EXPERIMENTAL AND CLINICAL TRANSPLANTATION

Objectives: In irreversible acute liver failure, liver transplant is the only viable treatment option. In this study, our aim was to evaluate and determine the factors related to mortality in patients who received liver transplants in accordance with King’s College criteria for acute liver failure in order to prevent futile operations.

Materials and Methods: Our study included 65 adult patients with acute liver failure who received liver transplant according to King’s College criteria. Factors related to mortality, including demographic and operative data, causes of acute liver failure, severity of encephalopathy, and laboratory data, were retrospectively analyzed. Patients who received living-donor liver grafts had donations from first-degree to fourth-degree relatives.

Results: Of 65 patients analyzed, 55.3% were women. Ninety-day mortality rate was 36.9%. Preoperative bilirubin levels in survivor and nonsurvivor groups were 16.3 ± 9.6 and 21.3 ± 10.7 mg/dL, respectively (P = .03). Mortality rates of patients with bilirubin above and below 9 mg/dL were 31.8% and 8.3%, respectively (P = .03). Of patients who died, 75% were women (significantly more women than men, P = .015). Patients who had deceased-donor liver transplants had a significantly higher mortality rate than those who had living-donor liver transplants (52% vs 27.5%; P = .046). At 3 days posttransplant, bilirubin, creatinine, aspartate aminotransferase, phosphorus, sodium, and ammonia levels were significantly different between survivor and nonsurvivor groups (P < .05).

Conclusions: We found living-donor liver transplant to be superior versus deceased-donor liver transplant with regard to development of acute liver failure. Reasons could include the long wait period for deceased donors and liver grafts coming from marginal donors. Bilirubin level and presence of grade 4 encephalopathy had predictive values for poor prognosis of patients.

Key words : Bilirubin, Encephalopathy, Liver dysfunction, Risk factors

Acute liver failure (ALF) is defined by massive hepatocellular damage, severe liver dysfunction, and coagulopathy with or without encephalopathy in those without prior history of liver disease.¹ The clinical signs of liver failure develop rapidly, and mortality can reach 70% with little chance of spontaneous recovery of patients.² The annual incidence of ALF has been reported to be 1 to 6 cases per 1.5 million population and constitutes 10% of the indications for liver transplant (LT) worldwide.^3-5 There are a number of causes, including hepatitis A and B, pregnancy toxic hepatitis, mushroom poisoning, and accidental ingestion of other rare chemicals, as well as indeterminate causes.^6-10

In the modern era of LT, together with improvements in surgical techniques, anesthesia, and intensive care management, patients with ALF now have a chance for cure.^11-13 There are 2 major points to be defined during the initial evaluation of patients with ALF. First is to determine whether the patient really needs LT and to avoid unnecessary LT in patients who have a chance of spontaneous recovery.¹² The reason for this is because LT patients require lifelong immunosuppressive therapy and surgical problems may be encountered in the early postoperative period, which can have significant risks. The second point that should be clarified is determination of the posttransplant prognosis of patients.¹² Because ALF is a spectrum of diseases, the affected individuals are heterogenous and many factors can affect outcomes of patients.¹¹

In patients with ALF, there is multisystemic involvement, including loss of functional liver volume and its metabolic capacity, acute lung injury due to disrupted clearance of circulatory toxins, bone marrow suppression, increased risk of sepsis, increased catabolic state, pancreatitis (more prominent in drug toxicities), myocardial injury, and portal hypertension in subacute cases.⁴ In addition, patients with ALF who are candidates for urgent LT are actually in a state of emerging multiorgan failure, and major surgery in these patients has the potential for high mortality and morbidity.^3,11,12 Nevertheless, currently, 5-year patient survival after orthotopic LT has steadily increased and has reached 70% to 75%.¹⁴

The Clichy criteria¹⁵ and King’s College criteria (KCC)¹⁶ have been extensively used to predict outcomes of patients with ALF. These and similar classifications that have been developed can predict the preoperative mortality rate of patients with ALF due to the disease process. However, these criteria are not accurate, especially in the era of advanced extracorporeal liver support systems and newly developed treatment alternatives.¹⁷ Furthermore, variables determining patient outcomes are not uniform, with variations according to country and region. The main clinical dilemma is to choose the patients who would actually benefit and survive a complex procedure like LT. This is a complex disease process, and often the physician and the family have to make prompt decisions regarding treatment options. Therefore, further definitive, easily applicable, sensitive, and specific criteria to predict the post-LT prognosis of patients with ALF are needed.

