EASL Clinical Practice Guidelines

Liver transplantation

Liver transplantation is the first treatment choice for patients with small multinodular tumors (⩽3 nodules ⩽3 cm) or those with single tumors ⩽5 cm and advanced liver dysfunction. Theoretically, transplantation may simultaneously cure the tumor and the underlying cirrhosis. The broad selection criteria applied two decades ago led to poor results in terms of recurrence (32–54% at 5 years) and survival (5-year survival <40%), but allowed the identification of the best candidates for this procedure [[210], [211]]. Following this concept, some pioneering groups selecting ''optimal candidates'' reported 70% 5-year survival with a recurrence rate below 15% [[161], [212], [213], [214], [215]]. In a landmark manuscript the so-called Milan criteria were established for patients with a single HCC ⩽5 cm or up to three nodules ⩽3 cm [212]. Following these criteria and according to modern standards, peri-operative mortality, 1 and 5-year mortality are expected to be 3%, ⩽10%, and ⩽30%, respectively. Data on 10-year survival is scarce, and the panel endorses the practice of reporting these figures for surgical interventions following the intention-to-treat principle, in order to better discriminate differences in outcome between resection and transplantation not apparent at the conventional 5-year cut-off.

A recent systematic review including 90 studies, comprising a total of 17,780 patients over 15 years, identified the Milan criteria as an independent prognostic factor for outcome after liver transplantation [177]. Overall 5-year survival of patients within the Milan criteria (65–78%) was similar compared with non-HCC indications according to European (ELTR) and American registries (OPTN) (65–87%) [[177], [216], [217]]. ELTR reports 10-year survival rates of around 50% in more than 12,000 cases performed [216]. As a consequence of their success, the Milan criteria have been integrated in the BCLC staging system [[148], [149]] and in the UNOS pre-transplant staging for organ allocation in the US [218], and remain the benchmark for any other prognostic criteria proposed for expanding the indication to liver transplantation in cirrhotic patients with HCC [219].

The major drawback of liver transplantation as a treatment of HCC is the scarcity of donors. Increases in waiting time have led to 20% of transplant candidates dropping out of the lists before receiving the procedure, thus jeopardizing the outcome if analyzed according to intention-to-treat [[161], [220]]. Four concepts have been addressed by the panel in the context of transplantation for patients with HCC: (1) priority and delisting policies; (2) neoadjuvant treatments in the waiting list; (3) extension of criteria and downstaging for transplantation; and (4) living donor liver transplantation. The recently reported International Consensus Conference on Liver Transplantation has been instrumental in complementing the current guidelines [219].

Priority and delisting policies

UNOS developed a priority system to manage waiting lists for transplantation based on the MELD score [218], which was originally generated to predict 3-month survival in patients with End-Stage Liver Disease [221]. Since MELD is unable to predict the drop-out rate of patients with HCC, several priority scores have been assigned to these patients ranging from 24 (single <2 cm) and 29 points (single 2–5 cm or 3 nodules each <3 cm) in early proposals to none and 22 points respectively in the current era. The main difficulty for establishing priority policies is to define the at-risk patients for drop-out, which in some studies are identified as those patients with multinodular tumors, neoadjuvant treatment failures or those with baseline serum AFP levels >200 ng/ml or steady increase of >15 ng/ml/month [140]. At the opposite end of the spectrum, some patients with UNOS-T1 tumor (single <2 cm) may benefit from alternative non-transplant treatments and avoid futile transplantation, at least until recurrence occurs [222].

Strategies advocating “salvage transplantation” approaches in low-risk populations should be investigated in prospective studies focused on intention-to-treat analysis and survival benefit, as they also depend on waiting time and local scenarios of donor availability. Similarly, patients undergoing resection with pathological high risk of recurrence have been proposed to be enlisted for liver transplantation [223]. Since waiting times vary significantly worldwide, it is recommended that policymakers modulate priority policies along with these variables.

There is even less information available about policies of delisting. The current panel recommends putting on hold those patients whose HCC progressed beyond Milan while on the waiting list and explore neo-adjuvant therapies for them. The panel recommends delisting those patients developing macrovascular invasion or extrahepatic spread.

Neo-adjuvant treatments in the waiting list

Adjuvant therapies for patients within the Milan criteria while on the waiting list are used in most centers to prevent tumor progression. Robust data from RCT are lacking and thus, the potential benefits advocated for local ablation or chemoembolization are derived from observational studies and cost–effectiveness analyses. The main studies assessing neo-adjuvant treatments are case series, case–control studies and cohort studies showing that RFA achieves the higher rates of complete necrosis (12–55%) [[224], [225]] compared with TACE (22–29%) [[226], [227], [228]].

