• Open Access

Evaluation of a Fiber Reinforced Drillable Bone Cement for Screw Augmentation in a Sheep Model—Mechanical Testing


TP Schaer (tpschaer@vet.upenn.edu)


We evaluated the mechanical properties of a novel fiber reinforced calcium phosphate at time zero and after 12 weeks in vivo using a sheep long bone osteotomy model. Time zero data were obtained and compared by pullout testing of 4.5 mm bone screws from bone proper and overdrilled defects of 4.5 and 8 mm diameter. Defects were augmented with: polymethylmethacrylate (PMMA), calcium phosphate, and fiber reinforced calcium phosphate using cadaveric sheep tibiae. Twelve-week data were obtained from explanted tibiae of sheep that underwent unilateral tibial osteotomy surgery repaired with a locking compression plate. The most distal hole was overdrilled to 4.5 or 8 mm diameter, filled with fiber reinforced cement, drilled, tapped and a 4.5 mm screw was placed. Screw holding strength at t= 0 was significantly higher for reinforced when compared to nonreinforced cement, but not different from bone or PMMA in 4.5 mm defects. There was no difference in pullout strength for the 8 mm defect data. After 12 weeks fiber reinforced pullout strength increased by 45% and 8.9% for 4.5 and 8 mm defects, respectively, when compared to t= 0 testing. Fiber reinforced calcium phosphate bone cement can be drilled and tapped to support orthopedic hardware for trauma applications. Clin Trans Sci 2010; Volume 3: 112–115


The requirement for augmentation of bone during orthopedic surgery is a common scenario.1–6 Internal stabilization of fractured or otherwise mechanically destabilized bone often occurs in the presence of compromised bone quality. Placement of hardware into bone of reduced mechanical competency can significantly affect the performance of the fixation hardware, and ultimately the clinical outcome. Additionally, the stripping of screws during insertion is another common clinical reason for the use of bone augmentation materials.6–11

Currently available bone augmentation materials include: bone grafts, polymethylmethacrylate (PMMA), calcium phosphate implants, and calcium phosphate cements.1,5,10,12–23 None of these materials can be drilled and tapped after setting or placement into a defect. This hampers the applicability of many of these materials in the clinical scenario. PMMA is not specifically approved by the FDA for screw augmentation but it is commonly used off-label to reinforce or support screws that are either stripped or placed in very weak bone. PMMA, however, is nonbiodegradable, which may compromise healing, complicate implant removal, act as a nidus for infection and may raise concerns regarding biotoxicity of debris.24,25 Bioactive materials such as calcium phosphate can not currently be drilled and tapped once placed into bone defects or voids but can be advantageous as they are slowly replaced by new bone.

The Norian skeletal repair system (SRS) is a reabsorbable bone cement that has been used extensively in both human and veterinary medicine as an adjunct to fracture stabilization.6,10,18,20–22,26–29 A newly developed fiber reinforced Norian (FRN) biomaterial is a calcium phosphate material that can be utilized to fill a bone void and be subsequently drilled and tapped. This flexibility of application is a considerable advantage to this novel ceramic biomaterial.

Two differently sized overdrilled defects were evaluated in this study in order to evaluate the performance of these biomaterials under two different loading conditions. Using the 4.5 mm overdrilled defect, the 4.5 mm cortex screw will easily glide through the drill hole. Therefore, this situation attempts to mimic what would occur with a stripped cortex screw in a clinical patient. With this smaller size defect, a minimal amount of the biomaterial will remain in the cortex after the material is drilled and tapped. Resultantly only biomaterial in the medullary cavity will interface with the cortex screw. With the 8.0 mm overdrilled defect, the greater diameter of implant material in the cortical drill holes will be able to withstand drilling and tapping, and therefore create a mantle of cement that completely envelopes the cortex screw placed in the defect. In this scenario, the 8.0 mm contructs mimic filling of a bony void to anchor a critical screw in a clinical patient.

The objective of this research was to evaluate the performance of FRN to support orthopedic hardware in ovine cadaveric tibiae at time zero (t= 0) and after 12 weeks in vivo (t= 12) using a unilateral tibial osteotomy model in sheep. A “stripped screw” scenario overdrilled to 4.5 mm diameter and a “bone void” scenario overdrilled to 8 mm diameter were utilized for both arms of the study. Our hypotheses were: (1) screws placed in FRN would show comparable pullout strength when compared to PMMA or screws placed in intact cortical bone and (2) FRN after t= 12 in vivo would have similar pullout strength when compared to t= 0 testing.

