Injury (1991)22, (5), 377-382

Monitoring

Printed in Great Britain

377

the response to injury

U. Schiiffel, F. Bonnaire, B. U. van Specht and E. H. Kuner Department

of Surgery and Traumatology,

University

of Freiburg, FRG

Tissue trauma leads to a seue&y-dependent activation of plasma and cellular systems. This response can be recorded by determining parameters which represent the activation state of these systems. In severely injured patienfs with multiple trauma rhree auf of 14 parameters measured at the time of admission proved to be indicators of subsequent septic complications wilh a high degree of accuracy: Fibrinopeptide A (FPA - the first split prod& of jibrinogen), the C3 split product C3a, and the elastase-a, proteinase inhibitor-complex (ELx,PI/. In a second serk of multiply-injured patients wifh femoral fractures who did not develop clinical sepsis (NN=25) these parameters were measured continuously to evaluate the influence of injury sevm’fy and of therapeutic strategy on the further coursz. We found a strong correlation between injury severity (ISS) anA the degree of activation. The signs of activation decreased rapidly following immediate operafive fiation, and remained elevated or euen increased afier primary femoral traction and seconday stabilization. The operative procedure did not cause any additional activation. Complications such as infection or the formation of haematomas were reflected by raised parameter levels

Introduction The traditional conservative approach to the treatment of multiply-injured patients with skeletal traction and plaster cast immobilization has often been challenged during the last two decades. Until now, several studies have demonstrated the beneficial effects of immediate fracture fixation in preventing post-traumatic complications such as adult respiratory distress syndrome (ARDS) and multiple organ failure (Riiedi and Wolff, 1975; Riska et al., 1976; Meek et al., 1981; Goris et al., 1982; Johnson et al., 1985; Seibel et al., 1985; Lozman et al., 1985). The main cause of later death in accepted these patients is sepsis. Despite a generally treatment of early antibiotics for open fractures (Dellinger et al., 1988), ‘gut origin’ septic states (Border et al., 1987; Saadia et al., 1990) during the later course certainly can not be prevented, even if selective gut decontamination may promise some benefits with respect to nosocomial respiratory tract infections (Stoutenbeek et al., 1984). However, since a significant focus of invasive infection does not seem to be a necessary component for the development of sepsis, sepsis or the ‘sepsis-like syndrome’ might be related to some kind of mediator disease (Meakins et al., 1980; Goris et al., 1986; Schiiffel et al., 1989a). The use of trauma scales is generally accepted either to compare injured individuals or to compare series of injured 0 1991 Butterworth-Heinemann 0020-1383/91/050377-06

Ltd

F,br,r, 1 FDP

Figure 1. Activation

of plasma systems by tissue damage.

patients on the basis of the severity of the respective injury (Baker et al., 1974). A further possibility is an additional monitoring of the organism’s response to the traumatic event (Cheadle et al., 1989). Tissue trauma leads to a severity-dependent activation of the major plasma and cellular systems. This response can easily be recorded by determining parameters which represent the activation state of these systems as well as the inhibitory capacity of the plasma (F&we I). The aim of the present study was to evaluate whether the degree of this activation correlates to the trauma severity as assessed by a current trauma score (ISS) and whether a later development of sepsis may be characterized by an overwhelming inflammatory response. In a second series of patients without clinical sepsis, the influence of the therapeutic strategy on the actual activation state was evaluated.

Patients and methods The first part of the study was conducted in a selected series of severely injured patients (N= 16) with several long bone fractures, including at least one fractured femur (Hospital Trauma Index 4 and 5; ACS, 1980). Only the patients who were admitted no longer than 90 minutes after the initial trauma (motor vehicle accidents in all cases) and who had not yet been given blood transfusions or protein solutions were included. This group consisted of 11 men and five women with a mean age of 31.4 years (range 16-53 years) and a mean Injury Severity Score of 37.06 (range 11-57). Eight patients died.

