Introduction

Partial liver resection (LR) is a safe, effective treatment for early hepatocellular carcinoma (HCC) in patients with nondecompensated cirrhosis [1, 2]. However, there is a new tumor or recurrence in a high percentage of patients due to various factors, such as liver cirrhosis persistence [3].

Treatment of recurrence includes the use of antineoplasic drugs, locoregional therapies (radiofrequency, chemoembolization, alcoholization), re-resection, and liver transplantation (LT). The advantage of LT as a treatment for HCC recurrence over other treatments is its efficacy in treating HCC with total hepatectomy, removing underlying cirrhotic tissue and restoring normal liver function [4]. However, owing to organ shortage and other medical reasons, LT cannot always be applied; consequently, liver grafts must be optimized and LT performed only in selected cases [5]. Thus, careful selection of patients who may benefit from LT and identifying those whose risk of recurrence is as low as possible must be done preoperatively. Results regarding LT outcome in patients who underwent LT because of HCC and who have undergone partial LR for a previous tumor are controversial [6, 7].

The aim of our study was to ascertain the outcome of patients with LT due to HCC who had undergone partial LR for a previous tumor compared with that of patients who underwent LT for a primary HCC. We also wanted to determine if there is a group of patients in whom LT would be contraindicated after LR.

Materials and methods

Study design

The study group comprised cirrhotic patients who underwent LT for HCC recurrence who had undergone previous LR for HCC. A case-control (1:2) study was designed to compare these patients with those who underwent LT for primary HCC without previous LR (control group). Patients in the control group had undergone transplantation before and after each patient in the study group and with the same cirrhosis origin.

Preoperative evaluation

Written informed consent was obtained from all patients prior to surgery or to inclusion on the waiting list for LT.

Imaging study

Ultrasonography (US) was used as the first diagnostic approach for early detection of HCC in our surveillance program. HCC was diagnosed by US and abdominal computed tomography (CT) or abdominal nuclear magnetic resonance (MRI) of the abdomen. Extrahepatic tumoral spread was evaluated by abdominal and chest CT and bone scintigraphy. Since the publication of the AASLD Practice Guideline of Management of Hepatocellular Carcinoma, the diagnosis of HCC has been based on this guideline [8]. US-guided percutaneous needle liver biopsy of the liver nodule was performed when the diagnosis was not confirmed by two imaging studies.

Liver function tests and portal hypertension

Liver function was evaluated with liver function tests, and patients were classified according to the Child-Pugh classification. Significant portal hypertension was defined as a portosystemic gradient >10 mmHg. The presence of esophageal varices, splenomegaly, and/or a platelet count <100,000 were also accepted for diagnosis. An α-fetoprotein (AFP) value ≥100 ng/ml was also used as an aid to diagnosis [9].

Liver resection criteria and surgical techniques

Criteria for LR changed over the study period (before 1999 patients who underwent LR were mostly those with a single tumor and preserved liver function). Basically, candidates for surgery were patients with good general status and a feasible, complete, safe resection (Child-Pugh A or B) of the liver tumor, with no extrahepatic disease. Patients have been treated according to the HCC management guideline since its publication [8].

The surgical techniques and postoperative treatment were described previously by our group [10]. Intraoperative US examination was performed on all occasions to rule out other foci of HCC and to perform radical surgery with safe margins. Minor hepatectomy was defined as LR of fewer than three liver segments and major hepatectomy as three or more liver segments [10].

Liver transplantation criteria

Criteria for LT also changed during the study period (before 1996 candidates for LT were mostly those with fewer than three tumors and the tumors being <5 cm). Cirrhotic patients were considered for LT who were <70 years of age, had no contraindications for major surgery, and had a single unresectable central tumor or other tumors in patients with significant portal hypertension or bilirubin >1 mg/dl. During the early years of the study period, patients with large HCCs or with more nodules were selected to undergo transplantation. Since the Milan criteria were established, however, only patients fulfilling these criteria have been included on the waiting list for LT [11].

