This systematic review aimed to identify clinical studies on the short-term and long-term survival of implants placed in the pterygoid region. A structured literature search was conducted using PubMed, Scopus and Cochrane databases. Relevant studies were selected according to predetermined inclusion and exclusion criteria. Data from the final included studies could only be extracted for calculating interval survival rate (ISR) and cumulative survival rate (CSR) of implants for different time intervals. The initial database search yielded 693 titles. After filtering, 32 abstracts were selected culminating in 17 full text articles. Three additional articles were added through a hand search to obtain a total of 20 articles. Application of exclusion criteria led to elimination of 11 articles. Pooled data from the final 9 articles showed a first year ISR of 92%. The CSR over a 10 year period, largely due to data from one study was 91%. The minimum follow-up period reported in various studies was less than a year. There is insufficient data about failures that occurred beyond the first year interval, making it difficult to draw conclusions about long-term survival of these implants. More studies with longer follow-up periods involving adequate number of pterygoid implants are needed.
Placement of implants in the posterior maxilla is known to be challenging due to the quality and quantity of available bone and the presence of the maxillary sinus . In an attempt to solve these problems, pterygoid implants were introduced. They were first described by T ulasne in 1989, who credited Paul Tessier with the idea of placing implants in this region . The pterygoid implant is intended to pass through the maxillary tuberosity, pyramidal process of palatine bone and then engage the pterygoid process of the sphenoid bone .
Owing to their long path, the length of pterygoid implants ranges from 15 to 20 mm . The implant enters in the region of the former maxillary first or second molar and follows an oblique mesio-cranial direction proceeding posteriorly, towards the pyramidal process. It subsequently proceeds upwards between both wings of the pterygoid processes and finds its encroachment in the pterygoid or scaphoid fossa of the sphenoid bone . The pyramidal and pterygoid processes are composed of dense cortical bone and the average thickness of bone at their juncture is 6–6.7 mm . If an implant is passed through this juncture at an angle of 45 degrees, it can incorporate up to 8–9 mm of dense cortical bone and its apex protrudes 2 mm into the pterygoid fossa .
The primary reason for using pterygoid implants is the availability of dense cortical bone for engagement of the implant . It also helps to overcome the need for maxillary sinus lift and grafting procedures . This can shorten the treatment time and may allow immediate loading of the pterygoid implant . It allows a prosthesis to have sufficient posterior extensions, which eliminates the need for detrimental distal cantilevers . The disadvantages of the pterygoid implant are the learning curve and technique sensitivity associated with the procedure, proximity to vital anatomic structures and access difficulty for clinicians and patients . It is radiographically difficult to assess the marginal bone loss around these implants due to the nature of their position .
The literature is puzzling because of the use of terminology associated with implants placed in the pterygoid region. The terms ‘pterygoid implants’, ‘pterygomaxillary implants’ and ‘tuberosity implants’ are used interchangeably. The term pterygoid implant has been defined by the Glossary of Oral and Maxillofacial Implants (GOMI) as ‘implant placement through the maxillary tuberosity and into the pterygoid plate’ and maxillary tuberosity is defined as ‘the most distal aspect of the maxillary alveolar process’ . Authors using the term ‘pterygomaxillary implant’ probably imply that implants are placed in this complex, which involves the maxillary tuberosity, pyramidal process of palatine bone and pterygoid plates .
Significant differences exist between pterygoid and tuberosity implants ( Table 1 ). Pterygoid implants originate at the tuberosity and a major portion of its body and apex is embedded in dense cortical bone of the pterygoid plates and pyramidal process of palatine bone . The angulations of these implants ranges from 45 to 50 degrees to the maxillary plane .
