Literature review

528 THE JOURNAL OF PROSTHETIC DENTISTRY VOLUME 87 NUMBER 5

system have been relatively successful, screw loosening continues to be a frequent complication.2-4 Sutter et al5 suggested that an internally connected abutment design transfers occlusal loads to the implant body, with less impact on the abutment screw threads. Several studies have reported success with the use of these systems for single-tooth restorations.6-8

The use of individual implant-supported restora- tions can facilitate and simplify laboratory procedures. In contrast, the fabrication of passive splinted frame- works with conventional methods has been only moderately successful.9 Several techniques for mini- mizing framework misfit for multiple units are available, such as sectioning and soldering, specialized impression techniques, selective internal adjustments, electrical discharge machining (spark erosion), and

Effect of splinting and interproximal contact tightness on load transfer by implant restorations

David L. Guichet, DDS,a Diane Yoshinobu, DDS,b and Angelo A. Caputo, PhDc School of Dentistry, University of California, Los Angeles, Calif.

Statement of problem. To circumvent the difficulty of achieving a passive framework fit, some authors have suggested that multiple adjacent implants be restored individually. This protocol requires that each unit be able to withstand mastication forces. Non-splinted restorations have numerous interproximal con- tacts that require adjustments prior to placement, with an unknown outcome relative to load transfer. Purpose. This in vitro simulation study examined the effect of splinting and interproximal contact tight- ness on passivity of fit and the load transfer characteristics of implant restorations. Material and methods. A photoelastic model of a human partially edentulous left mandible with 3 screw-type implants (3.75 × 10 mm) was fabricated. For non-splinted restorations, individual crowns were fabricated on 3 custom-milled titanium abutments. After the units were cemented, 5 levels of interproxi- mal contact tightness were evaluated: open, ideal (8 µm shim stock drags without tearing), light (ideal +10 µm), medium (ideal + 50 µm), and heavy (ideal + 90 µm). For splinted restorations, five 3-unit fixed partial dentures were fabricated, internally adjusted with silicone disclosing material, and cemented to the model. Changes in stress distribution under simulated non-loaded and loaded conditions (6.8 kg) were analyzed with a polariscope. Results. In the simulated alveolar structures, non-splinted restorations with heavier interproximal con- tacts were associated with increased tensile stresses between implants; occlusal loads tended to concentrate around the specific loaded implant. Splinted restorations shared the occlusal loads and distributed the stresses more evenly between the implants when force was applied. The load-sharing effect was most evi- dent on the center implant but also was seen on the terminal abutments of the splinted restorations. Conclusion. The results of this in vitro study suggest that excessive contact tightness between individual crowns can lead to a non-passive situation. In this experiment, splinted restorations exhibited better load sharing than non-splinted restorations. (J Prosthet Dent 2002;87:528-35.)

Several prosthetic options are available for the restoration of multiple adjacent implants. Because it is difficult to fabricate a passively fitting prosthesis on multiple implants, some authors have suggested that adjacent implants be restored individually.1 It is believed that separating adjacent units allows the resulting restorations to seat passively. Although sin- gle-tooth restorations supported by an external hex

CLINICAL IMPLICATIONS

Within the limitations of this study, the results suggest that meticulous verification of interproximal contact tightness should be performed to ensure the passive fit of non- splinted restorations. If compromises exist in the position, the load-carrying capacity and/or biomechanical stability of the proposed implant restorations may be affected. Splinting may foster the reduction or sharing of loads and is therefore recommended in these circumstances.

Presented at the annual meeting of the Pacific Coast Society of Prosthodontics, Seattle, Wash., June 29-July 1, 2000.

aLecturer, Division of Advanced Prosthodontics, Biomaterials, and Hospital Dentistry.

bStaff Prosthodontist, VA Medical Center, West Los Angeles, Calif. cProfessor of Biomaterials Science, Division of Advanced

Prosthodontics, Biomaterials, and Hospital Dentistry.

laser welding. All of these procedures are technique- sensitive, and results are limited by the accuracy of the indexing method. It has been suggested that adhesive- based correction procedures, in which machined cylinders or discs are bonded underneath implant frameworks, can improve the fit of the suprastructure to the implants.10 Cement-retained fixed partial den- tures (FPDs) have been shown to achieve better passivity than screw-retained FPDs.11 Non-splinted restorations facilitate oral hygiene and may engender less stress development in the restoration and alveolar structure during mandibular flexure.12-15

