Abstract
Background/Aim: Balloon-occluded infusion of fragmented gelatin particles of transarterial embolization (BOIG-TAE) is a chemotherapy-free technique combining selective balloon occlusion with gelatin particle embolization.
Patients and Methods: This prospective, single-arm study (June 2022–June 2025) enrolled patients with hypervascular hepatic metastases refractory or intolerant to standard systemic therapy, with safety as the primary endpoint and preliminary efficacy and hepatic function as secondary endpoints. Adverse events were evaluated using Society of Interventional Radiology (SIR) criteria. Tumor response was independently assessed two months after treatment using modified Response Evaluation Criteria in Solid Tumors (mRECIST) and hepatic function was monitored using Child–Pugh and albumin–bilirubin (ALBI) scores.
Results: Six patients were enrolled (median age, 77 years; 4 male patients). One patient experienced transient abdominal pain lasting one week (SIR grade C); no other major complications occurred. Transient elevations in aspartate aminotransferase and alanine aminotransferase occurred in all patients and resolved within one month. No biliary or renal complications were recorded. The objective response rate at two months was 66.7% (4/6) and Child–Pugh and ALBI scores remained stable during follow-up.
Conclusion: BOIG-TAE may be feasible and generally well tolerated in selected patients, with potential for short-term tumor control while preserving hepatic function.
Introduction
Treatment strategies for hepatic metastases vary depending on the primary tumor, but chemotherapy and surgical resection are the primary options (1, 2). However, for hypervascular hepatic metastases such as neuroendocrine neoplasms (NEN) and gastrointestinal stromal tumors (GIST), transarterial chemoembolization (TACE), hepatic artery embolization (TAE), or transarterial radioembolization are also valid choices (3, 4). Reports on TACE drugs are diverse, with documented use of doxorubicin, mitomycin C, and streptozocin (5), while TAE/TACE embolization agents (e.g., iodized oil, gelatin, or polyvinyl alcohol particles) have also been studied (6). However, conclusive evidence on the most effective procedure, as well as which drugs or embolization materials are superior, remain scarce (7).
Although TACE is an established locoregional therapy for hepatocellular carcinoma (HCC), outcomes are influenced by multiple methodological variables (most notably the choice, size, and delivery of embolic materials), making standardization challenging (8). Balloon-occluded alternative infusion of transarterial chemoembolization (BOAI-TACE), which combines transient balloon occlusion with intra-arterial infusion of cisplatin solution and fragmented gelatin particles, further extends this concept and has been reported to improve local tumor control while reducing post-procedural deterioration of hepatic function compared with conventional techniques (9). In addition, BOAI-TACE has demonstrated favorable outcomes, even in patients with HCC beyond the up-to-7 criteria, highlighting the clinical relevance of hemodynamic modulation and embolic delivery strategy (10).
These observations support the hypothesis that balloon-mediated flow control can enhance selective intratumoral delivery of embolic material and mitigate non-target embolization.
We therefore developed balloon-occluded infusion of fragmented gelatin particles of transarterial embolization (BOIG-TAE) as a chemotherapy-free evolution of balloon-occlusion techniques. In BOIG-TAE, mechanically fragmented gelatin particles are infused under balloon occlusion to induce ischemia in hypervascular metastatic lesions while aiming to preserve surrounding hepatic parenchyma through improved selectivity and the temporary nature of gelatin embolization. Because therapeutic options are limited for patients with hypervascular hepatic metastases who are refractory or intolerant to standard systemic therapy, this prospective pilot trial evaluated the short-term safety of BOIG-TAE as the primary endpoint, with preliminary efficacy and hepatic function as secondary endpoints.
Patients and Methods
Ethics. This prospective, single-arm study was performed at a single institution from June 2022 to June 2025. This study was approved by the University of Tsukuba Clinical Research Review Board (CRB3180028) and written, informed consent was obtained from individual participants. This study complied with the tenets of the Declaration of Helsinki. This study was registered in the Japan Registry of Clinical Trials. The clinical trial registration number is jRCTs032220172, and the date of first registration was 30/08/2022.
