J. BIOMED. MATER. RES. SYMPOSIUM
No. 7, pp. 485492 (1976)
Bone Ingrowth into Dynamically Loaded Porous-Coated Intramedullary Nails J. P. COLLIER and G. A. COLLIGAN, Thayer School of Engineering, Dartmouth College, Hanover, New Hampshire 08755, and S . A. BROWN, Department of Surgery, Dartmouth Medical School, Hanover, New Hampshire 03755
Summary Research to determine whether porous-coated Vitallium intramedullary rods could be used to bridge segmental defects in long bones was performed using rabbit tibias as a suitable model for human bone growth. One-centimeter segments of the midshaft of tibias of mature rabbits were removed surgically and replaced with equal-sized segments of Silastic tubing to maintain leg length. A coated rod was inserted through the proximal tibia1 segment, through the tubing, and into the distal bone segment. The legs were taken out of plaster a t 3 weeks, and x-rays were taken periodically until sacrifice. At sacrifice, 30 weeks postoperatively, the mechanical properties of both tibias from each rabbit were measured destructively in a torsional testing machine. The load at failure of the tibia with the segmental defect averaged 90% of the contralateral control tibia. Rod pullout data and electron-probe x-ray microanalysis indicated that a substantial rod-endosteal bone bond existed due t o ingrowth into the porous coating. The torsional data may have been somewhat obscured by the presence of a bony callus which grew over the Silastic tubing and bridged the segmental defect exosteally in every case.
INTRODUCTION There is a need for a material to replace bone lost due to accident, disease, or birth defect. While solid metal implants have proven useful, there is no current method of obtaining a satisfactory bond between bone and the implant. One limiting characteristic of a solid metal implant is that its modulus of elasticity is much greater than that of bone, causing large amounts of energy t o be transferred across small areas of the implant during loading. This can eventually lead to necrosis of adjoining bone followed by loosening of the implant 486 @ 1976 by John Wiley & Sons, Inc.
486
COLLIER, COLLIGAN, AND BROWN
as the necrotic bone is resorbed. Porous metal coatings have a lower modulus of elasticity than solid metal, and also permit bony tissue ingrowth for mechanical stabilization. Such coatings should help to distribute loads more evenly, thereby reducing the possibility of necrosis.
BACKGROUND
It has been shown that bone will grow into a porous coating if the pore size is greater than 100 microns' and the amount of motion of the implant is limited.2 I n 1973, Cameron et al.3 showed that cortical bonc would not grow into porous-coated Vitallium staples used to fix unstable osteotomy sites in canine tibias. They believed that relative macromotion of the bone segments was present and proved to be too great for bony ingrowth to occur. The following experiment was designed to determine whether bone would grow into a porous coating subjected to the large loadings present a t the site of a segmental defect. The rabbit tibia was chosen as a model because of its size, the availability of rabbits of known species and age group, and ease of caring for them. METHODS Implant Fabrication The porous-coated implants were made using a nonpressed sintering method. A high-temperature sintering technique was used to produce localized melting and wclding of the Vitallium spheres rather than a lower temperature technique which takes longer and relies on diffusion. The rods were made one a t a time by sintering the coating onto a solid core. The core was a 10 cm length of a .25 cm Vitallium rod. The core was placed in a .36-cm i.d. quartz tube and the annulus area packed with 100 to 200 micron Vitallium spheres. The ends of the tube were sealed and the tube inserted into a U-tube through which dry hydrogen flowed to prevent oxidation. The U-tube was inserted in a 1350°C furnace for 15 min. The resulting porous-coatcd rod was cut t o 7.5 cm length and one end was ground round for ease in surgical insertion. The implants were passivated in dilute nitric acid and autoclaved prior to use. Implant coatings subjected to testing showed a shear strength of 9.6 x l o y N / m 2 (14,000 psi) a t a density of 6575 of theoretical.
