Introduction

Hepatocellular carcinoma (HCC) is the cause of the second most common cancer-related death worldwide. It is especially common in Southeast Asia, and its incidence is rising in Western countries [1, 2]. Surgical resection, local ablative therapies, and liver transplantation are potentially curative treatment modalities in selected patients. Liver resection has now been established as the first-line therapy for HCC. Unfortunately, the postoperative recurrence rate is high, with a 3-year recurrence rate of up to 60 % [3]. Intermittent hepatic pedicle clamping (HPC) is commonly used to reduce blood loss during hepatectomy. Recent animal studies showed that hepatic ischemia–reperfusion (I/R) injury due to hilar vascular clamping caused accelerated tumor growth, stimulated tumor cell adhesion, and promoted metastases [4, 5]. The underlying mechanism is that I/R injury causes cellular damage by inducing free-radical formation, upregulating inflammatory cytokines, dysregulating mitochondrial calcium handling, and upregulating matrix metalloproteinases. These events have been shown to promote metastases. However, in humans, three recent retrospective studies showed no evidence that HPC adversely affected long-term overall and disease-free survival after hepatectomy for colorectal liver metastases [6–8]. At present, it is still unknown whether I/R injury in general, and HPC in particular, affect recurrence and prognosis after hepatic resection for HCC [9, 10], especially in patients with cirrhosis. The answers to these questions will determine whether HPC should be used.

In this study we evaluated the effect of intermittent Pringle’s maneuver on the overall and disease-free survivals of HCC patients after curative resection, taking into account the diversity of patient subpopulations. Matrix metalloproteinases (MMPs) and E-selectin were measured to study hepatic I/R injury.

Patients and methods

Between January 2001 and December 2006, a total of 565 consecutive patients underwent hepatic resection for HCC in our institute. The study was approved by the local review board, and written informed consent was obtained from all patients before treatment. The clinicopathological features and serum/tissue samples of 386 patients were collected prospectively in a specially designed computerized database. Patients who satisfied the following criteria were entered into this study: (1) histologically proven hepatocellular carcinoma with curative resection. The curative nature of the resection was established by a clear tumor resection margin (>0 cm) on postoperative histology and by a negative chest radiograph, ultrasonography, triphasic abdominal contrast-enhanced spiral computed tomography (CT) scans, and abdominal magnetic resonance imaging (MRI) carried out within 2 months after surgery. (2) Either HPC was not used or the total HPC time was more than 30 min. (3) There was no evidence of residual lymph node metastases. Sixty-nine patients were excluded from this study for the following reasons: liver resection had been carried out for recurrent or spontaneously ruptured HCC, the patients dropped out for follow-up within 1 month of the operation, or HPC was done but the occlusion time was >30 min. The remaining 386 patients were divided into two groups: HPC (n = 224) and non-HPC (n = 162).

Preoperative management

All patients followed the same preoperative evaluation protocol, which included blood biochemistry, percutaneous ultrasonography, spiral CT and CT angiography of the abdomen, chest X-ray, and electrocardiogram (ECG). Liver function was assessed by the Child-Pugh grading, an oral glucose tolerance test (OGTT), and/or an indocyanine green clearance test (ICG-R15). Major liver resections were carried out when ICG-R15 was less than 14 %, as we have previously reported [11, 12].

Surgical management

Liver resection included nonanatomical resection, subsegmentectomy, segmentectomy, hemihepatectomy, and trisecmentectomy. A major hepatectomy is defined as resection of three or more of Couinaud’s segments.

