Introduction

Hepatocellular carcinoma (HCC) is the sixth most prevalent cancer and the third most frequent cause of cancer-related death in the world [1]. Although the majority of cases are still found in Asia and Africa, recent studies have shown that the incidences and mortality rates of HCC are increasing in North America and Europe [2].

Hepatic resection provides a potentially curative outcome for HCC patients who are indicated for surgery [3–5]. However, owing to the high recurrence rate, long-term survival after hepatic resection of HCC is far from satisfactory. Patients with large HCC are known to have poorer long-term outcomes than those with small HCC after curative resection [6–10]. This is often because large HCC has other adverse clinicopathological factors affecting long-term survival and recurrence, such as multiplicity, satellite nodules, macroscopic vascular invasion, or distant metastasis [11–13]. Nevertheless, when HCC is solitary, especially when it is not associated with macroscopic vascular invasion, hepatic resection is still carried out by many surgeons even when the tumor is very large size, provided that the tumor is resectable [4, 14–16]. Until now, the relationship between tumor size and long-term prognosis after curative resection remains unclear for this subset of patients with solitary HCC.

To address this issue, we conducted a single-center retrospective study to evaluate the prognostic value of tumor size on the overall survival (OS) and recurrence-free survival (RFS) after curative resection of solitary HCC without macroscopic vascular invasion, and to appraise the prognostic importance of tumor size on the decision-making of hepatic resection for those patients.

Patients and methods

Patients

From January 2002 to December 2010, all consecutive patients who received R0 curative resections for HCC, which was defined as macroscopically complete removal of tumor with a histopathologically tumor-free surgical margin, from the Fourth Department of Hepatic Surgery at the Eastern Hepatobiliary Surgery Hospital were included into this study. Patients excluded were those with multiple HCC (but not those with microsatellites), macroscopic vascular invasion (defined as a tumor invasion or a tumor thrombus in the vessels visible on radiological imaging prior to surgery or during surgery), and distant metastasis. For this retrospective study, tumor size, cirrhosis, microvascular invasion, tumor encapsulation, and tumor differentiation (evaluated by Edmondson–Steiner grade: I = well, II = moderately, and III–IV = poorly) were obtained by histological examination of resected specimens. The study protocol was approved by the Clinical Research Ethics Committee of the hospital. Written informed consent was obtained from all patients for the patients’ data to be used for clinical research before the operation.

Peri-operative procedures

Pre-operative evaluation of the extent of HCC, liver functional status, and general condition of the patient was carried out before any decision for surgery was made. The tumor was assessed by ultrasonography, computed tomography (CT), and/or magnetic resonance imaging (MRI). Liver function was evaluated by biochemistry and Child–Pugh classification. An upper gastrointestinal endoscopy was performed routinely on all HCC patients who were scheduled for operation. Patients older than 60 years and those with significant co-morbid illnesses underwent formal cardiopulmonary evaluation. A liver functional status of Child–Pugh C was considered as an absolute contraindication for HCC resection. The resection criteria used were consistent over the study period, and have previously been reported [17, 18]. Portal hypertension was defined to be present when endoscopy revealed esophageal varices, in patients with splenomegaly with a platelet count of <100 × 109/L, or when the hepatic venous pressure gradient was >210 mmHg.

The Pringle maneuver was used with cycles of clamping/unclamping times of 15/5 min. Transection of hepatic parenchyma was carried out by the clamp-crushing technique, and hemostasis was achieved on the raw liver surface with an argon beam coagulator. The plane of liver transection was determined by intraoperative ultrasonography. Anatomical resection was the method of choice, but non-anatomical resection was adopted for tumors situated at the junction of several liver segments, or for small and peripherally located tumors, or in patients with a high extent of cirrhosis. Major hepatectomy was defined as resection of three or more Couinaud’s liver segments, and minor hepatectomy as resection of fewer than three.

Follow-up

All patients were investigated for postoperative residual tumors using serum alpha-fetoprotein (AFP), ultrasonography or CT, and chest X-ray at 1 month after surgery. The patients were then followed-up for recurrence or distant metastasis at a 2-monthly interval for the first 6 months, and at a 3-monthly interval thereafter. CT, MRI, angiography, bone scan or positron emission tomography were performed when recurrence or distant metastasis was suspected. Further treatment decision was based on the pattern of recurrent tumor, residual hepatic functional reserve, and general condition of the patient. The OS was calculated from the day of the operation to either the day of death or the day of the last follow-up visit. The disease-free survival was calculated from the date of the operation to the date when recurrence/metastasis was diagnosed. This study was censored on 30 June 2012.

