Implant Therapy in the Rehabilitation of Treated Mandibular Arteriovenous Malformations Open Access

Arteriovenous malformations (AVMs) are rare congenital malformations characterized by the presence of a rapid-flowing arteriovenous shunt (nidus), without a capillary system, and draining into one or several veins.¹ 

They make up 1.5% of vascular malformations; 50% of AVMs have an orofacial location.²  The majority (70%) of maxillofacial AVMs affect the cheek, nose, upper lip, ear, and scalp.^2,3  Intraosseous AVMs are rare, and this location is surgically and prosthetically more difficult because of the associated risk of morbidity and mortality.⁴ 

Treatment of AVMs most often includes embolization and/or resection surgery. Implant-supported rehabilitation of this bone defect is addressed minimally in the literature. There are multiple implant-related therapeutic considerations: (1) difficulty in surgical planning due to radiological artifacts, (2) lack of knowledge of implant osseointegration capacity in contact with embolization products, and (3) prosthetic management of the vertical bone defect. In this article, we describe the implant treatment with a 3-year postplacement follow-up of a patient who had previously undergone embolization and surgery for a mandibular AVM. The different and specific problems with this pathology are explained and discussed.

A 24-year-old patient was taken into care for implant rehabilitation. The patient presented with an antecedent stage 3 AVMs of the right mandible. It was diagnosed following massive hemorrhage and treated with embolization and surgery in 2008. Surgery by osseous curettage with a noninterruptive mandibulectomy (Figure 1) was carried out, and the nidus was totally eliminated. An annual checkup was carried out in order to survey possible late recurrence. Nine years after the initial management of the AVM, rehabilitation with implant therapy was considered after objectification of the absence of late recurrence or residual AVM on magnetic resonance imaging (MRI) and arteriography. Then an osseous graft with iliac extraction was carried out in order to compensate for the vertical bone defect. After graft healing, the clinical showed a persistent osseous deficit on the vertical plane with a fine strip of keratinized gingiva with partial modified right labio-mental sensitivity (paresthesia). Interpretation of the radiological images was made difficult by the presence of the embolization material, and therefore it was not possible to locate the position of the mandibular canal and the mental foramen. In fact, surgically, the placement of implants with a length of 8.5 mm was not possible due to the risk of damage to V3 and the potential interference with the osteosynthesis material.

The surgical installation of the implants was carried out under local anesthetic with the help of a surgical guide. The implants were placed (Figure 2A and B). No pre- or postoperative bleeding was noticed. The postoperative period was without complications. Three months after installation of implants, phase II surgery was complete. After an additional 2 weeks, 3 multiunit abutments were placed and tightened to 30N/cm followed by a positional impression. A plaster validation key and a screw-retained occlusion model were constructed. These allowed verification of the impression (fit passivity) and the recording of intermaxillary records (Figure 3).

A chrome-cobalt armature was then machined and clinically validated through radiological and occlusal control. The prosthesis was then placed. The prosthesis screw was tightened to 15 N/cm. Access wells were sealed with the PTFE (Teflon) to protect the screw heads followed by the placement of photopolymerizable composite to ensure sealing (Figure 4). Hygiene techniques were explained, and a prescription for interdental brushes was given to the patient.

Intraosseous maxillo-mandibular localization of AVMs is rare, constituting less than 5% of all AVMs.⁵  They are normally situated in the posterior mandibular areas.

Clinically, the symptoms recorded in the literature are facial asymmetry, tooth mobility, swelling, discolorization of the mucous membranes, a raised pulse or breathlessness, gingival bleeding, pain, and paresthesia. The radiological appearance is not pathognomonic. Vascular lesions (low or fast flowing) create bone erosion and demineralization. We find radiolucent, multilocular lesions (similar to a soap bubble form) or absent trabecular bone.⁶  On the scanner, the cortical bone is not affected, but it can be perforated when there is an extraosseous component to the malformation.³ 

This absence of clinical and radiological specificity could be the reason for a massive “iatrogenic” hemorrhage either after a biopsy or during an excision when the lesion is mistaken for a cyst or an osteoma. These hemorrhages are also found in children after the loss of a temporary tooth or after a dental avulsion. They can therefore be life threatening.^6–8  The warning signs of these malformations (raised heart rate and breathlessness) must be known in order to guide the diagnosis and limit the risk of an iatrogenic hemorrhage.

Treatment varies according to the stage of the AVM, its location, and its architecture. Because of its anatomy, the maxillofacial region poses major difficulties, and surgical interventions must be weighed in relation to functional and aesthetic consequences. These interventions with consequences pose reconstructive problems all the more so because the patient is often young. It can be achieved with autologous bone grafts or by microanastomosis flaps.⁹  As in our observation, bone curettage preceded by embolization is a less damaging alternative to bone resection.

Bone reconstruction of AVMs is widely discussed in the literature;^9,10  however, there is a lack of detail concerning associated prosthetic rehabilitation. Bagherzadegan et al⁶  and Qu et al¹⁰  describe implant-supported treatment in a maxillo-mandibular AVM rehabilitation.

Implant treatment is described in the work of Qu et al,¹⁰  where 4 implants supported a removable prosthesis in a 52-year-old adult with an AVM treated several times by embolization (coils + ethanol) and surgery. The implants were located in the right mandibular body. The left mandibular body showed mucosal, submucosal, and crestal extension of the AVM. Due to the risk of hemorrhage, the potentially traumatic removable prosthesis was perforated in this area. The follow-up over 3 years revealed no implant loss or peri-implantitis.¹⁰  No prosthetic fractures or modifications were made, and the patient was very satisfied.

