Journal of Visceral Surgery (2016) 153, 33—43

Available online at

ScienceDirectwww.sciencedirect.com

REVIEW

Management of blunt hepatic trauma

C. Letoublon ∗, A. Amariutei, N. Taton, L. Lacaze,J. Abba, O. Risse, C. Arvieux

hirur Gre

16

yeapolicatus

effeCT)

(DS).. Whe

Clinique Universitaire de Cde Grenoble, BP 217, 38043

Available online 9 August 20

KEYWORDSLiver trauma;Perihepatic packing;Abdominalcompartmentsyndrome

Summary For the last 20(BHT) has been the initial

where the hemodynamic stcoexistence of three highlycomputerized tomography (and finally delayed surgery

part of the surgical strategy

option has seen its effectivenes(damage control) laparotomy anThe effectiveness of these two

evidenced by current experiencethe most severe grades of injury© 2016 Elsevier Masson SAS. All

Introduction

‘‘Miss nothing and fix everything’’ has long been thedogma for emergency management of visceral trauma,which imposed obligatory emergency laparotomy for anyhemoperitoneum. For blunt hepatic trauma (BHT), thatattitude has been profoundly but gradually transformedbeginning in the 1970’s, moving toward avoidance of emer-gency laparotomy whenever possible. This approach hasbeen supported by the contribution of contrast-enhancedCT [1—3]. This dogma was also upended by the conceptof abbreviated laparotomy (damage control) where controlof active liver bleeding is obtained by perihepatic packing(PHP); this practice has transformed the management ofmost severe BHT when hemodynamic instability imposes the

∗ Corresponding author.E-mail address: CLetoublon@chu-grenoble.fr (C. Letoublon).

http://dx.doi.org/10.1016/j.jviscsurg.2016.07.0051878-7886/© 2016 Elsevier Masson SAS. All rights reserved.

gie Digestive et de l’Urgence, Université Grenoble Alpes, CHUnoble cedex, France

rs, nonoperative management (NOM) of blunt hepatic traumay whenever this is possible (80% of cases), i.e., in all casesdoes not demand emergency laparotomy. NOM relies upon thective treatment modalities: radiology with contrast-enhancedand hepatic arterial embolization, intensive care surveillance,

DS is not a failure of NOM management but rather an integraln imposed by hemodynamic instability, the immediate surgical

s transformed by development of the concept of abbreviatedd wide application of the method of perihepatic packing (PHP).conservative and cautious strategies for initial management is, but the management of secondary events that may arise with

must be both rapid and effective.rights reserved.

need for an emergency surgical response [4—6]. This conceptrelies on a better understanding of the pathophysiology ofmajor uncontrolled bleeding. Mastery of the techniques ofurgent arterial embolization by interventional radiologistshas added an additional weapon to the armamentarium[7]. The concept of DS (even possibly by laparoscopy) alsodeserves its place in modern methods of treatment of BHT[8,9].

In practice, two effective therapeutic trios can bedescribed, depending on whether the trauma victim requiresurgent laparotomy or not.

Option 1 — nonoperative management andthe first therapeutic trio: CT + intensivecare surveillance + delayed surgery

This option can be applied to about 80% of patients with BHT[1—3]. To avoid immediate operation, the main requirement

3 C. Letoublon et al.

irltr

C

Cvodccir

iopamadoto

C

Fdhlotogi[ar

bntestheids

HFtp‘batos

Ftondpap

4

s hemodynamic stability or a satisfactory response to initialesuscitation: once the patient is hemodynamically stabi-ized, contrast-enhanced CT is the next necessity. ‘‘Therauma patient who does not go directly to the operatingoom goes to CT.’’

T and arterial embolization

T with three-stage contrast injection including a lateenous phase is part of a whole body CT. It allows definitionf anatomic severity according to Mirvis, but particularly theetection of active vascular leakage [10]. This leak usuallyonsists of blood: CT must define its size, from a small stableollection contained within a deep lesion to obvious ongo-ng extravasation bleeding freely into the abdominal cavityesulting in a large hemoperitoneum.

If active bleeding is detected, immediate embolizations indicated, but it is also essential to look for other sitesf bleeding that also require treatment (kidney, spleen,elvis). Hepatic arterial embolization should be as selectives possible to minimize the risk of gallbladder or parenchy-al ischemia [11] and, if possible, should be performed with

bsorbable material. In rare cases, a major venous leak isetected: in this case, whether the bleeding is from a portalr suprahepatic vein, its hemodynamic consequences dic-ate the need for surgery since it is known that compressionf the liver controls the majority of venous bleeding (Fig. 1).

areful and specifically adapted surveillance

ollowing radiologic diagnosis and/or intervention, hemo-ynamic status may require surgical intervention ifemodynamic status deteriorates (e.g., uncontrolled venouseak) or an abdominal compartment syndrome (ACS) devel-ps during the procedure. In most cases, NOM is theherapeutic choice and the intensity of surveillance dependsn both the severity of the associated injuries and the Mirvisrade itself: the risk of secondary events is much higher fornjuries Grade III or higher, or if embolization was required12]. Kozar et al., in a study of 453 BHT treated nonoper-tively, noted a global complication rate of 14%, while the

ate was 52% for Grade V injuries [12].

