—  HANS POPPER HEPATOPATHOLOGY SOCIETY   —

Grading and Staging of Hepatocellular Carcinoma


David E. Kleiner
National Cancer Institute
NIH, Bethesda, MD


Staging Hepatocellular Carcinoma
Staging is a method of classifying disease according to severity at a particular point in time in order to guide therapy, to prognosticate and to organize disease information for research or public health reasons. Staging is generally used in one of two situations. A stage may be a particular point along the natural history of the disease progression. For example, in chronic liver disease, the stage is generally defined as the degree of fibrotic progression, which, in the natural course of the disease, continuously progresses from no fibrosis to cirrhosis. Similarly, a malignant neoplasm will generally progress from low stage to high stage in the natural course of events. The other use of stage is to measure the extent of disease following an acute insult. For example the staging of burn injuries is used to classify the degree of the injury following the burn and one does not expect the stage to worsen following the acute event.

In oncology, three pieces of information are generally required before definitive treatment can be planned. These are the primary location of the neoplasm (e.g. breast, prostate, liver, lymph node), its histologic subtype (e.g. adenocarcinoma, lymphoma, melanoma, small cell carcinoma) and its stage. The first two elements are usually highly linked, since not all histologies occur with equal frequencies at all sites. The stage is independent of the other two elements in that tumors may present at any stage, but dependent in that staging criteria vary by location and histology. Although the earliest systematic oncology staging classifications were made in the early part of the 20th century, it was not until 1950 that the International Union against Cancer (UICC) proposed a system of staging tumors based on uniform criteria [1]. This proposal was organized along measuring the extent of tumor spread in three ways: the size and extent of the primary tumor (T stage), the involvement of regional lymph nodes (N stage) and the metastatic involvement of distant sites (M stage). These three scales were then organized into a 4-tiered composite stage (TNM stage). For most epithelial tumors this staging system breaks down in the following way: stage 1 tumors are small and organ confined, stage 2 tumors are larger, but generally still organ confined, stage 3 tumors usually have nodal spread and stage 4 tumors have distant metastases. Staging then has direct implications for the treatment modalities that may be used. The TNM system is a clinico-pathologic system in that most tumors will require both a clinical evaluation and a pathologic evaluation for complete classification. Furthermore, this system is limited to an evaluation of the tumor only, which is a significant caveat when dealing with hepatocellular carcinoma. While staging is most commonly done at the point in time when the tumor diagnosis is first made, it is possible to restage patients at various times in the course of their disease. However, with the exception of restaging following neo-adjuvant therapy, staging is usually too crude a tool to use in following patients for response to therapeutic intervention.

During the period from 1950 to 1967 the UICC developed TNM staging systems for most tumors. In 1954, the International Federation of Gynecology and Obstetrics (FIGO) began developing TNM staging classifications for gynecologic malignicies and in 1959, the American Joint Commission for Cancer Staging and End Results Reports (later shortened to the American Joint Commission for Cancer or AJCC) began developing TNM systems for use in the United States that were similar to, but not exactly the same as, the TNM systems the UICC was publishing. This led to much confusion, but it was not until 1987 that the UICC and AJCC systems were unified into a single system. Today, all three organizations (UICC, AJCC and FIGO) work together to produce a single set of rules for staging all malignancies. The latest edition of the TNM classification was published in 2002 [2, 3] .

