INCIDENCE

Hepatic injury occurs in approximately 5% of all trauma admissions. Nationwide, there has been a steady decline in the incidence of penetrating liver injuries. However, blunt injuries seem to be on the rise predominantly because their presence has been more readily detected by the almost routine use of CT scanning in patients sustaining blunt trauma. The incidence of complex hepatic injuries, however, has remained relatively stable over the past 25 years, ranging from 12% to 15%.¹

Motor vehicle crashes (MVCs) continue to account for most (approximately 80%) blunt hepatic injuries, followed by pedestrian and car collisions, falls, assaults, and motorcycle crashes. Most patients with blunt hepatic trauma have associated injuries, both intra-abdominal and extra-abdominal. Concomitant chest trauma is the most common associated injury encountered with blunt hepatic trauma, occurring in over 50% of patients. Patients with right-sided lower rib fractures, particularly ribs 9–11, have at least a 20% chance of sustaining an underlying hepatic injury. In spite of the high aforementioned incidence of associated chest trauma, injury to the brain remains the single most significant determinant in overall survival outcome.² In the era of nonoperative management of blunt trauma, the risk of a missed injury, especially to the diaphragm or small bowel, is of major concern. Adherence to meticulous interpretation on imaging studies by experienced personnel should limit this pitfall to 1%–2%.

Penetrating thoracoabdominal trauma has been noted to be associated with injuries to the liver in 30%–40% of such injuries. The extent of the injury is directly related to the type of weapon used. Associated intra-abdominal injuries (e.g., stomach, duodenum, colon, and pancreas) are common but rarely detected preoperatively.

MECHANISM OF INJURY

Blunt Hepatic Injury

In MVCs, those most susceptible to hepatic injury are unrestrained front-seat passengers. These passengers are particularly vulnerable to a compression injury from the steering wheel especially during periods of rapid deceleration.³ Although the anterior abdominal wall stops, the posterior abdominal wall continues to move forward, and the intra-abdominal organs are “trapped” and compressed resulting in stretching/tearing of the liver at its vascular and structural attachments. As the liver is only partially protected by the rib cage, liver injury from steering wheel contact is one of the most important contributing factors to driver injury.

In lateral impact (“T-bone”) collisions, the target vehicle is hit on its side and accelerated rapidly at 90 degrees to its previous direction of travel. The unrestrained passenger is subject to both compression and shear injuries that cause stretching/tearing and, at times, result in avulsion of the liver. Furthermore, in lateral impact injuries, because the spine/posterior abdominal wall is not in the line of impact in contradistinction to frontal impact injuries, more relative motion of the intra-abdominal organs ensues resulting in a greater likelihood of injury.

Penetrating Hepatic Injury

Damage caused by a penetrating injury is based on the kinetic energy of the projectile and the density and elasticity of the tissue. Lowenergy weapons such as knives only cut and do not create a temporary cavity. Medium-energy and high-energy firearms damage not only the tissue directly in the path of the missile but also the tissue on each side of the missile’s path. As a missile passes through the relatively inelastic liver parenchyma, a temporary cavity (three to six times the size of the missile’s front surface area, lasting for a fraction of a second) and a permanent cavity (visible to the examiner) are created. The higher-energy firearms create larger temporary and permanent cavities, resulting in far more extensive tissue damage; the vacuum created by this larger cavity pulls clothing, bacteria, and other debris from the surrounding area into the wound as well.

DIAGNOSIS

Hemodynamically Unstable Patients

Patients who arrive with hemodynamic instability (systolic blood pressure <90 mm Hg) and who do not immediately respond to appropriate fluid resuscitation are expeditiously taken to the operating room without delay, irrespective of mechanism of injury. Further diagnostic evaluation at this point is contraindicated, as unnecessary delays inevitably follow and are often responsible for the ensuing fatalities.

In the hemodynamically unstable patient with pelvic fractures from blunt trauma, diagnostic peritoneal lavage (DPL) and focused abdominal sonography for trauma (FAST) are currently the diagnostic modalities used to detect the presence of intraperitoneal blood. DPL is 96% accurate in detecting intraperitoneal blood and a grossly positive aspiration (>10 ml) mandates immediate operative intervention. In most trauma centers, FAST has replaced DPL as the preferred diagnostic modality for the determination of hemoperitoneum in the unstable bluntly injured patient. Although FAST has a 97% sensitivity for hemoperitoneum greater than 1 liter, the location of the parenchymal injury often cannot be reliably identified. One’s ability to assess the source of the hemorrhage will continue to increase as experience with FAST increases. The sensitivity of FAST drops precipitously when the quantity of intraperitoneal fluid is less than 400 ml.

