A Histological Comparison in the Dog of Porous-coated vs. Threaded Dental Implants D.A. DEPORTER, P.A. WATSON, R.M. PILLIAR, M.L. CHIPMAN1, and N. VALIQUETTE Medical Research Council Program in Dental Implantology, Faculty of Dentistry, University of Toronto, 124 Edward Street, Toronto, Ontario, Canada M5G 1G6; and 'Department of Preventive Medicine and Biostatistics, Faculty of Medicine The histological findings of an 18-month trial, in the dog, of a partially porous-coated endosseous dental implant made of Ti-6A1-4V, with a truncated conical shape, are described and compared with those for a cylindrical, threaded, endosseous implant made of commercially pure Ti. Six beagle dogs each received two porous-coated implants on one side of the mandible and two threaded implants on the contralateral side. Each set of two implants supported a two-unit fixed bridge for an 18-month functional period. Methylmethacrylate sections of both the buccolingual and mesiodistal aspects of each implant were examined qualitatively and by computer-assisted morphometiy. The morphometric measurements were used for determination of the length of implant surface in direct contact with bone on each aspect of each implant. The data were expressed both as an absolute length and as a fraction of the maximum length available for contact (contact length fraction or CLF). On the buccal and lingual aspects of the implants, both the absolute lengths and CLF were significantly smaller for the porous-coated design. For the mesial and distal aspects, the absolute lengths and CLF were less for the porous-coated design, but the differences were not significant. However, when the absolute contact length was related to the corre-
sponding vertical bone height, significant differences were observed, the absolute contact length being greater for any given bone height for the porous-coated design. Taken together, the data suggest that shorter implants may be used with the porous-coated design. J Dent Res 69(5):1138-1145, May, 1990
Introduction. Numerous authors have proposed the use of a porous surface to achieve stabilization of both orthopedic and dental endosseous implants (see Pilliar, 1986, for a review). Earlier publications from our laboratory (Deporter et al., 1986a, b, 1988) described a truncated conical-shaped dental implant made of Ti-6A1-4V that incorporated a porous surface geometry (pore size 50-200 lLrm) on both the root implant component and on the apical third of the transgingival collar. The latter was meant to encourage gingival connective tissue attachment. In fact, it provided the weak link in the system, resulting in contamination with bacterial plaque and what was considered to be progressive implant failure by seven months of function in 22 of 32 implants placed in dog mandibles. However, for the remaining ten non-contaminated implants, it was obvious that when the porous coat of both root and collar components was submerged largely or entirely within bone, the implants functioned as intended, resulting in implant stabilization and effective stress transfer between the root implant component and the bone. Therefore, the design of the system was changed so that the porous surface configuration would be limited to the apical two-thirds of the root component. The present report describes the histological findings of an 18-month (in function) clinical trial of this modified porouscoated implant in the dog. The experimental design included Received for publication July 3, 1989 Accepted for publication November 1, 1989 This work was supported by Program Grant #26 from the Medical Research Council of Canada.
1138
comparison with a threaded implant design similar, but not identical, to the Brdnemark implant (Branemark et al., 1977), since a morphometric histological assessment of threaded, endosseous dental implants has not been previously published. It was anticipated that these two designs with differing surface configurations would result in different patterns of stress distribution and resultant bone remodeling at the bone-implant interface.
