systemic responses to trauma - semantic scholar...systemic responses to trauma b a foex emergency...

Systemic responses to trauma

B A FoexEmergency Medicine, University of Manchester, Hope Hospital, Salford, UK

The systemic responses to trauma can be divided into cardiovascular, immuno-logical, and metabolic. The cardiovascular responses are seen immediately after atraumatic insult. The pattern of response depends on whether the insult is mainlyhaemorrhagic, tissue damage, or a combination of the two. The response may bequite different for penetrating vascular trauma, compared with a crush injury to alimb. The immunological, or inflammatory, consequences of trauma usuallybecome apparent several hours or days after the initial insult, although it isincreasingly clear that they may be triggered by the very early cardiovascularchanges. These have been implicated in the development of multiple organfailure. The metabolic responses are of greatest importance in the longer term:after successful resuscitation and after the definitive treatment of the patient'sinjuries. The metabolic responses need to be taken into account during therecovery from treatment and during the rehabilitation of the patient.

Correspondence toMr Bernard A FoSx,

Emergency Medicine,Clinical Sciences

Building, Hope Hospital,Salford M6 8HD, UK

The aims of this chapter are to present an overview of the physiologicalresponses to trauma. These can be divided for practical purposes intothree: cardiovascular, immunological, and metabolic. This chapter willcover the first two, while the third: the metabolic response to trauma iscovered in a separate volume of the British Medical Bulletin}.

In the 1970s and early 1980s, mortality from trauma was described asfollowing a trimodal distribution2-3. Deaths occurred either immediately,or within the first few hours, or much later (days, even weeks, after theaccident). A better understanding of the early pathophysiology oftrauma and improvements in the early management of trauma patientsseems to have resulted in a reduction in the early deaths45. In someareas, the trimodal distribution of deaths has been lost6-7. Continuedadvances in the management of trauma depend on a good understandingof the systemic responses.

Studying the response to trauma in a clinical setting is often difficult: theurgency of resuscitation makes extensive physiological assessmentimpossible and the resuscitation itself will alter the responses observed. Forthis reason, much of our understanding of the responses to trauma comesfrom laboratory studies using human volunteers and animal models.

British Medical Bulletin 1999, 55 (No 4) 726-743 C The British Council 1999


The cardiovascular response to trauma

In a physiological sense, trauma is not a single insult: it is a combinationof haemorrhage, tissue injury, pain, and fear. To understand the physio-logy of trauma these components have often been studied in isolation.

The cardiovascular response to haemorrhage

The biphasic response to haemorrhageSystematic research into the cardiovascular responses to haemorrhagebegan in the Second World War, when extensive blood donor programmeswere established. It was found that a small percentage of blood donorslost consciousness following even a small haemorrhage, and that thepercentage increased as the volume of venesection increased. Thisobservation was studied by venesecting volunteers and monitoring heartrate, blood pressure, cardiac output, right atrial pressure and fore-armblood flow8. Initially, the response to haemorrhage was an increase inheart rate and total peripheral vascular resistance so that, despite a fall incardiac output, blood pressure was maintained (Fig. 1). However, onceabout 1000 ml of blood had been removed, there was a sudden fall inblood pressure associated with a bradycardia and syncope. This wasfound to coincide with a profound increase in fore-arm blood flow and areduction in systemic vascular resistance. These changes could be largelyreversed by re-infusion of the shed blood.

BAROREFLEX

Fig. 1 Biphasicresponse to haemor-rhage. Faint induced

by venesectionshowing the biphasic

response in heart rate(HR), blood pressure

(SBP) and systemicvascular resistance

(SVR) with a gradualreduction in cardiac

output (CO). Adaptedfrom Barcroft etal"

170%

100%

100 |mmHg

(1/min)

DEPRESSORREFLEX

Syncope

SVR (%)

HR(bpm)

SBP (mmHg)

CO (1/min)

0 5 10 15Time (min) venesection

British Medical Bulletin 1999,55 (No 4) 727

Trauma

It appeared that the response to simple haemorrhage consisted of twodistinct phases: an initial phase of tachycardia and increased vascularresistance maintaining arterial pressure, followed by a phase of decom-pensation with hypotension and bradycardia. These observations havebeen confirmed clinically9'10, in simulated hypovolaemia11'12, and in animalexperiments13"15. The bradycardia of haemorrhagic shock is reviewed bySecher and Bie16. The precise mechanisms controlling this biphasicresponse to simple haemorrhage have been extensively reviewed17. Only abrief outline of two of the reflexes involved will be provided here.

The arterial baroreceptor reflexA relatively small haemorrhage (up to 10-15% of total blood volume(TBV)) results in an unloading of arterial baroreceptors located in theaortic arch and carotid sinuses. The result is a reduction in cardiac vagalactivity and an increase in sympathetic stimulation of the heart. At thesame time, there is an increase in sympathetic vasoconstrictor tone andan increase in total peripheral vascular resistance so that arterial bloodpressure and perfusion to tissues critically dependent on oxygen supplyare maintained despite a reduction in cardiac output.

The 'depressor" reflexWhen blood loss exceeds a critical volume (usually around 20% TBV),hypotension and bradycardia are seen. This is due to the activation ofanother reflex, referred to as the 'depressor' reflex, which appears to over-ride the baroreceptor reflex. The efferent limb of this 'depressor' reflex iscarried in the vagus (increased activation leading to bradycardia, whichcan be blocked by atropine11) and the sympathetic vasoconstrictor nerves(decreased activation leading to vasodilatation). The precise nature of theafferent limb of the 'depressor' reflex is still unclear.

The triphasic response to haemorrhageA recent animal study showed that after the bradycardic phase ofhaemorrhage there was a massive increase in heart rate as blood lossreached 44% of total blood volume. This was associated with a furtherreduction in mean arterial pressure14.

Jacobsen and Secher observed a similar triphasic response in a seriesof patients in haemorrhagic shock18. No patients in the bradycardicphase died, suggesting that this phase of haemorrhagic shock isreversible with prompt fluid resuscitation. However, once patients gointo a phase of tachycardia and hypotension, shock may be irreversible.This third and potentially irreversible phase of haemorrhagic shock isassociated with increased sympathetic activity. The precise mechanisms

728 British Medical Bulletin 1999,55 (No 4)


triggering this have not been clearly elucidated, but it may be related tothe development of cerebral hypoperfusion and cerebral hypoxia.

