Unicompartmental Versus Total Knee Replacement for Medial Compartment Knee Osteoarthritis (FLEX-UNI)

FLEX-UNI is a single-centre, prospective, randomized, double-blinded, controlled superiority trial comparing medial unicompartmental knee arthroplasty (UKA) with total knee arthroplasty (TKA) in patients with anteromedial knee osteoarthritis. Forty participants will be randomized 1:1. The trial is conducted at a high-volume elective arthroplasty centre and participants are followed for 12 months.

Participants are identified from the clinic lists of participating consultant orthopaedic surgeons and screened against the eligibility criteria by a member of the research team and by the treating consultant. Following written consent and baseline assessment, participants are randomized using sequentially numbered, opaque, sealed envelopes prepared and opened by a member of the research team not involved in screening, recruitment, or outcome assessment.

Participants allocated to UKA receive a cemented medial fixed-bearing Persona Partial Knee (Zimmer Biomet, Warsaw, Indiana, USA), implanted using a spacer block technique that resurfaces the medial compartment without deliberate correction of limb alignment. Participants allocated to TKA receive a cemented Triathlon cruciate-retaining total knee system without patellar resurfacing (Stryker, Mahwah, New Jersey, USA), implanted using a measured resection technique with a mechanical alignment target. Both procedures use conventional manual instrumentation, without robotic assistance or computer navigation, and are performed by the same group of consultant surgeons, each of whom performs both operations.

Suitability for medial UKA is confirmed intraoperatively at arthrotomy in all participants, including those allocated to TKA. Participants allocated to UKA who are found at arthrotomy to be unsuitable, principally because of disease progression beyond anteromedial osteoarthritis or a deficient anterior cruciate ligament, undergo TKA.

Participants are blinded to allocation. A standard-size sterile dressing is applied to the surgical wound by an independent healthcare support worker before biomechanical testing to conceal scar length, and participants are not shown their postoperative radiographs unless there is a clinical need. Research staff performing biomechanical assessment, outcome scoring, and statistical analysis are blinded to allocation, with unblinding only after analysis is complete. The treating surgeon cannot be blinded.

Assessments take place preoperatively, immediately postoperatively, and at six weeks, six months, and 12 months. Biomechanical assessment is performed in the centre's Motion Analysis Laboratory using a Vicon motion capture system (Vicon Motion Systems Ltd, Yarnton, UK) and two AMTI force plates (Advanced Mechanical Technology Inc., Watertown, Massachusetts, USA). Tasks are level walking, incline walking, decline walking, step ascent, step descent, sit-to-stand-to-sit, and single-leg balance.

A matched healthy control group is drawn from an existing motion analysis database held at the study centre, comprising individuals without known knee pathology previously assessed using an identical protocol under separate ethical approval. Control data do not contribute to the primary comparison between arms and are used only to contextualize findings against values expected in a healthy population. Healthy controls are not prospectively enrolled in this trial.

The sample size was calculated for the primary outcome, assuming 95 degrees for the UKA group and 89 degrees for the TKA group with a pooled standard deviation of 5.5 degrees, two-sided alpha of 0.05, and 80% power, giving 15 participants per group for an analysis of covariance with baseline as a covariate. Allowing for an estimated 10% intraoperative conversion from UKA to TKA and 15% loss to follow-up, the recruitment target is 20 per group. The trial is powered for the primary outcome alone, and all secondary outcomes are exploratory.

The primary analysis is performed on the per-protocol population, defined as participants who received their randomized implant and completed the 12-month biomechanical assessment. Participants found intraoperatively unsuitable for UKA at arthrotomy are excluded from the primary analysis regardless of allocation.

The trial is funded by Zimmer GmbH (Winterthur, Switzerland).