High pathologic risk of recurrence after liver transplantation

Patients with HCC who presented with microvascular invasion and/or satellitosis on pathologic study after LT or LR for HCC were considered to have a high pathologic risk of recurrence [12].

Treatment of HCC recurrence

Patient follow-up after LR included US and AFP every 3 months during the first year and every 6 months after the first year. A new complete evaluation was undertaken to detect extrahepatic metastases when liver HCC recurrence was diagnosed. Treatment of recurrence was repeat liver resection when technically feasible and the patient had nondecompensated cirrhosis, normal bilirubin, and no portal hypertension. In cases of decompensated cirrhosis or abnormal bilirubin and significant portal hypertension, LT was indicated if the patient’s age and HCC characteristics permitted such a procedure. In the remaining cases, ablative techniques such as radiofrequency thermal ablation (RFA) and transarterial chemoembolization (TACE) were applied. When HCC recurrence was systemic, other palliative treatments were administered.

Postoperative morbidity and mortality

Morbidity and mortality were considered as those incidents occurring during the postoperative hospital stay prior to discharge or within the first 30 days after surgery.

Statistical analysis

Data were expressed as the mean ± SD and median (range) when an abnormal distribution of data was identified. A nonparametric test was used (Mann-Whitney U-test) for numeric variables; and the chi-squared test with Fisher’s correction for categoric variables was used to compare the groups. Patient survival rates were estimated using the Kaplan-Meier method and compared by the long-rank test. p < 0.05 was considered statistically significant. Statistical analyses were performed with SPSS 15.0 softward (SPSS, Chicago, IL, USA).

All data were prospectively collected and retrospectively analyzed. Median follow-up was 70 months (12.7–203.0 months).

Results

Patient population

From January 1990 to December 2007, a total of 303 cirrhotic patients with primary HCCs were evaluated for surgical treatment at our department. Of these patients, 203 were included on the waiting list for LT. During follow-up, 11 patients dropped out and 1 remains on the waiting list; ultimately, 191 underwent LT for primary HCC (Fig. 1). In all, 100 patients underwent LR. When HCC recurrence was observed in the LR group (69/100), only 17 of the 69 patients (25%) could be treated by LT (study group). The other 52 patients were not eligible for LT at the time of HCC recurrence, of whom 30 had oncologic reasons (15 had out of Milan local recurrence and 15 extrahepatic disease) (Fig. 1).

Fig. 1
figure 1

Patient distribution. HCC hepatocellular carcinoma, LT liver transplantation

Full size image

Characteristics of the study group versus HCC recurrence in the patients noneligible for liver transplantation group

At the time of LR, the characteristics of the 17 patients who underwent LT due to HCC recurrence (study group) are summarized in Table 1. No differences were observed between the study group and the group of patients with HCC recurrence after LR who could not undergo transplantation for oncologic reasons (Table 2).

Table 1 Characteristics of the study group at the time of liver resection (n = 17)
Full size table
Table 2 Characteristics at the time of liver resection in the study group versus patients with HCC recurrence not eligible for LT for oncologic reasons
Full size table

Liver transplantation for HCC recurrence in patients with previous liver resection (study group) versus the control group

No differences were found between the study group and the control group except for preoperative portal hypertension (Table 3). Although no significant differences were found regarding HCC recurrence and mortality after LT between groups, the recurrence rate in patients in the study group was twice as high (Table 3). No differences were found regarding actuarial survival between groups, although better disease-free survival was observed in patients of the control group (Fig. 2).

Table 3 Study group and control group characteristics at the time of LT and outcome after LT
Full size table
Fig. 2
figure 2

a Actuarial patient survival of the control group versus the study group. b Patient cumulative recurrence rate in the control group versus the study group

Full size image

Survival according to time to recurrence after liver resection

Patients of the study group were divided into two groups according to the time to recurrence after the partial LR. The first group comprised 4 patients whose recurrence post-LR occurred during the first year of follow-up (<12 months). The other 13 patients formed the second group, whose recurrence was diagnosed after the first year (>12 months). Characteristics of the two groups are shown in Table 4. Patients with a recurrence during the first year had significantly poorer tumor cell differentiation than patients with recurrences after the first year. Although no significant differences were found, patients in the first group had three times more pathologic risk of recurrence at the time of LR. After LT, recurrence was higher in the first group than in the second group (75% vs. 15.4%, respectively, p < 0.03) (Table 4). Moreover, actuarial 1-, 3-, and 5-year patient survivals were better in the second group (25%, 0%, 0% vs. 69%, 69%, 69%; p < 0.02) (Fig. 3).

