' I For reprint orders. please contact: reprints@futuremedicine.com

Biomarker use in tailored combat casualty care

Modern war wounds are complex and primarily involve extremities. They require multiple operative interventions to achieve wound closure and begin rehabilitation. Current assessment of the suitability of surgical wound closure is based upon subjective methods coupled with a semiquantitative determination of the wound bacterial burden. Measurement of the systemic and local response to injury using inflammatory biomarkers may allow for accelerated wound closure and treatment of other combat-related morbidity. This article presents the introduction of personalized medicine into combat casualty care.

KEYWORDS: biomarker combat inflammation personalized medicine trauma war surgery wound

Biomarkers Modern day war surgery is dominared by d1e need ro manage open wounds that require delayed closure and consume signi6canr resources in the effons coward approprime closure. The abiliry co .~ubjecrivdy predicr which wounds wiU fai l is his­torically poor. Some wouuds that appear desrined for proper healing after debridement and remion­frce closure insread continue on ro dehiscence and require a return co surgical care. Similarly, wounds thar appear grossly abnormal, but arc biologically capable of healing, cndttre unnecessary surgical procedures wirh rheir corresponding risks and costs. Currendy, rhe decision ro dose a wound is based on the surgeon's subjecrive evaluation of rhe wound, including appearance of granulation tis­sue, assessmenr of vascttlarization and an absence of gross infeaion. These subjective measurements of wounds arc nororiously poor. Ideally, one would be able ro monitor a parienr's, or even an indi­vidual wouJ1d 's, progression through the phases of wound healing using biomarkers as objective measures predictive of wound outcome ro guide the riming of successful surgical closure. The role of biomarkers in wound healing has primar­ily been in the systemic evaluarion of .1 patient's overall clinical status. Serum biomarkers have been described for wound healing in d1ronically malnourished elderly or d 1ronically dcbilirared parienrswirh modest success and h:~vc been found inappropriate or ineffect ive in rraumaric acure injuries, as seen wirh combat casualries. For sucl1 patiems wirh complex trauma, appropri<~lc bio­markers oflocal (via wound efAuenr and wound rissue biopsies) and sysremic (vi:t serum) wound healing are on rhc horizon.

10.2217/BMM.10 13

In order ro undcrsrand rhe currem status of biomarkers in war-associated injuries , ir is imperative rhat one undersrands rhe hisrory of the rypes of wounds experienced during various wus, the ea rly work on biomarkers and, fin ally, rhe basic components oflocal wound healing.

Wartime injuries The number of wartime casualt ies dying from their wounds has steadily declined, from Bo/o in World War I (WW[) . 4.5% in World War II (\XIWU), 2.5% in Korea, 3.6% in Viern:lm and 2.1 % in Deserr Srorm [1] . The drastic decline in lethal wounds has been traded for a signifi­canr increase in rhe extremiry wounds seen by war surgeons. Exrremiry wounds account for the majoriry of rhe injuries (65%) followed by head and neck (15%), rhorax (10%), and abdo­men (7%) {2] . Warrime blast injuries can be classified as (J.4]:

• Primary: injury caused by blasr wave:

• Secondary: ball istic trauma rcsuliing from fragmenrarion of rhe weapon or envi ronmcm;

• Terriary: resulring from the displacemem of rhe vicrim or environmemal sr rucwres:

• Quau:m ary: burus, tQlcjns and radiologi.:. conraminarion.

In Iraq and Afghanistan, the majority of currcnr war wounds are caused by improvised explosive devices rhat predominarel}' induce secondary and teniary blast injuries, which contribute ro ext remity wounds being rhe mosr common combat injury in the modern warrime era (S] .

Biomarkers Mad. (2010) 4 (3). 465- 473

f.uture . .port of fil!lll mediCine...,

ISSN 1752-0363 465

Report Documentation Page Form ApprovedOMB 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 andmaintaining 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, ArlingtonVA 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 itdoes not display a currently valid OMB control number.

1. REPORT DATE 2010 2. REPORT TYPE

3. DATES COVERED 00-00-2010 to 00-00-2010

4. TITLE AND SUBTITLE Biomarker use in tailored combat casualty care

5a. CONTRACT NUMBER

5b. GRANT NUMBER

5c. PROGRAM ELEMENT NUMBER

6. AUTHOR(S) 5d. PROJECT NUMBER

5e. TASK NUMBER

5f. WORK UNIT NUMBER

7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) Naval Medical Research Center,Regenerative Medicine Department,503Robert Grant Avenue,Silver Spring,MD,20910

8. PERFORMING ORGANIZATIONREPORT NUMBER

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 ABSTRACT Same as

Report (SAR)

18. NUMBEROF PAGES

9

19a. NAME OFRESPONSIBLE PERSON

a. REPORT unclassified

b. ABSTRACT unclassified

c. THIS PAGE unclassified

Standard Form 298 (Rev. 8-98) Prescribed by ANSI Std Z39-18

- ' . ~

PERSPECTIYE~ . --

466

Brown, Safford, Caramanica & Elster

High-energy penetrating wounds, an example of secondary injuries, cause extensive mult i­sysrem rrauma wirh concomi tanr gross bacre­rial comaminar ion. Up ro 75% of these injuries are colonized or infccred with environmcnral bacrcria when the pariems arrive ar tcniary care facil irics f6) . T he level ofbacrerial conraminarion significantly alrers rhe abiliry of wounds ro heal or quali fy for flap coverage.

The managemenr of wa r-<~ ssociated injuries has been addressed throughour medical his­rory. O ne of rhe eaiiiesr paHerns comes from the Greek lirerarure describing a penerraring wou nd, where wound rnanagemem consisred of removing rhe arrow, cleansing rhe wound wirh warer, and applying analgesic and herbal medicines. j ohn Hunrer, a disringuished surgeon in the !8rh cenrury, who served as rhe British Surgeon General, lldvocared minimal su rgical debridement ofin tected wounds (i] . During rhe US Civil War, the concepr of removal of nccroric and infected rissuc was add ressed rhrough earlr amputation. Given rhe lack of anribiorics and the srrong associa tion of infccrion with morraliry, surgeons understood the risk associated wirh an infected limb. In WW!, the initial undem anding of infection and wound healing began ro surface when primary repair was :mempred in 224 gun­shor wounds at a casualry deari ng srarion {8] . A porrion of rhesc "\.'IOUnds were culrure negarive and healed. In rhe remaining infecred wounds, a variety ofbacreria were idenrified. Although a few of chose culturc-posirive wounds were able ro undergo primary wound closure, rhe presence of hemolytic srreproc~ccus was associated with fai led closure in 95% of rbem. During WWII, and wirh rlw advenr of ami microbia Is, the lir­erarure began ro focus on wound bacreriology, suscepribiliry of various 1nicrobes to rhe av:~ ilablt

anrimicrobial agenrs, and rhc critical nanrre of rime. These concepts of the inrerpl~y of infection and wound healing conrinucd to progress and direcr primary wound-closure decisions.

