**5. "Liver transplantation due to abdominal trauma" (Discussion)**

The aim of this retrospective investigation was to evaluate possible differences in the characteristics early and late mortality, sepsis and multi-organ failure as a function of the area of organ injury. Consideration of purely isolated organ injuries would not do justice to the complexity of a polytrauma, and may possibly lead to conclusions of no clinical relevance. The selection criteria "great severity of injury" of a specific organ system, with no attention paid to the average frequency and severity of additional injuries, would inaccurately illustrate the information value regarding organ-specific characteristics. It is well-known that liver injuries almost always accompany injuries to other organ systems. To consider only isolated liver injuries would lead to the description of a group that does not occur in this form in reality. The present study illustrates a patient group with a most severely injured organ system and the approach chosen was meant to investigate the impact on an organ system, in view of additional injuries, on the development of early mortality, transfusion requirement, sepsis, organ failure and late mortality.

To date, the effects of an isolated or primary liver injury on immunological function parameters has not to date been examined in either humans or animals. Only a retrospective evaluation weighted according to organ system can contribute to a more precise understanding of their significance for outcome, sepsis and MOF.

The results presented here show a clear increase in the incidence of sepsis from an MOF and early and late mortality with a severe liver injury. This increase seems to be liver-specific and stands out from the other organ systems investigated. Publications by Strong and Turnkey, which reported a mortality of over 11% of in isolated liver injuries, show a significantly lower mortality after liver trauma. However, these were not assessed in a comparably severely injured collective (31, 32). This stresses the significantly higher survival rates in patients with isolated liver injuries in comparison to poly-traumatized patients.

A review of the literature shows that the classification of more specific e.g. immunological consequences to different organ systems subsequent to polytrauma has not yet been examined. This applies both to experimental and clinical investigations and therefore the results presented here seem debatable, since they are only limited, given the low amount of literature in this regard. Despite the small amount of data, it seems beyond question that the participation of the liver in a traumatic event leads to an increase in mortality. However, there are some indirect references that characterize the liver as being a key organ after a trauma. At the beginning of the 1990s once Tinkhoff et al. had pointed out for the first time a connection between cirrhosis and outcome after trauma, this hypothesis was confirmed by numerous authors. In a matched pairs study, Dangleben et al. proved that cirrhosis of the liver is an independent prognosis marker of mortality, and with this they were able to demonstrate a correlation between mortality and the degree of the cirrhosis (definition according to Child-Turcotte-Pugh). These results were also verified by Christmas et al.: in addition to an increase in mortality and length of hospital stay, they showed a significant increase in the sepsis rate after trauma. Altogether 55% of the patients with cirrhosis of the liver in their study population died from sepsis. 33% of the patients with cirrhosis died compared to only 1% in the non-cirrhosis control group. These studies on cirrhosis of the liver and polytrauma show a close association between liver function and outcome after trauma.

In animal experiments, depending upon the quantity of the liver tissue removed, a liver resection leads to a clear restriction of synthesis efficiency, particularly for coagulation products (33). Furthermore, the clearance function for bacterial endotoxins is drastically reduced. The consequences can be expressed in a decompensated coagulation system, through to a Disseminated Intravascular Coagulation (DIC) in a spontaneous multi-organ failure after sepsis or in refractory shock to the extent that the effects of a liver resection resemble those of traumatic liver destruction (34-36).

However, traumatic liver damage is not necessarily associated with a measurable reduction in liver function. This is why, for example, Perdrizet et al. were able to demonstrate a clear increase in early mortality after reperfusion using a pig model, in which a blunt liver trauma was combined with a hemorrhagic shock. The increase in mortality resulted from continuous post-ischemic shock (37).

The significance of the liver in early trauma events was also demonstrated for example by Perl et al. after a thorax trauma in a mouse model. They showed for the first time a response to thorax trauma by Kupffer cells within 30 minutes. In so doing, the liver formed IL-6, TNF-alpha and IL-10 in high concentrations, without the liver itself being traumatized (38).

It has been proven that a tissue trauma leads to a significant reduction in immunological strength. The liver is a central organ of the reticuloendothelial system (RES) and its significance to the defense against infection has been described several times.

The results presented here show a clear increase in the incidence of sepsis from an MOF and early and late mortality with a severe liver injury. This increase seems to be liver-specific and stands out from the other organ systems investigated. Publications by Strong and Turnkey, which reported a mortality of over 11% of in isolated liver injuries, show a significantly lower mortality after liver trauma. However, these were not assessed in a comparably severely injured collective (31, 32). This stresses the significantly higher survival rates in patients with isolated liver injuries in comparison to poly-traumatized patients.

