Clinical and Radiographic Evaluation of Short Implants Placed in the Posterior Mandible: A 1-Year Pilot Split-Mouth Study

Implant placement is a very effective and common approach that has been performed during the past decades with reliable long-term results.¹  Since the discovery of osseointegration by Brånemark, the replacement of missing teeth has propelled implant therapy into a new era of reconstructive therapy.² 

At that time, the longest possible implants were used, based on the assumption that the longer the implant was, the more bone-to-implant contact (BIC) and the more stability could be achieved by the implant, leading to higher survival rates. This statement is now being revised, since short implants have shown promising results.^3,4 

There are some anatomical limitations, such as the alveolar nerve in the mandible and the sinus on the maxilla, which make the placement of the implant difficult in some cases due to the insufficient bone height availability. In these cases, advanced surgical procedures for vertical bone augmentation such as maxillary sinus floor elevation, onlay bone grafts, guided bone regeneration (GBR), and nerve transpositioning, among others, are required prior to placing a regular-length implant.⁵  Although regular-length implants show very good survival rates in augmented bone,^6,7  these procedures are time-consuming, technically demanding, and present more morbidity and increased costs.^6,8–10  In order to overcome these disadvantages, short implants were introduced as an alternative treatment.⁶ 

Multiple definitions of short implants have been proposed over the past years,¹¹  varying from ≤10 mm^11–18 to 8 mm.^6,19–21  During the European Association for Osseointegration Consensus Conference 2015, short implants were defined as implants with an intrabony length of 8 mm or less.²¹  In this study, we consider a device measuring 8 mm or less to be a short implant.

Regarding regular-length implants, long-term data is available, and shows very good results (98.1 % after 5 years).⁷  However, long-term data of short implants is still lacking. Short implants are associated with some controversy and many questions have been proposed concerning their length. Suggesting that short implants, have less BIC than their regular-length counterparts, contributing to a lower survival rate.¹¹  Also, bone quality is another factor that has been associated with lower implant survival rates. Although short implants are indicated in areas where the bone quantity is reduced, quality of bone has been suggested as a factor contributing to the success of these implants. Reduced bone density and poor bone quality leads to implant failure.²²  These implants are usually placed in posterior areas of the maxilla or mandible where the bone quantity and quality are often poor.^3,11,22  Although bone quality is considered an important parameter affecting implant survival, the nature and mechanism of bone remodeling around dental implants is largely not understood. In fact, recent studies describe osseointegration as a result of a foreign body response rather than a bone healing process.^22–24  Also, the marginal bone loss (MBL) around these implants is an issue that should be analyzed compared to regular-length implants, because of their inferred “inferior” mechanical conditions. It has been proposed that the higher crown-to-implant ratio (C/I)²⁴  of short implants could lead to increased MBL. Due to bone resorption following tooth loss, the crown height of short implants is usually greater than for average crowns to reach the occlusal plane, leading to a higher C/I ratio, so that the lever arm is higher, which could lead to more peri-implant stress and, consequently, greater marginal bone loss.¹¹ 

Noticeably, there are many open questions regarding short implants, such as long-term outcome, MBL around these implants, C/I ratio, and prosthetic outcomes, that must be answered.

So far, most studies involving short implants compared their outcome to regular-length implants placed in different subjects. An intrasubject comparison would facilitate more accurate and direct comparisons between the different implant lengths, thereby eliminating the effect of different subjects and providing more precise answers to the aforementioned questions.

The aim of this study was to ascertain whether short implants are comparable to regular-length implants in the posterior mandible in clinical and radiographic terms after 1 year.

The primary objective was to compare the mean marginal bone resorption between short (7 mm) and regular-length implants (≥10 mm) after 1 year.

The secondary objectives were:

• (1) To compare soft tissue parameters, such as peri-implant bleeding on probing, probing depth, and clinical attachment level between short- and regular-length implants after 1 year.

