used to diagnosesepsis in the burn patient (Table 1). Although the patientswere alive, the presence of organisms in burn wound, blood,or tissue biopsies was confirmed by colony counts; at the timeof autopsy, colony counts from organ cultures or histologicdetection were used to confirm presence of invasive organ-isms. Most patients were found to be infected with a combi-nation of fungi and gram-negative and -positive bacteria.

Serum CytokinesBlood was collected from the burn patients at the time of

admission, preoperatively, and 5 days postoperatively for 4weeks for serum cytokine analysis. Blood was drawn in aserum-separator collection tube and centrifuged for 10 min-utes at 1,320 rpm; the serum was removed and stored at �70°C until assayed. The Bio-Plex Human Cytokine 17-Plexpanel was used with the Bio-Plex Suspension Array System(Bio-Rad, Hercules, CA) to profile expression of seventeeninflammatory mediators (interleukin [IL]-1�, IL-2, IL-4,IL-5, IL-6, IL-7, IL-8, IL-10, IL-12p70, IL-13, IL-17, gran-ulocyte colony-stimulating factor, granulocyte-macrophagecolony stimulating factor, interferon-gamma, monocyte che-moattractant protein-1, macrophage inflammatory protein1-beta, and tumor necrosis factor [TNF]). The assay wasperformed according to the manufacturer’s instructions. Se-rum samples were briefly thawed and then centrifuged at4,500 rpm for 3 minutes at 4° C. Serum samples were then

Table 1 Criteria Used to Diagnose Sepsis

At least three of the following:DeliriumTemperature �38.5 or �36.5° CTachycardia (�120 BPM in adults, �150 BPM in children)TachypneaWhite blood cell count �12,000 or �4,000Refractory hypotension (systolic blood pressure �90 mm Hg)Thrombocytopenia (platelets �50,000 mm3)Hyperglycemia (serum glucose �240 mg/dl)Enteral feeding intolerance (residuals �200 mL/h or diarrhea

�1 L/d) and pathologic tissue source identified and �105

organisms/gram tissue

Propranolol and Infections

Volume 62 • Number 3 677

incubated with microbeads labeled with specific antibodies toone of the aforementioned cytokines for 30 minutes. After awash step, the beads were incubated with the detection anti-body cocktail, with each antibody specific to a single cyto-kine. After another wash step, the beads were incubated withstreptavidin-phycoerythrin for 10 minutes, washed, and theconcentrations of each cytokine were determined using thearray reader.

Statistical AnalysisOne-way analysis of variance, with Bonferroni’s post

hoc correction (cytokine analysis) and paired and unpairedStudent’s t tests (demographics, REE, infections) were usedto compare the two groups. Data are expressed as percentagesor means � standard error of the mean, where appropriate.Significance was accepted at p � 0.05.

RESULTSTwo hundred forty-five patients (143 controls, 102 pro-

pranolol) were included into the study. There were no differ-ences between the control and propranolol groups for age,gender distribution, burn size, third degree burn, inhalationinjury, and length of stay. Mortality was 6% in the controlgroup and 5% in the propranolol group (Table 2).

Resting Energy ExpenditurePatients receiving propranolol demonstrated a significant

change in REE and predicted REE compared with controlpatients (p � 0.05) (Fig. 1). Although control patients had anincrease in REE and percent predicted REE during the acutestay, propranolol patients showed a decreased change in REEand percent predicted REE (p � 0.05) (Fig. 1).

Infection Episodes and SepsisForty-three patients (30%) had infectious episodes with

CFUs �105 during the acute hospital stay, whereas 21 pa-tients (21%) in the propranolol group showed signs of infec-tion, which was not significantly different (Fig. 2). Similarly,

in 14 (10%) control patients, the diagnosis of sepsis wasmade, whereas the incidence of sepsis was 7% in the pro-pranolol group. There was no significant difference betweenthe two groups (Fig. 2). Furthermore, incidence of specificinfections did not vary between groups.

Serum CytokinesAnalysis of the cytokine expression profile in 20 patients

from each group revealed only minor differences. Proprano-lol administration significantly decreased serum TNF andIL-1�, at one time point when compared with controls (p �0.05) (Fig. 3). We found no differences between the propran-olol and control groups for IL-6, IL-8, IL-10, monocyte

Table 2 Demographic Data for Control andPropranolol Patients

Control Propranolol

Number 143 102Age (years) 7.8 � 0.4 7.2 � 0.6Gender (F/M) 83/119 43/59Time from burn to admit (days) 6 � 1 6 � 1TBSA (%) 55 � 1 54 � 23rd degree burns (%) 43 � 2 44 � 3Length of ICU stay (days) 32 � 2 30 � 2Length of ICU/TBSA (days/%) 0.56 � 0.02 0.54 � 0.02Inhalation injury (%) 35 39Multi Organ Failure (%) 9 6Mortality (%) 6 5

Data presented as mean � SEM.TBSA, total body surface area; ICU, intensive care unit.

