DTIC ADA627866: Patterns of Exogenous Insulin Requirement Reflect Insulin Sensitivity Changes in Trauma

0.18 MB·English

by  Defense Technical Information Center

TheAmericanJournalofSurgery194(2007)798–803
Presentation
Patterns of exogenous insulin requirement reflect insulin sensitivity
changes in trauma
Heather F. Pidcoke, M.D.a,b,*, Jose Salinas, Ph.D.a, Sandra M. Wanek, M.D.a,
Marybeth Concannon, B.S.a, Florence Loo, B.S.a, Kelly L. Wirfel, M.D.b,
John B. Holcomb, M.D.a,b, Steven E. Wolf, M.D.a,b, Charles E. Wade, Ph.D.a,b
aUnitedStatesArmyInstituteofSurgicalResearch,BrookeArmyMedicalCenter,3400RawleyE.ChambersAve,FortSamHouston,
TX78234-6315,USA
bUniversityofTexas,HealthScienceCenter,SanAntonio,7703FloydCurlDr,SanAntonio,TX78229,USA
ManuscriptreceivedMay15,2007;revisedmanuscriptAugust13,2007
Presentedatthe59thAnnualMeetingoftheSouthwesternSurgicalCongress,RanchoMirage,CA,March25–29,2007
Abstract
Introduction: We investigated patterns of blood glucose and exogenous insulin requirement in the
intensivecareunit,andquestionedwhethertheyreflectfluctuationsininsulinactivity.
Methods: Recordsforburnintensivecareunitpatientswith7daysofglucosecontrolwithinsulinwere
reviewed.Hourlybloodglucoseandinsulindosewerematchedfortimecollectedandanalyzedwithlinear
andcosineregression.Frequencyanalysisidentifiedrecurringpatterns.
Results: Diurnalpatternsofbloodglucoseandinsulinrequirementwerenoted(insulintroughs(cid:1)noon;
insulinpeaks(cid:1)midnight;glucosetroughs(cid:1)5AM;glucosepeaks(cid:1)5PM).Averageinsulinrequirement
increasedataconstantlinearrate(slope(cid:1).013,r2(cid:1).57,P(cid:1).001).
Conclusions: Diurnalpatternsinbloodglucoseandinsulinrequirementmirrorthoseofhealthysubjects
andmayreflectpersistenceofnormalvariabilityininsulinactivity.The5-houroffsetinpeaksandtroughs
is suggestive of complex interplay between insulin availability and receptor sensitivity. The insulin
requirementtobloodglucoseratioincreased,evidencethatinsulinresistanceprogressesovertime.©2007
ExcerptaMedicaInc.Allrightsreserved.
Keywords:Hyperglycemia;Trauma;Diurnalvariation;Circadianrhythm;Insulinsensitivity;Insulinresistance
Multiple studies have demonstrated the association in crit- Glucose regulation in healthy subjects is characterized
ical injury of hyperglycemia with increased morbidity and bymaintenanceofbloodglucosewithinnarrowparameters
mortality [1–8], and this has led in turn to widespread [16];however,withinthistightrange,glucosefluctuatesin
adoptionofintensiveinsulinprotocolsaimedatmaintaining adiurnalpattern[17].Thecircadianrhythmischaracterized
blood glucose within tight parameters [9–12]. However, byanadirinthemorning,followedbyapeakintheevening
such protocols are difficult to implement because glucose [17]. Other molecules involved in glycemic regulation dis-
andinsulinpatternsintraumaarepoorlyunderstood.Recent play similar circadian rhythms, including insulin, cortisol,
work in mixed intensive care units (ICUs) found that glu- and leptin [17–20]. Plasma insulin clearance is typically
cose variability is also associated with poor outcomes [13– loweratnight,andhigherduringtheday,reflectingchanges
15], thus merely controlling hyperglycemia may not be in insulin receptor sensitivity [19]. During the day, when a
sufficient. A clear understanding of normal and abnormal majorityofcaloriesareingested,insulinactivityisatitshigh-
variance in glucose and insulin in critical injury may be est,subsequentlydecliningatnight[17–20].Thisismanifested
necessary to adequately maintain glucose homeostasis. in diabetics as the “dawn phenomenon,” a term used to de-
scribetheobservedincreaseininsulinrequirementatnightin
thesepatients[17].
