Surgical management of AAST grades III-V hepatic trauma by Damage control surgery with perihepatic packing and Definitive hepatic repair–single centre experience

Despite the great advances in surgical treatment and resuscitation of trauma patients with liver injuries in the last decades, severe liver trauma still accounts for significant morbidity and mortality [1‐4]. Major liver injury is the leading cause of death in patients with abdominal trauma, and their treatment continues to challenge surgeons [5, 6]. The main cause of early liver injury-related death is uncontrolled bleeding, and it is associated with a mortality rate of 50–54 % in the first 24 h after admission, with 80 % of operative deaths [1, 2, 5].

Early diagnosis of the extent of liver trauma with adequate treatment adapted to the severity of the injury and the physiological condition of the patient, may result in significant reduction of morbidity and mortality [5]. Hemodynamic stability is key for diagnostic and therapeutic approach to the severe liver injuries. The diagnosis of hepatic trauma starts simultaneously with reanimation, immediately after admission, which implies targeted clinical examination, laboratory blood tests, abdominal ultrasound (Focused assessment with sonography for trauma, FAST) followed by Multislice Computed Tomography (MSCT) [6‐10]. Elevation of the serum aspartate aminotransferase (AST) and alanine aminotransferase (ALT) is the laboratory indicator of liver injury [9].

Menagment of a severe trauma patient involves systematic sequence of actions [8‐15]. Haemodynamically unstable patients with major liver injuries require rapid manoeuvers to control bleeding [10]. Accordingly, exsanguinating patients require substantial blood transfusions [11]. Uncontrolled bleeding leads to new adverse events that announce catastrophe–coagulopathy, as a result of depletion and dilution of coagulation factors, acidosis, and hypothermia [10, 11]. The decision for an emergency laparotomy is usually based upon the presence of the “lethal triad” with coagulopathy, acidosis and hypothermia [5, 11]. Surgical treatment in bleeding liver trauma is required in cases of progressive hemodynamic instability due to hemorrhage shock [11]. Operative techniques in liver trauma are some of the most challenging. They include the broad and complex area from Damage control surgery (DCS) to the liver resection [2, 5]. Surgical control of bleeding is the main goal in damage control strategy as well as the prevention of biliary complications which are specific for liver injuries. In unstable patients with severe physiological derangement surgical procedures such as direct vessel repair of juxtahepatic venous injuries and early perihepatic packing with the correction of hypothermia, coagulopathy, and acidosis may lead to improved outcome [5, 10, 11].

The purpose of this study was to determine the predictors of morbidity and mortality in trauma patients that underwent surgery for severe hepatic injury, as well as to identify a better approach for exsanguinating patients.

The general characteristics of all 121 patients with severe liver trauma who were included in our study with comparison between the survivors and non-survivors summarizes in Table 1. In this study 81(66.9 %) patients survived, while 40(33.1 %) of them died. In this study there were 90(74.4 %) males and 31(25.6 %) females (Table 1). Blunt hepatic injury was the leading mechanism of trauma, seen in 98(80.9 %) patients (Table 1). Road traffic accident was the leading cause of blunt trauma recorded in 80 (66.1 %) patients, and among them were 37(30.6 %) drivers, 36 (29.7 %) pedestrians and 7(5.8 %) passengers (data not shown). The remaining 10 (8.2 %) patients with blunt liver trauma were injured by falling from a roof and eight were hiting by assailant (6.6 %). Pentrating liver injury was recorded in 23 (19.0 %), with equal distribution between the groups: 15(19.0 %) suffered stab wounds, while 8(6.6 %) injured by firearms (Fig. 1). A total of 108(89.2 %) had liver trauma associated with injury >2 body regions, there was no difference between survivors and non-survivors (Table 1). Total of 82(67.8 %) patients had a ISS>34 (Table 2).

