Introduction

An aberrant left hepatic artery (ALHA) arising from the left gastric artery (LGA) is a common anomaly of the LGA and hepatic artery that is occasionally encountered during gastric surgery [1,2,3]. ALHA is classified as either a replaced left hepatic artery (RLHA), a substitute for the normal LHA, or as an accessory LHA in addition to the normal LHA [3]. In gastric surgery, LGA is usually divided at the root level to ensure lymph node dissection of station number 7 of the Japanese Classification of Gastric Carcinoma (JCGC) [4]. However, this procedure may cause hepatic ischemia in RLHA patients, as it might be the only artery supplying blood to the left hepatic lobe. In several case reports, serious postoperative complications, including abscess formation, cholangitis, and left hepatic lobe necrosis, occurred during RLHA division [5,6,7,8]. Contrarily, several studies showed that ALHA division, including RLHA division, did not result in severe hepatic ischemia but only elevated the liver enzymes [9, 10]. However, none of these studies divided the RLHA or accessory LHA using preoperative computed tomography (CT) images. Furthermore, whether radical lymph node dissection can be performed with RLHA preservation in gastric surgery is unclear. Therefore, there is no consensus of the treatment strategy for RLHA during gastric surgery.

To identify the ALHA preoperatively, multidetector-row CT (MDCT) with three-dimensional CT angiography (3DCTA) is useful. Recently, several studies have reported that the number of hepatic vascular arterial anomaly detected by 3DCTA was comparable to that of classical studies using cadaver or conventional angiography [11,12,13,14]. Thus, MDCT with 3DCTA is thought to be useful for detecting ALHA and its subtypes preoperatively.

Laparoscopic distal gastrectomy (LDG) for clinical stage I gastric cancer is reportedly safe as open distal gastrectomy, and the non-inferiority of LDG to open distal gastrectomy in terms of relapse-free survival has been demonstrated [15, 16]. Generally, compared to open surgery, laparoscopic surgery has the disadvantage of limited exposure and visualization of the entire surgical field and limited tactile sensation of the surgeon while holding laparoscopic forceps. Thus, in laparoscopic surgery, the narrow view and limited palpation of arterial pulsation make intraoperative ALHA confirmation more difficult compared to open surgery. Moreover, it is impossible to identify whether an ALHA is either a replaced or an accessory type only through visual information during laparoscopic surgery, similar to open surgery. Therefore, in laparoscopic surgery, recognizing the existence of RLHA on preoperative MDCT images is necessary.

We investigated the short-term outcomes of gastric cancer patients with RLHA who underwent LDG to evaluate the efficacy of RLHA preservation.

Material and methods

Patients

Altogether, 513 primary gastric cancer patients who underwent LDG at our institute from January 2012 to December 2018 were included. Among them, the surgical records of 106 patients (20.7%) with ALHA were retrieved; of these, after a retrospective review of preoperative MDCT according to this study’s definition and determination of the presence of RLHA or accessory-type ALHA, those who underwent combined resection of organs besides the gallbladder and spleen and were converted to open laparotomy and those with accessory-type ALHA were excluded. Finally, 55 RLHA patients were evaluated (Fig. 1). This study was approved by the Institutional Review Board of our hospital (J2019-42–2019-1–3). Informed consent was waived by the ethical committee owing to the retrospective nature of the study.

Fig. 1
figure 1

Flowchart of the study population

Full size image

Retrospective detection method of ALHA types and RLHA diameter measurement.

ALHA types were accessory (with arterial communication to the LHA in the extra-hepatic area) and replaced (RLHA) (without LHA from the proper hepatic artery and without apparent arterial communications between the right and left hepatic lobes in the extra-hepatic area) types (Fig. 2).

Fig. 2
figure 2

Schematic diagram of the subtypes of an aberrant left hepatic artery. RLHA, replaced left hepatic artery; LGA, left gastric artery; GDA, gastroduodenal artery; SA, splenic artery; CHA, common hepatic artery; PHA, proper hepatic artery; MHA, middle hepatic artery; RHA, right hepatic artery

Full size image

Imaging techniques are shown in supplementary material 1. RLHA was identified using MDCT with 3DCTA (Fig. 3a). RLHA’s intraluminal diameter was measured using axial slices at the lesser omentum before running into the liver (Fig. 3b).