Splinting generally has been used to stabilize mobile teeth and restore edentulous spaces.16 With the excep- tion of Ante’s law,17 which has been questioned in several studies,18,19 specific clinical guidelines for splinting are lacking. Implant-supported restorations are splinted primarily to aid in the distribution of occlusal forces.20 Bending overload has been shown to lead to coronal bone loss, as well as screw and implant fracture.21-24 Based on clinical findings, finite element models, and photoelastic studies, it has been recom- mended that multiple implants be rigidly connected.25-28 Limited experimental models have demonstrated that splinting abutments can reduce the occlusal forces transferred to the periodontium and concentrate tensile and shear stresses in the connector regions.29-33

The restoration of multiple units with single-tooth implants is an attempt to circumvent the problem of achieving passive fit with splinted restorations. However, the restoration of individual adjacent implants requires careful adjustment of interproximal contacts. Campagni34 suggested that, for natural den- tition, interproximal contacts should be adjusted to allow an 8-µm metallic shim to drag without tearing. Other authors have been less specific about the method and measurement needed to establish ideal contact.35,36 Without the aid of the periodontal liga- ment around implants, contact adjustment may be more critical.

What constitutes an acceptable level of fit varies among clinicians,37 and the biological and mechanical effects of misfit have been debated. Retrospective clin- ical and animal studies failed to demonstrate any differences in bone response with varying degrees of prosthetic misfit.38,39 Rubin and Lanyon40 hypothe- sized that because no evidence of new bone formation was found under static loads, the bone essentially ignored constant loads as osteoregulatory stimuli. Adaptive bone remodeling did occur in animal ulnas under dynamic loading. The authors concluded that bone is “genetically programmed” to accept a particu- lar level and distribution of functional strain within the bone. Deviations may result in an adaptive increase or decrease in the bone mass.

GUICHET, YOSHINOBU, AND CAPUTO THE JOURNAL OF PROSTHETIC DENTISTRY

MAY 2002 529

Static strains from misfitting prostheses may not directly cause adverse bone reactions, but non-passive restorations may magnify dynamically applied occlusal loads (additive force). Studies have shown that load magnification leads to bone loss.22-24 In animal stud- ies, excess occlusal heights of 180 µm or more (occlusal overload) led to an increase in bone loss around implants with controlled oral hygiene.41,42 Skalak20 reported that misfit may fatigue implant com- ponents and reduce failure loads.

The purpose of this in vitro simulation study was to examine the effect of splinting and interproximal con- tact tightness on passivity of fit and the load transfer characteristics of implant restorations.

MATERIAL AND METHODS

A quasi-3-dimensional photoelastic stress analysis of the mandibular left posterior quadrant was designed for this study.43 An anatomically correct skeletal form was used to create a master cast, mold, and photoelastic resin “patient cast.” Three 3.75- × 10-mm screw-type implants (Nobel Biocare, Goteborg, Sweden) were embedded in the first and second premolar and first molar positions to simulate complete integration. The anterior implant was placed anatomically with a 6% mesial inclination to parallel the canine tooth angulation. A medium-modulus photoelastic resin designed to simulate healthy bone was used (PL-2; Measurements Group Inc, Raleigh, N.C.).

Accepted clinical and laboratory procedures were used to fabricate and deliver the restorations. An anatomically correct wax gingival form overlaid the photoelastic resin to simulate the thickness of gingival tissues during impression-making procedures. Implant-level pick-up impression posts (EP Square Impression Copings; Implant Innovations Inc, Palm Beach Gardens, Fla.) were oriented on the implants. A custom tray was fabricated to adapt loosely to the impression posts, and a light-bodied vinyl polysilox- ane impression was made (Extrude; Kerr, Romulus, Mich.). Implant analogs were secured in the impres- sion, and silicone that represented soft tissue (GI Mask; GC America Inc, Alsip, Ill.) was added to the internal aspect of the impression surrounding the implant analog necks. The impression was poured in vacuum-mixed die stone (Die Keen; Heraeus Kulzer, South Bend, Ind.), and a working cast was created. Teeth were fashioned in inlay wax in an anatomically correct manner with a 1-mm buccal cutback for porcelain application. A condensation silicone mold (Coltene/Whaledent, Mahwah, N.J.) was placed over the FPD wax pattern to allow for multiple replica- tions.