Materials and Methods

This study was approved by our Institutional Animal Care and Use Committee.

Time = 0 experiments

Twenty-four (n= 24) tibiae from 3- to 5-year-old female Polypay sheep were harvested intact and stored at −20°C until further testing. Tibiae were thawed for 24 hours in a balanced electrolyte solution bath maintained at 38°C before testing. Twelve tibiae were randomly allocated to either 4.5 mm or 8 mm overdrilled defect groups. All screws used in this component of the study were 10 mm longer than required, protruding above the bone surface to allow for mechanical testing. Three sites were identified for each tibia, starting 3 cm proximal to the most distal aspect of the medial malleolus and every 3 cm proximally. For all tibiae a 4.5 mm cortical bone screw was placed using standard technique of The Association for the Study of Internal Fixation (AO-ASIF) at the most proximal site. The remaining two sites were overdrilled to the required diameter using sequentially larger drill bits. Each tibia then randomly received PMMA and either conventional Norian or FRN to augment the remaining two defects.

PMMA, Norian, and FRN were prepared as per manufacturer's guidelines and 3 mL injected into the respective defects. Consistent defect filling was confirmed via fluoroscopy. For PMMA and Norian a 4.5 mm cortical bone screw was inserted prior to setting. For all materials handling ceased 2 minutes after mixing. FRN was incubated in a water bath at 38°C for 20 minutes and was subsequently drilled and tapped to accept a 4.5 mm cortical bone screw. All tibiae were placed in a water bath for 24 hours prior to mechanical testing to allow for complete curing of all materials.

Time = 12 weeks experiments

Eight 3- to 5-year-old female Polypay sheep were enrolled in this component of the study. All sheep were weighed and randomly allocated into either the 4.5 mm or 8 mm overdrilled defect group. All sheep received perioperative antimicrobials and analgesics. Sheep were placed into left lateral recumbency under general anesthesia and a medial subperiosteal approach to the tibia was made. A 7 hole 4.5 mm narrow Locking Compression Plate (LCP) (Synthes, West Chester, PA, USA) was positioned on the medial aspect of the intact tibia and the 5 proximal holes, skipping hole #4, were drilled and tapped to accept 4.0 mm locking head screws (LHS). A pilot hole of 4.5 mm diameter at the most distal screw (7th) was then made, followed by removal of the plate. The tibia was then osteotomized using an oscillating bone saw at the 4th plate hole. By random allocation half (n = 4) of the distally located (7th) treatment holes were overdrilled using progressively larger drill bits to a final diameter of 8 mm. Then holes were meticulously cleaned, dried and surrounding marrow contents were removed from the drill holes. The treatment defect was subsequently filled with FRN as per the cadaveric experimental design of the study with the exception that the biomaterial was allowed to cure under constant saline irrigation at 38°C for 20 minutes intraoperatively. The tibial osteotomy was reduced leaving a 0.6 mm gap and the plate was secured using 5 LHS inserted into the previously prepared holes. The Norian augmented site was then drilled and tapped using standard technique to accept a 4.5 mm cortex screw which was tightened to 2.5 Nm using a torque limiting screwdriver. After hardware placement, the surgical sites were irrigated and sof tissues were closed in layers using nonabsorbable monofilament suture material followed by stainless steel staples and an aseptic skin dressing. Hardware placement was documented via fluoroscopy at the end of the procedure. Sheep were recovered and returned to a 4 × 4 m stall for 7 days before being placed in a larger 15 × 5 m stall for the remaining 11 weeks of the study. All sheep were monitored twice daily for the duration of the study and no attempt was made to limit their activity. At 12 weeks postoperatively the sheep were euthanized and all treated tibiae harvested. All softtissues and hardware were removed. A new cortical and LHS, each 10 mm longer than previously, were reinserted into the treatment (7th) and control (1st) holes, respectively, to allow for mechanical testing.

Mechanical testing

Testing was conducted in accordance with the American Society for Testing and Materials (ASTM) standard F-54330,31 for metallic bone screws using a servo-hydraulic mechanical testing apparatus (Instron 8500, Canton, MA, USA). Data acquisition was by a 32-bit data acquisition and signal interface between a computer and the load cell. A 4,000 N load cell was used to measure loads during pullout tests. The specimen were mounted in the fixture and kept moist throughout testing. Axial pullout strength was determined as the maximum force (N) reached during the test run. Stiffness of the constructs (N/mm) was calculated from the linear portion of the load-displacement curves prior to pullout of the screws.

