Maxillary sinus floor augmentation: a review of current evidence on anatomical factors and a decision tree

Maxillary sinus floor augmentation using lateral window and crestal technique is considered as predictable methods to increase the residual bone height; however, this surgery is commonly complicated by Schneiderian membrane perforation, which is closely related to anatomical factors. This article aimed to assess anatomical factors on successful augmentation procedures. After review of the current evidence on sinus augmentation techniques, anatomical factors related to the stretching potential of Schneiderian membrane were assessed and a decision tree for the rational choice of surgical approaches was proposed. Schneiderian membrane perforation might occur when local tension exceeds its stretching potential, which is closely related to anatomical variations of the maxillary sinus. Choice of a surgical approach and clinical outcomes are influenced by the stretching potential of Schneiderian membrane. In addition to the residual bone height, clinicians should also consider the stretching potential affected by the membrane health condition, the contours of the maxillary sinus, and the presence of antral septa when evaluating the choice of surgical approaches and clinical outcomes.


INTRODUCTION
After tooth loss, the alveolar ridge can be compromised by atrophy and pneumatization of the maxillary sinus, which might limit the success of rehabilitation. 1,2Maxillary sinus floor augmentation (MSFA) involves Schneiderian membrane elevation to increase the residual crestal bone height (RBH) in the posterior maxilla, thereby increasing the bone volume to install dental implants in this region, including elevation through the lateral and transcrestal approaches. 37][8] The less invasive transcrestal approach first proposed by Tatum and adapted by Summers, 5,9 is a blind technique, with advantages such as a small wound, short operation time, and high patient satisfaction. 9Evidence supports the view that MSFA through both of the above approaches is effective and safe. 5,10][13][14][15][16][17] MSFA comprises the following steps: elevation of a flap, accessing the sinus cavity through a window, and Schneiderian membrane elevation above the maxillary floor and underneath the Schneider membrane to increase the alveolar bone height and create a "confined space". 18Observation of this confined space from the coronal plane shown in Fig. 1, reveals that it consists of three walls: the maxillary sinus lateral wall, the crest of the alveolar ridge, and the stretched and lifted maxillary sinus membrane.Research suggested that an average tension of 7.3 N/mm³ caused sinus membrane perforation, and the stretching of the membrane can be achieved in one dimension up to 132.6% of its original size and in two dimensions up to 124.7%. 19In spite of the predictability of MSFA techniques, the Schneiderian membrane might be perforated when the local tension exceeds its stretching potential, [20][21][22][23][24] which is also closely related to anatomical variations of this "confined space".
This article aims to: (1) Review the impact of the abovementioned three walls of the confined space over the maxillary floor on sinus augmentation techniques; and (2) Propose a decision tree on the choice of surgical approaches.

