Beyond Milan

Abstract

Download video file ^{(19.9MB, mp4)}

INTRODUCTION

Liver transplantation (LT) is the preferred treatment for patients with unresectable hepatocellular carcinoma (HCC) and cirrhosis because it addresses both cancer and the underlying liver disease. Early experience with LT for HCC had been marred by poor survival (30%‐40% at 5 years after LT) and frequent tumor recurrence because of lack of careful patient selection. In 1996, Mazzaferro et al.¹ in Milan, Italy, showed that outcomes after LT for unresectable HCC are acceptable (83% recurrence‐free survival [RFS] rate at 4 years) and comparable with cirrhosis if the HCC tumor burden is one tumor ≤5 cm or two to three tumors ≤3 cm each and without extrahepatic, lymph node, or vascular invasion. In the United States, post‐LT survival for HCC within Milan criteria (MC) was 89% at 1 year and 74% at 5 years,² validating Mazzaferro et al.'s findings. Since its inception, MC has become the worldwide standard defining the upper limit of HCC tumor burden that has acceptable post‐LT outcomes.

EXPANDED CRITERIA

Some patients with larger or more numerous tumors than allowed by MC are potentially deprived of curative LT because of the stringent and dichotomous nature of MC. Expanded transplant criteria have been proposed that permit LT in selected patients with HCC beyond MC to improve patient access to LT. These criteria allow for a modest increase in size and number of tumors for LT and used markers of tumor biology such as α‐fetoprotein (AFP). There is no consensus on preferred expanded HCC criteria for LT, but outcomes of expanded criteria should be comparable with MC and non‐HCC LT indications.^3,4 Table 1 outlines these expanded transplant criteria and their associated outcomes.

TABLE 1. Milan Criteria and Expanded Criteria for LT for HCC.

Selection Criteria	Criteria	Outcomes	Refs.
Milan	1. One tumor ≤5 cm	4‐year RFS: 83%	Mazzaferro et al.1
	2. Two to three tumors ≤3 cm each	HCC recurrence: 8.3%
	3. No vascular invasion
	4. No lymph node or distant metastases
UCSF	1. One tumor ≤6.5 cm	1‐year RFS: 92.1%	Yao et al.5,6
	2. One to three tumors ≤4.5 cm each and total tumor diameter ≤8 cm	5‐year RFS: 80.7%
	3. No vascular invasion	5‐year HCC recurrence: 6.4%
	4. No lymph node or distant metastases
Up‐to‐seven	Seven as the sum of:	5‐year OS: 71.2%	Mazzaferro et al.7
	1. Size of the largest tumor (in cm)
	2. Number of tumors
	No microvascular invasion
French‐AFP	Largest diameter (cm)	≤2 points:	Duvoux et al.8
	≤3 cm: 0 points	5‐year OS: 67.8%
	3–6 cm: 1 point	5‐year recurrence: 8.8%
	>6 cm: 4 points	>2 points:
	Number of nodules	5‐year OS: 47.5%
	1–3: 0 points	5‐year recurrence: 50.6%
	≥4: 2 points
	AFP level (ng/ml)
	≤100: 0 points
	100–1000: 2 points
	>1000: 3 points
Metroticket 2.0	AFP < 200 ng/ml: sum of tumor size and number ≤7	5‐year OS: 79.7%	Mazzaferro et al.9
	AFP 200–400 ng/ml: sum of tumor size and number ≤5	5‐year RFS: 89.6%
	AFP 400–1000 ng/ml: sum of tumor size and number ≤4
NYCA score	Largest diameter at diagnosis (cm)	5‐year recurrence:	Halazun et al.10
	≤3 cm: 0 points	Low risk, 0–2 points: 9.5%
	3–6 cm: 2 points	Acceptable risk, 3–6 points: 20.5%
	>6 cm: 4 points	High risk, ≥7 points: 40.5%
	Number of nodules at diagnosis
	1: 0 points
	2–3: 2 points
	≥4: 4 points
	AFP response (maximum AFP minus final AFP before LT)
	AFP always < 200 ng/ml: 0 points
	Responders:
	Maximum 200–1000 and final <200 ng/ml: 2 points
	Maximum >1000 and final <1000 ng/ml: 2 points
	Nonresponders:
	Maximum 200–400 and final >200 ng/ml: 3 points
	Maximum 400–1000 and final >200 ng/ml: 4 points
	Maximum >1000 and final >1000 ng/ml: 6 points

