Hepatocellular carcinoma (HCC) is one of the most common malignancies. Jaundice usually occurs in the later stages of HCC and is caused by infiltrative tumor growth, severe cirrhosis, progressive liver failure, and/or other factors. The prognosis is usually poor and patients often die within a short period of time after disease onset. However, bile duct thrombosis (BDT) due to biliary tract invasion by HCC is very rare. Previous studies have shown that only 1%-12% of patients with HCC exhibit obstructive jaundice as the initial complaint.1 Surgical treatment of HCC with BDT has increasingly been reported in recent years, and hepatectomy with the removal of BDT can achieve satisfactory long-term palliation, and sometimes cure.2–4 However, few reports have evaluated the efficacy of re-operation for relapse after initial surgery for HCC with BDT. In this study, we analyzed the surgical treatment of 28 HCC patients with BDT, among whom six patients received re-operation, and compared their outcomes with those of 852 patients without BDT.
In total, 880 patients underwent radical resection of HCC at our department between January 1998 and August 2008, and were included in this study. Of these, 28 patients also presented with BDT, including six patients who received re-operation for post-operative relapse. During the same period, due to poor liver function (Child-Pugh grade C), accompanying extrahepatic metastases, or diffuse intrahepatic metastases, 14 HCC patients with BDT received palliative treatment only, which included medical treatment (n=2), percutaneous transhepatic biliary drainage (PTBD) (n=3), endoscopic retrograde cholangiopancreatography (ERCP) (n=2), and thrombectomy through choledochotomy with T-tube drainage (n=7). The surgical resection rate for HCC with BDT was 66.7% (28/42).
Preoperative examination and surgical procedures
All patients underwent color Doppler ultrasound, computerized tomography (CT)/hepatic artery angiography, and magnetic resonance imaging (MRI) to determine the location of tumors, and color Doppler ultrasound, CT/percutaneous transhepatic cholangiography (PTC), ERCP, and magnetic resonance cholangiopancreatography (MRCP) to determine the location and extent of the tumor emboli. The clinical typing of BDT was performed using the criteria proposed by Ueda et al.2 Hepatic function was evaluated using the Child-Pugh grading criteria. The surgical procedures were selected based on the preoperative diagnosis of the primary lesions, liver function grade, and types of tumor emboli.
The number and size of tumor(s) were measured by examining the resected specimen(s). Pathological diagnosis on the resected specimens was performed by two pathologists, and included the pathological type of the tumors, differentiation level, involvement of the portal vein, hepatic vein, and/or bile duct, growth pattern of tumors, residual tumor at the cut margins, and invasion of tumor emboli into the extrahepatic biliary duct wall.
All patients were followed up at our department every 1–2 months during the first post-operative year, and then every 1–6 months thereafter. At each follow-up, we performed serum alpha-fetoprotein (AFP) test and color Doppler ultrasound or CT. Patients who experienced post-operative tumor relapse were managed with ethanol injection, cool-tip radiofrequency ablation, transhepatic arterial chemotherapy embolization (TACE), or underwent re-operation. For patients with BDT recurrence, a re-operation was performed if the liver function was relatively good (Child-Pugh grade A or B), the tumor had no extrahepatic metastasis, and if the tumor was localized in one hemi-liver.
Data were analyzed using SPSS software (version 15.0, Chicago, IL, USA). The associations between clinicopathological parameters and accompanying BDT were analyzed with χ2 tests. Survival analysis was performed using the Kaplan-Meier method and inter-group differences were analyzed with Log-rank tests. A value of P <0.05 was considered statistically significant.
The 28 HCC patients with BDT comprised 22 men and six women, with a mean age of (49.0±7.7) years. Patients initially presented with jaundice (n=17, 60.7%), abdominal pain (n=10, 35.7%), fever (n=6, 21.4%), or were asymptomatic (n=5, 17.9%). Twenty-one patients (75.0%) had cirrhosis (Child-Pugh grade A, n=20; Child-Pugh grade B, n=8). Serological tests were positive for HBsAg in 22 (78.6%), anti-HCV in two (7.1%), and AFP (≥400 ng/ml) in 19 (67.9%) patients. The mean serum total bilirubin level was 126.8 mmol/L before the first surgery.