• (2) To evaluate the success of short implants in relation to the C/I ratio.

The tertiary objective was to compare the prosthetic outcome between restorations placed on short- and regular-length implants after 1 year.

This study was as a prospective pilot clinical study in a split-mouth design according to the World Medical Association “Declaration of Helsinki.” All procedures and materials were approved by the ethical committee of the University Hospital Freiburg, Germany. Following approval from ethics committees, 10 subjects were selected. Each subject received 2 two-piece short titanium implants (7 mm) and 2 two-piece regular-length titanium implants (10 and 12 mm) (Superline Fixture, Dentium, Seoul, Korea) by means of guided surgery. Implant dehiscence, defined as a lateral exposure of the implant threads at the time of implant placement, was augmented by covering with autogenous bone chips obtained from adjacent areas and with bone substitutes (Osteon I or Osteon II, Dentium, Seoul, Korea) and a collagen membrane (Collagen Membrane, Dentium, Seoul, Korea).

The implants were distributed according to the split-mouth design in the posterior mandible, in which one side received short implants and the other side received regular-length implants. The implants were restored by means of splinted screw-retained metal-ceramic crowns. Two months post-placement, the implants were exposed by means of a second-stage surgery, and titanium abutments and metal-ceramic crowns were fabricated and delivered. The study subjects were followed for 1 year after receiving the restorations and possible dropouts and withdrawals were carefully monitored during the entire investigation period. Follow-up visits were scheduled at 6 and 12 months after delivery of the splinted screw-retained metal-ceramic crowns.

The data evaluation of the investigation was performed 1 year after delivery of the final restorations. The study was performed in a controlled design. Data of the short implants were compared with those of the regular-length implants.

Any implant removed after implant placement was considered a failure. A “successful implant” was defined as an implant that did not cause allergic, toxic, or infectious reactions either locally or systematically. The implant had to show anchorage to a functional prosthesis, no signs of fracture or bending, as well as no mobility. A mean MBL of more than 1 mm was considered an unsuccessful implant after the first year of function.

A “successful restoration” was considered a prosthetic restoration that was stable and functional and did not show any signs of chipping or delamination of the veneering ceramic. Patients were selected according to the inclusion and exclusion criteria reported in the Table.

The implants were provided in lengths of 7, 10, and 12 mm, and in diameters of 4.0 mm.

A diagnostic setup was used to fabricate an imaging appliance (radiographic template) that, in combination with cone-beam computerized tomography (CBCT) and guided surgery software (Med3D, C.Hafner GmbH, Pforzheim, Germany), served in determining the available amount of bone, as well as the exact position of the implants to be placed. Using the data computed by the software, the imaging appliance was transformed in the dental laboratory into a surgical guide to be used for implant placement.

Reproducible periapical radiographs were taken at the time of implant placement, delivery of final restoration, and at the 12-month follow-up visit. A custom-made acrylic stent around the film holder was used. The radiographs were taken with the long-cone parallel technique and read and evaluated by 2 independent radiologists. The measurement consisted of the distance from the implant shoulder to the first radiographically observable BIC mean measurements mesially and distally using an image processing program, ImageJ (imageJ 1.49v, Wayne Rasband, National Institutes of Health, Bethesda, Md).

A sample size of 10 will have 80% power to detect an effect size of 0.996 using a paired t test with a 0.05 two-sided significance level. Each subject received 4 implants resulting in 40 implants. There were a total of 40 observations per time point, therefore achieving a power gain.

In circumstances where there is little prior information available, Cohen²⁵  proposed a standardized effect size D. For cases in which the difference between two groups 1 and 2 is expressed by the difference between their means d = (m2 − m1) and s is the standard deviation of the endpoint variable, which is assumed to be a continuous measure, then D = (m2 − m1)/s = d/s. A value of D ≤ 0.2 is considered a “small” standardized effect, D ≈ 0.5 as “moderate,” and D ≥ 0.8 as “large.” Experience has suggested that in many areas of clinical research these can be considered as a good practical guide for design.

