Several studies have suggested that laparoscopic liver resection (LLR) is associated with fewer postoperative complications than open liver resection (OLR) for hepatocellular carcinoma (HCC). However, this issue remains controversial since the data may have been attributable to an imbalance in patients’ background.
We retrospectively analyzed 290 hepatectomies for HCC undertaken between 2011 and 2019. Liver resection difficulty was based on the 3 levels of the Institut Mutualiste Montsouris classification. Resection ratio was calculated using computed tomography volumetry. Patient characteristics were compared between the LLR and OLR groups. Propensity score matching (PSM) was adopted to adjust the imbalance between the cohorts, and the incidence of postoperative complications was compared.
The difficulty and resection ratio were significantly lower in LLR (n = 112) than in OLR (n = 178) (difficulty grade I/II/III: 84/10/18 vs. 43/39/96, p < 0.001; resection ratio: 11.4 ± 12.7 vs. 22.7 ± 17.2%, p < 0.001). The incidence of postoperative complications (Clavien–Dindo grade III or more) was lower in LLR (2.7% vs. 21.9%, p < 0.001), which was mainly attributable to fewer incidences of ascites and pleural effusion. PSM generated 68 well-matched patients in each group. The lower incidence of postoperative complications in LLR was also maintained in the PSM cohort (2.9% vs. 16.2%, p = 0.017). On multivariate analysis, LLR was the independent predictor of postoperative complications (OR 0.184, 95% CI 0.051–0.672, p = 0.010).
The present study demonstrated that a laparoscopic approach reduces the incidence of postoperative complications in liver resection for HCC.
Hepatectomy is an effective treatment for hepatocellular carcinoma (HCC). Perioperative outcomes of hepatectomy for HCC have greatly improved due to progress in preoperative evaluation, surgical technique, and postoperative management. However, postoperative complications occur in some highly invasive or complicated surgeries such as major hepatectomies . Reducing the incidence of postoperative complications should always be the priority in liver surgery as they can adversely affect not only short-term outcomes but also long-term prognosis in hepatectomy [2, 3].
Laparoscopic surgery in the gastrointestinal tract has gained wide acceptance as a minimally invasive surgery [4, 5] and has also been introduced in hepatectomy. Some comparative studies of open and laparoscopic liver resection (LLR) reported that LLR is associated with improved perioperative outcomes such as reductions in intraoperative blood loss and postoperative complication rates, as well as shorter postoperative hospital stays [6,7,8]. However, most studies were retrospective studies that inevitably involved patient selection bias, and thus the results should be interpreted cautiously. Even in a study  using propensity score matching (PSM), the results may be due to an imbalance in unconsidered factors. Resection ratio and liver resection difficulty were not considered in many previous reports. These factors can have a major impact on the decision between laparoscopic and open surgery, as well as the development of postoperative complications. Therefore, it is plausible that in previous studies, an imbalance in these factors may have led to reductions in the incidence of postoperative complications in LLR. The purpose of this study was to clarify whether laparoscopic hepatectomy for HCC can contribute to reductions in the risk of postoperative complications using the PSM balancing resection ratio and the difficulty of the involved procedures.
We retrospectively investigated 344 cases of hepatectomy for HCC undertaken from 2011 to 2019. Among these, 54 cases were excluded for the following reasons: presence of macroscopic tumor thrombus (n = 26), invasion in the inferior vena cava (n = 3), necessity of vascular reconstruction (n = 5), portal vein embolization performed before surgery (n = 1), simultaneous surgery for other disease (n = 6), dissection of lymph nodes (n = 2), and combined resection of other organs (n = 15). Finally, 290 cases were analyzed. This research was approved by the ethics committee and registered in the UMIN Clinical Trials Registry.
All patients underwent blood tests including a routine liver function test. The presence of hepatitis virus was checked, and indocyanine green (ICG) clearance tests at 15 min and computed tomography (CT) were performed for preoperative assessment. Postoperative blood tests were routinely performed on 1, 2, 3, 5, and 7 days after hepatectomy. The liver volume was measured using CT volumetry with a volume analyzer system (Synapse Vincent, Fujifilm, Tokyo, Japan) [10, 11]. The liver resection ratio was defined as resection volume (mL)/total liver volume (mL). For anatomical resections, the resection volume was calculated using volumetry, whereas it was substituted by the weight of the resection for non-anatomical resections. The extent of three or more liver segments was defined as major, and excisions below that as minor.
