2.1 Design

Six distinct models (Modification #1–6) were generated using a CAD program (Inventor Professional 2019, Autodesk©, San Rafael, California, US) by adapting a pre-existing short stem system (AIDA®, Implantcast, Buxtehude, Germany). These adaptations, aimed at enhancing elastic deformability, were modelled after the original AIDA® design, which is notably more flexible than other short stem systems.¹³ Refer to Fig. 1 for a visual representation of the six foundational models.

Close

Fig. 1 Overview of the different designs.

Export to PPT

These are all made of Ti6AL4V (Ti64) titanium alloy and have a CT length of ≤120 mm which is the length of the femoral head center (C) to the distal tip of the prosthesis (T). According to the CT length, they were loaded with 1200 N according to ISO 7206-4¹⁴ and also with 2300 N.¹⁷ After assessing the deformation behavior at five specified measurement points (Fig. 2), the unsuitable models were to be sorted out. Too large a deformation (>5 mm) or too small a deformation compared to the original model were considered unsuitable.

Close

Fig. 2 Detailed sagittal view of the measurement points used to calculate point C.

Export to PPT

2.2 Displacement point C

To ensure comparability of the deformation of the individual stem models, a total of five measuring points are defined on their surface. The deformation at these points is measured, with particular attention to the deformation at point C. If a model deforms more than 5 mm, it is discarded because it cannot meet the minimum specifications of ISO 7206-4.¹⁴ Point C describes the spatial center of the coordinates of the measuring points 1, 2 and 3. The measuring point (2; 3) is a calculated coordinate, which describes the center of the measuring points 2 and 3. The measuring points were defined on the sagittal plane of the prosthesis stem.

2.3 FE analysis

The stems were aligned according to ISO 7206-4 and frictionally anchored in an embedding mass. The embedding mass has the properties required by the standard in terms of Young's modulus.¹⁴ The materials used are listed in Table 1. Autodesk© Inventor Professional 2019 software (San Rafael, California, US) was also used for the FE analysis. The femoral head ceramics are generally alumina (Al2O3) ceramics. Since this was not available in the material library of the software, silicon nitride was used. This ceramic has nearly identical mechanical properties in terms of Young's modulus and Poisson's ratio as aluminum oxide (Al2O3) that is used in ceramic heads for THA.¹⁸

A mesh with an average element size of 0.10 mm was defined on the short stem model with a pitch factor of 1.00 at a rotation angle of 15°. The model has a minimum number of 100,000 nodes. Local mesh refinements were omitted because stress peaks in particularly delicate areas were not the object of observation. The contact definition between femoral head and femoral stem is assumed to be rigid and frictionless. Forces arising due to “incorrect seating” of the femoral head were to be excluded, as these were not to be investigated in the measurements. The mechanism for force application was defined in such a way that the degrees of freedom of movement correspond to a cylindrical joint and it can only move in the z-direction (vertically). The contact definition of the stem is defined as a friction interference fit. This means that the stem has an interference of 340 μm compared to the embedding mass. This value is derived from the anchoring method used in the operation of the role model short stem.¹⁶

2.4 Load definition according to ISO 7206-4

The loads according to ISO 7206-4¹⁴ are CT length dependent. ISO 7206-4 divides hip stems into three classes depending on the total length of the stem. The Standard divides hip stems in <120 mm total length, 120–250 mm total length and >250 mm total length. For the classification, the length between the ball center (point C) and the distal tip (point T) is determined. Depending on the determined length, the embedment height or the free length D as well as the maximum force to be applied must be determined. The loads to be applied in this test correspond to 1200 N for a CT length <120 mm. Since this value is only to be understood as a minimum value according to the standard, the stem models are also tested with the next higher load of 2300 N according to Bergmann et al.¹⁷

2.5 Loading protocol

The FE model should be critically examined and finally validated in vitro. In a first step, the steps and results were critically reviewed and evaluated using the guidelines from the published checklist for verification and validation of finite element analysis in orthopedic and trauma.¹⁹ In a second step, in vitro tests were carried out to verify whether the results of the FE simulation were in agreement with them.

