Reperfusion is the most critical period during orthotopic liver transplantation (OLT) for the anesthesiologist. It is associated with serious hemodynamic instability, which could be caused by a massive release of free radicals, endotoxin and inflammatory cytokines.1, 2 Whether more severe postreperfusion syndrome (PRS) causes a poorer long-term outcome or merely reflects other patient factors that lead to a generally poorer outcome, it seemed that a link existed.3 The purpose of this study was to investigate that link between the severity of PRS and patient and liver allograft short-term outcome.
The greatest part of liver allograft injury occurs during reperfusion, not during the cold ischemia phase. The aim of this study, therefore, was to investigate how the severity of postreperfusion syndrome (PRS) influences short-term outcome for the patient and for the liver allograft. Over a 2-year period, 338 consecutive patients who presented for orthotopic liver transplantation (OLT) were included in this retrospective study. They were divided into 2 groups according to the severity of the PRS they experienced. The first group comprised 152 patients with mild or no PRS; the second group comprised 186 patients with significant PRS. Perioperative hemodynamic parameters, coagulation profiles, blood product requirements, incidence of infection, incidence of rejection and outcome data for both groups were collected and analyzed. There was no demographic difference between the groups except for age; group 2 had older patients than group 1 (54.94 ± 9.07 versus 51.52 ± 9.91, P = 0.001). Compared to group 1, group 2 patients required more red blood cell transfusions (11.31 ± 10.90 versus 8.08 ± 7.89 units, P = 0.002), more fresh frozen plasma transfusions (10.25 ± 10.96 versus 7.03 ± 7.64 units, P = 0.002), more cryoprecipitate (1.88 ± 4.72 units versus 0.61 ± 1.80 units, P = 0.001), and were more likely to suffer from fibrinolysis (52.7% versus 41.4%, P = 0.041). Interestingly, group 2 had a shorter average warm ischemia time than group 1 (33.19 ± 8.55 versus 36.21 ± 11.83 minutes, P = 0.01). Group 2 also required longer, on average, mechanical ventilation (14.95 ± 29.79 versus 8.55 ± 17.79 days, P = 0.015), remained in the intensive care unit longer (17.65 ± 31.00 versus 11.49 ± 18.67 days, P = 0.025), and had a longer hospital stay (27.29 ± 32.35 versus 20.85 ± 21.08 days, P = 0.029). Group 2 was more likely to require retransplantation (8.6% versus 3.3%, P = 0.044). In conclusion, the severity of PRS during OLT appears to be related to the outcome of patient and liver allograft. Liver Transpl 14:504–508, 2008. © 2008 AASLD.
PATIENTS AND METHODS
After the approval of Institutional Review Board, the clinical and laboratory data of 338 patients who underwent primary OLT during a 2-year period (2002 and 2003) were retrospectively reviewed. We classified PRS as mild when the decrease in blood pressure and/or heart rate was less than 30% of the anhepatic levels and was short-lived (≤5 minutes); for this classification, it also needed to respond to calcium chloride (1 g intravenously) and/or epinephrine intravenous boluses (≤100 μg) without requiring continuous infusion of vasopressor agents. We defined significant PRS as severe hemodynamic instability such as persistent hypotension (more than 30% of the anhepatic level), asystole, or hemodynamically significant arrhythmias; patients who required a vasopressor infusion during the intraoperative period were considered to have significant PRS, whether that infusion continued through the postoperative period. Other presentations of significant PRS included prolonged (defined as greater than 30 minutes) or recurrent (defined as reappearing within 30 minutes after resolution) fibrinolysis that required treatment with antifibrinolytic agents. The patients were divided into 2 groups according to the severity of PRS: group 1 comprised patients with mild PRS; group 2 was made up of patients with significant PRS.
Patient demographics, quality of the liver allograft, the use of veno-venous bypass, hemodynamic data, incidence of fibrinolysis, intraoperative blood transfusion, incidence of postoperative infection and rejection, duration of ventilatory support, postoperative renal complications, duration of intensive care unit (ICU) and hospital stay, and patient and graft survival were compared between the groups. The quality of the liver allograft is classified as a traditional versus an Extended Donor Criteria (EDC) graft. The criteria that were used to define EDC allografts were the following: a non–heart beating donor, age > 65 years, serum sodium level >155 mEq/L, donor liver macrosteatosis ≥ 30% on biopsy, cold ischemia time > 16 hours, and warm ischemia time > 90 minutes. The Mann-Whitney Test was used to compare groups. P < 0.05 was considered statistically significant. For survival, the Kaplan-Meier test was used. Data are presented as mean ± standard deviation and median with range.
