Short implants versus longer implants with maxillary sinus lift. A systematic review and meta-analysis

This study compared the survival rate of dental implants, amount of marginal bone loss, and rates of complications (biological and prosthetic) between short implants and long implants placed after maxillary sinus augmentation. This systematic review has been registered at PROSPERO under the number (CRD42017073929). Two reviewers searched the PubMed/MEDLINE, Embase, LILACS, and Cochrane Library databases. Eligibility criteria included randomized controlled trials, comparisons between short implants and long implants placed after maxillary sinus augmentation in the same study, and follow-up for >6 months. The Cochrane Collaboration’s tool for assessing the risk of bias in randomized trials was used to assess the quality and risk of bias of the included studies. The search identified 1366 references. After applying the inclusion criteria, 11 trials including 420 patients who received 911 dental implants were considered eligible. No significant difference was observed in the survival rate [p = 0.86; risk ratio (RR): 1.08; 95% confidence interval (CI): 0.46–2.52] or in the amount of marginal bone loss (p = 0.08; RR: −0.05; 95%CI: −0.10 to 0.01). However, higher rates of biological complications for long implants associated with maxillary sinus augmentation were observed (p < 0.00001; RR: 0.21; 95%CI: 0.10–0.41), whereas a higher prosthetic complication rate for short implants was noted (p = 0.010; RR: 3.15; 95%CI: 1.32–7.51). Short implant placement is an effective alternative because of fewer biological complications and similar survival and marginal bone loss than long implant placement with maxillary sinus augmentation. However, the risk of mechanical complications associated with the prostheses fitted on short implants should be considered.


Introduction
Dental implants are considered an option for oral rehabilitation, particularly in the posterior maxillary region. 1 However, it is not possible to place dental implants with an adequate length in some clinical situations, primarily due to the lack of sufficient bone. 2 One option to overcome this limitation is using short implants. 3Another option is augmentation of the bone height through techniques such as maxillary sinus augmentation with or without bone grafting, which enables long implant placement. 4hort implant placement has been considered as a less invasive alternative, and it is associated with greater simplicity, a shorter surgical duration, and lower morbidity rates and costs. 5,6Furthermore, the clinical outcomes of short implants are reportedly similar to those of long implants in the posterior maxillary region. 7,8However, the bone-to-implant contact area with short implants may be small, impairing the primary stability and osseointegration process 9 and eventually leading to implant failure. 1 In addition, the discrepancy in the crown-to-implant ratio (C/I ratio) may increase the risk of marginal bone loss and other complications such as screw loosening, prefabricated abutment fracture, retention loss, and crown debonding. 10,11Maxillary sinus augmentation, which is performed using the lateral window technique or Summers technique, has shown favorable outcomes regarding implant survival. 6,12However, these are complex surgical procedures because they can result in postoperative complications that increase the morbidity rate and increase the patient's reluctance to undergo the procedure. 13,14The cost and duration of treatment are greater than that in conventional implant placement without bone grafting. 15ifferent reviews have reported the use of short implants with available alternative treatments. 1,15,16,17owever, more randomized controlled trials (RCTs) have reported comparisons between short implants and long implants placed after maxillary sinus augmentation. 8,18,19,20,21,22,23Thus, the choice of technique (short or long implants with maxillary sinus augmentation) should be based on recently published literature, including the relative risks of each technique.
The present systematic review aimed to compare short implants and long implants (length > 8.5 mm) placed after maxillary sinus augmentation for survival rates, amount of marginal bone loss, and biological and prosthetic complications.The null hypotheses were as follows.First, there is no difference in the survival rate between short implants and long implants placed after maxillary sinus augmentation.Second, the implant length does not influence the amount of marginal bone loss.Third, the implant length does not change the occurrence of biological and/ or prosthetic complications.

Methodology
This systematic review was registered in the PROSPERO database (CRD42017073929) and structured according to the PRISMA checklist. 24The protocol was established according to models proposed in the relevant literature. 16,25

Eligibility criteria
Studies meeting the following criteria were included: a. RCTs; b. comparisons between short implants without maxillary sinus augmentation and long implants with maxillary sinus augmentation in the same study; and c. follow-up for > 6 months.No restrictions on language or date of publication for searching in the electronic databases were made.Studies meeting at least one of the following criteria were excluded: a. animal studies; b. in vitro studies; c. case series or case reports; d. retrospective studies; e. patients or data repeated in other articles included; f. computer simulations; g. studies that presented only short implants without a comparison group; h.studies that considered short implants longer than 8.5 mm; and i. studies with short implants associated with maxillary sinus augmentation technique.