Surgery was performed as we have previously reported [11, 12]. The operation started with a right subcostal incision, or a midline incision with a right horizontal extension. A low central venous pressure was routinely used. If inflow occlusion was used, it was achieved by tightening a 4-mm tape around the portal triad. Intermittent HPC consisted of cycles of occlusion/unocclusion of 15/5 min. Liver transection was carried out using a Kelly clamp or a CUSA (Cavitron Ultrasonic Surgical Aspirator System 200; Valleylab Inc., Boulder, CO, USA) with a 23-kHz standard tip, 70 W, 4 mL/min flush, and sensitivity of 100 %. A TissueLink (Force FX-8C; Valleylab, Tyco Healthcare, Boulder, CO, USA) was used to coagulate the vessels and bile ducts, using an output power at 70 W. The radiofrequency (480 kHz) energy was focused at the tips of the instrument through a channel inside the device and conveyed to the liver tissue by a low flow of saline solution (one drop per second) to induce thermocoagulation. HPC was not routinely started at the beginning of liver transection but only when significant bleeding occurred in the operative field. Blood transfusion was given only when the hemoglobin dropped below 8.0 g/L.

Postoperative management

All patients received the same postoperative care by the same team of surgeons, as reported previously [11, 12]. After discharge from hospital and on follow-up, the patients were monitored by ultrasonography, contrast CT scan, or MRI every 2 months in the first year and every 3 months thereafter. Recurrence was defined as the appearance of a new lesion with radiologic features typical of HCC on two or more imaging modalities. Disease-free survival was defined as the interval between the operation and the date of diagnosis of the first recurrence. Overall recurrence was defined as the interval between the operation and the date of death or the last date of follow-up.

Serum MMP-2, MMP-9, and E-selectin levels

Human blood samples were collected at three time points: before ischemia and 24 and 48 h after initiation of reperfusion. The samples were centrifuged at 1,500×g for 15 min. Total serum MMP-2, MMP-9, and E-selectin protein concentrations were determined using highly specific enzyme-linked immunosorbent assays (ELISA) (R&B Systems). All samples were analyzed in triplicate. Standard curves were generated according to the manufacturer’s instructions.

Quantitative real-time PCR analysis for mRNA expression

Liver tissue samples were collected before liver transection (the Control group) and at the end of the operation (the Liver I/R group). Total RNA was extracted from the liver specimens with the use of Trizol (Invitrogen, Carlsbad, CA, USA) following the manufacturer’s instructions. First-strand cDNA synthesis and amplification were performed according to the protocol of the PrimeScript™ RT Reagent Kit (perfect real-time) (TaKaRa, Shiga, Japan). Quantitative determination of gene expression levels was done on a Bio-Rad CFX96 Real Time PCR System (Bio-Rad, Hercules, CA, USA), with the use of the SYBR® premix Ex Taq (TaKaRa).

Western blotting analysis

Liver tissue samples were collected before liver transection (the Control group) and at the end of the operation (the Liver I/R group). Total proteins were separated using sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) on a 10 % gel. Rainbow molecular weight markers were used to visualize protein migration. When electrophoresis was over, the specimen was then transferred onto a 0.45-μm nitrocellulose membrane (Millipore, Billerica, MA, USA) in a buffer containing 25 mmol/L Tris-HCl (pH 8.3), 192 mmol/L glycine, and 20 % methanol and blocked with 5 % fat-free dry milk in PBS for 2–4 h. The membranes were incubated with primary antibodies anti-MMP-2 and anti-GAPDH, followed by the addition of a secondary antibody anti-rabbit IgG or anti-mouse IgG (Santa Cruz Biotechnology, Santa Cruz, CA, USA). The bound antibodies were detected using a chemofluorescence detection kit (Amersham, Piscataway, NJ, USA) following the manufacturer’s instructions.

Statistical analysis

Statistical analysis was performed using the χ2 test or the Fisher exact test to compare discrete variables, and Student’s t test was used to compare continuous variables. The survival curves of the two groups, including the cumulative survival and the disease-free survival, were generated by the Kaplan–Meier method and compared using log-rank tests. Multivariate analysis for 17 variables in Table 1 using the Cox proportional hazards regression model was used to identify independent prognostic factors in predicting disease-free survival and overall cumulative survival. A p value <0.05 was considered to be significant. Statistical analyses were performed using the SPSS 10.0 for Windows (SPSS Inc., Chicago, IL, USA).

