Management of hepatocellular carcinoma patients with portal vein tumor thrombosis: A narrative review

Zi-Wen Tao, Bao-Quan Cheng, Tao Zhou, Yan-Jing Gao

Department of Gastroenterology, Qilu Hospital, Cheeloo College of Medicine, Shandong University, Jinan 250012, China

Keywords: Hepatocellular carcinoma Portal vein tumor thrombosis Transarterial chemoembolization Radiotherapy Surgery Sorafenib

ABSTRACT Background: Hepatocellular carcinoma (HCC) is one of the main reasons for malignancy-related death. Portal vein tumor thrombosis (PVTT) is the most common form of macrovascular invasion related to HCC occurring in 10%-60% of patients. HCC with PVTT is usually characterized by worsening liver function, vulnerability to blood metastasis, higher incidence of complications associated with portal hypertension, and intolerance to treatment when compared with that without PVTT. If only treated with supportive care, the median survival of HCC with PVTT is about 2.7 months. In the past, sorafenib was the only recommended therapy by guidelines with limited effectiveness. This narrative review aimed to describe the current management options for HCC with PVTT.Data sources: We have reviewed literature from PubMed on the treatment of HCC with PVTT and com- piled evidence-based facts on effective therapies available for different types of PVTT.Results: Sorafenib monotherapy is not much effective, but combining it with other methods can improve survival. Each type of PVTT can benefit from the combination of transarterial chemoembolization and so- rafenib than sorafenib monotherapy. The tumor downstaging can be realized possibly after transarterial chemoembolization, but tumor invasion into the main trunk of the portal vein greatly impairs efficacy. Although surgery is a curative approach, it is often not recommended for Vp4 PVTT. Some new methods can broaden the indication, but further explorations are needed. Radiotherapy can decrease the possi- bility of Vp3 progression to Vp4, but building a forecast model of best radiation dose and response is necessary. Systemic chemotherapy, hepatic arterial infusion chemotherapy, radiofrequency ablation, por- tal stenting, and traditional Chinese medicine are also beneficial in Vp3-4 PVTT. The accurate diagnosis of PVTT can be made by radiomics, and prognostic classification models can be used to design personal- ized treatments. The application of new treatment methods such as the atezolizumab plus bevacizumab scheme may increase survival.Conclusions: HCC with PVTT is still a thorny problem, and effective therapeutics need to be explored.

Introduction

Hepatocellular carcinoma (HCC) accounts for 75%-85% of pri- mary liver cancer and is the fourth leading cause of global cancer- related deaths annually [1] . It tends to invade the portal vein (PV) system and form portal vein tumor thrombosis (PVTT), which oc- curs in 10%-60% of patients [2] . About 70% of HCC patients are di- agnosed at an advanced stage [3] . HCC with PVTT is usually clas- sified as Barcelona Clinic Liver Cancer (BCLC) stage C. Also, PVTT is a strong negative prognostic factor with very short overall sur- vival (OS) that ranges from 2 to 4 months, when only treated with supportive care [4] . The poor prognosis is probably due to the in- tensified risk of tumor spread, increased portal pressure, and re- duced portal flow, which can lead to further complications, includ- ing variceal bleeding, ascites, and even liver failure [5] .

Additionally, therapeutic options for HCC with PVTT are limited. Based on the result of phase III clinical trial, sorafenib could pro- long OS and time-to-progression (TTP) [6] , and it is also recom- mended by the BCLC, American Association for the Study of Liver Diseases, and European guidelines [7] . However, the efficiency of sorafenib treatment for HCC with PVTT was modest, with a me- dian OS of 9.5 months in BCLC stage C patients [8] . According to the previous experience, PVTT contradicts a variety of therapeu- tic methods, such as liver transplantation [9] . Recently, the com- bination of atezolizumab and bevacizumab was expected to break the status quo of sorafenib as the only treatment included in the guidelines [ 10 , 11 ]. However, further studies are still needed.

In recent years, good results have been observed in the re- search on different managements of HCC with PVTT. A novel point of view indicates that HCC patients with PVTT include a hetero- geneous population whose prognosis depends on the extension of PVTT, degree of portal hypertension, the severity of liver dysfunc- tion, clinical characteristics of patients, and response to treatments. Therefore, some selected patients may gain survival benefits from some aggressive treatment methods [12] . Hence, it is remarkably important to apply individualized treatment to each group of PVTT with a multidisciplinary approach. This narrative review describes the current management options for HCC with PVTT and provides possible directions for future research.

Classification of PVTT

The extension of PVTT is closely related to the prognosis of HCC, but discrepancies are observed in recommended treatments for different types of PVTT. At present, many classification types for PVTT exist, but the better-known and more conventional one is proposed by the Liver Cancer Study Group of Japan [13] , known as Japanese Vp classification, which according to the severity of tu- mor thrombus and anatomic structures of PV is divided into five types: Vp0, no PVTT; Vp1, PVTT distal to but not involved in the second-order branches of PV; Vp2, PVTT invasion to the second- order branches; Vp3, PVTT presented in the first-order branches; Vp4, PVTT extended to the main portal trunk and/or a contralat- eral PV branches.

Another classification system is called Cheng’s classification, which is identified into four categories: type I, PVTT involved in segmental branches of PV or above, which combines Vp1 and Vp2; type II, PVTT involved in the left or right PV, which corresponds to Vp3; type III, PVTT involved in the main PV; type IV, PVTT involved in the superior mesenteric vein [14] .

Treatment methods for HCC with PVTT

Although the only recommended treatment for HCC with PVTT is sorafenib, many other therapeutics have also been studied, such as transarterial chemoembolization (TACE), radiotherapy (RT), cu- rative surgery, hepatic arterial infusion chemotherapy (HAIC), ra- diofrequency ablation (RFA), etc.

