LAUSR

To the Editor: We read with great interest the article by Lu and colleagues on total hilar en bloc resection for perihilar cholangiocarcinoma (PHC) with a left-sided predominance, published in the Journal of Gastrointestinal Surgery. The study showed that eight patients with advanced PHC underwent left hemihepatectomy with simultaneous resection and reconstruction of portal vein and hepatic artery, using a novel approach termed the Bno-touch technique.^ The concept of no-touch technique for surgical management of PHC was first introduced in 1999 by Neuhaus et al., involving right trisectionectomy with routine portal reconstruction. Subsequently, a similar procedure, combining en bloc portal vein resection with right hemihepatectomy, was described by Hirano et al. The oncologic rationale is that this technique avoids dissection of the portal bifurcation and the right hepatic artery on the dorsal side of the tumor, thereby obviating spillage of malignant cells. Moreover, several anatomic advantages favor the right-sided hepatectomy as follows: (1) the left-sided hilar plate is approximately 10 mm longer than the right-sided hilar plate; (2) there is a distinct demarcation of the left caudate lobe (the Spiegel lobe), and; (3) the proper and left hepatic arteries run along the left-most portion of the hepatoduodenal ligament, and are rarely involved. Subsequently, skeletonization resection of the hepatic hilum along the left border is farthest away from the tumor. Although conceptually interesting, several issues regarding the surgical anatomy and technique applied to a left-sided resection remain to be resolved in Lu and colleagues’ study. First, combined left-sided hepatectomy and vascular resection is more complicated than combined right-sided hepatectomy and vascular resection, as anatomic variations of the right lobe are common. Isolation of segmental hepatic arteries using a transhepatic approach is technically demanding and at increased risk of arterial injury, especially with a supraportal right posterior hepatic artery. Additionally, hepatic hilum dissection described by Lu and colleagues was in close proximity to the biliary confluence. In general, portal vein reconstruction following hilar en bloc resection for the left-sided tumor requires an interposition graft because of a short branch. However, the authors reported on end-to-end anastomosis in all venous reconstructions after segmental resection. Thus, this is technically not feasible without compromise of notouch principles. Another issue is that left-sided hepatectomy bears the risk of tumor dissemination at parenchymal transaction because the caudate lobe is readily involved by direct tumor infiltration, and there is no anatomic landmark between the caudate lobe and the right liver. Taking these anatomic features into account, the no-touch concept may no longer be valid for left-dominant tumor, apart from technical concerns and postoperative morbidity associated with routine hepatic artery reconstruction. As for the survival associated with the no-touch technique, Neuhaus and colleagues reported better long-term outcomes as compared with conventional procedure. Nevertheless, these results remain contentious and have yet not to be validated by other centers. Furthermore, the survival benefit from concomitant portal vein resection is validated by numerous studies while the role of simultaneous hepatic artery resection for PHC remains controversial. Accordingly, routine vascular resection is not recommended by consensus guidelines. Remarkably, negative longitudinal and vertical margins are needed for a curative resection while positive radial margin is a powerful predictor of local recurrence and dismal prognosis. However, a multicenter survey from France revealed that the ductal margin was generally assessed in PHC whereas only 10% of the pathology reports specified the radial margins. Pathologically, perineural and lymphangitic spread of PHC * Xiangcheng Li [email protected]