Background
Gastric cancer is the fifth most common malignancy worldwide. It is one of the major causes of mortality in Central Asia, Eastern Europe, and Eastern/Southeastern Asia, including Japan [
1]. In the era of multidisciplinary therapy with advanced chemotherapy and immunotherapy for advanced diseases, gastrectomy with lymphadenectomy plays important roles in the treatment strategies for early and advanced gastric cancer [
2]. Laparoscopic gastrectomy has been gradually accepted in recent years because of its advantages over open surgery, including minimal invasiveness, less pain, and lower rate of overall complications, as well as its non-inferiority to open gastrectomy for survival [
3‐
7]. Although laparoscopic gastrectomy is minimally invasive and enables precise manipulation with enlarged visual field, some technical limitations apply for more advanced surgical procedures or for the treatment of locally advanced cancers. As the indication of laparoscopic gastrectomy has gradually increased in recent years, the importance of preoperative evaluation of perigastric vascular anatomy has also increased [
8]. However, comprehensive analyses of perigastric vascular anatomy have not been well conducted.
Recent advances in three-dimensional (3D) computed tomography have enabled the examination of vascular anatomy without percutaneous catheter angiography. Previous studies reported the usefulness of 3D angiographic analysis of perigastric vessels [
9‐
15]. In contrast to conventional angiography, multidetector-row computed tomography (MDCT)-based 3D angiography enables angle-free observations to be performed. In our earlier report, we classified the ramification pattern of the right gastric artery (RGA) into distal, caudal, and proximal types. We also showed that RGA ramification points can be misdiagnosed under conventional angiographic anterior views because of the lack of 3D information [
15]. Therefore, perigastric vascular anatomy should be re-evaluated using 3D angiography in the era of laparoscopic surgery.
In this study, we aimed to comprehensively evaluate perigastric vascular anatomy using a 3D angiography reconstructed from enhanced MDCT data and to discuss the usefulness of preoperative 3D angiography from the perspective of laparoscopic gastrectomy.
Discussion
We evaluated the vascular anatomies in detail for the CA, LGV, LGA, LIPA, RGA, IPA, PGA, and LGEA using a 3D angiography reconstructed from enhanced MDCT data. To our knowledge, this study is the first to conduct a comprehensive analysis of perigastric vessel anatomies using a 3D angiography.
Laparoscopy-assisted distal gastrectomy was first reported in 1994. It has been widely performed for early gastric cancer in recent years in Asian countries including Japan [
8,
19,
20]. A series of randomized clinical trials confirmed the non-inferiority of laparoscopy-assisted distal gastrectomy to open distal gastrectomy in terms of adverse events, short-term clinical outcomes, and relapse-free and overall survivals for Stage I gastric cancer [
3‐
7]. The non-inferiority of laparoscopic distal gastrectomy (LDG) for advanced gastric cancers was also reported by several randomized clinical trials, which suggested that experienced surgeons can safely perform LDG with D2 lymphadenectomy for advanced gastric cancer [
21‐
24]. In contrast to the multicenter trials conducted in high-volume centers, retrospective cohort studies based on a Japanese nationwide registry database revealed a higher incidence of pancreatic fistula in LDG than in open distal gastrectomy, but wound infection and dehiscence were less common in the LDG group [
25,
26]. In general practice, LDG seems to be a feasible therapeutic alternative for gastric cancer, but further improvements in surgical quality are warranted. On the contrary, the indication of laparoscopic gastrectomy has gradually increased in Asian countries. Laparoscopic total gastrectomy (LTG) and proximal gastrectomy (LPG) are more commonly performed for early and advanced gastric or esophagogastric junction cancers, which require advanced surgical skills [
8,
27]. Recent advances in laparoscopic surgery and the development of robotic surgery have the potential to overcome the technical difficulties of performing LTG [
28]. However, preoperative simulation including the evaluation of perigastric vessel anatomies may help surgeons safely perform LDG and LTG for gastric cancers.