Fig. 1. Resting energy expenditure (REE) and predicted REE. Al-though control patients had an increase in REE and percent pre-dicted REE during the acute stay, propranolol patients showed adecreased change in REE and percent predicted REE. *Significantdifference between control and propranolol (p � 0.05).


678 March 2007

chemoattractant protein-1, and macrophage inflammatoryprotein 1-� during acute hospital stay (Fig. 3).

DISCUSSIONThe systemic inflammatory response after burn is trig-

gered by catecholamines and leads to hypermetabolism, andthus to protein degradation and catabolism. Consequently, thestructure and function of essential organs such as the muscle,skin, heart, immune system, and liver are compromised, con-tributing to multiple organ failure and mortality.19,20 Uncon-trolled release of proinflammatory mediators such as IL-6,IL-8, and acute-phase proteins trigger and enhance protein

wasting and organ dysfunction.21–25 Organ function break-down can then lead to increased incidence of infection andsepsis, ultimately leading to multiple organ failure and death.This circulus vitiosus is very difficult to break and successfultherapy for sepsis has yet to be defined.

There is evidence in the literature that catecholaminesare beneficial to fighting infections and enhancing the im-mune system. For example, it has been recently shown thatepinephrine enhances platelet-neutrophil adhesion, which iscrucial for survival of critically ill patients.26 Ortega et al.27

showed that Norepinephrine stimulated phagocytosis inducedby moderate exercise. Garcia et al.28 showed in 2003 thatnoradrenaline modulates the phagocytic process of macro-phages and maintains the phagocytic functions at physio-logically optimal levels. Modulation of chemotaxis is mainlymediated by �-receptors and phagocytosis needs both �- and�-receptor stimulation. Based on these effects of catecholamineson immune cells, the question arises as to whether nonselective�-blockade has a detrimental or adverse effect on the immunesystem and function in severely burned pediatric patients. Wedid not measure immune function directly; however, we sug-gest that the incidence of infections or sepsis can be used asa marker for immune function. We found that there was nodifference in the incidence of infection or sepsis in severelyburned pediatric patients receiving placebo or propranololthroughout the hospital stay. Both groups had an incidenceof infection of approximately 25% to 30% and incidence ofsepsis of 10% and 7%, respectively, and the incidence ofspecific infections did not vary between groups. This dataindicate that propranolol has no adverse effects on infec-tions and sepsis. There was also no difference betweengroups regarding mortality, multiple organ failure, or lengthof hospital stay. We did not look at pneumonia in the presentstudy and we might have missed differences in the incidenceof pneumonia between groups.

That propranolol was administered in the correct dosecan be seen in a decreased hypermetabolic response. In agree-ment with previous data, we found that propranolol decreasedREE and predicted REE in severely burned pediatric patientswhen compared with controls. We hypothesized that an at-tenuated hypermetabolic response is associated with a de-creased inflammatory response. We showed that propranololdid not affect the inflammatory response as compared withcontrol patients. Propranolol decreased TNF and IL-1� con-centration in the serum, but the biological significance isquestionable because levels were only decreased at one timepoint. Therefore, we suggest that propranolol did not alter theinflammatory reaction compared with controls.

In summary, we showed that a nonselective �1/�2 an-tagonist does not affect the incidence of infections or sepsis inseverely burned pediatric patients, indicating that propranololhas no adverse effect on the immune system and function.Propranolol decreases REE, which is associated with a de-crease in serum TNF and IL-1�. We suggest that propranololtreatment is beneficial in burn victims and may improve

Fig. 2. Forty-three control patients (30%) had infectious episodeswith colony forming units �105 during the acute hospital stay,whereas 21 patients (21%) showed signs of infections in the pro-pranolol group, which was not significantly different. In 14 controlpatients (10%), the diagnosis of sepsis was made; the incidence ofsepsis was 7% in the Propranolol group. There was no significantdifference between the two groups.



Fig. 3. Analysis of the cytokine expression profile. Propranolol administration significantly decreased serum tumor necrosis factor andinterleukin (IL)-1� at one time point when compared with controls. We found no differences between propranolol and control for IL-6, IL-8,IL-10, monocyte chemoattractant protein-1, macrophage inflammatory protein 1-� during the acute hospital stay. *Significant differencesbetween propranolol and control (p � 0.05).


680 March 2007

survival. To detect differences in mortality, a large, multi-center trial would have to be performed, since mortality ratesin children are 4% to 6%.

REFERENCES1. Wilmore DW, Long JM, Mason AD Jr, Skreen RW, Pruitt BA Jr.

Catecholamines: mediator of the hypermetabolic response to thermalinjury. Ann Surg. 1974;180:653–669.

2. Goodall M, Stone C, Haynes BW Jr. Urinary output of adrenalineand noradrenaline in severe thermal burns. Ann Surg. 1957;145:479–487.

