行政院國家科學委員會補助專題研究計畫成果報告
※ ※※※※※※※※※※※※※※※※※※※※※※※
※
※
※
※
脊椎內固定器疲勞壽命之改善
※
※
※
※ ※※※※※※※※※※※※※※※※※※※※※※※
※
計畫類別：▇個別型計畫
□整合型計畫
計畫編號：NSC
89
－
2213
－
E
－
002
－
136
－
執行期間：八十八年八月一日至八十九年七月三十一日
計畫主持人：陳博光 教授
共同主持人：
本成果報告包括以下應繳交之附件：
□赴國外出差或研習心得報告一份
□赴大陸地區出差或研習心得報告一份
執行單位：國立台灣大學醫學院骨科部
一、中文摘要 由於脊椎固定器所使用的 關鍵詞： 脊椎固定器、骨螺絲、疲勞壽命 Abstract Keywords: 二、緣由與目的
Spinal fixation is currently a preferred surgical procedure for stabilization of the vertebral column in degenerative spinal dis-orders. Many innovative pedicle-fixation systems have been widely adopted for clini-cal applications in recent years.29,30 Prior to clinical use, new devices need to be tested in vitro, either in isolation or in conjunction
with artificial or cadaver-derived spinal specimens.26 The results of these biome-chanical tests provide the information about construct stiffness, fatigue resistance, and failure mechanisms for further improve-ment.24
Both titanium alloy (Ti-alloy) and 316L stainless-steel (316L-SS) spinal implants are quite commonly used for clinical applications. Ti-alloy is considered superior in terms of biocompatibility, enhanced corrosion resis-tance, and fewer resultant artifacts on CT scans and MRI images.10,11,19,23 Acceptance of Ti-alloy spinal implants for clinical appli-cations has been hampered, however, by the sensitivity of the material to surface condition (notch effect). Even with relatively superfi-cial surface damage (eg. laser inscription), the fatigue strength of Ti-alloy is drastically reduced.12 Hence, Scuderi et al.28 have con-cluded that Ti-alloy is a poor choice for cer-vical-spine fixation systems due to increased notch sensitivity compared with 316L-SS. Further, Pienkowski et al.26 have reported
size and design. To date, refinement of de-sign and optimization of relevant factors to produce superior rigidity for Ti-alloy im-plants (over 316L-SS) still remains a chal-lenge and has not been extensively studied.
The Formosa posterior spinal implant (FPI, Syntec Scientific Inc., Taiwan) was de-signed in this laboratory for the stabilization of the spine and correction of deformity. The FPI implant consists of three smooth rods of 6.0 mm in diameter, with polished surfaces, pairs of pedicle screws, and two connecting plates as cross linkages. The final assembly is depicted schematically in Fig. 1. The design of the connecting plates subjectively ensures the formation of a more rigid rectangular frame, thus increasing implant stiffness. Liu et al.18 have reported that a 24% increase in pedicle-screw internal diameter resulted in an 104% increase in static strength. Therefore, two types of pedicle screw were incorporated in the design to investigate the effects of in-ner diameter of screw and notch sensitivity for Ti-alloy: a type A screw with uniform in-ner (3.7 mm) and outer (6.0 mm) diameters, and a tapering type B screw with inner and outer diameters equal at the screw hub (6.4 mm) in Fig. 2. These new implants were made from both 316L-SS and Ti-alloy (Car-penter Technology, Reading, PA), with iden-tical dimensions.
The purpose of this study was two-fold. Firstly, to test the mechanical performance of the FPI system by fixing prototypes to UHMWPE blocks as described in previous reports.3,7,17 Secondly, to investigate the in-fluence of materials, effect of screw design (at hub) and connecting-plates on construct stiffness and fatigue resistance.
三、結果
Failur e Mechanisms: All prototypes
incor-porating screw A (FPISS-Plate-A and FPITi-Plate-A)
consistently failed at the cephalic screw hub (Table 1). By contrast, all FPI ,
tensile stress. Typical fatigue striations were observed at the fracture surface and no sign of permanent rod and screw deformation was observed.
Multicycle Stiffness: The derived statistics
for multicycle stiffness and fatigue life of the five FPI implant groups are detailed in Table 2. The mean multicycle stiffness for the FPI implants are revealed in Fig. 3. On average, the stiffness of the 316L-SS prototypes, FPISS-Plate-A and FPISS-Plate-B, were respectively
77.5% (p<0.001) and 64.9% (p<0.001) greater than for the Ti-alloy prototypes, FPITi-Plate-A and FPITi-Plate-B. The use of
con-necting plates for FPISS-Plate-B variants raised
the stiffness for FPISS-B variants by about
22.2% (p<0.001). With the lower inner di-ameter, the FPISS-Plate-A and FPITi-Plate-A
sys-tems demonstrated lower bending stiffness of
138.8 and 78.2 N/mm, respectively f- a di
ference of 41.4% and 29.55% when com-pared to the FPISS-Plate-B and FPITi-Plate-B
sys-tems, respectively.