The cardiovascular response to tissue injury

The response to tissue injury consists mainly of a pressor response: anincrease in blood pressure mediated by an increase in sympatheticvasoconstrictor tone producing an increase in peripheral vascularresistance19. The tissue injury response is accompanied by a tachycardia:there is a suppression of the baroreflex20 which would normally result ina bradycardia.

A similar response is seen when the 'defence area' of the brain isstimulated electrically21"22. Stimulation of the 'defence area' provokes the'defence reaction' the features of which are: (i) a tachycardia (n) anincrease in blood pressure; (ni) an increase in sympathetic efferentactivity; and (iv) a relative vasodilatation to skeletal muscle withvasoconstriction in the splanchmc vasculature. In contrast to theresponse to haemorrhage, which preserves blood flow to vital organs,the response to injury prepares the organism for 'fight or flight' and sodiverts blood away from the visceral organs to skeletal muscle. Thesimilarities between these two responses has resulted in the suggestionthat the response to tissue injury is mediated via the same pathways asthe defence reaction. A more detailed explanation of the centralpathways can be found in Kirkman and Little17.

The cardiovascular response to combined haemorrhage and tissue injury

Historically, haemorrhage was regarded as a complication of woundshock. When Parsons and Phemister realised that blood loss couldaccount for wound shock they also noticed that a minor degree ofhaemorrhage could be fatal if associated with tissue injury23. Thisinteraction was studied later by Wang et aP4. In a canine model ofhaemorrhage and haemorrhage and soft tissue injury, there was a highermortality in the haemorrhage and soft tissue injury group. As thecirculating volume lost was the same in both groups they concluded thatanother factor was involved. In further studies, they found similarincreases in mortality when sciatic nerve stimulation or soft tissue injurywere added to then" haemorrhage model25. From this, they deduced thatafferent nerve stimulation was an important factor in the increasedmortality of traumatic shock compared with simple haemorrhagic shock.

A variety of models has been used to study the interaction betweenhaemorrhage and tissue injury. Figure 2 compares the cardiovascular


Trauma

response to haemorrhage with and without concomitant tissue injury,modelled using hind-limb ischaemia13. A number of differences areapparent. The background limb ischaemia results in a higher resting HRand MAP. It appears that there is a re-setting of the baroreceptor reflexto a higher MAP and there is also a reduction in the sensitivity of thereflex. Coote et al showed that somatic afferent nerve stimulation canmodulate the baroreceptor reflex21-26.

A similar phenomenon has been recorded clinically by Anderson etaP°, who found that patients suffering moderate injuries, such as longbone fractures, showed a reduction in baroreflex receptor sensitivitywithin a few hours of their injury. The reflex had not returned to itsnormal sensitivity even after 14 days. Driscoll also suggested that therewas a reduction in baroreflex sensitivity and a right shift in baroreflexsetting in patients with significant tissue injury in addition to bloodloss10. Figure 2 also shows that, in those animals subjected to hind-limbischaemia, there is no bradycardia as haemorrhage progresses. Such abradycardia is normally mediated by the 'depressor' reflex and anincreased vagal tone to the heart. Thus, tissue injury also seems tointerfere with this reflex pathway.

It is clear that the cardiovascular responses to haemorrhage and tissueinjury are quite different. However, central integration of the variousreflex pathways is such that the cardiovascular response, when the twoconditions occur simultaneously, resembles the response to tissue injurymore than haemorrhage.

550

Fig. 2 Cardiovascularresponse to haemor-rhage and hind-limb

ischaemia. The figureshows the effects ofprogressive haemor-

rhage (Haem), andthe effects of

haemorrhage in thepresence of bilateralhind-limb ischaemia(H+LI), on heart rate

(HR), in the upperpanel, and mean

arterial bloodpressure (MAP), in the

lower panel, in theconscious rat. Values

are expressed asmeans ± SE. Adaptedfrom Kirkman etaP1.

R(b

pm

X

'SB

1

400

250

130

90

50

H+LI

Haem

H+LIHaem

H 1 1 1 1 1 1

0 15 35

Haemorrhage (% TBV)



Consequences of haemorrhage and tissue injury: oxygen transport

The cardiovascular responses are accompanied by important changes inoxygen transport. Rady et aP7 found that, in the anaesthetised pig,oxygen delivery (DO2I) was halved by a 40% TBV haemorrhage whileoxygen consumption (VO2I) was maintained. This was achieved by anincrease in oxygen extraction ratio (OER) from 30% to 67% and a fallin mixed venous oxygen saturation (SvO2) from 72% to 34%. Whenhaemorrhage occurred on a background of somatic afferent nervestimulation (SANS - to mimic tissue injury) DO2I was reduced to 27%.At the same time, VO2I fell to 53% despite an increase in OER from37% to 81% and a fall in SvO2 from 66% to 20%. Somatic afferentnerve stimulation itself was found to have no effect on either DO2I orVO2I, although it did produce the expected pressor response to tissueinjury (tachycardia, increase in arterial blood pressure andj increase insystemic vascular resistance).

A later study28 showed that skeletal muscle injury (SMI) had aqualitatively similar detrimental effect. The precise mechanism for thedeleterious effects of SANS and SMI are not clear. It may be that SANSor tissue injury causes an increase in critical oxygen delivery (the level ofDO2 below which VO2 becomes supply-dependent) as has beendemonstrated by Kirkman et al29, which, in turn, may be dependent onthe redistribution of blood flow away from certain metabolically activeorgans, such as the gut, towards more inactive skeletal muscle.

Consequences of haemorrhage and tissue injury: regional blood flows

It was apparent that, in the initial phase of haemorrhage, there was anincrease in systemic vascular resistance (SVR), which helped maintainarterial blood pressure in the face of a fall in cardiac output8. Fore-armblood flow was unchanged, suggesting that the increase in SVR wascaused by vasoconstriction in other vascular beds.