Search
The following question was elaborated for the PICO process (population, intervention, comparison, outcomes): Do short posterior maxillary implants exhibit the same clinical predictability as long implants placed after maxillary sinus augmentation?The population (P) was patients rehabilitated with dental implants in the posterior maxilla; the intervention (I) was short implant (≤ 8.5 mm) placement; the comparison (C) was long implant (> 8.5 mm) placement after maxillary sinus augmentation; and the outcomes (O) were the survival rate of implants, amount of marginal bone loss, and biological and prosthetic complication rates.
Two researchers independently searched the PubMed/Medline, Embase, LILACS, and Cochrane Library databases for articles published up to January 2018 according to the eligibility criteria.Studies comparing the survival rate of short implants (≤ 8.5 mm) placed in the posterior maxilla with that of long implants (> 8.5 mm) placed after maxillary sinus augmentation were selected.The search terms included (short implant and maxilla) OR (short implant and sinus lift) OR (short implant and sinus elevation) OR (short implant and maxilla and augmentation) OR (short implant and sinus floor augmentation) OR (short implant and maxilla and dental implant) OR (short implant and sinus lift and dental implant) OR (short implant and sinus elevation and dental implant) OR (short implant and maxilla and augmentation and dental implant) OR (short implant and maxilla and dental implant) OR (short implant and sinus floor augmentation and dental implant).
No filters and limits were used in the database

Risk of bias
One author evaluated the risk of bias in the included studies using the Cochrane Collaboration's tool for assessing the risk of bias in randomized trials.The assessment criteria were separately prepared for different domains: random sequence generation, allocation concealment, blinding (patients and/or outcome assessment), incomplete outcome data, and other bias.For each domain, the risk of bias was graded as high, low, or unclear based on criteria described in the Cochrane Handbook for Systematic Reviews of Interventions version 5.1.0.A second author was responsible for checking the risk of bias, and a consensus was obtained with another author in case of discrepancies.

Data extraction
One researcher extracted the data from articles (quantitative or qualitative), and another check these data.The following data were recorded: author/year, number of patients, mean age, length of short and long implants and number of implants, diameter (mm) of implants, implant system, insertion bone graft/system, technique performed, follow-up (months), amount of marginal bone loss, complications (biological and prosthetic), and survival rates of implants (short and long implants).

Summary measures
The meta-analysis was based on the Mantel-Haenszel (MH) and Inverse Variance (IV) weighting methods.The outcome measures evaluated by risk ratio (RR) included the survival rates of implants and biological and prosthetic complications.The amount of marginal bone loss was evaluated by mean difference (MD) and the corresponding 95% confidence interval (CI).The RR and MD values were considered significant when the P-value was < 0.05.Reviewer Manager 5 software (the Cochrane Collaboration) was used for meta-analysis.The I 2 statistic was used to analyze the percentage of variations due to heterogeneity.I 2 values > 75% (range: 0-100) indicated high heterogeneity.Because the meta-analysis showed significant heterogeneity (p < 0.10), a random-effects model was adopted, whereas the fixed-effect model was used when heterogeneity was not statistically significant. 26,27,28
The inter-investigator agreement (Kappa) was calculated by evaluating the selected titles and abstracts.The values derived for the articles selected from PubMed/MEDLINE (kappa = 0.96), Embase (kappa = 0.95), LILACS (kappa = 0.93), and Cochrane Library (kappa = 1.00) suggested a high level of agreement between investigators.
The length of short implants ranged from 4 mm to 8.5 mm, while that of conventional, long implants ranged from 10 mm to 15 mm.The diameter of all implants ranged from 3.75 mm to 7.0 mm.In addition, five studies reported the use of internal connections, 8,19,20,22,29 four used external connections, 21,23,31,32 and two did not report the type of connection. 18,30The follow-up period of included studies varied from 9 to 36 months (Table 2).

