1. Risk assessment for outcomes following potentially curative treatments for HCC
Francis Yao, M.D
Professor of Clinical Medicine and Surgery
Director, Hepatology
Medical Director, Liver Transplantation
University of California, San Francisco

2. Potentially curative treatments for HCC
Liver Transplant
Liver Resection
Thermal Ablation
5-year survival
70-80% (Milan)
50-60%
Tumor recurrence
10-15% (Milan)
50-70%
Time to Treatment
May be prolonged: Risk for waitlist removal due to tumor progression
Immediate
At risk for hepatic decompensation No
Yes

3. EASL 2018 algorithm for curative treatments of HCC
Very early stage HCC
Single < 2 cm
Preserved liver function, PS 0
Early stage HCC
Single or 2-3 nodules < 3 cm
Preserved liver function, PS 0
Solitary
2-3 nodules
- Child’s class
- MELD score
- portal hypertension
- Residual liver
Optimal surgical candidate
Transplant candidate
Yes No
Yes No
Ablation
Resection
Liver Transplant
Ablation
EASL Clinical Practice Guidelines. J Hepatol 2018;69:

4. Liver Resection: Assessing risks of liver decompensation
Portal Hypertension
Extent of hepatectomy
Extent of hepatectomy
Minor
(< 3 segment)
Major
(> 3 segment)
Minor
(< 3 segment)
Major
(> 3 segment)
MELD score
< 10
> 10
Low risk
5% risk of liver decompensation
Liver related morality 0.5%
Intermediate risk
<30% risk of liver decompensation
Liver related morality 9%
High risk
>30% risk of liver decompensation
Liver related morality 25%
Citterio D et al. JAMA Surg 2016;15:

5. Liver Resection: Assessing risks of liver decompensation
Indocyanine green (ICG) routinely used in the East in clinical practice since criteria proposed by Makuuchi et al.1
ICG remains an important factor in risk stratification of post-operative liver decompensation in the East.2
Liver Cancer Study Group of Japan
1. Makuuchi M et al. Sem Surg Oncol 1993;9:
2. Kudo M et al. Dig Dis 2011;29:

6. Predictors of early HCC recurrence < 2 years of resection
ERASL-pre
ERASL-pre
low
Intermediate
high
1.00
0.75
0.50
0.25
0.0
ERASL-pre
Gender
ALBI grade
AFP
Tumor size
Tumor number
Recurrence-free survival (months)
ERASL-post
ERASL-post
low
Intermediate
high
1.00
0.75
0.50
0.25
0.0
ERASL-pre
Gender
ALBI grade
Microvascular invasion
AFP
Tumor size
Tumor number
Recurrence-free survival (months)
Chan AWH et al. J Hepatol 2018;69:

7. Pattern and risk of HCC recurrence after resection
No recurrence
Recurrence
> Milan or
(+) metastasis
resection
What % of HCC recurrence?
Recurrence
within Milan
“Salvage” liver transplant

8. Resection and salvage transplant: Risk of HCC recurrence > Milan
Total of 304 patients, median follow-up 47 months
Pre-op Milan (n=94) – 19% HCC recurrence > Milan
Pre-op > Milan (n=210) – 46% HCC recurrence > Milan
Cumulative Incidence
CRS O
CRS 1
CRS 2
CRS 3
Clinical Risk Score
Initial tumor > Milan (pre-operative)
Microsatellites or multiple tumors
Lymphovascular disease
Years
Lee SY et al. HPB 2014;16:

9. Pattern and risk of HCC recurrence after resection
No recurrence
“preemptive” liver transplant
Recurrence
> Milan or
(+) metastasis
resection
What % of HCC recurrence?
Recurrence
within Milan
“Salvage” liver transplant

10. Preemptive transplant after resection in high risk patients
Microvascular invasion in resection specimen:
17/37 high risk patients underwent LT (10 of 17 before HCC recurrence), with reported 5-year post-transplant survival of 82%.1
At least 3 high risk factors for HCC recurrence > Milan.2
Cirrhosis present
Poorly differentiated grade
Microvascular invasion
Tumor diameter > 3 cm
Satellite nodules
1. Ferrer-Fabrega J, et al. Hepatology 2016;63:
2. Fuks D, et al. Hepatology 2012;55:

11. Liver resection: Minimally invasive/ laparoscopic
EASL recommendations: Liver resection via minimally invasive/ laparoscopic approach may be considered in properly trained centers, especially for tumors located in anterolateral and superficial locations.1
No differences between laparoscopic versus open approaches in overall survival and disease-free survival.2-4
Whether laparoscopic approach reduce incidence of hepatic decompensation and perioperative morbidity has not yet been established.2-4
A multi-center Japanese study reported shorter hospital stay and lower complication rate with laparoscopic approach.5
EASL Clinical Practice Guidelines. J Hepatol 2018;69:
Parks KR, et al. HPB (Oxford) 2014;16:
Kasai M, et al. Surgery 2018;163:
Abu Hilal M, et al. Ann Surg 2018;268:11-18
Takahara T, et al. J Hepatobiliary Pancreat Sci 2015;

12. Randomized controlled trials of resection versus RFA
Inclusion Criteria
Overall Survival
Other outcomes
Chen et al
Resection (n=90)
RFA (n=71)
1 lesion ≤ 5 cm
No difference
More complications with resection
Huang et al
Resection (n=115)
RFA (n=115)
Milan criteria
Better with resection
Lower HCC recurrence with resection
Feng et al
Resection (n=84)
RFA (n=84)
Up to ≤ 4 cm
and ≤ 2 lesions
Fang et al
Resection (n=60)
RFA (n=60)
1 lesion ≤ 3 cm
Ng et al
Resection (n=109)
RFA (n=109)
Trend for better disease-free survival with resection
Chen MS et al. Ann Surg 2006;243: Huang J et al. Ann Surg 2010;252:
Feng K et al. J Hepatol 2012;57: Fang Y et al. J Gastroenterol Hepatol 2014;29:
Ng KKC et al. Br J Surg 2017;104:

13. Randomized controlled trial of resection versus RFA
All Patients; 1 lesion ≤5 cm
Overall Survival
Disease-free Survival
P = 0.531
P = 0.072
Resection
RFA
Resection
RFA
Ng KKC et al. Br J Surg 2017;104:

14. Randomized controlled trial of resection versus RFA
Very early HCC; 1 lesion ≤2 cm
Overall Survival
Disease-free Survival
P = 0.95
P = 0.90
Resection
RFA
Resection
RFA
Ng KKC et al. Br J Surg 2017;104:

15. Ablation: Factors in outcome assessment
RFA
Tumor Location
Tumor Diameter ≤
3 cm
vs 3-5 cm
AFP

16. Ablation: Factors in outcome assessment
RFA
Tumor Location
Tumor Diameter ≤
3 cm
vs 3-5 cm
AFP
Procedure-related complications
Liver function
Coagulopathy
Thrombocytopenia
Ascites
Tumor location

17. Thermal Ablation: Very early HCC
Single Tumor < 2 cm
A multi-center study on 218 patients with single lesion < 2cm, median follow-up 31 months.
Sustained complete response in 97% after 1 (86%) or 2 (12%) sessions.
5-year survival 55%, perioperative mortality 0% and major complication rate 1.8%.
5-year disease free survival rate 26%.
Livraghi T, et al. Hepatology 2008;47:82-89

18. Thermal Ablation: Very early HCC
Single Tumor < 2 cm
Systematic review and meta-analysis of 17 studies (3996 treated with resection and 4424 with ablation), with cost-effectiveness using a Markov model.
Very early HCC < 2 cm in Child’s class A patients: RFA provides similar life expectancy and quality-adjusted life expectancy at a lower cost compared to resection.
Cucchetti A, et al. J Hepatol 2013;59:

19. Thermal Ablation for HCC > 3 cm
≤ 3 cm versus > 3 cm
Treatment response rate 70-95% for lesions < 3 cm versus around 50% for lesions > 3 cm
In lesions > 3 cm, overall 5-year survival 30-35%, 5-year recurrence rate up to 80%.
Sala M, et al. Hepatology 2004;40:
Lencioni R, et al. Radiology 2005;234:961-7
N’Kontchou G, et al. Hepatology 2009;50:
Santambrogio R, et al. Ann Surg Oncol 2009;16:

20. Resection vs ablation for HCC
Well compensated cirrhosis, no portal hypertension
< 2 cm
2-3 cm
> 3 and ≤ 5 cm
Liver Resection √
Ablation
* Resection may be preferred over RFA for high alpha-fetoprotein

