Increase of Portal Vein Pressure Gradient After Hepatectomy Predicts Post-operative Liver Dysfunction

Background. Post-hepatectomy liver failure (PHLF) is an important cause of mortality and morbidity. Whether Child–Pugh A patients with varying degrees of cirrhosis are good candidates for hepatectomy is disputed. The purpose of this study was to analyze the impact of portal venous pressure gradient (PVPG) variation during surgery on PHLF. Methods. PVPG, the pressure gradient between the portal vein and central vein, was measured in consecutive patients before and after liver resection. The optimal cutoff of PVPG to predict PHLF was determined by receiver operating characteristic curve analysis. Risk factors for PHLF were subjected to univariate and multivariable analysis. Results. Sixty Child–Pugh A patients were recruited. The mean PVPG was increased from 5.17 ± 4.78 mm of mercury (mmHg) to 6.37 ± 4.44 mmHg after liver resection. The optimal cutoff value of PVPG increments to predict PHLF was 1.5 mmHg. Multivariable analysis showed prothrombin time (PT), post-hepatectomy PVPG increments of 1.5 mmHg or greater, and resected liver segments of 3 or more to be independent predictors of PHLF. Conclusions. Acute PVPG increase after hepatectomy is associated with a higher risk of PHLF in Child–Pugh A patients.


Background
Hepatectomy is currently the main treatment for focal liver diseases in patients with Child-Pugh A cirrhosis. 1 Despite progress in surgical techniques over the past decades, post-hepatectomy liver failure (PHLF) is still an important cause of mortality and morbidity, especially in patients with liver cirrhosis. 2,3 Child-Pugh A patients have various degrees of cirrhosis, which are associated with different levels of risk for PHLF. 4 Thus, it is of great clinical value to identify the predictive factors of PHLF in patients with Child-Pugh A cirrhosis.
Portal hypertension (PHT) is one of the main features of cirrhosis. 5 PHT in patients that have undergone hepatectomy is associated with a higher risk of PHLF. 6 The Barcelona Clinic Liver Cancer classification guidelines for hepatocellular carcinoma (HCC) suggest that a hepatic venous pressure gradient (HVPG) greater than 10 mmHg is a contraindication for hepatectomy 1,6 ; however, many doctors do not agree that PHT precludes hepatectomy in all cases. 7 Several studies have shown that increased portal vein pressure (PVP) affects postoperative safety in patients who have undergone hepatectomy. [8][9][10][11] However, very few studies have measured the gradient between caval pressure and portal pressure. Many studies have demonstrated that increases of central venous pressure (CVP) are transmitted to PVP. [12][13][14][15][16][17][18] Thus, measurement of PVP could be affected by CVP. Portal vein pressure gradient (PVPG) is the gradient between PVP and CVP. Conducted in Child-Pugh A patients with future liver remnant (FLR) >30% of total liver volume, this prospective study is aimed at exploring the predictive value of acute PVPG change during surgery on postoperative liver dysfunction and to further enhance the safety of liver resection.

Patients
From April 2018 to March 2019, patients who meet the following criteria are eligible for this study: (1) patient age ≥18 and ≤75; (2) preoperative liver function of Child-Pugh A; (3) major resection (defined by removal of 3 or more liver segments) planned or liver stiffness (LS) higher than 12 kilopascals (kPa) measured before hepatectomy (our previous study showed that patients with LS higher than 12 kPa were more likely to be associated with higher risk of PHLF 19 ); and (4) if a major hepatectomy is planned, liver volumetry on CT scan was conducted to confirm that the patients had an FLR of >30% of total liver volume.
Clinically significant portal hypertension was defined according to the Sixth Baveno Consensus on Portal Hypertension. 20 The spleen diameter was defined as its longest axis. Liver fibrosis was staged using the Scheuer classification. 21 PHLF was defined as postoperative deteriorated liver function with an increase in the international normalized ratio and concomitant hyperbilirubinemia on or after postoperative day 5, as proposed by the International Study Group of Liver Surgery (ISGLS). 22 Postoperative complications were recorded according to the modified Clavien-Dindo classification. 23 Perioperative mortality was defined as 90-day mortality.
Written consent was obtained from all patients. Ethics approval and consent to participate this prospective study were approved by the Ethics Committee of Zhongshan Hospital, Fudan University. All methods were conducted in accordance with the approved guidelines.