Sorafenib and target therapy

Sorafenib is an oral multikinase inhibitor that inhibits angiogen- esis and tumor cell proliferation by inhibiting serine-threonine ki- nases, Raf-1 and B-Raf, platelet-derived growth factor receptorβ, and receptor tyrosine kinase of vascular endothelial growth fac- tor receptor [15] . Besides, many tumor models have found that sorafenib could promote tumor cell apoptosis [16] . Sorafenib was the first drug to improve survival in patients with advanced HCC based on two landmark clinical trials [17]. The Sorafenib Hepato- cellular Carcinoma Assessment Randomized Protocol Trial showed that patients with macrovascular invasion who took sorafenib had an OS of 8.1 months. In contrast, the placebo group had an OS of 4.9 months [hazard ratio (HR) = 0.68; 95% confidence inter- val (95% CI): 0.49-0.93] [6] . In the Asia-Pacific Trial, similar results were obtained with OS of 5.6 months in the sorafenib group ver- sus 4.1 months in the placebo group, respectively (HR = 0.75; 95% CI: 0.54-1.05) [18] .

Jeong et al. [15] verified the efficacy of sorafenib monother- apy for Vp3-4 PVTT through a retrospective study, where the median OS was 3.1 months (95% CI: 2.70-3.50) and the median progression-free survival (PFS) was 2 months (95% CI: 1.96-2.05). Three patients with partial response (PR) achieved portal venous revascularization after 4 months of treatment. Fatigue and hand- foot skin reactions were the most common side effects (43.3% and 30%, respectively). Although the number of cases was limited, the therapeutic effects were remarkable.

Yang et al. [19] indicated that the survival benefits of metro- nomic chemotherapy using cisplatin and 5-fluorouracil (5-FU) were better than that of sorafenib monotherapy (median OS: 158 vs. 117 days,P= 0.029). They observed no difference in grades 3-4 adverse events (AEs) (P= 0.488), suggesting that Vp2-4 PVTT patients with poor liver function could select metronomic chemotherapy rather than sorafenib. In a systematic review, Liu et al. [20] showed that HAIC was superior to sorafenib in treat- ing types III-IV PVTT concerning OS (HR = 0.29,P＜0.001), PFS (HR = 0.39,P＜0.001), and disease control rate (DCR). In terms of treatment of HCC with major PVTT, a propensity score analy- sis [21] found that no statistical difference was observed in me- dian survival between the sorafenib group (4.3 months) and the RT group (5.9 months,P= 0.115). However, a better survival rate was noted in the latter after propensity score matching (median 10.9 vs. 4.8 months,P= 0.002), suggesting that RT was better in treating Vp4 PVTT.

Even when the four famous global phase 3 clinical trials could not confirm the non-inferiority or superiority of other therapies to sorafenib, including sunitinib, brivanib, linifanib, and erlotinib plus sorafenib as the first-line of treatment for advanced HCC, in terms of OS, the study of Kudo et al. [22] affirmed it. The study showed that the ability of levatinib to extend median OS was non-inferior to sorafenib (13.6 vs. 12.3 months; HR = 0.92; 95% CI: 0.79-1.06), and that levatinib had better PFS, objective response rate, and TTP along with tolerable AEs. Therefore, levatinib has been listed as the first-line of treatment for advanced HCC by the China Food and Drug Administration [23] . However, this study excluded Vp4 PVTT patients.

Due to the inefficiency of sorafenib monotherapy, many stud- ies were conducted to verify the effectiveness of combination ther- apy [24–26] . Yu et al. [24] confirmed that sorafenib could promote radiation-induced apoptosis, inhibit radiation-induced proliferation and DNA repair, and perhaps also have anti-angiogenesis effects, thus, enhance the effect of RT for HCC cellsinvivo. A prospec- tive randomized controlled trial (RCT) [25] showed that an addi- tional combination of cryotherapy could significantly improve the clinical outcomes compared to sorafenib monotherapy, where ei- ther the PV branches or the main trunk was involved, and the me- dian OS was 12.5 versus 8.6 months (P= 0.01). Another RCT also found similar results [26] that sorafenib combined with percuta- neous RFA significantly improved 3-year survival rate in Vp4 pa- tients in contrast to sorafenib monotherapy (26% vs. 0%, HR = 2.87; 95% CI: 1.61-5.39;P＜0.001), and multivariate analysis showed that the combination therapy was the only independent predictor for the prognosis (HR = 2.89; 95% CI: 1.55-5.43;P＜0.001).

TACE

Precise delivery of chemotherapeutic drugs to tumor target ar- teries through TACE can prolong the residence time of drugs and cause tumor ischemia through embolization [27] . However, HCC with PVTT is generally considered a contraindication to TACE be- cause of the concerns of disruption of hepatic artery supply, which may lead to necrosis of the liver since PV blood supply is already impaired and may cause liver failure [28] . However, a large num- ber of studies confirmed that selected PVTT patients could tolerate TACE as long as liver function and collateral circulation around the blocking site are sufficient [ 29 , 30 ].

Comparison of monotherapy and other treatments

A retrospective study [31] found that 1-, 3-, and 5-year survival rates of patients with segmental PV invasion treated with TACE were 44.9%, 16.0%, and 12.0%, respectively. High tumor load, extra- hepatic metastasis, poor tumor response to TACE, and high Child- Pugh score were the risk factors for poor prognosis. Patients with 2-4 risk factors were not recommended TACE because the expected OS of the patients with 0, 1, and 2-4 risk factors were 29.1, 15.1, and 5.3 months, respectively. Le et al. [32] also recommended ob- taining TACE for getting maximum survival benefits and minimal adverse reactions for patients who only had 0-1 risk factor.

Tawada et al. [33] retrospectively studied the efficacy and tolerability of TACE in Vp2-4 patients and found that PVTT response-positive cases had better median OS than negative ones (14.0 vs. 5.8 months,P＜0.01) with both peripheral and major PVTT response-positive groups showing this outcome (18.2 vs. 5.9 months,P= 0.04; 11.1 vs. 5.5 months,P= 0.04; respectively). Liu et al. [34] found that median OS after TACE was 6.1 months (95% CI: 5.6-6.5), and those with no invasion into the main PV showed better survival.

Zhang et al. [35] firstly compared the survival benefits of TACE with hepatectomy and found better results in the surgery group in types I-II PVTT, but the difference was not observed in types III-IV. A retrospective study [36] compared surgical resection versus RT combined with TACE for resectable HCC (tumor size was about 9.5 cm) with types I-III PVTT, which revealed that the combination of RT and TACE was superior to surgical resection plus TACE (median OS: 12.3 vs. 10 months,P= 0.029). Patients who achieved down- stage after 3-dimensional conformal radiotherapy (3D-CRT) then received salvage operations had better survival.