Information regarding complex disease processes such as ALF can be deduced from the experience of high-volume centers. At our center, we perform 200 to 250 orthotopic LTs annually, with nearly 10% of our cases due to ALF. Therefore, in this study, our aim was to evaluate factors affecting prognosis, 90-day mortality, and survival outcomes of patients who underwent LT for ALF in our center.

Materials and Methods

Patient allocation and study design
Between January 2012 and November 2018, 1548 patients received LT for various causes at our center. Of these, 148 patients (9.6%) received an LT for ALF according to KCC. Patients who received retransplants (n = 7) and pediatric transplant recipients (n = 75) were excluded from this study. As a result, 65 adult orthotopic LT recipients due to ALF were included in our study analyses. Deceased-donor organ allocation was performed according to the emergency status of the patients and was decided by the National Coordination Center of Ministry of Health. Living-donor liver grafts were donated by first-degree to fourth-degree relatives of patients, with all donors undergoing multidisciplinary evaluations at our center.

Patients were grouped into 2 groups: survivors (n = 41) and nonsurvivors (n = 24). Patient medical records were retrospectively analyzed for demographic characteristics, clinical data (such as KCC compatibility), intensive care unit (ICU) admission, ventilatory support before and after LT, presence and grade of hepatic encephalopathy, determination of cerebral edema, the preoperative and postoperative laboratory variables, operative data (including graft type), and 90-day mortality and survival data. None of the 65 patients were excluded from the analysis.

Study parameters
Demographic data evaluated in the present study included age, sex, cause of ALF, and operative parameters, including graft type used (that is, deceased-donor LT [DDLT] or living-donor LT [LDLT]). Status of deceased-donor liver grafts (marginal or not), which were defined according to Attia and associates,¹⁸ was also retrospectively analyzed. Laboratory data included parameters related to KCC. These parameters included arterial blood analysis for pH and lactate and ammonia concentrations; renal function tests consisting of serum creatinine, serum sodium, and potassium concentrations; liver function tests consisting of aspartate aminotransferase (AST), alanine aminotransferase (ALT), and bilirubin levels (in this study, we present total bilirubin levels); and coagulation tests, including international normalized ratio evaluated in the preoperative and postoperative periods.

Clinical data included the most severe stage of the disease at the time of enlisting for LT. These parameters included Model for End-Stage Liver Disease (MELD) score development and severity of hepatic encephalopathy, presence of cerebral edema, pre-LT need for ICU admission, and pre-LT and post-LT need for mechanical ventilation support. Furthermore, 90-day mortality and survival data of patients were also retrospectively evaluated. The highest laboratory data levels before and after LT were used for our analyses. The reason for this was the high variance in the pre-LT follow-up interval of patients and heterogeneity of the use of extracorporeal liver support systems.

Preoperative follow-up
A number of problems may be encountered and should be treated by the physician during the preoperative period of patients with ALF. These problems include circulatory failure, coagulopathy, gastrointestinal bleeding, encephalopathy, cerebral edema, metabolic problems (such as acidosis), and hypoglycemia.19 At our center, we actively resuscitate these patients using liver support systems, renal replacement therapy, vasopressor therapy, and mechanical ventilation. For those who require resuscitation, determination of possible causative factors and prognostic risk assessment using KCC and Clichy criteria are performed. We are a tertiary referral center; therefore, most of the treated patients are scheduled for LT, with patients having bridging therapies until LT.

Postoperative follow-up
During postoperative follow-up, patients receive both standard and more aggressive posttransplant follow-up procedures. The former includes daily evaluation of vascular inflow and outflow of the graft using Doppler ultrasonography for the first postoperative 10 days. Other follow-up postoperative procedures may include more aggressive vasopressor and fluid replacement therapy. Surveillance of neurologic complications such as development of cerebral edema is closely monitored. Liver support systems, renal replacement therapy and mechanical ventilation, are performed when necessary.

