Patellofemoral parameters and anterior knee pain after medial unicompartmental knee arthroplasty

Abstract Background Medial unicompartmental knee arthroplasty (UKA) corrects varus alignment, but whether this correction is accompanied by measurable changes in patellofemoral (PF) geometry and whether such changes are associated with postoperative patellofemoral symptoms and function remain unclear. Methods In this prospective single-centre cohort study, 82 patients underwent robot-assisted medial UKA. CT-based varus angle, Q-angle, minimal medial and lateral patellofemoral joint spaces (MMPFJS and MLPFJS), patellofemoral index (PFI), and patellar tilt angle (PTA) were measured preoperatively and on postoperative day 1. The Anterior Knee Pain Scale (AKPS) was recorded preoperatively and at 6 months. Preoperative-to-postoperative comparisons used paired analyses. Associations between varus correction, Q-angle change, and AKPS improvement were evaluated using change-score analyses and exploratory multivariable models. Results Varus angle decreased from 8.38° (SD 3.11) to 3.65° (SD 3.20), with a mean change of − 4.73° (95% CI − 5.28 to − 4.17; P < 0.001). Q-angle increased from 6.39° (SD 6.74) to 9.74° (SD 5.87), with a mean change of 3.35° (95% CI 2.05 to 4.65; P < 0.001). MLPFJS increased and PFI decreased significantly, whereas PTA did not change significantly; the MMPFJS decrease did not remain significant after Holm correction. AKPS improved by 16.05 points (95% CI 14.20 to 17.90; P < 0.001). Greater varus correction was associated with a larger Q-angle increase, including after covariate adjustment (β = 0.76° per 1° correction, 95% CI 0.34 to 1.19; P < 0.001), but was not associated with AKPS improvement. Postoperative PF parameters were not associated with 6-month AKPS. Conclusions Robot-assisted medial UKA was followed by correction of varus alignment and an increase in the CT-based Q-angle, with greater varus correction associated with a larger Q-angle increase. However, early postoperative PF geometric changes were not associated with the magnitude of improvement in patellofemoral symptoms and function at 6 months. Part of the observed association between coronal correction and the CT-based Q-angle may reflect geometric coupling rather than a causal relationship with clinical improvement. Clinical trial number Not applicable.