The TNM staging of hepatocellular carcinoma underwent significant revision in the latest edition. Table 1 shows the system as it existed in the 5th edition of the TNM classification published by the UICC in 1997. Stratification of the primary tumor into the various T stages was dependent on size (with 2 cm being the dividing point), involvement of one or both lobes, single or multiple tumor nodules and the presence or absence of vascular invasion or spread beyond the liver. Following the publication of the 5th edition TNM system, several large scale studies of hepatocellular carcinoma were undertaken to examine the different elements of the staging system and to validate the stage groupings that were made [4, 5, 6, 7] . Two major observations resulted from these studies(8}. The first was that single nodule tumors that did not show vascular invasion had basically the same prognosis regardless of size. The second was that tumors that involved both lobes with multiple nodules did not have a prognosis different from that of similar multinodular tumors that were confined to one lobe. Thus, in the 6th edition of the TNM classification (Table 2), T1 tumors are single nodules of any size that do not demonstrate vascular invasion and classification based on involvement of one or both lobes has been dropped. The translation of individual T, N and M stages to the composite stage has also been simplified, with T3, T4 and N1/M0 tumors all grouped together as stage III. The 6th edition of the TNM classification also recommended the routine inclusion of Edmondson grading (G scores, discussed below) and a simplified fibrosis classification for the non-tumorous liver (F scores). Systematic reporting of G and F scores along with the T, N, and M stages will allow for evaluation of these features as part of future staging systems.

Hepatocellular carcinoma is nearly unique in that it most often arises in the presence of a particular medical condition—advanced chronic liver disease—which itself has significant morbidity and mortality. As noted above, the TNM staging of hepatocellular carcinoma has not incorporated any aspects of concurrent liver disease in the stage of the cancer. Since cirrhosis and its sequelae significantly affect patient management and treatment options, a number of other staging systems have been developed to incorporate prognostic information from concurrent liver disease. The three systems discussed here are the Okuda system, the Barcelona system and the CLIP system. All of them incorporate prognostic features of end-stage liver disease along with the evaluation of tumor extent. Tumor extent is entirely based on evaluation of the primary tumor—none of the systems incorporates information about regional or distant metastases. Pathologic examination of resected specimens has little impact on any of these systems as most of the staging of tumor extent can be done on imaging studies.

The Okuda system (Table 3) is the oldest of the three and is least dependent on pathologic evaluation [9]. The tumor extent is gauged on an imaging study by taking the ratio of tumor area on the slice with the greatest involvement to the area of the liver on the slice with the greatest liver area. The simple determination of more than 50% involvement is made and combined with three other prognostic factors: albumin level, bilirubin level and the presence or absence of ascites. Absence of all four poor prognosis features is stage I with stage II and III having 1 to 2 and 3 to 4 features respectively. The Okuda system has shown good prognostic ability. Among a cohort of 229 patients who received no specific therapy the median survival times were 8.3, 2.0 and 0.7 months for stages I, II and III.

The Barcelona system (Table 4) was proposed as a tool for stratifying patients into treatment pools and to identify patient populations that could be recruited for experimental protocols [10]. It too is heavily dependent on clinical features, both general performance status according to the ECOG system and liver disease specific features. Patients with single nodule disease or those with several small nodules as well as good performance status and low Child's class are directed toward surgical or local ablative therapies combined with orthotopic liver transplantation with the goal of cure. Patients with multiple tumors and moderate performance status are directed towards palliative and or experimental therapies. Finally patients with poor performance status or advanced decompensated liver disease are directed towards supportive care regardless of tumor extent.

The Cancer of the Liver Italian Program (CLIP) Score (Table 5) represents a refinement of the idea used in the Okuda system [11]. It was based on multivariate analysis of a large number of potential prognostic factors. It incorporates an image-based estimation of tumor extent along with the Child-Pugh stage (itself a composite score of 5 clinical variables), alpha-feto protein levels and the presence or absence of portal vein thrombosis. In a prospective validation study done on a cohort of patients treated heterogeneously, there was good stratification of patients into 4 groups, with total CLIP scores of 0, 1, 2-3 and 4-6 [12].

