Cranial bone as a source for intra-sinus grafts

Introduction

Since the first publication on the subject by ^{Boyne and James (1980)}, the frequency of use and indications for the autogenous sinus floor bone graft have increased to the point that it can be considered a routine technique for the restoration of maxillary posterior segments by implants. The widespread use of sinus bone grafts results from the reliability and lack of complications of the procedure :

- reliable, because after eight years in function, the bridges are still in place and the grafted area remains stable, so much so, that in cases of extreme atrophy of the maxilla it is the sinus floor reconstruction which has the best prognosis ;

- lack of postoperative complications. When they do occur, they are associated with the graft donor sites : cranial bone can be removed without pain, haematoma and with the minimum of discomfort to the patient.

The first cranial graft appears to have been undertaken by ^{Dandy in 1929} but the regular use of the calvarium for the reconstruction of the face and cranium is the work of ^{Tessier (1982)}. From the end of the 1980s, we are led to believe that the cranium is the preferred donor site for pre-implant bone reconstructions. Our first report on the results of maxillary sinus bone grafts were published in 1993, four years after the first patient was grafted (^{Tulasne, 1993}). This study involved 75 sinuses, of which 53 (70 %) were reconstructed with cranial bone. An update of cases four years later indicates the stability of these first results and reviews 243 subsequent sinus graft operations (Tulasne, 1998). This paper presents our experience of 369 sinus operations of which 306 (83 %) were grafted with cranial bone (^{table 1}).

Materials and method

Between September 1990 and the end of November 1998, 306 sinus graft operations were undertaken using cranial bone in 187 patients, of whom 147 (76 %) were female. The mean age of the subjects was 49 years (25 to 77 years).

A bone graft is indicated when the height of residual bone is equal to, or less than 7 mm, taking into account bone density, the width of the ridge, the number of teeth missing and the occlusion of the individual patient. The condition of the residual teeth, maxillary bone and of the sinus is assessed clinically and radiographically by orthopantomogram and CT scan. Any pathology due to infection is treated in advance and, if necessary, appropriate periodontal therapy undertaken prior to the operative intervention.

The surgical treatment is carried out in a minimum of two stages with a six month interval :

- reconstruction of the floor of the sinus with a bone graft ;

- placement of the implants and possibly the attachments if it is decided not to bury the implants.

The bone graft reconstruction

It is undertaken under general anaesthesia. This was bilateral in 119 patients, whilst in the other cases, there was no clear predominance of one side over the other. The reconstruction was comprised of four successive stages :

Dissection of the sinus mucosa

The antero-lateral surface of the maxilla was approached via an incision either at the depth of the vestibule or, by preference, by an incision along the crest of the ridge, in order to preserve the periosteum covering the entry to the antrum. A direct approach via the alveolar ridge, essential in the case of simultaneous or recent extractions, greatly facilitates its augmentation with bone grafts. It allows the possible placement of implants at a neighbouring site at the same operative visit.

The antero-lateral wall of the maxilla is fenestrated under abundant irrigation until the sinus mucosa is exposed over an area of 3 to 4 cm² (approximately 15 mm high and 25 mm wide). The mucosa is progressively elevated from the wall of the sinus as far as the ostium, thus creating a cavity of 10 to 15 cm³ which has a roof of mucosa and whose bony walls are carefully scored to receive the graft (^{fig. 1}). The dissection is much easier when the mucosa is thick and when the floor is smooth. Perforation of the mucosa, almost inevitable when there are bony ridges within the sinus, generally has no effect on bone healing.

