A comparison by multivariate analysis between the hospital mortality group and the no hospital mortality group found high donor age, high blood loss and annual surgery volume <30 liver transplantations to be statistically significant risk factors. Annual LDLT volume was less than 10 cases in the first 5 years of our transplantation center (Figure 1). It was a little surprising that the outcome for the period with <20 annual cases and the outcome for the period with ≥ 20 annual cases were not statistically different. When the case volumes were small, we were more cautious in that selected recipients had lower MELD scores, the GRWR was kept above 0.8% and safe donor grafts were used (normal vessel or biliary tract anatomy). The first 15-20 LDLT cases are associated with a significant surgery learning curve [16, 17]. Annual surgery volume grew to 20-30 cases in the 6th and 7th years; there were more urgent patients with acute liver failure conditions, including some recipients with rapid development of hepatic dysfunction associated with encephalopathy or renal failure. By putting the recipient in a positive pressure isolation room with 2 beds in the intensive care unit, the nosocomial infection risk was reduced. Over the period with 30 to 50 annual cases, we used soft power of critical care training and continuing education of the staffs. An important point is that multidisciplinary characteristics such as integration and organization management add to the operative learning curve (multiple anastomoses of vessel or biliary tract anatomy). The training of the multidisciplinary staffs covered color Doppler techniques, anesthesia, critical care, rejection identification and infection treatment. We have a combined intensivist in infectious diseases and critical care, and we established a specialized liver transplantation ward. Thus, we had put infrastructure in place to ensure favorable outcomes. Operative time, blood loss, re-operation, intensive care unit days, length of stay and hospital mortality decreased significantly after annual surgical volume reached 30 cases. After improving soft power and care quality, our physical selection criteria became less strict in regard to senior recipients, donor age, the GRWR lower bound (range 0.55-0.8), ABO-incompatible liver transplantation, and multiple hepatic duct or portal vein anastomoses, which were not limiting factors or difficult techniques. However, neither portal vein complications nor early allograft dysfunction led to a significantly higher mortality rate. In addition, the hospital mortality rate when the annual surgical volume was less than 20 liver transplantations was 10.5%, and it decreased to 5.4% and 5.1% when annual surgical volume was ≥ 30 and ≥ 50 liver transplantations, respectively.
The complications with the highest mortality rates were cerebrovascular problems, including subarachnoid and intracerebral hemorrhages [18, 19]. Common early postoperative complications following LDLT include thrombosis in reconstructed major blood vessels (portal vein or hepatic artery reconstructed with an artificial vascular graft or cryopreserved vein grafts) [20]. In the absence of ongoing bleeding after operation, our center considered maintaining an INR between 1.5 and 2, a platelet count >50,000/μL and a fibrinogen level >100 mg/dL as satisfactory. Hypertension occurs usually in the initial treatment when the systolic blood pressure is greater than 160mmHg or the diastolic blood pressure is greater than 100mm Hg. In the general population, intracranial hemorrhage may occur in association with coagulopathy, acute hypertension or chronic hypertension. An intracranial or subarachnoid hemorrhage after liver transplantation that requires immediate craniotomy and removal of a hematoma may be combined with nosocomial infections and result in high mortality. Postoperatively, blood pressure and fibrinogen levels can be monitored closely to help prevent post-transplant intracranial hemorrhages [18].
In our center’s policy, when old age, cardiomegaly, history of coronary artery disease (CAD), or massive ascites is a trait in an alcoholic cirrhosis patient, the patient undergoes regular electrocardiograms and echocardiographies for pre-operative cardiovascular assessment of LT. If the patient is an LT candidate, then dobutamine stress myocardial perfusion scanning is performed for detection of CAD. Active coronary artery disease is a relative contraindication to liver transplantation and at a minimum should be treated as aggressively as possible preoperatively (stenting, angioplasty). For high cardiopulmonary risk patients, we evaluated their hemodynamic measurements by pulmonary artery catheter. Cardiac dysfunction and moderate to severe portopulmonary hypertension (mean pulmonary arterial pressure ≥ 35 mmHg and elevated pulmonary vascular resistance) were diagnosed and were considered contraindications of liver transplantation [21]. Cardiovascular complications occurred in 8 patients after LDLT. Two patients survived. One survivor had arrhythmia with atrial fibrillation, paroxysmal supraventricular tachycardia, ventricular tachycardia and ventricular fibrillation in the first week after liver transplantation. The recurrent arrhythmia was poorly controlled by anti-arrhythmia treatment and defibrillation. Careful laboratory monitoring and supplementation were warranted; electrolytes were provided to maintain a normal level. An echocardiography showed moderate mitral regurgitation and tricuspid regurgitation. Still, anti-arrhythmia treatment did not prevent the arise of severe bradycardia with atrioventricular block. The discontinuation of antiarrhythmic drug therapy showed no significant improvement, and a cardiologist suggested inserting temporary pacemakers. The other surviving patient’s electrocardiogram presented ST elevation and increased levels of myocardial enzymes. Percutaneous coronary intervention was done to exclude an obvious problem.