THREE WALLS OF THE CONFINED SPACE OVER THE MAXILLARY FLOOR Residual crestal bone height
The RBH is an important factor informing the choice of surgical approach.The criteria for the minimum effective implant osseointegration height of implants have been explored, along with the range of RBH to elevate the transcrestal sinus floor.
According to the academy of Osseointegration Consensus Conference on sinus grafts (1996), elevation of the MSFA can be carried out according to the category of the classification below: Category A (RBH ≥ 10 mm): classic implant procedure; Category B (RBH ≥ 7-9 mm): osteotome technique with simultaneous placement of implants ; Category C (RBH ≥ 4-6 mm): maxillary sinus elevation with lateral access and bone graft and immediate or deferred placement of implants; Category D (RBH ≥ 1-3 mm): maxillary sinus elevation with lateral access and bone graft and deferred placement of implants. 25The impact of anatomical factors and the choice of surgical approaches has been continuously updated and several studies reported that the RBH did not appear to affect osteogenesis after MSFA, 12,13,26 suggesting that more emphasis should be placed on surgical difficulty and complications, rather than osteogenesis, when considering effects of the RBH on the surgical approaches.
Although the survival rate for longer (>8 mm) implants was 0.5% higher during long-term follow-up of 16-18 months, the insertion of longer dental implants into the augmented sinus is associated with biological complications, higher morbidity, increased costs, and longer surgery, and it has been suggested that alternative treatment using shorter dental implants (≤8 mm) might be preferrable. 27,28The advantages of fewer complications and disadvantages of low survival rate have been compared and discussed in different length definitions of "short implants". 29,30or example, it is concluded that short implants (≤6 mm) are a valid option in situations of reduced bone height to avoid possible morbidity associated with augmentation procedures; however, they reveal a higher variability and lower predictability in survival rates; 31 while according to Group 1 ITI Consensus Report: for atrophic posterior maxilla, short implants (≤6 mm) are a promising alternative to sinus floor elevation, with comparable survival rate, less MBL (marginal bone loss) and post-surgery reactions. 32ubsequently, it was suggested that elevation of the transcrestal sinus floor could be extended to patients with an RBH of 4-6 mm. 335][36] Alternatively, good clinical results were observed for posterior mandibles treated using single extra-short (4 mm) implants and Pommer et al., using multifactorial analysis of the maxillae of human cadavers, reported no significant influence of RBH on the stability of the primary implant in atrophic sinus floor. 37,38but initial RBH of less than 4 mm was reported to influence the success rates of implants inserted in combination with sinus floor elevation using osteotomes. 39Although elevation of the sinus using the transcrestal window technique in a patient with residual alveolar bone in the posterior maxilla of 1-2 mm was reported recently 11,25 and new bone formation differences were non-significant for residual bone height ≤2 and >2 mm, 12 the evidence was insufficient and further long-term follow-up studies were needed and data reviewed from literature suggested a higher implant survival predictability as available residual bone increases. 40teral wall Besides the RBH, evaluation of the lateral wall thickness should also be carried out when choosing the lateral approach.The mean maxillary lateral wall thickness has been reported as (1.98 ± 1.87) mm at the first molar and (1.21 ± 1.07) mm at the second molar. 41A retrospective study reported that the overall mean lateral wall thickness was (1.59 ± 0.84) mm at 4 mm coronal to the sinus floor and (1.58 ± 0.83) mm at 6 mm. 21Also, A. Monje et al. reported mean sinus lateral wall thickness for partial edentulous atrophic maxilla was (1.71 ± 0.12) mm, and for complete edentulous atrophic maxilla, (1.57± 0.07) mm. 42irstly, membrane elevation requires good access and vision; however, a thick lateral wall can hinder instrument maneuverability during the lateral window sinus augmentation. 43Secondly, vascularization of cancellous bone is more extensive than that of cortical bone, and increased bleeding might obstruct visibility, thus prolonging surgery. 44,45Thirdly, membrane perforation is affected by maxillary sinus lateral wall thickness. 21,46At a lateral wall thickness of ≥2 mm at 4 mm coronal to the sinus floor, a perforation rate of 56.4% was observed, which decreased to 12.1% if the lateral wall thickness was ≤1 mm. 21And the alteration of the lateral approach sinus elevation technique is recommended if complications such as membrane perforation or bleeding are expected. 47Meanwhile, when accessing the antral cavity from a lateral wall of more than 2 mm, considering the vision, bleeding, and membrane perforation risk, a transcrestal approach might be a favorable alternative.