Materials and methods. The components of both endosseous dental implant systems were made by machining (Strite Industries, Cambridge, Ontario), with the porous coat being added subsequently in our laboratory, through procedures developed by one of us (Pilliar, 1987). The porous-coated root component (Fig. la) was fabricated from Ti-6A1-4V (ASTM F 136), rather than from commercially pure Ti (CP Ti), because the former is known to have greater strength. It consisted of a truncated conical shape (taper angle = 50) with the porous coat limited to the apical two-thirds, the coronal third having a machined surface, with a surface finish better than 1.75-km cla (center line average). Figs. lb and ic show scanning electron micrographs of this coronal region. Fine machining lines can be seen running circumferentially. A thermal etching pattern can also be seen superimposed on the machining lines. The high magnification view (Fig. lc) of the sample edge profile confirms the small cla surface finish. (Examination of the threaded implants showed similar fine machining lines on the surfaces of the threads and on the non-threaded coronal portion.) For achievement of a continuous contour at the machined-to-porous-coated surface junction, the apical two-thirds was reduced with a 0.25-mm in-step relative to the coronal third. This permitted the subsequent placement of a porous coat on the outer surface, which was approximately flush with the coronal third. The self-tapping threaded implant was made from CP Ti (ASTM F 67)so that it would be similar to the Brdnemark implant-was cylindrical and had an effective thread angle of 600. The pitch of the threads was 0.635 mm, the number of threads per cm was 15.74, and the thread depth was 0.787 mm. The diameter of the implant shoulder (Fig. la) was equal to 3.53 mm and was slightly larger than the collar diameter (3.50 mm). The other components of both implant systems were identical in design (Fig. 2) and were machined as required from either Ti-6AI-4V or pure titanium. All of the components were cleaned and sterilized as described previously (Deporter et al.,
1986b). Six in-bred beagle dogs, two males and four females (Laboratory Research Enterprises, Kalamazoo, MI), initial weights 9.8 to 17.3 kg, were used. Implants were placed bilaterally in the region of the mandibular third and fourth premolars, which had been rendered edentulous at least six months prior to implantation. Two porous-surfaced implant root components were placed on the left side, and two threaded components were placed on the right side of the mandible of each dog. (The two
Downloaded from jdr.sagepub.com at University of Manitoba Libraries on June 5, 2015 For personal use only. No other uses without permission.
Vol. 69 No. 5
POROUS-COATED vs. THREADED DENTAL IMPLANTS
1139
a
l1"mm
Fig. 1-(a) The porous-coated (left) and threaded (right) root implant components used in the study. (b) Low (x5O) and (c) high (x5OO) magnification scanning electron micrographs of the machined coronal region of the partially porous-coated implant. The lower-magnification micrograph shows circumferential machining lines. The higher-magnification view shows an edge profile of these lines. A thermal etching pattern is also seen on this micrograph.
implants on each side were referred to as the mesial and distal implants, respectively.) The porous-coated implants were inserted, as previously described (Deporter et al., 1986b), following elevation of a fullthickness mucoperiosteal flap. A preliminary channel was cut with a twist drill (1.524-mm diameter; Whaledent Inc., New York, NY), and this was subsequently enlarged with a custommade, tungsten-carbide bur (Beaver Dental Products Ltd., P.O. Box 900, Morrisburg, Ontario) in a speed-reducing handpiece operated at a maximum speed of 120 rpm under external saline
irrigation. After thorough irrigation of each prepared site, implants were tapped into position with a mallet and a Teflontipped punch. The threaded implants were placed by means of a similar flap approach and initial site preparation with the same twist drill. Each site was then enlarged with a customfabricated carbon-steel twist drill of larger diameter (2.362 mm) corresponding to that of the implant without its threads. Finally, each site was threaded with a manual tap of standard carbide steel so that the threaded implant would be received. All of these manipulations were done under copious saline irrigation. The flaps were sutured with 4-0 Vicryl (Ethicon Sutures Ltd., Peterborough, Ontario). The animals were maintained on a soft diet (Pep dog food, Quaker Oats Co., Peterborough, Ontario) for the first two weeks post-operatively, and were given Penlong XL at appropriate intervals during this period. After an initial healing period of six weeks, all root components were uncovered, and each received a collar and collarretaining screw. At this time, one porous-coated root component in one dog had not become fixed by bone ingrowth, for some unknown reason, and was removed. This implant was not replaced, and the remaining porous-coated implant was left to function as a free-standing implant. The other 22 implants were used as support for two-unit fixed bridges for an 18month functional period. Following the 18-month functional period, the animals were
Downloaded from jdr.sagepub.com at University of Manitoba Libraries on June 5, 2015 For personal use only. No other uses without permission.
J Dent Res May 1990
DEPORTER et l.
