The treatment of non-metastatic hepatic malignancies (hepatocellular carcinoma [HCC] and intrahepatic cholangiocarcinoma [ICC]) consists primarily of surgical resection when possible and only if the cancer is limited to the liver.1 As the overall prognosis for breast and colorectal metastases to the liver (and only the liver) improves with advances in systemic therapies, more and more patients are undergoing liver resection for limited or oligometastatic disease.2–4 Often, even in this setting, patients will not have adequate future liver remnant (FLR) to survive after resection and die from liver failure, especially in the setting of cirrhosis or chronic liver disease.5 Therefore, it is critical to ensure adequate FLR in this patient population.6

There are two options to increase the size of the FLR, minimize the risk of post-operative liver failure and death after liver resection - portal vein embolization (PVE) and trans-arterial radioembolization (TARE), with PVE being the most common approach.7–9 TARE is a minimally invasive percutaneous procedure, where yttrium-90 filled microspheres are injected intra-arterially as a means of internal radiation.10 The microspheres deliver a high dose of radiation to the tumor. Retrospective clinical studies have shown that the use of radioembolization can help extend the lives of some patients with unresectable tumors and simultaneously improve their quality of life, but it is not a cure and is less effective than resection.4,11 TARE injection is also associated with 35-40% increase in mass of the contralateral liver lobe, and marked ipsilateral lobe atrophy, a phenomenon termed “radiation lobectomy”.9,12–16

Liver resection after PVE is standard of care when the FLR grows adequately.17–19 Liver resection after TARE is not standard of care, however, is sometimes performed on a highly selective basis when PVE fails in this setting, or patients have an uncharacteristically good response.20 Importantly, no published studies have investigated the efficacy or safety of the use of TARE followed by liver resection for HCC or ICC in large series. Similarly, there have been no articles published that investigate TARE followed by resection as a treatment option for patients diagnosed with CRLM or NET.

The purpose of this study was to evaluate the use of TARE and PVE as options for increasing FLR in patients with unresectable non-transplantable hepatic malignancies. We hypothesize that TARE is an effective option as a bridge to resection in highly selected patients where PVE has failed.


Subjects. This study was a retrospective case-control study, from 2008 to 2018. The population consisted of non-liver transplant eligible patients who underwent TARE, PVE, or liver resection in Memphis, TN. Patients were propensity matched based on co-morbidities, etiology of liver disease, Child-Pugh Score, and ECOG performance status. Additionally, patients were divided into subpopulations based on principal diagnosis. This work was approved by the UTHSC (University of Tennessee Health Science Center) Institutional Review Board.

Statistical analyses. Statistical analyses were performed with chi-square, using SPSS (Statistical Package for the Social Sciences) (Statistical Package for the Social Sciences), in a histology-specific fashion. Survival analysis was completed in R version 4.0.4 using the packages survival and survminer.21,22 Statistical significance in survival analysis was determined by the log-rank test to compare the different treatment groups.

To understand the racial variations in the use of TARE and PVE, a chi-square test of independence was performed to examine the relation between race and embolization procedure. Similarly, to better comprehend how the principal diagnosis affects the choice of embolization procedure, a second chi-square test of independence was performed to study the relation between histology and embolization procedure.


Patients. Of the 196 patients, the average age was 57.8±12.6 years and 148 underwent liver resection. There were 119 male patients (60.7%). There were 137 Caucasian patients (CC, 70%) and 59 African American patients (AA, 30%). There were 61 HCC, 31 ICC, 65 CRHM, and 12 NET patients. 28 patients had other diagnoses (other). Of the 11 patients who received PVE, all but 1 eventually underwent liver resection (Figure 1). Of the 56 patients who received TARE, 9 (16%) patients eventually underwent liver resection.

Figure 1
Figure 1.Schematic used to subdivide patients based on treatment modality
Table 1.Patient distribution according to diagnosis within the “Other” category
Diagnosis Number of Patients
Acinar cell pancreatic carcinoma 1
Adrenocortical Carcinoma 1
Angiomyolipoma 1
Breast Cancer 3
Desmoplastic Small Cell Tumor 1
Fibrous histiosarcoma 1
Gastrointestinal Stromal Tumor 1
Hepatic Sclerosing Carcinoma 1
Intraductal Papillary Mucinous Neoplasm 1
Leiomyosarcoma 1
Metastatic Melanoma 5
Ovarian cancer 3
Pheochromocytoma 1
Metastatic Pancreatic Adenocarcinoma 2
Renal Cell Carcinoma 2
Spindle Cell Tumor 1
Squamous Cell Carcinoma of the gallbladder 1
Telangiectatic Adenoma 1
Total 28

Pre-operative embolization before resection. Of the 148 patients who underwent resection, 20 (13.5%) received pre-operative embolization procedures (Figure 2). 10 patients underwent PVE, 9 patients underwent TARE, and 1 patient underwent both TARE and PVE (Figure 2). Patients with primary liver malignancies (HCC and ICC) were more likely to undergo pre-operative embolization procedures (Figure 3). Overall, CC patients were less likely than AA patients to undergo pre-resection embolization (P=0.03, Figure 4). TARE-only treatment served as a bridge to resection for 9 patients: HCC (n=3), ICC (n=2), CRHM (n=2), NET (n=1), and other (n=1). All of these 9 patients were Caucasian. All 9 patients that underwent preoperative TARE had some degree of cirrhosis, the most common cause being hepatitis virus infection (n=6). The single patient that underwent both PVE and TARE preoperatively did not have a good response to PVE, and thus underwent TARE in hopes of increasing FLR.