*Corresponding author. Tel.: (cid:2)1-210-916-3301; fax: (cid:2)1-210-916- Inacuteinjury,diurnalvarianceofcortisolandleptinare
2942. knowntobebluntedorabsent[21],andasaresult,similar
E-mailaddress:[email protected] attenuationofglucoseandinsulindiurnalpatternsmightbe
0002-9610/07/$–seefrontmatter©2007ExcerptaMedicaInc.Allrightsreserved.
doi:10.1016/j.amjsurg.2007.08.022
Report Documentation Page Form Approved
OMB No. 0704-0188
Public reporting burden for the collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and
maintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information,
including suggestions for reducing this burden, to Washington Headquarters Services, Directorate for Information Operations and Reports, 1215 Jefferson Davis Highway, Suite 1204, Arlington
VA 22202-4302. Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to a penalty for failing to comply with a collection of information if it

does not display a currently valid OMB control number.
1. REPORT DATE 2. REPORT TYPE 3. DATES COVERED
01 DEC 2007 N/A -
4. TITLE AND SUBTITLE 5a. CONTRACT NUMBER
Patterns of exogenous insulin requirement reflect insulin sensitivity
5b. GRANT NUMBER
changes in trauma
5c. PROGRAM ELEMENT NUMBER
6. AUTHOR(S) 5d. PROJECT NUMBER
Pidcoke H. F., Salinas J., Wanek S. M., Concannon M., Loo F., Wirfel K.
5e. TASK NUMBER
L., Holcomb J. B., Wolf S. E., Wade C. E.,
5f. WORK UNIT NUMBER
7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) 8. PERFORMING ORGANIZATION
United States Army Institute of Surgical Research, JBSA Fort Sam REPORT NUMBER
Houston, TX 78234
9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSOR/MONITOR’S ACRONYM(S)
11. SPONSOR/MONITOR’S REPORT
NUMBER(S)
12. DISTRIBUTION/AVAILABILITY STATEMENT
Approved for public release, distribution unlimited
13. SUPPLEMENTARY NOTES
14. ABSTRACT
15. SUBJECT TERMS
16. SECURITY CLASSIFICATION OF: 17. LIMITATION OF 18. NUMBER 19a. NAME OF
ABSTRACT OF PAGES RESPONSIBLE PERSON

a. REPORT b. ABSTRACT c. THIS PAGE SAR 6
unclassified unclassified unclassified
Standard Form 298 (Rev. 8-98)
Prescribed by ANSI Std Z39-18
H.F.Pidcokeetal./TheAmericanJournalofSurgery194(2007)798–803 799
Table1 Point-of-care glucometer readings were collected and a
Populationdemographics correctionformulaappliedtoadjustforhematocriteffectin
Variable Value Numberofsubjects this profoundly anemic population [29]. Hourly blood glu-
cose and insulin dose requirement were matched between
TBSAburned(%) 35(cid:5)23 129
subjects for time of day given, and the glucose values
ISS 18(cid:5)11 92
averagedtoobtaina24-hourcurve.Averagebloodglucose
Inhalationinjury(%) 26 153
andinsulinrequirementforthestudypopulationwereplot-
Polytrauma(%) 9 147
Averagecorrectedglucuse(mg/dL) 120(cid:5)24 156 ted with application of linear and cosine regression to de-
Age(y) 46(cid:5)20 156 termine daily and overall trends. Data filtration and fre-
Percentfemale 24 156 quency analysis identified periodicity of recurring patterns.
Comparison was then made between blood glucose and
TBSA(cid:1)totalbodysurfacearea;ISS(cid:1)InjurySeverityScore.
exogenous insulin levels in the population.
