Evaluation of three impression techniques for osseointegrated oral implants

Evaluation of three impression osseointegrated oral implants

techniques

for

Jose A. Inturregui, DDS,a Steven A. Aquilino, DDS, MS,b Jeffrey S. Ryther, DDS, MS,C and Peter S. Lund, DDS, MSC University of Iowa College of Dentistry, Iowa City, Iowa The purpose of this in vitro investigation was to determine the accuracy of gypsum casts produced from impressions made with polyether, polyether and impression plaster, or polyether and acrylic resin for the fabrication of osseointegrated implant prostheses. Strain gauges were attached to a master framework to determine the passivity of fit of the framework to sample casts made by the three impression techniques. Strain values were statistically compared by one way analysis of variance and Duncan’s multiple range test. A statistically significant difference was found between the three impression techniques tested (p < 0.05). The results of this investigation revealed that none of the impression techniques resulted in an absolutely passive framework fit. However, of the techniques tested, the polyether alone resulted in the closest duplication of the master cast. (J PROSTHET DENT 1993;69:503-9.)

M

ost of the scientific literature about dental implants has dealt with the processof osseointegration.1-3 Most research efforts have been directed toward investigating the tissue/titanium interface.4 To date, there has been little research on the biomechanical aspectsof implant prosthodontics.5 The implant literature emphasizesthe importance of a passively fitting fixed prosthesisto prevent prosthodontic complicationsor even lossof fixture integration.6 Failure to achieve a passively fitting prosthesisand force tightening of the superstructure may result in complications such as abutment, framework, and gold screw looseningor fracture.7.s In addition, the increasedstressmay result in microfractures of bone, marginal ischemia,and healing with a nonmineralized attachment to the implant fixture.g There is, however, little if any scientific evidenceto support these hypotheses. The impressionmaterials and techniques usedfor master cast fabrication play a key role in the accuracy of fit and passivity of the cast metal framework.lOal1 Nevertheless, only two studies were found that compared different impressiontechniques. Humphries et a1.12found no statistically significant difference between three different impressiontechniques.Spector et a1.,13on the other hand, reported that there waspotential for distortion with the three impressiontechniques they investigated. They concluded

that further work wasnecessaryto isolate the most reliable and predictable impressionprocedure. This investigation objectively tested the passivity of fit of a master framework, via strain gauges,on stone cast replicas of osseointegratedimplants made by three different implant impressiontechniques. The strain values were compared and analyzed statistically to determine which impressionprocedure most accurately recorded the interabutment relationship.

MATERIAL

AND

METHODS

A master metal framework wascast in a silver palladium alloy (Palliag M, DegussaDental Inc., Long Island City, N.Y.). The framework consistedof two 4 mm gold cylinders (DCA 075, NobelpharmaUSA Inc., Chicago,Ill.) connected by a 30 mm x 6 mm x 3 mm bar. In addition, two fiat surfaces (5 mm x 5 mm), one in the horizontal plane and one in the vertical plane, were prepared midway between the abutments for strain gaugeplacement (Fig. 1). The finished master framework wasusedfor fabrication of the master cast to assurea passivefit. Two brassabutment replicas (DCA 015, Nobelpharma USA Inc.) and associated guide pins were tightened to 10 Ncm on the framework with a torque driver (DIA-250, Blue, Nobelpharma USA Inc.). A baseplatewax mold was constructed. The framework waspositioned in the mold with sticky wax. Improved dental stone (Die-keen, Modern Materials, Columbus Dental, Miles Inc., St. Louis, mixed according to the manufacturer’s

aGraduate student, Department of Prosthodontics. bAssociate Professor and Graduate Program Director, Department of Prosthodontics. CAssistant Professor, Department of Prosthodontics. OF Copyright ‘3 1993 by The Editorial Council of THE JOURNAL PROSTHETIC DENTISTRY. 0022-3913/93/$1.00 +.lO. 10/l/45266

MAY

1993

MO.) was vacuumrecommendations

and poured into the mold to make the mastercast. The cast was allowed to set for a period of 24 hours before the framework was removed. The master cast was trimmed and stored at ambient conditions until needed for the impression procedures (Fig. 2). A duplicate of the master cast was modiHed and used as

503

THE

JOURNAL

Fig.