It is difficult to make objective comparisons between these different staging systems, partly because they were developed with somewhat different objectives and because the patient populations vary considerably with respect to etiology of HCC and with the treatment modalities that are available. While it seems important to incorporate clinical prognostic information related to the patient's liver disease or performance status, this has the effect of obscuring or devaluing the contribution of the tumor itself to prognosis. Currently patients with end-stage liver disease make poor candidates for either resection or transplantation but that may not always be the case. It therefore seems prudent to continue to give the tumor a stage separate from other clinical characteristics and then to combine that stage with information about the patient's liver disease in order to offer the best therapeutic options.

Grading Hepatocellular Carcinoma
Grading a disease is an attempt to gauge how rapid the disease will evolve from its current stage. Tumor grade should therefore be a measure of the aggressiveness of the tumor and is usually estimated by the evaluation of surrogate markers like the degree of differentiation, mitotic rate etc. In some tumors, such as breast carcinoma, the grade of the tumor is an independent predictor of outcome when tested against other predictors such as stage. The availability of an effective therapy may reduce or eliminate the predictive value of grade or may even reverse the usual prediction if the high grade tumor turns out to be more responsive to therapy.

There is only one well-accepted system of grading hepatocellular carcinomas—the method proposed by Edmondson and Steiner in their 1954 landmark paper in Cancer [13]. This system is based on an overall assessment of tumor architecture and cytologic characteristics. The current edition of the UICC/AJCC staging system recommends that hepatocellular carcinomas be routinely graded using the Edmondson grade. Since the best way to describe the difference in grades is to quote the original monograph, the following descriptions are quoted from that paper:

Grade 1 – "probably best reserved for those areas in Grade II carcinomas where the difference between the tumor cells and hyperplastic liver cells is so minor that diagnosis of carcinoma rests upon the demonstration of more aggressive growths in other parts of the neoplasm."

Grade 2 – "cells showed a marked resemblance to normal hepatic cells. The nuclei were larger and more hyperchromatic than normal, but the cytoplasm was abundant, and acidophilic… Acini were frequent, their lumina varying in size from tiny canaliculi to large thyroid-like spaces. The lumina were often filled with bile; otherwise they contained a small amount of protein precipitate."

Grade 3 – "nuclei were usually larger and more hyperchromatic than those of Grade II. These nuclei occupied a relatively greater proportion of the cell. The cytoplasm was granular and acidophilic as in Grade II but it usually was less so… Bile and acinar formation were noted less frequently. Tumor giant cells were most numerous in this group."

Grade 4 – "nuclei were intensely hyperchromatic and occupied the greater part of the cell. The cytoplasm was variable in amount, was often scanty, and contained fewer granules. The growths were more medullary in character, the trabeculae were difficult to find…"

Table 7 attempts to organize some of these features across the grades. In univariate analysis, Edmondson grading has been correlated with overall survival, however it frequently drops out of multivariate analyses of prognostic indicators. Part of the problem may be that because the Edmondson grade is a complex cyto-architectural evaluation, individual features that have powerful predictive value are diluted by features that have little or no predictive value. In one examination of this issue, Lauwers et al. examined mitotic activity, tumor necrosis, invasive growth pattern, microvascular invasion, nuclear grade and predominant architectural pattern [14]. Only microvascular invasion and nuclear grade were significantly associated with overall survival in a multivariate analysis of these six features. Since microvascular invasion is already incorporated into the existing TNM staging system, it may be worth defining a grading system for hepatocellular carcinoma that is just based on nuclear features.

Other Factors to Consider – Margin status and Fibrosis
Margin status of surgical resections is a typical predictor of outcome that is usually most related to the risk of local recurrence. Because many hepatocellular carcinomas arise from cirrhotic livers that may contain other premalignant or sub-clinical malignant foci, local recurrence may frequently result from this field effect rather than a close margin. Several studies have examined margin status in multivariate analysis and the results have been mixed [15, 16, 17] . However, because examination of the surgical margin is an ingrained part of the specimen examination and is expected by the surgeon, it is still should be a required part of the specimen evaluation and surgical pathology report.