Fatigue Life: The FPISS-Plate-B and FPITi-Plate-B
prototypes achieved means of 290,000 and 400,000 cycles fatigue life, respectively, however standard deviations were as high as 14.5% and 13.1% of the mean (Table 2 and Fig. 4). The FPISS-Plate-A and FPITi-Plate-A
pro-totypes achieved means of 140,000 and 87,000 cycles fatigue life, with comparatively high standard deviations of 18.7% and 28.3%, respectively. The fatigue life for FPITi-Plate-B
was significantly greater (p<0.05) than that for FPISS-Plate-B, but the result for FPITi-Plate-A
and FPISS-Plate-A was reversed (p<0.05). On
average, the increases in fatigue life for FPISS-Plate-B and FPITi-Plate-B were 107.4% and
364.2% greater than FPISS-Plate-A and
FPITi-Plate-A, respectively. The removal of the
connecting plate in the FPISS-B variants
avoided the stress concentrated on the rod/plate junctions and increased the fatigue life by 38.2% compared to the FPISS-Plate-B
analogs (p<0.001).
四、討論
The purpose of the spinal implant is to provide better conditions for bony fusion. At the very least, the implant is expected to re-main intact until fusion is achieved. Other-wise, if any part of the implant breaks or comes loose after surgery, the stability of the fusing segments will be lost and/or the de-formity may recur, marking the failure of the surgical intervention. Although optimal im-plant design has yet to be established, evolu-tion of the design of the individual compo-nents and selection of the most appropriate spinal-implant materials are mandated as part of that process of refinement, and should be investigated thoroughly. In this study, we adopted a new design to compare the me-chanical performance of two metals, two screw shapes and the effect of connecting plates.
Failur e Mechanisms: As for previous
re-ports,14,25,31-33 failure mechanisms for our spinal-implant prototypes were bend-ing/fracture of pedicle screw and longitudinal rod, with failures consistently occurring at the cephalic, rather than caudal end. This may be because the plastic block on the ce-phalic side had a greater range of actuation than the block on the caudal side. The type A
screw reveals a fairly uniform thread depth throughout the threaded portion, with a rela-tively small inner diameter close to the hub compared with the type B screw. Conse-quently, FPISS-Plate-A and FPITi-Plate-A screws
typically broke/fractured at the hub, indicat-ing that the load was concentrated on the weaker part of the screw. For FPISS-Plate-B and
FPITi-Plate-B variants, the hub site was much
stronger due to the greater inner diameter. The cephalic fracture/bending of FPISS-Plate-B
and FPITi-Plate-B variants was as a result of the
use of connection plates, with fracture sites shifting from the screw hubs to rod/plate junctions in response to the concentration of induced stress in this area.32,33
Multicycle Stiffness: Implant stiffness was
measured to verify the ability of the implant to stabilize the spinal segments and enhance spinal fusion in vivo16,20. Clinically, the im-plant stiffness varies with the number of load cycles as a result of intraspecimen slippage and microcracks in the implant material. In this study, the multicycle stiffness at the steady-state phase was selected for its clinical values and it might not need implants for single load-to-failure test.
The construct stiffness of the stressed structure was influenced by two factors － structure design and material properties. For the type B screw, inner and outer screw di-ameters at the hub are equal resulting in 1.4-and 1.3-times greater multicycle stiffness compared to implants using type A screws, for both titanium alloy and stainless steel, respectively. As for previously published data,25,32 the bending stiffness of 316L-SS systems was definitely greater (p<0.001) than for the dimensionally identical Ti-alloy counterparts (Fig. 3). This is attributable to the fact that the elastic modulus for 316L-SS is 1.84 times greater than for Ti-alloy.13 Johnston et al.16 have suggested that in-creased implant stiffness enhances the incor-poration of fusion mass of the short segments.
some of the stress to the instrumented seg-ments and the stress-shielding effect may be minimized and implant life prolonged, even-tually promoting spinal fusion.5,20,34 In an evaluation of long-bone fracture, Uhthoff et al.35 conducted a study using the femoras from 36 beagle dogs, recommending titanium alloy as a promising material for bone plate with superior bone-remodeling characteristics (at the osteotomy level) compared with 316L-SS. Given this complex interaction, the increased stiffness of the 316L-SS implant may have no clinical advantage in practice.