The single-unit and FPD designs studied were cement-retained restorations based on 2-piece abut-

ment posts (EP; Implant Innovations Inc). The abut- ments were delivered to the working cast and modified according to the manufacturer’s recommendations to allow for occlusal and axial clearance, cervical emer- gence, taper, and path of draw. A uniform total taper of 6 degrees was achieved with a precision-milling machine (F-1; Ney/Degussa, Bloomfield, Conn.). The abutments were polished to a medium-high finish with rubber wheels. Resin copings (Pattern Resin; GC Corp, Tokyo, Japan) were made for each abutment without the use of a die spacer. The copings were posi- tioned on the abutments and then assembled in the silicone mold. Wax was subsequently injected into the mold to form the wax patterns. Five replications were completed for the FPDs and one set for the single-unit restorations.

An implant cast verification index was fabricated to verify the implant orientation and abutment relation- ships and to orient the individual restorative units prior to investing and casting. The index was fabricat- ed as follows: Three square implant-level impression posts were secured to the stone cast, joined with den- tal floss and acrylic resin (Relate; Parkell, Farmingdale, N.Y.), and placed in boiling water for 10 minutes prior to sectioning. After the solid index was separated, lead film foil was placed between the sections, and acrylic resin was applied on each side of the foil. The separator was removed, and the 3 sec- tions were boiled for an additional 10 minutes to ensure complete polymerization. The 3 resin indices were placed without contact on the photoelastic model, and the thin separation was luted with cyano- acrylate gel and accelerator (Rocket-heavy; Dental Ventures of America Inc, Corona, Calif.). Implant replicas were added to the impression posts, and the verification assembly was seated in low-expansion mounting stone (Whip Mix Corp, Louisville, Ky.) The stone was allowed to harden. The index was verified visually with the 1-screw test (Scheffield fitting test),9

THE JOURNAL OF PROSTHETIC DENTISTRY GUICHET, YOSHINOBU, AND CAPUTO

530 VOLUME 87 NUMBER 5

and stress analysis testing on the photoealstic model was used to determine that the index accurately pro- duced the implant relationships.

FPD wax patterns were sectioned interproximally and rejoined on the implant verification index with cyanoacrylate gel (Zap-it; Dental Ventures of North America) prior to spruing and investing. For the non- splinted, single-unit restorations, contacts between the implants were separated and waxed to form a “heavy” contact that would allow polishing procedures. Margins (margin wax; Metalor, North Attleborough, Mass.) were finished under ×10 microscopy.

The investment, burnout, and casting techniques were standardized. Patterns were sprued and invested individually in a phosphate-bonded investment (Cera- Fina; Whip Mix Corp) with the casting technique described by White.9 After bench polymerization and burnout, the investment rings were cast in a gold- palladium alloy (Silhouette 550SL; Argen Alloys Inc, San Diego, Calif.). Devesting was completed in the usual manner with minimum use of aluminum oxide air abrasives on critical interfaces. Burs were used under laboratory microscopes to eliminate internal casting inaccuracies.

Further internal adjustments were made by painting a thin layer of die lubricant (Belle de St. Clair, Chatsworth, Calif.) on the abutments and removing any wet, shiny areas on the lightly air-abraded internal surfaces of the castings.44 Adjustments were made until the best seating of the castings was achieved and confirmed under microscopy. The cement-retained FPDs were adjusted as described above with a silicone disclosing medium (Fit-Checker; GC Corp) (Fig. 1).

The single-unit and FPD restorations were deliv- ered to the patient model with a standard protocol. The specimens were seated with zinc oxide–eugenol temporary cement (Temp-Bond; Kerr) under a 4.5-kg load for 1 minute followed by a 0.9-kg load for 2 min- utes and then bench set. Excess cement was removed prior to testing. Each specimen was evaluated under the same conditions.