Luo et al. [37] found that both segmental and major PVTT could benefit from TACE through a prospective study. Here, 9 patients who received TACE were downstaged to receive partial hepatec- tomy (n= 6) or local ablative therapy (n= 3), which were consid- ered to be curative treatments with longer median OS than the re- maining 75 patients who only received TACE (23.0 vs. 6.3 months,P＜0.01). However, Xiang et al. [38] had different views and found that median OS of types I, II, III, IV PVTT were 19, 12, 9, and 6 months, respectively, in the TACE group compared to 12, 7, 7, and 5 months in the best supportive care group. Thus, TACE was not recommended for type IV PVTT patients.

Combination therapy

Zheng et al. [39] confirmed that TACE combined with RFA was an effective method to treat types I-III PVTT with 1-, 3-, and 5-year OS rates of 63%, 40%, and 23%, respectively. Patients with low vas- cular endothelial growth factor levels had relatively higher survival (P= 0.031). The functional remnant liver volume and remnant liver volume/total liver volume were included in the influencing factors analysis (HR = 1.008,P= 0.039; and HR = 0.976,P= 0.010; re- spectively) since the liver volume was related to hepatic reserve, and could reflect the possibility of liver failure.

An increased number of studies have reached a consensus on the efficacy of TACE combined with sorafenib for PVTT. TACE can induce extensive ischemic and hypoxic changes in intrahepatic tu- mor cells while sorafenib targets PVTT and inhibits revasculariza- tion induced by TACE, thus, improving local tumor control. There- fore, the combined application of two methods can delay the pro- gression of PVTT and reduce the risk of tumor metastasis [40] . A retrospective study [41] found that the branch PVTT showed better survival than the main PVTT when TACE plus sorafenib was used (median OS: 9.7 vs. 4.3 months,P= 0.001). Patients who got early disease control showed longer OS than those who did not (15.5 vs. 9.1 months,P＜0.001). However, Lencioni et al. [42] found that it was technically feasible to combine sorafenib with doxorubicin- eluting beads-TACE for intermediate-stage HCC, but TTP was not longer than that of the TACE alone. However, PVTT patients were not included in this RCT. Another study [43] also found that TACE combined with sorafenib did not extend OS compared to sorafenib monotherapy in advanced HCC (12.8 vs. 10.8 months,P= 0.29) and increased the incidence of serious AEs (33.3% vs. 19.8%,P= 0.006), but it benefited the tumor response rate (60.6% vs. 47.3%,P= 0.005) and PFS (5.2 vs. 3.6 months,P= 0.01). For patients with Vp3-4 PVTT, the combination therapy tended to ex- tend survival, but it did not show a statistically significant differ- ence [43] .

The TACE treatment for HCC with PVTT has not yet reached a consensus, and more studies are needed in the future to screen out patients suitable for TACE to obtain maximum survival benefits.

RT

External beam radiation

RT is not considered suitable for HCC because the liver is a radiation-sensitive viscera, where high-dose radiation can cause se- rious liver toxicity [12] , while a lack of precise irradiation on the target area often involving surrounding normal liver tissues can in- crease the occurrence of radiation-induced liver disease. However, with the emergence of new radiation technology, such as 3D-CRT, stereotactic body radiotherapy (SBRT), and proton beam therapy (PBT), RT has gradually become an effective means of treatment for HCC.

3D-CRT

3D-CRT is the foundation of modern RT, which is based on three-dimensional anatomical information and uses computed to- mography for conformal radiation [44] . Kim et al. [45] retrospec- tively explored the relationship between 3D-CRT dose and the re- sponse of Vp3-4 PVTT, where 71.2% of patients had cirrhosis. Pa- tients with a dose of ≥58 Gy 10 RT had a higher response rate than those with＜58 Gy 10 (54.6% vs. 20%,P= 0.034), which suggested that there was a dose-response effect, and the recommended dose was＞58 Gy 10 . The median OS of responders was longer than that of the non-responders (10.7 vs. 5.3 months,P= 0.05). How- ever, Rim et al. [46] found that using high-dose RT (median total dose was 61.2 Gy) for HCC could cause grade 3 late complications; hence, the safety of high-dose RT therapy still needs to be consid- ered carefully.

Su et al. [47] compared 3D-CRT with surgery and found that the median OS of two groups were 13 versus 18 months (P= 0.003). The proportion of tumor size ≥10 cm in the RT group was signifi- cantly higher than that found in the surgery group. The 1-, 2-, and 3-year OS rates were 65%, 39%, 19% in RT versus 83%, 53%, 42% in surgery for type I PVTT (P＜0.001) while these rates were 52%, 35%, 11% versus 55%, 42%, 25% for type II PVTT (P= 0.612), and 16%, 3%, 0% versus 11%, 0%, 0% for type III PVTT (P= 0.041). The OS was longer in the surgery group than in the RT group before com- bined TACE (P= 0.018), but there was no statistical significance after combined TACE (P= 0.108), indicating that TACE was a pro- tective factor while 3D-CRT was an effective therapeutic measure for types II and III PVTT.

SBRT

SBRT can accurately deliver a larger single fraction size with fewer fractionations and is more conformal [48] . Xi et al. [49] found that volumetric modulated arc therapy-based SBRT treatment for types I-III PVTT showed a good local control rate with low toxicity. The complete response (CR) and PR rates were 36.6% and 39%, respectively, with no grades 4-5 AEs. They also found that combining SBRT with sorafenib showed a tendency to prolong median OS (16.6 vs. 10.9 months,P= 0.755), but no statistical difference was observed. Thus, further prospective stud- ies are needed to confirm it. Another study [50] suggested that SBRT could be used as a bridge treatment for liver transplantation so that patients do not lose the opportunity while waiting for a liver donor due to disease progression. The 5-year OS and disease- free survival (DFS) rates are both 100% after SBRT followed by liver transplantation.