1140
BUCCAL Buccal- Lingual
Dist~al Buccal,'
i
Mesial Buccal
DISTAL
MESIAL Distal Lingual
C_;
AO_
Fig. were
(B),
transgingival components used for both implant systems design, and included a collar (A), a collar-retaining screw coping (C), and a coping-retaining screw (D). 2-The
identical in a
killed under general anesthesia by bilateral carotid perfusion with a solution of 37% formalin and 99% methanol in distilled water (1:1:1.5 V/V). Segments of mandible containing the implants were removed, processed, and embedded in methylmethacrylate, and subsequently sectioned for histological assessment, as previously described (Deporter et al., 1986b, 1988). Briefly, blocks were sectioned with a Buehler Isomet lowspeed saw (Tech-Met Canada) equipped with a diamond wafering blade. The sectioning technique was designed so that each implant would have two buccolingual, two mesial, and two distal sections (Fig. 3). The initial section thickness of 120-150 ALm was reduced to approximately 40 Elm by means of petrographic grinding techniques, and the sections were stained with a mixture of Stevenel's blue and van Gieson's picro-fuchsin. All of the sections were examined both qualitatively and by computer-assisted morphometry (Deporter et al., 1986b, 1988). This allowed for the determination of the length of implant surface in direct contact with the bone on each aspect (i.e., buccal vs. lingual vs. mesial vs. distal) of each implant. Since the porous-coated implants were designed to be stabilized by bone in-growth into the porous surface, measurements for these implants did not include data from the coronal third region. For these implants, the length of implant surface in direct contact with bone was determined on a digitizing tablet by tracing the absolute length of bone contact with the outer surfaces of the spheres of the outermost layer of porous coat. In contrast, since the threaded implants were designed and inserted so that all of the threads for implant stabilization would be used, the
4.Mesial Lingual
Fig. 3- -A diagram depicting the sectioning technique used.
full length of the threaded implant surface was included, and the length of implant surface in contact with bone was determined by the absolute length of bone contact with the surfaces of the threads being traced. These data were expressed in two ways: first, as an absolute length, and second, as a fraction of the maximum length of implant surface available for contact with bone. The fraction figure was referred to as the contact length fraction (CLF), and is expressed as follows: (absolute length contacted by bone) CLF (length available for contact with bone) For each aspect, we examined the characteristics of the distributions of measures of absolute contact length and contact length fraction by type of implant and aspect of implant, calculating the mean, standard deviation, and median. Because these descriptive data suggested that the two implant designs behaved differently on buccal and lingual aspects, the two designs were compared (separately), as described below, for each of these aspects of the implants. Likewise, the mesiobuccal-distobuccal and mesiolingual-distolingual aspects for the two designs were compared separately. Analyses of variance were done for comparison of absolute contact length and contact length fraction for the relevant aspects of the two implant designs with variation between dogs controlled for. In addition, for the mesial and distal aspects, analyses of variance were done for comparison of the ratio of absolute contact length with the corresponding bone height for each design. Bone height was defined and measured as the effective vertical (as opposed to the absolute) length of bone contact with implant, and appeared to differ qualitatively for the two designs on the mesial and distal aspects. Analyses of covariance were also done so that differences in either bone height or total available contact surface could be controlled for when the mesial and distal aspects of the two designs were compared. In each analysis, estimates of the adjusted mean contact length were obtained and compared by use of the t test.
Results. With the exception of one porous-coated implant, which did not become osseo-integrated during the initial healing period and was removed, all of the implants became securely fixed by bone ingrowth and remained so throughout the 18-month trial. This included the single porous-coated implant left in the animal from which the implant mentioned above was removed. This single implant was not included in the morphometric analyses, since it would have been subjected to different levels of occlusal stress than were the other implants, which in each case functioned as part of a two-unit bridge. It was also not possible that one of the threaded implants from another dog be used in the morphometric analyses, since this implant frac-
Downloaded from jdr.sagepub.com at University of Manitoba Libraries on June 5, 2015 For personal use only. No other uses without permission.
