- Email: [email protected]
The radiographic resolution of beads from porous coated joint prostheses
The Radiographic R e s o l u t i o n of Beads From Porous Coated Joint Prostheses
D a r r e l W . H a y n e s , P h i ) , a n d Eric L. F r e e d m a n ,
BA
Abstract: While there have been reports of the 750 I.tm chrome-cobalt beads disassociating from the PCA type prosthesis, failure of the 200 i.tm bead has not been documented despite reports of failed prosthesis, which use the smaller bead. As a result of the negative reports, the hypothesis was posed as to the limit of resolution in the image of the standard clinical radiographical film. To test the hypothesis, chrome-cobalt beads, ranging in size from 200 to 700 Ixm, in differing configurations were placed adjacent to a femoral prosthesis that had been implanted into a cadaveric specimen. Using standard clinical radiographic technique, differing exposures were made of the configurations. The results demonstrated that beads or a conglomerate of beads smaller than 300 Ism could not be resolved and were therefore invisible and unrecordable. The clinical significance is to question whether all porous-coated prostheses shed metal allo,, particles, some of which are radiographically undetectable due to the limitatic s of their resolvable size. Key words: joint arthroplasty, porous coating, bead size, radiographic resolution.
Pore sizes in the range of 1 5 0 - 4 5 0 g.m are currently used, which were determined to be optimal based on limited animal studies (10). To achieve pore sizes in this range with sintered bead porous-coated implants, various bead sizes have been used, ranging from 100 to 750 i.tm. Initially, a porous-coated M o o r e type prosthesis (DePuy) was reported, which used 100 Ixm beads to obtain a pore size of approximately 100 p.m (6). The prosthesis was later m o d ified to achieve pore sizes within the 1 5 0 - 3 0 0 I.tm range. The D e p u y Anatomic Medullary Locking (AML) and Low Contact Stress (LCS) prostheses presumably use the s a m e porous coating as that on the modified M o o r e prosthesis. The Howmedica PCA prosthesis uses larger beads of 7 0 0 - 7 5 0 i~m, to create an average pore size of 425 ttm (8). There h a v e been several reports in the literature of loosening of beads from the porous-coated anatomical (PCA) prostheses (4, 1 I, 12), but there appear to be no reports of individual bead loosening from other porous-coated prostheses. Although the
The long-term stable fixation of total hip and total k n e e prostheses has assumed increased importance in the field of joint arthroplasty. Cemented fixation techniques are being replaced by cementless biological fixation due to the rate of loosening associated with cemented fixation, which has been related to " c e m e n t b o n e disease"(9). Cementless fixation relies on the ability to achieve b o n y ingrowth into the prosthesis by various types of porous coating, including sintered fiber-metal, sintered beads, and plasma flame-spray (13). There appear to be m a n y necessary conditions for b o n y ingrowth, which include a pore size of 5 0 - 4 0 0 p.m (1), intimate apposition to bone, and minimal implant m o v e m e n t relative to adjacent bone (3).
From the Department of Orthopaedic Surgery, Johns Hopkins University School of Medicine, Bahimore, Maryland.
Reprint requests: Darrel W. Haynes, Phl), Department of Orthopaedic Surgery, Johns Hopkins University School of Medicine, 5601 Loch Raven Boulevard, Baltimore, MD 21239. 117
118
The Journal of Arthroplasty Vol. 5 No. 2 June 1990
techniques of high-temperature sintering of metal alloy powders on the substrate surface are constantly being improved, it would appear likely that occasional beads may separate from all beaded porouscoated surfaces. In attempting to explain the apparent absence of smaller bead separation it was hypothesized that the smaller beads could not be resolved due to the limitations of radiographic techniques. To test this hypothesis, we performed the following experiment.
Materials and Methods Chrome cobalt beads similar to those used to coat prostheses, with diameters confirmed by scanning electron microscopy of 200, 300, 400, 500, 600, and 700 I-tm (obtained from Howmedica), were arranged between two pieces of transparent tape in two configurations. The first configuration was comprised of a single layer of multiple numbers of each bead size between 200 and 700 Ixm (Figs. 1-5). The second configuration was a five-bead pattern of each bead size within the same 2 0 0 - 7 0 0 Ixm range (Figs. 6 10). To simulate in vivo conditions, the femoral component of the PCA total hip prosthesis was implanted into a cadaveric hip, with skin and soft tissue intact, using the standard surgical procedure of the direct lateral approach. After implantation, a slit was cut the thickness of a saw blade with an oscillating bone saw in the lateral femoral cortical bone adjacent to the proximal porous-coated portion of the prosthesis. Care was taken to ensure that the window was in the plane perpendicular to both the anterior/posterior plane and the radiographic beam. The prepared bead configurations were placed sequentially in the window and radiographs were taken using a clinical mobile radiographic apparatus at 70 kV, 50 mA, at a distance of 40 inches from the tube to the film (Figs. 1, 6). These settings were chosen since they are used as the standard technique for hip radiography. High-resolution CRONEX intensifying screens were used with Kodak X-OMAT film. With time acting as the dependent variable, films were exposed at ~o, ~, and ~o second, with ~ being the optimum exposure. The two other exposures were used for bracketing purposes since it is recognized that clinical films are not ahvays of the optimal exposure. To compare the effect of radiographs without bone or soft tissue, exposures of both configurations were made outside the cortical window at 70 kV, 50 kV, and 35 kV at 50 mA with exposure times ranging from ~ to ~ second (Figs. 3, 8).
Fig. 1. Single-layer configuration of beads laying in window adjacent to a PCA femoral prosthesis implanted in a cadaver with soft tissue and skin intact (70 kV, 50 mA, second).
Results The limits of apparent radiographic resolution differed between the two configurations. With the second configuration of five beads, the entire range of sizes was radiographically apparent when placed in direct contact with the film cassette and exposed for second at 35 kV and 50 mA (Fig. 8). However, with the single-layer multiple-numbered beads, the 200 p,m sizes appeared as an amorphus patch and the individual beads could not be resolved (Fig. 3).
Porous Coated Beads
200
I
•
Haynes and Freedman
119
"~ .~'..',..,"
, :':~.:
.,..
-
,'2'
:
3O0 },
..'"
,:..
_..'.~:.,..: .:.: ~ . . . .
400
*o
e= . " . ~
o61 °
L~..,~'v'4L; ,% :'.J: ""
".'~.-.
.....
..:-\~i~
500 ~I. tl
600
IN
700
Figs. 2-4. Fig. 2 (Left). Optical view of the single-layer configuration of 200-700 p,m beads. Fig. 3 (Middle). Single-layer configuration radiographed in contact with the film cassette (35 kV, 50 mA, ~ second). Note the inability to individually resolve the 200 and 300 gm beads. Fig. 4 (Right). Enlargement of Figure 1 with less exposure (70 kV, 50 mA, -~ second).
W h e n the configurations were placed in the window adjacent to the prosthesis and radiographed with the optimum exposure, the proximal 200 and 300 i.tm size beads appeared as an opaque patch in the single-layer multiple-numbered configuration, and individual beads could not be resolved (Figs. 4, 5). The same results were observed in the five-bead configuration, where the proximal 200 and 300 p.m beads could not be resolved (Figs. 9, 10). By decreasing the exposure, the resolution could not be
improved., and increasing the exposure resulted in a gradual loss of overall resolution (Figs. 4, 5, 9, 10).
Discussion The results clearly demonstrate that there are limits in the ability to resolve differing sizes of metal alloy beads radiographically. Using clinical radiographic techniques and standard exposure time for
120
The Journal of Arthroplasty Vol. 5 No. 2 June 1990
E
Fig. 5. Enlargement of Figure 1 with same exposure (70 kV, 50 mA, ~ second). The 200 and 300 Fm beads appear as a patch, but individual beads ate impossible to resolve.
human hips, bead sizes of less than 400 p.m cannot be resolved and are therefore invisible. Even with the beads placed directly on the film cassette with reduced exposures, individual beads of less than 300 ~m cannot be clearly resolved. Resolution is the smallest distance between two dots or lines that can be defined separately. The resolution in a radiographic image is dependent upon image clarity, which is the result of both image quality and contrast (5). Factors that affect the image quality are the defuse light transfers from the phos-
Fig. 6. Five-bead configuration of beads lay~u~,, . . . . . . . . . dow as in Figure 1 (70 kV, 50 mA, ~ second).
phorous layer of the intensifying screen to the film. The light is further defused by the double-sided photographic emolution on the film. Factors that affect the contrast ate differences in the thickness and density of the object. The difference between a 200 p.m and a 700 p.m bead is not only their width but also their thickness. As a 700 Ixm bead is almost 4 times thicker than a 200 v m bead, more x-rays will be absorbed and the image will appear sharper, affording better image quality, and therefore more resolvability than the 200 I-ttn bead. It has been reported
Porous Coated Beads * Haynes and Freedman 121
200
300
400
500
600
0 •
•
P
700
•
"" o
* •
•
•
I
Figs. 7-9. Fig. 7 (Left). Optical view of the five-bead configuration of 200-700 gtm beads. Fig. 8 (Middle). Five-bead configuration radiographed in contact with the film cassette (35 kV, 50 mA ~ second). Note that all the beads are radiographically apparent, but it is difficult to resolve the 200 and 300 p.m beads as spheres. Fig. 9 (Right). Enlargement of Figure 6 with less exposure (70 kV, 50 mA, ~ second).
that in the ideal radiograph with the best image clarity, five line pairs per millimeter (200 p.m) is the m a x i m u m resolution (5). Since the sintering techniques used among the different porous-coated prostheses are similar, it must be assumed that some beads fall off all prostheses at an equal rate, h o w e v e r minute the-number may be. The inability to resolve beads less than 400 I~m may account for the fact that Ranawat et al. (11) cited loose beads only from the PCA (Howmedica) prosthesis and not the UCI (DePuy) prosthesis. The lack of resolution below 400 IJ-m applies not only to loosened beads but also to all miscellaneous loosened debris smaller than the resolvable limit. This fact was
demonstrated by Buchert et al. (2), w h o reported sintered panicles detached from a non-bead-coated prosthesis, which were radiographically invisible but histologically evident. Although there has been no clinical evidence that bead detachment is a causative agent in prosthesis loosening, there have been m a n y reports recording the n u m b e r of loose beads associated with radiolucent lines only in the PCA prosthesis (4, 11, 12). As the PCA porous coating utilizes the larger 750 lzm beads, which are radiographically visible, one must question whether there are loose beads or debris in the radiolucent lines of other prostheses, which use beads that are too small to resolve radiographically.
122
The Journal of Arthroplasty Vol. 5 No. 2 June 1990
References
J
t o
.L Fig. 10. Enlargment of Figure 6 with the same exposure (70 kV, 50 mA, -h second).The 200 and 300 ttm beads are radiographically invisible.
1. Bobyn JD, Pilliar RM, Cameron HU, Weatherly GC: The optimum pore size for the fixation of porous-surfaced metal implants by the ingrowth of bone. Clin Orthop 150:263, 1980 2. Buchert PK, Vaughn BK, Mallory TH, et al: Excessive metal release due to loosening and fretting of sintered panicles on porous coated hip prostheses. J Bone Joint Surg 68A:606, 1986 3. Cameron HU, Pilliar RM, Macnab I: The effect of movement on the bonding of porous metal to bone. J Biomed Mater Res 7:301, 1973 4. Cheng CL, Gross AE: Loosening of the porous coating in total knee replacement: J Bone Joint Surg 70B:377, 1988 5. Christensen EE, Curry TS III, Dowdey JE: An introduction to the physics of diagnostic radiology. Lea and Febiger; Philadelphia, 1978 6. Engh CA: Hip arthroplasty with a Moore prosthesis with porous coating. Clin Onhop 176:52, 1983 7. Engh CA, Bobyn JD, Glassman AH:Porous coated hip replacement: tile factors governing bone ingrowth stress shielding and clinical results. J Bone Joint Surg 69B:45, 1987 8. Hungerford DS, Kenna RV: Preliminary experience with a total knee prosthesis with porous-coating used without cement. Clin Orthop 176:95, 1983 9. Jones LC, Hungerford DS: Cement disease. Clin Orthop 225:192, 1987 10. Pilliar RM: Powder metal-made orthopaedic implants with porous surface for fixation by tissue growth. Clin Orthop 176:42, 1983 11. Ranawat CS, Johanson NA, Rimnac CM et al: Retrieval analysis of porous-coated components for total knee anhroplasty. Clin Onhop 209:244, 1986 12. Rosenqvist R, Bylander B, Knutson K et ah Loosening of the porous coating of bicompanmental prostheses in patients with rheumatoid arthritis. J Bone Joint Surg 68A:538, 1986 13. Turner TM, Sumner DR, Urban RM et al: A comparative study of porous coatings in a weight bearing total hip arthroplasty model. J Bone Joint Surg 68A: 1396, 1986
D a r r e l W . H a y n e s , P h i ) , a n d Eric L. F r e e d m a n ,
BA
Abstract: While there have been reports of the 750 I.tm chrome-cobalt beads disassociating from the PCA type prosthesis, failure of the 200 i.tm bead has not been documented despite reports of failed prosthesis, which use the smaller bead. As a result of the negative reports, the hypothesis was posed as to the limit of resolution in the image of the standard clinical radiographical film. To test the hypothesis, chrome-cobalt beads, ranging in size from 200 to 700 Ixm, in differing configurations were placed adjacent to a femoral prosthesis that had been implanted into a cadaveric specimen. Using standard clinical radiographic technique, differing exposures were made of the configurations. The results demonstrated that beads or a conglomerate of beads smaller than 300 Ism could not be resolved and were therefore invisible and unrecordable. The clinical significance is to question whether all porous-coated prostheses shed metal allo,, particles, some of which are radiographically undetectable due to the limitatic s of their resolvable size. Key words: joint arthroplasty, porous coating, bead size, radiographic resolution.
Pore sizes in the range of 1 5 0 - 4 5 0 g.m are currently used, which were determined to be optimal based on limited animal studies (10). To achieve pore sizes in this range with sintered bead porous-coated implants, various bead sizes have been used, ranging from 100 to 750 i.tm. Initially, a porous-coated M o o r e type prosthesis (DePuy) was reported, which used 100 Ixm beads to obtain a pore size of approximately 100 p.m (6). The prosthesis was later m o d ified to achieve pore sizes within the 1 5 0 - 3 0 0 I.tm range. The D e p u y Anatomic Medullary Locking (AML) and Low Contact Stress (LCS) prostheses presumably use the s a m e porous coating as that on the modified M o o r e prosthesis. The Howmedica PCA prosthesis uses larger beads of 7 0 0 - 7 5 0 i~m, to create an average pore size of 425 ttm (8). There h a v e been several reports in the literature of loosening of beads from the porous-coated anatomical (PCA) prostheses (4, 1 I, 12), but there appear to be no reports of individual bead loosening from other porous-coated prostheses. Although the
The long-term stable fixation of total hip and total k n e e prostheses has assumed increased importance in the field of joint arthroplasty. Cemented fixation techniques are being replaced by cementless biological fixation due to the rate of loosening associated with cemented fixation, which has been related to " c e m e n t b o n e disease"(9). Cementless fixation relies on the ability to achieve b o n y ingrowth into the prosthesis by various types of porous coating, including sintered fiber-metal, sintered beads, and plasma flame-spray (13). There appear to be m a n y necessary conditions for b o n y ingrowth, which include a pore size of 5 0 - 4 0 0 p.m (1), intimate apposition to bone, and minimal implant m o v e m e n t relative to adjacent bone (3).
From the Department of Orthopaedic Surgery, Johns Hopkins University School of Medicine, Bahimore, Maryland.
Reprint requests: Darrel W. Haynes, Phl), Department of Orthopaedic Surgery, Johns Hopkins University School of Medicine, 5601 Loch Raven Boulevard, Baltimore, MD 21239. 117
118
The Journal of Arthroplasty Vol. 5 No. 2 June 1990
techniques of high-temperature sintering of metal alloy powders on the substrate surface are constantly being improved, it would appear likely that occasional beads may separate from all beaded porouscoated surfaces. In attempting to explain the apparent absence of smaller bead separation it was hypothesized that the smaller beads could not be resolved due to the limitations of radiographic techniques. To test this hypothesis, we performed the following experiment.
Materials and Methods Chrome cobalt beads similar to those used to coat prostheses, with diameters confirmed by scanning electron microscopy of 200, 300, 400, 500, 600, and 700 I-tm (obtained from Howmedica), were arranged between two pieces of transparent tape in two configurations. The first configuration was comprised of a single layer of multiple numbers of each bead size between 200 and 700 Ixm (Figs. 1-5). The second configuration was a five-bead pattern of each bead size within the same 2 0 0 - 7 0 0 Ixm range (Figs. 6 10). To simulate in vivo conditions, the femoral component of the PCA total hip prosthesis was implanted into a cadaveric hip, with skin and soft tissue intact, using the standard surgical procedure of the direct lateral approach. After implantation, a slit was cut the thickness of a saw blade with an oscillating bone saw in the lateral femoral cortical bone adjacent to the proximal porous-coated portion of the prosthesis. Care was taken to ensure that the window was in the plane perpendicular to both the anterior/posterior plane and the radiographic beam. The prepared bead configurations were placed sequentially in the window and radiographs were taken using a clinical mobile radiographic apparatus at 70 kV, 50 mA, at a distance of 40 inches from the tube to the film (Figs. 1, 6). These settings were chosen since they are used as the standard technique for hip radiography. High-resolution CRONEX intensifying screens were used with Kodak X-OMAT film. With time acting as the dependent variable, films were exposed at ~o, ~, and ~o second, with ~ being the optimum exposure. The two other exposures were used for bracketing purposes since it is recognized that clinical films are not ahvays of the optimal exposure. To compare the effect of radiographs without bone or soft tissue, exposures of both configurations were made outside the cortical window at 70 kV, 50 kV, and 35 kV at 50 mA with exposure times ranging from ~ to ~ second (Figs. 3, 8).
Fig. 1. Single-layer configuration of beads laying in window adjacent to a PCA femoral prosthesis implanted in a cadaver with soft tissue and skin intact (70 kV, 50 mA, second).
Results The limits of apparent radiographic resolution differed between the two configurations. With the second configuration of five beads, the entire range of sizes was radiographically apparent when placed in direct contact with the film cassette and exposed for second at 35 kV and 50 mA (Fig. 8). However, with the single-layer multiple-numbered beads, the 200 p,m sizes appeared as an amorphus patch and the individual beads could not be resolved (Fig. 3).
Porous Coated Beads
200
I
•
Haynes and Freedman
119
"~ .~'..',..,"
, :':~.:
.,..
-
,'2'
:
3O0 },
..'"
,:..
_..'.~:.,..: .:.: ~ . . . .
400
*o
e= . " . ~
o61 °
L~..,~'v'4L; ,% :'.J: ""
".'~.-.
.....
..:-\~i~
500 ~I. tl
600
IN
700
Figs. 2-4. Fig. 2 (Left). Optical view of the single-layer configuration of 200-700 p,m beads. Fig. 3 (Middle). Single-layer configuration radiographed in contact with the film cassette (35 kV, 50 mA, ~ second). Note the inability to individually resolve the 200 and 300 gm beads. Fig. 4 (Right). Enlargement of Figure 1 with less exposure (70 kV, 50 mA, -~ second).
W h e n the configurations were placed in the window adjacent to the prosthesis and radiographed with the optimum exposure, the proximal 200 and 300 i.tm size beads appeared as an opaque patch in the single-layer multiple-numbered configuration, and individual beads could not be resolved (Figs. 4, 5). The same results were observed in the five-bead configuration, where the proximal 200 and 300 p.m beads could not be resolved (Figs. 9, 10). By decreasing the exposure, the resolution could not be
improved., and increasing the exposure resulted in a gradual loss of overall resolution (Figs. 4, 5, 9, 10).
Discussion The results clearly demonstrate that there are limits in the ability to resolve differing sizes of metal alloy beads radiographically. Using clinical radiographic techniques and standard exposure time for
120
The Journal of Arthroplasty Vol. 5 No. 2 June 1990
E
Fig. 5. Enlargement of Figure 1 with same exposure (70 kV, 50 mA, ~ second). The 200 and 300 Fm beads appear as a patch, but individual beads ate impossible to resolve.
human hips, bead sizes of less than 400 p.m cannot be resolved and are therefore invisible. Even with the beads placed directly on the film cassette with reduced exposures, individual beads of less than 300 ~m cannot be clearly resolved. Resolution is the smallest distance between two dots or lines that can be defined separately. The resolution in a radiographic image is dependent upon image clarity, which is the result of both image quality and contrast (5). Factors that affect the image quality are the defuse light transfers from the phos-
Fig. 6. Five-bead configuration of beads lay~u~,, . . . . . . . . . dow as in Figure 1 (70 kV, 50 mA, ~ second).
phorous layer of the intensifying screen to the film. The light is further defused by the double-sided photographic emolution on the film. Factors that affect the contrast ate differences in the thickness and density of the object. The difference between a 200 p.m and a 700 p.m bead is not only their width but also their thickness. As a 700 Ixm bead is almost 4 times thicker than a 200 v m bead, more x-rays will be absorbed and the image will appear sharper, affording better image quality, and therefore more resolvability than the 200 I-ttn bead. It has been reported
Porous Coated Beads * Haynes and Freedman 121
200
300
400
500
600
0 •
•
P
700
•
"" o
* •
•
•
I
Figs. 7-9. Fig. 7 (Left). Optical view of the five-bead configuration of 200-700 gtm beads. Fig. 8 (Middle). Five-bead configuration radiographed in contact with the film cassette (35 kV, 50 mA ~ second). Note that all the beads are radiographically apparent, but it is difficult to resolve the 200 and 300 p.m beads as spheres. Fig. 9 (Right). Enlargement of Figure 6 with less exposure (70 kV, 50 mA, ~ second).
that in the ideal radiograph with the best image clarity, five line pairs per millimeter (200 p.m) is the m a x i m u m resolution (5). Since the sintering techniques used among the different porous-coated prostheses are similar, it must be assumed that some beads fall off all prostheses at an equal rate, h o w e v e r minute the-number may be. The inability to resolve beads less than 400 I~m may account for the fact that Ranawat et al. (11) cited loose beads only from the PCA (Howmedica) prosthesis and not the UCI (DePuy) prosthesis. The lack of resolution below 400 IJ-m applies not only to loosened beads but also to all miscellaneous loosened debris smaller than the resolvable limit. This fact was
demonstrated by Buchert et al. (2), w h o reported sintered panicles detached from a non-bead-coated prosthesis, which were radiographically invisible but histologically evident. Although there has been no clinical evidence that bead detachment is a causative agent in prosthesis loosening, there have been m a n y reports recording the n u m b e r of loose beads associated with radiolucent lines only in the PCA prosthesis (4, 11, 12). As the PCA porous coating utilizes the larger 750 lzm beads, which are radiographically visible, one must question whether there are loose beads or debris in the radiolucent lines of other prostheses, which use beads that are too small to resolve radiographically.
122
The Journal of Arthroplasty Vol. 5 No. 2 June 1990
References
J
t o
.L Fig. 10. Enlargment of Figure 6 with the same exposure (70 kV, 50 mA, -h second).The 200 and 300 ttm beads are radiographically invisible.
1. Bobyn JD, Pilliar RM, Cameron HU, Weatherly GC: The optimum pore size for the fixation of porous-surfaced metal implants by the ingrowth of bone. Clin Orthop 150:263, 1980 2. Buchert PK, Vaughn BK, Mallory TH, et al: Excessive metal release due to loosening and fretting of sintered panicles on porous coated hip prostheses. J Bone Joint Surg 68A:606, 1986 3. Cameron HU, Pilliar RM, Macnab I: The effect of movement on the bonding of porous metal to bone. J Biomed Mater Res 7:301, 1973 4. Cheng CL, Gross AE: Loosening of the porous coating in total knee replacement: J Bone Joint Surg 70B:377, 1988 5. Christensen EE, Curry TS III, Dowdey JE: An introduction to the physics of diagnostic radiology. Lea and Febiger; Philadelphia, 1978 6. Engh CA: Hip arthroplasty with a Moore prosthesis with porous coating. Clin Onhop 176:52, 1983 7. Engh CA, Bobyn JD, Glassman AH:Porous coated hip replacement: tile factors governing bone ingrowth stress shielding and clinical results. J Bone Joint Surg 69B:45, 1987 8. Hungerford DS, Kenna RV: Preliminary experience with a total knee prosthesis with porous-coating used without cement. Clin Orthop 176:95, 1983 9. Jones LC, Hungerford DS: Cement disease. Clin Orthop 225:192, 1987 10. Pilliar RM: Powder metal-made orthopaedic implants with porous surface for fixation by tissue growth. Clin Orthop 176:42, 1983 11. Ranawat CS, Johanson NA, Rimnac CM et al: Retrieval analysis of porous-coated components for total knee anhroplasty. Clin Onhop 209:244, 1986 12. Rosenqvist R, Bylander B, Knutson K et ah Loosening of the porous coating of bicompanmental prostheses in patients with rheumatoid arthritis. J Bone Joint Surg 68A:538, 1986 13. Turner TM, Sumner DR, Urban RM et al: A comparative study of porous coatings in a weight bearing total hip arthroplasty model. J Bone Joint Surg 68A: 1396, 1986