- Email: [email protected]
Prognostic Index for Patients with Hepatocellular Carcinoma Combined with Tumor Thrombosis in the Major Portal Vein
Prognostic Index for Patients with Hepatocellular Carcinoma Combined with Tumor Thrombosis in the Major Portal Vein Iwao Ikai, MD, Etsuro Hatano, MD, Suguru Hasegawa, MD, Hideaki Fujii, MD, Kojiro Taura, MD, Naoki Uyama, MD, Yasuyuki Shimahara, MD This study sought to analyze prognostic factors in patients with hepatocellular carcinoma and tumor thrombosis in the first branch or trunk of the portal vein, and to provide a prognostic index. STUDY DESIGN: We performed a retrospective cohort study of 78 consecutive patients with hepatocellular carcinoma and tumor thrombosis in the first branch or trunk of the portal vein who underwent liver resection. Multivariate analysis of survival used the Cox’s proportional hazard model. RESULTS: Median survival time and 3-year survival rate were 0.74 years and 21.7%, respectively. Six factors, ie, absence of ascites, average elimination rate constant of indocyanine green, prothrombin activity, serum albumin level, maximal tumor diameter, and blood loss at operation were univariately related to survival time. By multivariate analysis, absence of ascites (hazard ratio, 2.23; 95% confidence interval, 1.10 to 4.52; p ⫽ 0.027), prothrombin activity ⱖ 75% (hazard ratio, 2.37; confidence interval, 1.30 to 4.32; p ⫽ 0.005), and maximal tumor diameter ⬍ 5 cm (hazard ratio, 2.37; confidence interval, 1.14 to 4.94; p ⫽ 0.021) were independent prognostic factors with similar hazard ratios. We calculated a prognostic index from these factors as follows: (ascites: absent ⫽ 0, present ⫽ 1) ⫹ (prothrombin activity: ⱖ 75% ⫽ 0, ⬍ 75% ⫽ 1) ⫹ (maximal tumor diameter: ⬍ 5 cm ⫽ 0, ⱖ 5 cm ⫽ 1). This index provided good stratification ability (log-rank, p ⬍ 0.001). Median survival times for patients with prognostic index 0, 1, 2, and 3 were 5.6, 1.6, 0.5, and 0.1 years, respectively. CONCLUSION: This prognostic index is a useful for making appropriate treatment strategy decisions for patients with hepatocellular carcinoma and tumor thrombosis in the major portal vein. ( J Am Coll Surg 2006;202:431–438. © 2006 by the American College of Surgeons) BACKGROUND:
Hepatocellular carcinoma (HCC), one of the most common malignancies worldwide, can invade the intrahepatic vasculature, commonly affecting the portal vein. Portal vein invasion is a critical prognostic factor for HCC patients.1-3 On extension of the tumor thrombosis to the major portal vein, prognosis is extremely poor because of the likely spread of HCC intrahepatically through the portal vein.4 Obstruction of the portal trunk because of tumor thrombosis can also result in
portal hypertension, which can induce uncontrolled esophageal variceal bleeding. A recent report by the Liver Cancer Study Group of Japan demonstrated that the incidence of macroscopic tumor thrombosis of the portal vein was 16.7% in surgical cases and 65.8% in autopsy cases.5 Nonsurgical treatment modalities, such as local ablation therapy and transcatheter arterial chemoembolization (TACE), are unsuitable for patients with major vascular invasion. Because of the severe donor shortage and poor prognosis, these patients are not candidates for liver transplantation.6,7 Systemic chemotherapy often has disappointing outcomes.8 Llovet and colleagues9 reported that the median survival of untreated patients with portal thrombosis was only 2.7 months. On the other hand, recent improvements in surgical procedures and perioperative management have refined hepatic resection for patients
Competing Interests Declared: None. Received June 27, 2005; Revised September 28, 2005; Accepted November 8, 2005. From the Department of Gastroenterological Surgery, Kyoto University Graduate School of Medicine, Kyoto, Japan. Correspondence address: Iwao Ikai, MD, Department of Gastroenterological Surgery, Kyoto University Graduate School of Medicine, 54 Kawahara-cho, Shogoin, Sakyo-ku, Kyoto 606-8507, Japan. email: [email protected]
© 2006 by the American College of Surgeons Published by Elsevier Inc.
431
ISSN 1072-7515/06/$32.00 doi:10.1016/j.jamcollsurg.2005.11.012
432
Ikai et al
Hepatocellular Carcinoma and Tumor Thrombosis in Major Portal Vein
Abbreviations and Acronyms
HCC ⫽ hepatocellular carcinoma ICGK ⫽ average elimination rate constant of indocyanine green PT ⫽ prothrombin activity TACE ⫽ transcatheter arterial chemoembolization
with highly advanced HCC, and operative mortality decreased considerably in 1990.10 Liver resection with direct removal of tumor thrombosis in the portal vein might allow postoperative TACE. In our previous report detailing the results of liver resection for these patients, their outcomes were better than those other treatment modalities.10,11 Outcomes of HCC patients with tumor thrombosis in major vessels are still disappointing. To improve the outcomes of these patients, it is important to evaluate the potential prognostic factors for adequate surgical indication. In this study, we analyzed patients with HCC combined with tumor thrombosis of the first branch or trunk of portal vein, and provided a prognostic index to assess treatment strategies effectively. METHODS This study included patients with HCC accompanied by tumor thrombosis in the first branch or trunk of the portal vein, who had undergone liver resection between 1990 and 2002 at the Department of Gastroenterological Surgery, Kyoto University Hospital. Seventy-eight consecutive patients were enrolled in this study, including patients with extrahepatic metastases. Indication for liver resection was determined by preoperative imaging, according to these criteria: all tumors, including tumor thrombosis and extrahepatic metastasis when present, could be resected macroscopically or the residual tumors could be treated by perioperative modalities, such as intraoperative ablation therapy; or the primary tumor and tumor thrombosis in the portal vein could be removed to avoid life-threatening variceal breeding as a palliative operation. Written informed consent for the operation was obtained from all patients. Hepatic resection operative procedures consisted of 11 trisegmentectomies, 65 lobectomies, and 2 segmentectomies. Segments were defined according to Healey’s classification system.12 Tumor thrombosis in the first branch of the portal vein was removed by anatomic hepatic resection. Tumor thrombosis in the portal trunk was removed by open method under simple occlusion of
J Am Coll Surg