Maxillary sinus membrane
Schneiderian membrane perforation is the most frequently reported surgical complication. 48,49An intact Schneiderian membrane is crucial to maintain the postoperative osteogenic space.1][52][53][54] Perforation might occur when the local tension exceeds the stretching potential of the Schneiderian membrane (for example, a mean perforation tension of 7.3 N•mm −3 ), 19 which is closely related to membrane health and thickness, and anatomical characteristics, such as the maxillary sinus width and contours. 55nus health.Given the maxillary sinus diseases present in some patients, the application of sinus floor elevation is restricted. 25,56,57he presence of sinus diseases might affect the tensile potential of the Schneiderian membrane and increase the difficulty of surgery and the risk of postoperative complications. 58Besides, Schneiderian membrane thickened caused by inflammation might decrease elasticity and thus a reduced stretching potential.Sinusitis, polypoid (cystic) lesions, and mucosal thickening are the most frequently noted radiographical indications of sinus diseases. 59Small antral pseudocysts lacking clinical symptoms and slight flat thickening (>2 mm and <5 mm) are not considered as contraindications for osteotome sinus floor elevation surgery. 55owever, pre-existing conditions that might abrogate drainage patency must be addressed. 60hneiderian membrane thickness.An appropriate membrane thickness has an important and beneficial effects on the tensile potential of the Schneiderian membrane. 61,62he sinus membrane comprises a ciliated pseudostratified columnar epithelium and an average thickness = 0.5 mm. 63tudies of the risk factors for membrane perforation, identified that perforation was more frequent for thinner membranes. 64,65In a retrospective study reviewing the records of 551 patients, a thinner membrane was observed in patients who suffered membrane perforation compared with those that did not. 49In those that suffered perforation, the average membrane thickness was (0.84 ± 0.67) mm, whereas it was (2.65 ± 4.02) mm in the patients that did not suffer perforation. 49y contrast, a Schneiderian membrane thickened because of inflammation, such as from odontogenic infections, particularly apical infections, 63 and smoking 66 might have decreased elasticity and thus a reduced stretching potential. 67,68And thicker maxillary sinus membrane was indeed observed in smokers compared to non-smokers, 69,70 and smokers (46.2%) versus nonsmokers (23.4%) presented with at least a 10% difference in rates of perforations. 71ertain types of irritation, e.g., allergies, are associated with Fig. 1 Three walls of confined space over the maxillary floor: the lateral wall of the maxillary sinus, the alveolar ridge crest, and the stretched and lifted maxillary sinus membrane mucosal thickening. 59Park et al. reported that perforation occurrence and preoperative sinus lesions correlated significantly, 51 possibly because of the gelatinous structures of the pathologically thickened membranes. 72,73nus width.][76][77][78] Chan et al. defined sinuses as narrow (<8 mm), average (8-10 mm), and wide (>10 mm) on the basis of a mean distance of 2.3 mm from the sinus floor, and as narrow (<14 mm), average (14-17 mm), and wide (>17 mm) on the basis of a mean distance of 15 mm from the alveolar crest. 79Histological analysis indicated that a narrower maxillary sinus, a sinus floor with a tapered morphology, and a higher proportion of native mineralized bone would lead to a higher level of osteogenesis after MSFA. 13Similarly, Stacchi C al. represented newly formed bone after transcrestal sinus floor elevation was strongly influenced by sinus width and occurred consistently only in narrow sinus cavities (sinus width <12 mm). 80And graft bone resorption in elevated sinus showed a positive correlation with the sinus width. 81urgically, the chance of membrane perforation during elevation and the difficulty of surgery are increased by the presence of a maxillary sinus cavity with a narrow-tapered shape. 33,82This surgical difficulty might result from the acute angles encountered.However, the local tension increases with wider maxillary sinus floors when lifting the maxillary sinus membrane.
Sinus contours.Sinus contours have a vital function in procedures to elevate the sinus floor, [74][75][76] and special structures such as a maxillary sinus septum and tooth roots, might increase the membrane perforation risk. 24,83,84Notably, when the Schneiderian membrane is raised to the same height, its different contours affect its the stretching percentage.The maxilla sinus floor could have a complicated morphology, in which the width and contours of the sinus are closely related. 85inus contours were classified into five categories by Niu et al. in 2018: Narrow tapered, tapering, ovoid, square, and irregular.Niu et al. recommended a modified lateral sinus for a narrow-tapered sinus; both lateral and transcrestal approaches for tapering and ovoid sinuses; and for irregular sinuses, a lateral sinus with a wider window or a lateral sinus with double-window floor elevation were proposed. 33Compared with that for a U-shaped sinus, the risk of perforation is higher for an acute angled sinus because it is more difficult to angulate the instruments. 22Similarly, in a review of 407 augmentation procedures, anatomical variations, including a V-shaped sinus cavity and the presence of underwood septa, were identified as potential risk factors for membrane perforation because they limit access to the antral space and obscure the surgeon's view. 24nus septum.Membrane perforation risk and surgical difficulty during elevation might be increased by special structures. 24,83tudies have reported much higher perforation rates of MSFA in the presence of septa. 86,87bout 20-35% of maxillary sinuses contain an antral septum. 88 single septum is more commonly observed than multiple septa.Shih-Cheng et al. proposed that septa could be classified as easy (E), moderate (M), or difficult (D) according to the size, orientation, number, and location of antral septa. 86Treatment approaches for each category have been suggested; however, antral septa complicate sinus elevation surgery.
Assessment of membrane stretching potential: stretch-favorable type (SFT) and stretching-unfavorable type (SUT) Perforation might occur when the local tension exceeds the intrinsic or stretching potential of the Schneiderian membrane, which is also closely related anatomical factors, such as membrane health and the width of the maxillary sinus.The stretching potential of the Schneiderian membrane, involving the sinus width, sinus contours, sinus/membrane health, and membrane thickness, allows sinuses to be classified as the following two types (Fig. 2