We have previously reported a new system of classifying RGA ramification patterns using a preoperative 3D angiography [
15]. Preoperative simulation of RGA ramification patterns is useful for performing LDG or LTG. However, more information of perigastric vessel anatomy is needed for the precise and safe LTG and LPG that have been increasingly performed in recent years [
8,
27]. Therefore, in this study, we comprehensively evaluated 8 perigastric vessel anatomies for the LGV, LGA with rLHA, LIPA, RGA, IPA, PGA, LGEA, and the omental branch of the LGEA from the viewpoint based on laparoscopic gastrectomy (Table
1). In our previous study, the image resolution was not satisfactory. In some cases, we were unable to trace the blood vessels to the organs such as the stomach wall. In the present study, we improved the protocols of 3D angiography; not only the RGA but also the perigastric vessels to each organ could be traced, and more accurate data were obtained.
Table 1
Anatomy of perigastric vessels in 127 patients who underwent MDCT followed by gastrectomy
Left gastric vein (LGV) | 127 (100%) | 122 (96.1%) | CHA, SA | 4 | Dorsal of CHA, 59 (46.5%); dorsal of SA, 8 (6.3%); ventral of CHA, 27 (21.3%); ventral of SA, 28 (22.0%) |
LGA with replaced LHA | 18 (14.2%) | 18 (14.2%) | LGA branch | 3 | One LGA branch, 7 (38.9%); two LGA branches, 9 (50.0%); three LGA branches, 2 (11.1%) |
Left inferior phrenic artery (LIPA) | 127 (100%) | 124 (97.6%) | Ao, CA, LGA | 6 | Ao-I, 40 (31.5%); CA-I, 52 (40.9%); LGA-I, 6 (4.7%); Ao-C, 10 (7.9%); CA-C, 15 (11.8%); LGA-C, 1 (0.8%) |
Right gastric artery (RGA) | 127 (100%) | 127 (100%) | PHA, GDA, CHA | 3 | Distal type, 93 (73.2%); caudal type, 23 (18.1%); proximal type, 11 (8.7%) |
Infra-pyloric artery (IPA) | 127 (100%) | 127 (100%) | ASPDA, RGEA, GDA | 3 | Distal type, 82 (64.6%); caudal type, 33 (26.0%); proximal type, 12 (9.4%) |
Posterior gastric artery (PGA) | 103 (81.1%) | 103 (81.1%) | Pancreatic tail end, SPA | 4 | Prox-I, 22 (31.1%); Prox-C, 39 (37.9%); Dist-I, 13 (12.6%); Dist-C, 19 (18.4%) |
Left gastroepiploic artery (LGEA) | 127 (100%) | 127 (100%) | SA, ITA | 2 | SA, 23 (18.1%); ITA, 104 (81.9%) |
Gastric branch of the LGEA proximal to the omental branch | 127 (100%) | 126 (99.2%) | SA, ITA, omental branch of the LGEA | 3 | No gastric branch, 46 (36.5%); one gastric branch, 62 (49.2%); two gastric branches, 18 (14.3%) |
Kawasaki et al. classified the LGV location into five types: (i) dorsal to the CHA, (ii) ventral to the CHA, (iii) ventral to the SA, (iv) dorsal to the SA, and (v) others using an MDCT without a 3D angiography [
29]. Based on this classification, Yuasa et al. examined the joining pattern of the LGV using a 3D CT angiography [
30]. Our results in the running aspects of the LGV relative to the CHA and SA were consistent with those of Kawasaki et al. Notably, of the 67 cases with LGVs running along the dorsal side of the arteries, 59 cases (88.1%) were running along the dorsal side to the CHA, all of which flowed into the PV. By contrast, the joining pattern of the LGV running along the ventral side to these arteries varied and did not depend on which artery the LGV was running to.
rLHA resection can cause serious liver damage or necrosis [
31,
32]. Because the rLHA does not communicate with the hepatic artery in the liver, it needs to be preserved in cases with a common trunk with the LGA. In 60% of cases with a common trunk between the rLHA and the LGA, multiple branches of LGAs require attention during gastrectomy.