|Pterygoid implant||Tuberosity implant|
|1.||Defined as ‘implant placement through the maxillary tuberosity and into the pterygoid plate’ . Originates in the tuberosity region and proceeds posteriorly and superiorly through the pyramidal process to engage the pterygoid plates .||Implants involving ‘the most distal aspect of the maxillary alveolar process’ ; may occasionally engage the pyramidal process of palatine bone if specified .|
|2.||Quality of bone surrounding the implant is dense cortical bone of the pyramidal process and the pterygoid plate .||Quality of bone surrounding the implant is predominantly cancelleous (Type 3 or Type 4) from the maxillary tuberosity .|
|3.||Angulation of the implants ranges from 45 to 50 degrees to the maxillary plane .||Angulation is predominantly vertical or depends upon the planned prosthesis.|
|4.||Vital anatomic structures involved with implant placement are internal maxillary artery, greater palatine artery, posterior superior alveolar nerve, pterygoid muscles, infratemporal fossa, pterygopalatine fossa, nasopharynx and sphenoid sinus .||Vital anatomic structures involved with implant placement are maxillary sinus and greater palatine artery.|
|5.||Technique sensitive and potentially higher risk of encroachment of vital structures .||Less technique sensitive and potential risk of encroachment to vital structures is lower.|
|6.||Visualization of the entire surgical site is not practically possible. Part of the surgery is semi-blinded .||Visualization of entire surgical site is possible.|
|7.||Radiographic assessment of marginal bone loss around implants is difficult due to nature of location .||Radiographic assessment of marginal bone loss is easier.|
|8.||Length of implants is usually long and ranges from 15 to 20 mm .||Length of implant varies.|
Tuberosity implants originate at the most distal aspect of the maxillary alveolar process and may occasionally engage the pyramidal process if specified . It is well known that the tuberosity region is predominantly composed of Type III or Type IV cancelleous bone . Also, if an implant encompasses only the tuberosity region, it does not necessarily have an angulation of 45–50 degrees like its pterygoid counterpart.
Using cadaver dissections, R eiser investigated whether implants placed in the pterygomaxillary region might be supported by the maxillary tuberosity, pyramidal process of the palate, or the pterygoid process of the sphenoid bone. He stated that when the quality and quantity of the tuberosity is adequate, an implant can be placed completely within the tuberosity; if not, implants may be angled so that the apex can engage the pyramidal or pterygoid process . Therefore, by definition, all ‘pterygoid implants’ encompass the tuberosity region but all ‘tuberosity implants’ do not necessarily engage the pterygoid plates. It is necessary to clarify these differences, as the survival outcomes of these two implants may be significantly different. This is due to differences in quality of bone, potential effects of off-axial loading and complications due to anatomic structures.
The purpose of this article is to systematically review the existing literature on implants placed in the pterygoid region, and analyse their short-term and long-term survival. Short-term survival was defined as existence of pterygoid implant in function after 1-year of implant surgery and long-term survival was defined as existence of pterygoid implant in function after 10 years of implant surgery.
Materials and methods
Two investigators searched the English language literature using PubMed, Scopus and the Cochrane Library databases. Specific terms used for search were ‘pterygoid implants’, ‘pterygomaxillary implants’, ‘tuberosity implants’ and ‘posterior maxillary implants’. Bilingual articles including English language were also included in the search. The only search limits applied were ‘humans’. The years searched were from 1974 to August 2010. Review articles were excluded but their reference list carefully screened to add potential relevant material to the electronic search. Similarly, the reference lists of included studies were also screened.
The process of selection involved 3 stages in a hierarchical order. At stage I, a list of titles was obtained from the electronic databases, and each examiner independently analysed pertinent titles based on predetermined inclusion criteria. The examiners then debated the exclusion of these titles and any disagreement was resolved by discussion. In case of doubt, the title of the article was included for consideration in the next stage. At stage II, both examiners independently screened the abstracts of all selected titles. Abstracts to be included for further scrutiny were independently selected by the two authors. Any discrepancies between the authors were discussed until a consensus was reached. When in doubt, an abstract was incorporated for the next stage of full-text analysis of articles. At stage III, both examiners carefully studied the full text of all included articles. A hand search complemented this stage by inclusion of additional full text articles from before exclusion criteria were applied. The hand search for the time span January 2009 to August 2010 was conducted for the following journals: Journal of Prosthetic Dentistry, Journal of Oral and Maxillofacial Surgery, International Journal of Prosthodontics, International Journal of Oral & Maxillofacial Implants, Clinical Implant Dentistry & Related Research and Clinical Oral Implants & Related Research. The inclusion of additional full text articles was based on related articles and references cited by other articles that were reviewed. The inclusion criteria were: English language article in a peer-reviewed journal; and any clinical study of implants on humans involving the ‘pterygoid’, ‘pterygomaxillary’ ‘tuberosity’ or ‘posterior maxillary’ region. The exclusion criteria were: articles that did not pertain to implants in the pterygoid or tuberosity region; review/technique articles without associated clinical trial and data; case reports, series or studies with less than 5 patients with endosseous implants in the pterygoid region; patients or data repeated in other included articles; articles describing implants that were not endosseous in nature; article description that did not allow extraction of the required data; and articles that did not clarify whether tuberosity implants engaged the pterygoid plates as defined by GOMI .