OF PROSTHETIC

DENTISTRY

1. Master framework prepared for master cast fabri-

INTURREGUI

Fig.

ET AL

3. Polyether impression,technique I.

cation.

Fig.

2. Master cast with master framework in place.

a template for fabrication of custom acrylic resin impressiontrays. This technique ensureduniform thicknessof the impressionmaterial. A total of 30 trays weremade, 10trays for each technique group. Three different techniques were used to make the impressions: Technique I (‘lyether alone). Two guide pin-retained squareimpressioncopings(DCB 026, Nobelpharma USA Inc.) were carefully tightened to 10 Ncm with the torque driver. A polyether impressionmaterial (Impregum F, Premier Dental Products Co., Norristown, Pa.) was mixed accordingto manufacturer’s recommendations.The material was syringed around the impressioncopingsand the filled customtray wasplaced over the master cast. The impressionswere separatedfrom the mastercast 6 minutes after placement (Fig. 3). Technique II (polyether and plaster). Two guide pin-retained impression copings were placed as in technique I. An initial impressionof the master cast wasmade

504

with a customtray and polyether impressionmaterial. The impressionmaterial wasmixed and handled as previously described.However, only the border of the master cast was recorded. After separation from the master cast, excess impressionmaterial was removed from the window area and around the impressioncopings. The impressionwas again seated on the master cast, and the transfer coping relationship was recorded with impressionplaster (Snow White ImpressionPlaster No. 2, Kerr Mfg. Co., Emeryville, Calif.). The plaster wasmixed accordingto manufacturer’s recommendationsand allowed to set for 10minutes before the impressionwasremovedfrom the master cast.One coat of tinfoil substitute (Coe-Sep,Coe Laboratories Inc., Chicago,Ill.) waspainted on the plaster surface of the impression before pouring (Fig. 4). Technique III (polyether and autopolymerizing resin). Two guide pin-retained impressioncopings were again placed with the torque driver. Autopolymerizing acrylic resin (Duralay, Reliance Dental Mfg. Co., Worth, Ill.) wasusedto join the transfer copingsrigidly before the impressionwas made. A mold of vinyl polysiloxane putty (Reprosil Putty, L. D. Caulk Div., Dentsply Intl. Inc., Milford, Del.) was made to standardize the autopolymerizing acrylic resin splint. The acrylic resin wasmixed according to manufacturer’s recommendations,poured into the mold, and allowedto set for 15 minutes. The splinted impression copings were removed from the master cast. The excess resin wastrimmed and the splint wassectionedequidistant from the two copings with a thin Carborundum disk. The two segmentswere again seated on the master cast and retightened with the torque driver. The segmentswere joined together with autopolymerizing acrylic resin by a bead-brushing technique. The acrylic resin was again allowed to set for 15 minutes before the polyether pick-up impressionwasmade(Fig. 5). The impressionmaterial was injected around and under the splinted copings, and the filled custom impression tray was seated over the entire assembly.

VOLUME

69

NUMBER

6

INTIIRREGI-I

ET AL

THE

JOURNAL

OF PROSTHETIC

nENTISTRY

Fig. 4. Polyether and plaster impression, technique II. 6. Horizontal and vertical strain gaugescementedto the master framework. Fig.

5. Polyether and autopolymerizing resin, technique II, before pickup impression. Fig.

Impressionswere poured in sets of five, within a time period of lessthan 3 ‘/z hours from the first and at least 30 minutes after the last impressionwas made. Brass abutment replicas were securedwith guide pins to the impression transfer copings. The impressionswere poured with improved dental stone (Die-Keen). One hundred gramsof stone were added to 22 ml of distilled water, hand-spatulated for 30 seconds,and then vacuum-mixed for an additional 30 seconds.The stone wasallowed to set for a minimum of 2 hours before the cast was separated from the impression.Thirty working casts were stored at ambient conditions for 2 weeks until strain measurementswere made. Two strain gauges(EA-06-015SE-120, Micro-Measurements Division, MeasurementsGroup Inc., Raleigh, N.C.) were cementedonto the master framework with a modified methyl-2-cyanoacrylate adhesive (M-Bond 200 Adhesive, MeasurementsGroup Inc.), one in the horizontal plane and the other in the vertical plane (Fig. 6). The strain gauge leadsweresolderedto an l&gauge copper wire, 45 cm long,

MAY

1993

Fig.

