Surgical outcome of laparoscopic hepatic resection for hepatocellular carcinoma: A matched case-control study with propensity score matching.

BACKGROUND: Although the number of reports on laparoscopic hepatic resection (LHR) has increased, studies of long-term outcomes regarding tumor recurrence and patient survival compared to the conventional open approach are limited. We evaluated the long-term survival and feasibility of LHR in patients with hepatocellular carcinoma (HCC). PATIENTS AND METHODS: A retrospective analysis was performed on the clinical data of patients who underwent hepatic resection for primary HCC between August 2000 and December 2013. The patients were divided into the LHR or open hepatic resection (OHR) groups. To control for selection bias in the two groups, propensity score matching was used at a 1:1 ratio based on the following covariates: Child-Pugh grade, tumour size, tumour number and tumour location. Following propensity score matching, thirty patients were included in the LHR group and thirty were included in the OHR group.
RESULTS: The respective disease-free survival rates at 1 year, 3 years and 5 years were 78.4%, 61.1% and 38.9%, respectively, for the LHR group, and 89.3%, 57.5% and 47.9%, respectively, for the OHR group (P = 0.89). Also, the overall survival rates at 1 year, 3 years and 5 years were 96.4%, 68.2% and 62.5%, respectively, for the LHR group and 100.0%, 95.8% and 72.3%, respectively, for the OHR group (P = 0.44).
CONCLUSIONS: According to our study, using propensity score matching, LHR for HCC is safe, feasible and comparative, with good oncologic results.

INTRODUCTION

Hepatocellular carcinoma (HCC) is the most common primary liver cancer, and hepatic resection is the gold standard therapy with the best curative treatment option.[1] Laparoscopic hepatic resection (LHR) was reported first in 1991 by Reich et al.[2] for the treatment of patients with benign liver tumours. The first anatomical liver resection was reported in 1996 by Azagra et al.[3] Due to the technical and technological improvements achieved within the last 20 years, LHR has been considered to be a feasible and safe technique.[4567] A review by Nguyen et al.[4] in 2009 reported outcomes following LHR for 2804 patients, citing reduced intraoperative blood loss, post-operative morbidity and mortality and shorter hospital stay. Several studies have reported comparative results of LHR versus open hepatic resection (OHR).[567] However, few studies of long-term outcomes regarding tumour recurrence and patient survival compared to the conventional open approach have been reported.[48910111213141516] We compared the long-term outcome and safety of LHR for HCC versus the open approach during the same period.

PATIENTS AND METHODS

In August 2000, we began to perform LHR in the Department of Gastroenterological Surgery, Aichi Medical University School of Medicine. With technological and technical advances, LHR (including partial resection and left lateral segmentectomy) has become the established surgical procedure for HCC and the focus of our investigation in this study.

To eliminate operative bias in our study, we excluded the following: patients who had percutaneous radiofrequency ablation or transarterial chemoembolisation during the previous operation or the intraoperative period, combined HCC noted on pathologic results, non-curative resection, synchronous other organ resection (e.g. Hassab's operation and gastrectomy) and a medical history of other malignant tumour. Finally, we studied retrospectively 110 patients who underwent hepatic resection for primary HCC between August 2000 and December 2013. The patients were divided into the LHR (30) and OHR (80) groups. Both groups were compared with respect to operative time, blood loss, post-operative complications, length of post-operative hospital stay and long-term outcome. Morbidity was classified according to the Clavien–Dindo classification.[17]

Statistical analysis

We performed a 1:1 propensity score-matched analysis to control for selection bias between the OHR and LHR groups, and we compared the thirty patients who underwent laparoscopic surgery to thirty matched patients who underwent open surgery. To estimate the propensity score, a multivariate logistic regression analysis was used based on the following covariates: Child–Pugh grade, tumour size, tumour number and tumour location. This matching was performed without replacement and using a caliper width of 0.2 standard deviations of the logit of the estimated propensity score. Overall survival (OS) and disease-free survival (DFS) rates were calculated by the Kaplan–Meier survival curve and the log-rank test was performed. Values of P < 0.05 were considered to indicate statistically significant differences.

Propensity score matching was performed using SAS version 9.3 (SAS Institute, Inc., Cary, NC, USA). The other statistical analyses were performed with EZR (Saitama Medical Center, Jichi Medical University), which is a graphical user interface for R (The R Foundation for Statistical Computing). More precisely, it is a modified version of R commander designed to add statistical functions used frequently in biostatistics.[18]

Operative techniques of hepatic resection

All hepatic resections via the OHR and LHR approaches were performed by the same experienced team of hepatobiliary surgeons. Before surgery, each patient was individually assessed by our multidisciplinary team of surgeons, hepatologists and radiologists. Patients were followed up by computed tomography, magnetic resonance imaging or ultrasonography every 3 or 4 months postoperatively.