Quantitative bacterial counts as the initial biomarker of wound healing Causes of parhologic or clinical in fection include exogenous and endogenous factors (9]. O ne such col11 ribming f:tctor is simply the degree ofbacre­rial coloni1..1rion. Traumat ic wounds are noro­riouslr contaminarcd, especially in rhe setring of militllry tr~uma. For example, over 70o/o of wmbar wounds from Viernam were clini­cally infecred, in rl1at rhey comai ned bacterial coun ts greater than 105 colony-forming unirs per gram of rissue (CFU/g), where greater than

Biomarkers Med. (2010) 4(3)

105 CFU/g is considered clinical infec tion (10). A second key faccor is rhe lime from injury to trear­mem (I I! . When rhe mean rime from injury ro rrearmem is 2.2 h, wounds had bacrerial counrs ofl 01 CFU/g whereas a rimeof3 h led ro higher bacterial counrs of 102- 105 CFU/g, and a rime of 5.17 h conrained grearer rha n 105 CFU/g. T herefore, it has been shown th~r the greater rhe amount of ri me from injury ro rrearmenr, the more bacterial coloni1.ar ion occurs and, subsequently, the greater rhe risk of imminenr wound infection.

A direct relat ionsh ip between bacterial counts and wound healing has been established for many years (12]. T he direcr association was described in pressu re ulcers where hea li ng occurred only when the bacterial coums were less rhan 106/mJ IJ2! . The associat ion has also been demonstrated in many wound models. In a 1967 study of 50 granular ing wounds, grafr survival was 94% when bactt"ria l counrs were less rhan 101 CFU/g and 19o/o when bacterial counrs were greater rhan 10s CFU/g !13] . Similar resulrs were found in a 1968 delayed wound­closure study where topical :un.ibiorics were used for COJilrolling bacteria in the wound lr·il · In a review of the b:1crerial counrs of the 40 wounds. 28 of rhe 30 woumls conraining fewer rhan 10$ CfoU/g h:~d uncomplicarecl heal ing during delayed closu re; however, none of rhe wounds rhat had grearcr than 10$ CFU/g experienced successful closure {IS]. Following these swd ies, in a prospective evalu;~r ion of wound closure with quanritarive bncreriology, 89 ow of 93 wounds having less r h~ n 10) CFU/g progressed ro rapid uncomplicarecl he:~Jing {16] .

These studies led ro rhc umlersrandi ng of rhe effect ol' wound comaminarion on \vound lte-J.ling aud presented tht· quanritarion of bac­terial CFU/g as a predictor of a wound's :~bi l­

iry ro heal. However, we ,He now le:m1ing th:~r or her factors may also influence wound healing. Jnflammarion, rhe body's narural response ro injury, is caused by rhc signali ng of cyrokines and d1emokincs. These inlhmmarory rncdiarors c:w also be measured as a predictor of wound healing. T herefore, evalu:uion of rhc narural response to injury, in addirion ro bacrerial con­t ami n~t: ion, would porenr i:1lly further reline rhe undersranding of successful wound heal ing.

Wound healing Wound healing can be described as occurring in fou r phases: hemosrasis, in fl ammation, prolif­erarion and rnarurarion (17]. During hemosrasis, plarcler-med iared acrivarion of the i nr ri nsic

ft~ur~ SCience group IE

cloteing cascade and vasoconsrncuou resulr in clor form;uion. The clor forms the support for infihrating inflammarory cells :1s the plare­lers release inflammarory rnediawrs, including cyrokincs, chemokincs and growth facrors. Over rhc ensuing inflammatory phase, neurrophils, macrophages and .6broblam migrate inro rhe rissue guided by v-drious c.hc:moamactanrs and cywkines. Chcmoatrracranrs rhar a.re imporranr to rhe inflammatory phase include CXC and CC chemokines. and associared proinflammarory cyrokines include IL-l, TNF-a, TGF-P. plare­let factor 4 and leukorriene B4. The orcllcma­rion of rhcse facrors leads 10 rhe release of nirric oxide, oxygen free radicals, seriue proteases and m:m ix merallopr01einases (MMPs) that destroy bacn:ria, and clear damaged exu-Jcellubr matrix molecules and inflammarory debris [IS). In rhc proliferative phase, rhe wound begins epirheliza­rion, angiogenesis and m:urix form:uion. Finally, in m:uu rar ion, rhe wound u ndcrgoes com racrure and migrarion of epithelial cells from rhe periph­ery. The wound rnarri.x undergoes a change ro become a more organized srrucr urc of colbgc:n, proreoglyca n and fibronectin. The understand­ing and evalua£ion of rhese factOrs and rheir substrates, as well as the systemic response ro injury, will lend them uriliry as poremial bio­markers in rhe prediction of wound he:lling in wanime-associared injuries.

Current efforts to predict war wound outcome In our research program, our approach ro co moor casualty care is ro rake Ollr bedside underst.llld­ing of rhe wound healing process inro rhe labora­tory where we can qu:tnritare rhe biomolecular mediarors of rhe aforementioned phases. We then rerum ro the bedside with char knowledge as a decision-supporrive diagnosric wol for use by rhe surgical ream for din~cred parienr care in rbe spirit of translational medicine )191. Our inirial c:£forrs have focused on rhc idcmificarion of biomarkers for timing surgical wound closure to ensure h~ling and avoidance of dehiscence, Stratifying patient riskofheterowpic o sification and idcnrifying derrimenral microbial coloni­uuion. For rhe pu.rpose of rhis arridc. we will focus on rhe use of biomarkers of wound heal­ing ro illustrate a frontrunner rechnology for personalized care of our wounded war fighters.