Vatner found that, in conscious dogs, mild haemorrhage (resulting in atachycardia but no fall in blood pressure) resulted in increases inmesenteric and iliac vascular resistance, but a reduction in renal vascularresistance30. With moderate hypotensive haemorrhage, there was a greaterreduction in mesenteric blood flow than in coronary blood flow,suggesting redistribution of flow in favour of the heart at the expense ofother internal organs.

When haemorrhage or hypovolaemia is taken further so that there isprofound hypotension, regional blood flows may alter again. Barcroft etal found that when their volunteers became hypotensive and fainted,there was a sudden fall in SVR and a sudden increase in fore-arm blood


Trauma

flow, suggesting a sudden skeletal muscle vasodilatation8. A similarskeletal muscle vasodilatation occurred when sheep were haemorrhageduntil an abrupt fall in blood pressure was seen31.

Regional vascular responses have also been studied using lower bodynegative pressure (LBNP), in volunteers, to simulate hypovolaemia32'33.The first study showed that splanchnic blood flow was more sensitive tosimulated hypovolaemia than renal blood flow32, while the secondshowed that splanchnic blood flow remained depressed until the end ofthe study period 60 min after reversal of LBNP33.

Another indication of the vulnerability of gut blood flow came froman examination of microvascular blood flow following haemorrhage34.Rats haemorrhaged to a MAP of 40 mmHg showed reductions inmicrovascular blood flow in the liver, spleen, kidney, small intestine andskeletal muscle, with the greatest reductions being seen in the smallintestine and the kidney. Volume replacement with Ringer's lactatesolution could only partially restore microvascular blood flow, whichcontinued to deteriorate in all organs, even after resuscitation.

Parallels have been drawn between the responses to injury and thedefence reaction26. The essential components of the regional vascularchanges during the 'defence reaction' are an increase in splanchnicvasoconstrictor tone and a reduction in vasoconstrictor tone to skeletalmuscle, thus preparing the organism for 'fight or flight'22.

When haemorrhage, in a pig model, occurred on a background ofsomatic afferent nerve stimulation (SANS, to mimic tissue injury) therewas an alteration in the pattern of vascular response35. In haemorrhagealone (30% of TBV), vascular resistance in the femoral bed more thandoubled, while resistance in the splanchnic bed hardly changed. When30% TBV haemorrhage was superimposed on SANS, there was a muchsmaller increase in femoral resistance. Splanchnic resistance increased,showing that with the additional insult of SANS (mimicking tissue injury),there was a relative diversion of blood away from the gut towards skeletalmuscle (Fig. 3).

Consequences of haemorrhage and tissue injury: gut mucosa

Do the changes in gut blood flow, which are seen with both simplehaemorrhage and haemorrhage with tissue injury have any majorfunctional consequences? There now appears to be ample evidence thatreduced gut blood flow, or haemorrhagic shock, can result in mucosalischaemia and eventually to mucosal damage36-37. The gut mucosa playsan important role as a barrier between the host and the gut microflora,which contains potential pathogens and toxins. Even in resting conditionsthe gut mucosa appears hypoxic compared to the serosa38. Gut mucosal



Fig. 3 Changes inregional vascular

resistances in responseto haemorrhage and

somatic afferent nervestimulation. The

figure shows thechanges in femoralvascular resistance

(upper panel), and gutvascular resistance

(lower panel) beforeand after a 30% TBV

haemorrhage (Haem),and after the same

haemorrhage in thepresence of brachial

nerve stimulation(HNS) Values are

expressed as means ±SE. Adapted from

Mackway-Jones etaP1

P-H

220

120

I

Gut

^

150

100

I HNS

r/̂ _______ ^ Haem

1 1

Pre End haem

oxygen tension (PmO2) was found to be very sensitive to changes inblood volume: there was a fall in PmO2 before any change in MAP orserosal oxygen tension (PsO2). By the time blood loss reached 15%,PmO2 had fallen by 40% with minimal changes in MAP and PsO r Thisprovides further evidence for potential hypoxia of the gut mucosa in theface of haemorrhage.

Any ischaemic insult to the mucosa may result in a loss of this barrierfunction and the translocation of organisms or toxins from the gutlumen into the host lymphatics or circulation.

Consequences of haemorrhage and injury: translocation of bacteria and endotoxin

As early as 1952, the suggestion was made that: 'tissue anoxia of shockmight stimulate invasion of the liver by intestinal flora'39. It was not untilthe 1970s that the term translocation was applied to the movement oforganisms from the gut lumen into either the systemic circulation, thelymphatics, or the peritoneum. By the late 1980s, the gut had been heavilyimplicated in the development of multiple organ failure (MOF)40-41.

Bacterial translocation is well documented in rodents, in which entericorganisms have been cultured after haemorrhagic shock42"*6. Feedingrats [14C]-labelled Escherichia colt showed that the organisms originatedin the gut47.


Trauma

In large mammals, the evidence is mixed but a number of investigatorshave demonstrated translocation in swine48'49, sheep5OrS1, and baboons52.Other investigators found no evidence of translocation in swine53"54. Adirect link between gut blood flow and bacterial translocation has beenshown in pigs subjected to a 40% BSA burn55.

Translocation in humans remains controversial. Moore et al insertedportal vein catheters in 20 trauma patients undergoing emergency laparo-tomy to take serial portal venous blood samples56. Endotoxin assaysperformed over the first 48 h were negative. Although 30% of thesepatients went on to develop MOF, the investigators found no evidencethat MOF was caused by gut-derived organisms. In another study, noevidence of bacterial translocation from mesenteric lymph nodes, excisedat laparotomy, was found in 25 blunt trauma patients, despite the fact that28% suffered major infectious complications with enteric pathogens57.

In contrast, two studies, which used immunofluorescence to look forevidence of the breakdown products of translocated organisms inmesenteric lymph nodes of trauma patients, both gave positive results58*59.Positive blood cultures were found m 13 of 50 trauma patients very shortlyafter their injury. The incidence of positive cultures was related to bloodpressure on admission. For 10 of these, the average admission bloodpressure was 45 mmHg. Whether these patients provide evidence forbacterial translocation as part of the pathophysiology of trauma isdebatable: bacterial translocation may simply have been an agonal event60.