The statistical evaluation was performed by the Institute of Medical Biometry and Medical Informatics, University of Freiburg, Germany and considered all collected data from surgery and follow-up procedures. A linear mixed model is fitted with random intercepts with a normal distribution for each patient to evaluate implant length effects on response variables. This is done separately for different response variable (medial, distal, mean bone loss) and different time points. For the bleeding on probing (BOP) values the corresponding mixed-effects logistic regression model is used.

The analysis was carried out using the statistical software STATA 14.1 (StataCorp LP, College Station, Tex).

Ten patients received a total of 40 implants (20 short, 20 regular length). No implants presented any allergic, toxic, or infectious reactions either locally or systemically. The short implants had a length of 7 mm, while regular-length implants had lengths of 12 mm (5 implants) and 10 mm (15 implants). Regarding surgery, all the flaps were performed with a single incision without releasing. Additionally, small bone grafting procedures, exclusively in the buccal aspect, were performed around 16 implants (7 short and 9 regular-length implants). However, all implants were inserted in healed sites and presented a cortical anchorage, being stable without showing any mobility. None of the implants failed, all offering anchorage at prosthetic delivery, 6 months, and 12 months, as well as not showing any signs of radiolucency. Implants of both short and regular lengths showed a survival rate of 100%.

From implant placement to prosthetic delivery, remodeling occurred. At the prosthetic delivery, there was already an existing mean MBL of −0.85 mm (SD ± 0.59) in the short implant group and −0.87 mm (SD ± 0.37) in the regular-length implant group. Twelve months after prosthetic delivery, short implants presented an MBL of −0.56 mm (SD ± 0.51) measured from the implant shoulder to the first radiographic bone-to-implant contact, and regular-length implants presented a mean MBL of −0.67 mm (SD ± 0.23). Twelve months after prosthetic delivery, bone gain occurred in both types of implants: short implants showed a bone gain of +0.29 mm (SD ± 0.39), while regular-length implants showed a bone gain of +0.19 mm (SD ± 0.38). Although the overall result after 12 months was an existing marginal bone loss, the greatest bone loss occurred during the first 3 months after implant placement. After this remodeling process, and when the prosthetic delivery followed, a bone gain occurred (Figures 1 through 5).

In this clinical evaluation, short implants showed a mean BOP of 0% at prosthetic delivery, 70% at 6 months, and 85% after 12 months of follow-up. On the other hand, regular-length implants showed the same mean BOP as short implants at prosthetic delivery and at 6 months; however, at 12 months regular-length implants showed a BOP of 75%. Thus, no statistically significant differences were found between short and regular length implants (P = .299) (95% CI −3.67–1.13).

Probing depth was evaluated at prosthetic delivery, after 6 months, and after 12 months. The measurements were performed buccally, lingually, mesially, and distally for each implant. Short implants showed a mean PD of 1.89 (SD ± 0.62) mm at prosthetic delivery, 2.62 mm (SD ±0.43) after 6 months, and 3.21 mm (SD ± 0.73) after 12 months.

Regular-length implants showed a mean PD of 1.67 mm (SD ± 0.54) at prosthetic delivery, 2.7 mm (SD ± 0.60) after 6 months, and 3.2 mm (SD ± 0.85) after 12 months. After 12 months, no statistically significant difference was found between the 2 groups (P = .447) (95% CI −0.40–0.17).

Short implants presented mean recession values of +0.23 mm (SD ± 0.40) at prosthetic delivery, +0.35 mm (SD ± 0.38) after 6 months, and +0.1 mm (SD ± 0.18) after 12 months. Regular-length implants showed a mean recession of +0.15 mm (SD ± 0.32), +0.18 mm (SD ± 0.65), and +0.13 mm (SD ± 0.32), respectively. This means neither short nor regular-length implants presented recessions. Accordingly, no statistically significant difference between the 2 implant types after 12 months was found (P = .461) (95% CI −0.13–0.06).

