In favor of the Jarnagin-Blumgart classification.

e would like to congratulate Ito et al for their excellent study that emphasized the important topic of extrahepatic bile duct resec- tion combined with hepatectomy in patients with hilar cholangiocarcinoma.1 We would also like to highlight some aspects of their report and relate it to our own experience. On the basis of our own clinical in- vestigation we fully agree that right or left sided hepatectomy associated with extrahe- patic bile duct resection represent the only surgical procedure that can achieve a com- plete cure with improved survival.2 The Au- thors described an excellent outcome in pa- tients treated with major hepatic resection combined with en-bloc extrahepatic bile duct resection, that were associated with higher R0 resection rate (P 0.0006) and improved disease specific survival and disease free survival (P 0.0005). Moreover it is extremely important, and for this the authors deserve sincere con- gratulations, the demonstration that with op- timal surgical management, a decreased in- cidence of local recurrence can be obtained in patients who undergo bile duct resection combined to hemihepatectomy (P 0.031). Ito et al emphasize the ability to obtain an R0 resection and improved disease specific and disease free survival in patients who undergo concomitant major liver resection. The conclusions reported by Ito et al, about the need for “routine� combined hep- atectomy with bile duct resection are con- vincing. This procedure can be performed safely and is more likely to be associated with histologically negative resection mar- gin. We therefore strongly agree with Ito et al while we are concerned about the curative role of the local hilar resection without hep- atectomy. We believe that no patients can be effectively cured with R0 resection by this limited procedure. This is due to the strong propensity of this highly locally invasive tumor to infiltrate the surrounding tissues, vascular structures and proximal bile duct along the perineural perivascular spaces and lymphatics. For these reasons it is inevitable that residual tumor frequently remains when local resection for hilar bile duct is not asso- ciated to major hepatic resection. However to draw the conclusions above, there are major issues to be discussed. The first issue concerns the appropriate definition of microscopic margin involvement that in the study by Ito et al was done “gener- ally� and not “routinely� by frozen section