If endotoxaemia is taken as an index of translocation, then it has beenshown to occur on the day of injury and to increase until day four inburns patients61. No endotoxaemia was found in two series of traumapatients56-62, while only 7% of patients in haemorrhagic shock hadraised endotoxin concentrations63.

There can be little doubt that bacterial translocation occurs in rodents.From the evidence in man one may suggest that finding live organismsin the circulation or the reticulo-endothelial system is a rare, possiblyagonal, event. Immunofluorescence and microscopy have suggested thatorganisms may translocate across the gut mucosal barrier but that theyare then killed by the secondary defences.

Immunological consequences of trauma

Late deaths from trauma continue to be a problem6-7'64. Most of theseare the result of MOF. This late development of a systemic illness is nota new phenomenon: after the battle of El Alamein, a surgeon wrotethat65: 'severe wounding was often followed by an illness, more or lessserious and lasting at least several days, in which many factors other



than blood-loss or its late effects operated'. The precise nature of the'many factors' has been a source of debate ever since. The evolution ofthis debate was summarised in 199666.

It is clear that the immune system is intimately involved in theresponse to trauma. The production of various cytokines after traumawill be used to illustrate this. It would be impossible to review all aspectsof the immune response to trauma in such a short review. Two bookswill provide a very valuable introduction to this field67-68.

Consequences of haemorrhage and tissue injury: cytokine production

Inflammatory responseInflammatory cytokine production has been described in models ofhaemorrhagic and traumatic shock52'69"71. Ayala et al have suggested, asa result of their studies with rodent models of haemorrhage and tissueinjury, that the two insults act as slightly different triggers for cytokineproduction69. It seemed that tumour necrosis factor a (TNFa) was onlyproduced after haemorrhage, but that interleukin-6 (IL-6) productioncould be stimulated by tissue injury before any haemorrhage70.

It is clear from a number of studies of elective surgery that the tissueinjury caused by surgery will stimulate the production of interleukin-1(IL-1), and/or IL-672"80. Indeed, it appears that the production of EL-6 islinked to the severity of the tissue injury inflicted73-77. In these studies,IL-6 production was detectable within 2 h of the initial incision and thepeak was seen within a few hours, unlike the peak in C-reactive protein,which tended to occur the following day75.

Deitch et al suggested that the gut acted as a cytokine-producing organsince levels of TNFa and IL-6 were higher in portal venous blood thanin systemic blood71.

Thus, it appears that the vascular responses to haemorrhagic andtraumatic shock can result in functionally important reductions in gutblood flow so that there is a breakdown of the gut mucosal barrier. Oncethis has happened, bacterial or endotoxin translocation may occur withpotential triggering of the production of inflammatory mediators, suchas cytokines. This may represent the initiation of a chain of eventsleading to a process of systemic inflammation which may ultimatelyproduce multiple organ failure.

Clinical studiesStudies of elective surgery have shown that there seems to be somecorrelation between the degree of tissue injury and the production of EL-6.This has been confirmed in several trauma series62-81, and in burns82.


Trauma

Raised inflammatory cytokine concentrations were found on the dayof admission, after trauma, and then gradually declined over the next 48h in patients making a good recovery62-81. However, there was goodevidence for a secondary increase in cytokine production, later, in thosepatients developing septic complications62'81"83. Initial cytokineconcentrations were not predictive for those patients who woulddevelop septic complications.

The role of TNFa in the cytokine response to trauma remains veryunclear. Many studies have found that TNFa was either undetectable, oronly detectable in a minority of patients after trauma62-81'84"87.

It has been suggested that the TNFa response is too transient to bedetected in most studies. A study using blood taken at the scene of theaccident concluded that TNFa was active in the inflammatory responsedue to trauma and that it might be activated early and may havemodulated subsequent cytokine activity88.

Raised plasma TNFa concentrations have been found inhaemorrhagic shock patients63. There was no correlation between TNFaconcentrations and the degree of haemorrhage (as assessed by thevolume of blood transfusion), nor was there a difference in TNFaconcentrations between those patients developing MOF and those whodid not. However, most studies have shown that TNFa concentrationsare higher in those patients who do become septic after trauma, than inthose who do not62-81"83-89.

Anti-inflammatory responsesTrauma not only results in the production of various inflammatorycytokines. Interleukin-10 (EL-10) has been characterised as an anti-inflammatory cytokine90"93, and its production has been measured intrauma patients94. Plasma IL-10 was associated with hypotension onadmission, and the development of sepsis, but was not related to theseverity of injury. In contrast, in another study, higher IL-10concentrations were seen in patients with an ISS greater than 25 than inthose with a lower ISS95. In this same group of 401 patients, higher IL-10 concentrations were measured in those patients who became septicand those who developed multiple organ dysfunction syndrome(MODS). The authors suggested that IL-10 might actually be involvedin the pathogenesis of sepsis and MODS after injury. This seems to runcounter to the ideas generated by laboratory studies.

Other endogenous anti-inflammatory mediators have been examined.There is good laboratory evidence for a protective effect of theendogenous inhibitor of IL-1, interleukin-1 receptor antagonist (IL-lra),in endotoxaemia96'97. Elective surgery has been shown to result in anincrease in IL-1 inhibitory activity over a 24 h period, but there was still

736 Bntish Medical Bulletin 1999,55 (No 4)


an increase in IL-6 and in the C-reactive protein, mostly after IL-1inhibitory activity had returned to baseline75. An increase in IL-lra wasseen in 22 trauma patients, but the concentration of EL-lra and othersoluble cytokine receptors, was higher in non-survivors than insurvivors98. The authors suggested that these receptors might be markersof the severity of the insult, rather than contributors to the mortality.Although inflammatory cytokines were undetectable, it does not appearthat the anti-inflammatory mechanisms had a very protective function.

Soluble TNF receptors (sTNF-Rs) are produced in trauma patients98-99,but, in both studies, concentrations were higher in non-survivors than insurvivors. If they had a protective anti-inflammatory function one mightexpect that the non-survivors would be those patients unable to producethe receptors, rather than those who produced the most. Interestingly,sTNF-R concentrations decreased during episodes of infection orhypoxia in individual patients99. Maybe these reductions in sTNF-Rconcentrations do indicate a loss of endogenous anti-inflammatoryactivity during secondary episodes of sepsis. This may form part of thebasis of the 'two-hit' model for the development of MOF.

