The focused question was: What are the differences among techniques presented for simultaneous closed sinus floor elevation with implant placement in fresh extraction sockets?

We included all human clinical studies of the immediate placement of implants in fresh extraction sockets in the posterior maxilla with sinus floor elevation via the same extraction sockets. The search was limited to English-language studies. Abstracts, letters, and reviews were excluded.

All patients who needed the extraction of a tooth/teeth in the posterior maxillary region, sinus floor elevation through the extraction socket(s), and immediate implant placement in the fresh socket(s) were included.

Studies were included if they mentioned simultaneous closed sinus floor elevation with implant placement in fresh extraction sockets in the posterior maxillary region. Studies that did not perform sinus floor elevation through a fresh socket in the posterior maxilla or did not place implants immediately were excluded.

The primary outcome was the survival/success rate of the installed implants with a minimum of six months of follow-up. The survival rate represents the persistence of implants, with no implant removal during follow-up, whereas the success rate represents criteria such as soft/hard tissue level, probing pocket depth, and bleeding on probing²². Also, the following variables were mentioned in some studies:

• Healing time between the surgical and prosthetic phases

• Detailed surgical protocols/materials

The information source were PubMed, Web of Science, and Scopus databases. The search was limited to English-language studies published before the end of August 2020. Hand searching was also conducted.

An electronic search was performed using the following keywords: Maxillary sinus AND (augmentation OR lift OR elevation) AND extraction, Maxillary sinus AND (augmentation OR lift OR elevation) AND osteotome sinus, Maxillary sinus AND (augmentation OR lift OR elevation) AND immediate implant placement.

Two reviewers conducted the search using those keywords and then performed an initial screening of the titles and abstracts they found. Any disagreements between the reviewers were resolved by discussion. The reviewers then reviewed the full-text of the selected articles.

Data were extracted from the full-text of the articles and tabulated for comparison. The following variables were evaluated: technique of tooth extraction, extraction with/without elevating a flap, distance from the tooth apex to the sinus floor, instrument and technique for elevating the sinus floor, magnitude of the elevation of the sinus membrane, use or no use of graft materials and bone substitutes, implant height, method of wound closure in the coronal implant region, success rate and survival of implants, and the duration of follow-up.

Review and data extraction were performed according to the PRISMA flow diagrams. After the full-text of the selected articles were reviewed, data were extracted and tabulated. Disagreements between reviewers were resolved by consensus.

The quality of the methodologies used in all the included studies was assessed. The criteria used were based on the ROBINS-I tool²³: bias due to confounding, bias in selecting participants, bias in classifying interventions, bias due to deviation from the intended intervention, bias due to missing data, bias in outcome measurements, and bias in selecting the reported results. The criteria used for the one randomized controlled trial were obtained from the Cochrane Center: selection bias, performance bias, detection bias, attrition bias, and reporting bias. The degree of bias was categorized as low risk, moderate risk, and serious risk. In cases with no data about a criterion, ‘no information’ is stated. Two individual investigators (L.K. and A.K.) performed the assessment and resolved conflicts through discussion.

Of the 13 articles reviewed, six were retrospective case series^12,24-28, three were case reports^21,29,30, two were prospective cohorts in the form of case series^31,32, one was a prospective case series³³, and one was a randomized controlled trial with one year of follow-up¹⁵.

Of the 13 articles reviewed, 12 provided the number of patients evaluated^{12,15,21,24-27,29-33}, and one did not²⁸. A total of 245 patients were evaluated. Six studies reported the mean age of the patients^{15,24,27,31-33}. Four studies mentioned only the age range of patients (youngest and oldest)^12,25,26,28. Three studies evaluated only one patient and provided their exact age^21,29,30. In the 13 reviewed studies, the youngest patient was 18 years, and the oldest was 74 years.