Harvesting the bone graft from the cranium

The technique has been described in detail by Paul ^{Tessier in 1982}, using osteotomes which have been especially designed and which greatly facilitate the taking of the grafts (^{Kline and Wolfe, 1995}). The graft is harvested from the parietal region, behind the coronal suture and at a distance from the sagittal suture (^{fig. 2}). After scoring the external cortical surface with an acrylic bur, the area to be removed is superficially defined with an oscillating saw. In this way, the outline of the graft is traced in the shape of a rectangular patch approximately 45 × 15 mm. An initial small graft is outlined just behind the coronal suture and is removed with a chisel. This is to estimate the density of the bone (as judged by the ease of cleavage) and to allow the orientation of the chisel to be tangential to the vault for the lifting of the next graft. The grooves which were previously made with the saw are enlarged and deepened as far as the diploe, using a fissure bur (^{fig. 3}). After having removed several rectangular grafts from the external cortex (a minimum of three grafts is needed for each sinus), flakes of spongy bone are removed from the diploe, and the sides of the donor site smoothed down with a chisel and restored with chips of the cortical/spongy bone (^{fig. 4}). The sides are then smoothed with a drill in order to erase all projections from the area. After closure of the scalp in two layers with an vacuum drain, a compressive dressing is applied to complete the operation.

Grafting the floor of the sinus

In order to ensure that the reconstructed bone is of uniform density, it is necessary to pack the graft down tightly to the floor and walls of the sinus. In order to achieve this, a large graft composed of both cortical and spongy bone is placed horizontally in the sinus, 10-15 mm above the floor. This forms the roof of the cavity to be filled (^{fig. 5}). The cortical surface is thinned and perforated in several places to encourage revascularisation. It is then pressed against the posterior wall of the sinus and, with the cortical surface towards the mucosa, it is located anteriorly into a groove made in the canine pillar (^{fig. 6} and ⁷). The cavity thus formed is grafted with bone particles prepared by passing the grafts through a bone crusher (Tessier microtome) and forcibly packing the particles into the space in order to produce a homogeneous structure. (^{fig. 8}).

Augmentation of the alveolar crest

It is not essential to close the hole in the wall of the antrum, but it goes hand in hand with augmentation of the external surface of the alveolus which is usually resorbed following extraction of the teeth (^{fig. 9} and ¹⁰). In cases of complete resorption of the alveolus, two grafts placed side by side, one buccal and the other palatal, permits more favourable bone conditions for an implant-supported prosthesis (^{fig. 11}). On the other hand, the quality of the mucosa deteriorates in proportion to the volume of the bone grafts because it is necessary to cover them with vestibular mucosa, which is a less suitable tissue around the implants.

Surgical follow-up

The patient can be discharged the day after the operation. The after effects of the surgery are usually minor, limited to oedema over the maxilla which has been operated on, with a feeling of pressure and, in some cases, a general feeling of weakness which usually disappears after a week. Anaesthesia of the upper lip and of the sub-orbital regions occurs frequently and disappears after several weeks or months.

The placement of the fixtures and abutments

It is usually undertaken after an interval of six months, and can be done in a single surgical visit if it is decided not to bury the implants.

The position of the implants is determined by the results of the follow up scan taken at least six months after the bone grafts (^{fig. 12a} and ^12b). In all, a total of 456 implants have been placed in 306 sinus grafts ; 390 implants have been followed up clinically and radiographically after a minimum of six months of bone healing, mostly after burial under the gingiva. Resistence to attempts to tighten or slacken the healing cap and the absence of mobility are the most reliable criteria for osseointegration.

Results

Implants

The results are summarised in ^{table II} . Of the 390 implants which were monitored after six months of burial, the percentage which osseointegrated was 94 %. Of the 5.4 % failures, half failed to osseointegrate and half lost their osseointegration after exposure.

Marginal peri-implant bone

No abnormal resorption was observed.

Intra-sinus bone graft

Follow up scans were carried out on 7 patients between 4 and 6 years after the grafts. The images were, in all respects, comparable to those obtained at the 6th month postoperatively (^{fig. 13}, ^13b and ^13c and ¹⁴, ^14b and ^14c).

Complications

Implants

Several had to be removed, either because of failure to integrate (11 implants) or because they lost their osseointegration after being exposed (10 implants).

