A FOUR AND ONE HALF YEAR FOLLOW-UP OF A BIORESORBABLE BONE CEMENT IN A CANINE METAPHYSEAL MODEL

Participants: E.P. Frankenburg, P.V. Patil, C.M. DeBano, K.A. Sweet, J.A. Baker, M.R. Moalli, B.T. Nolan, D.C. Kayner, M.W. Stock, R.L. Taylor, S.A. Goldstein, T.W. Bauer*, R.D. Poser+

*The Cleveland Clinic Foundation, Cleveland, OH

+Norian Corporation, Cupertino, CA

Keywords: bone cement, fracture healing, allograft, calcium phosphate ceramic

Introduction

Previous studies have shown that bioresorbable calcium phosphate cements may be viable substitutes for bone or enhance fracture fixation constructs. A calcium phosphate cement produced by Norian Inc. (FractureGrout, SRS) provided sufficient mechanical integrity up to 16 weeks post-op when used in bounded metaphyseal defects. This cement was shown to have high compressive strength and be strongly osteoconductive. Histologic data up to 78 weeks post-op reveal that the cement continues to resorb and be replaced by bone. Although the histologic data at 78 weeks was very promising, significant amounts of cement were still present at that time. The purpose of this study was to mechanically and histologically evaluate the cement in two previously established metaphyseal canine models after long term implantation. Specifically, this study investigated the response to the calcium phosphate cement 4.5 years after implantation.

Materials and Methods

Four adult male mongrel dogs (approximately 25 kgs each) were obtained to study the histologic and mechanical properties of the cement over time. All procedures performed on the animals were approved and performed in accordance with the University['s Committee on Use and Care of Animals. A model previously described which resulted in a 3.5 mm defect in the proximal tibia and two 8 mm cylindrical defects in the contralateral distal femur of each dog was followed. Each defect was filled with Norian FractureGrout. The dogs were monitored clinically, as well as with planar radiographs and computed tomography taken every six months.

At approximately 4.5 years post-op, the animals were sacrificed and the operated limbs were mechanically tested with a MTS servo-hydraulic testing machine. Both the experimental and contralateral tibias were tested in torsion until failure at 100°/second. The femoral defects were located and removed from the metaphysis such that a 1.5 mm sheath of bone encompassed the original defect. The specimens were then cut to a constant height of 6 mm. In the same manner, control specimens were obtained from similar regions in the contralateral femurs. All femoral specimens were mechanically tested in compression at .015"/sec. Half of the femoral specimens were tested to failure, while the other half were tested pre-yield. Stiffness values from all specimens were determined, as were failure data where applicable (yield load and energy to yield in the femoral specimens, maximum torsional load, displacement at maximum load, and energy to failure in the tibias).

After mechanical testing, the proximal tibias were fixed in 70% EtOH, infiltrated with PMMA, sectioned into 500 micron thick slices along the coronal plane, and then either submitted for quantitative analysis using a scanning electron microscope or ground and polished to a thickness of less than 50 µm and viewed with a light microscope. The quantitative analysis using the SEM images was performed using NIH Image 1.61 PPC running on an Apple Power Macintosh. After a region of interest was defined (the original defect site), histograms were generated which revealed the number of pixels per area over a gray scale. Finally, after thresholding for cement and bone, the cement/bone ratio could be determined. A similar process was followed for the femoral specimens.

Results

Clinically, all animals demonstrated excellent health, ambulation, and mobility on physical exam. Mechanical testing of the tibias in torsion resulted in the failure of each experimental and control tibia through the central diaphysis. Only small differences in values were detected between the experimental and control tibias for all of the dogs. Femoral control tissue from the 4.5 year dogs revealed similar results to those control tissues at time zero. However, at 4.5 years, the properties of the cylinders filled with FractureGrout were not different from those of the control specimens at 4.5 years. This result is different from the 16 week data found in an earlier study, in which the mean properties of the cylinders filled with FractureGrout were higher than the average of allograft control specimens.

Scanning electron microscopy and light microscopy both revealed that the initial defect region could usually be defined. However, there was a marked decrease in the amount of cement compared to the previous tibial specimens, obtained at 78 weeks post-op. It is also clear that the cement seems to be resorbed and replaced by bone more quickly in the anterior regions of the defects compared to the posterior regions. All cortical regions appeared to be normal and nearly completely replaced by bone. At higher magnifications, it was evident that where FractureGrout does exist, it is often completely surrounded by trabecular bone. Scanning electron micrographs of the 4.5 year specimens differed from animal to animal. A SEM image of one of the experimental tibias shows that nearly all of the cement has been resorbed and replaced by bone. The new bone appears to be normal, and except for only very small amounts of cement remaining, no former defect was evident.

Femoral SEM images showed similar patterns as the tibial specimens, with increased cement resorbed and replaced by bone in h 4.5 year post-op dog specimens. The defects were usually well defines, although they had very scalloped edges, and often bone infiltrated into the inner most sections of the cement. Again, the femoral specimens from one dog had an increased amount of bone resorption compared to the other specimens, although the resorption rate of the cement in this defect was slower than that in its contralateral tibia.

The quantitative analysis of the SEM images supported the qualitative analysis. Within the defined areas, the femurs had a relatively greater amount of cement than the tibias, almost twice as much. IT was also interesting to note that the quantitative data of one animal had a significantly smaller amount of cement remaining in both defects compared to the other animals.

As in the earlier studies, the use of Norian FractureGrout to augment compromised trabecular bone proved to be a viable solution for use in a bounded canine metaphyseal defect over a 4.5 year period. The rates at which the cement resorbed was site and animal dependent. One of four animals showed nearly complete resorption in its tibial defect, although the remaining animals had some residual cement in all defects. In addition, FractureGrout continued to show mechanical competence over 4.5 years, while histologic features of resorption and cell-mediated replacement by bone took place.

Progress

All animals have undergone the operative procedure and have been sacrificed at 4.5 years post-op. Tibia and femoral plug testing has been completed, as have histologic and SEM analyses of all specimens.