Statistical analysis

The clinicopathologic and operative data were collected in a computerized database. Continuous variables were expressed as mean ± standard deviation (SD) or median (range) and compared using the Mann–Whitney U test. Categorical variables were compared using the χ 2 test with Yates correction or the Fisher’s exact test, as appropriate. Hospital deaths were included in calculating the OS and RFS rates. Cumulative OS and RFS rates and curves were analyzed by the Kaplan–Meier method, and differences were compared by the log-rank test. After univariate analysis, only significant variables with p < 0.1 were used in the multivariate analysis, which was performed using a stepwise logistic regression analysis. All statistical analyses in this study were performed with the software package SPSS 18.0 (SPSS Inc., Chicago, IL, USA). A p value < 0.05 was defined as statistically significant.

Results

Patient characteristics and operative variables

Of the 1,324 patients who received an R0 resection for HCC during the study period, 535 patients were excluded from this study because they had multiple HCC. Of the 789 patients with a solitary HCC, 174 patients were excluded because of macroscopic vascular invasion. Only 615 patients were shown to have a solitary HCC on pre-operative radiological imaging. There were 547 males and 68 females, and the median age was 51 years (range 5–81 years). The patients’ characteristics and operative variables are shown in Table 1. The median tumor size was 4.0 cm (range 0.9–22.0; mean ± SD 5.4 ± 3.6) (Fig. 1). Liver function was relatively good in the majority of all patients at the time of surgery (553 with Child A; 62 with Child B), and 421 (68.5 %) had liver cirrhosis histopathologically. A total of 105 patients were found to have microsatellite nodules on histopathology.

Table 1 Patient characteristics and operative variables
Full size table
Fig. 1
figure 1

Male, 54 years old, diagnosed as solitary huge hepatocellular carcinoma in the right side of the liver, who had a 22-cm tumor, underwent right hemi-hepatectomy in March 2004. After curative resection, this patient was still alive and disease-free up to June 2012

Full size image

Postoperative outcomes

Of 615 patients, six died within the hospital stay, with an in-hospital mortality of 1.0 %. The primary complications that led to death included hepatic failure (N = 4), intra-abdominal hemorrhage (N = 1), and pneumonia (N = 1). Postoperative complications occurred in 137 patients, with an overall morbidity of 22.3 %. The grades and types of postoperative complications according to the Clavien–Dindo classification [19] are shown in Table 2. A total of 82 (13.3 %) patients developed only one complication, and 55 (8.9 %) patients developed two or more complications. The most common complications were ascites and pleural effusion, which were usually resolved with diuretics or paracentesis.

Table 2 Postoperative complications
Full size table

Long-term OS and RFS

Within a median follow-up of 61.7 months (range 2.4–134.5), recurrences had developed in 287 patients (46.7 %) and deaths in 216 patients (35.1 %). The OS rates of 1-, 3-, and 5-years were 96.0, 79.8, and 69.9 %, and the corresponding RFS rates were 83.6, 72.7, and 57.2 %, respectively.

Comparison of OS and RFS according to different cut-off values of tumor size

Using 2.0, 3.0, 4.0, 5.0, 8.0, and 10.0 cm as cut-off values of tumor size, patients were divided into groups 1–6, the 1-, 3-, and 5-year OS and RFS rates between the different groups of patients with tumor size up to a certain cut-off value were compared with the groups of patients with tumor size above that cut-off value, and are shown in Table 3. The corresponding OS and RFS curves are shown in Fig. 2. There was a significant difference between each of these groups (all p < 0.05).

Table 3 Univariate analysis of various cut-off values of tumor size for overall and recurrence-free survival
Full size table
Fig. 2
figure 2figure 2

Comparisons of overall survival and recurrence-free survival between two groups according to various cut-off values of tumor size

Full size image

Prognostic factors for OS and RFS

Univariate analyses for the prognostic factors of OS and RFS after curative resection of solitary HCC are shown in Table 4. When putting all the variables with p < 0.1 in Table 4, as well as using tumor size as a continuous variable into the multivariate analysis (Table 5), only microvascular invasion, absence of tumor encapsulation, resection margin <1 cm and intra-operative transfusion were independently associated with poorer OS and RFS. In addition, portal hypertension and serum albumin<35 g/L were also independent factors for OS, with poor tumor differentiation for RFS. However, the logistic regression analyses revealed that when putting as a continuous variable into multivariate analyses, tumor size was not an independent prognostic factor of OS, or of RFS after curative resection.

Table 4 Univariate analysis of prognostic factors for overall and recurrence-free survival
Full size table
Table 5 Multivariate analysis of risk factors for poor overall survival and recurrence-free survival
Full size table

Discussion

In the last few decades, improvements in patient evaluation, surgical techniques, and peri-operative management have significantly reduced the mortality and morbidity rates of hepatic resection. As a result, surgical indications for HCC have extended to resection of tumors even of very large size [20]. Currently, the Barcelona Clinic Liver Cancer (BCLC) staging treatment strategy has become widely accepted, and it is recommended by the American Association for the Study of Liver Diseases [5] and the European Association for the Study of the Liver [21]. In the BCLC staging treatment strategy, hepatic resection is only recommended for BCLC A stage, i.e. early stage HCCs (solitary/single nodule with preserved liver function [Child–Pugh A or B] and asymptomatic [performance status = 0]). For solitary/single HCCs with preserved liver function (Child–Pugh A or B) and without macroscopic vascular invasion, the data from the current study showed that favorable outcomes were achieved after curative hepatic resection, with 5-year OS and RFS rates of 69.9 and 57.2 %, respectively. Thus, the study partly provides support to the BCLC staging treatment strategy on hepatic resection for HCC.

