Cumulative disease-free survival of patients who underwent extended right-sided hepatic resection for hepatocellular carcinoma (median survival, 8.9 months).
Cumulative overall survival of patients who underwent extended right-sided hepatic resection for hepatocellular carcinoma (median survival, 34.5 months).
Customize your JAMA Network experience by selecting one or more topics from the list below.
Chik BH, Liu CL, Fan ST, et al. Tumor Size and Operative Risks of Extended Right-Sided Hepatic Resection for Hepatocellular Carcinoma: Implication for Preoperative Portal Vein Embolization. Arch Surg. 2007;142(1):63–69. doi:10.1001/archsurg.142.1.63
Major hepatic resection for hepatocellular carcinoma (HCC) is associated with high operative morbidity and mortality, especially in patients with underlying chronic liver disease. The present study evaluated the factors associated with increased operative risks in patients who underwent extended right-sided hepatic resection for HCC.
Tertiary referral center.
A retrospective study was performed on 172 patients who underwent extended right-sided hepatic resection of more than 4 Couinaud segments for HCC during a 16-year period (January 1, 1989, to December 31, 2004) to evaluate the clinical factors associated with operative morbidity and mortality.
Main Outcome Measure
Risk factors associated with hospital mortality and major operative morbidity.
The overall major morbidity and hospital mortality rates were 14.0% and 8.1%, respectively. On multivariate analysis, small tumor size, conventional-approach hepatectomy, Child-Pugh grade B cirrhosis, and preexisting tumor rupture were the independent factors significantly associated with an increased risk of operative mortality. Discriminant analysis showed that a tumor size smaller than 10 cm significantly increased the risk of operative mortality compared with larger tumors (17.2% vs 3.5%; P = .046).
Anterior approach is the preferred technique for extended right-sided hepatic resection for HCC. Increased risk of operative mortality was identified in patients who had a small tumor, which was associated with the resection of a large volume of functioning liver parenchyma. Preoperative portal vein embolization should be considered in this group of patients to enhance atrophy of the right lobe and hypertrophy of the future liver remnant to minimize the operative risk.
With improved perioperative and operative care of patients undergoing hepatic resection for hepatocellular carcinoma (HCC), a significant improvement in the associated operative morbidity and mortality has been observed in recent years.1,2 However, extended right-sided hepatic resection remains a major surgical challenge, especially when underlying chronic liver diseases are present, because of the resection of a large volume of liver. In addition, the extent of underlying liver diseases often limits the options of treatment because advanced cirrhosis precludes major hepatic resection.3 One of the most severe complications of major hepatic resection for HCC is the development of liver failure, which is associated with a difficult postoperative course if not operative mortality. Although a large HCC may pose technical difficulties for hepatic resection, a smaller HCC does not necessarily equate to better operative and survival outcomes. In fact, major right-sided hepatic resection for a small HCC results in resection of a large volume of nontumorous functioning liver parenchyma, leaving a small remnant of liver.4,5 Previous studies have evaluated the risk factors associated with increased operative risks in patients undergoing major hepatic resection for HCC,6,7 and preoperative liver function assessment using the Child-Pugh grading system8 and the indocyanine green (ICG) clearance test9,10 have been used as important criteria for selection of patients for major hepatic resection. The importance of the tumor size as a risk factor for extended right-sided hepatic resection has not been previously addressed. The aim of the present study was to evaluate the factors associated with an increased risk of operative morbidity and hospital mortality in patients undergoing extended right-sided hepatic resection for HCC and, hence, to determine a strategy to lower its associated mortality or even to expand resectability of HCC in our patients.
A retrospective study was performed on 179 consecutive patients who underwent extended right-sided hepatic resection for HCC at the Department of Surgery of the University of Hong Kong at Queen Mary Hospital from January 1, 1989, to December 31, 2004. To evaluate the importance of tumor size and the associated resected volume of nontumorous functioning liver parenchyma, 7 patients with preoperative right-sided portal vein embolization (PVE) were excluded from the present analysis. The remaining 172 patients were the subjects of the present study. Extended right-sided hepatic resection was defined as the resection of more than 4 right-sided Couinaud segments.11 All preoperative and postoperative data were collected prospectively in a computerized database by a research assistant.
Preoperative investigation of the patients included blood biochemical analysis, serum α-fetoprotein assay, chest x-ray film examination, percutaneous ultrasonography, computed tomography (CT) of the abdomen, and hepatic angiography or magnetic resonance imaging in selected patients. Bilobar disease, extrahepatic metastases, and inferior vena cava or main portal vein tumor thrombus precluded curative hepatic resection. Hepatic function was evaluated with a combination of Child-Pugh classification,8 liver biochemical analysis, and ICG clearance test. Although on ICG clearance testing a retention at 15 minutes of less than 14% was found to be the safety limit of major hepatic resection in our previous evaluation,9 patients' general condition and the presence of significant comorbidity were also taken into consideration in the decision to perform hepatic resection.