The effects of cross-linkage on stiffness performance for posterior spinal implants has been measured by various researchers.8,9,33 These studies have revealed that cross-linkage did not significantly increase the stiffness of posterior implants for com-pression/flexion motion. In the present study, the increase in stiffness of FPISS-Plate-B
im-plants was 22.2% compared to the FPISS-B
variant (p<0.001). This discrepancy and the favorable results produced by adding cross-links may be attributable to the more rigid frame formed by the FPI connecting plates.
Fatigue Life: The fatigue life of the stressed
structure was also determined by structure and material. McKinley et al.22 and Yerby et al.36 have determined that the greatest bend-ing moment occurs at the screw hub. Hence, by minimizing the notch effect and enlarging the inner diameter, prototypes incorporating screw B demonstrated an fatigue life from 2.1 (316 L-SS) to 4.6 (Ti alloy) times higher than screw-A implants (Fig. 4). The endur-ance limit of Ti-alloy is about 1.68 times that of 316L-SS.13 Therefore, the superiority of Ti-alloy for the FPITi-Plate-B variant in terms of
fatigue limits emerged when the notch effect at the screw hub was minimized, producing a 38.5% increase in fatigue life compared to the 316L-SS counterparts. For Ti-alloy vari-ants incorporating lower-diameter type A
centration from the screw hub to the rod/plate junctions, reducing the fatigue life by 27.6% compared to FPISS-B, which was without
supplemental cross-links (p<0.001).
Interestingly, the increase in inner hub diameter from screw A to screw B resulted in significantly greater increases for fatigue life than for stiffness. For stainless steel, proto-types incorporating screw B were 31.8% stiffer than screw-A equivalents, while for ti-tanium alloy the screw-B variant was 41.9% more rigid. However, the same increase in fa-tigue life for the previous type of stainless steel was 107.4%, while for the titanium al-loy, screw-B implants demonstrated 364.2% longer fatigue life than screw-A analogs. This demonstrates that the structural change had a more pronounced influence on fatigue life than on stiffness, especially in the case of ti-tanium alloy and should be considered during the design process for new implants.
In addition to structure design and mate-rial properties, the biomechanical perform-ance of the spinal construct was additionally influenced by loading and boundary condi-tions. Normal physiological loading on the L3-L4 disc may involve compressive forces of about 200% body weight (≈1700 N for a 70-kg man)21. Also, Rohlmann et al.27 and Stambough et al.32 have demonstrated that the loading rate and small changes in relative position of the implants have the marked in-fluences on fatigue cracking site and the dis-tribution of the loads across the implants. In the worst-case scenario, without adequate load sharing through anterior bone- or de-vice-related support, the testing load (60 N ~ 600 N) in this study was below the aforemen-tioned physiological loads and only the fre-quency (5 Hz) was within the critical range. Hence, results for fracture site and fatigue life may not be directly applicable to other testing conditions. Also, this study was not able to establish the optimal trade-off between pull-out strength and pedicle-screw fatigue strength because no variations for screw-thread shape were systematically tested.
Johnston et al.15 have used bovine implants to investigate the contour effect for fatigue life of 316L-SS Luque rods. Correlating the
measured strain at the rod apex with the de-gree of deformity, it was determined that relatively low axial loads produced enough tensile stress to induce implant fatigue. With the surface damage that inevitably results from surgical contouring by bent bars, the vulnerability of the contoured Ti-alloy rod implants to fatigue failure may be greater than for the 316L-SS analogs. Very little is known, however, about the effect of rod contouring on fatigue resistance of Ti-alloy implant. Hence, we suggest that future stud-ies, involving the corpectomy model and utilizing contoured Ti-alloy rod, are ducted to explore the effects of surgical con-touring on component strength.
五、參考文獻
1. Ashman R, Birch JG, Bone LB. Me-chanical testing of spinal instrumentation. Clin Orthop. 1988;227:113-125.
2. Ashman R, Galpin R, Corin J. Biome-chanical analysis of pedicle screw in-strumentation systems in a corpectomy model. Spine 1989;14:1398-405.
3. ASTM: Standard test method for static bending properties of metallic bone plates. American Society for Testing Materials, Standard F382-86. Philadel-phia, PA.
4. Car son WL, Duffield RC, Arendt M. In-ternal forces and moments in transpe-dicular spine instrumentation. The effect of pedicle screw angle and transfixation: the 4R-4bar linkage concept. Spine 1990;15:893-901.
5. Cheal EJ, Spector M, Hayes WC. Role of loads and prosthesis material properties on the mechanics of the proximal femur after total hip arthroplasty. J Orthop Res. 1992;10:405-22.
6. Cotr el Y, Dubousset J, Guillaumat M. New universal instrumentation in spinal surgery. Clin Orthop. 1988;227:10-23. 7. Cunningham BW, Sefter JC, Shono Y.
Static and cyclic biomechanical analysis of pedicle screw spinal constructs. Spine 1993;18:1677-88.