Participant population. The inclusion criteria for participants in this study were as follows: (i) patients with hepatic hypervascular metastases [i.e., hyperenhancing or hyperintense compared to liver parenchyma in the arterial phase on contrast enhanced computed tomography (CT) or magnetic resonance imaging (MRI)]; (ii) patients refractory or intolerant to standard treatment as indicated in the guidelines for the primary tumors; (iii) age 18 years or more; (iv) Eastern Cooperative Oncology Group performance status of 0 or 1; (v) and ability to give informed consent. All primary tumors were histologically confirmed by biopsy and/or surgical specimens. Hepatic metastases were diagnosed radiologically based on contrast-enhanced CT/MRI findings and clinical course; biopsy of liver lesions was not mandatory. The exclusion criteria for participants in this study were: (i) serum bilirubin levels of 2.0 mg/dl or higher; (ii) an estimated glomerular filtration rate (eGFR) of less than 40 ml/min/1.73 m2; (iii) platelet counts of 50,000/μl or lower; (iv) a prothrombin time international normalized ratio (PT/INR) of 1.5 or higher; (v) tumor invasion into the portal vein or hepatic vein; (vi) a tumor volume that accounts for more than 70% of liver volume; (vii) dilatation of bile ducts larger than the accompanying portal vein; (viii) severe arterial-portal or arterial-venous shunts that could interfere with treatment; (ix) severe mental disorders; (x) severe allergy to iodine contrast media or other drugs; (xi) and/or those deemed unsuitable by the principal investigator or a sub-investigator of the study. Note that presence or absence of extrahepatic metastases was not considered a criterion for inclusion or exclusion.
BOIG-TAE. BOIG-TAE was performed according to the following protocol: A 4Fr hook-shaped catheter was placed in the celiac artery then CT angiography was performed through the celiac artery to identify tumor feeding arteries. Contrast material was injected at 4 ml/s, for a total volume of 24 ml, and CT angiography acquisition was initiated 6 s after the start of injection. The obtained images were used to identify the segmental branches that served tumors. The need for treatment of extrahepatic parasitic arteries was determined by referring to CT angiography and pre-treatment CT images.
Next, fragmented gelatin particles were prepared. Fine fragmentation was used to reduce proximal occlusion. Gelatin particles sized to 1 mm (Gelpart; Nippon Kayaku, Tokyo, Japan) and moistened with contrast medium were placed in a 2.5-ml syringe and crushed by the pumping method using a three-way stopper as this technique can pulverize gelatin into approximately 200 micrometer particles (11).
Specifically, a second 2.5-ml syringe was connected via a 3-way stopcock, the lumen of which was gradually narrowed during repeated pumping to achieve fine fragmentation. A total of 80 mg of fragmented gelatin particles were suspended in 20 ml contrast medium. Because fragmented gelatin particles in contrast medium are prone to precipitation, the suspension was stirred immediately before use and injected promptly after mixing.
Then, a microcatheter with a balloon (1.9Fr Logos GrandMaster, Piolax, Tokyo, Japan) was placed in a segmental branch of the feeding artery. Fragmented gelatin particle suspension was then administered under balloon occlusion. After confirming stable positioning, the occlusion balloon was inflated to achieve flow control and minimize reflux, after which the suspension was injected slowly under fluoroscopic guidance. Administration was continued until near-stasis was achieved with cast-like filling of the feeding artery on fluoroscopy. All feeding arteries were also treated in this manner.
As this study targeted participants refractory or intolerant to standard systemic therapy, no chemotherapeutic agents were administered.
Participant characteristics and laboratory data. Participant data obtained included age, sex, performance status, histological type of primary tumor, comorbidities, previous medical history, Child-Pugh score, and warfarin history.
Serum white blood cell count, hemoglobin, platelet count, albumin, total bilirubin, aspartate aminotransferase (AST), alanine transaminase (ALT), γ-glutamyl transferase (GTP), alkaline phosphatase (ALP), lactate dehydrogenase (LDH), serum creatinine, eGFR, amylase, and PT/INR were measured within two weeks before treatment and then on the day after treatment, seven days after treatment, one month after treatment, and two months after treatment.
Primary and secondary endpoints. The primary endpoint in this study was safety for BOIG-TAE. Adverse events were evaluated according to the Society of Interventional Radiology (SIR) guidelines (12). Post-procedure laboratory data were assessed using Common Terminology Criteria for Adverse Events (CTCAE) version 5.0. Adverse events were assessed up to three months after treatment.
The secondary endpoint was objective response rate (ORR) two months after treatment according to Modified Response Evaluation Criteria in Solid Tumors (mRECIST) guidelines (13). Another secondary endpoint was evaluation of Child-Pugh and albumin-bilirubin (ALBI) scores two months after treatment (14).
Imaging assessment. Dynamic contrast-enhanced CT or MRI was used for imaging evaluation of hypervascular hepatic metastases; both CT and MRI slices were less than 2 mm thick. Anonymized images were evaluated using the central-review style by two board-certified radiologists (26 and 10 years of experience) who did not conduct BOIG-TAE. Both size and number of tumors were quantified. The number of tumors was counted to a maximum of 20. Pre- and post-treatment images were evaluated by mRECIST, which was prespecified because eligibility required hypervascular lesions and the expected early treatment effect was reduced arterial-phase enhancement.