the portal trunk and both first branches of the portal vein, as tumor infiltration to the vascular endothelial layer is rare. The variables examined were classified as host-, cancer-, or operation-related factors. The host-related factors evaluated in this study were age, gender, hepatitis B virus surface antigen, antihepatitis C virus antibody, presence or absence of ascites, average elimination rate constant of indocyanine green (ICGK), prothrombin activity (PT), platelet count, serum bilirubin level, serum albumin level, serum ␣-fetoprotein level, and associated liver disease (normal liver, chronic hepatitis, or liver cirrhosis) assessed microscopically using the noncancerous regions of resected specimens. The evaluated cancer-related factors included tumor multiplicity, maximal tumor diameter, serosal invasion, tumor growth pattern (well or poorly demarcated border), capsular formation, portal vein invasion (either the first branch or portal trunk), hepatic vein invasion, bile duct invasion, tumor histologic grade, lymph node metastasis, and distant organ metastasis. We also examined the operationrelated factors of blood loss at operation; operation time; residual tumor, including tumors treated by ablation therapy at the time of operation; and a surgical-free margin ⬎ 1 cm. Cumulative survival rates were obtained using the Kaplan-Meier method. Differences in survival between groups were compared for each variable using a log-rank test. The starting point in the calculation of survival time was the date of operation. Deaths, including operative deaths and cancer- or liver-related deaths, were considered to be the end point. After univariate analysis of the factors affecting survival, only those variables that were pronounced were included in the multivariate analysis using Cox’s proportional hazard model. All patients were assigned a prognostic index value determined by a multivariate analysis of the independent prognostic factors. Stratification of the patients was performed based on the prognostic index. Median survival time and survival rates were calculated for patients in each index group. P values ⬍ 0.05 were considered to be statistically significant. Statistical analysis was carried out using SPSS software, version 11.0 (Statistical Package for the Social Sciences). RESULTS Patient profiles, detailed in Table 1, indicate that the ratio of male patients to female patients 2.7:1. Mean age
Vol. 202, No. 3, March 2006
Ikai et al
Hepatocellular Carcinoma and Tumor Thrombosis in Major Portal Vein
433
Table 1. Univariate Analysis of Patients with Hepatocellular Carcinoma and Tumor Thrombosis in the First Branch or Trunk of Portal Vein Variables/Category
n
Median survival time (y)
1y
Survival rate (%) 3y
5y
Overall Age (y) ⬍ 60 ⱖ 60 Gender Male Female HBVAg Negative Positive HCVAb Negative Positive Ascites Absent Present ICGK ⱖ 0.1 ⬍ 0.1 Prothrombin activity (%) ⱖ 75 ⬍ 75 Platelet (⫻103/mm3) ⱖ 150 ⬍ 150 Total bilirubin (mg/dL) ⱕ 1.0 ⬎ 1.0 Serum albumin (g/dL) ⱖ 3.5 ⬍ 3.5 AFP (ng/mL) ⱕ 1,000 ⬎ 1,000 No. of tumors Solitary Multiple Maximal tumor diameter (cm) ⬍5 ⱖ5 Serosal invasion Absent Present Growth pattern (tumor border) Well demarcated Poorly demarcated
78
0.74
45.7
21.7
10.9
42 36
0.92 0.62
46.6 44.7
18.8 27.7
7.5 18.5
0.67
57 21
0.62 1.01
43.6 52.2
21.8 17.1
9.3 17.1
0.48
54 24
0.74 0.86
49.4 37.1
21.5 21.2
21.5 0.0
0.27
42 36
0.86 0.74
45.7 45.8
22.7 20.3
4.5 20.3
0.80
61 17
1.01 0.47
50.6 29.4
27.0 0.0
13.5 0.0
0.012
52 25
1.30 0.41
58.1 20.0
26.3 0.0
13.1 0.0
0.002
53 25
1.22 0.38
56.8 24.0
23.9 20.0
15.9 —
0.003
41 37
0.52 0.86
47.5 43.4
25.2 18.1
12.6 9.0
0.92
51 27
1.01 0.57
50.9 37.0
23.9 18.5
13.6 —
0.32
61 17
1.18 0.49
54.8 12.9
26.7 —
13.3 —
0.001
42 36
0.74 0.86
49.4 41.5
31.7 8.4
18.1 0.0
0.17
21 57
0.57 0.86
48.5 44.8
26.9 19.2
20.2 4.8
0.47
21 57
2.76 0.57
66.0 38.1
45.3 13.2
30.2 5.3
0.009
51 27
1.05 0.52
51.8 34.6
24.5 18.5
12.2 9.2
0.15
38 40
0.58 0.86
40.8 50.0
20.4 22.7
15.3 5.7
0.88
Log-rank p value
434
Ikai et al
Hepatocellular Carcinoma and Tumor Thrombosis in Major Portal Vein
J Am Coll Surg
Table 1. (continued) Variables/Category
Capsule formation Absent Present Portal vein invasion First branch Portal trunk Hepatic vein invasion Absent Present Bile duct invasion Absent Present Histologic grade Well/moderate Poor/undifferentiated Lymph node metastasis Absent Present Distant organ metastasis Absent Present Blood loss at operation (mL) ⬍ 1,500 ⱖ 1,500 Operation time (min) ⬍ 400 ⱖ 400 Residual tumor No Yes Surgical margin (cm) ⱖ1 ⬍1 Associated liver disease NL/CH LC
n
Median survival time (y)
1y
Survival rate (%) 3y
5y
38 40
0.74 0.86
47.9 43.7
20.9 23.3
10.5 11.7
0.85
35 43
0.95 0.74
47.5 44.1
25.3 18.0
10.1 12.0
0.87
53 25
0.74 1.01
43.3 50.7
18.2 26.8
12.1 10.7
0.68
62 16
0.95 0.35
49.6 31.3
23.7 12.5
13.5 0.0
0.067
45 33
0.74 1.05
40.7 52.8
21.2 23.5
12.7 7.8
0.63
70 8
0.92 0.33
48.2 25.0
22.5 12.5
12.9 —
0.074
68 10
0.86 0.49
46.6 40.0
18.9 30.0
11.3 10.0
0.83
16 62
2.17 0.57
67.3 40.0
42.1 16.5
21.0 8.2
0.048
28 50
0.92 0.74
48.8 44.1
18.1 26.1
6.0 15.6
0.77
57 21
0.58 0.95
44.1 50.0
23.3 20.3
13.8 6.8
0.79
29 43
0.62 0.58
49.0 40.5
24.5 18.2
16.4 9.1
0.18
43 35
1.01 0.57
50.5 40.1
17.3 28.0
11.6 11.2
0.51
Log-rank p value
AFP, ␣-fetoprotein; CH, chronic hepatitis; HBVAg, hepatitis B virus surface antigen; HCVAb, antihepatitis C virus antibody; ICGK, average elimination rate constant of indocyanine green; LC, live cirrhosis; NL, normal liver.
of patients was 60.0 years old, ranging from 31 to 75 years old. Thirty-five of the 78 patients exhibited tumor thrombosis in the first portal branch, and 43 had tumor thrombosis in the portal trunk. Twenty-five and 16 patients had tumor thrombosis of the hepatic vein and bile duct, respectively. Sixteen patients had extrahepatic metastasis (both lymph node and adrenal gland in two patients, lymph node in six patients, adrenal gland in four patients, lung in three patients, and diaphragm in one
patient). Median followup period was 6.9 months (range 0.6 to 116.1 months). Twelve patients with metastases in the lymph node or adrenal gland, or both, underwent liver resection combined with extirpation of extrahepatic metastasis. Four patients with metastasis in the lung or diaphragm underwent liver resection as a palliative operation to avoid life-threatening variceal breeding. Twenty patients were without recurrence at the end of
Vol. 202, No. 3, March 2006
Ikai et al
Hepatocellular Carcinoma and Tumor Thrombosis in Major Portal Vein
Figure 1. Survival curve of patients with hepatocellular carcinoma accompanied by thrombosis in the first branch or trunk of the portal vein between 1990 and 2002 (n ⫽ 78). The 1-, 3-, and 5-y overall survival rates of hepatocellular carcinoma patients were 45.7%, 21.7%, and 10.9%, respectively.
