- Email: [email protected]
The ART-strategy: Sequential assessment of the ART score predicts outcome of patients with hepatocellular carcinoma re-treated with TACE
Research Article
The ART-strategy: Sequential assessment of the ART score predicts outcome of patients with hepatocellular carcinoma re-treated with TACE Florian Hucke1, Wolfgang Sieghart1,⇑, Matthias Pinter1, Ivo Graziadei2, Wolfgang Vogel2, Christian Müller1, Harald Heinzl3, Fredrik Waneck4, Michael Trauner1, Markus Peck-Radosavljevic1,⇑ 1
Division of Gastroenterology and Hepatology, Department of Internal Medicine III, AKH and Medical University of Vienna, Austria; Department of Gastroenterology and Hepatology, LKH and Medical University of Innsbruck, Austria; 3Section for Clinical Biometrics, Medical University of Vienna, Austria; 4Department of Interventional Radiology, AKH and Medical University of Vienna, Austria
2
Background & Aims: Recently, we developed the ART score (assessment for re-treatment with TACE) to guide the decision for a second transarterial chemoembolization (TACE-2) in patients with hepatocellular carcinoma (HCC). Patients with an ART score of 0–1.5 points gained benefit from a second TACE session, while patients with an ART score P2.5 points did not. Here, we investigated (1) the prognostic significance of the ART score prior to the third (TACE-3) and fourth TACE (TACE-4), and (2) the feasibility of an ART score guided re-treatment strategy by sequential assessment of the ART score in HCC patients treated with multiple TACE sessions. Methods: 109 patients, diagnosed with intermediate stage HCC and treated with P3 TACE sessions between January 1999 and December 2009 at the Medical Universities of Vienna and Innsbruck, were included. The ART score prior to each TACE session was assessed in comparison to the TACE naïve liver. The prognostic performance of the ART score before TACE-3 and 4 was evaluated with and without stratification based on the ART score prior to the respective last intervention. Results: The pre-TACE-3 ART score discriminated two groups with different prognosis and remained a valid predictor of OS independent of Child-Pugh score (5–7 points), CRP-levels and tumor characteristics. Even in patients with an initially beneficial ART score (0–1.5 points) before TACE-2, repeated ART score assessment before TACE-3 identified a subgroup of patients with dismal prognosis (median OS: 27.8 vs. 10.8 months, p <0.001). Similar results were observed when the ART score was applied before TACE-4. Conclusions: The sequential assessment of the ART score identifies patients with dismal prognosis prior to each TACE session.
Keywords: Hepatocellular carcinoma; TACE; Prognosis; HCC; ART score. Received 2 April 2013; received in revised form 11 July 2013; accepted 20 August 2013; available online 3 September 2013 ⇑ Corresponding authors. Address: AKH & Medizinische Universität Wien, Klinik Innere Medizin III, Abteilung Gastroenterologie & Hepatologie, Währinger Gürtel 18-20, A-1090 Wien, Austria. Tel.: +43 1 40400 6589; fax: +43 1 40400 4735. E-mail addresses: [email protected] (W. Sieghart), markus.peck@ meduniwien.ac.at (M. Peck-Radosavljevic).
Ó 2013 European Association for the Study of the Liver. Published by Elsevier B.V. All rights reserved.
Introduction Hepatocellular carcinoma (HCC), is the 5th most common cancer worldwide, and develops predominately in patients with liver cirrhosis [1]. The most widely adopted staging system for HCC is the Barcelona Clinic Liver Cancer (BCLC) staging system [2,3], which was endorsed by the European Association for the Study of the Liver (EASL) [4,5]. It distinguishes five tumor stages and links them to different treatment options. Curative therapies, like liver transplantation (OLT), liver resection, percutaneous ethanol injection (PEI) and radiofrequency ablation (RFA) are recommended only for early stages. However, most patients are diagnosed at advanced tumor stages [6]. Based on a meta-analysis of randomized controlled trials [7], transarterial chemoembolization (TACE) is the recommended treatment modality for asymptomatic, large or multifocal HCC without macrovascular invasion or extrahepatic metastasis (intermediate HCC, BCLC stage B). The outcome results of several TACE studies vary significantly [7,8]. Since the group of patients at BCLC stage B is very heterogeneous [9] in terms of liver function and tumor load (Child-Pugh stage A/B, tumor burden), there are patients who may not profit from repeated TACE sessions. In this context, we recently developed the ART score (assessment for re-treatment with TACE) (Table 1) [10]. This new and simple point score integrates radiologic tumor response (present vs. absent) and impairment of liver function (presence vs. absence of Child-Pugh score increase by 1 or P2 points and AST increase by 25% from pre TACE-1, respectively) after the first TACE to predict patient survival if re-treated with a second TACE. The ART score selects two distinct patient groups with significantly different prognosis and is able to identify patients who may probably not benefit from a second TACE sessions. These results were confirmed in an independent external validation cohort.
Journal of Hepatology 2014 vol. 60 j 118–126
JOURNAL OF HEPATOLOGY Table 1. ART-score.
Points Radiologic tumor response Absent Present AST increase >25% Present Absent Child-Pugh score increase 1 point ≥2 points Absent
1 0 4 0 1.5 3 0
AST, aspartate transaminase.
In our initial analysis [10], the prognostic performance of the ART score was only validated at one time point: prior to the second TACE session. However, most intermediate stage HCC patients are treated with more than two TACE sessions in a sequential manner [7]. Therefore, confirmation of the prognostic performance of repeated, sequential assessment of the ART score, would provide a strong rationale for a prospective study evaluating an ART score guided TACE strategy. The aim of this study was to investigate (1) the prognostic significance of the ART score prior the third (TACE-3) and fourth TACE (TACE-4) and (2) the feasibility of an ART score guided retreatment strategy by sequential assessment of the ART score in HCC patients treated with multiple TACE sessions.
Baseline (before TACE-1) imaging (triphasic CT/MRI scan) was performed 5–7 d before the first TACE session. HCC was staged according to the BCLCclassification [2,3] and by the International Union Against Cancer (UICC) tumor node metastasis (TNM) classification 6th edition [13]. In both institutions, radiologic tumor response was assessed either by CT (96%) or MRI (4%) scan prior to each TACE session (maximal 90 d after the first or second TACE, respectively) according to EASL criteria [4]. The same modality was used in single patients before and after each TACE session. Overall 7 patients (6%) received their 1st TACE session between 1999 and August 2001 (time before the implementation of the EASL criteria) and were originally evaluated according to WHO criteria [4]. These patients were re-evaluated according to EASL criteria for this study. Objective tumor response following each TACE was defined as complete response (CR) and partial response (PR), while stable disease (SD) and progressive disease (PD) were judged as absence of objective tumor response. All laboratory values including alpha-fetoprotein (AFP) and CRP levels [14] as well as liver function parameters including the Child-Pugh score [15] were determined one day before each TACE session. AFP response prior to the third TACE was defined as an AFP decrease by 50% from pre-TACE-1 values of P200 kU/L [16]. Three AFP groups were formed for univariate analysis: pre-TACE-1 AFP P200 kU/L with response vs. pre-TACE-1 AFP P200 kU/L and no response vs. pre-TACE-1 AFP levels <200 kU/L. Adverse events, which occurred within four weeks after any TACE or were unequivocally TACE related, were documented according to the Common Terminology Criteria for Adverse Events version 3.0 (CTCAE) [17]). ART score determination Details about the ART score development have recently been reported in [9]. In short, the ART score consists of three variables (Table 1): AST increase by 25% (ART score: 4 points), Child-Pugh increase by 1 or P2 points (ART score: 1.5 and 3 points, respectively) as well as absence of radiologic tumor response (ART score: 1 point). Based on the sum of all three variables, the ART score is calculated. A cut-off of 0–1.5 points and P2.5 points [10] was used to stratify patients for all survival analysis. In this study, the ART score was determined 1 day prior to TACE-2, TACE-3, and TACE-4, respectively. For this purpose, all ART score forming variables were determined with reference to the TACE naïve liver (before TACE-1).
Materials and methods Study design Eligibility Similar to our recent study [10], we used the following inclusion criteria (Fig. 1A): all patients (>18 yr), who were diagnosed with HCC by histology or dynamic imaging (CT/MRI scans) according to EASL diagnostic criteria [4] and treated with cTACE, TAE or DEB-TACE (thereafter summarized and referred to as TACE), at the Department of Gastroenterology and Hepatology of the Medical University of Vienna and Innsbruck between January 1999 and December, 2009 were included in this study. Patients were eligible if they had HCC at BCLC-stage A (not suitable for OLT or local ablative therapy) or B, preserved liver function (Child-Pugh stage A or B) and received at least three TACE sessions (time between the TACE cycles each 690 days). Patients were excluded, if they received TACE before OLT or resection, or if patients received TACE for HCC recurrence after OLT. Additionally, patients who received TACE despite poor liver function (Child-Pugh C) and patients at BCLC-stage C were excluded. In order to examine the effect of sequential assessment of the ART score in patients prior to the third and fourth TACE, we furthermore excluded patients in which the ART score prior to the respective last TACE was not determinable due to missing variables. The selection of the final study population is given in Fig. 1A. TACE procedure The procedure of transarterial embolization (TAE), conventional lipiodol-doxorubicin TACE (cTACE) and TACE with drug-eluting beads (DEB-TACE) in both centers was described in the original ART score publication [10]. In short, both institutions used a ‘‘treatment on demand’’ [11] TACE-schedule and no further TACE session was performed in presence of complete radiologic response. Additionally, presence of Child-Pugh C cirrhosis, portal vein thrombosis or ECOG >1 was considered as contraindication for re-treatment with TACE [12]. Collection of data
Based on the inclusion criteria of this study, we identified all patients at the Medical University of Vienna and Innsbruck, who underwent multiple TACE sessions and for which the ART score was determinable prior to TACE-2, TACE-3, and TACE-4. We pooled the two patient cohorts in order to achieve higher statistical power for further analysis. Given that the ART score was developed and validated prior to TACE-2, we first aimed to validate the prognostic value of the ART score if assessed prior to TACE-3 or TACE-4 (Fig. 1B). In a second step, we analyzed the prognostic significance of the sequential assessment of the ART score. For this purpose we focused on patients who were supposed to be in a beneficial prognostic group based on an ART score of 0–1.5 points prior to the respective last TACE session and investigated their outcome based on their ART score prior to the following TACE session (Fig. 1C). Statistical analyses Patient characteristics are presented with descriptive statistics. Overall survival was defined as the time from the day of the respective ART score assessment until death or last follow-up. Survival curves were calculated using the Kaplan–Meier method. Median survival times (OS) and their 95% confidence intervals (CIs) were reported. The log-rank test and a multivariate Cox regression model were used to assess the effects of patient variables prior to each TACE on OS. All reported p-values are results of two-sided tests. A significance level of 0.05 was applied throughout. Statistical analyses were performed using IBM SPSS version 20.0 (SPSS Inc., Armonk, NY) and SAS 9.3 (SAS Institute Inc., Cary, NC).
