Takahashi et al. found that liver function is closely related to blood flow [12]. It is important to ensure sufficient blood supply to the liver, but the patency of blood outflow from the liver is also very important. The major hepatic veins are often infiltrated by intrahepatic cholangiocarcinoma for the anatomical and pathological characteristics. If extensive liver resection for R0 resection results in liver blood outflow obstructed which leads to residual liver congestion, liver function is likely to diminish. So it is necessary to evaluate the area of hepatic vein preoperatively to decide whether and how to reconstruct the infiltrated hepatic vein.
The area of congested liver can be confirmed during surgery. Murata et al. found that when hepatic veins were occluded, the hepatic artery was the sole vessel for supplying blood to the liver, and the portal vein drained the congested portion of the liver [13]. Therefore, clamping a hepatic vein prior to resection and clamping the hepatic artery at the same time during surgery reveals the boundary of congested on the surface of the liver surface within 5 min [14]. The non-congested liver remnant volume can also be revealed during surgery with indocyanine green fluorescence imaging and hepatic vein clamping technique [15]. But this technique does not work sometimes. We have only one chance to use it during surgery [15]. When the hepatic veins are incompletely clamped, we cannot get the correct results. So, more liver remnant volume analysis methods are required, such as preoperative three-dimensional image reconstruction [14, 16]. The non-congested liver remnant volume can be determined by analyzing the volume of blood that each hepatic vein and its branches drains from each segment. With these methods, we can perform surgery safely by retaining enough liver volume which was normal on blood inflow and outflow.
Based on a study of living donor liver transplants, Sano et al. devised a criterion for hepatic vein reconstruction: the relevant volume of the congested liver should be subtracted from the remnant liver or the graft liver volume. When the remaining liver volume is less than 30% of the standard liver volume in normal liver resection or less than 40% in liver transplantation, reconstruction of the hepatic vein or its tributaries should be considered [14]. Based on a study of liver tumor resection, Mise et al. devised a criterion for hepatic vein reconstruction: venous reconstruction is recommended in patients with non-congested liver remnant volume smaller than 40% of total liver volume (TLV) when ICGR15 is less than 10% or non-congested liver remnant volume smaller than 50% of total liver volume when ICGR15 is 10–20% [16]. In some cases, liver volume increased because of continuous obstructive jaundice. So, for safety, we set 40% of both SLV and TLV as the cut-off value to reconstruct hepatic vein or its tributaries in liver tumor resection operation. When future non-congested liver remnant volume is smaller than 40% of SLV or TLV, hepatic vein reconstruction is needed to ensure blood outflow from the liver.
In Case 1, if the remnant liver is drained only by the right hepatic vein without reconstruction of the middle hepatic vein, the non-congested liver remnant volume is 630 mL, accounting for 37.8% of the total liver volume and 39.6% of the standard liver volume (Fig. 5). Diminished liver function can likely occur, so middle hepatic vein reconstruction must be factored into preoperative analysis. Blood outflow from the liver was ascertained postoperatively using ultrasound (Fig. 7), and liver function was normal. In Case 2, the volume drained by the right hepatic vein was 378 mL, accounting for 39.5% of the total liver volume and 36.4% of the standard liver volume (Fig. 9), so middle hepatic vein reconstruction was needed according to preoperative analysis. In case 2, we also confirmed the congested liver volume during surgery. The common trunk of the left and middle hepatic vein was clamped and then the right hepatic artery was clamped, revealing an area of congestion (Fig. 12) that was drained by the middle hepatic vein and that coincided with the area indicated by preoperative analysis. Hepatic vein reconstruction involving preoperative and intraoperative evaluation was required in both cases described in this report.
The vessel grafts used for hepatic veins reconstruction include autologous saphenous veins [1, 2], the internal jugular veins [3], the ovarian veins[4,5,6], the umbilical veins [6,7,8], artificial blood vessels [9], and allogeneic iliac vessels [10, 11]. When using a saphenous vein, the internal jugular vein, or the ovarian vein, a normal vessel needs to be harvested from the patient, thus increasing the number of surgical sites. An artificial blood vessel initially allows reconstruction of the hepatic vein, but it is less effective later on in comparison to a harvested vessel [9, 17]. Allogeneic blood vessels take time to acquire and cryopreserve. In the cases reported here, the vessels on the resected side of the liver were used as grafts to reconstruct the hepatic vein. The vessels used included the portal vein and a hepatic vein. The availability of blood vessels was assessed preoperatively, and the vascular grafts were harvested intraoperatively.
In Case 1, the tumor had not infiltrated the left hepatic vein and the tumor was sufficiently distant from the bifurcation of the left hepatic and middle hepatic veins, so the left hepatic vein was used to reconstruct the middle hepatic vein. Makuuchi et al. first reported a technique in which the left hepatic vein flap was used to repair the wall of the middle hepatic vein after the left hepatic vein and part of the lateral wall of the middle hepatic vein were removed because the left and middle hepatic veins had been infiltrated by a tumor [18]. The approach used in the current case differed slightly. Because the MHV was infiltrated by tumor, but the left hepatic vein was not infiltrated, so about 3 cm of the MHV trunk was resected from its root, and the trunk of the left hepatic vein was used to repair this part of MHV. In Case 2, the tumor had infiltrated the middle and left hepatic veins but was distant from the portal vein, so the left branch of the portal vein was used as a vessel graft to reconstruct the middle hepatic vein (Fig. 10). Ikegami et al. [19] and Junrungsee et al. [20] had reported MHV reconstruction with recipient's explanted portal vein in living donor liver transplantation. In this case, we reconstructed MHV with the trunk of portal vein in the resected liver itself.
In intrahepatic cholangiocarcinoma, the middle hepatic vein is usually infiltrated by a tumor at its central part. Therefore, the middle hepatic vein was first dissected from its root to the portion infiltrated by tumor and then dissected from its distal end to the tumor-infiltrated portion; this facilitates determining the length of middle hepatic vein trunk infiltrated by tumor and the specific location of the tumor, which is also good for bleeding control when dissociating liver parenchyma.
This report describes a method of harvesting a vascular graft from the portion of the liver that will be excised. Advantages of this approach include less damage than when using self-saphenous veins and self-internal jugular veins, highly similar characters than other grafts, obviates the need to wait for an allogeneic blood vessel. A limitation of this approach is that available blood vessels in the liver need to be assessed before removal.
With the development of surgical techniques, more extreme liver resection can be performed safely, especially for tumor which infiltrates key vessels. There are different vessel grafts with special advantages. We report two cases where vessel grafts from resected portion of liver were used, and find there are more advantages with good result. In future, more cases are needed to verify its effect.