Maxillary sinus vascular anatomy and itsrelation to sinus lift surgery

Gabriele RosanoSilvio TaschieriJean-Francois GaudyTommaso WeinsteinMassimo Del Fabbro

Authors’ affiliations:Gabriele Rosano, Silvio Taschieri, Tommaso Weinstein,Massimo Del Fabbro, Dental Clinic, Department ofHealth Technologies, Galeazzi Orthopaedic Institute,University of Milan, Milan, ItalyJean-Francois Gaudy, Department of CranialCervicofacial Anatomy, Faculty of Medicine,University Rene Descartes – Paris 5, Paris, France

Corresponding author:Dr Gabriele RosanoUniversita degli Studi di MilanoDipartimento di Tecnologie per la SaluteIRCCS Istituto Ortopedico GaleazziVia R. Galeazzi, 420161 – MilanoItalyTel.:þ39 02 50319950Fax:þ 39 02 50319960e-mail: gabriele.rosano@unimi.it

Key words: haemorrhage risk, maxillary sinus vascularisation, sinus lift surgery

Abstract

Objectives: To investigate the prevalence, location, size and course of the anastomosis between the

dental branch of the posterior superior alveolar artery (PSAA), known as alveolar antral artery (AAA),

and the infraorbital artery (IOA).

Material and methods: The first part of the study was performed on 30 maxillary sinuses deriving from

15 human cadaver heads. In order to visualize such anastomosis, the vascular network afferent to the

sinus was injected with liquid latex mixed with green India ink through the external carotid artery. The

second part of the study consisted of 100 CT scans from patients scheduled for sinus lift surgery.

Results: An anastomosis between the AAA and the IOA was found by dissection in the context of the

sinus anterolateral wall in 100% of cases, while a well-defined bony canal was detected

radiographically in 94 out of 200 sinuses (47% of cases).

The mean vertical distance from the lowest point of this bony canal to the alveolar crest was

11.25 � 2.99 mm (SD) in maxillae examined by CT. The canal diameter was o1 mm in 55.3% of cases,

1–2 mm in 40.4% of cases and 2–3 mm in 4.3% of cases.

In 100% of cases, the AAA was found to be partially intra-osseous, that is between the Schneiderian

membrane and the lateral bony wall of the sinus, in the area selected for sinus antrostomy.

Conclusions: A sound knowledge of the maxillary sinus vascular anatomy and its careful analysis by CT

scan is essential to prevent complications during surgical interventions involving this region.

Sinus augmentation using autogenous bone or

bone substitutes is a safe procedure with high

predictability (Wallace & Froum 2003; Aghaloo

& Moy 2007; Del Fabbro et al. 2008; Pjetursson

et al. 2008) for the rehabilitation of severely

atrophic posterior maxillae.

However, given the extensiveness of the max-

illary vascular network, it is not infrequent to run

into vascular complications that may compro-

mise the outcome of surgery. For example, severe

haemorrhage may occur as a result of arterial

injury (Chanavaz 1996).

A sound knowledge of the arterial supply of the

maxillary sinus is mandatory for surgical proce-

dures involving this area, such as sinus floor

elevation and implantation of grafting materials.

The vascularization of the antero-lateral wall of

the sinus, which is involved in sinus lift surgery

when the lateral approach is carried out, is

characterized by the presence of an intra-osseous

anastomosis between the dental branch of the

posterior superior alveolar artery (PSAA), also

known as alveolar antral artery (AAA) (Gaudy

2003; Rosano et al. 2009), and the infraorbital

artery (IOA).

Such anastomosis, although radiographically

evident in almost 50% of cases (Elian et al.

2005; Mardinger et al. 2007), courses intra-os-

seously halfway up the lateral sinus wall and is

reported in the width of the cortical bone of the

lateral wall of the maxillary sinus in 100% of

cases (Solar et al. 1999; Traxler et al. 1999;

Rosano et al. 2009).

The AAA, whose reported diameter is up to

2.5–3 mm (Mardinger et al. 2007; Ella et al.

2008), supplies the sinus membrane and the

antero-lateral wall of the sinus, and as a

consequence, has the potential to cause bleeding

complications during lateral window osteo-

tomies.

