Prevention of variceal rebleeding in cirrhotic patients with advanced hepatocellular carcinoma receiving molecularly targeted therapy: a randomized pilot study of transjugular intrahepatic portosystemic shunt versus endoscopic plus β-blocker

Although transjugular intrahepatic portosystemic shunt (TIPS) is recommended for secondary prophylaxis of variceal bleeding if standard therapy fails and for patients with high risk of recurrent bleeding, no guidelines for the treatment of symptomatic portal hypertension in HCC patients are available. This study aimed to compare the efficacy and safety of TIPS with endoscopic + β-blocker for prevention of the rebleeding in such patients. 106 consecutive advanced HCC patients receiving tyrosine kinase inhibitor (TKI) who had been treated with vasoactive drugs plus endoscopic therapy for variceal bleeding were randomly assigned to receive either TIPS (n = 52) or endoscopic + β-blocker therapy (n = 54) for the prevention of rebleeding. The primary endpoint was variceal rebleeding after randomization. During a median follow-up of 16 months, rebleeding occurred in 14 patients in the endoscopic + β-blocker group and 3 patients in the TIPS group (p < 0.001). Forty-nine patients died (38 in endoscopic + β-blocker group and 11 in TIPS group, p < 0.001). The 6-, 12-, and 18-month overall survival rates were 95, 81, and 73% for TIPS group and 35, 21, and 15% for endoscopic + β-blocker group, respectively (p < 0.001). Eight patients in endoscopic + β-blocker group received TIPS as rescue therapy, but two died. TKIs was discontinued in 32 patients, including 24 in the endoscopic + β-blocker group and 8 in the TIPS group (p < 0.001). No significant differences were observed between the two groups with respect to serious adverse events. In advanced HCC patients receiving TKIs and presented with variceal bleeding, the use of TIPS was associated with significant reduction in rebleeding, improved a higher adherence to TKIs therapy, and prolonged survival.


Introduction
Hepatocellular carcinoma (HCC) is the third most common cause of cancer death worldwide [1]. The majority of patients with HCC are diagnosed late in its course, rendering them carries a grim prognosis due to lack of systemic therapies [2]. Recently, the treatment landscape for advanced HCC [i.e., tumor extension into the hepatic vasculature, usually the portal or hepatic veins, patients with spread beyond the liver (including extrahepatic nodal metastasis), and/or patients with cancer-related symptoms (performance status 1 or 2)] has expanded rapidly, with the approvals of several oral tyrosine kinase inhibitors (TKIs, e.g., sorafenib, lenvatinib, donafenib, regorafenib, and cabozantinib), as well as immune check point inhibitors (e.g., atezolizumab, nivolumab and pembrolizumab) and tumor angiogenesis inhibitors (e.g., bevacizumab, ramucirumab) [3][4][5]. Ideal use of these agents in clinics can delay HCC progression and prolong overall survival (OS) of patients with advanced HCC as well as introduced fatal adverse events (AEs), such as hepatic failure and hemorrhage [6][7][8]. TKIs in advanced HCC were associated with increased risk of all grade and high-grade variceal bleeding [6], which resulted in some advanced HCC patients losing the opportunity for further tumor treatment with TKIs. Therefore, it is a critically important issue to manage variceal bleeding, especially the prevention of rebleeding in advanced HCC patients who are receiving TKIs and presented with portal hypertension.
The current treatment approaches for variceal bleeding include vasoactive drugs, endoscopic therapy, surgical management and transjugular intrahepatic portosystemic shunt (TIPS) [9]. Endoscopy can be used to identify the source of bleeding, as well as a hemostatic treatment for actively bleeding lesions [9,10]. However, endoscopic therapy is limited in value for variceal bleeding due to a lack of reducing the portal pressure. TIPS effectively reduces intravascular pressure in the portal venous system, and it is thus an ideal therapy in the management of portal hypertensive and prevention of rebleeding [11]. Although preemptive TIPS has become the standard of care in patients with high-risk acute variceal bleeding [12], TIPS is contraindicated in HCC patients with portal hypertension due to the risk of postoperative liver failure, severe complications, and low survival rate for HCC patients [3,4,9,11,12]. Based on our previous clinical experience [13][14][15], TIPS combined with anti-tumor therapy is safe and effective for the prevention of rebleeding in HCC patients with portal hypertension. However, these data were from single-center studies. Whether TIPS or endoscopic + β-blocker therapy is more beneficial for the prevention of rebleeding in advanced HCC patients who are receiving TKIs, especially regarding incidence of continuation of TKIs and achievement of a survival benefit, is still unknown. The aim of this study was to compare the efficacy and safety of TIPS with endoscopic + β-blocker for the prevention of variceal rebleeding in in such patients.