Statistical analysis

Pullout strength of each specimen was reported as mean ± standard deviation. Pullout loads in cadaveric tibiae for augmented screws compared to their normal parent bone was calculated as a percentage as previously reported by Grifton et al.10 Similarly, FRN after 12 weeks in vivo was compared to LHS loads in vivo (t= 12 weeks) using: Augmentation (%) =[Pullout load (augmented/t= 12)/Pullout load (parent bone/t= 0)]× 100.

Strength and stiffness of each pullout-testing specimen were separately analyzed for statistical significance using ANOVA with treatments as repeated measures. Student t -tests were used to compare between t= 0 and t= 12 weeks for each of the FRN sized defects. A statistical significance of p < 0.05 was used throughout unless otherwise stated.


All pullout tests had a typical load-deformity behavior.

Time = 0 pullout testing

Seventy-two test samples were prepared and 69 tests were completed and the data recorded and analyzed (Figures 1, 2 and Table 1). Three samples (2 PMMA and 1 bone specimen) were not tested due to preparation errors. For the 4.5 mm augmented defects, FRN (1,478 ± 699 N) had similar loads at failure compared to bone (2,206 ± 409 N), and PMMA (2,410 ± 960 N) (p > 0.05) and greater load to failure than Norian (239 ± 116 N) (p= 0.01). For the 8 mm defects there was no signif cant difference between FRN (1,200 ± 405 N), Norian (912 ± 271 N), PMMA (2,317 ± 643 N), or bone (2,206 ± 409 N) (p > 0.05). There was an increase in the pullout strength of Norian from the 4.5 mm defect (239 ± 116 N) to the 8 mm defect (912 ± 271 N) (p < 0.01) but no difference for FRN or PMMA (p > 0.05). There was no difference in the material stiffness between any of the groups (Figure 3).

Figure 1.

Cadaveric pullout testing peak load for parent bone and 4.5 and 8 mm defects augmented with Norian, FRN, and PMMA (mean ± SD). Letters denote significant differences within the 4.5 mm defects and stars denote significant differences between 4.5 and 8 mm defects.

Figure 2.

Cadaveric pullout stiffness (N/mm) for parent bone and 4.5 and 8 mm defects augmented with Norian, FRN, and PMMA (mean ± SD).

Table 1.  Summary mechanical testing data for cadaveric tibiae and percentage augmentation compared to parent bone
MaterialDefect size (mm)NPeak load (N)Stiffness (N/mm)% AUG
Bonen/a242,206 ± 4091,385 ± 447n/a
Norian4.56239 ± 1161,150 ± 46910.9%
FRN4.561,478 ± 6991,318 ± 29867.0%
PMMA4.5122,410 ± 9601,355 ± 521109.3%
Norian8.06912 ± 2711,351 ± 48741.3%
FRN8.061,200 ± 4051,520 ± 64454.4%
PMMA8.092,317 ± 6431,481 ± 510105.0%
Figure 3.

Twelve-week in vivo pullout testing of 4.5 mm cortical screws from FRN augmented 4.5 and 8 mm defects. Compared to 4.0 mm LHS pulled from bone.

Time = 12 weeks pullout testing

All sheep underwent uncomplicated surgery, had uneventful recoveries from general anesthesia, and were immediately partially weight bearing. They all were fully weight bearing within 72 hours postoperative and for the remainder of the study. At time of sacrifice, 16 samples were prepared and 15 tests completed and analyzed (Figures 3, 4, and Table 2). One LHS pullout test was not recorded due to technical error during data acquisition. After 12 weeks in vivo, there was no difference between pullout strength and stiffness in either of the LHS or FRN screw pullout data sets. There also was no difference between pullout strength and stiffness for the 4.5 mm or 8 mm defects.

Figure 4.

Twelve-week in vivo stiffness (N/mm) of 4.5 mm cortical screws from FRN augmented 4.5 and 8 mm defects. Compared to 4.0 mm LHS pulled from bone.

