HabibTM 4X-assisted resection versus clamp-crush resection for hepatocellular carcinoma: a propensity-matching study

Long term outcome of ablation-assisted hepatic resection is unclear for hepatocellular carcinoma (HCC) patients. This study was scheduled to compare the outcome of Habib 4X ablation assisted resection (Habib group) with clamp-crush resection (CC group) for HCC. In this study, we retrospectively enrolled 81 patients from the Habib group and 103 patients from the CC group. Oncologic outcomes were analyzed using a propensity score matching (PSM) method. Compared with the CC group, the Habib group had higher levels of γ-glutamyltransferase (P=0.044) and albumin (P=0.001), larger tumor sizes (P=0.007), shorter operation times (P=0.001), less blood loss (P=0.005), and less blood transfusions (P=0.038). There were no significant differences in complications (P=0.310), recurrence-free survival rates (RFS, P=0.112), or overall survival rates (OS, P=0.203) between the two groups. For the 67 patient pairs selected from the PSM analysis, the Habib group had better RFS and OS (P=0.033 and P=0.014, respectively). A Cox proportional hazards analysis revealed that Habib-assisted resection was an independent factor for RFS and OS (P=0.008 and P=0.016, respectively). Furthermore, for the 42 patients with central and large tumors, the Habib group had better RFS and OS than the CC group (P=0.035 and P=0.038, respectively). However, the differences of RFS and OS (P=0.117 and P=0.126, respectively) were not significant among 92 patients with peripheral or small tumors. Hence, HabibTM 4X-assisted resection is safe and provides better survival for HCC patients, particularly those with central and large tumors.


INTRODUCTION
Hepatocellular carcinoma (HCC) is the fifth most common cancer in the world. [1] It is well recognized that hepatic resection is the primary radical treatment for HCC. [2] Recently, many new instruments, such as the TissueLink and Cavitron Ultrasonic Surgical Aspirator, have been developed to improve hepatic resections. [3] However, these techniques often require intraoperative maneuvers, including hepatic pedicle clamping, hypotensive anesthetics , which increased ischemic injury and risk of air embolism. [4] The Habib TM 4X, a newer bipolar radiofrequency ablation introduced in 2007, creates a plane of coagulative necrosis along the intended line of parenchymal transection. [5,6] The Habib TM 4X is a promising device for decreasing blood loss [7], without performing a Pringle maneuver. [8] To date, the safety and outcomes of patients who have undergone hepatic resection with the Habib TM 4X have been controversial. [9,10] To further address this issue, we conducted this retrospective study to compare the Habib TM 4X with clamp-crush in terms of complications, recurrence-free survival (RFS), and overall survival (OS) of patients. Furthermore, as a retrospective study, we used a propensity-scoring matched (PSM) model to balance the effects of variables before resection.

Patient characteristics
A total of 183 HCC patients who underwent hepatic resections were enrolled. Eighty-one (44.3%) patients were underwent resections with the Habib TM 4X (Habib group), and 102 (55.7%) patients underwent clamp-crush resection (CC group). Overall, there were 159 (86.9%) male and 23 (13.1%) female patients. The median age was 51.0 years (range: 22-79 years). During a median followup of 41.5 months (range: 4.0-79.5 months), 91 patients (49.7%) experienced recurrence, and 62 patients (33.9%) died. The 2-and 5-year OS rates were 76.9% and 52.3%, and the 2-and 5-year RFS rates were 55.4% and 43.7%, respectively.  Outcomes before matching Table 1 summarizes preoperative and operative patient characteristics. Among the Habib group, 2 patients suffered bile leakage, 1 patient suffered hemorrhage, and 1 patient suffered liver failure. These were all major complications (Clavien-Dindo grade III-IV). Among the CC group, 3 patients developed bile leakage, 3 patients suffered hemorrhage, 2 patients suffered liver failure, and 1 patient suffered a severe lung infection, which were assigned as Clavien-Dindo grade III-IV (major complications). Neither group had dead within 30 days after resection. No significant differences in postoperative complications were observed between the two groups (P = 0.310).