Study design and participants
This is a multicenter randomized controlled trial (RCT) was conducted between October 2018 and December 2020 in three tertiary hospitals, China. Patients would be eligible for the study if they were advanced HCC patients receiving TKI and presented with variceal bleeding. The inclusion criteria were: (1) Barcelona Clinic Liver Cancer stage C patients who are receiving TKI presented with episode of endoscopically documented variceal bleeding; (2) hemostasis and stabilization were successfully achieved with vasoactive drugs or endoscopy; (3) liver cirrhosis (diagnosed based on clinical presentation, laboratory tests, images) with Child-Pugh-Turcotte score ≤ 13. Exclusion criteria were: (1) uncontrolled bleeding before randomization; (2) Child-Pugh-Turcotte scores > 13; (3) bleeding from isolated gastric or ectopic varices; (4) a history of serious or refractory hepatic encephalopathy (HE); (5) a history of significant heart failure (New York Heart Association class III and IV); (6) pulmonary hypertension previous; (7) previous use of a portosystemic shunt or TIPS; (8) sepsis and/or multiorgan failure.

Classification of portal vein tumor thrombosis
We classified portal vein tumor thrombosis (PVTT) into four grades based on the Liver Cancer Study Group of Japan (LCSGJ)'s classification [16], as follows: (1) Vp1: presence of a tumor thrombus distal to the second-order branches of portal vein (but not involving them directly); (2) Vp2: invasion of the second-order branches of portal vein; (3) Vp3: presence of the thrombus in the first-order branches; (4) Vp4: tumor thrombus in the main trunk of the portal vein and/or a portal vein branch contralateral to the primarily involved lobe.

Management of patients
After stabilization and successful vasoactive drugs or endoscopic hemostasis, eligible patients were randomly assigned in a 1:1 ratio to receive TIPS placement or endoscopic + β-blocker. The randomization sequence was generated by a computer with the use of a concealed block size of four. The sample size was based on the assumption that the rate of development of variceal rebleeding would be 29% in long-term endoscopic + β-blocker group at the end of 1 year [17], and 7.8% after TIPS placement [18]. The sample size was determined to be 106 patients, with a dropout rate of 15%, an alpha level of 0.05 and a power of 80%.

Treatment
Hemodynamic stabilization and sustained endoscopic hemostasis were in the Supplementary Materials.
In the endoscopic + β-blocker group, endoscopic variceal ligation (EVL) for esophageal varices and N-butyl cyanoacrylate (HistoAcryl; B. Braun, Melsungen, Germany) injection for gastric varices in combination with β-blocker (preferably slow-release propranolol, titrated to the maximum tolerated dose at rest by 25%, with a lowest limit of 50 beats per minute) was started at day 5 after the index bleeding, unless a contraindication was present (severe arrhythmia, severe obstructive chronic obstructive pulmonary disease, or known intolerance). Elective EVL sessions, which were started 2 weeks after the index bleeding, were performed every 2-4 weeks thereafter until eradication of varices, followed by endoscopic surveillance and retreatment, if indicated, every 6 months.
In the TIPS group, the TIPS placement was performed under sedation anesthesia with propofol and remifentanil. The right jugular vein was punctured, and a 10-Fr Ring Transjugular Intrahepatic Access Set (Arrow, Reading, PA, United States) was advanced into the vessel. The needle (16-gauge Colapinto, Optimed, Germany) direction, which was used to puncture into the portal system, was estimated according to the location of the tumor thrombus. For tumor thrombus in the main portal vein and both its branches, we preferred the bifurcation of the portal vein as the puncture site. TIPS with covered stents (Fluency; Bard, Tempe, AZ, United States), were used with initial balloon dilatation to 8 mm, aiming to decrease the portosystemic pressure gradient (PPG) to less than 12 mmHg (or decline of 50%). After simultaneous needle withdrawal and contrast agent injection, the appearance of a "grid-like outline" in the tumor thrombus interspaces, or occasionally, small branches, was considered to indicate successful puncture of the portal system ( Supplementary Fig. 1). TIPS function was primarily monitored as follows: ultrasound was performed at 1 week and 4 week after TIPS and then at 3 months and 6 months, and at 6-month intervals thereafter, or in case of recurrent bleeding or ascites. Angioplasty or/and another covered stent placement was performed when TIPS had been confirmed dysfunction.