In both groups, intraoperative ultrasonography was performed routinely. After tumour location was confirmed, we preferred to apply a previous coagulation using a microwave tissue coagulator (Microtaze, Azwell, Osaka, Japan) along the incision line before the actual liver excision to reduce blood loss during liver parenchymal transection. The superficial hepatic parenchyma was resected using the Harmonic scalpel (Ethicon Endo-Surgery, USA), and the deeper portion of the parenchyma was transected using a Cavitron ultrasonic surgical aspirator (CUSA system, Valleylab, USA). In case of bleeding, we used either the TissueLink (TissueLink Medical Inc., Dover, NH, USA) or monopolar soft-coagulation (VIO system, ERBE, Germany) systems. Due to the previous coagulation to reduce blood loss, Pringle's manoeuvre was performed in a selected few patients in each group.

Selection criteria for LHR at our centre included the following: tumours were smaller than 10 cm in the left liver and smaller than 5 cm in the right liver, located in accessible segments of the liver and were at the free margin of Glisson's pedicle, and patients had no history of previous major laparotomy at the upper abdomen. LHR was performed with three to five ports (5–12 mm in diameter) and the patient's position was decided on the basis of tumour location. The first periumbilical port for the laparoscopic camera was inserted via an open technique, and the flexible laparoscope (LFT type, Olympus, Tokyo, Japan) was used in all cases.

On the other hand, for the open approach, an upper midline incision with subcostal extension was standard. Patients were operated on under continuous CO2 pneumoperitoneum at 10 mmHg pressure. The resected liver was placed into a protective bag and extracted, with the least fragmentation possible, through a slightly enlarged periumbilical port site. After specimen removal, the resection surface was inspected and a flat drain was placed close to it. If there was difficulty locating the tumour via laparoscopy alone due to limited visibility or difficult-to-control bleeding, we used a hand-assisted laparoscopic surgery (HALS) or hybrid procedure, also called laparoscopic-assisted open resection.

Therefore, the LHR technique could be classified into three main categories based on the Louisville statement:[1] pure laparoscopy, HALS and the hybrid technique. The pure laparoscopic procedure involves performance of the entire hepatic resection using laparoscopic ports only. HALS is defined as laparoscopy with the addition of a hand port for the purpose of facilitating the procedure. The hybrid technique, called laparoscopy-assisted open resection, is started laparoscopically to mobilise the liver, followed by a mini-laparotomy to complete parenchymal transection.

RESULTS

Analysis of unmatched patients

Patient and tumor characteristics and perioperative outcomes of unmatched patients in the two groups are summarised in Table 1. We studied thirty patients (18 males and 12 females; median age, 70 years; range, 19–86 years) who underwent LHR (partial resection in 24 cases, left lateral segmentectomy in 6 cases) and eighty patients (60 males and 20 females; median age, 68 years; range, 28–84 years) who underwent OHR (partial resection in 68 cases, left lateral segmentectomy in 12 cases). No significant differences between the two groups were observed in the unmatched patient characteristics: age, sex, Child–Pugh grade, laboratory test results, tumour number, tumour size, histology, positive vascular invasion, positive surgical margin, liver cirrhosis and tumour staging (International Union Against Cancer [UICC] 7).[19] However, only tumour location (Couinaud's segment) was significantly different between the two groups.

Table 1

Patients, tumour characteristics and perioperative outcomes of all unmatched patients

The median operative time was 234 min (range, 72–493 min) in the LHR group and 189 min (range, 70–424 min) in the OHR group (P = 0.17). Median blood loss was 262 mL (range, 1–1791 mL) and 377 mL (range, 20–3032 mL), respectively (P = 0.02). Seven patients in the OHR group and three in the LHR group required blood transfusion (P = 1.00). Of the thirty patients in the LHR group, two (6.7%) underwent conversion to an OHR for bleeding control. The median hospital stay was 12 days (range, 4–25 days) in the LHR group and 16 days (range, 9–100 days) in the OHR group (P < 0.05). In the LHR group, there was no mortality and eight patients experienced complications, compared to 1 and 28 patients, respectively, in the OHR group. There was no significant difference in the complication rate between the two groups (P = 0.50).

Median follow-up was 37.4 months in the LHR group and 55.5 months in the OHR group (P = 0.10). The respective DFS rates at 1, 2, 3, 4 and 5 years were 78.4%, 70.5%, 61.1%, 55.6% and 38.9% in the LHR group, and 82.7%, 65.9%, 53.7%, 38.1% and 34.5% in the OHR group [P = 0.45; Figure 1a], respectively. Also, respective OS rates at 1, 2, 3, 4 and 5 years were 96.4%, 85.6%, 69.2%, 68.2% and 62.5% in the LHR group, and 94.8%, 89.4%, 85.0%, 74.0% and 67.2% in the OHR group [P = 0.97; Figure 1b] respectively. The DFS and OS rates showed no significant difference between the two groups of all unmatched patients.