Results Patient characteristics
The retrospective analysis of data was approved by the Ethics Committees of the Medical Universities of Vienna and Innsbruck.
Patient characteristics are shown in Table 2. Prior to the first TACE, the majority of patients was at BCLC-stage B (82%) and
Journal of Hepatology 2014 vol. 60 j 118–126
119
Research Article
Patients treated with TACE with determinable ART-score before TACE 3 n = 109
Patients treated with TACE with determinable ART-score before TACE 4 n = 43
Excluded, n = 279 • Liver transplantation prior or post-TACE/resection post first TACE, n = 95 • Child-Pugh C, n = 32 • BCLC-C, n = 50 • Patients with less than 2 TACE, n = 55 • Interval TACE 1-TACE 2 >90 days, n = 29 • ART-score before TACE 2 not determinable due to missing values, n = 18
Patients treated with ≥3 TACE cycles
TACE 1 ≤90 days
Patients treated with TACE with determinable ART-score before TACE 2 n = 204
B
TACE 2 Excluded, n = 95 • Patients with less than 3 TACE, n = 52 • Patients with ≥3 TACE but interval TACE 2-TACE 3 >90 days, n = 34 • ART-score before TACE 3 not determinable due to missing values, n = 9 Excluded, n = 67 • Patients with less than 4 TACE, n = 38 • Patients with ≥4 TACE but interval TACE 3-TACE 4 >90 days, n = 26 • ART-score before TACE 4 not determinable due to missing values, n = 3
≤90 days
Patients treated with TACE at the Medical University of Vienna and Innsbruck n = 483
ART-score before TACE 3
TACE 3 ≤90 days
A
ART-score before TACE 4
TACE 4
Overall survival
Death or censored
C
Patients treated with multiple TACE cycles
TACE 1 ART-score before TACE 2
≥2.5 points 0-1.5 points TACE 2 ART-score before TACE 3
≥2.5 points 0-1.5 points TACE 3 ART-score before TACE 4
≥2.5 points 0-1.5 points TACE 4
OUTCOME Fig. 1. Patient selection and study design. (A) Patient selection and final study population. (B) Study design for ART score validation pre-TACE-3 and pre-TACE-4. (C) Study design for sequential assessment of the ART score.
had a Child-Pugh A cirrhosis (61%). TACE was the first-line therapy in 88 (81%) of patients, while 21 (19%) received a prior antitumor therapy (liver resection (n = 9), PEI (n = 7), RFA (n = 5)).
120
Chemoembolization was the most common treatment technique (n = 95, 87%) (cTACE vs. DEB-TACE n = 85 vs. n = 10), while in 13% of cases (n = 14) TAE only was performed.
Journal of Hepatology 2014 vol. 60 j 118–126
JOURNAL OF HEPATOLOGY Table 2. Patient characteristics.
N Before TACE-1 Age (yr) Mean ± SD Range Sex Male Female Etiology Alcohol Viral (HBV, HCV) Others Unknown Child-Pugh stage A B 7 points ≥B 8 points Tumor size Mean ± SD Range Tumor extent Unilobar Bilobar Tumor number Unifocal Multifocal BCLC-stage A B TNM stage I II III Prior anti-tumor therapy Absent Present α-Fetoprotein (kU/L) <200 ≥200 Before TACE-2 Child-Pugh stage A B 7 points ≥B 8 points Before TACE-3 Child-Pugh stage A B 7 points ≥B 8 points Before TACE-4 Child-Pugh stage A B 7 points ≥B 8 points
% 109
63.8 ± 9.3 42-82 99 10
91 9
52 37 11 9
48 34 10 8
77 14 18
61 13 17
5.1 ± 2.6 1.3-12.4 51 58
47 53
33 76
30 70
20 89
18 82
19 44 46
17 40 42
88 21
81 19
84 25
77 23
75 16 18
69 15 17
71 20 18
65 18 17
29 4 10
63 9 22
HBV, hepatitis B virus infection; HCV, hepatitis C virus infection; BCLC, Barcelona Clinic Liver Cancer; TNM, tumor nodes metastasis.
Between TACE-1 and TACE-3, 36 patients (33%) suffered from a Child-Pugh increase by at least 1 point, while 57 patients (52%) showed no change and 16 patients (15%) showed a decrease of the Child-Pugh score by at least 1 point. Between TACE-1 and TACE-4, 17 patients (40%) suffered from a Child-Pugh increase by at least 1 point, while 23 patients (53%) showed no change and 3 patients (7%) showed a decrease of the Child-Pugh score by 1 point. Overall, the median number of TACE interventions was 4 (range 3–20). The median time interval between the TACE sessions was 41 d (range 13–85) between the first and second TACE, 42 d (range 19–89) between the second and third TACE, and 46 d (range 21–85) between the third and fourth TACE. Radiologic tumor response after TACE-2, 3, and 4 is given in Supplementary Table 1. Univariate analysis of prognostic factors prior to the third and fourth TACE During follow-up, 86% of patients (n = 94) died, while 14% were still alive (n = 10) or lost to follow-up (n = 5). The median OS of the whole population was 17.1 months (95% CI, 11.4–22.7) (as calculated from the date of TACE-1 until death or date last seen). The median time of follow-up was 62.7 months. In the univariate analysis of prognostic factors prior to the third TACE, Child-Pugh stage (OS: CP A vs. CP B7 vs. CP PB8: 23.0 vs. 12.6 vs. 7.2 months, p <0.001), tumor extent (OS: unilobar vs. bilobar: 24.1 vs. 12.6 months, p = 0.040) and tumor number (OS: unifocal vs. multifocal: 31.7 vs. 12.6 months, p = 0.040), CRP levels (OS: normal vs. elevated: 24.1 vs. 10.0 months, p = 0.003) and AFP response (OS: AFP response vs. no AFP response vs. baseline AFP <200 kU/L: 15.1 vs. 8.2 vs. 19.0 months, p = 0.008) had a significant impact on overall survival (Table 3). Of note, the single ART score variables AST increase >25% (OS: no AST increase vs. AST increase >25%: 20.3 vs. 8.9 months, p = 0.001), a Child-Pugh increase by 1 or >2 points (OS: no CP increase vs. CP increase 1 Point vs. CP increase P2 Points: 26.5 vs. 10.2 vs. 5.2 months, p <0.001), and absence of radiologic tumor response (OS: response vs. no response: 21.8 vs. 11.3 months, p = 0.048) had a significant impact on overall survival upon univariate analysis. Prior to the fourth TACE, only higher Child-Pugh stage and elevated CRP levels were associated with worse overall survival (data not shown), likely due to the small sample size.
The ART score prior to the third and fourth TACE is an independent predictor of overall survival Supplementary Tables 5 and 6 outline the different constellations of how patients gained their ART score points prior to TACE-3 and 4. Overall, patients with an ART score before TACE-3 of 0–1.5 points had a median OS of 28.1 months [95% CI: 21.7–24.4] while those with P2.5 points had a median OS of 8.5 [95% CI: 7.4–9.6] (p <0.001) (Fig. 2A). Similar results were observed if the ART score was assessed before TACE-4 (Fig. 2C). Additionally, these results could be confirmed upon separate analysis of patients treated at the Medical University of Vienna (Fig. 2B left and 2D left) and patients who were treated at the Medical University of Innsbruck (Fig. 2B right, Fig. 2D right). The ART score remained a significant prognostic factor independent of the TACE technique (Supplementary Fig. 1A–C,
Journal of Hepatology 2014 vol. 60 j 118–126
121
Research Article Table 3. Univariate analysis of prognostic factors before TACE-3 in HCC patients treated with TACE.
