Material and methods

Nine fresh-frozen, unembalmed, dentate, human cadaver mandibles donated for anatomic research purposes were selected for implant placement and testing. The mandibles came from 3 female and 6 male white donors between 48 and 81 years of age. No donors were known to have suffered from primary bone diseases.

For mounting purposes, impressions were taken of the test mandibles, and models were poured. Up to 3 custom acrylic (methylmethacrylate) bases were fabricated for each test mandible. To compensate for lingual surface variations, the acrylic base was adapted to the cortical surface of the entire lingual corpus and inferior ramus. The opposite side of the acrylic base was ground to a level plane. The acrylic base allowed rigid fixation of the mandible to the test equipment while maintaining the test site surface orientation perpendicular to the line of force in the pull-out tests and parallel to the line of force in the shear tests. Then the custom fit to each lingual surface allowed for uniform distribution of reaction forces during the application of test loads. It prevented the development of internal stress/strain induced by flexure secondary to fixation of the mandible to the test apparatus. An anterior hole near the mental foramen and a posterior hole in the inferior ramus were drilled through the bone sample and the custom acrylic support base to allow rigid fixation of the mandible to the test equipment.

The tensile and shear testing was completed with a universal testing machine (model DDL 200RT, TestResources, Shakopee, Minn), outfitted with a 112-pound calibrated load cell used in tensile mode. Adjustable x-axis and y-axis sliding tables (Sherline Products, Vista, Calif) were mounted to the base of the testing machine. A 360° rotational table (Sherline Products) was mounted on top of the sliding tables horizontally for pull-out tests and vertically for shear tests. A mounting plate was fabricated from a 0.5 × 3 × 9-in aluminum bar and attached to the 360° rotational table. Bolts were used to secure the test mandible and its acrylic base to the mounting plate.

A custom implant holder was attached to the load cell of the testing machine. It was designed to allow for rotational and x-y freedom during attachment of the test implants to the tensile machine. The bottom part of the implant holder was fabricated from a 6.4-mm diameter aluminum cylinder with a 2-mm key way cut in its base to a depth of 4 mm. This space gave the head of the implant 0.4 mm of horizontal and 0.5 mm of vertical clearance. A 1.0-mm hole was drilled perpendicular to the key way at a point 2.0 mm from the base. A 0.028-in piece of stainless steel orthodontic wire was run through the implant holder and the hole in the head of the implant to secure the implant to the implant holder.

MSIs (IMTEC, Ardmore, Okla), 6 mm long and 1.8 mm in diameter, were placed in the buccal cortex of the mandible, at angles of either 90° or 45° to the buccal surface. They were tested to failure by using pull-out and shear tests directed along the axes of maximum and minimum bone stiffness. The design test matrix consisted of 9 subgroups of 10 implants each ( Fig 1 ). Three subgroups were pull-out tests. The remaining 6 were shear tests. Three shear tests were sheared parallel to the maximum axis of bone stiffness, and 3 were sheared parallel to the minimum axis of bone stiffness, which is perpendicular to the axis of maximum bone stiffness.

Test matrix with 9 subgroups.

The pull-out tests consisted of MSIs in 1 of 3 orientations: placed orthogonally to the buccal surface, angled at 45° along the maximum axis of bone stiffness, and angled at 45° along the minimum axis of bone stiffness. All 3 implant orientations were tested to failure in pull-out tests in which the line of force was at 90° to the bone surface.

In both sets of the shear tests, along the axis of either maximum or minimum bone stiffness, the MSIs were in 1 of 3 positions: orthogonal to the buccal surface, angled at 45° toward the same direction as the line of shear force, and angled at 45° opposite the line of shear force. All MSI orientations were tested to failure in a shear test in which the line of shear force was along the bone surface and in a direction parallel to the maximum or minimum axis of bone stiffness. The maximum force at failure in the pull-out and shear tests and the cortical thickness were recorded for each sample.

Bone-implant test sites were selected on the buccal cortex of the corpus of the mandibles so that no proximate defects could adversely affect the test results. A 1.1-mm pilot drill was used to prepare the MSI site; the pilot holes were drilled only through the cortex. A guide was fabricated and used with the pilot drill to ensure the proper orientation for the MSIs oriented at 45° to the bone surface. After the pilot holes were drilled, the MSIs were placed according to the manufacturer’s specifications. Once the MSI was placed, the test mandible was set in the custom acrylic base and fixed to the mounting plate anteriorly and posteriorly. The mandible was then oriented to obtain the correct position of the test site relative to the actuator, as previously described.