3. Goodall MG. Sympathetic nerve and adrenal medullary response tothermal burn. Clinical analysis of adrenaline and noradrenalinedepletion. Am Surg. 1966;32:448–452.

4. Wilmore DW, Aulick LH. Metabolic changes in burned patients.Surg Clin North Am. 1978;58:1173–1187.

5. Asch MJ, Feldman RJ, Walker HL, et al. Systemic and pulmonaryhemodynamic changes accompanying thermal injury. Ann Surg.1973;178:218–221.

6. Wilmore DW. Hormonal responses and their effect on metabolism.Surg Clin North Am. 1976;56:999–1018.

7. Reiss E, Pearson E, Artz CP. The metabolic response to burns.J Clin Invest. 1956;35:62–77.

8. Wolfe RR, Durkot MJ, Allsop JR, Burke JF. Glucose metabolism inseverely burned patients. Metabolism. 1979;28:1031–1039.

9. Wolfe RR, Herndon DN, Peters EJ, Jahoor F, Desai MH, HollandOB. Regulation of lipolysis in severely burned children. Ann Surg.1987;206:214–221.

10. Wolfe RR, Miller HI, Spitzer JJ. Glucose and lactate kinetics in burnshock. Am J Physiol. 1977;232:E415–418.

11. Yurt RW, McManus AT, Mason AD Jr, Pruitt BA Jr. Increasedsusceptibility to infection related to extent of burn injury. Arch Surg.1984;119:183–188.

12. Herndon DN, Tompkins RG. Support of the metabolic response toburn injury. Lancet. 2004;363:1895–1902.

13. Minifee PK, Barrow RE, Abston S, Desai M, Herndon DN.Improved myocardial oxygen utilization following propranololinfusion in adolescents with postburn hypermetabolism. J PediatrSurg. 1989;24:806–810; discussion 810–811.

14. Baron PW, Barrow RE, Pierre EJ, Herndon DN. Prolonged use ofpropranolol safely decreases cardiac work in burned children. J BurnCare Rehabil. 1997;18:223–227.

15. Herndon DN, Hart DW, Wolf SE, Chinkes DL, Wolfe RR. Reversalof catabolism by beta-blockade after severe burns. N Engl J Med.2001;345:1223–1229.

16. Mangano DT, Layug EL, Wallace A, Tateo I. Effect of atenolol onmortality and cardiovascular morbidity after noncardiac surgery.Multicenter Study of Perioperative Ischemia Research Group. N EnglJ Med. 1996;335:1713–1720.

17. Oberbeck R. Therapeutic implications of immune-endocrineinteractions in the critically ill patients. Curr Drug Targets ImmuneEndocr Metabol Disord. 2004;4:129–139.

18. Oberbeck R, Schmitz D, Wilsenack K, et al. Adrenergic modulationof survival and cellular immune functions during polymicrobialsepsis. Neuroimmunomodulation. 2004;11:214–223.

19. Arnold J, Campbell IT, Samuels TA, et al. Increased whole bodyprotein breakdown predominates over increased whole bodyprotein synthesis in multiple organ failure. Clin Sci (Lond). 1993;84:655– 661.

20. Rennie MJ. Muscle protein turnover and the wasting due to injuryand disease. Br Med Bull. 1985;41:257–264.

21. Jeschke MG, Barrow RE, Herndon DN. Extended hypermetabolicresponse of the liver in severely burned pediatric patients. ArchSurg. 2004;139:641–647.

22. Jeschke MG, Einspanier R, Klein D, Jauch KW. Insulin attenuatesthe systemic inflammatory response to thermal trauma. Mol Med.2002;8:443–450.

23. Jeschke MG, Herndon DN. Effect of growth factors as therapeuticdrugs on hepatic metabolism during the systemic inflammatoryresponse syndrome. Curr Drug Metab. 2004;5:399–413.

24. Jeschke MG, Rensing H, Klein D, et al. Insulin prevents liverdamage and preserves liver function in lipopolysaccharide-inducedendotoxemic rats. J Hepatol. 2005;42:870–879.

25. Moshage H. Cytokines and the hepatic acute phase response.J Pathol. 1997;181:257–266.

26. Horn NA, Anastase DM, Hecker KE, Baumert JH, Robitzsch T,Rossaint R. Epinephrine enhances platelet-neutrophil adhesion inwhole blood in vitro. Anesth Analg. 2005;100:520–526.

27. Ortega E, Marchena JM, Garcia JJ, Barriga C, Rodriguez AB.Norepinephrine as mediator in the stimulation of phagocytosis inducedby moderate exercise. Eur J Appl Physiol. 2005;93:714–718.

28. Garcia JJ, del Carmen Saez M, De la Fuente M, Ortega E. Regulationof phagocytic process of macrophages by noradrenaline and its endmetabolite 4-hydroxy-3-metoxyphenyl-glycol. Role of alpha- and beta-adrenoreceptors. Mol Cell Biochem. 2003;254:299–304.

Fig. 3. Continued.



burns & propranolol

Documents