|Home | About | Journals | Submit | Contact Us | Français|
Hepatocellular carcinoma (HCC) recurs in approximately 70% of cases after resection. Vascular invasion by tumor cells can be classified as gross or microscopic (mVI) and is a risk factor for recurrence. We examined a large cohort of patients with HCC that were treated by resection to identify features of microvascular invasion that correlated with recurrence and survival.
We reviewed the records of all HCC resections performed at the Mount Sinai School of Medicine between January 1990 and March 2006 to identify those with mVI, established by histological analysis. The numbers and sizes of vessels invaded, invasion of a vessel with a muscular wall, distance from the tumor, and satellite nodules were recorded.
Of the 384 patients that underwent resection for HCC, 131 (34.1%) met the entry criteria. The median follow-up period was 28.9 mo. There were 68 recurrences and 54 deaths. In multivariate analysis, invasion of a vessel with a muscular wall predicted recurrence (hazard ratio 1.8, P=0.02); invasion of a vessel with a muscular wall (hazard ratio=2.2, P=0.018) and invasion of a vessel that was more than 1 cm from the tumor (hazard ratio=2.1, P=0.015) predicted survival. A risk score that assigned point for the presence of each variable was correlated with recurrence (p=0.028) and survival (p<0.0001).
A novel classification system that includes invasion of a vessel with a muscular wall and invasion of a vessel that is more than 1 cm from the tumor can accurately predict risk of recurrence and survival of patients with mVI after HCC resection.
Resection of hepatocellular carcinoma (HCC) is an accepted first line treatment for patients with well preserved liver function and no evidence of portal hypertension.1 Unfortunately, resection is associated with a significant chance of recurrence with many series reporting recurrence rates of 60% to 70% at 5 years. 2-6 Vascular invasion has repeatedly been identified as a risk factor for recurrence and death after resection of HCC. 3,5-10
Vascular invasion is usually identified either as macroscopic, when the invasion of the vessel is visible on gross examination, or as microscopic, when the invasion is visible only on microscopy. Consequently, microscopic vascular invasion (mVI) can encompass a wide spectrum ranging anywhere from invasion of a single small vessel near the tumor capsule to just short of gross vascular invasion. As a result, one would also expect a wide range of outcomes in patients with mVI undergoing surgical resection. While staging systems have been proposed for patients with gross vascular invasion, there have been no robust studies stratifying patients with mVI in terms of recurrence and survival.11
In the present study, we examine a large cohort of patients with HCC and mVI treated with surgical resection. We examine the factors that correlate with recurrence and survival in this group of patients. A classification system for vascular invasion based on a risk score linked with prognosis is proposed.
We analyzed the prospectively collected data of a cohort of patients undergoing hepatic resection for HCC between January 1990 and March 2007 after obtaining approval from the Institutional Review Board. In order to have the outcomes affected primarily by tumor related factors and not by the degree of underlying liver disease or portal hypertension, a homogenous study population was selected out. Consequently, the inclusion criteria required patients to have Child's-Pugh A liver function and platelet count >100,000/μl with HCC and no evidence of extrahepatic disease on contrast enhanced CT scan of the chest and abdomen or contrast enhanced MRI of the abdomen and non-contrast CT of the chest. There were no strict exclusion criteria based on tumor size or number.
A minimum of 10 sections of non-tumoral hepatic parenchyma were examined microscopically by two expert pathologists (S.T and M.F.). At least one section of non-tumoral hepatic parenchyma 1cm away from the tumor was also examined. Patients without a margin of non-tumoral tissue ≥1cm were excluded. Microscopic vascular invasion (mVI) was defined as tumor within a vascular space lined by endothelium that was visible only on microscopy.
In patients with mVI, the size of the largest vessel invaded (<5mm, 5-10mm, >10mm), number of vessels invaded (1, 2-5, >5), and furthest distance on invasion from the tumor capsule (≤1cm, >1cm) were recorded as categorical variables. The presence of a muscular wall in any vessel with invasion as well as adherence of the intravascular tumor to the vessel wall were also noted. The presence of muscular wall was confirmed by trichrome stain, in which the muscle cells stained red. The presence of muscle in the wall was further assessed by immunoreactivity for anti-smooth muscle actin antibody (Figure 1). In addition to histological characteristics assessing extent of mVI listed above, other clinical variables such as tumor size, number, and grade, as well as presence of satellites were included in the analysis. Satellites were defined as tumors ≤2cm in size and located ≤2cm from the main tumor. Several clinical variables including liver function, platelet count, and need for transfusion were also recorded and analyzed.
