Increased radiolucency on conventional radiographs around tibial components does not imply more load-induced displacement of a total knee arthroplasty.

Abstract

After total knee arthroplasty (TKA), radiolucent lines on routine conventional radiographs are commonly interpreted as impaired tibial component fixation, although their relation to implant motion remains unclear. This necessitates a comparison with measured component displacement. Recently, a computed tomography (CT)-based method measuring load-induced tibial component displacement has been introduced. This study aimed to determine whether the extent of radiolucent lines observed on radiographic imaging truly implicates displacement of tibial components. Conventional radiographs and CT-scans of 50 hybrid posterior stabilised TKA's with a cemented tibial baseplate were included in a post-market 10-year follow-up study. The extent of radiolucent lines on conventional radiographs was classified according to the Modern Knee Society, by two independent raters, creating three subgroups: 'none', 'partial', and 'complete'. Subsequently, load-induced implant displacement was quantified using CT-scans under valgus-varus loading. Relative implant displacement was quantified using rotation about the helical axis, Maximum Total Point Motion (MTPM), and mean target registration error (mTRE). Hypothesising similar displacement between subgroups despite differences in radiolucency, a Two One-Sided Tests (TOST) procedure, using the Smallest Detectable Change (SDC) as equivalence margins, was conducted. Statistical equivalence would indicate the subgroups showed similar displacement. Median displacement values were comparable across subgroups. Median rotation ranged from 0.60 to 0.79 degrees, median MTPM from 0.59 to 0.85 mm, and median mTRE from 0.44 to 0.53 mm across the subgroups. Statistical equivalence was confirmed in eight of nine comparisons, indicating that the subgroups were similar. The comparison of mTRE between the 'partial' and 'complete' subgroups did not meet statistical equivalence (p = 0.053). The inter-observer Cohen's Kappa for subgroup classification was 0.95 (95% confidence interval: 0.84-1.00). These findings suggest that a greater number of radiolucent lines does not represent more load-induced displacement of tibial components of TKA. This limits the diagnostic value of radiolucent lines for assessing tibial component fixation. Level III.