):
Type A: Stretch-favorable type (SFT).An SFT occurs when the maxillary sinus/membrane is in a healthy state; the membrane thickness is within the normal range, the maxillary contours tend to be oval; and there is no special structure, such as a maxillary sinus septum.Under these conditions, the stretching potential of the maxillary sinus membrane is relatively favorable, with relatively low surgical difficulty and risk (Fig. 2a).
Type B: Stretching-unfavorable type (SUT).By contrast, an SUT presents when the maxillary sinus/membrane is in a diseased state; the maxillary membrane is too thin or thickened because of inflammation; the maxillary contour tends to be wide, sharp, or angular; or there are special structures, such as a maxillary sinus septum.Under these circumstances, the stretch potential of the maxillary sinus mucosa is relatively poor, and the surgical difficulty and risks are relatively high.Table 1 lists the effects of anatomical variations on the stretching potential of Schneiderian membrane, based on the combined consideration of the sinus width, sinus contours, sinus/membrane health, and membrane thickness, which can be used to assess surgical risk and guide surgical approaches (Fig. 2b).

DECISION TREE AND CLINICAL CASES Decision tree
Ultimately, the goal of sinus elevation is to augment the available bone height to place implants, meanwhile lowering the risk of surgery.However, the success of the procedure depends on the technique chosen and treatment indications must be strictly adhered to.The RBH, lateral wall thickness, maxillary sinus contours, and the health of the Schneiderian membrane and sinus should be assessed when considering the choice of surgical approach and clinical outcomes.After a review of the literature concerning anatomical factors, and considering clinical findings, we propose the following decision tree for choosing the optimal surgical approach (Fig. 3): 1.If the RBH exceeds 6 mm: the transcrestal approach is the more favorable alternative because it is minimally invasive, and its morbidity, duration, and cost are limited.2. If the RBH is between 4 and 6 mm, the transcrestal approach is the more favorable alternative when the sinus and membrane are in a relatively healthy state, which is the most common situations, whereas the lateral approach is preferred when the sinus and membrane are in an unhealthy state.The RBH does not appear to affect osteogenesis, indicating that surgical difficulty and complications should be considered rather than osteogenesis in this situation.The presence of sinus diseases might have an important effect on the tensile potential of Schneiderian membrane, thus increasing the difficulty of surgery and the risk of postoperative complications.3.If the RBH is between 2 and 4 mm, the lateral approach is the more favorable alternative when the sinus wall is less than 2 mm, which is the most common situations, otherwise, the transcrestal approach should be chosen.Membrane elevation requires good vision and access, and the incidence of membrane perforation correlates with the thickness of the maxillary sinus lateral wall.Thus, when accessing the antral cavity from a lateral wall more than 2 mm away, the transcrestal approach is more favorable alternative.Health state of Schneiderian membrane and sinus should also be considered when necessary.4. If the RBH is less than 2 mm, we suggest that the stretching potential of Schneiderian membrane and risk of the surgery should be assessed comprehensively (Table 1).Histologically, the smaller the width of the maxillary sinus, the higher the level of osteogenesis can be anticipated after MSFA.Surgically, during elevation, the membrane perforation risk is increased by the presence of a narrow and tapered maxillary sinus cavity; however, the wider the floor of the maxillary sinus floor, the greater local tension stretches when lifting the maxillary sinus membrane.If a patient has a healthy maxillary sinus, and the maxillary contours tends to be oval and no septa are present in the lifting region, the stretching potential of the maxillary sinus membrane is relatively favorable (Type A: SFT).For a stretch-favorable case, the transcrestal approach is the more favorable alternative.If a patient has an unhealthy maxillary sinus, or the maxillary contours tend to be wide (or too sharp) and there is a septum within the lifting region, the stretching potential of the maxillary sinus membrane is relatively unfavorable (Type B: SUT), and the surgical difficulty and risks are relatively high.For a stretch-unfavorable cases, the lateral approach is recommended.