Initial and average blood glucose means (cid:5) SD were
assumed to be present. However current research in mixed calculated.Periodicitywasdeterminedwithfrequencyanal-
ICUs changed this view, as preservation of glucose circa- ysis. Diurnal glycemic and insulin requirement variability
dian rhythms was demonstrated [22]. Circadian rhythms of was analyzed by simple regression and cosine regression,
glucose metabolism in trauma patients remain to be char- and significance evaluated with Student t test. Quantitative
acterized, and insulin patterns have not been examined in variableswithnormaldistributionwereanalyzedusingStu-
any ICU population. dent t test; when distribution was not normal, the nonpara-
Glycemicdysregulationincriticalinjuryiscommon,and metric Mann-Whitney test was used. Simple chi-square
manifests as acute and long-term hyperglycemia. Insulin testswereusedforqualitativevariables.Testswere2-sided
resistance is thought to play a role, perpetuating hypergly- in all cases. Significance level was set at .05.
cemia and, in consequence, exacerbating catabolism [23–
25].Weinvestigatedwhethervariabilityinbloodglucoselev- Results
els and exogenous insulin requirement reflect fluctuations in Review of admissions to our burn unit over 54 months
underlyinginsulinsensitivity.Wewonderedwhether:(1)glu- identified 156 eligible subjects. Total body surface area
cose diurnal patterns are preserved in trauma patients; (2) burned was 35% (cid:5) 23%, Injury Severity Score was 18 (cid:5)
insulin diurnal patterns are preserved in trauma patients; and 11,andagewas46(cid:5)20years(Table1).Amongoutcome
(3)increasedinsulinresistanceisdemonstratedbycomparison measures evaluated were mortality, incidence of pneumo-
ofglucoseandinsulinpatterns. nia, ARDS, bacteremia, and sepsis, and these were com-
bined to provide the percent of subjects that experienced
Methods singleandmultiplecomplications(Table2).Averageinitial
After regulatory approval was obtained, records were glucose was 148 (cid:5) 61 mg/dL, and trended down sharply
reviewedforallpatientsadmittedtoasingleburnICU,and over the first 24 hours to a mean of 120 (cid:5) 24 with insulin
data were collected from January 1, 2002 to December 31, treatment. On days 2 through 7, blood glucose levels and
2004forthosewhoreceivedaminimumof7daysofinsulin exogenousinsulinrequirementeachexhibitedanunderlying
treatment to maintain glucose control. The insulin nomo- cyclic pattern (Fig. 1). Frequency analysis demonstrated
gram was determined by physician preference until August that both were diurnal in nature (Fig. 2), with peaks at 24
2003, when a strandardized protocol was adopted (Appen- hours. However, cosine regression of the insulin require-
dix A). Demographic information included total body sur- ment and blood glucose data demonstrated that the respec-
face area burned, injury severity score, polytrauma, age, tive troughs and peaks were offset by 5 hours. Those for
gender, and presence of inhalation injury. Initial and aver- blood glucose were at 5 AM and 5 PM, while troughs and
age glucose was calculated for the population. Outcome peaks for exogenous insulin dose were at noon and mid-
variables were collected, including incidence of mortality, night,respectively(Fig.3).Cosineregressionconfirmedthe
pneumonia, bacteremia, sepsis, and acute respiratory dis- diurnalglucoseandinsulinpatternsforboth;apairedttest
tress syndrome (ARDS), and these data were combined to between the glucose cosine wave and our blood glucose
determine the percent of subjects with single and multiple data renders an r2 of .65 (P (cid:1) .001), while the correlation
complications. Pneumonia was defined as a new, persistent fortheinsulinrequirementcosinewaveandourdatawasr2(cid:1)
radiographic infiltrate with 2 of the 3 following clinical .64 (P (cid:1) .001).
criteria:fever(cid:2)38°C,leukocytosis(cid:2)12(cid:3)109/Lorleuko-
penia(cid:1)4(cid:3)109/L,andpurulentrespiratorysecretions[26].