Fifteen patients underwent single-session surgical resection and experienced postoperative relapse, among whom 11 patients were positive for AFP preoperatively, and nine were positive for AFP after relapse. Of the four patients whose AFP was negative preoperatively, one patient was positive for AFP after relapse. Of the six patients who underwent re-operation, four patients were positive for AFP preoperatively and before re-operation. In the other two patients who were negative for AFP preoperatively, AFP remained negative before re-operation.
Comparing the two groups of patients with HCC (i.e, those with versus those without BDT), there were no significant differences between these groups in terms of age, sex, HBsAg positive rate, serum AFP level, liver function (Child-Pugh grade), or presence of cirrhosis (P >0.05) (Table 1).
In 17 patients with obstructive jaundice, the preoperative serum total bilirubin levels were >100 mmol/L in 13 patients, of whom 11 received PTBD and two received ERCP. For patients with preoperative serum total bilirubin levels ≤100 mmol/L, PTBD or ERCP was performed non-routinely. In two PTBD-treated patients and one ERCP-treated patient, the serum total bilirubin level increased after a transient decrease. In nine PTBD-treated and one ERCP-treated patient, the serum total bilirubin level was ≤100 mmol/L after treatment.
Out of 28 HCC patients with BDT, the diagnosis of the primary lesions was confirmed by preoperative ultrasound and CT in 25 patients, while in two patients, suspicious lesions were confirmed by hepatic artery angiography. In the remaining one patient, the primary lesion was not detected in either preoperative examinations or by intra-operative exploration. Two patients did not have BDT at the first operation. Of the remaining 26 patients, BDT was found in 13 patients (11 by PTC and 2 by ERCP). Color Doppler ultrasound was performed in all patients preoperatively and detected tumor emboli in 23 patients. MRCP was performed in 14 patients preoperatively and detected BDT in all of these patients. The predominant imaging feature of tumor emboli was a biliary filling defect or bile duct stenosis accompanied with proximal biliary dilatation. Accompanied portal vein tumor emboli were confirmed by preoperative color Doppler ultrasound in nine patients. Before surgery, one patient was misdiagnosed as having bile duct stones, and two patients were misdiagnosed as having portal cholangiocarcinoma.
Surgical procedures and follow-up results
Twenty-two patients underwent single-session surgery, including left hemihepatectomy (n=4), left hemihe- patectomy plus caudate lobe resection (n=2), right hemihepatectomy (n=4), hepatic lobe resection (n=8), hepatic segmentectomy or irregular liver resection (n=4), thrombectomy through choledochotomy (n=11), thrombectomy through bile duct stumps (n=6), complete resection of the invaded bile duct and tumor thrombi (n=3), and resection of the extra-hepatic bile duct plus Roux-en-Y-type hepatic cholangiojejunostomy (n=2). In nine patients with HCC and tumor thrombi inside the portal veins, the thrombi were removed through the opening of the involved portal vein stump. Of 28 patients with HCC and BDT, 21 experienced postoperative relapse, including relapse of HCC and tumor emboli in 11 patients, HCC relapse in seven patients, lung metastasis in two patients, and HCC relapse with lung metastasis in one patient. Intra-abdominal implantation metastasis was not found in any patient. The survival time of patients who underwent single-session surgery was 8–70 months.
Six patients received re-operations due to relapse of HCC or tumor emboli (Table 2). One patient received five sessions of surgery. In this patient, the first session, segmentectomy (segments VI and VII) was performed for HCC at segments VI and VII without BDT. In the second session, thrombectomy through choledochotomy was performed for tumor emboli at the lower end of common bile duct (CBD) in the absence of a primary hepatic lesion. Because the emboli were located at the lower end of CBD and were difficult to remove, the posterior peritoneum was dissected at the lateral side of duodenum to dissociate the duodenum and pancreatic head. By pressing the tumor emboli upward, the emboli were successfully separated from the bile duct wall and removed. Before the third session, the patient developed HCC in segments V and VIII, accompanied with tumor emboli in the right bile duct and CBD. ERCP was performed before surgical resection due to jaundice and acute cholangitis. However, because the jaundice was not satisfactorily relieved, right hemihepatectomy and thrombectomy through choledochotomy were performed. In the fourth session, thrombectomy through choledochotomy was performed because tumor emboli were detected in the common hepatic duct (CHD) and CBD, although no primary hepatic lesion was found. Before the fifth session, the patient developed HCC at the stump of segment IV, which was accompanied by tumor emboli in the CHD and CBD. ERCP was performed to remove the emboli from the bile duct, followed by endoscopic nasal biliary drainage, before surgical resection to treat the jaundice and acute cholangitis; however, because the symptoms were not improved by this approach, HCC resection, thrombectomy through choledochotomy, and right hepatic duct stump resection were performed (Figure 1). The patient has survived for 37 months after the first treatment.