the followup period. Intrahepatic recurrent tumor(s) were observed in 34 patients, distant organ metastases in 6 patients, and both of intrahepatic recurrence and distant organ metastases in 18 patients. Sixteen patients were alive, and 62 patients died at the time of analysis. Three patients died within 30 days; one as a result of pulmonary bleeding, and the other two from growth of extrahepatic metastases. Cancer- and liver-related deaths were in 46 and 10 patients, respectively. Three patients died from miscellaneous causes. The longest surviving patient, who exhibited HCC with metastasis of the left adrenal gland with tumor thrombosis in the inferior vena cava, died without recurrence 9.5 years after operation. The 1-, 3-, and 5-year overall survival rates of HCC patients were 45.7%, 21.7%, and 10.9%, respectively (Fig. 1, Table 1). Median survival time was 0.74 years (Table 1). Differences in survival between groups with each prognostic factor were evaluated using a log-rank test. Absence of ascites (p ⫽ 0.012), an ICGK ⱖ 0.1 (p ⫽ 0.002), PT value ⱖ 75% (p ⫽ 0.003), serum al-
435
bumin level ⱖ 3.5 g/dL (p ⫽ 0.001), a maximal tumor diameter ⬍ 5 cm (p ⫽ 0.009), and blood loss ⬍ 1,500 mL during the surgical procedure (p ⫽ 0.048) notably favored a longer survival. We conducted a Cox multivariate regression analysis to identify independent prognostic factors. Absence of ascites (hazard ratio, 2.23; 95% CI, 1.10 to 4.52, p ⫽ 0.027), PT (hazard ratio, 2.37; 95% CI, 1.30 to 4.32, p ⫽ 0.005) and the maximal tumor diameter (hazard ratio, 2.37; 95% CI, 1.14 to 4.94, p ⫽ 0.021) were notably related to survival (Table 2). ICGK, serum albumin levels, and blood loss at operation were not markedly associated with survival time in the multivariate analysis. To facilitate the clinical application of these findings, we calculated a prognostic index based on the hazard ratio derived from these three independent prognostic factors. We proposed a simple prognostic index, as each of three independent prognostic factors exhibited a similar hazard ratio, ranging from 2.2 to 2.4, as follows (ascites: absent ⫽ 0, present ⫽ 1) ⫹ (PT: ⱖ 75% ⫽ 0, ⬍ 75% ⫽ 1) ⫹ (maximal tumor diameter: ⬍ 5 cm ⫽ 0, ⱖ 5 cm ⫽ 1). This index was used to classify the patients into four groups, the survival curves for which are shown in Figure 2. Median survival times and 5-year survival rates for each prognostic index group are 5.6 years (95% CI, 0.69 to 10.43 years; n ⫽ 12) and 51.9% in patients with prognostic index 0, 1.6 years (95% CI, 0.76 to 2.43 years; n ⫽ 38) and 10.3% in patients with prognostic index 1, 0.5 years (95% CI, 0.31 to 0.68 years; n ⫽ 23) and 0% in patients with prognostic index 2, and 0.1 years (95% CI, 0.06 to 0.11 years; n ⫽ 5) and 0% in patients with prognostic index 3. We identified substantial differences in the survival between these four groups (p ⬍ 0.001). DISCUSSION Local ablation therapy is a widely used nonsurgical therapeutic modality. Such therapy is beneficial for treat-
Table 2. Multivariate Analysis Using Cox’s Proportional Hazard Model Variables
Ascites ICGK PT (%) Serum albumin (g/dL) Maximal tumor diameter (cm) Blood loss at operation (mL)
Category
Hazard ratio
95% CI
p Value
Present versus absent ⬍ 0.1 versus ⱖ 0.1 ⬍ 75 versus ⱖ 75 ⬍ 3.5 versus ⱖ 3.5 ⱖ 5 versus ⬍ 5 ⱖ 1,500 versus ⬍ 1,500
2.23 1.90 2.37 1.88 2.37 1.28
1.10–4.52 0.98–3.65 1.30–4.32 0.95–3.71 1.14–4.94 0.47–3.49
0.027 0.056 0.005 0.071 0.021 0.63
ICGK, average elimination rate constant of indocyanine green; PT, prothrombin activity.
436
Ikai et al
Hepatocellular Carcinoma and Tumor Thrombosis in Major Portal Vein
Figure 2. Survival curve of patients with hepatocellular carcinoma accompanied by thrombosis in the first branch or trunk of the portal vein for each prognostic index. Median survival times and 5-y survival rates for each prognostic index group are 5.6 y (95% CI, 0.69 to 10.43; n ⫽ 12) and 51.9% in patients with prognostic index 0, 1.6 y (95% CI, 0.76 to 2.43; n ⫽ 38) and 10.3% in patients with prognostic index 1, 0.5 y (95% CI, 0.31 to 0.68; n ⫽ 23) and 0% in patients with prognostic index 2, and 0.1 y and 0% (95% CI, 0.06 to 0.11; n ⫽ 5) in patients with prognostic index 3. A considerable difference was seen in survival between these four groups (p ⬍ 0.001).