PBT

PBT can achieve precise treatment and protect surrounding normal tissue. A retrospective study [51] found that PBT could improve local control and survival in HCC with Vp1-4 PVTT, where all 12 cases showed a response and had a median PFS of 2.3 years. Two patients lived for 4.3 and 6.4 years, respectively, with no evidence of recurrence and no third-level AEs. Sugahara et al. [52] concluded that both the main trunk and major branch in PVTT could benefit from PBT; the OS of both groups showed no statistical difference. In general, the 2- and 5-year local PFS rates were 46% and 20%, respectively, and the median local PFS was 21 months. They also suggested expanding the radiation target appro- priately to get a better treatment effect.

Gamma knife radiosurgery (GKR)

GKR can also aid in designing precise treatment. Lu et al. [53] compared GKR with palliative treatment and found that the median OS were 6.1 and 3.0 months, respectively (P= 0.003). Most of the AEs were mild or moderate; the subgroup analysis found that many types of PVTT could gain survival benefits from GKR. Multivariate analysis found that the performance status 0- 1 (HR = 0.412; 95% CI: 0.209-0.791;P= 0.014), alpha-fetoprotein (AFP) ≤400 ng/mL (HR = 0.477; 95% CI: 0.276-0.823;P= 0.008) and Child A (HR = 0.386; 95% CI: 0.196-0.758;P= 0.006) were predictors of good prognosis of GKR.

Intensity-modulated radiotherapy (IMRT)

Hou et al. [54] retrospectively found that Vp1-4 PVTT patients treated with high-dose IMRT had longer median OS than 3D-CRT (15.47 vs. 10.46 months,P= 0.005) without increasing radiation toxicity. Tumor thrombi response (P= 0.044) was found to be higher in the IMRT group, while the responses were similar in the overall and intrahepatic tumors. However, the optimal radia- tion dose for good local control is still not determined.

Combination therapy

Li et al. [55] suggested that for types II-III PVTT, both the tumor and PVTT response rates were higher in the TACE plus RT group than in the TACE group. The tumor response rates were 72.3% ver- sus 54.7% in type II (P= 0.027), 61.1% versus 45.2% in type III (P= 0.039). The PVTT response rates were 78.7% versus 38.2% in type II, 70.4% versus 30.0% in type III (bothP＜0.001). However, no difference was observed in types I and IV PVTT; it suggested that the location of PVTT could guide the combination therapy. Chu et al. [56] found no statistical difference in the OS (13.2 vs. 12 months,P= 0.299) and PFS (5.9 vs. 4.8 months,P= 0.258) between TACE combined with RT and TACE combined with so- rafenib in Vp1-4 PVTT after propensity score matching. A prospec- tive study conducted by Yoon et al. [57] revealed that compared to sorafenib monotherapy, HCC with macroscopic vascular inva- sion treated using TACE plus RT showed longer OS (55.0 vs. 43.0 weeks,P= 0.04) and TTP (31.0 vs. 11.7 weeks,P＜0.001). Five cases (11.1%) achieved tumor downstage in the TACE plus RT group and then successfully underwent surgery.

Fujino et al. [58] found that HAIC plus 3D-CRT provided survival benefits to HAIC non-responders with Vp3-4 PVTT. The median OS, time-to-treatment failure, and post-progression survival were longer in the combination group than in the HAIC alone group (8.6 vs. 5.0 months,P= 0.0 0 02; 5.0 vs. 2.7 months,P= 0.0024; 5.3 vs. 1.5 months,P= 0.0 0 01, respectively).

For the first time, Lu et al. [59] verified the efficacy of gamma knife surgery plus TACE, the median OS was 9.7 months, and the main AEs were abdominal pain, nausea/vomiting, anorexia, fatigue, and liver function impairment, most of which were tolerable. Sub- group analysis showed that the combination therapy had survival benefits no matter what the position of PVTT was.

Selective internal radiation therapy

Selective internal radiation therapy can deliver a therapeutic dose of 90 Y, 131 I, 125 I or other radioactive substances to the liver artery via resin or glass particles [12] . Since the range of radiation and the size of the particle are small, postoperative isolation is not required, and the risk of liver ischemia is greatly reduced [60] .

Mazzaferro et al. [61] prospectively revealed that yttrium-90- radioembolization (Y90-RE) was effective for Vp1-4 PVTT, the me- dian OS was 15 months (95% CI: 12-18), and no difference was found in median OS between patients with PVTT and without (13 vs. 18 months,P＞0.05). The tumor response was significantly cor- related with radiation dose in the target area (r= 0.60; 95% CI: 0.41-0.74;P＜0.001), with 500 Gy being the threshold to predict response. Spreafico et al. [62] established a prognostic score model to predict the response of Y90-RE in Vp1-3 PVTT patients, where the prognosis was divided into three categories: favorable, inter- mediate, and dismal (0, 2-3,＞3 points, respectively). The median OS of the three groups was 32.2, 14.9, and 7.8 months, respectively (P＜0.0 0 01); screening is helpful to predict the therapeutic out- comes.

Edeline et al. [63] retrospectively compared the efficacy of Y90- RE with sorafenib for the first time. The median OS was 18.8 months in Y90-RE cohort versus 6.5 months in sorafenib group (P＜0.001). The difference in OS between the Y90-RE and so- rafenib groups was more pronounced in cases of branch PVTT, with a respective median OS of 25.3 versus 7.0 months (P= 0.001), than in cases of main PVTT, with a respective median OS of 12.0 versus 6.5 months (P= 0.195). A systematic review and meta-analysis of the HCC patients with PVTT showed higher 6-month and 1-year OS rates in the RE group (76% vs. 54%; 47% vs. 24%; respectively), and a longer TTP (46% vs. 87%; 63% vs. 94%; respectively) when com- pared to the sorafenib group [64] . An RCT [65] found that Y90-RE and sorafenib result in similar OS for advanced HCC (8.0 vs. 9.9 months,P= 0.18), while the incidence of treatment-related AEs in the sorafenib group was two times higher than that of RE group. This suggested that RE is more tolerable in terms of the quality of life.

Akinwande et al. [66] compared the doxorubicin drug elut- ing beads-TACE with Y90-RE in Vp4 PVTT and found higher DCR (67% vs. 20%,P= 0.0014), longer median OS (10 vs. 3 months,P= 0.037), fewer AEs (11% vs. 39%,P= 0.03), and lower mortality rate (25% vs. 75%,P= 0.0011) in the TACE group. However, since the sample size was small, further studies are needed to verify the results.

