|Home | About | Journals | Submit | Contact Us | Français|
This prospective open pilot study was designed to assess the suitability of flexible composite sheets of polymer and hydroxyapatite (HA) for the reconstruction of limited lateral skull base defects through a conservative transmastoid approach.
Seven patients with a petrous bone dehiscence less than 3 cm in diameter, either iatrogenic or caused by chronic otitis media or temporal bone trauma, underwent a mastoidectomy. The defect was repaired with the new material and a connective tissue graft. All bone defects were detected by computed tomography (CT) of the temporal bone except one that was found at surgery in an asymptomatic patient.
Outcome was evaluated in terms of anatomical integrity of the tegmina, absence of cerebrospinal fluid leaks, side effects, and extrusion and complication rates. None of the patients suffered from immediate side effects related to the implant or the operation. With a minimum follow–up of 18 months (maximum, 62 months), neither extrusion nor a foreign body reaction occurred. Postoperative CT confirmed a satisfactory anatomic contour. Although the number of patients is limited, these preliminary results are encouraging and indicate a need for further clinical trials with a material that allows a minimally invasive approach to selected skull base defects.
Cerebrospinal fluid (CSF) leaks, meningoceles, and meningitis are worrisome complications of skull base defects, regardless of whether they are congenital, traumatic, neoplastic, or iatrogenic. These defects are usually repaired through a limited subtemporal craniotomy with elevation of the temporal lobe through a combined transmastoid and subtemporal approach. Combined cranioplasty techniques are used to treat larger bone defects.1
Several materials have been used to reconstruct bony defects of the lateral skull base: autologous and homologous cartilage or bone; lyophilized homologous dura mater; pre–formed hydroxyapatite (HA) blocks; titanium or carbon fiber mesh; and myofascial pedicled or distant free flaps. Each material has unique advantages, but each also has disadvantages that can lead to failure of the implantation, adverse reactions, or complications.
To minimize operative complications, we initiated a prospective clinical study using a new composite membrane obtained in the laboratory by mixing HA powder and an organic polymer in appropriate proportions. The latter adds malleability to the typical features of HA (resistance and biointegration), allowing the material to be inserted easily through a transmastoid approach alone. The implant can be bent and accurately trimmed to the dimensions of the defect. We present the preliminary results of using this flexible HA sheet to reconstruct small to medium tegmental defects in seven patients. The surgical pitfalls and the properties of the material are highlighted.
HA is a bioactive ceramic with a molecular structure analogous to that of human bone, dentin, and enamel. It is composed of 58% calcium and phosphates in the form of submicroscopic crystallites. Its Ca:P ratio is 1.67:1 compared with 1.77:1 of normal bone. It is biocompatible, highly resistant to stress forces, nontoxic, and noncarcinogenic. It has marked osteoconductive and osteoinductive properties that lead to the regeneration of lamellar bone within 4 to 10 months of implantation. This characteristic is exploited in many surgical areas (orthopedics, maxillofacial surgery, dentistry, plastic surgery, neurosurgery). It also has been used to coat other materials to increase their stability and osteointegration. It is sintered (heat treated in specific thermal cycles) in the laboratory under forms of blocks (raw or presculptured), granules, and powder (cement). In otolaryngology it has been used to reconstruct defects of the jaw and other facial bones, to obliterate mastoidectomy cavities, and to reconstruct the ossicular chain and wall of the outer ear canal. Its main application in otoneurosurgery is cranioplasty procedures.
The compactness of HA blocks or prostheses is responsible for their minimal elasticity and tensile strength45 and requires time–consuming intraoperative remodeling with high–speed drills and diamond burs. To improve the flexibility of this material for prostheses, a composite ceramic membrane made of HA blended with an organic polymer was developed after many in vitro and in vivo trials at the laboratories of FIN–Ceramica, Faenza, Italy, in cooperation with the Italian National Council for Research.
The flexible HA sheet is a polymeric–ceramic composite with a controlled degree of resorption. It is constituted by an organic matrix (polycaprolactone) embedded with the granular porous HA. It is manufactured in rectangular plates (30 × 20 mm, Fig. 1), and its thickness varies from 0.3 mm to 1.2 mm (Table 1). The morphologic structure of the material, based on scanning electron microscopy, appears homogeneous (Fig. 2).
The HA/polymer plate is easily fitted to the size of a defect by trimming it with a scalpel or delicate scissors in a sterile fashion (Fig. 3). If bending is required, it is impressed by hand or with delicate anatomic forceps.