One patient received three sessions of surgery. In the first session, segmentectomy (segment VII) was performed for HCC at segment VII. In the second session, segmentectomy (segments II and III) was performed for HCC at segments II and III. Before the third session, the patient developed HCC at segment IV, which was accompanied by tumor emboli in the CHD and CBD. ERCP was performed and a metal stent was implanted to treat the jaundice and acute cholangitis before surgical resection. However, because the symptoms were not improved, HCC resection and thrombectomy through choledochotomy were performed (Figure 2). This patient has survived for 23 months after the first treatment. Four patients received two sessions of surgery. Of these, two patients twice underwent hepatectomy plus BDT removal and they have survived for 61 and 44 months, respectively. In one patient whose primary lesion was not detected during the preoperative examinations, only thrombectomy through choledochotomy was performed in the first session. In this patient, the tumor relapsed 5 months after the initial surgery, prompting hepatectomy plus BDT removal. This patient has survived for 48 months. In the remaining patient, thrombectomy through choledochotomy was performed during the first session because of his poor hepatic function. Three months later, his hepatic function was restored and he underwent hepatectomy, removal of the tumor emboli in portal vein and BDT removal. About 22 months later, he died of diffuse intrahepatic metastasis and BDT recurrence.
All patients tolerated the surgery well and no severe complications were noted. All patients discharged uneventfully. The mean operation time was (206±47) minutes and (215±63) minutes during the first and second sessions of surgery, respectively. The mean intra-operative blood loss was (735±320) ml and (983±231) ml, respectively. The mean operation time and intra-operative blood loss were not significantly different between the first and second sessions of surgery (P >0.05).
Among 22 patients who underwent single-session surgery and six patients who underwent multiple operations, the sizes of tumors during the first session of surgery ranged from 0 to 18 cm. In 21 cases, a single nodule was found and in six cases, multiple nodules were found, while no primary lesion was detected in one patient. All patients were pathologically diagnosed as HCC, including 13 cases of poorly differentiated HCC and 15 cases of moderately differentiated HCC. The size and number of HCC nodules were not significantly different between the 28 HCC patients with BDT and 852 HCC patients without BDT (P >0.05). However, the differentiation level, growth pattern (infiltration or expansive growth), and invasion of the portal veins were significantly different between the two groups (P <0.05, Table 1).
In HCC with BDT, the histological examinations revealed that the tumor tended to grow around the intra-hepatic small bile ducts and invade the small bile ducts; however, no invasion to the large bile duct at the hilum was observed. Similarly, histological examinations did not show clear tumor invasion to the extra-hepatic bile duct wall in two cases or the right hepatic duct stump in one patient who underwent re-operation (Figure 1). Most of the tumor emboli were mixtures of hepatoma cells, red blood cells, necrotic tissue, and fibrous tissue (Figures 1 and 2). However, in one patient, the biliary emboli were caused by thrombi, but no malignant tumor cells were identified. The smooth surface of these biliary emboli meant they could be easily removed from the bile duct wall. By applying the BDT clinical typing criteria proposed by Ueda et al,3 there were three cases of type I, eight cases of type II, 12 cases of type III, and three cases of type IV. No BDT was detected during the first session of surgery in two patients (Table 2).
The 1-, 3- and 5-year survival rates were 89.3%, 46.4%, and 21.4%, respectively, for the 28 patients with HCC and BDT, and were 91.4%, 52.9%, and 20.9% for the 852 patients with HCC without BDT (P >0.05, Figure 3). As stated above, six patients with HCC and BDT underwent multiple operations; For the patients with HCC and BDT, the survival time was significantly longer for those who underwent re-operation after relapse than for those who did not undergo re-operation after relapse (P <0.05) (Figure 4). During the same period, 14 patients with HCC and BDT received palliative care only. Of these, seven patients received non-surgical treatments and their mean survival time was 3.1 months. The other seven patients received single thrombectomy through choledochotomy with T-tube drainage and their mean survival time was 9.4 months.