Surgery

Although PVTT was previously considered a contraindication to surgery, more and more studies have confirmed that multidisci- plinary therapies based on surgical treatment could be used in ad- vanced HCC with PVTT.

Monotherapy and comparison with other treatments

Some retrospective studies found that hepatectomy was the first choice when PVTT was confined to the first-order branch with good liver functions. However, hepatectomy should not be consid- ered the first option for Vp4 PVTT unless PVTT needs to be re- moved urgently [ 67 , 68 ]. Ban et al. [69] showed diverse results, and they found that hepatectomy and thrombectomy could not only avoid PV obstruction caused by tumor thrombi for Vp4 patients but also provide them with similar survival benefits as Vp3 pa- tients (P= 0.821).

Some studies tried addressing this question. Vennarecci et al. [70] reported two HCC cases with major PVTT, who acquired adequate future liver remnant after associating liver partition and PV ligation in staged hepatectomy (ALPPS) then were successfully treated with classical surgery, suggesting ALPPS ability to expand surgical indications for HCC. Peng et al. [71] conducted an innova- tive surgical procedure and found that thrombectomy before hepa- tectomy could reduce the incidence of metastases due to the com- pression of PVTT, which was beneficial to the prognosis of types III-IV PVTT. Back-flow thrombectomy was also considered to ex- pand the therapeutic time window for Vp4 PVTT, but the OS was similar in both the back-flow thrombectomy group and the control group (14 vs. 15 months,P= 0.76) [72] .

Several studies have identified prognostic predictors. Zhang et al. [73] conducted a nomogram to predict early postoperative HCC recurrence after R0 resection, which contained PVTT, HBV- DNA, hepatitis B surface antigen, AFP, tumor diameter, and satel- lite nodules. This nomogram could help in screening out high- risk patients considered for adjuvant therapy. Eastern Hepatobiliary Surgery Hospital established EHBH-PVTT score to predict progno- sis of R0 resection, which contained tumor diameter, total biliru- bin, AFP, and satellite lesions. The median OS was longer in pa- tients with a score of ≤3 than those with scores＞3 (17.0 vs. 7.9 months,P＜0.001). Also, they recommended the first-order branch or above PVTT to receive surgery [74] . Yu et al. [75] found that although the 2-year PFS rate for Vp1-2 PVTT was higher in the liver resection group than in the TACE-RT group (41.9% vs. 15.7%,P= 0.02), the 2-year OS rate was similar (75.8% vs. 61.5%,P= 0.09), the tumor size was similar between the two groups (5.5 vs. 6.0 cm,P= 0.11). Cirrhosis prevalence was different but not statistically significant (45.2% vs. 68.1%,P= 0.06). However, these results need to be verified further. Wang et al. [76] proposed that the best therapy for types I-II PVTT was surgery, while for types III and I/II PVTT patients unsuitable for surgery, TACE-RT was the best. TACE plus sorafenib could be the best choice for those unsuitable for surgery or RT.

Combination therapy

Sun et al. [77] prospectively compared the efficacy of hep- atectomy and resection combined postoperative IMRT for the first time and found that the latter had significantly longer OS (18.9 ± 1.8 vs. 10.8 ± 1.3 months,P= 0.005) and DFS (9.1 ± 1.6 vs. 4.1 ± 0.5 months,P= 0.001) than the former. While a better prog- nosis was seen in types I-II PVTT receiving postoperative IMRT, no such trend was found in types III-IV PVTT. Another RCT [78] found that neoadjuvant 3D-CRT combined with hepatectomy was more effective than hepatectomy alone in types II-III PVTT while high expression of interleukin-6 might indicate RT resistance.

Yamamoto et al. [79] found no difference in OS of Vp2-4 PVTT patients between those who received preoperative TACE and those who did not (5-year OS rate: 29.3% vs. 11.3%,P= 0.747). Bai et al. [80] found that postoperative adjuvant TACE or RT showed sim- ilar median OS in types I-II PVTT patients with cirrhosis. The me- dian OS was 21.91 ± 3.60 versus 14.53 ± 1.61 months (P= 0.716), and median DFS was 13.98 ± 3.38 versus 14.03 ± 2.40 months (P= 0.078).

Gao et al. [81] found that a combination of surgical resection and PV perfusion chemotherapy for types I-III PVTT with liver cir- rhosis was associated with a lower rate of early recurrence ( ≤1 year) (40.3% vs. 64.2%,P= 0.006) and longer OS (19.0 vs. 13.4 months,P= 0.037) than surgical resection. Li et al. [82] observed that in Vp3-4 PVTT patients with cirrhosis, compared to those with only surgical resection, the one who received sorafenib after surgi- cal resection had longer OS (37 vs. 30 months,P= 0.01) and TTP (29 vs. 22 months,P= 0.041). The main AEs were hand-foot-skin reaction (91.67%), diarrhea (83.3%), and hypertension (83.3%), all of which were easily treated; hence, surgical resection plus sorafenib was considered tolerable and effective.

Systemic chemotherapy

Although the usage of traditional systemic chemotherapy in HCC patients with PVTT is limited due to the high-risk of toxicity, it is often used as a second-line of treatment or in combination therapy [83] .

Obi et al. [84] , in a prospective study, observed that a combina- tion of systemic interferon-α(IFN-α) and intra-arterial 5-FU could improve the OS in Vp3-4 PVTT patients with 12- and 24-month OS rates being 34% and 18%, respectively. For patients who achieved CR, the rates were 81% and 59%, and for patients who achieved PR, the rates were 43% and 18%. Nausea and vomiting were the main reported AEs. A retrospective study found that TACE plus apatinib might be an option for Cheng’s types I-II PVTT when contrasted with TACE alone (median OS: 13.7 vs. 7.2 months,P= 0.006; 12.2 vs. 7.5 months,P＜0.001; respectively) [85] .

A combination of subcutaneous IFN-αand intra-arterial 5-FU can be used as a postoperative adjuvant therapy for Vp3-4 PVTT with 1-, 3-, and 5-year OS rates being higher than surgical resec- tion alone (100%, 69%, 44% vs. 30%, 15%, 5%,P＜0.0 0 01) [86] .