7. Digital strain indicator,

and attached to a digital strain indicator that measuredthe strain in the masterframework. The masterframework was attached to each samplecast with gold screws(DCA 075, Nobelpharma, USA Inc.), which were tightened with the torque driver to ensurea consistent torque application of 10 Ncm. A P-3500 digital strain indicator (MeasurementsGroup Inc.) wasusedin the 120ohm quarter-bridge configuration (Fig. 7). The gaugefactor wasset to 2.021asprescribedfor the strain gaugesused.The strain indicator recorded the amount of microstrain (PC),which equaled the change in length of the strain gauge divided by its original length (0.38 mm) multiplied by 10-s.The apparatus wasbalanced to t- 0000 strain with no load on the master framework. The sampleswere securedon a vise during testing procedures for maximum stability. Samples were tested in a random fashion. Two readingswere taken from eachstrain gaugeand an averagept was calculated for the horizontal gaugesand for the vertical gauges. In a separateexperiment, the loads required to produce

THE

JOURNAL

OF PROSTHETIC

DENTISTRY

INTURREGUI

ET AL

I. Calculated data for the horizontal and vertical strain gauges

Table

Mean w

Group number

Horizontal I II III Vertical I II III I, Polyether

Standard deviation

Coeff. variance

of

42.41

13.41

-22.50

+96.00

14.40

-124.50

t27.00

t351.98 -73.61

-70.75

31.53

9.97

-137.00

-37.50

-44.57

10.93 12.48 9.64

3.46

-28.50

+3.00

3.95

-12.50

t30.00

3.05

-2.00

+26.50

plane -8.50

t7.95 t10.50 alone; II, polyether

and plaster;

III, polyether

and autopolymerizing

Mean squares

F value*

3

13946.999 1621.977

Error Vertical

27

3

784.683

Error

27

122.638

strain gauge Model

-128.58

t156.92 t91.76

resin.

0.9999for both groups.Therefore, the samplesizeof 10was more than adequateto determine significant differences (o( P

strain

Model

8.60

0.0004

6.40

0.0020

value 2.96 (p 5 0.05).

horizontal and vertical framework strains similar to those recorded on the sample castswere investigated. A digital force gauge (Accuforce III, Basic Service Corp., Detroit, Mich.) was usedto measurethe applied forces to the system. Force wasapplied to the framework in two directions to place the upper surface of the framework in tension or compression,and the force (kg) applied and strain (PC) produced were recorded. The master framework was also manipulated to correlate the strain state (tension or compression) with the type of flexure applied to the system.

RESULTS Two pcreadingswere averagedfor the horizontal gauges and two for the vertical strain gaugeson the mastercast and eachof the 30 samplecasts.The mean strain values on the master cast were -17 PChorizontal plane and -12 pt vertical plane. The strain values for the master cast were recorded at the beginning of the experiment and checked six times throughout the data collection proceduresto ensure consistency. The meanstrains, standard deviations, standard error of the mean, minimum and maximum values, and coefficient of variance for the experimental groups are listed in Table I. Given the high standard deviations, a power analysiswas conducted to validate the samplesize (n = 10) usedin this study. Calculations for both the horizontal (Z = 6.03) and vertical (Z = 6.38) strain gaugesresulted in power valuesof

506

value (P4

45.53

kww

*F Critical

Maximum

+12.05 -61.85

Degrees of freedom

Horizontal

value (44

plane

II. Analysis of variance (ANOVA)

Table

Minimum

Standard error of mean

0.05).‘4

A one way analysis of variance (ANOVA, (YI 0.05) revealed statistically significant differences in both the horizontal plane (p 0.0004)and the vertical plane (pO.002) (Table II). Duncan’s multiple range tests (a! 5 0.05) demonstrated that technique I (polyether alone) had mean strain values that were statistically different from techniques II and III in both the horizontal and vertical planes (Table III). In addition, technique I demonstrated mean strains that were closestin value to those found on the master cast (Figs. 8 and 9). The forces required to produce the strains recorded with each impressiontechnique were also investigated. A force of <1.7 kg in the vertical plane and <2 kg in the horizontal plane produced a strain equal to the maximum recorded with any of the impressiontechniques investigated. The type of strain (+ or -) wasrelated to the theoretical types of framework distortion and is summarizedin Table IV.

