Surgical Anatomy of the left liver: Considerations relevant to the surgery of Klatskin’s tumours, split and living-donor liver transplantation

AbstractThe surgical approach to hilar tumours is still not perfectly codified, despite some very eminent contributions in the surgical literature, such as the works by Prof. Y. Nimura. The aim of the presentation is to illustrate some anatomical details that can simplify the operation and can favour both a wider tumour clearance and a no-touch technique. In this abstract the anatomy will be described as it is encountered during the operation.The left hilar plate. As described (but poorly represented) in Couinaud’s seminal work of 1957, “Le Foie : Études anatomiques et chirurgicales” the hilar plate is part of a tube that has to be opened to expose (and understand) the plate fully. This tube is composed on its parenchymal side by the hilar plate, a thickening of the connective tissue in which the bile ducts are embedded. The peritoneal side of this tube is composed by the peritoneum of Glisson’s capsule covering the left portal pedicle and the round ligament.Once the peritoneum of is opened, generally at the level of the umbilical portion of the portal vein, one can find two separate planes in the fat of the round ligament/umbilical pedicle: a superficial plane where the branches of the hepatic artery lie, and a deeper plane lying on the portal vein. The plane on the portal vein is the one that has to be followed to free the umbilical and transverse portions of the left portal vein, by cutting all P4 branches and P1 branches. The access is made easier by cutting the A4 artery that runs in the more superficial plane. This is generally possible as the artery giving A3 and A2 runs in the left of the hepatic pedicle, and once A4 is cut, A3 and A2 can be displaced to the left, exposing the portal vein completely. Once all P4 and P1 branches are cut, the last structure that has to be cut to allow full mobilization of the left portal vein is the portal attachment of Arantius’ ligament.Arantius’ ligament. This ligament runs from the elbow of the left portal vein and the axilla between the left and the middle hepatic vein. The ligament is the remnant of Arantius’ duct that in foetal life channelled the umbilical blood from the umbilical portal vein into the inferior vena cava and the right atrium. Cutting of the ligament on the portal side increases markedly the freedom of the left portal vein, and is the key to its mobilisation. Cutting the ligament on the hepatic vein side, with a long stump, allows to pull on the ligament and to tease open a passage between the left and the middle hepatic vein to perform a hanging manoeuvre to lift the left lobe in plane of the resection that is appropriate for the parenchymotomy. The middle part of the ligament can be left on the left lobe (e.g. for a split procedure, to favour the vascularization of the left lobe) or on S1 (to increase tumour clearance in a Klatskin’s tumour).The transection of the hilar plate. Once all the above have been done, in Klatskin’s tumours the hilar plate can be transected on the left of the left portal vein, generally exposing two ducts. In LDLT or split procedures the hilar plate can be cut on the right of the portal vein, or even on the right of the arrival of B4 to obtain a wider and thicker duct for reconstruction. Indeed carrying out a full umbilical plate dissection is not necessary in LDLT and splits if one aims to obtain an anastomosis at the junction of B4, B3 and B2 (i.e. in the left duct). The disadvantage with the intraparenchymal dissection is, probably, a higher incidence of S4 necrosis, which is theoretically less of a problem if one leaves the hilar plate of B4 intact as it can be done in the true plane of the umbilical plate. The other disadvantage of the intraparenchymal plane is the potential injury of a right duct inserting in the left duct, which can be avoided by careful probing of the junctions through the main bile duct in split liver procedures (a manoeuvre that is well advised also to detect the path of the B3 duct, at risk in the umbilical plane). Regardless of the dissection plane that is chosen, practice has proven that Couinaud was wrong in describing a split procedure leaving S4 in place as a scandal: S4 necrosis is seldom a clinically relevant problem.