Injury: the British Journal of Accident Surgery (1991) Vol. 22/No. 5

378

Table I. Severity of trauma and activation state at the time of admission (first series) lnjwy Severity Score (ES)

FPA x (range) 18 30 43.6

27

Blood sampling was started at the time of admission and was continued every 24 h for the first week on the intensive care unit (KU). Fibrinopeptide A (FPA), the complement factor 3 split products C3a, C3d, and C3c, the elastase-a, proteinase inhibitor complex (Ea,PI), fibronectin, prekallikrein (PKK), plasminogen, coagulation factor XII, fibrinogen, and the proteinase inhibitors antithrombin III antigen (ATIIIa) and functional activity (ATIIIf), a,-macroglobulin, and ol,-antiplasmin were determined by use of commercially available assay systems in appropriately prepared serum and plasma samples. C3d was measured by rocket-immunoelectrophoresis, using antibodies to C3c and C3d in the gel preparation. ‘Sepsis’ or ‘Septic Syndrome’ were later diagnosed in 10 patients in accordance with the commonly accepted clinical criteria of at least five of the following: bacteraemia, presence of an infectious focus, acute deterioration of mental status, oliguria (urine output less than 30 ml/h for at least 2 consecutive hours), body temperature above 38.5”C, white blood cell count of more than 15,000 cells/mm3, thrombocytopenia of less than 100,000/mm3, or depression of h4AP requiring catecholamines despite an adequate resuscitation. For three parameters (FPA, C3a, Ea,PI), which showed a significant difference in levels between the survivors and fatalities during the clinical course, cut-off points were

0’) O/2 O/3 8/l 1

225 (194-256) 338 (304-396) 442 (308-700)

11.5 (10-13) 14.0 (11-19) 24.6 (13-42)

(9-27) (17-39) (25-86)

Died

Ei ,PI x (range)

C3a x (range)

calculated which predicted the occurrence of post-traumatic complications with great accuracy. In the second series 25 consecutive patients with closed femoral fractures (HTI 3-5) were included, using the same exclusion criteria as for the first group; 17 men and eight women with a mean age of 34.2 years (range 15-52 years) had a mean Injury Severity Score of 23.0 (range 4-59). All patients survived and no development of a ‘septic syndrome’ was observed. Of these patients, 20 were treated by primary fracture fixation (within 6 h after admission) while five patients were treated with initial skeletal traction and secondary osteosynthesis of the femur (three nails, two plates) 4 or 5 days after trauma. Complications involving major infections (pneumonia, deep soft tissue infection with or without osteitis, or bacteraemia) were observed in eight patients and major bleeding requiring operative revision in two patients (one patient with both, infection and haematoma). The primary management of the traumatized patients with fluid resuscitation and transfusion of blood or fresh frozen plasma did not reveal any significant differences, neither between the groups with immediate or with secondary operative fracture treatment nor between the groups with or without later complications. There were also no significant differences between the groups with respect to

ATIKCf

FPA 46

. ng ml-’

fibrinogen

fayf~m-septic U ml-’

fi bronectin

a

PKK

FXII plasminogen

day 3 day6

______ ===_.

Schaffel et al.: Monitoring of the response to injury

379

.: .-..-.r.

a2 ontiplasmin c D/,

I A/\\

\

65

;j1

2 ” 6 lo?’ *..,/ _..__.. ... .. ________----160 50 ... . . .s - -

-7s. \

-

..,

_\ _ _

, 30 ok * . . . . . . . . . . . . . . . . :‘:‘-“-‘s __--

. . . . . . .-’ ._.._.k.’ _ - -

FPA . ; .,’ ._’ . . . , . --.. 4f&___..,.+.-

Posttroumatlc, ISS 33

100 %

CO

og litre’

--.-..‘Y,.\._.;...

fibrinogen

-I

\-

ngml-I

septic course,

day 0 . . . . . . day 3 --day 6...., patient

died at day 6

%

/

pgml-’ UrnI-’

->

fibronectln

b

PKK

%

FXII

plasminogen

C3d

ATlEo

FPA

At odmusslon

--GO

pg lltre

/ fibronectin

C

pg ml-’

I

Uml-’

