Liver injury following blunt abdominal trauma: a new mechanism-driven classification

In Western countries, blunt liver injuries are caused by traffic accidents in approximately 70 % of cases [1]. In the case of polytraumatized patients with open or blunt abdominal trauma, the liver is the most frequently injured abdominal organ [2‐4]. Thirty-one percent of polytrauma patients have abdominal injuries, and lesions to the liver are found in 16 % of patients [5]. The main cause of liver injury-related death is uncontrolled bleeding, and it is associated with a mortality rate of 54 % [6].

However, the management of traumatic liver injuries has changed during recent years, and the outcome of patients has markedly improved [3, 7, 8]. Surgical treatment was the standard procedure for all kinds of trauma-related liver injuries, based on the idea that surgery was necessary to control the bleeding and prevent biliary complications. However, an improved understanding of the natural course of liver injuries and the development of new interventional radiological techniques have changed the paradigm toward a more non-surgical patient management [8‐10]. In the literature, more than 80 % of patients with blunt hepatic trauma are treated in a non-surgical fashion [7, 11, 12].

In addition, the comprehensive introduction and use of CT scanning enabled a reliable diagnosis of liver injuries within a short time after admission to the emergency room, and has become the gold standard for assessing trauma patients [4]. CT scan-based classifications of liver injuries [13] allowed for the selection of patients who could be managed conservatively. A prerequisite for a non-surgical approach is a haemodynamically stable patient with no further need for a laparotomy [11]. This conservative approach, however, should only be used in centers with an appropriate infrastructure providing capabilities for intensive care monitoring and instantaneous surgery [12]. However, these centers do not necessarily have to be high volume centers [11].

A variety of classifications for traumatic liver injuries have been described in the literature. The most accepted scoring system is the Moore score [14], which is based on the Organ Injury Scale (OIS) of the American Association for Surgery of Trauma (AAST) which was published in 1989 [15]. The Moore score is considered a gold standard to describe liver injuries. Another well-established scoring system is the Mirvis score [13], which is based on CT-graphic findings and gives the first hints about the necessity of surgery for patients with traumatic liver injuries.

However, the current scoring systems do not incorporate the localization of liver injury or the mechanism of trauma. Therefore, we developed a new classification for liver injuries and analyzed our patient cohort regarding the mechanism of trauma and corresponding pattern of liver injury.

Based on the results of this retrospective analysis, a new classification for blunt liver injuries was presented. The new classification is based on the localization of liver disruption by CT scanning or intraoperative findings, and correlates with the mechanism of trauma. Our analyses show that type A injuries resulted from a frontal impact of energy, e.g., in cases of frontal car accidents. This kind of energy transfer causes severe injury of the left liver lobe, i.e., segments II, III, IVa, and IVb. Therefore, type A injuries develop along the falciform ligament. Interestingly, all patients in our analysis who had a type A liver injury required immediate surgery due to haemodynamic instability. Despite immediate laparotomy for bleeding control, type A injuries were associated with a 25 % mortality rate, independent of the severity of liver injury, as assessed by the Moore score or Mirvis score. According to the literature, gender or age were also not the risk factors for type A liver injury-associated mortality [4].

In contrast, type B injuries resulted from more complex mechanisms of trauma, with an impact of energy from other directions than straight frontal, such as due to horse kicks, being crushed between vehicles or a high speed collision. Type B injuries are—according to our new classification—localized in the right liver lobe, i.e., segments V–VIII. Interestingly, the mortality of patients with type B injuries was not significantly different from that of patients with type A injuries, although type B injuries were associated with more complex trauma and more concomitant injuries. Again, gender, age, and the severity of liver injury were not determinants for type B injury-associated death. In contrast to type A injuries, significantly, more patients with type B liver rupture (six out of 36 patients) were haemodynamically stable and survived without surgery. The severity of liver injury in these six patients was not significantly different from that of patients undergoing surgery for type B liver rupture. Due to the higher energy transfer, additional injuries were more frequently observed after type B than after type A liver rupture.

