Favorable short-term radiographic outcomes seen after TKA performed with posterior-stabilized implant

Denis Nam , MD, MSc

The study by Beason and colleagues is a retrospective, radiographic evaluation of 37 patients who received a computer-navigated total knee arthroplasty (TKA) using the Vega posterior-stabilized knee prosthesis (Aesculap Implant Systems).  Weight-bearing anterior-posterior and lateral radiographs were obtained at each follow-up visit and at a minimum of 2 years postoperatively, and comparisons were made between the first and last available radiographs using the Knee Society Roentgenographic Evaluation Scoring System (KSRESS).  Based on their radiographic analysis, the authors concluded the Vega prosthesis system to have stable femoral and tibial implant interfaces at a mean of 3 years postoperatively. 

                The authors should be commended for their radiographic surveillance of a recently introduced TKA prosthesis, but there are significant methodological concerns with this investigation. First, despite reviewing a 4-year period from September 2009 to May 2013, only 37 “consecutive” patients could be identified who were included in this study. Furthermore, the term “consecutive” should be clarified, as it is unknown if all patients who received this prosthesis were included or only those with adequate follow-up. Second, two independent observers used the KSRESS system to review all patient radiographs, but intra- and inter-observer correlation coefficients were not provided. Third, the measurement of radiolucent lines at multiple zones of the bone-prosthesis interface is an important aspect of the KSRESS radiographic system and can provide useful information regarding component fixation and potential subsidence over time. It is unclear why only radiographic alignment, and not the presence of radiolucent lines, was evaluated. Fourth, and perhaps most importantly, I question the manner in which the results of Table 1 are interpreted. The authors conclude minimal component migration was present in this series of patients based on the small mean angular difference in component alignment over time. However, the range for the difference in the tibial angle (β) over time was from -2.54° to 3.67°, and for femoral flexion (γ) was -9.30° to 3.79°. Thus, presentation of the mean difference only masks the fact that an individual patient might have had a significant angular difference over subsequent radiographs. Despite the limitations of the KSRESS with regard to standardization of lower extremity rotation, it would be difficult to argue against the significance of these large changes in radiographic alignment over time for individual patients.

                Lastly, it remains unclear why clinical or functional outcome measures were not included in this analysis. The authors have noted their focus was on radiographic outcomes, but clearly the functional outcomes of these patients are equally important in determining if any of the purported design advantages of this system (such as enhanced joint biomechanics and implant stability as written in the manuscript) are seen clinically. While the authors should be commended for continued follow-up of their patients, a larger cohort of patients with continued clinical and radiographic follow-up is necessary to draw any conclusions regarding the success of this prosthesis.