Even if the transection of such artery is not life

threatening because its haemorrhage mostly re-

solves itself owing to a reactive contraction

(Rosano et al. 2009), impairment in visualization

of the Schneiderian membrane may occur,

especially when the AAA diameter is relevant,

Date:Accepted 7 July 2010

To cite this article:Rosano G, Taschieri S, Gaudy J-F, Weinstein T, Del FabbroM. Maxillary sinus vascular anatomy and its relation to sinuslift surgery.Clin. Oral Impl. Res. 22, 2011; 711–715.doi: 10.1111/j.1600-0501.2010.02045.x

c� 2010 John Wiley & Sons A/S 711

making its elevation far more difficult and inter-

fering with placement of the graft material.

In such a context, the purpose of this cadaveric

and CT scan study was to investigate the pre-

valence, location, size and course of the AAA

located on the anterior lateral wall of the max-

illary sinus, so as to provide indications for

improving the safety of sinus floor elevation

procedure, especially in cases of extreme atrophy

of the alveolar process.

Material and methods

The first part of the study was performed on 30

maxillary sinuses, deriving from 15 human

cadaver heads. The specimens belonged to sub-

jects with an age range of 59–90 years (mean age

76 years) and equal sex distribution, who had

donated their body for research purpose. The

study obtained ethical approval from the Depart-

ment of Anatomy at the Faculty of Medicine

Rene Descartes of Paris 5 (Paris 5 University,

Paris). Direct visualization of the AAA was

obtained by fenestrating the anterior lateral

wall of the sinus cavity and its dissection was

carried out as far as the IOA and the PSAA were

visible at its extremities (Fig. 1), in order to

determine its course with respect to both the

Schneiderian membrane and the buccal antral

wall.

To detect such an artery, the vascular network

afferent to the sinus was injected with liquid latex

mixed with green India ink through the external

carotid artery.

The second part of the study consisted of 100

CT scans from 100 patients scheduled for sinus

lift surgery at the Dental Clinic of the IRCCS

Istituto Ortopedico Galeazzi, Universita degli

Studi di Milano. The age range was 29–78

(mean: 53.5) years.

The CT scans were performed using a 2000

SOMATOM Volume Zoom 4 slice CT scanner

(Siemens AG, Medical Solutions, Forchheim,

Germany) with slices of 0.5 mm thickness. CT

images were investigated for the presence of a

bony canal, housing the AAA, in the context of

the sinus anterolateral wall.

Coronal, axial and sagittal views of the max-

illary sinus were obtained by means of a software

for 3D reconstruction (OneScan 3D, 3D-MED

s.r.l., Brescia, Italy), offering a photorealistic

rendering quality and able to import Dicom

formatted CT images.

The route of the AAA was assessed with

respect to the Schneiderian membrane and to

the bony wall, as well as its diameter and the

distance from its lowest point to the alveolar

crest, with a precision of 0.1 mm. The corre-

sponding ridge height was measured. The corre-

lation between the residual ridge height and the

distance between the AAA and the crest was also

analysed.

Only edentulous or partially edentulous max-

illae displaying Class V or VI resorption of

the alveolar process, according to Cawood &

Howell’s classification (1988), were taken into

consideration.

Results

The anatomical dissection confirmed that the

PSAA divides into two branches along its course:

an external (gingival) branch is directed towards

the superior buccal fornix and the maxillary

tuberosity; the other branch is internal (dental)

and, after coursing below the zygomatic process,

was found to point towards the inside of the

orbit making a circular anastomosis with the

IOA.

An intra-osseous anastomosis between the

AAA and the IOA was found by dissection in

100% of the anatomical cases (30/30 sinuses),

while a well-defined bony canal, located in the

context of the sinus anterolateral wall, was de-

tected radiographically in 94 out of 200 sinuses

examined (47% of cases).

The diameter of such bony canal was o1 mm

in 52 sinuses (55.3% of 94 cases), 1–o2 mm in

38 sinuses (40.4%) and � 2 mm in four sinuses

(4.3%).

The AAA displayed three different courses: (1)

within the buccal antral wall cortex; (2) between

the Schneiderian membrane and the lateral bony

wall of the sinus, in which a small concavity was

often visible (Figs 2 and 3); and (3) under the

periosteum of the sinus lateral wall.

In particular, the AAA course was found to be

(1) completely intra-osseous at its extremities in

100% of cases (Fig. 4); (2) partially intra-osseous

in the area usually involved with sinus antrost-

omy (from second premolar to second molar) in

100% of cases (Fig. 4). In such an area, the AAA

was strictly close to the Schneiderian membrane

and partially encased in the lateral sinus wall

in all specimens. No bony layer interposed

between the AAA and the sinus membrane

could be identified by dissection (Figs 1 and 5);

and (3) variable (either intra-osseous or intra-

Fig. 1. View of the anterolateral wall of the maxillary sinus by transillumination: the alveolar antral artery dissection is carried

out in the area selected for sinus antrostomy as far as the infraorbital artery (a) and the posterior superior alveolar artery (b) are

visible respectively at its medial and distal extremity. The bony vessel is strictly stick to the Schneiderian membrane.