ment of a limited number of small HCC cases without vascular invasion, but was ineffective in a population of patients with large lesions and portal vein tumor thrombosis.13 TACE, another option for a nonsurgical modality, is usually contraindicated for patients with portal obstruction because of the high risk of hepatic insufficiency.14 Recently, Lee and colleagues15 reported that TACE can be a safe treatment modality for nodular HCC with portal trunk obstruction when patients exhibit good hepatic functional reserve and sufficient collateral circulation around the portal trunk. This report could not identify any improvements in survival for patients treated by TACE. In the treatment algorithm proposed by the Barcelona Clinic Liver Cancer staging system, the only available treatments for patients with vascular invasion are either palliative or investigational.3 We previously reported the results of surgical resection with removal of tumor thrombosis10,11; the 5-year survival rate of HCC patients with tumor thrombosis was 26%, 12%, and 7% for patients with a tumor thrombosis in the second branch of the portal vein, the first branch of the portal vein, or the portal trunk, respectively, with low operative mortality.11 In a subset of patients, surgical resection lead to increased survival, extending more than 5 years in some patients, although the degree of portal vein tumor thrombosis considerably af-
J Am Coll Surg
fected their survival.11 We proposed surgical resection as a standard therapy for HCC patients with tumor thromboses in the major portal vein.10 This study sought to aid in the selection of appropriate candidates for hepatic resection, as the survival time of patients with macroscopic portal vein invasion is still limited. In this study, 6 factors of the 27 potential prognostic factors, ie, absence of ascites, ICGK, PT, serum albumin level, maximal tumor diameter, and blood loss at operation, were univariately related to prognosis. Neither extrahepatic metastasis, including distant organ and lymph node, nor residual tumor affected their prognosis. Intrahepatic recurrence was observed in 66.7% (52 of 78) of patients. These results suggested that tumor thrombosis in the portal vein might induce intrahepatic spreading of HCC. By multivariate analysis, only three factors—ascites, PT, and maximal tumor diameter— were independent prognostic factors. We proposed, as each displayed a similar hazard ratio, a simple prognostic index using these factors. This prognostic index exhibited good stratification ability. The 5-year survival of patients with a tumor ⬍ 5 cm in diameter and a PT ⱖ 75% without ascites (prognostic index ⫽ 0) was 51.9%. Such patients with the most favorable index are good candidates for surgical modality. In addition, 1- and 3-year survival rates of patients with a prognostic index of 1 were 60.5% and 27.7%, respectively, and some patients survived for more than 5 years. These patients are also candidates for surgical modality. On the other hand, prognosis of patients with an unfavorable prognostic index was extremely poor. Mean survival time was ⱕ 0.5 years. New therapeutic strategies should be established for these patients. So far, only a number of extensive studies have evaluated the prognostic factors for patients with HCC accompanied by tumor thrombosis in the major portal vein.16-19 Minagawa and colleagues16 reported better survival in patients with tumor thrombosis in or beyond the second branch of the portal vein treated in the subset of patients with TACE: no more than two primary nodules, no portal trunk occlusion, and an indocyanine green retention rate at 15 minutes of better than 20%. The selection of patients for hepatic resection is a critical process; in Minagawa’s report, the 5-year survival of patients who met the criteria and underwent hepatic resection was 42%. Sixty percent of patients did not undergo hepatic resection; their mean survival was only 4.3 months. Outcomes for these patients must be improved. In our study, 55% of patients exhibited tumor
Vol. 202, No. 3, March 2006
Ikai et al
Hepatocellular Carcinoma and Tumor Thrombosis in Major Portal Vein
thrombosis in the portal trunk, and there were no marked differences in survival between patients with tumor thrombosis in the first branch and the portal trunk. TACE is not applicable for patients with tumor thrombosis in the portal trunk because of portal obstruction. Using univariate analysis, Ohkubo and colleagues17 reported that, for a patient with a tumor diameter ⬍ 10 cm and no intrahepatic metastasis, curative resection can lead to favorable prognosis in patients with HCC infiltrating the second branch or beyond of the portal vein. Severity of liver disease had little effect on their prognosis. In that study, it might be difficult to evaluate severity of liver disease, because ⬎ 90% of patients exhibited Child-Pugh class A disease. Current study excluded patients with tumor thrombosis in the second branch of the portal vein, because survival of these patients was considerably better than that of patients with tumor thrombosis in the first branch or the portal trunk.11 And patients in Child-Pugh class A were 73% of the total patients (number of patients in ChildPugh classification: A, 57; B, 20; C, 1). In this study, tumor size was one of the independent prognostic factors, but tumor multiplicity did not affect patient’s prognosis. Recently, Pawlik and colleagues18 reported only moderate to severe fibrosis was an independent predictor of poor survival for patients with HCC and major portal or hepatic vein invasion. Our series excluded patients with hepatic vein invasion only, because prognosis of patients with tumor thrombosis invading the inferior vena cava was extremely poor and their survival was much worse than those with hepatic vein tumor thrombosis.10 We considered these patients should be evaluated as another patient’s subset. Inconsistency between these three reports and our study might depend on patient population subjected to surgical resection. In this study, absence of ascites , PT, and tumor diameter were the independent prognostic factors that represented portal hypertension, liver function, and tumor status, respectively. The simple prognostic index using these three factors had a good discriminate ability for patients with major portal vein invasion. In conclusion, we analyzed the prognostic factors for patients with tumor thrombosis in the major portal vein and proposed a prognostic index. The prognostic index can be useful to select adequate treatment modalities for, and improve survival of, patients with HCC accompanied by a major portal invasion.
437
Author Contributions
Study conception and design: Ikai Acquisition of data: Ikai, Hatano, Hasegawa, Fujii, Taura, Uyama, Shimahara Analysis and interpretation of data: Ikai Drafting of manuscript: Ikai Critical revision: Hatano Statistical expertise: Hasegawa REFERENCES 1. Ikai I, Arii S, Kojiro M, et al. Re-evaluation of prognostic factors for survival after liver resection in patients with hepatocellular carcinoma in a Japanese nationwide survey. Cancer 2004;101: 796–802. 2. Arii S, Tanaka J, Yamazoe Y, et al. Predictive factors for intrahepatic recurrence of hepatocellular carcinoma after partial hepatectomy. Cancer 1992;69:913–919. 3. Llovet JM, Bru C, Bruix J. Prognosis of hepatocellular carcinoma: the BCLC staging classification. Semin Liver Dis 1999; 19:329–338. 4. Tsai TJ, Chau GY, Lui WY, et al. Clinical significance of microscopic tumor venous invasion in patients with resectable hepatocellular carcinoma. Surgery 2000;127:603–608. 5. Ikai I, Itai Y, Okita K, et al. Report of the 15th follow-up survey of primary liver cancer. Hepatol Res 2004;28:21–29. 6. Mazzaferro V, Regalia E, Doci R, et al. Liver transplantation for the treatment of small hepatocellular carcinomas in patients with cirrhosis. N Engl J Med 1996;334:693–699. 7. Iwatsuki S, Dvorchik I, Marsh JW, et al. Liver transplantation for hepatocellular carcinoma: a proposal of a prognostic scoring system. J Am Coll Surg 2000;191:389–394. 8. Okada S, Okazaki N, Nose H, et al. Prognostic factors in patients with hepatocellular carcinoma receiving systemic chemotherapy. Hepatology 1992;16:112–117. 9. Llovet JM, Bustamante J, Castells A, et al. Natural history of untreated nonsurgical hepatocellular carcinoma: rationale for the design and evaluation of therapeutic trials. Hepatology 1999;29:62–67. 10. Ikai I, Yamaoka Y, Yamamoto Y, et al. Surgical intervention for patients with stage IV—A hepatocellular carcinoma without lymph node metastasis: proposal as a standard therapy. Ann Surg 1998;227:433–439. 11. Ikai I, Yamamoto Y, Yamamoto N, et al. Results of hepatic resection for hepatocellular carcinoma invading major portal and/or hepatic veins. Surg Oncol Clin N Am 2003;12:65–75. 12. Healey JE Jr, Schroy PC. Anatomy of the biliary ducts within the human liver: analysis of the prevailing pattern of branching and the major variations of biliary duct. Arch Surg 1953;66:599–616. 13. Allgaier HP, Deibert P, Olschewski M, et al. Survival benefit of patients with inoperable hepatocellular carcinoma treated by a combination of transarterial chemoembolization and percutaneous ethanol injection—a single-center analysis including 132 patients. Int J Cancer 1998;79:601–605. 14. Yamada R, Kishi K, Sato M, et al. Transcatheter arterial chemoembolization (TACE) in the treatment of unresectable liver cancer. World J Surg 1995;19:795–800. 15. Lee HS, Kim JS, Choi IJ, et al. The safety and efficacy of transcatheter arterial chemoembolization in the treatment of patients with
438
Ikai et al
Hepatocellular Carcinoma and Tumor Thrombosis in Major Portal Vein
hepatocellular carcinoma and main portal vein obstruction. A prospective controlled study. Cancer 1997;79:2087–2094. 16. Minagawa M, Makuuchi M, Takayama T, et al. Selection criteria for hepatectomy in patients with hepatocellular carcinoma and portal vein tumor thrombus. Ann Surg 2001;233:379–384. 17. Ohkubo T, Yamamoto J, Sugawara Y, et al. Surgical results for hepatocellular carcinoma with macroscopic portal vein tumor thrombosis. J Am Coll Surg 2000;191:657–660.