HAIC

HAIC can directly deliver chemotherapy drugs to the tumor site, thus improving efficacy and limiting the incidence of systemic AEs [12] . Ikeda et al. [87] prospectively found that using cis- platin, HAIC could achieve a moderate activity in Vp3-4 PVTT and markedly prolong the survival time of patients who responded to it. Median PFS and OS were 3.6 and 7.6 months, respectively, while patients showing CR or PR exhibited more than 3 years of survival time.

Ahn et al. [88] compared sorafenib and HAIC for main PVTT and found no difference in the OS (6.4 vs. 10.0 months,P= 0.139). However, HAIC has a longer TTP (6.2 vs. 2.1 months,P= 0.006) and higher DCR (76% vs. 37%,P= 0.001), suggesting that HAIC could be used as an alternative therapy for sorafenib. Chong et al. [89] found that performing surgery after downstaging with con- current chemoradiotherapy followed by HAIC showed more bene- fits than surgery alone in Vp2-3 PVTT, with disease-specific sur- vival of 62 versus 15 months (P= 0.006). However, the inci- dence of complications was also markedly increased. Katamura et al. [90] also found that in Vp3-4 PVTT, HAIC combined with 3D- CRT could significantly improve the treatment response rate (75% vs. 25%,P= 0.012) in PVTT and reduce the incidence of portal hy- pertension related events (P= 0.0195) compared to HAIC using 5- FU/IFN-α. However, the combination had no significant advantage over the OS (7.5 vs. 7.9 months,P= 0.871).

Othermethods

RFA

Table 1 Different treatments for hepatocellular carcinoma with portal vein tumor thrombosis. Studies Therapeutic method Type of PVTT ( n ) Median survival (mon) Survival rate Yang et al. [19] Sorafenib Vp2 (12), Vp3 (17), Vp4 (24) 3.9 - - - - - - - - Yang et al. [19] MET Vp2 (3), Vp3 (33), Vp4 (18) 5.3 Nakazawa et al. [21] Sorafenib Vp3 (21), Vp4 (7) 4.8 Nakazawa et al. [21] RT Vp3 (19), Vp4 (9) 10.9 Kudo et al. [22] Lenvatinib - - 13.6 Kudo et al. [22] Sorafenib 12.3 Yang et al. [25] Sorafenib + cryoRx Vp1-3 (36), Vp4 (16) 12.5 Yang et al. [25] Sorafenib Vp1-3 (37), Vp4 (15) 8.6 Giorgio et al. [26] Sorafenib + RFA Vp4 (49) - - 10.7 1-, 2-, 3-year: 60.0%, 35.0%, 26.0% Giorgio et al. [26] Sorafenib Vp4 (50) 1-, 2-year: 37.0%, 0% Kim et al. [31] TACE Segmental PV (331) 1-, 3-, 5-year: 44.9%, 16.0%, 12.0% Le et al. [32] TACE - 8.4 1-, 3-year: 36.0%, 1.3% Tawada et al. [33] TACE Vp2 (12), Vp3 (17), Vp4 (4) 9.1 - Liu et al. [34] TACE Type I (90), type II (98) 6.1 1-, 2-year: 21.3%, 5.5% Tang et al. [36] TACE + 3D-CRT Type I (72), type II (64), type III (49) 12.3 1-, 2-, 3-year: 51.6%, 28.4%, 19.9% Tang et al. [36] TACE + surgery Type I (80), type II (66), type III (40) 10.0 1-, 2-, 3-year: 40.1%, 17.0%, 13.6% Luo et al. [37] TACE Vp1-2 (40), Vp3-4 (44) 7.1 3-, 6-, 12-, 24-month: 85.6%, 56.4%, 30.9%, 9.2% Xiang et al. [38] TACE Type I (26), type II (107), type III (96), type IV (121) 11.0 1-, 2-, 3-year: 45.3%, 27.7%, 19.3% Zheng et al. [39] TACE + RFA Type I (55), type II (57), type III (22) 29.5 1-, 3-, 5-year: 63.0%, 40.0%, 23.0% Zhu et al. [40] TACE + sorafenib Vp1-2 (17), Vp3 (19), Vp4 (10) 11.0 - - Zhu et al. [40] TACE Vp1-2 (13), Vp3 (21), Vp4 (11) 6.0 Meng et al. [41] TACE + sorafenib Vp1-3 (36), Vp4 (10) 12.7 6-month, 1-, 2-year: 89.5%, 51.3%, 16.2% Park et al. [43] Sorafenib Vp1-2 (121), Vp3-4 (48) 10.8 - - Park et al. [43] TACE + sorafenib Vp1-2 (122), Vp3-4 (48) 12.8 Kim et al. [45] 3D-CRT Vp3-4 (59) 10.7 for responders, 5.3 for non-responders 1-, 2-year: 40.7%, 20.7% for responders, 25.0%, 4.7% for non-responders Rim et al. [46] 3D-CRT Vp1-3 (26), Vp4 (19) 16.7 for responders, 8.0 for non-responders 1-year: 63.7% for responders, 28.2% for non-responders Su et al. [47] 3D-CRT Type I (23), type II (49), type III (62) 13.0 1-, 2-, 3-year: 54.0%, 33.0%, 18.0% Su et al. [47] Surgery Type I (75), type II (77), type III (37) 18.0 1-, 2-, 3-year: 62.0%, 47.0%, 43.0% Xi et al. [49] SBRT Vp1-3 (16), Vp4 (17) 13.0 1-year: 50.3% Hata et al. [51] PBT Vp1-3 (9), Vp4 (3) - 22.0 2-, 5-year: 88.0%, 58.0% Sugahara et al. [52] PBT Vp1-3 (15), Vp4 (20) 2-, 5-year: 48.0%, 21.0% Lu et al. [53] GKR Vp1-3 (36), Vp4 (28) 6.1 - Hou et al. [54] IMRT Vp1-3 (18), Vp4 (25) 15.47 1-year: 59.3% Hou et al. [54] 3D-CRT Vp1-3 (28), Vp4 (22) 10.46 1-year: 35.8% Li et al. [55] TACE Type I (55), type II (338), type III (330), type IV (12) 4.8 - Li et al. [55] RT + TACE Type I (8), type II (47), type III (54), type IV (3) 11.0 - Chu et al. [56] RT + TACE Vp1-2 (90), Vp3 (93), Vp4 (20) 16.4 1-, 3-year: 65.3%, 19.0% Chu et al. [56] TACE + sorafenib Vp1-2 (33), Vp3 (59), Vp4 (12) 12.0 1-, 3-year: 50.0%, 10.5% Yoon et al. [57] EBRT + TACE - 13.75 - - - Fujino et al. [58] 3D-CRT + HAIC Vp3 (18), Vp4 (23) 12.1 Lu et al. [59] GKS + TACE - 9.7 Mazzaferro et al. [61] Y90-RE Vp1-2 (29), Vp3 (5), Vp4 (1) 15.0 3-year: 16.0% Spreafico et al. [62] Y90-RE Vp1 (53), Vp2 (37), Vp3 (30) 14.1 1-, 3-year: 53.2%, 18.5% Edeline et al. [63] SIRT Vp1-3 (18), Vp4 (16) 18.8 12-, 24-, 36-month: 61.8%, 44.1%, 25.2% Edeline et al. [63] Sorafenib Vp1-3 (62), Vp4 (55) 6.5 12-, 24-, 36-month: 27.1%, 13.9%, 8.7% Vilgrain et al. [65] SIRT - - 8.0 1-year: 39.5% Vilgrain et al. [65] Sorafenib 9.9 1-year: 42.1% Akinwande et al. [66] Y90-RE Vp4 (20) 3.0 - ( continued on next page )