Fig. 2. Computed tomography scan 3D view of the lateral wall of the maxillary sinus which shows the point of emergence of

the infraorbital artery (IOA) (1), the point of anastomosis between the IOA and the alveolar antral artery (AAA) (2) as well as

the route of the AAA (3) forming a small concavity (white arrow).

Rosano et al �Haemorrhage risk during sinus surgery

712 | Clin. Oral Impl. Res. 22, 2011 / 711–715 c� 2010 John Wiley & Sons A/S

sinusal or sub-periosteal) in the maxillary tuber-

osity area.

The vertical distance from the lowest point of

the vessel, corresponding to the first molar area,

to the alveolar crest averaged 11.25 � 2.99 (SD)

mm (range between 7.2 and 17.7 mm).

The residual ridge height ranged from 0.7 to

5.1 mm (mean height 3.60 � 1.28 mm). A slight

positive correlation between such a distance and

the ridge height was observed (r¼0.38). When

considering a threshold of 3 mm for the residual

ridge height, the AAA-to-alveolar crest distance

averaged 9.33 � 2.41 (n¼39) and 12.45� 2.71

(n¼55) for cases with ridge height o3 mm and

� 3 mm, respectively.

Discussion

The anastomosis between PSAA and IOA pro-

vides blood supply to the sinus membrane, to the

periosteal tissues, and especially, to the antero-

lateral wall of the sinus (Solar et al. 1999; Rosano

et al. 2009).

The scientific literature reports that this vessel

is located at an average distance of 19 mm (Solar

et al. 1999; Traxler et al. 1999), 16.4 mm (Elian

et al. 2005) and 16.9 mm (Mardinger et al. 2007)

from the alveolar crest of the posterior maxilla.

Nevertheless, such data can be misleading

because the height of the residual bony ridge,

the maxillary atrophy class and the presence of

teeth play a relevant role in determining the

location of the vessel.

In the present study, the average distance of the

AAA from the alveolar ridge in atrophic maxillae

of Cawood & Howell class V and VI was

11.25 mm. For the most atrophic cases, in which

the ridge height is inferior to 3 mm, such a

distance was significantly lower with respect to

lesser atrophic cases. This would confirm that

the more resorbed the bone crest, the higher the

risk of violation of such a vessel during sinus

augmentation procedure.

These results are substantially in agreement

with the study by Mardinger et al. (2007), which

found that this vessel was located at a mean

distance of 10.9 mm from the crest in classes

D, E (Lekholm & Zarb 1985) and at a distance

greater than 15 mm in classes A, B and C.

Differences concerning the mean distance from

the vessel to the crest, with the studies by Solar

et al. (1999), Traxler et al. (1999) and Elian et al.

(2005) are probably due to the more strict inclu-

sion criteria considered in the present study,

where only highly atrophic ridges have been

examined.

Moreover, because a well-distinguished bony

wall between the intra-osseous maxillary anasto-

mosis and the maxillary sinus has never been

found by anatomic dissection (Fig. 1), it could be

speculated that the lowest border of such a vessel

Fig. 3. Internal view of the maxillary sinus: the arrow A shows the alveolar antral artery, the endosseous branch of the

posterior superior alveolar artery (PSAA), partially encased in the lateral sinus wall, while the arrow B shows the infraorbital

artery deriving from the maxillary artery and forming a vascular arcade with the PSAA.

Fig. 4. Computed tomography scan transversal views of the anterior lateral wall of a sinus where it is possible to evidence the course of the alveolar antral artery from the infraorbital artery (1)

to the posterior superior alveolar artery (2): completely intra-osseous at its extremities, between the Schneiderian membrane and the bony wall in the sinus antrostomy area, sub-periosteal in

the maxillary tuberosity area.

Rosano et al �Haemorrhage risk during sinus surgery

c� 2010 John Wiley & Sons A/S 713 | Clin. Oral Impl. Res. 22, 2011 / 711–715

could often be completely adherent to the sinus

membrane (that means not radiographically visi-

ble) instead of being located inside the buccal wall

cortex.

This would justify the contradiction between a

100% prevalence of this artery found by dissec-

tion and an only 47% prevalence detected by CT

scan in the present study.