The main outcome of interest was the short-term and long-term survival of implants placed in the pterygoid region. The description of each included article was analysed to ascertain whether the description of implants in each study conformed to the definition of a pterygoid implant . Thereafter, data were extracted from each article and classified into qualitative or quantitative forms. Qualitative data included information on the nature of the study (prospective or retrospective), study setting (university or private office), surgical complications, implant manufacturer, surface type of implant, partial or completely edentulous patients, nature of the prosthesis used (fixed or removable) and nature of loading (immediate, early or delayed). Quantitative data extracted were number of patients in each study who received pterygoid implants, range of follow-up, total number of pterygoid implants placed, number of failures before loading and number of failures after loading. Thereafter, data pertaining to number of implants in each time interval, and number of implant failures in a respective time interval were extracted. These data were used for statistical analysis.
Inter-reviewer agreement at stages I (titles) and II (abstracts) of article selection was determined using Cohen’s kappa coefficients. Quantitative data extracted from the final 9 included studies was used for calculation of interval survival rate (ISR) during each follow-up period and cumulative survival rate (CSR) over a 10-year period. ISR is an entity that represents the proportion of surviving items in a group, during a specific time interval only . CSR represents the proportion of items existing at the beginning of a time interval that survive till the end of the interval being studied . A life table survival analysis for the total number of implants surviving over a 10-year period was performed to study these two elements.
The electronic search from the 3 databases yielded 693 titles. Of these, only 32 titles were relevant to the study, which culminated in a total of 17 full text articles after application of the predetermined exclusion criteria. The inter-reviewer agreement for the titles (kappa: 0.85) and the abstracts (kappa: 0.93) were deemed ‘almost perfect’ according to Cohen’s kappa norm ( Table 2 ). Three additional full text articles were included from a hand search , resulting in a total of 20 full text articles that were investigated. Application of exclusion criteria eventually resulted in 9 articles that were considered for final analysis ( Fig. 1 ). The abstracts and full text articles that were excluded are shown in Tables 3 and 4 with the reason for exclusion.
|Criterion||Kappa||Standard error||95% CI interval|
|Reason for exclusion||Year|
|Article did not pertain to implants in the pterygoid or tuberosity region.|
|1. Block et al.||2009|
|2. Pi-Urgell et al.||2008|
|4. Galan et al.||2007|
|5. R osen and G ynther||2007|
|6. B edrossian et al.||2006|
|7. F arzad et al.||2006|
|8. E sposito et al.||2005|
|9. C alandriello et al.||2005|
|10. A paricio et al.||2001|
|11. L ee et al.||2001|
|12. D rago||1992|
|Review/technique articles without associated clinical trial and data.|
|13. B alshi et al.||2006|
|14. S orni et al.||2005|
|Less than 5 patients with endosseous implants in the pterygoid region.|
|15. P enarrocha et al.||2005|
|Reason for exclusion||Year|
|Patients or data being repeated in other included articles.|
|B alshi et al.||2005|
|F ernandez Valeron et al.||1997|
|Articles describing implants that were not endosseous in nature.|
|Article description that did not allow extraction of the required data.|
|P ark and C ho||2010|
|R oumanas et al.||1997|
|K hayat and N ader||1994|
|Articles that did not clarify whether tuberosity implants engaged the pterygoid plates as defined by GOMI.|
|R idell et al.||2009|
|A gliardi et al.||2008|
There were 3 prospective and 6 retrospective studies and most studies (6/9) were carried out in a private office setting ( Table 5 ). Four studies reported some form of complication related to the surgical procedure. The complications reported were slight venous bleeding , minor trismus , misplacement of implant and a unique case of continuous episode of pain and discomfort . Most studies (7/9) reported the use of implants from the same manufacturer (Nobel Biocare – Branemark System). Five studies used rough surface implants, two studies used machined surface implants and the remaining two studies did not report the surface type of the implant. Three studies reported use of pterygoid implants in both partially and completely edentulous patients , three studies were exclusive to edentulous patients and two studies were exclusive to partially edentulous patients . All but one study reported the use of pterygoid implants for fixed prosthesis. Only one study reported the use of these implants for removable prosthesis as well . One study did not report the nature of patients (partial or completely edentulous) nor the type of prosthesis used . Six studies exclusively followed a traditional delayed implant loading protocol in which the prosthesis was fabricated after a second stage procedure, that was performed 6 months after implant placement; 3 studies incorporated immediate as well as early loading protocols. Prosthetic complications that were reported only by one study included fractures of prosthetic anterior teeth .