Hemodynamically Stable Patients

The hemodynamically stable blunt-trauma patient, on the other hand, may undergo further diagnostic studies. Hemodynamic stability, however, should not lull the trauma surgeon into a false sense of security, as significant intra-abdominal injuries may be present despite normal vital signs and a normal abdominal exam. The ability to accurately assess the presence or absence of significant intra-abdominal injuries by physical examination alone in the blunt trauma patient is notoriously poor, as up to 20%–30% of patients with a “benign” abdomen on physical examination have been shown to subsequently have significant intra-abdominal injuries on imaging or at laparotomy.

CT scanning is the preferred initial diagnostic modality in the hemodynamically stable patient with blunt abdominal or lower thoracic cage injuries. High-speed resolution scanning with a spiral scanner is employed after the administration of intravenous and oral (when feasible) contrast. Five-millimeter cuts are obtained after 120 ml of noniodinated contrast (Omnipaque) is injected at a rate of 2 ml/sec. Scanning commences 50 seconds after injection, a delay that corresponds to the portal venous phase of liver imaging.

Scans should immediately be interpreted and classified according to the American Association for the Surgery of Trauma liver injury scale ⁴ (Table 1) by the CT fellow or attending radiologist, always in the presence of the chief trauma resident and trauma attending. The senior trauma attending in presence makes the final decision as to the appropriateness of nonoperative therapy. It should be noted that the grade of injury or degree of hemoperitoneum on CT does not determine the need for operative intervention, as this decision is based primarily on the patient’s hemodynamic stability and the absence of peritoneal signs. Instead, the CT scan merely provides the surgeon with a general anatomic overview of the injury, identifies associated abdominal injuries requiring operative intervention, and can be used as a base for comparing future healing of the hepatic injury and resorption of intraperitoneal blood. CT can also identify injuries involving the bare area of the liver, which commonly present with minimal intra-abdominal bleeding, a paucity of abdominal signs, and often a negative DPL.

Table 1 American Association for the Surgery of Trauma Liver Injury Scale

The role of FAST as a screening exam in hemodynamically stable patients is evolving and in the near future may eliminate the need for CT scan. Currently, many trauma centers forgo CT scanning in stable patients with negative initial FAST exams and merely repeat the FAST in 6 hours. However, scanning for only free fluid has its diagnostic limitations because not all blunt hepatic injuries result in hemoperitoneum. In a recent study looking specifically at sonographic detection of blunt hepatic trauma, Richards et al.⁵ determined the overall sensitivity of FAST for blunt hepatic injuries (all grades) to be 67%, based on the detection of free fluid alone. On the other hand, it is clear that most solid organ injuries without intraperitoneal fluid on FAST are, in general, of minimal clinical significance.⁶ At present, most trauma surgeons agree that those patients who are hemodynamically stable and who have either intraperitoneal blood on their initial FAST exam or positive findings on physical exam over the lower chest and upper abdomen should have a CT to specifically identify a hepatic or splenic injury that can be managed nonoperatively. Once identified, the hepatic injury may be followed with ultrasound if necessary.

Diagnostic laparoscopy (DL) is a safe procedure that has had a major impact in avoiding unnecessary abdominal explorations in patients with stab wounds or gunshot wounds that may not have penetrated the peritoneal cavity. The role of diagnostic laparoscopy in patients with blunt hepatic injury is less clear. DL should allow for an accurate assessment of most hepatic injuries and, as advances in laparoscopic instrumentation progress, perhaps allow for repair of some liver injuries. However, reports of missed enteric and other intra-abdominal injuries with DL are sufficiently significant to warrant further evaluation of DL in blunt trauma patients.

ANATOMIC LOCATION OF INJURY AND INJURY GRADING—AAST-OIS

Comprehensive knowledge of hepatic anatomy is essential to the proper management of traumatic liver injuries. Couinaud has described the functional anatomy of the liver, based on the hepatic venous drainage (Figure 1). The ligamentous attachments of the liver are depicted in Figure 2.

Figure 1 Functional division of the liver, according to Couinaud’s nomenclature.

(From Mattox KL, Feliciano DV, Moore EE, editors: Trauma, 4th ed. New York, McGraw-Hill, 1999, Figure 30-1. Originally appeared in Blumgart LH, editor: Surgery of the Liver and Biliary Tract. New York, Churchill Livingstone, 1988.)

Figure 2 Surgical anatomy of the liver.

(From Mattox KL, Feliciano DV, Moore EE, editors: Trauma, 5th ed. New York, McGraw-Hill, 2004, Figure 30-2.)

The American Association for the Surgery of Trauma (AAST) created the Organ Injury Scaling (OIS) Committee to standardize injury severity scores for individual organs to facilitate clinical investigation and outcomes research. The liver injury scale devised by the AAST is shown in Table 1.