Vol. 69 No. 5
POROUS-COATED vs. THREADED DENTAL IMPLANTS
-50iAmm Fig. 4-A buccolingual section of a porous-coated implant after 18 months of function. The interfacial bone (arrows) has remodeled to a level at (L = lingual side) or just coronal to (B = buccal side) the machined surface-porous surface junction.
tured during the final two months of function. However, the contiguous, bridge-supporting implant in this animal was used in the morphometric assessment. Qualitative observations.-The preparation of histological sections was fraught with technical difficulties. While it was possible for us to obtain reproducibly-good-quality sections of bone and implant system with the porous-coated design, it was more difficult with the threaded design, where in some instances, particularly with mesial or distal sections, the implant would become separated and dislodged from the associated bone during section preparation. This technical complication never occurred with the porous-coated design, presumably because of the three-dimensional interdigitation with bone possible with these implants, compared with the two-dimensional threaded implant-bone interlock (Pilliar, 1986). The greatest technical difficulty, however, was encountered in the obtaining of reproducibly good sections of the gingival tissues associated with the suprabony components of the implant systems. Thus, very often, because of limitations in the petrographic grinding technique, the gingiva in the sections was too thick or too thin to permit detailed interpretation, for example, of the nature and extent of inflammatory infiltrate. However, recognition of the overall relationships of implant, gingiva, and bone was generally possible.
,
1141
9t;s
Fig. 5-A higher magnification of the buccal aspect of the implant in Fig. 4. Bone (B) has remodeled to a point (open arrow) just coronal to the machined surface-porous surface junction (curved arrow). While there may be some artefactual separation of the gingival tissues (G) and bone from the machined implant surface (C), it can be seen that a band of dense fibrous connective tissue (large arrows) separates the bone crest from the overlying gingival epithelium (E).
A representative buccolingual section of a porous-coated implant is presented in Fig. 4. The interfacial bone had remodeled to a level at (lingual, Fig. 4) or just coronal to (buccal, Fig. 4) the machined surface-porous surface junction of the implant. The gingival epithelium had migrated along the coronal third of the implant to a variable extent, but was always separated from the underlying bone crest by a dense band of fibrous connective tissue, the fibers and cells of which were oriented parallel to the implant surface (Fig. 5). The porous coat was ingrown with bone to a variable extent, depending upon the aspect of implant examined (Figs. 4 and 6a). Those pores not occupied by bone were filled with fibrous connective tissue that appeared to encapsulate some of the particles of the porous layer and to be oriented parallel to the substrate surface where it juxtaposed the solid implant core. Sections of a threaded implant are shown in Figs. 6b and 7. While threaded implants were submerged fully within bone at the time of implantation, in all cases, bone remodeling resulted in the exposure of the shoulder and associated undercut, and the top one or two threads. The gingival epithelium had migrated along the shoulder-associated undercut, and the exposed threads were in contact with dense fibrous tissue (Fig. 8), the cells and fibers of which were oriented parallel to the surfaces of the affected threads.
Downloaded from jdr.sagepub.com at University of Manitoba Libraries on June 5, 2015 For personal use only. No other uses without permission.
J Dent Res
DEPORTER et al.
1142
I.
0. 5 mm
]-.5 mm
Fig. 6-Distal sections of the porous-coated implant (a) shown in Fig. 4 and the threaded implant (b) shown in Fig. 7. For the porous-coated implant, bone (B) has remodeled to a point (open arrow) just coronal to the machined surface-porous surface junction. Only about half the length of the implant was required for it to become in-grown with bone on this aspect, in order to provide effective osseo-integration. For the threaded implant, bone (B) has remodeled to the level of the first thread (curved arrow), resulting in the exposure of the shoulder (S) and associated undercut. The gingival epithelium has migrated along the shoulder-associated undercut, while the exposed threads were in contact with fibrous connective tissue (black arrow). This tissue is shown in higher magnification in Fig. 8.
The remaining threads were embedded in bone to a variable extent, depending on the aspect of implant observed. Thus, in buccolingual sections (Fig. 7), all remaining threads were fully embedded in bone, whereas in mesial and distal sections, some of the threads were covered with fibrous tissue similar to that seen in relation to the top one or two threads (Fig. 6b). Morphometric observations. -As indicated in the "methods" section, we designed the sectioning technique to provide two buccolingual, two mesial, and two distal sections of each implant. Technical complications resulted in loss of some sections, as follows: There were 20 buccolingual sections from the 10 porouscoated implants. Eight of these implants were represented by two sections each, while one implant had only one buccolingual section, and another had three. The same implants were represented by 20 mesial and 19 distal sections. Twenty buccolingual sections were available from the 11 threaded implants. Three of the implants had only one buccolingual section, while one implant had three sections. The
m
May 1990
AE;0i,
Fig. 7-A buccolingual section of a threaded implant after 18 months of function. Bone has remodeled to the level of the first thread on the lingual aspect (L) and the second thread on the buccal aspect (B), as indicated by the arrows.
