Impact of pretransplantation transarterial chemoembolization on survival and recurrence after liver transplantation for hepatocellular carcinoma



The actual impact of transarterial chemoembolization before liver transplantation (LT) for hepatocellular carcinoma (HCC) on patient survival and HCC recurrence is not known. Between 1985 and 1998, 479 patients with HCC in 14 French centers were evaluated for LT. Among these 479 patients, this case-control study included 100 patients who received transarterial chemoembolization before LT (TACE group) and 100 control patients who did not receive chemoembolization (no-TACE group). Patients and controls were matched for the pre-LT tumor characteristics, the period of transplantation, the time spent on the waiting list, and pre- and posttransplantation treatments. Kaplan-Meier estimates were calculated 5 years after LT and were compared with the log-rank test. The mean waiting time before LT was 4.2 ± 3.2 months in the TACE group and 4.3 ± 4.4 months in the no-TACE group. The median number of TACE procedures was 1 (range: 1-12). Demographic data, median alpha-fetoprotein level (21.6 ng/mL and 22.0 ng/mL, respectively), and pre- and post-LT morphologic characteristics of the tumors did not differ in the TACE and no-TACE groups. Overall 5-year survival was 59.4% with TACE and 59.3% without TACE (ns). Survival rates did not differ significantly between the two groups with respect to the time on the waiting list, the tumor diameter, or the type of TACE (selective or nonselective). In the TACE group, 30 patients had tumor necrosis ≥80% on the liver explant with a 5-year survival rate of 63.2%, compared with 54.2% among their matched controls (P = 0.9). In conclusion, with a mean waiting period of 4.2 months and 1 TACE procedure, pre-LT TACE does not influence post-LT overall survival and disease-free survival. (Liver Transpl 2005;11:767–775.)

Hepatocellular carcinoma (HCC) is one of the most common cancers, and its incidence is rising worldwide.1–3 HCC occurs usually in cirrhotic livers and less than 30% of cases qualify for resection.4 Liver transplantation (LT) is the only potentially curative treatment for patients with HCC and cirrhosis. The 5-year survival rate of patients after LT ranges from 33% to 74%, depending on the tumor stage. When LT for HCC is restricted to patients with tumors meeting the widely used Milan criteria,5 the 5-year overall survival rate is 70% and the recurrence rate is below 15%.5–7

Lengthy waiting periods due to a shortage of donors can allow the tumor to progress to stages that may contraindicate LT. As reported by Yao et al.,8 a 6-month waiting period is associated with a 7.2% cumulative dropout probability, and this rate rises to 37.8% and 55.1% at 12 and 18 months, respectively. Such patients are often offered treatments aimed to control tumor growth pending LT in order to reduce the dropout rate and to improve survival after transplantation. Transarterial chemoembolization (TACE) is one of the most widely used treatments of this type.

The efficacy of TACE in patients with HCC who are not candidates for LT is controversial,9–18 despite the positive results of some recent randomized controlled trials and a meta-analysis.19

Likewise, there is no proof that pretransplantation TACE improves long-term survival after LT.20–28

Univariate analysis of a large multicenter database of patients who underwent transplantation for HCC suggested a beneficial effect of perioperative treatment including TACE29 on recurrence and disease-free survival. We thus decided to evaluate the effect of TACE on overall 5-year survival, disease-free survival, and risk of recurrence after LT.


LT, liver transplantation; HCC, hepatocellular carcinoma; TACE, transarterial chemoembolization.

Patients and Methods

Study Population

We reviewed all patients listed for LT for HCC between 1985 and 1998 in 14 French liver transplant centers (Besançon-Jean Minjoz, Bordeaux-Pellegrin, Clichy-Beaujon, Créteil-Henri Mondor [investigating center], Lille, Lyon-Croix Rousse, Lyon-Edouard Herriot, Marseille-La Conception, Montpellier–Saint-Eloi, Nice-L'Archet 2, Paris-Cochin, Paris-Saint Antoine, Rennes-Ponchaillou, and Strasbourg-Hautefeuille).

HCC was histologically proven on the liver explants in all the patients. Patients with incidental HCC (HCC diagnosed only on the explanted liver) or fibrolamellar carcinoma were not included.

On the basis of these criteria, 479 patients were identified in the database. Twelve patients (2.5%) were removed from the waiting list because of tumor progression and/or death. Among the remaining 467 patients who underwent transplantation, 185 patients (39.6%) had received TACE, once listed. According to four criteria, i.e., the period of transplantation, the time spent on the waiting list, the pre-LT tumor characteristics according to the modified TNM stage for LT,30 and the pre- and posttransplantation types of treatments (surgery, percutaneous ethanol injection, post-LT chemotherapy, and/or tamoxifen), we matched 100 patients who had received TACE after listing and before LT (TACE group) with 100 patients who had not received TACE before LT (no-TACE group). For 85 patients who had been treated with TACE, matched control patients could not be identified; these patients were therefore not considered suitable for the study.


The charts of the 200 patients were individually reviewed and the following data were collected.

Pretransplantation Data

Demographic data, the etiology and severity of liver disease, liver biochemical tests, alpha-fetoprotein level and Karnofsky index31 were recorded. Patients with cirrhosis were classified according to Child-Turcotte-Pugh classification. Morphological features of HCC and preoperative staging were derived from abdominal imaging studies: ultrasonography, computed tomography, and/or magnetic resonance imaging when available. The number, size and location of the tumors, and vascular involvement, were determined from this preoperative evaluation. According to the number and size of tumors prior to LT, patients were classified as proposed by Mazzaferro et al.,5 and to the modified TNM staging system for LT (Table 1).30

Table 1. Modified TNM Staging for LT30
T0Tumor not found
T11 nodule ≤1.9 cm
T21 nodule 2–5 cm; 2 or 3 nodules, all <3 cm
T31 nodule >5.0 cm; 2 or 3 nodules, at least 1 >3.0 cm
T4a4 or more nodules, any size
T4bT2, T3, T4a plus gross intrahepatic portal or hepatic vein involvement as indicated by US, CT or MRI
N1Involvement of regional (porta hepatitis) nodes
M1Metastatic disease, including extrahepatic portal or hepatic vein involvement
Stage IT1
Stage IIT2
Stage IIIT3
Stage IVA1T4a
Stage IVA2T4b
Stage IVBAny N1, any M1

Transarterial Chemoembolization

Conventional mesenteric arteriography was performed first, and the celiac artery was catheterized. After assessing the hepatic vascular anatomy, a mixture of 10 mL of lipiodol (Lipiodol Ultrafluide; Guerbet, Aulnay-sous-Bois, France) and 1 mg/kg doxorubicin (Adriamycin; Pharmacia-UpJohn, Saint-Quentin-en-Yvelines, France), 1 mg/kg cisplatin (Cisplatyl; Lilly, St. Cloud, France) or 1 mg/kg epirubicin (Farmorubicin; Pharmacia-UpJohn, Saint-Quentin-en-Yvelines, France) was given before mechanical obstruction.

Posttransplantation Management and Follow-Up

The following data were collected:

  • Surgical procedure: technique (piggy-back vs. conventional), duration of surgery, duration of cold ischemia, number of red cell packs, and number of fresh frozen plasma concentrates.

  • Explanted liver: pathological data on the explanted liver were obtained by reviewing the histopathological reports (number of nodules, maximum diameter of nodules, sum of nodule diameters, nodule locations, micro- or macrovascular involvement, tumor differentiation (WHO classification), nodal involvement, and metastasis.

  • Immunosuppressive treatment: immunosuppressive drugs (calcineurin inhibitors vs. antilymphocyte antibodies), and histologically proven acute rejection and its treatment (methyl-prednisone pulses, antilymphocyte antibodies).

Patients were screened for tumor recurrence by alpha-fetoprotein assay and thoracic and abdominal CT every 3 months for the first 2 years and/or when clinically indicated. Additional imaging techniques (bone scan, magnetic resonance imaging) were used if necessary.

Causes of death (postoperative death, death due to HCC recurrence, and other late causes of death), HCC recurrence, and length of follow-up from LT to death, HCC recurrence or last news, were also determined.

Statistical Analysis

Baseline characteristics were expressed as means ± standard deviation or medians and ranges as appropriate. Comparisons between groups were based on Student's t test or the Mann-Whitney test for continuous variables, and the X2 test for categorical variables. Follow-up and survival times are expressed as medians (range). Kaplan-Meier estimates of overall survival, disease-free survival, and recurrence were calculated and compared with the log-rank test. Patients who died during the first three postoperative months and who were not exposed to the risk of recurrence were not included in the disease-free survival calculation. A P value less than 0.05 was considered statistically significant. All calculations were done with the BMDP package.


Baseline Characteristics of the Patients and Preoperative Staging of HCC

The baseline characteristics of the 200 patients are summarized in Tables 2 and 3. The two groups were similar regarding demographic data, underlying liver disease, pretransplantation biochemical results, and alpha-fetoprotein levels. In contrast, the Child-Turcotte-Pugh class distribution and performance status differed significantly between the two groups: more patients had Child-Turcotte-Pugh class C disease, fewer had Child-Turcotte-Pugh class A disease (P = 0.01), and more had a Karnofsky index <80% in the no-TACE group than in the TACE group (P = 0.009).

Table 2. Patient Characteristics
DemographicsTACE (n = 100)No TACE (n = 100)P
Age (years)54 ± 8.552 ± 9.0ns
Performance status   
Karnofsky index <80%3050< 0.05
Cause of hepatitis   
 Hepatitis C virus3024ns
 Hepatitis B virus2114ns
 Alcohol and B or C virus108ns
 Hepatitis B and C virus31ns
Cirrhosis/no cirrhosis92/895/5ns
Child-Turcotte-Pugh classification A/B/C63/27/1044/33/23< 0.05
 Serum albumin (g/L)34 ± 6.532 ± 7.0ns
 Prothrombin activity (%)69 ± 1461 ± 19ns
 Serum bilirubin (μmol/L)30 ± 32.841 ± 55ns
 γ-glutamyltranspeptidase (× N)2.7 ± 2.52.5 ± 1.8ns
 Alkaline phosphatase (× N)1.3 ± 0.81.5 ± 0.6ns
 α-fetoprotein (ng/mL) (range)21.6 (1.5-3420.0)22.0 (0.5-245,240.0)ns
Distribution of α-fetoprotein   
 <10 ng/mL3538ns
 10-400 ng/mL4647ns
 >400 ng/mL1915ns
Table 3. Preoperative Tumor Characteristics Based on Pre-LT Morphological Assessment
 TACE (n = 100)No TACE (n = 100)P
Median number of nodules1 (1-10)1 (1-7)ns
Mean diameter of the largest nodule (cm)3.6 ± 1.53.9 ± 3.3ns
Bilobar location2127ns
Macrovascular thrombosis58ns
Number of nodules   
 Single nodule652ns
 2 or 3 nodules3240ns
 ≥4 nodules78ns
Diameter of the largest nodule   
 ≤3 cm5456ns
 3 to 5 cm3528ns
 >5 cm1116ns
Modified TNM for LT29   
 Stage I55ns
 Stage II6161ns
 Stage III2020ns
 Stage IVA177ns
 Stage IVA277ns
Mazzaferro's criteria5   

Owing to the case-control matching, the modified LT TNM stage distribution was identical in the two groups, with stage I tumors in 5% of cases, stage II in 61%, stage III in 20%, stage IVA1 in 7%, and stage IVA2 in 7%. In addition, the median number of nodules [1, range: 1-10; mean: 1.7 ± 1.3) vs. 1 (range: 1-7; mean: 1.8 ± 1.1)], the mean diameter of the largest nodule (3.6 ± 1.5 cm vs. 3.9 ± 3.3 cm), the bilobar distribution of the nodules (21% vs. 27%), the proportion of patients with lobar portal obstruction (5% vs. 8%) and the proportion of patients meeting Mazzaferro's criteria (74% vs. 68%) (Table 3) did not differ significantly between the TACE and the no-TACE groups.

Transarterial Chemoembolization

A median of one TACE procedure (range: 1-12; mean: 1.76 ± 1.1) was performed during a mean waiting period of 4.2 ± 3.2 months (range: 0.1-18 months). The mean waiting period did not differ in the no-TACE group: 4.3 ± 4.4 months (P = 0.94).

Doxorubicin was used in 75 patients, epirubicin in 14 patients, and cisplatin in 11 patients. Embolization was always performed with gelatin foam (Spongel; Hoeschst Houde, Paris-La Defense, France). Embolization was nonselective in 13 centers (92 patients) and “as selective as possible” in the tumor vascular bed in 1 center (8 patients).

Other pretransplantation treatments for HCC were received by 28% of patients in both groups, and consisted of surgical resection in 5 patients in both groups (5%), ethanol injection in 12 patients in both groups (12%), postoperative systemic chemotherapy in 10 patients in both groups (10%), and another treatment in 1 patient in both groups (1%).

Surgical Data

The piggy-back technique was used in 61.0% of patients in the TACE group and 61.8% of patients in the no-TACE group (ns). The duration of surgery was shorter in the TACE group than in the no-TACE group (6.37 ± 2.5 vs. 7.35 ± 2.2 hours, respectively, P = 0.03). There was no difference in the number of red cell packs and fresh frozen plasma concentrates used.

Postoperative Data

The histopathological characteristics of the excised tumors are summarized in Table 4. As expected,32 preoperative staging underestimated tumor size and spread. However, from examination of explanted livers, the two groups remained very similar regarding the number, size, and distribution of nodules, macro-/microvascular invasion, and tumor differentiation. The median number of nodules was 2 (range: 1-50; mean: 4.0 ± 6.5) in the TACE group and 2 (range: 1-30; mean: 3.7 ± 4.3) in the no-TACE group (ns). The mean diameter of the largest nodule was 3.7 ± 2.2 cm and 4.4 ± 3.9 cm, respectively (ns). The nodule distribution was bilobar in 40% of patients in the TACE group and 46% in the no-TACE group (ns). Microvascular invasion was found in 12% and 9% of patients, respectively, and macrovascular tumor obstruction in 17% and 24% (both ns). The distribution of WHO tumor differentiation stages was similar in the two groups (well differentiated: 53% vs. 61%; moderately: 24% vs. 23%; poorly: 6% vs. 5%; and unknown: 17% vs. 11%, respectively, in the TACE and no-TACE groups).

Table 4. Pathological Examination of the Explanted Liver
 TACE (n = 100)No TACE (n = 100)P
Median number of nodules2 (1-50)2 (1-30)ns
Mean diameter of the largest nodule (cm)3.7 ± 2.24.4 ± 3.9ns
Bilobar location4046ns
Microvascular invasion129ns
Macrovascular invasion1724ns
Number of nodules   
 Single nodule3732ns
 2 or 3 nodules3740ns
 ≥4 nodules2628ns
Diameter of the largest nodule   
 ≤3 cm5252ns
 3 to 5 cm2828ns
 >5 cm2020ns
Tumor differentiation   
Tumor necrosis   
 Subtotal (80 to 99%)146< 0.05
  50 to 80%207 
  Less than 50%5084 

Tumor necrosis was evaluated by dividing the sum of the diameters of necrotic tissue by the sum of the nodule diameters. Necrosis was total in 16% and 3% of patients (P = 0.02), subtotal (80% to 99% necrosis) in 14% and 6% (P = 0.04), partial (50 to 80%) in 20% and 7% (P = 0.005) and minor (less than 50%) in 50% and 84% (P < 0.001), respectively, in the TACE and no-TACE groups.


Immunosuppressive therapy after orthotopic liver transplantation consisted of a dual- or triple-drug regimen including either ciclosporine A (60 patients in the TACE group and 70 patients in the no-TACE group; ns) or tacrolimus (40 and 30 patients, respectively; ns), combined with steroids and/or azathioprine. The incidence of acute steroid-treated rejection was similar in the TACE and no-TACE groups (24% vs. 23%, P = 0.84). Antithyroglobulin was used in 10 and 8 cases, respectively (P = 0.14).

Mean follow-up after LT was 49.7 ±35.8 months in the TACE group and 53.7 ±41.1 months in the no-TACE group (ns). The median follow-up after LT was 49 months (range: 0-184) and 47 months (range: 0-140) in the TACE and no-TACE group respectively (ns). HCC recurred in 13% and 23% of cases in the TACE and no-TACE groups, respectively (ns). The median time from LT to recurrence was 15.7 (range: 3.2-39.3) and 19.3 (range: 1.9-94.7) months, and the median time from recurrence to death was 6.3 (range: 0.5-39.1) and 5.0 (range: 0.2-34.4) months in the TACE and no-TACE groups, respectively (ns). Ninety-two patients died during follow-up, including 47 in the TACE group and 45 in the no-TACE group. Causes of death were similar in the two groups, and included postoperative death in 15 and 7 cases (ns), late death of causes other than HCC (sepsis, recurrent viral hepatitis, and miscellaneous) in 20 and 15 cases (ns) and HCC recurrence in 13 and 23 cases, respectively (ns). At the end of follow-up, 53 and 55 patients were alive in the TACE and no-TACE groups; 50 and 45 patients respectively were alive and tumor free (ns).

Survival Rates

Overall 5-year survival after LT (59.4% vs. 59.3%, P = 0.70, Fig. 1), overall 5-year survival of patients who survived after the third postoperative month (69.8% vs. 63.8%, P = 0.56) and 5-year disease-free survival (after exclusion of postoperative deaths; see Patients and Methods) (67.5% vs. 64.1%, P = 0.72, Fig. 2) did not differ significantly between the TACE and no-TACE groups. Disease-free survival was similar between TACE and no-TACE according to tumor stage (Fig. 3). Survival rates were also similar after adjustment for the waiting period (data not shown) and tumor diameter (data not shown). Stratification of the study population on the basis of Milan criteria showed that overall survival (Fig. 4) and disease-free survival (data not shown) were similar in the TACE and no-TACE groups, whether or not patients met Milan criteria.

Figure 1.

Overall survival of the cohort. Solid line represents TACE (n = 100). Dotted line represents no TACE (n = 100).

Figure 2.

Disease-free survival of the cohort. Solid line represents TACE (n = 85). Dotted line represents no TACE (n = 93).

Figure 3.

Disease-free survival of patients according to tumor stage. Solid line represents TACE. Dotted line represents no TACE. Stage I: n = 5; P= 0.93. Stage II: n = 61;P= 0.76. Stage III: n = 20;P= 0.97. Stage IV: n = 14;P= 0.47

Figure 4.

Overall survival of patients who meet the actual criteria of liver transplantation for HCC (Milan's criteria). Solid line represents TACE (n = 66). Dotted line represents no TACE (n = 66).

In the TACE group, 30 patients had total or subtotal tumor necrosis (≥80%) on histopathological analysis. In this subgroup of patients, overall 5-year survival was slightly higher in the TACE group (63.2%) than in controls (54.2%) but the difference was not significant (P = 0.9) (Fig. 5). Similar results were also obtained for the subgroup of 16 patients with total tumor necrosis (data not shown), as well as in the subgroup of patients with tumor down-staging or/and tumor necrosis (data not shown).

Figure 5.

Overall survival of TACE patients with tumor necrosis superior to 80% in total hepatectomy specimen matched with their controls. Solid line represents TACE (n = 30). Dotted line represents no TACE (n = 30).

Finally, neither antimitotic drug used during the TACE procedure nor the number of TACE procedures or the type of TACE procedures (selective vs. nonselective TACE) influenced overall or disease-free survival (data not shown).


The results of this study suggest that TACE performed during waiting period in candidate to LT for HCC has no impact on tumor recurrence or patient survival after LT for HCC :

  • 1Although the case and control patients were similar regarding known predictive factors for HCC recurrence, such as tumor size, the number of nodules, a bilobar nodule distribution, tumor differentiation and macrovascular invasion, the overall survival and recurrence-free survival rates did not differ between the patients who received or not received TACE.
  • 2Adjustment for other factors that might have influenced recurrence, such as the LT waiting period, Milan criteria, and the extent of tumor necrosis on the explanted liver did not change the overall results, and overall and disease-free survival rates remained similar in different patient subgroups.
  • 3Despite matching for tumor features, more patients had Child-Turcotte-Pugh class A cirrhosis and fewer had Child-Turcotte-Pugh class C cirrhosis in the TACE group than in the control group; in addition, more patients in the control group than in the TACE group had poor performance status (Karnofsky index <80%). These two features suggest that the control patients were at higher risk for postoperative death than the TACE patients. Nevertheless, postoperative and overall survival did not differ significantly in the two groups.

Some previous studies suggested a beneficial effect of TACE in patients waiting for LT21–28 but most of them were open uncontrolled studies (Table 5). One study suggested a survival benefit in a specific subgroup of patients in whom downstaging or total necrosis had been achieved by TACE.21 However, this subgroup had an improved 5-year disease-free survival rate only in comparison with patients in whom no tumor reduction or necrosis had been achieved, and not in comparison with patients who did not receive TACE. Another retrospective study with historical controls showed better survival in the TACE group, but only when TACE was combined with transcutaneous ethanol injection.22 Finally, five open uncontrolled studies23–27 also suggested a benefit of TACE.

Table 5. Published Results of Pre-LT TACE
Study typeNo. of patientsResults
  • *

    TEI: transcutaneous ethanol injection.

Historical controls
 Oldhafer et al. (1998)2021No effect
 Veltri et al. (1998)2229Effect if TEI* is associated
Majno et al. (1997)2154No significant effect compared with no TACE(even in subgroup with downstaging or necrosis)
Open studies without control group
 Graziadei et al. (2003)2341Suggested benefit
 Roayaie et al. (2002)2943Suggested benefit
 Harnois et al. (1999)2424Suggested benefit
 Venook et al. (1995)2513Suggested benefit
 Spreafico et al. (1994)2621Suggested benefit
 Morino et al. (1993)2736Suggested benefit

Our results are in keeping with those of another retrospective study with historical controls, in which 21 patients were compared to 21 historical controls and which showed no benefit of TACE.20

A recent study suggested that TACE could reduce the LT waiting list dropout rate.23 This possible benefit was not examined here, as our aim was to assess the effect of TACE on tumor recurrence and disease-free survival and as during the study period our drop out rate was only 2.5%. Our study population therefore consisted of patients who actually underwent transplantation.

The reasons why TACE did not improve survival or the recurrence rate are unclear. A trend to a lower rate of recurrences in the TACE group than in the non-TACE group (5-year post-LT recurrence rate: 13% vs. 23%, P = 0.32) was observed. This difference did not reach statistical significance but a type II error cannot be excluded. These results suggest nevertheless, that subject to the small number of events, TACE did not increase the risk of distant metastasis, a concern that some investigators have raised. A second explanation could be the small proportion of patients in whom TACE achieved major tumor necrosis since only 30% of patients had tumor necrosis greater than 80%. Such a figure is in keeping with previous TACE studies indicating a similar rate of total or subtotal tumor necrosis following TACE in candidates for LT (median: 36%, range: 24.1%-58.0%).21, 22, 26, 27 However, we found that TACE had no impact on overall or disease-free survival, regardless of the extent of necrosis on total hepatectomy specimens. Another explanation could be the number of procedures within a short waiting time (median of 1 TACE in a mean of 4.2 months). Indeed, in studies showing a benefit of TACE on tumor size and survival in patients with unresectable HCC, a mean of 3 to 4.5 courses of treatment were applied.12 We cannot therefore exclude that a higher number of TACE procedures before LT might have been beneficial.

A fourth explanation could be related to the relatively short waiting period in this study (4.2 months) that might have not been sufficient enough to demonstrate a beneficial effect of TACE. However, overall and disease-free survival rates were still identical after adjustment for this variable (<6 months, 6-12 months, or >12 months). Such a waiting time was also very similar to that currently reported in France, which is 4.5 months.33 It is also similar to that currently observed in United States, where the waiting period has been considerably shortened for candidates to LT for HCC since the implementation of the MELD scoring system.

In conclusion, this case-control study shows that, with a waiting time of 4 months, pre-LT TACE has no beneficial impact on overall or disease-free survival, in either the whole population of patients who undergo transplantation for HCC or in specific patient subsets. A role for TACE in limiting tumor growth during the waiting period, thus reducing the dropout rate, cannot yet be excluded, and this issue deserves further specific study.