Immunosuppressive and antimicrobial therapy
In patients without a history of ongoing infection, 1000 mg ampicillin/sulbactam is administered intravenously every 4 hours. If any sign of infection is observed, the same antimicrobial therapy is continued for another 48 hours postoperatively at a twice daily dose.

Our immunosuppressive therapy protocol includes initiation of 100 mg/day methylprednisolone in the ICU, with dose tapered gradually according to the graft function. In addition, mycophenolate mofetil is also initiated at a dose of 3 g/day. On postoperative day 3, in patients without any contraindication, administration of tacrolimus is initiated at a dose of 2 mg/day.

Statistical analyses
Continuous variables are expressed as means and standard deviation. Discrete variables are expressed as number of patients affected together with percentages. Associations between dependent and independent variables were analyzed using parametric tests, including chi-square test. P ≤ .05 was considered as statistically significant. Cutoff values for laboratory parameters to predict mortality were determined by receiver operating characteristic curve analyses. Data of patients who received marginal versus nonmarginal deceased-donor LTs were analyzed using Mann-Whitney U test. All statistical evaluations were performed with SPSS software (SPSS: An IBM Company, version 20.0, IBM Corporation, Armonk, NY, USA).

Demographic and clinical characteristics of the study groups
Our study included 65 adult patients who underwent orthotopic LT for ALF. Mean age of the patients was 36.6 ± 14.3 years. Twenty-nine patients (44.6%) were men, and 36 patients (55.4%) were women. Demographic and clinical characteristics of patients are summarized in Table 1.

The distribution of causes of ALF in our cohort was as follows: toxic hepatitis in 24 patients (36.9%), hepatitis B virus in 18 patients (27.7%), indeterminate causes in 17 patients (26.2%), hepatitis A virus in 1 patient (1.5%), Wilson disease in 2 patients (3.2%), and miscellaneous causes in 3 patients (4.6%). The distribution of mortality rate according to the causes of ALF is summarized in Figure 1A. When causes of ALF in patients were analyzed, acute decompensated Wilson disease and indeterminate causes showed higher mortality rates versus other causes; however, this difference did not reach statistical significance (P = .09; Figure 1A).

Patients with indeterminate causes of ALF had the worst prognosis, with a mortality rate of 64.7%. Patients with ALF causes due to hepatitis A or B, Wilson disease, and toxic hepatitis showed mortality rates ranging from 50% to 100%. The mean MELD score in our patient cohort was 30.3 ± 8.2. Encephalopathy was observed in 49 patients (75.4%), with grade 2/3 and grade 4 encephalopathy present in 38 (58.5%) and 11 patients (16.9%), respectively. Distribution of mortality rates according to grade 2/3 encephalopathy and grade 4 encephalopathy is summarized in Figure 1B. Although the mortality rate in grade 4 encephalopathy was higher than in grade 2/3 encephalopathy, it did not reach statistical significance (mortality of 63.6% vs 36.8%; P = .11). Mortality was observed in 24 patients (36.9%).

Survivor and nonsurvivor groups were similar in terms of age and MELD scores. However, 75% of patients who died were female, with difference versus men showing statistical significance (P = .015). Postoperative neurologic dysfunction and cerebral edema were more frequently encountered among female patients (7/36 women [19.5%] vs 2/29 men [6.8%]). The type of liver graft had a significant effect on patient survival. This will be discussed in the next section.

Impact of type of liver graft on patient survival
Forty patients (61.5%) received LDLT and 25 patients (38.5%) received DDLT. Twenty-nine of the 40 patients (72.5%) who received LDLT survived, with mortality rate of 27.5% among patients who had LDLT. However, in patients who had DDLT, the mortality rate was 52% (13/25 patients with DDLT died during the first 90-day postoperative period). The distribution of mortality according to the graft type is summarized in Figure 1C. Only one of the deceased donors was a split right lobe graft, which had been transplanted to a 53-year-old male patient who died on postoperative day 2 due to neurologic complications. This difference in mortality among LDLT and DDLT recipients was statistically significant (P = .046).