Surgical Procedures and PVPG Measurement
The procedure for hepatectomy has been previously described. 24 Briefly, hepatectomy was conducted using a clamp crushing method with cycles of porta hepatis clamping/unclamping of 15/5 minutes. The extent of liver resection was defined according to the number of Couinaud liver segments removed. Major hepatectomy was defined as resection of 3 liver segments or more; minor hepatectomy was defined as fewer than 3 segments. 25 A central venous catheter was inserted into the jugular or subclavian vein before operation. After exploration of the abdomen and decision to proceed with the hepatectomy, a 16-Fr polyvinyl chloride catheter was inserted into the main portal vein through the right gastroepiploic vein, with all air purged from the extension tubing beforehand.
The position of the tip of the tube was verified by intraoperative ultrasonography. PVP was measured electronically using a transducer fixed at the level of the right atrium and set to atmospheric. The readings were standardized to zero when the transducer was connected to the atmosphere. The PVP and CVP measurements were displayed on a monitor (Dräger, Infinity C700). PVP and CVP were recorded before liver resection and 15 minutes after hepatectomy when hemodynamics were stable. All measurements were performed twice. In cases where the difference of 2 measurement values is greater than 1 mmHg, a third measurement was performed. PVPG was defined as the pressure gradient between the portal vein and the central vein.

Postoperative Management and Follow-up
Routine blood, liver, and coagulation function tests and serum electrolyte detection were performed once every other day during the first postoperative week. Patients were followed up every 2 months for 1 year after discharge from hospital and then every 3 months thereafter.

Statistical Analysis
For PVP and PVPG data, delta values (ΔPVP and ΔPVPG) were generated by subtraction of posthepatectomy and pre-hepatectomy values. Continuous variables are reported as mean (standard deviation, SD) or median (interquartile range, IQR) and were compared using Student's t-test or Mann-Whitney U test as appropriate. Categorical variables are reported as numbers and percentages and compared using Pearson's χ 2 analysis or Fisher's exact test. Variables shown to be statistically significant in univariate analysis (P < .05) were included in multivariable logistic regression analysis to identify independent predictors of PHLF. Receiver operating characteristic (ROC) curve analysis was used to determine the optimal cutoff for PVPG variation predicting PHLF. Calculations were done with SPSS 22.0 (IBM, New York, NY, USA). Graphical illustrations were carried out with GraphPad Prism v7 (GraphPad Software, Inc., La Jolla, CA, USA). A P-value less than .05 indicates statistical significance.

Patient Demographics
A total of 60 eligible patients were recruited for the study from April 2018 to March 2019. The characteristics of the patients are listed in Table 1. There were 54 men (90%) and 6 women (10%), with a mean age of 57.5 years. Two patients were diagnosed with hepatic cyst, 48 patients were diagnosed with HCC, 4 patients were diagnosed with intrahepatic cholangiocarcinoma (ICC), and 6 patients were diagnosed with mixed type (HCC-ICC). None of the patients received preoperative chemotherapy, radiotherapy, or interventional treatment. When patients were grouped by a binary classification of PVPG increment (illustrated by the ROC curve in Figure 1), all variables were similar between the two groups except for prehepatectomy PVPG.
The classifications of PVP (≤/>12 mmHg) before hepatectomy were set referring to Chen's work. 9 In the pre-hepatectomy PVP >12 mmHg group, PVP decreased after hepatectomy, while PVPG remained stable ( Figure 2B). While there were significant increments of PVPG in the patients with major resection or PHLF, no remarkable changes of PVP were seen ( Figures 2D and F).

PVPG Increment Cutoff Before and After Hepatectomy
The ROC curve shows that a ΔPVPG ≥1.5 mmHg is the optimal cutoff for the occurrence of PHLF, with a sensitivity of 65% and specificity of 70% (AUC = .67, P = .038) as shown in Figure 1. Thus, we adopt a binary classification of "ΔPVPG </≥1.5 mmHg."