Our swdies have enrolled wounded US service members who had susrained penetraring injuries ro one or more e.xucmirics, who did not have confounding immunologic comorbid conditions, and who had been evacuared ro the N:~riona l

Biomarker use in railored combar casualty care

Capiral Area from Iraq and Afghanistan. Surgical debridernenr, pulse lavage wound irrigarion and negarive pressure wound therapy w:~s repeated every 4S- 72 h unril wound closure or coverage, according ro currcnr insrirurional srandards of pracrice 120-22). In general. these parienrs were young (average age of23 years old, wirh a range between 18 and 37 years) and fi t (with an :wer­age BMI under 25). Therefore, we were afforded a fairly homogeneous, athletic population within which ro sllldy the biological response ro rraumatic, acme injury and idemify the early hall marks of wound heaJing versus dehiscence.

Today's cornbar casualty care typically involves rbe aggressive surgical care of complex injuries caused by blasrs and high-energy ballistics thar inAicr devastating r r:utm:t, often violating sofr tissue, bone and neurovascular srruccures, espe­cially in rhe exrremitic:s [5.lJ-17) . With the shift in modern war tactics rowards an increased use of imprO\·ised explosive devices, high-powered munitions and close-qu:~ner cornbar, war inju­ries and combat care increasingly differs from classical trauma managemem 14] . ln our recenr repon , we monitored 52 exrremiry wounds in 33 parients wi rh up ro rhree wounds per paricnr. The primary clinica l ourcomc: measure was suc­cessful wound healing within 30 days of defini ­tive closure or coverage with skin grafr wirhour necessi rar ing a rewrn to the operaring room, spomaneous partial or complcre wound disrup­rion afrer primary closure, or over 90% skin-grafr loss [20]. The wounds of that srudy covered an average surf.1ce area of241 cm1 (ranging berween 25 and 1729 cmZ) and had an aver-age volume of 369 em> (ranging berwecn 1.45 and 2119 em·'). There was a mean patient injury severity score (ISS) of20 (ranging berween 8 and 45). These represent rhe typical modern combar wound that reaches reniary care.

Trearmenr ofw~r wounds requires a signifi­cant arnoum of rime :tnd resources direcred roward cirher limb ~~~lvage or length preserva­tion in ampurarions. In addirion, up to 75% of rhese injuries arrive ar rerriary care military medical faci lities, such as rhe National Naval Medical Cenrer (MD, USA) or Walrer Reed Army Medical Center (DC, USA), colonized or infecred with fastidious environmemal bacte­ria such :IS Acillcttobnrter bnumnnnii flS.29J. Local amibioric delivery used in concc:rr with irriga­tion, debridcmenr and ncgarivc-prcssure wound rherapy have adv;tnced rhe rre.atmem of these injuries, but the basic surgical decision regarding rhc [iming of wound closure or Aap coverage of ~wound remains subjective 1211.

www.futurcmed<C <ne.com 467

PERSPECTrv(

468

Brown, Safford, Caraman ica & Elster

In the following sections, we describe our efforts ro define rhe role of biomarkers :moci­ated wich combat wound healing in this p:nienr population. Our initial pilot SttH.ly of procalcito­nin {ProCD and cytokinc expression at the rime of surgical wound closure is described, followed by a discussion of expanded studies invcsrig:H­ing rhe longitudinal inflammatory response and. finally, rhe role of remodeling prorcascs in com bar wounds.

• Role of ProCT & cytokines at the time of wound closure The utiliry of a well-defined biomarker profile of traumatic acme wounds rhar progress ro success­ful healing afrer surgical closure was introduced by our pilor work, reponed by Forsberg (t rrl., which invesrigated rhe correlation of ProCT and cyrokine expression wirh war-wound dehis­cence l2tJ. In that work, we posited til:ll a diag­nost ic biomarker pa nel predicrive.ofhealing, as opposed ro dehiscence, would be of great use in reducing the rate of wound fa ilu res caused by premarure surgical closures. In addirion, such a predictive diagnostic would reduce rhe number of surgical wound debridcmcnrs a parient would undergo. In either case, a diagnostic biomarker panel rhar was predicrive of healing would rr.sult in lower patient risk, reduced lengrh of stny in hospiral and a ElSfcr ret urn to dury. In an cfforr to provide an objective me:tsure of appropri:ne riming of rraumaric wound surgical closure, the Forsberg t:t nl. pilot smdy detected and quanti­fied ProCT and cytokines in wou nd effluent as well as serum w preemptively differenriarc rhc wouuds thar will hc~lunevcnrfully from those: at risk of deh iscencc.

Proc.Jcitonin and C-reactivc protein have: been investigarcd as correlates of inflamma­tion, complications and outcome in pa.rients wirh mulriple t rauma (30.3t]. Those srudies were primarilr concerned with determining rhe onset of septic complications w direct rrearmem. They reponed rhat C-reacrive pro­rein had lirde 10 no diagnostic value as ir was d evared owing ro rhe initial rrauma, demon­srr:ued only :1 slow decline in conccnrration over the course of treatmenr, and did not indiCl tc a significant difference between survivn l nnd nonsmvival groups. However, ProCT dclllon­srrarcd a slighr-to-moderare increase, which did correlate wirh severiry of mntm;l and followed a relatively rapid return tO baseline level in the absence of sepsis while remaining signi fi cantly ele\·ared or further incre:tsing at the momem or sepric complications.

Biomarkers Med. (2010) 4(3)

In agreemenr with rhese findings, and fu r­thering their applicarion ro the evaluation of rraumaric wound healing, Forsberg (t a/. lim reponed I h:~t elevated serum and emueut ProCT levels measured at rhe rime of wound closure correlated wirh later wound dehiscence (21J . As reponed, no wound failed wirh an effiuem ProCT concemration of less rhan 220 pg/ml, IL-13 concemrarion greater rhan 12 pg/ml or Regulated on Acrivation, Normal T Expressed and Secreted (RANTES) protein concenrrarion greater than I 000 pg/ml. lr was a l~o observed that there w;1s a la rger ratio of serum:efflucm ProCT concenrrarions lor wounds th:u healed (mean 114.2 pg/ml vs me:tn 36.01 pg/ml) com­pared wirh those rhat eventually clehisced (mean 560.R pg/ml vs 338.6 pg/ml). T hese inirial results highlighted rhe imporr:1ncc of both rhe local (effluem) and sysremic (se rum) response ro rrauma, and warranted an expanded study ro furrher explore effluem cytokine and chemo­kine an3lysis <tS objecrive, dccision-supponi\·c biomarkers predictive of wound healing.