Fig. 3
figure 3

Preoperative classification and measurement of the RLHA diameter on CT images a RLHA did not originate from the PHA in the extra-hepatic area on 3DCTA. b The intraluminal diameter of the RLHA was measured between the white arrow heads using axial slices at the lesser omentum before running into the liver. RLHA, replaced left hepatic artery; CT, computed tomography; PHA, proper hepatic artery; 3DCTA, three-dimensional computed tomography angiography; LGA, left gastric artery; RGA, right gastric artery; SA, splenic artery; CHA, common hepatic artery; RHA, right hepatic artery; RGEA, right gastro epiploic artery.

Full size image

Surgical procedures

LDG was performed using 5 trocar methods. The Nathanson liver retractor was used for liver retraction, and mini laparotomy of < 4 cm at the umbilical lesion was performed for specimen retrieval and anastomosis. Extent of lymph node dissection was based on the Japanese gastric cancer treatment guidelines [17]. Reconstruction methods, either Billroth I or Roux-en Y were determined by the surgeon’s preference, patient’s conditions such as the presence or absence of hiatal hernia and remnant stomach size. The surgical technique of RLHA preservation was performed as follows (Supplementary Video). After lymph node dissection of the greater curvature side, the duodenum was transected distally to the pylorus ring, and the supra-pyloric lymph node was dissected. Subsequently, the lesser omentum was divided from the cranial aspect of the right gastric artery to the esophagogastric junction with confirmation of the presence of RLHA (Fig. 4a). If necessary, the RLHA was taped. Next, the supra-pancreatic lymph nodes at station numbers 8a and 11p of the JCGC were dissected. During this procedure, the left and right sides of the celiac artery were sufficiently exfoliated (Fig. 4b). Additionally, after exposing LGA root, the surrounding tissues were cut open on the right side of the artery from the root toward the RLHA running into the liver to completely expose the artery (Fig. 4c). Arterial branches arising from the LGA to the stomach were identified and divided using clips (Fig. 4d). Routinely, 2 or 3 arterial branches to the stomach were identified. Finally, the lymph nodes of lesser curvature side of the stomach were retrieved, and lymph node dissection with RLHA preservation was completed (Fig. 4e).

Fig. 4
figure 4

Practical techniques of RLHA preservation during laparoscopic gastrectomy. a Identifying an RLHA in the lesser omentum, b exfoliating the left and right sides of the celiac artery and LGA, c cutting open the right side of the artery from the root of the LGA toward the RLHA running into the liver (black dotted line), d confirming the arterial branch to the stomach, e complete view after the preservation procedure of the RLHA. RLHA, replaced left hepatic artery; LGA, left gastric artery; SA, splenic artery; CHA, common hepatic artery

Full size image

In patients diagnosed with RLHA preoperatively, the RLHA was basically preserved. However, RLHA division was performed in patients with hepatic artery communications from the proper hepatic artery to the left medial and/or lateral segment according to the Healey and Schroy classification on preoperative MDCT images [18], or without hepatic ischemia by performing temporary clamping of the RLHA.

Analyzed factors

Clinicopathological data, including age, sex, the Eastern Cooperative Oncology Group performance status, body mass index, comorbidities, internal RLHA diameter, tumor location, and clinical and pathological TNM classifications according to the JCGC were collected for all patients. Preoperative confirmation rates of ALHA and RLHA were also collected from the medical records. Surgical outcomes included operative time, blood loss, extent of lymph node dissection, number of harvested lymph nodes as a whole and at each station around the LGA according to the JCGC, length of postoperative hospitalization, and postoperative complications according to the Clavien-Dindo classification (CD) [19]. Liver function test results, including alanine aminotransferase (ALT), aspartate aminotransferase (AST), and bilirubin levels, preoperatively and on postoperative days 1 and 3 were collected and graded according to the National Cancer Institute Common Terminology Criteria for Adverse Events (CTCAE) version 4.0. The incidence of CD grade II or higher postoperative complications during the postoperative hospitalization was investigated. Hepatic infarction was diagnosed by radiologists when enhanced MDCT images showed an area of non-contrasted hepatic parenchyma. Long-term outcomes were also evaluated.

Statistical analysis

All statistical analyses were performed using R version 3.4 software package (R Foundation for Statistical Computing, Vienna, Austria). Continuous data are presented as median (range), and continuous variables were nonparametrically analyzed using the Mann–Whitney test. Categorical variables were compared using Fisher’s exact test. All variables with p-values < 0.15 in the univariate analyses were included and analyzed in the multivariate logistic regression analysis. P-values < 0.05 were considered statistically significant.