MANAGEMENT

Nonoperative Management/Blunt Hepatic Trauma

Currently, nonoperative management of adult blunt hepatic injuries is the standard of care. Initial hemodynamic stability or hemodynamic stability achieved and maintained with moderate fluid resuscitation is the single most crucial prerequisite qualifying patients for nonoperative management. Once hemodynamic stability has been ascertained, the following criteria must be met:

Absence of peritoneal signs

Precise CT scan delineation and AAST grading (see Table 1)

Absence of associated intra-abdominal or retroperitoneal injuries on CT scan that require operative intervention

Avoidance of excessive hepatic-related blood transfusions

Previously cited inclusion criteria such as neurological integrity are no longer valid, as neurologically impaired patients can be safely managed nonoperatively in a monitored setting.⁷ Furthermore, mandatory repeat CT scans to document improvement or stabilization of injury are unnecessary and contribute little to patient outcome. Rather, the patient’s clinical course should dictate the need for additional evaluation.

Most (80%–90%) blunt hepatic trauma patients can be successfully managed nonoperatively. Although nonoperative management was initially limited to AAST grades I–III injuries, it is now clear that the hemodynamic status of the patient, rather than AAST grade of injury, is the most significant factor in determining the need for operative intervention. Up to 20% of select patients with grades IV and V injuries can be managed nonoperatively. However, many grade IV and most grade V injuries will usually present with hemodynamic instability or concomitant injuries mandating surgery, thus precluding nonoperative intervention. In a multi-institutional study, grades IV and V injuries were responsible for 67% of all patients who failed nonoperative management and subsequently required operative intervention.⁸ Therefore, although hemodynamic stability determines which patients can be managed nonoperatively, the subgroup of patients with complex hepatic injuries (grades IV and V) are at substantially higher risk for treatment failure and should therefore be closely monitored in a critical care unit.

Conversely, the same basic standards apply to patients with lower AAST-grade injuries (i.e., I–III). In these instances, the initial injury may be deemed as “not significant,” and thus it becomes tempting to avoid surgical intervention despite hemodynamic instability or a decreasing hematocrit, relying instead on further fluid and blood transfusions. This course of action is fraught with pitfalls and should be avoided to minimize the morbidity and mortality of nonoperative management. To summarize, of all the variables monitored, hemodynamic stability appears to be the most crucial and is considered the watershed for nonoperative or operative intervention.

Contrast “Blush” on CT

Specific cause for concern is the presence on the initial CT scan, after the administration of intravenous contrast, of a contrast “blush” or “pooling” of contrast material within the hepatic parenchyma. This finding indicates active bleeding. Even in the context of hemodynamic stability and irrespective of AAST grade of injury, preparation for possible surgical intervention should promptly be made, as patients can suddenly and unpredictably decompensate clinically. If the patient remains hemodynamically stable, angiography with the intent of embolizing the lacerated vessel should be attempted (with an operating room on standby secured). An experienced interventional radiologist will usually have little difficulty in selectively catheterizing and embolizing the injured vessel, most often with stainless steel coils and/or Gelfoam. Successful embolization will then permit further nonoperative management. As the natural history of these lesions is unknown, they are best dealt with immediately so that sudden bleeding, false aneurysm formation, and late hemobilia may be avoided.

Persistent and prolonged attempts at controlling the bleeding vessel through angiographic means should be discouraged. In the rare event in which angioembolization fails to control ongoing bleeding, surgical intervention using the angiogram as an anatomic marker to more rapidly achieve intrahepatic hemostasis should promptly be undertaken.

Nonoperative Management/Penetrating Hepatic Trauma

Most penetrating civilian injuries to the liver result in a lesser degree of parenchymal damage than do those incurred by blunt trauma, at least by AAST criteria. Therefore, it seems logical that nonoperative management of a penetrating isolated hepatic injury would be successful in hemodynamically stable patients without evidence of peritonitis.

Renz et al.⁹ nonoperatively managed 13 patients with penetrating right thoracoabdominal gunshot wounds. The authors stressed the importance of serial abdominal exams and contrast-enhanced CT scanning in their successful nonoperative management. Demetriades et al.¹⁰ substantiated this concept with their successful management of select patients with isolated gunshot injuries to the liver. These authors concluded that hemodynamically stable patients with grades I and II liver injuries and no evidence of peritonitis can be safely managed nonoperatively. However, it should be noted that this approach failed in nearly one-third (5/16) of the patients in the “observed group” who eventually required delayed laparotomy. More recently, Omoshoro-Jones et al.¹¹ described successful nonoperative management in 31 of 33 patients with gunshot wounds to the liver, including grades III–V injuries. Although the higher-grade injuries were associated with more complications (most of which were managed nonoperatively), the overall success of nonoperative management did not depend on the AAST grade of liver injury.

Clearly, the most difficult aspect in the nonoperative management of penetrating hepatic trauma is patient selection, as only up to 30% of those with gunshot wounds to the liver are eligible for nonoperative management to begin with. At the very least, hemodynamic stability, an intact level of consciousness to allow serial abdominal exams, absence of peritoneal signs, and no evidence of active bleeding on CT are required for successful nonoperative management.