Risk of bias
The Cochrane Collaboration's tool for assessing risk of bias in randomized trials indicated that all studies showed a low risk of bias for random sequence generation, allocation concealment (selection bias), incomplete outcome data (attrition bias), selective reporting (reporting bias), and other sources of bias.Regarding blinding of participants and personnel, one study reported the blinding of patients without the blinding of surgeons, 18 whereas two studies reported no blinding of surgeons. 31,32All other studies were unclear about this parameter.This may have been due to difficulty in blinding surgeons and/or patients scheduled for auxiliary surgical, particularly in split-mouth design studies. 29,31,32Blinding of outcome assessment was reported for almost all included studies. 18,21,22,23,29,31,32However, a few studies were unclear regarding this. 8,20,30(Figure 2).

Marginal bone loss
Ten studies reported data in terms of the amount of marginal bone loss, which was reported in millimeters.However, only eight of these studies 19,21,22,23,29,30,31,32 were used for meta-analysis because one study reported bone loss separately (mesial and distal), 20 and another study reported marginal bone loss without standard deviation. 8The mean amount of marginal bone loss for short implants was 0.86 mm (range: 0.10-1.41mm), whereas for long implants placed after maxillary sinus augmentation, it was 0.99 mm (range: 0.10-1.74mm).The meta-analysis based on MD found no significant difference between short implants and long implants (p = 0.08; RR: −0.05; 95%CI: −0.10 to 0.01) (Figure 4).