Outcomes after matching
As mentioned before, the two groups have different baseline levels of several important preoperative clinical factors, including GGT and albumin levels as well as tumor size. Thus, we used the PSM method to balance bias. Using PSM, 67 paired patients were selected. The variable balance in the matched cohort was markedly improved, and no significant difference was observed between the two groups with respect to preoperative demographics (Table 1 and Figure 2). For the Habib group, 1 patient suffered bile leakage, and 1 patient suffered hemorrhage. For the CC group, 2 patients suffered bile leakage, 2 patients suffered hemorrhage, and 1 patient suffered liver failure. No significant difference in complications was observed between the two groups (P = 0.437).
The associations of clinicopathological factors with RFS and OS are presented in Table 2. The 2-and 5-year RFS rates were 73.2% and 60.6%, respectively, in the Habib group, and 51.2% and 40.2%, respectively, in the CC group (P = 0.033, Figure 3A). The 2-and 5-year OS rates were 86.5% and 62.9%, respectively, in the Habib  In addition, resections of central tumors [11] and large tumors [12] were always associated with more blood loss and poor survival. Thus, we further stratified patients according to tumor location and tumor size. Central tumors were defined as previous: tumor located central segments (Couinaud's segments I, IV, V and VIII) and with a distance of ≥ 2 cm from the liver capsule. [11] For the 42 patients with central and large tumors (tumor size more than 5 cm), the Habib group had better RFS and OS (P = 0.035 and P = 0.038, respectively, Figures  4A and 4B). While, for the 92 patients with peripheral and small tumors (tumor size no more than 5 cm), two groups had similar RFS and OS (P = 0.117 and P = 0.126, respectively, Figure 4C and 4D).

DISCUSSION
Although the safety and survival of patients who undergo ablation-assisted resections has been evaluated,     [8,13] there is still lack of results regarding the oncological outcome of Habib TM 4X-assisted resections. As the best of our knowledge, this is the first study to investigate the long term oncological outcome due to the technique of resection. In the current study, patients in the Habib group had similar survival rates as those in the CC group before matching. After one-to-one PSM analysis, the Habib group was significantly associated with less blood loss, lower rates of blood transfusion, and shorter operation times, compared with the CC group. The Habib group had better RFS and OS compared to the CC group. In addition, patients with central or large tumors are better candidate for Habib TM 4X-assisted resections, compared to the clamp-crush resection. This study supports the Habib TM 4X providing safer resections and favorable survival for HCC patients.
During the past decade, there has been a significant increase in the number of liver resections. [14,15] Still, resection comes with significant risks and the high frequency of complications. [16] In our study, the rates of severe complications before and after matching were 4.9% and 3.0%, respectively, for the Habib group, which were both relatively lower than patients in the CC group and previous studies. [17,18] This indicates that Habib TM 4X-assisted resection is a safe.
CC is recognized as a standard method of liver parenchymal transection over the past decades. However, CC often lead to excessive blood loss and more blood transfusions, which further influenced the safety and survival. [19] The Habib TM 4X releases energy and produces a plane of coagulative necrosis along the intended line of parenchymal transection. [20] As this process reduce the risk of bleeding of hepatic parachymal, it thus ensured rapid hepatic parachymal division and shorted operation times. This may benefit cirrhotic patients. The rate of required blood transfusions was 12.3% (10/81), which was much lower than that for most previous studies. [21,22] Our data revealed that Habib TM 4X-assisted resection provided better survival than the traditional clamp-crush method. One of the potential reasons is the coagulative necrosis zone formed by the Habib TM 4X. This ablation zone was a tumor cell-free zone, which could kill potentially metastatic cells and decrease the risk of local recurrence. A clinical trial showed that narrow surgical margins increased intrahepatic recurrence [23]. Furthermore, a coagulative necrosis zone was made before the hepatic resection could reduce the risk of tumor cell intravasation and dissemination into the circulation. In addition, recently studies showed that after thermal ablation-mediated necrosis, tissue debris remains in the treated area, which may help active tumor-specific T cell response and increase the likelihood of tumor control. [24][25][26] Studies from the team of Curro and Habib opened an eye for liver surgeon to recognize the Habib TM 4X device, [27,28] while there are still many problems need to be discussed, including long term oncological outcome. [29] An Italian study showed that ablation-assisted resection provides a clean surgical field, but it was associated with a higher rate of complications than clamp-crushing. [9] Recently, another study showed that ablationassisted resection resulted in lower blood loss and fewer complications. [10] These two studies have controversial outcomes, and a meta-analysis was also unable to reach convincing outcomes. [30] While both of these studies had small sample sizes, lacked oncologic outcomes, and were based on monopolar ablation. [9,10] We surmised that the HabibTM 4X, a bipolar ablation device, produce a better coagulative necrosis zone than with monopolar ablation. [31] It is interesting to note that the Habib TM 4X may provide favorable survival rates for central and large tumors. Resection for central and large tumors are always associated more blood loss and longer operation time. The advantage of Habib TM 4X-assisted resection is more obviously for central and large tumors. While, the benefits of Habib TM 4X over clamp-crush was unclear for peripheral or small tumors Thus, patients with central and large tumors were good candidate for Habib TM 4X-assisted resection.
This study had limitations. Thermal damage from ablation may injure the neighboring vessels. Thus, the Habib TM 4X requires skilled surgical techniques. Use of ultrasound guidance greatly decreases the risk of injury to large vessels. The Habib TM 4X device should be operated by experienced surgeons. In our study, all resections were performed by Haiying Liu, who has performed at least 80 hepatic resection per year. Furthermore, with increased availability and practice of laparoscopy, laparoscopic HabibTM 4X procedures could greatly control blood loss. [20,32] We have performed several hepatic resections with laparoscopic Habib TM 4X. A large sample size study could provide more specific information on laproscopic Habib TM 4X in the near future. In addition, a comparative study of Habib TM 4X with other surgical tools, such as bipolar scissors and LigaSure diathermy, may broaden our understanding of Habib TM 4X and increase hepatic resection device options. Finally, it should be noted that this is a retrospective and single-institution study. A prospective, multi-center study is needed to validate our findings.