ISS(%)

lethal

38

non-lethal

25

fibrinogen Later ._._._

course

(day

5+ 6

)

lethal non- lethal

F XII

PKK

Figure 2. a, Post-traumatic non-saptic course. b, Post-traumatic septic course. c, Activation states in fatal and non-fatal courses. Showing the parameters in a radial presentation with their scales dependent on the normal range, indicated for each parameter by the ring surrounding the centre. The indented figure within the inner circle results from connecting the actual zeros which touch the normal range in the case of FPA and C3a. The connecting lines between the values obtained for each parameter will result in a star-like figure, characteristic of the actual inflammatory state. Values obtained from controls were always shown to lie upon the ring.

380

Injury: the British Journal of Accident Surgery (1991) Vol. 22/No.

C3a A hg dl

1000

-

800

-

FPA . (ngml-‘I

Ea,PI 0 (ngml-‘I

-

-

620

-

500

22

-

,6

-

+ _.___.__

600_-_-_--~

____.__________

_.____

_________...___

. . . ..__

-

400-

200

.,,-

-

FPA l (ngml-‘I

620

480

-

480

.-

360

--

360

IO

-

240

-

240

4

-

-

120

EXWBATION-

4o

120

-0

o-

-0

-0 I 0

I I

I 2

a

I 3

I 4

I

I

I

5

6

7

Ea,PlO (ngml-‘1 -

40

-2.3

OP

c5a * hgml-‘)

5

Days after +r0um0 Ea,Pl 0 (ngml-‘I

FPA l (ngml-‘)

c3or (ngml-‘I OP

-

40

-

620

-

28

-

500

-

22

-

480

TRACTION LAPAROTOMY c3a hgd

.

REVISION

OP

*

+

JI

1

h

+ 1000

-

800

-

\

~,,_~~~~~~~~~~~~~~~~__~_._~~~~~~ . . ..__......- -~. ..------ ah

400-

-

,6

----_

,O

-

,. .~

.

_

oI 0

I

I

I 2

b

I 3

-

i

-4

~,j _‘

-

240

-

120

.. -0

I 5

I 6

OP

>:360ngml-‘I

.

Early ostecsynthesls

wlthout

n = I I ISS 20 4

C3a > 600

0

Delayed

.

Controls

exceedmg

FPA>

cutoff

points

ng ml-’

Ibngml-’

n=

0 Days after

lSS317

n=5

osteasynthesls

I

I

compkatlons

I

I

3

7

ISS242 2

IS54

tmuma

Figure&

Activation pattern in different patient groups (mean values). Two patients with uncomplicated, isolated fracture of one femur served as ‘controls’.

Number exceedmg (Ea,PI>

of parameters cutoff

points

360ngml-‘I

C3a>600ngml-’ FPA > 16ngmlK 0

3-

0

2-

-

oe

cm0

COocmo

.a

l

0

COO

cxx

0

o-

I 0

I

I

I

I

I

I

I

5

IO

20

30

40

50

60

lqury

seventy

score

Figure 5. Correlation

between Injury Severity Score and activation state at the time of admission. Black circles indicate cases with complications during the further course.

dependent activation could be observed at the time of admission (day 0). There was a constant decrease in the degree of activation after an early operative stabilization. Skeletal traction alone did not markedly influence the situation, whereas secondary osteosynthesis on day 5 did (F&we 3~). Complications led to a secondary increase of the parameter levels, reflecting reactivation by inflammatory processes. Neither nail-osteosynthesis nor osteosynthesis with femoral plates resulted in additional activation, not even if evaluated during the first postoperative hours. These observations are basically confirmed by the comparison of different treatment modalities and different courses (Figure4). There was no significant difference in the occurrence of complications between the early and the secondary osteosynthesis group. On the other hand, the values of three patients with delayed osteosynthesis without complications did not differ from those two patients who developed complications during the post-traumatic course. Figure 5 shows the correlation between the Injury Severity Score and the activation state for the second series. Spots mark the cases involving complications.