J Am Coll Surg
18. Pawlik TM, Poon RT, Abdalla EK, et al. Hepatectomy for hepatocellular carcinoma with major portal or hepatic vein invasion: results of a multicenter study. Surgery 2005;137: 403–410. 19. Wu CC, Hsieh SR, Chen JT, et al. An appraisal of liver and portal vein resection for hepatocellular carcinoma with tumor thrombi extending to portal bifurcation. Arch Surg 2000;135: 1273–1279.
JACS ONLINE CME-1 PROGRAM CME-1 COURSE EVERY MONTH 2 CREDITS AVAILABLE EACH MONTH 48 CREDITS AVAILABLE AT ALL TIMES www.jacscme.facs.org
Hepatocellular carcinoma (HCC), one of the most common malignancies worldwide, can invade the intrahepatic vasculature, commonly affecting the portal vein. Portal vein invasion is a critical prognostic factor for HCC patients.1-3 On extension of the tumor thrombosis to the major portal vein, prognosis is extremely poor because of the likely spread of HCC intrahepatically through the portal vein.4 Obstruction of the portal trunk because of tumor thrombosis can also result in
portal hypertension, which can induce uncontrolled esophageal variceal bleeding. A recent report by the Liver Cancer Study Group of Japan demonstrated that the incidence of macroscopic tumor thrombosis of the portal vein was 16.7% in surgical cases and 65.8% in autopsy cases.5 Nonsurgical treatment modalities, such as local ablation therapy and transcatheter arterial chemoembolization (TACE), are unsuitable for patients with major vascular invasion. Because of the severe donor shortage and poor prognosis, these patients are not candidates for liver transplantation.6,7 Systemic chemotherapy often has disappointing outcomes.8 Llovet and colleagues9 reported that the median survival of untreated patients with portal thrombosis was only 2.7 months. On the other hand, recent improvements in surgical procedures and perioperative management have refined hepatic resection for patients
Competing Interests Declared: None. Received June 27, 2005; Revised September 28, 2005; Accepted November 8, 2005. From the Department of Gastroenterological Surgery, Kyoto University Graduate School of Medicine, Kyoto, Japan. Correspondence address: Iwao Ikai, MD, Department of Gastroenterological Surgery, Kyoto University Graduate School of Medicine, 54 Kawahara-cho, Shogoin, Sakyo-ku, Kyoto 606-8507, Japan. email: [email protected]
© 2006 by the American College of Surgeons Published by Elsevier Inc.
431
ISSN 1072-7515/06/$32.00 doi:10.1016/j.jamcollsurg.2005.11.012
432
Ikai et al
Hepatocellular Carcinoma and Tumor Thrombosis in Major Portal Vein
Abbreviations and Acronyms
HCC ⫽ hepatocellular carcinoma ICGK ⫽ average elimination rate constant of indocyanine green PT ⫽ prothrombin activity TACE ⫽ transcatheter arterial chemoembolization
with highly advanced HCC, and operative mortality decreased considerably in 1990.10 Liver resection with direct removal of tumor thrombosis in the portal vein might allow postoperative TACE. In our previous report detailing the results of liver resection for these patients, their outcomes were better than those other treatment modalities.10,11 Outcomes of HCC patients with tumor thrombosis in major vessels are still disappointing. To improve the outcomes of these patients, it is important to evaluate the potential prognostic factors for adequate surgical indication. In this study, we analyzed patients with HCC combined with tumor thrombosis of the first branch or trunk of portal vein, and provided a prognostic index to assess treatment strategies effectively. METHODS This study included patients with HCC accompanied by tumor thrombosis in the first branch or trunk of the portal vein, who had undergone liver resection between 1990 and 2002 at the Department of Gastroenterological Surgery, Kyoto University Hospital. Seventy-eight consecutive patients were enrolled in this study, including patients with extrahepatic metastases. Indication for liver resection was determined by preoperative imaging, according to these criteria: all tumors, including tumor thrombosis and extrahepatic metastasis when present, could be resected macroscopically or the residual tumors could be treated by perioperative modalities, such as intraoperative ablation therapy; or the primary tumor and tumor thrombosis in the portal vein could be removed to avoid life-threatening variceal breeding as a palliative operation. Written informed consent for the operation was obtained from all patients. Hepatic resection operative procedures consisted of 11 trisegmentectomies, 65 lobectomies, and 2 segmentectomies. Segments were defined according to Healey’s classification system.12 Tumor thrombosis in the first branch of the portal vein was removed by anatomic hepatic resection. Tumor thrombosis in the portal trunk was removed by open method under simple occlusion of
J Am Coll Surg
the portal trunk and both first branches of the portal vein, as tumor infiltration to the vascular endothelial layer is rare. The variables examined were classified as host-, cancer-, or operation-related factors. The host-related factors evaluated in this study were age, gender, hepatitis B virus surface antigen, antihepatitis C virus antibody, presence or absence of ascites, average elimination rate constant of indocyanine green (ICGK), prothrombin activity (PT), platelet count, serum bilirubin level, serum albumin level, serum ␣-fetoprotein level, and associated liver disease (normal liver, chronic hepatitis, or liver cirrhosis) assessed microscopically using the noncancerous regions of resected specimens. The evaluated cancer-related factors included tumor multiplicity, maximal tumor diameter, serosal invasion, tumor growth pattern (well or poorly demarcated border), capsular formation, portal vein invasion (either the first branch or portal trunk), hepatic vein invasion, bile