Statistical power and analysis. For Phase I safety testing, the target number of cases was set at 10. If four cases of serious adverse events (as defined in the SIR guidelines) had occurred during the trial, the trial would have been discontinued and phase II testing would have been halted. Two consecutive, serious adverse events would have forced a safety evaluation committee to convene and determine whether to continue the trial.
Worsening of Child-Pugh and/or ALBI scores before and two months after treatment was evaluated using the Wilcoxon signed-rank test. Statistical analyses were performed using SPSS for Windows version 26 (IBM SPSS Inc., Chicago, IL, USA). The significance level was set at 0.05.
Results
Participant characteristics. Six participants were enrolled during the study period. Enrollment was lower than planned because few patients met the eligibility criteria and extension of the enrollment period was not expected to substantially increase accrual. Informed consent was obtained from all six cases and there were no cases of ineligibility or withdrawal of consent. Safety and efficacy evaluations were performed on all enrolled patients.
Characteristics of the participants enrolled in this study are summarized in Table I. The median participant age was 77 years (range=65-82, male/female: 4/2). The primary tumors were two neuroendocrine neoplasms, one mixed neuroendocrine non-neuroendocrine neoplasm, one gastrointestinal stromal tumor, one intrahepatic cholangiocarcinoma, and one colon cancer. Four participants were refractory, one was intolerant, and one was both refractory and intolerant to standard systemic therapy. The median maximum tumor diameter was 41 mm (range=23-61 mm). Pretreatment Child-Pugh scores were 5 in five participants and 6 in one participant; all were classified as Child–Pugh class A. The median pretreatment ALBI score was −2.77.
Participant characteristics.
BOIG-TAE and treatment efficacy outcomes. The median number of treated segmental arteries of the liver was 5 (range=3-8), while the mean amount of Gelpart used was 112 mg (range=48-160 mg).
The mRECIST assessments by the two readers were concordant for all participants. ORR of BOIG-TAE for hepatic metastases in this study was 66.7% [95% confidence intervals (CI)=22.3%-95.7%] (Table II). Representative cases are shown in Figure 1 and Figure 2.
Treatment efficacy of balloon-occluded infusion of fragmented gelatin particles of transarterial embolization (BOIG-TAE) for hepatic metastases.
Arterial phase of contrast-enhanced computed tomography before (A, B) and after (C, D) balloon-occluded infusion of fragmented gelatin particles of transarterial embolization (BOIG-TAE) in a 79-year-old man with gastrointestinal stromal tumor (GIST), refractory to imatinib, sunitinib, regorafenib, and pazopanib. Six metastases were found in the liver, including a 44 mm mass in S4/8 (A, arrow) and a 16 mm mass in S6 (B, arrowhead). The contrast effect of metastases disappeared after treatment and complete response (CR) was achieved (C, D). During preparation for proton beam therapy, a hepatic subcapsular hemorrhage developed while attempting to place metallic fiducial markers (curved arrow).
Arterial phase of contrast-enhanced computed tomography before (A, B, C) and after (D, E, F) balloon-occluded infusion of fragmented gelatin particles of transarterial embolization (BOIG-TAE) in a 76-year-old woman with intrahepatic cholangiocellular carcinoma (ICC). For postoperative recurrence of ICC, gemcitabine plus cisplatin combination therapy, durvalumab, and S-1 (tegafur/gimeracil/oteracil potassium) were administered, but all were either ineffective or intolerable. Seven metastases were found in the liver, including a 23 mm mass in S8 (A, arrow), 21 mm mass in S5/6 (B, arrowhead), and a 19 mm mass in S1 (C, curved arrow). The contrast effect of metastases disappeared after treatment and complete response (CR) was achieved (D, E, F).
Hepatic function outcomes. All paired Child-Pugh scores were identical; therefore, no statistical differences were observed. The median ALBI score two months after BOIG-TAE was −2.72. No significant differences were observed between ALBI scores before or after BOIG-TAE (p=0.6875). The Child-Pugh scores before and after BOIG-TAE are presented in Table III and the ALBI scores in Table IV.
Child-Pugh scores before and after balloon-occluded infusion of fragmented gelatin particles of transarterial embolization (BOIG-TAE).
Modified albumin-bilirubin (ALBI) grades before and after balloon-occluded infusion of fragmented gelatin particles of transarterial embolization (BOIG-TAE).
