Because hilar cholangiocarcinomas may send metastases to the caudate lobe via minute bile duct branches [1], total caudate lobectomy, along with hepatic resection, is necessary [23]. Although the importance of the caudate lobe is being increasingly accepted and a copious amount of research carried out, the anatomy of the caudate lobe lacks consensus.

Couinaud [4] suggested a clinically useful anatomy of the liver, but the concept of caudate lobe anatomy has been changing since then. In 1954, Couinaud defined the caudate lobe as segment I according to its portal ramification. This portion of the liver corresponds to the Spiegel lobe in the Kumon classification, which is currently used worldwide. Kumon classified the caudate lobe into 3 parts: Spiegel's lobe, the paracaval portion, and the caudate process portion [5]. Many researchers have agreed with the Kumon classification, and it is being used clinically.

For the right margin of the caudate lobe, Kogure et al. [6] advocated that the hepatic vein is the boundary between the caudate process and the right liver, but it is more important to find the right margin of the paracaval portion when searching for the right border of the caudate lobe. The paracaval portion is located inside the liver parenchyma, in neighboring segments 7 and 8, making it difficult to identify the margin. Kitagawa, of the Murakami group, initially thought that the right hepatic vein (RHV) was the most important landmark of the right margin of the paracaval portion. Later, however, he accepted the concept of the right paracaval plane [7]. Recently, Maki et al. [8] found that the paracaval vein of a branch of the RHV is the boundary between the caudate lobe and segments 7 and 8.

The ventral margin of the caudate lobe was believed to correspond to the route of the middle hepatic vein (MHV). Couinaud [9] opposed this idea, however, and thought that the margin was situated in the vicinity of the confluence of the middle and RHV drainage area into the inferior vena cava (IVC). Later, Kwon et al. [10] reported the existence of the precaudate plane.

The current study focuses on identifying the right and ventral margins of the caudate lobe, which is poorly recognized in the surgical field. So far, there are no reports on the right and ventral margins of the caudate lobe that can be applied clinically. As caudate anatomy has been analyzed by Synapse 3D software (Fujifilm, Tokyo, Japan) to visualize the 3-dimensional (3D) anatomy of the caudate lobe, we set out to delineate the right and ventral margins of the caudate lobe preoperatively using the same system.

The characteristics of the caudate lobe in 94 cases are shown in Table 1. For the right margin, 17 subjects (18%) exceeded the RHV-IVC plane by a mean of 10.2 mm (range, 2.4–27.2 mm). Among them, 11 were <10 mm, 3 were 10–20 mm, and 3 were >20 mm (Fig. 3A, B). For the ventral margin, 28 subjects (30%) exceeded the MHV-RHV plane by a mean of 17.4 mm (range, 1.2–49.1 mm). Among them, 12 were <10 mm, 5 were 10–20 mm, and 11 were >20 mm (Fig. 3C, D). In all, 10 subjects were found to exceed both the MHV-RHV and RHV-IVC planes.

Hilar cholangiocarcinoma is known to invade the parenchyma and bile duct of the caudate lobe rapidly during early stages of the disease. In the literature, microscopic tumor involvement in caudate lobe branches was found in 42% of patients at the time of surgery [1]. Thus, caudate anatomy is of great importance when performing total caudate lobectomy in patients with hilar cholangiocarcinoma. Prognosis reportedly improved in cholangiocarcinoma patients when caudate lobectomy was performed [12]. In cases of caudate lobe tumor, isolated caudate lobectomy can be performed, as introduced by Lerut et al. [13] in 1990. Surgery included partial or total caudate lobectomy. Partial caudate lobectomy was performed by resecting Spiegel's lobe or the caudate process only, each of which was easily identified intraoperatively [14]. In other previous research, total isolated caudate lobectomy was performed by approaching the caudate lobe from the right side of the liver, using the right posterior Glisson pedicle as the anterior border plane of the caudate lobe and reaching the dorsal plane by transecting the parenchyma between the MHV and RHV. There was no mention of the right margin of the caudate lobe [1516]. Although it is known that the anatomy of the caudate lobe is important when performing total caudate lobectomy, the exact locations of the right and ventral margins remain obscure.