Table 1 ( continued ) Studies Therapeutic method Type of PVTT ( n ) Median survival (mon) Survival rate Kokudo et al. [67] LR Vp1 (893), Vp2 (528), Vp3 (466), Vp4 (206) 34.44 1-, 3-, 5-year: 70.9%, 43.5%, 32.9% Ban et al. [69] LR Vp3 (26), Vp4 (19) 20.0 1-, 3-, 5-year: 69.6%, 37.4%, 22.4% Fukumoto et al. [72] LR Vp4 (46) 15.0 1-, 3-year: 58.5%, 17.1% Yu et al. [75] LR Vp1-2 (31) - - 18.9 2-year: 75.8% Yu et al. [75] TACE + RT Vp1-2 (47) 2-year: 61.5% Sun et al. [77] LR + IMRT Type I (5), type II (15), type III (5), type IV (1) 1-, 2-, 3-year: 76.9%, 19.2%, 11.5% Wei et al. [78] Neoadjuvant 3D-CRT + LR Type II (41), type III (41) - - - 6-, 12-, 18-, 24-month: 89.0%, 75.2%, 43.9%, 27.4% Wei et al. [78] LR Type II (51), type III (31) 6-, 12-, 18-, 24-month: 81.7%, 43.1%, 16.7%, 9.4% Yamamoto et al. [79] LR Vp1 (63), Vp2 (10), Vp3 (6), Vp4 (4) 5-year: 58.3% Bai et al. [80] LR + TACE Type I (17), type II (14) 21.91 6-, 12-, 24-month: 80.0%, 53.3%, 31.8% Bai et al. [80] LR + RT Type I (6), type II (4) 14.53 6-, 12-month: 88.9%, 71.1% Gao et al. [81] LR + PVC Types I-II (46), type III (21) 19.0 1-, 2-, 3-, 5-year: 63.8%, 37.9%, 24.4%, 18.3% Li et al. [82] SR + sorafenib Vp3 (10), Vp4 (2) 37.0 - Obi et al. [84] Intra-arterial 5-FU + systemic IFN- α Vp3 (88), Vp4 (28) 6.9 6-, 12-, 24-month: 53.0%, 34.0%, 18.0% Fan et al. [85] Apatinib + TACE Vp1-2 (18), Vp3 (51), Vp4 (16) 12.0 - - Fan et al. [85] TACE Vp1-2 (19), Vp3 (54), Vp4 (30) 7.0 Nagano et al. [86] SR + subcutaneous IFN- α + intra-arterial 5-FU Vp3 (6), Vp4 (24) - 1-, 3-, 5-year: 100.0%, 69.0%, 44.0% Nagano et al. [86] SR Vp3 (10), Vp4 (10) - 7.6 1-, 3-, 5-year: 30.0%, 15.0%, 5.0% Ikeda et al. [87] HAIC Vp3 (6), Vp4 (19) 1-, 2-, 3-year: 40.3%, 36.0%, 20.0% Ahn et al. [88] HAIC Vp4 (38) 10.0 - - - - - Ahn et al. [88] Sorafenib Vp4 (35) 6.4 Chong et al. [89] HAIC + CCRT + LR Vp2 (12), Vp3 (14) 62.0 Katamura et al. [90] HAIC + 3D-CRT Vp3 (8), Vp4 (8) 7.5 Katamura et al. [90] HAIC Vp3 (8), Vp4 (8) 7.9 Lu et al. [91] Percutaneous laser ablation Vp3 (58), Vp4 (50) - 5.0 1-, 2-, 3-year: 55.56%, 33.58%, 22.38% Vibert et al. [92] PVS Vp2-3 (41), Vp4 (13) 6-, 12-, 24-month: 47.0%, 44.0%, 36.0% Li et al. [93] PVS combined with 125 I particle chain implantation followed by TACE with As 2 O 3 Vp2 (6), Vp3 (14), Vp4 (10) 10.0 3-, 6-, 9-, 12-month: 83.1%, 69.2%, 43.7%, 31.2% Li et al. [94] PVS- 125 I-TACE-S Vp3 (12), Vp4 (6) 10.0 - - Liu et al. [95] FFBM capsule Vp3-4 (95) 4.0 PVTT: portal vein tumor thrombosis; MET: metronomic chemotherapy; RT: radiotherapy; cryoRx: cryotherapy; RFA: radiofrequency ablation; TACE: transarterial chemoembolization; PV: portal vein; 3D-CRT: 3-dimensional conformal radiotherapy; SBRT: stereotactic body radiotherapy; PBT: proton beam therapy; GKR: gamma knife radiosurgery; IMRT: intensity modulated radiation therapy; EBRT: external beam radiotherapy; HAIC: hepatic arterial infusion chemotherapy; GKS: gamma knife surgery; Y90-RE: Yttrium-90 radioembolization; SIRT: selective internal radiation therapy; LR: liver resection; PVC: portal vein perfusion chemotherapy; SR: surgical resection; 5-FU: 5-fluorouracil; IFN- α: interferon- α; HAIC: hepatic arterial infusion chemotherapy; CCRT: concurrent chemoradiotherapy; PVS: portal vein stenting; As 2 O 3 : arsenic trioxide; PVS- 125 I-TACE-S: portal vein stent combined with iodine- 125 seed-strips followed by transcatheter arterial chemoembolization with sorafenib; FFBM: Fufang Banmao.