Overall, 306 implant fixtures were placed in 245 patients. The number of patients matched the number of inserted implants in eight studies^{12,21,25,26,30-33}. In one study³³, 10 implants were placed for 10 patients; out of which 8 were considered to be fresh socket implant placements simultaneous with sinus floor elevation, whereas in 2 cases, radiographs revealed that (in contrast to the researcher’s belief at the time of surgery) the sinus membrane had not actually been elevated. In four studies, the number of implants placed was greater than the number of patients^15,24,27,29. In one of those studies²⁹, despite the immediate placement of two implants in one patient, only one was accompanied by sinus lifting. Another study did not report the number of implants placed²⁸. In one study²⁷, both premolar and molar sites underwent sinus lifting. Four studies^15,24,32,33 evaluated only molar sites, and four studies^12,29-31 evaluated only premolar sites. In the randomized controlled trial, implant placement was performed for 33 patients immediately after tooth extraction and sinus lifting (test group), whereas the other 35 patients received implant placement and sinus lifting three months after tooth extraction (control group)¹⁵. Even though sinus floor elevation is more commonly performed in the premolar and molar regions, one study reported performing sinus floor elevation at the canine site, in addition to the premolar site²⁶.

Six articles did not mention the probable existence of infection in the sockets^{15,21,25,28,30,31}; four articles did not have any infected teeth in their patients^15,21,24,29, and three studies excluded infected cases^15,32,33. In three studies, prophylactic antibiotics were recommended^12,15,27. In the two of them that included periapical radiolucencies or endodontal lesions, the survival rates were similar in infected and non-infected sites¹². Among the 70 dental implants installed in patients with a possible history of endo/periodontal disease, one fixture failed, but the exact etiology of that failure was not mentioned²⁷.

Three studies reported using a Periotome for atraumatic tooth extraction^12,27,30. Four studies did not mention the instruments used for tooth extraction^21,25,28,29. Routine instruments available for tooth extraction, such as elevators and forceps, were used in the remaining studies. Among the 13 studies, one did not provide any information about the type of flap²⁸, and 6 studies adopted the flapless technique^{21,24,27,29,30,32}. In the six other studies, a full-thickness flap was elevated^{12,15,25,26,31,33}.

An osteotome with a hand mallet was used for sinus floor elevation in nine studies^{12,15,25,26,28,29,30,31,33}. However, in one study, piezosurgery was performed first to reach the Schneiderian membrane, and then the sinus membrane was elevated using an osteotome³¹. In another study, an osteotome was used with an electrical mallet to control the force²⁷. In one study, a bone expander was used with a mallet, which was believed to be similar to traditional osteotomes²⁴. However, in two studies, rounded drills were used to reach the intact sinus membrane, and then various instruments were used to detach and elevate it^15,32.

Three of the 13 studies did not mention whether or not they used grafts or biomaterials^27,28,30. Of the remaining 10 articles, two studies reported placing no graft material beneath the sinus membrane^15,26. One study used plasma rich in growth factors (PRGF) as an autograft for sinus membrane elevation³¹. One study placed a collagen sheet beneath the sinus membrane²⁴, and two studies used collagen compounds along with a xenograft^25,32. One study used only a xenograft²⁹. Another study used a xenograft or synthetic beta-tricalcium phosphates³³, whereas another article reported using a xenograft in some cases and an allograft in others¹². Only one article placed a xenograft with platelet rich fibrin (PRF) in the space between the implant and the socket walls²¹.

Five of the 13 articles used periapical radiography^{21,26,29,30,31}, and four used periapical radiography along with an occlusal or surgical template to increase the accuracy of measurements^24,25,27,33. One study used periapical or panoramic radiography¹². One study used cone-beam computed tomography (CBCT)¹⁵, and another one used both CBCT and periapical radiography³². In one study, one patient had a panoramic radiograph in his records, but the type of radiography used for the other patients was not specified²⁸.

Four studies did not specify the primary bone height at baseline^12,26,28,30. Table 4 shows the mean primary bone height at baseline in the other studies.

Three studies mentioned the magnitude of sinus floor elevation^15,25,31, of which two measured it directly and reported that it was approximately 2.9 mm²⁵, 3.6 mm (case group), and 3.4 mm (control group)¹⁵. The other study³¹ estimated the magnitude of sinus floor elevation to be 4.2 mm by measuring the approximate distance from the implant apex to the sinus floor. Seven studies mentioned the implant height^{12,21,24,25,27,30,33}. The most common implant height in all the reviewed studies was 13 mm; the shortest implant was 10 mm, and the longest was 15 mm.