Certainly, the localization of liver rupture, i.e., type A or type B rupture can never be the sole parameter used to decide on whether surgical or conservative management should be used for the liver rupture. Instead, the decision should be made based on the haemodynamic stability and lack of other injuries requiring abdominal surgery. To what extent the localization of liver disruption, and thus, the type of liver injury according to our new classification, will contribute to the decision to perform surgical or conservative management cannot be answered based on our data presented herein.

Irrespective of the type of liver injury, there has been a trend toward a more conservative and “watchful waiting” management of patients during the past two decades [5, 7, 11, 17]. Several studies have demonstrated that up to 80 % of the patients with blunt liver trauma can be managed conservatively [11], but these reports do not nominate other factors than the haemodynamic stability for making the decision on whether to use a surgical or non-surgical approach. In our study, only 9.4 % (6/64) of the patients could be managed conservatively within the last 11 years. During this time, the treatment strategies have changed, and angioembolization has been introduced for the treatment of blunt liver injuries and has been proven to be an effective treatment option even for patients with severe liver injuries [18, 19]. Since this technique has not been performed during the observation period at our center, our data include those of patients who could likely have been treated conservatively with angioembolization but underwent surgery due to the unavailability of this technique, which may explain the high percentage of patients requiring surgical therapy.

The current classifications for liver rupture are the Organ Injury Scale proposed by the American Association for Surgery of trauma (AAST-OIS) [20] or the Moore score [14, 15] and the Mirvis score, which describe the extent of liver damage either morphologically (AAST-OIS) or based on CT findings [13]. These scaling systems are widely accepted due to their long-term use, and in the case of the Mirvis score, due to the comprehensive use of CT scanning for trauma diagnostics. In addition, these scoring systems are validated and offer clear prognoses for different grades of injury severity [20]. However, these scoring systems do not incorporate the localization of liver injury, i.e., left or right liver lobe, or the mechanism of trauma and vector of power impact. One might argue that the site of the liver or direction of impact is irrelevant when making the decision whether to perform a laparotomy in a patient with blunt liver rupture or to choose a conservative approach. While this is somewhat reasonable, since haemodynamic stability, additional injuries, and other factors confound the decision, our data clearly demonstrate that more complex trauma mechanisms mainly affect the right liver lobe. Injuries in this liver lobe can be better tolerated, in the sense that they do not necessarily cause haemodynamic instability and are more often self-limiting. This knowledge might allow for the choice of a conservative treatment approach without an increased risk of mortality. Nevertheless, the localization of liver disruption alone can never be the sole parameter for deciding on surgical or conservative treatment.

As a possible explanation for the more severe injury in Type A cases, we suppose that a frontal energy impact first leads to an acceleration of the liver toward the spine. When the liver is decelerated by the falciform ligament, which is very stable and rigid, the liver is disrupted along this ligament, and large intrahepatic vessels are destroyed leading to severe bleeding toward the abdominal cavity. In contrast, the large volume of the right liver lobe and the direct covering of the right liver by the diaphragm (area nuda) allow the right liver to compress and self-limit even a severe disruption of large blood vessels.

In conclusion, we herein propose an additional classification that can be used in combination with the current and established scoring systems and treatment algorithms for blunt liver injuries, which is based on the localization of the liver injury and which represents the mechanism of blunt liver trauma. According to our new classification, type A injuries, which occur along the falciform ligament, are associated with a relevant haemodynamic instability requiring immediate surgical therapy. Type B injuries involving the right liver lobe are more likely to be self-limiting, and a watchful waiting strategy is justified if there are no other factors requiring abdominal surgery. Importantly, this classification is neither a decision aid nor a treatment algorithm with regard to surgical or non-surgical management, nor is it capable of predicting the outcome or mortality. Since our data were obtained from a retrospective analysis, a prospective validation and evaluation of the practicability of this classification for the emergency room is necessary.