In the LDLT recipient group, there were 29 patients with various levels of encephalopathy, with 5 patients (17.2%) having severe encephalopathy. In the DDLT recipient group, there were 20 patients with various levels of encephalopathy, with 6 patients (30%) having severe encephalopathy. This difference did not reach statistical significance (P = .3; chi-square test).

Among preoperative values, only phosphate and potassium levels were significantly different between the LDLT and DDLT groups. Preoperative serum phosphate levels in the LDLT and DDLT groups were 2.9 ± 1.3 and 4.1 ± 2.6 mg/dL, respectively (P = .007). Preoperative serum potassium level was lower in the LDLT group than in the DDLT group (3.9 ± 0.7 vs 4.5 ± 1.2 mEq/L; P = .04). The only postoperative parameter that showed significant difference between DDLT and LDLT was AST level, which was higher in the DDLT group than in the LDLT group (3134 IU/L vs 1222.6 IU/L; P = .03).

Deceased donors were evaluated according to graft status, which revealed that there were 13 patients who received marginal liver grafts and 11 patients who received nonmarginal liver grafts. There were no differences in the frequency of patients with marginal versus nonmarginal deceased liver grafts (P = .77; Pearson chi-square test). The status of the liver graft in 1 patient was not determined because the graft characteristics were normal; however, the deceased donor had lupus nephritis and arteritis. Reasons of marginality of deceased-donor liver grafts included advanced donor age (≥ 80 y), hypernatremia (≥ 160 mEq/L), administration of high doses of noradrenaline and dopamine, elevated liver enzyme levels, split liver grafts, and infection.

Mortality rates of patients who received marginal versus nonmarginal liver grafts were 46.2% (n = 6) versus 54.5% (n = 6), respectively (P = .76; Mann Whitney U test).

Laboratory results in the survivor versus nonsurvivor groups
Preoperative and postoperative laboratory variables of patients are summarized in Table 1. Among the preoperative parameters, patients who did not survive had significantly higher mean bilirubin levels than patients who survived (21.9 ± 10.7 vs 16.3 ± 9.6 mg/dL; P = .035). Although the standard deviation of bilirubin was high, receiver operating characteristic curve analysis did not give a cutoff value for bilirubin. However, mortality rates for different cutoff values that were arbitrarily defined (20, 15, 10, and 9 mg/dL) were analyzed, with mortality rate for bilirubin levels of < 9 mg/dL of 8.3% and mortality rate of level ≥ 9 mg/dl of 31.8% (P = .03).

With regard to postoperative laboratory variables, bilirubin levels were significantly higher in the nonsurvivor group than in the survivor group (13.3 ± 9.5 vs 7.9 ± 4.3 mg/dL; P = .003). Mean postoperative creatinine levels in the nonsurvivor and survivor groups were 1.6 ± 1.1 and 0.9 ± 0.6 mg/dL, respectively (P = .004). Mean postoperative ammonia in the nonsurvivor group was 357.1 ± 2701 μg/dL, whereas it was 215.9 ± 189.9 1 μg/dL in the survivor group (P = .016). Mean postoperative AST was significantly higher in the nonsurvivor group (3541.6 ± 4327.2 vs 1030.6 ± 899 IU/L; P = .001). Mean postoperative phosphate levels in the nonsurvivor and survivor groups were 5.7 ± 3.3 and 4 ± 1.4 mEq/L, respectively (P = .006). Mean postoperative sodium concentration was slightly but statistically higher in the nonsurvivor group (144.9 ± 5.3 vs 142.1 ± 4.4 mEq/L; P = .027).

We analyzed the outcomes of adult orthotopic LT recipients with ALF who were seen over the past 5 years. Acute liver failure comprised 9.6% of the indications for orthotopic LT in our series, which is similar to the rate of 7.2% reported in the European liver transplant registry data.20 There are major differences between our practice and those of Western centers, as most of our transplant procedures involve LDLT (approximately 62%), which includes grafts obtained from up to fourth-degree relatives because of the limited deceased-donor organ supply. Therefore, factors that differ among survivors and nonsurvivors posttransplant must be analyzed to define patients in whom orthotopic LT is futile. For this purpose, we evaluated standard preoperative demographic characteristics and laboratory data and postoperative laboratory data. The present study is unique because we analyzed factors affecting the prognosis of patients with ALF who underwent LT and the endpoint of the study was either survival in the early postoperative period (first 90 days) or mortality.