BONE INGROWTH INTO POROUS-COATED NAILS
487
Surgical Procedure Surgery was performed on mature New Zealand white rabbits under general anesthesia and full sterile cover. Preoperative x-rays were taken to insure that there were no abnormalities in the test tibia. Animals were anesthetized using a combination of intramuscular and intravascular injections of Thorazine, sodium pentobarbitol, and Ketamine (see Appendix). The surgical team was scrubbed and gowned. The shaved right leg of the rabbit was scrubbed with Betadine and the animal draped with sterile linens. A 5-cm incision was made over the anterior-medial aspect of the right tibia1 diaphysis extending distally from the fibular anklyosis. The skin and muscles were retracted and the periosteum elevated. A 1 cm segment of the tibia was removed by making two parallel cuts with a sterilized cast saw and saline coolant. A piece of Silastic tubing was cut to length and used to replace the bone segment and maintain leg length. The medullary canal was reamed and a hole drilled through the proximal medial metaphysis. The rod was inserted through an incision over the hole through proximal bone segment, through the tubing, and into the distal bone segment. The periosteum and the skin were closed in layers, the leg was wrapped in Webril, and a plaster cast was applied. The cast was removed after three weeks and x-rays taken a t threeweek intervals until sacrifice a t 30 weeks.
Testing The rabbit tibias were subjected to torsional loading and implant pullout tests following sacrifice. Care was taken t o keep the bones moist throughout this testing. The torsion testing machine was similar to the device used by Burnstein and F ~ - a n k e l . ~It utilized a swinging pendulum to produce high strain rates to break the tibias which had been dissected free of soft tissue and had their ends mounted in methyl methacrylate cups to fit the machine. A load cell and variable potentiometer were used to measure load and deflection values which were displayed on an oscilloscope and photographed for analysis. The torsion testing provided data on energy absorption prior to failure modulus of elasticity of the bones, ductility (angular deflection), and load a t failure. In each case, the contralateral tibia was used as the control and the fracture strengths and other measure-
COLLIER, COLLIGAN, AND BROWN
488
ments were all expressed as a percentage of the control. This eliminated the effect of variations in animal size and strength on the results. The torsion test, while providing much valuable data, resulted in shattered experimental and control tibias. I n cases where the implant was still bonded to one end of the tibia, a pullout strength test was performed. I n this test, the implant was fitted tightly through a hole in a plate connected to the crosshead of an Instron tensile testing machine and clamped to a stationary holder. As the crosshead moved downward against the cut bone surface, it pulled the rod out of the intramedullary cavity. Cross sections of tibias with implants which were not subjected to torsional testing were mounted and polished and viewed with a microscope and a n electron-probe x-ray microanalyzer. The microscope sections were analyzed for bone structure and type. The electron probe was used t o analyze the samples for calcium levels t o determine the density of the ingrown bone as well as cobalt to dctermine the extent of bone ingrowth.
RESULTS Table I is a summary of the data obtained from eight rabbits sacrificed a t 30 weeks postoperatively. The average load a t failure of the leg with the implant was 91% of the contralatcral control tibia. TABLE I Summary of Postoperative D a t a a ~~~
Rabbit 1 2 3 4
5 6 7 8 Avg.
Energy absorbed
Ductility
Load a t failure
Torsional stiffness
Pullout strength
86 50 65 50 46 63 54 110 65
77 67 71 55 53 75 63 100 70
112 76 92 90 84 85 81 10'3 91
146 113 130 163 158 112 129 109 129
140 kg 22 kg
-
-
~
a
Values given are percentage of control except for pullout strength.
BONE INGROWTH INTO POROUS-COATED NAILS
489
The ductility, i.e., the angular deflection of leg with implant, was 70% of control; the energy absorbed was 65% of control, and the torsional stiffness was 129y0of the control side tibia. Optical analysis of the mounted and polished cross sections showed bone growth into the coatings of the implants a t both proximal and distal ends. Haversian systems were seen with the aid of a microscope in the pores of the coatings, and electron-probe analysis confirmed calcium levels in the pores were equal to those of the surrounding cortical bone.
DISCUSSION Examination of the radiologic appearance of the tibia in response t o the surgical procedure provides some insight into the implications of the mechanical results. Figure 1 is an x-ray of a rabbit six weeks postoperative. The rabbit has been moving about and weight bearing for three weeks. The callus over the segmental removal site has bridged the defect. Physical examination showed satisfactory stability of the leg. No callus growth anteriorly or medially can be seen.
Fig. 1.
X-Ray of rabbit taken six weeks postoperative.
490
COLLIER, COLLIGAN, AND BROWN
Fig. 2 . X-Ray of the rabbit taken 23 weeks after surgery.