Hepatic resection was performed using the standard technique described previously.12,13 The operation started with a bilateral subcostal incision with an upward midline extension, and an ultrasonic dissector was used for parenchymal transection.14 Intraoperative ultrasonography was performed routinely to delineate the extent of tumor involvement, detect tumor nodules in the contralateral lobe and invasion of the tumor to major blood vessels, and plan and mark the plane of parenchymal transection. The anterior approach technique was adopted for selective patients with large HCCs that were difficult to mobilize prior to parenchymal transection before 1999.15 The technique was also used in 28 patients who were recruited in a prospective randomized study from 1999 to 2004. The Pringle maneuver with intermittent clamping of the portal inflow for 15 minutes and release for a 5-minute interval was applied in selected patients, mainly when difficult parenchymal transection with excessive bleeding was encountered. Total vascular exclusion had not been used. The bile leakage test was performed routinely using methylene blue injection via a cannula placed inside the common bile duct through the cystic duct after transection. An intra-abdominal drain had not been placed since 2002.16
All patients received the same perioperative care by the same team of surgeons and were treated in the intensive care unit during the early postoperative period after hepatic resection. All patients received broad-spectrum antibiotics for 3 days; parenteral nutrition in the form of branched-chain amino acid–enriched solution, low-dose dextrose, and medium- and long-chain triglycerides was provided until oral feeding was well tolerated. An oral preparation of branched-chain amino acid–enriched nutritional supplement (Aminoleban EN; Otsuka Pharmaceutical Co Ltd, Seoul, Korea) was also given postoperatively for approximately 1 month. All intraoperative complications and postoperative morbidities were recorded prospectively. Hospital mortality was defined as death during the same period of hospitalization as for the hepatic resection. Major operative morbidity was defined as the occurrence of 1 or more of the following potentially life-threatening perioperative events: liver failure, renal failure, subphrenic abscess, intra-abdominal hemorrhage, esophageal variceal bleeding, peptic ulcer bleeding, biliary fistula, infected ascites, empyema thoracis, intestinal obstruction, bowel fistula, myocardial infarction, or pulmonary embolism.
Continuous data were expressed as medians with their ranges unless otherwise stated. Proportions were given as number and percentage. Survival analysis, including cumulative overall survival and disease-free survival, was estimated by the Kaplan-Meier survival method. Differences in clinical variables were compared using the Mann-Whitney U test for continuous data. Categorical variables were compared using the χ2 test or Fisher exact test where appropriate. Multivariate analysis was performed using stepwise logistic regression analysis to identify significant independent risk factors associated with hospital mortality or operative morbidity. P<.05 was considered to indicate statistical significance. All statistical analyses were performed using SPSS statistical software, version 11.0 (SPSS Inc, Chicago, Ill).
Among the 172 patients who underwent extended right-sided hepatic resection for HCC during the study period, 137 (79.7%) were male and 35 (20.3%) were female, with a median age of 52 years (range, 13-79 years). The serum samples from 146 patients (84.9%) tested positive for hepatitis B surface antigen. Details of the clinical variables are listed in Table 1. Four patients (2.3%) had received transarterial oily chemoembolization before surgery.
The extent of hepatic resection according to the Brisbane 2000 terminology17 is indicated in Table 2. Forty-one patients (23.8%) underwent concomitant complete caudate lobectomy. The intraoperative and postoperative data of the patients are listed in Table 3. The median intraoperative blood loss was 1.7 L (range, 0.2-20.0 L), and 79 patients (46.0%) did not require blood transfusion. Preexisting tumor rupture was evident on laparotomy in 19 patients (11.0%). The intermittent Pringle maneuver and anterior approach hepatic resection were used in 69 patients (40.1%) and 82 patients (47.7%), respectively. The median postoperative intensive care unit and hospital stays were 1 day and 14 days, respectively.
Histological examination findings of the resected specimens showed that the median tumor size was 11.3 cm (range, 2.5-28.0 cm). One hundred twenty-one patients (82.0%) had evidence of underlying chronic liver diseases, including 44 patients (25.6%) who had underlying cirrhosis and 97 patients (56.4%) who had chronic hepatitis. The details of tumor staging according to the American Joint Committee on Cancer18 are listed in Table 4. Hospital mortality occurred in 14 patients (8.1%; 13 men and 1 woman; median age, 60 years; age range, 38-79 years), and the causes of death are presented in Table 5. Major operative morbidity occurred in 24 patients (14.0%). Most of these patients (n = 13) were affected by 2 or more operative complications (Table 6). The median disease-free survival was 8.9 months (Figure 1), and the median overall survival was 34.5 months (Figure 2) on follow-up.