The patients were followed with serial contrast enhanced CT scan of the chest and abdomen or contrast enhanced MRI of the abdomen and non-contrast CT of the chest every 3 months for the first year, every 4 months for the second year, and subsequently every 6 months. Solitary pulmonary recurrences were treated with lung resection. Patients with a solitary liver recurrence and Child's-Pugh A liver disease and no evidence of portal hypertension underwent a second hepatic resection. Patients with multiple intrahepatic recurrences or compromised hepatic function were treated with radiofrequency ablation and/or transarterial chemoembolization. Patients with recurrence confined to the liver and without significant comorbidities were also referred for liver transplantation. Patients undergoing liver transplantation were censored at the time of transplant for the purposes of this study.
Categorical data were compared using χ2 or Fisher's exact test as indicated. Continuous variables were compared using Student's t test or Man-Whitney test for variables with an abnormal distribution. Survival curves were calculated using the Kaplan-Meier method and compared using the log-rank test. All variables found to be significant on univariate analysis (P ≤ .05) were entered into a step-down Cox proportional hazard regression analysis. Receiver operating characteristic (ROC) curve analysis was used to determine the optimal cut-offs of continuous variables. SPSS version 16.0 for Windows was used.
During the study period, 385 patients underwent hepatic resection for treatment of HCC. Of these, 14 had either platelet count <100,000/μl or Child's B liver function and they were excluded from analysis (Figure 2). No vascular invasion was detected in 109 (29%) of patients, while 151(40%) had mVI and 111 (30%) had gross vascular invasion on pathologic examination. The patient demographics, liver function and tumor characteristics are provided in table 1. Survival and recurrence were significantly different based on presence and degree of vascular invasion (Figure 3).
Of the 151 patients with mVI found on pathology, 20 did not have adequate margin to assess non-tumoral tissue at 1cm from the tumor. These patients were excluded leaving 131 patients as the final mVI study population. Median follow-up of these patients was 29 months. During the study period, there were 54 deaths, including 4 (3%) perioperative deaths, and 68 recurrences. Fifty five (81%) of the recurrences were limited to the remaining liver.
Among the histological variables assessing extent of mVI, invasion of a vessel with a muscular wall, invasion of a vessel ≥1cm from the tumor capsule and invasion of >5 vessels were significantly associated with recurrence (Table 2). Multivariate analysis including these three variables found only invasion of a vessel with a muscular wall as a significant and independent predictor of recurrence (Table 2).
Invasion of a vessel with a muscular wall and invasion of a vessel ≥1cm from the tumor capsule were significant predictors of survival on univariate analysis (Table 2). On multivariate analysis, both variables were significant and independent predictors of survival and had similar hazard ratios and 95% confidence intervals (Table 2). A risk score assigning one point for invasion of a vessel with a muscular wall and one point for invasion of a vessel ≥1cm from the tumor capsule was calculated for each patient. This risk score was found to significantly and independently correlate with risk of recurrence and death (Figure 4).
Of the other clinical variables analyzed, presence of satellite nodules and tumor size, with a cut-off of 10cm as determined by ROC curve analysis, were also found to be significant predictors of recurrence on univariate analysis (Table 3). The presence of satellites, presence of >1 tumor nodule, tumor size >10cm, and need for blood transfusion were found to be significantly associated with survival on univariate analysis (Table 3). Variables including liver function, platelet count, patient age, AFP, tumor grade, and underlying liver disease were not significantly associated with either survival or recurrence in patients with mVI on univariate analysis in this cohort of patients.
In order to better understand the overall relevance of the risk score, multivariate analyses of recurrence and survival including the histological risk score as well as the other clinical variables found to be significant on univariate analysis were conducted. The risk score was found to be an independent predictor of both recurrence and survival after including these clinical variables (Table 3).
Finally, we aimed to further classify survival of the overall cohort according to a novel classification based on the absence of vascular invasion (class A), presence of microscopi vascular invasion (class B), which was further subdivided in B1-2 and 3 according to the results of the study, and presence of macroscopic vascular invasion (class C). (Table 4)
Hepatic resection for HCC can result in excellent long-term survival in properly selected patients. Nevertheless, resection continues to be plagued by a high rate of tumor recurrence. As our current study demonstrates, the large majority of recurrences are within the remaining liver. This is primarily due to intrahepatic dissemination of the tumor via the portal circulation. Consequently, it is not surprising that vascular invasion has been repeatedly identified as a predictor of recurrence and survival in series examining the outcomes of hepatic resection for HCC.