Each response to tissue injury as well as every inflammatory response leads to the activation of plasma and cellular systems so that certain parameters can be used as indicators (Schoffel, 1989b). Fibrinopeptide A is an indicator of thrombin-induced fibrinogen proteolysis and thus, indirectly, of the activation of coagulation factor XII (Fareed et al., 1983). The early increase of FPA observed in this study seems to prove the importance of this particular pathway also in the response to injury. The role of the complement system during an inflammatory response is rather complex. Apart from its cytotoxic effect, its chemotactic (C5a, C567), opsonizing (C3b), and vasoactive properties (C3a, C5a, C&kinin) make it one of the most important factors in the response to injury and during any further septic development (Nuytinck et al., 1986); Slotman et al., 1986; Schoffel, 1989b). In Figure&-r the increase of C3a correlates with an increase in C3d and corresponds to a relatively low level of C3c, reflecting the total amount of C3. A consumption of complement factors did not occur within the observation periods in our study. The activation of leukocytes is invariably followed by a release reaction and can thus be recorded by the complex of leukocytic elastase and its main inhibitor a,-proteinase inhibitor (Duswald et al., 1985; Dittmer et al., 1986). The observed variations in the levels of the other parameters were not statistically significant with respect to survivors and fatalities. Whether these variations are due to an altered synthesis rate, specific consumption, or unspecified degradation and elimination must remain open. Within a certain range they are paralleled by the total protein content of the plasma. However, whereas declining prekallikrein (PKK) levels might be explained best in terms of consumption, the depressed fibronectin levels have to be interpreted more carefully since the mechanisms of synthesis, of exchange between free and surface-bound molecules, and of their control are not yet fully understood (Chadwick et al., 1984; Kiener et al., 1986). This uncertainty concerning the influence of the effects of wash-out, dilution, consumption, and degradation points out the advantage of inflammatory markers like FPA, C3a, and Ea,PI, which are less influenced by altered synthesis rates or substrate depletion. Since all these influences are difficult to evaluate, correction factors can hardly be used and statistical analysis may also be open to doubt under these circumstances. We can state, however, that there are some parameters which may permit very early recognition of the intensity of the actual response to injury and sensitively to monitor its further course. The fact that the intensity of the activation correlated with the injury severity in both series was not surprising and tends to confirm the current pathophysiological concepts. The secondary increase in complicated courses also fits into the concept of an overwhelming inflammatory response as the central feature in the development of sepsis. That is why the recording of the degree of activation offers the advantage of simultaneously determining the severity of trauma and monitoring the clinical course. As shown in the second series, any operative attempt at stabilization reduces the degree of activation. Primary skeletal traction alone did not seem to stabilize the posttraumatic situation sufficiently. It may be assumed that a focus of activation remained in those cases, and that this focus might become harmful in some instances. On the other hand, the additional traumatizing effect of the operative procedure seems to be negligible. The presented numbers

382

Injury: the British Journal of Accident Surgery (1991) VoI. 22/No. 5

are clearly not large enough to have subsets where statistical analysis can be carried out and which definitely confirm the necessity of an immediate operative fracture fixation. In conclusion, the described response to injury cannot only be used to determine the severity of trauma and to monitor the clinical course, but might also be useful in the further assessment of the efficacy of therapeutic strategies.

References American College of Surgeons (1980) Hospital trauma index. Bull. Am. Coil. Surg. 65, 32. Baker S. P., O’Neill B., Haddon W. et al. (1974) The injury severity score: a method for describing patients with multiple injuries and evaluating emergency care. J. Trauma 14, 187. Border J. R., Hassett J., LaDuca J. et al. (1987) The gut origin septic states in blunt multiple trauma (ISS = 40) in the ICU. Ann. Surg. 206, 427. Chadwick S. J. D., Mowbray J. F. and Dudley H. A. F. (1984) Plasma fibronectin and complement in surgical patients. BY.1. Surg. 71,718. Cheadle W. G., Wilson M., Hersman M. J. et al. (1989) Comparison of trauma assessment scores and their use in prediction of infection and death. Ann. Surg. 208,541. Dellinger E. P., Caplan E. S. Weaver L. D. et al. (1988) Duration of preventive antibiotic administration for open extremity fractures. Arch. Surg. 123,333. Dittmer H., Jochum M. and Fritz H. (1986) Freisetzung von granulozytarer Elastase und Plasmaproteinveranderungen nach traumatisch-hamorrhagischem Schock. Unfallckirurg. 89, 160. Duswald K.-H., Jochum M., Schramm W. et al. (1985) Released granulocytic elastase: an indicator of pathobiochemical alterations in septicemia after abdominal surgery. Stlrgoy 98,