Prior to any evaluation, the photoelastic resin cast was determined to be free of residual stress. The model was immersed in mineral oil to minimize surface refraction and then placed in the field of a circular polariscope (Measurements Group Inc), as described previously.28 The non-splinted, single-unit restora- tions were tested for stress generation within the supporting structures with varying contact tightness. After cementation of the non-splinted units, inter- proximal contacts were adjusted by placing metallic shims (Shim stock; Almore International Inc, Beaverton, Ore.) of varying thickness into the inter- proximal spaces to simulate different degrees of contact tightness. The following thicknesses were

Fig. 1. Silicone disclosing medium used to identify interfer- ences for internal adjustment of fixed partial denture.

GUICHET, YOSHINOBU, AND CAPUTO THE JOURNAL OF PROSTHETIC DENTISTRY

MAY 2002 531

used: open, ideal (8-µm tin foil shim could drag between contacts without tearing) (Fig. 2), light (ideal + 10 µm), medium (ideal + 50 µm), and heavy (ideal + 90 µm). Stresses generated by the various degrees of contact were observed and photographed within the polariscope.

Vertical loads of 6.8 kg were applied in a straining frame by means of a calibrated load cell mounted on the movable head of a loading frame. Loads were monitored with a digital read-out (Models 2130 and 2120A; Measurements Group Inc). For each of the FPDs and contact conditions, the loading point loca- tions were over the anterior, middle, and posterior implants. The model was immersed in a tank of min- eral oil to minimize surface refraction and thereby facilitate photoelastic observation. The resulting stresses in all areas of the supporting structure were monitored and recorded photographically in the field of the circular polariscope. Each loading and observa- tion sequence was repeated at least 2 times to ensure reproducibility of the results. The model was allowed to rest between each test series to ensure that the observed stresses were not residual but the result of applied conditions.

RESULTS

The results for non-loaded test conditions are pre- sented in Figure 3. No stresses were observed in non-splinted restorations with open contacts. A low level of coronal stress (<1⁄2 fringe order) developed between the implants when the interproximal contacts were adjusted to an ideal contact. With the addition of a 10-µm shim in both interproximal spaces (light con- tact), coronal stresses became more evident among the 3 implants (1⁄2 fringe order). With a 50-µm shim (medi- um contact), an increase in stress among the implants was identified (1 fringe order), as were stresses along the implant threads and apices (1⁄2 fringe order). Placement of a 90-µm shim (heavy contact) resulted in

high levels of stress in the coronal region (11⁄2 fringe orders) and increased stress along the implant threads and apices (1 fringe order).

Splinted FPDs generally exhibited low stress levels in the alveolar structure. The FPDs were ranked in order of increasing stress (Fig. 3). After cementation, FPDs 1 and 2 were essentially free from stress, with the exception of low interproximal stress between the dis- tal implants (<1⁄2 fringe order). FPDs 3 and 4 exhibited

Fig. 2. Eight-µm tin foil shim dragged between interproximal contacts without tearing to establish ideal contact tightness.

Fig. 3. Non-loaded data. Non-splinted restorations trans- ferred increased stress as contact tightness increased. Open = least stress, heavy = most stress. Splinted restora- tions transferred stresses generated upon placement. FPDs 1 and 2 exhibited least stress; FPD 5 exhibited most stress.

THE JOURNAL OF PROSTHETIC DENTISTRY GUICHET, YOSHINOBU, AND CAPUTO

532 VOLUME 87 NUMBER 5

low levels of stress along the distal threads of the ante- rior implant and interproximally between the posterior implants (1⁄2 fringe order). Only FPD 5 exhibited a sig- nificant amount of stress (1 fringe order) in the same areas mentioned for FPDs 3 and 4. Most of the FPDs consistently demonstrated low stress distribution.

The results for loaded test conditions are presented in Figures 4 through 6. For each of the FPDs and con- tact conditions, peak stresses were manifested under the loaded implant; remote stress concentrations were

of equal or lesser intensity. For this reason, the loaded implant was isolated for analysis.

Anterior implant data are presented in Figure 4. For both splinted and non-splinted restorations, stress concentrated on the mesial aspect of the implant (Fig. 4). In the non-splinted group, stress increased as con- tact tightness increased (open to heavy = 2-3 fringe orders). Splinted restorations exhibited the same load- ing characteristics but with slightly less mesial stress (FPDs 1 and 3-5 = 11⁄2 fringe orders, FPD 2 = 2 fringe orders). Splinted restorations also distributed loads to

Fig. 4. Anterior load, 6-degree mesial implant inclination. Non-splinted restoration concen- trated stress on mesial aspect of loaded implant. Stress increased as contact tightness increased. Splinted restorations exhibited same loading characteristics but with slightly less mesial stress.

