Survival Analysis of Patients With Transplantable Recurrent Hepatocellular Carcinoma: Implications for Salvage Liver Transplant

Hypothesis  There is no difference in the survival benefit between salvage liver transplant and nontransplant therapies for recurrent hepatocellular carcinoma (HCC).

Design  Retrospective study.

Setting  Tertiary referral center.

Patients  Sixty patients developed transplantable intrahepatic recurrent HCC after curative resection. Twelve patients received salvage liver transplant, whereas 48 received nontransplant therapies, including a second surgical resection, radiofrequency ablation, transarterial chemoembolization, and percutaneous ethanol injection.

Main Outcome Measures  The overall survival rates were compared between the 2 groups. Clinicopathologic variables were evaluated by univariate and multivariate analyses for their influence on overall survival.

Results  There was no significant difference in overall survival rates between the salvage transplant and nontransplant groups. In the nontransplant group, pTNM (pathologic TNM) staging at primary resection and the time from primary resection to tumor recurrence were identified as independent prognostic factors affecting overall survival. These 2 factors carried no prognostic value in the salvage transplant group. Patients in the salvage transplant group with stage II tumors before the primary resection or intrahepatic recurrence within 12 months of the primary resection had significantly better overall survival than did the nontransplant group with corresponding poor prognostic factors.

Conclusions  Patients with transplantable intrahepatic recurrence can be treated effectively by salvage transplant or nontransplant therapies. Salvage transplant may be more beneficial to patients with stage II tumors before the primary resection and those with early intrahepatic recurrence.

The incidence of hepatocellular carcinoma (HCC) is increasing worldwide because of the high prevalence of hepatitis B and C viral infection in areas where such infections have become endemic to the region. Hepatocellular carcinoma accounts for 5.6% of all human cancers.¹ For patients with early HCC (solitary tumor ≤5 cm, or ≤3 nodules, each ≤3 cm), orthotopic liver transplant (OLT) is the curative treatment of choice. However, under the pressure of donor organ shortage, the application of OLT for early HCC is limited.^2-4 A long waiting time and hence a high dropout rate because of tumor progression have been identified as the major poor prognostic factors for patients with early HCC listed for OLT.⁵ Therefore, a strategy of primary resection and salvage transplant for intrahepatic tumor recurrence has been proposed to be a reasonable approach for early HCC with preserved liver function.^6,7 The outcome of this approach has been shown to be comparable to that of primary liver transplant.⁸

Although salvage transplant is gaining much enthusiasm among transplant centers, there are several other nontransplant therapeutic modalities for intrahepatic recurrent HCC in which favorable outcome has been reported. These include a second hepatic resection (hereinafter referred to as re-resection),⁹ radiofrequency ablation (RFA),¹⁰ transarterial chemoembolization (TACE),¹¹ and percutaneous ethanol injection (PEI).¹² However, most of these retrospective studies involved a heterogeneous group of patients with recurrent HCC. To our knowledge, no data in the literature document the long-term outcome of these nontransplant therapies for patients with transplantable recurrent HCC. Such information is crucial in the era of liver transplant, which has an increasing applicability for patients with HCC. Whether or not to enlist patients with transplantable recurrent HCC for salvage transplant requires detailed analysis of the efficacy of salvage transplant and other nontransplant therapies. Hence, the allocation of deceased donor liver grafts among patients with HCC or cirrhosis can be optimized, and it is not absolutely necessary to risk a live donor if live donor liver transplant is contemplated. The aim of this study was to evaluate the survival outcome of patients with transplantable recurrent HCC receiving salvage transplant or nontransplant therapies. Based on the results of subgroup analysis, stratification of patients with transplantable recurrent HCC to salvage transplant or nontransplant therapies could be better defined.

From January 1, 1989, through December 31, 2003, 711 consecutive patients received curative hepatic resection for HCC in the Department of Surgery of the University of Hong Kong, Queen Mary Hospital. Clinicopathologic data including patients' demographic characteristics, liver functional status, tumor status, and histopathological findings of the resected tumors were prospectively collected in a computerized database. From this database, we found 277 patients 65 years or younger with transplantable early HCC according to the widely accepted criteria consisting of (1) a solitary tumor 5 cm or smaller, or no more than 3 nodules, each 3 cm or smaller; (2) absence of extrahepatic metastasis; and (3) absence of macroscopic tumor invasion to the portal or the hepatic vein.^13,14 Patients with a history of tumor rupture, positive tumor resection margin, and/or mixed cholangiohepatocellular carcinoma were excluded from the study. All patients were followed up by the same surgical team with a standard protocol of surveillance for recurrence, which consisted of contrast-enhanced computed tomography (CT) 1 month after resection, followed by a serum α-fetoprotein (AFP) assay, ultrasonography, or contrast-enhanced CT and chest radiography every 3 months. Among the 277 patients with transplantable early HCC during the primary hepatic resection, 75 patients developed tumor recurrence in the liver remnant only during follow-up. Sixty patients constituted the subjects of the present study, in whom the intrahepatic recurrent HCC fulfilled the criteria for salvage transplant, as suggested in a previous study (solitary tumor ≤5 cm, or ≤3 nodules, each ≤3 cm; absence of extrahepatic metastasis; and absence of macroscopic tumor invasion to the portal or the hepatic vein).⁶ Fifteen patients were excluded from the study because the recurrent tumor did not fulfill the criteria for salvage transplant.

