Staging and prognostic systems: beyond BCLC? Alessandro Vitale, MD, PhD, FEBS U.O.C. di Chirurgia Epatobiliare e dei Trapianti Epatici, Department of Surgery, Oncology and Gastroenterology, University of Padua, Padua; Italy
ALESSANDRO VITALE, MD, PhD Azienda Ospedaliera e Università di Padova Il sottoscritto dichiara di non aver avuto negli ultimi 12 mesi conflitto d interesse in relazione a questa presentazione e che la presentazione non contiene discussione di farmaci in studio o ad uso off-label alessandro.vitale@unipd.it
Importance of HCC prognostic systems (1) Prognosis for individual patients. (2) Common scale for treatment selection for individual patients. Avoiding under and overtreatment. (3) Common scale for the comparison of outcomes among treatment methods and institutions and for RCT design. (4) A graph contrasting outcomes of transplantation to longterm outcomes of preexisting treatment methods for deciding indication of liver transplantation (Transplant benefit). Kudo M, et al. Dig Dis 2011;29:339 364
HCC Prognostic Systems 1. PROGNOSTIC SCORES are conventional prognostic scores that incorporate variables that were significant in multivariable (Cox, parametric models) survival analyses. The prognostic weights of the variables are used to construct the score (DATA BASED). 2. STAGING SYSTEMS typically based on systematic reviews of the literature and/or expert opinions (EVIDENCE BASED). These systems stratify the HCC population in evolutionary stages exclusively or mainly defined by tumor characteristics. Farinati F, et al. PLoS Med 2016; 13(4):e1002006
Staging and prognostic systems: beyond BCLC? Data based prognostic scores Evidence based staging systems Combined prognostic systems The issue of treatment allocation
Data based Prognostic Scores Liu PH, et al. J Hepatol 2016; 64: 601
Data based Prognostic Scores OKUDA FRENCH CLIP TOKYO Faria SC, et al. Abdominal Imaging 2014
Data based Prognostic Scores Model to Estimate Survival In Ambulatory HCC patients (MESIAH) Yang JD, et al. HEPATOLOGY 2012;56:614-621
Data based Prognostic Scores Johnson PJ, et al. JCO 2015; 33: 550 Chan WHA, et al. Liver Int 2016. In press
Data based Prognostic Scores PRO: objective and reproducible variables and rigorous statistical methodology. Accurate survival prediction CONS: often they are not suitably generalizable to populations different from the one that generated the score, and they don t define tumor stages for treatment selection Farinati F, et al. PLoS Med 2016; 13(4):e1002006
Data based Prognostic Scores Staging system (literature based) Prognostic score (multivariate analysis)
Staging and prognostic systems: beyond BCLC? Data based prognostic scores Evidence based staging systems Combined prognostic systems The issue of treatment allocation
Evidence based staging systems T definition originally refers only to HCC pathological characteristics of patients receiving liver resection There are 3 TNM surgical staging systems: 1) Liver Cancer Study Group of Japan (LCSGJ) 2) American Joint Committee on Cancer (AJCC) and International Union against Cancer(UICC) 3) United Network for Organ Sharing (UNOS) Minagawa M, et al. Ann Surg 2007; 245: 909
AJCC-UICC TNM 5th edition, 1997 Evidence based staging systems AJCC-UICC TNM 6th edition, 2002 UNOS TNM, 2002
Evidence based staging systems The Barcelona Clinic Liver Cancer (BCLC) Staging Classification for HCC BCLC stage 0 Very early Tumor volume, number and invasiveness Single < 2 cm Carcinoma in situ Performance status Child-Pugh 0 A A Early Single or 3 nodules < 3 cm 0 A B B Intermediate Multinodular 0 A B C Advanced Portal invasion N1M1 1 2 A B D Terminal Any of above > 2 C Cillo U, Vitale A, et al. J Hepatol 2004 Cillo U, Vitale A, et al. J Hepatol 2006
Evidence based staging systems PRO: They are useful to link stages to guidelines for the management of patients with HCC and the design of clinical trials. CONS: However, these systems are not based on a strong statistical methodology (variables are not weightened) and they often lack prognostic power Farinati F, et al. PLoS Med 2016; 13(4):e1002006