A review of the literature shows that the classification of more specific e.g. immunological consequences to different organ systems subsequent to polytrauma has not yet been examined. This applies both to experimental and clinical investigations and therefore the results presented here seem debatable, since they are only limited, given the low amount of literature in this regard. Despite the small amount of data, it seems beyond question that the participation of the liver in a traumatic event leads to an increase in mortality. However, there are some indirect references that characterize the liver as being a key organ after a trauma. At the beginning of the 1990s once Tinkhoff et al. had pointed out for the first time a connection between cirrhosis and outcome after trauma, this hypothesis was confirmed by numerous authors. In a matched pairs study, Dangleben et al. proved that cirrhosis of the liver is an independent prognosis marker of mortality, and with this they were able to demonstrate a correlation between mortality and the degree of the cirrhosis (definition according to Child-Turcotte-Pugh). These results were also verified by Christmas et al.: in addition to an increase in mortality and length of hospital stay, they showed a significant increase in the sepsis rate after trauma. Altogether 55% of the patients with cirrhosis of the liver in their study population died from sepsis. 33% of the patients with cirrhosis died compared to only 1% in the non-cirrhosis control group. These studies on cirrhosis of the liver and polytrauma show a close association between liver function and outcome after trauma. In animal experiments, depending upon the quantity of the liver tissue removed, a liver resection leads to a clear restriction of synthesis efficiency, particularly for coagulation products (33). Furthermore, the clearance function for bacterial endotoxins is drastically reduced. The consequences can be expressed in a decompensated coagulation system, through to a Disseminated Intravascular Coagulation (DIC) in a spontaneous multi-organ failure after sepsis or in refractory shock to the extent that the effects of a liver resection

However, traumatic liver damage is not necessarily associated with a measurable reduction in liver function. This is why, for example, Perdrizet et al. were able to demonstrate a clear increase in early mortality after reperfusion using a pig model, in which a blunt liver trauma was combined with a hemorrhagic shock. The increase in mortality resulted from

The significance of the liver in early trauma events was also demonstrated for example by Perl et al. after a thorax trauma in a mouse model. They showed for the first time a response to thorax trauma by Kupffer cells within 30 minutes. In so doing, the liver formed IL-6, TNF-alpha and IL-10 in high concentrations, without the liver itself being traumatized

It has been proven that a tissue trauma leads to a significant reduction in immunological strength. The liver is a central organ of the reticuloendothelial system (RES) and its

significance to the defense against infection has been described several times.

resemble those of traumatic liver destruction (34-36).

continuous post-ischemic shock (37).

(38).

The results shown here from the trauma registry indicate that in the group with severe liver trauma, there is a clear increase in the number of ECs in the early and late phases after trauma. This observation after liver trauma is also supported by other research groups. Thus, for example, the number of transfused ECs constitutes an independent prognosis factor in the post-traumatic period after liver trauma. The authors argue that the blood products possibly lead to an increase in the incidence of sepsis due to their antigenicity (39). Both Moore et al. and Malone et al. showed a clear connection between the number of transfused ECs and the occurrence of post-traumatic organ failure; Malone et al. even showed this correlation within the first 24 hours after trauma (40, 41). Critical in this respect, however, it should be fair to pose the question whether and to what extent the administration of erythrocytes causes immunoparalysis, particularly since trauma patients can develop sepsis and MOF without erythrocytes being administered. Hence, it should be discussed whether the correlation between ECs and mortality must possibly be considered as only an epiphenomenon, e.g. an extended tissue ischemia period. So the number of transfused blood products is also always a marker for injury severity, incidence of shock and length of ischemia time. This cannot be obviously separated by a multivariance analysis. In order to examine this question more closely, two subgroups were formed in the present analysis. Here it shows up remarkably that despite a similar ISS and number of transfused ECs, the patients with severe liver participation continue to predominate, with regard to mortality, sepsis and MOF. In this context, immune modulating substances contribute to a considerable reduction in infectious complications. After polytrauma, proteins such as granulocyte-macrophage colony-stimulating factor (GM-CSF) and interferon gamma can contribute to an improvement in post-traumatic immunoparalysis (42, 43). Patients with immune insufficiency, e.g. also due to liver damage, could benefit from the early use of immune modulating substances.