RFA conducts the radiofrequency to the tumor site through a probe, causing the tumor to heat up and necrotize. Percutaneous laser ablation delivers laser beam via an optic fiber to tumor site guided by color Doppler ultrasonography, which then causes necrosis. Lu et al. [91] confirmed its efficacy in Vp3-4 PVTT and found that the 1-, 2-, and 3-year OS rates were 55.56%, 33.58%, and 22.38%, respectively. It could also relieve portal hypertension- related symptoms. Electrochemotherapy can increase cell mem- brane permeability through electroporation to facilitate entry of chemotherapy drugs. Tarantino et al. [5] confirmed that elec- trochemotherapy was safe and feasible for Vp3-4 PVTT, where two of six patients achieved PV recanalization.

Portal stenting

Portal stenting could improve the patency of PV, making subse- quent treatment possible. The 1-, 6-, and 12-month patency rates were 95%, 75%, and 75%, while 6-, 12-, 24-month OS rates were 47%, 44%, and 36%, respectively. Also, patients without pre-stenting jaundice had a better prognosis [92] . A recent study confirmed that PV stent combined with125I particle chain implantation followed by TACE using As2O3was safe and effective in Vp2-4 PVTT, and the median survival time was 10.0 ± 5.1 months with no AEs ( ≥3 grade) [93] . Li et al. [94] showed that PV stent combined with iodine-125 seed-strips followed by TACE-sorafenib was more effec- tive than TACE-sorafenib alone, especially for Vp3-4 PVTT.

Traditional Chinese medicine

In recent years, the research on traditional Chinese medicine has been increasing gradually. Liu et al. [95] found that the Fu- fang Banmao capsule could prolong the OS of Vp3-4 PVTT patients compared to only the supportive treatment (4.0 vs. 2.2 months,P= 0.0 0 04). Chen et al. [96] proved that Huaier granule could sig- nificantly prolong the recurrence free survival (62.39% vs. 49.05%,P= 0.0 0 01) and reduce the extrahepatic recurrence rate (8.6% vs. 13.61%,P= 0.0018) in patients who had recurrence after surgery when contrasted with no further treatment.

Different treatments for HCC with PVTT from literatures are summarized in Table 1 .

Perspectives

Enhancing diagnostic accuracy of PVTT

PVTT is mainly diagnosed by computed tomography, magnetic resonance imaging, etc. The potentially missed diagnosis and mis- diagnosis can affect the choices of treatment methods. In recent years, the field of artificial intelligence has been developing rapidly. Lambin et al. [97] proposed the concept of radiomics, which refers to high-throughput extraction of imaging features to establish a predictive model, which can be used to diagnose HCC with mi- crovascular invasion before treatment [ 98 , 99 ], and more studies are expected to focus on it.

Establishing a prognostic classification model for personalized treatment

HCC patients with PVTT are a heterogeneous group. The tumor information, PVTT type, and liver function, can help in determining the appropriate treatment methods. While different choices have different effects on the same population, it is necessary to select patients with a good prognosis for specific therapeutic methods. Cao et al. [100] established a prognostic classification model for HCC with macroscopic vascular invasion, which combined a tree- based classification system and HVTT-PVTT score to guide person- alized therapy. Liu et al. [101] used a nomogram to predict the sur- vival time of HCC patients with PVTT treated by conservative treat- ment, suggesting that the lower-risk group should consider further active treatment.

Innovation on treatment methods

Aggressive treatment strategies and various modalities of com- bination therapy have achieved good results in HCC with PVTT cases. Finn et al. [10] recently proposed a new regimen for un- resectable HCC, namely the T plus A scheme (atezolizumab and bevacizumab), which had longer PFS than sorafenib (6.8 vs. 4.3 months,P＜0.001). It is expected to replace sorafenib as the first- line of treatment for HCC. However, there is a need to further explore whether its success is due to the synergistic or superim- posed effects of two drugs and understand its therapeutic effect on PVTT. Future research should also focus on balancing the in- creased efficacy of combination therapies while reducing toxicities along with exploring the possibility of immunotherapy combined with targeted therapy.

Conclusions

HCC with PVTT is still a thorny problem, and the current treat- ment choices are unable to extend the survival effectively. The existing researches show that although the effect of sorafenib monotherapy is poor, combining it with other methods can im- prove survival. Each type of PVTT can benefit better from TACE and sorafenib than sorafenib monotherapy. Tumor downstage can pos- sibly be realized after TACE, but the invasion of the tumor into the main trunk of PV greatly impairs the efficacy of TACE. Surgery is a curative choice but is often not recommended for Vp4 PVTT. Its indications can be broadened by new methods that need further exploration. RT can decrease the possibility of Vp3 progression to Vp4 but building a forecast model of the best radiation dose and response is necessary. Systemic chemotherapy, HAIC, RFA, portal stenting, and traditional Chinese medicine have also proven benefi- cial in implementing aggressive choices of Vp3-4 PVTT. Radiomics, a new technique, makes accurate diagnosis of PVTT before treat- ment feasible. Establishing some prognostic classification models directs the selection of personalized treatment to achieve the best prognosis. The good survival benefits of the T plus A scheme in- dicate the importance of exploring new treatment methods. How- ever, a lot remains to be done in the treatment of HCC with PVTT.

Acknowledgments

None.

CRediTauthorshipcontributionstatement

Zi-WenTao:Data curation, Formal analysis, Writing - original draft.Bao-QuanCheng:Supervision, Writing - review & editing.TaoZhou:Supervision, Writing - review & editing.Yan-JingGao:Conceptualization, Supervision, Writing - review & editing.

Funding

None.

Ethicalapproval

Not needed.

Competinginterest

No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the sub- ject of this article.