Very careful monitoring is necessary because a num-er of events may occur, which, while not deserving theame of ‘‘complications’’, should trigger timely and effec-ive responses. Patients should be monitored in a surgicalnvironment or intensive care unit (ICU) depending on theeverity of the hepatic injury and other associated mul-iple trauma: a grade III or higher liver injury, and/or aemoperitoneum that is obvious on CT, and/or the need formbolization all require ICU surveillance. Search for hepaticnjuries or other associated lesions that may not have beenetected in patients who did not undergo initial laparotomyhould be ongoing (Fig. 2).

epatic hemorrhageor massive uncontrolled bleeding, emergency laparo-omy is indicated. Otherwise, contrast-enhanced CT iserformed, which is superior to initial arteriography. If a‘blush’’ is detected, embolization of the involved arterialranch should be performed. Failure of embolization torrest hemorrhage should lead to laparotomy. Sometimeshe clinical picture is more ‘‘chronic’’, with a gradual fallf hemoglobin, while CT images do not reveal any bleedingite: transfusion, in such cases, may avoid recourse to

shishiobw

BOdTs(

PTaop

igure 1. 60-year old female. Skiing accident. Computerizedomography (CT) on arrival (a): severe extravasation due to rupturef the middle hepatic vein with blush extending to the abdomi-al cavity (arrows). Patient developed shock in the radiology suiteuring CT scan, leading to urgent intervention with perihepaticacking. Repeat CT prior to removal of packing on day 5 (b) shows noctive bleeding and the stump of the middle hepatic vein. Uncom-licated course thereafter.

urgical intervention or repeated arterial embolization inalf the cases [13]. Subcapsular hematoma is rarely anndication for surgery, even when this reaches an impres-ive volume or is painful. But one must verify that theematoma is not causing parenchymal ischemia by a sort ofntrahepatic compartment syndrome [14,15] or by torsionf the suprahepatic veins: elevated transaminases are theest sign of this entity. In such cases, surgical interventionith decompression and hemostasis is required [14—17].

ile peritonitisngoing bile leakage into the peritoneum often becomes evi-ent between day 2 and day 5, presenting as abdominal pain.he diagnosis is confirmed by ultrasound-guided paracente-is: positive diagnosis should lead to DS (described below)Fig. 3).

eritoneal inflammatory syndromehis typically presents as progressive abdominal pain aftern interval of one to six days, fever without obvious signsf infection, and elevated C-reactive protein (CRP). Theresence of free peritoneal fluid, which can be of moderate

Management of blunt hepatic trauma 35

theithouto suse th

Figure 2. 28-year old female. Fall from an ATV. Transferred totomography (CT): grade IV fracture of the right hepatic lobe, whemoperitoneum. Surveillance in ICU for 24 hours, then transfer

treatment with low molecular weight heparin. Uncomplicated cour

volume, is an indication for DS. The syndrome is due toperitoneal irritation caused by a mixture of bile and blood;it is usually sterile, and typically occurs after the mostsevere lesions on CT (grade IV or V) [8].

Intra-abdominal hypertension with abdominalcompartment syndromeIntra-abdominal hypertension (IAH) is a threat whenrelatively major bleeding has resulted in major hemoperi-toneum, often after embolization, and multiple transfu-

sions. HIA is defined by intra-abdominal pressure exceeding15 cm H2O, as measured by intravesical pressure. IAH cantrigger abdominal compartment syndrome (ACS), compa-rable to muscular compartment syndromes, which canlead to multi-organ failure. Decompression is required, bylaparotomy or laparoscopy (beginning with a decompressivemini-laparotomy) [18,19].

Rarer complicationsBilomaBiloma is caused by a slow and contained bile leak; the col-lection is typically juxtahepatic and asymptomatic. Rarely,this may result in local compression. The diagnosis ismost often established by ultrasound-guided needle aspira-tion, followed by drainage and drain placement to createa controlled bilio-cutaneous fistula. Intrahepatic bilomamay mimic a parenchymal hematoma, but no treatment isrequired unless there are associated signs of infection [20].

AbscessAbscess is usually due to a biloma and/or hepatic parenchy-mal necrosis that becomes infected. Percutaneous drainagedoes not always successfully avoid the need for DS with moreor less extensive resection and wide drainage.

university hospital center. Stable blood pressure. Computerizedt extravasation in either the arterial or venous phase. Moderaterveillance on the surgical service. Pulmonary embolus leading toereafter. Discharge on day 20.

Vascular complicationsHemobilia is arterial bleeding into the biliary tract; thediagnosis should be suspected in the presence of thesymptomatic triad of pain, jaundice, and gastrointestinalbleeding. It can usually be treated by arterial embolization;classic right hepatectomy can usually be avoided [21].

Intrahepatic arterial false aneurysms are treated byselective embolization of large lesions or smaller lesions thatpersist for 1—2 weeks.

Arterioportal fistulas, a rare occurrence, may result inportal hypertension. High-flow lesions require embolization.

However, they are most often distal and asymptomatic, andoften heal spontaneously.

