Introduction: For prognostication of hepatocellular carcinoma (HCC), the optimal size cutoff value and the significance of microvascular invasion (mvi) remain inconclusive. Objective: This study sought to revisit the inconclusive components of current American Joint Committee on Cancer (AJCC) staging system and to develop a new practical prognostication model for patients with HCC based on 2 large cohorts. Method: In 1,175 patients who underwent resection for HCC (training cohort), prognostic significance of mvi, and optimal size cutoff value were revisited and a new staging model was established. Then, its performance was validated using 5,249 patients derived from a population-based database (validation cohort). Results: The optimal size cutoff value was 2 cm in multiple HCC, similar to that for solitary HCC. Multivariate analyses confirmed that mvi, a size >2 cm, and multiple lesions were independent predictors with similar weights for disease-specific survival. Based on these results, earlier stages of HCC were reclassified according to the number of the following factors: multiple tumors, >2 cm in diameter, and presence of mvi. Also, given the similar prognosis of Stages IIIB and IVA according to the AJCC 8th edition, these groups were reclassified into the same stage. This new staging model had a better performance than the AJCC 8th edition in both training cohort (c-statistics, 0.648 vs. 0.629) and the validation cohort (c-statistics, 0.646 vs. 0.645) regardless of the presence of cirrhosis. Conclusions: Inclusion of a size cutoff value of 2 cm and mvi could reclassify the current version of the AJCC staging system and offer an alternative prognostication model using a single size-cutoff value of HCC.

Prognostication of patients with hepatocellular carcinoma (HCC) remains a challenge because the patients suffer from both HCC itself and underlying liver disease, both of which are prognostic factors. Incorporating components of the anatomical extent of the tumor and the degree of the underlying liver disease has led to multiple staging and scoring systems with differing options [1]. However, the oncological extent of the tumor and underlying liver disease should ideally be assessed independently, since a mixture of tumor stages and a treatment algorithm that is influenced by the hepatic functional reserve could result in a significant discrepancy in the prognosis of patients with the same cancer stage among regions and countries because of geographic heterogeneity in cancer management and preferences and/or expertise in specific treatments.

Among several staging systems, the American Joint Committee on Cancer (AJCC) staging system [2] and the Liver Cancer Study Group of Japan (LCSGJ) staging system [3] have conventionally been used to describe the pure anatomical extents of the diseases. By incorporating clinical factors strongly associated with the prognosis of HCC, these staging systems have regularly been revised to reflect the updated clinical outcomes of populations [4, 5] and/or published evidence [6, 7]. While the AJCC staging system and the LCSGJ staging system include similar clinical factors for prognostication, there have been slight differences in the weights of the clinical factors and the cutoff values for tumor size between these 2 staging systems [8]. These differences are partly derived from differences in the baseline characteristics of patients, basic policies for patient management, and treatment algorithms between Japan and Western countries. However, for the comparison of clinical outcomes and the standardization of optimal patient management, a universally available staging model is needed. Therefore, this study sought to revisit the inconclusive components of the current AJCC staging systems and to develop a new simple and practical prognostication model for patients with HCC based on 2 large cohorts.

Study Population

From a prospective hepatobiliary database maintained by the Department of Gastroenterological Surgery, Toranomon Hospital, Tokyo, Japan, a total of 1,175 consecutive patients who had undergone curative resection for HCC between January 1995 and October 2017 were identified as a training cohort. In addition, 5,249 patients aged ≥20 years old with histologically confirmed HCC undergoing surgical resection were identified using the surveillance, epidemiology, and end-results (SEER) database from 1995 to 2015 inclusive as a validation cohort. The International Classification of Disease 3rd edition was used to identify HCC using site code C22.0 and the corresponding histology code of 8170 hepatocellular carcinoma. All the analyses in the current study were performed in accordance with the ethical guidelines for clinical studies in Japan, SEER research data use agreement, and Declaration of Helsinki under approval of Institutional Review Board of Toranomon Hospital (No. 1823). Written informed consent was waived because of the retrospective design.

