The study, led by Claudio Sallemi of the Interventional Radiology Unit at Fondazione Poliambulanza Istituto Ospedaliero, describes four consecutive patients with ruptured hepatocellular carcinoma treated with an unusual embolic material: a low-viscosity ethylene vinyl alcohol copolymer marketed as Squid Peri12. These agents, known as non-adhesive liquid embolic agents, consist of the EVOH copolymer dissolved in dimethyl sulfoxide, a solvent that diffuses rapidly into the bloodstream once the mixture is injected. As the solvent leaves, the copolymer precipitates in a slow, controlled fashion, solidifying into a soft, cohesive cast that conforms precisely to the twisted architecture of tumor vessels. The behavior is often described by operators as a lava effect, a gradual filling rather than the near-instantaneous gluing seen with older cyanoacrylate adhesives. Squid Peri12 is notable as the only EVOH-based agent with a kinematic viscosity of just 12 centistokes, a watery consistency that allows it to travel deep into fragile, tortuous tumor feeders that thicker agents might never reach.

All four patients arrived with the classic signature of rupture: acute anemia, hemodynamic decompensation, and free blood in the abdomen visible on non-contrast computed tomography. None was a candidate for emergency surgery, whether because of the sheer extent of the tumors, serious comorbidities, or outright instability. Conservative management with fluids and transfusions had failed to stabilize them. The tumors themselves were formidable, with a median diameter of 8.2 centimeters and a range stretching to 12.7 centimeters, all intensely hypervascular and all meeting the strictest LI-RADS 5 imaging criteria for hepatocellular carcinoma. One patient carried two bleeding lesions, one in each lobe. Two showed ominous local extension, one with suspected invasion of the adrenal gland and inferior vena cava plus segmental portal vein thrombosis, the other with a bulky mass displacing the gastric wall. In two patients alpha-fetoprotein, the blood marker that often tracks liver cancer activity, was elevated to as much as 3500 nanograms per milliliter.

The technical execution reveals how much finesse modern embolization demands. Procedures were performed under local anesthesia with mild sedation, using a slender 5 French sheath inserted through the radial or femoral artery. After mapping the hepatic arterial tree with a 4 French catheter, the operators advanced a DMSO-compatible microcatheter barely wider than a human hair into the subsegmental or even intralesional vessels feeding the tumor. Squid Peri12 was then injected at a deliberately slow rate of roughly 0.2 milliliters per minute under continuous fluoroscopy, allowing the team to watch the polymer cast grow in real time and stop the instant it threatened to escape into healthy territory. When reflux toward a larger artery appeared, the operators paused briefly to let the material partially polymerize, then resumed injection to push the cast deeper, a maneuver they call cast and push, adapted from the neurointerventional suites where EVOH agents have long been used for brain vascular malformations.

The results were striking on every measure the team recorded. Technical success was achieved in all four cases, with immediate hemostasis confirmed on post-procedural angiography and total embolic volumes of only 1 to 2.5 milliliters per patient. Fluoroscopy times averaged just 12 minutes despite the vascular complexity of the lesions. Crucially, no procedure-related adverse events occurred at all: no non-target embolization, no catheter entrapment, no post-embolization syndrome, no rebleeding, and no portal vein thrombosis. Liver function remained stable in three patients, while the fourth showed a transient rise in the liver enzyme AST from 63 to 358 units per liter, formally graded as a CTCAE version 5.0 grade 3 event but without any clinical deterioration or symptoms. Every patient was discharged home within three to five days, an outcome that would have been unthinkable a generation ago for patients bleeding to death from a burst liver tumor.

What followed over the next six months is what elevates this report from a hemostasis story to an oncology story. Follow-up CT scans at one, three, and six months, analyzed with modified RECIST criteria and semi-automatic volumetric software, showed sustained devascularization and consistent tumor shrinkage in every patient, with an average volume reduction approaching 60 percent and no additional transarterial or systemic therapy of any kind. Two patients, including the one whose tumor had appeared to invade the adrenal gland and vena cava, became eligible for curative-intent surgery after the embolization alone. In the operating room the suspected adrenal invasion was confirmed, an adrenalectomy was performed, and the resection was completed with preservation of the kidney. A third patient, initially inoperable because of gastric wall displacement and an alpha-fetoprotein level of 3500 nanograms per milliliter, experienced such marked regression that her marker fell to 27 nanograms per milliliter, rendering her a surgical candidate, though she ultimately declined the operation. The fourth patient achieved a complete radiologic response and remained disease-free without any further treatment.

The mechanistic explanation for this unexpected anti-tumor effect likely lies in the depth and permanence of the embolic cast. Particulate agents, the workhorses of conventional embolization, occlude vessels at a certain caliber but leave distal vascular networks patent, allowing tumors to reconstitute their blood supply over time. Cyanoacrylate glues polymerize on contact with blood, achieving rapid occlusion but demanding considerable operator skill and carrying a real risk of gluing the catheter in place or showering adjacent vessels with unintended emboli. The gradual precipitation of EVOH, by contrast, permits progressive, controlled filling of the entire intratumoral vascular bed, from feeding artery down to the smallest sinusoidal channels. A tumor stripped of its vasculature at that level cannot easily recover, and the ischemic injury may extend well beyond what a temporary particulate occlusion inflicts. The authors suggest this could explain both the sustained devascularization and the volume reductions observed without any adjunctive chemotherapy or immunotherapy.