Arterio-hepatic vein fistulas are rare but tumultuous,often resulting in heart failure; embolization is often dif-ficult.

Fistula between the bile ducts and the hepatic veins oftenresults in hyperbilirubinemia due to direct flow of bile intothe bloodstream; jaundice is characterized by discordancebetween marked hyperbilirubinemia and paradoxically lowlevels of alkaline phosphatase. This lesion is very rare andheals spontaneously in most cases, but nasobiliary drainagehas been proposed to speed resolution [22].

Hepatic parenchymal necrosis occurs rarely, and typicallyensues in high-grade lesions requiring embolization, particu-larly if trauma has affected a large area of the liver, and evenmore so if a portal lesion is associated with hepatic arterialinjury. Diagnosis and management are similar to those forgeneral complications of embolization after surgery [23,24].

Delayed surgery

Delayed laparotomy or laparoscopy deserves a genuine placein the arsenal for management of BHT that do not requireimmediate surgery. The literature usually considers DS as

36 C. Letoublon et al.

Figure 3. 14-year old male skier. Fall from ski lift. Stable condition on arrival. Immediate computerized tomography (CT) (a to e):grade IV—V fracture with major extravasation (b). Double embolization with coils (f). At the 30th hour, development of a syndrome ofintra-abdominal hypertension with vesical pressure of 20 mmHg, and major cytolysis (ASAT = 640 mg/dL). Delayed intervention: laparoscopicdecompression-lavage-drainage of 3.5 liters of blood and bile. Respiratory improvement but marked leukocytosis, low-grade fever, elevatedC-reactive protein (CRP). CT on day 5: absence of vascularization of the upper three-quarters of the right hepatic lobe (g to i). Anatomicright hepatectomy on day 6. Pathology findings: necrosis of 90% of the resected liver. Uncomplicated course thereafter.

37

Management of blunt hepatic trauma

a failure of NOM. This is unjustified in our opinion. In ourexperience, DS was performed in 20% of NOM cases, andrarely for extrahepatic causes. Three-quarters of the caseswere for peritoneal inflammatory syndrome or for abdom-inal compartment syndrome. DS typically was performedbetween day 2 and day 6. In over half of cases, the procedureconsisted of peritoneal lavage combined with wide exter-nal drainage to control bile leakage and create a controlledexternal biliary fistula. Currently, the initial approach to DSis usually laparoscopic (Fig. 3). For us, this mini-invasiveinterventional option has a clear advantage in efficacy overthe option of percutaneous drainage and lavage, and alsodecreases hospital stay. If delayed laparotomy is required,it can be performed more simply and under more favor-able conditions than immediate laparotomy, and the ‘‘loss ofchance’’ for these patients ‘‘deprived of the initial abdom-inal exploration’’, alleged by some authors, is unfounded[9,25].

Option 2 — immediate surgery and thesecond therapeutic trio: perihepaticpacking + postoperativeembolization + scheduled reoperation

Abbreviated laparotomy (damage control)with perihepatic packing

Emergency laparotomy has become rare (20% of cases), andis used for only the gravest presentations. When an injuredpatient cannot be stabilized by fluid resuscitation or is insevere shock on arrival, the entire management strategymust focus on avoiding loss of time. Three signs, if presentupon arrival of a patient in shock, predict the need forabbreviated laparotomy [5.15,16] because of the risk ofuncontrolled bleeding due to the ‘‘bloody vicious cycle.’’These warning signs, the so-called ‘‘lethal triad’’, are:• pH < 7.3;• temperature < 35 ◦C;• > 5 units of blood transfusion (> 10 for some, multiple

trauma).

In these cases, the patient must be taken directly tothe OR where preparation for laparotomy is performedquickly on the operating table, and ongoing resuscitationis completed. Delay in transfer to the OR is one of the best-identified causes of ‘‘preventable deaths.’’

In most cases the initial operating strategy can be sum-marized as follows:• skin preparation from xiphoid to pubis;• if there is major shock with risk of cardiac arrest, sub-

xiphoid compression of the aorta should be performeduntil the incision opens the way for direct intraperitonealaortic compression;

• long midline incision, which can be easily extended witha right subcostal incision (Fig. 4).

Once the peritoneal cavity is open, the strategy isadapted to the situation.

The liver is no longer bleeding. . .

After evacuation of the hemoperitoneum, we proceed to fullexploration of the abdominal cavity in search of associatedinjuries. There is no valid reason to systematically liberatethe liver and explore for damage, because there is a real

Figure 4. Xiphoid to pubis skin preparation. Long midline incisionextended to a right subcostal incision because of operative findings.

risk of triggering bleeding. After irrigation of the abdominalcavity with warm saline, simple drainage and abdominal wallclosure are in order. Postoperative monitoring and CT areorganized.

The liver is actively bleeding. . .

The primary goal is to stop the bleeding, without trying todetermine the exact type of lesions. The liver is compressed

closing it on itself and pressing it against the diaphragm.The suspensory ligaments of the liver are not divided, otherthan some possible adhesions to the lower right surface.The assistant replaces the operator in applying compression.The hepatic pedicle is encircled with a tape. The presenceor absence of other bleeding sites is quickly verified. Theincision can be adapted to the exploration. Surgical and/oranesthesia reinforcements may then be recruited. At thisstage of the procedure, several situations can be describedthat will dictate the ongoing operative strategy.