The other 6 cases with cardiovascular complications resulted in hospital mortality. One case involved the rupture of an aortic mycotic aneurysm after transplantation. This patient had a diagnosis of abdominal mycotic aneurysms and infection by salmonella species in pre-transplantation image evaluation. Treatments of patients with mycotic aneurysms caused by salmonella should include antibiotic therapy and surgery [22]. This patient had a high MELD score and massive ascites; a cardiovascular surgeon recommended antibiotic therapy and endovascular stent repair after liver transplantation. On the 12th day after liver transplantation, a rupture of an aortic mycotic aneurysm resulted in emergency surgery; the cause of death was hemorrhage. Surgery as an early-stage prevention of aneurysm rupture may decrease morbidity or mortality [22]. In three cases, cardiac arrest occurred within 2-5 minutes after reperfusion in intraoperative status. They experienced high-quality cardiopulmonary resuscitation, but their hemodynamic conditions remained unstable. They were then placed on extra-corporeal membrane oxygenation (ECMO). The 2nd case had hyperkalemia combined with acidosis due to a massive intraoperative blood transfusion and renal failure history. In the 3rd and 4th cases, CAD and pulmonary embolism diagnoses were ruled out by percutaneous coronary intervention. Cardiac death within 5 minutes after graft reperfusion may result from many possible causes, including hyperkalemia, acidosis, pulmonary embolism, hypothermia, arrhythmia, cardiac tamponade, acute heart failure, and myocardial infarction [23, 24]. We finally reached the diagnoses of acute coronary syndrome. The patients’ conditions were hemodynamically stable, and their ECMOs were successfully removed after a few days. However, they suffered from cardiac arrest in the ICU. In the 5th case, we diagnosed acute coronary syndrome by percutaneous coronary intervention, and the 6th case showed mild pulmonary hypertension by echocardiography. Both cases developed septic shock. CAD was found to be the leading cause of early mortality, and it was followed by infection [25]. Cardiovascular complications are the main cause of non-graft-related mortality after LT.
Early allograft dysfunction occurred in 10 cases after LDLT. Successful liver function recovery occurred in 2 cases. Six cases resulted in additional liver transplantations (2 case deaths from severe sepsis), and 2 cases resulted in death while waiting for graft liver. An analysis revealed that causes of early allograft dysfunction are high MELD score (≥35) combined with low GRWR (range 0.67 ~ 0.69), hepatic steatosis (moderate) in the donor graft and senior donor age (59 and 65 years). The donor's age and moderate and severe steatosis were already established risk factors for early allograft dysfunction or primary graft dysfunction or non-function [26-30]. Donor age (≥ 50 years) was an independent risk factor to affect regeneration after transplant [26]. Effective graft regeneration may be associated with stem cells or progenitor cells in an elderly donor’s liver [9]. The best survival results in our study were observed when the MELD score was below 15. Those with low MELD scores (n=2, MELD 7 and 13) could better tolerate an initial graft dysfunction than the recipients with medium or higher MELD scores who were not achieving successful transplantations due to hepatic steatosis (moderate), senior donor age, or a small-sized graft (GRWR<0.7) succumbing to early allograft dysfunction or graft failure. Values above this limit constitute important factors associated with post-transplant hospital mortality [28-30]. This should also be taken into account when deciding whether to transplant.
Many studies have analyzed the association between hospital mortality and transplantation volume in centers [1, 5]. Our study analyzed the association between surgeon volume (LT cases) and hospital mortality. The early phase of a surgeon’s LDLT practice involves frustration and numerous hardships. Massive blood loss, long operative time, prolonged mechanical ventilation, and post-operative complications lead to sepsis and early hospital mortality. Our center is a non-metropolitan hospital; decisive transplant leadership and team staff centripetal force were absolutely essential. Transplant leadership made decisions on valid multidisciplinary integration and organization management. We found a statistically significant association between hospital mortality and surgical volume. This study has some limitations. This study excluded deceased donor liver transplants. Our center only performed 4 to 8 deceased donor liver transplants per year due to a shortage of available organs in Taiwan, and initial development was lower liver graft. An LDLT department needs multidisciplinary integration and organization management. In sharing the development experience of our center, we believe all emerging centers must have access to mentoring while developing an LDLT program.