Greig et al. examined the right inferior phrenic artery and the LIPA using 425 cadavers, and classified the origin of the inferior phrenic arteries into eight types [
33]. According to the Japanese Gastric Cancer Association, infra-diaphragmatic lymph nodes predominantly along the LIPA are categorized as no. 19 [
18]. Because the origin of the LIPA, which may branch from the LGA, is important than that of the right inferior phrenic artery in gastrectomy, we focused on the branching patterns of the LIPA. In this study, the LIPA ramified from the LGA in 5.5% of the cases.
The ramification patterns of the RGA were classified into three types in our previous study on 100 cases and re-evaluated in the present study on 127 cases [
15]. The two studies showed different branching rates of distal, caudal, and proximal types (68.8% vs. 72.4%, 16.9% vs. 19.5%, and 14.3% vs. 8.1%, respectively). In the present study, we were able to trace the RGAs to the stomach wall in all cases, therefore obtaining more precise data. Because caudally or proximally ramified RGAs may cause difficulty in dissecting supra-pancreatic lymph nodes, the confirmation of RGA branching rate and the high success rate of preoperative RGA visualization in this study seem to help us perform laparoscopic gastrectomy more safely.
Shinohara et al. reported laparoscopic techniques for dissection of no. 6 infra-pyloric lymph nodes and the anatomical importance of IPA [
34]. Accordingly, based on the intraoperative findings, Haruta et al. classified the origins of IPA into three types: distal (64.2%), caudal (23.1%), and proximal (12.7%) [
17]. In the present study, we re-evaluated the branching types of IPA using a 3D angiography: distal (64.6%), caudal (26.0%), and proximal (9.4%), which were consistent with a previous report based on intraoperative findings [
17].
Although numerous researchers reported the anatomy of PGA using cadavers, the definitions and names varied [
16,
35,
36]. In the present study, we defined the branch from the SA to the posterior side of the stomach as a PGA from the viewpoint of gastrectomy with lymphadenectomy. We then analyzed the PGA anatomy according to the Japanese classification of gastric carcinoma, which defined the middle from the origin of the SA to the pancreatic tail end as the boundary between nos. 11p and 11d [
18]. The PGA was clearly visualized using 3D angiography in 103 cases (81.1%). The PGA was present in the proximal region in 55.9% (n = 71) of the cases, and the PGA branched as a common trunk with the SPA in 39 cases (30.7%). Many of the proximal type PGAs were closely located at the boundary of nos. 11p and 11d, which could be used as a milestone for dissection of no. 11p lymph nodes.
The LGEA branching from around the splenic hilum is an artery with many anomalies. The branching rate from the SA itself was 26.0–36.7% in previous reports [
37], but it was 18.1% in the present study. Left greater curvature lymph nodes along the LGEA distal to its first gastric branch, and those along the first gastric branch of the LGEA are categorized as no. 4sb according to the Japanese Gastric Cancer Association [
18]. In laparoscopic gastrectomy for early gastric cancer, the omental branch of the LGEA may be preserved. In cases with no gastric branches between the roots of the LGEA and its omental branch (36.5%; Fig.
7c), no. 4sb lymphadenectomy can be performed with preserving the omental branch by dissecting the LGEA after the branching point of its omental branch. By contrast, in cases with gastric branches ramifying from the LGEA proximal to its omental branch (63.5%; Fig.
7d and e), one or two gastric branches need to be dissected to preserve the omental branch.
In this study, we sought to evaluate the running aspects of the SA relative to the SV using a 3D-CT angiography. In one case, the SA ran along the dorsal side to the SV, however, a larger number of cases would be needed to clarify the frequency of such rare cases.
This study has several limitations. First, the vessel anatomies identified using a 3D angiography were not completely validated during surgery. Second, the clinical values of the preoperative simulation of perigastric vessels for laparoscopic gastrectomy, such as shorter operation time and less complications, were not evaluated in clinical studies. However, our data showed detailed anatomical variations in perigastric vessels, understanding of which will likely help surgeons to perform laparoscopic surgery safely. In addition, this study revealed that a 3D angiography is potentially useful for precise visualization of small perigastric vessels that may be used for preoperative simulation as well as for education to young doctors and medical students.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.