‘‘Favorable’’ location: direct hemostasis?If hepatic compression is effective, the hepatic bleedingsite is anterior and easy to visualize, and above all, if thereis no shock, hypothermia or transfusions > 5 units, and if thewhole team is accustomed to this type of surgery, we canthen decide to attempt exploration of the injury and directhemostasis. Hepatic pedicle clamping is performed. If thiseffectively slows bleeding, the edges of the hepatic woundcan be gently opened, and hemostatic sutures directlyplaced (Figs. 5 and 6). Closure with large mass suturesshould not be performed (Fig. 7). The duration of pedicleclamping should be short in this situation: if it exceeds20 minutes, it is clear that the gesture is more complex thananticipated. If profuse bleeding appears (measurement of

38 C. Letoublon et al.

Fa

bb

cvd

hlhsepcl

‘Iw

Figure 6. Individual ligation of bleeding vessels during pedicularc

igure 5. Pedicular clamping. Always search for a left hepaticrtery arising from the celiac axis.

lood loss in the suction may be surprising. . .), we return toimanual compression and perform PHP.

Sometimes, a nearly complete traumatic hepatectomyan be quickly completed. But anatomical hepatectomy hasirtually no place in emergency surgery for BHT. It is tooangerous [1].

When hemostasis can be simply obtained, withoutemodynamic collapse or prolonged pedicle clamping, aayer-by-layer parietal closure can be performed. Theepatic region must be widely drained to externalize anyecondary bile leakage. But associated lesions (intestinaldema after clamping, retroperitoneal hematoma, severeelvic fractures, severe head trauma. . .) may be an indi-ation for either one-layer skin closure or vacuum-assistedaparostomy (VALS) to prevent HIA and ACS [26].

‘Unfavorable’’ situations: perihepatic packing!f the ‘‘lethal triad’’ is present in a patient with BHT,e must diagnose the components of the ‘‘bloody vicious

F

coPo(

tausffsmvsbpa

m

lamping.

igure 7. Wide mass sutures should not be used for hemostasis.

ycle’’, i.e., acidosis, hypothermia and coagulopathy, inrder to decide the need for abbreviated laparotomy withHP. The systematic performance of PHP is more rigor-us than the ‘‘rustic’’ nature of the process might suggestFig. 8).

A compressive mattress consisting of large-sized laparo-omy pads is gradually formed to replace the hands of thessistant who has been compressing the liver on itself andpward against the diaphragm. A compressive mattress ofufficient thickness must first be placed under the right lobe,ar enough posterior so that the liver is pushed upward andorward supported by the kidney, then beneath the left lobeupported by the lesser gastric curvature. Do not place toouch packing in direct contact with the subhepatic inferior

ena cava (IVC) to avoid the risk of caval compression withubsequent anuria. Do not pack the hepatic fracture itself,ut use the packs to close the fracture on itself. Also avoidlacing packs over the dome of the liver to avoid opening up

lesion that involves the right hepatic vein (Fig. 9).Simple one-layer skin closure is the rule; it is fast and

inimizes the risk of secondary IAH. It allows a parietal

Management of blunt hepatic trauma 39

Figure 8. Grade III or more hepatic fracture (a). Bimanual compressiothe hepatic fracture (d).

Figure 9. A very dangerous maneuver: retracting on the liver inorder to ‘‘see better’’. This may open up high lesions and aggravatethe bleeding (a). One should also avoid the dangerous practice ofplacing packing above the liver (b).

n (b). Perihepatic packing (c). Avoid placing packing material into

relaxation of more than 6 cm to allow for further subsequentdistension. If this closure is under tension, the option is forVALS attached to continuous suction at 100 mmHg (50 mmHgin patients with residual bleeding) [26,27]. The high price ofcommercial foams has led us to avoid their use, particularlywhen a reoperation is very likely within two days.

At this point in time, ongoing management raises a choicebetween pursuing resuscitation or immediate transfer to theCT suite for detection and assessment of active blood leak-age. Transfer to a specialized trauma center may also bediscussed at this point.

Pedicle clamping is ineffective. . .

One must first seek out technical reasons for failure of

pedicle clamping to control bleeding; clamp incompletelytightened down, an unappreciated left hepatic artery aris-ing from the left gastric artery. If this is not the case, thereis probably a leak from a hepatic vein. Only very rarely isthere a gesture that is feasible and simple (An obvious lefthepatic vein tear, for example). Most often the patient hasalready bled a lot and the lethal triad is developing: thesurgeon should remember that liver compression and PHPwill control the vast majority of venous injuries, even whenlocated high [28].

Perihepatic packing is ineffective. . . but a Pringlemaneuver improves the situationIn such cases, the surgeon must reinforce PHP by manuallycompressing the liver, adding more packs, wedged againstthe kidney posteriorly, or against the stomach on the leftside.