Outcomes and follow-up
The primary endpoint was rebleeding. After secondary prevention treatment, new melena or hematemesis occurs again after the stool color is normal, resulting in hospitalization, blood transfusion or hemoglobin decrease of at least 3 g/L, which was defined as rebleeding. Variceal rebleeding was further divided into failure to control bleeding (within 120 h after index endoscopic treatment) or failure of secondary prophylaxis (after 120 h) according to the Baveno VI criteria [19]. The secondary endpoints were occurrence of treatment failure (either switch to other therapy or death), discontinuation of TKIs therapy, HE, OS, and progression-free survival (PFS) based on clinical parameters. The TKIs dose reduction and drug discontinuation criteria are previously reported [20].

Safety evaluation
Major postoperative complications were defined as grade 3 and above according to the Clavien-Dindo classification [21]. Morbidity and mortality were defined as complications and death occurring within 90 days after the procedure. TKIrelated AEs were evaluated using CTCAE version 4.0. OS was calculated as the time from initial intervention to death or the last follow-up.

Statistical analysis
Patients were censored at the time of loss-to-follow-up or last outpatient visit before study closure. In addition to the censoring time points in the intention-to-treat analysis, patients were censored at the moment they switched therapy. Student's t tests and χ 2 tests were performed to compare continuous and categorical variables, respectively. Data are expressed as median (interquartile range) and number (percentage) when appropriate. Kaplan-Meier (event-free) survival analyses with log-rank tests and Cox's proportional hazard analyses were performed for the endpoints, including rebleeding, treatment failure, death, and HE. In case of zero events in one group, likelihood ratio test with Firth's correction and 95% hazard ratio profile with likelihood confidence limits were used. All variables with p < 0.05 evaluated by univariate analysis were subjected to multivariate analysis. All analyses were performed using the IBM SPSS statistics software v. 25 (IBM Corporation, Armonk, NY, USA). A two-sided p value < 0.05 was considered statistically significant.

Patients, recruitment, and follow-up
The flow diagram of patients enrolled in this study is displayed in Supplementary Fig. 2. A total of 106 patients were included and randomly assigned to either the TIPS group (52 patients) or the endoscopic + β-blocker group (54 patients) after a median of 5 days from index bleeding (interquartile range [IQR]: 3-8). Among them, the most common distribution of etiology was HBV-related HCC (78%), 30 (30/106, 28%) patients who had HCC recurrence after prior radical therapy (resection as 8, RFA as 22) at the time of initial diagnosis and progressed to advanced stage, 25 (25/106, 24%) patients who had progressed to advanced stage after prior TACE. PVTT classification, the proportion of Vp2, Vp3, and Vp4 were 52, 29 and 19%, respectively, in the TIPS group and 54, 33 and 13%, respectively, in the endoscopic + β-blocker group (p = 0.658, Table 1). The baseline variables were comparable between the two groups. Two (4%) of the 54 patients who were randomized to the endoscopic + β-blocker group crossed over to the TIPS group due to refractory ascites. The median followup time was 16.1 months (range, 4.2-24.6 months) for the endoscopic + β-blocker group and 16.2 months (range, 13.0-28.8 months) for the TIPS group (p = 0.186, Table 2). Three patients (3%; 2 in the endoscopic + β-blocker group, 1 in the TIPS group) were lost to follow-up after a median of 14.2 months (range, 5.2-20.3 months).