21. Survival outcome after RFA for HCC ≤ 3 cm
Recurrence-free Survival
Overall Survival v
Recurrence-free Survival
HCC > 2 & ≤ 3 cm
HCC ≤ 2 cm
Cumulative probabilities
HCC > 2 & ≤ 3 cm
HCC ≤ 2 cm
HCC > 2 & ≤ 3 cm
HCC ≤ 2 cm
Overall Survival
79%
P = 0.01
71%
P < 0.001
P = 0.001
Doyle A, et al. J Hepatol 2019;70:

22. HCC recurrence pattern after RFA for HCC ≤ 3 cm
Total
(n=301)
≤ 2 cm
(n=167)
> 2 and ≤ 3 cm
(n= 134)
P-value
HCC recurrence
199 (66%)
105 (63%)
94 (70.1%)
0.18
Beyond Milan
At first recurrence
83 (28%)
38 (13%)
36 (22%)
15 (9%)
47 (36%)
23 (17%)
0.01
0.03
Reasons > Milan
Tumor size/ number
Vascular invasion
Metastatic disease
29 (35%)
30 (36%)
24 (29%)
11 (31%)
15 (42%)
10 (28%)
18 (38%)
15 (32%)
14 (30%)
0.78
0.36
0.84
Doyle A, et al. J Hepatol 2019;70:

23. HCC recurrence pattern after RFA for HCC ≤ 3 cm
HCC Recurrence > Milan Criteria v
Cumulative probabilities
HCC > 2 & ≤ 3 cm
HCC ≤ 2 cm
Recurrence-free Survival
HCC > 2 & ≤ 3 cm
HCC ≤ 2 cm
Factors
(multivariate) HR
HCC size > 2 cm
(vs ≤ 2 cm)
1.94 (p=0.01)
AFP
2.05 (p=0.02)
AFP > 1000
2.06 (p=0.12)
P < 0.001
P = 0.01
Doyle A, et al. J Hepatol 2019;70:

24. Liver Transplant: Outcome risk assessment
Immuno-suppression
and complications
Liver disease
Psycho-social
issues
Liver
Transplant
Outcome
HCC
Tumor burden + others factors
Medical
Co-morbidities
Donor factors

25. Liver Transplant: Outcome risk assessment
Immuno-suppression
and complications
Liver disease
Psycho-social
issues
Liver
Transplant
Outcome
HCC
Tumor burden + others factors
Selection
Criteria
Medical
Co-morbidities
Donor factors

26. The HCC “Metro-ticket” – Tumor Size and Number
HCC Forecast Chart: Survey of 1112 patients > Milan (Pathology)
Number of nodules
Size of the largest nodule (in mm)
Courtesy of Dr. Vincenco Mazzaferro, with permission
Mazzaferro et al. Lancet Oncology 2009;10:35-43

27. The HCC “Metro-ticket” – Tumor Size and Number
HCC Forecast Chart: Survey of 1112 patients > Milan (Pathology)
Number of nodules
Milan Criteria
Size of the largest nodule (in mm)
Courtesy of Dr. Vincenco Mazzaferro, with permission
Mazzaferro et al. Lancet Oncology 2009;10:35-43

28. Liver Transplant for HCC Changing views on Selection Criteria
“Morphology”
Size
Number
Volume
“Biology”
Response to LRT?
Biomarkers?
Other surrogates?

29. French AFP Model - Tumor size, number and AFP
Variables Points
Largest tumor diameter, cm ≤
>
Number of tumor nodules ≥
AFP level, ng/mL ≤
>
Duvoux C et al. Gastroenterology 2012;143:986-94

30. national policy in France
French AFP Model - Tumor size, number and AFP
Variables Points
Largest tumor diameter, cm ≤
>
Number of tumor nodules ≥
AFP level, ng/mL ≤
>
Low risk
≤ 2 points
Within Milan but AFP > 1000
= High risk
Some HCC > Milan but AFP ≤ 100
= Low risk
national policy in France
Duvoux C et al. Gastroenterology 2012;143:986-94

31. Metro-ticket 2.0: AFP + Tumor Burden
Mazzaferro et al. Gastroenterology 2018;154:

32. AFP and Liver Transplant (UNOS) UNOS data 2005-2015 Within Milan or
100
AFP > < 100
Survival rate (%) 80
88%
AFP > 60
67%
P <
AFP > 1000
49% 40
UNOS data Within Milan or
UCSF Down-staging 20
P = 0.09 for AFP vs AFP < 100
Years after liver transplant
<
>
Mehta N et al. Hepatology 2019;69:

33. HCC-HALT: AFP, Tumor Burden score, MELD-Na
(1·27 × Tumer Burden Score (TBS)) + (1·85 × lnAFP) + (0·26 × MELD-Na) 
AUROC 0.61 79 75 72 62
Sasaki et al. Lancet Gastroenterol Hepatol 2017;2:

34. UCSF Down-staging protocol for liver transplant
Minimum observation 3 months
(n=41)
Meeting Milan criteria
Median f/u 3.8 years
Dropout
- 5 HCC recurrence (8%)
- 78% 5-yr survival
post-transplant
- 91% 5-yr recurrence
free probability
- 56% 5-yr intention-
to-treat survival
Transplant
Down-staging
(n=118)
(n=64)
Inclusion Criteria for Down-staging
1 tumor ≤ 8 cm
2-3 tumor ≤ 5 cm + total diameter ≤ 8 cm
4-5 tumor ≤ 3 cm + total diameter ≤ 8 cm
US national policy
Yao FY et al. Hepatology 2015;61:

35. Post-transplant survival after tumor down-staging
- - - Milan group
81%
Down-staged group
78%
p=0.69
Yao FY, et al. Hepatology 2015;61:

36. Predictors of Treatment Failure:
Multivariable Analysis
Treatment failure defined as dropout due to tumor progression, liver-related death without LT, or post-LT HCC recurrence
Predictor 
Multivariable HR (95% CI) p-
value
Child-Pugh B/C vs A
1.6 ( )
0.04
AFP* >1000 vs <1000
3.3 ( )
<0.001
* Before 1st down-staging procedure
Mehta N et al. Clin Gastroenterol Hepatol 2018;16:

37. Down-staging treatment failure: AFP and Child’s Class
100%
Risk factors
- Pre-treatment AFP > 1000
- Child-Pugh B/C
2 Risk Factors
46%
1 Risk Factor
33%
0 Risk Factors
p=0.001
Mehta N et al. Clin Gastroenterol Hepatol 2018;16:

38. The effects of initial tumor burden
Post-transplant survival after down-staging
The effects of initial tumor burden
100
Post-transplant Survival (%)
83%
79%
71% 80
Milan (n=3276)
UNOS-DS (n=422)
>UNOS-DS or “All-comers” (n=121) 60
UNOS-DS vs Milan, p=0.17
>UNOS-DS vs Milan, p=0.04 40 20
Months after Transplant
UCSF/ UNOS-down-staging Inclusion Criteria
1 tumor ≤ 8 cm
2-3 tumor ≤ 5 cm + total diameter ≤ 8 cm
4-5 tumor ≤ 3 cm + total diameter ≤ 8 cm
Mehta N, et al. Hepatology [Epub]

39. Pre-Transplant Selection
Pre-transplant Prognostic Models (selected)
Pre-Transplant Selection
Tumor Burden
Biomarkers
AUROC
US National Policy 1,2
Milan or Down-staged to Milan
No AFP > 1000
(reduced to < 500)
French AFP Model 3
Largest tumor Size and total number
AFP
0.7
Metro-ticket 2 4
0.72
HCC-HALT* 5
Tumor burden score (size and number)
0.61
TTV + AFP 6
TTV ≤ 115 cm3
AFP ≤ 400 ng/ml
Pre-MORAL 7
Largest tumor size
AFP, NLR
0.82
1. Yao FY, et al. Hepatology 2015;61:
2. Hameed B. et al. Liver Transpl 2014;20:
3. Duvoux et al. Gastroenterology 2012;143:986-94
4. Mazzaferro et al. Gastroenterology 2018;154:
5. Sasaki et al. Lancet Gastroenterol Hepatol 2017; 2:
6. Toso et al. Hepatology 2015;62:
7. Halazun KJ, et al. Ann Surg 2017;265:
*Include MELD-Na

40. Post-transplant Prognostic Models
Explant Pathology
Biomarkers
AUROC
US HCC Consortium 1
Tumor burden (radiologic)
> Milan, not down-staged
Largest tumor diameter
Non-incidental
Vascular Invasion
Macro-vascular
Micro-vascular
Tumor Differentiation
Poorly differentiated
Moderately differentiated
AFP
NLR
0.76
RETREAT 2
Tumor size and number (Viable)
0.77
Post-MORAL 3
Tumor size and Number
0.87
1. Agopian VG. et al. J Am Coll Surg 2015;220:
2. Mehta N. et al. JAMA Oncol 2017;3:
3. Halazun KJ, et al. Ann Surg 2017;265: