Abstract Hepatic trauma causes a signifi cant amount of morbidity and mortality. The decision to on whether or not to operate is a key step in managing this problem. Operative management carries with it a higher rate of morbidity and mortality than non-operative management. However, clear indications do exist when an operation is needed. These include hemodynamic instability, continued bleeding, peritonitis, and other abdominal injuries requiring operation. Operative management involves a graded response to the injury. Selected angioembolization can be a useful adjunct to both operative and non-operative approach to hepatic trauma. Keywords Hepatic trauma • Operative management • Non-operative management • Angioembolization Introduction Hepatic trauma is associated with a signifi cant amount of morbidity and mortality . It is a complex and challenging problem to manage. Hemodynamically unstable patients with a blunt mechanism of injury to the liver should undergo operative management [ 1 ]. Similarly, patients with penetrating hepatic injuries mostly need operative management to search for associated intra-abdominal injury [ 2 ]. Nonoperative management (NOM) of hemodynamically stable patients with blunt liver trauma has become the standard of care. This is a change in philosophy of management of these injuries began to evolve in the early 1990s. Spurred by case reports and results from the pediatric literature, several studies in this period showed a high rate of success of NOM for all American Association for the Surgery of Trauma (AAST) grades of blunt hepatic trauma [ 3 – 5 ]. More recently, several centers have C. Evans • M. A. Croce (*) Department of Surgery , University of Tennessee Health Science Center , 910 Madison Ave, 2nd fl oor , Memphis , TN 38163 , USA e-mail: [email protected]; [email protected] 64 been evaluating patients for non-operative management of penetrating hepatic trauma. These patients must be hemodynamically stable and carefully selected [ 6 – 8 ]. Patients who are managed with early operative management have higher Injury Severity Scores (ISS), higher rates of hypotension upon presentation, and higher mortality . Failure of NOM is also associated with higher overall mortality [ 4 ]. For this reason, the attempt will be made to manage most patients with blunt liver injury non-operatively. Angiography can be used as an adjunct to NOM or operative management of liver injury [ 1 ]. The failure of non-operative management can be due to the development of hemodynamic instability, failure to respond appropriately to transfusion or drop in hematocrit, suspicion of injury to other organs requiring operation, or clinical peritonitis [ 3 ]. Failures may be due to the liver injury itself or due to injuries to other intra-abdominal organs. Not surprisingly, higher grade injuries have a higher rate of failure. Grade IV and V injuries may have as high as two-third failure rate [ 2 , 3 ]. Risk factors for failure are older age, lower admission mean SBP, worse base defi cit, higher mean lactate, higher mean ISS, and hemoperitoneum extending into the pericolic gutter [ 3 ]. Operative management strategies can range from simple electrocautery and topical hemostasis to liver transplantation at the other extreme end of the spectrum. Most authors advocate a graded approach to the operative management of liver trauma. This begins with topical hemostasis for lower grade injuries, to suture ligation of bleeding vessels, to anatomic and non-anatomic liver resection s. In the 1990s the concept of damage control laparotomy was reintroduced. Injuries to the liver are usually to the veins, which are a low pressure system. These can often be controlled simply with temporary gauze packing. This limits blood loss and allows for resuscitation of the patient. The patient can then return to the operating room in 24–72 h and defi nitive control if needed can be achieved. Search Strategy A literature search from the PubMed Database of English language publications from the last 10 years was used to identity published data on the operative management of liver trauma using the PICO outline (Table 6.1 ). Terms used in the search were “operative liver trauma”, or “operative hepatic trauma ”. Articles were excluded if the full text was not available through the institution. Case reports were excluded. Table 6.1 PICO table for perioperative arrhythmia