A multivariable MRI-based risk model for medial meniscal extrusion progression and treatment failure

Medial meniscal extrusion is a frequent magnetic resonance imaging (MRI) finding in patients with medial knee pain and early degenerative changes. Despite its clinical relevance, extrusion is commonly interpreted as an isolated linear measurement using fixed thresholds, which limits its prognostic value and usefulness for guiding treatment decisions. The complex interaction between meniscal displacement, anchorage integrity, alignment, and the surrounding cartilage environment is not captured by current approaches.

To develop and internally validate a multivariable MRI-based risk model incorporating a novel biplanar, size-adjusted meniscal extrusion index to predict progression of medial meniscal extrusion and treatment failure.

Observational cohort study.

A consecutive cohort of 200 patients with medial knee pain and measurable medial meniscal displacement on MRI was analyzed. Medial meniscal extrusion was quantified using standardized coronal and sagittal MRI measurements and integrated into a biplanar Medial Meniscal Extrusion Index (bpMMEI), normalized to tibial dimensions. Candidate predictors were predefined based on biomechanical and clinical relevance and included bpMMEI, medial meniscal root injury, mechanical varus alignment, cartilage damage severity, age, and body mass index. Treatment failure was defined as the occurrence of major subsequent surgery, clinically relevant meniscal reintervention, or meaningful deterioration in patient-reported outcomes. A penalized multivariable logistic regression model was developed and internally validated using bootstrap resampling. Model discrimination, calibration, and clinical utility were assessed.

Treatment failure occurred in 64 patients (32.0%) at a mean follow up of 32.4 ± 8.1 months. Higher bpMMEI values were independently associated with treatment failure; however, medial meniscal root injury, mechanical varus alignment, and advanced cartilage degeneration demonstrated comparable or greater relative effects within the multivariable framework, supporting the concept that extrusion should be interpreted as part of a broader biomechanical failure pattern rather than as an isolated imaging marker. Medial meniscal root injury, mechanical varus alignment, and advanced cartilage damage were also significant predictors. The final model demonstrated good discrimination (optimism-corrected AUC = 0.80) and satisfactory calibration. Risk stratification based on bpMMEI identified low-, intermediate-, and high-risk groups with observed failure rates of 11.8%, 33.8%, and 64.0%, respectively. Decision curve analysis showed a net clinical benefit of the model across a wide range of threshold probabilities.

A multivariable MRI-based risk model integrating a biplanar, size-adjusted meniscal extrusion index with key structural and biomechanical modifiers enables reliable stratification of treatment failure risk. This approach moves beyond isolated extrusion thresholds and provides a clinically applicable framework to support decision-making in medial meniscal preservation surgery.