Failure of Proximal Constructs in Scoliosis: The Role of Metal Memory.
Purpose: The purpose of this study is to determine the difference in metal memory between 6 mm stainless steel rods and 6 mm titanium metal. Background: Proximal rod screw construct pull-out can be a severe problem when dealing with spinal instrumentation, especially with added forces from surgical deformity such as thoracic kyphosis are present. Titanium constructs are advocated for the ability to repeatedly view the soft tissues and neural tissue post operatively with MRI. Although the bone-metal interface with titanium has proven superior over that of stainless steel, proximal screw rod failure may be enhanced with titanium rods due to the metal's tendency to revert to its original shape (metal memory). Postoperative surgical balance may also be lost when using titanium rods as an implant versus stainless steel for the same reason.
Methods: The following 6 mm diameter rods supplied by Synthes (Paoli, PA) were used in this study: 300 mm SST (298.112), 100 mm SST, (298.104), 300 mm Ti (498.112), 100 mm Ti (498.104), and 85 mm pre-bent Ti (498.143). A three-point rod bender was used to contour the straight rods in as uniform a curve as possible, using a template with a radius of curvature of 30 cm. The rods were bent at room temperature, then placed in an incubator at temperature of 37±2º C and left for six months. At the time the rods were bent and every 2 weeks thereafter, the rods were photographed at high resolution (2560 x 1920 pixels) over graph paper with a Digital Camera (Nikon Coolpix 5000). The digital images were then processed using Jandel Sigma Scan, by outlining the covex side border of each rod, and digitizing those pints into an x-y graph. Then a best-fit regression polynomial was used to calculate the average curvature, defined as the change in angle from end to end divided by the rod length. After 36 weeks, the curvature of each rod was normalized against the initial curvature in order to account for variances in initial curvature, plotted against time, with a linear regression performed to assess changes in curvature in terms of percent change per year.
Results: Changes in both SST rods and the pre-bent Ti rod were within measurement error (0.52% increase for long SST, 0.26% decrease for short SST, 0.52% for Pre-bent Ti). In contrast, both manually bent Ti rods changed markedly (decreases of 6.76% and 5.2% for long and short Ti rods respectively).
Discussion: Titanium rods used for scoliosis constructs are of a high grade of titanium alloy combined of 0.5% nitrogen, .10% carbon, .01255% hydrogen, .5% iron, and .4% oxygen. This allows the rods to be contoured. Previous in-vitro testing has shown that heat conduction in titanium rods could possibly be used to correct scoliosis when a pre-bent rod has been placed. The temperatures needed for this deformation to occur and for the rods to obtain the original length makes the application unpractical for use in scoliosis correction. The metals have to be heated at transformation temperature. This study shows that continuous heat conduction equal to that of the body temperature may contribute to a considerable loss of curvature in the titanium metal, which would be mainly due to the memory in the titanium. The ideal implant should be strong and stiff, but most of all, retain its contoured shape. No compression or distraction forces were added to these rods and yet a significant change in the curvature was seen. The author postulates that this inherent ability of the metal of titanium might contribute to construct failure, especially as compression forces are added by underlying pathology. Although titanium has proven to have excellent compatibility with bone, this study points to an important fact that titanium should probably not be used in correcting scoliosis unless a pre-bent tempered rod could be used. Over time, this affect of titanium might contribute considerably to the loss of correction and sagittal imbalance in spine cases.