• Inf lammatory response to injury & combat wounds In an expanded scudy, Hawkswonh t:t nl. demon­strated rhat kinetic e.xprcssion of inAammatorr cyrokines and chemokines are objectively associ­ated wirh acute war wound healing, and iJellli­ficd cyrokine and chcmokinc protein, a.nd gene transcripts wirh potential utilit)' as biomarkers for predicting wound outcome (summarized in T•ac£ 1) 120]. This study involved :111al}•sis of the 52 wounds described previously in this article and reponed rhar nine (17.3%) wounds across live pat ienrs dehisced; three of rhese patients deh isced all or rhcir wounds. Significant clini­cal detenninams for wound dehiscence included elevated ISS (p " 0.03). larger wound su rface area (p = 0.005) and volume (p = 0.004), and associated vascular injury (p = 0.002) (20].

While rhe clinical association berween sysrt.mic illness and impaired wound healing is cstlb­lished (.H .J3J. 1-!awkswonh era/. reponed a panel of markers that seemingly report some of the biological mechanisms responsible for he:1ling rraum:11ic injuries.

Using mu.lriplexed protein quamirarion of scrutn and efflucm combined with 4ua11t itative real-time PCR of wound biopsies, significant and predictive differences berwecn wounds tlaJr prog­ress ro healing and 1 hose th3r FJil within 30 clays of surgical closure are described. Wirh an :tnalysis across serial wound debridemems. receiver opcr­aring charACterisric curves for serum !L-6. llr8.

macrophage inflammatory protein (M lPHa, effluent If-N-y inducible protein 10 and !L-6 protein biomarkers were each st:u isrically prcdic­rive of wound dehiscence wirh favorable receiver operating characrcrisric curves (a rca u ndcr rhe curve of 0.805, 0.755, 0.695. 0.660 and 0.632 respectively; p < 0.05). The tissue gene transcripts MCP-1, IL-Ia. TN F-a, IL-B.MIP-Ia, GM-CSF, IL-JP and -6 were also individually srarisrically predictive of wound dehiscence (0.845, 0.845, 0.845, 0.832, 0.755, 0.744 . 0.729 and 0.662, respectively; p < 0.05) [20).

• Wound remodeling proteases are associated with wound failure The equilibrium of inflammarory mediators is crucial for rhe progression of acll[c wound heal­ing. lo the srudy by Hawksworrh ctllf., a srarc of systemic inflammatory dysregulacion was sup­porred by consisremly clevared ~erum prorein IL-6, -8 and MIP-la in paticms wi rh wound dehiscence. Such a prolonged inflammatory response involving an imporra nr rnediawr of rhe acure-phase response, and porcnt chemo­amacrams of macrophages and granulocytes is consisrem wirh a wounds delayed advance imo the fibroproliferative phase, in which collagen deposirion and rapid gains in wound-breaking srrengrh typically occur [20.!4].

In addition, MMPs play crucial roles in rhe inflammation, proliferation and mant ra· rion phases by performing several funcrions relared ro in8arnmarory signal ing (establishing chemotacric gradicnrs that di rect immune cell migrarion) :111d wounci remodeling (proces~ing

collagens and elastin) 135-371 . Elevared levels of rhe proteases MMP-2 (gdarinase) .wd MMP-3 {srromelysi n) and low levels of rhei r respective inhibitors have been measmeci in wound efAu­enr of ch ronic pressure nlcers rrcared wid1 ncg­arive-pressure wound d1erapy IJ8) . Furrhcrmore, disproporrionare expression of MMPs and rheir inhibirors has been proposed as a cause of wound chroniciry [3".39- 41] . By .lnalrzing concemrarions ofMMP-2, ·3. -7 (marrilysin), -9 (gelarinase) and -13 (collagenase) in parcnr serum and traum:ltic exrrcmiry wonnd efAuenr throughout the rreat­menr process, the work presenred by H:1w~vonh er nl. was extended ro include these representative MMPs as porenrial objective markers of healing and, furd1ermorc, as indicarors of ri ming for suc­cessful surgical wound closure and avoidance of dehiscence in rraumaric 3ClJ(C wounds f22J.

In rhe subser of wounds analrud (38 wounds in 25 parienrs; up ro three wounds per parienr). MMP expression was associated with impaired

• fU1Utr.: screr.cesroup

:::

.lib _:y. Biomarker use in railored combar casualty care ......:....___...... ......

Biomarker AUC

Inflammatory cytokines'

Tissue MCP·I 0.845 Tissue IL-1a. 0.845 Tissue TNF 0.845 Tissue IL-8 0.832 Serum IL·6 0.805 Serum IL-8 0.755 Tissue MIP-Ia. 0.755 Tissue GM-CSF 0.744 Tissue IL-1 p 0.729 Serum MIP·1a. 0.695 Tissue /L-6 0.662 Effluent IP-1 0 0.660 Effluent IL-6 0.632 Effluent IL-2 0.612.