Indications for Liver Resection and Follow-Up

The indications for surgery were based on an algorithm that included the presence/absence of ascites, the serum total bilirubin level, and the results of indocyanine green clearance tests [9, 10]. Follow-up protocols were reported elsewhere [9] including regular screening for recurrence by monitoring the plasma levels of alfa-fetoprotein and des-gamma-carboxyprothrombin every 1–2 months, abdominal ultrasound every 3 months, and dynamic computed tomography or magnetic resonance imaging every 6 months. Recurrence was aggressively treated appropriately based on the recommendations at a multidisciplinary discussion.

Pathological Diagnosis

The histological grade of differentiation of the tumor, the degree of fibrosis in the background liver, and the presence/absence of vascular invasion were assessed microscopically based on the classification system proposed by the LCSGJ [3] for Japanese cohort. Tumor grade and degree of fibrosis were described according to Edmondson grade [11] and Ishak score [12], respectively in the SEER database. Diagnosis of liver cirrhosis was based on histopathologic evidence of grade 4 fibrosis according to Desmet’s classification [3, 13] in the training set and F5–6 according to Ishak score in the validation set.

Statistical Analysis

A statistical analysis was performed using JMP 14.0 software (SAS Institute Japan, Tokyo, Japan). The continuous variables were compared using the Mann-Whitney U test. Categorical variables were compared using the Pearson chi-square test or the Fisher exact test, as appropriate. Statistical analysis was two-tailed, and p values of <0.05 were considered to be statistical significance. Survival curves were generated using the Kaplan-Meier method and were compared using the log-rank test.

In addition to the currently adopted cutoff value for solitary HCC [6], an optimal size cutoff value for multiple HCC was also determined based on the results of the minimum p value approach for disease-specific survival after surgery. A new staging model was established based on the results in the training cohort, and its performance was then tested using the validation cohort. Harrel’s c-statistic was used to determine the performance of each staging model.

Baseline Characteristics

The baseline characteristics of the training cohort and the validation cohort are summarized in Table 1. The distributions of age and gender were similar between the 2 cohorts. However, the tumor size was significantly larger, and the presence of multiple lesions was more frequent in the validation cohort. Although the prevalence of microvascular invasion (mvi) was similar, the validation cohort included more cases of advanced disease, and a major hepatectomy was selected more frequently.

Table 1.

Baseline characteristics

Baseline characteristics
Baseline characteristics

Prognostic Impact of mvi

The prognostic impact of mvi was compared in Figure 1 according to size (≤2 vs. >2 cm) and tumor number (solitary vs. multiple). The rationale for adopting a cutoff value of 2 cm for multiple HCC was based on the results of a minimum p value approach to predict disease-specific survival after surgical resection. The hazard ratio (HR) was highest and the p value was lowest at 2.5 cm in the training cohort (HR 1.55; p = 0.132) and 1.7 cm in the validation cohort (HR 2.28; p = 0.0025), respectively, when sliding the cutoff value from 1.1 through 5.0 cm (Fig. 2). When focusing on the group of patients without direct invasion to surrounding organs, tumor rupture, major vascular invasion, nodal involvement, or distant metastases, the presence of mvi had prognostic value in each group in both the training cohort and the validation cohort.

Fig. 1.

Disease-specific survival rates according to the size, number, and presence of microvascular invasion in HCC. mvi, microvascular invasion.

Fig. 1.

Disease-specific survival rates according to the size, number, and presence of microvascular invasion in HCC. mvi, microvascular invasion.

Close modal
Fig. 2.

Optimal size cutoff values for multiple lesions estimated by minimum-p value approach. HR, hazard ratio.

Fig. 2.

Optimal size cutoff values for multiple lesions estimated by minimum-p value approach. HR, hazard ratio.

Close modal

Cutoff values of tumor size for predicting mvi estimated by receiver operating characteristics curve analysis were 53 mm for solitary HCC and 50 mm for multiple lesions, respectively. In a multivariate analysis of prognostic factors for disease-specific survival rates, however, a size >2 cm, multiple lesions, and mvi were independent prognostic factors in both the training cohort and the validation cohort, and their degrees of contribution to disease-specific survival were comparable (HRs 1.4–2.4; Table 2).