Variable Age Child-Pugh stage
Etiology Tumor extent Tumor number Tumor size (cm) BCLC-stage TNM stage
Prior therapy CRP values α-Fetoprotein (kU/L) AFP response¹
TACE type
AST increase >25% Child-Pugh increase
Radiologic tumor response
<65 ≥65 A B 7 points ≥B 8 points Viral Other Unilobar Bilobar Unifocal Multifocal <7.5 ≥7.5 A B I II III Absent Present <1 mg/dl ≥1 mg/dl <200 ≥200 AFP >200: no response AFP >200: response AFP <200 TAE cTACE DEB-TACE Absent Present 0 points 1 point ≥2 points Present Absent
N = 109 59 50 71 20 18 72 37 51 58 33 76 85 24 20 89 19 44 46 88 21 62 41 84 25 11 14 84 14 85 10 85 24 73 23 13 56 53
Overall survival (mo) Median 95% CI 18.6 7.0-30.2 17.1 13.3-20.8 23.0 16.7-29.3 12.6 6.1-19.1 7.2 3.3-11.1 17.1 9.5-24.6 18.0 7.7-28.4 24.1 13.8-34.4 12.6 7.8-17.3 31.7 16.3-47.0 12.6 6.0-19.2 18.2 12.4-24.0 12.6 4.1-21.1 24.9 2.7-47.0 15.1 9.1-21.0 24.9 2.0-47.7 18.2 7.6-28.8 14.9 9.7-20.1 17.1 11.2-22.9 18.0 10.9-25.2 24.1 17.4-30.8 10.0 7.2-12.8 19.0 11.8-26.2 11.3 9.7-12.9 8.2 3.8-22.7 4.5-25.6 15.1 19.0 11.8-26.2 14.9 2.0-27.8 15.6 9.9-21.4 19.0 8.2-29.8 20.3 13.4-27.2 8.9 7.7-10.1 26.5 22.1-31.0 10.2 7.2-13.3 5.2 2.5-7.9 21.8 12.0-31.6 11.3 6.1-16.0
p value (Log-rank) 0.921
<0.001 0.954 0.040 0.040 0.758 0.073
0.162 0.980 0.003 0.330
0.008
0.700 0.001
<0.001 0.048
BCLC, Barcelona Clinic Liver Cancer; TNM, tumor nodes metastasis; TAE, transarterial embolization; cTACE, conventional transarterial chemoembolization; DEB-TACE, drugeluting beads transarterial chemoembolization. 1 AFP response was defined by a decrease of >50% from baseline levels.
Supplementary Table 2), Child-Pugh stage (A-B7) prior to the respective TACE session, elevated CRP levels and presence of ascites. Additionally, an ART score of P2.5 points was associated with a higher frequency of serious adverse events after TACE-3 (Supplementary Table 4). Of note, patients at Child-Pugh stage PB8 prior to the third TACE had a similar dismal prognosis in both ART score groups (median OS: 0–1.5 points vs. P2.5 points: 7.9 vs. 6.0 [95% CI: 1.6–14.2 vs. 2.4–9.5], p = 0.404) (Supplementary Fig. 2C). Similar results were observed before TACE-4: while the ART score further differentiated survival in patients with Child-Pugh A cirrhosis (OS 122
[95% CI] for ART score 0–1.5 (n = 19) vs. P2.5 (n = 10): 25.1 vs. 4.4 months [19.2–31.1 vs. 0–11.1]), no discriminative capability was observed in patients with Child-Pugh B cirrhosis. Sample size before TACE-3 allowed inclusion of significant variables of the univariate analysis into multivariate analysis (Table 4). Only Child-Pugh stage prior to the third TACE (hazard ratio [95% CI]: CP B7 vs. CP PB8: 1.4 vs. 2.5 [0.8–2.5 vs. 1.2–5.2], p = 0.012) and the ART score P2.5 before TACE-3 (hazard ratio [95% CI], 3.1 [1.7–5.5], p <0.001) remained independent significant predictors of overall survival.
Journal of Hepatology 2014 vol. 60 j 118–126
JOURNAL OF HEPATOLOGY Medical University of Vienna
0-1.5 points (n = 67): 28.1 [95% CI: 21.7-24.4] ≥2.5 points (n = 42): 8.5 [95% CI: 7.4-9.6]
B
1.0
0-1.5 points (n = 35): 24.1 [95% CI: 16.1-32.1] ≥2.5 points (n = 23): 8.5 [95% CI: 7.5-9.5]
1.0
1.0
0.8 0.6 0.4
ART score 0-1.5 points ≥2.5 points
0.2 0.0
0.6 0.4
ART score 0-1.5 points ≥2.5 points
0.2 0.0
0
20
40
60 80 Months
100 120
0-1.5 points (n = 23): 23.9 [95% CI: 14.7-33.1] ≥2.5 points (n = 20): 8.6 [95% CI: 7.5-9.8]
D
20
40
60 80 Months
ART score 0-1.5 points ≥2.5 points
0.0
Cumulative survival
0.6
40
60 80 Months
0.2
0
20
40
100 120
60 80 Months
100 120
0-1.5 points (n = 13): 35.5 [95% CI: 22.8-48.3] ≥2.5 points (n = 10): 6.4 [95% CI: 0-16.8]
1.0 p = 0.033
0.8 0.6 0.4
ART score 0-1.5 points ≥2.5 points
0.2 0.0
20
ART score 0-1.5 points ≥2.5 points
p = 0.024
0.8
0
0.4
100 120
1.0
0.2
0.6
0-1.5 points (n = 10): 17.0 [95% CI: 13.8-20.2] ≥2.5 points (n = 10): 8.6 [95% CI: 7.8-9.4]
p = 0.003
0.4
0.8
0.0 0
1.0 Cumulative survival
p <0.001
0.8
Cumulative survival
p <0.001
Cumulative survival
Cumulative survival
p <0.001
C
0-1.5 points (n = 32): 39.1 [95% CI: 28.5-49.7] ≥2.5 points (n = 19): 8.0 [95% CI: 3.6-12.5]
Cumulative survival
A
Medical University of Innsbruck
0.8 0.6 0.4
ART score 0-1.5 points ≥2.5 points
0.2 0.0
0
20
40
60 80 Months
100 120
0
20
40
60 80 Months
100 120
Fig. 2. Prognostic significance of the ART score before TACE-3 and before TACE-4. (A) Prognostic significance of the ART score before TACE-3 in all HCC patients and (B) patients treated at the Medical University of Vienna (left) and Innsbruck (right). (C) Prognostic significance of the ART score before TACE-4 in all HCC patients and (D) patients treated at the Medical University of Vienna (left) and Innsbruck (right). Analysis was performed using the Kaplan Meier method (OS in months according to Logrank test).
Finally, to reassure the significance of all single ART score variables prior to TACE-3, we performed a sensitivity-multivariate analysis including the three ART score variables and all other significant variables of the univariate analysis (Table 3). Crucially, only AST increase >25%, absence of radiologic response and Child-Pugh increase of 1 or >2 points remained independent significant predictors of overall survival (Supplementary Table 3). Due to low sample size, no multivariate analysis could be performed before TACE-4. Sequential assessment of the ART score prior to each TACE session identifies patient subgroups with different prognosis Finally we analyzed the prognostic significance of the sequential application of the ART score (Fig. 1C). The ART score prior to TACE-2 identified 77 patients with good prognosis (median OS 28.0 months [95% CI: 24.0–31.9], Fig. 3). Crucially, the sequential assessment of the ART score prior to TACE-3 in this population (n = 77) identified two further subgroups with different outcome: patients with an ART score of 0–1.5 points (n = 62) prior to the third TACE had a median OS of 28.7 months [95%CI: 22.3–35.2], while one-fifth of patients (ART score P2.5 points, n = 15) showed a worse prognosis of 10.8 months [95% CI: 0–21.8]
(p = 0.001) (Fig. 3A). Similar results were observed if the ART score was sequentially assessed before TACE-4 (Fig. 3). Conversely, in patients who were in the dismal prognosis group prior to the second TACE (P2.5 points, n = 32, median OS [95% CI]: 7.9 [5.9–10.0]) and received a further TACE cycle, the sequential assessment of the ART score could not discriminate prognosis (Fig. 3): (median OS [95% CI]: 0–1.5 points vs. P2.5 points: 7.9 vs. 7.9 months [5.0–10.9 vs. 4.4–11.4], p = 0.159). The significance of the ART score before TACE-4 in patients who were in the dismal prognosis group prior to the third TACE (P2.5 points, n = 15, median OS [95% CI]: 10.8 [0–21.8]) could not be assessed due to low sample size (n = 5).
Discussion Most HCC patients allocated to TACE suffer from liver cirrhosis, which may limit the potential survival benefit of this procedure. This is further complicated by the fact that typically HCC patients suitable for TACE mostly need P3 TACE sessions for optimal radiologic tumor response [7,11]. Therefore repeated assessment of liver function and TACE derived liver damage is of paramount importance for re-treatment decisions. The optimal strategy for
Journal of Hepatology 2014 vol. 60 j 118–126
123
Research Article Table 4. Multivariable Cox regression analysis before TACE-3 of prognostic factors in patients with HCC treated with TACE.