Short implants showed a mean CAL of 1.55 mm (SD ± 0.77) at prosthetic delivery, 2.27 mm (SD ± 0.53) after 6 months, and 3.11 mm (SD ± 0.87) after 12 months. On the other hand, regular-length implants showed a CAL of 1.52 mm (SD ± 0.48) at prosthetic delivery, 2.45 mm (SD ± 0.6) after 6 months, and 2.92 mm (SD ± 0.95) after 12 months. After 12 months, no statistically significant difference was found between the 2 groups (P = .204) (95% CI −0.47–0.10).

One of the secondary objectives of our study was to evaluate the success of short implants in relation to the C/I ratio. The length of the crowns ranged from 6.54–12.51mm. The overall mean crown length was 9.97 mm. Crowns supported by short implants presented a mean length of 9.11 mm (min: 6.54 mm, max: 12.51 mm). Crowns on regular-length implants presented a mean length of 9.02 mm (min: 11.01 mm, max: 7.11 mm). C/I ratio varied from 0.59 to 2.28. The mean C/I ratio of the included restorations was 1.01. Short implants presented a mean C/I of 1.34 (min: 0.93, max: 2.28) while regular-length implants presented a mean C/I ratio of 0.86 (min: 0.59, max: 1.1). The C/I ratio did not affect marginal bone loss 12 months after prosthetic delivery (P = .421) (95% CI −0.65–0.27).

Prosthetic restorations were evaluated in terms of anatomical form, color match, retention, polishability, surface staining, soft tissue health, and proximal contact points over the 3 evaluation appointments and showed extremely suitable results. To gauge restoration success, various factors were evaluated, including stability, chipping, delamination, or any complication. All restorations were stable along the study except one crown that presented with chipping after 12 months. This chipping occurred at the distolingual side of a molar on a regular-length (12 mm) implant. The reason for failure was a laboratory problem in the construction of the framework. No other complications occurred on the prostheses after 12 months' follow-up, leading to a prosthetic survival and success rate of 95% for regular-length implants and a 100% prosthetic survival and success rate for short implants.

The survival rates observed in this clinical study were similar for short- and standard-length implants (100%). These results were in agreement with other published studies.^6,26,27  No difference between short implants and regular-length implants combined with bone augmentation respectively (7 short and 9 regular-length implants) was found. However, this is not the main concern for clinicians and, therefore, data regarding short implants in augmented bone is lacking. The focus is whether short implants depict similar clinical outcomes to that of regular-length implants whether these are augmented or not. Many clinical studies have been proposed combining regular-length implants in augmented bone compared to short implants placed in native bone^4,28,29  and have shown comparable outcomes.^21,30 

Going into the literature, in a recent systematic review including 4 randomized controlled clinical trials (RCT), similar implant survival rates between implants placed in vertically augmented posterior mandibles (95.09%) and short implants (96.24%) were found.²¹  However, more surgical complications were reported in the vertically augmented implant group. In 1 RCT including only implants placed in the posterior mandible, short implants (6.6 mm) were compared to regular-length implants inserted in the vertically augmented posterior mandible of a control group.³¹  Sixty partially edentulous patients were included, 30 of which received 1 to 3 short implants (6.6 mm), and the other group regular-length implants (9.6 mm) in the vertically augmented bone. No statistically significant differences for prosthesis and implant failures were found between the 2 groups. In addition, more complications in augmented implant sites were also reported in this study. The authors reported that short implants could be an interesting alternative to vertical augmentation in posterior atrophic mandibles since the treatment is faster, cheaper, and associated with less morbidity.³¹  In a randomized controlled split-mouth clinical trial, short implants were compared to longer ones in vertically augmented mandibles after a 5-year follow-up. Fifteen patients with bilateral atrophic mandibles and 15 patients with bilateral atrophic maxillaries were vertically augmented with interpositional bone blocks and maxillary sinuses with particulated bone via a lateral window. There were no statistically significant differences in the failure rates. Short implants (5 mm) achieved similar results to longer implants in augmented bone.³²  The authors recommended the use of short implants as well.