To identify the risk factors associated with hospital mortality, univariate analysis was performed for different clinical factors, including patient, tumor, and intraoperative factors (Table 7). Significant factors that were associated with an increased risk of hospital mortality included tumor size less than 10 cm, Child-Pugh grade B cirrhosis,8 preexisting tumor rupture, and conventional-approach hepatic resection instead of the anterior-approach technique. Discriminant analysis showed that a tumor size less than 10 cm significantly increased the risk of hospital mortality compared with a larger tumor (17.2% vs 3.5%; P = .046). The Mann-Whitney U test was used to analyze continuous variables, including age, serum α-fetoprotein level, serum albumin level, serum total bilirubin level, platelet count, hemoglobin level, prothrombin time, ICG clearance at 15 minutes, tumor size, and intraoperative blood loss. Tumor size was the only risk factor significantly associated with hospital mortality (P = .01).
Clinical factors that were significantly associated with hospital mortality on univariate analysis were then subjected to a stepwise forward multivariate logistic regression analysis. Four independent clinical factors were associated with an increased risk of hospital mortality. These factors included conventional-approach hepatic resection, Child-Pugh grade B cirrhosis,8 preexisting tumor rupture, and tumor size smaller than 10 cm (Table 8). Statistical study using univariate analysis showed that Child-Pugh grade B cirrhosis8 was the only significant clinical factor associated with an increased risk of major operative morbidity among the 172 patients undergoing extended right-sided hepatic resection for HCC (Table 9).
Although hepatic resection is now commonly performed for curative treatment of patients who have hepatobiliary malignant neoplasms with minimal hospital mortality and acceptable operative morbidity, extended right-sided hepatic resection remains a high-risk operation even when performed in specialized and high-volume centers.19 Previous reports on the safety and success of extended hepatic resection in the literature included a high proportion of patients with liver metastases5,20,21 having healthy, noncirrhotic livers with a much lower operative risk of liver failure, which was the main cause of most postoperative hospital mortality and major operative morbidity. On the other hand, most reports on hepatic resection in cirrhotic patients comprised small-volume hepatic resection, including wedge resection.22,23
What set extended hepatectomy apart was the technical complexity, which can result in significant intraoperative hemorrhage and a higher incidence of technical complications, including iatrogenic tumor rupture or biliary complications. In addition, the volume of liver remnant is an important contributing factor, especially in patients with HCC, that is often associated with underlying chronic liver diseases. As demonstrated in the present study, 82.0% of the study population had underlying chronic liver diseases, and 25.6% had established liver cirrhosis. Impaired function and regenerative capacities after extensive liver resection render such patients at higher risk for postoperative liver failure.
Among the 172 patients who underwent extensive hepatectomy with extended right-sided resection for HCC, the hospital mortality rate was 8.1%. The results of the present study identified 4 independent risk factors that were significantly associated with hospital mortality, including Child-Pugh grade B cirrhosis,8 preexisting tumor rupture, conventional-approach hepatic resection, and tumor size less than 10 cm. The hospital mortality was 50% among the 6 patients included in the present study with Child-Pugh grade B cirrhosis.8 This finding illustrates that extended right-sided hepatic resection might have been too extensive for this subgroup of patients with poor liver function reserve. Whether preoperative right PVE would induce hypertrophy of the future liver remnant and thus improve the operative risks of these patients is still a controversial question.24 Nevertheless, strict selection criteria and better assessment of the operative risks are essential to identify the best treatment regimen for these patients. In selected patients who have HCC with underlying liver cirrhosis and poor functional reserve, liver transplantation may provide a better option for the treatment of the malignancy together with underlying liver cirrhosis.25
Although HCC can present in a variety of ways, spontaneous rupture is the most dramatic and detrimental presentation.26 It often results in mortality, especially in patients with compromised liver function, and can be considered a terminal event in patients with advanced untreatable diseases. In our previous report26 on 33 patients with spontaneous rupture of HCC who underwent subsequent hepatic resection, the operative and survival outcomes were found to be inferior to those of patients without the history of tumor rupture. In the present study, the hospital mortality was 26.3% among the 19 patients with preexisting tumor rupture. In addition to the potential negative effects on hemodynamic, liver, and renal function status, the implication of preexisting tumor rupture on the technical difficulties during hepatic resection should not be underestimated. Concealed hematoma was often identified on laparotomy between the liver tumor and the hilar structures, hepatic flexure of the colon, or the right side of the diaphragm. Difficult mobilization of the tumor might be encountered, which could lead to intraoperative rerupture of the tumor and excessive operative blood loss. In patients with preexisting tumor rupture, we recommend the use of the anterior-approach technique, which we found to have favorable operative and survival outcomes in these patients in the latter part of our experience.15
The potential beneficial effects of anterior-approach hepatic resection on the operative and survival outcomes for HCC have previously been addressed.15 The technique involves initial vascular inflow control, completion of parenchymal transection, and complete venous outflow control before the right lobe is mobilized. It was considered to have beneficial effects in preserving the liver function of the liver remnant by avoiding warm ischemia of the latter related to pedicle torsion during mobilization of the huge tumor. Our previous retrospective analysis15 of 160 patients with large right lobe HCC suggested that the anterior-approach technique was associated with significantly less intraoperative blood loss and blood transfusion, a lower hospital mortality rate, and better disease-free and overall survival outcomes. The results of the present retrospective analysis also confirmed a lower hospital mortality rate in patients who underwent extended right-sided hepatic resection for HCC using the anterior-approach technique compared with the conventional approach. However, the theoretic advantages of the anterior approach over the conventional approach in patients with large HCCs have not been documented in a prospective manner. A prospective, randomized study is now being conducted in our institution to evaluate the potential benefits of the anterior approach compared with the conventional approach in major right-sided hepatic resection for large HCCs.