same implants were represented by 15 mesial and 20 distal sections. The means and standard deviations of both the absolute contact lengths and the contact length fractions (CLF) for each of the buccal and lingual aspects of both implant systems are shown in Table 1. For the porous-coated design only, there were significant differences for both the absolute values and for CLF (p
sponding vertical bone height, significant differences were observed, the absolute contact length being greater for any given bone height for the porous-coated design. Taken together, the data suggest that shorter implants may be used with the porous-coated design. J Dent Res 69(5):1138-1145, May, 1990
Introduction. Numerous authors have proposed the use of a porous surface to achieve stabilization of both orthopedic and dental endosseous implants (see Pilliar, 1986, for a review). Earlier publications from our laboratory (Deporter et al., 1986a, b, 1988) described a truncated conical-shaped dental implant made of Ti-6A1-4V that incorporated a porous surface geometry (pore size 50-200 lLrm) on both the root implant component and on the apical third of the transgingival collar. The latter was meant to encourage gingival connective tissue attachment. In fact, it provided the weak link in the system, resulting in contamination with bacterial plaque and what was considered to be progressive implant failure by seven months of function in 22 of 32 implants placed in dog mandibles. However, for the remaining ten non-contaminated implants, it was obvious that when the porous coat of both root and collar components was submerged largely or entirely within bone, the implants functioned as intended, resulting in implant stabilization and effective stress transfer between the root implant component and the bone. Therefore, the design of the system was changed so that the porous surface configuration would be limited to the apical two-thirds of the root component. The present report describes the histological findings of an 18-month (in function) clinical trial of this modified porouscoated implant in the dog. The experimental design included Received for publication July 3, 1989 Accepted for publication November 1, 1989 This work was supported by Program Grant #26 from the Medical Research Council of Canada.
1138
comparison with a threaded implant design similar, but not identical, to the Brdnemark implant (Branemark et al., 1977), since a morphometric histological assessment of threaded, endosseous dental implants has not been previously published. It was anticipated that these two designs with differing surface configurations would result in different patterns of stress distribution and resultant bone remodeling at the bone-implant interface.
Materials and methods. The components of both endosseous dental implant systems were made by machining (Strite Industries, Cambridge, Ontario), with the porous coat being added subsequently in our laboratory, through procedures developed by one of us (Pilliar, 1987). The porous-coated root component (Fig. la) was fabricated from Ti-6A1-4V (ASTM F 136), rather than from commercially pure Ti (CP Ti), because the former is known to have greater strength. It consisted of a truncated conical shape (taper angle = 50) with the porous coat limited to the apical two-thirds, the coronal third having a machined surface, with a surface finish better than 1.75-km cla (center line average). Figs. lb and ic show scanning electron micrographs of this coronal region. Fine machining lines can be seen running circumferentially. A thermal etching pattern can also be seen superimposed on the machining lines. The high magnification view (Fig. lc) of the sample edge profile confirms the small cla surface finish. (Examination of the threaded implants showed similar fine machining lines on the surfaces of the threads and on the non-threaded coronal portion.) For achievement of a continuous contour at the machined-to-porous-coated surface junction, the apical two-thirds was reduced with a 0.25-mm in-step relative to the coronal third. This permitted the subsequent placement of a porous coat on the outer surface, which was approximately flush with the coronal third. The self-tapping threaded implant was made from CP Ti (ASTM F 67)so that it would be similar to the Brdnemark implant-was cylindrical and had an effective thread angle of 600. The pitch of the threads was 0.635 mm, the number of threads per cm was 15.74, and the thread depth was 0.787 mm. The diameter of the implant shoulder (Fig. la) was equal to 3.53 mm and was slightly larger than the collar diameter (3.50 mm). The other components of both implant systems were identical in design (Fig. 2) and were machined as required from either Ti-6AI-4V or pure titanium. All of the components were cleaned and sterilized as described previously (Deporter et al.,
1986b). Six in-bred beagle dogs, two males and four females (Laboratory Research Enterprises, Kalamazoo, MI), initial weights 9.8 to 17.3 kg, were used. Implants were placed bilaterally in the region of the mandibular third and fourth premolars, which had been rendered edentulous at least six months prior to implantation. Two porous-surfaced implant root components were placed on the left side, and two threaded components were placed on the right side of the mandible of each dog. (The two
Downloaded from jdr.sagepub.com at University of Manitoba Libraries on June 5, 2015 For personal use only. No other uses without permission.