Table 2.

Multivariate analysis for disease-specific survival

Multivariate analysis for disease-specific survival
Multivariate analysis for disease-specific survival

Establishment of New Staging Models and Validation

When the AJCC 8th edition was applied to the training cohort, the survival outcomes were not well stratified with the overlapping survival curves for several stages. Based on these observations and the aforementioned clinical outcomes, a new staging model was established using (1) equivalent weighting for size, number, and mvi for the classification of earlier stages of HCC and (2) the reclassification of Stages IIIB and IVA into the same group (Fig. 3, Table 3). Because the purpose of the present study was to investigate the possibility of reclassification for the current version of the AJCC staging system, only clinical parameters used for the AJCC 8th edition were included in the new prognostication model, and the performance of the new model was then tested according to the presence of cirrhosis or tumor grade.

Table 3.

Definition of staging systems and survival outcomes

Definition of staging systems and survival outcomes
Definition of staging systems and survival outcomes
Fig. 3.

Disease-specific survival rates stratified by the AJCC 8th edition and the new classification in the training cohort. AJCC, American Joint Committee on Cancer.

Fig. 3.

Disease-specific survival rates stratified by the AJCC 8th edition and the new classification in the training cohort. AJCC, American Joint Committee on Cancer.

Close modal

The performance of the new classification was better than that of the AJCC 8th edition or the AJCC 7th edition in the training cohort (c-statistics, 0.648 vs. 0.629 vs. 0.632 for disease-specific survival; 0.606 vs. 0.596 vs. 0.594 for overall survival; and 0.592 vs. 0.586 vs. 0.583 for time-to-interventional failure [14]). This tendency was also confirmed in the validation cohort. The discriminatory power of the new classification for disease-specific survival (c-statistics, 0.646) was comparable or even better than that of the AJCC 8th edition (0.645) or the AJCC 7th edition (0.639; online suppl. Fig. 1; for all online suppl. material, see www.karger.com/doi/10.1159/000504193).

Performance of the New Classification According to Underlying Liver Status and Tumor Grade

Figure 4 shows disease-specific survival curves according to the presence of cirrhosis in the validation cohort. In cirrhotic patients, the new classification showed better performance than the AJCC 8th edition in both the training cohort (c-statistics, 0.643 vs. 0.615) and the validation cohort (0.643 vs. 0.642). This tendency was also confirmed in noncirrhotic patients (training set, 0.696 vs. 0.692; validation set, 0.691 vs. 0.686).

Fig. 4.

Disease-specific survival rates according to the presence of cirrhosis by the new classification model in the validation cohort.

Fig. 4.

Disease-specific survival rates according to the presence of cirrhosis by the new classification model in the validation cohort.

Close modal

Similarly, when stratified by tumor grade, the new classification showed comparable to better performance compared with the AJCC 8th edition in both the training cohort (0.620 vs. 0.598) and the validation cohort (0.634 vs. 0.631) for well to moderately differentiated HCC. Comparable performance between the AJCC 8th edition and new classification was also confirmed for poorly differentiated or undifferentiated HCC (training set, 0.709 vs. 0.693; validation set, 0.635 vs. 0.640; online suppl. Fig. 2).

In this study, we revisited the influence of tumor size and mvi on the disease-specific survival of 1,175 patients with HCC. The optimal size cutoff value for multiple HCC was around 2 cm, which was similar to that adopted in the current version of the AJCC staging system for solitary HCC. Our findings indicate that mvi is a significant prognostic factor irrespective of the size and number of tumors. By incorporating these new findings, a new prognostication model was established, and its discriminatory performance was confirmed in 5,249 patients derived from the SEER database.