If reinforcement of the packing is ineffective, pedi-cle clamping must be added to the PHP. If this decreasesthe bleeding, it is likely that arterial bleeding is involved.The common hepatic artery should be individually clamped(clamping of the individualized right hepatic artery is the-oretically better, if possible). Finally, if this fails to controlbleeding, one can consider hepatic artery ligation or place-ment of a small bulldog-type clamp that is left in place

4 C. Letoublon et al.

uWiptowb(tlPwecctiqtacscc

TtIa•

•

C

Iieooaoa

0

ntil reoperation since the laparotomy must be abbreviated.hile this gesture has been lifesaving for some patients,

t can also cause hepatic necrosis or increase the risk ofostoperative liver failure due to hypoxic damage to a liverhat is already stressed by hypotensive shock [6]. Becausef these risks, arterial ligation should be reserved for caseshere arterial embolization is impossible. Another possi-ility to consider is intra-operative arterial embolizationrarely possible), or immediate postoperative emboliza-ion, which can be more distal, more effective and causeess ischemia than ligation. One can opt for aggressiveHP followed by immediate transfer to the radiology suiteith ongoing maximal resuscitation for angiography andmbolization. In the experience of Asensio et al., this pro-edure stopped all bleeding uncontrolled by PHP [29]. Weurrently prefer the following maneuver, which we haveested: hepatic artery occlusion (or of the entire pediclef the artery cannot be isolated) by a vessel loop tourni-uet brought out through the incision; the patient is thenransferred in extreme urgency to the angiography suite forrterial embolization. The tourniquet occluding the arteryan then be released at a favorable moment (Fig. 10). Forituations of intermediate gravity, the vessel loop tourniquetan be put in place but tightened down only if the patient’sondition deteriorates during transfer.

he Pringle maneuver does not improve the situa-ion or the packing is ineffectivef the rate of bleeding is too heavy to allow transfer tongiography, the choice is then limited to two options:either the combination of PHP + closure + warming +maximal fluid resuscitation on the table, and reoperationif the situation seems to be improving;hepatic vascular exclusion. This is an aggressive but dan-gerous surgical approach: one of the surgeons maintainscompression of the liver against itself and against thediaphragm, while another obtains exposure by extend-ing the incision widely and placing suitable retractors.Liver vascular exclusion can then, in theory, have a role

to play in such exceptional conditions. In this situation,quadruple clamping is essential to avoid unpriming of thecardiac pump and cardiac arrest. In order to avoid lossof prime of the heart, venous return through the IVCmust be preserved. Many methods have been described:internal shunts to the IVC or veno-venous extracorporealcirculation (similar to that used in liver transplantation).While shunts are reputed to have had more articles writ-ten about them than patient lives they have saved, andwhile they have been useful for some open traumatic liverinjuries (that would probably have been better handled byPHP), we can say that shunts and emergency extracorpo-real circulation (ECC) have no place except for desperatecases that fail to respond to correctly applied safer meth-ods (PHP + embolization. . .) [28].

T and postoperative embolization

f the first intervention has effectively controlled bleed-ng, whole body contrast-enhanced CT should be performed,ither immediately upon leaving the OR or after 10—12 hoursf active resuscitation. But if the initial procedure fails tobtain satisfactory hemostasis, CT and then highly-selectiverterial embolization should be performed without delay. Inur experience, postoperative embolization was required in

third of cases.

Fiutvo

Er

RpApboc

ltbfrod

igure 10. Lifesaving tourniquet placement for hemorrhage thats not controlled by perihepatic packing (PHP) but responds to pedic-

lar clamping. If possible, the vessel loop should be placed aroundhe isolated common hepatic artery (window). If impossible, theessel loop should encircle the entire hepatic pedicle; there is riskf hepatic ischemia if pedicular clamping is prolonged.

arly postoperative resuscitation and eventualeoperation

esuscitation of the patient with perihepaticacking in placet this stage, one can envisage transfer the traumaatient to a specialized trauma center. Resuscitation isased primarily on volume replacement and correctionf hypothermia and acidosis to stop the bloody viciousycle.

For severe shock, vascular volume repletion may be veryarge in order to maintain adequate hemodynamics andissue perfusion while correcting other metabolic distur-ances. Many units of red blood cell transfusion, thawedresh frozen plasma and platelets may be required. The cor-ection of hypothermia is important and combines all meansf external and internal warming. The correction of acidosisepends primarily on correction of hypothermia and shock.

Management of blunt hepatic trauma

Correction of coagulation defects is usually obtained onlyafter several hours. While the importance of fresh bloodproducts in this situation has been highlighted, the role ofrecombinant factor VII remains to be validated, even thoughthis medication has sometimes been attempted [30]. Twoevents of different severity can occur.

Major bleedingMajor bleeding (exceeding one unit per hour) that persistsafter one to two hours of resuscitation and that makestransfer to radiology for CT impossible poses the problemof early reoperation. This becomes imperative even if thelethal triad has not been corrected, in hope of discoveringan undetected extrahepatic bleeding source or inadequatePHP compression. Otherwise, arterial embolization canrectify the situation either by stopping bleeding or bydetecting a bleeding source that requires operation. Somerecommend routine arteriography with embolization afterPHP if possible [29].

