Retrohepatic Vena Cava and Juxtahepatic Venous Injuries.
WHILE the overall trend in recent years has been toward nonoperative management of liver trauma, few injuries pose as formidable a challenge to the trauma surgeon as retrohepatic vena cava and juxtahepatic venous injuries. The ongoing massive hemorrhage in conjunction with the difficult exposure and visualization in the acute setting are enough to cause trepidation in the boldest of surgeons. Adding to the list of difficulties is the knowledge that not only must vascular control be obtained, but also it must be accomplished before irreversible coagulopathy, acidosis, and hypothermia occur. These facts taken together make the impressive mortality rates of these injuries understandable and have led to development of several surgical modalities for dealing with this type of injury. The following case report will serve as a useful segue into indications and techniques of each.
A 20-year-old African American woman was brought to the emergency room with gunshot wounds to the chest, abdomen, and extremity. She had an entrance wound at the tip of the right scapula, an exit wound in the left upper quadrant, and through-and-through wounds to the left thigh. In the field, her blood pressure was 60 mm Hg systolic and heart rate was 140/min. At the time of presentation, she was alert but combative. She was maintaining her own airway and was breathing at 28 times per minute, but with breath sounds that were difficult to assess. There were no obvious signs of fracture of the left leg, and femoral and pedal pulses on the affected side were good.
A chest tube placed in the right hemithorax yielded 300 mL of blood, and two large bore peripheral intravenous lines were started. After rapid infusion of 2 liters of crystalloid, blood pressure was 100 mm Hg systolic with a heart rate of 116/mm. The patient was then taken to the operating room.
The abdomen was entered through a midline incision, and a large amount of blood clot was evacuated as all four quadrants were packed. Once the blood pressure stabilized, careful inspection of bilateral lower quadrants and left upper quadrant showed no injuries. A large retroperitoneal hematoma in the epigastrium appeared to extend to the right upper quadrant. As the packing was removed from the area, massive hemorrhage was encountered from beneath the liver in the region of the porta hepatis and vena cava and persisted after manual occlusion of the portal triad. A large defect was visible in the liver parenchyma in this area, but further efforts to delineate the nature of the injury were unsuccessful since exsanguinating rates of hemorrhage were encountered with every attempt to further explore the region. Hemostasis was finally achieved by bimanually compressing the liver in the mediolateral plane and pushing it posteriorly. With the likelihood of a retrohepatic caval or juxtahepatic venous injury, a core b ody temperature of 34[degrees]C (93.2[degrees]F), a pH of 7.11, and a rising transfusion requirement, laparotomy pads were packed around the liver to hold it in its compressed position. Because of the extensive blood loss, the fascia was closed in hopes of promoting tamponade of the injury site. The patient was transported to the surgical intensive care unit, having had 8 units of packed red blood cells, 800 mL more from a cell saver, and 4 units of fresh frozen plasma. Her temperature was 33[degrees]C (91.4[degrees]F); prothrombin time was 17.9 sec, and partial thromboplastin time was 16.8 sec.
For the first 24 hours postoperatively, bladder pressure increased to 30 mm Hg, and peak inspiratory pressure rose to an average of 55 cm [H.sub.2]O. In view of adequate urine output and satisfactory tissue oxygenation, it was decided to continue to wait and follow the bladder pressure before re-exploration because the immediate danger posed by premature removal of packing probably outweighed the potential problems associated with increased intra-abdominal pressure. These values plateaued, and the patient remained hemodynamically stable without signs of decompensation or infection. The packing was removed in the operating room on postoperative day 5 without signs of rebleeding or intraperitoneal infection. The patient was returned to the intensive care unit. She initially did well as bladder pressure dropped to 12 mm Hg, peak inspiratory pressure dropped to 30 cm [H.sub.2]O, and urine output remained satisfactory. Attempts to wean her from the ventilator were unsuccessful, however, and on day 7 after initial operation, chest x-ray changes were consistent with adult respiratory distress syndrome (ARDS). Despite broad spectrum antibiotics, multiple ventilator manipulations designed to minimize barotrauma, and a decompressing bedside laparotomy with vacuum packing, pulmonary function continued to worsen. As the prognosis worsened, the family opted to withdraw support, and the patient died on postoperative day 18.
Before discussing the management of juxtahepatic venous trauma, a brief review of the anatomy is worthwhile. The retrohepatic cava is approximately 8 to 10 cm long, and it receives venous return from the large right, middle, and left hepatic veins. Cadaver studies have revealed that the average distance from the right hepatic vein to the diaphragm is 0.7 cm.  These veins course in an intersegmental distribution and have an extremely short extrahepatic course. All of these facts taken together make obtaining vascular control in this area intimidating, especially in an emergency setting.