In this study there was 42(34.7 %) patients with liver trauma AAST III, 53(43.8 %) AAST IV and 26(21.5 %) patients with liver trauma AAST V including 4(3.3 %) retrohepatic vena cava and 10(8.3 %) major hepatic veins injury (Table 2). There was a significant difference between survivors 49(60.5 %) with ISS>34 and non-survivors 33(82.5 %) with ISS>34 (p = 0.0001) (Table 2). In comparison with the survivors, non-survivors have significantly higher liver AAST grade of injury: there was a statistical significance for the AAST III and AAST V (p = 0.001; p = 0.0001), while the AAST IV showed the same distribution (p > 0.05) (Table 2). Non-survivors showed significant hypotension on arrival (p = 0.0001) and lower GCS (p = 0.0001) (Table 2). Non-survivors needed significalntly more RBC units transfusions (p = 0.001) (Table 2). Non-survivors had significantly higher serum AST and ALT level within first 24 h (p = 0.010; p = 0.033) (Table 2).

There was no statistically significant difference in the application of surgical approach (p= > 0.05) (Table 3). Range of blood removed from peritoneal cavity was 500–1500 ml. Definitive hepatic repair was performed in 62(51.2 %) patients (Table 3). Liver resection was performed in 12(9.9 %) patients: non-anatomic resection in 6(4.9 %) patients and major resection (≥3 Couinauds segments) in 6(4.9 %) (Fig. 1, 2). DCS with perihepatic packing and planned re-laparotomy after 48 h was used in 59(48.8 %) (Table 3). In DC strategy we used different additional procedures in combination with liver packing (Fig. 3).

Most common non-related liver complications were: right-sided exudative pleural effusion in 24(19.8 %) patients, ARDS in 20(16.5 %) and pneumonia in 12(9.9 %) (Table 4). As liver-related complications we recorded: prolonged hemorrhage in 18(14.9 %), bile leak 21(17.3 %), biloma 12(9.9 %), liver abscess 2(1.6 %) and liver failure 1(0.8 %) (Table 4). Regarding complications non-survivors had a significantly prolonged bleeding and higher rate of ARDS (p = 0.0001, for all) (Table 4). Survivors had a significantly higher biloma (p = 0.014) and pleural effusion (p = 0.0001) (Table 4). Eleven (9.0 %) of all patients required re-operation during hospitalization: 9(7.4 %) due to prolonged bleeding and 2(1.6 %) due to uncontrolled bile fistula (Table 4). Other complications were treated with non-surgical approach, in one case the liver failure, patients had associated severe brain and lung injury .

Mortality was 33.1 %. We recorded statistical significance in terms of ICU and hospital stay (p = 0.001; p = 0.001) (Table 5). The early trauma-related mortality within the first 24 h after admission was noted in 35 % (Table 5). The cause of mortality in “early period” was massive prolonged bleeding. Among non-survivals 62.5 % died within the first 7 days (Table 5). Patients died in the further course of hospitalization due to late respiratory complications: ARDS and pneumonia.

In hemodynamically unstable patients with hemorrhage from major liver injury and massive hemoperitoneum on abdominal imaging, the strategy and techniques for bleeding control can be extremely demanding and complex. We presented the results of surgical treatment of 121 trauma patients with severe bleeding liver injuries.

Since Pringle’s publication of inflow vascular control on liver, the primary focus of trauma surgeons was to find the best way to achieve hemo-stasis, bile-stasis and infection control in hepatic injuries [15‐23]. Today the focus of trauma surgeons is selection of appropriate patients for operative management, who are the candidates for surgery and when to operate [7, 8]. The general contraindications for none-operative management of liver trauma included the hemodynamic instability, extravasations of intravenous contrast on abdominal imaging, expanding hematoma and grade IV and V liver injury [6, 7]. According to the study conducted by Coimbra et al. nonoperative management has been accepted as treatment of choice only for stable patients with low grade of injury [23]. Fang et al. study of 214 patients with a hepatic injury showed that the independent predictors for the surgical treatment even in hemodynamically stable patients included intraperitoneal contrast extravasation and hemoperitoneum in six compartments on CT scan [22]. FAST is able to sensitively detect hemoperitoneum presented as free fluid in the abdomen and pelvis, but its numerous limitations have been recognized [8, 24]. MSCT is the imaging modality of choice in evaluating hemodynamically stable patients with suspected hepatic injury [7, 22]. Abdominal CT accurately defines the morphology and extent of the hepatic trauma, identifies associated visceral injuries and depicts the amount of hemoperitoneum [7, 25]. It is important to know that life- threatening liver injuries can be detected by MSCT with high sensitivity.