Intra-abdominal hypertensionIAH may develop early and is a particular threat to patientsafter PHP who have undergone vascular clamping, thosewho were in severe shock or required multiple transfu-sions, and those in whom a complete layered closure ofthe wall was performed. IAH is observed in 15% of casesof PHP [31]. IAH may result in ACS, a true Volkmann’s com-partment syndrome of the abdomen that directly threatensvital organ functions. Preventive measures such as limita-tion of crystalloid resuscitation, skin-only closure, or VALShave decreased its incidence [32]. Signs suggestive of ACS inthis post-PHP setting are abdominal wall tightness, anuria,ventilatory difficulty with high ventilation pressures, wors-ening hypotension, embarrassment of IVC venous returnand cardiac compression. IAH also has negative effects oncerebral perfusion. Diagnosis of ACS is based on measure-ment of intravesical pressure (or, if a bladder lesion oflarge pelvic hematoma prevents this, by measurement ofintragastric pressure) with a threshold of > 15 mmHg [19].Any delay in decompression is dangerous [32]. Abdominalwall tension must be relaxed by VALS. When the underly-ing causes of ACS have disappeared after two or three days,

it is often possible to close the abdominal wall and oftena multi-layer closure is possible at the time of removal ofpacking.

One must also keep in mind that, while PHP saves manypatients, it may also contribute to liver ischemia if it istoo compressive, or also if it is associated with prolongedpedicle clamping (or hepatic artery or segmental branchligation), embolization, or prolonged severe shock. In somecases, a clinical picture of postoperative hepatic failure isdescribed, usually accompanied by major cytolysis. The sus-picion of a mechanical hepatic complication is an indicationfor re-exploration, but some ‘‘shock liver’’ is due to severeischemia from major and prolonged hypotension. The pos-sibility of an authentic partial necrosis of the liver must bekept in mind (Fig. 3) [23,33].

Scheduled reoperation after perihepaticpackingPrior to reoperation, a comprehensive clinical and radiologicassessment must be completed. Whole body CT angiographyis essential, especially if the patient is unconscious. Sec-ondary elective laparotomy after PHP takes place betweenday 1 and day 6. It must address procedures performed inassociation with PHP as well as general conditions: intestinal

41

staple closures argue for early reoperation within two days,but severe neurosurgical lesions or pulmonary insufficiencymay require a delay in reoperation. A prolonged interval upto the 6th day is not deleterious. Patient transfer to a spe-cialized trauma center may also be considered during thisinterval.

This reoperation must be carefully prepared: full CTangiography is essential pre-operatively and all technicalpossibilities and skills must be available (triple or quadru-ple clamping, veno-venous ECC). But in most cases, there-intervention is simple and straightforward: after evacu-ation of hemoperitoneum and exploration of the abdominalcavity, the packs are soaked in warm saline and gentlyremoved. In the majority of cases, the bleeding has stopped.Sometimes a bile leak is obvious, escaping from a hepaticfracture or a more or less deep contusion: bile leaks shouldbe widely drained to direct a controlled bilio-cutaneous fis-tula. Neither cholecystostomy nor cholangiography shouldbe performed. If significant bleeding occurs, rare if oneis careful, the pedicle is clamped and search is made forthe bleeding site after gently exposing the surfaces of thehepatic fracture. If bleeding persists despite clamping, PHPmust once again be performed, although the packing is oftenless extensive. At this point, one has the choice of closingthe abdomen with a planned removal of PHP in two or threedays, or of embarking upon major surgery (mobilization ofthe hepatic ligaments while maintaining compression andavoiding tears, or vascular exclusion of the liver, or ECC),but, if the surgeon has not already ‘‘gone too far’’, a repeatPHP may achieve hemostasis and allow patient transfer to aspecialized trauma center.

When hepatic necrosis, consisting of pale tissue thatis more or less demarcated, is present and there areno CT signs or laboratory findings to suggest massiveparenchymal necrosis, one can, at this relatively early re-intervention, either drain this area or perform a limitednecrosectomy, but ongoing vigilance for massive necrosis isnecessary.

The management of complications that

may arise with either option

Whether or not the patient has undergone surgery, severalsimilar complications or events, which are consequences of aconservative and cautious mindset, may occur. These eventsare primarily biliary and vascular in nature, the latter inassociation with the onset of ischemia.

Biliary fistula

In all cases where external drainage has been placed, theprobability of bile leakage increases with the severity ofthe lesion. Maturation of a drainage tract and appreciationof its flow require 10—12 days without mobilization of thedrain. If the flow is large (more than 200—300 cm3/day), amagnetic resonance cholangio-pancreatography (MRCP) canprovide information on the source of the leak. But after thisinitial period, the drain can be gradually removed over aweek’s time in almost all cases. If a high-flow leak persists,endoscopic placement of a nasobiliary drainage or biliarystent can be proposed. This approach may speed healing[34]. In our experience, all biliary fistulas healed sponta-neously without any additional intervention, but the healingperiod could exceed two months.

