ADI Masterclass 2006: “Evidence-based decision making and practical surgical management for sinus augmentation surgery”
Stephen S Wallace DDS
It had been a number of years since Stephen Wallace last enthused on sinus augmentation in the UK. I remember his last lecture at the Eastman and how it essentially influenced my technique since. I was eager to see what had changed.
Stephen Wallace works with Drs Tarnow and Froum at New York University, running a postgraduate implant programme taking on 50 residents a year. Their sinus database is one of the biggest worldwide. This, plus the fact that Professor Wallace has conducted a systematic review of over 1200 published papers on the survival of implants in sinus bone grafts was obviously sufficient appetite-whetting for a packed theatre at The Royal College of Surgeons for his masterclass.
Completely unphased by an audiovisual glitch that saw him initially computerless for 15 minutes, Professor Wallace proceeded to start his presentation completely without notes and ran through an encyclopaedic knowledge base, explaining the first part of his presentation simply in words and numbers.
His first comment was that of not being affiliated financially with any product or company. This was singularly important in view of the following: There was no mention of the myriad products available for bone augmentation other than the use of xenograft materials only, in this case BioOss (Geistlich, Switzerland). Discussion of products such as irradiated bone, DFDBA, TCP, coral, etc were distinctly notable by their absence from the proceedings. It would appear that the literature came down heavily on the side of BioOss both in quality and quantity, and Professor Wallace, having gained most of his experience when autogenous bone was regarded as the gold standard suggested that BioOss was actually now the platinum standard. More on this later…
The second was a disclaimer on alteration of photographs, noted by a deliberately ham-fisted montage of him (not) kayaking in Rome’s Trevi fountain. However, with his “limited experience” of “only about 2,500 sinus lifts”, there really was little need for him to doctor any of his huge collection of clinical cases which were shown full screen in fine detail for the duration of the presentation.
The presentation was broken up into three parts:
Evidence-based decision making
Since Boyne’s introduction of the lateral window sinus augmentation technique, over 1200 randomised clinical trials have been published in the literature where implants placed in augmented sinuses have been followed up for one year or more after loading. The critical measurement was not so much “success” of the sinus augmentation, but of the functioning implants placed therein for one year or more. The literature showed a mean cumulative implant success of 91.8%. However this ranged from an extremely low 61.4% to 100%. The variables that seemed to make the man differences were:
The use of particulate rather than block graft in the sinus
Particulate grafts led to 10% higher implant success values than block grafts. The details of the reasons for this were not expanded upon, but rapid resorption and delay in full vascularisation of large blocks were likely causes.
Machined implants had an 11-14% greater failure rate when placed in particulate grafts. He noted that there was an average of 79% success only when machined implants were placed into blocks!
Thirdly, the placement of a barrier membrane over the sinus window led to a 5% improvement of implant success rates from 93.5% to 98.6% in three studies carried out by the New York department.
These three decisions alone made a cumulative 25-29% improvement on implant success in sinus augmentations after one year or more of loading.
Further data compared the success of implants in augmented sinuses with those placed in unaugmented posterior maxilla. The figures were eye-opening. Implants placed in unaugmented posterior maxilla had a cumulative success of 95.1%. The cumulative score for implant success in all sinus lifts in the systematic review as stated earlier was 91.8%. However, use of roughened implants raised this to 94.5% and additional use of a barrier membrane to 98.6%. This is a similar success rate to type 3 bone found in the anterior maxilla. The data implies that implants in sinus lifted sites perform better than in unaugmented normal posterior maxilla.
There were no statistically significant differences in the use of autogenous bone instead of or in addition to bone substitute materials, between different roughened surfaces by manufacturers or success beyond one year. There seemed to be a good steady state after one year such that implants surviving in function for one year in an augmented sinus tended not to show delayed or late failures outside the normal survival data.
There was insufficient data to determine differences in survival for simultaneous vs. delayed placement of implants in grafts since there were too many variables to measure, such as height and quality of remaining crestal bone, graft material, implant surface textures, smoking, PRP, loading times etc. for meaningful meta-analysis.
Professor Wallace than went on to discuss his rationale biologically for his treatment decision in using BioOss. His first comment was to dismiss the Creuzfeldt-Jakob/ spongiform encephalopathy fears by stating that a) there was no documented case yet reported in over 20 years of BioOss use and systematic follow up and b) the theoretical risk was recently computed to be in the order of 1:40 billion which is rather low.
