Figure 26.1 CT images of liver injury by grade. Arrows denote site of injury.
Liver injury, regardless of grade (I–V), should be initially managed nonoperatively in the haemodynamically stable child, as defined by age-specific heart rate and blood pressure [7].
There are, of course, no current randomised controlled trials regarding the NOM of isolated hepatic injuries in children, although guidelines have been proposed based on classification of liver injury or physiologic stability. Stylianos initially reviewed the practices of 32 paediatric trauma centres and proposed a set of five guidelines, and this was validated in a follow-up prospective study [16,17]. These were based on CT grade of liver injury, and they recommended an intensive care unit (ICU) stay of 1 day for grade IV injury only; lengths of stay of 2, 3, 5 and 5 days for grades I, II, III and IV, respectively; and no predischarge or postdischarge imaging and an activity restriction (normal age appropriate) of 3, 4, 5 and 6 weeks for grades I, II, III and IV, respectively.
McVay et al. proposed guidelines based on physiologic stability based on a retrospective review of 101 patients with isolated liver or spleen injuries [18]. The guidelines suggest an admission to an ICU for patients who were initially unstable but responded to resuscitation. In the ICU, patients were nil by mouth and on strict bedrest, and haematocrit levels were checked every 6 h, and prothrombin and partial thromboplastin time checked every 12 h for the first 24 h. If the haematocrit dropped below 21%, the patient was transfused and monitored for an additional 24 h in the ICU. If the haematocrit was above 21% after the first 24 h, the patient was transferred to a normal ward. For stable patients (or patients transferred to a normal ward from the ICU), patients were placed on bedrest and nil by mouth, and their haematocrit was checked every 8 h. If the haematocrit was <21%, the patient was transfused and monitored for an additional 24 h. If the haematocrit was >21%, the patient was advanced to a regular diet and allowed to ambulate. The patients were then followed up with an abdominal US in 4 weeks and activity restrictions were associated with the US results. Prospective randomised controlled studies are needed to assess the safest and most cost-efficient management algorithm for isolated hepatic injuries.
26.6 NONOPERATIVE MANAGEMENT COMPLICATIONS
NOM complications are rare but do occur, and rates ranged from 3% to 11%, with the cause being ongoing bleeding, biliary tract–related complications, abdominal compartment syndrome or liver abscesses [19,20,21]. The complication rate increases with grade of injury; Giss et al. [19] reported a 0% complication rate for grade I and II injuries, while grade III had an 11% complication rate and grade IV had a 20% complication rate. In one study, only 34% of complications required operative intervention [22]. Persistent bleeding is the most common complication, with an incidence of up 9% [4]. A recent study of angioembolisation of paediatric patients with abdominal trauma had an overall success rate of 87% and a major complication rate of 6%, with all of these resolving [23]. This study included all patients regardless of additional injuries or severity of injury. Other studies report success rates up to 100% [24].
Biliary tract–related complications are commonly biliary tract disruptions, biloma or biliary peritonitis. Bile leaks are more difficult to diagnose with NOM, but feeding intolerances or elevated liver function tests (LFTs) should alert the team to the possibility. Bile leaks may be diagnosed intraoperatively (possible incidental finding while operating for some other reason) or with magnetic resonance (MRCP) or endoscopic retrograde (ERCP) cholangiopancreatography. The latter is commonly used if there is a high suspicion for leak, as it allows for both diagnosis and treatment. Bile peritonitis may cause the need for operative intervention or may be treated with percutaneous drainage. Bilomas are intrahepatic collections of bile. Small bilomas usually regress without treatment, while larger (>3 cm) ones usually require percutaneous drainage.
Abdominal compartment syndrome is a rare finding that occurs if the intra-abdominal pressure is >20 cm of water with associated organ failure. Abdominal compartment syndrome should be managed with decompressive laparotomy.
Hepatic abscess are uncommon, with a rate of 3% among NOM patients, and can be treated with percutaneous drainage [22].
The need for follow-up imaging or the appropriate timing of imaging has not been defined, and such decisions should be made based on a patient’s clinical need. Stylianos recommends no predischarge or postdischarge imaging routinely and found that after adoption of their guidelines, the rates of follow-up CT were 7%, 9%, 11% and 23% for grades I, II, III, and IV, respectively. Before the adoption of the guidelines, the rates were 34%, 46%, 54% and 52%. This resulted in an overall rate of 87% of patients receiving follow-up imaging, and no adverse events were found [17].