prophylaxis for lung resection P (Patients) I (Intervention) C (Comparator group) O (Outcomes measured) Patients with traumatic liver injury Operative management or angioembolization Non- operative management Failure of non-operative management, morbidity and mortality, liver related morbidity and mortality C. Evans and M.A. Croce 65 Twenty-six articles were selected to be used for the writing of this chapter. All studies used data from retrospective analysis, or data that was prospectively entered into a trauma database. Other cited articles were review articles or book chapters. The data was classifi ed using the GRADE system. Results Non-operative Management Hemodynamically stable patients with blunt liver injury can be managed nonoperatively in up to 80 % of cases. All grades of liver injury except Grade VI can be managed non-operatively [ 2 ]. Cumulative results of reported NOM in the last 10 years shows a success rate of 91.3 %. This is including studies looking only at high grade liver injuries. Most common liver related reasons for the failure of NOM include continued drop in hematocrit or continued bleeding, hemodynamic instability, and biloma or bile leak (table). Non-liver related causes of failure of NOM are related to injury to other organs; mostly spleen, pancreas, or small bowel (Table 6.2 ). Higher grades of liver injury according to AAST grading system are more likely to be managed operatively and tend to have a higher failure rate of NOM. The results of fi ve studies were tabulated showing failure rates in Grade I of 0 %, Grade II of 0.8 %, Grade III of 3.0 %, Grade IV of 10.4 %, Grade V of 21.7 % [ 9 – 13 ]. Christmas et al. showed a similar trend of increasing operative management with higher grade injuries [ 14 ]. Cohn et al. also reported a poor sensitivity for predicting the need for operative management based on the AAST grading system for liver Table 6.2 Failure of non-operative management and reasons for failure Author (year) N op NOM Fail NOM HD unstable Hct. drop Peritonitis Other Fang (2006) 278 64 214 30 0 24 0 6 Gaarder (2007) 114 41 73 11 ns ns ns ns Ghnnam (2013) 56 20 36 0 ns ns ns ns Kozar (2005) 337 107 230 12 0 1 0 11 Morales (2014) 117 19 98 7 0 2 2 3 Norrman (2009) 46 11 35 4 0 2 1 1 Parray (2011) 152 ns 152 8 8 0 0 0 Prichayudh (2013) 152 92 60 6 0 4 2 0 van der Wilden (2012) 393 131 262 23 0 7 10 6 Zago (2012) 120 55 65 6 2 0 4 0 Combined 1765 540 1225 107 10 40 19 27 %Combined – – – 8.7 10.4 41.7 19.8 28.1 op operative management , NOM non-operative management , ns not studied 6 When Should You Operate on Major Hepatic Trauma? 66 trauma. They did, however, report that the fi ndings of lacerations involving more than two segments, lacerations extending into the hilum, and active extravasation correlate with a >90 % specifi city for either angiographic or operative intervention. Furthermore they report a specifi city of 85 % for intervention for >500 cc hemoperitoneum [ 15 ]. Interestingly, Fang et al. reported an operative rate of 100 % in 14 patients who had free extravasation of contrast into the peritoneal cavity [ 16 ] (Table 6.3 ). Several factors are signifi cantly associated with failure of NOM. These include hypotension on admission, worse base defi cit, higher lactate, higher ISS, vascular blush on CT , and hemoperitoneum on CT scan extending into at least the paracolic gutters [ 3 ]. Polanco retrospectively analyzed the National Trauma Data Bank for isolated liver injuries. In over 3000 patients, increasing age, higher ISS, and hypotension on presentation were identifi ed as risk factors in patients more likely to fail NOM [ 4 ]. In a smaller study, Norman et al. also found ISS and a lower presentation blood pressure in patients failing NOM [ 17 ]. One study identifi ed only lower average blood pressure as a predictor of failure [ 20 ], while other studies failed to identify any variables as risk factors for failure of NOM [ 13 , 18 ]. Penetrating trauma to the abdomen is generally managed with laparotomy, however case series do exist of the non-operative management of highly selective patients. Inaba et al. reported a series of eight patients with isolated liver injury and two patients with liver and kidney injury following gunshot wounds that were managed non-operatively. One patient with a isolated liver injury underwent nontherapeutic laparotomy and was discharged home after an 