Transvenous curative embolization of brain arteriovenous malformations using extra-low viscosity _ihtObtura versions: Initial experience and technical insights

Abstract

Background and purpose

This study presents our initial experience using _ihtObtura, a novel nonadhesive liquid embolic agent with extra-low viscosity variants, in transvenous curative embolization of brain arteriovenous malformations (bAVMs). We assess the agent's performance and compare its advantages with other extra-low viscosity options currently available.

Materials and methods

Five patients (three females, two males; mean age, 33 years; range, 20–55 years) with ruptured bAVMs were treated using the transvenous retrograde pressure cooker technique (TVRPCT). Three patients underwent staged embolization for large bAVMs, with the final session performed using TVRPCT, while two patients with smaller bAVMs were treated in a single session.

Results

Complete obliteration was achieved in 100% of patients. The mean volume of _ihtObtura used per transvenous session was 7.9 mL (range, 2–14 mL) ± 4.8. A combination of _ihtObtura 15 and _ihtObtura 12 was used in three cases, while _ihtObtura 15 alone was used in two cases. Complete angiographic occlusion was sustained at six and 24 months follow-up. Clinical outcomes improved in three patients and remained unchanged in two (mRS = 0).

Conclusions

The TVRPCT with extra-low viscosity _ihtObtura appears to be as safe and effective as other extra-low viscosity nonadhesive liquid embolic agents for curative treatment of bAVMs. The penetration, diffusion, and distribution of _ihtObtura through the transvenous route were optimal, with no observed vascular rupture or extravasation.

Introduction

Advances in embolization techniques and the development of specialized materials have improved curative outcomes and reduced morbidity and mortality in the treatment of complex brain arteriovenous malformations (bAVMs).^1,2 The transarterial pressure cooker technique, for example, has enhanced curative embolization rates. However, in certain cases, such as when arterial feeders are too small or tortuous for antegrade access, a retrograde transvenous approach may be preferred. ²

Currently, transvenous embolization is a well-established approach for bAVMs, showing high rates of total occlusion and an acceptable morbidity profile.^1–7 This approach is particularly valuable for ruptured bAVMs, where rerupture risk remains significant (6%–18% annually), contributing to increased disability and mortality rates. ⁴ Transvenous embolization may involves retrograde injection of an extra-low viscosity, nonadhesive liquid embolic agent via the venous route—a strategy favored in a substantial proportion of cases to facilitate complete AVM occlusion or selective targeting in sequential venous approaches. Key anatomical indications for this method include deep AVM location, challenging arterial access, small nidus size, or single draining vein. It is especially indicated when safer alternative treatments are unavailable.^1,2,6

Extra-low viscosity agents such as Squid 12 and PHIL LV exhibit unique embolic properties. Their lower viscosity enables faster, more extensive antegrade flow during embolization, which facilitates deeper vascular penetration and minimizes the risk of rupture or extravasation of small nidus vessels.^2,8–10 These properties enhance the likelihood of achieving complete bAVM occlusion through transarterial and transvenous retrograde pressure cooker techniques.

_ihtObtura is a novel nonadhesive liquid embolic agent, composed of EVOH copolymer and an iodinated compound dissolved in DMSO.^11,12 It is available in four formulations—_ihtObtura 12, 15, 20, and 35—designed for varying levels of viscosity. Notably, _ihtObtura loses 90% of its radiopacity within six weeks postembolization, enabling clearer visualization of AVM angioarchitecture during sequential sessions.^11–14 Unlike other agents such as PHIL, Onyx, and Squid, _ihtObtura minimizes artifact formation on CT and causes no artifacts in MRI, facilitating improved postprocedure imaging.^11–14 The extra-low viscosity _ihtObtura 12 and 15 variants allow for enhanced penetration from both arterial and venous routes, mitigating the risk of vascular rupture in small vessels.^11–12

This study aims to report our initial experience using _ihtObtura for transvenous embolization in curative treatment of bAVMs, emphasizing its effectiveness and unique advantages over other extra-low viscosity agents available.

Materials and methods

Five patients with ruptured bAVMs were treated with _ihtObtura (Iberhospitex, Lliçà de Vall, Spain) with curative intent at The National Institute of Neurology and Neurosurgery (Havana, Cuba) between 10 November 2021 and 10 September 2022, using the transvenous retrograde pressure cooker technique (TVRPCT). All patients were assessed by an interdisciplinary team of neuroradiologists, neurologists, and neurosurgeons, who evaluated the appropriateness of endovascular treatment for each case. Written informed consent was obtained from the patients or their families prior to each procedure.