Bone grafts

Infection occurring in the first weeks (3 cases) could be controlled with drainage and with antibiotics, possibly by intravenous infusion in cases where the local reaction was very severe. After the 2nd postoperative month, complications are most unlikely, except for the need to remove excess graft material which could show itself as an asymptomatic alveolar fistula. In our experience, perforation of the sinus mucosa is not related to the incidence of infection. Experience has shown that maxillary osteotomies which involve the sinus without taking into account its mucosa never suffer the complication of infection, even in those cases which incorporate bone grafts.

No complications in connection with the cranial donor site were observed, except for two cases of haematoma under the scalp. Several patients complained of a slight depression in the area operated upon.

Discussion

Why cranial bone ?

The advantages and disadvantages of bone grafts from various sites have been reviewed in a previous article (^{Tulasne, 1990}). Following the conclusions of ^{Tessier (1982)}, cranial bone already appears to be an ideal reconstructive material. It is easy to harvest, there is minimal postoperative discomfort and the quality and stability of the reconstruction offer excellent conditions for the anchorage of implants due to the high density of the graft (^{fig. 12a} and ^12b). Experience accumulated over the last 8 years, together with experimental studies, confirms more than ever the superiority of cranial grafts over iliac grafts for cranio-facial reconstructions (^{Hardesty and Marsh, 1990} ; ^{Donovan et al., 1993}). Generally speaking, it has been demonstrated that bone of membraneous origin resorbes less than bone of endochondral origin (^{Smith and Abramson, 1994} ; ^{Zins and Whitaker, 1983}), for which no explanation has been found.

Contradictory hypotheses…

It has been suggested that cranial bone resorbes less because early revascularisation occurs in membraneous bone, as has been demonstrated by ^{Zins and Whitaker (1983)} and later by ^{Kusiak et al. (1985)}. However, several investigations (by ^{Albrektsson, 1980}, as well as others) have demonstrated the reverse, that spongy bone is revascularised more rapidly and intensively, especially that from the iliac bone (^{Sullivan and Szwajkun, 1991}), in comparison with cranial bone.

A seemingly logical explanation

^{Hardesty and Marsh (1990)} have advanced the hypothesis that differences in the resorption and incorporation of grafts are due directly in their three dimensional structure. The predominantly spongy/cancellous structure of iliac bone permits more rapid penetration of blood vessels than is the case with the denser cranial bone and osteoclastic activity will be more pronounced, thus leading to collapse of the graft cortex. This structural explanation has been reinforced by the experimental studies of ^{Sullivan and Szwajkun (1991)} showing that cortical bone seemed to act as a barrier to vascular penetration. In addition, the relatively thin cortex of iliac grafts is probably more susceptible to resorption than the more robust cranial graft, before which, osteogenesis on the surface cannot commence (^{Hardesty and Marsh, 1990}). In summary, cranial bone resorbes less because of the predominance of cortex, although there is, to date, no scientific explanation as to the mechanisms of this.

Surgical considerations

The calvarium does not permit removal of blocks of bone, only slices of varying thicknesses together with chippings of cortical and spongy bone. The thickness of these slices is too thin (3-5 mm maximum) to adequately reinforce the floor of the sinus. We have used cranial grafts in two ways :

- the first 22 patients (1990-1991) were treated according to the techniques current at the time which involved rotating through 90in the sinus cavity, a plate of bone, shaped from the antero-lateral wall of the maxilla and hinged at its superior margin. The sinus mucosa was progressively lifted from the floor and lateral walls and then chips of cortico-spongy bone were placed on the floor of the sinus to a height of approximately 15 mm. Several follow up scans carried out a week postoperatively showed, as one would expect, very heterogeneous appearances. The grafts were scattered and more or less in contact with each other. After 6 months, the mass of bone had considerably reduced and appeared to be dense and relatively homogeneous. In the longer term (5 years or more), scans of some of these early cases did not show any changes in the mass of intra-sinus bone (^{fig. 13}, ^13b and ^13c and ¹⁴, ^14b and ^14c) ;

- since January 1992, we have used the technique described earlier in this paper which consists of creating a cavity above the floor of the sinus into which bone particles can be packed and thus obtain a homogeneous structure. This technique takes into account two fundamental principles of bone reconstruction : immobilisation of the grafts and elimination of dead spaces.