79 day recovery in the hospital [ 6 ]. MacGoey et al. presented a series of ten non-operatively managed patients with penetrating injuries with two failures of NOM due to hemodynamic instability [ 7 ]. Omoshoro-Jones et al. in a prospective study of 33 highly selected patients with gunshot wounds to the liver, had only two non-liver related failures of NOM [ 8 ]. These studies involved highly selected patients who were hemodynamically Table 6.3 AAST grade of injury and failure of non-operative management based on grade Non- operative management total Failure non-operative management Author (year) Grade I Grade II Grade III Grade IV Grade V Grade I Grade II Grade III Grade IV Grade V Ghnnam (2013) 4 11 12 9 0 0 0 0 0 0 Kozar (2005) na na 130 92 8 na na 0 7 4 Norman (2009) 24 11 3 1 Parray (2011) 30 63 26 33 0 0 0 2 6 0 Saltzherr (2010) 20 43 30 10 1 0 1 4 2 0 van der Wilden (2012) na na na 234 28 na na na 19 4 Zago (2012) 51 14 5 1 Combined 58 74 209 378 51 0 1 6 34 9 C. Evans and M.A. Croce 67 stable with right upper quadrant gunshot wounds and no clinical signs of peritonitis who had a reliable abdominal exam [ 6 – 8 ]. Based on the paucity of data, laparotomy is recommended for patients with penetrating hepatic trauma . Angiography and Embolization The EAST guidelines for management of liver trauma give a level 2 recommendation to the use of angiography and embolization as an adjunct to operative management of liver trauma or as a primary treatment modality for NOM in transient responders to resuscitation. They also give a level 2 recommendation to the use of angiography in patients who have an active blush on CT scan [ 1 ]. Sivrikoz et al. performed a retrospective analysis of over 6000 patients with isolated grade IV and V blunt hepatic injuries from the National Trauma Data Bank to investigate the role of angioembolization . Eleven percent of these patients underwent angiographic embolization. Angiographic embolization was shown to be an independent predictor of survival in both patients undergoing operation and patients managed non- operatively [ 19 ]. Saltzherr et al. showed a reduced mortality in high grade liver injuries and higher percentage of liver injuries successfully managed non- operatively after the introduction of angiographic embolization at their facility [ 11 ]. Christmas et al. showed embolization prevented failure of non-operative management in patients with persistent liver bleeding in 11 of 12 patients [ 14 ]. Van der Wilden et al. report successful non-operative management of grade IV and V while heavily relying on angiography and embolization. Embolization had a 93 % success rate for preventing FOM for persistent bleeding in 59 patients [ 20 ]. Dabbs et al. performed a retrospective analysis of 538 patients admitted with high grade liver injuries. One hundred sixteen of these underwent angiography and 71 had embolization. Forty- three of these patients had hepatic related complications including 100 % of patients with grade 5 injuries and thee were eight liver related deaths. The most common complication they observed was major hepatic necrosis which made up 42 % of the complications [ 21 ]. Outcomes Published studies over the last 10 years were combined to determine the morbidity and mortality , liver related (LR) morbidity and mortality of operative, NOM and failure of NOM for blunt hepatic trauma . Not all studies consistently differentiated LR morbidity and mortality from all cause morbidity and mortality. Some only report LR morbidity and mortality or combined morbidity and mortality. Some studies focus only on high grade liver injuries. Not surprisingly, patients who were managed with an operation, either on presentation or after failure of NOM trended towards having higher morbidity and mortality. Morbidity of NOM was 18 % with 6 When Should You Operate on Major Hepatic Trauma? 