Figure 2 is an x-ray of the same rabbit 23 weeks after surgery. Significant bone remodeling has occurred. The callus has been partially resorbed, and calcified bone of density (radio-opacity) equal to the surrounding bone can be seen. The remodeling extends proximally above the fibular anklyosis and distally a t least 1 em below the defect. A thin sheath of bone can be seen to completely cover the tubing which separated the bone segments. Bone growth has occurred in the medial and anterior planes and the tibia was beginning to appear structurally similar to the control tibia. Figure 3 is an x-ray of the rabbit 30 weeks after surgery. Bony bridges between endosteal bone and the implant can be seen a t several locations. This growth, in conjunction with the decreasing diameter of the bone sheath over the tubing, may indicate that an increased structural load was being carried by the implant. The increase in stiffness of the experimental tibia is believed to be related to two features seen on x-ray. The portion of tibia containing the implant acts as a composite material with a modulus close to that of bone. However, the segmental replacement region hehaves simply as a porous-coated rod with a modulus greater than twice that of bone. Thus the rod-tibia combination has a higher
BONE INGROWTH INTO POROUS-COATED NAILS
Fig. 3.
49 1
X-Ray of the rabbit taken 30 weeks after surgery.
modulus than control. The stiffness of the experimental tibia is further increased by the formation of a bony sheath or periosteal callus seen on the x-rays t o be bridging the gap between the two fragments. Since the diameter of the sheath is greater than the contralateral diaphyseal bone, it would have a greater structural stiffness. The increase in structural stiffness caused a decrease in ductility or angular rotation prior t o failure. As a result, the amount of energy absorbed prior t o failure was substantially less than control, even though the ultimate strength was nearly equal t o control. The ability of the rabbit t o bridge the 1 cm defect created b y the insertion of the silastic tubing may have obscured the rod-bone interface contribution t o the torsional testing data. However, the results of the pullout test indicated that a significant bond existed between the porous-coated rod and the endosteal surface of the bones.
CONCLUSIONS The radiologic appearance of segmental defects filled with Silastic tubing and stabilized with porous-coated Vitallium rods demonstrates that the rods provide sufficient stabilization for bony bridging
492
COLLIER, COLLIGAN, AND BROWN
of the defect, and that endosteal new bone forms t o create a solid bone-porous coating bond. Although torsional assessment of the bone-rod interface strength may have been masked by the perosteal bony sleeve, the torsional tests did demonstrate an ultimate strength of 91% of control. Rod pullout results demonstrated that a significant bond had developed between porous coating and bone. The microscopic and microprobe analysis demonstrated ingrowth of haversian bone. These results would suggest that bone can grow into a porouscoated implant in a dynamically loaded situation, and that porous coatings may prove to be a satisfactory solution to the implantloosening problem. Furthermore, the results of this rabbit model would indicate the porous-coated intramedullary rods may provide a satisfactory solution to the problem of segmental bone loss due to disease or trauma.
APPENDIX Rabbit Anesthesia Protocols 1. 2. 3. 4.
Thorazine i.m., 1-2 mg/kg; 45 min preoperatively. Sodium pentobarbitol i.v., 10 mg/kg. Ketamine, HCl i.v., 10 mg/kg. Ketamine, 10 mg/kg. i.m. during surgery (p.r.n.).
The authors wish to acknowledge their assistants, Mr. Victor Surprenant and Mr. Leo Ihuphinais. This work was supported in part by a gift from G. T. and E .
References 1 . S. F. Hulbert, J. J. Klawitter, C. I). Talbert, F. A. Young, R. S. Mathews, and F. H. Stelling, “Potential of ceramics as permanently implantable skeletal prosthesis,” presented a t the Second Materials Engineering and Sciences Conf., national meeting of the American Institute of Chemical Engineers, Atlanta, Georgia, February, 1970. 2 . P . R. Welsh, R. M. Pilliar, and I . Macnab, J . Bone Joint Sury., 53-A (5) 963 (1971). 3. H. Cameron, R . M. Pilliar, and I. Macnab, J . Biomed. Muter. Res., 7 , 301 (1973). 4. A. H. Burstein and V. H. Frankel, J . Biomeeh., 4, 155 (1971). 5. N. W. Klein and R. 0. Becker, personal communication, 1972.
No. 7, pp. 485492 (1976)
Bone Ingrowth into Dynamically Loaded Porous-Coated Intramedullary Nails J. P. COLLIER and G. A. COLLIGAN, Thayer School of Engineering, Dartmouth College, Hanover, New Hampshire 08755, and S . A. BROWN, Department of Surgery, Dartmouth Medical School, Hanover, New Hampshire 03755
Summary Research to determine whether porous-coated Vitallium intramedullary rods could be used to bridge segmental defects in long bones was performed using rabbit tibias as a suitable model for human bone growth. One-centimeter segments of the midshaft of tibias of mature rabbits were removed surgically and replaced with equal-sized segments of Silastic tubing to maintain leg length. A coated rod was inserted through the proximal tibia1 segment, through the tubing, and into the distal bone segment. The legs were taken out of plaster a t 3 weeks, and x-rays were taken periodically until sacrifice. At sacrifice, 30 weeks postoperatively, the mechanical properties of both tibias from each rabbit were measured destructively in a torsional testing machine. The load at failure of the tibia with the segmental defect averaged 90% of the contralateral control tibia. Rod pullout data and electron-probe x-ray microanalysis indicated that a substantial rod-endosteal bone bond existed due t o ingrowth into the porous coating. The torsional data may have been somewhat obscured by the presence of a bony callus which grew over the Silastic tubing and bridged the segmental defect exosteally in every case.
INTRODUCTION There is a need for a material to replace bone lost due to accident, disease, or birth defect. While solid metal implants have proven useful, there is no current method of obtaining a satisfactory bond between bone and the implant. One limiting characteristic of a solid metal implant is that its modulus of elasticity is much greater than that of bone, causing large amounts of energy t o be transferred across small areas of the implant during loading. This can eventually lead to necrosis of adjoining bone followed by loosening of the implant 486 @ 1976 by John Wiley & Sons, Inc.
486
COLLIER, COLLIGAN, AND BROWN
as the necrotic bone is resorbed. Porous metal coatings have a lower modulus of elasticity than solid metal, and also permit bony tissue ingrowth for mechanical stabilization. Such coatings should help to distribute loads more evenly, thereby reducing the possibility of necrosis.
BACKGROUND
It has been shown that bone will grow into a porous coating if the pore size is greater than 100 microns' and the amount of motion of the implant is limited.2 I n 1973, Cameron et al.3 showed that cortical bonc would not grow into porous-coated Vitallium staples used to fix unstable osteotomy sites in canine tibias. They believed that relative macromotion of the bone segments was present and proved to be too great for bony ingrowth to occur. The following experiment was designed to determine whether bone would grow into a porous coating subjected to the large loadings present a t the site of a segmental defect. The rabbit tibia was chosen as a model because of its size, the availability of rabbits of known species and age group, and ease of caring for them. METHODS Implant Fabrication The porous-coated implants were made using a nonpressed sintering method. A high-temperature sintering technique was used to produce localized melting and wclding of the Vitallium spheres rather than a lower temperature technique which takes longer and relies on diffusion. The rods were made one a t a time by sintering the coating onto a solid core. The core was a 10 cm length of a .25 cm Vitallium rod. The core was placed in a .36-cm i.d. quartz tube and the annulus area packed with 100 to 200 micron Vitallium spheres. The ends of the tube were sealed and the tube inserted into a U-tube through which dry hydrogen flowed to prevent oxidation. The U-tube was inserted in a 1350°C furnace for 15 min. The resulting porous-coatcd rod was cut t o 7.5 cm length and one end was ground round for ease in surgical insertion. The implants were passivated in dilute nitric acid and autoclaved prior to use. Implant coatings subjected to testing showed a shear strength of 9.6 x l o y N / m 2 (14,000 psi) a t a density of 6575 of theoretical.
