A16 Abstracts / Digestive and Liver Disease 41 (2009) A1–A19 PREDICTIVE RISK FACTORS OF RECURRENCE OF HEPATOCELLULAR CARCINOMA AFTER LIVER TRANSPLANTATION F. Gentili a , Q. Lai b , M. Giusto a , G. Mennini b , P.B. Berloco b , M. Rossi b , A. De Santis a , S.Ginanni Corradini a , A.F. Attili a , M. Merli a a II Division of Gastroenterology. La Sapienza University of Rome, Italy b Department of General Surgery and Organ Transplanta- tion. La Sapienza University of Rome, Italy Introduction. Liver transplantation (LT) is considered the optimal treatment for hepatocellular carcinoma (HCC). Since LT based on the Milan criteria (MC) has been shown to provide good disease-free survival, to excellent long-term survival and a low incidence of recurrence. We have ana- lyzed individual tumor characteristics applying the new score “up-to-seven criteria” [1]. Aim. We analyzed our experience with LT for HCC to iden- tify predictive risk factors of HCC recurrence. Outcomes were compared for patients with HCC recurrence and patients recurrence free after LT. Patients and methods. We retrospectively studied 109 con- secutive HCC patients transplanted from January 1988 to December 2007. We excluded all the patients who died from factors other than tumour recurrence within the first year (n = 24). Clinical data, eligibility to MC and San Francisco (UCSF) criteria were considered. Results. 85 patients were enrolled, 19 with recurrence (group A) and 66 without recurrence (group B). All patients in Group A deceased for HCC recurrence, with a disease-free survival mean follow-up of 12.3 ± 7.2 months and an overall-survival mean follow-up of 23.9 ± 19.8 Table 1 Characteristics of the patients with (Group A) and without (Group B) HCC recurrence. Group A (n = 19) Group B (n = 66) P Demographic data Age (year) 51.9 ± 11.8 55.1 ± 7.3 NS Gender (M/F) 12/7 57/9 0.05 * Clinical data No cirrhosis (n) 6 3 0.003 * HEV status (n) 1 14 NS HCV status (n) 9 39 NS Cirrhosis other causes (n) 4 13 NS Tumor data Single HCC (n) 7 33 NS Number of lesions 2.3 ± 1.3 1.7 ± 1 0.035 * Up-to 7 lesions (%) 68 12 <0.0001 * Microvascular invasion (%) 79 13 <0.0001 * G1-2 (%) 47 65 NS G3-4 (%) 52 18 0.006 * Exceeding MC (%) 74 18 <0.0001 * Exceeding UCSF (%) 68 9 <0.0001 * * P < 0.05 Chi-square test and Student’s t test. months. Demographic, clinical and tumour characteristics of groups A and B are shown in Table 1. Significant variables were included in multivariate logistic regres- sion analysis. Microvascular invasion (Wald = 2.663;95% [CI] = 1.751–4.935; P = 0.007) and exceeding of UCSF cri- teria (Wald = 2.912;95% [CI] = 3.182–372.853; P = 0.003) were selected as independent predictors of HCC recurrence. Conclusion. Microvascular invasion and exceeding of UCSF criteria are independent risk factors and highly specific pre- dictors of HCC recurrence (specificity > 86%). Reference [1] Mazzaferro V, Llovet JM, Miceli R, et al. Lancet Oncol 2008. doi:10.1016/j.dld.2009.02.040 THE ETHICAL EQUIPOISE IN LIVING AND DECEASED DONOR LIVER TRANSPLANTATION: TOWARDS DECISION PROCESSES BASED ON MATHEMATICAL MODELS A. Vitale a,b , M. Volk c , E. Gringeri a , M. Valmasoni a , P. Burra d , P. Angeli e , U. Cillo a a Unità di Chirurgia Epatobiliare e Trapianto Epatico, Azienda Ospedaliera-Università di Padova, Italy b Istituto Oncologico Veneto, IRCCS, Padova, Italy c Division of Gastroenterology, University of Michigan Health System, Ann Arbor, USA d Divisione di Gastroenterologia, Università di Padova, Italy e Clinica Medica V, Università di Padova, Italy Background. The decision process allocating a specific organ from a cadaveric or a living donor to a particular liver transplantation (LT) recipient is strongly influenced by ethical issues. Aims. (a) To effectively represent the potential equipoise achievable between the different ethical principles involved in LT. (b) To construct a mathematical decision model able to objectify and quantify these ethical aspects. Methods. The desirable LT ethical equipoise may be described by a triangle with the transplant benefit (life expectancy with LT minus that without LT) at its superior apex and, the potential harms to the waiting list and to the living donor at inferior apices. We then constructed a Markov model to objectify and quantify the ethical equipoise triangle. The data sources to construct and validate the model were: the online UNOS web-site, and a prospective database from Padua about a new allocation model. Results. Although our Centre was characterized by a higher proportion of HCC patients in the WL (25% versus 10%) and a lower proportion of high MELD score (>20) non-HCC patients (17% versus 27%) than the average US centre, these proportion were similar among transplanted patients.