The authors’ opinion is that such contradiction

may not depend on the AAA small diameter,

which makes it radiographically undetectable in

some cases, as suggested by Elian et al. (2005)

and Mardinger et al. (2007), but on an entirely

intra-sinusal location of the vessel.

The course of the AAA, as identified in this

study, is in agreement with the CT study by Ella

et al. (2008) where intra-osseous, intra-sinusal

and sub-periosteal courses of this artery were

detected.

As stated by Ella et al. (2008), a ‘‘superficial’’

location of such anastomosis, which is under

the periosteum of the sinus lateral wall

should also be considered. In the present study,

such sub-periosteal course was identified by

means of both CT scan analysis and anatomical

dissection in the maxillary tuberosity area but

never in the area usually selected for sinus

antrostomy.

When carrying out sinus lift surgery, the bony

window height should be almost 13 mm from the

ridge if the purpose is to place 11–13 mm dental

implants. Thus, in patients with severely

atrophic posterior maxillae (classes V, VI), the

possibility of lacerating the AAA must be con-

sidered, especially when the residual ridge is

o3 mm high.

The diameter of the anastomosis was

� 2 mm in a very few cases (3.3% by dissection

and 2% by CT scan); anyway, this eventuality,

even if infrequent, is worthy to be taken into

serious consideration.

As a matter of fact, if the damage of a bony

vessel o2 mm can be barely relevant under a

clinical point of view, the transection of an AAA

with a diameter over 2 mm is likely to produce

bleeding and impairment of vision, which may

lead to a potential membrane perforation, thus

prolonging the overall operation time, interfering

with the placement of bone graft and constituting

a true surgical complication.

In addition, the haemorrage from this artery (a)

may displace the grafting material due to a

‘‘washing’’ effect caused by the blood pressure,

thus reducing or compromising the filling of the

space below the Schneiderian membrane after

sinus floor elevation, and (b) may produce rele-

vant haematomas of the cheek area causing

discomfort to patients and creating an ideal ‘‘pa-

bulum’’ for bacteria growth and consequent in-

fection.

It is the authors’ opinion that the excision of a

large diameter AAA in combination with the

inadvertent tearing of the sinus membrane has

the potential to induce sinus mucosa swelling,

extrusion of blood into the sinus cavity as

well as a postoperative sinusitis as a major

drawback.

In fact, if the maxillary sinus is, even partly,

filled up by mucosal oedema, haematoma or

seroma, a delay of maxillary sinus clearance

may occur because of the reduction of maxillary

ostium patency, and maxillary sinusitis may

develop as well, compromising the success of

the grafting procedure (Timmenga et al. 2003).

The preservation of such anastomosis is im-

portant not only to avoid bleeding complications

but also to support bone graft neoangiogenesis

(Taschieri & Rosano 2010); in this perspective,

its concomitant reflection with the Schneiderian

membrane during sinus augmentation proce-

dures, if possible and especially when its dia-

meter is consistent, should be seriously

considered.

In conclusion, the authors recommend to rely

upon CT scan imaging, which has been proved to

be the most appropriate radiographic method for

detecting any anatomical variation within the

sinus (Schwarz et al. 1987; Quirynen et al.

1990; Alder et al. 1995; Dula et al. 2001), before

sinus lift surgery is performed, in order to pre-

surgically evaluate the location, size and thus the

clinical relevance of this anastomotic vessel.

Extreme caution should be taken when the re-

sidual ridge height is o3 mm.

References

Aghaloo, T.L. & Moy, P.K. (2007) Which hard tissue

augmentation techniques are the most successful in

furnishing bony support for implant placement? The

International Journal of Oral & Maxillofacial Im-

plants 22 (Suppl.): 49–70.

Alder, M.E., Deahl, S.T. & Matteson, S.R. (1995)

Clinical usefulness of two dimensional refor-

matted and three dimensionally rendered computer-

ized tomographic images: literature review and

a survey of surgeons’opinions. Journal of Oral and

Maxillofacial Surgery 53: 375–386.

Cawood, J.I. & Howell, R.A. (1988) A classification of

the edentulous jaws. The International Journal of

Oral and Maxillofacial Surgery 17: 232–236.

Chanavaz, M. (1996) Sinus grafting related to implan-

tology. Statistical analysis of 15 years of surgical

experience (1979–1994). Journal of Oral Implantol-

ogy 22: 119–130.

Del Fabbro, M., Rosano, G. & Taschieri, S. (2008)

Implant survival rates after maxillary sinus aug-

mentation. European Journal of Oral Sciences 116:

497–506.

