Fewer poor joint awareness results with cementless TKA compared with the same-design cemented implant

Comparisons of cemented and cementless TKA have focused on mean outcomes and generally reported equivalence. We hypothesized that a 3D-printed full-titanium cementless TKA and the same-design cemented implant would not differ significantly in average joint awareness but would differ in the proportion of patients with the poorest results, and we compared the two both at the mean and across the outcome distribution.

Prospective two-cohort study of 109 primary unilateral TKAs (54 cementless, 55 cemented) for end-stage knee osteoarthritis, with complete 24-month follow-up. Exclusion criteria were valgus deformity and bilateral surgery. The primary outcome was the FJS-12, analysed both at the mean (multivariable linear regression, with propensity score overlap weighting as sensitivity analysis) and across its distribution (quantile regression). Secondary outcomes were PASS achievement (FJS-12 ≥ 33.3), OKS, and time to functional satisfaction.

The cementless group was younger (P < 0.001) with more men (P = 0.008). Mean FJS-12 was not different between groups after adjustment (β = 7.4, 95% CI - 1.8 to 16.6, P = 0.115), although the confidence interval did not exclude a difference exceeding the minimal clinically important difference; this finding was reproduced by overlap weighting. On quantile regression, cementless patients scored significantly higher at the 10th and 25th percentiles (adjusted P = 0.028 and 0.021) but not at the median or above. A greater proportion of cementless patients achieved the PASS (96.3% vs. 76.4%; absolute difference 19.9%, 95% CI 7.1 to 32.9; number needed to treat 5), which persisted after adjustment (P = 0.005). No significant differences were observed in OKS (P = 0.91) or time to functional satisfaction (adjusted P = 0.39). No revisions or loosening occurred.

The 3D-printed full-titanium cementless TKA did not produce a significantly higher mean FJS-12 than the same-design cemented device at 24 months, but it significantly reduced the proportion of patients with poor joint awareness outcomes.

III.