Similar to the staging systems of other neoplasms, the TNM staging system [58] for HCC also includes tumor size, number, vascular invasion, lymphatic metastasis, and metastasis (Table 11.2). The TNM staging system emphasizes the characteristics of the tumor, which comprehensively and accurately describe the developmental history of HCC. Many studies have identified that the seventh edition of the AJCC TNM staging system is able to adequately stratify patients, and this system is one of the most widely accepted staging systems for HCC. The major modification from the sixth to the seventh edition was the separation of the T3 stage into T3a and T3b; this change indicates major vascular invasion of portal or hepatic veins as an important predictive factor for prognosis. AJCC recommends the TNM staging system for HCC patient staging. The West China Hospital of Sichuan University examined 774 cases of HCC from 2007 to 2009 according to the TNM staging system. The respective 1-, 3-, and 5-year survival rates were 89, 65.1, and 41.1 % for patients with stage I HCC; 78.5, 32.2, and 15.1 % for stage II HCC; 55.3, 13.4, and 10.1 % for stage III HCC; and 44.4, 5.6, and 0 % for stage IV [57]. However, there are several limitations of the TNM staging system, and the TNM system has limited stratification ability. (1) The TNM staging system only includes tumor-related morphology but not liver performance-related parameters, such as liver function and portal hypertension. (2) The TNM staging system is based on postoperative pathology results; therefore, its application has been limited because most patients with HCC are at an advanced stage that is surgically unresectable at the time of diagnosis. (3) The TNM staging system considers tumor size, number, and vascular invasion with equivalent prediction value for survival. This classification would underestimate the prognosis of patients with large, solitary tumors without vascular invasion and fail to adequately stratify patients.

The Okuda staging system was proposed by Okuda and is the first staging system to combine liver function and tumor characteristics (Table 11.3) [59]. This system is based on 850 cases of HCC patients with liver resection. At that time, early HCC diagnosis was relatively rare, and the staging system was therefore based on data from patients with advanced disease [60]. Thus, the median survival time for the 850 HCC cases was 4.1 months. This staging system includes an index for tumor characteristics (tumor size) and three indexes for liver function (ascites, albumin, and bilirubin). Many other HCC staging systems are based on the Okuda system and have been constantly improved. The Okuda system is primarily suitable for advanced patients, and several limitations also exist. First, the Okuda system is inadequate for contemporary HCC cases, particularly those that are diagnosed early. Thus, it has limited ability to stratify patients with early HCC. Second, this system did not consider several other tumor characteristics, such as single and multifocal tumors, AFP levels, vascular invasion, and metastasis. In addition, the stratification of tumor size is also very difficult. Third, the cutoff value of bilirubin is high and suitable for patients with severe liver function damage. Fourth, some subjective indices remain, such as ascites and liver size.

The system proposed by the Cancer of the Liver Italian Program (CLIP) in 1998 was based on a retrospective study with 435 patients in Italy and was subsequently prospectively validated [60]. The CLIP system has been widely demonstrated to be a more appropriate prognostic model for the late-stage HCC population and has been validated in case series from various parts of the world [61–62]. This system was proposed to overcome the disadvantage of the TNM staging system and is superior to the Okuda system. This system includes four indexes: Child-Pugh grade, tumor morphology, AFP, and portal vein thrombosis (Table 11.4). However, this system included a small number of advanced cases and many radically treated patients in their original studies. The main limitations are as follows. First, approximately 70–80 % of all patients have a CLIP score of 0–2. The CLIP score can discriminate patient populations with scores of 0–3, but it is not able to discriminate score between scores of 4–6. Second, the definition of tumor morphology in the best prognostic group is too advanced [63]. Third, other indexes for tumor characteristics are not included in this system, such as lymph node invasion and metastasis. Therefore, this system cannot identify the groups who would most benefit from curative and aggressive treatment.

The JIS staging system combines Child-Pugh grade and the TNM staging system based on the LCSGJ criteria [63]. It was proposed in Japan in 2003 (Table 11.5) based on 722 HCCs and is believed to have greater stratification ability than the CLIP scoring system and perform better than the CLIP scoring system in selecting the best prognostic patient group. Each patient with a Child-Pugh classification of A, B, or C was allocated scores of 0, 1, and 2, respectively. Based on the TNM staging of the LCSGJ, stage I (fulfilling the following three conditions: solitary, <2 cm, no vessel invasion), stage II (fulfilling two of the three conditions), stage III (fulfilling one of the three conditions), and stage IV (fulfilling none of the three conditions) were allocated scores of 0, 1, 2, and 3, respectively. The summation of the tumor staging score and the Child-Pugh classification score was defined as the JIS score [63].

This system is suitable for most HCC patients, especially for patients with good prognosis. However, the JIS system may be limited in its ability to stratify patients with advanced scores because it uniformly assigns tumor stage and liver function.

The CUPI score was the only system widely used for Chinese HCC patients with HBV infection. This system was based on a study cohort of 926 Chinese patients with primarily hepatitis B-associated HCCs in 2002 by Leung [64]. The CUPI score includes the conventional TNM system, a number of other liver functional factors, AFP level, and performance status (Table 11.6) [65]. The CUPI was more discriminant than the TNM staging system, Okuda staging systems, and CLIP prognostic score in classifying patients into different risk groups and was better at predicting survival. This system is primarily suitable for patients with HBV-related HCC, especially in China. The limitations are as follows. First, it is unclear whether this system can be used in other western counties. Second, this system includes an objective index. Third, most patients in the CUPI study were advanced patients. Therefore, the application for radical liver resection is limited.

Any staging system should classify patients into subgroups with significantly different outcomes and should simultaneously help to direct therapy. Clinical staging for cancers provides guidance for predicting survival outcome and deciding optimal treatment strategies. Although several staging systems have been proposed over the past several decades, there is no ideal staging system for patient stratification and survival prediction. Generally, the Okuda staging system, CLIP staging system, CUPI staging system, and JIS staging system are more appropriate for assessing advanced HCC patients without operation, with an overall median survival time of 4–5 months. The TNM system and BCLC system are suitable for patients with liver resection. Currently, most studies have acknowledged that the TNM, CLIP, and BCLC systems are better for patient stratification and survival prediction, especially the BCLC system. Several studies [60, 66–68] have suggested that the BCLC system is better than the Okuda, CLIP, CUPI, JIS, and TNM staging systems for predicting survival. The CLIP system is superior to the Okuda system [60], and JIS is superior to CUPI [68].