4

G

Hgac

ITrol

HhPeapifssnTbiaaarpes

issorobw[

mpooa

C

Higaiaralo

ic

D

T

R

[

[

2

allbladder and liver ischemia

epatic arterial embolization often increases the risk ofallbladder and liver ischemia. In our experience with 35rterial embolizations, two patients developed ischemicholecystitis and five had symptomatic hepatic necrosis.

schemic cholecystitishe frequency of ischemic cholecystitis varies in differenteports, but is generally in the range of 15% [24]. Occlusionf the cystic artery is probably responsible. This may developate and require cholecystectomy.

epatic parenchymal necrosis andepatectomy for liver traumaarenchymal necrosis is usually evident on CT. This canvolve aseptically with progressive atrophy, but there islways the risk of infected necrosis with suppuration. Therobability of developing necrosis increases with the sever-ty of the lesion, the performance of embolization, the needor PHP, or the presence of portal vein injury at the out-et [2]. Markedly elevated transaminases, jaundice, a septicyndrome, and CT images of persistent hypoperfusion oron-perfusion must raise the question of hepatic necrosis.he incidence of necrosis after arterial embolization haseen reported to reach 41% [33], but it has been much lowern our experience. Massive necrosis requires a resectionfter a few days, one of the few indications for hepatectomyfter BHT. One must choose between a formal anatomic hep-tectomy or necrosectomy, which is simpler but carries theisk of a prolonged postoperative course due to residual sup-uration or biliary fistula. Anatomic hepatectomy can bextended into healthy liver tissue, potentially leading to aafer postoperative course.

The need to perform secondary hepatectomy cannot,n our opinion, plead for a return of an initial aggressiveurgical attitude to the management of severe BHT; analy-is of the rare publications that celebrate resection show,n one hand, that mortality with the surgical approach

emains very high, and that, on the other hand, good resultsccurred in patients who were initially treated (and safely!)y PHP and then transported to specialized trauma centershere the decision for intervention was made without haste

35].Liver transplantation for BHT is exceptional, indicated

ainly for total hepatic necrosis. This rarely performedrocedure consists of total hepatectomy for control of hem-rrhage combined with portocaval anastomosis to allow ane or two day anhepatic interval while awaiting the avail-bility of a donor liver for transplant [36,37].

onclusion

emorrhagic shock remains the primary decisional elementn the management of BHT. Trauma victims who do noto immediately to the operating room should undergo CTssessment after stabilization. The most significant progressn management has been the adoption of a conservativettitude based on nonoperative management with vigorousesuscitation, control of hypothermia and coagulopathy,nd radiologic skill for arterial embolization. Delayedaparotomy is a very useful tool and is an integral partf the initial non-interventional strategy. When surgery

[

[

[

[

[

[

[

[

C. Letoublon et al.

s necessary, combined abbreviated laparotomy and PHPonstitute the safest approach.

isclosure of interest

he authors declare that they have no competing interest.

eferences

[1] Richardson JD. Changes in the management of injuries to theliver and spleen. J Am Coll Surg 2005;200:648—69.

[2] Letoublon C, Abba J, Arvieux C. Traumatismes fermes dufoie. Principes de tactique et de technique chirurgicales. EMCTechniques chirurgicales-Appareil Digestif 2012:1—23 [Article40-785].

[3] Letoublon CCD. Les traumatismes fermés du foie. Paris:Arnette; 1996. p. 158.

[4] Burch JM, Ortiz VB, Richardson RJ, et al. Abbreviated laparo-tomy and planned reoperation for critically injured patients.Ann Surg 1992;215:476—83 [discussion 483—474].

[5] Arvieux C, Letoublon C. [Abbreviated laparotomy]. J Chir(Paris) 2000;137:133—41.

[6] Calne RY, McMaster P, Pentlow BD. The treatment of majorliver trauma by primary packing with transfer of the patientfor definitive treatment. Br J Surg 1979;66:338—9.

[7] Ciraulo DL, Luk S, Palter M, et al. Selective hepatic arte-rial embolization of grade IV and V blunt hepatic injuries: anextension of resuscitation in the nonoperative management oftraumatic hepatic injuries. J Trauma 1998;45:353—8 [discus-sion 358—359].

[8] Carrillo EH, Reed Jr DN, Gordon L, et al. Delayed laparoscopyfacilitates the management of biliary peritonitis in patientswith complex liver injuries. Surg Endosc 2001;15:319—22.

[9] Letoublon C, Chen Y, Arvieux C, et al. Delayed celiotomy orlaparoscopy as part of the nonoperative management of blunthepatic trauma. World J Surg 2008;32:1189—93.

10] Mirvis SE, Whitley NO, Vainwright JR, Gens DR. Blunt hepatictrauma in adults: CT-based classification and correlation withprognosis and treatment. Radiology 1989;171:27—32.

11] Monnin V, Sengel C, Thony F, et al. Place of arterial emboliza-tion in severe blunt hepatic trauma: a multidisciplinary

approach. Cardiovasc Intervent Radiol 2008;31:875—82.

12] Kozar RA, Moore FA, Cothren CC, et al. Risk factors for hepaticmorbidity following nonoperative management: multicenterstudy. Arch Surg 2006;141:451—8 [discussion 458—459].

