O11 Prevalence and detection of acute pancreatitis in acute liver failure and clinical consequences after super-urgent liver transplantation

<h3>Introduction</h3> Acute pancreatitis (AP) contraindicates liver transplantation (LT), but pre-LT identification can be difficult in acute liver failure (ALF). Diagnostic criteria in patients with ALF and the consequences for those who proceed to LT are not well-defined. We aimed to describe AP prevalence, the relationship between biomarkers and intra-operative findings and the consequences for those who undergo LT. <h3>Methods</h3> Consecutive patients admitted for ALF to a specialist critical care unit who underwent super-urgent LT (1/2013–12/2023) or did not undergo LT (1/2018–1/2022) were included. In LT recipients, those with an intra-operative diagnosis of AP were matched by age, sex and disease aetiology to ALF LT recipients without AP (1:2 ratio). Univariate and multivariate analyses were undertaken, with missing data excluded. <h3>Results</h3> 227 patients admitted with ALF were identified (120 transplanted, 107 non-transplanted). Median age was 37 years; 65% were female. Aetiologies included paracetamol (37%), seronegative (16%) and drug-induced liver injury (13%). Hyperamylasaemia above 100 IU/L was present in 49% (92/189) and above 300 IU/L in 15% (28/189); a radiological diagnosis of AP was made on CT in 5% (10/191). AP was found intra-operatively in 8% of those transplanted (9/120). Serum amylase was significantly higher in those with radiological AP but not intra-operative AP (figure 1). CT and serum amylase over 300 IU/L had low sensitivity for intra-operative AP (11%;22% respectively), but high specificity (96%;89% respectively). Patients with intra-operative findings of severe AP (necrosis or saponification, 5% [6/120]) were more often men (67%, p=0.03). Age, ALF aetiology and clinical presentation were non-significant. Comparing intra-operative AP cases to matched controls, white cell count on admission was higher (median 17.0;12.3, p=0.03) with increased organ dysfunction at admission (median SOFA score 16;7, p=0.01) and at time of LT (SOFA 19;15, p=0.001). AP cases had increased rates of re-laparotomy (p=0.037), tracheostomy (p=0.037), days on mechanical ventilation (p=0.008), days requiring renal replacement therapy (p=0.003), ITU length of stay (p&lt;0.001), ITU readmissions (p=0.003) and hospital length of stay (p&lt;0.001). There was no significant impact on early allograft function, survival to hospital discharge or 1-year patient and graft survival. <h3>Conclusion</h3> Hyperamylasaemia is common in patients with ALF but standard diagnostic cut-offs for AP perform poorly. Radiology and biochemistry do not correlate well with intra-operative pancreas status, confounding the decision-making around super-urgent wait-listing in ALF patients. Marked increases in morbidity and resource usage follow an unexpected diagnosis of AP in patients transplanted for ALF, but without significant impact on survival.