The evaluation of the data from the trauma registry concerning liver trauma (AIS>2) and treatment before and after 2000 shows the paradigm shift starting in 2000 mentioned in the introduction. The reduction in the rate of laparotomies from 2000 to 11.2% in hospitals affiliated with the trauma registry proves a rethink in care after abdominal injury. This resulted in a reduction in mortality of 2.4% in similar patients (ISS: 39.7 vs. 39.8). In order to better support this advantage of conservative treatment, however, more detailed study is necessary given that both preclinical and clinical care have made progress in the same time period. While in former times an exploratory laparotomy was nearly always performed, now conservative therapy under hemodynamically stable conditions is increasingly being recommended (44). Therefore, the portion in an American (multicenter) study was 47%. With 404 patients, a success rate of 98.5% was reported, where hemorrhaging appeared in only 3.5% of other complications (45).

In another series of 495 conservatively treated patients, the success rate was 94% and the average hospital treatment was 13 days, where only 1.9 EC/patient had to be transfused. The complication rate was 6.2%, whereby there was only 2.8% with hemorrhages. Liverrelated deaths or overlooked intestinal injuries were not observed.

Both groups predominantly involved not so serious liver traumas, whereas Moore type IV and V injuries (14%) were rather rare. In a study from Germany up to 2004, only 14% of all patients were treated conservatively. Moreover, the not so serious Type I-III injuries were operated in 2/3 of the cases (31/44), where no liver-related mortality was observed. The authors came to the conclusion that in view of the convincing data from the multicenter studies mentioned and numerous other, at times large patient groups, laparotomy is probably an overtreatment in most patients with Type I-III injuries and seems to be of no real advantage regarding survival, morbidity and duration of treatment (46). Data from this study corroborates this statement.

The matter of the urgent criteria for operating on abdominal and liver trauma is not clearly answered in the literature. The criteria are not uniform and often refer to the term "unstable". It has been shown however by Clarke et al. that mortality increases by 1% every 3 minutes after a trauma involving hematogenic shock, so the time from arrival at the ER to the laparotomy has a crucial effect on the outcome (47).

In addition to acute trauma care following abdominal injury, the therapeutic option of transplant plays a role in chronic hepatic damage rather than in acute injuries. Persistent chronic hepatic damage is mostly seen in the form of "secondary sclerosing cholangitis". The option of transplantation for acute, inoperable hepatic damage also plays an admittedly minor role, but in times of scarce organ availability this should be exercised within reason.

Therefore, treatment of liver trauma has rapidly changed over the past decades. Thus, especially development of the intensive and emergency medicine as well as coagulation substitution reveal a more and more conservative therapy approach against the severity of the injury. To date, merely 10% of the liver trauma patients are surgically treated, 90% follow a conservative therapy regimen. In the process, the overall mortality of 60% could be reduced to about 6% over the past century (48-50).

However, in a few patients with liver injuries it may still occur that they cannot be treated adequately despite exploitation of all conventional surgical measures. Continuous noncontrollable acute bleeding, non-reconstructible liver injuries, like e.g. injuries of the liver's veins or the bile duct system, and a liver insufficiency caused by trauma, e.g. shock liver, allow for the consideration of LT (51, 52).

LT then remains the only available life-saving procedure for these patients. However, not all patients are suitable candidates for LT. Pre-transplant neurological status, severe sepsis, MOF, and accompanying severe injuries may all be contraindications to LT. Furthermore, there is a fundamental difference whether a patient is transplanted due to acute noncontrollable liver injury or due to subacute (e.g. shock liver) respectively chronic (e.g. secondary biliary cirrhosis) liver mutation after occurred trauma. Ultimately, only a fraction of patients with uncontrollable liver trauma are deemed to be candidates for transplantation. Like those patients who die before LT, mortality after LT is usually secondary to hemodynamically instability, infections and MOF (53, 54).

The underlying severity of the injury and the occasionally life-threatening other injuries are reflected by the results in our patients who received a LT due to trauma all from a motorvehicle accident. These patients differ fundamentally from the majority of our other liver transplant patients in the peri-operative prognosis. Based on our clinical experience, the most relevant preoperative prognostic factors negatively influencing the post-transplant outcome have been the hemodynamic, secondary injuries and the recipient age. There are diverging opinions about the role of the MELD score as a prognostic factor for the postoperative outcome in such cases.

The results following LT for uncontrollable traumatic liver injuries are substantially worse than those of LT for sub-acute/chronic and elective indications. In fact, the general patient

studies mentioned and numerous other, at times large patient groups, laparotomy is probably an overtreatment in most patients with Type I-III injuries and seems to be of no real advantage regarding survival, morbidity and duration of treatment (46). Data from this

The matter of the urgent criteria for operating on abdominal and liver trauma is not clearly answered in the literature. The criteria are not uniform and often refer to the term "unstable". It has been shown however by Clarke et al. that mortality increases by 1% every 3 minutes after a trauma involving hematogenic shock, so the time from arrival at the ER to