Dynamic Changes of Liver Function, PHLF, and Complications
Liver function was tested before and once every 2 days after the operation. Worse liver function was observed in patients in the "ΔPVPG ≥1.5 mmHg" group compared with the "ΔPVPG <1.5 mmHg" group ( Figure 3). Of the 60 patients, 13 (21.7%) developed one or more postoperative complications including subphrenic effusion in 2 patients (3.3%), pleural effusion in 2 patients (3.3%), ascites in 2 patients (3.3%), and wound infection in 1 patient (1.7%). Using the modified Clavien-Dindo classification, complications were categorized as grade I in 3 patients (5.0%), grade II in 6 patients (10.0%), grade IIIa in 4 patients (6.7%), and grade IV and grade V in 0 patients. Due to the low complication rate, grades 0-I and grades II-III were merged, respectively. The complication rate for the "ΔPVPG ≥1.5 mmHg" group was significantly higher than that of the "ΔPVPG <1.5 mmHg" group (Table 2).
Of the 60 patients, 17 (28.3%) developed PHLF, all of whom were defined as PHLF grade A according to ISGLS criteria. A PVPG increment ≥1.5 mmHg was associated with a greater risk of PHLF (P = .014) than a PVPG increase <1.5 mmHg (Table 2). There was no mortality within 3 months after hepatectomy.

Discussion
The present study demonstrated that the acute increase of PVPG after hepatectomy (ΔPVPG) could predict PHLF in patients with Child-Pugh A cirrhosis. To our knowledge, this is the first report to focus on acute changes of PVPG after hepatectomy and show a clear correlation between ΔPVPG and PHLF after liver resection. More importantly, ΔPVPG may be used as an indicator that will help to modify surgical or medical treatments, for example, splenectomy or somatostatin 26 or terlipressin, 27 to reduce portal pressure and prevent PHLF.
Inadequate volume of FLR is the leading adjustable predictor of PHLF. 28 Thus, the evaluation of functional FLR is extremely vital before liver resection. In our study, liver volumetry on CT scan was conducted to confirm that the patient had an FLR of >30% of total liver volume before major hepatectomy.
In the present study, we focused on the acute change of PVPG after liver resection and its effect on PHLF. The results showed that major liver resection and ΔPVPG ≥1.5 mmHg were independent risk factors for PHLF, which supports the pathophysiological mechanism of PHLF. 3,29 Major hepatectomy usually resulted in an acute increase of PVPG, which led to portal hyperfusion and compensatory decrease in hepatic arterial inflow (hepatic arterial buffer response 30 ), followed by sinusoidal endothelial cell injury, then activated sinusoidal endothelial cells generating increased reactive oxygen species and cytokines, causing the etiology of PHLF. There is another mechanism involved in PHT-related liver dysfunction: acute PVPG increase could induce excessive regeneration factors, leading to hyperplasia of hepatocyte without the support of an organized sinusoidal network, ultimately leading to liver dysfunction. 31 Many techniques have been applied to evaluate liver function reserve and liver injury before hepatectomy, including Child-Pugh classification, indocyanine green clearance test, HVPG measurement, and so on. 29 However, these means could not reflect the immediate changes in the liver's hemodynamic during operation, which is critically involved in the development of PHLF. 11 By contrast, portal pressure gradient could show the dynamic changes of portal pressure during surgery and serve as an immediate indicator for potential portal pressure adjustment.
The present study suggested PVPG to be a better indicator of PHT than the pure PVP value. PHT is pathologic increase in the pressure gradient between the portal vein and the inferior vena cava. 32 Previous clinical studies have demonstrated that postoperative PHT is associated with liver dysfunction and poor outcome. [8][9][10][11] Chen et al. 9 found that Child-Pugh A patients with postoperative PVP of 12 mmHg or above had a higher risk of PHLF than patients with PVP below 12 mmHg. In patients without cirrhosis, Allard et al. 8 reported that post-hepatectomy PVP of 22 mmHg was the best cutoff for predicting PHLF defined by ISGLS. 22 The study by Bogner et al. 11 suggested that increase of portal pressure during operation could serve as an immediate and independent predictor of PHLF after major hepatectomy, which underlines the importance of intraoperative dynamics rather than the preand post-resection PVP. However, the level of central vein pressure pre-and post-hepatectomy were not taken into consideration.
Interestingly, the evaluation of PHT severity could be misled by PVP. [12][13][14][15][16][17][18] Experimental study on canine showed portal pressure rose by .91 mmHg for every mmHg increase in caval pressure. 14   liver transplantation procedures performed on 77 patients. 18 PVPG is the pressure gradient between the portal vein and central vein. An experimental study based on PHT canines model after haemorrhagic shock showed that, when compared with PVP, PVPG increased earlier and more significantly in the large volume infusion group. 33 In the present study, we noticed that PVP did not significantly increase in the "major hepatectomy group" and "PHLF+ group" (Figures 1 D and F), in contradiction with previous findings. 8,9,11 In contrast, increased PVPG was observed in these 2 groups. Therefore, PVPG may be a more reliable indicator to evaluate portal pressure and predict PHLF than PVP.
In the study reported by Lan et al., PVP and CVP were measured every day for a week after hepatectomy through portal and central vein catheterization. 34 It was found that PVPG was affected by extent of resection only in patients with moderate and severe cirrhosis. However, the comparison between PVP and PVPG was not mentioned. Furthermore, Lan' s study provided an unique opportunity to observe the dynamic changes of PVPG after hepatectomy. However, it is impossible to measure PVPG after closure of the abdomen in most cases; therefore, we believe monitoring the early response of PVPG is more practical, in addition to which it may provide the opportunity for early portal pressure modulation in order to prevent PHLF. If intraoperative PVPG increases greatly, we may add some treatments, such as terlipressin, to modulate PVPG and improve postoperative liver function recovery. 27 There were some limitations in the present study. First, there was a risk that this study was underpowered because the number of patients was small, preventing a robust statistical analysis. Second, the cutoff value for increase of PVPG was made based on the study cohort. An external validation cohort is needed for further study. Third, the incidence of PHLF was low, and all of them were grade A PHLF, which may be not clinically critical. Because many HCC patients with severe cirrhosis were treated with radiofrequency ablation or liver transplantation, the patients enrolled were Child-Pugh A with FLR >30%.   (2) Therefore, a post-hepatectomy PVPG increment of 1.5 mmHg or greater only affected patients at risk of mild liver failure in our study. Fourth, the cohort in this study includes heterogeneous diseases. Among the 60 cases enrolled, 2 were hepatic cyst, 48 were HCC, 4 were ICC, and 6 were mixed HCC-ICC. The procedure of cyst resection is different from removing liver malignancy and may thus affect the result.

