Oncological benefits of portal vein embolization for patients with hepatocellular carcinoma

Abstract Portal vein embolization (PVE) for hepatocellular carcinoma (HCC) was first introduced in 1986 and has been continuously developed throughout the years. Basically, PVE has been applied to expand the indication of liver resection for HCC patients of insufficient future liver remnant. Importantly, PVE can result in tumor progression in both embolized and non‐embolized livers; however, long‐term survival after liver resection following PVE is at least not inferior compared with liver resection alone despite the smaller future liver remnant volume. Five‐year disease‐free survival and 5‐year overall survival were 17% to 49% and 12% to 53% in non‐PVE patients, and 21% to 78% and 44% to 72% in PVE patients, respectively. At present, it has proven that PVE has multiple oncological advantages for both surgical and nonsurgical treatments. PVE can also enhance the anticancer effects of transarterial chemoembolization and can avoid intraportal tumor cell dissemination. Additional interventional transarterial chemoembolization and hepatic vein embolization as well as surgical two‐stage hepatectomy and associated liver partition and portal vein ligation for staged hepatectomy can enhance the oncological benefit of PVE monotherapy. Taken together, PVE is an important treatment which we recommend for listing in the guidelines for HCC treatment strategies.

TA B L E 1 Long-term outcome after hepatectomy for hepatocellular carcinoma patients with or without portal vein embolization  been found to be significantly greater than that of the traditional volume ratio of the remnant liver. 16 In regard to patients with HCC, PVE can provide not only an increased remnant liver volume but also an enhanced effect of transarterial treatments [17][18][19][20][21][22][23] and the prevention of transportal metastases   to non-embolized areas. 24 PVE has a lower direct therapeutic effect   on HCC, thus it can cause tumor progression while waiting for liver regeneration. [25][26][27][28][29] Previous papers have confirmed comparable disease-free survival (DFS) and overall survival (OS) rates for HCC patients undergoing major hepatectomy with or without PVE. 3,5,[30][31][32] In contrast, two papers have demonstrated a better DFS or OS in patients that had received PVE as compared to the patients who did not. 33,34 Whether PVE might show a better influence on recurrence or long-term prognosis after major hepatic resection remains controversial.
There have been numerous review articles about PVE; however, articles specific for HCC remain limited and all of them mainly discussed the PVE procedure and liver regeneration effect. [35][36][37] In this review, we will summarize the role of PVE for HCC with special attention to oncological effects.

| INFLUEN CE OF PVE ON TUMOR PROG RE SS I ON
Several studies have suggested that tumor progression can occur after PVE in both embolized and non-embolized livers; however, data remain inconclusive. [25][26][27][28][29] Tumor progression after PVE has been reported to be influenced by the following factors: (a) malignant potential of the primary tumor, (b) alterations of hepatic blood supply to the tumor, (c) acceleration of inflammatory cytokines and growth factors, and (d) an enhanced cellular host response. 25,26,38 Unilateral reduction of portal blood flow after PVE causes a compensatory increase in hepatic artery blood perfusion (hepatic arterial buffer response). As HCC tumors are mainly fed by arterial blood supply, PVE can potentiate local tumor growth. 39 Using a rat portal vein ligation (PVL) model, it was found that hepatocyte growth factor (HGF) mRNA levels increased to a detectable level 6 to 24 hours after the operation in non-ligated lobes, but was only slightly elevated in ligated atrophic lobes. 40

| EFFEC TS OF PREOPER ATIVE PVE ON LONG -TERM OUTCOME
The DFS and OS for HCC after major hepatectomy with or without PVE are summarized in Table 1 even after propensity score-matching (PSM) analysis. After PVE, the %FLR was identical in the two groups (50% vs 52%).
To resolve these clinical questions, we conducted a multicenter study using PSM analysis for patients with HCC (≥5 cm) that underwent PVE followed by right-sided hemi-hepatectomy. 5 In the overall cohort of patients with or without PVE before PSM, RFS and OS in the PVE group were significantly greater than those in the non-PVE group (P < .005 for RFS and P < .037 for OS) ( Figure 1A,B); however, the application of PVE was not an independent prognostic factor for RFS and OS by multivariate analysis. In contrast, in the PSM cohort, patients treated with PVE showed at least a non-inferior long-term prognosis as compared to patients undergoing upfront hepatectomy despite the smaller FLR ( Figure 1C,D). Furthermore, 10 random PSM analyses (