Activation of cytokine productionBacterial and/or endotoxin translocation from the gastrointestinal tracthas been advanced as one possible mechanism for the activation ofcytokine production. Deitch et al have suggested that, even in theabsence of live translocating organisms or endotoxin in the portal andsystemic circulations, the gut can still act as a cytokine generator via theactivation of the gut associated lymphoid tissue71.

Alternatively, it could be that cytokine production is triggered directlyat the wound site. Evidence for this comes from both a rat model and aclinical situation100. Raised levels of EL-6 were found in both woundfluid and serum of rats subjected to polyvinyl alcohol sponge implant-ation, and in fluid draining from mastectomy scars. This suggested thatcytokines found in the systemic circulation might have originated at thewound site. Both IL-1 and IL-6 were found in the blisters of burnedchildren, although IL-1 was generally absent from the circulation, againsuggesting local production101.

It may be that two mechanisms act together or act independently ondifferent cytokines: IL-6 being triggered to a greater extent at the woundsite by tissue trauma while TNFa and IL-1 might be more sensitive tochanges in gut barrier function and activation of the gut associatedlymphoid tissue.

The question remains as to the exact trigger mechanism for thecytokine production seen in trauma and burns patients. The macro-phage hypothesis of activated macrophages producing cytokines and


Trauma

other mediators which then result in systemic inflammation leaves thequestion of what activates the macrophages unanswered102.

Models for the development of post-trauma multiple organ failure

A better understanding of the inflammatory responses after trauma hasresulted in interesting models for the development of MOF. Moore et alhave divided the inflammatory response into two phases: an initialhyperinflammatory response (the systemic inflammatory responsesyndrome - SIRS), followed by a phase of immunosuppression103'104.This model was a response to several studies104'105, which showed thatthere were two types of MOF patients: those who went into MOF'early', and those who went into MOF 'late'.

The other model is that of the 'one-hit' or 'two-hit' insult resulting inearly MOF103'106. In this model, patients may go into MOF, eitherbecause of one early and massive traumatic insult, which causes so muchinflammation that MOF ensues, or, if the initial insult is not massive,then this inflammatory response may be triggered by a secondary insultsuch as definitive surgery a few days after the trauma. Secondary surgeryhas been reported as a trigger for MOF107. The validity of these ideas hasaroused some controversy108-109. Even if there is debate about this modelit does serve as a warning that there may be a risk to delaying definitivesurgery for orthopaedic injuries, for example. There may be an argu-ment for either doing early "damage control surgery" or delaying for aweek or more, until after the worst of the hyperinflammation110'111.

Summary

In this review, I have tried to show that trauma elicits a variety ofcardiovascular responses. The responses to haemorrhage and to tissueinjury are different and the two insults interact to produce changes inorgan blood flow and oxygen delivery. These changes may haveimportant effects on the gut and gut mucosal integrity. The gut has beenheavily implicated in the development of MOF40-41.

Whether or not the gut and bacterial translocation are relevant inman, it is now clear that there is an inflammatory response after trauma.The development of that response seems to be responsible for thedevelopment of MOF, either by a 'one-hit' or by a 'two-hit' process inthe early stages, or later as the result of immunosuppression.

By understanding these responses, attempts can be made to optimisethe resuscitation and later care of trauma patients. As a recent review



stated112: 'patients in shock, if they are to survive the initial insult, arestill at risk of dying on a subacute basis from continuing globalhypoperfusion and the resultant multiple organ dysfunction, if they areinadequately resuscitated...It is clear that adequate resuscitation goesbeyond the restoration of a normal blood pressure and urine output.'

References

1 Griffiths R, Hinds C, Little R Manipulating the metabolic response to injury. Br Med Bull1999, 55: 181-95

2 Lowe D, Gately H, Goss J, Frey C, Peterson C. Patterns of death, complication, and error inthe management of motor vehicle accident victims: implications for a regional system of traumacare. / Trauma 1983; 23- 503-9

3 Trunkey DD. Trauma. Sa Am 1983; 249: 20-74 Regel G, Lobenhoffer P, Grotz M et al Treatment results of patients with multiple trauma an

analysis of 3406 cases treated between 1972 and 1991 at a German level 1 trauma centre. /Trauma 1995; 38: 70-8

5 Shackford S. The evolution of modern trauma care Surg Chn North Am 1995; 75: 147-566 Sauaia A, Moore FA, Moore EE et al. Epidemiology of trauma deaths: a reassessment. / Trauma

1995; 38: 185-937 Wyatt J, Beard D, Gray A, Busuttil A, Robertson C. The time of death after trauma BMJ 1995;

310- 15028 Barcroft H, Edholm OG, McMichael J, Sharpey-Schafer EP. Posthaemorrhagic fainting study

by cardiac output and forearm flow. Lancet 1944; i 489-919 Sander-Jensen K, Secher NH, Bie P, Warberg J, Schwartz TW. Vagal slowing of the heart during

haemorrhage observations from 20 consecutive hypotensive patients. BMJ 1986, 292: 364—610 Dnscoll PA Changes in systolic blood pressure, heart rate, shock index, rate pressure product

and tympanic temperature following blood loss and tissue damage in humans Leeds: Universityof Leeds, 1994

11 Secher NH, Sander-Jensen K, Werner C, Warberg J, Bie P. Bradycardia during severe butreversible hypovolaemic shock in man Circ Shock 1984, 14: 267—74

12 Matzen S, Perko G, Groth S, Friedman DB, Secher NH. Blood volume distribution during head-up tilt induced central hypovolaemia in man. Chn Physiol 1991; 11: 411-22

13 Little RA, Marshall HW, Kirkman E. Attenuation of the acute cardiovascular responses tohaemorrhage by tissue injury in the conscious rat Q ] Exp Physiol 1989, 74: 825-33

14 Jacobsen J, Sofelt S, Sheikh S, Warberg J, Secher NH. Cardiovascular and endocrine responsesto haemorrhage in the pig. Ada Physiol Scand 1990; 138: 167-73

