Text

Objectives and questions: Persistent patellofemoral pain, maltracking, and instability remain a common cause of dissatisfaction following total knee arthroplasty (TKA) and have been linked to mismatch between implant trochlear geometry and patient-specific anatomy. Three-dimensional (3D) descriptions of trochlear morphology may better capture clinically relevant variability than isolated two-dimensional measures.

Aim: To characterise 3D trochlear groove morphology in an arthroplasty population using statistical shape modelling (SSM) and to identify clinically relevant patterns of anatomical variability.

Material and methods: This retrospective case series included 46 patients (4 bilateral) with knee osteoarthritis undergoing primary TKA, resulting in 50 distal femur datasets. Preoperative computed tomography (CT) scans were segmented to generate 3D distal femur reconstructions. Femora were aligned to a common anatomical coordinate system, left-sided femurs were mirrored for consistency. An SSM was built using a point distribution model and principal component analysis (PCA). Principal components (PCs) explaining ≥80% of total variance were retained and visualised using ±3 standard deviation deformations from the mean. The outcome measures were: 1. The principal components describing independent patterns of trochlear groove morphology and their relative contribution to total shape variance. 2. Feature-based parameters quantified for the mean shape and across ±3 standard deviation deformations of each component. These parameters included sulcus angle, trochlear width, medial and lateral trochlear inclination, and mediolateral trochlear groove offset.

Results: Five principal components explained 83% of total trochlear shape variance.PC1 (28%) represented variation in trochlear groove flatness and mediolateral asymmetry, primarily affecting sulcus angle and trochlear groove offset. PC2 (21%) captured overall distal femoral size and trochlear width. PC3 (16%) primarily reflected variation in lateral trochlear inclination. PC4 (12%) represented more distributed higher-order shape variation without a single dominant feature in the trochlear region. PC5 (6%) reflected variation in medial trochlear inclination. Collectively, the outcome measures demonstrated substantial inter-patient variability across multiple independent anatomical dimensions.

Discussion and conclusions: Trochlear groove morphology exhibits substantial 3D variability in patients undergoing total knee arthroplasty, which is not adequately described by isolated 2D measurements or addressed by standardised femoral component designs. These findings provide anatomical context for implant-anatomy mismatch and persistent patellofemoral symptoms after TKA. Incorporating 3D trochlear shape information into implant design and surgical planning may help improve restoration of patellofemoral biomechanics.