In addition to acute trauma care following abdominal injury, the therapeutic option of transplant plays a role in chronic hepatic damage rather than in acute injuries. Persistent chronic hepatic damage is mostly seen in the form of "secondary sclerosing cholangitis". The option of transplantation for acute, inoperable hepatic damage also plays an admittedly minor role, but in times of scarce organ availability this should be exercised within reason. Therefore, treatment of liver trauma has rapidly changed over the past decades. Thus, especially development of the intensive and emergency medicine as well as coagulation substitution reveal a more and more conservative therapy approach against the severity of the injury. To date, merely 10% of the liver trauma patients are surgically treated, 90% follow a conservative therapy regimen. In the process, the overall mortality of 60% could be

However, in a few patients with liver injuries it may still occur that they cannot be treated adequately despite exploitation of all conventional surgical measures. Continuous noncontrollable acute bleeding, non-reconstructible liver injuries, like e.g. injuries of the liver's veins or the bile duct system, and a liver insufficiency caused by trauma, e.g. shock liver,

LT then remains the only available life-saving procedure for these patients. However, not all patients are suitable candidates for LT. Pre-transplant neurological status, severe sepsis, MOF, and accompanying severe injuries may all be contraindications to LT. Furthermore, there is a fundamental difference whether a patient is transplanted due to acute noncontrollable liver injury or due to subacute (e.g. shock liver) respectively chronic (e.g. secondary biliary cirrhosis) liver mutation after occurred trauma. Ultimately, only a fraction of patients with uncontrollable liver trauma are deemed to be candidates for transplantation. Like those patients who die before LT, mortality after LT is usually

The underlying severity of the injury and the occasionally life-threatening other injuries are reflected by the results in our patients who received a LT due to trauma all from a motorvehicle accident. These patients differ fundamentally from the majority of our other liver transplant patients in the peri-operative prognosis. Based on our clinical experience, the most relevant preoperative prognostic factors negatively influencing the post-transplant outcome have been the hemodynamic, secondary injuries and the recipient age. There are diverging opinions about the role of the MELD score as a prognostic factor for the

The results following LT for uncontrollable traumatic liver injuries are substantially worse than those of LT for sub-acute/chronic and elective indications. In fact, the general patient

secondary to hemodynamically instability, infections and MOF (53, 54).

study corroborates this statement.

the laparotomy has a crucial effect on the outcome (47).

reduced to about 6% over the past century (48-50).

allow for the consideration of LT (51, 52).

postoperative outcome in such cases.

survival rates are approximately 50-75%. Unfortunately, the few reported cases in the current literature are quite inhomogeneous, reflecting different transplant eras, clinical experience, LT techniques/procedures, and clinical conditions of the patients prior to undergoing LT. In addition these case reports mostly outline the clinical course of liver transplant patients following trauma. While accurate comparison of the clinical presentation of patients across various case reports is not always possible, we can say, based on the available data in 3 case series, that the clinical conditions of our patients appear to be similar to those reported (55-57).

Delis et al. also describe 4 patients with liver trauma in their work who were transplanted in the course of their disease. Non-uniform genesis of these patients are reflected in a range of relatively positive GCS scores. These may be explained by the fact that 3 of the abovementioned patients had gun-shot liver injuries and hence no, as common in blunt liver injuries, large-area, complex liver injuries. Furthermore, one patient was transplanted after two years due to secondary biliary cirrhosis caused by trauma. This explains the fairly good results in this group with a patient survival rate of 75% after more than 9 years.

Altogether 3 patients with liver injuries due to car accidents, that were hepatectomized preoperatively due to massive unsalvageable liver trauma, are described by Ringe et al. This quite more homogenous patient population is better comparable to our study and demonstrated a patient survival rate geared to our results. Thereby, Ringe postulates a bilateral approach in patients where no sufficient hemostasis after liver trauma is achievable. After an indication for total hepatectomy depending on hemodynamic parameters, a than obligatory liver transplantation is carried out as soon as possible. In his works, however, also patients are described that could not be allocated with an adequate organ in time due to the present lack of donor organs.

Also comparable with our results are those published in the 1980ies by Esquivel et al. on 2 traffic accident victims with nonreconstructable injuries to the portal vein and following nonfunctional hepatic remnants. In literature, these are the first published cases of liver transplantations after liver trauma.

The majority of our patients demonstrated one or more of negative prognostic factors. This study covers all recorded liver transplantations for otherwise uncontrollable liver trauma due to motor-vehicle accidents at our hospital. These cases often had poor general prognoses. Despite the acute condition of our patients, our results, patient survival rate is 50% with a corresponding graft survival of 25%, are among the first reports on survival rates in a homogenous series to date in the literature.