However, one factor influencing the survival rate of short implants is the location of the implants. This factor is also associated with the bone quality, since the posterior location of the maxilla and mandible show reduced bone density. Bone quality was demonstrated to be an important factor in implant survival^24,33 ; however, no clear physiological process has been presented to explain why this is the case. Some authors conclude that because no drilling protocols are adapted to the bone quality, a poor primary stability of the implant is achieved, leading to lower success and survival rates.²⁴  Indeed, an association of decreased primary stability with reduced survival rates has been proposed. Low primary stability can lead to micromovements that could lead to fibrous encapsulation and subsequent implant loss.³⁴  Additionally, comparisons of survival rates have been performed between implants placed in the maxilla and mandible. Nevertheless, bone quality and density is lower in the maxilla, which would suggest a higher implant failure rate than for implants placed in the mandible. Therefore, short implants were evaluated in a systematic review comparing implants placed in the maxilla and in the mandible. Short implants showed improved performance in the mandible than in the maxilla.¹⁶  However, in a recent systematic review including only RCTs, no difference in failure rates were found between mandible and maxilla.²⁰  The authors concluded that these good results may be explained by the adaptation of drilling protocols to the bone quality.

In our study, only implants in the mandible were evaluated; therefore, the maxilla and mandible were not compared. However, bone quality was assessed and implants placed in lower bone quality areas (Class II and III, Lekholm and Zarb) did not show lower survival rates than implants placed in Class I areas.

Another factor influencing the success of short implants is the surface configuration of these implants. Implants have been treated with different methods to improve the osseointegration process and achieve increased BIC.²⁰  In the early years of implant dentistry, machined implants were used. These have shown lower survival and success rates relative to rough-surface implants, based on the assumption that more microscopic surface area enables greater BIC, increased implant stability, and more reliable osseointegration.^{16,20,22,24,33,35,36}  Various methods, such as sandblasting, acid etching, a combination of both, or oxide-thickened surfaces, have been proposed to create a rough surface.³³  Furthermore, it has been suggested that the surface configuration may have more influence on implant survival than implant length does.²²  For this reason, we chose to use rough sandblasting with large-grit and acid etching (SLA) surface implants.

Many studies have reported that implant length does not seem to have any influence on the survival and success rates of the implants, as well as comparable outcomes to regular-length implants.^21,37 

The first objective of this study was to compare MBL between short and regular-length implants. Marginal bone loss is considered a biological complication⁷  and a consistent parameter with which to rate the success of an implant.³⁸  This is the first parameter included in the success criteria proposed by Albrektsson et al³⁹  with a modification, in which an implant losing 1 mm MBL after the first year from implant placement and 0.2 mm in the following years was considered successful.⁴⁰ 

The remodeling process around an implant can be classified in 2 phases; early and late remodeling. Early bone remodeling occurs in the first year of loading and consists of the remodeling of the woven bone. On the other hand, late remodeling occurs up to 5 years after loading and consists of the maturation and mineralization and remodeling of the bone.⁴¹  After both healing processes have taken place, the number of osteocytes is reduced and further bone remodeling is impaired.³⁴  Several theories have been proposed regarding marginal bone loss. A recent theory proposed by Albrektsson et al⁴¹  reported that some resorption occurs after the first year of implant placement due to bone remodeling in response to surgery. The marginal bone loss after the first year of loading is related to a complication with implant placement such as surgery or prosthodontics, and is in most cases unrelated to infectious disease reinforced by immunological reactions and adverse biomechanical situations.⁴¹  Factors influencing early bone resorption have been described as reflection of the periosteum during surgery, high insertion torque, microgap formation and consequent bacterial invasion, biological width establishment, and occlusal overload,⁴²  as well as patient-related factors such as smoking,⁴³  periodontal disease, consumption of pharmaceuticals, and bone quality.^34,41 