Variable Child-Pugh stage A B 7 points ≥B 8 points Tumor number Unifocal Multifocal Tumor extent Unilobar Bilobar CRP-values <1 mg/dl ≥1 mg/dl AFP response AFP <200 kU/L AFP >200: response AFP >200: no response ART-score 0-1.5 points ≥2.5 points
Overall survival p value HR 95% CI (Cox regression) 1 1.4 2.5
0.8-2.5 1.2-5.2
0.012
1 1.2
0.6-2.2
0.604
1 1.2
0.7-2.2
0.518
1 1.5
0.9-2.4
0.216
1 0.9 2.0
0.5-1.9 0.9-4.5
0.196
1 3.1
1.7-5.5
<0.001
HCC, hepatocellular carcinoma; AFP, a-fetoprotein.
re-treatment with TACE would weigh tumor response and liver deterioration after TACE to optimize the overall clinical outcome of these patients. In this context, we recently developed a score for the assessment of re-treatment with TACE (ART score) [10]. The ART score is a simple point score, which integrates radiologic tumor response, liver damage and impairment of liver function after the first TACE to predict patient’s prognosis after re-treatment with TACE (Table 1). However, the ART score was only developed and validated prior to the second TACE session. Yet, the prognostic significance of the ART score prior to the third and fourth TACE session and, in particular, the value of a repeated sequential assessment of the ART score within a defined patient population undergoing multiple TACE sessions remains unclear. In the current study, sensitivity analysis prior to TACE-3 again identified an AST increase by >25%, Child-Pugh increase by 1 or >2 points and absence of radiologic response as independent predictors of overall survival (Supplementary Table 3). Additionally, the ART score prior to TACE-3 and -4 discriminated two different groups of patients with distinct prognosis (Fig. 2A–D). The ART score retained significance in several clinically important subgroups (Supplementary Figs. 1–3) and remained an independent prognostic marker upon multivariate analysis before TACE-3 (Table 4). Therefore, the ART score and its prognostic significance were validated prior to the third and fourth TACE, and identified patients with poor survival despite, or maybe even due to, further TACE sessions. Of note, an ART score of 0–1.5 points prior to TACE-2 identified patients with good prognosis even in presence of more advanced liver dysfunction (Child-Pugh score B >8) [9]. This no longer holds true for further TACE-cylces: Child-Pugh stage remained an independent predictor of overall survival in addition 124
to the ART score (Table 4). Further subgroup analyses showed that patients with a Child-Pugh score B >8 who received a third TACE had a dismal prognosis, irrespective of their ART score before TACE-3 (Supplementary Fig. 2C). In line with this finding, the ART score lost its predictive capability in Child-Pugh B patients before TACE-4. These data suggest that there might exist a subgroup of Child B >8 patients defined by an ART score of 0–1.5 before TACE-2 who may profit from a second TACE session, but none of these patients profit from a third or fourth TACE session. On the other hand, these data underline the value of the ART score to differentiate prognosis in patients (Child-Pugh A5-B7) who would be considered suitable for a third TACE session based on their Child-Pugh score. Interestingly, there was no obvious difference in the proportion of patients at Child-Pugh stage A or B prior to TACE-2, 3, and 4 (Table 2). However, if analysed at an individual basis, 33–40% of patients suffered from a Child-Pugh score increase, which did not necessarily translate into a ChildPugh stage migration. It is one of the key advantages of the ART score that these prognostically relevant subclinical changes are noticed and considered for re-treatment decisions. Given the successful validation of the ART score prior to TACE-3 and 4, we next investigated the concept of an ART score guided re-treatment strategy (Fig. 4). For this purpose, we stratified patients according to their ART score prior to their respective last TACE session in a sequential manner (Fig. 1C). Patients with a beneficial ART score (0–1.5 points) prior to TACE-2 who had a promising OS of 28 months (Fig. 3), could be further divided into two prognostic subgroups by the ART score before TACE-3 (Fig. 3). One-fifth of those patients had an ART score of P2.5 prior to TACE-3 and showed a median OS of only 11 months. In patients with a dismal ART score (P2.5) prior to TACE-2, no subgroup that would benefit from further TACE cycles could be identified by the ART score before TACE-3 (Fig. 3). Crucially, similar results were observed with the sequential assessment of the ART score before TACE-4. This suggests that the sequential assessment of the ART score is reliable and that the first detection of a dismal ART score (P0.5 points) should prompt research into the optimal treatment options for these patients. Whether continuation of TACE, experimental treatments like systemic internal radiotherapy (SIRT), or treatment stage migration [5] to other evidence based treatments like, for example, systemic therapy with sorafenib, would be most appropriate in these ART score high patients remains to be elucidated in prospective comparative trials. In summary, we validated the use of the ART score prior to TACE-3 and 4 and demonstrated that the sequential assessment using the ART score each time identifies patients with dismal prognosis if re-treated with TACE. We therefore propose an ART score guided re-treatment strategy as basis for prospective clinical trials into the best treatment options for these patients.
Conflict of interest W. Sieghart received speaker and consulting fees and research grants from Bayer Schering Pharma; H. Heinzl received a research grant from Roche; M. Trauner received speaker fees and travel grants from Roche; M. Peck-Radosavljevic received speakers and consulting fees and research grants from Bayer, Schering Pharma, Lilly Pharma and Boehringer Ingelheim.
Journal of Hepatology 2014 vol. 60 j 118–126
JOURNAL OF HEPATOLOGY ART-score before TACE 2 0-1.5 points (n = 77): 28.3 [95% CI: 24.0-32.0] ≥2.5 points (n = 32): 9.1 [95% CI: 6.6-11.6] p <0.001
ART-score before TACE 3
ART-score before TACE 3
0-1.5 points (n = 62): 28.7 [95% CI: 22.3-35.2] ≥2.5 points (n = 15): 10.8 [95% CI: 0-21.8]
0-1.5 points (n = 5): 7.9 [95% CI: 5.0-10.9] ≥2.5 points (n = 27): 7.9 [95% CI: 4.4-11.4]
p = 0.001
p = 0.159
ART-score before TACE 4 0-1.5 points (n = 21): 25.1 [95% CI: 18.9-31.4] ≥2.5 points (n = 8): 8.6 [95% CI: 5.2-12.0] p = 0.033
Fig. 3. Sequential assessment of the ART score in patients treated with multiple TACE cycles. Right (vertical): sequential assessment of the ART score before TACE-3 and before TACE-4 in patients who were in the beneficial ART score group (0–1.5 points) prior to the respective last TACE. Left (vertical): sequential assessment of the ART score prior TACE-3 in patients who had an ART score of P2.5 points prior TACE-2. All analysis were performed using the Kaplan Meier method (OS in months according to Log-rank test).
Patient/disease characteristics BCLC-A*/B
Child-Pugh A/B
TACE 1
0-1.5 points
ART-score
≥2.5 points
TACE 2
0-1.5 points
ART-score
≥2.5 points1
Consider alternative therapies
TACE 3
0-1.5 points
ART-score
≥2.5 points1
TACE 4 Fig. 4. Proposal for an ART score guided re-treatment strategy with TACE. ⁄If not suitable for OLT or local ablative therapy. 1And/or Child-Pugh stage B >7.
Journal of Hepatology 2014 vol. 60 j 118–126
125
Research Article Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.jhep.2013.08. 022.
[9]
[10]
References [11] [1] Parkin DM, Bray F, Ferlay J, Pisani P. Global cancer statistics, 2002. CA Cancer J Clin 2005;55:74–108. [2] Llovet JM, Bru C, Bruix J. Prognosis of hepatocellular carcinoma: the BCLC staging classification. Semin Liver Dis 1999;19:329–338. [3] Forner A, Reig ME, de Lope CR, Bruix J. Current strategy for staging and treatment: the BCLC update and future prospects. Semin Liver Dis 2010;30:61–74. [4] Bruix J, Sherman M, Llovet JM, Beaugrand M, Lencioni R, Burroughs AK, et al. Clinical management of hepatocellular carcinoma. Conclusions of the Barcelona-2000 EASL conference. European Association for the Study of the Liver. J Hepatol 2001;35:421–430. [5] EASL-EORTC clinical practice guidelines. Management of hepatocellular carcinoma. J Hepatol 2012;56:908–943. [6] Hucke F, Sieghart W, Schoniger-Hekele M, Peck-Radosavljevic M, Muller C. Clinical characteristics of patients with hepatocellular carcinoma in Austria – is there a need for a structured screening program? Wien Klin Wochenschr 2011;123:542–551. [7] Llovet JM, Bruix J. Systematic review of randomized trials for unresectable hepatocellular carcinoma: chemoembolization improves survival. Hepatology 2003;37:429–442. [8] Raoul JL, Sangro B, Forner A, Mazzaferro V, Piscaglia F, Bolondi L, et al. Evolving strategies for the management of intermediate-stage hepatocellu-
126
[12]
[13] [14]
[15] [16]
[17]
lar carcinoma: available evidence and expert opinion on the use of transarterial chemoembolization. Cancer Treat Rev 2011;37:212–220. Bolondi L, Burroughs A, Dufour JF, Galle PR, Mazzaferro V, Piscaglia F, et al. Heterogeneity of patients with intermediate (BCLC B) hepatocellular carcinoma: proposal for a subclassification to facilitate treatment decisions. Semin Liver Dis 2012;32:348–359. Sieghart W, Hucke F, Pinter M, Graziadei I, Vogel W, Mueller C, et al. The ART of decision making: re-treatment with TACE in patients with hepatocellular carcinoma. Hepatology 2013. http://dx.doi.org/10.1002/hep.26256. Terzi E, Golfieri R, Piscaglia F, Galassi M, Dazzi A, Leoni S, et al. Response rate and clinical outcome of HCC after first and repeated cTACE performed ‘‘on demand’’. J Hepatol 2012;57:1258–1267. Peck-Radosavljevic M, Sieghart W, Kolblinger C, Reiter M, Schindl M, Ulbrich G, et al. Austrian Joint OGGH-OGIR-OGHO-ASSO position statement on the use of transarterial chemoembolization (TACE) in hepatocellular carcinoma. Wien Klin Wochenschr 2012;124:104–110. Sobin L, Wittekind C, editors. TNM classification of malignant tumours. New York: UICC; 2002. Sieghart W, Pinter M, Hucke F, Graziadei I, Schoniger-Hekele M, Muller C, et al. A single determination of C-reactive protein at the time of diagnosis predicts long term outcome of patients with hepatocellular carcinoma. Hepatology 2012. http://dx.doi.org/10.1002/hep.26057. Child CG, Turcotte JG. Surgery and portal hypertension. Major Probl Clin Surg 1964;1:1–85. Memon K, Kulik L, Lewandowski RJ, Wang E, Ryu RK, Riaz A, et al. Alphafetoprotein response correlates with EASL response and survival in solitary hepatocellular carcinoma treated with transarterial therapies: a subgroup analysis. J Hepatol 2012;56:1112–1120. Trotti A, Colevas AD, Setser A, Rusch V, Jaques D, Budach V, et al. CTCAE v3.0: development of a comprehensive grading system for the adverse effects of cancer treatment. Semin Radiat Oncol 2003;13:176–181.