The flexible HA sheet was implanted in seven patients with small bony defects in the tegmen antri or tympani (Table 2). In two patients the defect was the consequence of a chronic middle ear infection associated with osteoclastic resorption of the middle cranial fossa plate by granulation tissue in one and with a cholesteatoma in the other. Three cases were the result of iatrogenic injury by inadvertent drilling of the dural plate during previous canal wall down (CWD) mastoidectomies. One patient sustained multiple fractures of the temporal bone during a motor vehicle accident. The last case was apparently idiopathic. A meningocele plunging into the mastoid cavity was observed in the latter two patients and in the patient with the long–standing iatrogenic defect (Patient 2).
In four patients the middle fossa dura exposure was clinically silent. The three patients with meningocele experienced repeated episodes of liquorrhea either from the nose through the Eustachian tube (Patient 6) or through the ear canal (Patient 7). One patient (Patient 2) complained of an aspecific headache without meningeal signs. This had been diagnosed by a neurologist as “cluster headaches.” Two patients with purulent otorrhea underwent surgery only after their ear had been dry for at least 1 week after local cleansing and antibiotic treatment. In one case only (Patient 5), the bone defect in the tegmen was observed directly by otomicroscopic examination (the patient was otherwise asymptomatic). In the remaining patients, the presence of the bony defect was confirmed by CT.
The eligibility criteria for repairing the defect with the flexible HA lamina were as follows: (1) The size of the bone dehiscence was less than 2 × 3 cm. (2) There was no active purulent disease in the mastoid/middle ear. (3) The patient had no neurologic impairments or history of seizures.
Preoperative axial and coronal CT scans of the temporal bone from all patients allowed the extent of bony resorption to be assessed and helped in planning the reconstruction. Patients suspected of having a meningocele underwent MRI of the brain that clearly defined the pathology (Fig. 4).
The polymer and HA sheet must be grafted in a field that is as sterile as possible. The conditions associated with chronic otitis media (presence of anatomical recesses, submucosal atrophy, osteitic lesions) are unfavorable. It is therefore necessary to administer a systemic antibiotic in the immediate preoperative period, possibly based on bacterial cultures. Local debridement and suction of the mastoidectomy cavity under micro–otoscopy are recommended for a few days. When otorrhea is present, auricular irrigations with mild acid solutions (e.g., acetic acid 1.5%) and instillation of antiseptic drops (e.g., boric acid 2% in 60 volume alcoholic solution) are indicated until the middle ear is dry. All seven patients underwent this treatment before surgery.
A mastoidectomy through a retroauricular transmeatal approach is the procedure of choice. If a previous CWD mastoidectomy has been performed, its revision is mandatory. Eradication of the disease is enforced, and a tympanoplasty with reconstruction of either the eardrum or ossicular chain is performed as needed (this was not the case in our patients). After the bone work is complete and the pathology is removed, the bone defect is visible (Fig. 5). The operative field is rinsed with an antibiotic solution.
The HA/polymer plate of standard size and thickness (20 × 30 × 1.0 mm), previously removed from its sterile envelope and dipped in a bath of antibiotic solution at the beginning of the procedure, is fitted precisely to the size and contour of the bony defect to be repaired through subsequent trials “in situ.” The plate is trimmed with a blade or Fisch tympanoplasty scissors in a sterile fashion. The middle cranial fossa dura (Fig. 6A) is smoothly detached 3 or 4 mm from the bony borders of the defect with a duckbill elevator (Fig. 6B). The final dimensions of the implant are slightly larger (1 to 2 mm) than the defect itself to permit limited overlapping. An autologous fascia temporalis graft is wrapped on the smooth side of the HA sheet facing the dura.
The HA plate is inserted first elevating the meninges from one edge of the defect (Fig. 6C). The flexible sheet is bent slightly and adjusted over the other borders of the dehiscence in the intracranial compartment (Fig. 6D). The HA implant must be covered completely by soft tissue to seal the intracranial compartments because the HA sheet is porous. A graft of autologous or homologous temporalis fascia or lyophilized dura mater is interposed between the HA implant and the tegmen bone using the above technique, or it is applied to the mastoid side of the tegmen bone with fibrin glue. Gelatin sponges and an expandable Merocel® Oto–Wick plug the outer ear canal.
Postoperative management is similar to that of conventional tympanoplasties. The outer ear canal is instilled daily with antibiotic and anti–inflammatory solutions. On the second postoperative day, the endoauricular Merocel® is replaced and removed after 1 week. Antibiotic prophylaxis is prescribed during the same period. Two weeks after surgery, the residual Gelfoam® is suctioned under otomicroscopic control. The correct positioning of the HA implant at the skull base and the initial success of the fascial graft are verified.
Postoperatively, office visits with micro–otoscopy (Fig. 7) were scheduled at 7 days, 1, 3, 6, 9, and 12 months, and every 6 months thereafter. The mean follow–up was 27.5 months (minimum 18 months, maximum 62 months). High–resolution axial and coronal CT scans of the temporal bone were obtained in three patients at 2 months and in all seven patients 12 months after surgery.