In our present study, the mortality rate among patients with ALF due to indeterminate causes (ie, cryptogenic causes) was 64.7%. Patients with transplant due to hepatitis A showed the best prognosis versus other groups. This has been emphasized previously by other studies.^21–25 Fujiwara and associates²⁶ stated that most indeterminate causes of ALF show histologic characteristics that are compatible with autoimmune hepatitis. Furthermore, they also stated that indeterminate causes of ALF resulted in poor prognosis.²⁶ We hypothesize that the indeterminate causes actually involve idiosyncratic drug- or chemical-induced liver injury, which tend to cause slowly progressing liver failure that obscures the relationship with exposure to certain chemicals. Drug- or chemical-induced liver injury actually involves 2 mechanisms. The intrinsic mechanism is dose dependent, and the cause of the injury can always be identified; the best example of this type of injury is acetaminophen overdose, which is mainly observed in western countries. However, idiosyncratic drug reactions are dose independent, and biotransformation and immune mechanisms such as haptenization play major roles in the injury process.²⁷ With idiosyncratic drug-induced liver injury, it is almost never possible to pinpoint the chemical exposure. In addition, the immunologic mechanisms that underlie the pathophysiology are resistant to supportive therapies. In an report from Somasekar and colleagues²⁸ that analyzed 204 patients with ALF, 8 cases were due to infection from previously unrecognized viral pathogens such as herpes simplex, Epstein-Barr virus, cytomegalovirus, and parvovirus 19. The group concluded that uncommon viral causes should be investigated in patients with indeterminate ALF.

In our present study, patients with grade 4 encephalopathy had a higher mortality rate than those with grade 2/3 encephalopathy, although this did not reach significance, perhaps due to our low patient number. However, mortality was significantly higher in female patients (75% vs 25%). This may be because of the greater number of female patients in our study. It may also be due to higher incidences of postoperative cerebral edema and neurologic dysfunction observed in female versus male patients (19.5% vs 6.8%). In fact, the mortality rate in patients with postoperative cerebral edema and neurologic dysfunction was 100%. This result was also emphasized by Bernal and colleagues,²³ who reported that neurologic complications such as edema and intracranial hypertension were significantly higher in female patients (64.9%) with ALF. Ostapowicz and colleagues²⁹ found that women were more frequent in any of the etiologic categories of ALF (overall, 73% of patients were women). Although they and other researchers could not clearly explain this observation, they hypothesized that women may be more prone to development of ALF and stated that there was a high rate of prescribed or over-the-counter drug use among women.^21,29 Females have been found to be susceptible to traumatic brain injury, with sex steroids implicated in this phenomenon; however, results are conflicting because estrogen and progesterone are known to be neuroprotective factors.³⁰ Overall, female sex seems to be a risk factor for mortality in ALF.

We could not find any preoperative difference regarding serum ammonia levels among survivors and nonsurvivors in our study. This may be related to the use of liver assist support systems in the ICU, as we are a tertiary referral center for liver disease. However, postoperative ammonia levels of patients who did not survive were higher than those shown in patients who survived. Ammonia is a major substrate that is metabolized in the liver, and its accumulation in the serum is responsible for central nervous system complications observed in ALF. Associations between patient survival and serum ammonia levels have been previously emphasized.31,32 In our study, we observed that postoperative ammonia levels are an important factor to predict graft function and therefore patient survival. High postoperative ammonia may be the main reason for central nervous system complications observed in our patients and also may indicate primary nonfunction of liver grafts.