Journal of Hepatology 2014 vol. 60 j 118–126
The ART-strategy: Sequential assessment of the ART score predicts outcome of patients with hepatocellular carcinoma re-treated with TACE Florian Hucke1, Wolfgang Sieghart1,⇑, Matthias Pinter1, Ivo Graziadei2, Wolfgang Vogel2, Christian Müller1, Harald Heinzl3, Fredrik Waneck4, Michael Trauner1, Markus Peck-Radosavljevic1,⇑ 1
Division of Gastroenterology and Hepatology, Department of Internal Medicine III, AKH and Medical University of Vienna, Austria; Department of Gastroenterology and Hepatology, LKH and Medical University of Innsbruck, Austria; 3Section for Clinical Biometrics, Medical University of Vienna, Austria; 4Department of Interventional Radiology, AKH and Medical University of Vienna, Austria
2
Background & Aims: Recently, we developed the ART score (assessment for re-treatment with TACE) to guide the decision for a second transarterial chemoembolization (TACE-2) in patients with hepatocellular carcinoma (HCC). Patients with an ART score of 0–1.5 points gained benefit from a second TACE session, while patients with an ART score P2.5 points did not. Here, we investigated (1) the prognostic significance of the ART score prior to the third (TACE-3) and fourth TACE (TACE-4), and (2) the feasibility of an ART score guided re-treatment strategy by sequential assessment of the ART score in HCC patients treated with multiple TACE sessions. Methods: 109 patients, diagnosed with intermediate stage HCC and treated with P3 TACE sessions between January 1999 and December 2009 at the Medical Universities of Vienna and Innsbruck, were included. The ART score prior to each TACE session was assessed in comparison to the TACE naïve liver. The prognostic performance of the ART score before TACE-3 and 4 was evaluated with and without stratification based on the ART score prior to the respective last intervention. Results: The pre-TACE-3 ART score discriminated two groups with different prognosis and remained a valid predictor of OS independent of Child-Pugh score (5–7 points), CRP-levels and tumor characteristics. Even in patients with an initially beneficial ART score (0–1.5 points) before TACE-2, repeated ART score assessment before TACE-3 identified a subgroup of patients with dismal prognosis (median OS: 27.8 vs. 10.8 months, p <0.001). Similar results were observed when the ART score was applied before TACE-4. Conclusions: The sequential assessment of the ART score identifies patients with dismal prognosis prior to each TACE session.
Keywords: Hepatocellular carcinoma; TACE; Prognosis; HCC; ART score. Received 2 April 2013; received in revised form 11 July 2013; accepted 20 August 2013; available online 3 September 2013 ⇑ Corresponding authors. Address: AKH & Medizinische Universität Wien, Klinik Innere Medizin III, Abteilung Gastroenterologie & Hepatologie, Währinger Gürtel 18-20, A-1090 Wien, Austria. Tel.: +43 1 40400 6589; fax: +43 1 40400 4735. E-mail addresses: [email protected] (W. Sieghart), markus.peck@ meduniwien.ac.at (M. Peck-Radosavljevic).
Ó 2013 European Association for the Study of the Liver. Published by Elsevier B.V. All rights reserved.
Introduction Hepatocellular carcinoma (HCC), is the 5th most common cancer worldwide, and develops predominately in patients with liver cirrhosis [1]. The most widely adopted staging system for HCC is the Barcelona Clinic Liver Cancer (BCLC) staging system [2,3], which was endorsed by the European Association for the Study of the Liver (EASL) [4,5]. It distinguishes five tumor stages and links them to different treatment options. Curative therapies, like liver transplantation (OLT), liver resection, percutaneous ethanol injection (PEI) and radiofrequency ablation (RFA) are recommended only for early stages. However, most patients are diagnosed at advanced tumor stages [6]. Based on a meta-analysis of randomized controlled trials [7], transarterial chemoembolization (TACE) is the recommended treatment modality for asymptomatic, large or multifocal HCC without macrovascular invasion or extrahepatic metastasis (intermediate HCC, BCLC stage B). The outcome results of several TACE studies vary significantly [7,8]. Since the group of patients at BCLC stage B is very heterogeneous [9] in terms of liver function and tumor load (Child-Pugh stage A/B, tumor burden), there are patients who may not profit from repeated TACE sessions. In this context, we recently developed the ART score (assessment for re-treatment with TACE) (Table 1) [10]. This new and simple point score integrates radiologic tumor response (present vs. absent) and impairment of liver function (presence vs. absence of Child-Pugh score increase by 1 or P2 points and AST increase by 25% from pre TACE-1, respectively) after the first TACE to predict patient survival if re-treated with a second TACE. The ART score selects two distinct patient groups with significantly different prognosis and is able to identify patients who may probably not benefit from a second TACE sessions. These results were confirmed in an independent external validation cohort.
Journal of Hepatology 2014 vol. 60 j 118–126
JOURNAL OF HEPATOLOGY Table 1. ART-score.
Points Radiologic tumor response Absent Present AST increase >25% Present Absent Child-Pugh score increase 1 point ≥2 points Absent
1 0 4 0 1.5 3 0
AST, aspartate transaminase.
In our initial analysis [10], the prognostic performance of the ART score was only validated at one time point: prior to the second TACE session. However, most intermediate stage HCC patients are treated with more than two TACE sessions in a sequential manner [7]. Therefore, confirmation of the prognostic performance of repeated, sequential assessment of the ART score, would provide a strong rationale for a prospective study evaluating an ART score guided TACE strategy. The aim of this study was to investigate (1) the prognostic significance of the ART score prior the third (TACE-3) and fourth TACE (TACE-4) and (2) the feasibility of an ART score guided retreatment strategy by sequential assessment of the ART score in HCC patients treated with multiple TACE sessions.
Baseline (before TACE-1) imaging (triphasic CT/MRI scan) was performed 5–7 d before the first TACE session. HCC was staged according to the BCLCclassification [2,3] and by the International Union Against Cancer (UICC) tumor node metastasis (TNM) classification 6th edition [13]. In both institutions, radiologic tumor response was assessed either by CT (96%) or MRI (4%) scan prior to each TACE session (maximal 90 d after the first or second TACE, respectively) according to EASL criteria [4]. The same modality was used in single patients before and after each TACE session. Overall 7 patients (6%) received their 1st TACE session between 1999 and August 2001 (time before the implementation of the EASL criteria) and were originally evaluated according to WHO criteria [4]. These patients were re-evaluated according to EASL criteria for this study. Objective tumor response following each TACE was defined as complete response (CR) and partial response (PR), while stable disease (SD) and progressive disease (PD) were judged as absence of objective tumor response. All laboratory values including alpha-fetoprotein (AFP) and CRP levels [14] as well as liver function parameters including the Child-Pugh score [15] were determined one day before each TACE session. AFP response prior to the third TACE was defined as an AFP decrease by 50% from pre-TACE-1 values of P200 kU/L [16]. Three AFP groups were formed for univariate analysis: pre-TACE-1 AFP P200 kU/L with response vs. pre-TACE-1 AFP P200 kU/L and no response vs. pre-TACE-1 AFP levels <200 kU/L. Adverse events, which occurred within four weeks after any TACE or were unequivocally TACE related, were documented according to the Common Terminology Criteria for Adverse Events version 3.0 (CTCAE) [17]). ART score determination Details about the ART score development have recently been reported in [9]. In short, the ART score consists of three variables (Table 1): AST increase by 25% (ART score: 4 points), Child-Pugh increase by 1 or P2 points (ART score: 1.5 and 3 points, respectively) as well as absence of radiologic tumor response (ART score: 1 point). Based on the sum of all three variables, the ART score is calculated. A cut-off of 0–1.5 points and P2.5 points [10] was used to stratify patients for all survival analysis. In this study, the ART score was determined 1 day prior to TACE-2, TACE-3, and TACE-4, respectively. For this purpose, all ART score forming variables were determined with reference to the TACE naïve liver (before TACE-1).
Materials and methods Study design Eligibility Similar to our recent study [10], we used the following inclusion criteria (Fig. 1A): all patients (>18 yr), who were diagnosed with HCC by histology or dynamic imaging (CT/MRI scans) according to EASL diagnostic criteria [4] and treated with cTACE, TAE or DEB-TACE (thereafter summarized and referred to as TACE), at the Department of Gastroenterology and Hepatology of the Medical University of Vienna and Innsbruck between January 1999 and December, 2009 were included in this study. Patients were eligible if they had HCC at BCLC-stage A (not suitable for OLT or local ablative therapy) or B, preserved liver function (Child-Pugh stage A or B) and received at least three TACE sessions (time between the TACE cycles each 690 days). Patients were excluded, if they received TACE before OLT or resection, or if patients received TACE for HCC recurrence after OLT. Additionally, patients who received TACE despite poor liver function (Child-Pugh C) and patients at BCLC-stage C were excluded. In order to examine the effect of sequential assessment of the ART score in patients prior to the third and fourth TACE, we furthermore excluded patients in which the ART score prior to the respective last TACE was not determinable due to missing variables. The selection of the final study population is given in Fig. 1A. TACE procedure The procedure of transarterial embolization (TAE), conventional lipiodol-doxorubicin TACE (cTACE) and TACE with drug-eluting beads (DEB-TACE) in both centers was described in the original ART score publication [10]. In short, both institutions used a ‘‘treatment on demand’’ [11] TACE-schedule and no further TACE session was performed in presence of complete radiologic response. Additionally, presence of Child-Pugh C cirrhosis, portal vein thrombosis or ECOG >1 was considered as contraindication for re-treatment with TACE [12]. Collection of data
Based on the inclusion criteria of this study, we identified all patients at the Medical University of Vienna and Innsbruck, who underwent multiple TACE sessions and for which the ART score was determinable prior to TACE-2, TACE-3, and TACE-4. We pooled the two patient cohorts in order to achieve higher statistical power for further analysis. Given that the ART score was developed and validated prior to TACE-2, we first aimed to validate the prognostic value of the ART score if assessed prior to TACE-3 or TACE-4 (Fig. 1B). In a second step, we analyzed the prognostic significance of the sequential assessment of the ART score. For this purpose we focused on patients who were supposed to be in a beneficial prognostic group based on an ART score of 0–1.5 points prior to the respective last TACE session and investigated their outcome based on their ART score prior to the following TACE session (Fig. 1C). Statistical analyses Patient characteristics are presented with descriptive statistics. Overall survival was defined as the time from the day of the respective ART score assessment until death or last follow-up. Survival curves were calculated using the Kaplan–Meier method. Median survival times (OS) and their 95% confidence intervals (CIs) were reported. The log-rank test and a multivariate Cox regression model were used to assess the effects of patient variables prior to each TACE on OS. All reported p-values are results of two-sided tests. A significance level of 0.05 was applied throughout. Statistical analyses were performed using IBM SPSS version 20.0 (SPSS Inc., Armonk, NY) and SAS 9.3 (SAS Institute Inc., Cary, NC).