Since 2000, we have offered salvage transplant to selected patients with transplantable recurrent HCC. The selection criteria include poor liver function (Child-Pugh class B or C),¹⁵ the patient's preference, and the availability of liver donor in case of live-donor liver transplant. Twelve of 60 patients with transplantable recurrent HCC received salvage transplant. It was our policy that no priority for OLT be given to these patients over those on the waiting list with cirrhosis without HCC. Salvage transplant was performed using live donor liver grafts in 9 patients and deceased donor liver grafts in 3. The median waiting time for salvage transplant since tumor recurrence was 5 months (range, 8 days to 56 months). Seven patients received TACE as bridging therapy before OLT. The techniques of OLT have been described in our previous series.¹⁶ In the early period of the present series, the OLT recipients received a double immunosuppression regimen consisting of tacrolimus and a corticosteroid. Since 2002, we have supplemented the regimen with mycophenolate mofetil and interleukin 2 receptor antagonist during the perioperative period.

Forty-eight patients with transplantable recurrent HCC underwent various nontransplant therapies, including hepatic re-resection (12 patients), RFA (3), TACE (31), and PEI (2). The treatment decision was based on the pattern of recurrence and liver function reserve. In patients with compensated liver function, hepatic re-resection was considered the treatment of choice for resectable recurrent HCC. The extent of liver resection was based on the number and site of recurrent tumors and the general status of the patient. However, most of the patients had unresectable recurrent HCC, in which reasonable local tumor control could be achieved by means of locoregional therapies. These included TACE using cisplatinum as the chemotherapeutic agent, RFA using a lethal high local temperature for tumor ablation, and PEI using absolute alcohol as the cytotoxic agent. We have previously reported the efficacy of these locoregional treatment modalities for recurrent HCC.¹¹

For patients receiving salvage transplant, hepatic re-resection, and RFA, CT was performed 1 month after treatment to detect any residual tumor and every 3 months thereafter with the serum AFP assay to detect tumor recurrence. Chest radiography, CT, and, if indicated, a bone scan were used to detect any concurrent extrahepatic metastasis. Further treatment using locoregional therapies was offered to patients in case of intrahepatic recurrence. The same follow-up policy (CT and serum AFP assay every 3 months) was adopted in patients with TACE or PEI. Tumor response was categorized according to the Response Evaluation Criteria in Solid Tumors guidelines,¹⁷ in which the disappearance of all target lesions indicates complete response; at least a 30% decrease in the sum of the longest diameter of target lesions indicates partial response; a 20% increase in the sum of the longest diameter of target lesions or appearance of 1 or more new lesions indicates progressive disease; and insufficient shrinkage to qualify for partial response or insufficient increase to qualify for progressive disease indicates stable disease.

All epidemiologic and clinicopathologic data were prospectively collected in a computerized database. Long-term overall survival after recurrence was analyzed and compared between the salvage transplant and nontransplant groups. For patients receiving nontransplant therapies, overall survival after recurrence was correlated with 16 clinicopathologic variables to determine the prognostic factors. Host factors included age at recurrence, sex, hepatitis B surface antigen status, anti–hepatitis C virus antibody status, Child-Pugh grade¹⁵ at recurrence, histological status of the liver, and any perioperative blood transfusion during the initial hepatectomy. Tumor factors at primary resection included maximum tumor size, number of tumors, presence of lymphovascular permeation, and pTNM (pathologic TNM) staging according to the 2002 International Union Against Cancer classification.¹⁸ Tumor factors at recurrence included serum AFP concentration, maximum tumor size, number of tumors, and time from primary resection to recurrence. Finally, the type of treatment received was included in the analysis.