Adequate insertion torque (IT) has been suggested to prevent micromovements that could lead to fibrous encapsulation (early failure). However, a high IT (>50) is associated with an increase in pressure, enhancing bone necrosis and leading to greater bone loss.³⁴ 

Another hypothesis of primary peri-implant MBL that has been proposed is the consequence of the biological width adaptation, which has been reported to be 3.5 mm around implants.³⁴  Additionally, it has been suggested that in cases of 2-piece implants, the microgap (10–50 μm) between the implant and abutment facilitates the colonization of the implant with pathogens, leading to an inflammatory process that triggers osteoclast activity and leads to MBL.³⁴  The sealing of the microgap can be reduced with an internal implant connection.³⁴  Additionally, the microgap has been reported to initially produce an inflammatory reaction, followed by bone loss and, finally, peri-implantitis.³⁴  Therefore, this factor can be associated with early and late remodeling/bone loss.

In our study, the mean MBL measured from the implant shoulder to first radiographic bone-to-implant contact after 1 year of loading was −0.56 mm (SD ± 0.51) for the short implant group and −0.67 mm (SD ± 0.23) for the regular-length implant group. Of the 40 implants inserted, 10% of the short implants and 10% of the regular-length implants presented bone loss exceeding 1 mm. Therefore, the success rate of each implant length after 12 months was 90%. However, in this period mean bone gains +0.29 mm (SD ± 0.39) and +0.19 mm (SD ± 0.38), respectively, were observed. The overall results are however negative, due to the greater bone loss that occurred between implant placement and prosthetic delivery. The initial bone loss occurring during the 3 months after placement can be explained as a result of early remodeling. Here, patient-related factors, such as smoking, periodontal disease, and pharmaceutical consumption, were excluded from this study. Additionally, no type IV bone was present in any implant. Nonetheless, in this study, a bone gain occurred after the implants were loaded. This can be explained as a response to mechanical stress that strengthens the bone by increasing the bone density or apposition of bone.⁴² 

The reaction of bone to mechanical stress is explained as an adaptation of the bone by becoming more mineralized and dense.⁴²  With a slight strain, the bone becomes mildly overloaded and compensates by forming more bone. However, there is a threshold that exceeds the bone capacity and as a response, the bone can fatigue and fracture, leading to bone loss.⁴²  However, a moderately increasing continuous load reestablishes new optimal strains due to the mineralization of the bone.⁴²  During late remodeling, maturation and mineralization of the bone occurs throughout the 5 years post-placement.⁴¹  Nonetheless, occlusal overload, the presence of a microgap, bacterial invasion, and patient-related factors can lead to greater late marginal bone remodeling and, consequently, greater bone loss.³⁴  It has been reported that most of the peri-implant bone loss changes occur during the first year after implant placement.^38,41  However, in a review including internal connection implants with SLA surfaces, implants presented a MBL of −0.2 to −0.3 mm in the first year after placement.

In a further RCT, 30 patients were evaluated in a split-mouth design study. Fifteen patients received implants in the maxilla and 15 in the mandible.³²  One year after loading in the mandible, the mean MBL was -1.2 mm (SD ± 0.49) in both the short- and regular-length implant groups (SD 0.49 and 0.47, respectively). However, it should be emphasized that the regular-length implants were placed in augmented areas. Three years after loading, patients with mandibular implants lost on average −1.44 mm (SD 0.44) for short implants and −1.63 mm (SD ± 0.52) in the regular-length implant group. In an RCT performed by the same group, 60 partially edentulous patients receiving implants in the posterior mandible presented a mean MBL of −1 mm after 1 year of loading.³¹  A further report observed MBL of −0.22 ± 0.4 mm for short implants after 1 year of loading.⁸  In a clinical study, the MBL evaluated after 3 years of loading in the mandible was, on average, −1.44 mm for short implants and 1.63 mm for regular-length implants. This difference was not statistically significant (P = .059, 95% CI −0.01, 0.49). In maxillae, patients lost on average 1.02 mm for short implants and 1.54 mm for long implants. This difference was statistically significant (P = .001 95% CI 0.21, 0.60).³¹ 