Osseointegration of cranial grafts

Few articles have been dedicated to pre-implant reconstructions using cranial bone. The only article found in the English language literature was in connection with alveolar augmentation without opening the sinus. The proportion of 93 Brånemark implants placed in 24 patients which osseointegrated varied from 86 to 98 %, according to the technique used (^{Donovan et al., 1994}). ^{Zerbib (1992}, ¹⁹⁹⁴, ¹⁹⁹⁶) reported a failure rate of 4.3 % (22 implants lost out of 465 which were monitored) after iliac bone grafts from 1988 to 1992, whilst the cranium was the preferred source of bone for the grafts during the last three years of the study. ^{Philippe (1998)} reported 7 sinus grafts using crushed parietal bone, without specifying whether implants were eventually placed. ^{Daelemans and Malevez (1997)} noted a high percentage of osseointegration in implants placed into sinuses whether grafted with cranial or iliac bone.

An indisputable advantage of cranial bone

If the medium term stability of sinus floor reconstructions seems to be the same whether iliac or cranial bone is used, it is not the same for augmentation of the alveolus where resorption is much greater with iliac cancellous bone than with cortical bone obtained from the calverium or the mental region. In addition, the postoperative effects following harvesting from the cranium do not compare with the always painful after-effects of taking iliac bone. The only sequel to taking bone from the cranium is a palpable depression in the parietal region, which is generally invisible, unless the patient is bald. In five years, our percentage of grafts from the cranium has increased to 70 to 83 %. This source of graft can be used even in elderly patients, the density of bone and the ability to split off the external layer of bone is not in any way dependent on the age of the patient. As for the quantity of bone available it is, if not unlimited, at least amply sufficient to totally reconstruct an entire edentulous maxilla, even if harvested from only one side.

Several minor disadvantages

Even if taking the graft from the calvarium does not present any real difficulty, it is advisable to undergo training in the technique in order to avoid a number of major complications such as those described by ^{Cannella and Hopkins (1990)} and ^{Frodel et al. (1993)}. A multicentre study involving 13,000 grafts from the cranium revealed 11 cases with neurological complications (7 temporary and 4 permanent) which arose following procedures undertaken by surgeons having little or no experience of cranial or craniofacial surgery (^{Kline and Wolfe, 1995}). Before the intervention, it is always useful to assess the thickness of the cranial vault. Two or three frontal radiographs taken at various angles will provide excellent information on the thickness and density of the bone. Measurements made on 200 adult dry skulls by ^{Pensler and McCarthy (1985)}, showed a mean thickness of just over 7 mm, with a maximum value in the posterior parietal region.

Finally, the fact that cranial bone has a thick cortex could render it more resistant to revascularisation, especially in the case of large grafts. According to Tessier, there is a « critical mass » above which some of the graft cannot be revascularised.

Why separate the insertion of the implants from the bone grafting procedure ?

Even if it is technically possible to insert the implants into the residual bone and to reconstruct the alveolus with a bone graft at the same time, it is definitely preferable to begin with the reconstruction and, six months later, to carry out a CT scan in order to evaluate the amount of bone available for the implants. This way, the implants can be placed in the optimal position for the future prosthesis. Furthermore, the high density of the lamellar bone of cranial grafts makes it difficult to undertake the reconstruction and insertion of the implants at a single visit. In conclusion, the frequent need to build up the external surface of the alveolus with an appositional graft, is an additional argument for undertaking the procedure in two stages.

Conclusion

The reconstruction of the posterior maxillary segments by cranial bone grafts to the floor of the sinus seems to be a reliable procedure. The ease of harvesting bone, lack of postoperative complications, the large volume of bone available and the high density of the bone from the calvarium makes this an ideal source of bone, especially with a view to the use of implants. This technique must, in all cases, be undertaken only by surgeons with experience of cranial or cranio-facial surgery.