Vol. 69 No. 5
POROUS-COATED vs. THREADED DENTAL IMPLANTS
1139
a
l1"mm
Fig. 1-(a) The porous-coated (left) and threaded (right) root implant components used in the study. (b) Low (x5O) and (c) high (x5OO) magnification scanning electron micrographs of the machined coronal region of the partially porous-coated implant. The lower-magnification micrograph shows circumferential machining lines. The higher-magnification view shows an edge profile of these lines. A thermal etching pattern is also seen on this micrograph.
implants on each side were referred to as the mesial and distal implants, respectively.) The porous-coated implants were inserted, as previously described (Deporter et al., 1986b), following elevation of a fullthickness mucoperiosteal flap. A preliminary channel was cut with a twist drill (1.524-mm diameter; Whaledent Inc., New York, NY), and this was subsequently enlarged with a custommade, tungsten-carbide bur (Beaver Dental Products Ltd., P.O. Box 900, Morrisburg, Ontario) in a speed-reducing handpiece operated at a maximum speed of 120 rpm under external saline
irrigation. After thorough irrigation of each prepared site, implants were tapped into position with a mallet and a Teflontipped punch. The threaded implants were placed by means of a similar flap approach and initial site preparation with the same twist drill. Each site was then enlarged with a customfabricated carbon-steel twist drill of larger diameter (2.362 mm) corresponding to that of the implant without its threads. Finally, each site was threaded with a manual tap of standard carbide steel so that the threaded implant would be received. All of these manipulations were done under copious saline irrigation. The flaps were sutured with 4-0 Vicryl (Ethicon Sutures Ltd., Peterborough, Ontario). The animals were maintained on a soft diet (Pep dog food, Quaker Oats Co., Peterborough, Ontario) for the first two weeks post-operatively, and were given Penlong XL at appropriate intervals during this period. After an initial healing period of six weeks, all root components were uncovered, and each received a collar and collarretaining screw. At this time, one porous-coated root component in one dog had not become fixed by bone ingrowth, for some unknown reason, and was removed. This implant was not replaced, and the remaining porous-coated implant was left to function as a free-standing implant. The other 22 implants were used as support for two-unit fixed bridges for an 18month functional period. Following the 18-month functional period, the animals were
Downloaded from jdr.sagepub.com at University of Manitoba Libraries on June 5, 2015 For personal use only. No other uses without permission.
J Dent Res May 1990
DEPORTER et l.
1140
BUCCAL Buccal- Lingual
Dist~al Buccal,'
i
Mesial Buccal
DISTAL
MESIAL Distal Lingual
C_;
AO_
Fig. were
(B),
transgingival components used for both implant systems design, and included a collar (A), a collar-retaining screw coping (C), and a coping-retaining screw (D). 2-The
identical in a
killed under general anesthesia by bilateral carotid perfusion with a solution of 37% formalin and 99% methanol in distilled water (1:1:1.5 V/V). Segments of mandible containing the implants were removed, processed, and embedded in methylmethacrylate, and subsequently sectioned for histological assessment, as previously described (Deporter et al., 1986b, 1988). Briefly, blocks were sectioned with a Buehler Isomet lowspeed saw (Tech-Met Canada) equipped with a diamond wafering blade. The sectioning technique was designed so that each implant would have two buccolingual, two mesial, and two distal sections (Fig. 3). The initial section thickness of 120-150 ALm was reduced to approximately 40 Elm by means of petrographic grinding techniques, and the sections were stained with a mixture of Stevenel's blue and van Gieson's picro-fuchsin. All of the sections were examined both qualitatively and by computer-assisted morphometry (Deporter et al., 1986b, 1988). This allowed for the determination of the length of implant surface in direct contact with the bone on each aspect (i.e., buccal vs. lingual vs. mesial vs. distal) of each implant. Since the porous-coated implants were designed to be stabilized by bone in-growth into the porous surface, measurements for these implants did not include data from the coronal third region. For these implants, the length of implant surface in direct contact with bone was determined on a digitizing tablet by tracing the absolute length of bone contact with the outer surfaces of the spheres of the outermost layer of porous coat. In contrast, since the threaded implants were designed and inserted so that all of the threads for implant stabilization would be used, the
4.Mesial Lingual
Fig. 3- -A diagram depicting the sectioning technique used.