Results Patient characteristics
The retrospective analysis of data was approved by the Ethics Committees of the Medical Universities of Vienna and Innsbruck.
Patient characteristics are shown in Table 2. Prior to the first TACE, the majority of patients was at BCLC-stage B (82%) and
Journal of Hepatology 2014 vol. 60 j 118–126
119
Research Article
Patients treated with TACE with determinable ART-score before TACE 3 n = 109
Patients treated with TACE with determinable ART-score before TACE 4 n = 43
Excluded, n = 279 • Liver transplantation prior or post-TACE/resection post first TACE, n = 95 • Child-Pugh C, n = 32 • BCLC-C, n = 50 • Patients with less than 2 TACE, n = 55 • Interval TACE 1-TACE 2 >90 days, n = 29 • ART-score before TACE 2 not determinable due to missing values, n = 18
Patients treated with ≥3 TACE cycles
TACE 1 ≤90 days
Patients treated with TACE with determinable ART-score before TACE 2 n = 204
B
TACE 2 Excluded, n = 95 • Patients with less than 3 TACE, n = 52 • Patients with ≥3 TACE but interval TACE 2-TACE 3 >90 days, n = 34 • ART-score before TACE 3 not determinable due to missing values, n = 9 Excluded, n = 67 • Patients with less than 4 TACE, n = 38 • Patients with ≥4 TACE but interval TACE 3-TACE 4 >90 days, n = 26 • ART-score before TACE 4 not determinable due to missing values, n = 3
≤90 days
Patients treated with TACE at the Medical University of Vienna and Innsbruck n = 483
ART-score before TACE 3
TACE 3 ≤90 days
A
ART-score before TACE 4
TACE 4
Overall survival
Death or censored
C
Patients treated with multiple TACE cycles
TACE 1 ART-score before TACE 2
≥2.5 points 0-1.5 points TACE 2 ART-score before TACE 3
≥2.5 points 0-1.5 points TACE 3 ART-score before TACE 4
≥2.5 points 0-1.5 points TACE 4
OUTCOME Fig. 1. Patient selection and study design. (A) Patient selection and final study population. (B) Study design for ART score validation pre-TACE-3 and pre-TACE-4. (C) Study design for sequential assessment of the ART score.
had a Child-Pugh A cirrhosis (61%). TACE was the first-line therapy in 88 (81%) of patients, while 21 (19%) received a prior antitumor therapy (liver resection (n = 9), PEI (n = 7), RFA (n = 5)).
120
Chemoembolization was the most common treatment technique (n = 95, 87%) (cTACE vs. DEB-TACE n = 85 vs. n = 10), while in 13% of cases (n = 14) TAE only was performed.
Journal of Hepatology 2014 vol. 60 j 118–126
JOURNAL OF HEPATOLOGY Table 2. Patient characteristics.
N Before TACE-1 Age (yr) Mean ± SD Range Sex Male Female Etiology Alcohol Viral (HBV, HCV) Others Unknown Child-Pugh stage A B 7 points ≥B 8 points Tumor size Mean ± SD Range Tumor extent Unilobar Bilobar Tumor number Unifocal Multifocal BCLC-stage A B TNM stage I II III Prior anti-tumor therapy Absent Present α-Fetoprotein (kU/L) <200 ≥200 Before TACE-2 Child-Pugh stage A B 7 points ≥B 8 points Before TACE-3 Child-Pugh stage A B 7 points ≥B 8 points Before TACE-4 Child-Pugh stage A B 7 points ≥B 8 points
% 109
63.8 ± 9.3 42-82 99 10
91 9
52 37 11 9
48 34 10 8
77 14 18
61 13 17
5.1 ± 2.6 1.3-12.4 51 58
47 53
33 76
30 70
20 89
18 82
19 44 46
17 40 42
88 21
81 19
84 25
77 23
75 16 18
69 15 17
71 20 18
65 18 17
29 4 10
63 9 22
HBV, hepatitis B virus infection; HCV, hepatitis C virus infection; BCLC, Barcelona Clinic Liver Cancer; TNM, tumor nodes metastasis.
Between TACE-1 and TACE-3, 36 patients (33%) suffered from a Child-Pugh increase by at least 1 point, while 57 patients (52%) showed no change and 16 patients (15%) showed a decrease of the Child-Pugh score by at least 1 point. Between TACE-1 and TACE-4, 17 patients (40%) suffered from a Child-Pugh increase by at least 1 point, while 23 patients (53%) showed no change and 3 patients (7%) showed a decrease of the Child-Pugh score by 1 point. Overall, the median number of TACE interventions was 4 (range 3–20). The median time interval between the TACE sessions was 41 d (range 13–85) between the first and second TACE, 42 d (range 19–89) between the second and third TACE, and 46 d (range 21–85) between the third and fourth TACE. Radiologic tumor response after TACE-2, 3, and 4 is given in Supplementary Table 1. Univariate analysis of prognostic factors prior to the third and fourth TACE During follow-up, 86% of patients (n = 94) died, while 14% were still alive (n = 10) or lost to follow-up (n = 5). The median OS of the whole population was 17.1 months (95% CI, 11.4–22.7) (as calculated from the date of TACE-1 until death or date last seen). The median time of follow-up was 62.7 months. In the univariate analysis of prognostic factors prior to the third TACE, Child-Pugh stage (OS: CP A vs. CP B7 vs. CP PB8: 23.0 vs. 12.6 vs. 7.2 months, p <0.001), tumor extent (OS: unilobar vs. bilobar: 24.1 vs. 12.6 months, p = 0.040) and tumor number (OS: unifocal vs. multifocal: 31.7 vs. 12.6 months, p = 0.040), CRP levels (OS: normal vs. elevated: 24.1 vs. 10.0 months, p = 0.003) and AFP response (OS: AFP response vs. no AFP response vs. baseline AFP <200 kU/L: 15.1 vs. 8.2 vs. 19.0 months, p = 0.008) had a significant impact on overall survival (Table 3). Of note, the single ART score variables AST increase >25% (OS: no AST increase vs. AST increase >25%: 20.3 vs. 8.9 months, p = 0.001), a Child-Pugh increase by 1 or >2 points (OS: no CP increase vs. CP increase 1 Point vs. CP increase P2 Points: 26.5 vs. 10.2 vs. 5.2 months, p <0.001), and absence of radiologic tumor response (OS: response vs. no response: 21.8 vs. 11.3 months, p = 0.048) had a significant impact on overall survival upon univariate analysis. Prior to the fourth TACE, only higher Child-Pugh stage and elevated CRP levels were associated with worse overall survival (data not shown), likely due to the small sample size.
The ART score prior to the third and fourth TACE is an independent predictor of overall survival Supplementary Tables 5 and 6 outline the different constellations of how patients gained their ART score points prior to TACE-3 and 4. Overall, patients with an ART score before TACE-3 of 0–1.5 points had a median OS of 28.1 months [95% CI: 21.7–24.4] while those with P2.5 points had a median OS of 8.5 [95% CI: 7.4–9.6] (p <0.001) (Fig. 2A). Similar results were observed if the ART score was assessed before TACE-4 (Fig. 2C). Additionally, these results could be confirmed upon separate analysis of patients treated at the Medical University of Vienna (Fig. 2B left and 2D left) and patients who were treated at the Medical University of Innsbruck (Fig. 2B right, Fig. 2D right). The ART score remained a significant prognostic factor independent of the TACE technique (Supplementary Fig. 1A–C,
Journal of Hepatology 2014 vol. 60 j 118–126
121
Research Article Table 3. Univariate analysis of prognostic factors before TACE-3 in HCC patients treated with TACE.
