War Liver Injuries

Vojnomedicinska akademija, Klinika za opštu i vaskularnu hirurgiju, Beograd Aim. To provide a retrospective analysis of our results and experience in primary surgical treatment of subjects with war liver injuries. Methods. 204 subjects with war liver injuries were treated. A total of 82.8% of the injured were with the liver injuries combined with the injuries of other organs. In 93.7%, the injuries were caused by fragments of explosive devices or bullets of various calibers. In 140 (68.6%) of the injured there were minor lesions (grade I to II), treated with simple repair or drainage. There were complex injuries of the liver (grade III−V) in 64 (31.4%) of the injured. Those injuries required complex repair (hepatorrhaphy, hepatotomy, resection débridement, re-section, packing alone). The technique of perihepatic packing and planned reoperation had a crucial and life-saving role when severe bleeding was present. Routine peritoneal drainage was applied in all of the injured. Primary management of 74.0% of the injured was performed in war hospitals. Results. After primary treatment, 72 (35.3%) of the injured were with postoperative complications. Reoperation was done in 66 injured. Total mortality rate in 204 injured was 18.1%. All the deceased had significant combined injuries. Mortality rates due to the liver injury of the grade III, IV and V were 16.6%, 70.0% and 83.3%, respectively. Conclusion. Complex liver injuries caused very high mortality rate and the management of the injured was delicate under war circumstances (if the injured reached the hospital alive). Our experience under war circumstances and with war surgeons of limited knowledge of the liver surgery and war surgery, confirmed that it was necessary to apply compressive abdominal packing alone or in combination with other techniques for he-mostasis in the treatment of liver injuries grade III−V, resuscitation and rapid transportation to specialized hospitals.


Introduction
The liver is one of the most frequently injured organs in either war or general abdominal trauma (1−8).
Statistics obtained from field experience involving injuries of the liver showed a progressive decrease in the mortality rate: 60% in World War I, 27% in World War II, 14% in Korea and 8.5% in Vietnam (8).
Application of the principles established by military experience had greatly improved the management of hepatic trauma in civilian trauma centers.According to the data from recent civilian series, the mortality rate of 10 to 20% has been reported, with exsanguination as the leading cause of death (2−5).The majority of the hepatic injuries in civilians were minor (grade I to II), and required minimal or non-surgical treatment (3,4).However, the significant mor-tality rate, that often exceeded 50%, was associated with more complex and combined injuries (grades III−V).In the last decades, morbidity and mortality rates were decreased by better surgical techniques, perihepatic packing with planned re-exploration, and non-surgical treatment of hepatic injuries in selected, hemodynamically stable patients, followed up by computer tomography (CT).Severity of liver war injuries varied and prognosis depended on the grade of injury, the time interval from the injury infliction to the medical and surgical treatment, combined injuries and conditions of treatment.Penetrating abdominal trauma index (P.A.T.I.) score quantifies the risk of complications following the combined penetrating abdominal trauma (9).Although the basic principles (control of hemorrhage, removal of devitalized tissue, and perihepatic drainage) of management were well established, the optimal techniques for treatment of the liver injuries, especially under war conditious, remained controversial (1,3,5,6).
In most cases, war hospitals were organized in places with insufficient material and technical capabilities and with medical teams including one or two general surgeons or residents in general surgery.Surgical management of casualties depended on available material, conditions of the combat, dynamics of the injured influx, surgeon's experience, and the frequency of evacuation roads' disruption.
The aim of this retrospective study was to demonstrate our results and experience in surgical treatment of the liver injuries in war hospitals of poor technical capabilities.

