Fatal bleeding can occur with mandibular arteriovenous malformations (AVMs) during dental manipulation or biopsy if not diagnosed and treated appropriately, with bleeding being the main presentation and a predilection for the posterior jaw.1 The diagnostic approach may include, a panoramic x-ray film, computed tomography (CT) or magnetic resonance (MR) study, digital catheter angiography.2 Intra-arterial catheter angiography is the gold standard2 and the angiographic appearance of mandibular lesions has been classified into 5 types to aid therapeutic decision making.3 An arteriolovenous malformation is a type of AVM where plexiform appearing multiple feeding arteries shunt into a single draining vein.4 Treatment options include endovascular embolization, sclerotherapy, surgical intervention, or a combination of these treatment options.2,5,6
Diagnostic angiography was performed using a 5F guiding catheter (Envoy, Johnson and Johnson, New Brunswick, NJ) as a standard for external carotid artery diagnostic angiography with selective catheterization of the feeding arteries performed with a diagnostic catheter (Cobra, Cook Inc, Bloomington, IL) if vessels size appropriate or a microcatheter (Rebar, Medtronic, Dublin, Ireland) if smaller. The choice of embolic agent was determined by the degree of fistulous outflow and the size and number of the feeding arteries. Polyvinyl alcohol (PVA) particles (Contour, Boston Scientific, Marlborough, MA) of larger dimension (500–700 µm) were used to reduce flow and stasis confirmed with angiography. Since both of these lesions were classified as Liu type 2,3 the decision to use an occlusion balloon (Scepter XC, Microvention Inc., Tustin, CA) as an adjunctive technique served 2 purposes; one to reduce arterial flow while administering PVA particles and the second as the sole temporary occlusive agent in the artery when performing direct venous embolization. Percutaneous ethylene-vinyl alcohol copolymer (EVOH) (ONYX/ev3 Micro Therapeutic Inc., Irvine, CA) was injected under fluoroscopic guidance via a 21-g micropuncture needle connected to a slip tip T connector and a polyethylene syringe (3–5 mL) due to the volume of EVOH anticipated and the desire for constant injection. The sites chosen for percutaneous access were based on the fluoroscopic appearance of the lesion, that is, lucency and the intraoral sites were chosen based on the CT appearance of a soft tissue component and floating teeth. Prior bench testing of the slip tip T connector had been performed by the operator, with a dimethyl sulfoxide dwell time of approximately 1 hour (anticipated direct injection embolization time) to confirm the T connector did not degrade. The system was primed with dimethyl sulfoxide before EVOH. The EVOH was injected using subtraction roadmapping and at a constant slow rate of injection, until the entire nidus was filled and confirmed by angiography. Intermittent new subtraction roadmaps were obtained to confirm new lesional filling during injection. EVOH was chosen as the percutaneous embolic agent due to operator experience. Prophylactic preprocedural intravenous antibiotics and postprocedural oral antibiotics were administered for 10 days to reduce the risk of infection.
A 13-year-old male presented with painless loose teeth in the right posterior mandible, confirmed on physical examination and on a Panorex radiograph to be associated with roots 29 and 30. MR imaging described a mandibular arteriolovenous malformation with arterial feeders from the right facial, lingual, and inferior alveolar arteries (Figure 2A). Angiography confirmed the MR findings (Figure 1A) with the mandibular lucency occupied by the dilated outflow venous component of the arteriolovenous malformation (Figure 1B), draining to the internal jugular vein. The direct approach for EVOH delivery in this case was intraoral (Figure 1C). A 5-week follow-up angiogram revealed near-complete obliteration of the nidus with a possible small residual nidal component present along the anterior aspect of the EVOH too small for intervention (Figure 1D). At 5-year follow-up, CT imaging demonstrates remodeling of the mandible with realignment of the teeth, no residual mandibular lucency, a stable EVOH cast, and no recurrence (Figure 2B–D).
