LONGITUDINAL ELEMENT SIZE EFFECT ON LOAD SHARING, INTERNAL LOADS, ANDFATIGUE LIFE OF TRI-LEVEL SPINAL IMPLANT CONSTRUCTS

The effects of implant stiffness on load sharing and stress shielding, of vertebral column load sharing on implant fatigue life, and of inst rumenting two versus one level adjacent to a comminuted segment on imp lant internal loads were studied. Finite element models of six screw c onstructs with 4.76 mm rod; 6.35 mm rod, and VSP plate tri-level instr umentation of two motion segments (healthy vertebra case and comminute d) and an adjacent healthy motion segment with dimensions representati ve of the human lumbar spine were used. Also a simplified model was de veloped to predict the percent of axial load passing through the colum n, which is a function of ki/kv the ratio of implant axial stiffness t o instrumented vertebral column axial stiffness. For constructs with d imensions typical of the human lumbar spine, 77 to 80% of the axial lo ad was predicted to pass through one or two healthy motion segments wh en instrumented with either 6.35 mm rod or VSP plates, compared to 90% when instrumented with 4.76 mm rods. When instrumenting smaller motio n segments (in dogs) for comparison, 60% of the axial load was predict ed to pass through the column for 4.76 mm rod and 33% for 6.35 mm rod constructs due to increased implant stiffness ki as a result of decrea sed AP and longitudinal construct dimensions, and lower canine motion segment stiffness kv. Single level instrumentation adjacent to a commi nuted segment resisted the entire axial load which produced a max bend ing moment of 11.4 Nm per 445 N axial lumbar load at each correspondin g screw-longitudinal element junction as compared to less than 2 Nm wh en load sharing by a healthy motion segment existed. Thus finite impla nt fatigue life was predicted when instrumenting comminuted lumbar seg ments, with low probability of fatigue when load sharing equivalent to a healthy motion segment existed. Instrumenting two levels adjacent t o a comminuted human lumbar segment was predicted to reduce the flexio n moment at screw-longitudinal element interconnections by 16% when us ing 4.76 mm rods, 32% for 6.35 mm rods, 36% for VSP plates, and 50% wh en ki = kv. These results illustrate the clinical need to create load sharing when possible, to select an implant creating the desired ki/kv ratio depending upon the trade off between potential iatrogenic effec ts of stress shielding relative to the fatigue strength/life of the se lected implant, and to control the patient's type of activity and numb er of cycles until fusion has satisfactorily been achieved.