BONE INGROWTH INTO POROUS-COATED NAILS
487
Surgical Procedure Surgery was performed on mature New Zealand white rabbits under general anesthesia and full sterile cover. Preoperative x-rays were taken to insure that there were no abnormalities in the test tibia. Animals were anesthetized using a combination of intramuscular and intravascular injections of Thorazine, sodium pentobarbitol, and Ketamine (see Appendix). The surgical team was scrubbed and gowned. The shaved right leg of the rabbit was scrubbed with Betadine and the animal draped with sterile linens. A 5-cm incision was made over the anterior-medial aspect of the right tibia1 diaphysis extending distally from the fibular anklyosis. The skin and muscles were retracted and the periosteum elevated. A 1 cm segment of the tibia was removed by making two parallel cuts with a sterilized cast saw and saline coolant. A piece of Silastic tubing was cut to length and used to replace the bone segment and maintain leg length. The medullary canal was reamed and a hole drilled through the proximal medial metaphysis. The rod was inserted through an incision over the hole through proximal bone segment, through the tubing, and into the distal bone segment. The periosteum and the skin were closed in layers, the leg was wrapped in Webril, and a plaster cast was applied. The cast was removed after three weeks and x-rays taken a t threeweek intervals until sacrifice a t 30 weeks.
Testing The rabbit tibias were subjected to torsional loading and implant pullout tests following sacrifice. Care was taken t o keep the bones moist throughout this testing. The torsion testing machine was similar to the device used by Burnstein and F ~ - a n k e l . ~It utilized a swinging pendulum to produce high strain rates to break the tibias which had been dissected free of soft tissue and had their ends mounted in methyl methacrylate cups to fit the machine. A load cell and variable potentiometer were used to measure load and deflection values which were displayed on an oscilloscope and photographed for analysis. The torsion testing provided data on energy absorption prior to failure modulus of elasticity of the bones, ductility (angular deflection), and load a t failure. In each case, the contralateral tibia was used as the control and the fracture strengths and other measure-
COLLIER, COLLIGAN, AND BROWN
488
ments were all expressed as a percentage of the control. This eliminated the effect of variations in animal size and strength on the results. The torsion test, while providing much valuable data, resulted in shattered experimental and control tibias. I n cases where the implant was still bonded to one end of the tibia, a pullout strength test was performed. I n this test, the implant was fitted tightly through a hole in a plate connected to the crosshead of an Instron tensile testing machine and clamped to a stationary holder. As the crosshead moved downward against the cut bone surface, it pulled the rod out of the intramedullary cavity. Cross sections of tibias with implants which were not subjected to torsional testing were mounted and polished and viewed with a microscope and a n electron-probe x-ray microanalyzer. The microscope sections were analyzed for bone structure and type. The electron probe was used t o analyze the samples for calcium levels t o determine the density of the ingrown bone as well as cobalt to dctermine the extent of bone ingrowth.
RESULTS Table I is a summary of the data obtained from eight rabbits sacrificed a t 30 weeks postoperatively. The average load a t failure of the leg with the implant was 91% of the contralatcral control tibia. TABLE I Summary of Postoperative D a t a a ~~~
Rabbit 1 2 3 4
5 6 7 8 Avg.
Energy absorbed
Ductility
Load a t failure
Torsional stiffness
Pullout strength
86 50 65 50 46 63 54 110 65
77 67 71 55 53 75 63 100 70
112 76 92 90 84 85 81 10'3 91
146 113 130 163 158 112 129 109 129
140 kg 22 kg
-
-
~
a
Values given are percentage of control except for pullout strength.
BONE INGROWTH INTO POROUS-COATED NAILS
489
The ductility, i.e., the angular deflection of leg with implant, was 70% of control; the energy absorbed was 65% of control, and the torsional stiffness was 129y0of the control side tibia. Optical analysis of the mounted and polished cross sections showed bone growth into the coatings of the implants a t both proximal and distal ends. Haversian systems were seen with the aid of a microscope in the pores of the coatings, and electron-probe analysis confirmed calcium levels in the pores were equal to those of the surrounding cortical bone.
DISCUSSION Examination of the radiologic appearance of the tibia in response t o the surgical procedure provides some insight into the implications of the mechanical results. Figure 1 is an x-ray of a rabbit six weeks postoperative. The rabbit has been moving about and weight bearing for three weeks. The callus over the segmental removal site has bridged the defect. Physical examination showed satisfactory stability of the leg. No callus growth anteriorly or medially can be seen.
Fig. 1.
X-Ray of rabbit taken six weeks postoperative.
490
COLLIER, COLLIGAN, AND BROWN
Fig. 2 . X-Ray of the rabbit taken 23 weeks after surgery.