full length of the threaded implant surface was included, and the length of implant surface in contact with bone was determined by the absolute length of bone contact with the surfaces of the threads being traced. These data were expressed in two ways: first, as an absolute length, and second, as a fraction of the maximum length of implant surface available for contact with bone. The fraction figure was referred to as the contact length fraction (CLF), and is expressed as follows: (absolute length contacted by bone) CLF (length available for contact with bone) For each aspect, we examined the characteristics of the distributions of measures of absolute contact length and contact length fraction by type of implant and aspect of implant, calculating the mean, standard deviation, and median. Because these descriptive data suggested that the two implant designs behaved differently on buccal and lingual aspects, the two designs were compared (separately), as described below, for each of these aspects of the implants. Likewise, the mesiobuccal-distobuccal and mesiolingual-distolingual aspects for the two designs were compared separately. Analyses of variance were done for comparison of absolute contact length and contact length fraction for the relevant aspects of the two implant designs with variation between dogs controlled for. In addition, for the mesial and distal aspects, analyses of variance were done for comparison of the ratio of absolute contact length with the corresponding bone height for each design. Bone height was defined and measured as the effective vertical (as opposed to the absolute) length of bone contact with implant, and appeared to differ qualitatively for the two designs on the mesial and distal aspects. Analyses of covariance were also done so that differences in either bone height or total available contact surface could be controlled for when the mesial and distal aspects of the two designs were compared. In each analysis, estimates of the adjusted mean contact length were obtained and compared by use of the t test.
Results. With the exception of one porous-coated implant, which did not become osseo-integrated during the initial healing period and was removed, all of the implants became securely fixed by bone ingrowth and remained so throughout the 18-month trial. This included the single porous-coated implant left in the animal from which the implant mentioned above was removed. This single implant was not included in the morphometric analyses, since it would have been subjected to different levels of occlusal stress than were the other implants, which in each case functioned as part of a two-unit bridge. It was also not possible that one of the threaded implants from another dog be used in the morphometric analyses, since this implant frac-
Downloaded from jdr.sagepub.com at University of Manitoba Libraries on June 5, 2015 For personal use only. No other uses without permission.
Vol. 69 No. 5
POROUS-COATED vs. THREADED DENTAL IMPLANTS
-50iAmm Fig. 4-A buccolingual section of a porous-coated implant after 18 months of function. The interfacial bone (arrows) has remodeled to a level at (L = lingual side) or just coronal to (B = buccal side) the machined surface-porous surface junction.
tured during the final two months of function. However, the contiguous, bridge-supporting implant in this animal was used in the morphometric assessment. Qualitative observations.-The preparation of histological sections was fraught with technical difficulties. While it was possible for us to obtain reproducibly-good-quality sections of bone and implant system with the porous-coated design, it was more difficult with the threaded design, where in some instances, particularly with mesial or distal sections, the implant would become separated and dislodged from the associated bone during section preparation. This technical complication never occurred with the porous-coated design, presumably because of the three-dimensional interdigitation with bone possible with these implants, compared with the two-dimensional threaded implant-bone interlock (Pilliar, 1986). The greatest technical difficulty, however, was encountered in the obtaining of reproducibly good sections of the gingival tissues associated with the suprabony components of the implant systems. Thus, very often, because of limitations in the petrographic grinding technique, the gingiva in the sections was too thick or too thin to permit detailed interpretation, for example, of the nature and extent of inflammatory infiltrate. However, recognition of the overall relationships of implant, gingiva, and bone was generally possible.