Concerning other biological complications, most of them are patient-based and can be minimally influenced by modification of the implants.⁷  However, a systematic review reported that regular-length implants in augmented areas present more biological complications than short implants.⁶  Nevertheless, none of the implants included in this clinical study presented any biological complication. However, complications in short implants, such as peri-implantitis, mucositis, lack of osseointegration, greater marginal bone loss, hyperplastic peri-implant tissues, temporary mental nerve dysunction, and others have been described in the literature.^8,22,44  In a retrospective clinical study of 111 evaluated short implants, 1 short implant presented peri-implantitis and had to be removed. The authors suggest that the thin gingival biotype and an excessive accumulation of plaque due to poor hygiene could have been the cause of this perimplantitis.⁴⁵ 

Our secondary objective was to compare the soft tissue parameters between short and regular-length implants. These parameters, such as BOP, PD, and CAL represent important elements of implant diagnostics, and have been included in the success criteria of implants. However, a recent review reported that periodontal indices such as bleeding on probing and probing depth are irrelevant diagnostic tools in the evaluation of implants and that these should be avoided, as they cause unnecessary trauma to the peri-implant tissues.⁴¹  In this clinical study, no statistical significant differences were found in terms of soft tissue parameters between short and regular-length implants. Additionally, our measurements did not traumatize or affect the peri-implant tissues. Comparing these results with existing literature is quite difficult, since many clinical studies do not report soft tissue outcomes (Figure 6).

Another objective was to assess the success of implants in relation to their C/I ratio. In our study, no association between C/I ratio and MBL was found. The consideration that the crown-to-root ratio of teeth influences their survival is clear, but this statement has been extrapolated to implants, which lack a periodontal ligament.¹⁴  However, it has been reported that implants are more vulnerable to nonaxial forces. Furthermore, some authors suggested that a poor C/I ratio could lead to excessive occlusal loading, and that nonaxial loading could act like a lever arm, creating a bending moment that could lead to increased technical and biological complications.^42,46,47 

Usually in atrophied mandibles, the use of short implants is associated with greater crown height spaces due to the greater interarch distance. Poor bone quality has also been associated with a high C/I ratio.⁴⁸  It has been reported that C/I ratios greater than 2 do not affect the MBL around short implants,^44,47  and may even positively affect via marginal bone gain, which has been reported in the literature.⁴⁶ 

In a clinical prospective study, short implants with a length of 6.5 mm and differing C/I ratios were compared.⁴⁴  Fifty-one patients were restored with 65 short implants. Two groups were compared: 1 group with C/I ratios ≥ 2, and the other with C/I ratios < 2. MBL was not associated with higher C/I ratios. Additionally, biological complication rates were not significantly different, whereas prosthetic complications were more frequent for C/I ratios ≥ 2. However, these results were not statistically significant.⁴⁴  In this study, the authors conclude that abutment connection, specifically internal connections, could resist eccentric loading complexes and bending moments, ensuring a mechanical stability and reducing prosthetic complications at the implant–abutment interface.⁴⁴  A retrospective clinical study reported mean C/I ratios of 1.4 and ranging from 0.9 to 2.5,⁴⁵  and reported no association between C/I and MBL.⁴⁵  Additionally, in this clinical study, the clinical C/I ratio increased over time, from 1.5 upon prosthetic delivery to 1.8 after 2 years of loading.⁴⁹ 