Remodeling proteases'

Effluent MMP-3 0.805 Serum MMP-7 0.783 Serum MMP-2 0.744 Effluent MMP·9 0.661 Serum MMP·9 0.655 Effluent MMP·13 0.644 Effluent MMP-7 0.632 lnflommarory cyroktnes and remodeling pro ceases assodared with combat wound dehiscence. Tissue refers co Crunscripr allillysis for wound biopsies, serum refers ro protein analysis of serum samples and effluenr refers ro ptotein analysis of effluent from ncgarive-pressure WOIJnd therapy. 'Data taken from [U) 'Data taken from {2S]. AUC: Are<~ undl!r rhe curve; IP! lnterf£>ron·inducible protein; MIP. Macrophage inflammatory prore1i1; MMP: Matrix metalloprorein.l.Se.

wound healing, which was defined as wounds rhar deh isced or required delayed defini­rive wound closure of 21 days or more after injury (rwo srandard deviations ourside of rhe mean normal wound closure rime period of 10 days) (2£1 . h was found th:u proinflam· m:aory MMP-2 and -7 levels, sampled at each surgical debridemem, were staristically higher in patiems that demonsrrared impaired wound healing as opposed to pariems with wounds lhat healed normally (p < 0.001) (sec T•eL£ t).l n addi­tion. serum MMP-7 di ffered significanrly when grouped by vascular injury or ISS (2!J. Patienrs wirh impaired wound healing expressed sraristi­cally lowt:r levels of effluenr MMP-3, a wound resolution MMP, rhroughour rhe debridemem and healing process when compared wirh parienrs wirh wou nds thar healed nornully (p < 0.001). Finally, receiver operating charac[eristic curve a n<~ lysis revealed rhar serum MMP-2 ancl-7, and

www.fururemedicine.com 469

. . PERSPECTIVE,

>! t. ",:...~.

470

Brown, Safford, Caramanica & Elster

effluent MMP-3 were ead1 srarisrically predicrive of wound healing th roughour rhe iniri al rhree and final debridcmems prior ro surgicaJ wound dosure (area under rhe curvcof0.744, 0.783 and 0.805, respectively; p < 0.001) 122] .

In Utz ern/., rhe inrcrplar berween markers of prolonged sysremic response ro injury and local mediarors of the hea ling process is furrhcr i!lus­rrared. lr is poinred our thac while M MP-2's col­lagen and basement membwne deaving acriviry, as well as inmlUne cell rccruirmenr, are impor­ram ro this process, excessive anion has clear derrimenral consequences for wound healing. It is further elaborarcd rhar rhe consrirutively high expression of serum MMP-7 (involved in extra­cellular marrix processing, re-cpirhelializarion and neutrophil migration) in impaired heal­ing wounds is consistcnr wirh an exaggerarcd inflammarory response and a component of 'sralling' the healing process 122]. Ad mirredly, ir was nor clear wherher rhe elevared MMP-7 was directly linked ro impaired wound healing or if ir was insread reporcing rhe associated vascular injury. Nonerheless, it presents as a porenr ial classifier of normal healing wounds versus those rhar will require addirional medical am:nrion. Finally, rhe investigation also idenrificd MMP-3, an important componem of wound remodel­ing and resolurion, as elcvarcd in wounds with normal heal ing rares versus reduced levels in impaired wounds [22) .

• Using a biomarker panel for a bench-to-bedside approach Whereas the previously reviewed reporrs iden­tified bioma rkers rhar were individually pre­dictive of outcome, we would be remiss nor ro acknowledge rhar Statistically significant differ­ences in quanrifiahle biomarkers do not ncces· sari ly equ3re ro clinical pracricaliry or uriliry. As many of rhe ana lyres invc:srigared fUr usc as indices of wound healing or infecrion arc com­poncms of ovtrbpping biological processes, it can become exceedingly difficulr ro adequarely describe the par icm (or wound) for whom a siugle-analyte d i:~gnosric would funcrion prop­erly. I nsread, systems-biology approaches may be more appropriate in translaringuiomarkers from bench co bedside. One mer hod ro analyze dar:t in a sysrems-based approacl1 is by using advanced machine-le:trning algorithms.

Prel iminary work wirhin our rescard1 group has demonsrr:ued thar machine-learned Baresian belief networks, which ident ify condition;t l dependence berween variables and present rhi~ >uucrurc in a grapbic.1l format, ro pred ict

Biomarkers Med. (20t0) 4 (3)

acute war wound ourcome from an imegrared gene-expression panel could furrher enlnnce wound dassific:n ion IB~<o"'" T. EoLRitA!\DTj. E• srr.R f:.

u,.,..,,L,SIIr.u o ... r.,J. 1:-!awkswonh N nl. concludt:d rhar the cytokine and chemokine protein and gene uanscripr expression p:merns demonstrated a coudirion ofinflan1rnatory dysrcgu lation asso­ciated wirh war wound fn ilure [zo] . To move those findings closer ro rhe bedside, we hyporhesized thar a combined molecular biomarker panel may predict combat wound-healing ourcome when considered together in a single <llgori thm d~ssi­

fying the mulrivariare dependence relationships. Full Bayesian an,1lysis of cytokine and chemo­kinc expression allowed ror effective predictive model ing of rhe occurrence of" wound healing. In addition, rhe coordin:ned dependence reb ­rionships of our Bayesia n model corroborate the univarianr :111alysis wirh the addirion of an evi­dence-based risk of wound dehiscence predicrion (8110'11':< T. Ese><ttARUr j . l:u.1 bR E. LIN~UDLtSIIlD D., rA] .

Preliminary rnodding included rhe protein quamirarion for each wound ar lirsc. second, rhird and final surgical wound debridement as separate va riables, which were c;~regorized b)' disrriburion inro rhrec equal probabil iry densiry groups. Models were derived in stepwise irera­rions unci! rhe opri rn:1l network was idenri fied, as determined by cross-,·alidarion and quali­r:~rive assessmem aga inst cl inical experience and rhe lirerarure. To evaluate robusrness of rhe Bayesian models, each paricnr w;ts ser i~lly

excluded from rhe darascr in a leave-one-out cross-validarion 1421 .

Mulrivariare models classifying he:lling or microbial infection in complex t raumaric wounds will need ro be validated in random­ized clinical ~rials in which model-based patient care is performed a ncl rhe response of rhc model ro :tSSociared changes in care (e.g .• earl ier surgi­ca l wound closures or admi nistration of anri­biQ[ics) can be evaluated. Owing ro rhe com­plexity of r hese rypcs ofinjuries compounded by a prevalence of microbial infecrion, we bel ieve that advances in the field of wounded warrime ca re will come fro1n Slll:h inregrared panels of patient-specific markers of wound healing.