Patient demographic and clinical data were recorded, including age, sex, presenting symptoms, history of prior hemorrhage, and previous treatments. Angiographic characteristics such as bAVM volume, maximal diameter, location, venous drainage type, and eloquent cortex involvement were documented. For each embolization session, the number of arterial pedicles embolized, the volume of agent used, overall reduction in AVM volume, and any procedural or postembolization complications were noted.

Angiography and embolization technique

All procedures were performed under general anesthesia. Pre- and postembolization brain CT scans were acquired within the angiography suite. Embolizations were conducted under roadmap guidance using the “Glue” mode available in the roadmapping options of the Allura system, with intermittent resetting of the blank roadmap to accurately assess the diffusion pattern of the LEA. A single-plane angiography unit (Allura FD20, Philips Healthcare, the Netherlands) was utilized for all procedures. Systemic heparin was not administered.

A 7F guiding catheter was placed in the arterial access, followed by selective microcatheterization of each bAVM feeder using a Magic 1.2F microcatheter (Balt Extrusion, Montmorency, France) to assess arterial supply and venous drainage patterns. For the transvenous approach, a 7-8F guiding catheter (Fubuki, Asahi Intecc, Amsterdam, Netherlands; Envoy, CodmanNeuro, Colorado Springs, United States; Ballast, Balt Extrusion, Montmorency, France) was navigated from the jugular vein to the straight sinus or superior sagittal sinus.

Retrograde catheterization of the main draining vein was achieved using a Sonic 1.2F35 detachable tip microcatheter (Balt Extrusion, Montmorency, France) for _ihtObtura injection. A second microcatheter, Apollo 1.5F3–5 cm detachable tip (Medtronic, Dublin, Ireland), was positioned alongside the primary catheter per the “sheeping” technique, creating a plug using detachable coils (Barricade Coils, Blockade Medical, Balt USA, Irvine, CA) and glue (Histoacryl, Melsungen, Germany) with a 50% Glue–50% Lipiodol (Guerbet, Paris, France) dilution.^1,2

_ihtObtura 15 or _ihtObtura 12 was injected through the Sonic 1.2F microcatheter until the entire bAVM was filled. The bAVM was deemed occluded when the final cast of the embolic agent matched the preinterventional angiographic appearance in the working projection. Selective microcatheterization of all feeders was repeated postembolization to check for residual flow to the AVM, with transarterial embolization performed as needed.

During TVRPCT, arterial pressure was consistently maintained at a mean of 60–65 mmHg. Following each procedure, patients were transferred to the intensive care unit, where blood pressure was monitored and maintained between 65 and 7 mmHg for 2 h. Postprocedural CT was performed immediately after embolization, and patients received intravenous corticosteroids before and after the procedure. All patients were discharged home with follow-up care.

Follow-Up

Clinical follow-up was based on modified Rankin scale (mRS) scores. Neurological evaluations were conducted at admission, upon discharge, and at 1, 3, 6, 12, and 24 months postembolization. Complications were classified into five categories: technical complications without neurological deficit, transient neurological deficits, nondisabling permanent neurological deficits, disabling permanent neurological deficits, and death.

Complete bAVM occlusion was assessed by digital subtraction angiography (DSA) at six and 24 months. Angiographic cure was defined as complete obliteration of the AVM nidus and arteriovenous shunting, with no residual filling on final DSA and no evidence of recurrence on follow-up DSA performed six months after the last embolization session.

Results

Patient and AVM characteristics

Five patients (three females, two males; mean age 33 years, range 20–55 years) with ruptured bAVMs were treated with _ihtObtura using the transvenous retrograde pressure cooker technique. Three patients underwent TVRPCT as the final session of staged embolization for large bAVMs, while two patients with smaller bAVMs received a single-session treatment. All patients presented with intracerebral hemorrhage.

Based on the Spetzler-Martin grading scale, three bAVMs (60%) were high-grade, and two (40%) were low-grade. AVMs were supratentorial in all cases, with three located in the dominant hemisphere. Venous drainage involved both cortical and deep veins in three cases and was exclusively superficial in two. Patients had an average of two draining veins (range, 1–3). Two patients had three draining veins, two patients had two draining veins, and one patient had one draining vein. Two patients presented with an intranidal aneurysm.

