Primary Versus Salvage Living Donor Liver Transplantation for Patients With Hepatocellular Carcinoma: Impact of Microvascular Invasion on Survival

Abstract Objective Salvage liver transplantation (LT) has been proposed for patients with a small hepatocellular carcinoma (HCC) and preserved liver function. Few reports have been issued on salvage LT in a living-donor (LD) LT setting. Therefore, we performed this study to evaluate differences in tumor invasiveness and other risk factors on survival after salvage versus primary LDLT. Methods Between September 1996 and December 2008, 324 patients with HCC underwent LT. We excluded 138 patient from the analysis, leaving 186 HCC patients for analysis, including 17 (9.1%) who had undergone earlier resection, the salvage LDLT cohort. The other 169 patients underwent primary LDLT. Results Intrahepatic metastasis, Edmonson-Steiner histologic grade, microscopic vascular invasion, and preoperative serum alpha-fetoprotein levels significantly influenced tumor recurrence. Microscopic vascular invasion, intrahepatic metastasis, Edmonson-Steiner histologic grade, and treatment by salvage LDLT were significantly associated with poor patient survival univariate analysis. However, only microscopic vascular invasion was significant on multivariate analysis. The treatment modality (primary or salvage LDLT) was not observed to affect overall or disease-free survival significantly on multivariate analysis. Disease-free survival was significantly better in the primary than in the salvage LDLT group. Furthermore, patients in the primary LDLT group tended to show better survival. However, when stratified by the presence or absence of microscopic vascular invasion, no significant group difference was found for overall or disease-free survival among those without versus with microscopic vascular invasion. Conclusions Five-year overall survival after primary versus salvage LDLT were similar when differences in tumor pathologic features, such as microscopic vascular invasion, were taken into account. Multivariate analysis showed that the treatment itself was not a significant prognostic factor for survival.