Table 2.  Summary mechanical testing data after 12 weeks in vivo and percentage augmentation compared to 4.0 mm LHS
MaterialScrew size (mm)nPeak loadStiffness (N/mm)% AUG
LHS4.072,179 ± 9121,027 ± 280n/a
FRN4.542,144 ± 4941,596 ± 27498.4%
FRN8.041,309 ± 316952 ± 33060.1%

After 12 weeks in vivo, FRN was 45% stronger when compared to the t= 0 cadaveric pullout testing (p= 0.14) for the 4.5 mm defects (2144 ± 494 N and 1478 ± 699 N, respectively) and 9% stronger for the 8 mm defects (1309 ± 316 N and 1200 ± 405 N, respectively [p= 0.66]). FRN after 12 weeks in vivo was 21% stiffer than the t= 0 pullout testing (p= 0.14) for the 4.5 mm defects and 39% less stif for the 8 mm defects (p= 0.12) (Figures 3, 4, and Table 2)


This study examined FRN as a bone void filler and its ability to support metallic hardware without adequate bony purchase in a long bone osteotomy large animal model. Norian and PMMA augmented defects and cortical bone screws inserted in normal bone were evaluated for comparison. FRN was shown to have holding strength similar to PMMA or bone augmentation for small defects and was superior to Norian in this application. For larger defects FRN pullout was similar to Norian, PMMA, and bone. This study demonstrated FRN to be capable of supporting orthopedic hardware in a loaded long bone environment.

FRN was easily injected into defects during this study and once set allowed for uncomplicated drilling, tapping, and screw insertion. Screws inserted into FRN could be tightened in vivo to 2.5 N. In comparison, Norian and PMMA screws had to be simply held in place until either material fully polymerized without allowing tightening of the screw. The ability to tighten the cortical screws may provide increased compression of the plate to the bone interface. For both Norian and PMMA screws must be inserted into the material prior to setting. This requires a degree of estimation as to the volume required to augment defects as overflow can occur. FRN can be placed and sculpted for up to two minutes prior to allowing for setting. This increases accuracy and versatility during cement delivery. Furthermore, multiple defects can be augmented with FRN at one time without concern about placing screws before the material hardens. In comparison, PMMA and Norian can realistically only augment one defect at a time.

The pullout strength of Norian increased from the 4.5 mm defect to the 8 mm defect. This was likely due to a larger interface between the Norian and the screw thread. For the 8 mm defects the presence of a cement mantle at both cortices significantly increased the pullout strength for this material. For the FRN material there was a slight but not signif cant decrease in the pullout strength from the 8 mm defect to the 4.5 mm defect. The reinforcing fibers in the FRN reduced the stripping of the screws that had occurred in the nonreinforced Norian. This is important in this animal model as it demonstrates that for a stripped screw scenario the use of FRN will effect a repair that is superior to Norian and similar to PMMA and bone with regard to strength.

A large bone defect was simulated by an 8 mm defect in this study. Augmentation with Norian, FRN, and PMMA along with bone had similar pullout strengths comparatively. These defects were easily augmented intraoperatively and allowed for screws to be placed in these large voids.

The 12-week in vivo pullout strength for FRN was higher compared to the time zero testing. A statistical significance could not be determined between the time zero and 12-week data due to the relative small number of sheep enrolled in the in vivo component of the study. The pullout strength and stiffness of the 4.5 mm FRN augmented defects in vivo increased compared to the cadaveric tests. This is likely due to new bone formation and integration of the FRN. This new bone formation was evident in μ-CT images prior to mechanical testing (data not shown). The increased strength of screw pullout after 12 weeks in vivo allows for assurance in regard to the construct's long-term stability.

We conclude that FRN used in this manner is able to be drilled and tapped after augmenting a stripped screw or large bone void. This augmentation yields superior screw holding strength to Norian and similar to PMMA for a stripped screw scenario. For larger bone defects, FRN augmentation results in similar screw holding strength when compared to Norian and PMMA. This study suggests that FRN may have some benefit where bone or hardware augmentation is required. However, additional studies and clinical data will be required to further investigate this currently unapproved use of FRN.

Manufacturer Disclaimer

This is an experimental, non-clinical study to evaluate an unapproved use of Norian products. These materials are not indicated for screw augmentation or load bearing applications.


We acknowledge the support of Synthes (USA) for this preclinical trial. This study was in part f nanced by Synthes Inc., West Chester, PA, USA. The research was carried out at The Comparative Orthopedic Research Laboratory, University of Pennsylvania, New Bolton Center.