Table2
Sepsiswasdefinedasaknownorsuspectedinfectioninthe
Outcomemeasures
presence of 2 or more of the following criteria: (1) core
temperature(cid:2)38°C(100.4°F)or(cid:1)36°C(96.8°F);(2)heart Measure % n
rate(cid:2)90beats/min;(3)respiratoryrate(cid:2)20breaths/minor Mortality 21 156
PaCO of (cid:2)32 mm Hg; and (4) white blood cell count of Pneumonia 30 134
2
(cid:2)12,000/(cid:3)Lor(cid:1)4,000/(cid:3)Lor(cid:2)10%immatureneutrophils Bacteremia 13 120
[27]. ARDS was diagnosed for P/F ratio (cid:1)200, bilateral Sepsis 11 127
pulmonaryinfiltratesonchestradiograph,intheabsenceof ARDS 9 136
Complications 56 156
clinicalevidenceofleftatrialhypertension(orapulmonary-
capillarywedgepressureof(cid:4)18mmHg,ifmeasured)[28]. Multiplecomplications 22 156
800 H.F.Pidcokeetal./TheAmericanJournalofSurgery194(2007)798–803
insulin diurnal variation suggests that circadian changes in
insulin receptor sensitivity may persist.
Comments
Hyperglycemia has become a major target of therapy in
most ICUs after intensive insulin therapy was shown to
improve morbidity and mortality [9–12]. However, efforts
tomaintainglucosewithintightphysiologicparametersare
hampered by a paucity of knowledge regarding the impact
of critical injury on normal glucose and insulin curves.
Normal subjects display diurnal variation in several mole-
cules involved in glucose metabolism, including glucose
itself, insulin, cortisol, and leptin [17–20]. Glycemic regu-
lationintraumaisdisordered,evidencedbyhyperglycemia,
Fig.1.Diurnalpatternofrawversusfilteredbloodglucoseover24hours
whichiscommon[1–7].Historically,thiswasconsidereda
for the population studied, demonstrating an early morning trough and
eveningpeak. beneficial response related to the “fight or flight” stress
reaction; this view has since been replaced by the recogni-
tionthatoutcomesareimprovedwithtightglucosecontrol,
While cosine regression clarifies the pattern of glucose usuallywithexogenousinsulin.Inasimilarfashion,diurnal
and insulin regulation, simple regression of the data identi- glucoseandinsulinvariabilitywasassumedtobeattenuated
fies the overall trend in mean levels. The average insulin or absent, after demonstration of these changes in cortisol
dose to correct hyperglycemia in the study population was and leptin circadian rhythms [21].
lowestatadmissionandincreasedataconstantrate(slope(cid:1) Our data show that blood glucose diurnal patterns are
.013, r2 (cid:1) .57, P (cid:1) .001) during the first 7 days of admis- preserved after injury, findings that are similar to those
sion.Incontrast,aftertheinitialdeclineinbloodglucose,a previously reported in a mixed ICU population [22]. Mean
simple regression line expressing the average glucose is glucose, however, remains stable or decreases after an ini-
nearly a horizontal line (slope (cid:1) .001, Fig. 3). The combi- tial precipitous decline in the first 24 hours. It would be
nation of a constantly increasing average daily insulin re- temptingtoconcludethatinsulinactivityisnotreduced,but
quirement, with little to no change in mean glucose levels, comparison of the glucose simple regression line to that of
is evidence that insulin activity decreases over the ICU insulin requirement dispels this notion. Increasingly higher
course. Conversely, preservation of normal glucose and levels of insulin are required to maintain a constant blood
Fig.2.Diurnalpatternofbloodglucoseconfirmedbyfrequencyanalysisdemonstratingastrongpeakat24hours.Asimilarpeakat24hoursidentifiedthe
diurnalpatternofinsulinrequirement.
H.F.Pidcokeetal./TheAmericanJournalofSurgery194(2007)798–803 801
Fig.3.Diurnalpatternsovertimeofinsulinrequirement(top)andbloodglucose(bottom).Simpleregressionformulaforinsulinrequirement(largedash)
demonstratesthattheslopeincreasesovertime,whilethatofglucoseremainsnearlyconstant.Diurnalpeaksandtroughsareoffsetby5hours.
glucose level over time, evidence of diminishing insulin injury [13–15]. Inappropriate application of intensive insu-
activityand/orgrowinginsulinresistanceafterinjury.Clin- lintherapiesduringdecreasedinsulinreceptoractivitycom-
ical conditions other than severe burns and trauma are also bined with feeding during times of relative insulin resis-
known causes of increased insulin requirement, including, tance may drive blood glucose outside the target range,
but not limited to, sepsis, steroid use, and frequent trips to triggeringcounter-regulatorymechanisms,andthusincreas-
the operating room. Causative relationships are beyond the ingoverallvariability.Thismay,inturn,resultinworsened
scope of this study but merit further investigation to eluci- outcomes. Further research is needed to fully characterize
date underlying mechanisms and identify potential targets changes in insulin receptor sensitivity, and use the knowl-
of therapy. edgegainedtorefinebothinsulinandnutritionalstrategies.