Fig. 5. Middle load, non-splinted restoration. Stress concentrated around loaded implant and exhibited higher peak stress than in splinted group. Splinted restorations exhibited load shar- ing and transferred significantly less stress to loaded middle implant.

GUICHET, YOSHINOBU, AND CAPUTO THE JOURNAL OF PROSTHETIC DENTISTRY

MAY 2002 533

the distal aspect of the implant in the coronal half, but to a lesser degree than on the mesial aspect (FPD 2 = 1⁄2 fringe order, FPDs 1 and 3-5 = 1 fringe order). Stresses were distributed slightly more evenly in the splinted group than in the non-splinted group.

Middle implant data are presented in Figure 5. In the non-splinted group, a high level of stress (11⁄2 fringe orders) was identified along the threads and alveolar structure of the restoration with open con- tacts. Lower-intensity stresses were seen on the anterior and posterior implants (not depicted). With increasing interproximal contact tightness (open to heavy), peak stress levels continued to concentrate along the mesial aspect of the loaded implant (2 fringe orders). Stresses around the anterior and posterior implants also increased in complexity, but at a com- paratively reduced intensity. Conversely, in the splinted group, less force was transferred to the supporting structure. Stresses were distributed more evenly among the 3 implants, and all 5 FPDs exhibited low stress levels (1 fringe order).

Posterior implant data are presented in Figure 6. In the non-splinted group, high levels of stress concen- trated along the loaded implant. Less intense stress was identified in the anterior and middle implants. As interproximal contact tightness increased, an associat- ed shift in peak stress from the mesial aspect to the distal aspect of the loaded implant was observed (open to medium = 11⁄2 fringe orders). Only in the heavy-con- tact group were the peak stresses significantly higher on the distal aspect of the coronal portion of the

implant (21⁄2 fringe orders, the highest peak stress observed in this study). In the splinted group, occlusal loads applied to the posterior implant were not shared as evenly as loads applied to the middle implant. Stresses tended to concentrate on the coronal aspect of the posterior implant (11⁄2-2 fringe orders) but were not as intense as those associated with the non-splint- ed, heavy-contact posterior implant.

DISCUSSION

A passively fitting prosthesis has been considered a prerequisite for the success and maintenance of osseointegration. Passivity is a particular concern with multiple implants because of documented inaccuracies in the casting and soldering process. One way to avoid this problem is to restore the implants individually. High success rates with single-tooth implants and altered abutment/implant connections have fostered confidence in the individual-restoration protocol, which eliminates the need for procedures designed to minimize framework misfit.

The individual restoration of multiple adjacent implants also minimizes component loosening and/or fractures as the mandible undergoes flexure and tor- sion during function.12 With a finite element model of the human mandible, Korioth and Hannam13 record- ed a range of deformation (0.46 to 1.06 mm) during simulated tooth clenching. Hobkirk and Schwab14 observed a medial deflection of the mandible of up to 420 µm during active opening and protrusive jaw movements. Reports of significant amounts of

Fig. 6. Posterior load, non-splinted restoration. As contact tightness increased, higher stresses were noted. This was most evident in distal aspect of implants in heavy tightness group. Splinted restorations shared stresses to lesser degree than seen in middle load group (Fig. 4). Character of stress distribution in FPDs was similar, and magnitude was of equal or slightly greater value than seen in medium contact tightness condition.

mandibular flexure have engendered concerns about potentially harmful forces along the implant/bone interface with rigidly connected superstructures. Fischman15 suggested that a complete-arch restoration be avoided in favor of freestanding, short-span, implant-supported prostheses. Individually restored implants enable better oral hygiene access and improved axial and interproximal contours.

Multiple single-unit restorations do have disadvan- tages. This restorative design may not be possible if each implant is not placed in the correct location and angulation. A lack of hard tissue and anatomic land- marks may prevent appropriate spacing and the use of an adequate number of implants. Moreover, although individual restoration may simplify laboratory proce- dures, the adjustment of interproximal contacts with multiple adjacent implants is difficult. Fixed prostho- dontic textbooks suggest that floss or articulating paper be used to adjust and evaluate proximal con- tacts.35,36 These techniques are variable and do not allow a quantitative evaluation of contact tightness. Without the presence of periodontal ligaments around implants, the adjustment of interproximal contacts becomes more critical.