68 mortality of 3.6 %. LR NOM morbidity was 2.2 % with no mortalities. Failure of NOM had a morbidity and mortality of 55 % and 14.4 %, respectively. LR failure NOM had morbidity of 44.8 % and mortality of 14.1 %. Operative morbidity and mortality were 79.9 % and 43.6 %. LR operative morbidity and mortality were 34.9 % and 21.3 % [ 9 , 10 , 13 , 16 – 18 , 20 , 22 – 24 ] (Table 6.4 ). Christmas et al. showed a signifi cantly higher mortality in grade III–V injuries managed operatively vs. non- operatively [ 14 ]. Surgical Strategies Operative management of the liver fi rst involves proper mobilization to allow for suffi cient exposure to deal with the injury. This begins by dividing the ligamentum teres and mobilizing the falciform ligament off the abdominal wall. The right and left triangular and coronary ligaments should also be divided so that both lobes of the liver are freely mobile. The caveat is that the presence of a confi ned, nonexpanding retrohepatic hematoma should be a contraindication to mobilization of the liver in patients with blunt trauma. This often signifi es an injury to the retrohepatic veins or inferior vena cava. Mobilizing the liver can lead to uncontrolled hemorrhage from these injuries [ 2 ]. Bleeding from minor lacerations to the liver can usually be controlled with manual compression, electrocautery, or topical hemostatic agents. The argon beam coagulator or Aquamantis may also be used to control hemorrhage, but should not be used for deep lacerations [ 25 ]. In damage control situations, packing the liver with gauze laparotomy pads and compression are the fi rst line strategies to control bleeding [ 23 , 25 ]. These will be removed at a planned second look laparotomy after the patient has been properly resuscitated, usually between 24 and 72 h after the initial operation. Often, the liver is packed in conjunction with other hemostatic procedures [ 3 ]. If persistent bleeding or bile leak is present despite packing, the injury will need further exploration. The Pringle maneuver is useful to control infl ow while the wound is explored. In general, this should not be left in place for more than 30–45 min [ 2 ]. Intermittent clamping in a 15 min on 15 min off fashion is preferred [ 25 ]. If bleeding persists despite the Pringle maneuver, bleeding is most likely from a retrohepatic venous source. Visible bleeding vessels or bile duct s in the liver parenchyma should be suture ligated. Injuries to large veins can be repaired with 5–0 polypropylene suture. Often bleeding and bile leak is hidden deep within a laceration. To gain exposure, the fi nger fracture technique can be used to further open up the liver parenchyma, ligating and dividing vessels as they are exposed. However, this can be quite time consuming. Alternatively a laparoscopic stapler with vascular loads can be used to divide the liver more quickly [ 2 , 25 ]. Every effort should be made to control bleeding from a retrohepatic venous injury by packing alone [ 23 ]. If forced to repair this type of injury, total vascular insolation needs to be achieved. This is accomplished by controlling the suprarenal C. Evans and M.A. Croce 69 Table 6.4 Morbidity and mortality and liver related morbidity and mortality from liver trauma Author (year) N op NOM Fail NOM MB LR NOM MB NOM ML NOM LR ML Fail MB LR fail MB Fail ML Fail LR ML op MB LR op MB op ML Op LR ML Fang (2006) 278 64 214 30 3 3 ns ns 7 3 8 8 ns ns ns ns Gaarder (2007) 114 41 73 11 ns ns ns 0 ns ns ns 0 ns ns ns 0 Ghnnam (2013) 56 20 36 0 ns ns 0 0 na na na na ns ns 2 2 Kozar (2005) 337 107 230 12 ns 2 ns ns ns 4 1 1 ns ns ns ns Morales (2014) 117 19 98 7 10 5 2 0 ns ns 0 0 11 8 11 8 Norrman (2009) 46 11 35 4 9 ns 0 0 3 ns 1 1 7 ns 1 0 Parray (2011) 152 ns 152 8 ns ns 0 0 4 4 0 0 ns ns ns ns Pricha-yudh (2013) 152 92 60 6 13 6 1 0 ns ns ns ns 59 28 37 29 van der Wilden (2012) 393 131 262 23 91 ns 13 ns 23 ns 2 ns 112 ns 69 ns Zago (2012) 120 55 65 6 7 0 0 0 2 2 1 1 57 22 23 ns Combined 1765 540 1225 107 133 16 16 0 39 13 13 11 246 58 143 39 op operative management , NOM non-operative management , fail failure of NOM, MB morbidity , ML mortality , LR liver related 6 When Should You Operate on Major Hepatic Trauma? 70 IVC, and suprahepatic IVC and the porta hepatis. The suprahepatic IVC is readily controlled within the pericardium either via a median sternotomy or by incising the tendinous portion of the diaphragm. The suprarenal IVC can be approached by performing a Kocher maneuver or by mobilizing the left lobe of the liver and entering the lesser sac along the lesser curvature of the stomach. Historically the Schrock atriocaval shunt has been advocated for addressing these injuries. However, survival is uniformly poor and this technique has all but been abandoned [ 2 , 23 ]. Morales Uribe et al. discussed their operative management techniques for blunt liver trauma. Of 19 patients, three with grade I liver injuries required