Conclusion The assessmenr of rhe suitabi li ry for combat wound closure is evolving from subjccrive clini­cal observation and assessmem of mic.robial sta­nrs using early 20rlr cemury cul[lJ rC rnerhods ro an understanding of rhc infla mmarory envi ron­ment in which wound healing occttrs. T his shifr h.ts occurred wirh a change in combar-rdated

future sctenee group IE

morbidiry towards larger extrcmiry wounds in parieurs rhat are more acurcly ill. While our group has begun tO describe the rclarionship bcrween the inflammatory response ro trauma and rhe ability to prcdiCI wound healing in com­hal casuahics, i1 is nor 10 be assumed unique w rhis parienr popularion as orhers have dem­onsrrared a significant link between sysrcmic inflammation and semm cyrokines wirh clinica.l omcome (43j. Many of che cyrolcines seen in such studies of civilian rrauma patients overlap with ones measured in combat-wounded pariems, sug­gesting rhe miliry of this approach beyond 1hc military-specific injury.

The ability co assess a paticnr's inflamma­wry response, borh systemically and locaUy, and develop decision-support tools based on these bio­markers inrroduces the use of personalized medi­cine in rhe rreatment of rrauma pmienrs based on d1e individual's biology rather than the classical prescriprion rhat is 'bcsr for mosr' or popularion­ba.sed care. While still a developing approach to di:lgnosis and treacmem of illness and disease, persona.lized medicine is becoming a pivotal mechanism by wh icll docwrs 1 rear paricnrs. In the trcarmenr of disease, personalized medicine uses the generic make-up of an individual, both static (as in cheir DNA sequence) and dynamic (as in DNA modifications or expressed RNA and proteins), ro de1ermine whicll rrearmem approach wiU be mosr dfeclive for an ill ness. As we have presenred here, measuremenr of llfl individual's cycok.inc, chemokine, MMPs, and ProCT expres­sion in response ro injury may, in the ncar furure, direct rraum:uic wound-trc:mnenr decis ion~ .

T herefore, by usi ng a personalized medicine approach, clinicians can predetermine rhc mosr eiTccrive rrearmenr for rhe most beneficial our­come per palicm insread of tremmcnrs hased on be.nefi1s for a genera lized popul:uion.

\Yle concede rhat rhe currem me1hod of derer­nuning wound closure is often successful. Thar is. an experienced surgeon can observe a wound and objecrively evaluate cllaracrcrisrics such as size. color and efflucm ro subjec1ivcly prcdicr wherher more debridemenrs arc nccess:~ rr or if rbe wound will close successfully. Using biomarkers as~ pre­dictive mechanism, however, is expeCLed to nor only allow these decisions robe made by less expe­rienced surgeons, bur to both incrc:1se ch~ success race of wound closures and reduce the number of operaring room visi1s per wound prior to dccid· ing co perform sucl1 surgical closures. These ea rly seeps, with regards to wound closure. will allow for rhe assessmcm of several other aspccrs of trauma care thai are memioncd l:uer.

• futurr.5>eiertt:egroup

Biomarker use in railored combar casualry care

Future perspective An objective decision on wound closu re. based on changes in an individual 's own bio­logy, provides ~ clear benefi1 for rhe rrcaunem of trauma patienrs in rhe ncar rerm. But rhis mcrhod could prove successful for other indi­carions as well. As wound fa ilure has been associared wirh other patienr ourcomes, apply­ing this dara to orher clinical scenarios could guide rherapeurics. Pneumonia, for example, is an ourcome commonly associated wid1 parienrs wirh rraurmric wounds. If a biomarker p:wel of protein ~nd gene e.xpression can be linked with rhe devclopmenr of pneumonia, doctors will be empowered for early imervcmion by a predicrion of a pa1ienr's risk for comracri ng rhc disease. Orhcr possible disease processes rh:u can be addressed in a si111il:tr f<'lshion include rraumaric brain injury, hereroropic ossificarions and infecrious complicarions.

Ideally, this predictive mechanism wi ll bt applicable for rnulriplc clinical indicarors wirhin any single patienr. Wid1 mu ltiplexed biomarker panels associa1ed wid1 clinical omcomes, clini­cians could assay mulriple rargers wirh a si11gle paciem sample. Nor only will rh is reduce the burden on the paricnr by requiring fewer sample collections, bur coSIS wi ll also be decreased by requiring fewer resrs.

The abiliry ro scan for mulriple clinical problems could also lead ro improved care of wounded patienrs. By predicring wound our­come, some p:nicms rnay rxpc:rience ~ reduced number of operative procedures (i rrigarion and debridcmcms) prior to wound closure and :1 shorrcr hospir:tf stay. Conversely, predicr ing when a wound is not ready for closure willl llow hospital s1aff ro amici pate and prepare for a lon­ger sray in rhe intensive care unit for thai pariem. Similarly, by predicring 01her ourcomes in add i­lion ro successful wound closure, doctors can anricipatc wllcther specific clinical complications will inAuence a paricm's rreatrncnr plan.

In wars of rhe furure, paricnts wi ll experi­ence improved and accelerated care. Wounded pariems wi ll cmcr any hospira.l and clinicians wi ll be able co make a series of predicrions using only a cissue, seru m or cfAucnr sample. Analyzing a series ofbiornnrkcr panels in a single assay will allow doctors ro objectively decide to

close a wound in one palicnr or 10 use systemic 1 hcrapy to prevcm secondary cornplicarions for anorher paticnc. These rherapeutic advances will provide for improved paticnr care and quicker reLUrn ro dury for our injured fighrers. As with l ll previous conflicts, rhe lessons learned from

www.futuremedicln~.com 471

PERSPECTIVE Brown, Safford, Caramanica & Elster

currenr ancl furure wJrs wi ll provide a remplare for rhc uearmenr of civi l i:~n injuries and the next conflict.

Disclosure

Tht virws rxprrJud in this 11rriclr art r!Jolf oftlJr m11bors

trnd do nor riflm s!Jr oflirial policy or tlu pnsirion of r/;r

Drpamnrnr ofrhr Nnr7, Drptmmmr oftbr Amry. rhr

Drpartnunr ofD1tmc or rbr US Go11mrmr111. Tbis work

WIIJ pr~paud 1/J pnrr ofrhrl7 offirinl durin

Financial & competing interests dis closure

Th~ amborJ har·~ no u/~u.rnr uj)iliariom or finanrur/

im•ohwnmrwirb any organi.t.arion or mriry ruirb a finan ·

rial inrrrrst in or finanri,r/ ,•onftirr u:irh rJ,r mbjur m11rru

or martriab diuumd in rlu mrurusrripr. Tim inrludrr

rmploymmt . .-onmltann'n, h11noraria, ,u;rk ortmasl;ip or

optiotu. <'>-f!trr usrimony. ::rn1111 or pmnrtJ rarivrd or

pmding. or royalti(.<.