As noted above, the degree of fibrosis in the non-tumorous liver is currently part of the UICC/AJCC staging evaluation. In practice, the non-tumorous liver should receive the standard evaluation for chronic liver disease that is appropriate for the etiology of the disease. Thus, if viral hepatitis is present, then the report should contain a statement of inflammatory activity and fibrosis. Other chronic liver diseases should at least be evaluated for the degree of fibrosis, and qualitative statements can be made about other features, such as inflammation, steatosis, iron accumulation, etc.

Synoptic Reporting and Checklists
Beginning in January of 2004 the American Council of Surgeons Commission on Cancer required that cancer programs under its purview demonstrate that pathology reports for oncology specimens meet certain requirements. One of these requirements was that cancer-related pathology reports include all scientifically validated and regularly used data elements relevant to the particular neoplasm under study. Pathologists had already been moving slowly towards more standardized reports, so these requirements merely sped up the process. Both the College of American Pathologists (CAP) and the Association of Directors of Anatomic and Surgical Pathology (ADASP) have published reporting recommendations and checklists of required report elements for hepatic neoplasms. The lists of required elements are very similar:

CAP1 ADASP [18]
Specimen/Resection Type
Focality

Size of tumor in greatest dimension
Histologic Classification (WHO)
Histologic Grade (Edmonston for HCC)

Pathologic T Stage
Pathologic N Stage
Pathologic M Stage
Parenchymal Margin
Bile Duct Margin
Other Margin(s) 		Specimen/Resection Type
Focality
Extent of Invasion
Size of largest tumor in 3 dimensions
Histologic Classification (WHO)
Histologic Grade (Edmonston for HCC)
Detailed Nodal Evaluation by Node Group
Pathologic T Stage
Pathologic N Stage
Pathologic M Stage
Parenchymal Margin
Bile Duct Margin
Other Margin(s)

1 http://www.cap.org/apps/docs/cancer_protocols/protocols_intro.html

In both sets of recommendations there are optional elements that include evaluation of the non-tumorous liver. It is interesting that neither list includes fibrosis evaluation as a requirement although this is now a recommended part of the UICC/AJCC hepatocellular carcinoma recommendations. Since the degree and type of liver disease may have an impact on the patient's future therapeutic options, it would be prudent to include an evaluation of that liver disease according to accepted guidelines. At the end of this handout is a one page checklist that we use at our institution for primary hepatic neoplasms. If you find it useful, please feel free to copy and/or adapt to your own needs. Copies in MS Word format may be obtained via e-mail request (kleinerd@mail.nih.gov).

Table 1. TNM classification (5th edition, 1997)

T Stage Size of Largest Nodule Extent Multiplicity of Nodules Vascular Invasion1 Major Vein Invasion2 Adjacent Organ Invasion3
T1 <= 2 cm One lobe Single No No No
T2 <= 2 cm One lobe Single Yes No No
<= 2 cm One lobe Multiple No No No
> 2 cm One lobe Single No No No
T3 <= 2 cm One lobe Multiple Yes No No
> 2 cm One lobe Single Yes No No
> 2 cm One lobe Multiple Any No No
T4 Any Both lobes Multiple Any  
Any Any Any Any Yes Any
Any Any Any Any Any Yes

1 Microscopic invasion of portal or central veins
2 Macroscopic invasion of veins
3 Extension of tumor across the peritoneal capsule or into any adjacent organ except for the gall bladder

N Stage
N0   No positive regional nodes
N1   Positive regional nodes

M Stage
M0   No distant metastases
M1   Distant metastases

Overall Stage

Stage T Stage N Stage M Stage
I 1 0 0
II 2 0 0
IIIA 3 0 0
IIIB 1, 2 or 3 1 0
IVA 4 Any Any
IVB Any Any 1

Table 2. TNM Classification (6th edition, 2002)

