D. Lung, D.J Tarinelli, MS, L. Crossett, MD, J.M Lee, MD, T.M Vogrin, MS, S.L-Y. Woo, PhD., DSc

Musculoskeletal Research Center
Department of Orthopaedic Surgery
University of Pittsburgh Medical Center, PA

Surgeons retain the Posterior Cruciate Ligament (PCL) in Total Knee Arthroplasty (TKA) believing that it provides improved stability and allows femoral rollback, enabling greater flexion. There are two commonly used polyethylene tibial inserts used in PCL retaining TKAs. One is the relatively flat Posterior-lipped (PLI) liner, which is preferably used when the PCL is functionally intact. The other is a more curved liner (CVD) that is designed to provide greater conformity when the PCL has sufficient laxity. Both of the liners are designed to produce sufficient femoral rollback. My project this summer was to determine which liner provides greater rollback in a TKA. We hypothesized that since the CVD theoretically provides greater conformity, that more femoral rollback would occur in the PLI liner.

For this study we used five fresh-frozen human cadaveric knees on the robotic/universal force-moment sensor (UFS) system. Each knee was thawed to room temperature before testing. Radiographs were taken at set angles (full extension, 30, 90) and analyzed to obtain the mechanical and anatomical axes. Radiographic templates were overlaid on the films to estimate the approximate size of the prosthesis.

Figure 1: Radiograph of a CVD liner at 30 degrees.

First, a Cruciate-retaining procedure was performed on each specimen using the Johnson & Johnson P.F.C. Modular Total Knee System. The ACL and menisci were removed, while the PCL was left intact. The femoral and tibial trial components were placed to determine the polyethylene tibial insert thickness. (Range .8-12 mm) The knees were balance for both the PLI and CVD inserts. Both medial releases as well as partial PCL releases were performed to achieve balance in the knee at full extensions, 30, and 90 degrees. The tibia and femur were resected 20 cm from the joint line leaving all remaining soft tissue structures intact. The femur was rigidly mounted relative to the base of the robot while the tibia was mounted through the UFS to the end-effector of the robot. The path of passive flexion-extension of the knee was determined between full extension and 90 degrees. The testing protocol was repeated for the second liner with an alternate testing procedure, switching the order of testing. The data collect that was using in this project consisted radiographs taken at full extension, 30 and 90 degrees, as well as passive path files for each liner. Using the positions from each passive path of flexion-extension, a screw axis can be calculated. These axes represent the amount of femoral rollback that occurred for each liner. Since both liners were tested on the same knee, comparisons can be made as to which provides greater rollback. Femoral Rollback was not analyzed under these those loading conditions however. Only passive path data was analyzed.

I analyzed the radiographs to determine the femoral-tibial contact point. From that point I would measure the percentage of the liner from that point to the edge at the anterior side. This percentage will increase with increasing femoral rollback throughout flexion.

Figure 2: Line graph showing the CVD vs PLI liner

When I analyzed the radiograph data I found that both shows some signs of rollback. Most of the knees did not show signs of rollback from angle to angle. Figure 2 shows the average of all five knees. The results from the radiographs show that the CVD liner provides slightly more rollback than PLI.

Figure 3: Screw Axes for an intact knee.

Figure 3 shows the screw axes for an intact knee as it goes from full extension to 90 degrees. There was a slight inconsistency at the 40-50 degree axes in which the axes shifted out of the excepted location. Axes after that returned back to showing rollback. As shown in Figures 4 and 5, both the CVD and PLI liners both showed rollback. The axes followed the same general motion as in the intact going from full extension to 90 degrees. The CVD and PLI liners showed more difference in the earlier flexions than later ones, at least in terms of axis direction. Whether or not one shows "more" rollback than the other is hard to determine since they do not occupy the same 3D space when you overlay them.

Figure 4: Screw Axes for a CVD liner

Figure 5: Screw Axes for a PLI Liner.

The current data shows that the CVD liner provided slightly more femoral rollback than the PLI liner. Unfortunately the data from knee to knee was very inconsistent, (some knees did not show rollback at all) which seems to point to the fact that these results are not very reliable. This inconsistence showed in both the radiographs and the screw axes. The screw axes did show more precisely the femoral rollback (3D as opposed to 2D), and is more feasible than taking 9 radiographs every 10 degrees.

To develop a way to quantify femoral rollback using screw axes. Difficult to quantify since each knee has a different paths. Maybe a comparison with the rollback that occurs in the exact knee before the TKA is performed. Calculating rollback under loading may also be a possibility.

I would like to thank Danyel for her guidance, and all the residents along with Ted Rudy who helped during the experiments. I would also like to thank Dr. Gilbertson and Dr. Woo for giving me the opportunity to work at the MSRC this summer. Lastly I would like to thank my fellow summer students who made the MSRC a fun and enjoyable place to work.