Diurnalvariationininsulinrequirementscontinuestobe This study has several limitations; it is a retrospective
seen and may reflect persistence of normal changes in in- review, and cannot show causality. The interplay between
sulin receptor sensitivity; in normal subjects, receptors are bloodglucoseandinsulinrequirementinthispopulationhas
relatively resistant at night. This is perhaps a conserved enhanced our understanding of glucose metabolism in the
protectivemechanismtopreventhypoglycemiaduringfast- critically injured; however, the analysis would be enriched
inghours.Peaksandtroughsforexogenousinsulinrequire- by the addition of nutritional intake, total blood insulin
ment follow those for blood glucose by 5 hours. The half- concentration,C-peptide(ameasureofendogenousinsulin),
life of intravenous regular insulin is 7 minutes; thus the and cortisol levels. Ideally, this information could be cou-
offset likely reflects complex interactions between glucose pled with isotope studies providing data regarding hepatic
production and insulin receptor activity. This is further glucoseproduction,rateofglucosemetabolism,andinsulin
supported by the fact that the nocturnal exogenous insulin receptor function. Despite these limitations, this study pro-
peak reflects a similar nighttime insulin insensitivity in the vides evidence that glucose metabolism in the critically
healthy subjects. This finding has clinical implications for injuredischaracterizedbyahighdegreeofcomplexity,and
how patients are fed. Currently, continuous feeding strate- mayberesponsiblefordifficultiesencounteredinmaintain-
gies predominate; however, if insulin receptor sensitivity ing physiologic euglycemia.
continues to wax and wane in the critically injured, such a
schedule may hamper attempts at insulin control, by pro-
viding calories when the body is less able to metabolize Conclusions
them. A return to bolus feeding during daylight hours may Diurnal patterns of insulin requirement in critical injury
be beneficial and allow for better glucose control. Further mirror insulin levels in healthy subjects, and may reflect
studies are needed to define and test this notion. persistence of normal variability in insulin sensitivity. In-
Normal diurnal variation in glucose metabolism persists sulin requirement peaks and troughs (midnight and noon,
after injury; however abnormal variability also is present. respectively) are likely inversely related to normal fluctua-
StudiesconductedinmixedICUsettingsdemonstratedthat tions in insulin sensitivity, which peaks during daylight
increasedvariabilityisassociatedwithhighermortality,and hours, when calorie consumption is high. Conversely, fast-
our center found that this association is also present after ing at night is physiologically normal and corresponds to
802 H.F.Pidcokeetal./TheAmericanJournalofSurgery194(2007)798–803
decreased insulin sensitivity. The zenith and nadir of glu- [17] CauterEV.Endocrinerhythms.In:BeckerKL,editor.Principlesand
cose levels follow that of insulin; the 5-hour time lag is Practice of Endocrinology and Metabolism, 3rd ed. Philadelphia:
LippincottWilliams&Wilkins;2001:57–68.
suggestive of the complex interplay between insulin avail-
[18] Ahren B. Diurnal variation in circulating leptin is dependent on
ability and receptor sensitivity. The normal sinusoidal pat-
gender, food intake and circulating insulin in mice. Acta Physiol
tern of insulin and glucose regulation appears to be pre- Scand2000;169:325–31.
served in the critically injured, and indicate that the use of [19] BodenG,RuizJ,UrbainJL,andChenX.Evidenceforacircadian
continuousfeedingstrategiesinthispopulationmayneedto rhythmofinsulinsecretion.AmJPhysiol1996;271:E246–52.
