Fatigue Load of Current Tibial Intramedullary Nail Designs: A Simulated Study

Materials and Methods

The goal of this biomechanical study was to evaluate the fatigue strength of several different statically locked tibial nail constructs in an axial compression device that would approximate the normal physiological stress placed on a nail over 16 weeks with full weight bearing. Sixteen weeks was chosen because closed and low-grade open tibial shaft fractures require approximately 16 to 20 weeks for fracture healing to occur. ^1,7 In this model, we assumed that the bone shared none of the load. To achieve reproducible results and assure fatigue failure at the intramedullary device, 2 sections of polyvinyl chloride (PVC) pipe were used to simulate a comminuted tibial fracture model.

Four different nail designs were tested: T2 (Stryker, Mahwah, New Jersey), M/DN (Zimmer, Warsaw, Indiana), Synthes (Paoli, Pennsylvania), and Trigen (Smith & Nephew, Memphis, Tennessee). Each implant was tested with 2 proximal and 2 distal statically locked bolts. In addition, the Trigen nail was also tested with 3 proximal and 3 distally locked bolts. The number, diameter, and position of the bolts in the nail were placed as recommended by the manufacturer. Ten-millimeter-diameter by 360-mm-length intramedullary nails were used in all tests. Interlocking was achieved with 50-mm locking bolts.

The tested designs were:

1. ST-2, a 10-mm Stryker T2 tibial nail with 2 mediolateral 5.0-mm proximal interlocks, and 2 mediolateral 5.0-mm distal interlocks;

2. Z-2, a 10-mm Zimmer M/DN tibial nail with 1 mediolateral and 1 oblique 4.5-mm proximal interlock, and 2 mediolateral 4.5-mm distal interlocks;

3. S-2, a 10-mm Synthes tibial nail with 2 mediolateral 4.9-mm proximal interlocks, and 2 mediolateral 4.9-mm distal interlocks;

4. SN-2, a 10-mm Smith & Nephew Trigen nail with 1 mediolateral and 1 oblique 5.0-mm proximal interlocks, and 2 mediolateral 5.0-mm distal interlocks; and

5. SN-3, a 10-mm Smith & Nephew Trigen nail with 2 oblique and 1 mediolateral 5.0-mm proximal interlocks, and 2 mediolateral with 1 anteroposterior distal interlocks.

The specifications of the interlocking screws are outlined in Table . Approximately 7 constructs were used in each simulation.

Table 1. Major and Minor Bolt Diameter

We followed the testing setup previously described by Brumback et al ⁸ for femoral shaft fractures. Each nail was placed inside 2 pieces of 41-mm-diameter PVC pipe section with 2-mm-thick walls (). The sections were separated by a 10-cm gap to simulate the proximal and distal portions of a segmental defect in a reamed tibia. Interlocking bolts were inserted in the manufacturer’s suggested orientation in the proximal and distal pieces of PVC and through the nail. The proximal interlocks were inserted with the aid of the outrigger specifically designed for the nail, while the distal interlocks were placed in a freehand fashion. Plastic spacers and/or washers that had no attachment to the nail or PVC were used to keep the nail centered in the PVC during the testing. The proximal end of the PVC pipe was potted in methylmethacrylate cement at an angle of 20° to simulate the proximal bend of tibial nails. The distal end was supported by a single transverse bolt acting as a hinge to allow bending of the nail. Once the nail was secured to each piece of the PVC pipe, it was placed in an Instron materials testing system (Model 8500; Instron Corp, Canton, Massachusetts) and loaded in axial compression according to the protocol as outlined below. Thus, the stress from the Instron was transmitted to the proximal piece of PVC, through the proximal screws, down through the tibial nail, through the distal interlocking screws, through the distal piece of PVC, and finally back to the load cell of the Instron.

Figure:. Sections of 2-mm PVC Pipe (A) Were Set to Transfer the Load from the Instron Machine to the Nail Samples (B).

The maximum amplitude of the cyclic loading reached 1200 N, and each construct was tested for 500,000 cycles. To shorten testing time, we cycled implants according to their material properties. Stainless steel returns to a resting state faster than titanium, which allowed for testing of the Z-2 construct at 12 Hz, in contrast to 8 Hz for the others. A modified staircase method loading algorithm was then applied. ⁹ If the construct withstood 500,000 cycles at a certain load, the load was increased by 200 N, and a new construct was tested. Similarly, if a device failed, a new construct was tested at a load 200 N less. Failure was defined by breakage or visible deformity of any of the components, or increased translation of the Instron machine by 5 mm. Once the apparent fatigue limit was identified, a new construct was tested at the load just before failure. The fatigue strength was defined at a load at which the construct passed 3 times just below it and failed twice just above it. This effectively bracketed the fatigue strength.