Should symptomatic HCCs (performance status 1–2) be excluded from surgical resection? Many HCCs become symptomatic with increasing size, even though some remain solitary. In the present study, patients with solitary HCC but with performance status 1–2 accounted for more than one-third of all patients. The 5-year OS and RFS rates were 55.7 and 46.3 % after curative resection, which are still quite favorable. Our results cast doubt on the BCLC staging that put these symptomatic HCC patients (performance status 1–2) into an advanced stage (BCLC stage C), and thus deny these patients the chance of a ‘curative’ hepatic resection. Currently, some other authors also question the BCLC staging on the selection criteria for hepatic resection of HCC [4, 22, 23].

Previous studies identified that there was a negative correlation between HCC tumor size and survival rate, and poor outcomes were observed after hepatic resection for large HCCs [6–10]. Nevertheless, in these studies, although conducted on patients with solitary HCCs, some tumors had macroscopic vascular invasion. The present study identified that tumor size was not an independent predictor of long-term survival and recurrence after curative resection of solitary HCC without macroscopic vascular invasion. Recent studies [14, 24–29] also suggested that patients with large solitary HCCs should be considered for surgical resection. Cho et al. [25] reported that patients with a solitary large HCC without vascular invasion had a 5-year OS and RFS rate of 52.9 and 31.7 %, respectively. Some studies even suggested that solitary large HCC is a specific subtype of HCC that has a good outcome after hepatic resection, as these HCC exhibited specific molecular characteristics [14, 26, 27]. Among all the prognostic clinicopathological factors of long-term survival outcomes, macroscopic vascular invasion is well known to be associated with poor prognosis and a high possibility of tumor recurrence after resection or liver transplantation of HCC [30–32]. In the present study, those solitary HCCs with macroscopic vascular invasion were excluded, thus resulting in relatively better OS and RFS rates than other studies [33–36].

On the other hand, among those solitary HCCs without macroscopic vascular invasion, the occurrence rate of microvascular invasion was 45.5 %. Generally, as the tumors grow in size, progression of their clinicopathological characters usually follows, such as the occurrence of microvascular invasion, multiplicity, satellite nodules, macroscopic vascular invasion, and even distant metastasis. In contrast, it is a sign of ideal clinicopathological characters if a solitary tumor grows only in size without those bad features such as multiplicity, macroscopic vascular invasion, and distant metastasis, which partly explains the more satisfying prognosis after curative resection. Therefore, it is easy to understand that microvascular invasion, instead of tumor size alone, is an independent prognostic factor of OS and RFS after curative resection for these patients with solitary HCCs.

In the present study, a resection margin <1 cm was identified as a risk factor of OS and RFS. A previously reported prospective randomized controlled trial revealed that a resection margin aiming grossly at 2 cm efficaciously and safely decreased postoperative recurrence rates and improved survival outcomes when compared with a gross resection margin aiming at 1 cm for solitary HCC [35]. Our study also supported the significance of resection margin status in long-term prognosis of solitary HCC after curative resection. Therefore, attaining a wide enough resection margin is important. On the other hand, the surgeon has to preserve enough volume of the hepatic remnant, which is relatively more difficult for a very large HCC.

Anatomical resection is the preferred procedure in our center, but non-anatomical resection is used for tumors situated at the junction of several liver segments, or for small and peripherally located tumors, or in patients with a high extent of cirrhosis. In the present study, there were no significant differences in OS and RFS between anatomical resection and non-anatomical resection, which have also been reported by Tanaka et al. [37] and Kaibori et al. [38]. However, anatomical resections are more likely to give better outcomes than non-anatomical resections for small solitary HCC [39], T1–T2 HCC [40], and non-cirrhotic patients [41].

There are some limitations to this study. First, this is a retrospective, single-center study; thus, the results may not be generalized. A multi-center prospective study may have to be performed to validate our results. Second, 90 % of HCC patients in the current study had hepatitis B virus infection. This feature is apparently different from the Western countries, where hepatitis C virus infection is the most prevailing etiology. Third, although there are patients with tumor sizes larger than 10 cm, which bring higher tendencies for recurrence, no further cut-off points were set because the results would not be convincing due to insufficient patient volume.

In conclusion, this study demonstrated that tumor size was not an independent predictor of long-term survival and recurrence after curative resection of solitary HCC, and there was no size limit to preclude hepatic resection for solitary HCC without macroscopic vascular invasion, provided the tumor is resectable.