None of the patients suffered immediate side effects related to the implant or operation. Patient 2 continued to have episodes of ipsilateral temporoparietal headache that gradually subsided 3 months after surgery.
Twelve months after surgery, CT confirmed a satisfactory anatomic outcome in all patients. The contour of the mastoidectomy was smooth and re–epithelialized. In the three patients who underwent a CWD mastoidectomy, its roof was easily accessible through the external ear canal, and the reconstruction could easily be inspected.
Postoperative CT scans obtained 1 year after surgery showed the implant and its relationship to the bony borders (Fig. 8). In five patients radiographic signs of osteointegration were observed. A subtle rim of discontinuity was still present between the petrous bone and the implant in two patients.
At a minimum follow–up of 18 months (maximum, 62 months), no complications related to the implant have been encountered. At a maximum follow–up of 60 months, there have been no instances of extrusion of the implant or foreign body reactions. No implant has become dislodged, and no revision surgery has been necessary.
Many methods and materials have been used to obtain a waterproof seal of bone defects in the lateral skull base. These materials include autologous and homologous bone or cartilage grafts,4 alloplastic materials such as bone substitutes cements,5, 6, 7 local pedicled soft tissue flaps,8, 9 and distant free revascularized flaps.3, 10
Autologous bone grafts have almost been abandoned worldwide because they are prone to infection and difficult to harvest and fit to the defect, leading to instability and a lack of osseointegration. The process of remodeling usually reduces their size 15 to 25%, thus impairing the outcome of the reconstruction.
Cartilaginous grafts are sufficiently elastic but offer little resistance to deformation. They are not osteogenic. They are often reabsorbed to a variable extent or converted to fibroid tissue 1 year after implantation. Both cartilage and bone homografts and xenografts share the common drawbacks of availability, reabsorption, and infection. Furthermore, they are unavailable in some countries because of governmental regulations and medicolegal issues.
The use of alloplastic material gained widespread acceptance during and after World War I when many metals (aluminium, vitallium, tantalum) were introduced. Later they were abandoned in favor of less radiopaque meshes of stainless steel11 or titanium.12 At the beginning of the 1950s, acrylics were developed, and excellent cosmetic outcomes were achieved with methylmethacrylate13 because it is easily molded at the operating table. Later studies showed that acrylics had a low tensile strength and tended to crack under pressure. Improvement has been obtained by combining titanium meshes with methylmethacrylate paste.14
Alloplastic materials (i.e., plastics, ceramics, and metals) share the common problems of biocompatibility and extrusion.15 Implants can fail immediately from exposure and infection or in a delayed fashion from graft retraction, implant fragmentation, and extrusion. Most failures are related to limited blood supply and incomplete coverage of the prosthesis with soft tissue.16, 17 Possible reasons for exposure are the presence of sharp angles and failure to fit the implant exactly to the defect because of the difficulty of modeling the material. The current trend is toward lightweight ceramics and other synthetic composites that promote osteogenic ingrowth at the edges of the implant.
HA has optimal osteoconductive properties that influence its speed of reabsorption and are primarily regulated by its porosity.18, 19, 20 Direct and stable contact with the host's bone stimulates osteogenesis and therefore integration of the material.21 A mucous coating usually surrounds the implant with no signs of reabsorption or foreign body reaction as occurs with organic grafts.22
The consistency and compactness of HA are responsible for its considerable resistance to compression and rigidity. Pre–formed rigid blocks of HA have already been used to repair petrous bone defects through a subtemporal craniotomy23, 24, 25, 26, 27 or combined transmastoid and middle cranial fossa approach.28, 29 Unfortunately, presculptured HA blocks require tedious intraoperative reshaping, and the result is not always satisfactory.
The material used in the present study is a composite ceramic whose main feature is flexibility obtained by the addition of an organic polymeric matrix (polycaprolactone) to the HA powder. It was originally developed for orthopedic applications to add a certain degree of malleability to rigid prostheses. It has also been used by dental surgeons to fill bone defects caused by cystectomies, odontectomy cavities, and bone atrophy or to increase the thickness of bone in reception sites of prosthetic implants.30, 31, 32, 33 Experimental studies in rabbits have shown that the organic component is gradually reabsorbed within 6 to 10 weeks of implantation. The HA structure is initially invaded by fibroblasts and neoangiogenesis is marked.2 The fibrous ingrowth is replaced by regeneration of lamellar bone 2 to 10 months after surgery and leads to firm integration with the surrounding autogenous bone.
To our knowledge this is the first report of using HA sheets to repair lateral skull base defects. The characteristics of the material help minimize operative complications by allowing a shift from a craniotomy or combined approach to a less invasive transmastoid approach. The main limitation is the size of the dehiscence. Large cranial defects, usually post–traumatic, require a combined cranioplasty with wide surgical access. They are unsuitable for reconstruction with the flexible HA sheets described here, whose standard dimensions are 20 × 30 mm. Although larger sheets are manufactured for maxillofacial purposes, there is no guarantee than their tensile strength is sufficient to withstand intracranial pressure because their thickness is limited (1 mm). High failure rates have been reported with a transmastoid approach alone.