A 12-year-old Caucasian male was evaluated by an oral and maxillofacial surgeon for a painful expansile left mandibular mass of 6 months duration (Figure 3A). Biopsy of the lesion resulted in significant oral bleeding treated with packing for 3 weeks. The child was referred to our institution with packing in place. Magnetic resonance imaging demonstrated findings consistent with intraosseous arteriolovenous malformation (Figure 3B). Angiography and embolization performed using the technique described above, revealed arterial feeders from the left inferior alveolar and left lingual arteries (Figure 4A, C). The mandibular component again was comprised of the venous outflow component (Figure 4B). EVOH delivery in this case was direct via a percutaneous approach (Figure 4D–F). At 5-year follow-up, the child had achieved both clinical and radiologic resolution of the mandibular intraosseous arteriolovenous malformation. Mandibular remodeling is evident on serial radiographs in the last 3 years. The teeth have realigned and the EVOH cast is stable without recurrence (Figure 3C, D).
Neither of the 2 patients had cutaneous vascular lesions or underwent genetic screening at the time for syndromic vascular malformations. There was also no family history of vascular malformations. The peak skin dose for both procedures was as follows; case 1, 1264 mGy, and case 2, 1198 mGy. There were no procedure-related complications or delayed complications for these procedures.
Mandibular AVMs require a multidisciplinary approach with the goal of nidal obliteration. A meta-analysis of 25 publications with a total of 50 mandibular AVMs treated with either endovascular or combined endovascular and direct embolization describe response rates of 50% for clinical stability and reported ossification responses in 36%.2 The complication rate reported was 50%, mostly minimal and self-limited. Comparing our experience with a similar subset of 29 cases (15 reports) in this meta-analysis, it is important to note that none of the reports described EVOH for direct embolization. Bone reossification was reported as an outcome in 45% (n = 12/29) with follow-up ranging from 6 months to 8.5 years, higher rates than reported in the overall meta-analysis. The complication rate in this cohort was higher compared with the overall complication rate reported (52% [n = 15/29] versus 50%). Major complications reported included massive intraprocedural bleeding, pulmonary migration of embolization materials, tongue necrosis, recurrence of bleeding, infection, and mucosal dehiscence with oral perforation by coil.
The use of PVA as a flow reduction embolic agent not a therapeutic agent was deemed useful in this experience to reduce the risk of direct injection nontarget EVOH embolization. Stasis was achieved on angiography before EVOH embolization for a more controlled result.
This series is the only series to our knowledge reporting direct EVOH injection for nidal obliteration. There were no immediate or long-term procedure-related complications; in particular, case 1 underwent intraoral injection without evidence of infection or dehiscence. Our deviation from embolization and surgical curettage was successful with long-term follow-up of 5 years. Remodeling to normal appearance was noted for case 1 as early as 6 months and for case 2 at 2 years (Figure 2A, 3C).
Vascular anomalies of the mandible are rare, challenging to diagnose and treat. Careful evaluation is needed to avoid life-threatening bleeding during biopsy and tooth extraction. To our knowledge, this is the first report of direct intraoral/percutaneous injection of EVOH in combination with arterial embolization for mandibular arteriolovenous malformations and may be a useful embolic agent for consideration for this approach.
1. Li X, Su L, Wang D, Gui Z, Jiang M, Yang X, et al. Clinical and imaging features of intraosseous arteriovenous malformations in jaws: a 15-year experience of single centre. Sci Rep. 2020;10:12046.
2. Monteiro JLGC, de Arruda JAA, Figueiredo Leal JL, Batista LL, Célia de Aguiar Soares Carneiro S, do Egito Vasconcelos BC. Embolization
as the primary treatment for mandibular arteriovenous malformations: an analysis of 50 literature reports and of an illustrative case. J Oral Maxillofac Surg. 2018;76:1695–1707.
3. Liu DG, Ma XC, Zhao FY, Zhang JG. A preliminary study of angiographic classification and its correlation to treatment of central arteriovenous malformation in the jaw. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 2005;100:473–480.
4. Houdart E, Gobin YP, Casasco A, Aymard A, Herbreteau D, Merland JJ. A proposed angiographic classification of intracranial arteriovenous fistulae and malformations. Neuroradiology. 1993;35:381–385.
5. Giaoui L, Princ G, Chiras J, Guilbert F, Bertrand JC. Treatment of vascular malformations of the mandible: a description of 12 cases. Int J Oral Maxillofac Surg. 2003;32:132–136.
6. Ravula P, Rangachari S, Susarla R, Sambari L, Jammula SS. Primary free fibula reconstruction in life-threatening haemorrhage from high flow arteriovenous malformation of mandible. Indian J Plast Surg. 2018;51:222–230.