CONCLUSIONS
In conclusion, Habib TM 4X-assisted hepatic resection provides safety and survival benefits for patients with HCC who undergo hepatic resections, particularly for those with central and large tumors. www.impactjournals.com/oncotarget

Study population
Between January 2009 and December 2013, 183 patients underwent hepatic resections at Guangzhou Medical University Cancer Center. This retrospective study was approved by the Institutional Review Board at Guangzhou Medical University Cancer Center. All the patients were diagnosed with HCC. None had undergone liver transplantation, ablation, trans-arterial chemoembolization (TACE), or radiotherapy before resection.

Operative techniques
Patients underwent liver resections with general anesthesia. Intraoperative bimanual liver palpation and ultrasonography were then performed to confirm tumor and major vessel locations. If a resection plane was close to a secondary major vessel, resection was performed using the clamp-crush technique (CC group), which divides and seals the hepatic duct. If tumors were distal ( ≥ 1 cm) to the secondary major or major vessels, the Habib TM 4X device was used to develop a coagulation plane and create a resection line with an argon diathermy 1-2 cm from the edge of the tumor (Habib group). For each application, 100 W were delivered, and the procedure took 5-15 s. According to the thickness and vascular distribution of the liver tissue, we repeated the above steps until a fully ablated zone of desiccation was created. The number of ablations required to obtain a zone of necrosis depended on the depth of the liver parenchyma to be resected. This zone of desiccation was related to the size of the cut resection margin surface.
Once ablation was completed, a scalpel was used to divide the parenchyma between the pair of needles. Following that, we ligated the large blood vessels and bile ducts with diameters larger than 7 mm, leaving only the coagulated liver parenchyma behind. These processes were repeated until the entire tumor was removed.

Follow-up
In this study, the Clavien-Dindo classification was used to accurately and objectively grade the severity of postoperative complications. [33,34] Patients underwent contrast-enhanced CT or magnetic resonance imaging (MRI) of the abdomen and chest x-ray every 3 months for the first two years after resection and every 5-6 months subsequently for follow-up. Serum alphafetoprotein (AFP) levels and liver function were tested simultaneously. Recurrence diagnoses were made based on imaging alone if the tumor displayed typical enhancement characteristics. Extrahepatic tumors or those with atypical imaging characteristics were biopsied to confirm HCC. The data in this study were censored on February 1, 2016. Whenever possible, salvage treatment was administered to patients with recurrence or metastases. Repeat resections and ablations were the primary treatment choices for patients with solitary lesions or up to 3 lesions in total. Other non-radical treatments included transarterial chemoembolization (TACE).

Statistical analysis
Statistical analyses were performed using SPSS 20.0 (IBM, New York, NY, USA) and R version 2.12.1. OS was defined as the time from the date of resection to the date of death or the last follow-up. RFS was defined as the time from the date of resection to the time of recurrence, metastasis, or the last follow-up. RFS and OS rates were generated using the Kaplan-Meier method with log rank analyses. The Cox regression model was built using a stepwise variable selection.
PSM analysis was performed as described in our previous study [35]. Age, sex, tumor size, tumor number, tumor location, portal vein tumor thrombosis (PVTT), levels of alanine aminotransferase (ALT), aspartate aminotransferase (AST), γ-glutamyltransferase (GGT), albumin, total bilirubin, and ascites were the variables analyzed. Subsequently, 1:1 matching between the CC group and the Habib group was performed using the nearest neighbor matching. Once patients were matched, conditional logistic regression was used to compare survival. All P value calculations were 2-sided, and P < 0.05 was considered statistically significant.

CONFLICTS OF INTEREST
None.