13] Pachter HL, Knudson MM, Esrig B, et al. Status of nonoperativemanagement of blunt hepatic injuries in 1995: a multicenterexperience with 404 patients. J Trauma 1996;40:31—8.

14] Ye B, Miao YD. Acute liver failure secondary to hepatic com-partment syndrome: case report and literature review. UlusTravma Acil Cerrahi Derg 2014;20:136—8.

15] Nissen NN, Geller SA, Klein A, et al. Percutaneous liver biopsyafter living donor liver transplantation resulting in fulminanthepatic failure: the first reported case of hepatic compartmentsyndrome. J Transplant 2010;2010:273578.

16] Markert DJ, Shanmuganathan K, Mirvis SE, et al. Budd-Chiarisyndrome resulting from intrahepatic IVC compression sec-ondary to blunt hepatic trauma. Clin Radiol 1997;52:384—7.

17] Pretre R, Mentha G, Robert J, et al. Acute segmental arterio-portal shunt with reversal of portal blood flow after atypicalresection for hepatic injury. Eur J Surg 1993;159:313—5.

18] Chen RJ, Fang JF, Chen MF. Intra-abdominal pressure monitor-ing as a guideline in the nonoperative management of blunthepatic trauma. J Trauma 2001;51:44—50.

19] Cheatham ML, White MW, Sagraves SG, et al. Abdominal per-fusion pressure: a superior parameter in the assessment ofintra-abdominal hypertension. J Trauma 2000;49:621—6 [dis-cussion 626—627].

[

[

[

[

[

[

[

[

[

[

[30] Smith JE. The use of recombinant activated factor VII (rFVIIa)

Management of blunt hepatic trauma

20] Tamura N, Ishihara S, Kuriyama A, et al. Long-term follow-upafter nonoperative management of biloma due to blunt liverinjury. World J Surg 2015;39:179—83.

21] Schouten van der Velden AP, de Ruijter WM, Janssen CM,

et al. Hemobilia as a late complication after blunt abdominaltrauma: a case report and review of the literature. J EmergMed 2010;39:592—5.

22] Glaser K, Wetscher G, Pointner R, et al. Traumatic bilhemia.Surgery 1994;116:24—7.

23] Dabbs DN, Stein DM, Scalea TM. Major hepatic necrosis: a com-mon complication after angioembolization for treatment ofhigh-grade liver injuries. J Trauma 2009;66:621—7 [discussion627—629].

24] Mohr AM, Lavery RF, Barone A, et al. Angiographic emboliza-tion for liver injuries: low mortality, high morbidity. J Trauma2003;55:1077—81 [discussion 1081—1072].

25] Polanco PM, Brown JB, Puyana JC, et al. The swingingpendulum: a national perspective of nonoperative manage-ment in severe blunt liver injury. J Trauma Acute Care Surg2013;75:590—5.

26] Letoublon C, Reche F, Abba J, Arvieux C. Damage controllaparotomy. J Visc Surg 2011;148:e366—70.

27] Barker DE, Kaufman HJ, Smith LA, et al. Vacuum packtechnique of temporary abdominal closure: a 7-year experi-ence with 112 patients. J Trauma 2000;48:201—6 [discussion206—207].

28] Beal SL. Fatal hepatic hemorrhage: an unresolved problemin the management of complex liver injuries. J Trauma1990;30:163—9.

29] Asensio JA, Roldan G, Petrone P, et al. Operative manage-ment and outcomes in 103 AAST-OIS grades IV and V complex

[

[

[

[

[

[

[

43

hepatic injuries: trauma surgeons still need to operate, butangioembolization helps. J Trauma 2003;54:647—53 [discussion653—644].

in the management of patients with major haemorrhagein military hospitals over the last 5 years. Emerg Med J2013;30:316—9.

31] Pottecher T, Segura P, Launoy A. [Abdominal compartment syn-drome]. Ann Chir 2001;126:192—200.

32] Balogh ZJ, Lumsdaine W, Moore EE, Moore FA. Postinjuryabdominal compartment syndrome: from recognition to pre-vention. Lancet 2014;384:1466—75.

33] Letoublon C, Morra I, Chen Y, et al. Hepatic arterial emboliza-tion in the management of blunt hepatic trauma: indicationsand complications. J Trauma 2011;70:1032—6 [discussion1036—1037].

34] Yuan KC, Wong YC, Fu CY, et al. Screening and managementof major bile leak after blunt liver trauma: a retrospectivesingle center study. Scand J Trauma Resusc Emerg Med 2014;22:26.

35] Li Petri S, Gruttadauria S, Pagano D, et al. Surgical manage-ment of complex liver trauma: a single liver transplant centerexperience. Am Surg 2012;78:20—5.

36] Patrono D, Brunati A, Romagnoli R, Salizzoni M. Liver trans-plantation after severe hepatic trauma: a sustainable practice.A single center experience and review of the literature. ClinTransplant 2013;27:E528—37.

37] Kaltenborn A, Reichert B, Bourg CM, et al. Long-term outcomeanalysis of liver transplantation for severe hepatic trauma. JTrauma Acute Care Surg 2013;75:864—9.