Continuous data were expressed as the median with ranges and were compared using the Mann-Whitney test. Categorical data were compared using the χ² test with Yates' correction or the Fisher exact test where appropriate. The primary end point of this study was survival time after tumor recurrence. The overall survival rates were calculated by the Kaplan-Meier method and compared using the log-rank test. Hospital deaths were included in the overall survival analysis. Multivariate analysis was performed using the Cox's proportional hazards model to identify independent prognostic factors of overall survival in the nontransplant group. All statistical analyses were performed using SPSS statistical software (version 11.0 for Windows; SPSS Inc, Chicago, Illinois). A P < .05 was considered statistically significant.

The demographic and clinicopathologic data of patients undergoing salvage transplant and nontransplant therapies are presented in Table 1. There were no significant differences between groups in age, sex, or proportions of patients with chronic hepatitis B and C virus infection. Patients in the salvage transplant group had significantly poorer liver function than did those in the nontransplant group. Among the 12 patients in the salvage transplant group, 2 patients (17%) and 4 (33%) had liver function of Child-Pugh classes C and B,¹⁵ respectively, whereas 47 of the 48 patients in the nontransplant group (98%) had Child-Pugh class A¹⁵ liver function. Most of the patients in both groups (100% in the salvage transplant group and 40 [83%] in the nontransplant group) had liver cirrhosis proved by histological examination results at the primary resection of HCC. The tumor size, number of tumors, presence of microvascular invasion, and pTNM staging¹⁸ were similar between groups at the primary resection. Fewer patients in the salvage transplant group (1 [8%]) required perioperative blood transfusion than those in the nontransplant group (9 patients [19%]), but the difference was not statistically significant. Tumor characteristics at recurrence (tumor size, number of tumors, and serum AFP concentration) were similar between groups. The median time from primary resection to recurrence was longer in the salvage transplant group (34 months) than in the nontransplant group (17 months), although the difference was not statistically significant.

The median operating time was 14 (range, 8-17) hours. Five of the 12 patients (42%) did not require blood transfusion during the perioperative period, and the median blood transfusion requirement was 5 (range, 0-18) U. There was no hospital mortality and no patient required reoperation. Postoperative complications occurred in 5 patients (42%), which included symptomatic pleural effusion (3 patients) and bleeding peptic ulcer and biliary anastomotic site stricture (1 patient each). At a median follow-up of 31 months, 5 patients (42%) developed tumor recurrence within the liver graft, and no patient had distant metastasis. At the time of analysis, 7 patients had died; the causes of death included terminal malignancy (4 patients), uncontrolled sepsis (2), and gastrointestinal tract bleeding (1).

Nine liver donors were involved in the donor hepatectomy for salvage liver transplant. No hospital mortality or major postoperative complication occurred. One patient developed a minor wound infection after the operation and was treated conservatively. All liver donors were able to resume normal daily lives within 6 to 8 weeks after the operation.

The patient outcomes after each type of nontransplant therapy and at a median follow-up of 61 months are described in the following paragraphs.

Most of the patients (10 of 12) underwent minor hepatic resection, defined as the resection of 2 or fewer liver segments according to the Couinaud classification.¹⁹ Only 2 patients received major hepatic resection, defined as the resection of 3 or more liver segments. All patients had margin-negative curative resection for liver tumor, which was confirmed histologically. There was no evidence of residual disease on the postoperative 1-month CT in any of these patients. The median operating time was 3.5 (range, 1.5-7.0) hours, and median blood loss during operation was 400 (range, 150-1200) mL. The mortality and morbidity rates were 0% and 25%, respectively. The postoperative complications included symptomatic ascites, symptomatic pleural effusion, and chest infection in 1 patient each. Six patients (50%) developed intrahepatic recurrence, and no patients had distant metastasis.

Radiofrequency ablation was performed using a percutaneous approach in 2 patients and a laparoscopic approach in 1 patient. The complete ablation rate was 100%. There was no hospital mortality. One patient developed hemoglobinuria shortly after RFA, which subsided spontaneously after adequate hydration and administration of intravenous mannitol. Local recurrence at the ablation site and intrahepatic recurrence occurred in 1 patient each, and no patient had distant metastasis.

The median number of sessions of TACE and PEI was 6, with an overall complication rate of 12%. These complications included liver failure (2 patients) and acute pancreatitis and acute cholecystitis (1 patient each). Complete and partial tumor response was achieved in 6 patients (18%) and 10 patients (30%), respectively. Three patients (9%) had stable disease and 13 patients (39%) had tumor progression after treatment. The overall tumor response rate to TACE and PEI was 48%.