Extended right-sided hepatic resection for HCCs smaller than 10 cm have often implied the resection of a large amount of nontumorous liver parenchyma, leaving a relatively small liver remnant. It has been suggested that a small liver remnant can be related to adverse surgical outcomes.27 Yigitler et al28 demonstrated that the postoperative course after major hepatectomy was more difficult in patients with a small liver remnant, although a direct causal relationship between small liver remnant and postoperative complication rate was not demonstrated in their study. In a retrospective analysis of 434 patients undergoing hepatic resection for HCC and metastatic liver cancer during a 10-year period, Yamanaka et al10 suggested a predictive score for discriminating survivors from nonsurvivors. The authors identified the liver function status in terms of ICG clearance at 15 minutes, age, and parenchymal hepatic resection rate calculated by CT volumetry as the significant contributing factors that affect hospital mortality. Unfortunately, our study was limited by the retrospective nature of such a long study, and most of the liver resections were performed before the advent of CT volumetry. It was therefore not possible for us to include the future liver remnant volume on CT volumetry as a factor that affects hospital mortality or operative morbidity during analysis of our patients. In our current daily practice, most of the patients referred to us have had CT or magnetic resonance imaging performed in other hospitals, making it often difficult to obtain data on the future liver remnant volume or parenchymal hepatic resection rate unless the investigation is repeated in our institution. With the results of the present study, it would be helpful to identify patients with an increased risk of hospital mortality (tumor size <10 cm) and consider preoperative PVE to enhance the safety of the postoperative course.
Preoperative PVE was first advocated by Makuuchi et al29 to enhance the postoperative safety of patients who underwent extended hepatic resection for hepatobiliary malignant neoplasms. Although it has commonly been used in many centers, the exact role and indication of preoperative PVE for HCC still remain controversial.30,31 It appears that PVE is a safe procedure, effectively reducing operative mortality especially after extended hepatectomy or concomitant resection of major organs.32 However, diminished regenerative potential of the future liver remnant in patients with chronic liver diseases is still a concern. Farges et al24 suggested that the absence of hypertrophy of the contralateral lobe after PVE should be regarded as a contraindication for major hepatic resection.
In conclusion, the anterior approach is the preferred technique for extended right-sided hepatic resection of HCC. Increased risk of operative mortality was identified in patients who had a small tumor, which was associated with resection of a large volume of functioning liver parenchyma. Preoperative PVE should be considered in this group of patients to enhance atrophy of the right lobe and hypertrophy of the future liver remnant to minimize the operative risk, thereby making preoperative PVE a plausible strategy to expand the resectability of the liver tumor in patients with Child-Pugh grade B cirrhosis.8 Further prospective studies on the risk factors associated with extended hepatic resection for HCC should be conducted to evaluate the effects of the liver remnant and parenchymal resection volumes on the operative outcomes.
Correspondence: Chi Leung Liu, MS, MD, Department of Surgery, The University of Hong Kong, Queen Mary Hospital, 102 Pokfulam Rd, Hong Kong, China (firstname.lastname@example.org).
Accepted for Publication: November 11, 2005.
Author Contributions:Study concept and design: Liu, Fan, and Wong. Acquisition of data: Chik, Liu, Lo, Poon, and Lam. Analysis and interpretation of data: Chik and Liu. Drafting of the manuscript: Chik and Liu. Critical revision of the manuscript for important intellectual content: Fan, Lo, Poon, Lam, and Wong. Statistical analysis: Chik. Administrative, technical, and material support: Liu, Fan, and Wong. Study supervision: Fan and Wong.
Financial Disclosure: None reported.
Funding/Support: This study was supported by the Sun C.Y. Research Foundation for Hepatobiliary and Pancreatic Surgery of the University of Hong Kong.
Create a personal account or sign in to: