Bone apposition to plasma-sprayed cobalt-chromium alloy Hugh A. Luckey* Orthopedic Products Division/3M, 270-4N-093M Center, St. Paul, Minnesota 55144 Elden G . Lamprecht Surgical G. Histopathological Services/3M, 270-3s-06 3M Center, St. Paul, Minnesota 55144 Michael J. Waltt Orthopedic Products Division/3M, 270-4N-093M Center, St. Paul, Minnesota 55144 The use of porous metallic coatings for fixation of total joint prostheses by bone ingrowth has become a widespread altern a t i v e t o f i x a t i o n w i t h PMMA b o n e cement. However, concerns about such coatings include long-term effects of metal ion release, potential coating loss, and decreased substrate fatigue strength. T h e biological fixation capability of a nonporous, h i g h - i n t e g r i t y plasmasprayed CoCr coating with low surface area was compared to a conventional sintered bead coating in goat cortical and cancellous bone sites after 8 and 16 weeks of implantation. Histological evaluation showed substantial variations in fixation quality between individual animals and between surgical sites with no consistent difference between implant types. Shear testing of bonelimplant interfaces showed
that although conventional porous coati n g exhibited higher overall average shear strengths in cortical bone sites at both time periods, the differences were not statistically significant. In cancellous sites, the average shear strengths achieved with conventional porous and plasmasprayed coatings were essentially equal. Analysis using average paired differences, however, revealed that when porous and plasma-coated implants are placed in identical sites of contralateral limbs, the plasma coatings consistently yielded higher shear strengths in cancellous bone sites at the later time period. Since current design theory for biological fixation favors metaphysical fixation, this surface may offer potential advantages over conventional porous coatings.
INTRODUCTION
Clinical dissatisfaction with (po1y)met hylme t hacry late (PMMA) bone cement has led to the increased use of implants with porous coatings to anchor total joint prostheses. A number of concerns have been raised, however, regarding the long-term performance of implants with porous surface coatings. These include the effects of increased metal ion release due to increased *Current address: Materials Engineering, Ltd., 1386 Pine View Trail, St. Joseph, WI 54082. 'To whom correspondence should be addressed. Journal of Biomedical Materials Research, Vol. 26, 557-575 (1992) 0 1992 John Wiley & Sons, Inc. ccc 0021-9304/92/050557-19$4.00
LUCKEY, LAMPRECHT, A N D WALT
558
surface area,'" a reduction in substrate fatigue strength due to high temperature sintering,"".'" and the potential for coating loss.'z7 A rough, dense CoCr (ASTM F75 alloy) coating using plasma arc spraying techniques has been developed which addresses these concerns." This work investigates the capability of this coating to provide a surface for direct biological fixation with bone in an animal model.
METHODS A N D MATERIALS
Specimen fabrication Cylindrical test plugs (6.3 mm diameter x 12.7 mm length) fabricated from CoCrMo alloy, conforming to ASTM F75, were prepared with the following three surface structures for in vivo implantation: (a) a dense plasma-sprayed (PS) F75 alloy coating, (b) a porous-beaded (PC) F75 alloy coating, and (c) a smooth (S) abrasive grit blasted surface. Different initial substrate diameters were used in order to produce specimens with approximately the same finished diameter. Plasma-sprayed specimens were made as previously described.' The coatings were created by plasma arc spraying 300-420-pm spherical powder on to hot isostatically pressed (HIP) cylindrical substrates. These specimens were subsequently heat treated at conventional solution treating temperatures to bond the coating to the substrate. A final plasma-sprayed coating thickness of 100 pm was obtained, which exhibited an RMS surface roughness of approximately 25 pm. No interconnected porosity was evident, although isolated microcrevices of up to 50 pm were noted (Fig. l). The porous-beaded specimens were prepared by vacuum sintering 300400-pm R.E.P. powder (Nuclear Metals, Inc., Concord, Ma) to cast F75 alloy substrates at approximately 1300°C.A two-layer beaded structure was created with a resulting pore size estimated from scanning electron microscopy (SEM) to be around 250-300 pm. Pore volume was estimated to be 30% (Fig. 2). Specimens with a smooth surface were fabricated by normal implant manufacturing techniques. Final abrasive grit blasting produced an RMS surface roughness of 1-2 pm (Fig. 3). All test specimens were cleaned and passivated per ASTM F86 prior to implantation.
Surface characterization Characterization of the resulting dense plasma-spray-coated surface and processing effects on the substrate have been reported previously.' No decrease in substrate fatigue strength was evident. The increase in the surface area as a result of the coating roughness was estimated from scanning electron micrographs to be 200-300%. This compares with typical surface area
BONE APPOSITION
Figure 1. SEM of plasma-sprayed (PS) implant surface.
Figure 2. SEM of porous-beaded (PC) implant surface.
559
LUCKEY, LAMPRECHT, A N D WALT
560
Figure 3. SEM of smooth (S) implant surface.
increases for sintered bead coatings of 700-800%. Corrosion rate increases of only 20% were seen for the dense plasma sprayed coating compared to 270% increases for the sintered bead coating3 Specimen implantation The implant plugs were placed transversely through a single cortex in both hind legs of eight adult goats. Three cortical and three cancellous nonloaded sites were chosen (Table I). The cancellous sites were at the distal femoral condyle (DFC), the anterior proximal tibia (APT), and the posterior proximal tibia (PPT). The cortical sites were at the proximal tibia diaphysis (PTD), the mid-tibia1 diaphysis (MTD), and the distal tibia1 diaphysis (DTD) (Fig. 4). Precision reamers were used to prepare each site to achieve a uniform 2% interference fit between the implant and bone. The plug was inserted with the outer end flush with the periosteum. After recovery from surgery, the animals had unrestricted ambulation in an outdoor environment. Sacrifice was performed at 8 and 16 weeks postoperatively. NIH guidelines for the care and use of laboratory animals (NIH Publication #85-23 Rev. 1985) have been observed. Mechanical testing At sacrifice, each hind limb designated for mechanical push-out testing was stripped of soft tissue and frozen at -20°C. Prior to testing, each limb
561
BONE APPOSITION TABLE I Implantation Schedule in Goat Hind Limbs Animal Number and Limb 1&5
Cancellous Sites APT PPT DFC Cortical Sites PTD MTD DTD
3&7
2&6
4&8
R
L
R
L
R
L
R
L
S*
PS S PC
PC PS S
S PC PS
PC PS PC
PS PC PS
PS PC PS
PC PS PC
PS S PC
PC PS S
S PC
PC PS PC
PS PC PS
PS PC PS
PC PS PC
PC" PS* S*
PC* PS*
PS
PS
= Plasma-sprayed; PC = porous-coated; S = smooth. *Histology specimens.
Figure 4. Radiograph of implant sites in goat hind limbs.
LUCKEY, LAMPRECHT, AND WALT
562
was thawed and the bone around each implant sectioned transversely. Any periosteal bone which had formed over the plugs was carefully dissected away, and each section containing an implant was bisected perpendicular to the axis of the implant, exposing each end of the implant cylinder. Care was taken to leave a minimum thickness of 6 mm-10 mm between the implant and the adjacent transverse section cut. Axial push-out testing was performed using the method of Bobyn' at a displacement rate of 25.4 mm/min. After push-out testing, the bone was sectioned at the implant site using a low-speed diamond saw. Measurements of bone thickness taken at 90" intervals around the circumference of the implant site were used to estimate the area of bone contact with the implant. Several modifications were made to the above test method to assure axial alignment of the specimen with the load application axis. Each cylindrical implant was fabricated with a 1.6 mm diameter axial hole and a 6.4 mm radius, partially spherical indentation at one end. Upon completion of test specimen preparation, the implant was placed over a centering pin in the test fixture with the longitudinal cut resting on the fixture. A small donut of bone cement was used to fill the gaps between the fixture and the bone, and the cement cured in place. The centering pin was removed, a 4.75-mm ball bearing placed in the indentation of the implant, and the load applied through the ball bearing. All testing was performed using an Instron 1321H with calibration traceable to National Institute of Standards and Technology (NIST).
Histological examination One limb was selected from each time period for histological examination as denoted in Table I. In addition, two cortical and two cancellous specimens each were selected from the mechanical testing group after push-out testing. The post-mechanical specimens selected were those which exhibited the highest measured shear strengths for the PS and PC surface types. Implants and attached bone were fixed in formaldehyde, dehydrated in graded alcohol, and embedded in PMMA. The embedded specimens were subsequently thick sectioned transversely, polished to approximately 100 pm, stained with toluidine blue and picrofuchsin, and viewed by transmission light microscopy.
RESULTS
Mechanical testing The attachment values for the smooth implants (S) were substantially lower than those for either the plasma-sprayed or the porous-coated implants. The shear strengths for the smooth surfaced specimens ranged from 0.0 to 3.3 MPa with average values of 1.07 & 0.59 MPa in cancellous bone and 1.22 k 1.62 MPa in cortical bone. Because of the low interface strength, some
BONE APPOSITION
563
specimens may have been dislodged during preparation for mechanical testing. The low shear strength values also indicate that these specimens acted as negative controls. Therefore, the results from the smooth specimens will be excluded from further discussion. Although the shear strengths obtained compare well with those reported in the literature,”’ the values ranged from 0.3 to 31.9 MPa. Interfacial shear strengths in cancellous bone ranged from 3.5 to 9.2 MPa at 8 weeks and ranged from 1.7 to 12.3 MPa at 16 weeks (Tables 11, I11 and Fig. 5). In cortical bone, shear strengths of 0.3 to 10.7 MPa were recorded for the 8-week specimens, while shear strengths of 4.7 to 31.9 MPa were seen at 16 weeks (Tables IV, V and Fig. 6). The average interfacial shear strengths for all sites, surface types, and time periods are also shown (Fig. 7).
Time effects In cancellous bone, there was a trend toward increasing average shear strengths with time at most sites and with both surface types (Table VI and Fig. 7). The average shear strength for all cancellous sites with porous-coated specimens increased from 6.0 MPa at 8 weeks to 7.2 MPa at 16 weeks. A similar increase from 5.5 MPa to 7.2 MPa was seen for the plasma-sprayed specimens. However, the differences in the average strengths for the cancellous sites between time periods were not statistically significant for either surface type ( p 5 .27) or individual site. TABLE I1 Eight-Week Mechanical Testing Data in Cancellous Bone Animal 5L
6L 6R
7L 7R 8L 8R
Implant Site
Surface Treatment
Push-Out Force ( N )
Contact Area (mm’)
Shear Strength (MPa)
APT PPT DFC APT PPT DFC APT PPT DFC APT PPT DF C APT PPT DFC APT PPT DFC APT PPT DFC
PS S PC S PC PS PC PS S PS PC PS PC
1138 245 1080 89 1468 916 334 1112 210 1001 1334 1334 734 845 1500 613 801 1458 712 970 1201
203 165 190 189 184 173 178 179 173 175 181 168 144 181 164 184 179 188 203 184 199
5.60 1.48 5.68 0.47 8.00 5.30 1.88 6.23 1.21 5.72 7.37 7.96 5.10 4.66 9.14 3.34 4.48 7.77 3.50 5.28 6.05
rs
PC PC PS PC PS PC PS
LUCKEY, LAMPRECHT, AND WALT
564
TABLE I11 Sixteen-Week Mechanical Testing Data in Cancellous Bone Animal
Implant Site
Surface Treatment PS S PC S PC PS PC PS S PS PC PS PC PS PC PS* PC* PC PS PC PS
APT PPT DFC APT PPT DFC
1L
2L 2R
APT
3L
DFC APT
3R
DFC APT
4L
DFC APT
4R
DFC APT
PPT
PPT PPT
PPT
PPT DFC
Push-Out Force ( N )
Contact Area (mm2)
Shear Strength (MPa)
845 222 1735 44 1201 1535 801 2091 356 1059 1015 2594 855 1401 1245 317 890 1312 396 1700 2317
181 190 185 163 182 194 159 188 194 173 161 211 166 20 5 166 188 141 167 203 167 195
4.68 1.17 9.38 0.28 6.60 7.93 5.05 11.17 1.83 6.12 6.32 12.28 5.15 6.83 7.52 1.68 6.32 7.86 1.95 10.14 11.93
~
*%sface types switched at time of implantation.
15
1
T
APT
PPT
DFC
Implantation Sites PC-8
PC-16
PS-8
a
PS-16
Figure 5. Average interfacial shear stresses with standard deviations for cancellous sites.
BONE APPOSITION
565
TABLE IV Eight-Week Mechanical Testing Data in Cortical Bone Animal
Implant Site
Surface Treatment
Push-Out Force ( N )
Contact Area (mm2)
Shear Strength (MPa)
PTD MTD DTD PTD MTD DTD PTD MTD DTD PTD MTD DTD PTD MTD DTD PTD MTD DTD PTD MTD DTD
PS S PC
1312 614 1592 0 858 378 1134 912 543 58 165 67 1432 872 467 823 734 734 614 925 734
123 188 160 148 184 86 180 163 74 167 195 156 188 157 98 195 152 82 145 189
10.69 3.26 9.96 0.00 5.79 2.06 13.24 5.07 3.34 0.79 0.99 0.34 9.17 4.64 2.97 8.41 3.77 4.82 7.52 6.37 3.88
~
5L
6L 6R
7L
7R 8L 8R
S PC PS PC PS S PS PC PS PC PS PC PC PS PC PS PC PS
TABLE V Sixteen-Week Mechanical Testing Data in Cortical Bone Animal 1L 2L 2R
3L 3R 4L 4R
Implant Site
Surface Treatment
Push-Out Force ( N )
Contact Area (mm’)
Shear Strength (MPa)
PTD MTD DTD PTD MTD DTD PTD MTD DTD PTD MTD DTD PTD MTD DT D PTD MTD DTD PTD MTD DTD
PS S PC PS PC S PC PS S PS PC PS PC PS PC PC PS PC PS PC PS
623 111 1815 1379 2246 0 1357 2002 22 1970 3060 2313 1059 2260 3336 383 2335 1873 801 4448 1277
132 193 197 128 176
4.71 0.58 9.24 10.83 12.76 0.00 14.14 9.86 0.11 12.21 18.28 11.59 13.24 13.17 21.38 6.90 11.59 11.45 7.66 31.93 6.30
-
96 203 199 162 168 199 80 172 156 55 201 163 105 139 203
LUCKEY, LAMPRECHT, AND WALT
566
Y
01
25
m
L
tj
20
5
-
f
15
Q)
2
10
0
> a
5 0
PTD
MTD
=
nPC-8
DTD
Implantation Sites
PC-16
PS-8
PS-16
Figure 6. Average interfacial shear stresses with standard deviations for cortical sites.
T
Q)
0
E 5
a
0 Cancellous
aPC-8
=
Cortical Implantation Sites PC-16
PS-8
PS-16
Figure 7. Average interfacial shear stress with standard deviations for all sites, surface types, and time periods.
In cortical bone, there was a statistically significant increase in the overall average shear strength for the porous-coated (PC) and the plasma-sprayed (PS) implants in all sites between the 8-week and 16-week time periods. (Table VI and Fig. 7). The overall average shear strengths increased from 6.9 MPa to 15.5 MPa for the PC specimens ( p 5 .011*), while the overall aver-
*“g
Overall
DTD
MTD
Overall A% Cortical Sites PTD
DFC
PPT
Cancellous Sites APT
0.23 (k0.31)
9.77 (22.90) 4.31 (23.26)
2.20 (k1.90)
0.06 ( 2 .08) 3.34
15.48 (27.55)
0.58 3.26
6.86 (k3.76)
-
0.00
8.85 (k3.35) 11.54 (k1.66) 8.94 (23.74)
6.33 (k5.06) 4.49 (50.66) 2.09 (21.77)
11.43 (k3.95) 20.99 (k9.87) 14.02 (k7.47)
10.27 (k2.60) 4.38 (k2.95) 5.92 (k3.62)
1.83
1.21
1.09 (50.78)
1.17
0.28
S [16 wk]
1.48
0.47
S [8 wk]
1.05 (k0.52)
5.50 (21.26)
7.15 (k1.76)
5.96 (k2.36)
3.61 (22.15) 9.00 (23.06) 10.71 (22.42)
PS [16 wk]
7.17 (k4.03)
4.94 (k1.25) 5.12 (k0.96) 6.44 (21.37)
PS [8 wk]
5.10 (20.07) 7.35 (21.87) 8.25 (k0.99)
PC [16 wk]
3.44 (k1.61) 6.88 (k1.42) 7.55 (21.74)
PC [8 wk]
TABLE VI Average Shear Strengths (MPa) and Standard Deviations by Implantation Site
E
CJ
c
2 m
568
LUCKEY, LAMPRECHT, AND WALT
age shear strengths for PS specimens increased from 4.31 to 9.77 MPa ( p 5 ,002). An increase in shear strength with time was also seen at each site for both surface types. However, the only statistically significant increase was at the MTD site for the plasma-sprayed surface ( p 5 .02).
Surgical site effects Some site-to-site differences in the push-out results were noted. In cancellous bone, the APT site was significantly lower in strength than either the PPT or the DFC sites for all porous-coated samples at both time periods ( p 5 .05). The trend was similar for plasma-sprayed specimens, but only the APT site was significantly lower than the DFC site at the 16 weeks (p 5 0.16). In the cortical sites, there was no clear trend, although the MTD site tended to have a lower strength than the PTD or DTD sites at 8 weeks, and a higher strength at 16 weeks.
Surface type effects There were no statistically significant differences between surface types at the same time period. However, the porous-coated specimens exhibited higher average strengths at two of three cortical sites at 8 weeks and at all 16-week cortical sites.
Animal effects Marked differences in shear strengths occurred among animals for both surface types. Most animals gave either consistently high or low values for both surface types. For any given animal, substantial site-to-site variability in shear strength was also observed for each type of bone. Implants in the MTD sites tended to give the highest shear strengths among cortical sites. Among cancellous sites, the implants in the DFC sites gave consistently higher values.
Paired site analysis The surgical implantation schedule was designed to randomize specimens of the two surface types with regard to the various surgical sites in each type of bone. Although insufficient data existed for a precise determination, interactions of time, site, and coating type seemed to exist. Because of the greater than expected variability among animals and surgical sites, any actual differences in bone response between surface types may have been masked. In order to directly compare the response between surface types, data from animals with PS and PC specimens in identical sites in contralateral limbs were examined separately. The average percentage shear strength differences be-
BONE APPOSITION
569
tween paired sites were calculated according to the formula shown below and are listed in Table VII.
% paired difference = [(PS - PC)/PC] x 100 PS = average shear strength for plasma-sprayed specimens PC = average shear strength for porous-coated specimens With the exception of the 16-week APT site, at least two pairs existed for each site at each time period. Push-out strengths were generally lower in cortical bone for PS specimens than for PC specimens at both time periods resulting in a negative percent paired difference. At 8 weeks, percent paired differences ranged from +2.5% to -52.7%. At 16 weeks, the percent paired differences ranged from -10.5% to -45.5%. This agrees with the overall averages. Agreement with overall averages for cancellous bone was limited to the 8week time period. The overall average shear strengths were similar for both surface types at both time periods. The percent paired differences were lower for the PS specimens than for PC specimens at 8 weeks (+9.2% to -25.6%) while at 16 weeks, the PC specimens exhibited substantially higher percent paired differences (+18.8% to 57.3%).Thus, the relative strengths were reversed at the longer time period.
+
Histology results Substantial variations in the extent of new bone growth between animals and among surgical sites were found in the region between preexisting bone and the implant surface. These variations appeared to result from differences in the initial wave of bone formation rather than subsequent bone resorption or remodeling. All surface types, however, exhibited areas of apparent direct bone contact as well as areas with interposed fibrous tissue. The amount of bone apposition varied among the specimens. The 8-week exposure specimens had prolific callus and new bone formation at all sites. New woven bone filled the space between the preexisting TABLE VII Average Percent Shear Strength Differences for Contralateral Paired Sites by Time Period Cancellous Site 8 week
APT PPT DFC 16 week APT PPT DF C
Cortical
% Diff.
N Pair
Site
% Diff.
N Pair
+ 9.2 -25.6 -17.3
2 3 2
PTD MTD DTD
-52.7 + 2.5 -45.8
2 3 2
i-18.8
1 2 2
PTD MTD DTD
-10.5 -45.0 -45.5
3 3 2
+39.3 +57.3
LUCKEY, LAMPRECHT, AND WALT
570
bone and the PS (DFC) (Fig. 8) and PC (PPT) (Fig. 9) implants in the cancellous sites. Direct bone apposition was also observed on one-third of the surface of the smooth implant (APT). Bone contact was limited to areas of callus around the PC (MTD) and PS (DTD) implants in the cortical sites. The 16-week exposure individual was not as prolific in forming bone, New bone formation led to direct bone contact at only pinpoint locations on three of the six implants. All other sites exhibited a collagen repair tissue layer interposed between the preexisting bone, or the new bone, and the implant (Fig. 10). Bone ingrowth into the porous implants was limited to less than one-half the available space.
Histology of push-out specimens The four push-out specimens exhibited freezing and thawing artifacts of soft tissue but had excellent preservation of bone repair results. Bone adhered to the surfaces of one PC and one PS implant to varying degree. Bone was intimately apposed to the bulk of the free surface of a plasmasprayed plug in a cancellous DFC site (Fig. 11).Bone made periodic direct contact with the metal surface. At its greatest extent, up to 4 mm of bone remained undisturbed in close apposition to the implant. The apposed bone exhibited occasional microscopic bends or fractures, but remained largely intact.
Figure 8. Bone apposition to plasma-sprayed (PS) surface in cancellous bone (DFC) at 8 weeks (X10).
BONE APPOSITION
Figure 9. Bone ingrowth into porous bead (PC) surface in cancellous bone (PPT) at 8 weeks ( ~ 1 0 ) .
Figure 10. Collagen repair tissue filling the space between original bone and porous bead (PC) surface in cancellous bone (PPT) at 16 weeks (original magnification ~ 1 0 ) .
571
572
LUCKEY, LAMPRECHT, AND WALT
Figure 11. Bone fragment remaining attached to the plasma-sprayed (PS) surface after push-out from cancellous bone (DFC) at 16 weeks (original magnification ~ 1 0 ) .
Bone sheared at the outer bead surface of the porous-coated plug in the cortical MTD site, thus leaving negligible bone attached to the outer surface (Fig. 12). However, bone had invaded approximately one-half of the available pore space below the outer circumference of the implant.
Figure 12. Bone sheared at the outer edge of the porous bead (PC) surface after push-out from cortical bone (MTD) at 16 weeks (original magnification ~ 1 0 ) .
573
BONE APPOSITION DISCUSSION
The subgross histological observation of prolific callus and new bone formation correlated well with the high push-out strength seen in the contralateral limb of one 8-week individual. The highest shear strength for smooth implants was also recorded at 3.3 MPa (MTD) and 1.5 MPa (PPT) for that individual. Shear strengths for PS and PC surfaces were nearly equal within the same type of bone. In the cortical sites, the PC strength was 10.7 MPa (PTD), while the PS strength was 9.9 MPa (DTD). In the cancellous sites, the PC strength was 5.7 MPa (APT) with the corresponding PS strength at 5.6 MPa. The lack of observed bone apposition in the 16-week individual corresponded to lower measured shear strengths. The shear strength values for the smooth implants for this 16-week individual (0.7 MPa at the MTD site) were lower than for the 8-week individual (1.2 MPa at the PPT site). Also, the PS implants in the PTD and APT sites gave strengths approximately 50% that of the PC implants in the DTD and DFC sites (4.7 MPa vs. 9.2 MPa). The paired comparisons between identical implant surfaces in other animals in the above sites suggest that these differences are related to site and not implant surf ace morphology. Gross and subgross findings of the selected push-out specimens suggested that apposition of the bone to the implant may be controlled by interactions of bone type and implant surface morphology. Adherent bone observed on the highest strength cortical bone porous-coated specimens and on the cancellous bone plasma sprayed specimens indicate that, at least over part of the surface, bone-implant shear strength was greater than intrinsic bone shear strength. In cortical bone, bone and fibrous repair tissue ingrowth into a conventional porous coating gave higher shear strengths than ongrowth to a plasma-sprayed coating. We believe this was due to the higher intrinsic strength of cortical bone. Adaptation patterns of cancellous bone apposed to the plasma-sprayed coating and the higher percent paired differences for cancellous bone suggest superiority for the plasma-sprayed surface in cancellous bone. Even though total surface area of the plasma-sprayed surface is less than that for the porous coating, it appears to offer a better surface for bone interlock with cancellous bone.
SUMMARY
Large variations due to interactions of site and time with coating type were observed. Shear strength values ranged from 0.0 to 31.9 MPa. Smooth implants exhibited the lowest shear strengths. Overall average shear strengths increased with time for both plasma-sprayed and porous surfaces in both cancellous and cortical bone. In cortical bone, the overall average shear strengths were higher for the porous surface specimens at both the 8- and 16-week time periods. In addi-
LUCKEY, LAMPRECHT, AND WALT
574
tion, analysis of the averaged percent paired differences between the bead porous and plasma-sprayed implants in the identical sites of contralateral limbs also showed that porous-coated implants gave higher bone shear strengths at both time periods. In cancellous bone, the overall average shear strengths for the porous and plasma-sprayed surfaces were statistically indistinguishable at either time period. Based on the percent paired differences, the average shear strengths at the initial 8-week time period were generally lower for plasma-sprayed implants than for porous-coated implants. However, the average shear strengths at 16 weeks were notably higher for the plasma-sprayed implants than for the porous surface implants. CONCLUSIONS
We have shown previously that plasma arc spraying of a dense rough CoCr surface can be done in a manner which does not cause lowering of fatigue strength due to elevated temperature exposure and which increases implant surface area much less than porous coating. The dense nature of such a coating offers substantially higher structural integrity than other typical porous coatings. This work demonstrates that biological fixation to such a dense, rough coating is roughly equivalent to that seen with a typical beaded pcrous coating, and may even be stronger than such porous coatings in cancellous bone. Since current design theory for biological fixation favors metaphyseal fixation, this surface offers potential advantages over conventional porous coatings. The authors thank Henry Hahn of Artech Corporation, Falls Church, VA for assistance in the development of dense plasma arc sprayed coatings and for preparation of the specimens used in this study. We are also grateful to John Grahm of AstroMet Associates, Cincinnati, OH for preparation of the sintered beaded specimen. The invaluable efforts of H.V. Mendenhall, DVM in the implant surgery and Kathleen M. Laska in subsequent histological preparation (both of the 3M Biosciences Laboratory), and Paul R. Bushey in the mechanical testing (3M Orthopedic Products Laboratory), are gratefully acknowledged.
References P. K. Buchert, B. K. Vaughn, T. H. Mallory, C.A. Engh, and I. D. Bobyn, “Excessive metal release due to loosening and fretting of sintered particles on porous-coated hip prostheses,” J. Bone joint Surg., 68-A(4), 606-609 (1986). 2. T. J. Jorgensen, F. Munno, T.G. Mitchell, and D. Hungerford, “Urinary cobalt levels in patients with porous Austin-Moore prostheses,” Clin. Orfhop., 176, 124-126 (1983). 3. L.C. Lucas, J. Lemons, J. Lee, and P. Dale, ”In V i f r o corrosion evaluations of porous titanium-base and cobalt-base alloys,” Tuans. Soc. Biomaterials, 8, 80 (1985). 4. S. Margolian, R. M. Pilliar, and G.C. Weatherly, “Corrosion characteristics of porous surfaced cast CO-base implants-in v i m and in v i f r o studies,” Trans. Sac. Riomafcrials, 4, 82 (1981). 1.
BONE APPOSITION 5. 6. 7.
8. 9.
10. 11.
575
R. M. Pilliar, "Powder metal-made orthopedic implants with porous surface for fixation by tissue ingrowth," Clin. Ortkop., 176, 42-51 (1983). ES. Georgette, S. D. Cook, H. B. Skinner, A.M. Weinstein, and R. Yapp, "Fatigue behavior of coated and uncoated Ti-6A1-4V surgical implant material," Trans. Soc. Biomaterials, 6, 6 (1983). R. Rosenquist, B. Bylander, K . Knutson, U. Rydholm, B. Rooser, N. Egund, and L. Lidgren, "Loosening of the porous coating of biocompartmental prostheses in patients with rheumatoid arthritis," J. Bone Joint Surg., 68-A(4), 538-542 (1986). H. A . Luckey, and M. J. Walt, "Dense plasma sprayed F75 alloy for bone apposition," Trans. SOC.Biomaterials, 10, 93 (1987). J. D. Bobyn, R. M. Pilliar, H.U. Cameron, and G.C. Weatherly, "The optimum pore size for the fixation of porous-surfaced metal implants by bone ingrowth," Clin. Ortkop., 150, 263-270 (1980). T. Kilner, R. M. Pilliar, G.C. Weatherly, and C. Allibert, "Phase identification and incipient melting in a cast Co-Cr surgical implant alloy," J. Biomed. Muter. Res., 16, 63-79 (1982). H. A. Luckey, M. J. Walt, and E.G. Lamprecht, "Conservative biological fixation by bone ongrowth to F75 Alloy," Trans. Soc. Biomaterials, 10, 46 (1987).
Received June 20,1991 Accepted October 31, 1991
that although conventional porous coati n g exhibited higher overall average shear strengths in cortical bone sites at both time periods, the differences were not statistically significant. In cancellous sites, the average shear strengths achieved with conventional porous and plasmasprayed coatings were essentially equal. Analysis using average paired differences, however, revealed that when porous and plasma-coated implants are placed in identical sites of contralateral limbs, the plasma coatings consistently yielded higher shear strengths in cancellous bone sites at the later time period. Since current design theory for biological fixation favors metaphysical fixation, this surface may offer potential advantages over conventional porous coatings.
INTRODUCTION
Clinical dissatisfaction with (po1y)met hylme t hacry late (PMMA) bone cement has led to the increased use of implants with porous coatings to anchor total joint prostheses. A number of concerns have been raised, however, regarding the long-term performance of implants with porous surface coatings. These include the effects of increased metal ion release due to increased *Current address: Materials Engineering, Ltd., 1386 Pine View Trail, St. Joseph, WI 54082. 'To whom correspondence should be addressed. Journal of Biomedical Materials Research, Vol. 26, 557-575 (1992) 0 1992 John Wiley & Sons, Inc. ccc 0021-9304/92/050557-19$4.00
LUCKEY, LAMPRECHT, A N D WALT
558
surface area,'" a reduction in substrate fatigue strength due to high temperature sintering,"".'" and the potential for coating loss.'z7 A rough, dense CoCr (ASTM F75 alloy) coating using plasma arc spraying techniques has been developed which addresses these concerns." This work investigates the capability of this coating to provide a surface for direct biological fixation with bone in an animal model.
METHODS A N D MATERIALS
Specimen fabrication Cylindrical test plugs (6.3 mm diameter x 12.7 mm length) fabricated from CoCrMo alloy, conforming to ASTM F75, were prepared with the following three surface structures for in vivo implantation: (a) a dense plasma-sprayed (PS) F75 alloy coating, (b) a porous-beaded (PC) F75 alloy coating, and (c) a smooth (S) abrasive grit blasted surface. Different initial substrate diameters were used in order to produce specimens with approximately the same finished diameter. Plasma-sprayed specimens were made as previously described.' The coatings were created by plasma arc spraying 300-420-pm spherical powder on to hot isostatically pressed (HIP) cylindrical substrates. These specimens were subsequently heat treated at conventional solution treating temperatures to bond the coating to the substrate. A final plasma-sprayed coating thickness of 100 pm was obtained, which exhibited an RMS surface roughness of approximately 25 pm. No interconnected porosity was evident, although isolated microcrevices of up to 50 pm were noted (Fig. l). The porous-beaded specimens were prepared by vacuum sintering 300400-pm R.E.P. powder (Nuclear Metals, Inc., Concord, Ma) to cast F75 alloy substrates at approximately 1300°C.A two-layer beaded structure was created with a resulting pore size estimated from scanning electron microscopy (SEM) to be around 250-300 pm. Pore volume was estimated to be 30% (Fig. 2). Specimens with a smooth surface were fabricated by normal implant manufacturing techniques. Final abrasive grit blasting produced an RMS surface roughness of 1-2 pm (Fig. 3). All test specimens were cleaned and passivated per ASTM F86 prior to implantation.
Surface characterization Characterization of the resulting dense plasma-spray-coated surface and processing effects on the substrate have been reported previously.' No decrease in substrate fatigue strength was evident. The increase in the surface area as a result of the coating roughness was estimated from scanning electron micrographs to be 200-300%. This compares with typical surface area
BONE APPOSITION
Figure 1. SEM of plasma-sprayed (PS) implant surface.
Figure 2. SEM of porous-beaded (PC) implant surface.
559
LUCKEY, LAMPRECHT, A N D WALT
560
Figure 3. SEM of smooth (S) implant surface.
increases for sintered bead coatings of 700-800%. Corrosion rate increases of only 20% were seen for the dense plasma sprayed coating compared to 270% increases for the sintered bead coating3 Specimen implantation The implant plugs were placed transversely through a single cortex in both hind legs of eight adult goats. Three cortical and three cancellous nonloaded sites were chosen (Table I). The cancellous sites were at the distal femoral condyle (DFC), the anterior proximal tibia (APT), and the posterior proximal tibia (PPT). The cortical sites were at the proximal tibia diaphysis (PTD), the mid-tibia1 diaphysis (MTD), and the distal tibia1 diaphysis (DTD) (Fig. 4). Precision reamers were used to prepare each site to achieve a uniform 2% interference fit between the implant and bone. The plug was inserted with the outer end flush with the periosteum. After recovery from surgery, the animals had unrestricted ambulation in an outdoor environment. Sacrifice was performed at 8 and 16 weeks postoperatively. NIH guidelines for the care and use of laboratory animals (NIH Publication #85-23 Rev. 1985) have been observed. Mechanical testing At sacrifice, each hind limb designated for mechanical push-out testing was stripped of soft tissue and frozen at -20°C. Prior to testing, each limb
561
BONE APPOSITION TABLE I Implantation Schedule in Goat Hind Limbs Animal Number and Limb 1&5
Cancellous Sites APT PPT DFC Cortical Sites PTD MTD DTD
3&7
2&6
4&8
R
L
R
L
R
L
R
L
S*
PS S PC
PC PS S
S PC PS
PC PS PC
PS PC PS
PS PC PS
PC PS PC
PS S PC
PC PS S
S PC
PC PS PC
PS PC PS
PS PC PS
PC PS PC
PC" PS* S*
PC* PS*
PS
PS
= Plasma-sprayed; PC = porous-coated; S = smooth. *Histology specimens.
Figure 4. Radiograph of implant sites in goat hind limbs.
LUCKEY, LAMPRECHT, AND WALT
562
was thawed and the bone around each implant sectioned transversely. Any periosteal bone which had formed over the plugs was carefully dissected away, and each section containing an implant was bisected perpendicular to the axis of the implant, exposing each end of the implant cylinder. Care was taken to leave a minimum thickness of 6 mm-10 mm between the implant and the adjacent transverse section cut. Axial push-out testing was performed using the method of Bobyn' at a displacement rate of 25.4 mm/min. After push-out testing, the bone was sectioned at the implant site using a low-speed diamond saw. Measurements of bone thickness taken at 90" intervals around the circumference of the implant site were used to estimate the area of bone contact with the implant. Several modifications were made to the above test method to assure axial alignment of the specimen with the load application axis. Each cylindrical implant was fabricated with a 1.6 mm diameter axial hole and a 6.4 mm radius, partially spherical indentation at one end. Upon completion of test specimen preparation, the implant was placed over a centering pin in the test fixture with the longitudinal cut resting on the fixture. A small donut of bone cement was used to fill the gaps between the fixture and the bone, and the cement cured in place. The centering pin was removed, a 4.75-mm ball bearing placed in the indentation of the implant, and the load applied through the ball bearing. All testing was performed using an Instron 1321H with calibration traceable to National Institute of Standards and Technology (NIST).
Histological examination One limb was selected from each time period for histological examination as denoted in Table I. In addition, two cortical and two cancellous specimens each were selected from the mechanical testing group after push-out testing. The post-mechanical specimens selected were those which exhibited the highest measured shear strengths for the PS and PC surface types. Implants and attached bone were fixed in formaldehyde, dehydrated in graded alcohol, and embedded in PMMA. The embedded specimens were subsequently thick sectioned transversely, polished to approximately 100 pm, stained with toluidine blue and picrofuchsin, and viewed by transmission light microscopy.
RESULTS
Mechanical testing The attachment values for the smooth implants (S) were substantially lower than those for either the plasma-sprayed or the porous-coated implants. The shear strengths for the smooth surfaced specimens ranged from 0.0 to 3.3 MPa with average values of 1.07 & 0.59 MPa in cancellous bone and 1.22 k 1.62 MPa in cortical bone. Because of the low interface strength, some
BONE APPOSITION
563
specimens may have been dislodged during preparation for mechanical testing. The low shear strength values also indicate that these specimens acted as negative controls. Therefore, the results from the smooth specimens will be excluded from further discussion. Although the shear strengths obtained compare well with those reported in the literature,”’ the values ranged from 0.3 to 31.9 MPa. Interfacial shear strengths in cancellous bone ranged from 3.5 to 9.2 MPa at 8 weeks and ranged from 1.7 to 12.3 MPa at 16 weeks (Tables 11, I11 and Fig. 5). In cortical bone, shear strengths of 0.3 to 10.7 MPa were recorded for the 8-week specimens, while shear strengths of 4.7 to 31.9 MPa were seen at 16 weeks (Tables IV, V and Fig. 6). The average interfacial shear strengths for all sites, surface types, and time periods are also shown (Fig. 7).
Time effects In cancellous bone, there was a trend toward increasing average shear strengths with time at most sites and with both surface types (Table VI and Fig. 7). The average shear strength for all cancellous sites with porous-coated specimens increased from 6.0 MPa at 8 weeks to 7.2 MPa at 16 weeks. A similar increase from 5.5 MPa to 7.2 MPa was seen for the plasma-sprayed specimens. However, the differences in the average strengths for the cancellous sites between time periods were not statistically significant for either surface type ( p 5 .27) or individual site. TABLE I1 Eight-Week Mechanical Testing Data in Cancellous Bone Animal 5L
6L 6R
7L 7R 8L 8R
Implant Site
Surface Treatment
Push-Out Force ( N )
Contact Area (mm’)
Shear Strength (MPa)
APT PPT DFC APT PPT DFC APT PPT DFC APT PPT DF C APT PPT DFC APT PPT DFC APT PPT DFC
PS S PC S PC PS PC PS S PS PC PS PC
1138 245 1080 89 1468 916 334 1112 210 1001 1334 1334 734 845 1500 613 801 1458 712 970 1201
203 165 190 189 184 173 178 179 173 175 181 168 144 181 164 184 179 188 203 184 199
5.60 1.48 5.68 0.47 8.00 5.30 1.88 6.23 1.21 5.72 7.37 7.96 5.10 4.66 9.14 3.34 4.48 7.77 3.50 5.28 6.05
rs
PC PC PS PC PS PC PS
LUCKEY, LAMPRECHT, AND WALT
564
TABLE I11 Sixteen-Week Mechanical Testing Data in Cancellous Bone Animal
Implant Site
Surface Treatment PS S PC S PC PS PC PS S PS PC PS PC PS PC PS* PC* PC PS PC PS
APT PPT DFC APT PPT DFC
1L
2L 2R
APT
3L
DFC APT
3R
DFC APT
4L
DFC APT
4R
DFC APT
PPT
PPT PPT
PPT
PPT DFC
Push-Out Force ( N )
Contact Area (mm2)
Shear Strength (MPa)
845 222 1735 44 1201 1535 801 2091 356 1059 1015 2594 855 1401 1245 317 890 1312 396 1700 2317
181 190 185 163 182 194 159 188 194 173 161 211 166 20 5 166 188 141 167 203 167 195
4.68 1.17 9.38 0.28 6.60 7.93 5.05 11.17 1.83 6.12 6.32 12.28 5.15 6.83 7.52 1.68 6.32 7.86 1.95 10.14 11.93
~
*%sface types switched at time of implantation.
15
1
T
APT
PPT
DFC
Implantation Sites PC-8
PC-16
PS-8
a
PS-16
Figure 5. Average interfacial shear stresses with standard deviations for cancellous sites.
BONE APPOSITION
565
TABLE IV Eight-Week Mechanical Testing Data in Cortical Bone Animal
Implant Site
Surface Treatment
Push-Out Force ( N )
Contact Area (mm2)
Shear Strength (MPa)
PTD MTD DTD PTD MTD DTD PTD MTD DTD PTD MTD DTD PTD MTD DTD PTD MTD DTD PTD MTD DTD
PS S PC
1312 614 1592 0 858 378 1134 912 543 58 165 67 1432 872 467 823 734 734 614 925 734
123 188 160 148 184 86 180 163 74 167 195 156 188 157 98 195 152 82 145 189
10.69 3.26 9.96 0.00 5.79 2.06 13.24 5.07 3.34 0.79 0.99 0.34 9.17 4.64 2.97 8.41 3.77 4.82 7.52 6.37 3.88
~
5L
6L 6R
7L
7R 8L 8R
S PC PS PC PS S PS PC PS PC PS PC PC PS PC PS PC PS
TABLE V Sixteen-Week Mechanical Testing Data in Cortical Bone Animal 1L 2L 2R
3L 3R 4L 4R
Implant Site
Surface Treatment
Push-Out Force ( N )
Contact Area (mm’)
Shear Strength (MPa)
PTD MTD DTD PTD MTD DTD PTD MTD DTD PTD MTD DTD PTD MTD DT D PTD MTD DTD PTD MTD DTD
PS S PC PS PC S PC PS S PS PC PS PC PS PC PC PS PC PS PC PS
623 111 1815 1379 2246 0 1357 2002 22 1970 3060 2313 1059 2260 3336 383 2335 1873 801 4448 1277
132 193 197 128 176
4.71 0.58 9.24 10.83 12.76 0.00 14.14 9.86 0.11 12.21 18.28 11.59 13.24 13.17 21.38 6.90 11.59 11.45 7.66 31.93 6.30
-
96 203 199 162 168 199 80 172 156 55 201 163 105 139 203
LUCKEY, LAMPRECHT, AND WALT
566
Y
01
25
m
L
tj
20
5
-
f
15
Q)
2
10
0
> a
5 0
PTD
MTD
=
nPC-8
DTD
Implantation Sites
PC-16
PS-8
PS-16
Figure 6. Average interfacial shear stresses with standard deviations for cortical sites.
T
Q)
0
E 5
a
0 Cancellous
aPC-8
=
Cortical Implantation Sites PC-16
PS-8
PS-16
Figure 7. Average interfacial shear stress with standard deviations for all sites, surface types, and time periods.
In cortical bone, there was a statistically significant increase in the overall average shear strength for the porous-coated (PC) and the plasma-sprayed (PS) implants in all sites between the 8-week and 16-week time periods. (Table VI and Fig. 7). The overall average shear strengths increased from 6.9 MPa to 15.5 MPa for the PC specimens ( p 5 .011*), while the overall aver-
*“g
Overall
DTD
MTD
Overall A% Cortical Sites PTD
DFC
PPT
Cancellous Sites APT
0.23 (k0.31)
9.77 (22.90) 4.31 (23.26)
2.20 (k1.90)
0.06 ( 2 .08) 3.34
15.48 (27.55)
0.58 3.26
6.86 (k3.76)
-
0.00
8.85 (k3.35) 11.54 (k1.66) 8.94 (23.74)
6.33 (k5.06) 4.49 (50.66) 2.09 (21.77)
11.43 (k3.95) 20.99 (k9.87) 14.02 (k7.47)
10.27 (k2.60) 4.38 (k2.95) 5.92 (k3.62)
1.83
1.21
1.09 (50.78)
1.17
0.28
S [16 wk]
1.48
0.47
S [8 wk]
1.05 (k0.52)
5.50 (21.26)
7.15 (k1.76)
5.96 (k2.36)
3.61 (22.15) 9.00 (23.06) 10.71 (22.42)
PS [16 wk]
7.17 (k4.03)
4.94 (k1.25) 5.12 (k0.96) 6.44 (21.37)
PS [8 wk]
5.10 (20.07) 7.35 (21.87) 8.25 (k0.99)
PC [16 wk]
3.44 (k1.61) 6.88 (k1.42) 7.55 (21.74)
PC [8 wk]
TABLE VI Average Shear Strengths (MPa) and Standard Deviations by Implantation Site
E
CJ
c
2 m
568
LUCKEY, LAMPRECHT, AND WALT
age shear strengths for PS specimens increased from 4.31 to 9.77 MPa ( p 5 ,002). An increase in shear strength with time was also seen at each site for both surface types. However, the only statistically significant increase was at the MTD site for the plasma-sprayed surface ( p 5 .02).
Surgical site effects Some site-to-site differences in the push-out results were noted. In cancellous bone, the APT site was significantly lower in strength than either the PPT or the DFC sites for all porous-coated samples at both time periods ( p 5 .05). The trend was similar for plasma-sprayed specimens, but only the APT site was significantly lower than the DFC site at the 16 weeks (p 5 0.16). In the cortical sites, there was no clear trend, although the MTD site tended to have a lower strength than the PTD or DTD sites at 8 weeks, and a higher strength at 16 weeks.
Surface type effects There were no statistically significant differences between surface types at the same time period. However, the porous-coated specimens exhibited higher average strengths at two of three cortical sites at 8 weeks and at all 16-week cortical sites.
Animal effects Marked differences in shear strengths occurred among animals for both surface types. Most animals gave either consistently high or low values for both surface types. For any given animal, substantial site-to-site variability in shear strength was also observed for each type of bone. Implants in the MTD sites tended to give the highest shear strengths among cortical sites. Among cancellous sites, the implants in the DFC sites gave consistently higher values.
Paired site analysis The surgical implantation schedule was designed to randomize specimens of the two surface types with regard to the various surgical sites in each type of bone. Although insufficient data existed for a precise determination, interactions of time, site, and coating type seemed to exist. Because of the greater than expected variability among animals and surgical sites, any actual differences in bone response between surface types may have been masked. In order to directly compare the response between surface types, data from animals with PS and PC specimens in identical sites in contralateral limbs were examined separately. The average percentage shear strength differences be-
BONE APPOSITION
569
tween paired sites were calculated according to the formula shown below and are listed in Table VII.
% paired difference = [(PS - PC)/PC] x 100 PS = average shear strength for plasma-sprayed specimens PC = average shear strength for porous-coated specimens With the exception of the 16-week APT site, at least two pairs existed for each site at each time period. Push-out strengths were generally lower in cortical bone for PS specimens than for PC specimens at both time periods resulting in a negative percent paired difference. At 8 weeks, percent paired differences ranged from +2.5% to -52.7%. At 16 weeks, the percent paired differences ranged from -10.5% to -45.5%. This agrees with the overall averages. Agreement with overall averages for cancellous bone was limited to the 8week time period. The overall average shear strengths were similar for both surface types at both time periods. The percent paired differences were lower for the PS specimens than for PC specimens at 8 weeks (+9.2% to -25.6%) while at 16 weeks, the PC specimens exhibited substantially higher percent paired differences (+18.8% to 57.3%).Thus, the relative strengths were reversed at the longer time period.
+
Histology results Substantial variations in the extent of new bone growth between animals and among surgical sites were found in the region between preexisting bone and the implant surface. These variations appeared to result from differences in the initial wave of bone formation rather than subsequent bone resorption or remodeling. All surface types, however, exhibited areas of apparent direct bone contact as well as areas with interposed fibrous tissue. The amount of bone apposition varied among the specimens. The 8-week exposure specimens had prolific callus and new bone formation at all sites. New woven bone filled the space between the preexisting TABLE VII Average Percent Shear Strength Differences for Contralateral Paired Sites by Time Period Cancellous Site 8 week
APT PPT DFC 16 week APT PPT DF C
Cortical
% Diff.
N Pair
Site
% Diff.
N Pair
+ 9.2 -25.6 -17.3
2 3 2
PTD MTD DTD
-52.7 + 2.5 -45.8
2 3 2
i-18.8
1 2 2
PTD MTD DTD
-10.5 -45.0 -45.5
3 3 2
+39.3 +57.3
LUCKEY, LAMPRECHT, AND WALT
570
bone and the PS (DFC) (Fig. 8) and PC (PPT) (Fig. 9) implants in the cancellous sites. Direct bone apposition was also observed on one-third of the surface of the smooth implant (APT). Bone contact was limited to areas of callus around the PC (MTD) and PS (DTD) implants in the cortical sites. The 16-week exposure individual was not as prolific in forming bone, New bone formation led to direct bone contact at only pinpoint locations on three of the six implants. All other sites exhibited a collagen repair tissue layer interposed between the preexisting bone, or the new bone, and the implant (Fig. 10). Bone ingrowth into the porous implants was limited to less than one-half the available space.
Histology of push-out specimens The four push-out specimens exhibited freezing and thawing artifacts of soft tissue but had excellent preservation of bone repair results. Bone adhered to the surfaces of one PC and one PS implant to varying degree. Bone was intimately apposed to the bulk of the free surface of a plasmasprayed plug in a cancellous DFC site (Fig. 11).Bone made periodic direct contact with the metal surface. At its greatest extent, up to 4 mm of bone remained undisturbed in close apposition to the implant. The apposed bone exhibited occasional microscopic bends or fractures, but remained largely intact.
Figure 8. Bone apposition to plasma-sprayed (PS) surface in cancellous bone (DFC) at 8 weeks (X10).
BONE APPOSITION
Figure 9. Bone ingrowth into porous bead (PC) surface in cancellous bone (PPT) at 8 weeks ( ~ 1 0 ) .
Figure 10. Collagen repair tissue filling the space between original bone and porous bead (PC) surface in cancellous bone (PPT) at 16 weeks (original magnification ~ 1 0 ) .
571
572
LUCKEY, LAMPRECHT, AND WALT
Figure 11. Bone fragment remaining attached to the plasma-sprayed (PS) surface after push-out from cancellous bone (DFC) at 16 weeks (original magnification ~ 1 0 ) .
Bone sheared at the outer bead surface of the porous-coated plug in the cortical MTD site, thus leaving negligible bone attached to the outer surface (Fig. 12). However, bone had invaded approximately one-half of the available pore space below the outer circumference of the implant.
Figure 12. Bone sheared at the outer edge of the porous bead (PC) surface after push-out from cortical bone (MTD) at 16 weeks (original magnification ~ 1 0 ) .
573
BONE APPOSITION DISCUSSION
The subgross histological observation of prolific callus and new bone formation correlated well with the high push-out strength seen in the contralateral limb of one 8-week individual. The highest shear strength for smooth implants was also recorded at 3.3 MPa (MTD) and 1.5 MPa (PPT) for that individual. Shear strengths for PS and PC surfaces were nearly equal within the same type of bone. In the cortical sites, the PC strength was 10.7 MPa (PTD), while the PS strength was 9.9 MPa (DTD). In the cancellous sites, the PC strength was 5.7 MPa (APT) with the corresponding PS strength at 5.6 MPa. The lack of observed bone apposition in the 16-week individual corresponded to lower measured shear strengths. The shear strength values for the smooth implants for this 16-week individual (0.7 MPa at the MTD site) were lower than for the 8-week individual (1.2 MPa at the PPT site). Also, the PS implants in the PTD and APT sites gave strengths approximately 50% that of the PC implants in the DTD and DFC sites (4.7 MPa vs. 9.2 MPa). The paired comparisons between identical implant surfaces in other animals in the above sites suggest that these differences are related to site and not implant surf ace morphology. Gross and subgross findings of the selected push-out specimens suggested that apposition of the bone to the implant may be controlled by interactions of bone type and implant surface morphology. Adherent bone observed on the highest strength cortical bone porous-coated specimens and on the cancellous bone plasma sprayed specimens indicate that, at least over part of the surface, bone-implant shear strength was greater than intrinsic bone shear strength. In cortical bone, bone and fibrous repair tissue ingrowth into a conventional porous coating gave higher shear strengths than ongrowth to a plasma-sprayed coating. We believe this was due to the higher intrinsic strength of cortical bone. Adaptation patterns of cancellous bone apposed to the plasma-sprayed coating and the higher percent paired differences for cancellous bone suggest superiority for the plasma-sprayed surface in cancellous bone. Even though total surface area of the plasma-sprayed surface is less than that for the porous coating, it appears to offer a better surface for bone interlock with cancellous bone.
SUMMARY
Large variations due to interactions of site and time with coating type were observed. Shear strength values ranged from 0.0 to 31.9 MPa. Smooth implants exhibited the lowest shear strengths. Overall average shear strengths increased with time for both plasma-sprayed and porous surfaces in both cancellous and cortical bone. In cortical bone, the overall average shear strengths were higher for the porous surface specimens at both the 8- and 16-week time periods. In addi-
LUCKEY, LAMPRECHT, AND WALT
574
tion, analysis of the averaged percent paired differences between the bead porous and plasma-sprayed implants in the identical sites of contralateral limbs also showed that porous-coated implants gave higher bone shear strengths at both time periods. In cancellous bone, the overall average shear strengths for the porous and plasma-sprayed surfaces were statistically indistinguishable at either time period. Based on the percent paired differences, the average shear strengths at the initial 8-week time period were generally lower for plasma-sprayed implants than for porous-coated implants. However, the average shear strengths at 16 weeks were notably higher for the plasma-sprayed implants than for the porous surface implants. CONCLUSIONS
We have shown previously that plasma arc spraying of a dense rough CoCr surface can be done in a manner which does not cause lowering of fatigue strength due to elevated temperature exposure and which increases implant surface area much less than porous coating. The dense nature of such a coating offers substantially higher structural integrity than other typical porous coatings. This work demonstrates that biological fixation to such a dense, rough coating is roughly equivalent to that seen with a typical beaded pcrous coating, and may even be stronger than such porous coatings in cancellous bone. Since current design theory for biological fixation favors metaphyseal fixation, this surface offers potential advantages over conventional porous coatings. The authors thank Henry Hahn of Artech Corporation, Falls Church, VA for assistance in the development of dense plasma arc sprayed coatings and for preparation of the specimens used in this study. We are also grateful to John Grahm of AstroMet Associates, Cincinnati, OH for preparation of the sintered beaded specimen. The invaluable efforts of H.V. Mendenhall, DVM in the implant surgery and Kathleen M. Laska in subsequent histological preparation (both of the 3M Biosciences Laboratory), and Paul R. Bushey in the mechanical testing (3M Orthopedic Products Laboratory), are gratefully acknowledged.
References P. K. Buchert, B. K. Vaughn, T. H. Mallory, C.A. Engh, and I. D. Bobyn, “Excessive metal release due to loosening and fretting of sintered particles on porous-coated hip prostheses,” J. Bone joint Surg., 68-A(4), 606-609 (1986). 2. T. J. Jorgensen, F. Munno, T.G. Mitchell, and D. Hungerford, “Urinary cobalt levels in patients with porous Austin-Moore prostheses,” Clin. Orfhop., 176, 124-126 (1983). 3. L.C. Lucas, J. Lemons, J. Lee, and P. Dale, ”In V i f r o corrosion evaluations of porous titanium-base and cobalt-base alloys,” Tuans. Soc. Biomaterials, 8, 80 (1985). 4. S. Margolian, R. M. Pilliar, and G.C. Weatherly, “Corrosion characteristics of porous surfaced cast CO-base implants-in v i m and in v i f r o studies,” Trans. Sac. Riomafcrials, 4, 82 (1981). 1.
BONE APPOSITION 5. 6. 7.
8. 9.
10. 11.
575
R. M. Pilliar, "Powder metal-made orthopedic implants with porous surface for fixation by tissue ingrowth," Clin. Ortkop., 176, 42-51 (1983). ES. Georgette, S. D. Cook, H. B. Skinner, A.M. Weinstein, and R. Yapp, "Fatigue behavior of coated and uncoated Ti-6A1-4V surgical implant material," Trans. Soc. Biomaterials, 6, 6 (1983). R. Rosenquist, B. Bylander, K . Knutson, U. Rydholm, B. Rooser, N. Egund, and L. Lidgren, "Loosening of the porous coating of biocompartmental prostheses in patients with rheumatoid arthritis," J. Bone Joint Surg., 68-A(4), 538-542 (1986). H. A . Luckey, and M. J. Walt, "Dense plasma sprayed F75 alloy for bone apposition," Trans. SOC.Biomaterials, 10, 93 (1987). J. D. Bobyn, R. M. Pilliar, H.U. Cameron, and G.C. Weatherly, "The optimum pore size for the fixation of porous-surfaced metal implants by bone ingrowth," Clin. Ortkop., 150, 263-270 (1980). T. Kilner, R. M. Pilliar, G.C. Weatherly, and C. Allibert, "Phase identification and incipient melting in a cast Co-Cr surgical implant alloy," J. Biomed. Muter. Res., 16, 63-79 (1982). H. A. Luckey, M. J. Walt, and E.G. Lamprecht, "Conservative biological fixation by bone ongrowth to F75 Alloy," Trans. Soc. Biomaterials, 10, 46 (1987).
Received June 20,1991 Accepted October 31, 1991