892. Fareed J., Bickl R. L., Squillaci G. et al. (1983) Molecular markers of hemostatic disorders. C/in. Ckem. 29, 1641. Goris R. J. A., Gimbrere J. S. F., van Niekerk J. L. M. et al. (1982) Early osteosynthesis and prophylactic mechanical ventilation in the multitrauma patient. J. Trauma 22, 895. Goris R. J. A., Boekholtz W. K. F., van Bebber I. P. T. et al. (1986) Multiple-organ failure and sepsis without bacteria. Arch. Surg.

Kiener J. L., Cho E. and Saba T. M. (1986) Comparative effect of circulating bacterial or non-bacterial particulates on plasma fibronectin. Circ. Shock 19,357. Lozman J., Deno C., Feustel P. J. et al. (1986) Pulmonary and cardiovascular consequences of immediate fixation or conservative management of long-bone fractures. Arch. Suq. 121,992. Meek R. N., Vivoda E., Crichton A. A. et al. (1981) A comparison of mortality of patients with multiple injuries according to method of fracture treatment. 1. Bane ]oint Stlrg. 63B, 456. Meakins J. L., Wicklund B., Forse R. A. et al. (1980) The surgical intensive care unit: current concepts in infection. Surg, Clin. North Am. 60,117. Nuytinck J. K. S., Goris R. J. A., Redl H. et al. (1986) Post-traumatic complications and inflammatory mediators. Arch. Surg. 121,

886. Riska E. B., Von Bonsdorff H., Hakkinen S. et al. (1976) Prevention of fat embolism by early internal fixation of fractures in patients with multiple injuries. Injury 8, 110. Riiedi T. and Wolff G. (1975) Vermeidung posttraumatischer Komplikationen durch friihe definitive Versorgung von Polytraumatisierten mit Frakturen des Bewegungsapparats. Helv. Chit-.Acta 42,507. Saadia R., Schein M., MacFarlane C. et al. (1990) Gut barrier function and the surgeon. Br. J Surg. 77,487. Schiiffel U., Lausen M., Ruf G. et al. (1989a) The overwhelming inflammatory response and the role of endotoxin in early sepsis. Prog. Clin. Biol. Res. 308, 371. Schoffel U., Zeller T., Lausen M. et al. (1989b) Monitoring of the inflammatory response in early peritonitis. Am. 1. Surg. 157,

567. Seibel R., LaDuca J., Hassett J. M. et al. (1985) Blunt multiple trauma (ISS 36), femur traction, and the pulmonary failureseptic state. Ann. Surg. 202,283. Slotman G. J., Burchard K. W., Williams J. J. et al. (1986) Interaction of prostaglandins, activated complement, and granulocytes in clinical sepsis and hypotension. Surgery 99, 744. Stoutenbeek C. P., Van Saene H. F. K., Miranda D. R. et al. (1984) The effect of selective decontamination of the digestive tract on colonization and infection rate in multiple trauma patients. Intensive Care Med. 10,185.

Paper accepted

14 February

1991.

121,897. Johnson K. D., Cadambi A. and Seibert G. B. (1985) Incidence of ARDS in patients with multiple musculoskeletal injuries: effect of early operative stabilization of fractures. J. Trauma 25, 375.

RequesLsfor reprints should be addressed to: Ulrich Schiiffel MD, Chirurgische Universititsklinik, Hugstetterstrasse 55, D-7800 Freiburg, FRG.