Recent papers support the existing literature that bone density and bone implant contact (BIC) with BioOss alone are actually at least as good and usually superior to autogenous bone and mixtures of autogenous bone and BioOss. The previous gold standard of having autogenous bone in the site to provide growth factors and “seeding” simply is not supported by the data, hence the allusion to BioOss alone now being the platinum standard. After one year of healing, pure BioOss placed in the sinus reveals a reproducible histomorphometric result of approximately 25% of living bone, 25% of BioOss in direct communication with this living bone and 50% connective tissue in samples taken from the part of the graft furthest away from the patient’s existing live bone. This is better than the normal type 3-4 bone found here which is often 60-80% connective tissue. BioOss can still be found after years or even decades in situ, so it maintains it structural integrity whilst allowing remodelling and autogenous bone infiltration.
Histology has failed to ever reveal an area of direct BioOss / implant contact at any site. Autogenous living bone always intervenes between BioOss particles and the functioning integrated implant surface showing its ostoconduction ability.
BioOss is also a relatively inexpensive and freely available material and can be used in large volumes (Prof Wallace feels that 4-5g per sinus is routine!).
The particle size of the BioOss (0.25-1mm vs 1-2mm particles) was discussed. Professor Wallace unofficially quoted his friends and colleagues Valentini who always uses the smaller particle size and Testori who always uses the larger particle size and stated that he uses half and half to keep them both happy but gives him an opportunity to blame either if there is a non-ideal result! The potential problems were stated as a possible tendency to over pack the smaller particle size, leading to impenetrable areas that did not vascularise and therefore were powder-like after even a year of healing, versus the potential risk of tearing the thinner membranes with the more jagged edges of the larger particle sizes.
The overriding caveat to the use of pure BioOss was the need to WAIT for full healing. Bone grows at the rate of 1mm per month from a bone surface on average and even use of PRP or growth factor concentrates such as PDGF (Gem 21) did not appreciably speed up the quantity or quality of the bone growth. Thus, the greatest dimension of the graft has to be measured and biology respected. Professor Wallace regularly waits for 10-12 months for a sinus graft prior to placement of implants. Whilst he admitted some colleagues felt this was longer than necessary, he also admitted being conservative and safe due to his “limited experience with only 2,500 sinus elevations”! The message was rather clear: risk a shorter healing time if you dare but don’t complain when there are unnecessary failures that require retreatment.
Biologically based surgical technique
Prof Wallace suggested that the formula for flap design was simple – always keep the incision lines as distant as practicable from the intended window. For improved access, visibility and passive flap retraction during surgery, he favoured a large flap, incising towards the palatal aspect in areas where inferior margin of the window was likely to be near the crest of the alveolus. This also left margin for error where the window needed to be widened or repositioned during surgery.
In cases where there had been previous sinus access, it was recommended that a split thickness flap was created over the window, leaving a layer of connective tissue over the previously created bone window to be elevated into the sinus with the deeper attached sinus membrane.
The margins of the window should be:
The window can be:
There is no evidence to suggest that the bone of the window serves any additional benefit to bone formation if left attached to the membrane or introduced into the sinus en bloc or piecemeal. Professor Wallace suggested routinely removing and discarding the bone since keeping it may impede access to the sinus during membrane elevation. However, if access was adequate, there was also no harm in keeping the bone as long as it had not been contaminated in any way.
He felt that the hinge method which involves perforation of the superior margins of the bone plate and elevation in a superiorly-hinged postage stamp style was probably obsolete due to greater risk of membrane perforation during creation of the perforations and thereafter due to the jagged margins of bone left remaining.
His currently favoured technique was that of piezosurgery, which allows rapid and clean bone removal but with minimal risk of damage to both the membrane and any blood vessels running across the window. A paper recently accepted by IJPRD revealed a 5% perforation rate by hand and 0% by piezosurgery in a controlled study.