In our present study, we found that bilirubin was significantly higher in patients in our nonsurvivor group. In fact, once bilirubin levels reach above 9 mg/dL, the mortality rate increased by approximately 4 times. This observation was also shown by Farmer and associates³³ who found a significant impact of bilirubin over 5.65 mg/dL on reduced graft survival, although no effect was shown on patient survival. Bernal and associates²³ have shown that bilirubin levels were higher in patients who did not survive compared with those who survived (13.2 vs 12.3 mg/dL). Although the study by Bernal and associates²³ has been the only one that has evaluated the prognostic factors for patients undergoing liver transplant, they did not show total bilirubin as an independent risk factor for determination of patient survival. In our study, preoperative bilirubin levels were higher in our nonsurvivor than in our survivor group. A probable explanation for this is that bilirubin reflects the disease process before LT and that patients with elevated levels are more critically ill. This idea is also supported by the work of Farmer and associates, who also stated that bilirubin was a marker of disease severity.³³ Posttransplant bilirubin seems to reflect the function of the graft. Bolondi and associates³⁴ stated that bilirubin showed the functional activity of the liver in the postoperative period, which reflected both donor-related factors and recipient condition posttransplant. Because we observed that our survivor group had significantly lower bilirubin levels than the nonsurvivor group, we may assume that graft function in survivors was better than in nonsurvivors. Hence, bilirubin can be a surrogate marker for follow-up of transplant success.

In our present study, postoperative creatinine (but not preoperative) levels also determined patient survival. In our opinion, creatinine reflects the functional reserve of the kidneys and any deterioration in renal function that is observed by elevated creatinine means that the patient is experiencing multiorgan failure. Jin and colleagues³⁵ considered creatinine and vasopressor use to be factors reflecting multiorgan failure. They also stated that ALF creates a state of vasodilatation that is a factor potentiating circulatory and renal insufficiency.

Sodium is already used as a component of certain disease severity and in physiologic assessment scores, including in APACHE-II score (acute physiology and chronic health evaluation-II) and POSSUM score (patient operative severity scores for enumeration of mortality and morbidity). Hyponatremia has been defined as a marker for neurohumoral activation.36 Dawwas and colleagues³⁷ showed that preoperative hyponatremia was associated with in-hospital patient mortality and that hypernatremia was associated with poor survival in the first 3 years after LT. Furthermore, they stated that the overall complication rate was increased in patients with increased sodium levels.³⁷ Leise and associates³⁸ showed that pretransplant hypernatremia was associated with increased in-hospital mortality. Their study of patients with end-stage liver disease undergoing LT concluded that splanchnic vasodilation and hypoperfusion of kidneys were responsible factors for increased hyponatremia.³⁸ Therefore, from this point of view, a well-functioning graft would reverse these factors and would aid in the correction of hyponatremia. On the other hand, hypernatremia involves dysfunction of multiple systems and is an indicator of the quality of care.³⁹ Totsuka and associates⁴⁰ showed that donor hypernatremia was more frequently associated with primary graft failure. They explained this phenomenon as increased osmolarity due to hypernatremia leading to hepatocellular damage.⁴⁰ In our present study, preoperative sodium levels were not significantly different between the survivor versus nonsurvivor groups, although postoperative sodium levels were higher in nonsurvivors (144.9 ± 5.3 vs 142.1 ± 4.4 mEq/L). Furthermore, because 5 of our marginal DDLT grafts were due to hypernatremia, we checked serum sodium levels in LDLT and DDLT recipients and observed that they were similar. Because the postoperative values in both groups were within normal limits, we believe that the impact of sodium in patients with ALF is not as prominent as in patients with end-stage liver disease undergoing elective operations.

Chen and associates,⁴¹ in their study of hemodynamics after LT in pediatric patients, found that systolic and diastolic blood pressure increased after LT. Furthermore, they showed that potassium levels were slightly increased at postoperative day 3 and then normalized. Although hyponatremia and hypokalemia are the most frequent electrolyte imbalances after LT, they did not observe these phenomena. Potassium levels during LT can increase due to low cardiac output, metabolic acidosis, and type of preservation solution. The University of Wisconsin solution has by far the highest concentration of potassium (129 mEq/L), which could lead to hyperkalemia after graft perfusion.⁴² Furthermore, prolonged anesthesia, low-serum albumin, and blood transfusions can cause hyperkalemia.⁴³ Mikolasevic and associates⁴⁴ showed an increased mortality rate with pretransplant hyperkalemia. Nevertheless, our preoperative and postoperative serum potassium levels did not change significantly among survivors and nonsurvivors. We observed higher preoperative (but not postoperative) potassium levels in individuals who had DDLT than in those who had LDLT. As far as we know, this is the first study emphasizing this clinical phenomenon. We believe it may be due to the severity of hepatic necrosis and the interval between diagnosis and LT. Borderline hyperkalemia observed in patients scheduled for DDLT may cause an increase in postoperative cardiac events and may be responsible for the increased early postoperative death rate observed in these patients. Nevertheless, we do not have any data to backup such a hypothesis. Further research is needed to support our findings and hypothesis.