Methods
This retrospective study included 204 injured from the war zone in former Yugoslavia, with the liver injuries, treated in war hospitals and at the Clinic of General and Vascular Surgery of the Military Medical Academy (MMA), from July 1991 to December 1999.Data were used from medical records.For the purposes of this analysis, the injured were divided into two groups of simple or complex management, respectively, depending on the type and the extent of the liver management.
At our hospital, the initial surgical treatment was done in 54 injured (26.0%).Most of the other injured (74.0%) were treated in war hospitals near the war zone and after primary surgical treatment, evacuated to MMA for further treatment.All the injured suffered from combined injuries and their survival after primary treatment was uncertain.Most of them showed signs of hemorrhagic shock and were evacuated to MMA within 24 hours after primary surgery.
The average age in that group was 26.5 years.The cause of injury in 93.7% cases were fragments of explosive devices and bullets of various calibers, and blunt trauma, blast or blades in other cases.Only 35 (17.2%) of the injured had the isolated liver injury.Combined injuries of the liver and other organs (right colon, duodenum, pancreas, kidney, small bowel, lung, spleen, diaphragma, head, extremities) were present in other injured (82.8%).In the group of injured with penetrating abdominal wounds, the liver injuries were at the third place, after the injuries of the small and large bowel (Table 1).We identified the time interval from the injury to the initial surgical management only in 132 (64.7%) in that group of the injured.Sixty-three (47.7%) of them were operated on in the period of 1−6 hours after the injury.Resuscitation treatment of the injured was minimal before admission to the hospital.The data had shown that only 67 (32.8%) of the injured had received infusion of Ringer lactate before admission to the hospital.
In most of the injured, after necessary resuscitation procedures, war surgeon usually performed laparotomy, and further decision was based on general condition of the injured, cause and position of penetrating abdominal wounds.In those cases, the type of the liver lesion was confirmed during surgery.In the stable injured (blunt trauma), besides preoperative examination, abdominocentesis and peritoneal lavage were also done.Abdominal ultrasonography and CT and also selective angiography were performed in only 3 injured (MMA) before primary surgical treatment (Figures 1  and 2).
In most of the injured, the explorative laparotomy was done through medial incision.If necessary, it was prolonged sub-or bisubcostally.Thoraco-phreno laparotomy was done in 16 injured.According to Moore classification, minor liver injuries grade I or II were found in 140 (68.6%) of the injured.In 64 (31.4%) injured, complex liver injuries (grade III -V) were present (surgical procedures were presented in Table 2).Liver injuries of the grade I and II were initially treated with hemostatic agents, electrical cauterisation, suture and temporary packing with normal warm saline.In the injuries without bleeding, surgery was not done.Liver inju-    ries of the grade III and IV without injury of retrohepatic part of the inferior vena cava (IVC), were treated, after bleeding control (Pringle procedure or isolated clamping of vascular elements of hepatoduodenal ligament), with hepatorraphy (deep liver sutures), resectional debridement or resection, omental packing and perihepatic packing, often in combination.Resection -resectional debridement and suture of IVC were done in 3 injured with grade V lesions (in 1 injured in total vascular exclusion -MMA).
After management of the liver injury grade III and IV, common bile duct drainage through a T-tube was not used, but in all the injured it was necessary to do passive, active, or combined perihepatic drainage to eliminate the rest of the blood, necrotic tissue and bile.