The lateral wall maxillary artery branch is invariably present, sometimes of significant size and often visible within the wall both on CT scan section and visually at surgery. If this artery branch is severed, it will occasionally bleed briskly making visibility and control of the surgery more difficult. Isolation and elevation of the vessel with the membrane by piezosurgical bone removal invariably leads to a cleaner surgical site. Professor Wallace admitted to using a number of techniques from adrenaline to crimping to stop this bleeding if it occurred, but found that most were essentially unsuccessful. The recommended method was having the assistant place the aspirator tip over the cut end for 20 minutes until the vessel retracted into the bone and bleeding stopped naturally. By this time the graft material was usually packed and the procedure near completion!
The most common errors in window preparation were those of positioning in relation to remaining alveolus and septa. Part of the osteotomy line is thus left on thick crestal alveolus or at the base of a septum.
In these cases, the margin can simply be repositioned in the appropriate direction, and a kidney bean shaped incision developed to accommodate the septal base in the inner concavity of the osteotomy line. The membrane will appear blue grey as the bone is thinned to the point where it can be accessed.
If the osteotomy is carried out with a No.8 diamond bur, the Kramer No 3 curette (Hu Friedy) has a shape that fits perfectly into this concavity for commencement of bone elevation off the membrane. Interestingly, a deluxe set of 48 double-ended sinus instruments did not seem to be needed….Professor Wallace used at most 3 instruments.
Containment of the particulate graft is of paramount importance in the success of this procedure. Thus, sinus membrane integrity is a primary objective. The most common reasons for membrane tearing were lack of adequate access both visually and physically, no three dimensional knowledge of the interior architecture of the sinus, septation at the level of elevation, or if the membrane was very thin which is found in healthy sinuses. There is also a greater frequency of perforation with the reduction of the angle between the buccal wall and the medial wall in the coronal plane.
CT scanning for three dimensional analysis of the sinus is mandatory at NYU and medicolegally this was stated as being the guide for standard of care. Prof Wallace did not seem to offer his patients a choice in the matter if they wanted the procedure carried out at his practice.
CT scanning would of course give information that reduced the risk of complications so that the surgeon would be able to almost virtually carry out the elevation surgery in their own minds first and not be surprised by internal anatomy encountered during the procedure.
The necessity for good access was reiterated and visual access with magnification and illumination mandatory.
Membrane elevation is carried out with appropriate instruments with a sharp leading edge for effortless membrane elevation off the bone and smooth, broad backs for lifting the membrane upwards. It is important not to push too hard in one spot but rather develop a radial, gradually progressive technique for freeing and lifting the fragile tissue.
Window design must be continually evaluated, especially mesially and inferiorly. If the window margin is too far distal for the anterior chamber to be accessed easily or too far apical for access to the floor, further bone removal at the periphery of the window must be carried out rather than continuing with blind elevation.
The most vital factor after maintaining membrane integrity is the elevation of the medial aspect of the membrane off the lateral nasal wall to at least the height of the superior aspect of the buccal window. Since the lateral wall is perforated and the distal wall is too far distal for most grafts to have access to the blood supply, most of the blood supply comes from the mesial, inferior and medial walls of the sinus. The medial wall is the most important with the posterior superior nasal artery providing much of the blood supply. Thus complete elevation of the membrane from the medial wall is mandatory for adequate blood supply and healing.
In the event that only partial elevation takes place, an air pocket or re-pneumatisation of the sinus can intervene between the graft placed against the buccal aspect and the medial wall. This process can actually also push the graft material out into the tissues through the buccal window. In lateral radiographic views, this may still appear to show a complete sinus fill but without in fact being so.
Professor Wallace would suggest that the mean volume of sinus graft material is in the order of 4.5-5.5cc supported by Uchida’s paper in JOMI in 1998. This equated to up to three 2g vials of BioOss per sinus! One could hear the calculators in the audience coming out….general discussion amongst colleagues after this presentation seemed to suggest about half of this was the norm for most of us. This one statistic was rather disquieting – had we not been elevating or packing our sinuses properly?
The most common error in packing the bone seemed to be limited access to the anterior chamber of the sinus, which can often be left unpacked due to limited visibility. Hence Professor Wallace advised making the mesial edge of the window close to the anterior wall and packing the mesial chamber first.
Upon being asked how this reduction in sinus volume affected patients with sinusitis, it appears that the literature supports sinus elevation procedure as actually reducing the severity and incidence of sinusitis, presumably by allowing more efficient drainage and reducing sinus volume. Physiologically there appears to be no effect on normal function following sinus elevation procedures.