Results

All constructs failed with breakage or bending of 1 of the locking bolts at the junction of the bolt and the rod. There was no gross damage to the intramedullary rod in any of the tests. Eight of the 13 failures occurred at the distal interlock site. The ST-2 construct failed at 1200 N, passed twice each at 1000 N and passed both times at 800 N, resulting in a fatigue strength between 900 and 1100 N. The Z-2 construct failed at 1400 N, passed and failed twice each at 1200 N, and passed both times at 1000 N, resulting in a fatigue strength between 1100 and 1300 N. The S-2 construct passed 3 times at 1200 N and failed twice at 1400 N, resulting in a fatigue strength between 1200 and 1400 N. The SN-2 construct tolerated the highest loads passing 4 times at 1400 N and failing both times at 1600 N, resulting in a fatigue strength between 1400 and 1600 N. The SN-3 passed once at 1600 N, twice at 1800 N, and failed once at 1800 N, resulting in a fatigue strength between 1600 and 1800 N. The results are summarized in Table .

Table 2. Summary of Test Results

Discussion

Load across the knee has been reported to be from 2 to 4 times body weight during normal walking. ^10,11 The resultant load across the tibia has not been well studied, but joint reaction forces of the knee should be significantly higher than that of the tibial shaft due to decreased muscular and contractile forces that span the knee. Intramedullary nails should therefore bear less load during normal walking than 2 to 4 times body weight. Similarly, loads across the hip joint are estimated to be between 1 to 4 times body weight, yet only peaks at 2 times body weight at the femoral shaft. ¹² Load estimation across the tibial shaft is complicated by the fact that it is difficult to determine the effect muscles spanning the tibia, including the gastrocneumius/soleus complex, have on forces generated across the fracture site. In this study, we determined the fatigue strength of 5 different tibial intramedullary constructs. The constructs had a fatigue strength ranging between 800 and 1600 N, equivalent to 1 to 2.5 times the weight of a 70-kg individual.

All implant failures occurred through the locking bolts and none through the nails, stressing the importance of multiple, large-diameter bolts to increase the strength of tibial nail constructs. Our study showed a trend that the core diameter, or minor diameter, of the interlocking bolt had a direct relationship to the fatigue strength (Table ). For instance, the Stryker T2 nail uses a 5.0-mm locking bolt (major diameter) that has a minor diameter of only 4.0 mm, whereas the 4.9-mm Synthes interlocking bolt has a minor diameter of 4.25 mm. This increase of .25 mm strengthens the construct significantly. This is not unlike the relationship between the bending rigidity of a cylinder and the fourth power of its radius or strength and the third power of its radius. This emphasizes the use of bolts with the largest minor diameter possible to prevent failure. The addition of a third locking bolt increased the fatigue strength in 1 of our nail designs by 13%. This observed strength may be valuable in the large patient, or with fractures that may take longer to heal due to other factors such as open fractures or tobacco use. ¹³

One concern for allowing immediate weight bearing is the risk of hardware failure. The study demonstrates that this risk is relatively small during a 16-week healing period. The constructs were loaded for 500,000 cycles, which simulates an individual walking 3 miles a day, 7 days a week, for 16 weeks. ⁸ The model assumed a worst case scenario and that no load sharing with the bone occurred. In vivo, however, increasing load should be borne by the tibia, and less by the nail as healing occurs. By 16 weeks, most comminuted tibial shaft fractures have progressive radiographic healing. If healing is not evident, then a secondary procedure such as exchange nailing and/or bone grafting is performed within the first 4 months to stimulate fracture healing. ⁷ Therefore, we believe that loading the implants for 500,000 cycles was sufficient.

This study has several limitations. The biomechanical model tests the axial strength of the implants and does not address torsional, shear, or bending moments seen in the tibia. Normal gait produces complex force patterns that cannot be accurately reproduced in the laboratory. The addition of these forces could decrease the fatigue strength of the hardware. Another limitation is the use of PVC to simulate the tibial shaft. We used PVC rather than cadaveric bone to minimize variability between specimens and to simplify the testing setup. In addition, we used the same method previously used by Brumback et al ⁸ to determine the viability of immediate weight bearing of femoral nails.

Despite these limitations, this study demonstrates that current tibial nails may allow patients to immediately bear weight with comminuted tibial shaft fractures. The potential benefits include increased mobility, increased union rate, decreased morbidity, and decreased hospital costs. These important advantages may warrant a clinical investigation into its safety and efficacy.