The implications reach beyond the emergency setting. Hepatocellular carcinoma is typically diagnosed in patients with chronic liver disease, where every milliliter of functioning hepatic reserve matters and where treatment options for large or locally advanced tumors are painfully limited. If a single embolic session can downstage locally advanced disease to resectability, as it apparently did in two of these four patients, the low-viscosity EVOH approach could earn a place in planned, non-emergent treatment algorithms, potentially in combination with or as an alternative to conventional chemoembolization and radioembolization. The authors are careful to frame this as a hypothesis rather than a conclusion, noting that prior multicenter work has already established the safety of EVOH agents in acute non-neurovascular abdominal bleeding, but that their role in ruptured hepatocellular carcinoma had never been explored until now.

Honest caveats temper the enthusiasm. This was a retrospective case series of four patients, the weakest tier of clinical evidence, formally classified as Level 4. The absence of a comparison group means the tumor shrinkage cannot be definitively attributed to the embolic agent rather than to the general effect of devascularization, and rare complications may simply not have surfaced in so small a cohort. The team itself, which also included Maria Parmigiani, Marco Garatti, and Massimo Graffeo, calls explicitly for larger prospective studies to confirm the findings and define the role of EVOH agents beyond emergent indications. Still, the combination of flawless technical outcomes, preserved portal vein patency, stable liver function, and two unexpected conversions to curative surgery is difficult to dismiss as coincidence.

For a complication that has long carried a grim prognosis, the message of this small study is quietly revolutionary: the material used to plug a bleeding liver tumor in the middle of the night may also be a weapon against the tumor itself. If prospective trials bear out what these four patients experienced, the emergency catheter lab could become an unexpected starting point on the road to cure for some of the sickest liver cancer patients in the hospital.

Subject of Research: Emergency transarterial embolization of ruptured hepatocellular carcinoma using a low-viscosity EVOH copolymer and its potential tumor-shrinking effects

Article Title: Transarterial embolization with a low-viscosity ethylene vinyl alcohol (EVOH) copolymer for ruptured hepatocellular carcinoma: an oncologic role beyond hemostasis

Article References: Sallemi, C., Parmigiani, M., Garatti, M., & Graffeo, M. (2025). Transarterial embolization with a low-viscosity ethylene vinyl alcohol (EVOH) copolymer for ruptured hepatocellular carcinoma: an oncologic role beyond hemostasis. CVIR Oncology, 1(1), Article 24. https://doi.org/10.1007/s44343-025-00023-1

Image Credits: AI Generated

DOI: 10.1007/s44343-025-00023-1

Keywords: hepatocellular carcinoma, ruptured HCC, transarterial embolization, EVOH copolymer, Squid Peri12, liquid embolic agent, interventional radiology, tumor downstaging, liver cancer, hemostasis, alpha-fetoprotein, DMSO-compatible microcatheter

The study, published in CVIR Oncology, draws on the CIRT registry spanning 27 sites across eight European countries and its French counterpart, CIRT-FR, which was conducted to help the French Health Authority evaluate reimbursement of yttrium-90 resin microspheres. Both registries were designed to capture real-world clinical practice rather than the tightly scripted conditions of a randomized trial, and that realism is precisely what makes the findings interesting. Physicians recorded their planned treatment intent at the start of therapy, but the researchers later reclassified each patient’s actual trajectory using predefined criteria covering transplantation, ablation, resection, or a sustained complete response lasting at least one year as measured by modified RECIST criteria.

The headline numbers tell a story about the gap between intention and outcome. Although 229 of the 615 patients, roughly 37 percent, were initially slated for curative-intent treatment pathways, only 40 patients, or 6.5 percent, ultimately met the retrospective curative-intent definition, while 93 percent were classified as palliative. Of those 40, half went on to liver transplantation after TARE, seven received ablation and two underwent resection. Most remarkably, 11 patients, 27.5 percent of the curative group, required no further treatment at all because TARE alone had produced a complete tumor response documented for at least a year.

That last figure carries particular weight given how the authors framed their expectations. Among 32 patients whose a priori plan was radiation segmentectomy, meaning planned curative yttrium-90 monotherapy, only one ultimately achieved a curative classification through subsequent transplantation. In other words, the patients who ended up cured by radiation alone were rarely the ones selected for that purpose in advance. Their complete responses were, in a sense, discovered rather than engineered, which the authors say underscores how little consensus currently exists on who will convert to a curative pathway after radioembolization and how urgently refined conversion criteria are needed.