Dula, K., Mini, R., Van der Stelt, P.F. & Buser, D.

(2001) The radiographic assessment of implant pa-

tients: decision making criteria. The International

Journal of Oral & Maxillofacial Implants 16: 80–89.

Elian, N., Wallace, S., Cho, S.C., Jalbout, Z.N. &

Froum, S. (2005) Distribution of the maxillary artery

as it relates to sinus floor augmentation. The Inter-

national Journal of Oral & Maxillofacial Implants

20: 784–787.

Ella, B., Sedarat, C., Da Costa Noble, R., Normand, E.,

Lauverjat, Y., Siberchicot, F., Caix, P. & Zwetyenga,

Fig. 5. Macro-anatomical dissection of the sinus lateral wall: the alveolar antral artery is close to the Schneiderian membrane

and no bony layer between such vessel and the membrane is visible after antrostomy.

Rosano et al �Haemorrhage risk during sinus surgery

714 | Clin. Oral Impl. Res. 22, 2011 / 711–715 c� 2010 John Wiley & Sons A/S

N. (2008) Vascular connections of the lateral wall of

the sinus: surgical effect in sinus augmentation. The

International Journal of Oral & Maxillofacial Im-

plants 23: 1047–1052.

Gaudy, J.-F. (2003) Anatomie Clinique. Rueil-Malmai-

son Cedex: Groupe Liaisons, Editions CdP, 11pp.

Lekholm, U. & Zarb, G.A. (1985) Patient selection.

In: Branemark, P.-I., Zarb, G.A. & Albrektsson, T.,

eds. Tissue Integrated Prosthesis. Osseointegration

in Clinical Dentistry, 199–209. Chicago: Quintes-

sence.

Mardinger, O., Abba, M., Hirshberg, A. & Schwartz-

Arad, D. (2007) Prevalence, diameter and course of

the maxillary intraosseous vascular canal with rela-

tion to sinus augmentation procedure: a radiographic

study. The International Journal of Oral and Max-

illofacial Surgery 36: 735–738.

Pjetursson, B.E., Tan, W.C., Zwahlen, M. & Lang, N.P.

(2008) A systematic review of the success of sinus

floor elevation and survival of implants inserted in

combination with sinus floor elevation. Journal of

Clinical Periodontology 35: 216–240.

Quirynen, M., Lamoral, Y., Dekeyser, C., Peene, P.,

van Steenberghe, D., Bonte, J. & Baert, AL. (1990)

The CT scan standard reconstruction technique for

reliable jaw bone volume determination. The Inter-

national Journal of Oral & Maxillofacial Implants 5:

384–389.

Rosano, G., Taschieri, S., Gaudy, J.-F. & Del Fabbro, M.

(2009) Maxillary sinus vascularization: a cada-

veric study. Journal of Craniofacial Surgery 20:

940–943.

Schwarz, M.S., Rothman, S.L.G., Rhodes, M.L. &

Chafetz, N. (1987) Computed tomography: part II.

Preoperative assessment of the maxilla for endosseous

implant surgery. Journal of Oral & Maxillofacial

Implants 2: 143–148.

Solar, P., Geyerhofer, U., Traxler, H., Windisch, A.,

Ulm, C. & Watzek, G. (1999) Blood supply to the

maxillary sinus relevant to sinus floor elevation

procedures. Clinical Oral Implants Research 10:

34–44.

Taschieri, S. & Rosano, G. (2010) Management of the

alveolar antral artery during sinus floor augmentation

procedures. The International Journal of Oral and

Maxillofacial Surgery 68: 230.

Timmenga, N.M., Raghoebar, G.M., Liem, R.S.B., van

Weissenbruch, R., Manson, W.L. & Vissink, A.

(2003) Effects of maxillary sinus floor elevation sur-

gery on maxillary sinus physiology. European Journal

of Oral Sciences 111: 189–197.

Traxler, H., Windisch, A., Geyerhofer, U., Surd, R.,

Solar, P. & Firbas, W. (1999) Arterial blood supply

of the maxillary sinus. Clinical Anatomy 12: 417–

421.

Wallace, S.S. & Froum, S.J. (2003) Effect of maxillary

sinus augmentation on the survival of endosseous

dental implants. A systematic review. Annals of

Periodontology 8: 328–343.

Rosano et al �Haemorrhage risk during sinus surgery

c� 2010 John Wiley & Sons A/S 715 | Clin. Oral Impl. Res. 22, 2011 / 711–715