Results
The majority (68.6%) of the injured was treated with simple management.
Advanced techniques for hemostasis such as extensive hepatorrhaphy (deep liver sutures) or hepatotomy with selective vascular ligation and omental packing, resectional debridement or resection with selective vascular ligation, selective hepatic artery ligation, omental packing, or perihepatic packing were required in 64 (31.4%) of the injuried (Table 2).The extended right hepatectomy with total vascular exclusion due to lesion of the right liver lobe was performed in one of the injured.The mortality rate in the group of 64 with complex injuries was 31.2% (20/64) (Table 3).Postoperative complications, depending on the liver lesion type and on surgical procedure, were present in 72 (35.3%) of the injured (Table 4).More than two complications were present at the same time in 36 injured.The complication frequency was directly correlated with the severity of the liver injuries, as well as with the severity of combined injuries.
Penetrating abdominal trauma index was 25 and higher in 92 (45.1%) of the injured.In most of the cases, complications that depended on the type of injury (grade III−V according to Moore) and surgical liver treatment, were the sequelae of inadequate primary treatment of the liver lesion (insufficient debridement of necrotic tissue, insufficient ligation of bile-vascular elements, missed lesions of hepatic veins and IVC, uncontrolled sutures of parenchyme in the place of the injury).Reoperation was done in 66 injured (in 16 injured more than twice): in 12 patients with lesion of grade III and IV, with only primary perihepatic packing, the planned reexploration was performed due to the management of the injury -resectional debridement and hepatic artery ligation.The pack was removed in the last 72 hours.No complication appeared due to the presence of the pack, but with packing or the temporary hemostasis (or if the bleeding was tolerable) the evacuation to the definitive surgical management was made possible.The only patient who survived with the liver lesion of the grade V was primarily managed by resectional debridement and packing, and the final management of the right hepatic vein injury to the wall of suprahepatic part of IVC was performed during reoperation.Due to that procedure all the injured with tamponade of the liver injury in combination with incomplete management of lesion (22 injured) survived up to the definitive management.In 39 injured hepatic bleeding was the sequela of incomplete vessels' ligation in the place where debridement was performed (21 injured), lesion of the hepatic artery that was not treated (8 injured), lesion of hepatic or portal vessels that was not treated (6 injured), bleeding after suture of IVC and the right hepatic vein (4 injured).
All the patients with deep liver sutures went through the reoperation due to necrosis of the liver tissue and abscess (12 injured).
In 4 patients with injury of the right liver lobe (grade IV), hemobilia was a sequela of unmanaged rupture of the right hepatic bile duct and the right hepatic artery (Figure 3).
A total mortality rate in the group of 204 injured was 18.1% (37/204) (Table 3).All those who died were with severe combined injuries (P.A.T.I. > 25).Sepsis and multiple organ failure sepsis (MOFS) were the cause of death in 17 injured.The liver injury was the cause of death in 20 injured.The mortality rate due to the liver injury of the grade III, IV and V was 16.6%, 70.0% and 83.3%, respectively (Table 3).
Five injured died of exanguination during the operation (fatal injuries were of the grade IV and V), due to severe uncontrollable hepatic hemorrhage.
The rest of the injured died in the first 30 days after the surgery and the cause of death was prolonged hemor-rhage, coagulopathy, hepatorenal failure, and lung thromboembolia.

Fig. 3 − A cholangiogram (ERCP) demonstrated the injury of
the right branch of biliary system causing hemobilia.This injury of the liver was treated operatively.
Pringle's procedure with duration of 20−45 minutes was a routine for most of the injured who required complex management.The injured in that group received 8-17 units of blood during the surgery.