In this study indications for emergency laparotomy were: hemorrhage shock on admission, refractory hemodynamic instability, signs of haemoperitoneum on ultrasound, MSCT findings of the severe liver trauma with contrast extravasation which indicate active hemorrhage. It is important to know that life-threatening liver injuries can be detected by MSCT with high sensitivity. The operative management of liver injuries requires the use of some of the most complex surgical techniques, including extensive hepatotomy with selective deep vessel ligation, hepatorrhaphy, selective hepatic artery ligation, non-anatomic resection and debridement and hepatectomy [6, 7]. The surgical treatments of severe liver injuries in this study were definitive procedure (hepatorrhaphy, hepatotomy with vascular ligation, debridement, selective hepatic artery ligation, liver resection) and DCS with liver packing. The incidence of trauma patients requiring liver packing varies from 5 to 62 % in the literature [6, 21, 26]. We performed perihepatic packing in 48.8 % patients. It is important to recognize that liver packing will not control arterial bleeding and that any bleeding artery should be suture/ligated prior to liver packing. A precise perihepatic packing technique starting with a Pringle manoeuvre and complete liver mobilization, with systematized placement of packs. Strategies for haemorrhage control are more efficient when DCS is associated with appropriate additional strategies such as an effective fluid resuscitation/transfusion protocols, a carefully selected angioembolization and an accurate ICU critical care [13, 19, 27]. According to the Asensio results, improvements in outcome can be achieved using damage control strategy to control hepatic bleeding in patients with severe liver injuries and compromised physiological stage [6]. However, interventional radiology with embolization may play an important role in cases of liver packing followed by re-bleeding. Angiography and angioembolization were not used in our study group which is the limitation of this study. Although this surgical technique was not a predictor of outcome in our study, the question is whether the use of DCS in combination with angioembolization for emergency control bleeding, would contribute to the lower rate of complications and deaths in this study. The reported survival rate associated with packing increased up to 65.5 % as reported by Peitzman et al [2]. Outcome could be improved by combining of venous bleeding control by liver packing and arterial haemorrhage control with angioembolization [2, 5, 6, 28].

Over half of patients surviving grade III–V liver injuries will be at risk for the development of complications [29]. Liver-related complications occur in approximately 20–45 % of patients and include hemorrhage, hemobilia, arteriovenous fistula, pseudo-aneurysm, biloma, bile leak and abscess formation [5, 10, 27, 29]. MSCT and ultrasaund was used in diagnosing specific post- traumatic postoperative complications such as hepatic or perihepatic abscesses or bilomas. Postoperative prolonged hemorrhage can be associated with coagulopathy [5, 10, 27, 29]. While blood transfusion is necessary because of massive bleeding in trauma, it carries many complications [29, 30].

Complex hepatic injuries have a high mortality rate (8,36,38). Our study shows that non-survivors had higher AAST grade of injury, higher AST and ALT level, significant hypotension, higher ISS score and lower GCS on arrival; and significalntly more RBC units transfusions within the first 24 h. Early liver injury-related death is typically secondary to uncontrolled bleeding (20–60 %), which is worsened with attendant coagulopathy, whereas late mortality is usually secondary to multiorgan failure (MOF), Acute Respiratory Distress Syndrom (ARDS) in 27(32.1 %), and Multiple Organ Dysfunction Syndrome (MODS) [6, 29‐31]. Previous studies have shown that increased risk of ARDS and MODS has been associated with massive transfusion, which can itself contribute to coagulopathy [30]. In blunt liver trauma mortality also appears to be higher in older patients, those with higher grade injuries, and those with hemodynamic instability on presentation [29‐31]. In a series of 144 patients with grade III–V hepatic injuries, uncontrollable bleeding from the liver injury and associated severe splenic injury favors early laparotomy and damage control strategy [31]. Asensio and colleagues showed that predictors of mortality in grade IV–V injuries are related to severe bleeding and include blood loss, number of red cell units transfused, hypothermia, acidosis, and dyasrhythmia [6].

Prolonged bleeding and amount of blood transfusions are statistically significant predictors of mortality in severe hepatic trauma. All efforts in the treatment of trauma patients with complex liver injuries AAST grade III–V should be directed to the rapid and effective control of liver hemorrhage and taking care of all associated life-threatening injuries.