With the mandible secured to the test holder and appropriately positioned, a digital photograph of the implant was taken (DXC-390P digital video, Sony, New York, NY; macro lens, Konica Minolta, Wayne, NJ; and imaging software, SCION, Frederick, Md). The 0.028-in stainless steel orthodontic wire was inserted through the holder and the MSI, and the load screw was activated to eliminate any play between the MSI and the load cell. The tensile machine position was then zeroed, and the test began with the machine moving at a rate of 2 mm per minute. Data collected digitally included time, load, and position during each test. At regular intervals during the testing, digital photographs were recorded and identified with load or position information to document the changes of the implant-bone interface during testing. At failure of the implant-bone interface, the test was discontinued, the test data were saved, and a final digital image of the postfailure implant-bone interface was recorded. For each test, maximum force at failure was recorded.

During testing, the mandibles were kept moist. All tests were conducted with the mandibles at room temperature (22°C). The mandibles were stored by wrapping them in wet paper towels, placing them in sealable plastic bags, and freezing them at –5°C. Freezing does not adversely affect the elastic properties of bone measured ultrasonically, although it might have some minor effects on mechanically determined material properties.

Results

The pull-out tests of the MSIs aligned at 90° to the cortical surface had a significantly higher maximum force at failure (342 ± 80.9 N; P <0.001) (mean ± standard deviation) than all other test groups ( Table , Fig 2 ). The loading curve for all 3 pull-out test groups was largely linear until immediately before failure ( Fig 3 , A ).

Maximum force at failure with 95% CI.

Representative loading curves of the 3 modes of implant failure: A, typical pull-out test or shear of 45° implants angled toward the shear force; B, typical shear test of 45° implants angled away from the shear force; C, typical shear test of 90° implants.

At failure, 1 of 2 bone-implant interface configurations was generally seen. In most cases, the bone remained in intimate contact with the MSI at failure. For MSI failures with intact bone, the 90° MSIs often had an elliptical surface outline with a conical cross-sectional shape. Generally, the greatest dimension of bone was in the direction of maximum bone stiffness ( Fig 4 , A ). In other cases, the MSI separated from the bone, with cortical splinters of bone projecting upward in the direction of the pull-out test.

A, Postfailure bone-implant section of a 90° pull-out test; B, postfailure bone-implant section of 45° pull-out test; C, buccal surface response when implant contacts lingual cortex.

The pull-out tests of the MSIs aligned at 45° to the bone surface—oriented in the direction of either the minimum or maximum bone stiffness—had a bone-failure wedge in the 135° angle between the MSI and the surface of the bone ( Fig 4 , B ). Similar to the pull-out tests of the MSIs aligned at 90° to the cortical surface, the widths of the bone fragments were smaller for the MSIs oriented along the direction of maximum bone stiffness than for the MSIs oriented along the direction of minimum bone stiffness.

The maximum forces at failure in tests of MSIs angled 45° toward the line of shear force in the direction of maximum bone stiffness (253 ± 74.1 N; P <0.001) and those in the direction of minimum bone stiffness (264 ± 21.0 N; P <0.001) were significantly higher than the other 4 test groups ( Table ).

The 45° MSIs opposing the shear line of force most often failed in a bimodal fashion ( Fig 3 , B ). The loading curve was initially similar to the other test subgroups until primary failure, which occurred as the MSI started to rotate from the original 45° placement. The loading curve generally flattened while the MSI “reoriented” to approximately 90° relative to the cortical surface; as the MSI continued to rotate from 90° to 45° in the same direction as the applied force, the loading curve usually increased until final failure of the MSI-bone interface.

Shear tests of the MSIs aligned 90° to the cortical surface also demonstrated a nonlinear loading curve ( Fig 3 , C ). Similar to the MSIs aligned at 45°and opposing the shear line of force, the loading curve fell after an initial linear loading response. The slope of the increasing load response decreased but remained positive until failure. Unlike the MSIs aligned 45° to the cortical surface and opposing the shear line of force, a single point of primary failure was not clearly identifiable.