There was no significant difference in overall survival between the 2 groups (P = .21). Patients in the salvage transplant group had a median overall survival of 61 (95% confidence interval [CI], 29-93) months, and patients in the nontransplant group had a median survival of 43 (95% CI, 38-49) months. The cumulative 1-, 3-, and 5-year overall survival rates were 100.0%, 80.4%, and 40.8%, respectively, for the salvage transplant group and 93.5%, 60.0%, and 53.8%, respectively, for the nontransplant group (Figure 1).

The overall survival of patients in the nontransplant group was further analyzed with respect to the associated prognostic factors. Univariate analysis showed that 5 of the 16 evaluated factors had a significant prognostic influence on overall survival (Table 2). Age, high serum AFP level, presence of lymphovascular permeation at primary resection, pTNM staging at the primary resection, and time from the primary resection to recurrence were associated with worse overall survival. On multivariate analysis, only pTNM staging at primary resection (risk ratio, 2.935; 95% CI, 1.057-8.147 [P = .04]) and time from the primary resection to recurrence of 12 months or less (risk ratio, 2.849; 95% CI, 1.114-7.282 [P = .03]) were independent predictors of overall survival.

When the 2 independent prognostic factors for the nontransplant group (pTNM staging at primary resection and time from primary resection to recurrence) were examined among the salvage transplant group using univariate analysis, there was no significant prognostic value on the overall survival of the latter group. An additional subgroup analysis focusing on patients with stage II tumors at the primary resection or intrahepatic tumor recurrence within 12 months of the primary resection was performed to compare the overall survival outcome between the salvage transplant (6 patients) and nontransplant (23 patients) groups. Under this condition, patients in the salvage transplant group had significantly better overall survival rates than did those in the nontransplant group (1 year, 100% vs 91.3%; 3 years, 100% vs 49.7%; 5 years, 50% vs 42.6%) (P = .046, Figure 2).

Because of the propensity for tumor spread along the tributaries of the portal venous system and the underlying diseased liver due to chronic viral infection, intrahepatic tumor recurrence after the primary resection for early HCC is frequently encountered. From our previous study,²⁰ intrahepatic recurrent HCC was considered transplantable in up to 79% of patients using the same criteria as for primary transplant for HCC. Management of transplantable recurrent HCC is always a dilemma to the clinician. On one hand, salvage transplant is definitely the best treatment option for this group of patients, in whom the recurrent tumors and diseased liver can be cured in a single operation. The risk for the liver donor involved in live donor salvage transplant is also minimal as shown in our study. However, the efficacy of salvage transplant is limited by the shortage of organ donors and the prolonged OLT waiting time, resulting in tumor progression during the waiting period. From an intention-to-treat analysis of primary OLT for HCC, a dropout rate of up to 23% was reported, and the 2-year intention-to-treat survival rate was only 54%.⁵ On the other hand, some nontransplant therapies, including hepatic re-resection, RFA, TACE, and PEI, are effective in achieving a reasonable tumor clearance and prolonging patients' survival. To optimize liver graft allocation among patients with HCC or cirrhosis, a better stratification of patients with transplantable recurrent HCC to salvage transplant or nontransplant therapies is mandatory.

This is, to our knowledge, the first study to compare the efficacy of salvage transplant and nontransplant therapies for patients with transplantable recurrent HCC. We found no significant difference in overall survival between the 2 treatment options. We have observed an effective tumor control in each of the nontransplant therapies. In particular, a tumor response rate of up to 48% was achieved in the TACE and PEI groups, and all patients in the hepatic re-resection group underwent curative surgery for recurrent tumors. In addition, the complete ablation rate was 100% in the RFA group. The high efficacy of these nontransplant therapies could be attributed to the underlying favorable biological characteristics of the tumors because patients with early-stage recurrence after curative resection for early HCC were subjected to these treatment options. In fact, the 5-year survival rate of 53.8% achieved in the nontransplant group is comparable to that of patients receiving hepatic resection for small HCC.²¹

Our study showed a 5-year posttransplant survival rate of 40.8% for patients undergoing salvage transplant. Considering a median interval from primary resection to recurrence of 34 months and a median waiting time of 5 months before transplant, the overall 5-year survival from the time of primary resection was 65%. This reflects the efficacy of our strategy of primary resection followed by salvage transplant for transplantable intrahepatic tumor recurrence. In a study by Belghiti et al,⁸ the 3- and 5-year overall survival rates for patients with salvage transplant were 82% and 61%, respectively, which were comparable to those of our series. Nonetheless, the tumor recurrence rate after salvage transplant was substantially higher in our series than that of Belghiti et al⁸ (41.8% vs 5%). In another study by Adam et al,²² the reported tumor recurrence rate after salvage transplant was up to 54%. This is mainly related to the biological aggressiveness of recurrent tumors. In our previous series,²³ 25% of patients had microvascular invasion at the primary resection, which was an important predictor for recurrence after transplant. In patients with long-term immunosuppression, such risk of recurrence should not be underestimated. In addition, 9 of 12 patients received live donor liver transplant, which was potentially another factor for high recurrence after transplant.²⁴ Theoretically, live donor liver grafts are often small for size. The subsequent acute-phase injury, regeneration, and resulting tumor angiogenesis might ultimately increase the chance of tumor recurrence.