Adverse events. Five out of six participants experienced abdominal pain and nausea consistent with post-embolization syndrome. In one patient, abdominal pain persisted for seven days and was classified as a grade C adverse event according to the SIR guidelines. No other grade C or higher adverse events were observed. No biliary complications, clinical liver failure, pancreatitis, or renal adverse events were observed during follow-up.
Overall, laboratory abnormalities peaked on the day after BOIG-TAE and were transient in most participants, with recovery to baseline or near-baseline levels within one month. No leukopenia was observed. Transient leukocytosis occurred in one participant on the day after BOIG-TAE. Anemia was present in all six participants (grade 1 in four and grade 2 in two); five had anemia before treatment and all improved to grade ≤1 within two months after BOIG-TAE. Platelet counts decreased in all participants compared with pretreatment levels. Grade 1 thrombocytopenia developed in four participants but all recovered to baseline levels after two months. AST elevation was noted in all six participants: grade 2 in two and grade 4 in two, peaking on the day after BOIG-TAE and returning to baseline within one month in all cases. ALT increased in four participants (grade 3 in two), also peaking within one day after BOIG-TAE and normalizing within one month. ALP elevation occurred in five participants (grade 1 in four and grade 2 in one), most of whom improved within one month. GTP increased in five participants (grade 2 in three and grade 3 in one) and one patient had persistent elevation at two months. Total bilirubin increased in three participants (grade 2 in two and grade 1 in one), peaking on the day after BOIG-TAE but returning to baseline levels thereafter. No elevation in serum creatinine or amylase was observed throughout the observation period. PT/INR prolongation was seen in two participants (grade 1 in one and grade 2 in one), with normalization in one participant by two months. Hypoalbuminemia was observed in five participants (grade 1 in two, grade 2 in two, and grade 3 in one) and all improved to grade ≤1 within two months after BOIG-TAE.
Upon post-protocol chart review for delayed adverse events through six months after BOIG-TAE, all participants were alive and no additional adverse events were identified.
Discussion
In this prospective pilot study, balloon-occluded infusion of fragmented gelatin particles (BOIG-TAE) achieved an objective response rate of 66.7 % in patients with hypervascular hepatic metastases refractory or intolerant to systemic therapy. Only one grade C (transient abdominal pain lasting one week) and no higher adverse events were observed according to SIR criteria. Neither Child-Pugh nor ALBI scores deteriorated during the 2-month follow-up. These findings suggest that BOIG-TAE may be a feasible and well-tolerated locoregional option for patients with limited therapeutic alternatives.
Transarterial embolization has long been an established approach for neuroendocrine tumor (NET) liver metastases. Madoff et al. and Vogl et al. demonstrated that both TACE and bland embolization provide symptomatic improvement in 60-95% of patients and objective tumor responses in 33-80%, with 5-year survival rates of 50-65% (5, 6). The NET-Liver-Metastases Consensus Conference subsequently recognized TAE, TACE, and radioembolization as standard locoregional options for unresectable NET liver disease (15). Nevertheless, these conventional approaches are often limited by cumulative hepatobiliary toxicity from chemotherapeutic agents and lipiodol, particularly in elderly or heavily pretreated patients (5, 16).
The therapeutic rationale of balloon-occluded techniques lies in transient flow control, enabling deeper penetration of embolic material and preferential delivery to tumor sinusoids while minimizing reflux to normal parenchyma. Previous studies of BOAI-TACE for hepatocellular carcinoma reported improved drug retention, higher local control, and reduced post-embolization dysfunction compared with conventional TACE (9, 10). In this pilot cohort of hypervascular hepatic metastases, the present results support the feasibility of chemotherapy-free embolization using mechanically fragmented gelatin particles under balloon occlusion. The potential hemodynamic advantages of balloon occlusion remain to be confirmed in larger studies.
Growing evidence supports the clinical value of chemotherapy-free embolization. Shimohira et al. conducted a multicenter prospective trial using microspheres for hypervascular metastases refractory to systemic therapy, achieving an ORR of 52% and DCR of 72% without grade ≥3 toxicity (17). Monfardini et al. histologically confirmed complete necrosis of NET metastases after bland embolization alone (18). Furthermore, the RETNET randomized trial recently reported bland TAE as non-inferior to drug-eluting bead (DEB)-TACE in progression-free survival, suggesting that arterial flow modulation is an important component of tumor control (7). The present findings are compatible with the concept that embolization-based flow modulation may contribute to tumor control, although mechanistic conclusions cannot be drawn from this study.