Liver resection difficulty was defined by the Institut Mutualiste Montsouris (IMM) classification . The difficulty grade was defined in the 3 levels of difficulty depending on the location and extent of liver resection. Grade I: partial resection and left lateral segmentectomy, Grade II: anterolateral segmentectomy and left hepatectomy, Grade III; posterosuperior segmentectomy, right hepatectomy, extended left hepatectomy, right posterior sectionectomy, central hepatectomy and extended left hepatectomy. Anterolateral segments were defined as Couinaud segments 2, 3, 4b, 5, and 6, whereas posterosuperior segments as 1, 4a, 7, and 8. The primary endpoint was the incidence of postoperative complications. The definition of the Clavien–Dindo classification  was adopted and grade 3 or higher complications were regarded as significant outcomes. The comprehensive complication index (CCI)  was calculated using a dedicated automated online calculator (https://www.assessurgery.com/index.php).
Post-hepatectomy liver failure (PHLF) is a specific complication of hepatectomy that has a specific criteria and grading system. For the definition of PHLF, we used the criteria proposed by the International Study Group of Liver Surgery . The textbook outcome (TO) [16, 17] was defined as fulfilling all of the following 6 parameters after surgery: negative surgical margin, no perioperative transfusion, no postoperative complications, no prolonged hospital stay (≦ 50% of the total cohort), no readmission, and no postoperative mortality within 30 days after surgery.
Propensity score matching
The PSM was used to minimize selection bias. A caliper of 0.20 was used in this model. The propensity score was calculated by preoperative characteristics such as sex, age, body mass index (BMI), hepatitis virus infection (HBV-/HCV-positive), alcohol dependency (ingestion amount of over 60 g/day of alcohol), history of hepatectomy, platelet count, alanine transaminase (ALT), ICG retention rate at 15 min (ICG R 15), Child–Pugh score, difficulty grade, tumor number, tumor size, proximity of major vessel, and resection ratio.
The continuous variables were expressed as the mean ± standard deviation (SD). Operation time and blood loss in perioperative outcomes were expressed as the median (range). The Chi-squared test or Fisher’s exact test were used to compare categorical variables, and the unpaired t test was used to compare continuous variables. The predictive factors of postoperative complications were assessed using multivariate logistic regression analysis. Continuous variables (age, BMI, Child–Pugh score, tumor number, tumor size, resection ratio, blood loss, operation time, and blood tests) were used to calculate the unit odds ratio (OR). A p-value of less than 0.05 was considered statistically significant. All analyses were performed using JMP software version 13 (SAS Institute Inc., Cary, NC, USA).
Comparison of patient characteristics in unmatched cohort
We compared the characteristics of all study subjects (n = 290) between the open liver resection (OLR) group and the LLR group (Table 1). There were 178 patients who underwent OLR and 112 who underwent LLR. The factors that differed between the two groups were ALT, tumor size, proximity to major vessel and des-gamma carboxyprothrombin. Importantly, in the LLR group, the difficulty grade was distributed in easier categories and resection ratio was significantly lower than that in the OLR group (difficulty grade I/II/III: LLR: 84/10/18 vs. OLR: 43/39/96, p < 0.001; resection ratio: LLR: 11.4 ± 12.7% vs. OLR: 22.7 ± 17.2%, p < 0.001) (Table 1), indicating that smaller and less difficult hepatectomies were preferentially performed in the LLR group.
Surgical outcomes in unmatched cohort
Surgical outcomes were compared between the LLR group and the OLR group in all study subjects (Table 2). A significant difference was observed in the incidence of postoperative complications between the groups (LLR: 2.7% vs. OLR: 21.9%, p < 0.001). The incidence of grade 3 or higher ascites and pleural effusion were significantly lower in the LLR group, which contributed to a reduction in the overall incidence of complications in the LLR group. CCI was also lower in LLR. The occurrence of PHLF grade B or C was significantly less frequent in LLR than in OLR. In addition, significant differences were observed in intraoperative blood loss, operation time, perioperative transfusion, and postoperative hospital stay. The intraoperative accidents in the OLR group included one case of uncontrollable hemorrhage, one case of diaphragmatic injury, one case of portal vein injury, and two cases of bile duct injury. One case with uncontrollable hemorrhage (intraoperative blood loss was over 4000 g) in the LLR group was converted to open surgery. TO was achieved in higher proportion of patients in LLR group (LLR: 45.5% vs. OLR: 21.9%, P < 0.001). The cutoff value for hospital stay was 14 days, which was the median hospital stay for total cohort.