For the 28 patients with HCC and BDT, univariate analysis showed that the tumor-free survival time and total survival time after the first session of surgery were not significantly correlated with age, sex, HBsAg positive rate, Child-Pugh grade, cirrhosis, serum AFP level, and types of tumor emboli (P >0.05), but were significantly correlated with tumor size, number, differentiation level, growth pattern and invasion to portal vein (P <0.05) (Table 3). Multivariate analysis revealed that invasion to the portal vein and tumor size were independent factors that influenced the tumor-free survival time and total survival time (P <0.05). Univariate and multivariate analyses showed that obstructive jaundice was not significantly associated with tumor-free survival time or total survival time (P >0.05) (Tables 3–5).
HCC rarely invades the biliary tract to cause BDT. Indeed, of 880 cases with HCC, only 28 developed BDT in our center over 10 years. HCC accompanied with jaundice is often suggestive of end-stage disease. At this point, the prognosis is usually poor. Except for tumor invasion or spreading into the liver parenchyma, progressive liver failure, severe cirrhosis, and tumor invasion or lymph node compression of the hilar bile duct and extrahepatic bile ducts, BDT is another cause of jaundice in patients with HCC. Recent studies have shown that appropriate treatment of HCC with BDT can relieve the symptoms and even cure the disease.2–4 Therefore, identifying the cause of jaundice is clinically important. In previous studies, the incidence of HCC with BDT ranged from 1%-5%.5–8
Because HCC with BDT is rare and there is often a lack of awareness of this condition, many cases are misdiagnosed, especially in patients without history of liver disease or in patients whose imaging examinations do not show primary liver lesions. Ultrasound, enhanced CT, hepatic artery angiography, and MRI can help to identify the primary liver lesions. Meanwhile, ultrasound, PTC, ERCP and MRCP are useful to observe the distribution of tumor emboli in the biliary tract. MRI, as a non-invasive examination, can identify the location and extent of primary tumor lesions. MRCP is superior to PTC and ERCP in observing the distribution of tumor emboli in biliary tract.9
It has been widely believed that the prognosis of HCC with BDT is poor. The low rate of surgical resection, which may be caused by fundamental liver diseases and obstructive jaundice, appear to be responsible for the low rate of surgical resection.10,11 Some authors3 have suggested that good prognosis is dependent on early diagnosis, before the biliary tract becomes completely obstructed. In recent years, along with the increased number of reports of HCC with BDT, the rate of hepatic resection has increased significantly.4 In this study, the surgical resection rate was 65.85% for HCC with BDT. However, it is still unclear whether preoperative PTBD and ERCP for jaundice reduction are necessary for HCC accompanied with obstructive jaundice. We believe that preoperative application of PTBD and ERCP should be based on the specific condition of each patient. If the tumor emboli are not removed or incompletely removed, the tumor emboli and blood clots can obstruct catheters, and the effectiveness of jaundice treatment will be poor. If PTBD and ERCP are applied preoperatively but with limited improvement in jaundice, surgery should be performed as early as possible to ensure prompt treatment. Of interest, in HCC patients with BDT, the survival rate is not significantly different between those with obstructive jaundice and those without obstructive jaundice.8,12 In our study, univariate and multivariate analyses indicated that obstructive jaundice was not significantly correlated with tumor-free survival time or total survival time (P >0.05). Therefore, obstructive jaundice caused by HCC with BDT should not contraindicate surgery and does not appear to influence prognosis.
Many therapies, including liver resection, tumor emboli removal, biliary tract drainage, TACE, intra-biliary duct stents, radiation, and ethanol injection, can be used for the management of HCC with BDT. However, hepatectomy with BDT removal remains the most effective treatment strategy. According to Peng et al,8 radical resection plus tumor emboli removal resulted in 1-, 3- and 5-year survival rates of 93.2%, 56.0% and 24.1%, respectively. In our study, 28 patients underwent this procedure, and the 1-, 3- and 5-year survival rates were 89.3%, 46.4% and 21.4%, respectively. Although the HCCs with versus without BDT were significantly different in terms of tumor differentiation, growth pattern (infiltration or expansive growth), and invasion of portal veins, the 1-, 3- and 5-year survival rates were not significantly different.
Satoh et al13 reported that the survival rate was not significantly different between patients who underwent bile duct resection and those who did not undergo bile duct resection. Meanwhile, Shiomi et al4 reported that BDT invasion of the large bile duct at the hilum was very rare. In our study, most tumor emboli contained hepatoma cells, red blood cells, necrotic tissue and fibrous tissue. Their smooth surfaces also meant they could be easily removed from the bile duct wall. According to our result, for most patients with HCC and BDT, hepatectomy plus BDT removal is usually sufficient to achieve radical treatment, and no bile duct resection or Whipple operation is usually required.
