A73 SURVIVAL PREDICTORS IN PATIENTS UNDERGOING LIVER TRANSPLANT FOR NON-ALCOHOLIC STEATOHEPATITIS: A POPULATION-BASED COHORT STUDY

Non-alcoholic steatohepatitis is a recognized cause of cirrhosis affecting 12% of word population. The proportion of orthotopic liver transplants (OLT) for NASH cirrhosis increased 8-fold (1.2 to 9.7%) between 2001–2009 and is anticipated to become the leading indication for OLT in the next 20 years. Surprisingly, disease-specific indications and mortality predictors for OLT in NASH are lacking. To develop a 5-years mortality prediction model applicable during the pre-surgical assessment of patients with NASH cirrhosis eligible for OLT. This single center retrospective cohort study included subjects undergoing OLT at the University of Alberta between 2002–2012 for NASH/cryptogenic cirrhosis. Clinical information was extracted from a dedicated computerized database (OTTR) and audited. The primary outcome was all-cause mortality at 5 years. Prediction models were constructed using Cox proportional hazard regression techniques. Model assumptions and discrimination capacity were tested. Ethics approval was obtained from the local ethics board. Of 524 OLT patients, NASH cirrhosis was the main indication in 41 (8%) patients (NASH-OLT). Compared to those receiving an OLT for other indications, the NASH-OLT cohort had comparable follow-up times (2.7 vs. 3.2y. p=0.4), 5yrs-mortality rates (11/41; 27% vs. 106/483; 22%, p=0.4) and MELD scores (18 vs. 20, p=0.1); a greater BMI (28 vs. 25 p=0.0006), higher prevalence of diabetes (39 vs. 20%, p=0.004) and renal insufficiency (32 vs. 16%, p=0.004), donors with lower Donor Risk Indexes (1.37 vs. 1.49, p=0.05) longer hospital stays (34 vs. 30d, p=0.01) and received two times more blood products during transplant (p<0.03). Previously published mortality estimation tools had poor discriminatory capacity in the NASH-OLT cohort: Charlson index for OLT [C-index = 0.54, p=0.6], United Network for Organ Sharing (UNOS) [C-index = 0.58, p=0.4] and Cardoso 2014 [C-index = 0.43, p=0.9]. We developed a 5yr mortality prediction model that included: extreme (<600 or >1200) modified-BMI (BMI kg/m2 * albumin g/L) [coef=1.62; p=0.02], pre-operative INR [coef=-2.76; p=0.02) and receiving a non-local transplant [coef=2.49; p=0.02). This new model had better mortality prediction capacity than any previously reported model (X2= 16.4, C-index=0.84, p=0.0008). Previously validated survival predictors for OLT perform very poorly in the NASH-OLT population. In this small cohort, 3 simple preoperative parameters (modified BMI, INR and receiving a non-local transplant) can be used to predict 5-yrs mortality in NASH-OLT recipients with excellent discriminatory capacity. Further validation of the proposed model needs to be done in a larger cohort. CAGAlberta Innovates Health Solutions (AIHS)

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.