|Study name||Year||Nature of the study||Setting of the study||Surgical complications reported||Implant manufacturer||Surface Type||Partial or completely edentulous patient||Type of prosthesis||Type of implant loading|
|P enarrocha et al.||2009||Retrospective||University||None||Implandent and Straumann||Rough||Both||Fixed||Delayed|
|A paricio et al.||2008||Prospective||Private||None||Nobel Biocare||Rough||Both||Fixed||Immediate and early|
|V aleron and V aleron||2007||Retrospective||Private||Slight venous bleeding and minor trismus||Nobel Biocare-Branemark||Not reported||Partially edentulous||Fixed||Delayed|
|B alshi et al.||2005||Retrospective||Private||Not reported||Nobel Biocare-Branemark||Rough||Completely edentulous||Fixed||Immediate and delayed|
|V rielinck et al.||2003||Prospective||University||Misplacement of implant||Nobel Biocare-Branemark||Rough||Completely edentulous||Fixed and removable||Delayed|
|K rekmanov||2000||Prospective||University||Misplacement of implant||Nobel Biocare-Branemark||Rough||Both||Fixed||Early and delayed|
|B alshi et al.||1999||Retrospective||Private||None||Nobel Biocare-Branemark||Machined||Completely edentulous||Fixed||Delayed|
|B alshi et al.||1995||Retrospective||Private||Continuous episodes of pain and discomfort||Nobel Biocare-Branemark||Machined||Partially edentulous||Fixed||Delayed|
|G raves||1994||Retrospective||Private||Not reported||Not reported||Not reported||Not reported||Not reported||Delayed|
The number of patients receiving pterygoid implants was reported in only 3 studies . The remaining 6 studies did not report this number, as they only described the total number of patients in their study who received various other implants . The range of follow-up varied for different studies, with the lowest being 0.1 month (3 days) and the highest being 169 months (14 years) after implant surgery . Only 3 studies reported a minimum follow-up period of all patients for at least 1 year . The follow-up period could not be extracted from two studies . In each study, different numbers of implants were followed for different time intervals. Only one study reported the number of implants that was followed-up beyond the third year interval, for up to 10 years . Most studies reported the number of implants and related failures only for the first year interval ( Fig. 2 ).
A life table survival analysis was reported only in 2 studies . Consequently, data had to be extracted by the authors, from the remaining 7 studies to be incorporated in a pooled life table survival analysis ( Table 6 ). Pooled data from all 9 studies showed a total of 79 failures over varying time periods. 70 of these failures occurred before the implants were loaded, and were reported to have occurred within the first year interval of implant placement in all 9 studies. Of the 9 post-loading failures, 4 occurred during the first year interval , 1 failure occurred during the fifth year interval and the remaining 4 failures occurred over varying time intervals, which were not reported. The lowest ISR was 92% due to 74 failures during the first year period, and the highest ISR was recorded as 100% at different time intervals.