| CLINIC AL B ENEFITS OF PVE IN NONSURG IC AL THER APY
The indication of PVE for HCC patients who required portal vein occlusion other than liver regeneration for major hepatectomy is summarized (Table 3). There were a few reports demonstrating total or subtotal pathological necrosis of HCC after PVE monotherapy. [54][55][56] These results were not predicted because HCC was fed by a predominant arterial flow. 38 This may be due to the fact that well-differentiated HCC could instead be mainly fed by portal flow.  There have been four papers comparing long-term outcomes between HCC patients treated with TACE plus PVE versus PVE alone followed by major hepatectomy (Table 4). 18 group, and 0%-38% and 20%-58% in the PVE-alone group, respectively. By univariate analysis, the PVE + TACE group provided better RFS and OS compared with the PVE-alone group. However, the data were inconclusive because there were no multivariate analyses, PSM studies as well as RCT. Recently, intent-to-treat analysis data were published investigating sequential TACE plus PVE (n = 27) versus PVE alone (n = 28) before major hepatectomy for patients with large HCC (≥5 cm). 23 Baseline characteristics were equivalent in the two groups. The number of dropout patients for liver resection were only two (9%) in the TACE + PVE group and nine (32%) in the PVE-alone group. OS was significantly better in the former as compared to the latter (3-year OS of 60% vs 20%; P = .01).

| Additional TACE on PVE
Sequential TACE and PVE have been carried out as an order of TACE followed by PVE with an interval of 2 or 3 weeks. [17][18][19][20][21][22][23] In contrast, our group performed with an inverted order. 34 The reasons for the "PVE-first approach" included: (a) the extent of liver regeneration depended on the interval period between PVE and liver resection, and (b) PVE was required to achieve complete obliteration, so regulation of the procedure was difficult. In the PVE-first approach, delicate TACE can be applied with minimal arterial obstruction of the surrounding liver tissue in the portal embolized liver.

| Additional hepatic vein embolization on PVE
Recently, staged and simultaneous preoperative portal and hepatic vein embolization (biembolization) have been introduced and described to create higher liver hypertrophy than PVE alone before major liver resection. 68,69 However, this approach involves a prolonged waiting period and thus further increases the risk of tumor progression. Unfortunately, oncological effects of biembolization for HCC have not been fully investigated. 68 It is noteworthy that additional HVE on PVE might be able to decrease both intra-and extrahepatic metastases.

| Advances in operative procedure using PVE
Two-stage hepatectomy has been developed for bilateral liver tumors. The first step includes tumor enucleation of the FLR followed by PVE or PVL, while the second step involves major hepatectomy. [70][71][72] Associated liver partition and PVL for staged hepatectomy (ALPPS) is a novel operative procedure consisting of two steps: (a) PVE or PVL and liver transection with or without tumor enucleation from the residual liver, and (b) major hepatectomy. [70][71][72] ALPPS technique has been described to obtain an increased volume of PVE and a decrease in dropout rates. 70 The 3-and 5-year DFS and OS rates were 34.9%, and 25.0%, and 41.8%, and 40.7%, respectively (P = .267), and 60.2% and 46.8%, and 73.5%, and 64.1%, respectively (P = .234), in the ALPPS and PVE groups. Recently, several types of modified ALPPS procedure including laparoscopic partial ALPPS have been developed mainly to decrease morbidity and mortality. 73,74 For patients with colorectal liver metastases, ALPPS can provide greater liver hypertrophy in a shorter period as compared to TSH; however, in some papers, early recurrence and poor OS is indicated. 75,76 Excessive production of various inflammatory cytokines and growth due to rapid liver regeneration is thought to be one of the reasons. Thus, further studies comparing ALPPS and TSH for HCC are strongly required.
In conclusion, PVE has been developed mainly to achieve hypertrophy of the non-embolized liver. We would like to emphasize that