Discussion
The findings of this systematic review and metaanalysis suggested that for posterior maxillary rehabilitation, short and long implant placement have similar implant survival rates.Thus, the first null hypothesis was accepted.Moreover, this finding agrees with the findings in studies reporting high success rates for short implants used for posterior maxillary rehabilitation. 1,8,18,44Furthermore, different systematic reviews evaluating only short implants showed survival rates between 93.1% and 99.1%, 45,46 which corroborate with the survival rate of 98.09% in the current review.
Among the selected studies, only five reported information pertaining to implant loss, including nine short implants and nine long implants.Accordingly, the failure rates for short and long implants were 1.90% and 1.95%, respectively.Some studies have reported mobility, chronic sinus infection, and history of periodontal disease and abscess as reasons for implant  loss. 8,22,23,29However, not all studies have reported the reasons for implant loss 31 .Some risk factors may influence the survival of implants, 47,48,49 including occlusal overload after prosthetic rehabilitation.From the biomechanical perspective, when comparing short implants with long implants, short implants may be associated with a trend for higher stresses within the implant and consequently on the cortical bone tissue. 50However, this can be overcome by using implants with splinted crowns. 50,51nother factor that may influence the increase in implant survival rate, regardless of length and diameter, is the fact that some studies report that implants with a wide diameter (≥ 5 mm) are more favorable than implants with a narrow and regular diameter. 52,53No such association was found in the present systematic review and meta-analysis.Studies that used implants with a 4-mm diameter showed 100% survival rates for both types of implants. 19,30,32Similarly, studies that analyzed implants with a diameter of 5 mm reported survival rates of 100% for long implants placed after maxillary sinus augmentation and 97.2% for short implants. 31The reason for this finding may be consideration of the implant diameter as a secondary factor for the long-term survival of implants placed in the posterior maxillary region. 54rimary stability is also essential for the success of dental implants. 49,55A lack of primary stability may compromise osseointegration and the longevity of osseointegrated implants. 56,57Analysis of the implant stability quotient (ISQ) using devices such as the Osstell ISQ has been documented in only two studies, 8,18 which stated high values indicating good primary stability after both short and long implant placement.
Resonance frequency analysis measurements result in a mean ISQ value of 68.68 for short implants and 70.69 for long implants placed after sinus augmentation.These findings are consistent with those of a previous study reporting the achievement of good primary stability for short implants. 58Furthermore, these values are within the limits previously established in the literature (54-74 ISQ). 59he literature suggests that modification of the implant design and surface may accelerate the process of osseointegration and influence the success of dental implants, 60 primarily in the analysis of short implants and machined implants. 47,49The surface properties of the implant have been identified as an important factor for osseointegration. 61,62Bechara et al. 8 performed a unique study that assessed implant surfaces reinforced with nanostructured calcium and reported a high success rate for both long implants placed after maxillary sinus augmentation (95.6%) and short implants (100%); however, the study did not perform comparisons with other surfaces.Therefore, it is difficult to conclude whether there is an actual benefit in terms of the success rate, particularly for short implants.
Marginal bone stability is another relevant factor for implant-supported rehabilitation because excess bone loss is considered as one of the secondary factors that may lead to implant loss. 63However, the current analysis finds no difference in the marginal bone loss between short implants and long implants placed after maxillary sinus augmentation.Thus, the second hypothesis was also accepted.Moreover, this finding is consistent with previous data showing similarity in marginal bone loss values for short and long implants placed in the posterior jaw. 1 Although there was no difference in marginal bone loss in the present systematic review, it is important to note that several studies observed greater marginal bone loss with long implants placed after maxillary sinus augmentation. 8,21,29,31,32This bone loss may have occurred because long implants are supported on grafted bone, which is considered to be of poorer quality compared with natural bone. 64In this way, it is indicated that the use of short implant is preferable considering only the bone loss aspect.However, no difference for bone loss between implant lengths was observed in this study.In addition, prosthetic complications are statistically more common for long implants and preclude a simple explanation.It is important to emphasize that bone loss may be influenced by several factors such as the implant geometry, 65 parafunctional habits, 66 crown fixation system, 67 biological factors, 68 systemic factors, 63 overheating during surgical preparation, 44,47 and the prosthesis loading condition. 25However, because of lack of data in the included studies, we could not perform sub-analyses based on these variables, and this may be considered as a limitation to our study.
The third hypothesis was rejected because significant differences were found in the biological and prosthetic complication rates between short implants and long implants placed after maxillary sinus augmentation.In the quantitative analysis of biological complications, a higher rate for long implants than for short implants was seen, which is consistent with previous findings showing an increased risk of biological complications after the performance of complementary surgical procedures for bone augmentation. 44,69he use of auxiliary techniques such as maxillary sinus augmentation is less accepted than conventional implant placement techniques because it results in increased morbidity, costs, and surgical duration 35,70 along with an increase in the time required for rehabilitation. 35The main complications reported in the included studies were membrane perforation, bleeding and sinusitis, 8,18,22,29,31,32 and pain and swelling .8,22,23However, these factors, even if documented as complications, cannot be considered limiting in terms of the decision to perform maxillary sinus augmentation for long implant placement, particularly if the dental surgeon has a good learning curve.
Regarding the rate of prosthetic complications, short implants are considered unfavorable compared with long implants.These results are in agreement with those reported in the literature, which indicate higher risks of prosthetic complications with the use of short implants because of an increase in the C/I ratio, which leads to mechanical failures such as loosening, pillar bolt, and ceramic fractures. 19,22,29,31,71,72erri et al. 73 reported that a C/I ratio of 1:1 may be considered more favorable in terms of lowering the risk of mechanical or prosthetic complications.
The results of the present review must be evaluated with care due to the limitations presented in the selected studies.One limitation of our review is the short follow-up period in some included studies and the small sample size in others.Moreover, some selected studies did not report failure rates for the different variables such as diameter, implantabutment connection, C/I ratio, and others.Finally, in some studies, the length of the implants was selected by the surgeon according to individual clinical circumstances.Further RCTs that include parameters that may influence the findings such as C/I ratio, splinting factor, implant geometry, and implant surface, should be conducted.

Conclusion
In conclusion, our findings suggest that short implant placement is an effective alternative to long implant placement with maxillary sinus aug ment at ion b e cau se of fewer biolog ica l complications and similar survival and marginal bone loss.However, the risk of mechan ical complications associated with the prostheses fitted on short implants should be considered.

Figure 2 .
Figure 2. Included studies assessed using Cochrane Collaboration's tool for assessing risk of bias in randomized trials

Figure 4 .
Figure 4. Forest plot for the event "marginal bone loss"

Figure 6 .
Figure 6.Forest plot for the event "prosthetic complications"

Table 1 .
Reasons for the exclusion of 15 articles.
IncludedFigure 1. Flow diagram of the literature search and results
Forest plot for the event "implant survival" Forest plot for the event "biological complications"