All AVMs were supplied by feeders from the anterior circulation, with one bAVM also receiving supply from the middle cerebral artery lenticulostriate branch and posterior circulation feeders.

The maximum AVM diameter before transvenous embolization was <3.2 cm. The mean initial bAVM size was 33.2 mm (range, 15–48 mm) ± 13.7, and pretransvenous size was 26.2 mm (range, 15–32 mm) ± 7.3. The mean initial bAVM volume was 20.5 mL (range, 1.4–27.6 mL) ± 10.0, and pretransvenous volume was 7.4 mL (range, 1.4–8.2 mL) ± 3.2.

Embolization procedure details

All patients were treated transvenously with _ihtObtura in the final session. Two patients underwent combined arterial and venous pressure cooker embolization; the remaining three underwent exclusively transvenous embolization. Cortical venous access was obtained via the superior sagittal sinus in three cases, the straight sinus in one case, and the vein of Labbé in one case.

All patients received at least one injection of _ihtObtura: two patients had four sessions, one had three sessions, and two had single-session treatments. A mean of 7.9 mL (range, 2–14 mL) ± 4.8 of _ihtObtura was used per session, with a combination of _ihtObtura 15 and 12 in three cases and _ihtObtura 15 alone in two. Optimal diffusion and penetration were observed in all cases, with no vascular rupture (Figures 1 and 2).

Figure 1.

(A) Lateral view of ruptured frontal bAVM. (B) _ihtObtura cast post third embolization session. (C) Carotid angiogram pre-fourth and final embolization; AVM nidus (white arrowheads). (D) Carotid angiogram postfinal embolization session. (E) _iht Obtura cast posttransvenous embolization (white arrowheads). (F) Carotid angiogram at 24-month follow-up showing complete bAVM occlusion and radiopacity loss. (G, H, I) T1, T2, and SWI MRI at 24-month follow-up with absence of artifacts and clear delineation of surrounding brain parenchyma.

Note: bAVM: brain arteriovenous malformation.

Figure 2.

(A) Lateral view of ruptured Sylvian fissure bAVM. (B) Carotid angiogram prior to the third and final embolization session. (C) Last embolization session; note the absence of the previous _ihtObtura cast, with detachable tip microcatheter and glue from the prior pressure cooker technique (black arrows); coils from the TVRPC (white arrow); Rebar 18 microcatheter used to deploy the Solitaire stent for MCA branch protection (black arrowhead). (D, E) Frontal and lateral views of carotid angiogram postfinal embolization session. (F) _ihtObtura cast posttransvenous embolization (white arrowheads); glue used in TVRPC (asterisk). (G, H) Complete bAVM occlusion in frontal and lateral views of carotid angiogram at 23-month follow-up. (I) Radiopacity loss at 23-month follow-up (white arrowheads), with coils from TVRPC (white arrow) and glue used in TVRPC (asterisk).

Note: bAVM: brain arteriovenous malformation; TVRPC: transvenous retrograde pressure cooker; MCA: middle cerebral artery.

Angiographic and clinical outcomes

Complete angiographic occlusion was achieved in all patients (100%), with mRS scores indicating favorable outcomes at discharge and follow-up.

The mRS data for preembolization, discharge, and follow-up are summarized in Table 1. Prior to embolization, five patients (100%) had mRS scores between 0 and 2. At discharge, all five patients (100%) remained in the 0–2 range. According to the mRS, two patients experienced nondisabling neurological deficits (mRS 1–2). At the six-month follow-up, all five patients (100%) achieved mRS scores between 0 and 1, with 3 patients scoring mRS = 1 and two patients scoring mRS = 0.

Table 1.

Demographic, clinical, and follow-up details of the patients treated with transvenous embolization approach.

