Perioperative and long-term outcomes of liver resection for hepatitis B virus-related hepatocellular carcinoma without versus with hepatic inflow occlusion: study protocol for a prospective randomized controlled trial

From January 2016 to November 2017, all patients with HBV-related HCC scheduled for elective partial hepatectomy in the Department of Hepatobiliary Surgery, Chinese People’s Liberation Army General Hospital (PLAGH), will be screened for eligibility for this trial.

Our retrospective data for 17 patients with HBV-related HCC undergoing liver resection with hepatic inflow occlusion (Pringle maneuver) and 22 patients without inflow occlusion showed that the mean ± SD values of serum TBil on POD 5 were 19.2 ± 6.9 μmol/L in the occlusion group and 15.0 ± 5.3 μmol/L in the nonocclusion group (J. C., unpublished data). The sample size for the prospective trial is estimated on the basis of an expectation of a 20 % reduction in serum TBil on POD 5 [26, 28, 29]. On the basis of a two-samples t test with a statistical difference of 0.05 and a power of 1 − β = 0.90, a sample size of 57 patients in each randomization arm will be required. Assuming an estimated withdrawal, protocol violation, and loss to follow-up rate of 15 % during the trial, a total number of 134 patients need to be enrolled. The time frame of patient enrollment will be approximately 24 months (Fig. 1).

The two-sided null hypothesis for the primary endpoint states that both interventions (liver resection with versus without hepatic inflow occlusion) lead to similar serum levels of TBil on POD 5. The alternative hypothesis states that one intervention performs better with lower TBil on POD 5 than the other. This null hypothesis will be tested by an analysis of covariance that adjusts for the baseline measures (e.g., presence of cirrhosis, preoperative TBil, and remnant liver volume estimated using CT). We will adjust for the differences in background characteristics by including them in a multivariable regression model. All background characteristics and surgical outcome measures will be presented as mean ± SD or median with range for the continuous variables, and as frequency or percentage for the categorical variables. A two-sided Student’s t test will be used to compare continuous parametric variables such as the serum TBil on POD 5, and a Mann–Whitney U test will be used to compare discrete or nonparametric variables, as appropriate. Categorical variables will be compared using the chi-square test or Fisher’s exact test. The homogeneity of patients in the two treatment arms will be described by comparing the demographic and baseline values. The serum parameters at serial time points will be analyzed using repeated-measures analysis of variance. The Kaplan-Meier method with log-rank test will be used for survival analysis. The effect of several risk factors on survival will be analyzed using Cox regression (or proportional hazards regression). A P value <0.05 will be considered statistically significant. All statistics will be processed using IBM SPSS version 22.0 software (IBM, Armonk, NY, USA) A fully specified statistical analysis plan will be written and approved at least before the database is unlocked.

The high prevalence of HBV infection imposes a huge burden of cirrhosis and HCC in Asia. The World Health Organization reported that the western Pacific and Southeast Asian regions have the highest rates of HBV infection worldwide. The prevalence in these areas represents >75 % of the world’s HBV carriers and accounts for >60 % of the world’s liver cancer cases [30, 31]. It has been projected that, in the next decade, approximately 80 % of the world’s HCC cases will be in Asia [32]. RCTs on the Chinese population with HBV-related HCC will therefore have a greater clinical impact. We are in a unique position to enroll the necessary number of eligible patients.

The study is directed at only the patients with HBV-related HCC rather than all patients with HCC. The high prevalence of HBV infection in certain local areas has been taken into consideration. More importantly, restricting the underlying disease to HCC will help us to stratify the patients, reduce confounding variables, and improve the comparability of the groups. We surmise that hepatic inflow occlusion may induce a more severe hepatic IR injury and perioperative inflammatory response in patients with HBV-related HCC, owing to the underlying proinflammatory disease (hepatitis and/or cirrhosis). The results of the study may provide an answer to this concern and change the current practice for treatment of HBV-related HCC.

Various techniques of vascular occlusion have been practiced to reduce blood loss during liver resection [33]. Hepatic inflow occlusion (the Pringle maneuver) is a long-standing method used because of its simplicity and proven efficacy [34]. A major concern with this maneuver is the IR injury to the remnant liver. As a common consequence of HBV infection, liver cirrhosis is detected in most cases of HBV-related HCC [10]. It is generally recognized that the cirrhotic liver is particularly sensitive to the IR injury [11, 12]. Given the technical advances in LDLT, the Pringle maneuver could be avoided in liver resection in experienced hands [17, 18]. All these considerations imply that liver resection without vascular occlusion could be a more preferred option for patients with HBV-related HCC. The clinical implications of the outcomes of the present study may change current clinical practice. Furthermore, the hemodynamic disturbance to the tumor-bearing liver by intermittent hepatic inflow occlusion remains an oncological concern [13, 14, 35, 36]. Long-term follow-up of patients with HCC in this study will provide clinical clues and perhaps fill current gaps in knowledge.

Serum TBil has been used widely as an important parameter for determining PHLF. The ISGLS defines PHLF as an increased serum TBil on or after POD 5 or an increasing TBil (compared with values at the previous time point) if TBil is increased preoperatively (INR is an equivalent parameter in this criterion) [26]. The “50-50” criterion defines PHLF with PT >50 % (or INR >1.7) and TBil >50 μmol/L on POD 5 [37]. As a widely agreed prognostic factor, serum TBil on POD 5 is selected as the primary endpoint in the present study. In our pilot study, a 20 % difference in serum TBil on POD 5 was found between the occlusion and nonocclusion groups. Although the average TBil levels of both groups were within acceptable range (15.0 ± 5.3 μmol/L vs. 19.2 ± 6.9 μmol/L), the preliminary data indicated a potential difference in perioperative outcomes between the two groups, and its clinical significance might be reflected by other endpoints. These conclusions are based on analysis of retrospective data, which could have been influenced by some confounding variables. The future prospective RCT will provide more analyzable data.

The secondary endpoints include most perioperative and long-term parameters. We also added serum inflammatory factors, such as the IL-1, IL-2, IL-8, IL-10, TNF-α, and CRP. Hepatic IR injury results in a complex release of inflammatory mediators [38]; thus, the perioperative inflammatory response would reflect the impact of vascular occlusion and link it to the subsequent outcomes.

This is to date the first prospective RCT comparing liver resection for HBV-related HCC with versus without hepatic inflow occlusion. The results with respect to the perioperative and long-term outcomes will be analyzed and discussed upon the completion of data collection.