It is difficult for us to define the right margin of the caudate lobe of the liver because the paracaval portion is located anterior to the IVC, to the right of the ligamentum venosum, and cranial to the hilar plate. The right and ventral margins were not identifiable from the exterior as their locations vary [17]. Couinaud [18] suggested that the right portion of the caudate lobe was a neighbor of segments 7 and 8, lying adjacent to the end point of the RHV. Kitagawa et al. [7] firstly thought that the RHV is the most important landmark of the right margin of the paracaval portion. Later, however, they suggested the concept of the right paracaval plane. These investigators suggested that the right paracaval plane may serve as an imaginary right margin of the caudate lobe. In fact, the right margin of the paracaval portion was located on the left side of the right paracaval plane in 65.4% (36 of 55) of cases. Therefore, if hepatic resection was carried out in this plane, total caudate lobectomy was complete with 60% anatomical reliability; if carried out at 10 mm to the right of this plane, 80% reliability was guaranteed; and if performed at 30 mm to the right of this plane, there was 100% reliability. This report was a milestone in the search for a definition of the right margin of the caudate lobe—which is poorly visualized from the exterior. Nevertheless, the concept had little effect on identifying the right margin during surgery. Recently, Maki et al. [8] identified the paracaval vein as the landmark for establishing the boundary between the caudate lobe and segments 7 and 8. The paracaval vein, however, cannot be found by intraoperative ultrasonography. In the current study, 17 cases (18%) exceeded the RHV-IVC plane. As landmarks, the RHV and IVC are easily found in the surgical field as well as with intraoperative ultrasonography.

The need to identify the individualized right margin of the caudate lobe is indisputable. The use of Synapse 3D is more likely to enable surgeons to perform total caudate lobectomy based on each patient's unique liver anatomy.

Regarding the ventral margin of the caudate lobe, in 2002 Kwon et al. [10] reported the existence of a precaudate plane. This plane, which they described as flat or slightly curved, was a continuum of the ligamentum venosum and the ventral margin of the hilar plate. According to this report, the ventral margin of the caudate lobe was restricted within the dorsal side of the precaudate line in 63.2% (48 of 76) of the liver specimens. In 36.8% of the cases where ventral extension of the caudate lobe was found, the ventral margin was extended to the dome-like area under the terminals of the RHV, MHV, and the left hepatic vein. The authors suggested that the precaudate plane could serve as an imaginary ventral margin of the caudate lobe, which was a milestone in the definition of the ventral margin of the caudate lobe. However, it is an illusory plane that could not be used as the foundation on which to base surgical decisions. In the current study, we used the MHV-RHV plane to facilitate identifying the ventral margin of the caudate lobe. The RHV and MHV could be easily identified during surgery because their locations were relatively fixed, allowing them to serve as landmarks. Traditionally, the caudate lobe was thought to be confined within the RHV and MHV. Based on this belief, surgeons performed total caudate lobectomy. Our results showed, however, that 28 cases (30%) exceeded the MHV-RHV plane. Kumon [5] noted that the paracaval portion reached the liver surface underneath the diaphragm in 53% of corrosion liver casts. Kwon et al. [10] found the same results in 50% of cadaveric studies, and Maki et al. [8] reported the same results in 30% of images of living patients. In this type of caudate lobe, it is impossible to perform total caudate lobectomy from the dorsal side of the liver.

Although our research has limitations—e.g., that small branches (<1-mm diameter) could not be easily identified on CT images—we learned that preoperative analysis of the caudate lobe using Synapse 3D is necessary to evaluate the anatomy of the caudate lobe. Synapse 3D provides the margins of the caudate lobe that 2-dimensional CT cannot.

In conclusion, regarding the right margin, 18% exceeded the RHV-IVC plane, whereas 30% exceeded the MHV-RHV plane for the ventral margin. Evaluating the anatomy of the caudate lobe using Synapse 3D preoperatively could be helpful in achieving more precise anatomical resection of the caudate lobe.