T Stage Size of Largest Nodule Multiplicity of Nodules Vascular Invasion1 Major Vein Invasion2Adjacent Organ Invasion3
T1 Any Single No No No
T2 Any Single Yes No No
< 5 cm Multiple Any No No
T3 >= 5 cm Multiple Any No No
Any Any Any Yes No
T4 Any Any Any Any Yes

1 Microscopic invasion of portal or central veins
2 Macroscopic invasion of veins
3 Extension of tumor across the peritoneal capsule or into any adjacent organ except for the gall bladder

N Stage
N0   No positive regional nodes
N1   Positive regional nodes

M Stage
M0   No distant metastases
M1   Distant metastases

Overall Stage

Stage T Stage N Stage M Stage
I 1 0 0
II 2 0 0
IIIA 3 0 0
IIIB 4 0 0
IIIC Any 1 0
IV Any Any 1

While not part of formal staging, the 6th edition TNM classification also calls for Edmondson grading (G1 to G4) and fibrosis:

F0   Fibrosis up to advanced bridging fibrosis (modified Ishak stage 0-4)
F1   Incomplete or complete cirrhosis (modified Ishak stage 5-6)

Table 3: Okuda System

Prognostic Features
Tumor Size >50% of cross-sectional area (CT, Radionuclide, Angiogram)
Ascites Present
Albumin < 3g/dL
Bilirubin > 3mg/dL

Stage # of Features
1 0
2 1-2
3 3-4

Table 4: Barcelona Clinic Liver Cancer Staging Classification (Barcelona/BLCL System)
 Liver Function
Stage Tumor Stage ECOG Score1 Okuda Stage Portal HTN Bilirubin
A (early)
A1 Single, any size 0 I Absent Normal
A2 Single, any size 0 I Present Abnormal
A3 Single, any size 0 I Present Abnormal
A4 2-3 nodules all <= 3 cm 0 I-II Child's class A-B
B (intermediate) > 3 nodules or > 3 cm 0 I-II Child's class A-B
C (advanced) Vascular Invasion or
Extrahepatic Spread 		OR 1-2I-II Child's class A-B
D (end stage) Any 3-4 OR III OR Child's class C

1 Eastern Cooperative Oncology Group Performance Status Scale

BCLC Treatment Recommendations

Stage Treatment Goal Treatment
A Cure Surgical resection +/- OLT, RFA, Ethanol inj.
B Palliation Embolization, Chemoembolization
C Palliation Experimental approaches
D Supportive Symptom management

Table 5. Cancer of the Liver Italian Program (CLIP) Score
 Points
Feature 0 1 2
Child-Pugh Stage A B C
Tumor Single, <= 50% Multiple, <=50% Massive or > 50%
AFP <400 >=400  
Portal vein thrombosis No Yes  

Total Score Ranges from 0-6

Table 6. Child-Turcotte-Pugh Classification

 Points
1 2 3
Bilirubin (mg/dL) <2 2-3 >3
for PBC/PSC <4 4-10 >10
Albumin (g/dL) <3.5 2.8-3.5 <2.8
PT (sec. prolonged) 1-3 4-6 >6
or INR <1.7 1.7-2.3 >2.3
Ascites None Slight/Controlled Moderate/Severe
Encephalopathy None Stage 1-2 Stage 3-4

Total Score Ranges from 5 to 15

Child's Class Total Score
A 5-6
B 7-9
C 10-15

Table 7: Edmondson Grading Summarized in Tabular Form

Grade N/C Ratio Cytoplasm Architecture
1 Near Normal Near normal Near normal
2 Greater than normal Abundant, eosinophilic Acini frequent, bile production seen
3 High Less cytoplasm but still acidophilic Acini less frequent
4 High, Anaplastic Scanty Acini rare

Surgical Pathology Cancer Case Summary (Checklist)
Click here to download or print the PDF file (122KB) for this checklist.

References

  1. Gospodarowicz M, Benedet L, Hutter R, Fleming I, Henson D, Sobin L. History and international developments in cancer staging. Cancer Prevent Control 1998;2:262-268.