duct invasion, tumor histologic grade, lymph node metastasis, and distant organ metastasis. We also examined the operationrelated factors of blood loss at operation; operation time; residual tumor, including tumors treated by ablation therapy at the time of operation; and a surgical-free margin ⬎ 1 cm. Cumulative survival rates were obtained using the Kaplan-Meier method. Differences in survival between groups were compared for each variable using a log-rank test. The starting point in the calculation of survival time was the date of operation. Deaths, including operative deaths and cancer- or liver-related deaths, were considered to be the end point. After univariate analysis of the factors affecting survival, only those variables that were pronounced were included in the multivariate analysis using Cox’s proportional hazard model. All patients were assigned a prognostic index value determined by a multivariate analysis of the independent prognostic factors. Stratification of the patients was performed based on the prognostic index. Median survival time and survival rates were calculated for patients in each index group. P values ⬍ 0.05 were considered to be statistically significant. Statistical analysis was carried out using SPSS software, version 11.0 (Statistical Package for the Social Sciences). RESULTS Patient profiles, detailed in Table 1, indicate that the ratio of male patients to female patients 2.7:1. Mean age
Vol. 202, No. 3, March 2006
Ikai et al
Hepatocellular Carcinoma and Tumor Thrombosis in Major Portal Vein
433
Table 1. Univariate Analysis of Patients with Hepatocellular Carcinoma and Tumor Thrombosis in the First Branch or Trunk of Portal Vein Variables/Category
n
Median survival time (y)
1y
Survival rate (%) 3y
5y
Overall Age (y) ⬍ 60 ⱖ 60 Gender Male Female HBVAg Negative Positive HCVAb Negative Positive Ascites Absent Present ICGK ⱖ 0.1 ⬍ 0.1 Prothrombin activity (%) ⱖ 75 ⬍ 75 Platelet (⫻103/mm3) ⱖ 150 ⬍ 150 Total bilirubin (mg/dL) ⱕ 1.0 ⬎ 1.0 Serum albumin (g/dL) ⱖ 3.5 ⬍ 3.5 AFP (ng/mL) ⱕ 1,000 ⬎ 1,000 No. of tumors Solitary Multiple Maximal tumor diameter (cm) ⬍5 ⱖ5 Serosal invasion Absent Present Growth pattern (tumor border) Well demarcated Poorly demarcated
78
0.74
45.7
21.7
10.9
42 36
0.92 0.62
46.6 44.7
18.8 27.7
7.5 18.5
0.67
57 21
0.62 1.01
43.6 52.2
21.8 17.1
9.3 17.1
0.48
54 24
0.74 0.86
49.4 37.1
21.5 21.2
21.5 0.0
0.27
42 36
0.86 0.74
45.7 45.8
22.7 20.3
4.5 20.3
0.80
61 17
1.01 0.47
50.6 29.4
27.0 0.0
13.5 0.0
0.012
52 25
1.30 0.41
58.1 20.0
26.3 0.0
13.1 0.0
0.002
53 25
1.22 0.38
56.8 24.0
23.9 20.0
15.9 —
0.003
41 37
0.52 0.86
47.5 43.4
25.2 18.1
12.6 9.0
0.92
51 27
1.01 0.57
50.9 37.0
23.9 18.5
13.6 —
0.32
61 17
1.18 0.49
54.8 12.9
26.7 —
13.3 —
0.001
42 36
0.74 0.86
49.4 41.5
31.7 8.4
18.1 0.0
0.17
21 57
0.57 0.86
48.5 44.8
26.9 19.2
20.2 4.8
0.47
21 57
2.76 0.57
66.0 38.1
45.3 13.2
30.2 5.3
0.009
51 27
1.05 0.52
51.8 34.6
24.5 18.5
12.2 9.2
0.15
38 40
0.58 0.86
40.8 50.0
20.4 22.7
15.3 5.7
0.88
Log-rank p value
434
Ikai et al
Hepatocellular Carcinoma and Tumor Thrombosis in Major Portal Vein
J Am Coll Surg
Table 1. (continued) Variables/Category
Capsule formation Absent Present Portal vein invasion First branch Portal trunk Hepatic vein invasion Absent Present Bile duct invasion Absent Present Histologic grade Well/moderate Poor/undifferentiated Lymph node metastasis Absent Present Distant organ metastasis Absent Present Blood loss at operation (mL) ⬍ 1,500 ⱖ 1,500 Operation time (min) ⬍ 400 ⱖ 400 Residual tumor No Yes Surgical margin (cm) ⱖ1 ⬍1 Associated liver disease NL/CH LC
n
Median survival time (y)
1y
Survival rate (%) 3y
5y
38 40
0.74 0.86
47.9 43.7
20.9 23.3
10.5 11.7
0.85
35 43
0.95 0.74
47.5 44.1
25.3 18.0
10.1 12.0
0.87
53 25
0.74 1.01
43.3 50.7
18.2 26.8
12.1 10.7
0.68
62 16
0.95 0.35
49.6 31.3
23.7 12.5
13.5 0.0
0.067
45 33
0.74 1.05
40.7 52.8
21.2 23.5
12.7 7.8
0.63
70 8
0.92 0.33
48.2 25.0
22.5 12.5
12.9 —
0.074
68 10
0.86 0.49
46.6 40.0
18.9 30.0
11.3 10.0
0.83
16 62
2.17 0.57
67.3 40.0
42.1 16.5
21.0 8.2
0.048
28 50
0.92 0.74
48.8 44.1
18.1 26.1
6.0 15.6
0.77
57 21
0.58 0.95
44.1 50.0
23.3 20.3
13.8 6.8
0.79
29 43
0.62 0.58
49.0 40.5
24.5 18.2
16.4 9.1
0.18
43 35
1.01 0.57
50.5 40.1
17.3 28.0
11.6 11.2
0.51
Log-rank p value
AFP, ␣-fetoprotein; CH, chronic hepatitis; HBVAg, hepatitis B virus surface antigen; HCVAb, antihepatitis C virus antibody; ICGK, average elimination rate constant of indocyanine green; LC, live cirrhosis; NL, normal liver.
of patients was 60.0 years old, ranging from 31 to 75 years old. Thirty-five of the 78 patients exhibited tumor thrombosis in the first portal branch, and 43 had tumor thrombosis in the portal trunk. Twenty-five and 16 patients had tumor thrombosis of the hepatic vein and bile duct, respectively. Sixteen patients had extrahepatic metastasis (both lymph node and adrenal gland in two patients, lymph node in six patients, adrenal gland in four patients, lung in three patients, and diaphragm in one
patient). Median followup period was 6.9 months (range 0.6 to 116.1 months). Twelve patients with metastases in the lymph node or adrenal gland, or both, underwent liver resection combined with extirpation of extrahepatic metastasis. Four patients with metastasis in the lung or diaphragm underwent liver resection as a palliative operation to avoid life-threatening variceal breeding. Twenty patients were without recurrence at the end of
Vol. 202, No. 3, March 2006
Ikai et al
Hepatocellular Carcinoma and Tumor Thrombosis in Major Portal Vein
Figure 1. Survival curve of patients with hepatocellular carcinoma accompanied by thrombosis in the first branch or trunk of the portal vein between 1990 and 2002 (n ⫽ 78). The 1-, 3-, and 5-y overall survival rates of hepatocellular carcinoma patients were 45.7%, 21.7%, and 10.9%, respectively.