[20] la Fleur SE. Daily rhythms in glucose metabolism: suprachias-
be revisited. Further research is needed to investigate
maticnucleusoutputtoperipheraltissue.JNeuroendocrinol2003;
whether a continuous nutritional schedule is optimal, or 15:315–22.
whether a return to bolus feeding would be beneficial. In- [21] Hobson KG, Havel PJ, McMurtry AL, et al. Circulating leptin and
sulinresistanceprogressivelyincreasesinearlytrauma,and cortisolafterburninjury:lossofdiurnalpattern.JBurnCareRehabil
2004;25:491–9.
knowledge of normal and abnormal patterns may improve
[22] Egi M, Bellomo R, Stachowski E, et al. Circadian rhythm of blood
efforts at maintaining euglycemia.
glucosevaluesincriticallyillpatients.CritCareMed2007;35:416–21.
[23] GoreDC,ChinkesDL,HartDW,etal.Hyperglycemiaexacerbates
muscle protein catabolism in burn-injured patients. Crit Care Med
Acknowledgments 2002;30:2438–42.
Supported by National Institutes of Health (NIH) 1 R01 [24] JahoorF,ShangrawRE,MiyoshiH,etal.Roleofinsulinandglucose
GM063120-04 and The Technologies for Metabolic Moni- oxidation in mediating the protein catabolism of burns and sepsis.
AmJPhysiol1989;257:E323–31.
toring (TMM)/Julia Weaver Fund, a Congressionally Di-
[25] GoreDC,WolfSE,HerndonDN,etal.Relativeinfluenceofglucose
rectedProgramJointlyManagedbytheUSAMRMC,NIH,
andinsulinonperipheralaminoacidmetabolisminseverelyburned
NASAandtheJuvenileDiabetesResearchFoundation,and patients.JPENJParenterEnteralNutr2002;26:271–7.
bytheCombatCasualtyCareDivisionUnitedStatesArmy [26] TorresA,el-EbiaryM,PadroL,etal.Validationofdifferenttech-
Medical Research and Materiel Command. niquesforthediagnosisofventilator-associatedpneumonia.Compar-
isonwithimmediatepostmortempulmonarybiopsy.AmJRespirCrit
CareMed1994;149:324–31.
References [27] LevyMM,FinkMP,MarshallJC,etal.2001SCCM/ESICM/ACCP/
ATS/SISInternationalSepsisDefinitionsConference.CritCareMed
[1] GoreDC,ChinkesD,HeggersJ,etal.Associationofhyperglycemia
2003;31:1250–6.
withincreasedmortalityaftersevereburninjury.JTrauma2001;51:
[28] TheARDSNetwork.Ventilationwithlowertidalvolumesascompared
540–4.
withtraditionaltidalvolumesforacutelunginjuryandtheacuterespi-
[2] LairdAM,MillerPR,KilgoPD,etal.Relationshipofearlyhyper-
ratorydistresssyndrome.NEnglJMed2000;342:1301–8.
glycemiatomortalityintraumapatients.JTrauma2004;56:1058–62
[29] PidcokeHP,MannE,SalinasJ,etal.Occulthypoglycemiaunmasked
[3] SungJ,BochicchioGV,JoshiM,etal.Admissionhyperglycemiais
withanalgorithmtocorrecthematocriteffectinpoint-of-care(POC)
predictive of outcome in critically ill trauma patients. J Trauma
glucometers.JBurnCareRes2007;28:S92(abstr).
2005;59:80–3.
[4] Bochicchio GV, Salzano L, Joshi M, et al. Admission preoperative
glucoseispredictiveofmorbidityandmortalityintraumapatientswho
requireimmediateoperativeintervention.AmSurg2005;71:171–4.