Patient characteristics after PSM
After performing PSM, each group consisted of 68 patients with comparable backgrounds. The patient characteristics of each group after PSM are shown in Table 3. There were no significant differences between the OLR group and LLR group in any of the parameters indicated. Difficulty grade was also matched well in both groups.
Surgical outcomes after PSM
Surgical outcomes were compared between the LLR group and the OLR group in the matched cohort (Table 4). Regarding the occurrence of postoperative complications, a significant difference was maintained between the OLR group (16.2%) and the LLR group (2.9%) (p = 0.017) in the matched cohort. Although the differences in the incidence of grade 3 or higher ascites and pleural effusion were no longer statistically significant after PSM, the tendency to decrease their frequency was maintained. LLR had significantly lower CCI than OLR. The incidence of PHLF grade B or C was not significantly different between the OLR group and the LLR group. The amount of blood loss was still less in the LLR group, but the difference between the groups became smaller and was no longer statistically significant. There was almost no difference between LLR and OLR in operation time. Postoperative hospital stay was significantly shorter in the LLR group. Higher percentage of the patients in LLR achieved the TO than in OLR.
Multivariate logistic regression analysis for the predictive factors of postoperative complications
Univariate and multivariate analyses of pre- and perioperative parameters in predicting postoperative complications were performed (Table 5). Univariate analysis demonstrated that the value of γ-glutamyl transpeptidase (GTP) and des-gamma carboxyprothrombin, Child–Pugh score, difficulty grade II and III, proximity to major vessel, resection ratio, extent of surgery (major or minor resection), surgical approach (LLR or OLR), blood loss, operation time, and intraoperative accidents were associated with postoperative complications. For multivariate analysis, blood loss, operation time, and intraoperative accidents were excluded because they are the consequences of surgery. Furthermore, the extent of surgery was excluded because it is strongly confounded with the resection ratio. As a result, multivariate logistic regression analyses were performed on the seven factors (the value of γ-GTP and des-gamma carboxyprothrombin, Child–Pugh score, difficulty grade II and III, resection ratio, proximity to major vessel, and surgical approach), which demonstrated that surgical approach and proximity to major vessel were the independent predictor of postoperative complications.
Since the introduction of LLR, its indications have gradually extended with advances in perioperative care and laparoscopic devices and techniques [18, 19]. The potential advantages of laparoscopic surgery include small operative incisions, minimum invasiveness, and early postoperative recovery. However, there are several potential disadvantages, including the difficulty of the surgery due to the restricted movement and the longer operation time associated with the difficulty.
Several studies have compared clinical outcomes between LLR and OLR and reported less intraoperative blood loss, shorter postoperative hospital stays, and fewer postoperative complications in LLR [20,21,22]. In fact, many previous reports highlighted the efficacy of LLR. Although there was one randomized controlled trial (RCT) comparing LLR and OLR for colorectal liver metastases,  most studies are retrospective studies, and patient selection bias may have greatly influenced the results. Thus far, there has been no RCT comparing LLR and OLR in HCC.
PSM is a method of minimizing patient selection bias in retrospective studies and has been adopted in studies comparing outcomes between LLR and OLR for HCC [9, 24,25,26,27]. In some of these studies, LLR decreased postoperative complications [9, 27]. However, even though PSM was performed, the patient background may still have been biased because factors that were not considered in PSM may not have been balanced. In most previous reports, there were factors that were not considered for PSM such as difficulty of LLR and what percentage of normal liver parenchyma was sacrificed. These factors are important factors for the surgeon to decide upon LLR or OLR, and can also affect the occurrence of postoperative complications. Therefore, performing PSM in consideration of these factors seems very significant. The IMM classification adopted in the present study was proposed by Kawaguchi et al.  It has been compared both in LLR and in OLR, and its validity has been demonstrated . In the present study, the occurrence of the postoperative complications was less frequent in the LLR group when analyzing the pre-matched cohort. However, the baseline characteristics of the patients had a significant bias between the LLR group and the OLR group; less difficult and smaller resections were performed for patients in the LLR group. It is reasonable to assume that the less difficult and smaller resections resulted in the lower incidence of postoperative complications in the LLR group. However, this was not the case; even when the difficulty and resection ratio were adjusted and matched by PSM, the significant difference in the incidence of postoperative complications was maintained. This result strongly suggested that the surgical approach (LLR or OLR) contributed to the decreased incidence of postoperative complications, regardless of the difficulty of liver resection or the resection ratio. The significantly better CCI in LLR was maintained after matching and was consistent with some previous reports [17, 23]. Although significant differences in the frequency of individual complications observed in the prematched cohort disappeared after PSM, this may be due to statistical β-error. Our findings were further supported by the multivariate logistic analysis, which showed that surgical approach was the independent predictor for postoperative complications. Longer operation time and more blood loss were observed in the OLR group in the unmatched cohort, but the operation time was not different between the two groups and the amount of blood loss tended to be small in LLR group in the matched cohorts. The postoperative hospital stay was significantly shorter in the LLR group than in the OLR group after PSM, which was consistent with many other previous studies [9, 24,25,26,27]. Fewer postoperative complications in the LLR group were reasonably considered to have contributed to the shorter postoperative hospital stays in the LLR group. For TO, LLR had significantly better results. Some previous reports indicated that TO contributed to long-term prognosis and TO was considered to be an important index for the perioperative period [16, 17]. In this study, this result was thought to be due to the low incidence of postoperative complications and the short postoperative hospital stay.