For this purpose, the stem was anchored in a container with epoxy resin (Rencast FC52 (Huntsman International LLC, Salt Lake City, USA) as in the simulation model and in accordance with the applicable standard and subjected to loads of 1200 N and 2300 N in compression. The hip stem (5) with attached 36 mm ceramic femoral head (4) was cast in epoxy resin in a container (6) as described in Fig. 3. The forces of 1200 N and 2300 N are applied to the hip stem by a universal testing machine (Inspekt Table Blue, Hegewald &Peschke, Nossen, Germany) (1) via a concave punch with a larger radius to prevent the punch from slipping off the head under loading but at the same time to realize a more punctual force application as in the FEA simulation and thus to ensure comparability (2). In order not to introduce additional stresses into the system, the punch was connected to the testing machine via a free bearing (2). This allows the punch to move in the frontal and sagittal planes (Fig. 3).

Close

Fig. 3 Illustration of the loading protocol using the H-beam as an example.

Export to PPT

The stem was then loaded in four steps. First, a pre-load of 20 N was applied and maintained for 10s. This was followed by a gradual increase to 1200 N load and 2300 N load, with a subsequent return to a 10 N holding load. All positions were maintained for 10s each. The displacement at point C was calculated from the x and y-displacement of the floating bearing, which was recorded using inductive probes (Millimar P2010 FA, Mahr GmbH, Göttingen). The z-displacement of point C, was recorded by the travel distance of the punch within the universal testing machine used.

The preload of 20 N was chosen as a reproducible starting point for the measurement. The in vitro measurements for the role model short stem were performed six times to compare the results with the FE analysis of the same stem.

3.1 Validation of the FE model

Of the six measurements performed for the role model short stem, five could be evaluated. One measurement could not be included in the evaluation due to a fracture of the embedding mass, as the ISO 7206-4 standard specifies that a test with failure of the embedding mass cannot be evaluated (see Table 2).

On average, there is a total displacement of the system of 1118.6 ± 59.1 μm under load (see Table 3).

On average, there is a total displacement of the system of 2175.6 ± 163.4 μm under load. The FE analysis of the short stem showed a displacement of point C of 1098 μm under 1200 N load, respectively 2027 μm under 2300 N load.

3.2 Derivation of the design

The initial analyses showed that all the basic modifications had better elastic deformation behavior than the role model short stem. It was also noticeable that none of the stems had a deformation >5 mm. With about 190% of the deformation of the role model short stem, modification #3 exhibited the best deformation behavior. On this basis, the basic principle of the “H-Beam” was worked out and this was improved via several optimization procedures and FE analyses. Taking into account implantability and good elastic deformability, the design shown in Fig. 4 was finally developed in several intermediate stages.

Close

Fig. 4 “H-Beam” in sagittal and isometric view.

Export to PPT

Since the H-Beam is a modification of the existing role model AIDA® short stem, it also has a double-conical, distally tapering, quasi-trapezoidal cross-section. In the proximal two-thirds, the H-beam principle has been consistently integrated. The area has a recess at the center running parallel to the sagittal plane, which in the following will be referred to as the “bar”. The lateral or medial area is used for the flat contact of the prosthesis with the bone. The two elevations are referred to below as “flanks”. The flanks taper progressively distally until they are at the same level as the bar. The proximal two-thirds exhibit the characteristic shape of an H-beam or double-T-beam principle. This area is designed to provide improved primary stability and enhanced rotational stability. The lower third is narrowly tapered in two stages to provide intramedullary guidance of the stem during implantation and stem apposition to the lateral femoral cortex. The tip of the prosthesis is rounded to prevent perforation of the lateral femoral cortices. The tapered “bar” of the prosthesis, which connects the two “flanks” in the upper region of the stem, has a polished surface in the further course of the stem below the flanks center roughness of Ra = 0.4 and serves only for distal lateral contact and intramedullary guidance. Due to the recess of the neutral fiber, the hip stem reacts significantly more flexibly under load than the established original model.