In the current version of the AJCC staging system, a solitary lesion measuring up to 2 cm in diameter is subclassified as T1a based on the results of a previous international multicenter study [6]. In this study, mvi or tumor differentiation showed no significant contribution to survival in patients with small HCC. Indeed, the prognostic difference was not significant when the “overall” survival of patients with solitary HCC measuring up to 2 cm was examined in the current SEER population (p = 0.0693, data not shown), while a significant difference in disease-specific survival was confirmed, as shown in Figure 1 (p = 0.0047). Because the population of patients with HCC has a relatively advanced age and may have underlying liver diseases, overall survival does not always reflect the oncological prognosis of patients because of the relatively high incidence of other causes of death [15]. As such, disease-specific survival would be more suitable for prognostication of this unique cancer.

Interestingly, while the median tumor size was approximately 5 cm in the validation cohort, the optimal size cutoff value for prognostication of multiple HCC was around 2 cm, which is similar to that for solitary HCC (Fig. 2). Although the current AJCC staging system uses 5 cm for the staging of multiple HCC, 5 cm is an empirical cutoff value for predicting the presence of mvi. In the current SEER population, size cutoff value for mvi was estimated around 5.0–5.3 cm in a receiver-operating characteristics curve analysis. Given the close association between a tumor size of >5 cm and mvi, the new cutoff value of 2 cm would be better to balance the weighting of the listed prognostic factors for the prediction of disease-specific survival in patients with HCC.

The new classification model uses a similar method for staging in earlier stages of HCC, compared with the LCSGJ staging system, while the LCSGJ uses “macrovascular invasion” instead of mvi for staging. As confirmed in the multivariate analysis, the degrees of the contributions to disease-specific survival among size, number, and mvi were similar (Table 2). Importantly, the current classification model showed acceptable performance in both cirrhotic and noncirrhotic patients (Fig. 4). Because presence of cirrhosis independently influences treatment outcomes and prognosis of patients with HCC [16], staging model should be effective apart from the presence of underlying liver disease. Although the actual performance of the new classification was almost comparable to or slightly better than the current version of the AJCC staging system, the current method could be an alternative model that effectively predicts survival outcomes using a “single” size cutoff value.

From a clinical standpoint, a 2 cm cutoff is important because various therapeutic options can be selected, including resection, ablation, and transplantation for solitary or small number of HCCs [17, 18]. Although the treatment of advanced stages of HCC remains challenging even in the era of multidisciplinary treatment, the new classification would better stratify patients who are candidates for curative treatment. As such, the current model could further improve the current version of the AJCC staging system in line with our actual clinical practice. Given the different policy in prognostication between the AJCC staging system (pathological staging system based on the surgical cohort) and the LCSGJ staging system (clinical staging system based on both surgical and medical cohorts), there remain several issues to be solved to create a universally acceptable oncological staging systems for HCC. However, current results indicate the possibility to merge these 2 staging models using the same size cutoff value and similar weight for prognostication and also highlight the next process of discussion including the issue of vascular invasion (macrovascular vs. microvascular), tumor grade (well to moderate differentiation vs. poor differentiation), or status of underlying liver (cirrhosis vs. noncirrhosis) to establish a universal prognostication model.

The limitations of this study include its retrospective nature and selected population. Also, several items of information were not accessible in the SEER database, including underlying liver disease, hepatic function, surgery detail, site of recurrence and repeated treatment, and so on, and accordingly, it is difficult to establish a completely new prognostication model using parameters not listed in the current version of AJCC staging system. However, the new model was established using prospectively collected data at a high-volume center in Japan, and the reproducibility of the results was confirmed using Western population-based data. In addition, the present validation cohort had similar baseline characteristics compared to the international multicenter database, which was used to establish the current version of the AJCC staging system: median tumor size, 50 vs. 52 mm; multiple tumors, 22.2 vs. 32.9%; mvi, 26.8 vs. 40.2%; cirrhosis 46.3 vs. 47.1%; and tumor grade (well differentiated/moderately differentiated/poor differentiated/ undifferentiated), 21.5/53.5/22.8/2.1 vs. 16.4/53.4/29.7/0.4% (personal communication). Therefore, the present results may warrant an international multicenter validation study to strengthen the current outcomes.