McVay et al. recommend an abdominal US at 4 weeks to assess resolution of injury, even for grade IV injuries. The authors state that the use of US allows for earlier return to activity and is cost-effective [18]. Further studies are required to determine if follow-up imaging is needed and if it is the most appropriate time for the study.
26.8 OPERATIVE MANAGEMENT OF LIVER TRAUMA
Most liver trauma is successfully managed nonoperatively, and as a result, the overall experience with operative management is decreasing. However, it remains essential for the paediatric trauma surgeon to recognise failing patients early and have the broad spectrum of operative treatment strategies available. Intimate knowledge of the perihepatic vasculature, as well as the segmental anatomy of the liver, is essential to safe and successful exploration and management. The two main indications for operative exploration after liver trauma include haemodynamic instability despite aggressive fluid resuscitation and peritonitis, which may not manifest until days after the injury due to associated bowel perforation or bile leak. The operating room should be equipped with a full set of vascular instruments, a large number of laparotomy pads, haemoclips, an argon beam coagulator and topical haemostatic agents prior to incision.
Figure 26.2 James Hogarth Pringle (1863–1941) and his manoeuvre. The original paper, published in 1908 in the Annals of Surgery [59], described a series of eight patients with significant liver injuries, all of which ultimately died. He developed his idea of temporary compression of the portal triad in two laparotomies with an assistant ‘holding the portal vein and the hepatic artery between a finger and thumb (which) completely arrested all bleeding’. (Portrait by William Dring reproduced. Held by the Royal College of Physicians and Surgeons of Glasgow.)
26.8.1 Management of the unstable patient
The abdomen is best explored through a generous midline incision, followed by evacuation of the intraperitoneal haematoma, manual compression of the liver and packing of all four quadrants.
Packs are ideally placed above and below the site of hepatic haemorrhage, using caution not to cause unintentional caval compression and impaired venous return. The initial goal is to temporarily control rapid volume loss and allow for correction of hypotension and coagulopathy. While resuscitation is ongoing, packs should be carefully monitored. Rapid soaking indicates the need for either better packing or additional blood products to correct the coagulopathy. In that case, temporary hepatic inflow occlusion should be performed by fenestration of the lesser sac or isolation of the hepatoduodenal ligament with vascular control using either a finger or a Rummel tourniquet known as the Pringle manoeuvre* (Figure 26.2). This control of hepatic inflow can be maintained up to 60 min or, with reperfusion intervals, up to 120 min and is without any significant long-term sequelae in the otherwise normal liver [25,26]. If coagulopathy cannot be corrected after initial adequate packing, further attempts to surgically control the source of bleeding may be futile and return to the ICU with packing and temporary closure for resuscitation is indicated [27].
* James Hogarth Pringle (1863–1941), Australian surgeon who worked in Glasgow. He is described as one of the most progressive British surgeons of his day for his pioneering surgery, for example, in fracture fixation and hindquarter amputations for tuberculosis of the pelvis.
26.8.2 Management of parenchymal haemorrhage
Patients successfully stabilising during intraoperative resuscitation must undergo a thorough exploration for definitive control of the bleeding source. Packs are sequentially removed from the quadrants, exploring the right upper quadrant last. Moderate bleeding from the liver surface may be controlled by argon beam coagulation, topical agents, placement of suture ligature with or without omental packing or a combination thereof [28,29]. Suture ligature with nonabsorbable material most often controls deep parenchymal venous or arterial haemorrhage. Deeper parenchymal bleeding may require temporary inflow occlusion to adequately localise the bleeding vessel. An arterial source should be suspected if the Pringle manoeuvre significantly reduces the bleeding and hepatic artery ligation at the hilum is the treatment of choice in the exsanguinating patient. However, selective angiographic hepatic artery embolisation may be considered in patients stabilised after damage control surgery and packing or with ongoing postoperative transfusion requirement.
Superficial penetrating injuries are often caused by a stabbing mechanism [30,31]. Such injuries can be managed by performing a tractotomy in order to expose transected parenchymal vessels. A stapling device such as an Endo GIA with vascular staples is fired sequentially to unroof the entire tract [32
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