No writing IIJJiJtanrt wtrs utiliud rrtt!u produrtion of

rbis mnnusrripr.

• As the monality rate in combat-wounded patients has decreased, the severiry of inJUries presented to military-care fac1lit1es has mcreased

m current conflicts.

• Past decisions regard1ng combat wound closure were made wnh a subjective assessment of the wound coupled w nh an assessment of

its bacterial burden.

• Wound failure 1n combat-wounded patients is dependent on the local and system1c inflamma10ry response to InJury coupled w1th the

bacterial burden. • Assessment of thrs response using serum and wound-effluent biomarkers may allow for the development of pred1ct1ve assays for the

timing of wound closure. • Development of personalized medtcine for combat-mjured patients has utility beyond wound care and will be deployable to C1V11ian

Bibliography Srokrs A, Tyler \VI: Primary and delayed 17 Srojadinovic: A, Ca rhtllr J, Sclruhl (;, prirnury sut ure ofb•un!lror wounds: report •)n Davis T, Elncr E: 'fi>pica l :~dv,>nc~s in

Pap~> I of spcci~l norc have hccn highlighted as: rhe bacteriology of wounds. Br.j . S11'1(. wound c;.1rc. Gyruro/. Onrol. Ill. S?O- SaO • of i ntcre>t •• of considerable interest

19. n-12t1 (1918). (20011) .

Mabry R, Holcomb), B;tkcr A tt nl.: ? Robson M: Wound infection: a fa ilure of wound IS Broughton G, )anis J, 1\ttingcr CE:

United St~tcs Army Rangers in Somalia: healing cuued bpn imb.tltncc ofbacwia. The buic science c>f wound hcal u•ll.·

an analy~is of conrb>t et>ualtic:s un an S11rg. Gin. Norr/J Am. 77(3). 637-650 ( 199n. l'um. Rrronstr. SnrK 117, Sll-534 (20011).

urban bmldidd.j. Tra:.ma49, 515-528 10 Kov31icJ, Mauumoto T, Dobek A, Ham it H: 19 Siuig OF. Wright A, O.hcrofT JA rr ttl.

(2000). Bacterial flor.t of one hundred and rwdvc Gllnd challenges in clinir:.tl deci,ion

1 .\lumy C· Epidemiology of infectinm comb.u wound>. MiL Mtd, 133, 622- 624 support.}. Biomtd. Inform. 41(2 ), 387-3!12

•.\SOCiatrJ with comlnt-rdatcd injuries in Iraq (1968). (2008).

and AfgiLmistan.}. Trrwour64, S232-S23S II Roluon M. Duke WI, Krirck T: R~pid Top-10 lisr of modern rC<JUIItments for

(200!1). bacterial screening in the trc:umcnt of civilian d inic:U decision supporr C3p2bilitics.

Ritenour A, Bukin T: Primary blast inju rt: wounru.}. Surs. Rn. 14(5), 426-430 (1973).

20 Hawksworrh JS, Stopdinov>c A, !.,;age FA

update on diagnosi> and ucarmcnt. SemiJ>:tl description of the rolr ofuioburdcn trn/. : lnfhrnrnJIOry hiorn.11kc11 in cuonh~t

Crit. Can Mttl. 31i(7), S311-S~l 7 (2008). ond wound healing. wound lw:rling. Amr. Surs.

.( Salroraf;u GH, Peru~ G: Principle• of '"u 12 Bently R, Nuccin P, Wolfe E tr a/.: 250(6), 1002-1007 (2009).

M11gcry: cu rrtnt cunccpt~ n nd fn ru rc Rc.bt ic)rl>loip of quantitative wound h•ctcri.\1 Report on thr relat ioJ>ship between the perspective's. Am.}. Emrrg. Mtd. 26(4). count.I IO hcalint: of ciccubiti. Effw of topical inflammatory resporue and wound healing. 480-489 (2()011). gentamicin. Amimirrob. Agmts Cltrmo. 71ur.

21 Fnr;bcr~JA, Elster E.i\, Ander~cn RC n nl.:

lncersting 3rticlc juxaposing rhe special 4, 147-155 (1964).

Correlation of proclciwnin and crtokinc

chall enge~ of war surgery with l:l Kri·rck T, RolbS<rn M, Khn E: (\actc:rial expression with ddtisc.:ne<: of warrime

clastiC2It rum~lology. j(mwth and skin gr:rft >urvival Surg. For11111 octremiry wound; } Bonr Jorm Surg. Am.

Pnel T, Wenner K, PrierS, Weber .\1, Ill, 511l 519 (I'J67). 90(3). 580- 5!18 (200tj).

Leveridge A, McAtee S: A U.S. Army I~ Rub<con M, Lea C, Dalton J, Hegge!') J· 22 Un ER. Ebrcr EA. TJdJt..i OK rt ,f.:

Forw:trd Surgic;,l T<::lon'$ ~xpcrience in Quantil'2tivr b3cteriology 3nd ddayrd wound MeullcJpwrrinJSc cxprc»inn i.> >lo.\c>ei•rcd

Opcr:rrion lr:rqi Fr<cdnm.} Trarmrn )7, dosurr. Surg, FommlO, 501-501 (1968). with tr:runuti~ WC>und (ailurc.j. Surg. Rti.

201-207 (2004). 15 I bwkswnrth J, K.upcr MIt Gage F ttal.: 159(2), 633-639 (2010)

~ Davis K, MorJn K. McAllister C, Gr:ry P: Quantirarivr bactcriolot;}'ond Wclund failure 2.1 Covey DC: nla.t and t'r:rgment injuries of

Multidrug-rcsi\IJnnt A.-itmo/Marrcxtrcmiry in high energy penetrating wu t.he mu...:ulo;kdetal >)'stem.}. o.,lt J•urr infectiurts in ~<>idiers. Enurg. !n[trr. Du. injurie>.}. Surgirrrl Rn 144(2), 343 (2008). Srorg. Aon. 84-A(?) , 122 1- 1234 (2002).

