Abstract
Summary Statement
Our data suggest that postoperative component positioning in TKA with PSPG is not consistent with pre-operative software planning. More studies are needed to rule out possible learning curve in this study.
Introduction
Patient specific positioning guides (PSPGs) in TKA are based on MRI or CT data. Preoperatively, knee component positions can be visualised in 3-dimensional reconstructed images. Software allows anticipation of component position. From software planning PSPGs are manufactured and those PSPGs represent intra-operative component alignment. To our knowledge, there are no studies comparing pre-operative software planning with post-operative alignment. Aim of this study is to investigate the correlation between pre-operative planning of component positioning and the post-operative achieved alignment with PSPG technique.
Patients & Methods
The first 25 TKA (cemented Vanguard® Complete Knee System, Biomet) with PSPG (Signature™ Biomet) performed at Telemark Hospital in 2009–2010 and the first 17 TKA with PSPG performed at Oslo University Hospital in 2010–2011 were included. A postoperative CT scanning and measurement protocol was used (Perth protocol). CT measurements were performed by 2 independent observers and comparative with pre-operative software (Materialise) planning. Component position angles of femur and tibia were measured. Mechanical axis for both femoral and tibial component angles in all planes was defined as zero degrees. Target angle for femoral component in sagittal plane was set to 2,8 degrees flexion on average and for the tibial tray to 3 degrees of posterior slope. Tibial rotation was in most cases obtained by using extra-medullary guide and therefore not included in this study.
Results
In respectively coronal, sagittal and axial plane the femoral component angle was on average 1.2° in varus, SD 1.6 (1.7° valgus −4.5° varus), 4.4° in flexion, SD 3.9 (17.3° flexion −1.6° extension) and 0.5° in external rotation, SD 0.1 (2.3° internal rotation −4.3° external rotation). For the tibial component angle the component was on average 0,5° in varus (3.5° valgus −7.3° varus) and 3.7° posterior slope, SD 2.3 (8.8° flexion −2.4° extension). Intra-class correlation (ICC) between the 2 independent observers was for femoral component in coronal, sagittal and axial plane 0.85, 0.93 and 0.63 and tibial component in coronal and sigittal plane 0.94 and 0.95.
Discussion/Conclusion
We expected that our measurements would be close to the pre-operative values. Although the mean values of post-operative measurements are close to pre-operative software planning, we found a considerable spread. Possible explanation might be error levels in pre-operative wrong identification of landmarks from MRI and/or different identification of bony landmarks on CT and intra-operative errors. All measurements were performed from the first Signatures performed in both hospitals. An early learning curve might explain some of the outliers. Time between manufacturing date and performed operation was in most cases several months, but less than the advocated 6 months. This time gap can theoretically provide a less proper fit in some cases due to slight change of anatomy in a progressive osteoarthritis. Our data suggest that postoperative positioning is not consistent with preoperative planning. This may be caused by the an early learning curve. It is uncertain whether this inconsistency is of clinical relevance. More data is necessary to prove any benefit of PSPG compared to existing procedures for TKA.