Appendix A
[5] BochicchioGV,SungJ,JoshiM,etal.Persistenthyperglycemiais
predictive of outcome in critically ill trauma patients. J Trauma Insulin Protocol Effective From August 15, 2003 to
2005;58:9214. October 1, 2003
[6] VogelzangM,NijboerJM,vanderHorstIC,etal.Hyperglycemia
Treatmentwithintravenouslyadministeredhumanregu-
hasastrongerrelationwithoutcomeintraumapatientsthaninother
criticallyillpatients.JTrauma2006;60:873–7. lar insulin (1 U/mL, Novo Nordisk, Princeton, NJ) was
[7] YendamuriS,FuldaGJ,TinkoffGH.Admissionhyperglycemiaasa initiated when blood glucose surpassed 150 mg/dL, at the
prognosticindicatorintrauma.JTrauma2003;55:33–8.
dose of 2 U/h for glucose 201–300 mg/dL, with 1-U/h
[8] MowlaviA,AndrewsK,MilnerS,etal.Theeffectsofhyperglycemia
on skin graft survival in the burn patient. Ann Plast Surg 2000;45: increases until glucose was between 151 and 200 mg/dL.
629–32. The intravenous insulin infusion was titrated according to
[9] Van den Berghe G, Wouters P, Weekers F, et al. Intensive insulin
the following parameters: insulin was held and 50 mL of
therapyinthecriticallyillpatients.NEnglJMed2001;345:1359–67.
[10] Pham TN, Warren AJ, Phan HH, et al. Impact of tight glycemic 50%percentdextrose(Hospira,Inc,LakeForest,IL)given
controlinseverelyburnedchildren.JTrauma2005;59:1148–54. forglucoselessthan60mg/dL;glucosebetween60and100
[11] JeschkeMG,KleinD,HerndonDN.Insulintreatmentimprovesthe
mg/dL resulted in holding of the infusion for one hour;
systemic inflammatory reaction to severe trauma. Ann Surg 2004;
239:553–60. whenglucosewasbetween101and100mg/dL,theinfusion
[12] HeuerJG,SharmaGR,ZhangT,etal.Effectsofhyperglycemiaand wasdecreasedbyoneunits;glucoselevelsbetween151and
insulin therapy on outcome in a hyperglycemic septic model of
200 prompted no change. Glucose levels from 201 to 300
criticalillness.JTrauma2006;60:865–72.
[13] EgiM,BellomoR,StachowskiE,etal.Variabilityofbloodglucose mg/dLand301to400mg/dLtriggeredarateincreaseof1
concentration and short-term mortality in critically ill patients. An- U, and a bolus of 2 and 4 U, respectively. Physician assis-
esthesiology2006;105:244–52.
tance was sought for glucose greater than 400 or less than
[14] WintergerstKA,BuckinghamB,GandrudL,etal.Associationofhy-
poglycemia,hyperglycemia,andglucosevariabilitywithmorbidityand 60 mg/dL.
deathinthepediatricintensivecareunit.Pediatrics2006;118:173–9.
[15] PidcokeHP,WanekSM,RohlederLS,etal.Glucosevariabilityis
associatedwithhighmortalityaftersevereburn.JTrauma2007;62: Insulin Protocol Effective From October 1, 2003 to
272(abstr). Present
[16] Kahn CR. Glucose homeostasis an insulin action. In: Becker KL,
Treatmentwithintravenouslyadministeredhumanregu-
editor. Principles and Practice of Endocrinology and Metabolism,
3rded.Philadelphia:LippincottWilliams&Wilkins;2001:1303–7. lar insulin (1 U/mL, Novo Nordisk) was initiated when
H.F.Pidcokeetal./TheAmericanJournalofSurgery194(2007)798–803 803
blood glucose surpassed 120 mg/ dL, at the dose of 1 U/h atedhepaticglucoseuptakeisknowntobedecreased,while
forglucose120–150mg/dL,2U/hifglucosewasbetween non–insulin-mediated glucose uptake in other tissues, such
151 and 200 mg/dL, and 4 U/h if initial glucose surpassed as the central nervous system, is increased. Hepatic insulin
200 mg/dL. The intravenous insulin infusion was titrated resistance has been shown to be associated with increased
accordingtothefollowingparameters:insulinwasheldand circulating levels of insulin-like growth factor (IGF)-bind-
25mLof50%dextrose(Hospira,Inc)givenforglucoseless ing protein-1, but of course IGF-binding protein-1 was not
than60mg/dL;glucosebetween60and79mg/dLresulted measured in this study.