DISCUSSION Although the passivity of fit of osseointegratedimplant prosthesescontinues to be emphasizedin the dental literature,6‘g no objective way to determine the fit of an implant-supported prosthesishas been described. Clinically the fit of an implant prosthesishas beenaccepted or rejected on the basisof the clinician’s judgment. It is possible that clinically unacceptablediscrepanciescould lead to unacceptably high stressesin the bone and the prosthesis. Previous studies used linear measurementsto evaluate the accuracy of various implant impressiontechniques.12,l3 In the studiesby Spector et al.13and Humphries et al.12no significant differences were noted in the impressiontechniques evaluated. This study was designedto investigate the strain produced in a master framework when it wassecuredon sample casts made by three impression techniques. Similar strain gaugetechnology has been usedin the past to measure occlusalloading of implant-supported prostheses.15-1g

VOLUME

69

NUMBER

6

INTURREGUI

THE

FX AL

JOURXAI,

OF PROSTHETIC

DENTISTRY

-80 -61.85

-80

Impression Technique 8. Mean horizontal strain: Impressiontechniques versusmaster cast. Technique I, polyether alone;technique II, polyether and plaster; technique III, polyether and autopolymerizing resin. Fig.

Two strain gaugesin the horizontal and vertical planes were incorporated on the master framework. Only two implant abutments were used,to simplify the force patterns produced. A torque driver was used to ensurethat a constant torque of 10Ncm wasapplied to all componentsof the system, except for placement of the brassabutment analogue before impression pouring. The tightening of the guide pin, impression coping, and abutment analogue complex was done with finger pressureonly. The 10 Ncm force applied by the torque driver causedrotation of the transfer copingsin the polyether impressions(technique I). Therefore, finger pressure was used to tighten the brass abutment analoguesfor all impressiontechniques. Even though the master cast was fabricated from the masterframework to assurea passivefit, strain waspresent in both the horizontal (-17 pclc) and vertical (-12 pt) planes. Therefore, the strains produced by the three impression techniques were compared with this level to determine which technique best duplicated the master cast. The strain present when the framework was secured to the master cast may have resulted from the setting expansion of the type V (ADA specification No. 25) improved dental stone (Die-Keen) used in its fabrication.20 The residual stresseswithin the master cast may have been released when the framework was first removed. When the framework wasreplacedon the master cast for strain gaugemeasurements,it was distorted in such a way as to place both horizontal and vertical strain gaugesin a state of compression. The master framework appearedto passivelyfit eachone of the sample casts, with no observable discrepancy be-

MAY

1993

Table

Duncan’s groupings

III. Group number*

Horizontal iwge I II III Vertical

Duncons

groupings?

strain 10 10 10

+12.05 -61.85 -70.75

A

+10.50 +7.95 -8.50

A 4

B R

strain

gauge III II I *I, Polyether; ing resin. tMeans with

Mean (PC)

N

10 10 10 II, polyether the same letter

and

plaster:

l3 III,

polyether

are not significantly

and

autopolymeriz-

difl‘rrent.

tween the framework and abutment analogues.However, measurablestrains wereproducedwith all three impression techniques. The sample casts made from polyether impressionmaterial alone(technique I) produced meanstrain values similar to the master cast in both the horizontal and vertical planes and statistically different from the other two impression techniques. In addition t,o the increased strain produced in techniques II and III, the direction of the strain (tension or compression)wasdifferent for these techniques than for technique I (Figs. 8 and 9). The increasedstrain and the change in direction of the strain with these techniques wasmost likely the result of a combination of factors. The setting expansionof the impression

507

THE

JOURNAL

OF PROSTHETIC

INTURREGUI

DENTISTRY

ET AL

15 ,

.