In conclusion, a size cutoff value of 2 cm and mvi were equally prognostic for both solitary and multiple HCC regardless of the presence of cirrhosis. The inclusion of these factors could be an alternative prognostication model using a single size-cutoff value for HCC.

None.

The Institutional Review Board of Toranomon Hospital approved this study protocol (No. 1823). This study was conducted according to the guidelines for human studies and the Declaration of Helsinki.

The authors have no conflicts of interest to declare.

This study is supported by the institutional research grand for clinical study at Toranomon Hospital.

J.S.: study conception. J.S., Y. Kobayashi., Y. Kawamura.: analysis and interpretation of data. All authors: acquisition of data, drafting of manuscript, and final approval.

1.
Huitzil-Melendez
FD
,
Capanu
M
,
O’Reilly
EM
,
Duffy
A
,
Gansukh
B
,
Saltz
LL
, et al
Advanced hepatocellular carcinoma: which staging systems best predict prognosis?
J Clin Oncol
.
2010
Jun
;
28
(
17
):
2889
95
.
[PubMed]
0732-183X
2.
Amin
M
,
Edge
S
,
Greene
F
,
Byrd
D
,
Washington
M
,
Gershnwald
J
, et al
AJCC Cancer Staging Manual
.
Springer International Publishing
;
2017
.
3.
Liver Cancer Study Group of Japan
.
The general rules for the clinical and pathological study of primary liver cancer
. 5th ed.
Tokyo
:
Kanehara
;
2008
.
4.
Kudo
M
,
Izumi
N
,
Sakamoto
M
,
Matsuyama
Y
,
Ichida
T
,
Nakashima
O
, et al;
Liver Cancer Study Group of Japan
.
Survival Analysis over 28 Years of 173,378 Patients with Hepatocellular Carcinoma in Japan
.
Liver Cancer
.
2016
Jul
;
5
(
3
):
190
7
.
[PubMed]
2235-1795
5.
Kudo
M
,
Kitano
M
,
Sakurai
T
,
Nishida
N
.
General Rules for the Clinical and Pathological Study of Primary Liver Cancer, Nationwide Follow-Up Survey and Clinical Practice Guidelines: The Outstanding Achievements of the Liver Cancer Study Group of Japan
.
Dig Dis
.
2015
Oct
;
33
(
6
):
765
70
.
[PubMed]
0257-2753
6.
Shindoh
J
,
Andreou
A
,
Aloia
TA
,
Zimmitti
G
,
Lauwers
GY
,
Laurent
A
, et al
Microvascular invasion does not predict long-term survival in hepatocellular carcinoma up to 2 cm: reappraisal of the staging system for solitary tumors
.
Ann Surg Oncol
.
2013
Apr
;
20
(
4
):
1223
9
.
[PubMed]
1068-9265
7.
Vauthey
JN
,
Lauwers
GY
,
Esnaola
NF
,
Do
KA
,
Belghiti
J
,
Mirza
N
, et al
Simplified staging for hepatocellular carcinoma
.
J Clin Oncol
.
2002
Mar
;
20
(
6
):
1527
36
.
[PubMed]
0732-183X
8.
Minagawa
M
,
Ikai
I
,
Matsuyama
Y
,
Yamaoka
Y
,
Makuuchi
M
.
Staging of hepatocellular carcinoma: assessment of the Japanese TNM and AJCC/UICC TNM systems in a cohort of 13,772 patients in Japan
.
Ann Surg
.
2007
Jun
;
245
(
6
):
909
22
.
[PubMed]
0003-4932
9.
Kobayashi
Y
,
Kiya
Y
,
Sugawara
T
,
Nishioka
Y
,
Hashimoto
M
,
Shindoh
J.
Expanded Makuuchi's criteria using estimated indocyanine green clearance rate of future liver remnant as a safety limit for maximum extent of liver resection.
HPB (Oxford).
2019
.
10.
Makuuchi
M
,
Kosuge
T
,
Takayama
T
,
Yamazaki
S
,
Kakazu
T
,
Miyagawa
S
, et al
Surgery for small liver cancers
.
Semin Surg Oncol
.
1993
Jul-Aug
;
9
(
4
):
298
304
.
[PubMed]
8756-0437
11.
Edmondson
HA
,
Steiner
PE
.