This study has several limitations. First, this study involved a relatively small number of patients at a single institution. In particular, the number of patients after PSM decreased significantly because the difficulty and resection ratio were the factors that had a particularly profound influence on the determination of the surgical approach. Second, the time at which patients were operated upon in this study was during the period in which the rate of laparoscopic liver resection was increasing in our department. As a result, it is difficult to exclude the possibility that there was a bias in the period of surgery or in the surgeons’ experience between the LLR and the OLR groups. Third, no matter how much the IMM classification was used to adjust the difficulty, it could not be ruled out that a bias still existed between the two groups. For example, surgery history, especially a history of hepatectomy, dramatically increases the difficulty of surgery, but this is not included in the IMM classification, although adhesions caused by past surgery may worsen the outcome of the current surgery by increasing blood loss and operation time. Our current study indicated that LLR for HCC is effective in reducing the incidence of postoperative complications. More careful analysis of the data, or ideally randomized controlled trials, is necessary to prove this concept.
Sato M, Tateishi R, Yasunaga H et al (2012) Mortality and morbidity of hepatectomy, radiofrequency ablation, and embolization for hepatocellular carcinoma: a national survey of 54,145 patients. J Gastroenterol 47:1125–1133
Law WL, Choi HK, Lee YM et al (2007) The impact of postoperative complications on long-term outcomes following curative resection for colorectal cancer. Ann Surg Oncol 14:2559–2566
Amisaki M, Saito H, Tokuyasu N et al (2018) Prognostic value of postoperative complication for early recurrence after curative resection of hepatocellular carcinoma. Hepatobiliary Pancreat Dis Int 17:323–329
Wu Q, Wei M, Ye Z et al (2017) Laparoscopic colectomy versus open colectomy for treatment of transverse colon cancer: a systematic review and meta-analysis. J Laparoendosc Adv Surg Tech A 27:1038–1050
Li Z, Zhao Y, Liu Y et al (2019) Laparoscopic versus open gastrectomy for high-risk patients with gastric cancer: a systematic review and meta-analysis. Int J Surg 65:52–60
Bhojani FD, Fox A, Pitzul K et al (2012) Clinical and economic comparison of laparoscopic to open liver resections using a 2-to-1 matched pair analysis: an institutional experience. J Am Coll Surg 214:184–195
Ito K, Ito H, Are C et al (2009) Laparoscopic versus open liver resection: a matched-pair case control study. J Gastrointest Surg 13:2276–2283
Welsh FK, Tekkis PP, John TG et al (2010) Open liver resection for colorectal metastases: better short- and long-term outcomes in patients potentially suitable for laparoscopic liver resection. HPB (Oxford) 12:188–194
Takahara T, Wakabayashi G, Beppu T et al (2015) Long-term and perioperative outcomes of laparoscopic versus open liver resection for hepatocellular carcinoma with propensity score matching: a multi-institutional Japanese study. J Hepatobiliary Pancreat Sci 22:721–727
Kubota KMM, Kusaka K et al (1997) Measurement of liver volume and hepatic functional reserve as a guide to decision-making in resectional surgery for hepatic tumors. Hepatology. https://doi.org/10.1053/jhep.1997.v26.pm0009362359
Ohshima S (2014) Volume analyzer SYNAPSE VINCENT for liver analysis. J Hepatobiliary Pancreat Sci 21:235–238
Kawaguchi Y, Fuks D, Kokudo N et al (2018) Difficulty of laparoscopic liver resection: proposal for a new classification. Ann Surg 267:13–17
Dindo D, Demartines N, Clavien PA (2004) Classification of surgical complications: a new proposal with evaluation in a cohort of 6336 patients and results of a survey. Ann Surg 240:205–213