Microscopic vascular invasion (mVI), as demonstrated by our results, encompasses a heterogenous population of patients with a wide range of potential outcomes. We have shown that a risk score based on histological features of mVI that includes invasion of a vessel with a muscular wall and invasion of vessels ≥1cm from the tumor capsule, is able to stratify patients into 3 distinct groups with significantly different risks of recurrence and death. Interestingly, the patients with mVI and no risk factors had outcomes similar to patients with no vascular invasion while patients with mVI and both risk factors behaved like patients with gross vascular invasion in terms of both recurrence and survival (Table 4).
In addition to vascular invasion, liver function and degree of portal hypertension have also been identified as predictors of survival after liver resection for HCC.1,12,13 The fact that clinical variables associated with liver function and platelet count (which can serve as a surrogate marker for portal hypertension) did not correlate with survival in our series is not surprising, as only patients with well preserved liver function and no portal hypertension were selected for the study. As a result, only tumor related variables influenced the outcomes.
Some investigators have proposed referring all patients found to have mVI after hepatic resection for liver transplantation due to the significant risk of recurrence.14 An alternate approach would be to selectively refer patients for transplantation based on their likelihood of recurrence as determined by the classification system proposed in this study.
The success of sorafenib in the treatment of advanced HCC has opened the door for the use of such small targeted molecules in the adjuvant setting (15). As more and more of such small molecular therapies are developed, there will be a growing demand to study their benefits in an adjuvant setting. The ability to accurately stratify patients found to have mVI after resection of HCC can help to ensure a homogeneous study population for such adjuvant trials.
The proposed classification system clearly requires external validation. However, if the results of this study are confirmed in an independent cohort of patients, the ability to stratify patients with mVI after resection can have obvious and significant advantages.
Grant Support: Josep Llovet was supported by grants from the U.S. National Institute of Diabetes and Digestive and Kidney Diseases (1R01DK076986-01), the Samuel Waxman Cancer Research Foundation, the Spanish National Health Institute (SAF-2007-61898) by Institut Catala de Recerca Avançada (ICREA). Myron Schwartz was supported by a grant from the National Institute of Health (K24 DK 60498-01)
Presented at the 1st Annual meeting of the International Liver Cancer Association (ILCA), Barcelona, Spain, October 2007
Financial disclosures: None
Writing assistance: None
Author contributions:Sasan Roayaie - study concept and design; analysis and interpretation of data; drafting of the manuscript; critical revision of the manuscript for important intellectual content; statistical analysis; study supervision
Iris Blume – data acquisition; analysis and interpretation of data; statistical analysis;
Swan Thung - study concept and design; data acquisition; technical support; critical revision of the manuscript for important intellectual content;
Maria Guido - study concept and design; data acquisition; technical support; critical revision of the manuscript for important intellectual content;
Maria-Isabel Fiel - study concept and design; data acquisition; technical support; critical revision of the manuscript for important intellectual content;
Spiros Hiotis - study concept and design; critical revision of the manuscript for important intellectual content;
Daniel Labow - study concept and design; critical revision of the manuscript for important intellectual content;
Josep Llovet - study concept and design; analysis and interpretation of data; critical revision of the manuscript for important intellectual content; study supervision
Myron Schwartz - study concept and design; analysis and interpretation of data; critical revision of the manuscript for important intellectual content; study supervision.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Sasan Roayaie, Mount Sinai Liver Cancer Program, Mount Sinai School of Medicine, New York, NY.
Iris N. Blume, Mount Sinai Liver Cancer Program, Mount Sinai School of Medicine, New York, NY.
Swan N. Thung, Mount Sinai Liver Cancer Program, Mount Sinai School of Medicine, New York, NY.
Maria Guido, Department of Pathology, University of Padua, Padua, Italy.
Maria-Isabel Fiel, Mount Sinai Liver Cancer Program, Mount Sinai School of Medicine, New York, NY.
Spiros Hiotis, Mount Sinai Liver Cancer Program, Mount Sinai School of Medicine, New York, NY.
Daniel M. Labow, Mount Sinai Liver Cancer Program, Mount Sinai School of Medicine, New York, NY.
Josep M. Llovet, BCLC Group. [HCC Translational Lab, Liver Unit and Pathology Department], Hospital Clínic, CIBERehd, IDIBAPS, Barcelona, Catalonia, Spain. Institució Catalana de Recerca i Estudis Avançats (ICREA). Generalitat de Catalunya. Barcelona, Catalonia, Spain.
Myron E. Schwartz, Mount Sinai Liver Cancer Program, Mount Sinai School of Medicine, New York, NY.