The statistical analysis offers some clues. In multivariable logistic regression, three factors independently predicted placement in the curative-intent group: an ECOG performance status of zero, corresponding to fully active patients, carried an odds ratio of 0.44; having one to three tumors yielded an odds ratio of 0.39; and a total tumor volume below 150 cubic centimeters produced an odds ratio of 0.40. That 150 cubic centimeter threshold emerged from receiver operating characteristic analysis as one of two optimal cut points for separating curative from palliative patients, and it was selected for its higher specificity of 0.725. Notably, 72.5 percent of curatively treated patients fell below this volume, which is substantially larger than the roughly 65 cubic centimeter ceiling imposed by the classic Milan criteria for transplantation eligibility.

This discrepancy is central to the paper’s argument. The Milan criteria, which have governed transplant selection for hepatocellular carcinoma for nearly three decades, rest almost entirely on morphometric measurements, and the authors found no significant differences between their curative and palliative groups on Milan-based eligibility before treatment. Around 80 percent of the curatively managed patients were outside Milan criteria before TARE. The implication is that radioembolization can shrink tumors enough to pull patients back within transplant boundaries, and that tumor volume thresholds well beyond Milan’s limits, combined with biological markers such as alpha-fetoprotein and the ALBI grade of liver function, may better identify who stands to benefit. The authors point to emerging tools like the Metroticket 2.0 model as the kind of finer-grained metrics the field needs.

Dosimetry adds another technical layer to the story. Because absorbed radiation doses were not prospectively collected in the original registries, the team recalculated them retrospectively using the simplified Medical Internal Radiation Dose formula, dividing administered activity by tumor mass. Their surrogate analysis found that absorbed doses above 400 grays correlated with longer survival, a threshold repeatedly associated with complete response in prior literature. Crucially, 92.3 percent of patients who received doses above that level had tumors under 150 cubic centimeters, suggesting that the concentration of radiation achievable in smaller, more selectively targeted volumes is what drives these dramatic responses. Yet the data also showed that doses exceeding 400 grays remained effective in tumors up to about 150 cubic centimeters, roughly 6.6 centimeters in diameter, larger than many earlier studies had considered treatable at such intensities.

Survival outcomes reinforce the signal. Median overall survival was not reached in the complete response, resection, and ablation subgroups by the end of follow-up, and in the transplantation subgroup, 75 percent and 85 percent survival percentiles were reached at 37.1 and 11.4 months respectively, with only one death recorded in the resection and ablation group and none in the complete response group. By contrast, the palliative group’s median survival of 16.1 months compares favorably with the 8.0 and 8.8 months reported in the landmark SARAH and SIRveNIB randomized trials of radioembolization versus sorafenib in advanced disease, and aligns with more recent prospective figures of 13.9 months. The real-world palliative performance alone is notable, but the theoretical cure fractions achieved in a minority of patients are what will spark debate.

The authors are careful about the limits of their evidence. This was an exploratory, retrospective classification layered onto prospective observational data, with no harmonized definition of curative intent when the studies began. Baseline characteristics were collected only before TARE, alpha-fetoprotein levels were missing in more than a fifth of curative patients and more than 40 percent of palliative patients, and dosimetry relied on a surrogate calculation. The heterogeneity of post-TARE curative procedures across different national practices further complicates generalization. Still, the conclusion is clear and consequential: yttrium-90 radioembolization belongs not just in the palliative toolbox but across the full spectrum of curative-intent strategies for hepatocellular carcinoma, whether as a bridge to transplantation, a downstaging tool before resection or ablation, or, in carefully selected patients with small tumor burdens and good performance status, as a standalone therapy capable of eradicating the disease outright. The task now facing the field, the authors argue, is to marry morphometry, tumor biology, and personalized dosimetry into reliable criteria that can identify, before the first microsphere is injected, which patients will be among the lucky few whose cancer never returns.

Subject of Research: Curative-intent transarterial radioembolization with yttrium-90 in hepatocellular carcinoma

Article Title: Hepatocellular carcinoma patients treated with TARE monotherapy or combination therapy with curative intent: a retrospective, exploratory analysis of two European prospective studies

Article References: Helmberger, T., Arnold, D., Loffroy, R., Ronot, M., Sangro, B., Kolligs, F., Pellerin, O., Maleux, G., Peynircioglu, B., Schaefer, N., Bilbao, J. I., de Jong, N., Geyer, M., Zeka, B., Urdaniz, M., & Vilgrain, V. (2026). Hepatocellular carcinoma patients treated with TARE monotherapy or combination therapy with curative intent: a retrospective, exploratory analysis of two European prospective studies. CVIR Oncology, 2(1), Article 5. https://doi.org/10.1007/s44343-026-00035-5

Image Credits: AI Generated

DOI: 10.1007/s44343-026-00035-5

Keywords: hepatocellular carcinoma, TARE, yttrium-90, radioembolization, liver transplantation, radiation segmentectomy, Milan criteria, dosimetry, tumor downstaging, locoregional therapy, complete response, CIRT registry