15 Geerdes BP, Frederick KL, Brunner MJ. Carotid baroreflex control during hemorrhage inconscious and anesthetized dogs. Am J Physiol 1993; 265: R195-202

16 Secher NH, Bie P. Bradycardia during reversible haemorrhagic shock - a forgotten observation?Chn Physiol 1985; 5: 315-23

17 Kirkman E, Little R. Central control of cardiovascular responses to injury. In: Rothwell N,Berkenbosch F (Eds) Brain Control of Responses to Trauma. Cambridge: Cambridge UniversityPress, 1994, 202-38

18 Jacobsen J, Secher NH. Heart rate during haemorrhagic shock Chn Physiol 1992; 12: 659-6619 Howard JM, Am CP, Stahl RR. Hypertensive response to injury. Ann Surg 1955; 141: 327-3620 Anderson ID, Little RA, Irving MH. An effect of trauma on human cardiovascular control:

baroreflex suppression. / Trauma 1990; 30: 974—8121 Coote JH, Hilton SM, Perez-Gonzalez JF Inhibition of the baroreceptor reflex on stimulation

in the brain stem defence centre / Physiol 1979; 288: 549-6022 Hilton SM, Marshall JM, Timms RJ. Ventral medullary relay neurones in the pathway from the

defence areas of the cat and their effect on blood pressure. / Physiol 1983; 345: 149-66


Trauma

23 Parsons E, Pherruster D. Haemorrhage and 'shock' in traumatized limbs Surg Gynaecol Obstet1930; 51: 196-207

24 Wang SC, Overman RR, Ferng JW, Root WS, Gregersen MI The relation of blood volume reduc-tion to mortality rate in hemorrhagic and traumatic shock in dogs. Am J Physiol 1947; 148 164-73

25 Overman RR, Wang SC. The contributory role of the afferent nervous factor in experimentalshock: sublethal hemorrhage and sciatic nerve stimulation. Am J Physiol 1947; 148 289-95

26 Quest JA, Gebber GL. Modulation of baroreceptor reflexes by somatic afferent nervestimulation Am J Physiol 1972, 222: 1251-9

27 Rady MY, Little RA, Edwards JD, Kirkman E, Faithfull S. The effect of nociceprive stimulationon the changes in hemodynamics and oxygen transport induced by hemorrhage in anesthetisedpigs. / Trauma 1991; 31: 617-22

28 Rady MY, Kirkman E, Cranley J, Little RA. A comparison of the effects of skeletal muscleinjury and somatic afferent nerve stimulation on the response to hemorrhage in anesthetisedpigs. / Trauma 1993; 35: 756-61

29 Kirkman E, Zhang H, Spapen H, Little RA, Vincent JL Effects of afferent neural stimulationon critical oxygen delivery a hemodynarruc explanation. Am J Physiol 1995; 269: R1448-54

30 Vatner SE Effects of haemorrhage on regional blood flow distribution in dogs and primates. /Cltn Invest 1974, 54: 225-35

31 Gunnarsson U, H]elmqvist H, Rundgren M Haemodynamic responses to hypotensivehaemorrhage in conscious sheep with emphasis on renal and femoral blood flow. Exp Physiol1994; 79. 957-65

32 Hirsch AT, Levenson DJ, Cutler SS, Dzau VJ, Creager MA Regional vascular responses toprolonged lower body negative pressure in normal subjects. Am J Physiol 1989; 257: H219-25

33 Edouard AR, Degremont A-C, Duranteau J et al Heterogeneous regional vascular responses tosimulated transient hypovolaemia in man. Intensive Care Med 1994; 20: 414-20

34 Wang P, Hauptman JG, Chaudry IH Hemorrhage produces depression in microvascular bloodflow which persists despite fluid resuscitation Circ Shock 1990; 32: 307-18

35 Mackway-Jones K, Foex B, Kirkman E, Little RA. Modification of the haemodynamic responseto blood loss by somatic afferent stimulation [abstract]. Shock 1994, 2 19

36 Chiu C-J, McArdle A, Brown R, Scott H, Gurd E Intestinal mucosal lesion in low-flow states.Arch Surg 1970; 101: 478-83

37 Haglund U, Hulten L, Ahren C, Lundgren O Mucosal lesions in the human small intestine inshock. Gut 1975, 16: 979-84

38 Zabel DD, Kramer K, Hunt TK. Transmural gut tissue oxygen tensions during gradedhemorrhage. Cnt Care Med 1994, 22 A185

39 Frank H, Jacob S, Schweinburg F, Goddard J, Fine J. Traumatic shock. XXI. Effectiveness ofan antibiotic in experimental hemorrhagic shock. Am J Physiol 1952; 168: 430—6

40 Border J, Hassett J, LaDuca J et al The gut origin septic states in blunt multiple trauma (ISS =40) in the ICU Ann Surg 1987; 206 427-48

41 Wilmore D, Smith R, O'Dwyer S et al. The gut: a central organ after surgical stress. Surgery1988; 104: 917-23

42 Baker JW, Deitch EA, Berg R, La M. Hemorrhagic shock impairs the mucosal barrier, resultingin bacterial translocation from the gut and sepsis. Surgery 1987; 38. 73-4

43 Baker JW, Deitch EA, Li M, Berg RD, Specian RD. Hemorrhagic shock induces bacterialtranslocation from the gut. / Trauma 1988; 28: 896-906

44 Deitch E, Berg R, Specian R. Endotoxin promotes the translocation of bacteria from the gut.Arch Surg 1987; 122 185-90

45 Deitch EA, Bridges W, Ma L et al. Hemorrhagic shock-induced bacterial translocation: the roleof neutrophils and hydroxyl radicals. / Trauma 1990; 30: 942-51

46 Koziol JM, Rush BF, Smith SM, Machiedo GW. Occurrence of bacteremia during and afterhemorrhagic shock / Trauma 1988; 28: 10-6

47 Son AJ, Rush BF, Lysz 1 W, Smith S, Machiedo GW. The gut as a source of sepsis afterhaemorrhagic shock. Am J Surg 1988; 155: 187-92