Conclusions
In summary, the present study showed that acute increase of PVPG after hepatectomy can predict PHLF in patients with Child-Pugh A cirrhosis and FLR >30%. A posthepatectomy PVPG increment of 1.5 mmHg or greater was an immediate and independent predictor for PHLF. The advantage of this evaluation index is early prediction of PHLF during operation and further improvement of safety for Child-Pugh A patients who underwent liver resection. Although these data require external validation, they support further study to investigate the role of PVPG as a therapeutic target that might be modified by surgical or medical treatments to prevent PHLF.

Acknowledgments
We are grateful to Dr Guo-Ming Shi and Dr Jia-Bin Cai for their assistance with data collection.

Author Contributions
N Xiao, X-L Li, and H-C Sun had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. N Xiao, X-L Li, and H-C Sun were responsible for the conception and design of the study, the acquisition, statistical analysis, and interpretation of the data, and the drafting of the manuscript. X-D Zhu, C Huang, Y-H Shen, J Zhou, and J Fan were responsible for the acquisition and interpretation of the data. All authors contributed to manuscript writing and final manuscript approval. The

Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article:

Availability of Data and Materials
The datasets used and analyzed during the current study are available from the corresponding author on reasonable request.

Ethical Approval and Consent to Participate
Written consent was obtained from all patients. Ethics approval and consent to participate this prospective study was approved by the Ethics Committee of Zhongshan Hospital, Fudan University. All methods were conducted in accordance with the approved guidelines.