Table 1 Characteristics of patients who underwent hepatic resection with or without intermittent HPC
Full size table

Results

Perioperative features of HCC patients

The background characteristics in the HPC and non-HPC groups are given in Table 1. Most patients (75.9 % of 386) were male. The platelet count in the HPC group was significantly lower than in the non-HPC group (p = 0.046). The number of patients with multiple tumors was also significantly higher in the HPC group than the non-HPC group (p = 0.034). On the other hand, gender, ICG-R15, total bilirubin, AST, hyaluronic acid, tumor size, portal vein invasion, and other variables showed no significant difference between the two groups. Table 2 gives the results of surgery in the two groups. The postoperative AST on day 3 was significantly lower in the HPC group than in the non-HPC group (p = 0.042). However, there was no significant difference in total bilirubin on postoperative day 3 between the two groups. The type of resection, intraoperative blood loss, red cell transfusion, and morbidity were similar between the two groups. The median of the total occlusion time of intermittent HPC was 50 min (range = 30–98 min).

Table 2 Surgical outcomes
Full size table

Analysis of prognosis of HCC patients after hepatectomy

At the time of censor of the study, 257 of 386 patients (66.6 %) had developed tumor recurrence. The mean time to recurrence was 26.1 months (median = 23; range = 3–111). Of the 257 patients with recurrence, 159 patients (61.9 %) developed hepatic recurrence only, 55 patients (21.4 %) extrahepatic recurrence only, and 43 patients (16.7 %) both intra- and extrahepatic recurrence. Overall, hepatic recurrence occurred in 202 patients (78.6 %). The overall recurrence rate and the liver-specific recurrence rate were not significantly different between the two groups (Table 3).

Table 3 Type of recurrence
Full size table

Multivariate analysis identified only one significant factor (cirrhosis) for overall survival rate and five significant factors (tumor size ≥5 cm, cirrhosis, Child–Pugh grade, tumor number, and portal vein invasion) for disease-free survival rate (Table 4). Cirrhosis was an independent factor for both overall and disease-free survival: the hazard ratios with 95 % confidence intervals and the p values were 1.54, 1.13–2.09, p = 0.009, and 0.64, 0.44–0.96, p = 0.03, respectively (Table 4).

Table 4 Multivariate analysis: predictors for disease-free survival and overall survival
Full size table

The overall and disease-free survival curves of the HPC and non-HPC groups are shown in Fig. 1a, b, respectively. Overall survival was similar in the two groups: the 1-, 3-, and 5-year survival rates were 95.6, 72.3, and 45.1 % in the non-HPC group vs. 86.4, 70.9, and 43.9 % in the HPC group (p = 0.146). The disease-free survival rates at 1, 3, and 5 years were 77.9, 31.7, and 13.8 % in the non-HPC group vs. 62.4, 44.5, and 17.9 % in the HPC group (p = 0.554). Similarly, there were no significant differences in disease-free survival (cirrhotic and noncirrhotic, p = 0.103, 0.351) and overall survival (cirrhotic and noncirrhotic, p = 0.275, 0.691) in patients with and without cirrhosis between the two groups (Fig. 2a–d). However, the postoperative long-term survivals for all of the patients were significantly better in the noncirrhotic group than in the cirrhotic group in terms of the overall and disease-free survival rates (p = 0.004 and 0.016, respectively). The overall 1-, 3-, and 5-year survival rates were 91.9, 76.8, and 57.8 % in the noncirrhotic group and 89.8, 70.2, and 40.9 % in the cirrhotic group, respectively. The 1-, 3-, and 5-year disease-free survival rates were 79.2, 51.7, and 25.4 % in the noncirrhotic group and 73.6, 36.4, and 14.8 % in the cirrhotic group, respectively (Fig. 3a, b).

