Hepatectomy Provides Good Long-Term Outcomes for Hepatocellular Carcinoma Patients With Portal Hypertension: A Propensity Score Matching Analysis


 Background: The role of hepatectomy in hepatocellular carcinoma (HCC) with portal hypertension (PH) remains controversial. This study aimed to evaluate the effect of hepatectomy on overall survival (OS) of HCC patients with PH.Methods: A total of 1651 HCC initially treated with hepatectomy were retrospectively reviewed and divided into PH group (n=157) or non-PH group (n=1494). Propensity score matching (PSM) was conducted to match the baseline characteristics of the PH group and non-PH group. Results: The PH group presented a similar OS (p=0.29) and recurrence free survival (RFS) (p=0.83) compared with non-PH group after initial hepatectomy before PSM. After PSM processing, the baseline characteristics were highly comparable for both groups (133 patients in each group). The PH group also presented a similar OS (p=0.81) and RFS (p=0.65) compared with non-PH group after initial hepatectomy. After PSM, multivariate analysis identified tumor size (>5 cm) (p=0.02), macro-venous invasion (p < 0.001), AST (>37 U/L) (p =0.008) as independent risk factors for OS.Conclusions: Hepatectomy provides good long-term outcomes for HCC patients with PH. PH should not be regarded as a contraindication for hepatectomy in HCC patients.


Introduction
Primary liver cancer is the second cause of cancer-related death and hepatocellular carcinoma (HCC) accounts for 90% of cases. HCC is common in China because of the high prevalence of hepatitis B virus (HBV) infection and cirrhosis [1]. Previous studies considered hepatectomy as a contraindication for HCC patients with portal hypertension (PH) [2,3].
Recently, studies showed that HCC patients with PH can bene t from hepatectomy [4][5][6][7], while both American Association for the Study of Liver Diseases (AASLD) and the European Association for Study of Liver (EASL) guidelines haven't accepted that opinion [8,9]. Different diagnosis criteria of PH might result in inconsistent results and increase selection bias. A meta-analysis conducted in 2015 reported that hepatectomy increased the risk of mortality and clinical decompensation for HCC patients with PH (evaluated by any method) [10]. Consequently, liver transplantation and ablation was considered as the rst-line treatment for HCC patients with PH [9,10]. However, only a small amount of HCC patients with PH could receive liver transplantation due to the shortage of liver donor and the expensive cost.
In this study, we retrospectively evaluated the effect of PH on survival of HCC patients initially treated with hepatectomy. Propensity score matching (PSM) was conducted to reduce the heterogeneity between PH patients and non-PH patients and made the results more convincing.

Materials And Methods
Patients A retrospective study about HCC patients with PH was conducted and was approved by the Ethics Committee of The First A liated Hospital of Sun Yat-sen University. The procedures used in the study adhere to the tenets of the Declaration of Helsinki and its later amendments or comparable ethical standards. All enrolled patients should meet the both inclusion criteria: 1) Pathological diagnosis of HCC after hepatectomy, 2) Underwent hepatectomy as initial treatment for HCC. Exclusion criteria were: 1) Underwent trans-arterial chemoembolization (TACE)/ radiofrequency ablation (RFA)/ chemotherapy before hepatectomy, 2) Complicated with other malignant tumors, 3) Patients without follow-up. There were 1651 HCC patients initially treated with hepatectomy enrolled in the study between February 2004 and November 2014.
Most accepted gold diagnosis criteria of PH was hepatic venous pressure gradient (HVPG) over 10 mmHg [10][11][12]. The presence of gastroesophageal varices (GEV) or platelet count <100,000/mL and splenomegaly (major spleen diameter >12 cm) were considered as standard surrogate criteria to de ne PH. HVPG was not routinely used in our hospital because of invasiveness. Therefore, in this study PH was de ned when one criteria was met:1) the presence of GEV by CT/MR or endoscopy, 2) platelet count <100,000/mL and splenomegaly (major spleen diameter >12 cm).
There were two groups in the study according to PH criteria: 1) non-PH group: HCC patients without PH before hepatectomy, 2) PH group: HCC patients with PH before hepatectomy. Patients' clinicopathological variables were collected from our HCC database.

Survival Analysis After PSM Processing
After PSM processing, the baseline characteristics were highly comparable for both groups (133 patients in each group) (

Discussion
We conducted a large retrospective study based on 1651 HCC patients with su cient follow-up data to determine the effect of PH on prognosis of HCC patients initially treated with hepatectomy. And our results demonstrated that PH was not an independent risk factor for OS and RFS. This is the largest sample size to evaluate the effect of PH on the OS and RFS of HCC patients initially treated with hepatectomy as we knew.
According to the treatment guidelines of BCLC staging system [15], 2001 EASL [16], 2011 AASLD [8], and 2012 EASL-EORTC [9], PH was a contraindication for hepatectomy. Two studies in 1996 and 1999 by Barcelona group concluded that PH was a risk factor of postoperative liver decompensation and poor OS after hepatectomy [2,3]. In these two studies, PH was de ned as HPVG ≥ 10 mmHg. PSM study and proved that hepatectomy was safe and had a survival advantage over ablation in HCC patients [23].
For HCC patients with cirrhosis who underwent TACE, Nam Hee Kim et al. found that PH was a major negative factor [24]. However, Xiao et al. found that hepatectomy provided better long-term prognosis for HCC patients with PH than TACE and ablation [25]. N. Takemura et al. found that perioperative prophylactic management could enhance the safety of hepatectomy and expanded the indications of hepatectomy in patients with PH [26]. These reports strongly suggest that PH is not an absolute contraindication for hepatectomy of HCC patients and hepatectomy should be considered as rst line treatment option for HCC patients with PH in selected cases.
HVPG was not routinely performed in our hospital since its invasiveness. In this retrospective study, GEV or platelet count < 100,000/mL and splenomegaly were used to diagnose PH. Research showed that the association between PH and clinical outcome was greater when HVPG was applied to diagnose PH [10]. HVPG could directly diagnose even small increases of portal pressure [27,28]. GEV was related to PH since study shown that GEV only form when HVPG was over 10 mmHg. However, platelet count < 100,000/mL and splenomegaly are not accurate enough to diagnose PH. These results strongly suggest that HVPG should be as gold diagnosis criteria of PH. Recently a non-invasive Computed Tomography based assessment was proposed and shown a high accuracy in diagnose of PH and could replace the HVPG assessment [29]. Liver stiffness assessed by transient elastography may be another choice to diagnose PH [30].
There are some limitations in our study. First, this is a single center study and need multiple-centers studies to validate our results. Second, as a retrospective study patients with no follow-up data were excluded in our study and these patients may in uenced the results. Third, we use standard surrogate criteria to diagnose PH and may cause selection bias since HVPG is gold diagnose criteria.
In conclusion, hepatectomy provides good long-term outcomes for HCC patients with PH. Our results suggest that PH should not be regarded as a contraindication for hepatectomy of HCC patients.