The influence of hip revision stem spline design on the torsional stability in the presence of major proximal bone defects

Abstract

Background: Despite the success of primary total hip arthroplasty, the number of revisions remains high. Infection, aseptic loosening, periprosthetic fractures and dislocations are the leading causes of hip revision. Current revision stem designs feature a tapered body with circumferential placed longitudinal thin metal splines that cut into the femoral cortex of the diaphysis to provide axial and rotational stability. Modifications to the spline design may help improve primary stability in various bone qualities. The purpose of this study was to evaluate whether the rotational stability of a revision hip stem can be improved by an additional set of less prominent, wider splines in addition to the existing set of splines. It is hypothesized that the additional splines will result in greater cortical contact, thereby improving torsional strength.

Methods and findings: The ultimate torsional strength of an established modular revision stem (Reclaim®, DePuy Synthes) was compared to a Prototype stem design with two sets of splines, differing in prominence by 0.25 mm. Five pairs of fresh-frozen human femurs (n = 10) were harvested and an extended trochanteric osteotomy was performed to obtain common bone defects in revision. Stems were implanted using successive droptower impacts to omit variability caused by mallet blows. The applied energy was increased from 2 J in 1 J increments until the planned implantation depth was reached or seating was less than 0.5 mm at 5 J impact. The ultimate torsional strength of the bone-to-implant interface was determined immediately after implantation. Image superposition was used to analyze and quantify the contact situation between bone and implant within the femoral canal. Cortical contact was larger for the Prototype design with the additional set of splines compared to the Reclaim stem (p = 0.046), associated with a higher torsional stability (35.2 ± 6.0 Nm vs. 28.2 ± 3.5 Nm, p = 0.039).

Conclusions: A second set of splines with reduced prominence could be shown to improve primary stability of a revision stem in the femoral diaphysis in the presence of significant proximal bone loss. The beneficial effect of varying spline size and number has the potential to further improve the longevity of revision hip stems.

Fig 1. Simulation of proximal bone defects by performing an ETO.

(A) Removal of the femoral head. (B) First ETO saw cut in distal direction following the linea aspera. (C) Completion of the ETO in proximal direction and removal of the 120° window including the majority of the greater trochanter.

Fig 2

(A) Cross-sectional view of the Reclaim implant. (B) Cross-sectional view of the Prototype design. The flat splines are 0.25 mm less prominent than the second set of splines.

Fig 3. Schematic drawing of the droptower used for stem impaction.

Dynamic impaction force was recorded with a load cell between stem adapter and impactor.

Fig 4

(A) Schematic drawing of the test setup. The top-mounted X-Y table allowed the bone-to-implant junction to be mounted in alignment with the axis of the two cardan joints. This allowed axial load and torque to be applied around the implant axis without detrimental limitations. (B) DIC markers were placed on the bone surface and the stem adapter to determine the relative torsion angle between bone and implant.

Fig 5. Alignment of the laser scan of a reclaim stem and the segmented cortical bone from the CT scan.

(A) Isolated meshes of the implant, the implant in the implanted state and the reamed cortical bone. (B) Aligned meshes based on the laser scan of the implant in the implanted state. The stopping criterion for ICP alignment was an RMSE of less than 0.2 mm.

Fig 6. Representative examples of the contact matrix revealing the intersections (yellow) between the mesh of the implant and the mesh of the reamed cortical bone.

(A) Contact matrix for a Prototype stem. (B) Contact matrix for a stem of the Reclaim design in the contralateral femur. Contact was evaluated between the distal incision of the ETO and the tip of the stem.

Fig 7

(A) Seating process for all specimens with fitted exponential curves. Some DIC measurements failed due to detached markers. (B) Stem seating coefficient for the two stem designs. Median is depicted in black and the mean is shown in red. The mean seating coefficient was higher (not significant, p = 0.305) for the Prototype design, indicating slower seating.

Fig 8

(A) The Prototype implants withstood a higher torsional moment to failure (p = 0.039). The Median is shown in black, the mean in red. (B) Torsional stability increased with maximum impaction forces (not significant, p = 0.326).

Fig 9

(A) Average cortical indentation depth of the longitudinal splines was not significantly different for the two designs (p = 0.502). (B) Number of matrix segments indicating cortical contact areas was higher for the Prototype design (p = 0.046). The median is shown in black, the mean in red.

Fig 10

(A) Cortical bone volume removed during cavity reaming. The median is shown in black, the mean in red. (B) The torsional stability increased with bone volume removed (p < 0.001, r² = 0.848).