Slankamenac K, Graf R, Barkun J et al (2013) The comprehensive complication index: a novel continuous scale to measure surgical morbidity. Ann Surg 258:1–7
Rahbari NN, Garden OJ, Padbury R et al (2011) Posthepatectomy liver failure: a definition and grading by the International Study Group of Liver Surgery (ISGLS). Surgery 149:713–724
Merath K, Chen Q, Bagante F et al (2019) A multi-institutional international analysis of textbook outcomes among patients undergoing curative-intent resection of intrahepatic cholangiocarcinoma. JAMA Surg 154:e190571
Hobeika C, Nault JC, Barbier L et al (2020) Influence of surgical approach and quality of resection on the probability of cure for early-stage HCC occurring in cirrhosis. JHEP Rep 2:100153
Sheen AJ, Jamdar S, Siriwardena AK (2019) Laparoscopic hepatectomy for colorectal liver metastases: the current state of the art. Front Oncol 9:442
Yamazaki S, Takayama T (2019) Current topics in liver surgery. Ann Gastroenterol Surg 3:146–159
Guro H, Cho JY, Han HS et al (2018) Outcomes of major laparoscopic liver resection for hepatocellular carcinoma. Surg Oncol 27:31–35
Kasai M, Cipriani F, Gayet B et al (2018) Laparoscopic versus open major hepatectomy: a systematic review and meta-analysis of individual patient data. Surgery 163:985–995
Tsai KY, Chen HA, Wang WY et al (2019) Long-term and short-term surgical outcomes of laparoscopic versus open liver resection for hepatocellular carcinoma: might laparoscopic approach be better in early HCC? Surg Endosc 33:1131–1139
Fretland AA, Dagenborg VJ, Bjornelv GMW et al (2018) Laparoscopic versus open resection for colorectal liver metastases: the OSLO-COMET randomized controlled trial. Ann Surg 267:199–207
Deng ZC, Jiang WZ, Tang XD et al (2018) Laparoscopic hepatectomy versus open hepatectomy for hepatocellular carcinoma in 157 patients: a case controlled study with propensity score matching at two Chinese centres. Int J Surg 56:203–207
Inoue Y, Ota M et al (2018) Short- and long-term results of laparoscopic parenchyma-sparing hepatectomy for small-sized hepatocellular carcinoma: a comparative study using propensity score matching analysis. Am Surg 84:230–237
Lanyun Luo HZ, Yao Y, Huang X (2015) Laparoscopic versus open hepatectomy for hepatocellular carcinoma: short- and long-term outcomes comparison. Int J Clin Exp Med 8:18772–18778
Wu X, Huang Z, Lau WY et al (2019) Perioperative and long-term outcomes of laparoscopic versus open liver resection for hepatocellular carcinoma with well-preserved liver function and cirrhotic background: a propensity score matching study. Surg Endosc 33:206–215
Kawaguchi Y, Hasegawa K, Tzeng CD et al (2020) Performance of a modified three-level classification in stratifying open liver resection procedures in terms of complexity and postoperative morbidity. Br J Surg 107:258–267
We thank H. Nikki March, PhD, from Edanz Group (www.edanzediting.com/ac) for editing a draft of this manuscript.
This research did not receive any specific Grant from funding agencies in the public, commercial, or not-for-profit sectors.
Conflict of interest
The authors of this manuscript have no conflicts of interest to disclose as described by this article.
This research was approved by the Institutional Review Board at Kyoto University Graduate School of Medicine (approval code: R1579).
The requirement for obtaining written informed consent from each patient was waived because of the retrospective study design.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Our study is registered with University Hospital Medical Information Network (UMIN, https://www.umin.ac.jp). The registration number is UMIN000041288 (https://upload.umin.ac.jp/cgi-openbin/ctr/ctr_view.cgi?recptno=R000047142).
About this article
Cite this article
Uemoto, Y., Taura, K., Nishio, T. et al. Laparoscopic Versus Open Liver Resection for Hepatocellular Carcinoma: A Case Controlled Study with Propensity Score Matching. World J Surg 45, 2572–2580 (2021). https://doi.org/10.1007/s00268-021-06115-2