  2. Greene F, Page D, Fleming I, editors. AJCC Cancer Staging Manual. 6th ed. New York: Springer; 2002.

  3. Sobin L, Wittekind C, editors. UICC TNM Classification of Malignant Tumours. 6th ed. New York: Wiley-Liss; 2002.

  4. Izumi R, Shimizu K, Ii T, Yagi M, Matsui O, Nonomura A, Miyazaki I. Prognostic factors of hepatocellular carcinoma in patients undergoing hepatic resection. Gastroenterology 1994;106:720-727.

  5. Tsai T-J, Chau G-Y, Lui W-Y, Tsay S-H, King K-L, Loong C-C, Hsia C-Y, et al. Clinical significance of microscopic tumor venous invasion in patients with resectable hepatocellular carcinoma. Surgery 2000;127:603-608.

  6. Bilimoria M, Lauwers G, Doherty D, Nagorney D, Belghiti J, Do K-A, Regimbeau J-M, et al. Underlying liver disease, not tumor factors predicts long-term survival after resection of hepatocellular carcinoma. Arch Surg 2001;136:528-535.

  7. Vauthey J-N, Lauwers G, Esnaola N, Do K-A, Belghiti J, Mirza N, Curley S, et al. Simplified staging for hepatocellular carcinoma. J Clin Oncol 2002;20:1527-1536.

  8. Greene F. TNM staging for malignancies of the digestive tract: 2003 changes and beyond. Semin Surg Oncol 2003;21:23-29.

  9. Okuda K, Ohtsuki T, Obata H, Tomimatsu M, Okazaki N, Hasegawa H, Nakajima Y, et al. Natural history of hepatocellular carcinoma and prognosis in relation to treatment. Study of 850 patients. Cancer 1985;56:918-928.

  10. Llovet JM, Bru C, Bruix J. Prognosis of hepatocellular carcinoma: the BCLC staging classification. Semin Liver Dis 1999;19:329-338.

  11. A new prognostic system for hepatocellular carcinoma: a retrospective study of 435 patients: the Cancer of the Liver Italian Program (CLIP) investigators. Hepatology 1998;28:751-755.

  12. Prospective validation of the CLIP score: a new prognostic system for patients with cirrhosis and hepatocellular carcinoma. The Cancer of the Liver Italian Program (CLIP) Investigators. Hepatology 2000;31:840-845.

  13. Edmondson H, Steiner P. Primary carcinoma of the liver: A study of 100 cases among 48,900 necropsies. Cancer 1954;7:462-503.

  14. Lauwers G, Terris B, Balis U, Batts K, Regimbeau J-M, Chang Y, Graeme-Cook F, et al. Prognostic histologic indicators of curatively resected hepatocellular carcinomas: A multi-institutional analysis of 425 patients with definition of a histologic prognostic index. Am J Surg Pathol 2002;26:25-34.

  15. Lauwers G, Vauthey J-N. Pathological aspects of hepatocellular carcinoma: a critical review of prognostic factors. Hepato-Gastroenterol 1998;45:1197-1202.

  16. Poon RT, Fan ST, Ng IO, Wong J. Significance of resection margin in hepatectomy for hepatocellular carcinoma: A critical reappraisal. Ann Surg 2000;231:544-551.

  17. Regimbeau J-M, Abdalla E, Vauthey J-N, Lauwers G, Durand F, Nagorney D, Ikai I, et al. Risk factors for early death due to recurrence after liver resection of hepatocellular carcinoma: results of a multicenter study. J Surg Oncol 2004;85:36-41.

  18. Dabbs D, Geisinger K, Ruggiero F, Raab S, Nalesnik M, Silverman J. Recommendations for the reporting of tissues removed as part of the surgical treatment of malignant liver tumors. Hum Pathol 2004;35:1315-1323.