Kinematic Alignment Does Not Result in Clinically Important Improvements After TKA Compared With Mechanical Alignment: A Meta-analysis of Randomized Trials.

Abstract

Background: There is debate as to whether kinematic TKA or mechanical alignment TKA is superior. Recent systematic reviews have suggested that kinematically aligned TKAs may be the preferred option. However, the observed differences in alignment favoring kinematic alignment may not improve outcomes (performance or durability) in ways that patients can perceive, and likewise, statistical differences in outcome scores sometimes observed in clinical trials may be too small for patients to notice. Minimum clinically important differences (MCIDs) are changes that are deemed meaningful to the patient. A meta-analysis of randomized trials that frames results on this topic in terms of MCIDs may therefore be informative to surgeons and their patients.

Questions/purposes: (1) Does kinematic alignment for TKA insertion improve patient-reported outcome measures (PROMs) by clinically important margins (for example, 5 points of 48 on the adjusted Oxford Knee Score [OKS] or 13.7 points of 100 on the Forgotten Joint Score [FJS]) compared with mechanical alignment? (2) Does kinematic alignment for TKA insertion improve ROM by a clinically important margin (defined as 3.8° to 6.4° in flexion) compared with mechanical alignment?

Methods: A systematic review of Medline and Embase databases was performed from inception to January 29, 2023, the date of search. We identified RCTs comparing mechanical alignment TKA with kinematic alignment TKA. All English-language RCTs comparing PROMs data in kinematic versus mechanical alignment TKAs performed in patients 18 years or older were included. Studies that were not in English, involved overlapping reports of the same trial, and/or utilized nonrandomized controlled trial methodology were excluded. Conference abstracts or study protocols, pilot studies, and review articles were also excluded. Two reviewers screened abstracts, full-text, and extracted data and assessed included studies for risk of bias using the Cochrane Risk of Bias tool, version 2. Twelve randomized controlled trials (RCTs) were identified, which included 1033 patients with a mean age of 68 years (range 40 to 94) from eight countries who were undergoing primary TKA. Six studies were determined to be low risk of bias, with the remaining six studies were determined to be of moderate-to-high risk of bias. As a result, we would expect that the included studies might overestimate the benefit of the newer approach. Outcomes included ROM and PROMs. Where feasible, pooled analysis was completed. PROMs data were extracted from nine pooled studies, with a randomized n = 443 in the kinematic alignment group and n = 435 in the mechanical alignment group. ROM data were extracted from six pooled studies, with randomized n = 248 in the kinematic alignment group and n = 243 in the mechanical alignment group. PROMS were converted to common scales where possible. Multiple versions of the OKS exist; therefore, OKS scores were converted if needed to a 0 to 48 Oxford scale, in which higher scores represent better clinical outcomes. WOMAC scores were converted to OKS using previously reported techniques. The OKS and converted WOMAC scores were represented as "functional scores" in our data set because of their conversion. An MCID of 5 was utilized as previously documented for the OKS. Heterogeneity was assessed using the I2 statistic, and for an I2 of > 25%, random-effects models were utilized.

Results: In nine pooled studies, we found no clinically important difference between the kinematic and mechanical alignment groups in terms of our generated functional score (mean difference 3 of possible 48 [95% confidence interval (CI) 0.81 to 4.54]; p = 0.005). The functional score included OKS and WOMAC scores converted to OKS. The difference did not exceed the MCID for the OKS. In three pooled studies, we found no difference between the kinematic and mechanical alignment groups in terms of FJS at 1 to 2 years (mean difference 4 of possible 200 [95% CI -1.77 to 9.08]; p = 0.19). In three pooled studies, we found no difference between the kinematic and mechanical alignment groups in terms of EuroQol 5-domain instrument VAS score at 1 to 2 years (mean difference 0.2 of possible 100 [95% CI -3.17 to 3.61]; p = 0.90). We found no clinically meaningful difference between kinematic TKA and mechanical alignment TKA for ROM (extension mean difference 0.1° [95% CI -1.08 to 1.34]; p = 0.83, and flexion mean difference 3° [95% CI 0.5 to 5.61]; p = 0.02).

Conclusion: This meta-analysis found no clinically important benefit favoring kinematic over mechanical alignment in TKA based on the available RCTs. Because patients cannot perceive advantages to kinematic alignment, and because it adds costs, time (if using advanced technologies), and potential risks to the patient that are associated with novelty, it should not be widely adopted in practice until or unless such advantages have been shown in well-designed RCTs.

Level of evidence: Level I, therapeutic study.