Variable Age Child-Pugh stage
Etiology Tumor extent Tumor number Tumor size (cm) BCLC-stage TNM stage
Prior therapy CRP values α-Fetoprotein (kU/L) AFP response¹
TACE type
AST increase >25% Child-Pugh increase
Radiologic tumor response
<65 ≥65 A B 7 points ≥B 8 points Viral Other Unilobar Bilobar Unifocal Multifocal <7.5 ≥7.5 A B I II III Absent Present <1 mg/dl ≥1 mg/dl <200 ≥200 AFP >200: no response AFP >200: response AFP <200 TAE cTACE DEB-TACE Absent Present 0 points 1 point ≥2 points Present Absent
N = 109 59 50 71 20 18 72 37 51 58 33 76 85 24 20 89 19 44 46 88 21 62 41 84 25 11 14 84 14 85 10 85 24 73 23 13 56 53
Overall survival (mo) Median 95% CI 18.6 7.0-30.2 17.1 13.3-20.8 23.0 16.7-29.3 12.6 6.1-19.1 7.2 3.3-11.1 17.1 9.5-24.6 18.0 7.7-28.4 24.1 13.8-34.4 12.6 7.8-17.3 31.7 16.3-47.0 12.6 6.0-19.2 18.2 12.4-24.0 12.6 4.1-21.1 24.9 2.7-47.0 15.1 9.1-21.0 24.9 2.0-47.7 18.2 7.6-28.8 14.9 9.7-20.1 17.1 11.2-22.9 18.0 10.9-25.2 24.1 17.4-30.8 10.0 7.2-12.8 19.0 11.8-26.2 11.3 9.7-12.9 8.2 3.8-22.7 4.5-25.6 15.1 19.0 11.8-26.2 14.9 2.0-27.8 15.6 9.9-21.4 19.0 8.2-29.8 20.3 13.4-27.2 8.9 7.7-10.1 26.5 22.1-31.0 10.2 7.2-13.3 5.2 2.5-7.9 21.8 12.0-31.6 11.3 6.1-16.0
p value (Log-rank) 0.921
<0.001 0.954 0.040 0.040 0.758 0.073
0.162 0.980 0.003 0.330
0.008
0.700 0.001
<0.001 0.048
BCLC, Barcelona Clinic Liver Cancer; TNM, tumor nodes metastasis; TAE, transarterial embolization; cTACE, conventional transarterial chemoembolization; DEB-TACE, drugeluting beads transarterial chemoembolization. 1 AFP response was defined by a decrease of >50% from baseline levels.
Supplementary Table 2), Child-Pugh stage (A-B7) prior to the respective TACE session, elevated CRP levels and presence of ascites. Additionally, an ART score of P2.5 points was associated with a higher frequency of serious adverse events after TACE-3 (Supplementary Table 4). Of note, patients at Child-Pugh stage PB8 prior to the third TACE had a similar dismal prognosis in both ART score groups (median OS: 0–1.5 points vs. P2.5 points: 7.9 vs. 6.0 [95% CI: 1.6–14.2 vs. 2.4–9.5], p = 0.404) (Supplementary Fig. 2C). Similar results were observed before TACE-4: while the ART score further differentiated survival in patients with Child-Pugh A cirrhosis (OS 122
[95% CI] for ART score 0–1.5 (n = 19) vs. P2.5 (n = 10): 25.1 vs. 4.4 months [19.2–31.1 vs. 0–11.1]), no discriminative capability was observed in patients with Child-Pugh B cirrhosis. Sample size before TACE-3 allowed inclusion of significant variables of the univariate analysis into multivariate analysis (Table 4). Only Child-Pugh stage prior to the third TACE (hazard ratio [95% CI]: CP B7 vs. CP PB8: 1.4 vs. 2.5 [0.8–2.5 vs. 1.2–5.2], p = 0.012) and the ART score P2.5 before TACE-3 (hazard ratio [95% CI], 3.1 [1.7–5.5], p <0.001) remained independent significant predictors of overall survival.
Journal of Hepatology 2014 vol. 60 j 118–126
JOURNAL OF HEPATOLOGY Medical University of Vienna
0-1.5 points (n = 67): 28.1 [95% CI: 21.7-24.4] ≥2.5 points (n = 42): 8.5 [95% CI: 7.4-9.6]
B
1.0
0-1.5 points (n = 35): 24.1 [95% CI: 16.1-32.1] ≥2.5 points (n = 23): 8.5 [95% CI: 7.5-9.5]
1.0
1.0
0.8 0.6 0.4
ART score 0-1.5 points ≥2.5 points
0.2 0.0
0.6 0.4
ART score 0-1.5 points ≥2.5 points
0.2 0.0
0
20
40
60 80 Months
100 120
0-1.5 points (n = 23): 23.9 [95% CI: 14.7-33.1] ≥2.5 points (n = 20): 8.6 [95% CI: 7.5-9.8]
D
20
40
60 80 Months
ART score 0-1.5 points ≥2.5 points
0.0
Cumulative survival
0.6
40
60 80 Months
0.2
0
20
40
100 120
60 80 Months
100 120
0-1.5 points (n = 13): 35.5 [95% CI: 22.8-48.3] ≥2.5 points (n = 10): 6.4 [95% CI: 0-16.8]
1.0 p = 0.033
0.8 0.6 0.4
ART score 0-1.5 points ≥2.5 points
0.2 0.0
20
ART score 0-1.5 points ≥2.5 points
p = 0.024
0.8
0
0.4
100 120
1.0
0.2
0.6
0-1.5 points (n = 10): 17.0 [95% CI: 13.8-20.2] ≥2.5 points (n = 10): 8.6 [95% CI: 7.8-9.4]
p = 0.003
0.4
0.8
0.0 0
1.0 Cumulative survival
p <0.001
0.8
Cumulative survival
p <0.001
Cumulative survival
Cumulative survival
p <0.001
C
0-1.5 points (n = 32): 39.1 [95% CI: 28.5-49.7] ≥2.5 points (n = 19): 8.0 [95% CI: 3.6-12.5]
Cumulative survival
A
Medical University of Innsbruck
0.8 0.6 0.4
ART score 0-1.5 points ≥2.5 points
0.2 0.0
0
20
40
60 80 Months
100 120
0
20
40
60 80 Months
100 120
Fig. 2. Prognostic significance of the ART score before TACE-3 and before TACE-4. (A) Prognostic significance of the ART score before TACE-3 in all HCC patients and (B) patients treated at the Medical University of Vienna (left) and Innsbruck (right). (C) Prognostic significance of the ART score before TACE-4 in all HCC patients and (D) patients treated at the Medical University of Vienna (left) and Innsbruck (right). Analysis was performed using the Kaplan Meier method (OS in months according to Logrank test).
Finally, to reassure the significance of all single ART score variables prior to TACE-3, we performed a sensitivity-multivariate analysis including the three ART score variables and all other significant variables of the univariate analysis (Table 3). Crucially, only AST increase >25%, absence of radiologic response and Child-Pugh increase of 1 or >2 points remained independent significant predictors of overall survival (Supplementary Table 3). Due to low sample size, no multivariate analysis could be performed before TACE-4. Sequential assessment of the ART score prior to each TACE session identifies patient subgroups with different prognosis Finally we analyzed the prognostic significance of the sequential application of the ART score (Fig. 1C). The ART score prior to TACE-2 identified 77 patients with good prognosis (median OS 28.0 months [95% CI: 24.0–31.9], Fig. 3). Crucially, the sequential assessment of the ART score prior to TACE-3 in this population (n = 77) identified two further subgroups with different outcome: patients with an ART score of 0–1.5 points (n = 62) prior to the third TACE had a median OS of 28.7 months [95%CI: 22.3–35.2], while one-fifth of patients (ART score P2.5 points, n = 15) showed a worse prognosis of 10.8 months [95% CI: 0–21.8]
(p = 0.001) (Fig. 3A). Similar results were observed if the ART score was sequentially assessed before TACE-4 (Fig. 3). Conversely, in patients who were in the dismal prognosis group prior to the second TACE (P2.5 points, n = 32, median OS [95% CI]: 7.9 [5.9–10.0]) and received a further TACE cycle, the sequential assessment of the ART score could not discriminate prognosis (Fig. 3): (median OS [95% CI]: 0–1.5 points vs. P2.5 points: 7.9 vs. 7.9 months [5.0–10.9 vs. 4.4–11.4], p = 0.159). The significance of the ART score before TACE-4 in patients who were in the dismal prognosis group prior to the third TACE (P2.5 points, n = 15, median OS [95% CI]: 10.8 [0–21.8]) could not be assessed due to low sample size (n = 5).
Discussion Most HCC patients allocated to TACE suffer from liver cirrhosis, which may limit the potential survival benefit of this procedure. This is further complicated by the fact that typically HCC patients suitable for TACE mostly need P3 TACE sessions for optimal radiologic tumor response [7,11]. Therefore repeated assessment of liver function and TACE derived liver damage is of paramount importance for re-treatment decisions. The optimal strategy for
Journal of Hepatology 2014 vol. 60 j 118–126
123
Research Article Table 4. Multivariable Cox regression analysis before TACE-3 of prognostic factors in patients with HCC treated with TACE.
Variable Child-Pugh stage A B 7 points ≥B 8 points Tumor number Unifocal Multifocal Tumor extent Unilobar Bilobar CRP-values <1 mg/dl ≥1 mg/dl AFP response AFP <200 kU/L AFP >200: response AFP >200: no response ART-score 0-1.5 points ≥2.5 points
Overall survival p value HR 95% CI (Cox regression) 1 1.4 2.5
0.8-2.5 1.2-5.2
0.012
1 1.2
0.6-2.2
0.604
1 1.2
0.7-2.2
0.518
1 1.5
0.9-2.4
0.216
1 0.9 2.0
0.5-1.9 0.9-4.5
0.196
1 3.1
1.7-5.5
<0.001
HCC, hepatocellular carcinoma; AFP, a-fetoprotein.