Clinical cases
Case 1.This case comprised a 67-year-old male patient with a pseudocyst on the right maxillary sinus antral floor.Clinical examination revealed an edentulous maxilla encompassing the region from right first molar to the second molar.Assessment using pre-operative Cone Beam Computed Tomography (CBCT) revealed an atrophied edentulous ridge with an RBH < 1-2 mm (Fig. 4a).The maxillary contours tended to be oval and a homogeneous radiopaque mass without clinical symptoms was observed on the antral floor.The stretching potential of the maxillary sinus membrane was assessed as relatively favorable (SFT).Intentional perforation of the sinus membrane was carried out and a fine needle was used to aspirate the fluid to reduce the volume of the pseudocyst.Following saline irrigation, the transcrestal approach was used to elevate the sinus membrane.Subsequently, anorganic bone graft (Bio-Oss, Geistlich Pharma, Switzerland) was inserted (Fig. 4b).Post-operative CBCT showed that the sinus membrane was elevated and the bone graft material was successfully emplaced (Fig. 4c).
Case 2. This case was a 49-year-old female without sinus pathology.Clinical examination revealed an edentulous maxilla encompassing the region from the left first molar to the second molar.Assessment using pre-operative Cone Beam Computed Tomography (CBCT) revealed an atrophied edentulous ridge with an RBH < 1-2 mm (Fig. 5a).The maxillary sinus of the patient was in a healthy state, while the maxillary contours tended to be wide and there was a maxillary sinus septum in the distal part.The stretching potential of the maxillary sinus membrane was assessed as relatively unfavorable (SUT), and the surgical difficulty and risks were relatively high.The lateral wall approach was used to elevate the sinus

Sinus health condition
Health sinus Lower perforation risk 96 Sinus diseases Managed with care before sinus lift procedure 59,96 pseudocysts in a small size without clinical symptoms Not contraindications to surgery 55 Membrane thickness Appropriate membrane thickness between 1-2 mm Predictable sinus augmentation procedure 65 Thinner membrane thickness of (0.84 ± 0.67) mm Higher perforation risk compared with (2.65 ± 4.02) mm group 49 Thicken due to inflammation Significant correlation between preoperative sinus lesions and perforation 51

Sinus width and contours
Tapering or ovoid Both lateral and transcrestal approaches are recommended 33,87 V-shape, Obscured visibility and limitd access to the antral space 22,24 Irregular Higher perforation risk 87 Square Lateral approach with a wider window 33 Septa Absence of septa Lower perforation risk 86 One or multiple septa Higher perforation risk 24,86   membrane (Fig. 5b).Subsequently, the sinus cavity was compacted using an anorganic bone graft (Bio-Oss, Geistlich Pharma, Switzerland).Post-operative CBCT showed that the sinus membrane was elevated and the bone graft material was successfully emplaced (Fig. 5c).