We found lower postoperative serum phosphate levels in patients who survived after LT. We also found lower preoperative phosphate levels in patients who had LDLT versus those who had DDLT. Studies regarding the predictive value of phosphorus levels are starting to accumulate. These studies have emphasized the importance of hypophosphatemia in predicting patient survival.^32,45,46 The regenerating liver is metabolically active, requiring high-energy phosphate substrates for proliferation, DNA synthesis, and other metabolic events. Acute consumption of phosphorus for the metabolic process causes an influx of phosphorus to the intracellular space, causing hypophosphatemia.⁴⁵ Therefore, patients with ALF who have preoperative hypophosphatemia are considered to have a regenerative effort and thus a better prognosis.⁴⁶ From this perspective, the postoperative hypophosphatemia that we observed in surviving patients may be attributed to the regeneration effort and function of the graft. Patients with LDLT undergo LT earlier and therefore have better preserved liver functions than their DDLT counterparts. For this reason, preoperative hyperphosphatemia in patients undergoing DDLT may reflect the severe liver failure or hepatic necrosis due to longer wait periods until an organ becomes available. The role of phosphate as a prognostic indicator after LT needs further research. When we consider our data regarding phosphate and ammonia, we may conclude that early mortality in patients may be due to primary nonfunction. Furthermore, postoperative serum AST levels of nonsurvivors and patients who received DDLT were higher than their counterparts. This supports the primary nonfunction that resulted in early postoperative mortality and also the increased mortality observed in our DDLT group, who usually have higher ALT and AST levels than those who receive living-donor grafts due to the long cold ischemia times. Robertson and associates⁴⁷ found that serum AST levels at postoperative day 3 predicted graft and patient survival. They postulated that peak AST levels suggested primary graft nonfunction.⁴⁷ Similarly, we observed that patients with DDLT had higher AST and ALT levels. Shortages of deceased donors can lead to longer wait periods, forcing surgeons to use marginal grafts, thus resulting in a higher primary nonfunction rate in these patients.

In our center, we mainly perform LDLT procedures for end-stage liver disease and for ALF because there is a shortage of deceased organ donations in our country. Because patients with ALF have poor prognosis, immediate care is required. In our opinion, LDLT provides such an emergency approach in the setting of ALF. Urrunaga and associates,⁴⁸ in a study of 21 patients who underwent LDLT for ALF, reported results comparable to DDLT in terms of 5-year graft and patient survival rates. Furthermore, other researchers have emphasized the efficacy of LDLT in the management of ALF in children and adults.^49,50 We observed better survival rates among our LDLT group, suggesting that patients with ALF should undergo LDLT if a living donor is available.

The retrospective nature of our study is the main limitation. We used rigorous selection criteria that limited the number of patients. This led to increased variance in the data collected. Therefore, a prospective study could be used to validate our study results, especially for the Turkish population. Although parameters were different in our survivor versus nonsurvivor group, our results did not have a predictive value. However, it is important to have an idea regarding the prognosis of patients to prevent evolving complications. Furthermore, surgeons should maintain interactions with the family of patients to inform regarding any dismal or other prognosis. Furthermore, we have provided some valuable information regarding the early course of patients after LT.

We found that LDLT was superior to DDLT in patients with ALF. The reason for this is that most of our DDLT recipients have had a longer wait period and grafts may be from marginal donors. We also found that bilirubin and grade 4 encephalopathy had predictive values for poor prognosis in our study patients.

EPUB Before Print

FULL TEXT