in holding of the infusion until glucose normalized and Regardlessofthemechanismsinvolved,thisstudydem-
restarting at 50% of the previous rate; when glucose was onstratesthatglucoselevelsvarydiurnallyinburnpatients,
between80and110mg/dL,theinfusionwasadjustedby0 and that insulin requirements increased over time.
to .5 U in response to increasing and decreasing trends; I have 3 questions:
glucose levels between 111–150, 151–200, and 201–250
1. Werethesepatientssubjectedtoburnwoundexcision
mg/dL triggered rate increases of .5, 1, and 2 U, respec-
andgraftingduringthe7-daystudyperiodthat,dueto
tively. Physician assistance was sought for glucose greater
increased levels of metabolic stress, might have ex-
than 250 or less than 60 mg/dL [9].
plained the increased insulin requirement?
2. Further,inarelatedquestion,didglucoseintake,due
to the initiation of, or increase in, nutrition support
Discussion
increase over the 7-day study period?
Richard Barton, M.D. (Salt Lake City, UT): The au-
3. Based on the diurnal variation that you demonstrated
thors report a retrospective study in which they recorded
in glucose levels and insulin requirements, and as-
hourly blood glucose measurements and exogenously ad-
suming that this variation in glucose level might be
ministered insulin in 156 burn patients. They demonstrated
clinically significant in terms of infectious complica-
3 significant findings. First, the glucose reached a nadir at
tions, how would you recommend that we alter the
approximately 5:00 in the morning and peaked at approxi-
administration of nutrition support to minimize this
mately 5:00 in the evening, and that this diurnal pattern
variation?
persistedoverthe7-daystudyperiod.Second,theydemon-
stratedthatexogenouslyadministeredinsulindosefollowed Heather Pidcoke, M.D. (Fort Sam Houston, TX): To
asimilardiurnalpattern,butwiththe5-houroffsetthatthey answer the first question, the subjects did undergo excision
described, and, finally, they showed that despite diurnal andgrafting.Thismayhavecontributedtoincreasedstress;
variations in glucose and insulin requirements, that insulin however, burn patients who are severely injured are maxi-
requirements tended to increase steadily over the 7-day mally stressed. In answer to your question regarding nutri-
period in order to maintain target levels of glucose. The tion, we typically start patients on nutrition as early as
authors note that the diurnal variation in glucose seen in possibleandinthefirst7days,itispossiblethatitmayhave
burn patients parallels that seen in normal subjects. They takentheentire7daystogetthemuptogo,however,weget
suggestthatthesteadyincreaseininsulinrequirementover them up to go sooner than that. Your final question was
the 7-day period reflects decreasing insulin activity and/or regarding what we should do in response to the study.
growing insulin resistance. I think there are alternative ex- Becauseitisaretrospectivereview,Ithinkmorestudiesare
planations for this increasing insulin requirement. In the needed before we change treatment; however, it certainly
stressed patient, glucose is produced via hepatic gluconeo- poses the question as to whether continuous feeding is the
genesis and glycogenolysis, under the influence of cat- rightway—whatrightwaytofeedthepatientsandperhaps
echolamines, cortisol, and glucagon. Van den Berghe [1] a 12-hour at the daytime feeding and 12-hour at night
recentlydemonstrateddecreasinginsulinrequirementsover feeding might be something to study in order to see if
timeinapopulationofcriticallyillsurgicalpatients.Inthis patients do better because there have been studies docu-
populationofburnedpatients,increasedlevelsofmetabolic mentingthatinhealthypatientsatleast,theinsulinreceptor
stress, caused by infection and perhaps repeated operative is more sensitive during daylight hours. Again, I am not
procedures, might have been responsible for the increased recommending this in response to this study. It is a retro-
glucose production and insulin requirements. Further, it is spective review and the question needs further study.
important to remember that glucose transport occurs down
References
its concentration gradient, is facilitated by carrier proteins,
[1] VandenBergheG,WoutersPJ,BouillonR,etal.Outcomebenefitof
and occurs via insulin-mediated as well as non–insulin-
intensiveinsulintherapyinthecriticallyill:insulindoseversusgly-
mediated pathways. In critically ill patients, insulin-medi- cemiccontrol.CritCareMed2003;31:359–66.