Advances in prehospital care have resulted in greater numbers of patients arriving at the hospital in extremis. Patients with juxtahepatic cava injuries who previously would have exsanguinated in the field are now surviving long enough to make it to the operating room, and it is incumbent on the trauma surgeon to be familiar with the variety of methods available to address this injury. At the time of exploration, juxtahepatic venous trauma should be suspected when occlusion of the portal triad (the Pringle maneuver) fails to stop bleeding from the liver parenchyma or from below a damaged hepatic lobe. 
Attempts to achieve vascular isolation of the liver led to the development of the atriocaval shunt. It consists of extending the laparotomy incision to include a sternotomy. This is followed by placement of a pursestring suture in the right atrial appendage and snaring the suprarenal and intrapericardial cava. Classically, a 36F chest tube with an extra side hole cut 20 cm from the most proximal port is inserted through the atrial hole, flushed with blood, and secured with the most proximal hole within the heart and the most distal holes below the level of the suprarenal snare. A recent amendment to the technique consists of using a cuffed endotracheal tube instead of a chest tub, since this obviates the need for suprarenal control. Both methods preserve venous return while isolating the area, which can then be expeditiously addressed. The most serious potential complications of the technique are the introduction of an air embolism and perforation of any of the involved vascular structures. The reported mort ality associated with atriocaval shunting ranges from 50% to 90%, and this is attributed to four causes.  First, there is often a delay in recognition of the juxtahepatic nature of the wound Second, the technical requirements for proper shunt insertion are often unfamiliar to the surgeon. Third, the decision to insert the shunt is often turned to only in desperation. Finally, the technique is sometimes applied in the setting of blunt trauma, which most authors believe to be almost exclusively fatal. The common theme running through most of these determinants of mortality is delay, because this time interval allows worsening of coagulopathy, acidosis, and hypothermia until they are irreversible. In this basic fact lies the controversy over the technique, because, while it is a drastic measure, its successful application would seem to lie in resorting to it quickly,
The depressing results seen with use of the atriocaval shunt led to the development of balloon shunts as a possible means of isolating the liver while preserving venous return. This method consists of threading a catheter with an inflatable balloon through the saphenofemoral junction and up into the vena cava, passing distal to the site of injury. Once in place, the balloon is blown up and provides caval occlusion, while a series of proximal side ports allow blood from below the kidneys to return to the heart.
Its advantages include the relative ease of insertion and the avoidance of a sternotomy. There are some disadvantages, however. The original shunt developed for this purpose had a 24F diameter, which decreased venous return to the point that mean arterial pressure dropped 50% and right atrial output by 30%.  The diameter was then increased to 28F, with resolution of these problems. However, the catheter was still prone to dislodging either superiorly, with a decrease in venous return, or inferiorly, with cessation of tamponade of the hepatic veins. Additionally, overinflation of the balloon can lead to occlusion of the outflow port and/or rupture of the balloon, with a potentially lethal air embolism, should air have mistakenly been used to inflate the a balloon. Finally, thrombus formation is always a potential concern as well.  Experience with the balloon shunt is still limited and makes up only 12% of the intracaval shunt experience in the literature. 
Sequential Vascular Clamping
A direct approach at control without shunting has been described. Known as sequential vascular clamping, it consists of the placement of occlusive vascular clamps across the suprarenal cava, suprahepatic cava, and the portal triad. The abdominal aorta frequently requires cross clamping as well. The common sequelae to caval occlusion include arrhythmias and cardiac arrest. Aortic cross clamping has obligate ischemia to the organs perfused below the level of the clamp; thus, the kidneys, viscera, and spinal cord are all at risk. The technique has been neither extensively used nor studied in the setting of trauma. Its biggest proponents in the literature are Khaneja et al,  who had 7 survivors among 10 patients with this injury, all treated with total vascular occlusion.
Liver injuries that are so massive they are not amenable to the previously mentioned techniques will occasionally be encountered. These patients are universally profoundly unstable, cold, and coagulopathic at the time of exploration. It is for this subset of patients with liver trauma that the method of perihepatic packing is reserved and recommended. The evolution of the technique is an interesting one. In the early part of the century, the placement of gauze packs in and around the liver was the mainstay of therapy. By the end of World War II, however, the complications of rebleeding at the time of pack removal and sepsis from prolonged placement were so significant that the procedure was widely denounced. In 1969, Walt (as quoted by Feliciano and Pachter ) wrote, "There is virtually no place in modern surgery for gauze packing of the liver as sepsis and recurrent bleeding are almost inevitable sequelae." The predominant school of thought shifted once again, though, as more seriously injured patients su rvived the field long enough to be evaluated by a surgeon. These patients are so grossly unstable that they are not candidates for definitive repair of their wounds at the time of exploration. Damage control is undertaken in which hemostasis is rapidly achieved using perihepatic packs, followed by resuscitation in the intensive care unit. Once acidosis, hypothermia, and coagulopathy are corrected, they are returned to the operating room for definitive exploration.