Table 4 Characteristics of the study group according to the time to HCC recurrence after liver resection
Full size table
Fig. 3
figure 3

Liver transplantation survival according to the time of recurrence after liver resection (n = 17)

Full size image

Patients of the study group recurring after the first year (n = 13) versus control group

No differences were found according to HCC recurrence and disease-free survival when patients of the study group were compared with the control group once patients with recurrence during the first year had been excluded from the study group. HCC recurrence was 15.4% for patients with recurrence after 1 year vs. 12% for the control group; and the 1-, 3-, and 5-year disease-free survivals were 100%, 88%, and 75% for patients with recurrence after 1 year vs. 97%, 93%, and 89% for the control group.

Discussion

Over the last decade, improvement in patient selection for LR, surgical skills, and postoperative period management, together with organ shortage, have led to an increase in the applicability of this treatment [13]. It is well known that 5-year survival rates after LR are at least 50% [14] with a high recurrence rate (up to 70%) [3]. At the time of HCC recurrence, the treatment we can offer our patients depends on various factors, such as age, size of the largest nodule, number of nodules, tumor location in the liver, and cirrhosis status, as well as the presence of extrahepatic recurrence. Poon et al. reported that almost 80% of patients with recurrence after a primary LR for HCC remain eligible for LT [15].

Our study focused on a group of patients who had undergone LR for HCC and, at the time of recurrence, were given a transplant. LT in patients with HCC is clearly limited by organ shortage worldwide. Various strategies, such as performing LR in patients with HCC who are eligible for LT, have been proposed in an attempt to cure these patients [16]. One consists of LR and “salvage” LT at the time of HCC recurrence; however contradictory results have been reported [6, 7]. Patients from the series of Belghiti et al. were favorable regarding HCC recurrence risk, and in some of those patients LT was not indicated for HCC recurrence [7]. On other hand, Adam et al. noted that the indication for LT was HCC recurrence in all patients [6]. Probably it is these and other differences between the cohorts that render the results so contradictory. The problem with this strategy and its applicability lies, first, in the type of HCC recurrence: advanced intrahepatic disease and extrahepatic recurrence are contraindications to transplantation. Moreover, it is not clear which patients, who are eligible for LT at the time of recurrence, will benefit from LT.

In our study group, LR was performed for primary HCC mainly in patients with a single tumor and no significant portal hypertension. However, some patients exceeded these criteria as our patient population belonged to a historical cohort of patients, some of whom had undergone LR before the Barcelona Clinic Liver Cancer (BCLC) staging classification and guidelines for HCC management were reported [8, 17]. The tumor size of a solitary nodule has not been reported as a clear-cut limiting factor for LR [8, 18]. Despite the risk of vascular invasion and satellitosis increasing with size [19, 20], some single tumors may grow as a large mass with no evidence of vascular invasion [8]. Most of our resections were minor hepatectomies with complete excision of the tumor in all cases; and even though our morbidity rate was high compared with other studies [21], our outcome was successful with nil mortality or liver failure. In our series, we found a low rate (25%) of patients who could have transplantation after HCC recurrence. Most of the patients had extrahepatic HCC recurrence or a local recurrence that exceeded the Milan criteria even though there were no significant differences in HCC pathology between the LR and study groups. These findings are probably explained by intrinsic tumor aggressiveness [22–25].

