Management of hepatocellular carcinoma on the waiting list before liver transplantation: Time for controlled trials?


  • This article is based on a review of the treatment of hepatocellular carcinoma while patients are on a liver transplant waiting list that our group published in Journal of Hepatology in 2005, and a similar text was provided in the manual of the 7th International Meeting on Therapy Liver Diseases, Barcelona, Spain, September 12-14, 2007. Although new concepts and references have been added, the views of the group are substantially unchanged, and similarities in the publications are unavoidable.

In the current situation of graft shortages, patients with hepatocellular carcinoma (HCC) spend a considerable time on the waiting list, facing the risks of not being transplanted because of death or contraindications developed while waiting. Management strategies have been devised to monitor or slow the progression of the tumor beyond the criteria accepted for transplantation: surveillance without treatment, transarterial chemoembolisation (TACE), radiofrequency thermoablation (RFTA), a combination of TACE and RFTA, and prioritization. This review summarizes the rationale behind each strategy and the studies concerning it. Although comparison between the different management modalities is difficult, some important information can be extracted from the literature, and some recommendations can be put forward.

Before reviewing each management strategy and the published experience summarized in Table 1, and Figure 1.1–14 a few general points need to be addressed.

Table 1. Main Series of Liver Transplantation for Hepatocellular Carcinoma Presenting Intention-to-Treat Results
AuthorStudyCriteriaPatients (transplanted)WT (d)Neoadjuvant treatmentDO criteriaDO tumorDO LF/medDO totalRecurrenceFollow-up (mo)ITT survivalComments
  • CL, contralateral lobe; DO, dropout; Downstaging+, decrease >50% in the maximum diameters of the tumors; ITT, intention to treat; LF, because of liver failure or complication of cirrhosis; MT, with Milan criteria; MS, Mount Sinai criteria (no involvement of main portal or hepatic veins, bilateral tumor <5 cm in the less affected lobe, tumor volume <75%, no diffuse infiltration); MS−, progression outside Mount Sinai criteria; NA, not available; P, prospective; PEI, percutaneous ethanol injection; R, retrospective; RFTA, radiofrequency thermal ablation; T0, no tumor found; T1, 1 nodule <2 cm; T2, 1 nodule 2-5 cm or 3 nodules <3cm; T3, 1 nodule >5 cm or up to 3 nodules with one >3 cm; T4, T4a, ≥4 nodules, any size, 4b, 4a criteria plus intrahepatic portal vein or hepatic vein involvement; TACE, transarterial chemoembolization; Tum, because of tumor; Tx, transplanted; VI/ES, vascular invasion or extrahepatic spread; WT, waiting time; UCSF, University of California, San Francisco, criteria (solitary tumor <6.5 cm, 3 or fewer nodules <4.5 cm, total tumor diameter <8 cm).

  • Statistics are presented as mean (SD) unless otherwise specified. Survival indicates the percentage of patients surviving at 1, 3, and 5 years.

  • *

    Updated series previously published by Graziadei IW, Sandmueller H, Waldenberger P, Koenigsrainer A, Nachbaur K, Jaschke W, et al. Chemoembolization followed by liver transplantation for hepatocellular carcinoma impedes tumor progression while on the waiting list and leads to excellent outcome. Liver Transpl 2003;9:557–563.

Llovet et al.1 1999 Milan50 (50)105 ± 111 1-485NoneVI/ES000?  No adjuvant treatment
 PMilan37 (29)242NoneVI/ES6223%3?Median 23, range 26-3054 (2 yr)AFP and WT predictors of dropout
Herrero et al.2 2001R 49 (47)117TACE 23, PEI 5, RFTA 3Progression?2NA5% NANA 
Roayaie et al.3 2002P>5 cm and MS80Tx:142 ± 168 DO:207 ± 306TACEMS-234?46% 55NADownstaging 33%
Hayashi et al.4 2004RT220 (12)333 ± 302TACEProgression4235% 66−/−/61No recurrences in transplant patients
Maddala et al.5 2004RT1 (2), T2 (45), T3 (7)54 (45)211 (28-1,099)TACEProgression6215%5?−/−/61% 
Yao et al.6 2003RT270 (50)186 ± 323TACE 11/PEI 7/RFTA 11UCSF18229%?1887/57/−1 death after TACE
Llovet et al.7 2003PMilan205 (172)159 ± 119NoneVI/ES20716%?3580/79/62Abstract presented AASL 2003
Yamashiki et al.8 2004RT1 (9), T2 (73), T3 (11)93104 (1-897)None/microwave 22/TACE 8Stage 4143+524% NASurvival 66%Dropout for tumor: AFP< 100 5% vs. >100 32%
Fisher et al.9 2004RT0-T1 (1-9), T2 (17), T3 (6)33 (28)278 ± 451MultimodalStage 43215% 32Survival 79%Predictors of include dropout AFP >400 and T3
Mazzaferro et al.10 2004P(40 M+)50 (50)289RFTA percutaneous/laparoscopic 29/21VI/ES000222 (after Tx)95/83/NARecurrences in partial responders gt;5 cm
Lu et al.11 2005P(42 M+)52387 (9-1,326)RFTAVI/ES3312%0Median 9 mean 15  
Martin et al.12 2006P 47253 (29-794)RFTAVI/ES55211Median 16 Short follow-up after OLT
Otto et al.13 2006RM+34 (?)250 (22-756)TACE 61?Total 7? Recurrences associated with progression
  M−62 (34)268 (129-818)  51?Total 7?  
Milloning et al.14* 2007PMilan68 (41)274 (36-1,037)TACEProgression2025 (8%)   
  UCSF33  Progression       
  Downstaging +, no VI/ES15 (10)254 Progression314 NA93/78/31Downstaging 42%, 3 noncancer deaths
GenevaPMilan+43203 (27-1,036)TACE (RF1, PEI 1)VI/ES4827%458 Dropout rate increasing recently with increasing waiting time
Figure 1.

Summary of dropout probabilities in different centers. Missing intermediate values were interpolated from the mean of the adjacent values.


Monitoring Tumor Progression

A careful surveillance protocol is mandatory to decide whether the patient needs to be removed from the waiting list. Progression to dropping off the waiting list may simply be missed because the monitoring protocol has not intervened to pick this up. Such protocols are seldom defined.

Waiting Times

In the studies, mean waiting times varied from 104 to 387 days, with wide ranges and standard deviations. This could be the result of differences in waiting due to blood groups (which is probably not a source of bias), or to fast-tracking patients with adverse tumor characteristics, rapid progression, or liver failure (which do introduce severe bias).8, 10

Dropping Off the Waiting List

Patients can drop off the waiting list for several reasons, as follows. First, the tumor may have progressed beyond commonly accepted criteria (e.g., metastatic spread). These patients would be dropped from any program. Second, the tumor may progress outside the criteria of a given program (e.g., outside Milan criteria: ≤3 nodule ≤3 cm, or 1 nodule ≤5 cm, no vascular invastion) or outside the criteria of an investigation protocol. Third, patients may develop non–tumor-related contraindications while waiting (e.g., alcohol relapse), or they may withdraw consent to the procedure. And finally, patients may die because of progression of the liver disease, or for other reasons.

Programs differ regarding the criteria that warrant exclusion from the waiting list, and these criteria are defined with varying precision in published papers. In some studies, from the United States in particular, patients were excluded once contemporary United Network for Organ Sharing (UNOS) listing criteria were no longer met (stage 4—more than 4 tumors, vessel invasion, or metastases; or stage 3—beyond Milan criteria but fewer than 4 tumors).4, 5, 8, 9 At the other end of the spectrum, unilateral portal branch invasion was not a reason for dropping off the waiting list in the study from Mount Sinai focusing on patients whose disease exceeded the Milan criteria.3 These differences have to be taken into account when interpreting data on dropping off the waiting list in the studies.

Outcome Measurement

Dropping off the waiting lists is not a meaningful event per se. Indeed, from the point of view of the community, the graft can be used for somebody else, and the worse scenario would indeed be one in which a treatment modifies the local appearance of the disease and not its metastatic potential, transforming dropouts into recurrences. Studies focusing on patients removed from the waiting list with a follow-up after transplantation that is too short to allow recurrences to manifest are therefore meaningless.

Analysis of the Different Treatment Strategies

Surveillance Without Treatment

Only one study reliably reports the outcome of observation with no additional treatment on the waiting list,1 and this was a landmark paper in evaluating and reporting the results of liver transplantation for HCC. In this study, the Barcelona Clinic Liver Cancer study group compared a group of 50 patients when the waiting time was 62 days to a group of 37 patients when the waiting time increased to 162 days. Two patients dropped off the list because of liver decompensation and 8 patients because of tumor progression. The recurrence rate after transplantation was 3 of 83 patients, with a relatively short follow-up after transplantation of 26 months. Also, in the BCLC study, it is not specified in how many of the patients the tumors progressed without being a cause of dropping off the waiting list, and the details of the surveillance protocol are not provided.

No other center has published (or implemented?) a strategy of surveillance alone for HCC patients on the waiting list, and the relatively high dropout rate of the BCLC study has not been observed in other studies. The fact that surveillance without treatment is not used in major units may also reflect the fact that both patients and caregivers are uncomfortable watching tumors progressing to the point where only palliative treatments are indicated, and then facing the questions why these treatments were started only after a delay. It is unlikely, therefore, that a prospective study comparing surveillance alone to no treatment will be feasible in the future.

Fast Tracking

The simplest way to avoid patients being dropped from the waiting list is to transplant patients without delay, obviating the need for treatment while on the waiting list. This, however, may be a double-edged sword because time on the waiting list is a privileged period to observe the biology tumor and to weed out patients with aggressive cancers that will progress to the stage of contraindications while waiting. Reducing the waiting time in these cases may transform dropouts into recurrences. The concern is pertinent to living donor liver transplantation (LDLT), in which the results in term of recurrences have been assumed—but not proven—to be equal to liver transplantation from deceased donors.16–18 Indeed, preliminary results from Northwestern University in Chicago reported a higher incidence in patients fast-tracked either by recent UNOS–Model for End-Stage Liver Disease (MELD) rules, or by the use of split–living donor grafts, compared with patients from a previous period with longer waiting times (T1-T2, 5 of 32 vs. 0 of 25; T3-T4, 4 of 9 vs. 1 of 8, respectively).19 The difference for the whole series was statistically significant.

In patients with small tumors, there is probably no reason to think that fast-tracking increases recurrences. The experience of units with short waiting times because of a deliberate policy of transplanting HCC within 3 months,20 or of landmark series when waiting times were shorter1, 21, 22 suggest that in patients whose disease is within Milan criteria, early transplantation is associated with recurrences at a rate of approximately 10%, corresponding to the proportion of tumors that have an aggressive histology and behavior despite the small size.20

For patients with disease at the margin of or outside the Milan criteria, however, the proportion of patients with unfavorable tumor behavior is higher, and stabilization of the tumors by treatment may allow adding observation time among the prognostic determinants. In marginal patients, fast tracking vs. a more conservative “stabilize and wait” approach reflects the choice between the willingness to maximize the chances of the recipient, or to avoid a futile procedure. Although no general recommendations can be made, cost-effectiveness models (and personal experience on how failures in the recipient are taken by living donors) would argue against fast tracking, even in LDLT.16 The optimum waiting time in these patients is still undefined, but 6 months appears to be a reasonable estimate.23

An additional concern about fast tracking appeared after the UNOS allocation policy assigned MELD points to HCC patients to compensate for the preserved liver function. In the early phase of the adjustment, 31% of patients transplanted with the diagnosis of early HCC (nodule <1.9 cm) were found to have no tumor in the explanted liver,24 and these patients are no longer given priority above the calculated MELD score. Strict adherence to radiological diagnostic criteria for HCC25 should be observed when listing patients, especially if they are likely to receive a graft before further assessment.26 The current UNOS allocation system attributes 24 points to stage 2 patients (1 nodule 2-5 cm or 2-3 nodules all <3 cm), equivalent to a chance of dropping off the waiting list of 15% at 3 months, and the addition of 2 points at 3-month intervals, corresponding to a 10% chance of dropping off the waiting list, regardless of the possible effect of pretransplantation treatment, and potentially discouraging controlled trials.27


TACE before transplantation has been used by several groups.2–6, 8, 12–14, 28, 29 The popularity of the procedure can be attributed to its relatively low invasiveness and to extrapolation to pretransplantation patients of the antitumor effect of TACE when used in the palliative setting.30–33 The use of TACE before transplantation has been extensively debated, and when posttransplantation survival figures are examined, there seems to be no benefit to treated patients.28 A recent review analyzing the possible role of TACE before transplantation stated that the evidence is too scanty to recommend its use, even if TACE appears to be safe.34 This point of view should be tempered.35 To systematically analyze the role of TACE before orthotopic liver transplantation, 3 aims of the procedure have to be taken into account: (1) to minimize dropping off the waiting list; (2) to improve the results after transplantation; and (3) to downstage HCC outside the criteria so that patients who are transplanted have an acceptable outcome.

As for decreasing the number of patients dropped from the waiting list, the first series with a systematic approach of TACE from the Mayo Clinic reported on 54 patients treated with TACE for a median of 211 days before transplantation.5 The criteria for monitoring and exclusion of patients on the waiting list were well defined. The dropout rate was 14% and the recurrences after transplantation were 10%. Follow-up after transplantation was probably short, but the time was unspecified. These figures compare favorably with the BCLC experience of surveillance alone. More recently, Millonig et al.14 updated a series from Innsbruck with a total of 116 patients (68 within Milan criteria) treated with TACE and a median waiting time of 274 days (range, 36-1,037 days). The dropout rate in patients within the Milan criteria was 2%, and recurrences after transplantation were 8, after what was probably a longer median follow-up than the Mayo Clinic series (follow-up time was unspecified).

To overcome the limitation of a short follow-up, we analyzed our own experience with TACE on the 43 patients within the Milan criteria treated between 2000 and 2004 for whom data on outcome with a follow-up >2 years are available. The tumor-related dropout rate was 10%, and recurrences were 10% after a median waiting time of 203 days (range, 27-1,036 days), showing that in patients within the Milan criteria, TACE seems to limit the pretransplantation dropout rate (compared with the BCLC, for instance) without transforming dropouts into recurrences (Table 1).

The second aim of TACE before transplantation could be to improve the overall results of transplantation in patients within the Milan criteria in terms of recurrences and of intention-to-treat (ITT) survival. Concerning recurrences, these events are so rare among patients within Milan criteria, whether treated or not, that it is unlikely that a prospective study will be considered meaningful and will be performed.35 Concerning the overall ITT results of transplantation, however, theoretical concerns on the safety of TACE before and after transplantation have to be addressed, starting from the waiting list and a risk of deterioration of liver function associated to the procedure. Deterioration of liver function was not reported in either the Mayo or Innsbruck series, although 2 patients in the Innsbruck series had hepatic abscesses, one of which required surgical drainage. In our center, the systematic use of TACE in the pretransplantation population since 1996 (with the only contraindications to a selective procedure being renal failure or hypovascular lesions) means that in a total of 84 patients listed, there has been one possible treatment-related death (intracerebral bleeding in a patient with a low platelet count), and 2 patients experienced decompensated cirrhosis from Child-Pugh class B to class C, in one case during the course of postprocedure variceal bleeding. Posttransplantation complications, and in particular arterial complications, do not seem to be more frequent, as deduced from more recent studies, and from historical series and series of patients submitted to TACE.5, 14, 21, 22, 28 Chemical pneumonia was reported only by one author.29

Overall, however, although the evidence in favor of systematic use of TACE before transplantation in patients within MC is still indirect, both the ITT and the dropout figures compare favorably to the results reported for surveillance alone.

The third possible aim of TACE is to select, among patients outside the criteria for transplantation, individuals who have a more benign form of the disease than expected by the number and size of tumor nodules. The possibility that such downstaging by TACE was associated with a better outcome than anticipated in apparently poor-risk cases was suggested by the fact that only 2 recurrences were observed among 19 patients, with tumors >3 cm having decreased the sum of 2 diameters by >50%.28 The concept that response to chemotherapy is associated with a more favorable histological (and possibly biological) behavior of the tumor, in the sense of vascular invasion and distant spread—indeed not applicable to a physical treatment, such as percutaneous ethanol injection (PEI) or RFTA—is tempting,13 but the picture is probably more complex than that. It is likely that observation time also plays a role, and that patients who respond well to TACE and who do not develop distant metastases or vascular invasion during the treatment period can be assumed to have a more favorable tumor biology, as discussed in the section on fast tracking. Time was indeed included in the expanded criteria suggested by the BCLC, where patients brought back into the criteria and stable for 6 months can still be considered for transplantation.23 In fact, downstaging of patients outside the criteria was associated with a poor survival among 15 patients in a subgroup of the TACE study from Innsbruck.14 The high non–tumor-related mortality among the 10 transplanted patients, however (3 of 5 deaths), weakens this argument, and further studies are needed.

Recently, Yao et al.15 prospectively analyzed a group of patients downstaged by TACE. Thirty patients were submitted to TACE and were accepted for transplantation if local eligibility criteria (University of California, San Francisco [UCSF]) were met. Sixteen patients were transplanted, with 81% experiencing 2-year survival and no recurrences. The authors concluded that the strategy was effective. The study has a prospective design, but the follow-up time after transplantation was short (median, 16 months), thus precluding robust conclusions.


There are few reports of PEI as a treatment before transplantation, generally in series where patients were treated with a variety of modalities, and none systematically evaluate the potential benefit.36, 37 From a theoretical point of view, the effectiveness (necrosis of 70-80% of nodules <3 cm,38 noninvasiveness, and low cost of the procedure would argue in its favor. Furthermore, the fine needle minimizes the risk of track seeding (generally reported at 0.6%39), and the reports available do not mention local recurrences, a concern of several transplant teams.40 Even if PEI cannot be recommended on the basis of direct experiences, from theoretical models, it seems a very reasonable treatment option in patients waiting longer than 6 months.41 Pragmatically, however, the technique is being replaced by RFTA for reasons of convenience (a single session, vs. several sessions required for PEI) and to some extent of fashion. It is therefore unlikely that PEI will be used in controlled studies in the future.


The rapid diffusion of the RFTA procedure has overtaken concerns of safety and efficacy. From the point of view of subsequent liver transplantation, the thicker needle and the reports of seeding, especially when superficial lesions are treated, would demand caution.42, 43 However, the fear of needle-track recurrences has been lessened by several reports,44, 45 and a recent well-designed randomized controlled trial proved percutaneous RFTA to be more effective than PEI in patients with tumors <3 cm,46 in which none of 52 patients had needle track seeding after a median follow-up of 23 months. Although the real incidence of seeding is still difficult to appreciate and may require several years to manifest,47 the risk can be estimated in the region of 1-2% from the caseload of experts (L. De Baere, personal communication, September 2004).

RFTA has been used in patients as a bridge to transplantation in a small uncontrolled study,48 and the Milan Istituto Nazionale Tumori team has recently reported a prospective experience of 60 nodules treated with percutaneous and laparoscopic RFTA before liver transplantation.10 The reader is referred to this study for a balanced approach on the use of RFTA before transplantation. In summary, 3 main points can be retained. The first is that the procedure was safe in terms of liver function, seeding, and recurrences. There was a modest progression in the Child-Pugh class scores (7.4 to 7.5), and only one case in which transplantation had to be anticipated because of worsening liver function. There was no seeding after a follow-up of 22 months after transplantation. Recurrences were detected only in 2 patients, both with tumors >5 cm. The second is that in the study from Milan, the frequency by which viable tumor cells were found on microscopic examination of the treated nodules increased with the interval from treatment to transplantation, suggesting that that the radiological techniques evaluating the success of the procedure overestimate its efficacy, and that nests of viable tumor cells persist and proliferate to the point of histological detectability as time goes on (an important message concerning RFTA—and possibly all percutaneous treatments—in general). Third, there were no dropouts with a mean waiting time of 9.5 months. Whether this result is due to the favorable characteristics of the cohort (mean size of disease, 27.5 mm, and 41 of 50 patients with solitary nodules, 9 of 50 with Child-Pugh class C) or to treatment is still debatable, but the study identifies a group of patients in which the dropout rate is—or can be kept—very low and can be compared with the BCLC.

More recently, a study from Los Angeles reported 52 patients (43 within the Milan criteria) waiting for 387 days (range, 9-1,326 days).11 The total number of patients who were dropped from the waiting list was 6; 3 of these were tumor related. Unfortunately, only a limited follow-up was available to interpret the results in terms of recurrences. On the basis of these studies, can RFTA be recommended for all patients in the waiting list? In our opinion, the problems of seeding by the percutaneous procedure (48% of the patients who met the Milan criteria were treated by the laparoscopic approach) and the efficacy of RFTA treatment in larger tumors need to be confirmed by other studies before the procedure can be accepted as standard of care. At present, the policy in our unit is to reserve RFTA only for patients included in a prospective randomized protocol designed specifically to assess the risks and benefits of the procedure when added to TACE, or to patients with hypovascular tumors likely to wait for >6 months (Table 2).

Table 2. Main Policies Used in Geneva for Patients on the Waiting List for Hepatocellular Carcinoma
PolicyMain policyProgrammed exceptions
  1. Abbreviations: TACE, transarterial chemoembolization; AFP, alfa-fetoprotein; RFTA, radiofrequency thermal ablation; CT, computed tomographic scan; HCC, hepatocellular carcinoma; LDLT, living donor liver transplantation.

  2. Minimal requirements for controlled study include stratification for high-low AFP; Child-Pugh class A and B patients; monitoring; and defined delisting criteria.

Inclusion criteriaMilanOutside Milan considered if <8 cm and: a) biopsy for tumors > 5 cm: excludes poor differentiation b) Downstaging by TACE and stable 6 months c) Complete response to TACE (no vascular uptake and normalized AFP) and marginal graft available
Treatment while waitingTACE or Study SASL 19 (TACE vs. TACE + RFTA)17RFTA considered if: Hepatorenal syndrome (high risk of) Hypovascular tumor likely to wait >6 months
MonitoringVisit every 3 months, ultrasound, and AFP 6 monthly CT with chest and bone scan ltrasound at admission for transplant, replaced by CT if last CT >3 monthsCT every 3 months in SASL 19 study
Exclusion criteriaVascular invasion and extrahepatic spreadTumor progression and/or increased AFP while receiving treatment prompts magnetic resonance imaging, bone scan, and multidisciplinary rediscussion
AllocationNo priority to HCC patients or patients with tumor progression 
LDLTSame criteria as for grafts from deceased donors 

Liver Resection

The use of resection as a tumor control strategy in Child-Pugh class A patients on the waiting list before transplantation has only been investigated in a Markov model,41 and it was found to be cost-effective for waiting times >1 year. This is consistent with the fact that liver resection is a demanding procedure in itself, and it undoubtedly complicates subsequent liver transplantation. Probably no group today would embark on resection merely as neoadjuvant treatment in patients who will be transplanted anyway.49 Rather, transplantation may be used as a salvage strategy in patients with recurrence after an intended curative resection. The decision analysis model by our group concluded that that primary resection and salvage transplantation was a reasonable approach with a similar ITT survival to primary resection, provided that at least 60% of recurrences could be detected at a stage when the patient was still transplantable, and that the outcome of transplantation is not more than 10% inferior to primary liver transplantation, resulting in a savings of approximately 50% of the grafts.50 Transplantability figures similar to ours were reported by Poon et al.,51 and a favorable retrospective experience on liver transplantation after primary resection was reported by the Beaujon team.52 On the other hand, the Villejuif group reported a poor outcome with salvage transplantation, a result possibly affected by the high mortality, high recurrence rate, and low transplantability of earlier patients.53

The Barcelona group contributed to the debate by a prospective study in which the subsequent management after primary resection was guided by the histology of the tumor.54 Patients with unfavorable prognostic factors such as satellite nodules, microscopic vascular invasion, poorly differentiated lesions, or disease-positive margins were scheduled for transplantation, whereas patients with a low pathological risk profile were followed up for recurrence and transplanted if needed. In total, 6 grafts were used for 17 patients, with a disease-free survival of 83% after a median follow-up of 55 months, a further evolution of the salvage approach. The feasibility and caveats of this approach were investigated by a study from another team in Barcelona, which confirmed that the approach is safe and feasible, and saves a considerable amount of grafts,55 as underlined by a commentary by our group.56


AFP, alfa-fetoprotein; BCLC, Barcelona Clinic Liver Cancer; CT, computed tomography; HCC, hepatocellular carcinoma; ITT, intention to treat; LDLT, living donor liver transplantation; MELD, Model for End-Stage Liver Disease; PEI, percutaneous ethanol injection; UCSF, University of California, San Francisco; RFTA, radiofrequency thermal ablation; TACE, transarterial chemoembolization; UNOS, United Network for Organ Sharing.


Analysis of the factors associated with progression or dropout is hampered by the fact that the results after transplantation are insufficiently taken into account. Possible factors related to dropping off the waiting list were investigated in 5 studies.5–9 The Mayo Clinic study did not identify any prognostic factors, but the numbers were too small for statistical significance.5 Larger tumor size and multicentricity were correlated to dropping off the waiting list in the UCSF series,6 and an alfa-fetoprotein (AFP) level of >100 was associated with an increased risk of dropping off the waiting list in the study from Miami8 and in the study from Richmond.9 It is difficult to draw conclusions from the latter 3 studies because there was no uniform treatment protocol in the cohort.

Recently the BCLC reported data on a cohort of 179 untreated patients who could be stratified as low or high risk of dropping off the waiting list according to tumor progression (50% increase in size during the previous 6 months) and AFP (cutoff of 200 ng/mL).7 A total of 153 low-risk patients had a stable 3% chance of dropping off the waiting list after 6 and 12 months (similar to the treated patients of the Innsbruck or Milan Istituto Nazionale Tumori cohorts), whereas 46 high-risk patients had a steep increase in being dropped from the waiting list, from 23% at 6 months to 52% at 12 months. Although further data are awaited, from analyzing the potential contribution of other markers, such as patient's age, cause and stage of the liver disease, and initial tumor size, it appears that an increased AFP and tumor progression despite treatment identify a high-risk category for dropout and recurrence.

Follow-up Protocol

In most series where the information is detailed, computed tomography (CT) of the liver every 3 months was used, accompanied by chest and technetium bone scans5, 10, 14 and repeated measurements of AFP. No information can be extracted on the efficacy or cost-effectiveness of more or less stringent monitoring. In Geneva, we have used triple-acquisition CT of the liver with nonenhanced acquisitions on the chest every 3 months and a technetium bone scan every 6 months. A raising AFP on the waiting list—with the disease apparently stable on protocol CT—prompts further search for vascular invasion and metastases with a liver magnetic resonance imaging and a repeated bone scan. With such a protocol, distant metastases and vascular invasion were detected in 4 cases, and the bone scan never intervened as a positive finding, but the number of observations is too small to change the monitoring policy.


At the time of this writing, the HCC transplant community is still at the stage of evidence gathering rather than of evidence-based practice.57 The main questions—which treatment to provide while waiting, which criteria should be used to remove patients from the waiting list, how to monitor patients during the waiting time—have come into focus but cannot be answered unequivocally. Recent experiences suggest that fast tracking patients at the edge of or beyond Milan criteria may be associated with higher probabilities of recurrence, and prioritizing HCC patients is probably a poor option in a concern for the optimal use of resources. TACE, RFTA, and TACE with RFTA have arguments in their favor, and the choice can still be tailored to local realities and expertise. Surveillance alone (i.e., no treatment) is not favored because TACE with or without RFTA at least do not harm; they bridge the majority of patients to transplantation; and they allow sufficient time to select poor risks during the observation time. Furthermore, in case of a given patient dropping off the waiting list, TACE and RFTA represent the most effective and timely palliative treatment, and this represents an argument facing a potential dropout.

On the basis of current knowledge, therefore, can a particular trial be recommended? Because the consensus in practice is now to treat patients, the most relevant question is not whether to treat or not to treat, but which treatment is most appropriate, and in particular which criteria should be used for removing a patient from the waiting list. The most topical question in our center was whether RFTA adds benefit to TACE alone, which we consider, for the reasons mentioned above, to be the standard treatment. A randomized trial comparing TACE alone vs. TACE with RFTA is underway as a Swiss multicenter study and is now open to all programs that want to include patients. Details of the protocol are available on the Web page of the Swiss Association for the Study of the Liver, and applications for participation are welcome. Knowledge will increase as the liver transplantation community acquires new data, and this will be possible only thanks to an approach that is structured and that allows sufficient follow-up to include recurrences after transplantation in the equation.58


The Geneva Liver Cancer Study Group acknowledges the contribution of Markus Heim, University of Basel, in initiating and promoting the Swiss Association for the Study of the Liver 19 study comparing transarterial chemoembolization to transarterial chemoembolization associated with radiofrequency thermal ablation while on the waiting list.

Members of the Geneva Liver Cancer Study group are as follows:

  1. Surgery: Philippe Morel, Gilles Mentha, Pietro Majno, Léo Buhler, Thierry Berney, Christian Toso.

  2. Hepatology: Antoine Hadengue, Emiliano Giostra, Laurent Spahr, Isabelle Morard.

  3. Radiology: Christoph Becker, Pierre-Alain Schneider, Sylvain Terraz.

  4. Pathology: Laura Rubbia-Brandt

  5. Clinical psychiatry: Pascale Gelez

  6. Decision analysis: François Sarasin.