-15

'

Impression Technique Fig. 9. Mean vertical strain: Impression techniques versus master cast. Technique I, polyether alone; technique II, polyether and plaster; technique III, polyether and autopolymerizing resin.

IV. Resulting strain (tension vs. compression) theoretical framework distortion

for

Table

Bending toward cast

Horizontal strain gauge Vertical strain gauge + Sign indicates

tension;

Bending away from cast

Torquing R-front L-back

Torquing R-back L-front

-

+

-

+

+

-

+

-

- sign indicates

compression.

plaster (technique II), the polymerization shrinkage of the acrylic resin (technique III), and the rigid fixation of the impression transfer copings in techniques II and III may have altered the setting expansion of the improved dental stone and, therefore, have distorted the interabutment relationship from that of the master cast. The directions of the mean strain values were similar to the theoretical framework distortion presented in Table IV. The horizontal and vertical strain gauges consistently had mean strain values in opposite directions (Table I). These factors and the inability to use the torque driver for brass analogue placement before impression pouring may have contributed to the high standard deviations and coefficients of variance. Davis et a1.17also reported large standard deviations with strain gauge recordings. They suggested that the fit and tightness of the coping screws had a marked effect on the deflection and consequent strain levels in the framework. In addition, forces of <2 kg were responsible for the maximum strains recorded with any of

the impression techniques tested. Therefore, the high standard deviations may have been an artifact of the high sensitivity of the strain gaugesusedas comparedwith the forces generated. Determination of the magnitude and exact patterns of distortion created by the impression techniques was beyond the scopeof this investigation. The magnitude of the forces recorded in this investigation may be significant in relation to the relatively low forces(30 to 100gms)required to move teeth orthodontically. Further studiesare necessary to determine the amount of strain that can be tolerated within the osseointegratedsystem.Although our ultimate goal should be zero strain this is probably impossible to achieve becauseof limitations created by the physical properties of the materials used. Determination of clinically acceptable strain limits are necessaryto allow us to further evaluate the biomechanicalproperties of implantsupported prostheses.

CLINICAL

IMPLICATIONS

As with any prosthodontic technique, a multitude of factors could contribute to the ultimate accuracy of the final prosthesis.Although the accuracy of each component is important, it is the accuracy of the total system that is imperative. Because the master framework appeared to passively fit eachone of the samplecasts,any of the techniques investigated should be clinically acceptable.There seemsto be no clinical advantage in splinting impression transfer copings with autopolymerizing acrylic resin or plaster, becausepolyether alone demonstrated the lowest mean strain values comparedwith the strain values of the

VOLUME

69

NUMBER

5

INTURREGUI

THE

ET AI,

master cast. Therefore, the use of a rigid elastomeric impression material alone should simplify the impression procedures for osseointegrated implants and make the appointment less time-consuming. It is uncertain what long-term effect strains of the magnitude recorded in this investigation may have on the bone and the prosthesis. That these strains were produced with no observable discrepancies between the master framework and abutment analogues indicates that a more objective method may be necessary to determine whether an osseointegrated implant framework fits passively. However, until a more objective method is established, the acceptance or rejection of the fit of an osseointegrated prosthesis will ultimately be based on the judgment of the clinician. CONCLUSIONS The results of this investigation suggest the following conclusions: 1. There were statistically significant differences between the mean strains recorded from the sample casts made with the three impression techniques. 2. Sample casts made from polyether alone (technique I) resulted in the closest duplication of the mean strain values recorded on the master cast in both the horizontal and the vertical planes. 3. On the basis of strain values recorded and the subjective fit of the master framework on the sample casts, any of the impression techniques investigated should be clinically acceptable. 4. There appears to be no clinical advantage in using the more time-consuming techniques of splinting impression transfer copings with either autopolymerizing acrylic resin or plaster. 5. Further research is needed to determine the effect a constant force of the magnitude found in this investigation (<2 kg) has on the osseointegrated implant, the bone, and the prosthesis. REFERENCES 1. Branemark PI, Breine U, Adell R, Hansson BO, Lindstrom J, Ohlsson A. Intra-osseous anchorage of dental prosthesis, I-experimental studies. Stand J Plaat Raconstr Surg 1969;3:81-100. 2. Adell R, Lekholm U, Rockier B, Branemark PI. A 15-year study of osseointegrated implants in the treatment of the edentulous jaw. Int J Oral Surg 1981;10:387-416. 3. Branemark PI. Osseointegration and its experimental background. J PROSTHET

MAY

1993

DENT

1983;50:399-410.