Six studies evaluated the effect of using or not using bone grafts on the amount of bone gain, and their results can be categorized into two groups^{12,24,25,31-33}. The first group, which reported bone gain, included the following studies. A study that used collagen sponges alone²⁴ and reported that the mean bone height at baseline (time of surgery) was 6.02±0.75 mm, which increased to 8.05±1.58 mm at the one-year follow-up and remained constant at 8.01±1.46 mm thereafter. Another study used a combination of collagen and xenograft bone powder²⁵ and reported 4.2±1.4 mm of bone gain at the 18-month follow-up. One study used an allograft in 4 out of 5 patients, and the remaining patient received a xenograft. The amount of bone gain was reported to be 1-4 mm at the 6- and 12-month follow-ups¹². Another study used xenograft or beta tricalcium phosphate, and postoperative radiographs showed 2.5-6 mm of bone gain, with a mean value of 4.3 mm. Those patients showed optimal clinical stability and support of the prosthesis at their 2-year follow-up visits³³.

The second group evaluated crestal bone loss. In 2008, Barone et al.²⁵ reported no significant bone resorption after the placement of 12 implants despite elevating a full-thickness flap. Artzi et al.³³ performed flap surgery for the placement of 12 implants in 2003 and reported no crestal bone resorption around the implant neck. In 2017, Chen et al.³² used the flapless technique to place 37 implants and comprehensively evaluated marginal bone resorption at the 6- and 12-month follow-ups. The mean crestal bone loss was 0.63 and 0.73 mm at the 6- and 12-month follow-ups, respectively³². Taschieri and Del Fabbro³¹ evaluated crestal bone loss in 2011 and reported that the peri-implant bone loss averaged 0.36±0.19 mm at the 12-month follow-up.

In two studies, the presented cases immediately received a temporary prosthesis that sealed the socket opening^28,29. Two other studies used collagen sponges as appropriate closures for the socket opening^27,32. One study comprehensively discussed the coronal seal²⁶ and used three methods for this purpose: (A) providing mucosal coverage by elevating a flap, (B) using an adhesive bridge, and (C) placing a healing abutment. In some cases, a combination of those techniques (A and B or B and C) was used. In the study that used PRGF³¹, all implants were semi-submerged in mucosal tissue with/without PRGF. In another article, the wound was covered with a PRF membrane and loose sutures²¹. In the remaining cases^{12,15,25,26,30,32,33}, the socket was partially sealed by suturing or the use of healing abutments. Of the studies that used sutures, one study reported using the palatal slipping flap technique for coronal coverage of the socket in some cases³². Another study released the flap and sutured it at the coronal part of the socket and then fabricated a temporary fixed prosthesis to prevent trauma to the surgical site; after 2 to 3 weeks, it was replaced with a removable prosthesis²⁵.

Seven studies did not report any peri- or postoperative complications^{12,21,26,28-31}. One study divided patients into three groups according to the distance from the apex to the sinus floor³² and reported that one patient in group 1 developed mild rhinosinusitis and cough in the first three days postoperatively, one patient from group 2 and one patient from group 3 developed fatigue, and other patients reported dental pain. Two studies reported sinus membrane perforation during the surgical procedure in two (test group) and one (control group)¹⁵ and three³³ patients. Nasal bleeding was also reported in two patients in the same study³³. Two other studies reported pain and swelling as well as nasal bleeding postoperatively^24,27, but they also reported that those symptoms resolved within 24 to 48 hours without significant complication. Another study reported only pain and swelling postoperatively²⁵.

Nine studies did not attach a temporary prosthesis to the implant platform before delivery of the final restoration^{12,15,21,25,26,30-33}. Two studies by a single group of researchers^25,27 fabricated a temporary prosthesis 70 days after implant placement and then delivered the final prosthesis 2 to 3 months after that. In two other studies, the treatment was accomplished by immediate delivery of a fixed prosthesis temporarily connected to the implant^28,29. In one of them, the crown was out of the occlusion²⁹ and in the other, it was loaded immediately²⁸. In terms of time lapse until restoration, one study reported 3 to 4 months³¹, another indicated 4 months²⁸, three studies mentioned 5 months^24,28,29, seven studies^{15,21,25,26,30,32,33} reported 6 months, and one study¹² reported 5 to 6 months between the surgical procedure and final restoration.