Evidence based staging systems Liu PH, et al. J Hepatol 2016; 64: 601
Prognostic pitfalls of BCLC classification ECOG PST 1 classified as BCLC B (in original BCLC stage C) Hsu CY, et al. Hepatology 2013
Prognostic pitfalls of BCLC classification Hsu CY, et al. Liver Int 2016. In press
Evidence based staging systems Prognostic score (multivariate analysis) Staging system (literature based)
Staging and prognostic systems: beyond BCLC? Data based prognostic scores Evidence based staging systems Combined prognostic systems The issue of treatment allocation
Combined prognostic systems Prognosis Treatment Prognostic score (multivariate analysis) Combined Prognostic System Staging system (literature based)
Combined prognostic systems From 2003 Japan TNM poor performance in western patients No PST No AFP Kudo M, et al. Dig Dis 2011;29:339 364
Combined prognostic systems 5183 HCC patients (database ITA.LI.CA. - Italian Liver Cancer) Training cohort (3628 pts) / Internal validation cohort (1555 pts) M.F = 3:1 Age (median): 68 yrs Child (median): 6 MELD (median): 11 HCV+ 60%, HBV+ 17%, Alcool 26% Diameter max (median): 3 cm Multifocal: 22% Metastases: 3% BCLC: 0 7%, A 33%, B 12%, C 42%, D 6% Main Treatment: Resection 11%, Transplant 2%, Ablation 30%,TACE 26%, Sorafenib 3%, Other 8%, BSC 20% External validation cohort: 2651 pts (database from Taipei Taiwan, 2000-2012) Tumor stage classification based on the literature Final score based on multivariate analysis Staging systems did not respect Proportional Hazard Assumption Kaplan-Meier, log-rank test Multivariate log-logistic parametric survival model: ITA.LI.CA staging construction Multivariate log-logistic parametric survival model: comparison between systems Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC)
Combined prognostic systems UNOS TNM, 2002 Hong Kong Liver Cancer Staging System BCLC B HCC: Proposal for a Subclassification ITA.LI.CA TUMOR STAGING Variables 0 A B1 B2 B3 C Diameter(cm) < 2 3 5 3-5 > 5 3-5 > 5 > 5 Any Any N nodules 1 2-3 1 2-3 1 > 3 2-3 > 3 Any Any Vascular invasion or metastases no no no no no no no no Intra Exta Farinati F, et al. PLoS Med 2016; 13(4):e1002006
Combined prognostic systems Farinati F, et al. PLoS Med 2016; 13(4):e1002006
Combined prognostic systems Farinati F, et al. PLoS Med 2016; 13(4):e1002006
Staging and prognostic systems: beyond BCLC? Data based prognostic scores Evidence based staging systems Combined prognostic systems The issue of treatment allocation
HCC Prognostic Factors Patient Prognosis of HCC 1 Most patients have underlying liver disease Key prognostic indicators are not clearly defined Prognostic indicators vary during the course of disease ECOG PS BCLC 4 CUPI 5 Factors affecting HCC prognosis 2,3 Tumour stage Liver function Health status Liver Child- Pugh GRETCH 6 Okuda 7 CLIP 8 JIS 9 TNM Tumour 1. Llovet JM, et al. Lancet 2003;362:1907 1917. 2. Marrero JA, et al. Clin Liver Dis 2006;10:339 351. 3. Marrero JA, et al. Hepatology 2005;41:707 716. 4. Llovet JM, et al. Semin Liver Dis 1999;19:329 338. 5. Leung T, et al. Cancer 2002;94:1760 1769. 6. Chevret S, et al. J Hepatol 1999;31:133 141. 7. Schafer DF, et al. Lancet 1999;353:1253 1257. 8. CLIP. Hepatology 1998;28:751 755. 9. Makuuchi M, et al. World J Gastroenterol 2006;12:828 829.
The issue of treatment allocation both the stage and the various types of intervention should ideally be built into the prognostic system.. There are four main factors affecting prognosis: (a) the stage, aggressiveness and growth rate of the tumor; (b) the general health of the patient; (c) the liver function of the patient; and (d) the specific intervention The.. optimal solution would be to develop a prognostic model for each relevant evolutionary stage of the disease (early, intermediate- advanced and terminal) and model into each stage the variables related to each specific intervention.