the followup period. Intrahepatic recurrent tumor(s) were observed in 34 patients, distant organ metastases in 6 patients, and both of intrahepatic recurrence and distant organ metastases in 18 patients. Sixteen patients were alive, and 62 patients died at the time of analysis. Three patients died within 30 days; one as a result of pulmonary bleeding, and the other two from growth of extrahepatic metastases. Cancer- and liver-related deaths were in 46 and 10 patients, respectively. Three patients died from miscellaneous causes. The longest surviving patient, who exhibited HCC with metastasis of the left adrenal gland with tumor thrombosis in the inferior vena cava, died without recurrence 9.5 years after operation. The 1-, 3-, and 5-year overall survival rates of HCC patients were 45.7%, 21.7%, and 10.9%, respectively (Fig. 1, Table 1). Median survival time was 0.74 years (Table 1). Differences in survival between groups with each prognostic factor were evaluated using a log-rank test. Absence of ascites (p ⫽ 0.012), an ICGK ⱖ 0.1 (p ⫽ 0.002), PT value ⱖ 75% (p ⫽ 0.003), serum al-
435
bumin level ⱖ 3.5 g/dL (p ⫽ 0.001), a maximal tumor diameter ⬍ 5 cm (p ⫽ 0.009), and blood loss ⬍ 1,500 mL during the surgical procedure (p ⫽ 0.048) notably favored a longer survival. We conducted a Cox multivariate regression analysis to identify independent prognostic factors. Absence of ascites (hazard ratio, 2.23; 95% CI, 1.10 to 4.52, p ⫽ 0.027), PT (hazard ratio, 2.37; 95% CI, 1.30 to 4.32, p ⫽ 0.005) and the maximal tumor diameter (hazard ratio, 2.37; 95% CI, 1.14 to 4.94, p ⫽ 0.021) were notably related to survival (Table 2). ICGK, serum albumin levels, and blood loss at operation were not markedly associated with survival time in the multivariate analysis. To facilitate the clinical application of these findings, we calculated a prognostic index based on the hazard ratio derived from these three independent prognostic factors. We proposed a simple prognostic index, as each of three independent prognostic factors exhibited a similar hazard ratio, ranging from 2.2 to 2.4, as follows (ascites: absent ⫽ 0, present ⫽ 1) ⫹ (PT: ⱖ 75% ⫽ 0, ⬍ 75% ⫽ 1) ⫹ (maximal tumor diameter: ⬍ 5 cm ⫽ 0, ⱖ 5 cm ⫽ 1). This index was used to classify the patients into four groups, the survival curves for which are shown in Figure 2. Median survival times and 5-year survival rates for each prognostic index group are 5.6 years (95% CI, 0.69 to 10.43 years; n ⫽ 12) and 51.9% in patients with prognostic index 0, 1.6 years (95% CI, 0.76 to 2.43 years; n ⫽ 38) and 10.3% in patients with prognostic index 1, 0.5 years (95% CI, 0.31 to 0.68 years; n ⫽ 23) and 0% in patients with prognostic index 2, and 0.1 years (95% CI, 0.06 to 0.11 years; n ⫽ 5) and 0% in patients with prognostic index 3. We identified substantial differences in the survival between these four groups (p ⬍ 0.001). DISCUSSION Local ablation therapy is a widely used nonsurgical therapeutic modality. Such therapy is beneficial for treat-
Table 2. Multivariate Analysis Using Cox’s Proportional Hazard Model Variables
Ascites ICGK PT (%) Serum albumin (g/dL) Maximal tumor diameter (cm) Blood loss at operation (mL)
Category
Hazard ratio
95% CI
p Value
Present versus absent ⬍ 0.1 versus ⱖ 0.1 ⬍ 75 versus ⱖ 75 ⬍ 3.5 versus ⱖ 3.5 ⱖ 5 versus ⬍ 5 ⱖ 1,500 versus ⬍ 1,500
2.23 1.90 2.37 1.88 2.37 1.28
1.10–4.52 0.98–3.65 1.30–4.32 0.95–3.71 1.14–4.94 0.47–3.49
0.027 0.056 0.005 0.071 0.021 0.63
ICGK, average elimination rate constant of indocyanine green; PT, prothrombin activity.
436
Ikai et al
Hepatocellular Carcinoma and Tumor Thrombosis in Major Portal Vein
Figure 2. Survival curve of patients with hepatocellular carcinoma accompanied by thrombosis in the first branch or trunk of the portal vein for each prognostic index. Median survival times and 5-y survival rates for each prognostic index group are 5.6 y (95% CI, 0.69 to 10.43; n ⫽ 12) and 51.9% in patients with prognostic index 0, 1.6 y (95% CI, 0.76 to 2.43; n ⫽ 38) and 10.3% in patients with prognostic index 1, 0.5 y (95% CI, 0.31 to 0.68; n ⫽ 23) and 0% in patients with prognostic index 2, and 0.1 y and 0% (95% CI, 0.06 to 0.11; n ⫽ 5) in patients with prognostic index 3. A considerable difference was seen in survival between these four groups (p ⬍ 0.001).
ment of a limited number of small HCC cases without vascular invasion, but was ineffective in a population of patients with large lesions and portal vein tumor thrombosis.13 TACE, another option for a nonsurgical modality, is usually contraindicated for patients with portal obstruction because of the high risk of hepatic insufficiency.14 Recently, Lee and colleagues15 reported that TACE can be a safe treatment modality for nodular HCC with portal trunk obstruction when patients exhibit good hepatic functional reserve and sufficient collateral circulation around the portal trunk. This report could not identify any improvements in survival for patients treated by TACE. In the treatment algorithm proposed by the Barcelona Clinic Liver Cancer staging system, the only available treatments for patients with vascular invasion are either palliative or investigational.3 We previously reported the results of surgical resection with removal of tumor thrombosis10,11; the 5-year survival rate of HCC patients with tumor thrombosis was 26%, 12%, and 7% for patients with a tumor thrombosis in the second branch of the portal vein, the first branch of the portal vein, or the portal trunk, respectively, with low operative mortality.11 In a subset of patients, surgical resection lead to increased survival, extending more than 5 years in some patients, although the degree of portal vein tumor thrombosis considerably af-
J Am Coll Surg
fected their survival.11 We proposed surgical resection as a standard therapy for HCC patients with tumor thromboses in the major portal vein.10 This study sought to aid in the selection of appropriate candidates for hepatic resection, as the survival time of patients with macroscopic portal vein invasion is still limited. In this study, 6 factors of the 27 potential prognostic factors, ie, absence of ascites, ICGK, PT, serum albumin level, maximal tumor diameter, and blood loss at operation, were univariately related to prognosis. Neither extrahepatic metastasis, including distant organ and lymph node, nor residual tumor affected their prognosis. Intrahepatic recurrence was observed in 66.7% (52 of 78) of patients. These results suggested that tumor thrombosis in the portal vein might induce intrahepatic spreading of HCC. By multivariate analysis, only three factors—ascites, PT, and maximal tumor diameter— were independent prognostic factors. We proposed, as each displayed a similar hazard ratio, a simple prognostic index using these factors. This prognostic index exhibited good stratification ability. The 5-year survival of patients with a tumor ⬍ 5 cm in diameter and a PT ⱖ 75% without ascites (prognostic index ⫽ 0) was 51.9%. Such patients with the most favorable index are good candidates for surgical modality. In addition, 1- and 3-year survival rates of patients with a prognostic index of 1 were 60.5% and 27.7%, respectively, and some patients survived for more than 5 years. These patients are also candidates for surgical modality. On the other hand, prognosis of patients with an unfavorable prognostic index was extremely poor. Mean survival time was ⱕ 0.5 years. New therapeutic strategies should be established for these patients. So far, only a number of extensive studies have evaluated the prognostic factors for patients with HCC accompanied by tumor thrombosis in the major portal vein.16-19 Minagawa and colleagues16 reported better survival in patients with tumor thrombosis in or beyond the second branch of the portal vein treated in the subset of patients with TACE: no more than two primary nodules, no portal trunk occlusion, and an indocyanine green retention rate at 15 minutes of better than 20%. The selection of patients for hepatic resection is a critical process; in Minagawa’s report, the 5-year survival of patients who met the criteria and underwent hepatic resection was 42%. Sixty percent of patients did not undergo hepatic resection; their mean survival was only 4.3 months. Outcomes for these patients must be improved. In our study, 55% of patients exhibited tumor