re-treatment with TACE would weigh tumor response and liver deterioration after TACE to optimize the overall clinical outcome of these patients. In this context, we recently developed a score for the assessment of re-treatment with TACE (ART score) [10]. The ART score is a simple point score, which integrates radiologic tumor response, liver damage and impairment of liver function after the first TACE to predict patient’s prognosis after re-treatment with TACE (Table 1). However, the ART score was only developed and validated prior to the second TACE session. Yet, the prognostic significance of the ART score prior to the third and fourth TACE session and, in particular, the value of a repeated sequential assessment of the ART score within a defined patient population undergoing multiple TACE sessions remains unclear. In the current study, sensitivity analysis prior to TACE-3 again identified an AST increase by >25%, Child-Pugh increase by 1 or >2 points and absence of radiologic response as independent predictors of overall survival (Supplementary Table 3). Additionally, the ART score prior to TACE-3 and -4 discriminated two different groups of patients with distinct prognosis (Fig. 2A–D). The ART score retained significance in several clinically important subgroups (Supplementary Figs. 1–3) and remained an independent prognostic marker upon multivariate analysis before TACE-3 (Table 4). Therefore, the ART score and its prognostic significance were validated prior to the third and fourth TACE, and identified patients with poor survival despite, or maybe even due to, further TACE sessions. Of note, an ART score of 0–1.5 points prior to TACE-2 identified patients with good prognosis even in presence of more advanced liver dysfunction (Child-Pugh score B >8) [9]. This no longer holds true for further TACE-cylces: Child-Pugh stage remained an independent predictor of overall survival in addition 124
to the ART score (Table 4). Further subgroup analyses showed that patients with a Child-Pugh score B >8 who received a third TACE had a dismal prognosis, irrespective of their ART score before TACE-3 (Supplementary Fig. 2C). In line with this finding, the ART score lost its predictive capability in Child-Pugh B patients before TACE-4. These data suggest that there might exist a subgroup of Child B >8 patients defined by an ART score of 0–1.5 before TACE-2 who may profit from a second TACE session, but none of these patients profit from a third or fourth TACE session. On the other hand, these data underline the value of the ART score to differentiate prognosis in patients (Child-Pugh A5-B7) who would be considered suitable for a third TACE session based on their Child-Pugh score. Interestingly, there was no obvious difference in the proportion of patients at Child-Pugh stage A or B prior to TACE-2, 3, and 4 (Table 2). However, if analysed at an individual basis, 33–40% of patients suffered from a Child-Pugh score increase, which did not necessarily translate into a ChildPugh stage migration. It is one of the key advantages of the ART score that these prognostically relevant subclinical changes are noticed and considered for re-treatment decisions. Given the successful validation of the ART score prior to TACE-3 and 4, we next investigated the concept of an ART score guided re-treatment strategy (Fig. 4). For this purpose, we stratified patients according to their ART score prior to their respective last TACE session in a sequential manner (Fig. 1C). Patients with a beneficial ART score (0–1.5 points) prior to TACE-2 who had a promising OS of 28 months (Fig. 3), could be further divided into two prognostic subgroups by the ART score before TACE-3 (Fig. 3). One-fifth of those patients had an ART score of P2.5 prior to TACE-3 and showed a median OS of only 11 months. In patients with a dismal ART score (P2.5) prior to TACE-2, no subgroup that would benefit from further TACE cycles could be identified by the ART score before TACE-3 (Fig. 3). Crucially, similar results were observed with the sequential assessment of the ART score before TACE-4. This suggests that the sequential assessment of the ART score is reliable and that the first detection of a dismal ART score (P0.5 points) should prompt research into the optimal treatment options for these patients. Whether continuation of TACE, experimental treatments like systemic internal radiotherapy (SIRT), or treatment stage migration [5] to other evidence based treatments like, for example, systemic therapy with sorafenib, would be most appropriate in these ART score high patients remains to be elucidated in prospective comparative trials. In summary, we validated the use of the ART score prior to TACE-3 and 4 and demonstrated that the sequential assessment using the ART score each time identifies patients with dismal prognosis if re-treated with TACE. We therefore propose an ART score guided re-treatment strategy as basis for prospective clinical trials into the best treatment options for these patients.
Conflict of interest W. Sieghart received speaker and consulting fees and research grants from Bayer Schering Pharma; H. Heinzl received a research grant from Roche; M. Trauner received speaker fees and travel grants from Roche; M. Peck-Radosavljevic received speakers and consulting fees and research grants from Bayer, Schering Pharma, Lilly Pharma and Boehringer Ingelheim.
Journal of Hepatology 2014 vol. 60 j 118–126
JOURNAL OF HEPATOLOGY ART-score before TACE 2 0-1.5 points (n = 77): 28.3 [95% CI: 24.0-32.0] ≥2.5 points (n = 32): 9.1 [95% CI: 6.6-11.6] p <0.001
ART-score before TACE 3
ART-score before TACE 3
0-1.5 points (n = 62): 28.7 [95% CI: 22.3-35.2] ≥2.5 points (n = 15): 10.8 [95% CI: 0-21.8]
0-1.5 points (n = 5): 7.9 [95% CI: 5.0-10.9] ≥2.5 points (n = 27): 7.9 [95% CI: 4.4-11.4]
p = 0.001
p = 0.159
ART-score before TACE 4 0-1.5 points (n = 21): 25.1 [95% CI: 18.9-31.4] ≥2.5 points (n = 8): 8.6 [95% CI: 5.2-12.0] p = 0.033
Fig. 3. Sequential assessment of the ART score in patients treated with multiple TACE cycles. Right (vertical): sequential assessment of the ART score before TACE-3 and before TACE-4 in patients who were in the beneficial ART score group (0–1.5 points) prior to the respective last TACE. Left (vertical): sequential assessment of the ART score prior TACE-3 in patients who had an ART score of P2.5 points prior TACE-2. All analysis were performed using the Kaplan Meier method (OS in months according to Log-rank test).
Patient/disease characteristics BCLC-A*/B
Child-Pugh A/B
TACE 1
0-1.5 points
ART-score
≥2.5 points
TACE 2
0-1.5 points
ART-score
≥2.5 points1
Consider alternative therapies
TACE 3
0-1.5 points
ART-score
≥2.5 points1
TACE 4 Fig. 4. Proposal for an ART score guided re-treatment strategy with TACE. ⁄If not suitable for OLT or local ablative therapy. 1And/or Child-Pugh stage B >7.
Journal of Hepatology 2014 vol. 60 j 118–126
125
Research Article Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.jhep.2013.08. 022.
[9]
[10]
References [11] [1] Parkin DM, Bray F, Ferlay J, Pisani P. Global cancer statistics, 2002. CA Cancer J Clin 2005;55:74–108. [2] Llovet JM, Bru C, Bruix J. Prognosis of hepatocellular carcinoma: the BCLC staging classification. Semin Liver Dis 1999;19:329–338. [3] Forner A, Reig ME, de Lope CR, Bruix J. Current strategy for staging and treatment: the BCLC update and future prospects. Semin Liver Dis 2010;30:61–74. [4] Bruix J, Sherman M, Llovet JM, Beaugrand M, Lencioni R, Burroughs AK, et al. Clinical management of hepatocellular carcinoma. Conclusions of the Barcelona-2000 EASL conference. European Association for the Study of the Liver. J Hepatol 2001;35:421–430. [5] EASL-EORTC clinical practice guidelines. Management of hepatocellular carcinoma. J Hepatol 2012;56:908–943. [6] Hucke F, Sieghart W, Schoniger-Hekele M, Peck-Radosavljevic M, Muller C. Clinical characteristics of patients with hepatocellular carcinoma in Austria – is there a need for a structured screening program? Wien Klin Wochenschr 2011;123:542–551. [7] Llovet JM, Bruix J. Systematic review of randomized trials for unresectable hepatocellular carcinoma: chemoembolization improves survival. Hepatology 2003;37:429–442. [8] Raoul JL, Sangro B, Forner A, Mazzaferro V, Piscaglia F, Bolondi L, et al. Evolving strategies for the management of intermediate-stage hepatocellu-
126
[12]
[13] [14]
[15] [16]
[17]
lar carcinoma: available evidence and expert opinion on the use of transarterial chemoembolization. Cancer Treat Rev 2011;37:212–220. Bolondi L, Burroughs A, Dufour JF, Galle PR, Mazzaferro V, Piscaglia F, et al. Heterogeneity of patients with intermediate (BCLC B) hepatocellular carcinoma: proposal for a subclassification to facilitate treatment decisions. Semin Liver Dis 2012;32:348–359. Sieghart W, Hucke F, Pinter M, Graziadei I, Vogel W, Mueller C, et al. The ART of decision making: re-treatment with TACE in patients with hepatocellular carcinoma. Hepatology 2013. http://dx.doi.org/10.1002/hep.26256. Terzi E, Golfieri R, Piscaglia F, Galassi M, Dazzi A, Leoni S, et al. Response rate and clinical outcome of HCC after first and repeated cTACE performed ‘‘on demand’’. J Hepatol 2012;57:1258–1267. Peck-Radosavljevic M, Sieghart W, Kolblinger C, Reiter M, Schindl M, Ulbrich G, et al. Austrian Joint OGGH-OGIR-OGHO-ASSO position statement on the use of transarterial chemoembolization (TACE) in hepatocellular carcinoma. Wien Klin Wochenschr 2012;124:104–110. Sobin L, Wittekind C, editors. TNM classification of malignant tumours. New York: UICC; 2002. Sieghart W, Pinter M, Hucke F, Graziadei I, Schoniger-Hekele M, Muller C, et al. A single determination of C-reactive protein at the time of diagnosis predicts long term outcome of patients with hepatocellular carcinoma. Hepatology 2012. http://dx.doi.org/10.1002/hep.26057. Child CG, Turcotte JG. Surgery and portal hypertension. Major Probl Clin Surg 1964;1:1–85. Memon K, Kulik L, Lewandowski RJ, Wang E, Ryu RK, Riaz A, et al. Alphafetoprotein response correlates with EASL response and survival in solitary hepatocellular carcinoma treated with transarterial therapies: a subgroup analysis. J Hepatol 2012;56:1112–1120. Trotti A, Colevas AD, Setser A, Rusch V, Jaques D, Budach V, et al. CTCAE v3.0: development of a comprehensive grading system for the adverse effects of cancer treatment. Semin Radiat Oncol 2003;13:176–181.
Journal of Hepatology 2014 vol. 60 j 118–126