no intervention to the liver, one with a grade II injury was treated by suturing. Of the remaining high grade injuries, seven were treated with packing alone, two with packing plus suturing, two with packing plus non-anatomic resection , one with packing plus hepatic artery ligation. They had eight liver related mortalities including both patients undergoing resection and the patient undergoing hepatic artery ligation [ 18 ]. Polanco et al. reported a much lower mortality rate for operative management . In 144 patients with liver trauma, resection was performed in 56 patients, packing in 30 patients, hepatorrhaphy in 15, and no liver intervention in the remaining 16. They had an overall series mortality of 9 % including non-operatively managed patients and those who succumbed in the trauma bay. Patients managed by resection had an overall morbidity of 62.5 % and mortality of 17.8 %, but liver related morbidity and mortality were 30 % and 9 % respectively. Mortality for hepatic venous injury was only 25 %. The authors advocate resection for continued venous bleeding, devitalized tissue for which the injury has done a large portion of the resection, and major bile leak s [ 26 ]. Recommendations Liver injuries should be operated on when a patient is hemodynamically unstable, develops peritonitis, continues to bleed, or has other abdominal injuries requiring an operation. Although some studies have shown successful NOM of penetrating liver injuries, this has been done in specialized centers with the resources to do so. Due to the high possibility of missing associated abdominal injuries, NOM of penetrating liver injuries is not recommended. NOM of blunt injuries should only be attempted in centers with units capable of continuous monitoring of patients with serial physical exams and laboratory studies. Angiographic embolization can be used as a modality to aid NOM or as an adjunct to operative management , but should not be used as a replacement for operative intervention in unstable patients. All hemodynamically stable patients with blunt liver injuries should undergo a CT scan. Although higher grade injuries are associated with higher failure of NOM, CT grading alone is not predictive of failure. Operative management of liver injuries should include a graded response to the injury using the simplest technique possible to manage the problem. Peri-hepatic packing should be used as a damage control strategy to allow time for appropriate resuscitation. Liver resection can be used for defi nitive management when required. C. Evans and M.A. Croce 71 A Personal View of the Data More recent studies confi rm earlier data which resulted in a paradigm shift from operative to selective NOM of blunt liver injury. Characteristics of the liver injury which result in failure of NOM are not based on the radiographic severity of the injury, and the decision to operate should be based on the patients clinical status. While there is clearly a lower morbidity and mortality rate for patients in whom successful NOM can be achieved, one should not hesitate to operate when the situation calls for it. Recommendations • Trial of NOM for all patients who are hemodynamically stable with blunt liver injuries • Operative management for penetrating liver injuries • Operative management of blunt liver injuries who are or become hemodynamically stable, patients who develop peritonitis, patients with ongoing bleeding, or patients with associated injuries requiring an operation References 1. Stassen NA, Bhullar I, Cheng JD, Crandall M, Friese R, Guillamondegui O, et al. Nonoperative management of blunt hepatic injury: an Eastern Association for the Surgery of Trauma practice management guideline. J Trauma Acute Care Surg. 2012;73(5 Suppl 4):S288–93. 2. Savage SA, Croce MA. The management of liver injuries. In: Cameron JL, editor. The management of liver injuries. Philadelphia: Elsevier Saunders; 2014. 3. Malhotra AK, Fabian TC, Croce MA, Gavin TJ, Kudsk KA, Minard G, et al. Blunt hepatic injury: a paradigm shift from operative to nonoperative management in the 1990s. Ann Surg. 2000;231(6):804–13. 4. Polanco PM, Brown JB, Puyana JC, Billiar TR, Peitzman AB, Sperry JL. The swinging pendulum: a national perspective of nonoperative management in severe blunt liver injury. J Trauma Acute Care Surg. 2013;75(4):590–5. 