University of California, San Francisco (UCSF) investigators retrospectively analyzed HCC in explants and found that having a single tumor ≤6.5 cm and two to three tumors ≤4.5 cm each with total tumor diameter ≤8 cm did not compromise transplant outcomes.⁵ UCSF criteria had been validated prospectively using preoperative imaging and been shown to have 1‐ and 5‐year RFS rates of 97% and 94%, respectively.⁶ Mazzaferro et al.⁷ identified a subset from an international registry of patients with HCC beyond MC who have comparable post‐LT survival to HCC within MC. Their up‐to‐seven criteria require that the sum of tumor number and largest tumor diameter (in cm) should be <7 and is associated with 5‐year overall survival (OS) of 71%.⁷

The French‐AFP and Metroticket 2.0 models included AFP as a marker of tumor biology with tumor size and number.^8,9 HCC beyond MC with low serum AFP is associated with acceptable post‐LT survival and tumor recurrence rates. For example, in the French‐AFP model, a patient with three tumors (0 points), with largest diameter up to 6 cm (1 point), will have an estimated 5‐year OS rate of 68% and 5‐year HCC recurrence rate of <10% (Table 1).⁸ Metroticket 2.0 (available online at www.hcc‐olt‐metroticket.org) provides individualized estimates of post‐LT outcomes based on tumor size, number, and AFP. Using the Metroticket 2.0 calculator, a patient with four tumors, with largest diameter of 5.5 cm, has a 76% predicted 5‐year HCC‐specific survival rate after LT.⁹ While French‐AFP and Metroticket 2.0 models used static AFP measurements, the New York/California (NYCA) score used dynamic AFP changes, which improved the model's predictive power for tumor recurrence compared with MC, French‐AFP, and Metroticket 2.0 models.¹⁰

DOWNSTAGING CRITERIA

Another approach to improve access to LT for HCC is downstaging (DS). DS involves the use of locoregional therapies (LRTs) to reduce HCC tumor burden from beyond MC to within MC. A retrospective multicenter study showed that transplant outcomes, including HCC recurrence, in downstaged tumors were, although not as good as in HCC within MC, acceptable and superior to HCC beyond MC that did not undergo DS.¹¹

UCSF investigators proposed upper limits of HCC tumor burden eligible for DS, which the United Network for Organ Sharing (UNOS) adopted as the standard national DS protocol (UNOS‐DS) in 2017 (Figure 1).^12,13 UNOS‐DS criteria include one tumor ≤8 cm, two to three tumors ≤5 cm each with total tumor diameter ≤8 cm, or four to five tumors ≤3 cm each with total tumors diameter ≤8 cm and without extrahepatic spread or vascular invasion on cross‐sectional imaging. HCC downstaged to within MC and pretreatment AFP < 1000 ng/ml (or posttreatment AFP < 500 ng/ml if pretreatment AFP ≥ 1000 ng/ml) automatically qualifies for standard Model for End‐stage Liver Disease (MELD) exception points according to UNOS policy.¹³ A minimum 3‐month observation period after LRT is recommended for downstaged tumors before LT to ensure tumor stability.¹²

FIGURE 1.

UNOS HCC DS protocol.¹⁰

UNOS data show that LT recipients downstaged from UNOS‐DS criteria to MC had comparable 3‐year post‐LT survival (79%) to recipients with HCC always within MC (83%).¹⁴ UNOS‐DS criteria were prospectively evaluated in a multicenter and multiregional cohort study.¹⁵ In this study, the 1‐year probability of successful DS from UNOS‐DS to MC was 83%. More than 95% of downstaged patients were listed for LT and, at the end of follow‐up, 30% had been transplanted and 34% remained active in the wait list. Tumor progression leading to wait‐list dropout occurred in 9% of downstaged patients. Downstaged patients had good post‐LT outcomes: 83% 3‐year survival probability and 8% HCC recurrence rate, despite a high frequency of explants with tumors beyond MC (43%) and microvascular invasion (18%).