Campagni34 suggested that interproximal contacts be modified until an 8-µm metallic shim can be dragged through the contact without tearing. In the present study, little to no stress developed within the supporting structure when the contacts were adjusted accordingly (Fig. 3). When a 10-µm shim was placed between the restorations (light contact), stresses developed in the coronal area between the implants. Higher levels of stress developed in the coronal areas and along the implant threads when medium and heavy contacts were tested. These data suggest that the individual restora- tion of multiple adjacent implants does not necessarily result in a passively restored prosthesis. Interproximal contacts must be carefully adjusted without creating an open contact, which increases in difficulty with a larger number of units and a screw-retained design.

Non-splinted restorations also must withstand mas- ticatory loads individually. In the present study, high levels of stress were concentrated on the loaded implant, with little transfer of load to the neighboring implants. Increased contact tightness was associated with increased stress intensity along the loaded implant (Figs. 4 and 5). Implant-supported FPDs are splinted to foster the distribution of occlusal forces and to pre- vent the transfer of detrimental force levels to the supporting implants, which may lead to bone resorp- tion and component failures.21-24 When multiple units are splinted, occlusal forces are “absorbed” within the framework: tensile and shear stresses are concentrated in the connector regions, which reduces the force transferred to the periodontium.32,33 A comparison of the non-splinted and splinted data in Figure 4 illus-

trates this phenomenon. It should be noted that forces were not as evenly distributed when anterior or poste- rior abutments were loaded as when middle implants were loaded (Fig. 5).

Occlusal stresses tended to concentrate on the loaded implant; these stresses were similar in magni- tude (11⁄2 fringe orders) to those observed for non-splinted restorations with either ideal or light interproximal contacts. These data suggest that the effect of splinting is limited, which is consistent with previous findings. An insubstantial cross-arch distribu- tion of load limited to 2 to 3 teeth or the implants closest to the applied load has been reported for splint- ed prostheses.29-31

Splinting may have a beneficial effect when off-axis implants are restored. The anterior implant evaluated in the present study had a 6% mesial inclination. When non-splinted implants were loaded, stress concentrated on the mesial threads. As contact tightness increased, the stress intensified. In the splinted group, loading was associated with stress concentrations in the same locations, but at a lower magnitude.

The design of the present study minimized the con- founding variables associated with animal and human studies. The applicability of the results is limited, how- ever, by the fact that photoelastic data were interpreted visually. It should be noted that all modeling systems (including finite element analysis, mathematical mod- els, and strain-gauge analysis) are limited when the biologic system is studied. Long-term clinical studies are needed to determine whether varied masticatory loads on non-splinted and splinted restorative designs affect implant survival. The decision to splint implants may be more important when no anterior guidance exists or when the patient has parafunctional habits. Implants placed in poor-quality bone or grafted areas also may benefit from splinting.

CONCLUSIONS

Within the limitations of this photoelastic stress analysis study, the following conclusions were drawn:

1. Splinting effectively reduced peak stresses in the loaded middle implant of a 3-unit FPD. When restora- tions were not splinted, stresses concentrated around the loaded implant.

2. Load sharing by splinted restorations was least significant when loads were applied to the terminal posterior implant of the FPDs. Axial loads in splinted and non-splinted restorations were similar.

3. Load sharing by splinted restorations was observed when loads were applied to the anterior ter- minal abutment, which was positioned slightly off-axis. This result suggests that splinting may be effective with off-axis implants.

4. As interproximal contact tightness increased, pas- sivity decreased. Heavier interproximal contact

THE JOURNAL OF PROSTHETIC DENTISTRY GUICHET, YOSHINOBU, AND CAPUTO

534 VOLUME 87 NUMBER 5

implant fracture: a retrospective clinical analysis. Int J Oral Maxillofac Implants 1995;10:326-34.

25. Wright KW, Yettram AL. Reactive force distributions for teeth when loaded singly and when used as fixed partial denture abutments. J Prosthet Dent 1979;42:411-6.