Vol. 202, No. 3, March 2006
Ikai et al
Hepatocellular Carcinoma and Tumor Thrombosis in Major Portal Vein
thrombosis in the portal trunk, and there were no marked differences in survival between patients with tumor thrombosis in the first branch and the portal trunk. TACE is not applicable for patients with tumor thrombosis in the portal trunk because of portal obstruction. Using univariate analysis, Ohkubo and colleagues17 reported that, for a patient with a tumor diameter ⬍ 10 cm and no intrahepatic metastasis, curative resection can lead to favorable prognosis in patients with HCC infiltrating the second branch or beyond of the portal vein. Severity of liver disease had little effect on their prognosis. In that study, it might be difficult to evaluate severity of liver disease, because ⬎ 90% of patients exhibited Child-Pugh class A disease. Current study excluded patients with tumor thrombosis in the second branch of the portal vein, because survival of these patients was considerably better than that of patients with tumor thrombosis in the first branch or the portal trunk.11 And patients in Child-Pugh class A were 73% of the total patients (number of patients in ChildPugh classification: A, 57; B, 20; C, 1). In this study, tumor size was one of the independent prognostic factors, but tumor multiplicity did not affect patient’s prognosis. Recently, Pawlik and colleagues18 reported only moderate to severe fibrosis was an independent predictor of poor survival for patients with HCC and major portal or hepatic vein invasion. Our series excluded patients with hepatic vein invasion only, because prognosis of patients with tumor thrombosis invading the inferior vena cava was extremely poor and their survival was much worse than those with hepatic vein tumor thrombosis.10 We considered these patients should be evaluated as another patient’s subset. Inconsistency between these three reports and our study might depend on patient population subjected to surgical resection. In this study, absence of ascites , PT, and tumor diameter were the independent prognostic factors that represented portal hypertension, liver function, and tumor status, respectively. The simple prognostic index using these three factors had a good discriminate ability for patients with major portal vein invasion. In conclusion, we analyzed the prognostic factors for patients with tumor thrombosis in the major portal vein and proposed a prognostic index. The prognostic index can be useful to select adequate treatment modalities for, and improve survival of, patients with HCC accompanied by a major portal invasion.
437
Author Contributions
Study conception and design: Ikai Acquisition of data: Ikai, Hatano, Hasegawa, Fujii, Taura, Uyama, Shimahara Analysis and interpretation of data: Ikai Drafting of manuscript: Ikai Critical revision: Hatano Statistical expertise: Hasegawa REFERENCES 1. Ikai I, Arii S, Kojiro M, et al. Re-evaluation of prognostic factors for survival after liver resection in patients with hepatocellular carcinoma in a Japanese nationwide survey. Cancer 2004;101: 796–802. 2. Arii S, Tanaka J, Yamazoe Y, et al. Predictive factors for intrahepatic recurrence of hepatocellular carcinoma after partial hepatectomy. Cancer 1992;69:913–919. 3. Llovet JM, Bru C, Bruix J. Prognosis of hepatocellular carcinoma: the BCLC staging classification. Semin Liver Dis 1999; 19:329–338. 4. Tsai TJ, Chau GY, Lui WY, et al. Clinical significance of microscopic tumor venous invasion in patients with resectable hepatocellular carcinoma. Surgery 2000;127:603–608. 5. Ikai I, Itai Y, Okita K, et al. Report of the 15th follow-up survey of primary liver cancer. Hepatol Res 2004;28:21–29. 6. Mazzaferro V, Regalia E, Doci R, et al. Liver transplantation for the treatment of small hepatocellular carcinomas in patients with cirrhosis. N Engl J Med 1996;334:693–699. 7. Iwatsuki S, Dvorchik I, Marsh JW, et al. Liver transplantation for hepatocellular carcinoma: a proposal of a prognostic scoring system. J Am Coll Surg 2000;191:389–394. 8. Okada S, Okazaki N, Nose H, et al. Prognostic factors in patients with hepatocellular carcinoma receiving systemic chemotherapy. Hepatology 1992;16:112–117. 9. Llovet JM, Bustamante J, Castells A, et al. Natural history of untreated nonsurgical hepatocellular carcinoma: rationale for the design and evaluation of therapeutic trials. Hepatology 1999;29:62–67. 10. Ikai I, Yamaoka Y, Yamamoto Y, et al. Surgical intervention for patients with stage IV—A hepatocellular carcinoma without lymph node metastasis: proposal as a standard therapy. Ann Surg 1998;227:433–439. 11. Ikai I, Yamamoto Y, Yamamoto N, et al. Results of hepatic resection for hepatocellular carcinoma invading major portal and/or hepatic veins. Surg Oncol Clin N Am 2003;12:65–75. 12. Healey JE Jr, Schroy PC. Anatomy of the biliary ducts within the human liver: analysis of the prevailing pattern of branching and the major variations of biliary duct. Arch Surg 1953;66:599–616. 13. Allgaier HP, Deibert P, Olschewski M, et al. Survival benefit of patients with inoperable hepatocellular carcinoma treated by a combination of transarterial chemoembolization and percutaneous ethanol injection—a single-center analysis including 132 patients. Int J Cancer 1998;79:601–605. 14. Yamada R, Kishi K, Sato M, et al. Transcatheter arterial chemoembolization (TACE) in the treatment of unresectable liver cancer. World J Surg 1995;19:795–800. 15. Lee HS, Kim JS, Choi IJ, et al. The safety and efficacy of transcatheter arterial chemoembolization in the treatment of patients with
438
Ikai et al
Hepatocellular Carcinoma and Tumor Thrombosis in Major Portal Vein
hepatocellular carcinoma and main portal vein obstruction. A prospective controlled study. Cancer 1997;79:2087–2094. 16. Minagawa M, Makuuchi M, Takayama T, et al. Selection criteria for hepatectomy in patients with hepatocellular carcinoma and portal vein tumor thrombus. Ann Surg 2001;233:379–384. 17. Ohkubo T, Yamamoto J, Sugawara Y, et al. Surgical results for hepatocellular carcinoma with macroscopic portal vein tumor thrombosis. J Am Coll Surg 2000;191:657–660.
J Am Coll Surg
18. Pawlik TM, Poon RT, Abdalla EK, et al. Hepatectomy for hepatocellular carcinoma with major portal or hepatic vein invasion: results of a multicenter study. Surgery 2005;137: 403–410. 19. Wu CC, Hsieh SR, Chen JT, et al. An appraisal of liver and portal vein resection for hepatocellular carcinoma with tumor thrombi extending to portal bifurcation. Arch Surg 2000;135: 1273–1279.
JACS ONLINE CME-1 PROGRAM CME-1 COURSE EVERY MONTH 2 CREDITS AVAILABLE EACH MONTH 48 CREDITS AVAILABLE AT ALL TIMES www.jacscme.facs.org