Treatments
In the endoscopic + β-blocker group, a total of 129 upper endoscopies (mean, 2.8 ± 2.4 per patient) were performed in the first year after randomization. The majority (71%) of procedures included EVL with placement of a mean 4.6 bands per procedure, 8% included injection N-butyl cyanoacrylate therapy, and in 21% no treatment was considered necessary. Propranolol slow release, titrated on heart rate and/or tolerance, was used for all but three patients.

Primary endpoints: rebleeding
During follow-up, 14 (27%) patients in the endoscopic + β-blocker group experienced a total of 20 variceal rebleeds compared to 3 (5.6%) patients in the TIPS group who experienced a total of 3 variceal rebleeds (p < 0.001; Fig. 1a and Table 2). Among the 14 patients with rebleeding in the endoscopic + β-blocker group, 6 patients experienced the first rebleed during follow-up, 8 patients experienced the second rebleed and received TIPS as a rescue therapy, of whom two died. Among the three patients with rebleeding in the TIPS group, all switched to receive a TIPS revision with embolization of varices. Univariate analysis revealed that two factors were significantly related to rebleeding after initial intervention. In the multivariate analyses, only the endoscopic + β-blocker (HR 7.548; 95% CI 2.513-22.668; p < 0.001) was an independent poor predictor of rebleeding after initial intervention (Table 3).

Changes in the Child-Pugh-Turcotte stages
One month after initial intervention, the distribution of the Child-Pugh-Turcotte improving, stabling, and worsening was 76, 9 and 15%, respectively, in the TIPS group and 12, 23 and 65%, respectively, in the endoscopic + β-blocker group ( Table 2). The TIPS group had a significantly higher rate of postoperative Child-Pugh-Turcotte stage improving

The specific adverse event of TKI
No differences in TKI-related AEs were observed in either group (Supplementary Table 1). In the TIPS group, a total of 70 complications were observed in 36 patients. Three patients (8%) stopped using TKI due to intolerance. Seventysix complications were observed in 39 patients in the endoscopic + β-blocker group. Three patients (8%) stopped using TKI due to intolerance.