In 10 articles, all the implants were stable during the healing and follow-up periods with no failures^{12,15,21,24,26,29-33}. One study did not specify the number of patients or implants and only mentioned the successful management of the cases²⁸. Failures were reported in two studies^25,27, which were both case-series. One of them reported the failure of 1 out of 12 implants due to infection and abscess²⁵ that occurred 6 weeks after the surgical procedure and was categorized as an early failure. In the other study²⁷, 1 out of 70 implants failed one month after placement, and it was successfully replaced with another implant after 6 months. Thus, those two studies reported a 91.7% success rate²⁵ and 98.57% survival rate²⁷ for the implants. In other studies, both those rates were 100%^{12,15,21,24,26,29,31-33} or no information about the survival or success rate was given^28,30.

Two studies reported regular recall sessions from the time of surgery until delivery of the final prosthesis^12,25. In one of them²⁵, the postoperative patients were followed up monthly for up to 6 months for prophylaxis. After delivery of the final prosthesis, all patients showed up for regular recall sessions after 2 to 4 months. The duration of follow-up was 18 months, and radiographic examinations were carried out at 6 and 18 months postoperatively. In the second study¹², the recall sessions were scheduled at 2 weeks, 2 months, and 5 months during the healing period. The final prosthesis was delivered 6 to 14 months postoperatively. After that, all patients were followed up at 6 and 12 months. One study mentioned only the time of delivery of the final prosthesis at 4 months postoperatively²⁸ and did not mention the timing of the other follow-up sessions. Three studies followed up with patients at 12 months^15,30,32 without mentioning the frequency of recall sessions during that time period. Two other studies reported a final follow-up visit with patients at 2 years^27,33 but did not mention recall sessions during that period. In two studies conducted by a single group of researchers^24,27, the recall sessions and radiographic examinations were scheduled at 70 days postoperatively. In one of them²⁷, follow-up and radiographic examinations were scheduled 1 and 2 years later, but in the second study²⁴, the follow-up sessions were scheduled annually after prosthesis delivery for a mean of 9.76±5.27 years (range, 4-17 years). In one study, radiography was performed at the time of prosthesis delivery. Of the 10 patients in that study, 1 was followed up at 12 months, 2 were followed up at 18 months, 4 were followed up at 24 months, 1 was followed up at 30 months, and 2 were followed up at 36 months²⁶. One study reported 4 years of follow-up but did not mention the time interval between recall sessions²⁹. In another study, the prosthesis was delivered 4 months postoperatively, and recall sessions were then scheduled 6 and 12 months after that and continued annually, with a mean follow-up time of 35.6 months (range, 24-50 months)³¹. In only one study, there were no recall sessions after definitive restoration delivery²¹.

Our assessment of the risk of bias is presented in Fig. 2. Most of the 11 non-randomized studies reported a low risk of bias in confounding. The three highest values with a moderate risk of bias were for deviation from the intended intervention (67%), classification of intervention and outcome measurements (50%), and missing data (42%). For the only randomized controlled trial, blinding of the outcome assessment was evaluated as high risk. Besides randomization bias, which presented a low risk, other criteria were reported to carry a moderate risk of bias. Considering all the values related to the risk of bias assessment, the studies included were categorized as being of moderate quality.