Figure 1

(a) Kaplan–Meier disease-free survival analysis of unmatched patients with laparoscopic hepatic resection and open hepatic resection (P = 0.45, log-rank test). (b) Kaplan–Meier overall survival analysis of unmatched patients with laparoscopic hepatic resection and open hepatic resection (P = 0.97, log-rank test)

Analysis of matched patients

We performed a 1:1 propensity score-matched analysis between the OHR and LHR patients, and we compared the thirty patients who underwent laparoscopic surgery to thirty matched patients who underwent open surgery. After case matching, patient characteristics were not significantly different between the two groups [Table 2]. In matched patients, compared to the OHR group, the LHR group demonstrated a significantly shorter hospital stay (17 vs. 12 days, P < 0.05), while the other perioperative outcomes showed no significant difference between the two groups.

Table 2

Patients, tumour characteristics and perioperative outcomes of matched patients

Median follow-up was 37.4 months in the LHR group and 52.9 months in the OHR group of matched patients (P = 0.08). The respective DFS rates at 1, 2, 3, 4 and 5 years were 78.4%, 70.5%, 61.1%, 55.6% and 38.9% in the LHR group, and 89.3%, 70.1%, 57.5%, 47.9% and 47.9% in the OHR group [P = 0.89; Figure 2a] respectively. On the other hand, OS rates at 1, 2, 3, 4 and 5 years were 96.4%, 85.6%, 68.2%, 68.2% and 62.5% in the LHR group, and 100.0%, 100.0%, 95.8%, 82.5% and 72.3% in the OHR group [P = 0.44; Figure 2b] respectively. Also, the DFS and OS rates showed no significant difference between the two groups of matched patients.

Figure 2

(a) Kaplan–Meier disease-free survival analysis of matched patients with laparoscopic hepatic resection and open hepatic resection (P = 0.89, log-rank test). (b) Kaplan–Meier overall survival analysis of matched patients with laparoscopic hepatic resection and open hepatic resection (P = 0.44, log-rank test)

DISCUSSION

Our study of long-term outcomes revealed that LHR for HCC is safe, feasible and comparative, with good oncologic results. In the matched and unmatched studies, the OS and DFS rates showed no significant difference between the LHR and OHR groups. These rates are the most important factors in the treatment of HCC with LHR. Positive surgical margin invasion may have a profound influence on long-term survival.[2021] In our unmatched study, one patient (3.3%) in the LHR group and eight (10%) in the OHR group had positive surgical margin invasion (P = 0.44).

With respect to short-term outcomes, there were no significant differences in operative time, blood transfusion or perioperative complications. The LHR group, treated with a minimally invasive technique, had significantly shorter hospital stays than the OHR group. In an unmatched study, blood loss was significantly less in the LHR than in the OHR groups (there were no significant differences in the matched study). We experienced only one case (3.3%) of severe complications in the LHR group (reoperation required for haemorrhage). Of the thirty patients in the LHR group, two (6.7%) underwent conversion to an OHR due to uncontrollable bleeding.[222324] In 18 cases (60.0%) with liver cirrhosis, LHR was performed with good results.[1625] Many studies have indicated better results with LHR than OHR with respect to short-term outcome.[316262728]

Japanese comprehensive screening systems, advances in diagnostic approach and careful application of recently developed treatment methods have improved the 5-year survival rate from 5% in 1978 to 43% in a nationwide follow-up survey conducted by the Liver Cancer Study Group of Japan between 1978 and 2005.[29] In addition, median survival time has improved from 4 to 50 months during the last 28 years. This is an exceptional treatment outcome that, to our knowledge, no other country in the world has come close to achieving.

Introduction of the laparoscopic approach in hepatic resection has been slower than in other fields of surgery because of technical difficulties and a high bleeding risk.[30] At our centre, we initiated our experience with LHR in 2000 and have gradually increased our expertise over the past 15 years. LHR has become a standard procedure for left lateral segmentectomy or partial resection. Also, the Louisville consensus statement concluded that laparoscopic left lateral segmentectomy should be considered standard practice, and described the currently acceptable indications for LHR as patients with solitary lesions, 5 cm or less, located in liver segments 2–6.[1]

Recent reports describing surgical outcome of LHR for HCC have demonstrated OS and DFS rates at 5 years ranging from 50.0% to 89.7% and 26.3% to 53.5%, and reviews of laparoscopic liver resection for HCC have shown good long-term outcomes.[4810] However, our retrospective study is limited by fewer patients. Yoon et al.[9] indicated that more detailed long-term studies, especially in the form of RCTs, will be needed.

We performed a propensity score-matched analysis to compare the long-term outcomes of open and laparoscopic surgery for primary HCC.

CONCLUSION

A larger number of patients are needed to obtain definitive results. LHR is a promising treatment option with minimally invasive surgery benefits for HCC patients.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.