All internal and external edges are rounded, which should reduce stress and provide a good hold in the bone. The model shown here has a caput-collum-diaphyseal (CCD) angle of 130°. Detailed specifications of the H-Beam can be found in the patent of the inventors' association Jahnke, Stiller, Harz (file number: EP18155665.5).

4.1 Possible surgical technique

The medial flank as well as the lateral flank can have different angles of inclination in order to correspond to the individual anatomy and femoral configuration of the patient intraoperatively. The “bar rasp” has a rasp structure in the proximal two thirds and is polished in the distal rasp area. It has a double conical, distally tapering shape, which is based on the bar profile of the final prosthesis. The “flank rasps” have the H-beam structure of the final H-beam prosthesis and are used successively from small to large. In contrast to the bar rasp, the flank rasps have a smooth polished surface structure of the bar. Only the flanks have a rasp structure. The brightly polished bar can thus only act as a guide for the flank rasp in the already prepared medullary canal without further rasping it. The procedure described as follows, for example, could be suitable for implanting the H-beam prosthesis using the bar rasp and flank rasp:

In a first step I, the bar rasp (Fig. 5a) is used to successively rasp the medullary canal, after osteotomy of the femoral neck, from small to large, until a lateral contact of the bar rasp tip with the femoral cortex occurs and the bar rasp is located centrally within the osteotomy, taking into account the necessary antetorsional position of the femur. The positioning and size of the bar rasp in step I are subsequently verified intraoperatively by anterior-posterior radiographic control by the surgeon in a step II. If adequate stem position of the bar rasp is determined in step II, it is removed from the medullary canal in a step III. Then, in step IV, the H-beam profile of the prosthesis is successively rasped in the proximal region of the femur with a “flank rasp” (Fig. 5d) until the H-beam prosthesis can be inserted. The surgeon can use different flank rasps with different flank pitches according to the individual anatomical conditions in order to achieve the best possible cortical contact of the prosthesis flanks in the proximal metaphyseal region of the femur while at the same time preserving the cancellous bone structure as much as possible.

Close

Fig. 5 Illustration of a possible surgical procedure with a) the “bar rasp” b) the “bar rasp” in the bone in the a.-p- view and c) the final seat of the “bar rasp” in the bone. Afterwards, the d) “flank rasp” is inserted in this canal e) until final preparation f) of the medullary canal for the final prosthesis.

Export to PPT

After rasping the flanks, the final flank rasp is again left in the medullary canal and, in step V, an X-ray check and a trial positioning of the joint are performed. After positive evaluation of the stem position and the trial repositioning, in step VI the H-beam prosthesis is inserted or cemented into the implant site rasped free in steps I to IV. In the case of the cementless version, this is done using a joining procedure, preferably with a hammer.

5.1 Limitations

In this project, only a few in vitro experiments were performed and evaluated. Due to the good agreement between FE model and in vitro experiments, further experiments were not performed. Since the results of the total displacements of point C over all tests show only small standard deviations, it can be assumed that the corresponding span width would not be significantly increased by further tests. Furthermore, the tests were performed quasi-statically and not dynamically because the appropriate testing machines are not available. Results in vitro never correspond to actual in vivo results. However, since experiments in bone, due to high differences in human material, cannot be performed in a standardized way, a reproducible material means a reproducible compromise according to standard and is a good basis for carrying out the first preclinical studies.

FE analyses have sources of error, such as simplification of the model and idealization of transition conditions. However, the analysis of the H-beam was performed with a validated simulation system, where no significant deviations occurred between the test series and the simulation. The only changing parameter was the prosthesis geometry, therefore it can be assumed that the obtained results have an acceptable validity.

However, further studies, especially with dynamic loading, need to be performed to get a full impression of the performance of the novel hip stem design.