48 Morales J, Kibsey P, Thomas PD, Poznansky MJ, Hamilton SM. The effects of ischaemia andischaemia reperfusion on bacterial translocation, hpid peroxidation, and gut histology: studieson haemorrhagic shock in pigs. ] Trauma 1992; 33: 221-6

740 British Medial Bulletin 1999,55 (No 4)


49 Turnbull R, Talbot J, Hamilton S. Hemodynamic changes and gut barrier function in sequentialhemorrhagic and endotoxic shock. / Trauma 1995; 38: 705-12

50 Morns SE, Navaratnam N, Townsend CM, Hemdon DN. Bacterial translocation andmesenteric blood flow in a large animal model after cutaneous thermal and smoke inhalationinjury Surg Forum 1988; 39 189-90

51 Navaratnam RLM, Morns SE, Traber DL et al. Endotoxin (LPS) increases mesentenc vascularresistance (MVR) and bactenaJ translocation (BT) / Trauma 1990; 30: 1104-15

52 Bahrami S, Schlag G, Yao Y-M, Redl H. Significance of translocation/endotoxin in thedevelopment of systemic sepsis following trauma and/or haemorrhage In: Levin J, Alvmg CR,Munford RS, Redl H (Eds) Bacterial Endotoxms: Ltpopolysaccbartdes from Genes to Therapy.New York: Wiley-Liss, 1995; 197-208

53 Gelfand GAJ, Morales J, Jones RL et al Hemorrhagic shock and bacterial translocation in aswine model. / Trauma 1991; 31 867-74

54 Foex BA. The haemodynamic response to haemorrhage and injury , and its consequencestn theanaesthetised pig North Western Injury Research Centre. Manchester: Manchester, 1998; 398

55 Saydjan R, Beerthuizen GIJM, Townsend CM, Herndon DN, Thompson TC. Bactenaltranslocation and its relationship to visceral blood flow, gut mucosal ornithine decarboxylaseactivity, and DNA in pigs / Trauma 1991; 31. 639-44

56 Moore FA, Moore EE, Poggetti R et al. Gut bactenal translocation via the portal vein: a clinicalperspective with major torso trauma. / Trauma 1991; 1991 629-36

57 Peitzman AB, Udekwu AO, Ochoa J, Smith S. Bacterial translocation in trauma patients. /Trauma 1991; 31: 1083-6

58 Brathwaite CEM, Ross SE, Nagele R et al Bactenal translocation occurs in humans aftertraumatic injury evidence usmg immunofluorescence. / Trauma 1993, 34 586-9

59 Reed LL, Martin M, Manglano R et al Bacterial translocation following abdominal trauma inhumans. Circ Shock 1994; 42- 1-6

60 Rush BF, Son AJ, Murphy TF et al. Endotoxemia and bacteremia during haemorrhagic shock.Ann Surg 1988; 207. 549-54

61 Winchurch RA, Thupari JN, Munster AM Endotoxemia in burn patients: levels of circulatingendotoxms are related to burn size. Surgery 1987; 102: 808-12

62 Hoch RC, Rodnguez R, Manning T et al Effects of accidental trauma on cytokine andendotoxin production. Cut Care Med 1993; 21: 839^5

63 Endo S, Inada K, Yamada Y et al. Plasma endotoxin and cytokine concentrations in patientswith hemorrhagic shock. Cnt Care Med 1994; 22: 949-55

64 Pape H-C, Remmers D, Kleemann W et al. Posttraumaac multiple organ failure - a report onclinical and autopsy findings. Shock 1994; 2: 228-34

65 Wilson W. State of men severely wounded in battle. Lancet 1944; i: 587-9166 Saadia R, Lipman J. Multiple organ failure after trauma. BMJ 1996; 313: 573—467 Gamelli R, Dnes D. Trauma 2000. Stategtes for the New Millenium Medical Intelligence Unit.

Austin/Georgetown: RG Landes Company, 199268 Geller E. Shock and Resuscitation. New York: McGraw-Hill, 199369 Ayala A, Pernn MM, Meldrum DR, Ertel W, Chaudry IH. Hemorrhage induces an increase in

serum TNF which is not associated with elevated levels of endotoxin. Cytokine 1990; 2: 170—470 Ayala A, Wang P, Ba ZF et al Differential alterations in plasma IL-6 and TNF levels after

trauma and haemorrhage. Am ] Physiol 1991; 260: R167-7171 Deitch EA, Xu D, Franko L, Ayala A, Chaudry IH. Evidence favoring the role of the gut as a

cytokine-generating organ in rats subjected to hemorrhagic shock. Shock 1994; 1: 141-672 Shenkin A, Fraser WD, Senes J et al. Serum interleukin 6 response to elective surgery.

Lymphoktne Res 1989; 8: 123-773 Cruickshank AM, Fraser WD, Burns HJG, van Damme J, Shenkin A. Response of serum

interleukin-6 in patients undergoing elective surgery of varying intensity. Clm Set 1990; 79:161-5

74 Baigne RJ, Lamont PM, Dallman M, Morns PJ. The release of interleukin-lB precedes that ofinterleukin-6 in patients undergoing major surgery. Lymphokine Cytokine Res 1991; 10: 253-6

75 Di Padova F, fozzi C, Tondre MJ, Tntapepe R Selective and early increase of H-l inhibitors,IL-6 and cortisol after elective surgery. Clm Exp Immunol 1991; 85: 137—42


Trauma

76 Baigrie R, Lamont P, Kwiatkowslu D, Dallman M, Morns P Systemic cytokine response aftermajor surgery. Br ] Surg 1992, 79: 757-60

77 Jons J, Cigarini I, Legrand M et al. Metabolic and respiratory changes after cholecystectomyperformed via laparotomy or laparoscopy. Br ] Anaestb 1992; 69: 341-5

78 Baigrie R, Lamont P, Whiting S, Morns P. Portal endotoxin and cytokine responses dunngabdominal aortic surgery. Am J Surg 1993; 166 248-51

79 Crozier TA, Muller JE, Quittkat D et al. Effect of anaesthesia on the cytokine responses toabdominal surgery. Br J Anaestb 1994; 72: 280-5