Discussion
During the thirty-year period after the World War II, the management of war and civilian injuries of the liver was relatively unchanged.The time of Pringle's procedure was limited to 15−20 minutes.Considering control of the haemorrhage, intrahepatic packing was replaced by deep sutures of the liver parenchyme, and resection was frequently used.
However, the experience of the last twenty years showed that deep sutures of the liver parenchyme were replaced by typical resection of the liver in the treatment of complex liver injuires.Modern surgical procedures gave the advantage to non-resection techniques, like hepatotomy with direct ligature of bilio-vascular elements, selective ligatures of the hepatic artery, limited resective debridement, omental packing and perihepatic packing (l−6, 10, 11).
The war liver injuries were usually caused by projectiles and fragments of explosive devices, which was confirmed by our experience (8,10,12,13).Isolated injuries were rare.According to various authors (2, 3, 8, 14−17), combined penetrating trauma of abdomen and/or extra-abdominal structures, including the liver injuries, were the most frequent ones, which was the proof of the complexity of surgical treatment of the injured.
Total of 82.8% of the injured in our study had the liver injuries combined with the injuries of other organs.
Considering the frequency of war injuries of the abdomen, the liver was at the third place, after the injuries of small bowel and colon.Pailler et al. reported similar results (8).
The Moore's anatomical classification of the liver injuries (19) was mostly used in the previous researches (5,6,18).
The liver injury is very time-sensitive, so the final outcome depends on the velocity and quality of the applied treatment.The number of the injured who survived and reached the field hospitals depended on the extent of preclinical help (resuscitation) and speed of the transport.In our research, all the wounded with injuries of the grade III−V (Moore's classification) were with hemorrhagic schock with highly changed hemodynamic parameters, prior to the primary surgical treatment in war hospitals.
Considering the critical state of the injured, according to our experience, as well as the experience from other wars (8,12,16,20,21), the mere physical examination and the reconstruction of the course of projectiles/fragments were often enough for the liver injuries to be suspected.The diagnosis in the stable patients should be complemented with abdominocentesis, peritoneal lavage and ultrasound examination (blunt trauma).
The results of our study as well as the results of other authors indicated that the war injuries of the grade II and III were the most frequent (8,16,20), while civilian injuries were mostly of the grade III and IV (2, 3, 22−24).
The liver injuries of the grade I and II were extremely significant from the point of view of the war surgeon, because the final outcome depended on the possibility of combined injuries and their results (1,8,16,25).The injuries of the above mentioned grades were verified and treated routinely, after the explorative laparotomy indicated for other reasons.Minor lesions of the liver parenchyme didn't result in morbidity and mortality specific for that kind of injury (20).
On the other hand, injuries of the liver parenchyme of the grade III and IV caused significant mortality.According to the results of civilian series (1,22,23), the mortality rate was from 6.5% to 54%.In cases of lesions combined with the injuries of retrohepatic veins, according to Buechter et al. (6), the mortality rate reached 100%, and it didn't depend on parenchymal lesions.The same authors suggested that the initial liver injury dictated the extent of surgical treatment, so that surgical trauma was also important for the final outcome.Some of the results published for civilian series (1−3, 7, 9, 10) indicated that Pringle's treatment combined with tissue-saving non-anatomic debridement of devitalised tissues using finger fracture technique with direct ligature of the open bilio-vascular vessels, and omental packing, were the optimal treatment of liver injuries of the grades III and IV.Using this procedure on large series Pachter et al. (1) achieved 6.5% mortality rate, Feliciano et al (2) 33.6%.In the multicenter study of Cogbill (23) the mortality rate was 7−30%.Beal (10) applied these techniques in 44% of the cases in the group of 121 injured with complex injuries.The outcome was successfully in 87% of the cases.
In our injured treated with complex procedures, the mortality rate caused by injuries of the grade III and IV was 16.6 and 70%.The period of warm ischaemia during Pringle's treatment in our series was 20−45 min.According to the experience of elective surgery it was up to 90 min (1,7,9,10).If that treatment did not control the hemorrhage, it indicated the lesion of hepatic veins and/or IVC.
In 12 of the injured in our field hospitals, the liver sutures set deep around the lobus ruptures or primary abdominal cavity caused prolonged bleeding (they were insufficient in the control of hemorrhage) and additional necrosis of the liver tissues.
According to our results as well as the results of other authors (2,3,23) hepatotomy with selective suture − ligature of bilio-vascular elements had an advantage over the use of deep sutures in controlling the hemorrhage in deep ruptures or temporary cavities.
We did not apply T-drainage of common bile duct as a supplementary treatment of the injured with complex injuries because it did not prevent the biliary fistula (7).
Selective ligature of hepatic artery was applied only in cases with the lesion of the hepatic artery branch.
Injuries of the liver parenchyme combined with the injurues of hepatic veins and/or IVC (grade V) caused very high mortality rate, and the management of the injured was very complicated under war circumstances.Quick orientation of the extent injuries was in that case obligatory, and it could be done by wide medial laparatomy extended to phrenolaparotomy (through the right seventh or eight intercostal areas).Fast control of hemorrhage under war circumstances, according to our experience is reduced to Pringle's treatment and the attempt to clamp directly the infra-and suprahepatic segment of IVC and after that to lesion management.The alternatives to that treatment were: perihepatic compressive tamponade with abdominal sponge, followed by resuscitation measures and fast transport to specialised hospitals.Other measures for bleeding control (10,23) are reserved for large trauma centres or hospitals specialized for the liver surgery.
Two of the injuried were transported to MMA after the treatment with perihepatic packing in the control of bleeding, and they were reoperated while in the state of serious hemorrhagic shock.One of them died during the operation.In the second patient, the injures of the right liver lobe, right hepatic vein and suprahepatic IVC segment were successfully managed after the removal of packing.The third one with the liver injuires grade V combined with thoracoabdominal injuries caused by explosion, including the right part of the liver, was initially treated at MMA after the quick helicopter transport.In that patient, we performed extended hepatectomy on the right side of the liver, and the suture of the IVC wall with the total vascular exclusion, as well as an intensive resuscitation.The patient died two days after the surgery in the state of cardiorespiratory failure.
In these two cases, typical resections (bisegmentectomies of segments II, III, VI, and VII) were dictated by the characteristic lesions of the liver parenchyme.Exanguination in the injured with complex liver injuries (grade III−V) required massive blood transfusions that could cause the refractory secondary coagulopathy and even to the liver failure.In the primary treatment, most of the injured received 8−17 units of blood.Pacher et al. (1), referred to the transfusions from 4 to 60 units, as an average.
In 14 of the injured from our group, the part of omentum at the vascular stem was placed in the area of hepatectomy after the ligature of biliary-vascular elements, and it also served for covering of the liver areas after the resection, or the resective debridement.
That procedure was effective in hemostasis and in the prevention of biliary fistula and septic complications.Pachter and Feliciano (1, 2) also reported good results of that procedure in the patients with severe liver injuries.
Perihepatic packing and planned re-operation were procedures that could save life of the injured with lifethreatening bleeding caused by severe hemorrhage and coagulopathy (1−4, 8, 23, 26−28).The packing of liver injuries goes back more than 90 years.It was abandoned as a technique for the control of liver hemorrhage, but it became popular again in World War II and in Vietnam (25).Many contemporary authors (1, 2, 9, 10, 12−14) reported the efficacy of this technique.Feliciano et al. (15) applied that technique in 5.3% of patients with the most severe injuries and the percentage of survival was above 57.1%.
Perihepatic packing combined with other techniques, including compressive packing of the primary managed liver injury, was applied in 22 injured of our group.Perihepatic packing of extensive injuries (in 6 cases combined with selective ligature of hepatic artery) was applied in 12 injured.That technique was applied in 16.6% injured (34/204) of our group.Reexplorations were done within 72 hours of its application.Temporary hemostasis was tolerable and there were no septic or other adverse effects of the packing.The reoperation included additional measures of war hospitals or revising of the primary treatment, but only in some cases.