The technique of perihepatic packing consists of manually approximating and compressing the liver parenchyma with placement of dry laparotomy pads around the liver to hold it in this position. Under no circumstances should the pads be placed in a parenchymal defect. The skin is then usually reapproximated, leaving the fascia open to allow for the inevitable increase in intra-abdominal pressure. Alternatively, the fascia can be initially closed to provide a better tamponade effect. Close monitoring of the intra-abdominal pressures by following the bladder pressures is mandatoiy. Should an abdominal compartment syndrome develop, the fascia can be reopened and the skin edges closed without removing the packing. This tamponade effect should provide enough hemostasis to allow for clotting off of the injury site. The technique is especially effective for addressing the small vessel ooze associated with a coagulopathy.  Refinements in the method have been described in which a sterile drape has been interposed be tween the laparotomy pads and the liver to prevent adherence of the pads to the parenchyma, as well as to preserve the hemostatic effects of the sponges.  The application of this technique to juxtahepatic cava injuries is most readily illustrated by a series by Beal.  In this series of 10 patients with atrialcaval shunts, 7 died. Six of these deaths were from exsanguination, and in 4 of these patients, hemorrhage was initially controlled by packing during shunt preparation and insertion. Beal concluded that if these patients had been treated with packing alone, some may have survived.
Controversy surrounds the timing of packing removal. A balance must be struck between the need to remove the packs early enough to prevent septic complications while allowing them to remain in place long enough to prevent rebleeding. In one retrospective series, waiting 36 to 72 hours for pack removal produced the same complication rates as early removal, but by waiting this long to remove the packs, the rebleeding rate dropped from 21% to 4%.  When reading the literature on the optimal timing, however, a range from 12 hours to several days will be encountered.
Potential complications associated with the technique are significant. As previously mentioned, intra-abdominal sepsis and renewed hemorrhage with packing removal are of concern. Additionally, the increased intra-abdominal pressure predisposes the patient to development of an abdominal compartment syndrome. Compression of the cava and renal veins will be present to some degree and must be balanced against the beneficial effect this has on promoting hemostasis. In our case, we closed the fascia to err on the side of tamponade. In one anecdotal report,  the top third of a fascial incision was closed (ie, the portion over the liver) and the bottom two thirds was left open to allow for decompression of the increased pressure. The defect was then controlled with a sheet of silastic. Regardless, it is a clinical situation that requires profound vigilance, since straying too far toward either extreme can be disastrous for the patient.
The problems facing those who undertake the management of juxtahepatic venous injuries are daunting The challenges begin with the need to rapidly and accurately recognize the nature of the injury, as well as the patient's physiologic tolerance to further intervention. After these assessments are made, the definitive treatment can be tailored accordingly.
(1.) Feliciano D, Pachter JL: Hepatic trauma revisited. Curr Probl Surg 1989; 7:459-524
(2.) Feliciano DV, Moore EE, Mattox EL: Trauma. Stamford, Conn, Appleton and Lange, 3rd Ed, 1996
(3.) McAnena OJ, Moore EE, Moore FA: Insertion of a retrohepatic vena cava balloon shunt through the saphenofemoral junction. Am J Surg 1989; 158:463-466
(4.) Khaneja SC, Pizzi WF, Barie PS, et al: Management of penetrating juxtahepatic inferior vena cava injuries under total vascular occlusion. J Am Coll Surg 1997; 184:469474
(5.) McHenry CR, Fedele GM, Malangoni MA: A refinement in the technique of perihepatic packing. Am J Surg 1994; 168:280-282
(6.) Beal SL: Fatal hepatic hemorrhage: an unresolved problem in the management of complex liver injuries. J Trauma 1990; 30:163-169
(7.) Caruso DM, Battistella FD, Owings JT, et al: Perihepatic packing of major liver injuries. Arch Stag 1999; 134:958-963
(8.) Sheridan R, Driscoll D, Felsen R: Packing and temporary closure in a liver injury. Injury 1997; 28:711-712
* In juxtahepatic venous Injuries, rapid identification of the nature of the injury is vital.
* Atriocaval shunting, balloon shunting, sequential vascular clamping, and perihepatic packing are all important maneuvers that should be considered a month the available treatments.
* Juxtahepatic venous trauma should be suspected when occlusion of the portal triad (the Pringle maneuver) fails to stop bleeding from the liver parenchyma or from beneath a damaged hepatic lobe.