Placement of barrier membrane
Placement of a barrier membrane ensures more complete bone formation in the sinus and access window site. Professor Wallace demonstrated this with clinical photographs and histology of trephines taken from the superior aspect of the grafts after healing. This simple membrane placement procedure has also now been shown to improve survival of the implants in sinus lifts by over 5 %! This was a significant figure. The only proviso was that the membrane should not shift during flap closure and Professor Wallace reiterated the need for adequate flap release and elevation around the access window to allow easy retraction and space to rest a membrane on at least 3mm of sound bone all around the access window without displacing it during suturing.
There appeared to be no difference in whether the membrane was a resorbable or ePTFE, though the latter should be removed at implant placement. Resorbable was recommended if concurrent implant placement was contemplated.
“Whatever is available” seemed to be good enough (within reason!). Since the suture line should be away from the window, the type of suture did not make a difference to end results as long as flap management was good and suturing technique acceptable.
“THESE ARE THE RULES – THEY CANNOT BE CHANGED” said the next slide. Nice to see a dogmatic, definitive, commanding statement occasionally rather than a “this requires further studies to determine….” kind of conclusion!
Experience-based complication management
Professor Wallace cheerfully reassured us that complications can be dealt with.
He started with the most common complication, one of membrane perforation. As stated, there is a correlation between perforations and narrowing sinus floor angle in the coronal plane. Cho in 2001 showed that if this angle was greater than 60 degrees, no perforations occurred. Between 30 and 60 degrees gave a 28.6% perforation rate and under 30 degrees gave a 62.5% perforation rate – very high! The literature showed overall perforation rates of 14% to 56%.
Membrane thickness also made a difference. Membranes over 1.5mm thick tore only 16.6% of the time compared with 41% tear rate for membranes under 1.5mm thick. Thus, healthy patients with no sinus membrane thickening are at greater risk of perforations. This variable was actually more significant than presence of septation.
Septation itself was rarely a great problem unless undiagnosed and unexpected and approached incorrectly. Velasquez-Plata in JOMI in 2002 showed that most septation was partial and mainly related to the medial wall. Complete division of a sinus into two distinct compartments by a septum was very rare. Three dimensional analysis pre-operatively would allow an approach to septae that would enable the operator to elevate the membrane from either side till the crest was reached without perforation. Access and visibility and good window design were important in the correct approach.
The use of various techniques such as further elevation, relocation of the window, use of barrier membranes with combinations of intricate folding, tacking and suturing procedures, PRP, fibrin glue and collagen tape were demonstrated in a series of ever more hair-raising and serious perforations. Some, it must be said, were probably more risky than most of us would be prepared to undertake in our practices but for the purposes of teaching, Professor Wallace seems to have presided over his fair share of seemingly hopeless salvage operations with very acceptable results.
However, the main message of “be careful and don’t perforate” was loud and clear and most practitioners present fidgeted somewhat at his use of 30x40mm BioGide or BioMend membrane to essentially act as a pseudo-sinus membrane following a virtual denudation-perforation of the membrane proper. I’m sure it was even more difficult than it looked but it did work in his hands!
As far as antimicrobial use was concerned, Augmentin 625mg or Azithromycin (Zitromax) seemed to be Professor Wallace’s choice for both prophylaxis and treatment of infections. Prolonged infection was rare and required debridement of the sinus but most minor infections associated with flap exposure or delayed healing responded to one of these after 5-7 days. Clindamycin seemed to be particularly not useful and he did not recommend its use in sinus elevation.
He also stated that even in cases where membrane thickening appeared to be so thick as to risk blocking drainage in the event it was elevated, a course of Augmentin would usually reduce the thickening to acceptable levels provided the aetiological factors had been removed. He stated that it was very rare for membrane removal to be required by ENT surgeons prior to a sinus lift simply for reasons of chronic hyperplasia of the membrane.
The afternoon was a very authoritative summary of what must be a huge amount of clinical experience and academic knowledge and Professor Wallace must be thanked for apparently having given us a definitive “How To Sinus Lift” – very refreshing in an age where a multitude of publications tend sometimes to confuse rather than solve our day to day clinical practice dilemmas.
I would be surprised if anyone who saw his lecture would elect to carry out future sinus lift procedures any other way.