The ORR of 66.7% in this study is comparable to or slightly higher than previous reports of TAE or TACE for NET or GIST hepatic metastases (40-60%) (5, 6, 19). The 2-month assessment was prespecified to capture early radiologic response and short-term safety; durability and long-term outcomes require longer follow-up. Importantly, no deterioration of hepatic reserve was detected by Child-Pugh or ALBI scores whereas conventional TACE series have reported ALBI worsening in 20-40% of cases (6, 16, 19). This functional stability may reflect the combination of selective balloon-occluded infusion, the transient nature of gelatin embolization, and careful patient selection excluding those with poor baseline reserve.
From a safety standpoint, transient AST and ALT elevations were nearly universal but resolved within one month, and no significant renal or coagulation abnormalities occurred. Although the overall frequency of adverse events may appear high in this small cohort, most events reflected expected post-embolization syndrome and transient laboratory abnormalities. Joskin et al. (2015) identified portal vein thrombosis and large (>300 μm) particles as risk factors for post-TACE hepatic necrosis in NET metastases (20). In contrast, the use of fine particles under balloon occlusion in the present study resulted in only transient biochemical changes and no biliary complications or liver failure, even in elderly participants (median 77 years). These findings support the short-term safety of selective, temporary embolization.
The precise and liver-sparing nature of BOIG-TAE also address the long-term treatment issues highlighted by Vogl et al., who reported that therapeutic efficacy of conventional TACE declines after the sixth session and that intervals greater than 90 days are associated with worse survival (16). Because BOIG-TAE uses temporary gelatin particles and selective balloon delivery, it may be suitable for repeated sessions while preserving hepatic function across multiple procedures. Furthermore, Hur et al. (2013) identified tumor burden and number of lesions as key prognostic determinants after TACE (19), indicating that embolization mainly serves a cytoreductive and palliative role in disease control. In this context, BOIG-TAE may function as a bridge or salvage therapy for patients who are refractory or intolerant to systemic treatment, providing symptom relief and temporary control without hepatic deterioration.
Limitations. It was a small, single-center pilot trial with a prespecified, short-term (2-month) imaging follow-up that precluded assessment of response durability, time to progression, repeatability, and survival. Enrollment fell short of the prespecified target (10 planned vs. 6 enrolled), resulting in wide confidence intervals and limiting the robustness and precision of safety and efficacy estimates; therefore, comparisons with prior studies should be interpreted cautiously. The cohort included heterogeneous primary tumor types while response assessment using mRECIST, albeit as prespecified and aligned with the perfusion-targeting mechanism of embolization, has limited validation in non-HCC metastases; thus, tumor-type–specific conclusions and long-term clinical benefit cannot be established. Larger prospective studies with longer follow-up and comparative designs are needed to define the role of BOIG-TAE among liver-directed therapies. Additionally, all procedures were performed at a single institution by experienced interventional radiologists, which may limit generalizability. Future multicenter studies with longer follow-up and histology-specific subgroup analyses are warranted to validate these findings and to define optimal parameters such as particle size, occlusion duration, and treatment interval.
Conclusion
BOIG-TAE showed favorable short-term safety and suggested preliminary local efficacy for hypervascular hepatic metastases refractory or intolerant to systemic therapy. By combining selective balloon occlusion and temporary gelatin embolization, this method may provide short-term tumor control while preserving hepatic function. BOIG-TAE may represent a feasible and liver-sparing alternative among transarterial therapies, warranting further evaluation in larger, controlled trials.
Acknowledgements
The Authors would like to thank Atsuko Sato and Miki Suzuki at Tsukuba Clinical Research and Development Organization (T-CReDO) for their support of EDC set up and monitoring. We thank all the practice staff who supported this clinical trial and the patients who participated.
Footnotes
Authors’ Contributions
Study conception and data analysis: S.H., T.Y., N.T, and T.I;. Acquisition of data: S.H., T.Y., M.K., T.A., N.T.,, and T.I; Interpretation of data: S.H., T.Y., M.K., T.A., N.T., B.M., T.S., T.N., and T.I; Drafting manuscript: all authors
Conflicts of Interest
Toshiyuki Irie receives royalties from Piolax Medical Devices, Inc. The remaining Authors have no conflicts of interest to disclose in relation to this study.
Funding
This work was supported by JSPS KAKENHI Grant Number JP22K07684, JP25K10930 and Grant for Implementation of Advanced Medicine (GIAM) of the University of Tsukuba Hospital.
Artificial Intelligence (AI) Disclosure
No artificial intelligence (AI) tools, including large language models or machine learning software, were used in the preparation, analysis, or presentation of this manuscript.
- Received January 16, 2026.
- Revision received February 9, 2026.
- Accepted February 11, 2026.
- Copyright © 2026 The Author(s). Published by the International Institute of Anticancer Research.
This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.