Since the introduction of laparoscopy for liver resection in the 1990s, the number of laparoscopic liver resections (LLR) has steadily increased.1-4 The use of this surgical technique has recently been expanded, with further applications to living-donor liver resection for transplantation.5 Despite this wide application, there is controversy regarding the use of LLR for tumors in areas (segments 1, 4a, 7, and 8) which are difficult to resect laparoscopically.6

For tumors located in the posterosuperior (PS) segment of the liver, the laparoscopic approach is challenging and risky owing to limited visualization and difficulty in controlling bleeding.7 Nevertheless, in recent years, LLR of the PS segment has been increasingly performed owing to the development of 3-D scope and surgical instruments.8,9 Furthermore, several studies have indicated that LLR has superior outcomes than does open liver resection (OLR).10

However, to the best of our knowledge, there are no data on the actual use of laparoscopy in liver resection. This study therefore aimed to investigate the extent to which laparoscopy is performed in liver resection, and to determine whether surgeons prefer laparoscopy and laparotomy when operating on tumors in the PS segment. In addition, the rate and causes of open conversion (OC) in LLR were investigated.

We conducted a retrospective observational study of patients who underwent liver resection at the Samsung Medical Center between January 2020 and December 2021. The exclusion criteria were as follows: 1) age ≤18 years, 2) living donor liver resection, and 3) non-mass-forming liver resection. This study was conducted in adherence with Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines (Supplementary Table 1).

Data on sex, body mass index, American Society of Anesthesiologists (ASA) score, liver function test results, previous surgical history, and postoperative course were collected from medical records. Tumor data regarding histologic subtype, surgical margin, maximal size, multifocality, and location were investigated through pathology records. The resection extent, cause of OC, time of OC decision, estimated blood loss, operative time, and transfusion rate were investigated through operation records. Operative techniques for LLR have been described previously.11 Data on hospital stay was collected including admission days before the actual operation. Usually, the patient is admitted to the hospital 2 days before the operation.

Proximity to the major vessels was evaluated using preoperative computed tomography, and the distance was calculated from the margin of the main tumor lesion to the major vessel (main or second branches of Glisson’s tree, major hepatic vein, and inferior vena cava).

Liver function was evaluated using the Child-Pugh classification.12 The prior surgical history, including the number and type of surgery, was reviewed and classified into four types (hepatectomy, upper gastrointestinal surgery, lower gastrointestinal surgery, genitourinary surgery). Patients who underwent various types of surgery were prioritized and classified in the following order: hepatectomy, upper gastrointestinal surgery, lower gastrointestinal surgery, and genitourinary surgery.

The tumor locations were classified into two groups (PS and anterolateral [AL]). The PS segment was defined as segments 1, 4a, 7, and 8, and the AL segment was defined as the remaining segments (2, 3, 4b, 5, 6).8 Patients with multiple tumor locations were classified as having PS segments when at least one tumor was located in the PS segment.

The extent of resection was classified into four groups: subsegmentectomy, segmentectomy, bisegmentectomy, and hemihepatectomy. Multiple wedge resections were classified as subsegmentectomy.

The selection criteria for the laparoscopic approach were surgeon-dependent. When considering LLR, the tumor location, history of portal vein embolization, and trisectionectomy were not considered. The indications for OLR included a tumor size >10 cm, except when the tumor was a pedunculated type, reconstruction of a vascular or biliary conduit was required, the tumor was close to an important vital structure making difficult to dissect laparoscopically, the tumor had invaded adjacent organs necessitating concomitant resection and reconstruction, future remnant liver was <25%, and Child–Pugh classification was class B. One surgeon performed laparoscopic surgery, while the remaining three surgeons used both approaches. The selection criteria for LLR were described in our previous article.13

Among 1,471 patients, 1,095 were eligible for inclusion after excluding living liver donors (n=375) and non-massforming liver lesions (immunoglobulin G4-related sclerosing cholangitis, n=1). There were 229 patients with OLR and 866 patients with LLR, and LLR accounted for 79% of total liver resections. A flow diagram showing the selection process of patients included in this study is given in Fig. 1.

With the advancement of surgical techniques, liver resection has become a safe procedure for liver tumors when performed by experienced surgeons in patients with adequate indications.14 In particular, LLR is considered a standard method because of the various advantages of postoperative outcomes in patients with tumors in areas such as the anterolateral segment, including S2, 3, 4b, 5, and 6, which are easily accessible by laparoscopy.6

In contrast to the AL location of most liver tumors, PS tumors are located deep below the right diaphragm and are surrounded by the rib cage. This induces several disadvantages for surgery, including poor visibility and difficult bleeding control during laparoscopic surgery. Thus, over the past decade, LLR of PS location tumors has been considered a challenging and dangerous procedure, and a scoring system has been developed to evaluate it.7,15 However, in recent years, the number of LLRs performed has rapidly increased, and with the development of flexible scopes and various surgical instruments, laparoscopy for tumors located in the PS is being actively applied.16-18 The aim of this study was to investigate the extent to which laparoscopy is actually performed in liver resection, and to determine whether surgeons prefer laparoscopy and laparotomy in the PS segment.

Laparoscopic liver resection was performed in 79% of patients. In the comparison between the OLR and LLR groups, previous hepatectomy, maximum tumor size, and wide surgical extent were significantly different. This suggests that surgeons prefer OLR if patients have a previous history of surgery in the liver, large tumors, or wide resection areas.

PS location was not an important factor in determining surgical procedures in the whole-group analysis. However, in the subgroup analysis of patients with a PS location, differences in the preference of surgeons according to the surgical method were identified. For example, surgeons preferred OLR over LLR in patients who needed to undergo hemihepatectomy or more extended resection with a large tumor in the PS location.

In our study, the incidence rate of OC during LLR (1.6%) was lower than that reported in previous studies (4.4-21.9%).3,19,20 The reason for this difference is that previous studies included patients with early LLR, and the development of technology also influenced the results. In the details of OC, the most common cause was adhesion. In contrast, previous studies have reported that bleeding is the major cause of OC.20 Although the incidence was low, adhesions have also been reported as a minor cause of OC.21-23 In addition, another feature of our results was that all the OC cases had a tumor located in the PS. This indicates that PS location may be a risk factor for OC, although this needs to be analyzed in future studies. The location is also important in cases with adhesions. The anterolateral location is covered by the anterior and right lateral abdominal walls, and gastrointestinal organs, including the greater omentum. These are covered with the peritoneum, and adhesiolysis can be performed relatively easily. However, the posterolateral location was covered with the diaphragm. Adhesiolysis of the posterolateral part can be complicated by bleeding from both the diaphragm and the liver, and opening of the diaphragm can occur.

This study has several limitations which should be discussed. Firstly, this was a single-center retrospective study, which limits the generalizability of the results. Additionally, the lack of details regarding tumor characteristics, such as proximity to major vessels or exact segment tumor location, made it impossible to evaluate the difficulty score. In future studies, research including the difficulty score index should be conducted. The OLR and OC rates may be higher in data from other centers; because of the low number of OC in our study, we did not perform multivariate analysis to analyze the risk factors for OLR or OC. However, these data could be a good guide not only for surgeons, but also for oncologists, when deciding on the treatment plan for patients with liver malignancy.

To the best of our knowledge, this is the first study to investigate the actual preference of surgeons for surgical techniques when performing liver resection. Although this study was limited by complications inherent to the single-center retrospective design, the results may nevertheless be helpful to surgeons who are inexperienced with few people to turn to for advice when deciding to perform liver resection.