JOURNAL

OF PROSTHETIC

DENTISTRY

4. Albrektsson

T, Branemark PI, Hansson HA. Lindstrom d. Osseointegrated titanium implants. Acta Orthop Stand 1981:52:155-70. Skalak R. Biomechanical considerations in osseointegrated prostheses. d PROSTHET DENT 1983:49:843-S. Zarb GA, Zarb FL. Tissue integrated dental prostheses. Quintessence Int 1985;1:39-42. Sellers GC. Direct assembly framework for ossr!ointegrated implant prosthesis. J PROSTHET DENT 1989:62:662-R. Sones AD. Complications with osseointegrated implants. J J’WsTHiT?

DENT 1989;62:581-5. P, Bolender 9. Worthington

CL, Taylor TD. The Swedish system of osseointegrated implants: problems and comphcations encountered during a 4.year trial period. Int J Oral Maxillofa~~ Implants 19X7:2:7764. 10. Tautin FS. Impression making for osseointegrarsad dentures. .I PIKK THET DENT

1965;54:250-1.

11. Henry PJ. An alternative method for the product ion of accurate cast,s and occlusal records in osseointegrated implant rehabilitation. .I PRO<‘WET

DENT

1987;58:694-7.

12 Humphries RM, Yaman P, Bloem TJ. The accuracy of implant master casts constructed from transfer impressions. Int .l Oral Maxillofac Implants 1990;5:331-6. 13 Spector MR, Donovan TE, Nicholls ,JI. An evaluation of impression techniques for osseointegrated implants. d PIVWHE’I’ DRNT 1990; 63444-7.

14. Remington RD, Schrok MA. Statistics with applicsations to the bioiogical and health sciences. Englewood Cliffs, NJ Prentice-Hall Inc., 1985;179. 15. Haraldson T, Carlsson GE. Bite force and oral function in patients wit b osseointegrated implants. Stand J Dent Res 1977:85:200-K. 16. Carr AB, Laney WR. Maximum occlusal force levels in patients with osseointegrated oral implant prostheses and patients with complete dentures. Int J Oral Maxillofac Implants 1987;2: 101-X. for osseointe17. Davis DM, Zarb GA, Chao YL. Studies on frameworks grated prostheses: part I. The effect of varying the number of supporting abutments. Int J Oral Maxillofac Implants 1988;3:197-201. 18. Falk H, Laurel1 L, Lundgren D. Occlusal force pattern in dentitions with mandibular implant-supported fixed cantilever prostheses occluded with complete dentures. Int J Oral Maxillofac Implants 1989:4:%X2. 19. Falk H, Laurel1 L, Lundgren D. Occlusal interferences and cantilever joint stress in implant-supported prostheses occluding with complete dentures. Int J Oral Maxillofac Implants 1990;5:70-7, Gypsum Products. 20. Revised ANSI-ADA Specification No. 25 UentG Chicago: American Dental Association Council on Dent.al Materials and Equipment, 1989. 21. Burstone CJ. Application of bioengineering to climcal orthodontics. In: Graber TM, Swain BF. Orthodontics: current :vinciplrs and techniques. St Louis: CV Mosby, 1986;193-227 Reprint

requests

to:

DR. STEVEN A. AQ~ILINO COLLEGE OF DENTIRTHV UNIVERSITY OF Ioti~ Iow,\ CITY, IA 52242

Contributing

author

Thomas E. Southard, DDS, MS, Assistant Professor, Department of Orthodontics, University of Iowa, College of Dentistry, Iowa City, Iowa.

509