Nevertheless, repairing lateral skull base defects smaller than 1 cm with this material is considered feasible.6, 27, 28, 34, 35, 36, 37, 38, 39, 40 It could be argued that there is no need to repair such small skull base dehiscences because they are harmless. However, CSF leaks and meningitis have been associated with microscopic fissuration in the petrous bone.29 These minor petrous bone defects are either iatrogenic (Table 3), including the planned surgical bone resection for neoplastic lesions, or the consequence of a suppurative temporal bone osteitis from otogenic foci. Although rare, a congenital defect can settle on a spontaneous meningocele or meningeoencephalocele over many years.
To date we have used HA sheets to repair the outer ear canal wall in 106 patients with chronic otitis media. A follow–up of more than 60 months is available for the first 42 patients.41, 42, 43 The extrusion rate of 16.6% has always been linked to postoperative infection at the implant site; no complications related to the HA implants have been observed. As with other alloplastic materials, the main problem is exposure of the implant, which leads to unavoidable infections. The implant must be coated meticulously with soft tissue to be successful.
Our recommendations for the transmastoid repair of skull base bone defects with the new material are as follows: (1) The operative field must be rinsed with an antibiotic solution. (2) The HA sheet must be bathed in an antibiotic solution for 15 minutes before implantation. (3) Connective tissue must coat the side of the implant in contact with the dura, and, possibly, the mastoid side. (4) The implant must fit the defect precisely to avoid sharp angles. In our seven patients, strict adhesion to these rules avoided both the extrusion of the implant and inflammatory reactions at the implant site.
In our experience it was unnecessary to reduce CSF pressure during the early healing stages to promote dural closure. The implant is placed in the intracranial compartment between the dura and the petrous bone and thus held in place by gravity and the physiologic intracranial pressure. Further sealing of the mastoidectomy side by temporalis muscle fascia and fibrin glue afforded a watertight dural closure without evidence of CSF leaks.
When a CWD procedure has been performed, the anatomic outcome of the tegmen repair can easily be assessed through the wide outer ear canal meatus. A feasible option in the presence of an intact canal wall mastoidectomy is to inspect the cavity by a fiberscope through a retroauricular puncture under local anesthesia.44
Although a combined or isolated intracranial procedure (i.e., subtemporal craniotomy or middle fossa approach) is essentially extradural, the more conservative transmastoid approach alone minimizes operative times and complications. The reported high failure rates associated with this approach27, 39 have been overcome by the pliability of the material that facilitates its insertion by trimming and bending. There have been no adverse reactions to the material so far, and none of the patients has manifested a CSF leak after the careful reconstruction described above. Although the number of patients is limited, the initial clinical trial with flexible HA sheets appears promising. Further clinical tests are needed to determine the long–term outcomes associated with these implants.
We thank Dr. Angelo Nataloni, FIN–Ceramica, Faenza, Italy, for the laboratory trials and development of the material. The present study was partially supported by a grant from Bielle Medicale, Brescia, Italy.
The authors report their experience with seven patients with minor skull base defects that were treated through a transmastoid approach with flexible hydroxyapatite sheets. These defects were less than 3 cm in diameter. There were no major complications, and all cerebrospinal fluid leaks and tegmental defects were treated successfully. The minimum follow–up was 18 months.
This approach represents a viable alternative for the treatment of tegmental defects. At our institution, we typically perform a middle fossa craniotomy, elevate the dura of the temporal lobe, and place a split–thickness skull graft in the defect. Depending on the size of the defect, we may place a lumbar drain. We have used a hydroxyapatite sheet for cranioplasty during closure of retrosigmoid craniotomies. After performing about 40 cases, however, we stopped using it because the hydroxyapatite sheet had eroded through the ear canal or caused infection in several patients.
The authors' recommendation to soak the product in an antibiotic solution is well taken. They should be congratulated on their innovative approach to this common problem.
Drs. Zanetti and Nassif present their initial experience with the use of flexible hydroxyapatite sheets to repair lateral skull base defects. They prospectively reviewed the outcomes of seven patients with mastoid defects, 3 cm or less, resulting from traumatic, iatrogenic, or infectious causes. After a mean follow–up of 27.5 months, none of the patients suffered any complications related to the graft such as infection, extrusion, or cerebrospinal fluid leaks. All patients exhibited a reasonable anatomic contour on postoperative computed tomographic scans. Although this experience is limited, flexible, osteoconductive hydroxyapatite sheets offer a viable alternative for the repair of skull base defects, and further clinical evaluation is warranted.