DISCUSSION
Sinus pneumatization and ridge atrophy represent challenges to the successful rehabilitation of patients with posterior maxilla endosseous implants.Elevating the sinus comprises forming a "mucoperiosteal-alveolar pocket" above the maxillary floor and underneath the Schneiderian membrane to increase alveolar bone height and create a "confined space". 181][92] With the improvement of implant surface modification and surgical equipment, the choice of MSFA approaches has been continuously updated and whether to choose immediate deferred placement of implants with anatomical variations is still controversial.8][39] Sinus elevation through the transcrestal window approach for a patient whose posterior maxilla had only 1-2 mm of residual alveolar bone was reported recently, and the incisions used in transcrestal window approach were shorter, compared with the lateral window approach, which could reduce discomfort of the patient after sinus elevation surgery. 11,93The evidence is insufficient and further long-term followup studies were needed.Meanwhile the transcrestal window approach requires a thorough assessment of the anatomy of sinus, elasticity, and deformation capacity of the Schneiderian membrane, the location of the intraosseous artery(which could be undetectable in CT/CBCT images), precise surgical approach, and so on. 93In addition, the crestal approach was used to elevate the sinus floor of 27 patients with residual bone heights of 2-3 mm. 94Moreover, a recent study revealed that the RBH and vital bone formation were not statistically correlated. 13No significant differences in the amount of osteogenesis in sinuses classified as having an RBH ≤ 2 mm or >2 mm were observed. 13When choosing the surgical technique, clinicians should assess the lateral wall thickness.A difference in the perforation rate was noted for a wall thickness measured at 6 mm coronal to the sinus floor. 21While other researchers also reported that lateral wall thickness had no effect on the perforation rate. 95][100] Sinus compliance comprises the intrinsic potential of the sinus mucosa to resume its homeostatic status after the surgical trauma caused by sinus lifting. 18A higher rate of perforation is associated with a thinner membrane, possibly because the tensile capacity of a thicker membrane is significantly higher. 19Sinus augmentation surgery can be carried out on a 1-2 mm thick healthy and resilient membrane; however, for a thin membrane (<1 mm), a more cautious approach should be adopted. 65The Schneiderian membrane has the potential to thicken during inflammation, such as during odontogenic infections, especially apical infections. 63rritations, such as allergies, can also thicken the mucosa. 59However, sinus augmentation is not contraindicated by the presence of mild mucosal thickening or pseudocysts in the absence of coexisting sinonasal symptoms. 101However, with a deeper understanding of the maxillary sinus disease, some researchers formed different opinions, 23 and diseases of the maxillary sinus should be diagnosed and managed carefully prior to sinus elevating surgery. 96he stretching potential of Schneider's membrane should be considered surgically and histologically. 13,33After lateral sinus floor elevation surgery, transient swelling of sinus membrane is observed, which reaching a peak value 7 days after surgery and completely resolves over months. 62The widths and contours of the sinus are closely related.For example, Wang et al. described the palate-nasalrecess (PNR) as the intersection point of the two imaginary lines following the lower part of the lateral nasal wall and the palatal wall in the maxillary sinus, 102 which would make elevation of the sinus membrane more complicated and enhance the risk of membrane perforation.Niu et al. considered the sinus width, sinus depth, the PNR, and angle A simultaneously. 33A flat or irregular maxillary sinus stretches more when lifted to the same height, which requires a better stretch potential of the Schneider membrane.For example, the presence of irregular bone walls, exostosis, and septa might contribute to the occurrence of sinus perforations. 87Perforation rates of MSFA when septa were present were much higher than in those without septa. 86,87However, at 6-9 months post-surgery, wider sinuses added with deproteinized bovine bone mineral (DBBM) alone showed a lower percentage of newly formed bone, while the sinus contours and the RBH and sinus contours did not appear to influence post-MSFA osteogenesis. 13NCLUSION Anatomical factors, including the remaining alveolar bone, maxillary sinus characteristics, the health condition of the Schneiderian membrane, and the lateral wall thickness, crucially affect successful lifting.The stretching potential affected by maxillary sinus characteristics and the health condition of Schneiderian membrane/sinus, as well as the residual bone height, influence clinical outcomes and the choice of surgical techniques, which should be taken into account by clinicians.

Fig. 3
Fig. 3 Decision tree and CBCT classification: a A decision tree for choice of surgical approaches; b Relevant 3D CBCT Classification a b c

Fig. 4 Fig. 5
Fig. 4 Case 1: a Pre-operative CBCT assessment, b surgical procedure, and c post-operative CBCT assessment of patient in case 1

Table 1 .
Impact of anatomical variations on the stretching potential of the Schneiderian membrane