Conclusions
Our experience of war hospitals and war surgeons with limited knowledge of liver surgery, suggested that it was necessary to apply compressive abdominal packing alone or in combination with other techniques of hemostasis, in the cases of complex liver injuries (grade IV and V).The injured were transported to MMA, applying all the resuscitation measures.Hemostasis up to reexploration was successful in 67.6% cases (23/34).One patient died during the surgery because of the extensive bleeding.Indications for the use of that procedure in the injured were: prolonged hemorrhage after leaving deep sutures in the area of the parenchymal injury; extensive injuries of one or both of the lobes alone or combined with another techniques for hemostasis; prolonged bleeding followed by coagulopathy due to blood transfusion after primary injury management with some other procedures; extensive subcapsular hematoma or capsular avulsion and injuries of the grade V as the only possibility of keeping the injured alive (successful in two cases).
The use of perihepatic packing in combined injuries of hollow organs, as well as the injuries of hepatic veins and ICV was not contraindicated in life-threatening situations where this procedure was the only option.The early removal of packing, detritus and necrectomies decreased the frequency of septic complications.
The survival of the injured depended not only on the degree of the liver injury, but also on the number and the severity of combined injuries.Exsanguination was the main cause of death in the casualties with combined injuries, including the liver injuries of the grade IV and V. On the other hand, the prognosis of combined injuries and liver injuries of the grade I−III depended on the severity and the number of other combined injuries.

Fig. 1 −
Fig. 1 − A computed tomographic scan in a young injured patient with a massive liver gunshot injury.

Fig. 2 −
Fig. 2 − These CT scans reveal a large intrahepatic hematoma and perihepatic blood collection.