Acknowledgements. We express our thanks to Madam Merri Scheitlin for her excellent diagrams and to Dr Alain Lacan, radiologist who carried out the CT scans of our patients.

Demande de tirés à part

J.F. TULASNE, 26, avenue Kléber, 75016 PARIS - FRANCE.

BIBLIOGRAPHIE

• ALBREKTSSON T. Repair of bone grafts. A vital microscopic and histological investigation in the rabbit. Scand J Plast Reconstr Surg 1980;14:1.
• BOYNE PJ, JAMES RA. Grafting of the maxillary sinus floor with autogenous marrow and bone. J Oral Surg 1980;38:613.
• CANNELLA DM, HOPKINS LN. Superior sagittal sinus laceration complicating an autogenous calvarial bone graft harvest : report of a case. J Oral Maxillofac Surg 1990;48:741-743.
• DAELEMANS P, MALEVEZ C. Communication personnelle, 1997.
• DANDY WE. An operative treatment for certain cases of meningocele (or encephalocele) into the orbit. Arch Ophtalmol 1929;2:123.
• DONOVAN MG, DICKERSON NC, HELLSTEIN JW, HANSON LJ. Autologous calvarial and iliac onlay bone grafts in miniature swine. J Oral Maxillofac Surg 1993;51:898-903.
• DONOVAN MG, DICKERSON NC, HANSON LJ, GUSTAFSON RB. Maxillary and mandibular reconstruction using calvarial bone grafts and Brånemark implants : a preliminary report. J Oral Maxillofac Surg 1994;52:588-594.
• FRODEL JL, MARENTETTE LJ, QUATELA VC, WEINSTEIN GS. Calvarial bone graft harvest. techniques, considerations and morbidity. Arch Otolaryngol Head Neck Surg 1993;119:17-23.
• HARDESTY RA, MARSH JL. Craniofacial onlay bone grafting : a prospective evaluation of graft morphology, orientation and embryogenic origin. Plast Reconstr Surg 1990;85:5.
• KLINE RM, WOLFE SA. Complications associated with the harvesting of cranial bone grafts. Discussion by Paul Tessier. Plast Reconstr Surg 1995;95:5-20.
• KUSIAK JF, ZINS JE, WHITAKER LA. The early revascularization of membranous bone. Plast Reconstr Surg 1985;76:510.
• PENSLER J, McCARTHY JG. The calvarial donor site : an anatomic study in cadavers. Plast Reconstr Surg 1985;75:648.
• PHILIPPE B. Critères de choix du site donneur osseux en chirurgie reconstructrice pré-implantaire. Cah Stomatol Chir Maxillofac 1998;1:3-14.
• SMITH JD, ABRAMSON M. Membraneous versus endochondral bone autografts. Arch Otolaryngol 1974;99:203.
• SULLIVAN WG, SZWAJKUN PR. Revascularization of cranial versus iliac crest bone grafts in the rat. Plast Reconstr Surg 1991;78:1105.
• TESSIER P. Autogenous bone grafts taken from the calvarium for facial and cranial applications. Clin Plast Surg 1982;9:531-538.
• TULASNE JF. Implants dentaires et greffes osseuses. Cah Proth 1990;71:81-102.
• TULASNE JF. Sinus grafting with calvarial bone. In : Jensen OT, ed. The sinus bone graft. Carol Stream : Quintessence, 1999:107-116.
• TULASNE JF, SAADE J, RIACHI A. Greffe osseuse du sinus maxillaire et implants de Brånemark. Implant mai 1993;101-114.
• ZINS JE, WHITAKER LA. Membraneous versus endochondral bone : implications for craniofacial reconstruction. Plast Reconstr Surg 1983;72:778.
• ZERBIB R. Implants et secteurs sous-sinusiens. Inform Dent 1992;3:2873-2881.
• ZERBIB R. Reconstruction maxillaire en chirurgie pré-implantaire. J Parodontol Implant Orale 1994;13:323-332.
• ZERBIB R. Greffes osseuses autogènes en chirurgie pré-implantaire. Rev Odonto-Stomatol 1996;25:437.