Preoperative gap phenotypes in functionally aligned robotic total knee arthroplasty: derivation and internal...

Area of Science:

• Orthopedic Surgery
• Robotics in Medicine
• Biomechanical Engineering

Background:

• Robotic total knee arthroplasty (TKA) offers precise intraoperative measurements but preoperative patterns for functional alignment are not fully understood.
• Characterizing preoperative drivers of deformity is crucial for optimizing TKA outcomes.

Purpose of the Study:

• To derive and validate a pragmatic, driver-based classification system for preoperative robotic TKA data.
• To assess the internal coherence and inter-observer reliability of this classification framework.

Main Methods:

• Analysis of 68 primary functionally aligned robotic TKAs using CT-based robotic-arm data.
• Derivation of a four-phenotype classification (bone-driven, ligament-driven, flexion-dominant, mixed/complex) based on preoperative variables.
• Evaluation of internal coherence and inter-observer reliability (Fleiss' kappa).

Main Results:

• The classification identified four distinct phenotypes: bone-driven (58.8%), ligament-driven (8.8%), flexion-dominant (14.7%), and mixed/complex (17.6%).
• High plan-to-final HKA fidelity (92.1%) and excellent mechanical (93.5%) and functional (98.4%) alignment balance were achieved.
• Inter-observer agreement was almost perfect (Fleiss' kappa 0.88), indicating robust reliability.

Conclusions:

• A four-phenotype classification derived from preoperative robotic TKA data demonstrates internal coherence and high inter-observer reliability.
• This framework aids in understanding preoperative deformity patterns for functional alignment planning in TKA.
• Further multicenter validation is needed to establish clinical utility.