Finally, our tertiary objective was to compare the prosthetic outcomes between short- and regular-length implants. Although one chipping occurred in the regular-length implant group, no statistically significant differences were found in terms of survival and success rates, regarding technical complications of the prosthesis supported by regular-length implants. The complication rate of fixed dental prostheses (FDPs) reported, in a systematic review of studies including a minimum of 5-year follow-up was 27.3% (95% CI 19.8–36.9).⁷  Screw loosening in FDPs occurred with a frequency of 4 % (95% CI 2.2–7.0), screw fracture at a frequency of 0.8% (0.4–1.6), and fracture of the veneering material at a frequency of 1.6% (0.77–3.3). The authors explain the fractures of the veneering material as a result of the delicate types of zirconia- or titanium-based FDPs.⁷  They also conclude that the risk of fracture of the veneering material increases with the size of the reconstruction.⁷ 

In the literature, technical complications reported concerning short implants were abutment or screw loosening, abutment fracture, loss of retention of the restoration, and porcelain fractures.^7,22  In a retrospective study including 231 short implants restored with single crowns, 2.6% of porcelain chipping was reported.²²  These results should not be compared with our study, because in our study we decided to splint the implants, and therefore cannot conclude whether splinting and non-splinting are comparable. However, several factors must be taken into account when considering whether to splint, for example: implant length, occlusion, hygiene, abutment connection design, and difficulty achieving a passive fit of the framework.⁴⁸  Some authors report that splinting implants helps to distribute functional loads and therefore reduces MBL specifically in short implants.^48,50  However, nonsplinted implants are preferable for many patients, being easier to floss and having improved aesthetics due to their more individual aspect and natural appearance.⁵⁰  Likewise, some authors report that by avoiding splinting, a passive fit problem can be prevented, as well as an adequate emergence profile and desirable interproximal hygiene facilitated.⁵⁰  However, there is still no consensus that describes when is indicated splinting or not, since both designs show advantages.

Several technical complications, such as screw loosening, screw breakage, or porcelain fracture have been reported in other clinical split-mouth studies comparing splinted and nonsplinted restorations on short implants.⁴⁸ 

In another clinical split-mouth study, no statistically significant difference in terms of MBL comparing splinted versus nonsplinted short dental implants was observed.¹⁷  However, lower success rates were reported for the non-splinted group, particularly on implants shorter than 10 mm.¹⁷  For this reason, in our study we decided to splint the implants in order to exclude possible effects of the substructure on the primary objective.

To explain this chipping, additional risk factors for porcelain chipping reported in other studies include fabrication error, bruxism, porcelain type, and opposing occlusion.⁵¹  Considering fabrication error, non-anatomical framework design, wrong firing temperature, excessive cooling speeds resulting in tensile stress in the ceramic, no cooling during occlusal adjustment, or no polishing after occlusal adjustment are some of the fabrication factors that influence chipping.⁵¹  Technical complications are mostly related to the materials and the design of the components.⁷  This event occurred at a frequency of 5% after one year. The framework was a nonprecious-alloy metal, which could be a risk factor, since some authors report that nonprecious alloy reconstructions are more difficult to produce.⁵¹ 

Within the limits of this study, short implants showed similar clinical and radiographic outcomes to those of regular-length implants after 1 year. Long-term clinical studies in a well-controlled design are needed to validate the clinical performance of short implants.

Abbreviations

REC

recession

SCs

single crowns

BIC

bone-to-implant contact

BOP

bleeding on probing

C/I

crown-to-implant ratio

CAL

clinical attachment level

CBCT

cone beam computerized tomography

FDP

fixed dental prosthesis

GBR

guided bone regeneration

IT

insertion torque

MBL

marginal bone loss

PD

probing depth

RCT

randomized controlled clinical trials

SLA

sandblasting with large-grit and acid etching

The authors thank the company Dentium for sponsoring this study and providing the required materials.

The authors declare no conflict of interest.