II. 1218- 1224 (2005). 16 Roll.\on M. Shaw R, Heg~;er. J: The rcclosure 24 OePalom RG , Burris DC, Champion HR.

Pruitt BA: CombJl cuua1ty care and surgical of postopcr.rtivr inci>ionalob>CC>~<S ua>cd On Hodgson MJ : OIHr injuries.

progrc£.~ . Ann. Surg. 243(11). 715- 719 bact< rial quancific.rt ion ol the wound. N. Engl.}. Mrd. 352(11), l.US-1342

(2006). Arm. Snrl!. 171, 279-2!12 (1970). (2005).

472 Biomarkers Med. (2010) 4(3) fc~urv scoenc: v. !Jfoup (II

~ .

Biomarker use in tai lored combat casualty care P.ERSHECTIVE J ' -· ,..

!S Montgomery SP, S"iccki C\YJ, Shriver CD: 30 C~>"tclli Gr . Pogruni C. Ciu M. Palldini R: 3i Li J, Chen J. Ki<>n<r R: Pothophysiology of

The cvalwuion of cosu31tin from Opcr:uiun PrOCIItitnnin as 3 prognostic lod di:lgnostic lCUtc wnund hclling. Clm. o~rmarol.

lmqi Fr<edom on return to the conti nenta l tool for septic compl ic:trion~ after major 25(1), 9-18 (2007).

United Stlr<s frum Mn rch ru Ju ne 2003. trnum>. Crit. G:rrMtd. 37(6), 1845- 1849 3H Stech miller JK. Kilpadi DV. Childrc.s B. }. Am. Cc/1 Surg. 201(1 ), 7-12: discussion (2009). Schuht GS: Effcct •lf,·ucuum·a>sistcd clo~ure 12-13 (2005). 31 Meisner M, Adina H, Schmidt J: Correlation thaapy on the exprc»ion t>f cyt<>kine. Jnd

R<vicw of the c3rly uptri<nce of combat of procalcitonin and C-rc:tctivc poncin 10 protc~scs in wound fluid of adult> With

cuuahy from Ope~t ion Iraqi Freedom ~nd inllammotion. complic:llicms, and ourcomc pres.<u re ulcer,. Wormd Rtpa;r R~rn

Ope~tion Enduring freedom at W:alr<r during rhc intensive C3r< unit course of 14(3). 37 1·-374 (2006).

Reed Army Medical Ccnter. multiplc-trnumn patients. Cru. Cllrt 3~ l3bkytny R, Jude E: The molecubr biology of

26 Peoples GE, Jczior JR, Shriver CD: Cllin& JO(I J, ru (2006). chronic wound~ •nd delayed htllin~; in

for the wounded in I rnq - a photo CS$:1)'. 32 Clark MA, Plank LD, Hill GL: W.Jund diab<:tcs. Dmbn. Mrd 23(6). 59-i (,()B (2006)

N. Engl.}. Mtd. 351 (24). 2476-2480 (2004). hC'31ing ;usoci3tcd with se,•erc sur~:icol illness. , o Erning SA. Krieg T. Da\•idson J.\1 ;

cxcdl~nt per~pecti \•e manuscript on the lVor/4}. Surg. 24(6), 648- 654 (200()). lnAa1111n3tiou in wound acpair: mokcubr

currem conRict. 33 Good$0n WH 3rd, Lopez·S~rm icmo A. ~nd cellular rncchJnism>. j. fmmt.

27 Owens BD, !\ugh JF Jr. Wenke JC, JenscnjA, Ww ] . Gr:~nja-Men~ L, Drrmntol. 127(3}, ) 14-525 (2007)

Macaitis J, Wade CE, Holcomb JB: Comblt Chavez..Emclla J: The inRuencc of • hricf 41 Menke NB, Ward tat, Wiuen T M,

wounru in Opcr:uion I rJqi Froedom and preoperative illncS! on pnnopcm•vc healing. l3onchc,· DC. Diegrlmonn RF: Impaired

Opcrnrion Enduring Freedom.}. Traumn Ann. Surg. 205(3). 250-255 (191!7). \\uund healing. Clm. Dmnu:ol. 25(1), 19-15

64(2). 295- 299 {2008). )4 Dubay DA. Fran~ MC: Acute wound (2007).

2M Centers for Disease Control • nd Prevention healing: the biology of acute wouncl fai lur<. 42 Koh.IVi R: A >tudr of cro.s-,-aliduion and

(CDC): Ar:nttobnrr" baum:muii Surg. C/in. NorthAm. 83(3). 463-481 lutomrop foroccu rncy e>timatit)n and rnudel

infections among patients at miliro ry mcdiC31 (2003). ,eJection. Pmcct-din.g\ of the 14th lnunumon11/

fJctlitics ll<•lling injured US sen ·icc mcmh~r.. , 35 Armstronl\ DG, Jude EB: T he rule of m21rix Joint Ccnfuro.·r nn Artifiriallnrdligmtr.

2002-2004. MM\fi R M orb. Morral. Wkl] mculloprotcin3:.e~ in wounc! healing.}. Am. 1-.•lontrtll, C2n;~da, 20-25 August 19~5.

Rtp. 53(15). 1063-101\6 (2004). Podi11rr. Mul. Assoc. ?2(1), 12-IH (2002). 4.~ )Ht row KM 3rJ, Cunt.tlc1 EA, McG uire NIF

29 Davis KA, Mur.n K t\ . MeA) liner CK, .~6 Gill SE, Parks \XIC: Mculloprotcinascs n ,;/,; Earl!· crwkinc production risk

Gray PJ : Multidrug-rcsistant Arirmnbamr 2nd thcir inhibito~ r<-gul~tCir.. of \\Ound "ratitio tr:wma p>tienu for mu ltiple urg>n

cxtr.:mit)' infections in soldiers. hc31ing.lru. j. Biodum. Cr/1 Bioi. 40(6-7) , f~ilurc.j. Am. Coli. Surg. 2M(3) . 320-331

Emtrr; lnfm. Dii. II (!!), 1211!- Jll4 (2005). I 334- 1 347 (2008). (2009)

www fut wern~diclne.Cofll 473