Results

Study subject

In 55 patients diagnosed with RLHA, 37 had RLHA preservation and 18 had RLHA division. The reasons for RLHA division were preoperative misdiagnosis of RLHA as accessory-type ALHA (n = 10, 55.6%), intrahepatic communication (n = 3, 16.7%), blood supply confirmation after temporary clamping (n = 3, 16.7%), suspected direct invasion of the lymph node to the artery (n = 1, 5.6%), and injury during operation (n = 1, 5.6%). Short-term and long-term outcomes were compared between the two groups.

Patients’ characteristics

Preoperative confirmation rate of RLHA was significantly lower in the divided group. The RLHA diameter was significantly larger in the preserved group. The other clinical findings were not significantly different between the two groups (Table 1).

Table 1 Patients’ characteristics
Full size table

Short-term outcomes and pathological findings

Operative time, blood loss, postoperative hospitalization, and number of harvested lymph nodes were not significantly different between the two groups (Table 2).

Table 2 Surgical outcomes and pathological findings
Full size table

Perioperative blood examinations

ALT and AST levels on postoperative days 1 and 3 as well as postoperative ALT and AST elevation grades according to CTCAE were significantly higher in the divided group (Fig. 5 and Table 3).

Fig. 5
figure 5

Sequential changes in perioperative blood examinations. a ALT, b AST, and c bilirubin. ALT, alanine aminotransferase; AST, aspartate aminotransferase; PRE, preoperative; POD, postoperative day. *Including two patients with very high ALT and AST liver enzyme levels (> 1000 IU/L)

Full size image
Table 3 Perioperative hematological examinations and postoperative CTCAE grading
Full size table

Postoperative complications

Although the incidence of postoperative bleeding was not significantly different between the two groups, the incidence of hepatic infarction in the medial and/or lateral segment (Fig. 6) was significantly higher in the divided group (n = 3; 16.7% vs. n = 0; 0%, p = 0.031) (Table 4). Postoperative hepatic transaminase values of patients with hepatic infarction in the divided group are shown in Supplementary material 2.

Fig. 6
figure 6

Enhanced-computed tomography images of hepatic infarction in the lateral segment after RLHA division. A large low-contrast area in the lateral segment is detected on both the (a) arterial and (b) venous phases. RLHA, replaced left hepatic artery

Full size image
Table 4 Postoperative complications according to the Clavien-Dindo classification (grade II or worse)
Full size table

Univariate and multivariate analyses for postoperative AST/ALT elevation

Univariate analysis showed that RLHA division was associated with postoperative ALT and/or AST increase of grade 3 or higher according to CTCAE. Multivariate analysis showed that RLHA division was identified as an independent risk factor for postoperative ALT and/or AST elevation (Table 5).

Table 5 Univariate and multivariate analyses of risk factors for the postoperative ALT and/or AST elevation (CTCAE grade 3 or worse)
Full size table

Survival and recurrence outcomes

The median follow-up period for all enrolled patients was 29.9 months (range, 5–72; interquartile range, 15–51). During this period, in the preserved group, one patient died of other diseases and another had bone metastasis relapse. There was no local recurrence around the LGA in both groups.

Discussion

The present study demonstrated that RLHA division during LDG is significantly associated with the elevation of ALT and AST levels on postoperative days 1 and 3. Moreover, in the divided group, three patients (16.5%) developed hepatic infarction of the lateral segment during postoperative hospitalization. RLHA division was demonstrated as an independent risk factor for postoperative transaminase increase of grade 3 or worse according to CTCAE.

Previously, the intraoperative identification rate of ALHA arising from the LGA has been reported to be 6%–22% during gastric surgery, and RLHA accounts for 45% of all ALHA cases from studies using cadaver [1,2,3]. For RLHA patients, the usual procedure of lymph node removal along the LGA during gastric cancer surgery includes RLHA division and blood flow disruption to the left lobe. Moreover, RLHA division results in hepatic infarction, which causes serious complications, including cholangitis and liver abscess [5,6,7,8]. However, ALHA division, including RLHA division, does not cause severe hepatic ischemia but only an elevation of liver enzymes [9, 10]. Despite this issue, no studies have evaluated the clinical impact of RLHA division among RLHA patients. Another concern of RLHA division is an oncological issue. For the RLHA preservation, the tissues surrounding the LGA and RLHA should be divided, and these vessels must be exposed. This additional procedure may increase the blood loss amount, operative time, and risk of inadequate lymph node dissection. The oncological safety of the procedure involving RLHA preservation, in terms of the accuracy of lymph node dissection around the LGA, has been scarcely reported. Therefore, there is still no consensus on whether to preserve or divide the RLHA during gastric cancer surgery.