For patients with poor hepatic reserve function and who cannot tolerate hepatectomy or for those without primary lesions in the liver, thrombectomy through choledochotomy with T-tube drainage was performed. However, if the primary lesion is not resected, the tumor can continue to grow and might invade the biliary tract or formed new tumor emboli, deeming treatment to be ineffective. As described by Peng et al,8 the mean survival time of four patients who underwent this procedure was 4.3 months while, in our study, seven patients underwent this procedure and their mean survival time was 9.4 months. Non-surgical treatment was applied preoperatively in these patients and in patients whose hepatic function was poor and could not tolerate surgery. However, the effectiveness of these non-surgical therapies was poor. Our result was consistent with Luo et al.14
Previously, the main treatment option for HCC relapse was TACE. However, in recent years, a second surgical resection is increasingly being performed. A second session of hepatectomy has also been considered as the most effective approach for postoperative relapse of HCC, particularly for HCC without vascular invasion.15–17 However, few reports evaluated the efficacy of re-operation for HCC with BDT. In this study, among 28 HCC patients with BDT, 21 patients experienced relapse. Univariate analysis showed that tumor size and number, differentiation level, growth pattern, and invasion to the portal vein were significantly associated with postoperative tumor-free survival time and total survival time. Meanwhile, multivariate analyses showed that invasion to the portal vein and tumor size were independently associated with tumor-free survival time and total survival time. Most relapses were of intrahepatic HCC with tumor emboli, which may be because HCC has a tendency to invade intrahepatic small bile ducts. All six patients who underwent re-operation experienced relapse of intrahepatic HCC with tumor emboli. The relapse is close related to the multi-centric HCC. In addition, large tumor size, accompanying microvascular invasion, and residual intrahepatic micro-lesions after surgery can increase the risk of postoperative relapse. To effectively prevent relapse of HCC with BDT, the operation should be gentle to avoid crushing the tumors and the cut margins should be free of residual tumor tissue.
During the re-operation of the patients with HCC and BDT, intraoperative exploration did not show any intra-abdominal implantation of the tumors, indicating that, when the adjacent tissues are well protected, bile duct dissection does not increase the likelihood of tumor metastasis or intra-abdominal implantation. In addition, we found that, among HCC patients with BDT, the survival time was significantly longer for those who underwent re-operation after relapse than for those who did not undergo re-operation after relapse. In the six patients who underwent re-operation, the mean survival time was 39 months, the longest being 61 months. Therefore, relapse after initial surgery for HCC with BDT is not a contraindication for surgical treatment. If hepatic function and physical conditions are good, re-operation should be considered because of its potential to increase the patient's quality of life and survival time.
BDT emboli are often accompanied by obstructive jaundice. Without timely intervention, the hepatic function will deteriorate progressively, which is accompanied by marked decreases in survival time. Tumor emboli arise from the primary hepatic lesion and only resection of the primary lesions can prevent their release. Therefore, hepatectomy plus BDT removal should be the preferred approach for HCC with BDT. The extent of surgical resection during re-operation should be based on the patient's specific conditions. If the hepatic function is good and if the residual liver has undergone compensatory enlargement, regular hepatectomy or hemihepatectomy can be considered. Among the six patients who underwent re-operation in our study, hemi-hepatectomy was performed in two, without severe postoperative complications. By contrast, if the patient has undergone hemihepatectomy or if the hepatic function is poor, partial hepatectomy or exenteration may be considered.
BDT can occur in patients with a very small primary hepatic lesion, which may even be undetectable before and during surgery. Indeed, in our study, the primary hepatic lesion was not found in the first operation in one patient who underwent re-operation or before two sessions of surgeries in another patient who underwent five surgical sessions. In both cases, re-operation (hepatectomy) was performed after relapse. Therefore, for BDT without a detectable primary lesion in liver, patients should be closely followed up after BDT removal. Once the tumor and/or tumor emboli recur, radical resection should be performed as soon as possible. The prompt diagnosis of relapse should be based on clinical manifestations, imaging, and serum AFP levels. Relapse should be considered if a patient experiences jaundice and cholangitis after surgery. In particular, the AFP test is very useful for the early detection of tumor relapse.