The clinical case of Qu et al¹⁰  different from the patient in this case report, as it concerns an adult presenting an uncured AVM and a removable implant-supported prosthesis. Our observation deals with problems not mentioned by Qu et al that are directly linked to prosthetic rehabilitation of AVMs.

The concerns considered in this case report are the following:

• The preoperative workup. Bone drilling during implant placement can be risky in the case of AVM. In our opinion, when treating a patient with a cured AVM and followed up annually, it is necessary to evaluate for the absence of recurrence. Evaluation should be included in addition to normal imaging (dentascanner, cone beam computed tomography [CBCT]); it is essential to utilize comprehensive imaging techniques. Techniques should include MRI angio (as in our clinical report) or computed tomography (CT) angio (or even arteriography) performed in association with radiologists The use of comprehensive techniques help in the consideration of the recurrence risk or aftereffects of the AVM before considering rehabilitation of the implant.

• Implant planning. Following CBCT images, we observed numerous artifacts linked to the embolization agents that made interpreting the scan difficult. Metallic agents (coils) or nonabsorbable radio-opaque agents like Onyx in our observation make it difficult or impossible for (1) interpretation of bone morphology, (2) localization of the mandibular canal, and (3) identification of the mental foramen. In this context, it was decided to place implants <8.5 mm in length. These short implants were selected to limit the risk of labio-mental anesthesia through damage to the inferior alveolar nerve. The use of the latest CBCT rather than a scanner could be considered in order to limit artifacts.¹¹ 

• Implant osseointegration is defined as the direct anatomical and functional junction between the bone and implant surface. There are numerous factors that influence osseointegration, including bone vascularization and bone type. The defining considerations of our observation are linked to the following:

• 1. Unstable vascularization, at least centrally, as the alveolar artery has been embolized. In our case, unstable vascularization, even when offset by anastomoses, could have contributed in part to the failure of the iliac bone graft. However, it did not compromise osseointegration of the implants.

  2. The presence of embolization product within the alveolar bone. One can evaluate osseointegration in different types of bone (from I to IV; Table 1) based on evidence-based literature;¹²  however, it is difficult to predict the behavior of an implant in a bone impregnated with exogenous material. It can involve the use of a solid material (coils), which can mechanically prevent the insertion of the implant, or a nonabsorbable plastic product, as in our case. It is logical to consider that the presence of this material will reduce the implant surface in contact with the bone and thus diminish the degree of osseointegration. However, we are not aware of the consequences of contact between ethylene-vinyl alcohol copolymers and titanium.

We find no data relating to this particular point in the literature. Computer-assisted surgery could be of interest in implant planning, as it allows access to the anatomical areas deprived of embolization material.

Prosthetically, the first problem to deal with is the height difference between the implant emergence and the occlusal plane resulting from an often significant vertical bone defect. It will sometimes be necessary to extend the height of transfer impression pins so that they can be seen and retained in the dental impression.

We have chosen to splint the implants, which is advantageous in several ways. Essentially, this means it is possible to distribute the load transmission generated by masticatory forces across all the implants and to their neighboring bone structures.¹³ 

This splint of the implants also means we are free of the difficulties associated with adjusting the contact points of nonsplinted implant-supported crowns. The adjustment of the contact points is indeed delicate, and Guichet et al¹⁴  and Tiossi et al¹⁵  demonstrated that the absence of proximal contact can increase the force by approximately 30% at the implant collar. Several authors have also shown that the splinting of adjacent implants is indicated in the case of short implants, small-diameter implants, or low bone density.^16,17 

Nevertheless, one must bear in mind that during the construction of an implant-supported prosthesis that splints two or more implants, the major problem is the creation of a totally passive prosthetic infrastructure. This absence of passivity imposes serious constraints, and this could be the reason for microfractures or ischemia of the peri-implant bone, which could compromise osseointegration and lead to cratering.¹⁸ 

In order to reduce this risk, special attention will be paid to the creation of the impression that will be adapted to the clinical indication. A plaster validation key will always be produced.

Finally, we will favor computer-aided design/computer-aided manufacturing (CAD/CAM) machining techniques, which eliminate the risk of deformation inherent to metal molds and improve the passivity and precision of the armature's adjustment.¹⁹ 

Several cases of AVM patients with diverse (1) initial medical management, (2) surgical treatments, and (3) implant-prosthetic rehabilitations should be assimilated, submitted to peer review, and published in order to improve knowledge and standardize treatment.

AVM is a rare pathology that is not well known by implantologists. Implant-supported treatment is entirely conceivable as long as the condition is surveyed closely. AVM treatments frequently result in bone defects that are difficult to reconstruct because of probable unstable vascularization due to embolization. Advanced imaging (MRI angio, CT angio, or even arteriography) will assist in making it possible to objectively determine the absence of disease recurrence before any surgical and prosthetic management.

Surgically, the preoperative workup takes on critical importance in order to avoid any remains that could make the surgery unsafe. Specific problems include (1) visualizing the critical anatomical elements due to radiological artifacts (induced by the embolization product) and (2) the lack of vascularization, which may affect implant osseointegration.

Prosthetic rehabilitation should compensate for vertical bone loss and poor keratinized gingival support. Implants should be splinted, and to ensure a passive fit, the prosthesis should be created by CAD/CAM technology.

The creation of a case series could be of interest in order to better understand implant treatment for patients with a history of AVM.

Andrew Jannetta is thanked for translation.