Fig. 1
figure 1

Disease-free survival (a) and overall survival (b) for all patients (Kaplan–Meier, log-rank test)

Full size image
Fig. 2
figure 2

Disease-free survival (a, b) and overall survival (c, d) for the patients divided into two subsets according to the presence of liver cirrhosis (Kaplan–Meier, log-rank test)

Full size image
Fig. 3
figure 3

Disease-free survival (a) and overall survival (b) for noncirrhotic and cirrhotic patients (Kaplan–Meier, log-rank test)

Full size image

Gene expression of MMP-2, MMP-9, and E-selectin

As shown in Fig. 4, the serum levels of MMP-2, MMP-9, and E-selectin were similar before surgery and 24 and 48 h after surgery. Hepatic expressions of MMP-2, MMP-9, and E-selectin were measured by real-time PCR in nontumorous liver tissues obtained from HCC patients before HPC and after HPC. There was no significant difference in the hepatic expressions of MMP-2, MMP-9, and E-selectin before and after occlusion (Fig. 5). The hepatic expression of MMP-2 was further analyzed at protein level by western blot analysis. The protein level of MMP-2 was similar in liver tissues before and after HPC (Fig. 6). Our data indicated that there was no upregulation of MMP-2, MMP-9, and E-selectin in patients who underwent hepatectomy with HPC.

Fig. 4
figure 4

Serum concentrations of E-selectin (a), MMP-2 (b), and MMP-9 (c) before and after surgery. Human blood samples were collected before surgery and 24 and 48 h after surgery. Serum concentrations of E-selectin, MMP-2, and MMP-9 were analyzed by enzyme-linked immunosorbent assay. The data are presented as mean ± SD (n = 14). No significant differences were observed at any time point

Full size image
Fig. 5
figure 5

Gene expression of E-selectin, MMP-2, and MMP-9 in nontumor livers before and after surgery. Nontumor liver samples were collected before and after hepatic pedicle clamping during operation. The mRNA levels of E-selectin (a), MMP-2 (b), and MMP-9 (c) in liver tissues were measured by real-time PCR. The data are presented as mean ± SD (n = 15)

Full size image
Fig. 6
figure 6

Protein level of MMP-2 in nontumor liver tissue samples before and after surgery. Nontumor liver samples were collected before and after hepatic pedicle clamping during surgery. The protein level of MMP-2 in liver tissues was measured by western blot

Full size image

Discussion

The present study showed that HPC did not affect overall and disease-free survival following curative liver resection for HCC. It also failed to show any significant rise in inflammatory mediators induced by hepatic ischemia–reperfusion injury caused by HPC using cycles of occlusion/unocclusion of 15/5 min.

Experimental and clinical studies have shown that even short clamping times produced some degree of ischemia–reperfusion injury, which resulted in hepatocellular damage and postoperative liver dysfunction. Hepatic I/R injury causes cellular damage by inducing free-radical formation, upregulating inflammatory cytokines, dysregulating mitochondrial calcium handling, and upregulating matrix metalloproteinases (MMPs) [13, 14]. Some experimental studies reported that ischemia–reperfusion injury stimulated outgrowth of pre-established micrometastases and consequently adversely influenced prognosis in patients with colorectal cancer. Yoshida et al. [15] reported that sustained ischemia and reperfusion in the liver accelerated growth of established hepatic metastases, and minimization of I/R injury through intermittent ischemia attenuated metastasis of colorectal cancer to the liver following surgical stress.

However, recent clinical studies showed that HPC did not affect survival of patients with liver metastases. Wong et al. [6] reported that there was no difference in the median disease-free survival after hepatectomy for colorectal metastasis (CRLM) between patients who received intermittent HPC and those who did not (21.1 vs. 19.9 months; p = 0.199). Ferrero et al. [7] showed that survival of patients with liver metastases who underwent hepatectomy with (n = 39 patients, the HPC group) or without (n = 41 patients, the NHPC group) pedicle clamping was the same, i.e., HPC did not affect survival. Similarly, a recent bi-institutional study showed that HPC did not affect long-term survival and liver-specific disease-free survival following hepatectomy for CRLM analyzed according to the use, type, and duration of HPC [8].

The other concerns of HPC are its safety and effectiveness in the early results of liver resections [16]. Makuuchi et al. [17] reported that HPC can be used for the donor hepatectomy in living-donor liver transplantation without inducing any ischemic injury attributable to the technique itself. Furthermore, the incidence of biliary complications and the 1-year survival rates were not influenced by whether or not the graft was harvested using HPC.