Case No.	Age (years)/sex	Spetzler Martin grade	Location/side/elocuent area	Arterial feeders	Venous drainage/number of veins	NA	Volume (mL)		Nidus size (mm)		Embolization session		Final session technique /occluded	Total ihtObtura mL	mRS			Follow-up angiogram results
							Initial	BTVE	Initial	BTVE	Trans arterial	Trans Venous			Pre	post	6MFU
1	20/F	IV	Parietal/left/Y	ACA, MCA, PCA, MMA,	SSS, RS/2	Y	27.6	8.2	48	32	3	1	TVRPCT + TAPCT	14	1	2	1	7- and 24-month occluded
2	39/F	II	Frontal/right/N	ACA, MCA	SSS,CS/3	N	13.4	7.4	42	32	3	1	TVRPCT + TAPCT	11	0	0	0	9-and 25-month occluded
3	55/M	IV	Sylvian fissure/left/Y	MCA, MMA, PCA	SSS, RS, TS/ 3	N	14.1	7.5	38	29	2	1	TVRPCT	8	0	1	1	8- and 23-month occluded
4	30/F	II	Frontal/left/Y	ACA, MCA	SSS/1	Y	1.4	1.4	15	15	-	1	TVRPCT	2	1	1	0	6- and 34-month occluded
5	21/M	III	Parietal/right/Y	MCA	SSS, RS /2	N	2.5	2.5	23	23	-	1	TVRPCT	4.6	2	2	1	6- and 24-month occluded
Mean range ± ED	33 (20–55) ±14.5	–	–	–	2 (1–3) ±1	–	20.5 (1.4 – 27.6) ±10.0	7.4 (1.4 – 8.2) ±3.2	33.2 (15–48) ±13.7	26.2 (15 –32) ±7.3	2.7 (2–3) ±0.6	1	–	7.9 (2–14) ±4.8	0.8	1.2	0.6	7.2- and 26-month occluded

Note: F: Female; M: Male; ACA: anterior cerebral artery; MCA: middle cerebral artery; PCA: posterior cerebral artery; MMA: middle meningeal artery; SSS: superior sagittal sinus; RS: rectus sinus; CS: cavernous sinus; TS: transverse sinus; TVRPCT: transvenous retrograde pressure cooker technique; TAPCT: transarterial pressure cooker technique; Y: yes; N: no; NA: nidal aneurysm; BTVE: before transvenous embolization.

The rate of complete angiographic occlusion was 100% at both the six- and 24-month marks. Clinically, four patients showed improvement following embolization, while the condition of one patient remained unchanged (mRS = 0). At the 24-month follow-up, all five patients (100%) had mRS scores between 0 and 1.

Radiopacity loss and safety profile

No systemic or local side effects related to _ihtObtura injection were observed. It is notable that the radiographic disappearance of _ihtObtura on fluoroscopy is delayed when bAVM occlusion is complete, either via the arterial or venous route, though more prominently through the latter one. This delay is associated with the mechanism of radiopaque agent clearance. At the six-month follow-up, 80–90% radiopacity loss was observed even in markedly dilated venous structures of fully occluded bAVMs. Importantly, this phenomenon was not associated with angiographic recanalization.

Discussion

Transvenous embolization has emerged as a safe and effective modality for treating bAVMs, either as a standalone approach or as part of a multimodal treatment strategy. This technique facilitates immediate nidus obliteration by targeting both the nidus and adjacent draining veins, minimizes arterial reflux, and reduces ischemic complications. Reported obliteration rates are high, and functional outcomes remain favorable, with low procedure-related neurological morbidity. ¹⁵

In the last decade, advancements in endovascular approaches have introduced novel nonadhesive embolic agents that enhance procedural control and increase rates of complete nidus eradication.^16–18 Slow-polymerizing nonadhesive agents, such as Onyx, have notably improved transvenous embolization by enabling a controlled injection process. ¹⁵ The development of extra-low viscosity agents, such as Squid 12, has further advanced the field. Squid 12's lower viscosity allows faster and more extensive antegrade flow of the liquid embolic agent (LEA) during embolization, facilitating effective and rapid penetration into the target vasculature. ⁸

Studies on Squid 12 usage for dural arteriovenous fistulas (DAVFs) demonstrated high rates of complete angiographic occlusion (89.5%) without major periprocedural adverse events.^8,9 The lower viscosity of Squid 12 supports deeper penetration of LEAs into the DAVF vasculature.