The elevation of a full-thickness flap can be associated with a higher risk of crestal bone resorption³⁷. On the other hand, flapless surgery can cause errors in correct implant placement due to inadequate visualization of the surgical site³⁸. Three sources, the PDL, bone marrow, and external periosteum, provide blood supply to the buccal bone around the teeth^39,40. Following tooth extraction, the blood supply provided by the PDL is no longer available. A thin buccal plate is mainly composed of cortical bone, so in those cases, the periosteum would be the only source of blood supply to the buccal bone plate⁴¹. Therefore, elevating a flap can impair the blood supply to the buccal bone plate provided by the periosteum³⁷, which can aggravate resorption of the buccal bone plate following tooth extraction⁴¹. Faster soft-tissue healing, decreased perioperative hemorrhaging, shorter surgical time, and less patient discomfort postoperatively are other advantages of flapless surgery. On the other hand, the surgical procedure in flapless surgery is performed blindly, which can lead to fenestration of the osteotomy site and implant mispositioning^42,43. A meta-analysis conducted in 2014 compared the flap and flapless surgical procedures for implant placement in healed extraction sockets and indicated that the risk of implant failure was higher in flapless surgery in most of the reviewed studies³⁸. It should be noted that in immediate implant placement, the coronal part of the socket allows adequate visualization of implant placement in the alveolar housing, and the soft and hard tissue conditions are favorable, so there would be no need to elevate a flap unless the surgeon wanted to reposition the flap coronally to cover the coronal part of the socket. In that case, there would be no need to elevate a full-thickness flap because a partial-thickness flap would serve that purpose. Therefore, it seems that crestal bone resorption following flap elevation in immediate implant placement is a more important issue than mispositioning of the implant.

In this review, an equal number of studies used the flap technique^{12,15,25,26,31,33} and flapless surgery^{21,24,27,29,30,32}, although more implants were placed with the flapless technique (173 implants) than via flap surgery (125 implants). Implant failure occurred with both techniques.

The correlation between crestal bone resorption and flap/flapless surgery was also previously reviewed⁴⁴. Those results showed that the mean magnitude of resorption was less than 1 mm or not significant^25,31-33. According to the implant success criteria, that amount of resorption is within the clinically acceptable range⁴⁵. Moreover, a study conducted in 2017 found no significant difference in ridge resorption following immediate implant placement with the flap or flapless technique⁴¹, which is in agreement with our findings.

Since the introduction of the closed sinus lift technique with the crestal approach⁸, osteotomes have been conventionally used for this purpose. In 2005, a meta-analysis reported 95.7% and 96% success/survival rates at 24 and 36 months, respectively, following implant placement with the osteotome technique, and those rates were similar to those with conventional implant placement in a partially edentulous posterior maxilla⁴⁶. However, another study reported that the risk of membrane perforation posed by techniques such as the hydraulic sinus lift procedure at the crestal site was less than that with the conventional use of osteotomes⁹.

Osteotome use can be accompanied by some complications. The mallet strokes used in the original crestal technique to create a “green stick” fracture and access the sinus membrane can cause discomfort and complications for patients⁴⁷. However, osteotomes are still popular for crestal sinus lifting, such that in 10 out of the 13 studies reviewed here, an osteotome was the main instrument used for sinus floor elevation^{12,15,25-31,33}. One study used an electrical mallet instead of a hand mallet because the authors believed that it enabled better control of the force applied to the osteotome and decreased postoperative patient discomfort by decreasing trauma to the bone and craniofacial structures²⁷. The authors in one of the reviewed studies used specific separating instruments to elevate the sinus membrane and reported no sign of membrane perforation during surgery³². Only two studies mentioned perforation of the sinus membrane during surgery, but they also reported that it caused no implant failure during one-year¹⁵ and two-years of follow-up¹². According to the available evidence, particularly a systematic review conducted in 2016, membrane perforation is the most common complication of sinus lifting, but no significant association has been found between the implant success rate and sinus membrane perforation⁴⁸. Therefore, use of osteotomes is still the gold standard for the crestal approach.

In 1994, Summers⁸ suggested that a minimum of 5-6 mm of residual bone height is required to ensure primary stability of the implant. Measuring the primary bone height beneath the sinus floor is easier in a healed edentulous ridge than in a fresh extraction socket. The references used for such measurements might be different in fresh extraction sockets, which is one of the limitations to accurately measuring bone changes when performing sinus lifting through an extraction socket. Our data indicate that the minimum residual bone height relative to the inter-radicular crest (as reference) was 4.54±0.64 mm³². That value was 5 mm when the interproximal crest served as the reference²⁹. Neither of those studies reported any implant failures. A systematic review in 2014 reported that a bone height less than 4 mm was associated with a lower success rate with the osteotome technique¹⁰. In contrast, a meta-analysis conducted in 2018 showed that the survival rate of implants placed in residual bone with ≤4 mm of height was only slightly lower than that of implants placed in residual bone with >4 mm of height⁴⁹. It seems that the amount of residual bone between the apex and the sinus floor and the amount of inter-radicular bone are more important to primary implant stability and the failure rate than the proximal bone level. Implant failure was reported in only two cases in the articles reviewed here. In one of the failed cases²⁵, the minimum bone thickness between the tooth apex and sinus floor was only 2 mm but was deemed adequate for primary stability of the implant, and the reason for failure after six weeks was infection and abscess, not inadequate primary stability. The other study that reported a failed case mentioned nothing about the distance from the tooth apex to the sinus floor, nor did it specify the reason for implant failure²⁷. However, it should be noted that in our study, similar to the meta-analysis conducted in 2018⁴⁹, measurements were made on periapical or panoramic radiographs; only one study used CBCT with periapical radiography for that purpose³², which could limit the accuracy of the reviewed results.