Discussion
Before our study, few data on the impact of the use of TIPS for a second prophylaxis for rebleeding in advanced HCC patients with variceal bleeding was reported. In most previous studies of primary and secondary prophylaxis [22][23][24], HCC was a criterion for exclusion, thus we were not in a position to fully understand whether strategies of treatment of portal hypertension in HCC-free patients do confer clinical benefits to advanced HCC patients who are receiving TKIs and presented with variceal bleeding. Therefore, well-designed prospective studies are needed to compare the efficacy and safety of TIPS with endoscopic + β-blocker for prevention of rebleeding in such patients. All in all, our data stressed TIPS placement was superior to endoscopic + β-blocker therapy for secondary prevention of rebleeding, reduced risks of failure to control bleeding and incidence of discontinuation of TKI, with no serious procedure-related complications, which was translated into improved survival.
First, we focused on the feasibility and safety of TIPS implantation. Importantly, all patients enrolled in this study had PVTT, especially tumor thrombus in the main portal vein or/and complete portal vein occlusion accounted for 48%. These patients are of particularly problematic as TIPS implantation may be associated with higher rates of technical failure, liver failure and HE. In the present study, after TIPS implantation, only one patient experienced HE of grade III. No TIPS-related liver failure occurred. A potential explanation for the feasibility results might be this study was markedly different from previous reported as follows: (1) the whole TIPS procedures were performed by expert interventional radiologists who have performed more than 1000 cases with complex and difficult TIPS placement. (2) Tumor thrombus in the trunk of the portal vein and its left or right branch, and the patent branches were successfully accessed with the conventional TIPS method [22]; (3) for tumor thrombus in the main portal vein and both its branches, we preferred the bifurcation of the portal vein as the puncture site, because the portal vein was enlarged due to expansion of the tumor thrombus; (4) to achieve long-term shunt patency, the tumor thrombus in the main portal vein should be completely covered; (5) contrast material is injected to determine whether access to a portal vein has been gained such as "grid-like outline". Our findings confirm that the use of TIPS for the prevention of rebleeding is also suitable for HCC with portal hypertension, especially in advanced HCC patients who are receiving TKIs therapy and presented with variceal bleeding.
Second, we found that the use of TIPS was associated with a reduction in the number discontinuation of TKI therapy cases, which can translate into improved survival of advanced HCC. Notably, no differences in serious AEs or TKI treatment-related intolerance discontinuation cases were observed between the TIPS and endoscopic + β-blocker groups in our study. However, endoscopic + β-blocker group resulted in a higher incidence of discontinuation of TKIs therapy because of failure to control variceal bleeding or rebleeding. Therefore, it is important to prevent rebleeding in these patients and in this setting TIPS implantation may contribute to the possibility for subsequent TKI treatment.
Third, and even more important, this study confirmed that the use of TIPS for the prevention of rebleeding in advanced HCC patients with portal hypertension was associated with a reduction in mortality. This beneficial effect on survival was observed even though rescue TIPS was used in patients in whom endoscopic + β-blocker treatment failed. Notably, mortality was very high among the patients who underwent the endoscopic + β-blocker treatment. Due to the lack of shunt function, the efficacy of endoscopic + β-blocker treatment for prevention of rebleeding in advanced HCC patients with portal hypertension is limited. In addition, the rebleeding leads to further deterioration of liver function, which is a more important reason for the high mortality after endoscopic + β-blocker treatment. These patients, in fact, the multiple endoscopic sessions required to curb the rebleeding risk can be not only unbearable for these frail patients, but can also delay TKIs treatment. However, TIPS can significantly improve Child-Pugh-Turcotte staging by reducing intravascular pressure in the portal vein system, preventing rebleeding, and reducing ascites, and thus patients could better tolerate subsequent TKIs therapy, ultimately affecting survival through changes a higher adherence to TKIs therapy. The multivariate analysis revealed that endoscopic + β-blocker treatment options, rebleeding, discontinuation of TKIs therapy, Child-Pugh-Turcotte class C were independent negative prognostic factors for OS. After TIPS implantation, improved liver function should increase the likelihood of TKIs therapy. Importantly, the use of TIPS for the improving survival may also be that TIPS can directly affect PVTT because the shunt stents were directly inserted into the PVTT. Hence, these findings provide the rationale to extend our current knowledge and practice by offering advanced HCC patients who are receiving TKIs therapy and presented with variceal bleeding can benefit from a preemptive (early) TIPS.
There were several limitations to the current study. First, we did not analyze the difference in length of hospital stay and cost between the two groups. Second, our study was based on multiple tertiary hospitals, the number of analyzed patients was relatively small, but the primary endpoint of the present study was met, which had the power of the study. Finally, due to different interventions, this study could not be conducted using a blind RCT, but an open-label RCT. However, under the randomized results hiding approach (i.e., data management committee), which greatly reduces the investigator selection bias. Despite these limitations, our data are clinically meaningful, as our results offer the first comparison of the feasibility and clinical value of TIPS versus endoscopic + β-blocker treatment in patients of with particular issues, such as TIPS implantation.
In conclusion, in advanced HCC patients who are receiving TKIs and presented with variceal bleeding, TIPS implantation is superior to endoscopic + β-blocker therapy for the prevention of rebleeding, reductions in the failure to control bleeding and mortality. Patients treated with TIPS implantation show the improved Child-Pugh-Turcotte stages and can better tolerate subsequent TKIs therapy, and translate into improved survival, thus TIPS is a feasible and effective management for the cirrhotic advanced HCC patients with portal hypertension.