PHIL LV is the extra-low viscosity version of PHIL. The difference between PHIL LV and PHIL 25% is the length of the polymer chains which are shorter for PHIL LV, resulting in a lower molecular weight of the single copolymer molecules, causing the lower viscosity.^8,10

Comparative experimental studies indicate that extra-low viscosity LEAs, including Squid 12 and PHIL LV, are more effective than their standard viscosity counterparts in achieving extensive embolization with minimal reflux events. These LEAs show promise for increasing total occlusion rates in AVMs. ⁸

_ihtObtura, a novel nonadhesive embolic agent composed of EVOH copolymer and an iodinated compound in DMSO, provides additional advantages for AVM embolization.^11–14 Developed in extra-low viscosity formulations (_ihtObtura 12 and 15), this agent allows deeper LEA penetration into vascular malformations from both arterial and venous approaches, reducing the risk of rupture in small nidus vessels. Our initial experience suggests that _ihtObtura offers superior penetration and diffusion capacity compared to established agents, particularly in transvenous embolization, with reduced risk of vascular rupture. ¹²

Moreover, no complementary transarterial injections were required to embolize residual stagnating portions of the nidus, as none were observed in the final supraselective angiograms following venous occlusion and transvenous injection. These findings are consistent with those observed in our initial experience using _ihtObtura in the treatment of AVMs. ¹²

However, similar to Squid 12, _ihtObtura's lower viscosity may delay plug formation, requiring prolonged waiting times to ensure sufficient proximal reflux and penetration, thereby making the retrograde pressure cooker technique (RPCT) essential to achieve complete retrograde occlusion without significant venous reflux.

Compared with tantalum-based EVOH agents (Onyx, Squid) and iodine-bound copolymers such as PHIL, _ihtObtura exhibits several procedural and imaging advantages.^11,12 Its iodine-based radiopacity provides reliable visualization of vascular structures during embolization without compromising precision, while its homogeneous formulation eliminates the need for preprocedural shaking. Unlike Onyx and Squid, which contain suspended tantalum particles prone to precipitation and microcatheter occlusion during prolonged injections, _ihtObtura enables sustained and uniform delivery without cast heterogeneity or flow interruption. Moreover, its peripheral precipitation mechanism leads to the formation of central low-density channels—known as the “tunnel effect”—which enhances the transparency of the cast and allows improved visualization of overlapping vascular structures and more accurate delineation of the malformation nidus during agent delivery. Additionally, the absence of tantalum avoids skin discoloration in superficial treatments and significantly reduces CT and MRI susceptibility artifacts.

An additional innovation of _ihtObtura is its gradual loss of radiopacity, which enables clearer visualization of the AVM angioarchitecture during sequential transvenous embolization sessions.^11–14 This feature may be advantageous in assessing potential complications and planning further interventions, as _ihtObtura produces minimal artifacts on conventional CT and no artifacts on MRI. ¹²

Radiopacity loss showed an inverse relationship with the injected volume and the diameter of the occluded vessels, suggesting a passive, diffusion-driven mechanism of clearance for the iodine-based radiopaque agent. The proprietary formulation is designed to gradually deconjugate from EVOH, implying that the radiopaque component diffuses out over time, thereby reducing the implant cast volume over time, without observed recanalization. ¹¹

Transvenous embolization has emerged as a safe and effective alternative for the treatment of complex AVMs, particularly those located in deep regions or classified as high-grade. ⁴ A systematic review by Chen et al., encompassing 13 studies and 70 AVMs treated via the transvenous approach, reported a complete obliteration rate of 93%. ¹⁹ Other studies have reported even higher efficacy, with angiographic obliteration achieved in up to 95% of cases when transvenous embolization is combined with transarterial embolization. ¹⁵ Over the past decade, the body of clinical evidence supporting the safety and efficacy of transvenous embolization for AVM management has grown substantially.^{2–6,18–25}

This report has several limitations that constrain the interpretability of its findings, particularly the small sample size and the absence of a comparison group. Nevertheless, it offers preliminary observations that may contribute to the emerging understanding of transvenous embolization of AVMs using this novel liquid embolic agent.

Conclusion

The transvenous retrograde pressure cooker technique using the extra-low viscosity _ihtObtura formulation appears to be a safe and effective approach for the endovascular treatment of brain arteriovenous malformations, showing favorable diffusion and penetration characteristics, with no cases of vascular rupture or extravasation observed in this series. The agent's gradual loss of radiopacity may assist in visualizing AVM angioarchitecture during staged procedures. However, these findings are preliminary and require confirmation through larger prospective studies.