In our study, there was no difference in the postoperative total bilirubin on day 3, although postoperative AST was higher in the nonocclusion group, which was attributed to the more extensive coagulation of liver parenchyma by the TissueLink. There was also no significant difference in intraoperative blood loss and blood transfusion between the two groups. Thus, it cannot be argued that the possible negative effect was compensated by a decrease in intraoperative blood loss when HPC was used. It is well known that an increased intraoperative blood loss is associated with an increase in postoperative complications and in hepatic tumor recurrence [18, 19].

Many previous studies on hepatic I/R injury demonstrated that cytokines are released through induction of adhesion molecules, and chemokines promote neutrophil activation and accumulation and thereby contribute to progression of parenchymal injury by releasing reactive oxygen species (ROS) and proteases. In particular, hepatic expression of MMP2 and MMP9 has been linked to hepatic I/R injury through recruitment of neutrophils and T cells [9, 10]. MMPs promote tumor growth and metastases by a variety of mechanisms, including proteolysis of extracellular matrix that allows initial migration and seeding of tumor cells [20–23]. MMP2 and MMP9 are widely implicated in the metastasis of colorectal carcinoma. Nicoud et al. [23] reported that hepatic I/R-induced elevation of MMP9 contributed to the growth of metastatic colorectal carcinoma in mouse liver and that postresection MMP9 inhibition was clinically beneficial in preventing recurrence following hepatic surgery. Similarly, E-selectin is an adhesion molecule that is transiently expressed on the surface of activated endothelial cells. It participates in the rolling of leukocytes upon the endothelial cell surface and has been implicated in the initial events of neutrophil extravasation in inflammatory response. Ligands for E-selectin, sLe-X and sLe-a, have been identified as markers of progression of cancer. The interaction between E-selectin and sLe-X or sLe-a affects tumor cell adhesion to sinusoidal endothelial cells. Doi et al. [24] indicated that hepatic I/R injury produced an increase in liver metastases of rat colon cancer with an increase in expression of E-selectin. Because a shorter duration of liver ischemia resulted in a smaller number of liver tumor nodules and a lower expression of E-selectin mRNA, shortening the ischemic period is worthwhile in liver resection carried out on patients with liver malignancy [24, 25]. However, we have not observed any significant difference in the serum concentrations and the tissue mRNA levels of E-selectin, MMP-2, and MMP-9 between the pre- and postoperative periods using HPC in human HCC patients. This study showed that hepatic changes in cytokine expression in animals did not occur in human. These data suggested that the changes in the expression of cytokines such as E-selectin, MMP-2, and MMP-9 were mild. As a consequence, the changes were not enough to lead to an increase in liver metastases or tumor recurrence in HCC patients when intermittent HPC was used in our clinical setting.

To the best of our knowledge, there have not been any clinical or experimental studies on the influence of HPC on the long-term prognosis of patients with hepatocellular carcinoma after partial hepatectomy, although there have been such studies on colorectal metastases to the liver. HPC is commonly used for liver resection and its influence on the outcome of surgery is important. In this study, although the baseline data in the two groups were not entirely comparable, the fact that there were more cases with multiple tumors in the HPC group did not influence the outcomes. Furthermore, the use of matrix metalloproteinases and E-selectin supported the conclusion that intermittent HPC was safe and efficacious. Several experimental studies showed that the degree of hepatic injury after I/R was significantly greater in cirrhotic livers than in normal livers [26]. The results from randomized controlled trials and from a meta-analysis of hepatic inflow clamping showed that chronically diseased livers are more susceptible to I/R injury than normal livers [27, 28]. Our subgroup analysis showed that HPC had no significant negative impact on the long-term prognosis of cirrhotic patients. However, a properly conducted randomized controlled trial is necessary to clarify the effects of HPC on the long-term prognosis of HCC patients after curative resection.

In conclusion, HPC was safe and efficacious in partial hepatectomy for HCC. It had no significant negative impact on disease-free survival and overall survival, even in patients with cirrhosis. Cirrhosis, however, was an independent risk factor for both the disease-free survival and overall survival for HCC patients.