26. Stegaroiu R, Sato T, Kusakari H, Miyakawa O. Influence of restoration type on stress distribution in bone around implants: a three-dimensional finite element analysis. Int J Oral Maxillofac Implants 1998;13:82-90.

27. Landry KE, Johnson PF, Parks VJ, Pelleu GB Jr. A photoelastic study to determine the location of the nonrigid connector in a five-unit interme- diate abutment prosthesis. J Prosthet Dent 1987;57:454-7.

28. Davidoff SR. Restorative-based treatment planning: determining ade- quate support for implant-retained fixed restorations. Implant Dent 1996;5:179-84.

29. White SN, Caputo AA, Anderkvist T. Effect of cantilever length on stress transfer by implant-supported prostheses. J Prosthet Dent 1994;71:493- 9.

30. Wylie RS, Caputo AA. Fixed cantilever splints on teeth with normal and reduced periodontal support. J Prosthet Dent 1991;66:737-42.

31. Itoh H, Caputo AA, Wylie R, Berg T. Effects of periodontal support and fixed splinting on load transfer by removable partial dentures. J Prosthet Dent 1998;79:465-71.

32. Yang HS, Lang LA, Felton DA. Finite element stress analysis on the effect of splinting in fixed partial dentures. J Prosthet Dent 1999;81:721-8.

33. el-Ebrashi MK, Craig RG, Peyton FA. Experimental stress analysis of den- tal restorations. VII. Structural design and stress analysis of fixed partial dentures. J Prosthet Dent 1970;23:177-86.

34. Campagni WV. The final touch in the delivery of a fixed prosthesis. CDA J 1984;12:21-9.

35. Shillingburg HT, Hobo S, Whitsett LD, Jacobi R, Brackett SE. Fundamentals of fixed prosthodontics. 3rd ed. Chicago: Quintessence; 1997. p. 391-2.

36. Rosenstiel SF, Land MF, Fujimoto J. Contemporary fixed prosthodontics. 3rd ed. St. Louis: Mosby; 1998. p. 747-9.

37. Kan JY, Rungcharassaeng K, Bohsali K, Goodacre CJ, Lang BR. Clinical methods for evaluating implant framework fit. J Prosthet Dent 1999;81:7- 13.

38. Jemt T, Book K. Prosthesis misfit and marginal bone loss in edentulous implant patients. Int J Oral Maxillofac Implants 1996;11:620-5.

39. Carr AB, Gerard DA, Larsen PE. The response of bone in primates around unloaded dental implants supporting prostheses with different levels of fit. J Prosthet Dent 1996;76:500-9.

40. Rubin CT, Lanyon LE. Kappa Delta Award paper. Osteoregulatory nature of mechanical stimuli: function as a determinant for adaptive remodel- ing in bone. J Orthop Res 1987;5:300-10.

41. Miyata T, Kobayashi Y, Araki H, Motomura Y, Shin K. The influence of controlled occlusal overload on peri-implant tissue: a histologic study in monkeys. Int J Oral Maxillofac Implants 1998;13:677-83.

42. Miyata T, Kobayashi Y, Araki H, Ohto T, Shin K. The influence of con- trolled occlusal overload on peri-implant tissue. Part 3: A histologic study in monkeys. Int J Oral Maxillofac Implants 2000;15:425-31.

43. Mahler DB, Peyton FA. Photoelasticity as a research technique for ana- lyzing stresses in dental structures. J Dent Res 1955;34:831-8.

44. White SN, Sorenson JA, Kang SK. Improved marginal seating of cast restorations using a silicone disclosing medium. Int J Prosthodont 1991;4:323-6.

Reprint requests to: DR DAVID L. GUICHET PROVIDENCE PROSTHODONTICS 1310 WEST STEWART DR, SUITE 202 ORANGE, CA 92868 FAX: (714)771-7557 E-MAIL: [email protected]

Copyright © 2002 by The Editorial Council of The Journal of Prosthetic Dentistry.

0022-3913/2002/$35.00 + 0. 10/1/124589

doi:10.1067/mpr.2002.124589

GUICHET, YOSHINOBU, AND CAPUTO THE JOURNAL OF PROSTHETIC DENTISTRY

MAY 2002 535

tightness in non-splinted restorations resulted in a non-passive situation.