In summary, hepatectomy plus BDT removal is the optimal surgical approach for HCC with BDT. Obstructive jaundice or relapse after the initial surgery for HCC with BDT should not contraindicate subsequent surgical treatment. Re-operation after relapse can provide good outcomes in patients with HCC plus BDT.
1. Tantawi B, Cherqui D, Tran van Nhieu J, Kracht M, Fagniez PL. Surgery for biliary obstruction by tumour thrombus in primary liver cancer. Br J Surg 1996; 83: 1522-1525.
2. Ueda M, Takeuchi T, Takayasu T, Takahashi K, Okamoto S, Tanaka A, et al. Classification and surgical treatment of hepatocellular carcinoma
(HCC) with bile duct thrombi. Hepatogastroenterology 1994; 41: 349-354.
3. Kew MC, Paterson AC. Unusual presentations of hepatocellular carcinoma
. Trop Gastroenterol 1985; 6: 10-22.
4. Shiomi M, Kamiya J, Nagino M, Uesaka K, Sano T, Hayakawa N. Hepatocellular carcinoma
with biliary tumor thrombi: aggressive operative approach after appropriate preoperative management. Surgery 2001; 129: 692-698.
5. Kojiro M, Kawabata K, Kawano Y, Shirai F, Takemoto N, Nakashima T. Hepatocellular carcinoma
presenting as intrabile duct tumor growth: a clinicopathologic study of 24 cases. Cancer 1982; 49: 2144-2147.
6. Lin TY, Chen KM, Chen YR, Lin WS, Wang TH, Sung JL. Icteric type hepatoma. Med Chir Dig 1975; 4: 267-270.
7. Lau WY, Leung JW, Li AK. Management of hepatocellular carcinoma
presenting as obstructive jaundice. Am J Surg 1990; 160: 280-282.
8. Peng BG, Liang LJ, Li SQ, Zhou F, Hua YP, Luo SM. Surgical treatment of hepatocellular carcinoma
with bile duct tumor thrombi. World J Gastroenterol 2005; 11: 3966-3969.
9. Gabata T, Terayama N, Kobayashi S, Sanada J, Kadoya M, Matsui O. MR imaging of hepatocellular carcinomas with biliary tumor thrombi. Abdom Imaging 2007; 32: 470-474.
10. Chen MF, Jan YY, Jeng LB, Hwang TL, Wang CS, Chen SC. Obstructive jaundice secondary to ruptured hepatocellular carcinoma
into the common bile duct. Surgical experiences of 20 cases. Cancer 1994; 73: 1335-1340.
11. Lau W, Leung K, Leung TW, Liew CT, Chan MS, Yu SC, et al. A logical approach to hepatocellular carcinoma
presenting with jaundice. Ann Surg 1997; 225: 281-285.
12. Peng SY, Wang JW, Liu YB, Cai XJ, Deng GL, Xu B, et al. Surgical intervention for obstructive jaundice due to biliary tumor thrombus in hepatocellular carcinoma
. World J Surg 2004; 28: 43-46.
13. Satoh S, Ikai I, Honda G, Okabe H, Takeyama O, Yamamoto Y, et al. Clinicopathologic evaluation of hepatocellular carcinoma
with bile duct thrombi. Surgery 2000; 128: 779-783.
14. Xiangji L, Weifeng T, Bin Y, Chen L, Xiaoqing J, Baihe Z, et al. Surgery of hepatocellular carcinoma
complicated with cancer thrombi in bile duct: efficacy for criteria for different therapy modalities. Langenbecks Arch Surg 2009; 394: 1033-1039.
15. Sugimachi K, Maehara S, Tanaka S, Shimada M, Sugimachi K. Repeat hepatectomy
is the most useful treatment for recurrent hepatocellular carcinoma
. J Hepatobiliary Pancreat Surg 2001; 8: 410-416.
16. Shimada M, Matsumata T, Taketomi A, Yamamoto K, Itasaka H, Sugimachi K. Repeat hepatectomy
for recurrent hepatocellular carcinoma
. Surgery 1994; 115: 703-706.
17. Itamoto T, Nakahara H, Amano H, Kohashi T, Ohdan H, Tashiro H, et al. Repeat hepatectomy
for recurrent hepatocellular carcinoma
. Surgery 2007; 141: 589-597.
Keywords:© 2010 Chinese Medical Association
bile duct thrombosis; hepatectomy; hepatocellular carcinoma