Varying degrees of PRS were observed in all 338 patients for the study period. Of these, 152 patients (45%) were classified as having mild PRS (Group 1), whereas 186 patients (55%) were classified as significant (group 2). The preoperative data (Table 1) of both groups showed that there was no significant difference in gender, weight, serum creatinine, incidence of chronic renal failure, hematocrit, international normalized ratio, platelet count, etiology of cirrhosis, and Model for End-Stage Liver Disease (MELD) scores. The only significant difference was the age; patients in group 2 were older than patients in group 1 (P = 0.001).
|Preoperative Data||Group 1 (n = 152)||Group 2 (n = 186)||P Value|
|Age (years)||51.52 ± 9.91||54.94 ± 9.07||0.001*|
|Weight (kg)||80.43 ± 16.99||78.33 ± 16.37||0.250|
|Serum Creatinine||1.18 ± 0.93||1.24 ± 0.96||0.530|
|Chronic Renal Failure||2.6%||4.3%||0.409|
|MELD score||15.09 ± 6.77||15.22 ± 6.18||0.852|
|INR||1.43 ± 0.40||1.38 ± 0.34||0.194|
|Hematocrit||31.19 ± 5.77||31.02 ± 5.64||0.786|
|Platelets count||81.79 ± 54.60||85.93 ± 52.91||0.481|
|PNC-HC + alcohol||11.2%||8.1%||0.330|
Intraoperative data (Table 2) showed that group 2 patients required more red blood cells and fresh frozen plasma (P = 0.002 for each); they also required more cryoprecipitate (P = 0.001), but platelet transfusion requirements were similar to group 1. The incidence of fibrinolysis was higher in group 2 (P = 0.041). No significant difference was noted in cold ischemia time (CIT); group 2 had a small but statistically significant decrease in warm ischemia time (WIT) (P = 0.010). The utilization of veno-venous bypass was distributed evenly between the 2 groups, and no association with severity of PRS could therefore be investigated.
|Intraoperative Data||Group 1 (n = 152)||Group 2 (n = 186)||P Value|
|CIT (hours)||11.00 ± 3.19||11.55 ± 3.34||0.124|
|WIT (minutes)||36.21 ± 11.83||33.19 ± 8.55||0.010*|
|RBC (units)||8.08 ± 7.89||11.31 ± 10.9||0.002 *|
|FFP (units)||7.03 ± 7.64||10.25 ± 10.96||0.002*|
|PLTS (units)||8.01 ± 6.64||9.31 ± 8.44||0.121|
|CRYO (units)||0.61 ± 1.80||1.88 ± 4.72||0.001*|
Postoperative data (Table 3) showed that group 2 trended toward a lower incidence of rejection within the first month (though statistical significance was not reached) but had a higher incidence of retransplantation due to primary graft nonfunction (P = 0.044). The number of patients who received EDC grafts was spread evenly between the 2 groups, and the utilization of EDC did not appear to affect the severity of PRS as defined by the aforementioned criteria. Overall 3-year survival rate of the 338 patients was 74%, with no significant difference between the 2 groups (Fig. 1).
|Postoperative Data||Group 1 (n = 152)||Group 2 (n = 186)||P Value|
|Days on Ventilator||8.55 ± 17.79||14.95 ± 29.79||0.015*|
|ICU Stay (days)||11.49 ± 18.67||17.65 ± 31.00||0.025*|
|Hospital Stay (days)||20.85 ± 21.08||27.29 ± 32.35||0.029*|
|Dialysis 1st month||16.3%||16.8%||0.728|
|Rejection 1st month||20.4%||13.1%||0.073|
Many studies have documented that hypoxia and ischemia are not the primary cause of the pathogenesis of tissue injury secondary to occlusion of arterial blood flow. A major cause may be due to the restoration of oxygen delivery to previously ischemic tissue. From these studies, one can conclude that most tissue injury actually occurs at the time of reperfusion.4 Hepatic injury due to ischemic reperfusion occurs in many clinical scenarios where a temporary interruption of liver blood flow (whether partial or complete) could happen during liver surgery for trauma or tumors, as well as during organ preservation before liver transplantation.5 In OLT surgery, blood flow to the liver is completely interrupted during the donor liver procurement and is not re-established until many hours after preservation techniques have been commenced.
The liver is quite resilient to ischemia and can tolerate warm ischemia of up to 90 minutes without serious cell necrosis.6 Ischemia that lasts greater than 90 minutes can induce irreversible hepatic cell necrosis at the time of the insult.7 However, ischemia that lasts less than 90 minutes can still lead to some degree of liver injury during the ischemia and more extensive lesion during the reperfusion.8, 9 Because the very same rules are applicable during OLT, keeping the WIT below 90 minutes appears crucial in preventing serious hepatic damage. In our study, the WIT was similar in both groups (Table 2) and compatible with the above limits in avoiding additional serious hepatic injury.
During cold ischemia, rapid depletion of high-energy molecules such as adenosine triphosphate occurs as the initial insult due to mitochondrial dysfunction,10 leading to cellular membrane damage. The cell membrane dysfunction is manifested by loss of intracellular ion homeostasis with intracellular accumulation of calcium and sodium.11, 12 As a result of this energy failure, activation of proteolytic enzymes (proteases, phospholipases, and endonucleases) occurs, leading to transformation of xanthine dehydrogenase to xanthine oxidase and degradation of adenosine triphosphate to hypoxanthine.12, 13 These changes in the micro-milieu form the basis for oxidative stress which will occur as soon as blood flow and oxygen are restored.14, 15 The reoxygenation which occurs at reperfusion leads to oxidative stress with activation of Kupffer cells; this is followed by microvascular disturbances (no-reflow) and activation of polymorphonucleocytes.16 Oxidative stress and the production of free radicals cause a biphasic pattern of reperfusion injury.17 An early phase of activation of Kupffer cells with mild initial injury (1–3 hours) is followed by more serious injury initiated by polymorphonucleocyte activation (6–24 hours). Both phases culminate in extensive hepatocyte and sinusoidal cell damage.17, 18 In this study, the CIT was not significantly different between the groups (P = 0.124). As a result, the effect of CIT may be excluded as the only influencing factor in patient and allograft outcome.
No standard criteria to score the severity of PRS exist. In this study, the criteria that were used to score the PRS relied on the hemodynamic changes that occurred at reperfusion. In addition, the severity of PRS was also scored on the basis of fibrinolysis which may have occurred after reperfusion. Accordingly, the severity of PRS is defined by its effects on cardiovascular parameters and/or its effects on thromboelastograph rather than by measuring and detecting the blood levels of different mediators which have been claimed to be key elements in the etiology of PRS. However, the cardiovascular presentation is an important element of PRS and was used to diagnose and define its occurrence.19 Relying on the clinical presentation of PRS is one limitation of this retrospective study, but this method of diagnosis of PRS has been used in multiple clinical studies that confirmed its practicality and usefulness.19
In this study, group 2 had a higher blood product transfusion requirement; the adverse effects of blood transfusion on the patient and allograft outcome are well documented.20 Although the effect of blood transfusion on the outcome of group 2 patients cannot be excluded, this increased transfusion requirement is itself linked to the fibrinolysis associated with severe PRS. The higher blood transfusion requirement in group 2 patients may explain the trend toward a reduced incidence of rejection in the first postoperative month, related to the immunosuppressant effect of blood transfusion.21
The increased allograft loss in group 2 and significantly higher retransplantation rate (P = 0.046) could be due to liver damage caused by prolonged use of high doses of vasopressor therapy, or alternatively could simply be the result of liver reperfusion injury. It may be difficult to explain the cause of a higher allograft loss in group 2 patients. However, the continuous use of high-dose vasopressor therapy in other clinical conditions without the development of liver failure may suggest that reperfusion injury plays an important role here. Finally, the use of vasopressors may have an additive effect on a multifactorial etiology of allograft failure in this group.
The use of EDC grafts has become more prevalent in many transplant centers to widen the donor organ pool, shorten the time spent on waiting lists, and lower the candidate death rate while awaiting transplant. Although there is no universal definition of EDC, the criteria used in this study to define EDC have been standard practice at the study center. Patients who received EDC grafts were evenly divided between the 2 groups (P = 0.353), and the use of these grafts did not affect the severity of the PRS. Ayanoglu et al.22 reported similar findings, documenting that PRS occurred in an unpredictable fashion unrelated to CIT or to recipient criteria. Another study by Nanashima et al.23 concluded that the age of the donor (>50 years) was associated with development of PRS, but that this did not adversely affect patient or graft outcome. These prior studies examined only one of several factors that now define the EDC and its effects on patient and graft outcome. In this study, EDC was strictly defined and the use of EDC did not affect the severity of PRS or seriously impact the patient and allograft outcome. It is interesting to note, however, that in another study, patients who received EDC grafts had a higher incidence of postoperative renal dysfunction or failure that required prolonged hemodialysis or continuous venovenous hemodialysis (25% versus 15% in traditional grafts).24
Because the severity of PRS may affect patient and allograft outcome, prevention of its occurrence or attenuation of the resultant hemodynamic and coagulation changes may theoretically improve outcome. One of the strategies used to lower the incidence of reperfusion injury has been ischemic preconditioning.25, 26 Although this strategy is still in the research phase, it could have clinical value with a potential therapeutic benefit in dealing with reperfusion injury. The use of therapeutic substances such as methylene blue, N-acetylcysteine, pentoxyfilline,27, 28 and anti-cytokines to attenuate the effects of reperfusion have shown some favorable results. However, their affect on the outcome needs to be verified.
In conclusion, the severity of PRS correlates with patient and allograft outcome in OLT. A clinical study of new therapeutic modalities to attenuate the PRS is warranted.