8. Dick JC, Jones MP, Zdeblick TA. A biomechanical comparison evaluating the use of intermediate screws and cross-linkage in lumbar pedicle fixation.
J Spinal Disord. 1994;7:402-7.
9. Dick JC, Zdeblick TA, Bartel BD. Me-chanical evaluation of cross-link designs in rigid pedicle screw systems. Spine 1997;22:370-5.
10. Ebraheim NA, Rupp RE, Savolain ER. Use of titanium implants in pedicular screw fixation. J Spinal Disord. 1994;7:478-86.
11. Ebraheim NA, Savolain ER, Stitgen SH. Magnetic resonance imaging after pe-dicular screw fixation of the spine. Clin Orthop. 1992;279:133-7.
12. Gaebler C, Stanzl-Tschegg S, Heinze G. Fatigue strength of locking screws and prototypes used in small-diameter tibial nails: a biomechanical study. J Trauma-Injury Infection & Critical Care 1999;47:379-84.
13. Goel VK, Martz EO, and Park JB. Mate-rial in spinal surgery. J Muscul Res. 1998; 2:73-88.
14. Goel VK, WinterBottom JM, Weinstein JN. A method for the fatigue testing of pedicle screw fixation devices. J Biomech. 1984;27:1383-88.
15. Johnston CE, Ashman RB, Sherman MC. Mechanical consequences of rod con-touring and residual scoliosis in sub-laminar segmental instrumentation. J Orthop Res. 1987;5:206-16.
16. Johnston CE, Welch RD, Baker KJ. Ef-fect of spine construct stiffness on short segment fusion mass incorporation. Spine 1995;20:2400-7.
17. Kotani Y, Cunningham BW, Kanayama M. Static and fatigue biomechanical properties of anterior thoracolumbar in-strumentation systems. Spine 1999;24:1406-13.
18. Liu YK, Unjust GO, Bahr PA. Fatigue life improvement of nitrogen-ion im-planted pedicle screws. Spine 1990;15:311-7.
19. Long M, Rack HJ. Titanium alloys in
to-1991;16(Suppl):S190-7.
21. Nachemson A. The load on lumbar discs in different positions of the body. Clin Orthop. 1966;45,107-22.
22. McKinley TO, McLain RF, Yerby SA. The effect of pedicle morphometry on pedicle screw loading. A synthetic model. Spine 1997;22:246-52.
23. Ortiz O, Pait TG, McAllister P. Postop-erative magnetic r esonance imaging with titanium implants of the thoracic and lumbar spine. Neurosurgery 1996;38:741-5.
24. Panjabi MM. Biomechanical evaluation of spinal fixation devices: I. A concep-tual framework. Spine 1988;13:1129-34. 25. Pienkowski D, Stephens GC, Doers TM.
Multicycle mechanical performance of titanium and stainless steel transpedicular spine implants. Spine 1998;23:782-8. 26. Puttlitz CM, Goel VK, Pope MH.
Biomechanical testing sequelae relevant to spinal fusion and instrumentation. Or-thop Clin North Am. 1998;29:571-89. 27. Rohlmann A, Bergmann G, Graichen F.
A spinal fixation device for in vivo load measurement. J Biomech. 1994;27:961-7.
28. Scuderi GJ, Greenberg SS, Cohen DS. A biomechanical evaluation of magnetic resonance imaging-compatible wire in cervical spine fixation. Spine 1993;18:1991-4.
29. Sidhu KS, Herkowitz HN: Spinal instru-mentation in the management of degen-erative disorders of the lumbar spine. Clin Orthop. 1997;335:39-53.
30. Stambough JL: Posterior instrumentation for thoracolumbar trauma. Clin Orthop. 1997;335:73-88.
31. Stambough JL, El-khatib F, Genaidy AM. Strength and fatigue resistance of thora-columbar spine implants: an experimen-tal study of selected clinical devices. J Spinal Disord. 1999;12:410-4.
Disord. 1997;10:473-81.
33. Stambough JL, Sabri EH, Huston RL. Effects of cross-linkage on fatigue life and failure modes of stainless steel pos-terior spinal constructs. J Spinal Disord. 1998;11:221-6.
34. Sumner DR, Galante JO: Determinants of stress shielding: design versus material versus interface. Clin Orthop 1992;274:124-34.
35. Uhthoff HK, Bardos DI, Liskova-Kiar M. The advantages of titanium alloy over stainless steel plates for the in-ternal fixation of fractures. An ex-perimental study in dogs. J Bone Joint Surg. 1981;63-B(3):427-84.
36. Yerby SA, Ehteshami JR, McLain RF.
Loading of pedicle screws within the vertebra. J Biomech. 1997;30:951-4.