5. Croce MA, Fabian TC, Menke PG, Waddle-Smith L, Minard G, Kudsk KA, et al. Nonoperative management of blunt hepatic trauma is the treatment of choice for hemodynamically stable patients. Results of a prospective trial. Ann Surg. 1995;221(6):744–53; discussion 753–5. 6. DuBose J, Inaba K, Teixeira PG, Pepe A, Dunham MB, McKenney M. Selective non-operative management of solid organ injury following abdominal gunshot wounds. Injury. 2007;38(9):1084–90. 7. MacGoey P, Navarro A, Beckingham I, Cameron I, Brooks A. Selective non-operative management of penetrating liver injuries at a UK tertiary referral centre. Ann R Coll Surg Engl. 2014;96(6):423–6. 8. Omoshoro-Jones JA, Nicol AJ, Navsaria PH, Zellweger R, Krige JE, Kahn DH. Selective nonoperative management of liver gunshot injuries. Br J Surg. 2005;92(7):890–5. 6 When Should You Operate on Major Hepatic Trauma? 72 9. Ghnnam WM, Almasry HN, Ghanem MA. Non-operative management of blunt liver trauma in a level II trauma hospital in Saudi Arabia. Int J Crit Illn Inj Sci. 2013;3(2):118–23. 10. Parray FQ, Wani ML, Malik AA, Thakur N, Wani RA, Naqash SH, et al. Evaluating a conservative approach to managing liver injuries in Kashmir, India. J Emerg Trauma Shock. 2011;4(4):483–7. 11. Saltzherr TP, van der Vlies CH, van Lienden KP, Beenen LF, Ponsen KJ, van Gulik TM, et al. Improved outcomes in the non-operative management of liver injuries. HPB (Oxf). 2011;13(5):350–5. 12. Kozar RA, Moore JB, Niles SE, Holcomb JB, Moore EE, Cothren CC, et al. Complications of nonoperative management of high-grade blunt hepatic injuries. J Trauma. 2005;59(5):1066–71. 13. Zago TM, Pereira BM, Calderan TR, Hirano ES, Rizoli S, Fraga GP. Blunt hepatic trauma: comparison between surgical and nonoperative treatment. Rev Col Bras Cir. 2012;39(4):307–13. 14. Christmas AB, Wilson AK, Manning B, Franklin GA, Miller FB, Richardson JD, et al. Selective management of blunt hepatic injuries including nonoperative management is a safe and effective strategy. Surgery. 2005;138(4):606–10; discussion 610–1. 15. Cohn SM, Arango JI, Myers JG, Lopez PP, Jonas RB, Waite LL, et al. Computed tomography grading systems poorly predict the need for intervention after spleen and liver injuries. Am Surg. 2009;75(2):133–9. 16. Fang JF, Wong YC, Lin BC, Hsu YP, Chen MF. The CT risk factors for the need of operative treatment in initially hemodynamically stable patients after blunt hepatic trauma. J Trauma. 2006;61(3):547–53; discussion 553–4. 17. Norrman G, Tingstedt B, Ekelund M, Andersson R. Non-operative management of blunt liver trauma: feasible and safe also in centres with a low trauma incidence. HPB (Oxf). 2009;11(1):50–6. 18. Morales Uribe CH, Lopez CA, Cote JC, Franco ST, Saldarriaga MF, Mosquera J, et al. Surgical treatment of blunt liver trauma, indications for surgery and results. Cir Esp. 2014;92(1):23–9. 19. Sivrikoz E, Teixeira GP, Resnick S, Inaba K, Talving P, Demetriades D. Angiointervention: an independent predictor of survival in high-grade blunt liver injuries. Am J Surg. 2015;209(4):742–6. 20. van der Wilden GM, Velmahos GC, Emhoff T, Brancato S, Adams C, Georgakis G, et al. Successful nonoperative management of the most severe blunt liver injuries: a multicenter study of the research consortium of New England centers for trauma. Arch Surg. 2012;147(5):423–8. 21. Dabbs DN, Stein DM, Scalea TM. Major hepatic necrosis: a common complication after angioembolization for treatment of high-grade liver injuries. J Trauma. 2009;66(3):621–7; discussion 627–9. 22. Gaarder C, Naess PA, Eken T, Skaga NO, Pillgram-Larsen J, Klow NE, et al. Liver injuries – improved results with a formal protocol including angiography. Injury. 2007;38(9):1075–83. 23. Kozar RA, Feliciano DV, Moore EE, Moore FA, Cocanour CS, West MA, et al. Western Trauma Association/critical decisions in trauma: operative management of adult blunt hepatic trauma. J Trauma. 2011;71(1):1–5. 24. Prichayudh S, Sirinawin C, Sriussadaporn S, Pak-art R, Kritayakirana K, Samorn P, et al. Management of liver injuries: predictors for the need of operation and damage control surgery. Injury. 2014;45(9):1373–7. 25. Peitzman AB, Marsh JW. Advanced operative techniques in the management of complex liver injury. J Trauma Acute Care Surg. 2012;73(3):765–70. 26. Polanco P, Leon S, Pineda J, Puyana JC, Ochoa JB, Alarcon L, et al. Hepatic resection in the management of complex injury to the liver. J Trauma. 2008;65(6):1264–9; discussion 1269–70.