REFERENCES

1. Solnit GS, Schneider RL. An alternative to splinting multiple implants: use of the ITI system. J Prosthodont 1998;7:114-9.

2. Laney WR, Jemt T, Harris D, Henry PJ, Krogh PH, Polizzi G, et al. Osseointegrated implants for single-tooth replacement: progress report from a multicenter prospective study after 3 years. Int J Oral Maxillofac Implants 1994;9:49-54.

3. Henry PJ, Laney WR, Jemt T, Harris D, Krogh PH, Polizzi G, et al. Osseointegrated implants for single-tooth replacement: a prospective 5- year multicenter study. Int J Oral Maxillofac Implants 1996;11:450-5.

4. Scheller H, Urgell JP, Kultje C, Klineberg I, Goldberg PV, Stevenson- Moore P, et al. A 5-year multicenter study on implant-supported single crown restorations. Int J Oral Maxillofac Implants 1998;13:212-8.

5. Sutter F, Schroeder A, Buser DA. The new concept of ITI hollow-cylinder and hollow-screw implants: Part I. Engineering and design. Int J Oral Maxillofac Implants 1988;3:161-72.

6. Moberg LE, Kondell PA, Kullman L, Heimdahl A, Gynther GW. Evaluation of single-tooth restorations on ITI dental implants. A prospec- tive study of 29 patients. Clin Oral Implants Res 1999;10:45-53.

7. Levine RA, Clem DS III, Wilson TG Jr, Higginbottom F, Solnit G. Multicenter retrospective analysis of the ITI implant system used for sin- gle-tooth replacements: results of loading for 2 or more years. Int J Oral Maxillofac Implants 1999;14:516-20.

8. Gomez-Roman G, Schulte W, d’Hoedt B, Axman-Krcmar D. The Frialit-2 implant system: five-year clinical experience in single-tooth and imme- diately postextraction applications. Int J Oral Maxillofac Implants 1997;12:299-309.

9. White GE. Osseointegrated dental technology. Carol Stream (IL): Quintessence; 1993. p. 61, 103-29.

10. Stumpel LJ III. The adhesive-corrected implant framework. J Calif Dent Assoc 1994;22:47-50, 52-3.

11. Guichet DL, Caputo AA, Choi H, Sorenson JA. Passivity of fit and mar- ginal opening in screw- or cement-retained implant fixed partial denture designs. Int J Oral Maxillofac Implants 2000;15:239-46.

12. English CE. Biomechanical concerns with fixed partial dentures involv- ing implants. Implant Dent 1993;2:221-42.

13. Korioth TW, Hannam AG. Deformation of the human mandible during simulated tooth clenching. J Dent Res 1994;73:56-66.

14. Hobkirk JA, Schwab J. Mandibular deformation in subjects with osseoin- tegrated implants. Int J Oral Maxillofac Implants 1991;6:319-28.

15. Fischman B. The rotational aspect of mandibular flexure. J Prosthet Dent 1990;64:483-5.

16. Serio FG. Clinical rationale for tooth stabilization and splinting. Dent Clin North Am 1999;43:1-6, v.

17. Ante IH. The fundamental principles of abutments. Mich State Soc Bull 1926;8:14-23.

18. Nyman S, Ericsson I. The capacity of reduced periodontal tissues to sup- port fixed bridgework. J Clin Periodontol 1982;9:409-14.

19. Nyman S, Lindhe J. Prosthetic rehabilitation of patients with advanced periodontal disease. J Clin Periodontol 1976;3:135-47.

20. Skalak R. Biomechanical considerations in osseointegrated prostheses. J Prosthet Dent 1983;49:843-8.

21. Quirynen M, Naert I, van Steenberghe D. Fixture design and overload influence marginal bone loss and fixture success in the Branemark sys- tem. Clin Oral Implants Res 1992;3:104-11.

22. Hoshaw SJ, Brunski JB, Cochran GV. Mechanical loading of Branemark implants affects interfacial bone modeling and remodeling. Int J Oral Maxillofac Implants 1994;9:345-60.

23. Rangert B, Jemt T, Jorneus L. Forces and moments on Branemark implants. Int J Oral Maxillofac Implants 1989;4:241-7.

24. Rangert B, Krogh PH, Langer B, Van Roekel N. Bending overload and