As previously reported, LT in patients with previous LR for HCC is technically feasible [7, 15]. This fact was confirmed in our case-control study as no differences were found regarding operating time, intraoperative transfusion, or postoperative morbidity and mortality. On the other hand, although both preoperative and pathologic findings of the tumor were similar in both groups, the recurrence rate and disease-free survival were better in the control group. Although vascular invasion and satellitosis are accepted worldwide as the best predictors of tumor recurrence [25, 26], the impact of these findings in patients fulfilling Milan criteria in the liver transplantation setting may not be as significant. As reported in the Adam et al. study and confirmed by our group, patients with a previous LR for HCC have an increased risk of recurrence and poorer disease-free survival [6]. One differential finding between the study and control groups was that the time from diagnosis of the first HCC to LT was clearly longer in the study group. Probably these patients had been exposed to tumor cells for a longer period of time, which would explain the higher recurrence rate.

One of the most salient results of our study was that patient survival post-LT in the study group differed depending on the time of HCC recurrence after LR. No patient in whom HCC recurrence was diagnosed during the first year of follow-up after LR survived more than 3 years post-LT. Although differences did not reach statistical significance, however, patients with early HCC recurrence after LR had a threefold higher pathologic risk of recurrence when the whole cirrhotic liver was examined at the time of LT. Minagawa et al. [25] reported a 5-year survival of 57% after repeat hepatectomy—but in selected patients who probably would have had a good prognosis regardless of the treatment. Interestingly, they found early recurrence (during the first year of follow-up) after the first hepatectomy to be a risk factor for poor outcome after repeated hepatectomy. In our series, when certain patients—those with early recurrence after LR—were excluded from the study, results were similar to those of the control group. These data strongly reinforce the finding that patients with early HCC recurrence after LR for HCC have poor prognosis after LT.

The fact that the tumor recurs early could be an indicator of advanced disease, with other intrahepatic HCC foci at the time of LR that were not detected with imaging techniques. In any event, early recurrence after LR appears to be a predictive factor of poor outcome when patients are included on the waiting list for LT.

Cillo et al. reported that tumor differentiation represents a direct marker of biologic tumor aggressiveness and that it may provide more accurate information on the risk of recurrence [27–29]. Even among the small number of patients in our series, all patients with recurrence during the first year after LR had poor cellular differentiation at the time of the resection. These findings indicate more aggressiveness of the tumors. Along the same lines, it has been suggested that the histologic grade of the HCC can be determined preoperatively with percutaneous needle biopsy, thus indicating the risk of microvascular invasion [30]. In association with new biohumoral factors that may predict the invasiveness of HCC, histologic grading may estimate the risk of posttransplant recurrence, thereby improving selection criteria [29].

To resolve this problem and reduce HCC recurrence after LT, Sala and colleagues from the BCLC study group proposed including on the LT waiting list all patients with unfavorable histopathologic signs (satellitosis and/or vascular invasion) after LR for HCC as a “prophylactic” LT before a hypothetical recurrence appears [12]. Although a good outcome was reported with this approach, further data are required to validate these results.

We consider LR an acceptable option as a first approach even though when HCC recurrence appeared in our patients who had undergone resection “salvage” LT could be applied in only one-fourth of them. Future biological HCC markers may help us decide which of these patients will benefit directly from LR and which will need a more aggressive treatment such as LT.

The MELD system [31] is currently being applied in most countries, and patients with HCC have a shorter waiting-list time than they had before; however, in some countries, dropout from the waiting list is high [32], and treatment to downstage HCCs that grow during the waiting-list time is applied. LR probably represents the best possible downstaging treatment for HCC. Despite being a more aggressive treatment that can be applied in only a small number of cases, it offers a 25% possibility of cure to our patients.

Conclusions

Liver transplantation can be safely performed in patients who have undergone previous LR for HCC, with outcomes similar to those in patients undergoing primary LT for HCC but with a high recurrence rate. In our study, patients with a disease-free interval between LR and HCC recurrence of <1 year had poor survival after LT. The fact that patients with recurrence during the first year after hepatectomy have such a poor prognosis after transplantation should induce us to consider whether, given the shortage of donors, these patients should indeed be chosen for a transplant, particularly if poor cellular differentiation is present in the tumor. Nevertheless, it should be taken into account that our series included a limited number of patients.