Primary carcinoma of the liver: a study of 100 cases among 48,900 necropsies
.
Cancer
.
1954
May
;
7
(
3
):
462
503
.
[PubMed]
0008-543X
12.
Ishak
K
,
Baptista
A
,
Bianchi
L
,
Callea
F
,
De Groote
J
,
Gudat
F
, et al
Histological grading and staging of chronic hepatitis
.
J Hepatol
.
1995
Jun
;
22
(
6
):
696
9
.
[PubMed]
0168-8278
13.
Desmet
VJ
,
Gerber
M
,
Hoofnagle
JH
,
Manns
M
,
Scheuer
PJ
.
Classification of chronic hepatitis: diagnosis, grading and staging
.
Hepatology
.
1994
Jun
;
19
(
6
):
1513
20
.
[PubMed]
0270-9139
14.
Shindoh
J
,
Kawamura
Y
,
Kobayashi
Y
,
Akuta
N
,
Kobayashi
M
,
Suzuki
Y
, et al
Time-to-Interventional Failure as a New Surrogate Measure for Survival Outcomes after Resection of Hepatocellular Carcinoma
.
J Gastrointest Surg
.
2019
Jun
.
[PubMed]
1091-255X
15.
Shindoh
J
,
Makuuchi
M
,
Matsuyama
Y
,
Mise
Y
,
Arita
J
,
Sakamoto
Y
, et al
Complete removal of the tumor-bearing portal territory decreases local tumor recurrence and improves disease-specific survival of patients with hepatocellular carcinoma
.
J Hepatol
.
2016
Mar
;
64
(
3
):
594
600
.
[PubMed]
0168-8278
16.
Sasaki
K
,
Shindoh
J
,
Margonis
GA
,
Nishioka
Y
,
Andreatos
N
,
Sekine
A
, et al
Effect of Background Liver Cirrhosis on Outcomes of Hepatectomy for Hepatocellular Carcinoma
.
JAMA Surg
.
2017
Mar
;
152
(
3
):
e165059
.
[PubMed]
2168-6254
17.
Llovet
JM
,
Burroughs
A
,
Bruix
J
.
Hepatocellular carcinoma
.
Lancet
.
2003
Dec
;
362
(
9399
):
1907
17
.
[PubMed]
0140-6736
18.
Kokudo
N
,
Hasegawa
K
,
Akahane
M
,
Igaki
H
,
Izumi
N
,
Ichida
T
, et al
Evidence-based Clinical Practice Guidelines for Hepatocellular Carcinoma: The Japan Society of Hepatology 2013 update (3rd JSH-HCC Guidelines)
.
Hepatol Res
.
2015
Jan
;
45
(
2
):
n/a
.
[PubMed]
1386-6346
Open Access License / Drug Dosage / Disclaimer
This article is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License (CC BY-NC-ND). Usage and distribution for commercial purposes as well as any distribution of modified material requires written permission. Drug Dosage: The authors and the publisher have exerted every effort to ensure that drug selection and dosage set forth in this text are in accord with current recommendations and practice at the time of publication. However, in view of ongoing research, changes in government regulations, and the constant flow of information relating to drug therapy and drug reactions, the reader is urged to check the package insert for each drug for any changes in indications and dosage and for added warnings and precautions. This is particularly important when the recommended agent is a new and/or infrequently employed drug. Disclaimer: The statements, opinions and data contained in this publication are solely those of the individual authors and contributors and not of the publishers and the editor(s). The appearance of advertisements or/and product references in the publication is not a warranty, endorsement, or approval of the products or services advertised or of their effectiveness, quality or safety. The publisher and the editor(s) disclaim responsibility for any injury to persons or property resulting from any ideas, methods, instructions or products referred to in the content or advertisements.