For vertical and horizontal bone expansion, osteotomes²⁵, bone expanders and mallets²⁴, and expander screws⁵⁰ were applied in premolar, molar, and both premolar and molar sites, respectively. The overall results with these methods showed survival and success rates comparable to those of standard surgery, ranging from 95% to 100%, indicating that the choice of devices does not have a notable effect on the outcomes. Also, a recent study showed that transcrestal sinus floor elevation has no negative effect on the long-term survival rate of implants⁵¹.

The application of graft materials below the sinus membrane following its elevation is a challenging topic in sinus lifting with the crestal approach⁵². Some of the studies reviewed here evaluated how using or not using graft materials affected bone gain, bone loss, and the survival and success rates of implants^48,53.

Two reasons have been suggested for new bone formation in a sinus cavity that has undergone sinus floor elevation using an osteotome: osteogenic activity after a mini-fracture of bone at the sinus floor and the role of mesenchymal cells present in the Schneiderian membrane⁵⁴. Therefore, maintaining the sinus membrane in an elevated position, either with graft materials and bone substitutes or by the implant apex, can trigger new bone formation beneath the membrane so long as the blood clot is preserved¹¹. A randomized clinical trial in 2013 evaluated the survival rates of implants with and without grafting and reported similar residual bone heights and survival rates of 95.2% and 95%, respectively⁵⁴. The marginal bone loss did not differ significantly between the two groups after 3 years, but the amount of bone gain in the graft group was significantly higher than that in the no graft group after 6 months. However, bone gain significantly decreased after 3 years in the graft group, whereas it continued to grow slowly in no-graft group. At the 3-year follow-up, the final volume of bone within the sinus was equal in the two groups (around 3 mm)⁵⁴. Nonetheless, a meta-analysis conducted in 2018⁴⁹ reported that the results of studies about bone gain supported the use of graft materials beneath the sinus membrane. In this meta-analysis⁴⁹, the amount of marginal bone loss in the bone graft group was slightly higher than that in the no-graft group, but that difference was not significant. However, the amount of bone gain did differ significantly between the groups. No significant difference was found in the implant survival rate with the use and non-use of a bone graft beneath the sinus membrane, which was in agreement with the findings of a meta-analysis conducted in 2017¹¹. In this review, we evaluated studies that performed sinus lifting with and without the placement of bone grafts, but more studies used an autogenous bone graft²⁹, bone substitute^{12,21,25,29,32,33}, or collagen sponges²⁴ than did not place any biomaterial beneath the sinus membrane²⁶. All studies that reported bone gain used some sort of collagen material or bone substitute beneath the sinus membrane. However, it is noteworthy that in the only study that used PRGF to keep the sinus membrane elevated, the elevation averaged 2.9±0.8 mm, but the amount of bone gain was not mentioned at the time of prosthesis delivery or the follow-up sessions³¹. Therefore, our results are in line with the findings of a recent meta-analysis⁴⁹ that showed that greater increase in bone height can be expected when using bone grafts.

Controlled clinical trials are seriously lacking on this topic and are therefore recommended. Calculating important data, including initial bone height, amount of sinus lifting, clinical and radiographic criteria pertaining to the success of implants, follow-up to the prosthetic and loading phases, and split mouth designs, are suggested for further studies. Also, some important factors such as patient-oriented outcomes, digital workflow, and presence of sinus septa in the apical socket should be emphasized in further research.