The impact of these treatments on drop-out rate, recurrence and survival is only estimated from non-randomized studies. From initial studies reporting drop-out rates, an actuarial probability of 15–30% at 1 year was established [[161], [220]]. Among the case series and cohort studies reported, some investigations suggest a favorable impact of treatment in decreasing the dropout rate to levels ranging from 0% to 25% [[222], [224]]. Similarly, since treatments on the waiting list have been studied in an uncontrolled fashion, their effects on survival after LT are difficult to assess. Since the publication of the seminal study [226], case–control studies including index treated cases and matched controls indicate similar survival rates as untreated individuals [[227], [228]]. Markov-based cost–effectiveness analysis, on the contrary, pointed to benefits for neo-adjuvant treatments when waiting times exceed 6 months [229]. The use of sorafenib for the treatment of UNOS-T2 patients in the waiting list is not recommended according to the small pilot studies and cost–effectiveness studies published so far [[230], [231]]. The real effect of loco-regional or molecular therapies on patient outcomes and on global gains of life expectancy from a societal perspective is uncertain. Therefore, considering the strength of evidence available, it is recommended to treat patients waiting for transplant with local ablation, and as a second choice with chemoembolization when waiting times are estimated to exceed 6 months.

Extension of indications and downstaging for liver transplantation

Analysis of the expansion of criteria beyond Milan and downstaging to Milan has been extensively explored. In summary, the main concept is that to establish a new policy allowing expansion of criteria for transplantation, it is essential to develop robust data for the specific category of patients included in the proposed expansion. Novel criteria might have a major impact on all transplant programs and the data needed to support any change should be impeccable. In addition, the impact of the expansion on the non-HCC patients waiting for liver transplantation should be taken into account.

The current understanding is that expansion to UCSF criteria (single nodule ⩽6.5 cm or 2–3 nodules ⩽4.5 cm and total tumor diameter ⩽8 cm) – which involves around 5–10% of all enlisted patients [[220], [232]] – has already been challenged from the pathological point of view by the up-to-seven criteria (i.e. those HCCs having the number 7 as the sum of the size of the largest tumor and the number of tumors) [233] This pathology-based proposal has been recently validated in an independent series [234]. The major concerns about the expansion proposals are the lack of specific data on overall survival and drop-out rate on the waiting list for the patients outside the current criteria but fulfiling the expanded criteria. Other recent studies challenging the Milan criteria have proposed different algorithms to optimize patient selection. Nonetheless, 5-year outcome prediction could vary from 70% to 40% according to the presence of microvascular invasion. Thus, preoperative markers of vascular invasion would be required prior to adopting these criteria. In a meta-analysis to evaluate tumor size and nodules, a cut-off of sum of diameters above 10 cm was considered to increase fourfold the risk of death [235], while a combination of tumor volume and AFP levels was considered the best strategy in other studies [[140], [141]]. Molecular markers, such as allelic imbalance reflecting chromosomal instability, have also been shown to predict recurrence after transplantation [236]. Considering the strength of the evidence, it is recommended not to allow the extension of the criteria for transplant eligibility, except in the context of research protocols.

Regarding downstaging, there is not a single RCT, large case–control study or large well-designed cohort study available on patients treated consistently and properly followed. Small prospective studies suggest that downstaging to Milan criteria from patients with liver-only disease treated by radiofrequency or chemoembolization achieves 5-year survival outcomes similar to those within Milan [[237], [238]]. It is unclear whether downstaging therapies yield measurable anti-cancer effects or only provide a time frame in which to evaluate the natural history of HCC, with the ultimate risk of transforming pre-transplant drop-outs into post-transplant recurrences [[239], [240]]. There is no clear upper limit for eligibility of downstaging [240].

Considering the current data, downstaging of patients beyond Milan criteria cannot be adopted as a tool to refine patient selection and further research is required. This research should be based on the principle that 5-year survival outcomes of patients undergoing transplantation after successful downstaging should be similar to those of patients transplanted following Milan criteria [219]. The panel considers, though, that a special policy should be adopted for patients already on the waiting list for liver transplantation with tumors progressing beyond Milan and liver-only disease. In this special circumstance, as stated above, it is recommended to place the candidate on hold until downstaging by local ablation or chemoembolization is achieved and maintained for a period of at least 3 months.

Living donor liver transplantation

Living donor liver transplantation (LDLT) using the right hepatic lobe of a healthy donor has emerged as an alternative to deceased liver transplantation [[241], [242]]. In 2000, there was great enthusiasm for LDLT, and it was estimated that it would represent a significant proportion of the patients transplanted with HCC [243]. Unfortunately, the associated risks of death (estimated in 0.3%) and life-threatening complications (∼2%) for the healthy donor have diminished the interest of the transplant community [[244], [245], [246]]. Currently, LDLT comprises less than 5% of adult liver transplants, significantly less than in kidney transplantation where living donors represent 40% of all cases performed [246]. The risks and benefits of LDLT should take into account both donor and recipient, a concept known as double equipoise [[219], [247], [248]]. Due to the complexity of the procedure, LDLT must be restricted to centers of excellence in hepatic surgery and transplantation.

Outcome results with LDLT compared with deceased LT have been controversial. Although some studies suggested that LDLT was associated with higher risk of recurrence, these data have not been confirmed [[249], [250]]. Cost–effectiveness studies suggested that LDLT can be offered to patients with HCC if the waiting list exceeds 7 months [248], a policy adopted by the panel. Some authors recommend a period of observation prior transplant of 3 months, in order to avoid transplanting potentially aggressive tumors, a proposition that needs to be confirmed in further investigations [[250], [251]]. LDLT has been proposed as an ideal setting to explore extended indications for HCC, considering the lack of graft allocation and priority policies [252]. Therefore, the panel does not recommend this procedure for any extended indication, except in the context of research studies.