80 Moore CM, Desborough JP, Powell H, Burnn JM, Hall GM Effects of extradural anaesthesiaon interleulun-6 and acute phase response to surgery. Br J Anaestb 1994; 72: 272-9

81 Svoboda P, Kantorova I, Ochmann J. Dynamics of interleukin 1, 2 and 6 and tumor necrosisfactor alpha in multiple trauma patients / Trauma 1994; 36: 336-40

82 De Bandt JP, Chollet-Martin S, Hernvann A et al. Cytokine response to burn injury:relationship with protem metabolism. / Trauma 1994; 36- 624-628

83 Abraham E. Reply to letter by Botha et al Crtt Care Med 1994; 22 89584 Focx BA, Lamb WR, Roberts TE et al Early cytokine response to multiple injury. Injury 1993;

24: 373-^85 Foex B, Lamb W, Roberts T et al. The inflammatory cytokine response to multiple injury.

Intensive Care Med 1994; 20: S6286 Rabinovici R, John R, Esser KM, Vernick J, Feuerstein G. Serum tumor necrosis factor-alpha

profile in trauma patients. / Trauma 1993; 35: 698-70187 Meade P, Shoemaker WC, Donnelly TJ et al. Temporal patterns of hemodynarrucs, oxygen

transport, cytokine activity, and complement activity in the development of adult respiratorydistress syndrome after severe injury. / Trauma 1994; 36: 651-7

88 Ferguson K, Cardellio A, Tonapi V, Rodnquez JL, Maio RF. TNF is involved m the systemiccytokine response to trauma. Acad Emerg Med 1994; 1: Poster

89 Endo S, Inada K, Inoue Y et al. Two types of septic shock classified by the plasma levels ofcytokines and endotoxin. Circ Shock 1992; 38: 264—74

90 Cavaillon JM. Cytokines and macrophages. Biomed Pharmacotber 1994; 48' 445-5391 Mane C, Pitton C, Fitting C, Cavaillon JM. Regulation by anti-inflammatory cytokines (IL-4,

IL-10, IL-13, TGFbeta) of interleukin-8 production by LPS- and/or TNFalpha-activated humanpolymorphonuclear cells. Mediators Inflamm 1996, 5: 334—40

92 Marie C, Cavaillon JM. Le retrocontrole neganf de l'lnflammation- role des cytokinesantiinflammatoires Bull Inst Pasteur 1997; 95: 41-54

93 Wang J, Okkenhaug C, Jorgensen P, Solberg R, Aasen A. Dose dependent dual effect of IL-4 orIL-13 on proinflammatory cytokine release in experimental endotoxemia. Shock 1998; 10: 18

94 Sherry R, Cu£ J, Goddard J, Parramore J, DiPiro J. Interleukin-10 is associated with thedevelopment of sepsis in trauma patients / Trauma 1996; 40: 613—7

95 Neidhardt R, Keel M, Steckholzer U et al. Interleukin-10 (IL-10) plasma levels from injuredpatients correlate with seventy of injury and clinical outcome. 56th Annual Meeting of theAmerican Association for the Surgery of Trauma. Houston, Texas Amencan Association forthe Surgery of Trauma, 1996; 424

96 Ohlsson K, Bjork P, Bergenfeldt M, Hageman R, Thompson R. Interleukin-1 receptorantagonist reduces mortality from endotoxin shock. Nature 1990; 348. 550-2

97 Fischer E, Marano M, Van Zee K et al. Interleukin-1 receptor blockade improves survival andhemodynamic performance m Eschenchta colt septic shock, but fails to alter host response tosublethal endotoxemia. / Clin Invest 1992; 89: 1551-7

98 Cinat M, Waxman K, Vazin N et al Soluble cytokine receptors and receptor antagonists aresequentially released after truama./ Trauma 1995, 39: 112-20

99 Cinat ME, Waxman K, Granger GA et al. Trauma causes sustained elevation of soluble tumournecrosis factor receptors. / Am Coll Surg 1994; 179: 529-37

100 Mateo RB, Reichner JS, Albina JE. Interleukin-6 activity in wounds. Am J Phystol 1994; 266:Rl 840-4

101 Childs C, Ratckffe R, Holt I, Little R, Hopkins S. The relationship between interleukin-1, mter-leukin-6 and pyrexia in burned children. In Dinarello C, Kluger M, Powanda M, Openheim J (Eds)The Physiological and Pathological Effects of Cytokines. New York Wiley-Liss, 1990; 295-300



102 Deitch EA. Cytokines yes, cytokines no, cytokines maybe? Crrt Care Med 1993; 21: 817-9103 Moore F, Moore E. Evolving concepts in the pathogenesis of postinjury multiple organ failure.

Surg Cltn North Am 1995; 75: 257-77104 Moore F, Sauaia A, Moore E et al. Postinjury multiple organ failure: a bimodal phenomenon.

/ Trauma 1996; 40: 501-12105 Faist E, Baue A, Dittmer H, Heberer G. Multiple organ failure in poly trauma patients. /

Trauma 1983; 23: 775-87106 Deitch E. Multiple organ failure - pathophysiology and potential future therapy. Ann Surg

1992; 216: 117-34107 Waydhas C, Nast-Kolb D, Trupka A et al. Posttraumatic inflammatory response, secondary

operations, and late multiple organ failure. / Trauma 1996; 40. 624-31108 Saadia R, Schein M. Multiple organ failure How valid is the 'two-hit' model? / Accid Emerg

Med 1999; 16: 163-7109 Foex B. Multiple organ failure. How valid is the 'two-hit' model? A commentary. / Accid

Emerg Med 1999; 16: 166-7110 Mattox K. Introduction, background, and future projections of damage control surgery. Surg

Clm North Am 1997; 77- 753-9111 Rotondo M, Zonies D. The damage control sequence and underlying logic. Surg Clm North

Am 1997; 77: 761-77112 Elliott D. An evaluation of the end pomts of resuscitation. / Am Coll Surg 1998; 187: 536-47

British Medial Bulletin 1999,55 (No 4) 743

systemic responses to trauma - semantic scholar...systemic responses to trauma b a foex emergency...

Documents