Few reports have clearly demonstrated ALHA subtypes based on preoperative MDCT images. Recently, the number of hepatic vascular arterial anomalies detected using 3DCTA was comparable to that of the classical autopsy series or conventional angiography, suggesting that preoperative evaluation using 3DCTA before gastrectomy can help in recognizing ALHA [11,12,13,14]. In this study, RLHA detection rate in the divided group was lower, which might be due to the preoperative misdiagnosis of RLHA as an accessory-type ALHA by preoperative MDCT image review. Conversely, 36 out of 44 patients diagnosed with RLHA by preoperative MDCT had RLHA preservation in our series. Therefore, surgeons should carefully find RLHA when using preoperative MDCT with 3DCTA images.

LDG has become the widely used standard treatment for early gastric cancer [15, 16]. Although RLHA preservation requires additional surgical procedures, several studies have demonstrated that surgical techniques for ALHA preservation do not affect blood loss, operative time, or number of harvested lymph nodes compared with the division techniques [20, 21]. Similarly, our study demonstrated no significant difference in blood loss, operative time, and number of retrieved lymph nodes around the LGA between the two groups. This result could be attributed to the arterial anatomy around the LGA that could be routinely confirmed using preoperative MDCT images, and the sharing of information of the anomaly among operators and assistant surgeons facilitated the surgery. Therefore, this surgical technique of RLHA preservation could be feasible in LDG and enable exfoliation of surrounding tissue and dissection of gastric arterial branches around the RLHA without taping several arteries owing to the magnifying effect of laparoscopic surgery.

Several studies have reported that ALHA division resulted in transient transaminase elevation without hepatic infarction [10, 20]. Generally, hepatic infarction can be caused by hepatic arterial interruption after hepatic artery division until the arterial flow is reconstituted. After hepatic artery ligation between the intrahepatic and adjacent arteries, collateralization through the inferior phrenic, pancreaticoduodenal, and intrahepatic arteries has been reported in angiographic examinations [22, 23]. In previous studies, ALHA subtypes could not be considered and classified. Contrarily, our study assessed only RLHA patients and demonstrated that RLHA division led to higher incidence of partial left hepatic lobe infarction. Therefore, when dividing the RLHA, the collateral arterial communications are occasionally insufficient to supply the arterial flow to the RLHA’s dominant area.

Several studies have reported the association between increased postoperative transaminase levels and immune system. In liver surgery, peak postoperative transaminase levels were associated with morbidity and mortality [24]. The liver plays an important role in innate immunity; its functions include reserving fixed-tissue macrophages, including Kupffer cells, and regulating the synthesis of proteins responsible for pathogen recognition and opsonization [25]. Previous studies have reported that innate immunity impairment due to liver dysfunction after liver surgery was associated with both changes in innate immune response and postoperative infection [26,27,28]. Meanwhile, we demonstrated that RLHA division was an independent risk factor for postoperative transaminase elevation, suggesting the possible presence of innate immune impairment in patients with RLHA division. Thus, we suggest that RLHA preservation should be performed owing to its comparable short-term outcomes and minimal postoperative hepatic transaminase elevation associated with innate immune impairment, unless the tumor directly invades the RLHA in LDG.

The present study has several limitations. First, this is a retrospective analysis of a small number of patients from a single institute. Furthermore, we did not perform a detailed imaging study on RLHA using 3DCTA, and the impact of CT-defined RLHA division on the postoperative hepatic arterial collateralization and transaminase elevation is not fully understood. Second, this study did not adequately demonstrate its oncological safety in terms of long-term survival because of the short follow-up period. However, as the number of harvested lymph nodes did not differ between the divided and preserved groups, the RLHA-preserving procedure is assumed to improve the survival outcomes by reducing postoperative complications [29].

Conclusions

The RLHA preserving procedure reduced the incidence of postoperative transaminase elevation and hepatic infarction in patients undergoing LDG. Surgeons should pay close attention to the RLHA using preoperative MDCT images and preserve the RLHA to prevent hepatic damage.