Figure 2
Figure 2.Distribution of patient population according to treatment

There are 196 patients, 11 of which underwent PVE, 148 of which underwent surgery, and 56 of which underwent TARE. Only 1 patient has undergone all 3 procedures. Out of the 56 that underwent TARE, 8 also underwent surgery. Out of the 11 that underwent PVE, 9 also underwent surgery.

Figure 3
Figure 3.Patient distribution according to histology and type of embolization procedure

There were 61 HCC patients, 30 ICC patients, 65 CRHM patients, 12 mNET patients, and 28 other.

Figure 4
Figure 4.Patient distribution according to race and embolization procedure

Of the 137 CC patients, 31% underwent some form of embolization procedure. Of the 59 AA patients, 48% underwent some form of embolization procedure.

PVE and TARE are associated with histology and diagnosis. The relation between histology and type of embolization procedure was significant, (p = 0.043. Figure 3). PVE as a preoperative strategy was used primarily in patients diagnosed with CRLM. TARE as a preoperative strategy was used in patients with all histologies.

Table 2.Cross tabulation of patients according to race and type of embolization procedure
Embolization Total
Race Caucasian 95 30 9 3 137
African American 31 25 1 2 59
Total 126 55 10 5 196
Table 3.Cross tabulation of patients according to principal diagnosis and type of embolization procedure
Embolization Total
Diagnosis HCC 36 20 1 4 61
ICC 18 11 1 0 30
CRLM 43 13 8 1 65
NET 7 5 0 0 12
Other 22 6 0 0 28
Total 126 55 10 5 196

Racial variations in the use of TARE and PVE. The relation between race and type of embolization procedure was significant with more AA patients undergoing any embolization procedure compared to none (p = 0.018, Figure 4). For the patients that underwent PVE preoperatively, 1 out of 10 were African American while all others were Caucasian.


Although TARE as a downstaging therapy has not been widely applied as a bridge to resection, our data supports its incorporation in currently employed strategies. The combined anti-tumor properties of the Y90 radiation and their ability to induce hypertrophy of the FLR are promising, particularly for those patients that do not meet the criteria for resection at the time of diagnosis or are deemed too high risk.

However, with any new intervention, long-term outcomes and complications must be considered to determine the therapies’ effectiveness and safety. In the case of TARE, limited data exists regarding its safety as a preoperative strategy. In the absence of guidelines for downstaging strategies for patients with unresectable status, it is difficult to determine which strategy provides the greatest reduction in tumor burden, while preserving liver function to become eligible for surgery at a latter point. There are theoretical advantages of TARE over PVE as an oncologic strategy. Likewise, PVE may have some technical advantages related to diversion of blood flow.

Our study introduces the concept of TARE as an efficacious preoperative strategy for patients whose tumor burden had previously been deemed unresectable or have otherwise very large, non-metastatic tumors. We identified 9 patients in which TARE was successfully used to bridge patients to resection. This equates to a 16% success rate in converting patients from unresectable to resectable status. Due to the retrospective nature of this analysis, it is highly likely that not all the patients were approached with an intent to bridge to resection. We believe that this would increase the success rate of safe bridge to resection strategy. The common thread among these patients was the presence of advanced liver disease and recurrence.

Overall, AA patients were more likely to undergo any type of embolization procedure. However, they were less likely to undergo PVE than Caucasian patients. Our expectation would be that since HCC is more common in AA patients,23,24 they would undergo embolization more often than their CC counterparts, which our data supports. This could be explained by disparate access to healthcare, which is well documented in the mid-South.25,26 Our data also suggests that as TARE became more widely available later, it could have been preferentially selected for these patients due to comorbidities or advanced disease.

HCC patients were more likely to undergo any embolization than patients diagnosed with other types of malignancies. Patients that present with HCC typically have other underlying liver diseases, such as cirrhosis and/or hepatitis, increasing the risk of post—operative liver failure. Thus, to preserve liver function, these patients are more likely to undergo an embolization procedure before being subjected to surgical management. Due to the ability to use TARE as a bridge to liver transplants some of these patients may also have been treated with this strategy.

There are several limitations to this study. First, the investigation is limited to a single-institutional experience and is retrospective in nature. Second, the analysis was performed in a highly select group of patients, yielding small population and sample sizes, particularly when looking at patients who underwent an embolization procedure before resection. Third, the retrospective nature and lack of true control group are also major limitations. Lastly, this study looks exclusively at demographic data, and thus all conclusions reached are associations.


Future studies should look at more causative relationships between TARE and achieving resectable status. Our data suggests a meaningful proportion of patients can be successfully bridged to resection demonstrating the importance of incorporating surgeons into the multi-disciplinary care of these patients. Similarly, an analysis of the survival and long-term outcomes in this population is long overdue. And to further complete the picture, a socioeconomic analysis of patients that have undergone TARE and/or PVE before surgery would be of interest to medical professionals.


Based on the retrospective analysis, TARE seems to be an effective bridging strategy in patients who have been deemed unresectable, or previously failed PVE. Long-term outcomes have yet to be determined with regards to TARE as a preoperative strategy and its impact on liver function, tumor recurrence, and mortality post-operative in these patients. As such, future studies are needed to define the role of TARE in patients with hepatic malignancies.