The issue of treatment allocation Bruix J and Sherman M, et al. Hepatology 2005
The issue of treatment allocation PROBLEM 1: PROGNOSTIC PROBLEM Pre-determined Staging System Treatment selection Treatment selection Staging System PROBLEM 1: Tumor, liver function, patient related variables differently influence Treatment selection and patient prognosis
The issue of treatment allocation PROBLEM 1: PROGNOSTIC PROBLEM?????? Bruix J and Sherman M. Gastroenterology 2016
Combined prognostic systems PROBLEM 1: PROGNOSTIC PROBLEM Farinati F, et al. PLoS Med 2016; 13(4):e1002006
The issue of treatment allocation PROBLEM 1: PROGNOSTIC PROBLEM 13 points score?? Yau T, et al. Gastroenterology 2014
The issue of treatment allocation PROBLEM 1: PROGNOSTIC PROBLEM Yau T, et al. Gastroenterology 2014
The issue of treatment allocation PROBLEM 2: THERAPEUTIC PROBLEM Pre-determined Staging System Treatment selection Treatment selection Staging System PROBLEM 2: Treatment selection criteria change with time and should be inclusive (indications better than algorithms)
The issue of treatment allocation PROBLEM 2: THERAPEUTIC PROBLEM The tempting simplicity of the BCLC classification came at a price of low clinical utility by compromising the importance of liver transplantation and locoregional therapies in medical management of HCC. Designed using data mostly acquired in small Western patient populations, the BCLC classification lacks universal applicability in terms of discriminatory ability and prognostic accuracy with regard to treatment recommendations. In fact, the BCLC system precludes patients with more advanced disease from receiving radical therapies out of safety considerations. Chapiro J, et al. Nat Rev Gastroenterol Hepatol 2014; 11: 334
?? The issue of treatment allocation PROBLEM 2: THERAPEUTIC PROBLEM????? Bruix J and Sherman M. Gastroenterology 2016
The issue of treatment allocation PROBLEM 2: THERAPEUTIC PROBLEM Roayaie S, et al. Hepatology 2015; 62: 440
The issue of treatment allocation PROBLEM 2: THERAPEUTIC PROBLEM 1302 BCLC A patients undergoing resection Roayaie S, et al. Hepatology 2015; 62: 440
The issue of treatment allocation PROBLEM 2: THERAPEUTIC PROBLEM A total of 3515 treatment-naıve, newly diagnosed HCC patients at a single centre were analyzed Kim KM, et al. Liver Int 2016.
The issue of treatment allocation SOLUTION 1: INDEPENDENT ALGORITHM Kudo M, et al. Dig Dis 2011;29:339 364
The issue of treatment allocation SOLUTION 2: TREATMENT INDICATIONS Bolondi L, et al. Sem Liv Dis 2012
New ITA.LI.CA 2015 database including 6669 HCC patients (database ITA.LI.CA. - Italian Liver Cancer) Inclusion criteria: - Cirrhotic patients - Complete follow-up data - Period 2002 2015 Study population: 4867 HCC patients The issue of treatment allocation SOLUTION 3: TREATMENT INDICATIONS+DATA BASED SURVIVAL BENEFIT External validation cohort: 2651 pts (database from Taipei Taiwan, 2002-2012) ITA.LI.CA treatment indications (no algorithm) were based on: TREATMENT SELECTION as end-point: multivariate logistic regression models SURVIVAL BENEFIT as end-point: multivariate loglogistic parametric survival models Treatment selection and survival benefit were combined using Inverse Probability Weight (IPW) EVIDENCE BASED approach (literature multiple societies document)
Variables 0 A B1 B2 B3 C Any Functional score (FS) FS 2: CTP ABandPST 0; CPT 7 and PST 2 FS > 2: CTP C/PST >2 Diameter(cm) < 2 3 5 3-5 > 5 3-5 > 5 > 5 Any Any Any N nodules 1 2-3 1 2-3 1 > 3 2-3 > 3 Any Any Any VI / meta no no no no no no no no Intra Extra Any Median survival 71 55 46 33 16 14 8 Therapy LT LR ABL IAT SOR BSC 120 102 77 64 120 76 61 50 120 64 46 33 120 50 40 33 25 28 15 18 16 7 102 5 Therapy LT LR ABL IAT SOR BSC 49 31 6 Neg 65 21 6 Neg 74 18 0 Neg 87 17 7 0 Neg 12 1 2 0 Neg 94 5
The issue of treatment allocation Variables 0 A B1 B2 B3 C Any Functional score (FS) FS 2: CTP ABandPST 0; CPT 7 and PST 2 FS > 2: CTP C/PST >2 Diameter(cm) < 2 3 5 3-5 > 5 3-5 > 5 > 5 Any Any Any N nodules 1 2-3 1 2-3 1 > 3 2-3 > 3 Any Any Any VI / meta no no no no no no no no Intra Extra Any Median survival 71 55 46 33 16 14 8 LIVER TRANSPLANTATION LT ABLATION LIVER RESECTION TACE/TARE SORAFENIB
The issue of treatment allocation SOLUTION 3: TREATMENT INDICATIONS+DATA BASE SURVIVAL BENEFIT Following ITA.LI.CA indications: 3153 pts (65%) vs 43% BCLC vs 55% HKLC algorithms In (MS 41 mo) Out (MS 41 mo)
CONCLUSIONS There is not worldwide consensus on the best prognostic system for HCC patients The ITA.LI.CA prognostic score showed the best predictive ability in large western and eastern cohorts Beyond BCLC?: - YES, for intrinsicprognosticpitfals(evidencebasedand treatment dependent system) - YES, for treatment relatedpitfals(distancefrom best clinical practice and personalized approach) ITA.LI.CA treatment indications could represent a potential solution??
Department of General Surgery and Organ Transplantation, Hepatobiliary Surgery and Liver Transplantation Unit, University Hospital of Padua, Padua; Italy Director: Prof. Umberto Cillo THANK YOU