,
1141
9t;s
Fig. 5-A higher magnification of the buccal aspect of the implant in Fig. 4. Bone (B) has remodeled to a point (open arrow) just coronal to the machined surface-porous surface junction (curved arrow). While there may be some artefactual separation of the gingival tissues (G) and bone from the machined implant surface (C), it can be seen that a band of dense fibrous connective tissue (large arrows) separates the bone crest from the overlying gingival epithelium (E).
A representative buccolingual section of a porous-coated implant is presented in Fig. 4. The interfacial bone had remodeled to a level at (lingual, Fig. 4) or just coronal to (buccal, Fig. 4) the machined surface-porous surface junction of the implant. The gingival epithelium had migrated along the coronal third of the implant to a variable extent, but was always separated from the underlying bone crest by a dense band of fibrous connective tissue, the fibers and cells of which were oriented parallel to the implant surface (Fig. 5). The porous coat was ingrown with bone to a variable extent, depending upon the aspect of implant examined (Figs. 4 and 6a). Those pores not occupied by bone were filled with fibrous connective tissue that appeared to encapsulate some of the particles of the porous layer and to be oriented parallel to the substrate surface where it juxtaposed the solid implant core. Sections of a threaded implant are shown in Figs. 6b and 7. While threaded implants were submerged fully within bone at the time of implantation, in all cases, bone remodeling resulted in the exposure of the shoulder and associated undercut, and the top one or two threads. The gingival epithelium had migrated along the shoulder-associated undercut, and the exposed threads were in contact with dense fibrous tissue (Fig. 8), the cells and fibers of which were oriented parallel to the surfaces of the affected threads.
Downloaded from jdr.sagepub.com at University of Manitoba Libraries on June 5, 2015 For personal use only. No other uses without permission.
J Dent Res
DEPORTER et al.
1142
I.
0. 5 mm
]-.5 mm
Fig. 6-Distal sections of the porous-coated implant (a) shown in Fig. 4 and the threaded implant (b) shown in Fig. 7. For the porous-coated implant, bone (B) has remodeled to a point (open arrow) just coronal to the machined surface-porous surface junction. Only about half the length of the implant was required for it to become in-grown with bone on this aspect, in order to provide effective osseo-integration. For the threaded implant, bone (B) has remodeled to the level of the first thread (curved arrow), resulting in the exposure of the shoulder (S) and associated undercut. The gingival epithelium has migrated along the shoulder-associated undercut, while the exposed threads were in contact with fibrous connective tissue (black arrow). This tissue is shown in higher magnification in Fig. 8.
The remaining threads were embedded in bone to a variable extent, depending on the aspect of implant observed. Thus, in buccolingual sections (Fig. 7), all remaining threads were fully embedded in bone, whereas in mesial and distal sections, some of the threads were covered with fibrous tissue similar to that seen in relation to the top one or two threads (Fig. 6b). Morphometric observations. -As indicated in the "methods" section, we designed the sectioning technique to provide two buccolingual, two mesial, and two distal sections of each implant. Technical complications resulted in loss of some sections, as follows: There were 20 buccolingual sections from the 10 porouscoated implants. Eight of these implants were represented by two sections each, while one implant had only one buccolingual section, and another had three. The same implants were represented by 20 mesial and 19 distal sections. Twenty buccolingual sections were available from the 11 threaded implants. Three of the implants had only one buccolingual section, while one implant had three sections. The
m
May 1990
AE;0i,
Fig. 7-A buccolingual section of a threaded implant after 18 months of function. Bone has remodeled to the level of the first thread on the lingual aspect (L) and the second thread on the buccal aspect (B), as indicated by the arrows.
same implants were represented by 15 mesial and 20 distal sections. The means and standard deviations of both the absolute contact lengths and the contact length fractions (CLF) for each of the buccal and lingual aspects of both implant systems are shown in Table 1. For the porous-coated design only, there were significant differences for both the absolute values and for CLF (p