LRT options for DS include chemical or thermal ablation, transarterial chemoembolization (TACE), yttrium‐90 (Y‐90) radioembolization, and stereotactic body radiation. A systematic review and a more recent multicenter prospective cohort study found that TACE and Y‐90 had comparable DS success and post‐LT outcomes (Table 2).^15,16 However, in comparing explants, Y‐90 had lower, but nonsignificant, frequency of tumors beyond MC (23% vs. 43%; p = 0.44) and microvascular invasion (8% vs. 21%; p = 0.29) than TACE.¹⁵ The American Association for the Study of Liver Diseases does not make a recommendation on the optimal form of LRT for DS.¹⁷

TABLE 2. Comparison of DS outcomes between TACE and radioembolization.

	TACE	Y‐90 Radioembolization	p
Systematic review by Parikh et al.16
DS success	Pooled rate: 48% (95% CI: 28%‐68%)	Pooled rate: 37% (95% CI: 15%‐61%)	0.51
Post‐LT HCC recurrence	Pooled rate: 17% (95% CI: 8%‐27%)	Pooled rate: 26% (95% CI: 6%‐53%)	0.40
Multicenter, prospective cohort study by Mehta et al.15
Radiographic tumor response (mRECIST criteria)	Complete response: 28.0%	Complete response: 27.4%	0.67
	Partial response: 52.3%	Partial response: 48.4%
	Stable disease: 10.6%	Stable disease: 11.3%
	Progressive disease: 9.1%	Progressive disease: 12.9%
Median SD time	2.9 months	2.4 months	0.73
DS success	85.6%	80.6%	0.38
Wait‐list dropout	36.4%	32.3%	0.58
LT cumulative incidence	33.3%	22.6%	0.18
Post‐LT survival	Combined data for TACE and Y‐90: 1‐year 100%, 2‐year 95%, 3‐year 83%
	No difference in intention‐to‐treat survival from receipt of first LRT between TACE and Y‐90

UPPER LIMITS OF TUMOR BURDEN FOR LT

The goal of LT is to provide all patients with liver disease with maximum benefit and equitable access to a limited pool of donor organs. Expanding transplant and DS criteria for HCC should improve survival of the individual patient without compromising outcomes and access to LT of other patients on the wait list. In 2010, a consensus conference endorsed by several international societies on liver disease and transplantation formulated guidelines on the indications and conduct of LT.⁴ The panel did not specify upper limits of HCC burden acceptable for LT but provided guidance on how transplant criteria for HCC should be scrutinized. Using MC as the benchmark, the panel recommended that expanded selection and DS criteria for beyond MC tumors should have comparable 5‐year post‐LT survival to HCC within MC and non‐HCC indications.

Due to high risk of tumor recurrence, extrahepatic spread and radiographic evidence of vascular invasion are generally accepted absolute contraindications for LT. Of interest, Italian investigators performed a pilot study of LT after Y‐90 DS of HCC with tumor thrombus up to the first portal vein branch and described 60% 5‐year post‐LT OS.¹⁸ The study was limited by infrequent DS success and high tumor recurrence. AFP > 1000 ng/ml is a strong predictor of vascular invasion that is associated with 5‐year RFS of only 53%, compared with 80% for AFP ≤ 1000 ng/ml, and may also be considered a contraindication to LT.¹⁹

Chapman et al.²⁰ showed comparable 5‐year OS of >70% in LT recipients with HCC within MC, HCC within UCSF, and HCC beyond UCSF, suggesting that there are no tumor burden restrictions for LT as long as tumors can be downstaged to within MC. In contrast, Sinha et al.²¹ concluded that UNOS‐DS criteria may be the upper limit of tumor burden that has acceptable DS and transplant outcomes. In their study, only 14% of patients with pretreatment tumors beyond UNOS‐DS made it to transplant after DS, with 50% survival and 40% recurrence‐free probability at 5 years after LT, compared with 79% and 86% for HCC within UCSF‐DS. UNOS data also showed inferior post‐LT survival (71% at 3 years) and higher rates of vascular invasion (24%) in pretreatment HCC beyond UNOS‐DS.¹⁴

In conclusion, LT is potentially curative and the preferred treatment in patients with unresectable HCC and cirrhosis. Appropriate patient selection is crucial for good individual patient outcomes and maximum societal benefit of LT. MC is the benchmark transplant selection criteria, although liberalized criteria that modestly increase the upper limits of tumor burden for LT have acceptable outcomes. DS allows patients with HCC beyond MC or UCSF criteria to safely undergo LT. To date, there is no absolute upper limit of tumor burden that precludes LT; however, UNOS‐DS criteria are a reasonable ceiling, provided that patients can be downstaged to within MC.