Our study is unique in its identification of poor prognostic factors in patients with transplantable recurrent HCC receiving nontransplant therapies, ie, pTNM staging at the primary resection and time from primary resection to recurrence. Pathologic tumor status in pTNM staging¹⁸ is closely linked to tumor size, the number of tumors, and the presence of microvascular invasion. These clinicopathologic variables were shown to carry a significant prognostic value in patients with HCC.²³ The time from primary resection to recurrence may signify the underlying pathogenesis of recurrent tumors. In our previous study of 126 patients with intrahepatic tumor recurrence,²⁵ 89% of early recurrences were classified as intrahepatic metastases by histological study results, whereas all of the late recurrences were multicentric occurrences due to underlying chronic liver disease. Unlike salvage transplant, each nontransplant therapy focuses on local tumor control by various mechanisms and the remaining liver remnant would be left untouched. Therefore, late recurrence would be better treated by nontransplant therapies because of its nature of de novo tumor. In a study by Minagawa et al,²⁶ late recurrence more than 12 months after the primary resection was identified as a favorable independent prognostic factor in 67 patients with hepatic re-resection for recurrent HCC. However, for early recurrence, nontransplant therapies are less effective at curing the disease, and the intrahepatic metastasis might be the main cause of treatment failure. In such a condition, salvage transplant might be beneficial in the elimination of both tumor clearance and the diseased liver. This is supported by the subgroup survival analysis on patients with a stage II tumor at the primary resection or with early recurrence (less than 12 months after primary resection). We have clearly demonstrated that salvage transplant contributed to better overall survival than did nontransplant therapies in this subgroup of patients. Nonetheless, one has to be cautious when interpreting this finding. The presence of intrahepatic metastasis and probable microvascular tumor invasion in early recurrent tumors signify a highly aggressive underlying tumor with relatively poor prognosis. The survival benefit of salvage transplant in this kind of tumor should be further evaluated by large-scale prospective studies.

Our study has the limitations of its retrospective analysis of the efficacy of 2 different treatment approaches for transplantable recurrent HCC and its small sample size. To provide clinical evidence of high scientific significance on this issue, a large-scale prospective clinical study based on an intention-to-treat basis would be desirable. This will help to eliminate the bias of dropout in patients with transplantable recurrent HCC listed for salvage transplant. Another shortcoming of the present study is the selection bias of the 2 studied groups. As expected, the salvage transplant group had significantly poorer liver function, and we tend to treat patients with preserved liver function with various nontransplant therapies. Furthermore, we have investigated the results of different nontransplant therapies as a group. Because the efficacy of individual nontransplant therapies may differ from each other, future studies should focus on the comparison of each of these treatment options with salvage transplant.

In conclusion, patients with transplantable intrahepatic recurrence can be treated effectively by salvage transplant or other nontransplant therapies, including hepatic re-resection, RFA, TACE, and PEI. Salvage transplant may offer a survival benefit to selected patients with stage II tumors at primary resection and those with recurrence within 12 months of the primary resection.

Correspondence: Sheung Tat Fan, MS, MD, PhD, FRCS(Glasg), FRCS(Edin), FACS, Department of Surgery, The University of Hong Kong, Queen Mary Hospital, 102 Pokfulam Rd, Hong Kong (stfan@hku.hk).

Accepted for Publication: September 8, 2007.

Author Contributions:Study concept and design: Ng, Lo, Liu, Poon, Chan, and Fan. Acquisition of data: Ng. Analysis and interpretation of data: Ng. Drafting of the manuscript: Ng. Critical revision of the manuscript for important intellectual content: Lo, Liu, Poon, Chan, and Fan. Statistical analysis: Ng. Study supervision: Lo, Liu, Poon, and Fan.

Financial Disclosure: None reported.

Funding/Support: This study was supported by the Sun C. Y. Research Foundation for Hepatobiliary and Pancreatic Surgery of The University of Hong Kong.