Figure 2 is an x-ray of the same rabbit 23 weeks after surgery. Significant bone remodeling has occurred. The callus has been partially resorbed, and calcified bone of density (radio-opacity) equal to the surrounding bone can be seen. The remodeling extends proximally above the fibular anklyosis and distally a t least 1 em below the defect. A thin sheath of bone can be seen to completely cover the tubing which separated the bone segments. Bone growth has occurred in the medial and anterior planes and the tibia was beginning to appear structurally similar to the control tibia. Figure 3 is an x-ray of the rabbit 30 weeks after surgery. Bony bridges between endosteal bone and the implant can be seen a t several locations. This growth, in conjunction with the decreasing diameter of the bone sheath over the tubing, may indicate that an increased structural load was being carried by the implant. The increase in stiffness of the experimental tibia is believed to be related to two features seen on x-ray. The portion of tibia containing the implant acts as a composite material with a modulus close to that of bone. However, the segmental replacement region hehaves simply as a porous-coated rod with a modulus greater than twice that of bone. Thus the rod-tibia combination has a higher
BONE INGROWTH INTO POROUS-COATED NAILS
Fig. 3.
49 1
X-Ray of the rabbit taken 30 weeks after surgery.
modulus than control. The stiffness of the experimental tibia is further increased by the formation of a bony sheath or periosteal callus seen on the x-rays t o be bridging the gap between the two fragments. Since the diameter of the sheath is greater than the contralateral diaphyseal bone, it would have a greater structural stiffness. The increase in structural stiffness caused a decrease in ductility or angular rotation prior t o failure. As a result, the amount of energy absorbed prior t o failure was substantially less than control, even though the ultimate strength was nearly equal t o control. The ability of the rabbit t o bridge the 1 cm defect created b y the insertion of the silastic tubing may have obscured the rod-bone interface contribution t o the torsional testing data. However, the results of the pullout test indicated that a significant bond existed between the porous-coated rod and the endosteal surface of the bones.
CONCLUSIONS The radiologic appearance of segmental defects filled with Silastic tubing and stabilized with porous-coated Vitallium rods demonstrates that the rods provide sufficient stabilization for bony bridging
492
COLLIER, COLLIGAN, AND BROWN
of the defect, and that endosteal new bone forms t o create a solid bone-porous coating bond. Although torsional assessment of the bone-rod interface strength may have been masked by the perosteal bony sleeve, the torsional tests did demonstrate an ultimate strength of 91% of control. Rod pullout results demonstrated that a significant bond had developed between porous coating and bone. The microscopic and microprobe analysis demonstrated ingrowth of haversian bone. These results would suggest that bone can grow into a porouscoated implant in a dynamically loaded situation, and that porous coatings may prove to be a satisfactory solution to the implantloosening problem. Furthermore, the results of this rabbit model would indicate the porous-coated intramedullary rods may provide a satisfactory solution to the problem of segmental bone loss due to disease or trauma.
APPENDIX Rabbit Anesthesia Protocols 1. 2. 3. 4.
Thorazine i.m., 1-2 mg/kg; 45 min preoperatively. Sodium pentobarbitol i.v., 10 mg/kg. Ketamine, HCl i.v., 10 mg/kg. Ketamine, 10 mg/kg. i.m. during surgery (p.r.n.).
The authors wish to acknowledge their assistants, Mr. Victor Surprenant and Mr. Leo Ihuphinais. This work was supported in part by a gift from G. T. and E .
References 1 . S. F. Hulbert, J. J. Klawitter, C. I). Talbert, F. A. Young, R. S. Mathews, and F. H. Stelling, “Potential of ceramics as permanently implantable skeletal prosthesis,” presented a t the Second Materials Engineering and Sciences Conf., national meeting of the American Institute of Chemical Engineers, Atlanta, Georgia, February, 1970. 2 . P . R. Welsh, R. M. Pilliar, and I . Macnab, J . Bone Joint Sury., 53-A (5) 963 (1971). 3. H. Cameron, R . M. Pilliar, and I. Macnab, J . Biomed. Muter. Res., 7 , 301 (1973). 4. A. H. Burstein and V. H. Frankel, J . Biomeeh., 4, 155 (1971). 5. N. W. Klein and R. 0. Becker, personal communication, 1972.
